MEDIATORS OF RADIATION RESISTANCE IN GLIOBLASTOMA
MULTIFORME
by
BRITTANY AGUILA
Submitted in partial fulfillment of the requirements for the degree of Doctor of
Philosophy
Dissertation Advisor: Scott M Welford, Ph.D.
Department of Biochemistry
CASE WESTERN RESERVE UNIVERSITY
May 2020
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the thesis/dissertation of
Brittany Aguila
Candidate for the degree of Doctor of Philosophy*
Committee Chair
Hung-Ying Kao, Ph.D.
Committee Member
Scott Welford, Ph.D.
Committee Member
Jason Mears, Ph.D.
Committee Member
Barbara Bedogni, Ph.D.
March 6, 2020
*We certify that written approval has been obtained for any proprietary material
contained therein.
2 Table of Contents List of Tables ...... 5
List of Figures ...... 6
Acknowledgements ...... 8
List of Abbreviations ...... 10
Abstract ...... 11
Chapter 1 - Introduction and Background ...... 13 Glioblastoma multiforme ...... 13 Standard of care ...... 14 Prognosis ...... 15 Molecular Subtypes ...... 15 Therapeutic Resistance ...... 17 Lentiviral mediated shRNA library negative selection and genomic hybridization functional knockdown screen ...... 22 PTGFRN ...... 28 CD9 ...... 30 Tetraspanin Enriched Microdomains ...... 28 AKT/PI3K signaling pathway ...... 32 SAT1 ...... 35 Statement of Purpose ...... 36 Chapter 2 - The Ig superfamily protein PTGFRN coordinates survival signaling in Glioblastoma multiforme ...... 39 Abstract ...... 40 Introduction ...... 41 Results ...... 43 PTGFRN is overexpressed in GBM and correlates with poorer patient outcome ...... 43 PTGFRN promotes cell proliferation and tumor growth ...... 44 PTGFRN inhibition sensitizes GBM cells to radiation ...... 46 PTGFRN is necessary for P-AKT signaling by increasing stability of p110β ...... 48 PTGFRN promotes DNA damage sensing through p110β ...... 51 Discussion ...... 52 Chapter 3 - Tetraspanin CD9 promotes tumorigenesis and radiation resistance in Glioblastoma multiforme ...... 74 Abstract ...... 75 Introduction ...... 76 Results ...... 77 CD9 is overexpressed in GBM and correlates with poorer outcome ...... 77 CD9 promotes cell proliferation and tumor growth in GBM tumors ...... 78
3 CD9 inhibition sensitizes GBM cells to radiation ...... 79 CD9 depletion hinders DNA damage repair ...... 80 CD9 signals through AKT ...... 81 Discussion ...... 81 Chapter 4 - A novel role for spermidine/spermine N1-acetyltransferase 1 as a gene-specific transcriptional regulator that promotes brain tumor aggressiveness ...... 94 Abstract ...... 95 Introduction ...... 96 Results ...... 97 SAT1 regulates gene programs controlling cell cycle and DNA dynamics ...... 97 SAT1 target genes are enriched in high-grade gliomas ...... 98 SAT1 regulates MELK and EZH2 by direct interaction with chromatin ...... 99 SAT1 regulates transcription of specific targets through polyamine catabolism102 Discussion ...... 103 Chapter 5 - Discussion and Future Directions ...... 116 The vast molecular basis for radioresistance ...... 116 Other signaling possibilities for PTGFRN ...... 118 PTGFRN and CD9: a possible scaffolding complex? ...... 124 Could SAT1 be regulating CD9? ...... 128 Concluding remarks ...... 133 Appendix A ...... 134
Appendix B ...... 139
Appendix C ...... 143
References ...... 148
4 List of Tables
Appendix C Supplementary Method Table 1: Antibodies ...... 145
Appendix C Supplementary Method Table 2: Primers for qPCR ...... 146
Appendix C Supplementary Method Table 3: Primers for Chromatin
Immunoprecipitations ...... 147
5 List of Figures
Figure 1-1: Negative Selection and Genomic Hybridization Functional Knockdown Screen……………………………………………………24
Figure 1-2: Results of negative selection and genomic hybridization functional knockdown screen…………………………………….26
Figure 2-1: Relevance of PTGFRN to GBM………………………………………56
Figure 2-2: PTGFRN expression in GBM tumors with specific mutation statuses and GBM molecular subtypes……………………………...58
Figure 2-3: PTGFRN is required for cell proliferation and tumor growth……………………………………………………………………………...... 60
Figure 2-4: PTGFRN depletion extends cell cycle progression……………..62
Figure 2-5: shRNA knockdown confirmation in GBM cell lines……………..64
Figure 2-6: PTGFRN reduction sensitizes GBM cells to IR…………………..66
Figure 2-7: PTGFRN depletion decreases p110β levels………………………68
Figure 2-8: PTGFRN depletion does not affect p110α protein levels and PTGFRN co-localizes with p110β……………………………………70
Figure 2-9: PTGFRN promotes DNA damage sensing………………………...72
Figure 3-1: Relevance of CD9 to GBM……………………………………………84
Figure 3-2: CD9 promotes cell proliferation and tumor growth……………..86
Figure 3-1: CD9 inhibition sensitizes GBM cells to radiation………………..88
Figure 3-2: CD9 depletion decreases the number of γH2AX foci and hinders DNA damage repair…………………………………………………..90
Figure 3-3: CD9 depletion decreases basal phospho-AKT protein levels…………………………………………………………………………………...92
Figure 4-1: SAT1 regulates cell cycle and DNA regulatory genes in GBM cells………………………………………………………………………….106
Figure 4-2: SAT1 target genes are elevated in aggressive brain tumors………………………………………………………………………………...108
6
Figure 4-3: MELK and EZH2 are SAT1 direct target nodes………………….110
Figure 4-4: Ectopic expression of MELK leads to overexpression of SAT1 target genes in an SAT1 dependent manner………………………..112
Figure 4-5: Polyamine catabolism is necessary for SAT1 transcriptional activity………………………………………………..……………114
Figure 5-1: Expression of FP receptor protein is not consistently altered after PTGFRN inhibition……………………………………………...... 122
Figure 5-2: Targeting CD9 and PTGFRN in combination decreases cell proliferation and radiosensitizes GBM cells……………………………...126
Figure 5-3: Inhibition of SAT1 decreases CD9 mRNA………………………..131
7 Acknowledgements
This work could not have been possible without the guidance and support from numerous people in my life. Thank you to my high school Biology teacher,
Mrs. Pam Zeigler, for introducing me to the excitement and mysteries of biology. I would also like to thank Dr. Gerri Hall, my Cleveland Clinic Science Internship mentor, and my undergraduate microbiology laboratory professor, Dr. Kathleen
Sandman, for introducing me to the world of science research. Further, I would like to thank Dr. Wenrui Duan, Li Gao, and Dr. Miguel Villalona-Calero for taking me under your wing and providing me with the necessary preliminary training and experience for my future graduate studies.
I would especially like to thank Dr. Scott Welford, my advisor, for his unwavering guidance and support through the good and the bad. I am also grateful for the guidance and mentorship provided by my thesis committee (old and new members), Dr. Barbara Bedogni, Dr. Hung-Ying Kao, Dr. Jason Mears and Dr. Junran Zhang. Your encouragement and insights were monumentally instructive and appreciated. Additionally, I would like to thank Dr. Nancy Oleinick, who went out of her way to support me from afar. I thank the members of my
“science lab family,” Dr. Vijay Thakur, Dr. Mariana Maier, Dr. Rutul Patel, Dr.
Yashi Gupta, Dr. Ravi Patel, Ayush Rana, Marco Dispagna, and Owen “Wilson”
Tan. I am thankful for your scientific insight but most importantly, for your friendship and laughter. I also thank my “science moms” Dr. Adina Brett-Morris and Varsha Thakur. Your friendships, love, and continuous support helped me during times of happiness and hardships and I am truly thankful for that.
8 To all the friends that I have made along this journey, you have become family and I am lucky for your friendships. You were there during times of scientific and personal struggle, always there to provide support and laughter.
Finally, I would like to thank my family, some of my biggest supporters. I am grateful for each and every one of you. To Sonia, Ramon, and Isabel, thank you for your support during this process. To my parents, thank you for always believing in me and encouraging me. Daddy, don’t worry, I’m not going to switch my major to business. To my sister, MacKenzie, you are an inspiration, and thank you for being a sounding board and for understanding the complexity of the scientific world. Finally, the biggest thank you to my husband, Frankie. You have encouraged me every single step of the way. I am thankful for the sacrifices you have made to allow me to fulfill my dreams and for the support you have given me throughout this long journey.
9 List of Abbreviations
BBB Blood brain barrier
ChIP Chromatin immunoprecipitation
CNS Central nervous system
DSB Double strand break
GBM Glioblastoma multiforme
Gy Gray
HR Homologous recombination
Ig immunoglobulin
IR Ionizing radiation
LGG Low grade glioma
MRI Magnetic resonance imaging
NHEJ Non homologous end joining
PCR Polymerase chain reaction
PI3K Phosphoinositide 3 kinase qRT-PCR Quantitative reverse transcription polymerase chain reaction
SPD spermidine
TCGA The Cancer Genome Atlas
TEM Tetraspanin enriched microdomain
10
Mediators of Radiation Resistance in Glioblastoma Multiforme
Abstract
by
BRITTANY AGUILA
Glioblastoma multiforme (GBM) is the most malignant primary brain tumor, occurring in every 3-4 per 100,000 adults. Despite aggressive treatment of surgical resection, chemotherapy and radiation therapy, the median survival remains approximately 14 months and only 3-5% of GBM patients surviving more than 3 years. Contributing to the poor survival rate, GBM is known to evade therapeutic intervention through multiple mechanisms of resistance, including resistance to radiation therapy. Therefore, in order to identify novel mediators of radiation resistance, a functional knockdown screen was conducted. The studies presented herein identify PTGFRN, CD9, and SAT1 as novel mediators of radiation resistance and drivers of tumorigenesis in GBM.
We show that elevated PTGFRN expression promotes glioblastoma tumorigenesis, doing so through regulation of cell survival signaling. We demonstrate that PTGFRN promotes AKT signaling through PI3K p110β stabilization, supporting AKT-driven survival signaling and tumor growth. Further, we show that PTGFRN inhibition decreases nuclear PI3K p110β leading to decreased DNA damage sensing and DNA damage repair, suggesting that
11 PTGFRN is necessary for efficient DNA damage repair signaling following radiation treatment.
Additionally, we show that CD9, a binding partner of PTGFRN, promotes tumorigenesis in GBM. Upon CD9 depletion, we find that cell proliferation and tumor growth is hindered and cells become sensitive to radiation treatment. Also, we demonstrate that targeting CD9 decreases phospho-AKT protein expression and hinders DNA damage repair. Furthermore, we demonstrate SAT1, which has been found to promote radiation resistance by regulating BRCA1 transcription through histone acetylation, to be a gene-specific transcriptional regulator through local polyamine acetylation. Importantly, we show that SAT1 regulates
MELK, and further EZH2 and FOXM1, all which are necessary for GBM stem cell maintenance, tumor growth, and radiation resistance.
Taken together, these data suggest critical roles for PTGFRN, CD9, and
SAT1 in mediating radiation resistance and driving GBM tumorigenesis and aggressiveness. Furthermore, the development of a therapeutic aimed at targeting these proteins, perhaps in combination with other targeted therapies and/or chemotherapeutics, could provide significant clinical benefit for glioblastoma patients.
12 Chapter 1 - Introduction and Background
Glioblastoma multiforme
Glioblastoma multiforme (GBM), as defined by the World Health
Organization, is a grade IV astrocytoma and the most common and aggressive primary brain tumor (1). Histopathologically, GBMs contain distinguishing features such as necrosis and/or microvascular proliferation, including endothelial hypertrophy, endothelial hyperplasia, and glomeruloid vessels (2).
GBMs are commonly located in the supratentorial region (temporal, parietal, frontal, and occipital lobes) and are rarely seen in the cerebellum and spinal cord
(3). In the United States, the incidence rate of GBM is approximately 3-4 per
100,000 adults per year, with more than 10,000 cases diagnosed annually (4, 5).
Clinically, GBMs can be classified into primary and secondary GBMs, which evolve through different genetic pathways, affect patients at different ages, and have differences in survival outcomes. Primary GBMs, comprising of about
80% of GBMs, arise de novo as aggressive, highly invasive neoplasias and occur in older patients (mean age of 62 years). Secondary GBMs are less common, derive from the transformation or progression of lower grade astrocytomas or oligodendrogliomas, and occur in younger patients (mean age of 45 years) (3).
Most common symptoms of GBM include persistent headaches, focal neurological deficits, memory loss, confusion, personality changes or seizures. A combination of neurological exams and imaging is used to diagnose GBM
13 including magnetic resonance imaging (MRI) and adjunct technology (functional
MRI, diffusion-weighted imaging, diffusion tensor imaging, dynamic contrast- enhanced MRI, perfusion imaging, proton magnetic resonance spectroscopy and positron-emission tomography) (6).
Standard of care
The current standard therapy for newly diagnosed GBM patients under 70 years old includes maximal surgical resection followed by radiotherapy and concomitant and adjuvant temozolomide or carmustin wafers (Gliadel) (5). While
GBM cannot be completely removed surgically because of its infiltrative nature, patients undergo maximal surgical resection whenever possible as it reduces the symptoms from mass effect and provides tissue for histologic diagnosis and molecular studies. The mainstay of treatment for GBM is radiotherapy, as the addition of radiotherapy increases GBM patient survival from approximately 3-4 months to 7-12 months. Conventional radiation therapy consists of 60 Gy of partial-field external-beam radiation delivered in 30 fractions of 1.8-2 Gy (6).
Temozolomide, a form of chemotherapy, was added to the standard of care for newly diagnosed GBM following a phase III trial. The trial, conducted by the
European Organization for Research and Treatment of Cancer and National
Cancer Institute, compared radiotherapy alone with radiotherapy and concomitant treatment with temozolomide, followed by adjuvant temozolomide therapy. The study revealed that the combination of radiotherapy and temozolomide had an acceptable side-effect profile and increased the median survival compared to radiotherapy alone (14.6 months versus 12.1 months).
14 Additionally, the survival rate at 2 years among the patients who received the combined treatment was significantly greater than the rate among patients who received radiation alone (26.5% versus 10.4%) (7).
For recurrent GBMs, treatment options include supportive care, reoperation, re-irradiation, systemic therapies, and combined modality therapy.
The role of reoperation remains unclear and the clinical and survival benefit is dependent on patient and tumor characteristics. Recurrent GBMs can receive a variety of methods for re-irradiation, including brachytherapy, fractionated stereotactic radiation therapy, radiosurgery, and conformal or intensity-modulated radiation therapy, with or without new chemotherapy agents. Additionally, the published data include a wide variety of radiation doses, emphasizing that no standard approach exists for recurrent GBMs (8).
Prognosis
Despite aggressive treatment, the vast majority of patients will have early tumor progression or recurrence (9, 10). In the majority (70-90%) of patients,
GBM recurs most often within 2 to 3 cm from the border of the original lesion and in 5% of cases, multiple lesions are seen after treatment (10). The prognosis of survival is approximately 12-15 months and a mean survival rate of 3.3% at found at 2 years and 1.2% at 3 years, making the survival prognosis of GBM patients one of the worst confronted in modern day oncology (5, 11, 12).
Molecular Subtypes
Even with identical histopathological classifications, GBMs may behave differently in terms of clinical outcomes and responses to treatment. Differences
15 in molecular patterns of GBM can partially explain clinical outcomes and predict responses to treatment (13). Therefore, understanding the molecular mechanisms underlying GBM tumorigenesis and progression can provide for improved diagnostic and prognostic information and may aid in the development of novel therapeutic treatments.
The Cancer Genome Atlas (TCGA) Research Network demonstrated that genomic profiling of GBM patient tumors outlined four GBM subtypes: classical, mesenchymal, proneural, and neural. The classical subtype is characterized by chromosome 7 amplification paired with chromosome 10 loss, EGFR amplification, no TP53 mutations, CDKN2A deletion thereby affecting the Rb signaling pathway, and high expression of nestin, Notch, and Sonic hedgehog signaling pathways. The mesenchymal subtype is characterized by decreased
NF1 expression levels (either by focal hemizygous deletions of a region at
17q11.2 or NF1 mutation), high expression of genes in the tumor necrosis factor super family pathway and the NF-κB pathway, and expression of mesenchymal markers such as CHI3L1, MET, CD44 and MERTK. The proneural subtype is characterized by concomitant focal amplification in conjunction with high levels of
PDGFRA gene expression, point mutations in IDH1, TP53 mutations and TP53 loss of heterozygosity. The proneural subtype also demonstrates high expression of oligodendrocytic development genes (PDGFRA, NKX2-2, and OLIG2), downregulation of CDKN1A expression, PIK3CA/PIK3R1 mutations, and contains several proneural development genes, such as SOX family genes, DCX,
DLL3, ASCL1 and TCF4. Finally, the neural subtype is characterized by the
16 expression of neuron markers, such as NEFL, GABRA1, SYT1, and SLC12A5
(14). Following the original four GBM-subtype classifications, additional studies have further refined the classifications into three GBM-subtypes, and their analyses demonstrated that the original proneural and neural subtypes are a single cluster, proneural/neural (15).
Clinically, the proneural/neural subgroup correlates with longer survival outcomes compared to the other subtypes. However, GBM patients with tumors of the proneural subtype were unlikely to benefit from aggressive therapies compared to patients with the classical and mesenchymal subtypes (14, 15).
Therefore, considering the various GBM molecular profiles, which can lead to different treatment efficacy among patients, it may be beneficial for therapy to be personalized to target each patient’s GBM subtype (5).
Therapeutic Resistance
The current therapies used to treat GBM have demonstrated limited effectiveness due to the anatomic location of GBM, the blood brain barrier, and intrinsic and acquired mechanisms of resistance (6, 7, 16). The blood brain barrier (BBB) functions as a protective boundary between the central nervous system (CNS) and systemic circulation to regulate CNS homeostasis and restrict the entry of circulating toxins, inflammatory cells, and macromolecules into the brain parenchyma. The BBB contains a layer of endothelial cells that line blood vessels creating a barrier restricting the transport of these substances through the use of tight junctions (17, 18). Pericytes and astrocytes, additional supportive cells found in the BBB, regulate the fidelity of these tight junctions and control
17 enzymatic, transport, and efflux mechanisms critical to BBB function (17). The barrier significantly limits distribution of many therapeutic drugs into the tumor tissue, including monoclonal antibodies, antibody-drug conjugates, and hydrophilic molecules that have difficulty crossing the lipid bilayer. Additionally, for lipophilic molecules that can readily cross lipid bilayers, various transmembrane efflux transporters, which are found in the endothelial cells, function as biochemical barriers by actively transporting therapeutic drugs back into the capillary lumen (blood vessel) and out of cells (19). Therefore, the BBB is a major challenge in treating GBM by limiting the accumulation of oncologic drugs in the brain tumor.
The stability of the BBB in GBM tumors is controversial. Some studies have suggested that GBM causes BBB dysfunction through a variety of anatomic and physiologic mechanisms. Tight junction proteins, including claudin and occludin, are found to be downregulated, and BBB transporter proteins are upregulated, resulting in increased BBB permeability. Additionally, GBM tumors are known to create new blood vessels, further disrupting the BBB physiologic barrier and promoting blood vessel leakiness (17). However, another study has demonstrated that all GBM patients have tumor regions with an intact BBB, and only portions of the tumor may have disrupted BBB. This study suggests that due to GBM having clinically significant regions of tumor with an intact BBB, patients will fail treatment because of the inability to deliver an effective therapy to all regions of the tumor (19). Therefore, focus should be on developing therapeutics
18 that can cross an intact BBB and deliver adequate drug distribution across the tumor.
GBM has also been characterized to contain a high degree of tumor heterogeneity, providing a major challenge to the efficacy of targeted therapy and leading to the development of rapid therapeutic resistance. A recent single-cell analysis of primary GBM patients demonstrated that cells from the same tumor have differential expression of genes involved in oncogenic signaling, proliferation, immune response, and hypoxia (20). The differences in oncogenic signaling pathways among intratumoral populations include the p53, EGFR,
PDGFRA, and mTOR pathways. This further suggests that different subtypes of
GBM can coexist within the same tumor and likely exhibit differential therapeutic responses (17, 20). GBM lesions also display vast intertumoral diversity, including lesions at different sites within the brain. Some studies demonstrate that 50% of recurrent disease samples share only half of its genetic mutations with the primary tumor (17). The different signaling pathways activated in the tumor, along with the possibility of an oncogenic signaling change in recurrent tumors, advocate for the development of a multi-target treatment modality.
Methylation of particular genes in GBM tumors can also promote therapeutic resistance. The MGMT gene is frequently silenced in GBM by promoter hypermethylation and is found in approximately 30-60% of GBM tumors
(5). The use of alkylating agents, such as temozolomide, as chemotherapeutic drugs produces an alkyl group on the O6-position of guanine, inducing a DNA mismatch, DNA double-strand break, and ultimately apoptosis in proliferating
19 cells (1). The MGMT protein repairs the DNA damage by transferring the alkyl group from the guanine to an internal cysteine residue found in the MGMT protein, therefore MGMT activity promotes resistance to temozolomide treatment
(21). MGMT promoter methylation is associated with better response to temozolomide as well as radiation therapy, and improves progression free and overall survival with combined treatment (temozolomide and radiation therapy) compared to either treatment modality alone (3).
While radiation therapy is a mainstay in GBM treatment, the therapeutic efficacy of radiotherapy is severely limited due to the high intrinsic radioresistance of GBM. The molecular network underlying radioresistance has yet to be fully defined, however several studies have provided some explanations for GBM’s radioresistant phenotype. The signaling pathways that may be involved in promoting radioresistance include the AKT, Notch, Wnt/β-catenin,
ATM/Chk2/p53, STAT3 and Hedgehog signaling pathways, with the majority of these pathways activating DNA damage repair, thereby minimizing the deleterious effects of radiation-induced DNA double strand breaks. Additional mechanisms of resistance to radiation therapy includes the presence of hypoxia, via DNA-PKcs regulating HIF-1α expression, the tumor microenvironment, and glioma stem cells (22).
Cancer stem cells are pluripotent with self-renewal capability, can proliferate continuously, form neurospheres, and possess characteristic biomarkers such as CD133, nestin and SOX2 (22). GBM tumors contain discrete populations of cancer stem cells, cancer-initiating cells, and cancer-propagating
20 cells. Generally, cancer stem cells comprise approximately 2-3% of a glioma tumor but can increase to 5% in GBM tumors (23). These cells are highly resistant to chemo- and radiotherapy and can recapitulate a large tumor mass following treatment (17). GBM stem cells have been shown to contribute to radiation resistance by increasing the DNA damage response machinery and containing more efficient pro-survival signaling (20, 24). Further, studies have shown that stem cell markers, such as L1CAM, Bmi-1, SOX2, and CD44, can regulate radioresistance by promoting DNA damage repair. L1CAM enhances radioresistance by increasing phosphorylation of ATM and Chk2. Bmi-1 is vital for the recruitment of DNA double-strand break response proteins to sites of damage. FOXM1 mediated transcriptional regulation of Sox2 promotes clonogenic growth, stem-like properties, and stem cell radioresistance. Finally,
CD44 expression correlates with cancer stem cell phenotypes, tumor aggressiveness, poor survival, and the mesenchymal subtype. GBMs with the mesenchymal subtype are enriched in CD44+ cancer stem cells and show limited benefit from radiation therapy, evidently through NF-κB activation (22, 25).
Several prosurvival signaling pathways have been shown to be important in promoting resistance to radiation in cancer stem cells. The PI3K/AKT/mTOR,
NF-κB, TGF-β, and Notch signaling pathways have all been implicated in promoting resistance to radiation therapy. Activation of the AKT pathway by radiation induces the repair of radiation induced DNA double strand breaks and single strand breaks (24, 26). The NF-κB signaling pathway has shown to play an essential function in the response of cells to ionizing radiation by mediating
21 mesenchymal differentiation in cancer stem cells, thereby promoting activation of checkpoint pathways, DNA damage repair, and unperturbed cell cycle progression (27). Additionally, a study has demonstrated that glioma stem cells containing TGF-β promote DNA damage repair, invasion, mesenchymal transition and angiogenesis (28). Finally, the Notch signaling pathway has also been implicated in radiation resistance in glioma stem cells, not by altering the
DNA damage response, but by regulating radiation-induced AKT activation and the levels of Mcl-1, an anti-apoptotic Bcl-2 protein (29). In response to multiple mechanisms promoting intrinsic resistance to radiation, developing novel therapies targeting these cell’s essential functions may reduce GBM recurrence and improve the overall survival rate (20).
Lentiviral mediated shRNA library negative selection and genomic hybridization functional knockdown screen
GBM has been shown to be resistant to radiotherapy, therefore, in order to identify novel mediators of radiation resistance in GBM, a negative selection and genomic hybridization functional knockdown screen utilizing a lentiviral mediated shRNA library on two different GBM cell lines was conducted. U87MG and
LN229 GBM cell lines were utilized in this screen; U87MG is p53 wild-type,
PTEN-null GBM cell line displaying a mesenchymal gene signature, and LN229 is p53 mutant, PTEN wild type GBM cell line that displays gene signatures of both proneural and mesenchymal GBM subtypes (30-36). The lentiviral shRNA library was comprised of approximately 30,000 barcoded shRNAs that targeted
22 more than 10,000 genes. Further, to ensure that each cell received only one shRNA lentivirus, the GBM cells were infected with pools of lentivriuses at an estimated multiplicity of infection of 0.3. After selection, the cells were divided into treatment (2 Gy of IR) and control groups. The cells were cultured for 72 hours post irradiation and then underwent genomic DNA isolation.
The knockdown of genes, via shRNA, that confer radioprotection should result in the sensitization of cells to radiation and therefore, their depletion from the population. Alternatively, knockdown of genes that promote radiation sensitivity should result in protection and increased survival in the population. By determining the relative abundances of the shRNA barcodes in the genomic DNA of the treatment group versus the control group, shRNAs that altered the sensitivity of cells to radiation could be identified. The shRNA barcodes were amplified by PCR, labeled with Cy3 or Cy5, and then hybridized onto barcode microarrays (flow chart of screen shown in Fig 1-1). The raw data of the U87MG screen is represented in Fig 1-2A. Of the 21,555 detected barcodes, 1,868 were decreased by 1.5-fold or more; 764 were decreased by 2-fold or more; and 126 were decreased by 4-fold or more (Fig 1-2B). As an internal positive control, 47 shRNAs targeting genes known to be involved in DNA repair were identified, such as ATM, PARP1, RAD9, and RAD51. When the results of both the U87MG and LN229 screens were combined, 79 shRNAs were found to be decreased by
1.5-fold or more, 10 were decreased by 2-fold or more, and no shRNAs were identified to be decreased by 4-fold (37). From this screen, PTGFRN, CD9 and
SAT1 were identified as possible novel mediators of radiation resistance in GBM.
23 Figure 1-1: Negative Selection and Genomic Hybridization Functional Knockdown Screen Flowchart demonstrating the experimental screen used to identify novel mediators of radiation resistance in U87MG and LN229 GBM cell lines.
24 Figure 1-1:
25 Figure 1-2: Results of negative selection and genomic hybridization functional knockdown screen A, Raw data of the U87MG shRNA negative selection and genomic hybridization functional knockdown screen. The abundance of individual shRNA barcodes are represented as log2-transformed ratios of irradiated over non-irradiated cells. B,
Quantification of shRNA barcode abundance ratios.
Note: Figure adapted from Brett-Morris A, Wright BM, Seo Y, Pasupuleti V,
Zhang J, Lu J, Spina R, Bar EE, Gujrati M, Schur R, Lu ZR, Welford SM. The polyamine catabolic enzyme SAT1 modulates tumorigenesis and radiation response in GBM. Cancer Res. 2014;74(23):6925-34.
26 Figure 1-2:
27 Tetraspanin Enriched Microdomains
Tetraspanins have the ability to associate among themselves and with various transmembrane proteins to form a distinct class of membrane domains, called the tetraspanin-enriched microdomains (TEMs) or the tetraspanin web
(38). TEMs are preorganized units present at the plasma membrane of cells that have great heterogeneity in size (an average area of 200-400nm2) and cell types.
TEM components diffuse as coordinated units as the ligation of a receptor present in TEMs affect the diffusion dynamics of other components of the microdomain. Additionally, their size allows them to accommodate numerous molecules thus regulating receptor avidity. Furthermore, the microdomains have a heterogeneous composition, allowing them to interconnect receptors involved in functionally related processes (39).
TEMS are organized into several different levels. Each level comprises of either tetraspanin-partner or tetraspanin-tetraspanin homo- or heterophilic interactions that are further accompanied by lipid-protein interactions involving palmitoylation-dependent cholesterol binding and gangliosides (39, 40).
Members of TEMs, along with tetraspanins, can include cell adhesion molecules
(e.g. integrins), Ig superfamily proteins, growth factor receptors and membrane bound growth factors, MHC class I and II proteins, proteoglycans, and proteases and ectoenzymes (41, 42). Due to the heterogeneous composition of TEMs, the microdomains are thought to serve as a regulated adaptor platform and are therefore involved in multiple cellular functions (43). For example, TEMs can modulate growth factor signaling. The tetraspanin CD9 has been show to
28 associate and modulate the functions of EGFR agonists, such as TGFα, HB-
EGF, and amphiregulin. Multiple tetraspanins have also been shown to associate with and modulate the function of growth factor receptors, such as GPR56 and
EGFR (44, 45). Another study demonstrated that tetraspanin CD82 can associate with EGFR causing an accelerated ligand-induced removal of EGFR from the cell surface resulting in attenuation of EGFR signaling (46). TEMs can also modulate integrin-dependent post-cell adhesion events, affecting integrin- dependent migration, spreading, and cell morphology (44, 45). As TEMs serve as an adaptor platform, the differences in function are a result of the wide diversity of TEM constituents present at the plasma membrane.
PTGFRN
PTGFRN (FPRP, CD315, CD9P-1, EWI-F), prostaglandin F2 receptor negative regulator, is an immunoglobulin (Ig) superfamily protein that was discovered when it co-purified with a prostaglandin F2α complex binding fraction
(47). From this, the predicted protein structure of PTGFRN includes a signal sequence, six glycosylated Ig loops, a single transmembrane domain, and a short, highly charged, intracellular carboxy tail (48). PTGFRN possesses no prostaglandin F2α binding activity and was found to down-regulate prostaglandin
F2α binding to the prostaglandin F2α receptor (FP receptor) by decreasing the receptor number rather than the affinity constant. Additionally, PTGFRN only has an inhibitory effect on prostaglandin F2α binding to the FP receptor when
29 PTGFRN and the FP receptor are co-expressed in the same cell. At present, it is unknown whether the FP receptor and PTGFRN interact directly or indirectly
(49). Further, PTGFRN, a member of the tetraspanin enriched microdomains
(TEMs) (40), has been discovered to possess partner proteins with which it associates. PTGFRN has been found to associate specifically, and at high stoichiometry, with tetraspanins CD81 and CD9, also members of TEMs, and modulators of tumorigenesis (47, 50).
The function of PTGFRN has been little studied. Thus far, PTGFRN has been found to be essential for human endothelial proliferation, migration, and in vitro and in vivo angiogenesis, through CD9 and CD151 stabilization at the human endothelial cell surface (51). Additionally, PTGFRN expression, at both the transcriptional and translational level, is found to correlate with the metastatic status of lung tumors, particularly at the migratory edge of tumors (52). Finally,
PTGFRN expression is associated with lipid accumulation in the 3T3-L1 preadipocyte cell line, suggesting that PTGFRN may be involved in lipid metabolism (53), and PTGFRN is found to interact with actin linking ezrin-radixin- moesin (ERM) proteins, not only connecting TEMs with the actin cytoskeleton, but acting as a regulatory protein to facilitate signal transduction pathways downstream from TEMs (54).
CD9
CD9 is a tetraspanin. Tetraspanins comprise a large family of transmembrane proteins that are expressed in all cell types. Tetraspanins
30 organize laterally with other membrane proteins and themselves to form TEMs, also termed the tetraspanin web (38). Members of the tetraspanin superfamily contain four hydrophobic transmembrane segments and two extracellular domains, known as the large and small extracellular loops (55, 56). Tetraspanins can associate with signaling enzymes including conventional Protein Kinase Cs
(PKCs), type II phosphatidylinositol 4-kinase (PI4K), and protein phosphatases
(38). Collectively, tetraspanins, acting as scaffolding proteins (57), can influence cell migration, invasion, cell-cell fusion, survival and signaling, all functions which are relevant in different stages of tumor development (58).
In cancer, CD9 is a context-dependent signaling molecule. In breast, prostate and lung cancers, elevated CD9 expression suppresses tumor progression and low CD9 expression correlates with poor patient prognosis (59).
In contrast, CD9 expression is increased in gastric cancer and is correlated with increased disease severity and poor prognosis (60, 61). In B cell acute lymphoblastic leukemia, high CD9 expression promotes cancer stem-cell like properties (62), and CD9 may be oncogenic in an ovarian cancer cell line (63).
The association of CD9 with other transmembrane and partner proteins within the TEM may explain and influence CD9 function in various cancers (38, 58).
Several studies have described the contribution of CD9 to GBM malignancy. One study defined CD9 as a potential novel biomarker for GBM cancer stem cells. The authors found CD9 to be upregulated in GBM tissues compared to normal brain and demonstrate that CD9 is involved in the regulation of GBM stem cell survival and invasion, possibly through receptor tyrosine kinase
31 signaling (64). Upon further investigation, CD9 was also found to stabilize gp130, a transmembrane IL-6 receptor, by blocking its lysosomal degradation to promote
STAT3 signaling and support GBM stem cell propagation and malignant progression (65). Both studies demonstrate CD9 to be a potential target for the development of new therapeutics against GBM stem cells in order to improve
GBM treatment.
AKT/PI3K signaling pathway
The PI3K pathway is one of the most potent pro-survival signaling pathways. Members of the PI3K family are involved in multiple cellular processes, including proliferation, differentiation, migration, metabolism, and survival. Upon ligand binding to a receptor tyrosine kinase or a G-protein coupled receptor, the activated receptor will bind to regulator PI3K subunit p85. The subsequent conformational change will release the catalytic PI3K subunit p110
(α, β, δ, and γ). Activated p110 then phosphorylates phosphatidy-linositol-3, 4- bisphosphate (PIP2) into phosphatidy-linositol-3, 4, 5-bisphosphate (PIP3) which recruits AKT to the inner membrane for phosphorylation on its serine/threonine kinase sites. Activated AKT is involved in numerous downstream signaling events (66). Generally, the PI3K p110α and p110β subunits are ubiquitously expressed, whereas p110δ and p110γ are primarily expressed in leukocytes. In cancer cells, the PI3K subunits exhibit distinct roles in various pathological processes. For example, the p110α subunit was found to promote tumor cell
32 proliferation, migration and invasion, and the p110β subunit is essential for cell survival and tumorigenesis (67).
In GBM, up to 88% of tumors have altered receptor tyrosine kinase/RAS/PI3K signaling (68). Genetic alterations such as EGFR amplification,
PIK3CA gain of function, or loss of PTEN all lead to continuous activation of AKT signaling (66, 69). In GBM cells, hyper-activation of the PI3K/AKT signaling pathway confers rapid growth, tumor progression, and multidrug resistance (67).
Indeed, multiple studies have linked abnormal PI3K/AKT signaling to radioresistance in GBM (22, 26, 70). In one study utilizing U251 glioma cells, downregulating AKT resulted in sustained, unrepairable DNA double strand breaks following radiation treatment (71). Similarly, another study demonstrated that activating AKT signaling promotes γH2AX foci resolution and enhances DNA damage repair through homologous recombination and non-homologous end joining (72). Further, EGFRvIII is a common EGFR mutation, characterized by the deletion of exons 2-7 resulting in a truncated extracellular domain with ligand- independent constitutive activity (73), occurring concurrently with EGFR amplification in GBMs. Experiments utilizing U87MG GBM cells have demonstrated that high levels of EGFRvIII increased radiation resistance by enhancing the PI3K/AKT signaling pathway and promoting the rapid repair of radiation-induced DNA double strand breaks (74). Therefore, these data demonstrate that activating the PI3K/AKT pathway can promote radioresistance by stimulating the DNA damage repair response thereby minimizing the lethal
33 effects of DNA double strand breaks. Therefore, targeting the aberrant PI3K/AKT pathway may result in sustained DNA damage and cell death in GBM cells.
While over 50 PI3K inhibitors have been designed and produced for cancer treatment, only a handful of them have successfully entered into GBM clinical trials. The first generation of pan-PI3K inhibitors included wortmannin and
LY294002, but their use is limited clinically due to their insolubility, short half-life, off-target effects and unacceptable toxicities. The next generations of pan-PI3K inhibitors that have made it into clinical trials include BKM120, XL147, PX-866, and GDC-0941 (67). BKM120, an orally bioavailable pan-PI3K inhibitor, induces
G2/M cell cycle arrest and apoptosis in GBM cells. BKM120, well-tolerated and permeable to the BBB, is currently the most frequently-used PI3K inhibitor in
GBM clinical trials (75, 76). XL147, also known as pilaralisib, is an orally bioavailable and reversible pan-PI3K inhibitor that has entered phase I/II clinical trials (77, 78). PX-866 is an irreversible inhibitor and recent studies have shown
PX-866 to promote pro-autophagic, anti-invasive, and anti-angiogenic activities in
GBM cells (79). Finally, GDC-0941, a derivative of a dual PI3K/mTOR inhibitor, shows sustained and significant inhibition of AKT phosphorylation and tumor growth. However, additional studies have demonstrated that GDC-0941 can barely penetrate through the intact BBB suggesting that it may be difficult for
GDC-0941 to reach distant parts of GBM tumors (80, 81). Despite promising in vitro experimental and in vivo preclinical data demonstrating effectiveness with
PI3K inhibitors, clinical data shows that the therapeutic effect of PI3K inhibitors
34 on GBM patients are not reaching expectations suggesting that targeting PI3K alone is not sufficient to treat GBM (67).
AKT inhibitors are also being explored as potential new therapeutics to treat GBM. Thus far, one of the most promising AKT inhibitors, perifosine, is being clinically tested in different cancers. Perifosine inhibits AKT activity by preventing its translocation to the plasma membrane (66). However, in a phase II clinical trial utilizing perifosine in recurrent GBM patients, perifosine was found to be tolerable but ineffective as a monotherapy, possibly due to its limited ability to penetrate the BBB. Therefore, future studies plan to look at perifosine in combination with other therapeutic approaches (66, 82).
Despite more and more PI3K/AKT targeted drugs being developed, targeted therapy for GBM has yet to demonstrate a considerable clinical survival benefit. Some possible reasons for the limited therapeutic effect include the BBB, the heterogeneity of GBM, and the activation of alternative pathways. A future therapeutic possibility for effective GBM therapy is the combination of multiple targeted therapy, thereby eliminating the effects of signaling cross-talk, and personalized therapy (66, 83).
SAT1
Spermidine/spermine N1-acetyltransferase, SAT1, is an imperative rate limiting catabolic enzyme in polyamine metabolism. Polyamines, small positively charged molecules that are present in millimolar amounts in cells, regulate functions such as replication, translation, and chromatin condensation.
Polyamines are essential for life and facilitators of cell death and are, therefore,
35 tightly regulated through synthesis, uptake, and degradation/secretion to maintain homeostatic levels (84, 85). SAT1 acetylates the aminopropyl ends of spermidine and spermine to reduce the charge on polyamines, thus altering their ability to bind acidic macromolecules and influencing their function.
SAT1, a highly regulated enzyme that is induced by elevated polyamines, plays a key role in maintaining polyamine homeostasis and influencing cellular processes. These processes include normal and neoplastic growth related to polyamine content, obesity/glucose tolerance, integrin function, stress response, and oxygen homeostasis. SAT1 has also been implicated in multiple disease states, including pancreatic cell death, blocking regenerative tissue growth, obesity and diabetes, mental health changes leading to suicide, cell migration through interaction with α9β1 integrin, gene regulation, and carcinogenesis (84).
In GBM, SAT1 can interact with α9β1 to affect cell migration (86) and is associated with radioresistance (37). SAT1 was found to be overexpressed in
GBM tumors compared to normal brain tissues and depleting SAT1 radiosensitized GBM tumors in vivo. SAT1 also promotes homologous recombination, controlling BRCA1 expression by regulating histone H3 acetylation (37).
Statement of Purpose
The studies highlighted herein were designed to assess the roles of radiation resistance facilitators PTGFRN, CD9, and SAT1 in GBM and to determine how these effects are mediated. The purpose of these studies was to
36 address the following: (i) what are the mechanisms by which PTGFRN and CD9 mediate radiation resistance and (ii) in what way does SAT1 control genes that regulate radiation resistance and tumor aggressiveness.
First, we identified PTGFRN, CD9, and SAT1 through a negative selection and functional knockdown screen by identifying loss of the shRNA barcodes corresponding to PTGFRN, CD9 and SAT1 in the genomic DNA of irradiated samples compared to non-irradiated samples. Initial screenings suggested that
PTGFRN, CD9, and SAT1 expression is upregulated in GBM tumor samples compared to normal brain and that the overexpression correlates with poor patient survival. Moreover, we show that these proteins promote radiation resistance in in vitro and in vivo assays. Individually, we described a novel role for PTGFRN as a promoter of GBM tumorigenesis and a potential signaling hub for GBM aggressiveness. We show that inhibiting PTGFRN expression leads to decreased tumor growth and increased radiation sensitization. Mechanistically, we find that PTGFRN promotes AKT signaling by stabilizing PI3K p110β at the membrane. Further, PTGFRN inhibition decreases nuclear PI3K p110β leading to decreased DNA damage sensing and DNA damage repair (Chapter 2).
Next, we investigated the role of CD9 in promoting resistance to radiation treatment. We find that decreasing CD9 expression radiosensitizes GBM cells as well as decreases cell proliferation and tumor growth. We further determined that
CD9 inhibition decreases the number of γH2AX foci following radiation treatment and hinders DNA damage repair. Mechanistically, CD9 is found to modulate AKT signaling as depleting CD9 decreases basal phospho-AKT protein levels. These
37 data demonstrate that CD9, a binding partner of PTGFRN, may play a role in
GBM tumorigenesis independent of PTGFRN (Chapter 3).
Finally, we examined SAT1 as a mediator of GBM stemness and aggressiveness by acting as a transcriptional regulator. Here, we showed that
SAT1 promotes the expression of DNA damage response pathways as well as cell cycle regulatory genes. Further, we find that SAT1 is able to regulate its target genes by physically associating with the promoters of transcription factors,
MELK and EZH2, and demonstrate the essentiality of the polyamine acetyltransferase activity of SAT1 for this regulation. These data suggest that
SAT1 is a critical mediator of GBM stemness and aggressiveness via transcriptional regulation of MELK and EZH2 (Chapter 4).
Altogether, these data highlight the critical roles for PTGFRN, CD9, and
SAT1 in GBM tumorigenesis and aggressiveness. Furthermore, the work presented herein demonstrates the vastness of molecular pathways promoting radiation resistance and suggest the potential value of PTGFRN, CD9, and SAT1 as therapeutics for the treatment of GBM.
38 Chapter 2 - The Ig superfamily protein PTGFRN coordinates survival signaling in Glioblastoma multiforme
The work in this chapter is originally published in Cancer Lett. Vol 462. Aguila B, Morris AB, Spina R, Bar E, Schraner J, Vinkler R, Sohn JW, Welford SM. “The Ig superfamily protein PTGFRN coordinates survival signaling in glioblastoma multiforme.” 462:33-42.
Copyright 2019 with permission from Elsevier
39 Abstract
Glioblastoma multiforme (GBM) is the most malignant primary brain tumor with a median survival of approximately 14 months. Despite aggressive treatment of surgical resection, chemotherapy and radiation therapy, only 3-5% of GBM patients survive more than 3 years. Contributing to this poor therapeutic response, it is believed that GBM contains both intrinsic and acquired mechanisms of resistance, including resistance to radiation therapy. In order to define novel mediators of radiation resistance, we conducted a functional knockdown screen, and identified the immunoglobulin superfamily protein,
PTGFRN. In GBM, PTGFRN is found to be overexpressed and to correlate with poor survival. Reducing PTGFRN expression radiosensitizes GBM cells and potently decreases the rate of cell proliferation and tumor growth. Further,
PTGFRN inhibition results in significant reduction of PI3K p110β and phosphorylated AKT, due to instability of p110β. Additionally, PTGFRN inhibition decreases nuclear p110β leading to decreased DNA damage sensing and DNA damage repair. Therefore overexpression of PTGFRN in glioblastoma promotes
AKT-driven survival signaling and tumor growth, as well as increased DNA repair signaling. These findings suggest PTGFRN is a potential signaling hub for aggressiveness in GBM.
40 Introduction
Glioblastoma multiforme (GBM) is a common and aggressive primary malignant brain tumor, and has one of the worst 5-year survival rates among all cancers despite aggressive treatment (12). Current therapies, including surgical resection followed by concurrent treatment with temozolomide and radiation therapy, have demonstrated limited effectiveness due to the anatomic location of
GBM, the blood brain barrier, and intrinsic and acquired mechanisms of resistance (6, 7, 16). Notably, GBM has been shown to exhibit radioresistance, especially in recurrent tumors. Despite identification of several mechanisms that contribute to the radioresistant phenotype (87-89), the molecular basis of radioresistance remains incompletely defined. Through a recently published functional knockdown screen (37), we identified PTGFRN, Prostaglandin F2 receptor negative regulator, as a novel mediator of radioresistance in GBM cells.
PTGFRN (EWI-F, FPRP, CD9P-1) is a cell surface transmembrane protein in the immunoglobulin superfamily (54). To date, the role of PTGFRN has been little studied in cancer. Published studies show that PTGFRN expression is essential for angiogenesis, and its expression at both transcriptional and translational levels correlates with the metastatic status of lung cancer (51, 52).
PTGFRN has also been shown to be associated with lipid accumulation in preadipocytes (53), involved in cell migration (90), and has been found to directly interact with ezrin-radixin-moesin proteins (54). PTGFRN self-associates but is also a component of tetraspanin-enriched microdomains (TEMs), which can include tetraspanins, growth factors, integrins, complement regulatory proteins,
41 signaling enzymes and proteoglycans (39, 54, 91). As a member of TEMs,
PTGFRN is primed to act as a scaffolding protein and associate with other proteins within the TEM to regulate downstream signaling events.
The AKT/PI3K signaling pathway is critical survival pathway in cells. PI3K contains a catalytic subunit p110 (α, β, and γ) and a regulatory subunit p85. Upon ligand binding to cell surface receptor tyrosine kinases, the phosphorylation of tyrosine residues creates binding sites for p85, resulting in a conformation change that releases the catalytic subunit p110. Activated p110 phosphorylates phosphatidy-linositol-3,4-bisphosphate (PIP2) into phosphatidy-linositol-3,4,5- bisphosphate (PIP3). PIP3 then recruits downstream AKT to the inner membrane for phosphorylation and activation (66). Studies have demonstrated increased activation of the PI3K/AKT pathway in GBM, among several other cancers, by a variety of mechanisms including homozygous deletion of the negative regulator of PI3K, PTEN (92). Crucially, pathway activation is significantly associated with radiation resistance (93, 94). Additionally, studies have also shown PI3K/AKT signaling to be altered by proteins found in TEMs, suggesting that multiple mechanisms exist for cancer cells to employ (95, 96).
In the present work, we demonstrate that PTGFRN supports tumorigenesis and survival signaling through increased PI3K/AKT signaling and p110β stability. We find PTGFRN to be elevated in primary GBM compared to normal brain tissue, and to significantly associate with poorer survival among patients under standard treatment paradigms. Our results indicate the inhibition of PTGFRN decreases tumor growth and sensitizes GBM cells to radiation
42 treatment by affecting the turnover of the PI3K catalytic subunit p110β, hindering
AKT signaling and DNA damage sensing. The findings support a novel role for
PTGFRN in GBM tumorigenesis and survival signaling, and provide a novel therapeutic target expressed on the cell surface.
Results PTGFRN is overexpressed in GBM and correlates with poorer patient outcome
To identify novel mediators of radiation resistance in GBM, we performed a functional knockdown screen using a lentivirus-mediated shRNA library on two
GBM cell lines (37) and filtered the expression levels of hits through in silico screens and patient data to determine relevance to GBM. In Oncomine, a public database of published gene expression studies (97), PTGFRN expression was found to be elevated in every study of GBM tumors versus normal brain tissue
(Fig 2-1A). In the Bredel, Liang, Sun and Murat studies of GBM, PTGFRN was overexpressed by factors of 2.443 (P=0.0014), 2.513 (P=0.0011), 2.804
(P<0.0001) and 3.032 (P<0.0001), respectively. We next looked at GBM patient survival data to evaluate its effect by PTGFRN expression. We inquired
Prognoscan, a public database for the meta-analysis of the prognostic value of genes (98), and found that GBM patients with elevated PTGFRN expression have a poorer outcome than patients with lower expression (15.66 months versus 8.67 months, P=0.0008, log-rank test) (Fig 2-1B). We also queried TCGA database where PTGFRN expression significantly correlates with poor outcome
43 in the low grade glioma and GBM cohort (Fig 2-1C). Additionally, PTGFRN was found to be significantly overexpressed in high grade GBM compared to subtypes of lower grade gliomas, including astrocytoma and oligodendroglioma
(Fig 2-1D). Further, we queried TCGA to determine if known oncogenic mutations, such as PTEN or PI3K mutations, found in GBM tumors correlate with
PTGFRN expression. We found no correlation of PTGFRN expression with the mutations analyzed (Fig 2-2A). Finally, as GBM has been classified into four major molecular subtypes: classical, mesenchymal, proneural, and neural (14), we queried the GBM Bio Discovery Portal to determine how PTGFRN expression correlates with the molecular subtypes. PTGFRN expression was found to be higher in classical and mesenchymal GBM subtypes compared to proneural and neural subtypes (Fig 2-2B). While a complete analysis of therapeutic response across the molecular subtypes is limited, literature has described mesenchymal
GBM tumors to exhibit increased resistance to therapy resulting in a worse patient survival outcome compared to proneural GBM subtype tumors which have a better prognosis (99). Together, these data suggest that PTGFRN is a novel mediator of radiation resistance and may be pro-tumorigenic in GBM tumors and predictive of poor prognosis, independent of common tumor mutation status.
PTGFRN promotes cell proliferation and tumor growth
Next, we wished to determine the phenotype of PTGFRN in GBM cells by investigating the effects upon PTGFRN depletion. First, we performed cell proliferation assays using A172 and U87MG GBM cell lines utilizing two different
44 shRNAs to target PTGFRN. We found that reduction of PTGFRN expression significantly reduced cell proliferation in vitro (Fig 2-3A, B, C, D). Additionally, to access growth in vivo, we performed a subcutaneous tumor assay using
PTGFRN knockdown U87MG cells. Tumors were measured twice weekly until the tumors reached greater than 1.5 cm3. Depleting PTGFRN drastically reduced tumor growth and increased mouse survival compared to shGFP control mice
(Fig 2-3 E, F).
Further, as primary brain tumors are thought to be maintained by self- renewing, tumorigenic cells (100-103), we also utilized neurosphere stem cell lines (100). An intracranial xenograft was performed after PTGFRN shRNA transduction into luciferase expressing GBM0913 neurosphere cells. Upon
PTGFRN depletion, as determined by qRT-PCR, cells were injected into the brains of nude mice and followed by bioluminescent imaging. We observed that similar to the established cell lines data, mice containing shPTGFRN neurospheres experienced decreased tumor growth rate (Fig 2-3G) and extended overall survival (Fig 2-3H) compared to control shGFP neurospheres.
Additionally, to determine whether the eventual tumor growth in the PTGFRN knockdown U87 subcutaneous and GBM0913 brain tumors correlated with regained expression of PTGFRN, tumors were excised and subjected to qRT-
PCR. We found that the expression of PTGFRN was returning in the U87 cells
(Fig 2-3I) and had returned in the GBM0913 cells (Fig 2-3J), compared to the expression of the cells injected suggesting a selective pressure exists for the expression of PTGFRN in the tumors.
45 Next, we wished to determine the cause of the significant decrease in cell proliferation seen in the GBM cell lines and GBM0913 neurospheres (Fig 2-4A,
B). In initial analyses of the cell proliferation, PTGFRN depletion increased the cell doubling time by roughly 20 hours compared to shGFP control cells (Fig 2-
4C). Also, upon performing an EdU pulse chase on GBM0913 shPTGFRN and shGFP neurosphere cells, we found that shPTGFRN cells are slower to progress through G2/M phases of the cell cycle compared to shGFP cells, as G1 peaks at
38 hours after release from growth factor starvation in shPTGFRN cells compared to shGFP cells where G1 peaks at 24 hours after release (Fig 2-4D).
These data suggest that the cause of decreased cell proliferation after depleting
PTGFRN is due to an extension of cell cycle progression. Finally, as tumor stem cells are defined by their ability to self-renew, we wished to determine if PTGFRN affects self-renewal (104). We performed sphere formation assays using
PTGFRN depleted GBM0821 and GBM0913 cells, and found no change in the number of spheres formed between shGFP and shPTGFRN neurospheres (Fig
2-4E), but did find the colonies to be smaller (Fig 2-4F), suggesting that PTGFRN does not affect the capability of the neurospheres to self-renew. Therefore, these results demonstrate that elevated expression of PTGFRN promotes cell proliferation and tumor growth by decreasing the time of cell cycle progression.
PTGFRN inhibition sensitizes GBM cells to radiation
We next sought to verify the functional knockdown screen and determine whether PTGFRN reduction can sensitize GBM cells to radiation. Clonogenic survival assays were performed with varying doses of ionizing radiation (IR)
46 targeting PTGFRN with two shRNAs in multiple GBM cell lines. qRT-PCR analysis was used to verify knockdown efficiency (Fig 2-5). Both shRNAs, shPTGFRN (1) and shPTGFRN (2), sensitized A172 cells, with dose enhancement factors at 10% survival of 1.67 and 1.4 respectively (Fig 2-6A).
Additionally, we observed that the depletion of PTGFRN also sensitized U87MG cells by a factor of 1.79 and 1.2 (Fig 2-6B) (10% survival). Further, in GBM neurosphere lines, PTGFRN knockdown sensitized GBM0821 (Fig 2-6C) by a factor of 1.77 and 3.28 (50% survival) and in GBM0913 (Fig 2-6D) by 1.45 (35% survival) and 1.52 (45% survival). Notably, the neurosphere lines are significantly more radioresistant than the established cell lines (87). Together, these results show that inhibiting PTGFRN is capable of sensitizing multiple GBM tumor lines to radiation.
To specifically analyze radiosensitization in vivo upon PTGFRN depletion without the previously observed decreased cell proliferation as a variable, we utilized an inducible Tet-On shRNA tumor model. GBM0913 neurosphere stem cells, which overexpress the Tet repressor, were infected with lentiviruses carrying a doxycycline-inducible pLKO-Tet-On-shRNA (either shGFP or shPTGFRN) construct and implanted intracranially into nude mice. We validated the effect of doxycycline on expression of PTGFRN in vitro (Fig 2-6E) and on six animals in vivo and saw a significant reduction in PTGFRN expression (Fig 2-
6F). To test radiosensitization, the animals received either GBM0913 Tet-On- shGFP cells or GBM0913 Tet-On-shPTGFRN cells. Two weeks after implantation, tumors were measured by bioluminescence and mice were
47 randomized into two treatment groups: doxycycline, or doxycycline plus 12 Gy
IR. Radiation was delivered using the GammaKnife clinical irradiator as a single dose (105), and doxycycline was given for 6 days to obtain gene silencing during
IR and then removed. As expected, control mice injected with GBM0913 Tet-On- shGFP and Tet-On-shPTGFRN showed no difference in overall survival (median survival: 64 days versus 69 days, respectively), given that PTGFRN was silenced for only 6 days. With the addition of IR, mice containing GBM0913 Tet-On- shPTGFRN neurosphere cells survived longer than the control, GBM0913 Tet-
On-shGFP with IR (median survival: 100 days versus 76 days, respectively,
P=0.0304) (Fig 2-6G). In addition, three of animals were cured and sacrificed with no evidence of tumors at 150 days. Together, the data suggest that
PTGFRN expression protects GBM cells from genotoxic IR in vitro as well as in vivo.
PTGFRN is necessary for P-AKT signaling by increasing stability of p110β
The molecular mechanisms of PTGFRN have been little studied. We found that PTGFRN inhibition decreases cell growth and sensitizes GBM cells to radiation, and as the PI3K/AKT pathway is a major survival signaling pathway found to be altered in GBM (68, 93), we hypothesized that PTGFRN modulates
AKT signaling. We thus performed western blots to examine the levels of phospho-AKT and total AKT in PTGFRN-depleted GBM cells. We found that
PTGFRN knockdown decreased basal phospho-AKT S473 levels compared to shGFP control cells in U87MG cells and both of the neurosphere lines (Fig 2-7A-
C). Upstream of AKT, Phosphatidyl inositol 3 kinase (PI3K) is a major mediator of
48 AKT phosphorylation, and we therefore assessed expression of the p110 and p85 subunits. We noted that PTGFRN knockdown-cells displayed decreased p110β protein levels compared to control, while p85α and p110α levels remained unchanged (Fig 2-7D, E, Fig 2-8A,B). We next wished to see if we could rescue the decreased phospho-AKT by overexpressing p110β following PTGFRN depletion. Upon PTGFRN knockdown, we saw decreased p110β and phospho-
AKT protein levels compared with shGFP control. Following overexpression of p110β (PIK3CB) utilizing an overexpression plasmid in the shPTGFRN cells, we noted increased levels of phospho-AKT compared to both shPTGFRN and shGFP (Fig 2-8C). Additionally, overexpression of p110β following PTGFRN knockdown was able to partially rescue cell viability and proliferation (Fig 2-8D), further supporting the proposed mechanism of PTGFRN signaling.
As altered protein levels could be due to corresponding mRNA levels, we determined by qRT-PCR that p110β mRNA levels were unaffected upon knockdown of PTGFRN (Fig 2-7F). With mRNA levels unaltered, we questioned whether the decreased p110β protein could be due to increased turnover. We treated shPTGFRN and shGFP cells with cycloheximide to block translation and collected the cell lysate at multiple time points. Over 24 hours, we observed in shPTGFRN cells that p110β is less stable compared to shGFP control (Fig 2-
7G). Further, we asked whether treatment with a proteasome inhibitor, MG132, could rescue p110β degradation in shPTGFRN cells. Cells were treated with cycloheximide in combination with MG132 and collected at specified time points
(Fig 2-7H). The amount of degradation seen at 24 hours was rescued in
49 shPTGFRN cells as well as shGFP control cells compared to cells treated with cycloheximide alone. The data thus suggest that PTGFRN plays a role in p110β stability.
Since both PTGFRN and p110β can be found at the plasma membrane and PTGFRN affects p110β turnover, we asked if PTGFRN is in close proximity with p110β. By performing a proximity ligation assay (PLA) in GBM0913 shGFP and shPTGFRN neurosphere cells, we found that PTGFRN and p110β are within close proximity to each other, as upon PTGFRN inhibition, the Texas Red signal decreased two-fold to levels similar to the negative control (Fig 2-7I). The close proximity of PTGFRN and p110β was further supported by evaluating the co- localization of PTGFRN and p110β by confocal imaging in GBM0913 shGFP and shPTGFRN cells (Fig 2-8E). In GBM0913 shPTGFRN cells, the fluorescence intensity of both PTGFRN and p110β protein significantly decreased compared to shGFP (Fig 2-8F). Furthermore, we determined that the majority of p110β co- localized with PTGFRN. Interestingly, not all of PTGFRN co-localized with p110β
(Fig 2-8G), suggesting that PTGFRN may have other roles in the cell.
Additionally, we hypothesized that if PTGFRN is necessary for p110β stability, then p110β expression should be decreased not just at the membrane, but also in the cytosol and nucleus of shPTGFRN GBM cells. When we fractionated proteins from shGFP and shPTGFRN neurosphere cells, we observed decreased levels of p110β at the membrane and in the cytosol and nucleus of shPTGFRN cells compared to control cells. Additionally, we found decreased phospho-AKT protein levels in the membrane and cytosol fractions
50 suggesting that the reduced phospho-AKT may be due to the decreased availability of p110β to catalyze PIP2 to PIP3, allowing for AKT to translocate to the membrane for activation (Fig 2-7J). With these results, we posit that
PTGFRN, acting as a scaffolding protein (54, 57), is important for p110β stability allowing for downstream AKT signaling.
PTGFRN promotes DNA damage sensing through p110β
Radiation therapy is a major component of the treatment for GBM patients, and the main deleterious damage caused by IR is DNA double strand- breaks (DSBs) (106). Our results demonstrate that depleting PTGFRN radiosensitizes GBM cells in vitro and in vivo, which led us to question how
PTGFRN depletion affects DNA damage repair. Interestingly, the literature has described a necessity for nuclear p110β in sensing DSBs and its depletion leads to a defect in DNA repair activation (107). Our cellular fractionation results also found decreased p110β in the nucleus (Fig 2-7J) suggesting that PTGFRN depletion could impair DSB sensing through p110β. Therefore, we irradiated
GBM0913 shRNA cells with 10 Gy and collected the cell lysates after 30 minutes to evaluate γH2AX activation, a canonical DSB marker, and DNA-PKcs activation, a canonical non-homologous end joining (NHEJ) factor. Upon
PTGFRN depletion, we find decreased γH2AX and phospho-DNA-PKcs following
IR compared to shGFP control (Fig 2-9A). Additionally, we irradiated U87 shRNA cells and performed immunofluorescence to analyze formation of γH2AX foci.
Similarly, we found decreased amounts of γH2AX foci in shPTGFRN cells
51 compared to shGFP 30 min post IR suggesting a decrease in DSB recognition
(Fig 2-9B, C). To confirm that DNA damage repair was also hindered by
PTGFRN depletion, we performed a homologous recombination (HR) DNA repair assay using a DR-GFP reporter in U87MG cells. Our results show that decreasing PTGFRN expression (Fig 2-9D) also hindered the efficiency of HR
DNA repair (Fig 2-9E), possibly due to decreased DSB sensing. Altogether, the data suggest that PTGFRN promotes DNA damage DSB sensing through p110β signaling.
Discussion
In the present study, we have defined a novel mediator of cell proliferation and radioprotection in GBM that associates with poor patient outcome. PTGFRN is overexpressed in GBM tumor samples compared to normal brain, as assessed bioinformatically in Oncomine and TCGA databases. Depletion of PTGFRN decreased cell proliferation and radiosensitized multiple cell lines, including neurosphere stem cell lines, and tumors in mice. Mechanistically, we found that depletion of PTGFRN decreases phospho-AKT signaling due to decreased stability of p110β at the cell membrane. Additionally, we found decreased nuclear p110β upon depletion of PTGFN, which decreased DNA damage sensing and reduced DNA repair. Together, our findings highlight a novel function for
PTGFRN as a scaffolding protein in regulating survival signaling responses in tumors, and identify a potential therapeutic target.
52 Our results indicated that PTGFRN depletion leads to decreased cell proliferation and basal phospho-AKT protein levels, in agreement with other published studies completed in GBM cells and tumors (108, 109). In this study, the effects are at least partially due to the reduction of p110β at the membrane of shPTGFRN cells. Furthermore, we observed that PTGFRN and p110β are within close proximity, and shPTGFRN cells had more p110β degradation after 24 hours compared to control, thus suggesting that PTGFRN impacts p110β turnover. It is known that p110β is stabilized by dimerization with p85, which also regulates p110β catalytic activity (110). Our results thus suggest that tethering p110b to the membrane is the mechanistic reason for promoting stabilization and signaling.
At the same time, other functions for p110β have been discovered that relate to radioresistance. In the nucleus, p110β has been found to be involved in
DNA damage sensing and cell survival by regulating recruitment of DNA damage proteins to DSB foci (107, 111). We found decreased p110β in the nuclear fraction of the shPTGFRN GBM cells compared to shGFP cells, suggesting that decreased p110β could also explain the observed radiation sensitivity upon
PTGFRN depletion. After DNA damage by radiation, PTGFRN depleted cells have reduced DNA damage sensing and delayed protein recruitment to the site of damage as demonstrated by the reduced activation of γH2AX and DNA-PKcs in shPTGFRN cells 30 minutes post IR. Furthermore, a decrease in DNA damage sensing could cause decreased DNA repair, as cells fail to detect the
53 initial damage, agreeing with the decrease in HR repair efficiency seen in shPTGFRN GBM cells.
Targeting PTGFRN could be beneficial in decreasing cell survival signaling in GBM tumors. Thus far, a truncated form of PTGFRN, GS-168AT2, was developed and found to function as a dominant negative protein that interferes with normal PTGFRN extracellular interactions. GS-168AT2 decreased tumor growth in vivo in lung xenograft mice. The authors suggest that
GS-168AT2 acts through preventing interactions with CD9 and CD151, two tetraspanin proteins implicated in tumorigenesis (51, 52, 58). Interestingly, CD9 has been found to promote tumor growth in GBM by stabilizing IL-6 receptor glycoprotein 130 from degradation (65), analogous to PTGFRN stabilizing p110β to promote survival signaling. Further, CD9 is a partner protein of PTGFRN (47) suggesting that CD9 and PTGFRN could regulate signaling as a complex, as well as individually. While not a focus of the present study, future analyses on the interaction between PTGFRN and CD9 in GBM and their possible dual implications to cell signaling would be attractive as both proteins, individually, seem to regulate turnover of signaling proteins.
Various PI3K inhibitors have been developed and studied in the last few decades since hyper-activation of the PI3K/AKT pathway confers rapid growth, tumor progression and multidrug resistance upon GBM cells. The first generation of pan-PI3K inhibitors, wortmannin and LY294002, are unable to be used clinically due to their high toxicities. The current PI3K inhibitors, which have improved safety, efficacy and pharmacokinetics, in clinical trials include BKM120,
54 XL147, and PX-866 (67). BKM120 is a pan-PI3K inhibitor that impedes intracerebral U87 GBM cell xenograft growth in preclinical studies and inhibits
AKT phosphorylation (75, 76). Currently, BKM120, which is well-tolerated and permeable to the blood-brain barrier, is the most-frequently used PI3K inhibitor in
GBM clinical trials (67). Another promising PI3K inhibitor is Pilaralisib (XL147) that exhibits dose dependent decrease in AKT phosphorylation, pS6K1, and Ki67 expression, suggesting moderate blood brain barrier penetration and inhibition of proliferation (77, 78). Finally, PX-866 is an irreversible wortmannin analogue that inhibits intracranial xenograft growth by selectively blocking the PI3K/AKT pathway (79). As depleting PTGFRN decreases AKT phosphorylation and tumor growth, similar to the above PI3K inhibitors, targeting PTGFRN, which is positioned on the cell surface, has the potential to be beneficial to GBM patients.
In summary, we have defined a novel function for PTGFRN in mediating survival signaling in brain tumors through regulation of p110β. Our findings describe a mechanism for PTGFRN to act as a scaffolding protein and regulate p110β protein stability, thus affecting cell proliferation and tumor growth, along with regulation of DNA DSB recognition and repair. Together, these findings contribute to our understanding of PTGFRN and the impact of scaffolding protein depletion on downstream signaling and suggest that inhibition of PTGFRN may prove to be a beneficial therapeutic target by affecting both cell proliferation and survival.
55 Figure 2-1: Relevance of PTGFRN to GBM A, Oncomine data of PTGFRN expression in brain tumors versus normal brain tissue in four studies. B, PrognoScan survival curves of GBM patients whose tumors express high versus low levels of PTGFRN. C, From TCGA LGG+GBM database, Kaplan Maier survival plot for low and high grade gliomas dichotomized by PTGFRN expression. D, The expression of PTGFRN in low and high grade gliomas from the TCGA database. Pairwise statistical comparisons of each subtype of low grade glioma to GBM are indicated (student’s t-tests).
56 Figure 2-1:
57 Figure 2-2: PTGFRN expression in GBM tumors with specific mutation statuses and GBM molecular subtypes A, TCGA data of PTGFRN expression in GBM tumors with PTEN, p53, PIK3CA, or PIK3R1 mutations. B, PTGFRN expression in the GBM molecular subtypes from the TCGA GBM BioDiscovery Portal. ***, p<0.001; ****, p<0.0001 by one- way ANOVA.
58 Figure 2-2:
59 Figure 2-3: PTGFRN is required for cell proliferation and tumor growth A, C, Cell proliferation assay in A172MG and U87MG GBM knockdown cells. B,
D, qRT-PCR for samples in A and C. E, Subcutaneous U87 shRNA tumor growth curve. F, Kaplan-Meier curve for mice in E. G, Bioluminescence of intracranial tumor growth of shRNA neurospheres. H, Kaplan-Meier curve of mice in G. I,
PTGFRN expression from U87 cells before injection and from U87 subcutaneous tumors. J, PTGFRN expression from GBM0913 cells before injection and from
GBM0913 intracranial tumors. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001 by two-way ANOVA (A, B, C, D, E, G, I, J) or log rank (F, H).
60 Figure 2-3:
61 Figure 2-4: PTGFRN depletion extends cell cycle progression A, qRT-PCR of GBM0913 cells in panel B. B, Cell proliferation assay of
GBM0913 shRNA cells. C, Cell doubling time of GBM0913 shRNA cells. D,
Histograms of EdU+ GBM0913 shRNA cells at designated time points. E,
Quantification of sphere formation assay in GBM0821 and GBM0913 shRNA cells. F, Representative images of neurosphere stem cell colonies following
PTGFRN knockdown and shGFP control *, p<0.05; **, p<0.01; ***, p<0.001, ****, p<0.0001 by one-way ANOVA (A, C) or student’s t-test (B).
62 Figure 2-4:
63 Figure 2-5: shRNA knockdown confirmation in GBM cell lines A-D, qRT-PCR of GBM shRNA cells used in clonogenic assays (Figure 3A-D).
***, p<0.001, ****, p<0.0001 by one-way ANOVA.
64 Figure 2-5:
65 Figure 2-6: PTGFRN reduction sensitizes GBM cells to IR Clonogenic assays of A172 (A), U87MG (B), GBM0821 (C) and GBM0913 neurospheres (D) cells after indicated doses of IR. qRT-PCR of expression normalized to GAPDH can be found in supplemental figure 3. E, qRTPCR of
GBM0913 Tet-on-shRNA cells with or without doxycycline in vitro. F, qRT-PCR of
PTGFRN expression in animals injected with GBM0913 Tet-On-shRNA cells treated with doxycycline. G, Kaplan-Meier survival curve of GBM0913 Tet-on- shRNA intracranial xenograft.*, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001 by student’s t tests (A, B, C, D), one or two-way ANOVA (E, F) or log rank (G).
66 Figure 2-6:
67 Figure 2-7: PTGFRN depletion decreases p110β levels A, B, C, Western blot evaluating basal P-AKT signaling in different GBM shPTGFRN cells compared to shGFP control. D, E, Western blot assessing p110β protein levels in GBM0821 and GBM0913 neurosphere cell lines. F, qRT-
PCR for p85α and p110β in 3 different GBM shPTGFRN cell lines. Expression normalized to GAPDH. ****, p<0.0001 by two-way ANOVA. G, H, Immunoblot analyses of GBM0913 shRNA cells treated with cycloheximide alone (G) or in combination with proteasome inhibitor, MG132 (H). I, Proximity ligation assay of
GBM0913 shPTGFRN or shGFP cells; negative control contained no primary antibody; positive control comprised two different antibodies targeting PTGFRN; * p<0.05 by student’s t test (above). Immunoblot of cells used for the PLA assay
(below). J, Cellular fractionation assay in GBM0913 shRNA neurosphere cells using α-tubulin, Histone H3 and Na/K-ATPase as cellular fractionation internal and loading controls.
68 Figure 2-7:
69 Figure 2-8: PTGFRN depletion does not affect p110α protein levels and PTGFRN co-localizes with p110β
A, B, Western blot evaluating p110α protein levels in GBM0821 and GBM0913 neurosphere cell lines. C, Western blot assessing AKT phosphorylation upon
PTGFRN knockdown followed by PIK3CB (p110β) overexpression. D, MTT assay of U87 cells with shPTGFRN and shPTGFRN+ p110β overexpression compared to shGFP control cells. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001 by student’s t-test. E, Confocal imaging evaluating co-localization of
PTGFRN and p110β after immunofluorescent staining. F, Fluorescence intensity of cells in E. **, p<0.01; ****, p<0.0001 by one-way ANOVA. G, Colocalization of p110beta with PTGFRN (p<0.0001) and PTGFRN with p110beta (p=0.004) of cells in E by two-tailed t-test of 30 cells.
70 Figure 2-8:
71 Figure 2-9: PTGFRN promotes DNA damage sensing A, Western blot of GBM0913 shRNA cells that received either no treatment (NT) or radiation (10 Gy) and collected after 30 minutes. B, Quantification of γH2AX foci following IR (2 Gy) in U87 shRNA cells. C, Immunofluorescence of γH2AX
(red) foci. D, qRTPCR of targeted genes in HR assay. Expression normalized to
β-actin. E, DR-GFP HR reporter assay in siPTGFRN, siGAPDH (negative control), siBRCA1 (positive control), or siRISC (control) U87 cells. ****, p<0.0001 by one-way ANOVA.
72 Figure 2-9:
73
Chapter 3 - Tetraspanin CD9 promotes tumorigenesis and radiation resistance in Glioblastoma multiforme
The work in this chapter is unpublished.
74
Abstract
Glioblastoma multiforme (GBM) is the most malignant primary brain tumor.
Despite aggressive treatment of surgical resection and radiochemotherapy, the median survival is approximately 14 months with only 3-5% of GBM patients surviving more than 3 years. Contributing to this poor therapeutic response, it is believed that GBM contains intrinsic and acquired mechanisms of resistance, including resistance to radiation therapy. In order to define novel mediators of radiation resistance, we conducted a functional knockdown screen, and identified the tetraspanin protein, CD9. In GBM, CD9 is found to be overexpressed and to correlate with poor survival. Additionally, reducing CD9 expression decreases the rate of cell proliferation and tumor growth, as well as radiosensitizes GBM cells.
Further, CD9 inhibition reduces phosphorylated AKT, and also reduces γH2AX foci formation correlating with decreased DNA repair efficiency. These findings suggest CD9 is critical for tumorigenesis and radioresistance in GBM and a potential therapeutic target.
75 Introduction
Glioblastoma multiforme (GBM) is a common and aggressive primary malignant brain tumor, and despite aggressive treatment, GBM has one of the worst 5-year survival rates among all cancers (6, 12). Current therapies have exhibited limited efficiency due to the anatomic location of GBM, the blood brain barrier, and intrinsic and acquired mechanisms of resistance (6, 7, 16). Markedly,
GBM has shown to exhibit radioresistance, and despite identification of several mechanisms that contribute to the radioresistant phenotype (87-89, 112), the molecular basis of radioresistance remains incompletely described. Through a recently published functional knockdown screen (37), we identified CD9 as a novel mediator of radioresistance in GBM cells.
CD9 is a member of the four-transmembrane tetraspanin superfamily.
Tetraspanins have been described to organize laterally with other membrane proteins and themselves to form tetraspanin-enriched microdomains (TEMs)
(38). Tetraspanins, collectively, act as scaffolding proteins and associate with membrane signaling enzymes to influence functions relevant in different stages of tumor development including cell migration, invasion, cell-cell fusion, survival and signaling (38, 57, 58). In cancer, CD9 is context dependent molecule. In gastric and ovarian cancers, CD9 expression is tumor promoting, leading to increased disease severity (58). However, in breast, prostate and lung cancers, increased CD9 expression suppresses tumor progression and low CD9 expression levels associate with poor patient prognosis (59). Particularly in GBM,
CD9 has been described to modulate cell proliferation through multiple signaling
76 pathways, including AKT, MAPK/ERK, and STAT3, and also affect tumor growth
(64, 65, 113).
In the present work, we demonstrate that CD9 supports tumorigenesis and radioresistance through increased PI3K/AKT signaling. We find CD9 to be elevated in GBM tumors compared to normal brain tissue, and to significantly associate with poorer survival among patients under standard treatments. Our results indicate that inhibiting CD9 decreases tumor growth and sensitizes GBM cells to radiation treatment. We also demonstrate that inhibiting CD9 hinders AKT signaling as well as γH2AX foci formation and DNA damage repair. Together, these findings support a novel role for CD9 in GBM tumorigenesis and radioresistance, and provide additional support for CD9 as a potential future therapeutic target.
Results
CD9 is overexpressed in GBM and correlates with poorer outcome
To identify novel mediators of radiation resistance in GBM, we performed a negative selection functional knockdown screen utilizing a lentiviral mediated shRNA library (37) and filtered the expression levels of hits through in silico analysis and patient data to determine relevance of CD9 in GBM. In Oncomine,
CD9 expression was found to be elevated in the Shai GBM study of cancer versus normal tissue by a factor of 3.54 (P=0.0038) (Fig 3-1A). Additionally, in a
Cancer Genome Atlas (TCGA) study of GBM versus normal tissue, CD9 was found to be overexpressed by a factor of 2.34 (P<0.0001) (Fig 3-1B). We next
77 looked at GBM patient survival data to evaluate its effect by CD9 expression by inquiring Prognoscan, a public database for the meta-analysis of the prognostic value of genes. We found that patients with elevated CD9 expression had a poorer outcome compared to patients with lower expression (11.35 months versus 16.38 months, P=0.0072, log-rank test) (Fig 3-1C). We also queried the
TCGA database where CD9 expression was found to significantly correlate with poor outcome in the low grade glioma and GBM cohort (Fig 3-1D). Furthermore,
CD9 was found to be significantly overexpressed in high grade GBM compared to subtypes of lower grade gliomas, including oligodendroglioma and astrocytoma (Fig 3-1E). Finally, as GBM has been classified into four molecular subtypes: neural, proneural, classical, and mesenchymal (14), we queried the
GBM Bio Discovery Portal to determine how CD9 expression correlates with these subtypes. We found that CD9 expression was higher in classical and neural GBM subtypes compared to proneural and mesenchymal subtypes (Fig 3-
1F). Altogether, these data suggest that CD9 is a potential mediator of radiation resistance and pro-tumorigenic in GBM tumors and predictive of poor prognosis.
CD9 promotes cell proliferation and tumor growth in GBM tumors
We next wished to determine the phenotype of CD9 in GBM cells by investigating the effects upon CD9 inhibition. First, we performed cell proliferation assays in A172 and U87MG GBM cell lines utilizing two different shRNAs to target CD9. We found that upon CD9 depletion, GBM cell proliferation was significantly reduced in vitro (Fig 3-2A, B, C, D). Additionally, to assess growth in vivo, we performed a subcutaneous tumor assay using CD9 knockdown U87MG
78 cells. Approximately one million cells were injected into the flank of 8-week-old nude mice and tumors were measured twice weekly until the tumors reached greater than 1.5 cm3. Depleting CD9 drastically reduced tumor growth and increased mouse survival compared to shGFP control mice (Fig 3-2E, F).
Additionally, to determine whether the eventual tumor growth in the CD9 depleted U87MG subcutaneous tumors correlated with regained CD9 expression, tumors were excised and subjected to qRT-PCR. We discovered that
CD9 expression had returned in the U87 tumors compared to the expression of cells initially injected (Fig 3-2G), suggesting a selective pressure exists for the expression of CD9 in the tumors. Finally, we wished to see if we could reduce cell proliferation and cell viability by targeting CD9 with an anti-CD9 antibody,
B2C11. We found that treatment with CD9 antibody at 5 µg/ml and 10 µg/ml was able to significantly reduce cell proliferation and cell viability compared to IgG control antibody and hybridoma media only (control) (Fig 3-2H). Therefore, these results demonstrate that elevated expression of CD9 promotes cell proliferation and tumor growth.
CD9 inhibition sensitizes GBM cells to radiation
We next sought to verify the functional knockdown screen and determine whether CD9 inhibition sensitizes GBM cells to radiation. To determine if depleting CD9 can sensitize GBM cells to radiation, we performed clonogenic survival assays with varying doses of ionizing radiation (IR) targeting CD9 with shRNAs in multiple GBM cell lines. qRT-PCR analysis was used to verify knockdown (insets). Both shRNAs, shCD9 (1) and shCD9 (2), sensitized A172
79 cells, with dose enhancement factors at 10% survival of 1.5 and 1.14, respectively (Fig 3-3A). Moreover, we observed that CD9 depletion also sensitized U87MG cells by a factor of 1.14 (both shRNAs at 10% survival) (Fig 3-
3B). Further, in GBM neurosphere stem cell lines, CD9 knockdown sensitized
GBM0821 (Fig 3-3C) by a factor of 1.25 and 1.2 (10% survival) and in GBM0913
(Fig 3-3D) by 1.2 (40% survival). Finally, we also treated U87MG cells with the
B2C11 anti-CD9 antibody to see if antibody treatment could sensitize the cells to radiation, and we found that pre-treatment of anti-CD9 antibody could sensitize cells to radiation with a dose enhancement factor of 1.25 at 10% survival (Fig 3-
3E). These data suggest that CD9 expression protects GBM cells from genotoxic ionizing radiation in vitro.
CD9 depletion hinders DNA damage repair
DNA double strand breaks (DSBs) are the main deleterious damage caused by ionizing radiation (IR) (106). Our results demonstrate that inhibiting
CD9 can radiosensitize GBM cells in vitro, which led us to question if depleting
CD9 affects DNA damage repair. Thus, we decided to evaluate the activation of
γH2AX foci, a canonical DSB marker, in U87MG shRNA cells following 2 Gy radiation treatment. Upon CD9 depletion and after irradiation, we found decreased amounts of γH2AX foci in shCD9 cells compared to shGFP control cells 30 min post IR (Fig 3-4A, B). Further, to determine whether DNA damage repair was also hindered by CD9 inhibition, we performed a homologous recombination (HR) DNA repair assay using a DR-GFP reporter in U87MG cells.
80 Our results show that deceasing CD9 expression (Fig 3-4C) hinders the efficiency of HR DNA repair (Fig 3-4D). Together, these data suggest that CD9 may play a role in DSB foci formation and DSB DNA repair.
CD9 signals through AKT
We demonstrate that CD9 inhibition decreases cell growth and sensitizes
GBM cells to radiation. As the PI3K/AKT pathway is a major cell survival signaling pathway found to be altered in GBM and correlates with tumor growth and radiation resistance (68, 93), we hypothesized that CD9 may modulate AKT signaling. Therefore, we performed a western blot to examine the levels of phospho-AKT and total AKT in U87MG shRNA cells. We found that inhibiting
CD9 decreased basal phospho-AKT S473 levels compared to shGFP control cells (Fig 3-5). Therefore, in our studies, CD9 mediates AKT signaling.
Discussion
In the present study, we defined a potential novel mediator of cell proliferation and radioprotection in GBM that associates with poor patient outcome. As assessed bioinformatically in Oncomine and TCGA databases, CD9 is overexpressed in GBM tumor samples compared to normal brain. CD9 inhibition decreased cell proliferation and radiosensitized multiple GBM cell lines, including neurosphere stem cells. Additionally, we found that CD9 may have a role in DNA damage repair, possibly through the promotion of DSB recognition or
DNA repair. Mechanistically, we found that depleting CD9 decreases the basal levels of phospho-AKT, which could explain the alterations seen in proliferation,
81 radiation sensitization and DNA repair but would require additional experimental support.
The role of CD9 in tumorigenesis has been studied in a variety of cancers.
Particularly in GBM, CD9 has been previously found to be essential for the self- renewal and proliferation of glioma stem cells (65), comparable to the results we saw in our cell proliferation and tumor growth studies. However, this study demonstrated a role for CD9 in promoting STAT3 activation through increased gp130 stability. Our study herein shows decreased basal phospho-AKT protein levels upon CD9 inhibition. As CD9 can interact with other proteins found in the
TEMs, our results demonstrate an additional signaling pathway that CD9 could be modulating through its various membrane protein associations.
Additionally, we describe a novel role for CD9 as a mediator of radiation resistance in GBM. We show that upon CD9 inhibition, GBM cells become more sensitive to radiation treatment. Additionally, we demonstrate that following radiation treatment, γH2AX foci formation is decreased in CD9 depleted GBM cells compared to shGFP control cells. Further, inhibiting CD9 also decreases the efficiency of cells to be able to repair themselves by homologous recombination compared to control cells. These data could lead us to hypothesize that the cause for radiation sensitization with CD9 inhibition may be due to the cells not sensing the DNA damage following radiation treatment. If the cells are not detecting the damage, then the cell’s repair machinery may not be activated resulting in unresolved DNA damage. However, further experimentation would have to be done in order to support this hypothesis.
82 In summary, we defined a novel function of CD9 in mediating radiation resistance in brain tumors. Our findings describe CD9 as a critical mediator of cell proliferation and radiation protection through increased AKT signaling and
DNA damage repair. Together, these findings contribute to our understanding of
CD9 and suggest that inhibiting CD9 may prove to be a beneficial therapeutic target.
83 Figure 3-1: Relevance of CD9 to GBM A, Oncomine data of CD9 expression in Shai GBM brain tumors versus normal brain tissue. B, The Cancer Genome Atlas study of CD9 expression in GBM brain tumors versus normal brain tissue. C, PrognoScan survival curves of GBM patients whose tumors express high versus low levels of CD9. D, From the
TCGA LGG (low grade glioma) +GBM database, Kaplan Maier survival plot for low and high grade gliomas dichotomized by CD9 expression. E, Expression of
CD9 in low and high grade gliomas from the TCGA database. Pairwise statistical comparisons of each subtype of low grade glioma to GBM are indicated
(student’s t-tests). F, CD9 expression in the GBM molecular subtypes from the
TCGA GBM BioDiscovery Portal. ***, p<0.001; ****, p<0.0001 by one-way
ANOVA.
84 Figure 3-1:
85 Figure 3-2: CD9 promotes cell proliferation and tumor growth A, C, Cell proliferation assay in A172MG and U87MG GBM knockdown cells. B,
D, qRT-PCR for samples in A and C. E, Subcutaneous U87 shRNA tumor growth curve. F, Kaplan-Meier curve for mice in E. G, CD9 expression from U87 cells before injection and from U87 subcutaneous tumors. H, An MTT analyzing cell viability and proliferation after anti-CD9 antibody treatment.
86 Figures 3-2:
87 Figure 3-3: CD9 inhibition sensitizes GBM cells to radiation Clonogenic survival assays of A172 (A), U87 (B), GBM0821 (C) and GBM0913 neurospheres (D) cells after indicated doses of ionizing radiation. qRT-PCR of
CD9 expression normalized to GAPDH are insets. E, Clonogenic survival assay in U87 cells following 24 hour pre-treatment with anti-CD9 B2C11 antibody.
88 Figure 3-3:
89 Figure 3-4: CD9 depletion decreases the number of γH2AX foci and hinders DNA damage repair
A, Quantification of γH2AX foci following ionizing radiation (2 Gy) in U87MG shRNA cells. B, Immunofluorescence of γH2AX (red) foci. C, qRT-PCR of targeted genes in homologous recombination (HR) reporter assay. Expression normalized to β-actin. D, DR-GFP HR reporter assay in siCD9, siGAPDH
(negative control), siBRCA1 (positive control), or siRISC (control) U87 cells. ***, p<0.001 by student’s t-test; ****, p<0.0001 by one-way ANOVA.
90 Figure 3-4:
91 Figure 3-5: CD9 depletion decreases basal phospho-AKT protein levels Immunoblot of shRNA U87 cells demonstrating a decrease in basal levels of phospho-AKT in shCD9 cells compared to shGFP control cells.
92 Figure 3-5:
93 Chapter 4 - Spermidine/spermine N1- acetyltransferase 1 is a gene-specific transcriptional regulator that drives brain tumor aggressiveness
The work in this chapter is originally published in Oncogene Vol 38. Thakur VS, Aguila B*, Brett-Morris A, Creighton CJ, Welford SM. “Spermidine/spermine N1-acetyltransferase 1 is a gene-specific transcriptional regulator that drives brain tumor aggressiveness.” 38(41):6794-6800.
Copyright 2019 with permission from Springer Nature Limited
For Supplementary Tables 1 and 2, please visit https:// doi.org/10.1038/s41388-019-0917-0
*Work contributed by the author of this dissertation are incorporated in Figures 1 and 3
94 Abstract
Spermine/spermine N1-acetyltransferase 1 (SAT1), the rate limiting enzyme in polyamine catabolism, has broad cell-regulatory roles due to near ubiquitous polyamine binding. We describe a novel function of SAT1 as a gene-specific transcriptional regulator through local polyamine acetylation. SAT1 expression is elevated in aggressive brain tumors and promotes resistance to radiotherapy.
Expression profiling in glioma cells identified SAT1 target genes that distinguish high and low grade tumors, in support of the prognostic utility of SAT1 expression. We further discovered mechanisms of SAT1-driven tumor aggressiveness through promotion of expression of both DNA damage response pathways as well as cell cycle regulatory genes. Mechanistically, SAT1 associates specifically with the promoter of the MELK gene, which functionally controls other SAT1 targets, and leads biologically to maintenance of neurosphere stemness in conjunction with FOXM1 and EZH2. CRISPR knockin mutants demonstrate the essentiality of the polyamine acetyltransferase activity of SAT1 for it function as a transcriptional regulator. Together, the data demonstrate that gene-specific polyamine removal is a major transcriptional regulatory mechanism active in high-grade gliomas that drives poor outcomes.
95 Introduction
Polyamines putrescine, spermidine, and spermine are small, positively charged molecules present in millimolar amounts in cells that bind acidic macromolecules, and broadly regulate functions such as replication, translation, and chromatin condensation. Polyamines are both essential for life, and facilitators of cell death, and are thus tightly regulated through synthesis, uptake, and degradation/secretion to maintain homeostatic levels (114). SAT1 is the global polyamine rheostat, serving as the rate-limiting catabolic enzyme that acetylates spermine and spermidine, and primes them for either back-oxidation
(spermine to spermidine, and spermidine to putrescine) or cellular efflux. SAT1 itself is highly regulated by polyamine concentrations, underscoring significant cellular investment in controlling polyamines (84).
Polyamines and SAT1 are present in all cellular compartments (115), but specific roles in different locations are not well-described. SAT1 has a role at the cell membrane, controlling motility by regulating a polyamine-sensitive potassium channel, and is recruited by interaction with α9β1 integrin (116). SAT1 has also been found to interact with the membrane-bound diamine transporter SLC3A2
(117), and cytoplasmic HIF1α (118) and eIF5A (119), the latter of which is a target of SAT1 acetylation controlling translation. While SAT1 is found in the nucleus, and polyamine catabolism is critical for unwinding DNA (115, 120), no localized nuclear functions have been this far ascribed.
SAT1 has been implicated in cancer, affecting cell migration (86), proliferation and response to ionizing radiation (IR) (37). Our group found SAT1
96 through an shRNA screen to cause resistance to radiation in glioblastoma (GBM), and showed that SAT1 is critical for tumor growth. We found that SAT1 controls expression of BRCA1 through a transcriptional mechanism. The observation helped explain why SAT1-elevated tumors respond less well to radiotherapy/temozolomide than SAT1-reduced tumors. However, the finding also led to the question of how SAT1 controls gene expression, and what other genes are controlled by SAT1. In the present study, we performed expression profiling on SAT1 competent and deficient brain tumor cells and identifi ed the breadth of SAT1-dependent genes. We confirmed regulation in multiple neurosphere lines, and evaluated expression of the genes in TCGA data of 677 brain tumors. Our studies delineate the mechanism of
SAT1-mediated gene expression, uncovering an unknown function of SAT1 in general, but also exposing SAT1 as a veritable therapeutic target in GBM that controls tumor cell stemness and aggressiveness.
Results SAT1 regulates gene programs controlling cell cycle and DNA dynamics
In order to assess the extent of genes regulated by SAT1 in brain tumors, we performed gene expression profiling of U87MG cells with two shRNAs to
SAT1 using Affymetrix microarrays covering 67,528 gene transcripts. U87MG cells were stably transduced with lentiviruses expressing shRNAs to either the green fluorescent protein (GFP) or SAT1, and RNA was harvested within 7 days of infections. RNA was reverse transcribed into cDNA, labeled with biotin and hybridized onto the arrays for analyses. Hierarchical clustering of duplicate
97 experiments correctly segregated samples according to sample identity (i.e. shGFP clustered together, shSAT1-1 clustered together, and shSAT1-2 clustered together; with the latter two separately branching from shGFP; Fig 4-1A). At a fold-difference of ≥2.5, and an ANOVA p-value of ≤0.07, 169 genes were differentially expressed between control and the shRNA cell lines
(Supplementary Table 1), including previously identified BRCA1. At a threshold of 2.0, 332 genes were identified.
To validate gene expression changes in U87MG cells, we tested SAT1 knockdown on a cohort of genes in two neurosphere stem cell lines (GBM0821 and GBM0913) that retain characteristics of tumor stem cells in vivo (100). All of the genes demonstrated significant reductions in expression upon knockdown of
SAT1 (Fig 4-1B). We also validated that reduction of transcript resulted in decreased protein for several of the genes by Western blot (Fig 4-1C). Finally, to gain insight into the functions of the genes affected by SAT1, we performed gene ontology-based functional classification and found significant over-representation of genes involved in both cell cycle/mitosis/cytokinesis and DNA metabolism/replication/repair over all other categories (Fig 4-1D). Since we found previously that SAT1 regulates DNA repair and tumor growth, the data support a broad effect of SAT1 on genes driving aggressive tumor biology.
SAT1 target genes are enriched in high-grade gliomas
To glean clinical relevance for SAT1 target genes, we queried TCGA and assessed expression in 677 gliomas. SAT1 is significantly overexpressed in high- grade glioblastoma compared to all subtypes of low-grade gliomas (Fig 4-2A).
98 Furthermore, SAT1 expression correlates with poor outcome in the LGG+GBM
(Low-Grade Glioma and Glioblastoma) cohort with a high level of significance (p
< 0.001, log rank test; Fig 4-2B), largely because when ranked by SAT1 expression levels, 80% of the glioblastomas segregate to the upper-third of the tumors, and only 4.9% fall to the lower-third. Considering only the low-grade gliomas, high SAT1 tumors again perform significantly worse than those with lower SAT1 levels (p = 0.001, log rank test; Fig 4-2C). Notably, SAT1 expression was only prognostic in IDH1 wild-type tumors, which tend to have poorer outcomes than IDH1 mutant (Fig 4-2D, E). Thus, SAT1 expression predicts poor outcome in human gliomas.
We next assessed the expression of the 152 SAT1 target genes (2.5-fold genes) that mapped to the TCGA dataset in each of the tumors in the LGG+GBM cohort that were clustered according to a recent integrative analysis producing four grade-independent subtypes (LGr1-4) (121). We noted strikingly that expression of the SAT1 genes was independent of subtype and distinctly higher in the GBM subset compared to the LGG subsets. Further among LGG, some of the LGr4 samples appeared more similar to the GBM samples, which are themselves comprised mostly of LGr4 (Fig 4-2D). Together, the data demonstrate that SAT1 and its target genes are expressed in more aggressive gliomas (i.e. GBM) and portend poor prognosis in the LGG groups.
SAT1 regulates MELK and EZH2 by direct interaction with chromatin
Intriguingly, MELK, FOXM1, EZH2, NEK2, and PLK1 are SAT1 target genes. Recent studies have demonstrated a role for the maternal embryonic
99 leucine zipper kinase (MELK) in brain tumors. MELK was shown to phosphorylate the FOXM1 transcription factor, leading to expression of EZH2, the catalytic subunit of the polycomb repressive complex PRC2 (122). All three proteins are necessary for stem cell maintenance, tumor growth, and resistance to IR. FOXM1 regulates proliferation by controlling polo-like kinase 1 (PLK1) during mitosis (123), and NIMA-related kinase 2 (NEK2) is a regulatory binding partner of EZH2 in glioma stem cells (124). Together, the MELK network promotes tumor aggressiveness.
We explored the interaction of MELK, FOXM1, and EZH2 at the gene expression level in three established cell lines (U87MG, LN229, and Gli36), and observed that knockdown of MELK caused reductions in expression of FOXM1,
EZH2, and BRCA1 (Fig 4-3A). We next targeted FOXM1 by shRNA and found similarly that MELK, EZH2, and BRCA1 were again reduced as a consequence
(Fig 4-3B). These data support published findings of a functional dependence/interaction between MELK and FOXM1 (123), and also that MELK and FOXM1 control each other at the transcriptional level (122). We next investigated how SAT1 regulates expression of MELK and FOXM1, and performed chromatin immunoprecipitation (ChIP) to ask if SAT1 interacts with the proximal promoter regions of several of its targets. Using shGFP U87MG and shSAT1 U87MG, we found as association of SAT1 with both the MELK and
EZH2 promoters, but not FOXM1, BRCA1, or NUSAP1 (Fig 4-3C). This observation led to the hypothesis that MELK and EZH2 are major nodes of SAT1 gene regulation, and the data in Fig 4-3A would suggest MELK precedes EZH2.
100 We then tested the effect of MELK knockdown in comparison to SAT1 knockdown over a range of identified target genes across the three cell lines. We found in all genes tested, MELK knockdown with two shRNAs phenocopied
SAT1 knockdown (Fig 4-3D). Interestingly, ectopic expression of MELK induced
SAT1 target genes, but in an SAT1-dependent manner (Fig 4-4); suggesting
MELK is necessary but not sufficient to drive SAT1 genes. To gain a broader view of the roles of MELK/FOXM1 and EZH2 in SAT1-driven gene expression programs, we analyzed recently published ChIPseq data for both FOXM1 (125) and H3K27me3 (126) (the EZH2-directed histone modification) in cancer cells, and compared their target genes to SAT1 target genes. Roughly one fifth (59 of
315, or 18.7%) of the SAT1 genes (≥2 change) were identified to have FOXM1 binding (Fig 4-3E, and Supplementary Table 2), and one twentieth (17 of 315, or
5.4%) of the published EZH2 targets were found on the SAT1 gene list. To verify the effect of SAT1 knockdown on FOXM1 occupancy, we performed FOXM1
ChIP on three of its identified targets (CCNB1, CCNB2, and CDC20) and found reduced association of FOXM1 (Fig 4-3F). Finally, because MELK, FOXM1, and
EZH2 have been associated with stemness of glioblastoma cells, we assess the ability of SAT1 knockdown primary glioblastoma lines to form neurospheres in culture. In both the GBM0821 and GBM0913 lines, we found that SAT1 depletion led to significant reduction in the ability to form neurospheres (Fig 4-3G, H).
Therefore, the data demonstrate that MELK and EZH2 are key drivers of SAT1- mediated gene regulation and biology, promoting brain tumor stemness.
101 SAT1 regulates transcription of specific targets through polyamine catabolism
SAT1 is a polyamine catabolic enzyme. To assess the role of polyamines in SAT1-directed gene expression, we treated U87MG cells with 100 µM spermidine (SPD) for 2, 4, or 6h, and harvested RNA for analysis. Notably, SAT1 itself is highly responsive to excess polyamines (84), and we observed rapid induction of SAT1 mRNA (Fig 4-5A). Similarly, we saw an induction of SAT1 targets BRCA1, MELK, and EZH2. To determine if SAT1 was required for induction of the genes, or whether polyamines were responsible independently to induce transcription, we tested SAT1 knockdown cells and found a loss of responsiveness for all genes. The data argue that polyamine stimulation of SAT1 is required, but do not delineate whether catabolism of polyamines is involved.
To determine if SAT1 polyamine acetylation activity is required for target gene activation, we created a conditional CRISPR/Cas9 knockin cell line in which wild type SAT1 remains expressed by its endogenous promoter until CRE recombinase induces excision of wild type exons 4-6, resulting in inclusion of mutant exons 4-6 with point mutation in both the acetyl Co-A and polyamine binding sites (R101A and E152K, respectively) (Fig 4-5C). The mutations have been shown to abrogate polyamine catabolism of SAT1 (127). To validate the cell line, adenoviral mediated expression of CRE recombinase was used, and genomic PCR was performed using primer sets A-B and A-C, as indicated in Fig
4-5C. The A-B primer pair only amplifies the target prior to recombination, while
A-C amplifies after recombination. As seen in Fig 4-5D, virtually complete
102 recombination was evidenced by the lack of signal in the A-B PCR post- recombination. We also validates that the resulting locus contained the designed mutation by Sanger sequencings the PCR products (Fig 4-5E). We then assessed the transcriptional activity of SAT1 mutants. As seen in Fig 4-5F, G,
CRE recombination resulted in a moderate increase in SAT1 expression compared to control cells, but a dramatic loss of expression of both MELK and
FOXM1 at both the mRNA and protein levels. The increase of SAT1 expression confirms the production of non-functional SAT1, to which the cells respond by attempting to elevate expression. The loss of MELK and FOXM1 expression demonstrates that polyamine catabolism is required for transcriptional activation.
Discussion
We have uncovered a novel function for SAT1 as a gene-specific transcriptional modulator. Depletion of SAT1 from cells revealed altered gene expression programs that include cell cycle regulation and DNA repair, which portend poor prognosis in brain tumors. Physical association of SAT1 with the
MELK and EZH2 promoters, and the requirement of SAT1 enzymatic function to regulate genes support a model in which localization of SAT1 to target genes, and acetylation and removal of polyamines from chromatin promote transcriptional activation. The studies highlight an expanding role for polyamine metabolism in cancer, and support pathway modulators as potential therapeutics.
How SAT1 is recruited to specific sites on DNA, and whether SAT1 transcriptional programs in different tissues are unique are open questions; analysis of SAT1 binding partners in different contexts will likely lend insight. The
103 only reported mechanism of functional localization of SAT1 is for cell motility, wherein SAT1 has been shown to interact with α9β1 integrin. As SAT1 does not have an identifiable DNA binding domain, protein-protein interaction is the likely mechanism to interact with chromatin. Additionally, recent studies have highlighted differential affinities of DNA sequences for polyamines, such that AT- rich regions bind more strongly than GC-rich regions; and polyamine binding stabilizes DNA duplexes (120). The significance of this is that SAT1-mediated acetylation and removal of polyamines could be sequence-dependent, offering a potential mode of specificity for transcriptional activity, and a differential role of
SAT1 on different genes. Interestingly, while we found MELK inhibition to phenocopy SAT1, and that ectopic expression of MELK elevated SAT1 target genes, however, MELK was insufficient to restore function in absence of SAT1.
This suggests SAT1 function is still required beyond the control of MELK.
Understanding exactly where SAT1 is needed in chromatin could illuminate a novel mechanism of epigenetic-like regulation of chromatin compaction and relaxation.
For brain tumors, the data presented argue that SAT1 function controls two major pathways, both of which are implicated in tumor aggressiveness and therapeutic resistance. Roughly 25% of SAT1 genes are direct targets of FOXM1 or EZH2. Some of the targets are transcription factors themselves, including
DEPDC1 (128) and TSHZ2 (129); thus it is conceivable that FOXM1/EZH2 account for virtually all of the SAT1 targets. With the negative association of
SAT1 expression and patient outcome, focusing therapeutic strategies on MELK,
104 FOXM1, and EZH2, for which inhibitors currently exist, may be therapeutically advantages. Whether inhibition of SAT1 rather than these three targets would be more efficacious will require further study, but expanding the list of viable therapeutic targets is an advance in a disease in need of novel approaches.
105 Figure 4-1: SAT1 regulates cell cycle and DNA regulatory genes in GBM cells A, Hierarchical clustering of differentially expressed genes from microarray data from U87MG cells expressing shGFP or one of two shSAT1 shRNAs. Gene names are included in supplementary table 1. B, Validation of SAT1 target genes in two neurosphere lines by qRT-PCR. p<0.05 for all genes in the knockdown experiments except for those indicated by “@.” C, Validation of SAT1 target genes by Western blot. D, Gene ontology classification of SAT1 target genes
(blue) into functional groups compared to all genes in the genome (red).
Statistically significant differences are indicated.
106 Figure 4-1:
107 Figure 4-2: SAT1 target genes are elevated in aggressive brain tumors A, The expression of SAT1 in low- and high-grade gliomas from the TCGA
LGG+GBM database. Pairwise statistical comparisons of low-grade glioma to
GBM are indicated (Student’s t-tests). B, Kaplan-Meier survival plot for low- and high-grade gliomas dichotomized by SAT1 expression. C, Kaplan-Meier survival plot for only low-grade gliomas dichotomized by SAT1 expression. D, Kaplan-
Meier survival plot for low- and high-grade IDH1 mutant gliomas dichotomized by
SAT1 expression. E, Kaplan-Meier survival plot for low- and high-grade IDH1 wild type gliomas dichotomized by SAT1 expression. F, Expression of SAT1 target genes in 677 brain tumors of both low and high grades, organized by LGr classifications.
108 Figure 4-2:
109 Figure 4-3: MELK and EZH2 are SAT1 direct target nodes A, Assessment of FOXM1, EZH2, and BRCA1 in control (shGFP) and MELK knockdown U87MG, LN229, and Gli36 cells by qRT-PCR. p<0.05 for all genes in the knockdown experiments. B, Assessment of MELK, EZH2, and BRCA1 in shGFP and shFOXM1 U87MG, LN229, and Gli36 cells by qRT-PCR. p<0.05 for all genes in the knockdown experiments. C, Input normalized SAT1 ChIP assay on U87MG shGFP and shSAT1 cells on the MELK, FOXM1, EZH2, BRCA1, and
NUSAP1 promoters. Histone H3 ChIP is a positive control. D, Heat map of expression of SAT1 target genes measured by qRT-PCR in shSAT1 U87MG,
LN229, and Gli36 cells compared to shMELK cells. E, Venn diagram comparing
SAT1 target genes with published FOXM1 and EZH2 targets. F, FOXM1 ChIP in shGFP and shSAT1 U87MG cells. G, Neurosphere formation assay in GBM0821 and GBM0913 neurosphere stem cell lines with shGFP or shSAT1-1. H,
Photomicrographs of neurospheres. Scale bar 100 µm.
110 Figure 4-3:
111 Figure 4-4: Ectopic expression of MELK leads to overexpression of SAT1 target genes in an SAT1 dependent manner U87MG cells were infected with shGFP or shSAT1 lentivirus, selected with puromycin, and then transfected with a MELK expression plasmid (pCSII-IB-
MELK). RNA was harvested after two days, and RT-PCR was performed on the indicated genes. SAT1 knockdown led to decreases in expression of the genes
(red bars vs. blue bars). Ectopic MELK led to induction of expression of each of the genes (green bars), but not in the context of SAT1 knockdown (purple bars).
*p≤0.05 by students t tests.
112 Figure 4-4:
113
Figure 4-5: Polyamine catabolism is necessary for SAT1 transcriptional activity A, B, Measurement of SAT1 target genes after exposure to spermidine (SPD) by qRT-PCR in shGFP cells (A) or shSAT1 cells (B). C, Schematic of the SAT1 locus on the X Chromosome (top) and the modified locus after CRISPR/Cas9 induced knock-in pre (middle) and post (lower) CRE recombination. D, PCR validation of recombination using primers indicated in the schematic. E, Sanger sequencing of the wild type and two mutation sites in SAT1 after CRE recombination. F, G, Measurement of SAT1, MELK, and FOXM1 by qRT-PCR
(F) and Western blot (G) in control or CRE adenovirus infected cells.
114 Figure 4-5:
115 Chapter 5 - Discussion and Future Directions
The vast molecular basis for radioresistance
Radiation therapy is a principal form of treatment for GBM patients.
However, GBM has been shown to exhibit resistance to radiation, especially in recurrent tumors. Despite identification of several mechanisms that contribute to radioresistance, the molecular basis remains incompletely defined. Here, our work has identified three proteins that contribute to radioresistance in GBM through alternative pathways and may prove to be beneficial therapeutic targets.
In chapter 2, we demonstrate that PTGFRN is overexpressed in GBM and this overexpression correlates with poor patient survival outcome. We show that
PTGFRN promotes cell proliferation by regulating AKT signaling through stabilization of PI3K p110β at the cell membrane. Additionally, we show that
PTGFRN promotes radiation resistance through its regulation of p110β stability.
Upon PTGFRN inhibition, p110β protein is degraded leading to decreased protein expression at the membrane, in the cytosol, and in the nucleus. The decrease in nuclear PI3K p110β expression disrupts DNA double strand break recognition and consequently DNA damage repair, thus leading to radiation sensitization. Therefore, we illustrate a novel role for PTGFRN in mediating cell survival signaling in GBM.
In chapter 3, we show that CD9 is also overexpressed in GBM tumors and elevated CD9 expression correlates with poor overall survival. We demonstrate
116 that CD9 promotes cell proliferation and may promote radiation resistance.
Additionally, CD9 inhibition decreases the number of γH2AX foci formed after ionizing radiation and hindered DNA damage repair. Further, we find CD9 to be regulating AKT signaling, and the observed decreased basal phospho-AKT expression could potentially explain the observed decreased cell proliferation and increased radiosensitization upon CD9 depletion. The data presented here illustrate a critical role for CD9 in promoting GBM progression by mediating cell proliferation and radiation resistance.
In chapter 4, we introduce polyamine catabolic enzyme SAT1 as a mediator of radiation resistance in GBM. Previously, SAT1 was found to mediate histone H3 acetylation to regulate BRCA1 gene expression. Inhibiting SAT1 decreased BRCA1 expression and BRCA1 foci formation following DNA damage and thereby inhibited homologous recombination DNA repair (37). Moreover, here we describe a novel function for SAT1 as a gene-specific transcriptional modulator, and demonstrate that depleting SAT1 from GBM cells altered gene expression programs, including cell cycle regulation and DNA damage repair, that contribute to tumor aggressiveness and therapeutic resistance.
Mechanistically, we show that SAT1 physically associates with the promoters of
MELK and EZH2, which functionally control SAT1 target genes, and that the polyamine acetyltransferase activity of SAT1 is essential for its function as a transcriptional regulator.
Taken together, these data demonstrate that the molecular basis for radiation resistance in GBM is vast, utilizing several proteins and multiple cellular
117 pathways. Specifically, our data suggest that SAT1, CD9, and PTGFRN could be potential future therapeutic targets but also suggest the need to develop multi- targeted therapies as radiation response is regulated through multiple pathways.
Furthermore, while these studies provide the framework for understanding the radioprotective mechanisms of SAT1, CD9, and PTGFRN in GBM cells, it also raises additional questions including the possibility of additional signaling pathways utilized by PTGFRN, whether PTGFRN and CD9 are modulating signaling together as a complex, and if SAT1 could be regulating CD9 or
PTGFRN transcription. These questions and future directions are outlined below.
Other signaling possibilities for PTGFRN
In chapter 2, we find that PTGFRN promotes cell proliferation and radioresistance in GBM cells through stabilization of PI3K p110β. However, interestingly, PTGFRN was first discovered to inhibit the binding of prostaglandin
F2α to the FP receptor, a G-protein coupled receptor (GPCR) that could selectively activate p110β (130), by decreasing the receptor number rather than the receptor affinity (49). Increased levels of prostaglandin F2α (PGF2α), and subsequently increased prostaglandin F2α:FP receptor signaling, have been found to increase cell proliferation and promote radiation resistance in prostate cancers (131, 132). Additionally, increased concentrations of PGF2α in GBM tumors correlate with poor patient surival (133). As PTGFRN was found to decrease the receptor number, decreasing PTGFRN would cause increased
118 levels of FP receptor, allowing for prostaglandin F2α binding and signal transduction. However, in our study, decreasing PTGFRN affects cell proliferation and radiation resistance through survival signaling that does not saffect protein levels of FP receptor in the GBM cells tested (Figure 5-1).
Therefore, in our study and possibly in the context of GBM, PTGFRN does not regulate prostaglandin signaling by depleting the number of FP receptor molecules.
If PTGFRN indeed does play an alternative role in cell signaling, it is possible that PTGFRN could be affecting another GPCR present on the cell membrane. In GBM, GPR56, a member of the adhesion-GPCR family, has been found to be important in hindering mesenchymal differentiation and radioresistance by inhibiting NF-κB signaling, and low expression of GPR56 was found to correlate with poor overall survival (134), an inverse relationship to
PTGFRN expression. Additionally, TNF-α treatment was found to downregulate
GPR56 mRNA expression in proneural glioma-initiation cells, thereby promoting mesenchymal differentiation (134). These data lead us to question if PTGFRN, particularly in mesenchymal GBM tumors, could be acting as a negative regulator of GPR56 through regulation of TNF-α/TNFR1 signaling. In GBM, we find
PTGFRN to be overexpressed, correlate with poor patient survival, and PTGFRN expression is elevated in mesenchymal GBM tumors compared to other subtypes
(Figure 2-1, 2-2). The high expression of PTGFRN could be inhibiting GPR56 expression by stabilizing TNF-α/TNFR1 signaling, thus promoting radioresistance and the mesenchymal GBM signature. Upon PTGFRN depletion, TNF-α/TNFR1
119 signaling may become hindered, the inhibition of GPR56 would be removed, and
GPR56 would then be able to inhibit NF-κB signaling further preventing mesenchymal differentiation and radioresistance. As such, we propose the following studies. Firstly, future studies need to confirm the negative relationship between PTGFRN and GPR56. Immunohistochemistry on mesenchymal GBM tumors examining PTGFRN and GPR56 expression and western blots analyzing
GPR56 protein after PTGFRN depletion would answer this inquiry.
Should GPR56 expression have an inverse relationship with PTGFRN expression, studies should be undertaken to next determine if PTGFRN is in close proximity of TNF-α receptor, TNFR1, tumor necrosis factor receptor 1.
Subsequently, a proximity ligation assay, and further co-immunoprecipitations to determine direct association, could provide information on a possible complex.
Secondly, protein signaling stimulation experiments need to be completed analyzing impeded TNF-α signaling upon PTGFRN depletion. For this, PTGFRN and GFP knockdown cells can be starved overnight and then stimulated with
TNF-α, followed by analysis of downstream NF-κB pathway activation. If upon
PTGFRN inhibition, activation of the NF-κB pathway is hindered, PTGFRN could be important in stabilizing the signaling complex and therefore, future studies should evaluate the stability of proteins found upstream of NF-κB and downstream of TNFR1, including but not limited to; TNFR1, TRADD, TRAF2, cIAP1/2, and RIP1 (135).
Further investigations into the mechanism(s) PTGFRN partakes in mediating cell survival signaling could provide insight into the importance of
120 scaffolding proteins in driving GBM tumorigenesis and aggressiveness. Indeed, changes in protein stability at the membrane due to shifts of corresponding scaffolding proteins could drastically alter cell survival signaling pathways, thereby either driving or inhibiting GBM progression.
121 Figure 5-1: Expression of FP receptor protein is not consistently altered after PTGFRN inhibition
Upon inhibition of PTGFRN in GBM0913 and GBM0821 cells, changes in FP receptor protein levels are not significantly altered and are inconsistent across
GBM neurosphere stem cell lines.
122 Figure 5-1:
123 PTGFRN and CD9: a possible scaffolding complex?
While PTGFRN was found to possibly associate with and regulate the FP receptor (49), PTGFRN was also found to be a highly stoichiometric, highly specific binding partner of CD9 (47, 50). Multiple studies have suggested that the phenotypic effects observed upon PTGFRN depletion is due to corresponding depletion of CD9 from the cell surface (51, 52). Interestingly, CD9 has been implicated in GBM tumorigenesis by promoting tumor growth through gp130 stabilization and STAT3 signaling (65), similar to PTGFRN stabilizing p110β.
Therefore, it raises the question on whether PTGFRN and CD9 are signaling together as a complex, as well as individually, to promote GBM tumorigenesis. In chapters 2 and 3, we show that upon individual PTGFRN and CD9 depletion, basal levels of P-AKT decrease. Further, we also demonstrate a decrease in protein expression also decreases cell proliferation and increases radiosenitization. As the phenotypic affects after CD9 or PTGFRN depletion appear similar and as both proteins, individually, seem to regulate turnover of signaling proteins, we could hypothesize that CD9 and PTGFRN could be signaling as a complex.
In order to test this hypothesis, initial experiments examining cell proliferation and radiation sensitization were done in GBM U87MG cells. We found that inhibition of both CD9 and PTGFRN decreased cell proliferation more than a single knockdown, producing an additive effect and suggesting that CD9 and PTGFRN promote cell proliferation through different pathways (Figure 5-
2A). However, depleting both CD9 and PTGFRN in a clonogenic survival assay
124 did not produce an additive radiosenitization effect compared to CD9 inhibition alone suggesting that CD9 and PTGFRN promote radioresistance through the same pathway (Figure 5-2D). It is plausible that CD9 and PTGFRN could be signaling in the same pathway for one phenotype and in a different pathway for another phenotype, as their ability to modulate signaling relies on interactions with proteins found within their proximity on the cell membrane. Future analyses are required to fully delineate the capacity of a CD9/PTGFRN complex, their surrounding proteins, and their possible dual implications to cell signaling.
125 Figure 5-2: Targeting CD9 and PTGFRN in combination decreases cell proliferation and radiosensitizes GBM cells A, Cell proliferation assay in U87MG GBM knockdown cells. B, C, qRT-PCR for samples in A. D, Clonogenic survival assay of shRNA U87MG cells after indicated doses of ionizing radiation. E, F, qRT-PCR for samples in D.
126 Figure 5-2:
127 Could SAT1 be regulating CD9?
In chapter 4 we describe a novel function for SAT1 as a gene-specific transcriptional modulator. Our model suggests that SAT1 localizes to target genes, driving acetylation and removal of polyamines from chromatin, thereby promoting chromatin relaxation and transcriptional activation. Specifically, we describe SAT1 to transcriptionally regulate EZH2 (enhancer of zeste homolog 2), a histone methyltransferase and a catalytic subunit of the polycomb repressive complex PRC2 (122, 136), by physically associating with the EZH2 promoter.
EZH2 catalyzes tri-methylation of histone H3 at lysine 27 (H3K27me) to regulate gene expression through epigenetic machinery and therefore, can function as both a transcriptional suppressor and a transcriptional co-activator (136).
Interestingly, EZH2 has also been demonstrated to regulate CD9 expression in gliomas and GBMs (137). Thus, we hypothesize that SAT1 could epigenetically regulate CD9 expression via EZH2. Interestingly, initial data analyzing mRNA through quantitative RT-PCR finds that CD9 mRNA expression is depleted in
SAT1 inhibited U87MG cells, supporting our hypothesis (Figure 5-3). Future studies should address if SAT1, or EZH2, binds to the promoter regions of CD9 to determine its transcriptional regulation. These studies can include additional western blots to confirm that CD9 protein levels are also decreased in SAT1 and
EZH2 knockdown cells and chromatin immunoprecipitations analyzing the capability of SAT1 and EZH2 binding to the proximal promoter regions of CD9 utilizing shGFP, shSAT1, and shEZH2 cells.
128 Additionally, in chapter 4 we perform gene ontology-based functional classification to gain insight into the functions of the genes affected by SAT1 and find cell-matrix adhesion genes to be over-represented. Integrins, a major family of cellular receptors for extracellular matrix proteins, have essential functions in a wide array of processes, including cell migration (116, 138). SAT1 has been shown to have a role at the cell membrane, accelerating cell migration by binding with the α9β1 integrin to control a polyamine-sensitive potassium channel (116).
Furthermore, integrins are found in tetraspanin-enriched microdomains and can form complexes with tetraspanins (139).
Literature has demonstrated that CD9 can form complexes with multiple different integrins to modulate integrin-dependent migration, spreading, and/or cell morphology (44, 138-140). Indeed, CD9, in a complex with CD81, has been found to associate with and regulate α3β1 integrin-dependent motility (138).
Additionally, CD9, in concert with integrin α5β1, was found to promote cell motility through a PI-3 kinase-dependent pathway (140). While literature has yet to demonstrate an interaction between CD9 and α9β1, it does not eliminate the possibility that CD9 may be in the proximity of SAT1 and α9β1 at the cell membrane. As CD9 has already been demonstrated to play the role of a scaffolding protein by stabilizing cell membrane proteins from degradation (65), we could hypothesize that CD9 may be important for the stability of the
SAT1/α9β1 complex and/or its downstream signaling proteins. Further studies should be undertaken in order to first validate the presence of a CD9/SAT1/α9β1 complex, through co-immunoprecipitations or proximity ligation assays, and
129 secondly, to understand its potential therapeutic capacity to target migration, radiation resistance and tumor growth in GBM.
130 Figure 5-3: Inhibition of SAT1 decreases CD9 mRNA Results for a qRT-PCR of analyzing CD9 mRNA levels in U87MG cells transfected with siRNA targeting SAT1.
131 Figure 5-3:
132 Concluding remarks
PTGFRN, CD9, and SAT1 have each been proven to have roles in promoting radiation resistance and aggressiveness in GBM through different molecular mechanisms including regulation of cell survival signaling and transcription. Therefore, these proteins appear to play a critical role in cancer progression and remain attractive targets for the development of targeted therapies. However, as PTGFRN, CD9, and SAT1 all promote radiation resistance through various pathways, a single targeted therapeutic approach may not prevent tumor recurrence. Identifying a multi-targeted therapeutic approach, while limiting drug toxicity by utilizing rational drug combination strategies or novel drug delivery mechanisms such as tumor directed nanoparticles, could combat GBM aggressiveness and recurrence. Additionally, further studies identifying additional mechanism(s) by which PTGFRN, CD9, and
SAT1 promote tumorigenesis and therapeutic resistance can enhance the fundamental understanding of the biology associated with the mechanisms of action and provide additional opportunities to inhibit tumor progression in glioblastoma multiforme patients.
133 Appendix A
Experimental Procedures Related to Chapter 2
134 Cell lines and reagents
U87MG and A172 cells were obtained from ATCC and authenticated by STR profiling by Genetica DNA Laboratories. The pLKO.1 Tet-ON construct was given by Dr. Eli Bar (Case Western Reserve University, Cleveland, OH).
Neurosphere stem cell lines GBM0821 and 0913 were a gift of Dr. Angelo
Vescovi (University of Bicocca, Milan) (100, 141, 142). D-luciferin came from
Gold Biotechnology. Stable PTGFRN inhibition was performed with lentiviral shRNA pLKO.1 clones: TRCN0000057452 and TRCN 0000057451 (Sigma). shRNA GFP was used as control. The PIK3CB overexpression plasmid was purchased from Addgene (plasmid #116555). Quantitative real-time PCR (qRT-
PCR) was performed using Power SYBR Green PCR Master Mix from Thermo
Fisher Scientific and normalized to either β-actin or GAPDH. Primer sequences:
PTGFRN F-5’ACAACAGCTGGGTGAAAAGC-3’, PTGFRN R-5’
TTTCATTGGGACTGGAGAGG-3’; Actin F-5’CATGTACGTTGCTATCCAGGC-3’,
R-5’CTCCTTAATGTCACGCACGAT-3’; GAPDH F-
5’AAGGTGAAGGTCGGAGTCAAC-3’, R-5’GGGGTCATTGATGGCAACAATA-3’; p85α F-5’AGTGGTTGGGCAATGAAAAC-3’, R-
5’GAAAAAGTGCCATCTCGCTTC-3’; p110β F-
5’GGGAAAGCTCATCGTAGCTG-3’, R-5’CTACTCTCCCGCTGACTTGC-3’;
BRCA1 F-5’TGGAAGAAACCACCAAGGTC-3’, R-
5’ACCACAGAAGCACCACACAG-3’.
135 Colony formation assay
Radiation was performed with a 137Cs irradiator (Shepherd). A total of 500 to
10,000 cells per plate were stained after 10 days with 0.1% crystal violet. Assays were done ≥ 3 times with individual samples in triplicate. Clonogenic assays of neurosphere lines were plated in 1 mL of neuro stem cell (NSA) media (100) containing 1.5% methylcellulose and fed every 3 days. Sphere formation was monitored and scored using GelCount (Oxford Optronix) after 10-12 days.
Cell Proliferation Assay
20,000 cells/well were plated into a 12 well dish and allowed to grow for 3 days.
On day 3, cells were counted, and 20,000 cells were replated. The process was repeated up to day 9 or day 10.
Tumor Formation Assay
Animal studies were performed in accordance with CWRU institutional guidelines. Subcutaneous U87MG knockdown tumors were produced by injection of 1x106 cells in the flanks of eight-week-old nude mice, and measured with calipers twice weekly. Eight-week-old nude mice were injected intracranially with 105 luciferase-expressing GBM0913 neurosphere stem cells into the right cerebral cortex at a depth of 3 mm. Two weeks following injection, Tet-on-shRNA injected mice were given doxycycline in their food for 6 days, and half of the mice were irradiated at 12 Gy using GammaKnife. Tumor growth was monitored and quantified using bioluminescent imaging. Animal appearance, behavior, and weight were monitored to evaluate tumor progression as per a Case Western
Reserve University approved IACUC protocol.
136 Cellular Fractionation
GBM shPTGFRN and shGFP neurosphere stem cell lysates were fractionated as described (143).
Western Blots
Western blots were performed using standard procedures. Antibodies against
PTGFRN (ab97567 Abcam), PI3K p110beta (Cell Signaling #3011), PI3K p110alpha (Cell Signaling #4249), PI3K p85alpha (Cell Signaling #4257), β-actin
(Sigma A1987), Phospho-AKT Ser473 (Cell signaling #9271), AKT (Cell signaling
#9272), Histone H3 (Santa Cruz sc-8654), α-Tubulin (Sigma T9026), Phospho-
Histone H2A.X (Ser139) (Millipore 05-636), Phospho-DNA-PKcs (PA5-78130,
Invitrogen), DNA-PKcs (MA5-13238, Invitrogen) or Na/K-ATPase (Cell signaling
#3010) were used to decorate the membranes, and developed by chemiluminescence on an Azure c300 imaging system. GBM0913 shPTGFRN or shGFP cells were pretreated with MG132 (20µM) or control vehicle (DMSO) for
30 mins. After 30 min, cycloheximide (50µg/ml) was added to each sample to block protein synthesis. Cells were collected at indicated time points.
Proximity Ligation Assay
The Sigma Duolink® In Situ Fluorescence (DUO92101) was used according to the manufacturer’s protocol. Antibodies: Anti-PIK3CB (SAB1404204-100UG),
Anti-PTGFRN (SAB2700379), monoclonal Anti-CD9 (SAB1402143-100UG), polyclonal Anti-CD9 (SAB4503606). Positive cells were detected by flow cytometry.
137 HR reporter assay
U87 DRGFP cells were transfected with siRNA using DharmaFECT. After 24 hours, cells were infected at 10 MOI with I-SceI-expressing adenovirus or control
(empty) virus (gift of Dr. Junran Zhang). Cells were incubated for an additional 48 hours and then analyzed by flow cytometry as described previously (144).
Immunofluorescent Staining
Cells were fixed in 3% paraformaldehyde and stained and analyzed as described previously (37). Antibody: Phospho-Histone H2A.X (Ser139) (Millipore 05-636), secondary Alexa-Fluor 594 anti-mouse (A11032; Invitrogen), Anti-PIK3CB
(SAB1404204-100UG; Sigma), Anti-PTGFRN (SAB2700379; Sigma), secondary
Alexa-Fluor 488 anti-rabbit (A11070; Invitrogen).
EdU pulse chase
Cells were treated with EdU (25uM) for 8 hours, cells were washed with HBSS and NSA complete media to remove the EdU and put back into culture to allow the cells to continue proliferating. A portion of cells were taken at indicated time points, fixed, permeabilized and stained according to Invitrogen’s Click-iT Plus
EdU flow cytometry assay (C10636) protocol. Cells were analyzed on a
CytoFLEX flow cytometer. Analysis of the cell cycle progression was completed as previously described (145).
Statistical Analysis
Student t tests and one or two-way ANOVA were used throughout the study to test the significance of differences between samples as indicated.
138 Appendix B
Experimental Procedures Related to Chapter 3
139 Cell lines and reagents
U87MG and A172 cells were obtained from ATCC and authenticated by STR profiling by Genetica DNA Laboratories. Neurosphere stem cell lines GBM0821 and 0913 were a gift of Dr. Angelo Vescovi (University of Bicocca, Milan) (100,
141, 142). Stable CD9 inhibition was performed with lentiviral shRNA pLKO.1 clones: TRCN0000296958 and TRCN 0000296953 (Sigma). shRNA GFP was used as control. Antibodies: Quantitative real-time PCR (qRT-PCR) was performed using Power SYBR Green PCR Master Mix from Thermo Fisher
Scientific and normalized to either β-actin or GAPDH. Primer sequences: CD9 F-
5’TGTTCTTCGGCTTCCTCTTG-3’, CD9 R-5’CAAACCACAGCAGTTCAACG-3’;
Actin F-5’CATGTACGTTGCTATCCAGGC-3’, R-
5’CTCCTTAATGTCACGCACGAT-3’; GAPDH F-
5’AAGGTGAAGGTCGGAGTCAAC-3’, R-5’GGGGTCATTGATGGCAACAATA-3’;
BRCA1 F-5’TGGAAGAAACCACCAAGGTC-3’, R-
5’ACCACAGAAGCACCACACAG-3’.
Colony formation assay
Radiation was performed with a 137Cs irradiator (Shepherd). A total of 500 to
10,000 cells per plate were stained after 10 days with 0.1% crystal violet. Assays were done ≥ 3 times with individual samples in triplicate. Clonogenic assays of neurosphere lines were plated in 1 mL of neuro stem cell (NSA) media (100) containing 1.5% methylcellulose and fed with NSA media every 3 days. Sphere formation was monitored and scored using GelCount (Oxford Optronix) after 10-
12 days.
140 Cell Proliferation Assay
20,000 cells/well were plated into a 12 well dish and allowed to grow for 3 days.
On day 3, cells were counted, and 20,000 cells were replated. The process was repeated for at least three times points.
Tumor Formation Assay
Animal studies were performed in accordance with CWRU institutional guidelines. Subcutaneous U87MG knockdown tumors were produced by injection of 1x106 cells in the flanks of eight-week-old nude mice, and measured with calipers twice weekly. Animal appearance, behavior, and weight were monitored to evaluate tumor progression as per a Case Western Reserve
University approved IACUC protocol.
HR reporter assay
GBM U87 DRGFP cells were transfected with siRNA using DharmaFECT. After
24 hours, cells were infected at 10 MOI with I-SceI-expressing adenovirus or control (empty) virus (gift of Dr. Junran Zhang). Cells were incubated for an additional 48 hours and then analyzed by flow cytometry as described previously
(144).
Western Blots
Western blots were performed using standard procedures. Antibodies against
CD9 (Abcam ab92726), β-actin (Sigma A1987), Phospho-AKT Ser473 (Cell signaling #9271), and AKT (Cell signaling #9272) were used to decorate the membranes, and developed by chemiluminescence on an Azure c300 imaging system.
141 Immunofluorescent Staining
GBM cells were fixed in 3% paraformaldehyde and stained and analyzed as described previously (37). Antibody: Phospho-Histone H2A.X (Ser139) (Millipore
05-636), secondary Alexa-Fluor 594 anti-mouse (A11032; Invitrogen).
Statistical analyses
Student t tests and two-way ANOVA were used throughout the study to test the significance of differences between samples. Survival analyses were performed by log-rank tests in GraphPad Prism.
142 Appendix C
Experimental Procedures Related to Chapter 4
143 Cells and neurospheres
U87MG and LN229 (ATCC), and Gli36 (gift of Dr. Susanne Brady-Kalnay) cells were maintained in DMEM 10% FCS. GBM0821 and GBM0913 (100) neurospheres were grown in neuro stem cell media (100). shRNAs were acquired from Sigma (below). For neurosphere quantification, the media contained 1.5% methylcellulose; spheres were scored after 12 days with
Metamorph software. pCSII-IB-MELK was a gift of Dr. Takeuchi (146).
Microarray analyses
RNA was processed and hybridized at the CWRU Integrated Genomics Shared
Resource. Analyses were performed using Transciptome Analysis Console
(Affymetrix), the PANTHER Classification System (www.patherdb.org), and the
Xena Functional Genomics Explorer (xenabrower.net).
Statistics
Statistical tests were performed with Graphpad Prism 7.0 as indicated for each figure. All assays were repeated with both technical and biological replicates
(n≥3).
ChIP/RT-PCR/western blotting
ChIP was described previously (147). For antibodies and RT-PCR primers, are below.
CRISPR
Knockin vectors were constructed in pBSKSII+ with a BSR cassette flanking mutated exons 4-6. The gRNA sequence targeting the PAM sequence
GGGAATACGTCAGGTTTACACGG within intron 3 of SAT1 was used to design
144 an sgRNA by in vitro transcription, and was transfected with CAS9 and linearized vector into U87MG cells. Individual blasticidin-resistant clones were screened by
PCR. Recombination was induced with Ad5CMVCre-eGFP (University of Iowa).
Appendix C Supplementary Method Table 1: Antibodies
Target Protein Company Catalogue number
SAT1 AbCam 105220
Histone H3 Cell Signaling 4620S
MELK Sigma HPA017214
FOXM1 Thermo Fisher Scientific PA5-27144
Β-Actin Sigma A1978
BRCA1 Santa Cruz SC-6954
FEN1 Novus Biologicale NB100-150
145 Appendix C Supplementary Method Table 2: Primers for qPCR
Primers Forward Reverse for qPCR
SAT1 GCTGATCAAGGAGCTGGCTA CAACAATGCTGTGTCCTTCC
GAPDH AAGGTGAAGGTCGGAGTCAA GGGGTCATTGATGGCAACAAT
C A
MELK CAGGCAAACAATGGAGGATT TCACTTGCGGTCACATCTTC
BRCA1 TGGAAGAAACCACCAAGGTC ACCACAGAAGCACCACACAG
EZH2 TTCATGCAACACCCAACACT GGGCCTGCTACTGTTATTGG
FANCI GGTTTTTGCTGCTCCTGAAG TGAGTCGAACATCAGCTTGC
USCP1 TGGGCATTACACTGCTTCTG TTGGCTGTGTATTTCCACCA
CLASP GTGGGGTCCATTCATTTGAG CGGGGTTTACGTTTGAAGAA
FOXM1 CCTCAAACCCAAACCAGCTA GAAGCCACTGGATGTTGGAT
FANCD2 CTGTCCCAATCCTGGATGTC GCTGCAGATCCAACTTCTCC
XRCC2 TTCCATAGGGCTGAGTCTGG AGGCCACCTTCTGATTTGG
KIAA154 AAGCTCTAGCCCTTGCACAG GTCCGTGCCTCTGTTTCAGC
2 G
PLK1 AAGAGATCCCGGAGGTCCTA GCTGCGGTGAATGGATATTT
BIRC3 ccgggtacagaaaacagtgg tgcttttgccagatctgttg
NEK2 gaatgccacagacgaagtga CGGGGTTTACGTTTGAAGAA
TTK gaacatgccaccacaagatg caaccaaatctcggcattct
NUSAP1 gtttgtctcgtcccctcaac cgtttcttccgttgctcttc
146
Appendix C Supplementary Method Table 3: Primers for Chromatin Immunoprecipitations
Primers Forward Reverse for ChIP
MELK TGGGGTTCTTATTCCTCGTG TGACTTGAGGCCTTGGTCTT
FOXM1 AGGGGAAGGAAAGAACCTTG CTCCGCTTTCTTCCATCTTG
EZH2 CTGGTTCAAACTTGGCTTCC CGCCGTCTCTTTGTTCTTTC
BRCA1 GGCAGGCATTTATGGCAAAC TTCGGAAATCCACTCTCCCACG
NUSAP1 CCTGGCCGTAAATATCAGGA TCCAGTGAGTGCAGTGAAGC
CCNB1 CCCTGGAAACGCATTCTCT GGACCTACACCCAGCAGAAA
CCNB2 TTGTCTTGGCCAATGAGAAC CGGACTGAAAAGGGAGGAC
CDC20 CCGCTAGACTCTCGTGATAGC GCTTTAACACGCCTGGCTTA
147 References
1. Crespo I, Vital AL, Gonzalez-Tablas M, Patino Mdel C, Otero A, Lopes
MC, de Oliveira C, Domingues P, Orfao A, Tabernero MD. Molecular and
Genomic Alterations in Glioblastoma Multiforme. Am J Pathol. 2015;185(7):1820-
33.
2. Miller CR, Perry A. Glioblastoma. Arch Pathol Lab Med. 2007;131(3):397-
406.
3. Thakkar JP, Dolecek TA, Horbinski C, Ostrom QT, Lightner DD,
Barnholtz-Sloan JS, Villano JL. Epidemiologic and molecular prognostic review of glioblastoma. Cancer Epidemiol Biomarkers Prev. 2014;23(10):1985-96.
4. Ostrom QT, Gittleman H, Liao P, Rouse C, Chen Y, Dowling J, Wolinsky
Y, Kruchko C, Barnholtz-Sloan J. CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2007-2011.
Neuro Oncol. 2014;16 Suppl 4:iv1-63.
5. Alifieris C, Trafalis DT. Glioblastoma multiforme: Pathogenesis and treatment. Pharmacol Ther. 2015;152:63-82.
6. Wen PY, Kesari S. Malignant gliomas in adults. The New England journal of medicine. 2008;359(5):492-507.
7. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ,
Belanger K, Brandes AA, Marosi C, Bogdahn U, Curschmann J, Janzer RC,
Ludwin SK, Gorlia T, Allgeier A, Lacombe D, Cairncross JG, Eisenhauer E,
Mirimanoff RO, European Organisation for R, Treatment of Cancer Brain T,
148 Radiotherapy G, National Cancer Institute of Canada Clinical Trials G.
Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. The
New England journal of medicine. 2005;352(10):987-96.
8. Fernandes C, Costa A, Osorio L, Lago RC, Linhares P, Carvalho B,
Caeiro C. Current Standards of Care in Glioblastoma Therapy. In: De
Vleeschouwer S, editor. Glioblastoma. Brisbane (AU)2017.
9. Bianco J, Bastiancich C, Jankovski A, des Rieux A, Preat V, Danhier F.
On glioblastoma and the search for a cure: where do we stand? Cellular and molecular life sciences : CMLS. 2017;74(13):2451-66.
10. Tykocki T, Eltayeb M. Ten-year survival in glioblastoma. A systematic review. J Clin Neurosci. 2018;54:7-13.
11. Stoyanov GS, Dzhenkov D, Ghenev P, Iliev B, Enchev Y, Tonchev AB.
Cell biology of glioblastoma multiforme: from basic science to diagnosis and treatment. Med Oncol. 2018;35(3):27.
12. Krex D, Klink B, Hartmann C, von Deimling A, Pietsch T, Simon M, Sabel
M, Steinbach JP, Heese O, Reifenberger G, Weller M, Schackert G, German
Glioma N. Long-term survival with glioblastoma multiforme. Brain : a journal of neurology. 2007;130(Pt 10):2596-606.
13. Lee E, Yong RL, Paddison P, Zhu J. Comparison of glioblastoma (GBM) molecular classification methods. Semin Cancer Biol. 2018;53:201-11.
14. Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, Miller
CR, Ding L, Golub T, Mesirov JP, Alexe G, Lawrence M, O'Kelly M, Tamayo P,
Weir BA, Gabriel S, Winckler W, Gupta S, Jakkula L, Feiler HS, Hodgson JG,
149 James CD, Sarkaria JN, Brennan C, Kahn A, Spellman PT, Wilson RK, Speed
TP, Gray JW, Meyerson M, Getz G, Perou CM, Hayes DN, Cancer Genome
Atlas Research N. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1,
EGFR, and NF1. Cancer Cell. 2010;17(1):98-110.
15. Teo WY, Sekar K, Seshachalam P, Shen J, Chow WY, Lau CC, Yang H,
Park J, Kang SG, Li X, Nam DH, Hui KM. Relevance of a TCGA-derived
Glioblastoma Subtype Gene-Classifier among Patient Populations. Sci Rep.
2019;9(1):7442.
16. Dunn GP, Rinne ML, Wykosky J, Genovese G, Quayle SN, Dunn IF,
Agarwalla PK, Chheda MG, Campos B, Wang A, Brennan C, Ligon KL, Furnari
F, Cavenee WK, Depinho RA, Chin L, Hahn WC. Emerging insights into the molecular and cellular basis of glioblastoma. Genes & development.
2012;26(8):756-84.
17. Noch EK, Ramakrishna R, Magge R. Challenges in the Treatment of
Glioblastoma: Multisystem Mechanisms of Therapeutic Resistance. World
Neurosurg. 2018;116:505-17.
18. Daneman R, Prat A. The blood-brain barrier. Cold Spring Harb Perspect
Biol. 2015;7(1):a020412.
19. Sarkaria JN, Hu LS, Parney IF, Pafundi DH, Brinkmann DH, Laack NN,
Giannini C, Burns TC, Kizilbash SH, Laramy JK, Swanson KR, Kaufmann TJ,
Brown PD, Agar NYR, Galanis E, Buckner JC, Elmquist WF. Is the blood-brain
150 barrier really disrupted in all glioblastomas? A critical assessment of existing clinical data. Neuro Oncol. 2018;20(2):184-91.
20. Lathia JD, Mack SC, Mulkearns-Hubert EE, Valentim CL, Rich JN. Cancer stem cells in glioblastoma. Genes & development. 2015;29(12):1203-17.
21. Fan CH, Liu WL, Cao H, Wen C, Chen L, Jiang G. O6-methylguanine
DNA methyltransferase as a promising target for the treatment of temozolomide- resistant gliomas. Cell Death Dis. 2013;4:e876.
22. Han X, Xue X, Zhou H, Zhang G. A molecular view of the radioresistance of gliomas. Oncotarget. 2017;8(59):100931-41.
23. Cho DY, Lin SZ, Yang WK, Lee HC, Hsu DM, Lin HL, Chen CC, Liu CL,
Lee WY, Ho LH. Targeting cancer stem cells for treatment of glioblastoma multiforme. Cell Transplant. 2013;22(4):731-9.
24. Rycaj K, Tang DG. Cancer stem cells and radioresistance. Int J Radiat
Biol. 2014;90(8):615-21.
25. Lee Y, Kim KH, Kim DG, Cho HJ, Kim Y, Rheey J, Shin K, Seo YJ, Choi
YS, Lee JI, Lee J, Joo KM, Nam DH. FoxM1 Promotes Stemness and Radio-
Resistance of Glioblastoma by Regulating the Master Stem Cell Regulator Sox2.
PloS one. 2015;10(10):e0137703.
26. Kao GD, Jiang Z, Fernandes AM, Gupta AK, Maity A. Inhibition of phosphatidylinositol-3-OH kinase/Akt signaling impairs DNA repair in glioblastoma cells following ionizing radiation. The Journal of biological chemistry. 2007;282(29):21206-12.
151 27. Bhat KP, Balasubramaniyan V, Vaillant B, Ezhilarasan R, Hummelink K,
Hollingsworth F, Wani K, Heathcock L, James JD, Goodman LD, Conroy S, Long
L, Lelic N, Wang S, Gumin J, Raj D, Kodama Y, Raghunathan A, Olar A, Joshi K,
Pelloski CE, Heimberger A, Kim SH, Cahill DP, Rao G, Den Dunnen WF,
Boddeke HW, Phillips HS, Nakano I, Lang FF, Colman H, Sulman EP, Aldape K.
Mesenchymal differentiation mediated by NF-kappaB promotes radiation resistance in glioblastoma. Cancer Cell. 2013;24(3):331-46.
28. Zhang M, Kleber S, Rohrich M, Timke C, Han N, Tuettenberg J, Martin-
Villalba A, Debus J, Peschke P, Wirkner U, Lahn M, Huber PE. Blockade of TGF- beta signaling by the TGFbetaR-I kinase inhibitor LY2109761 enhances radiation response and prolongs survival in glioblastoma. Cancer Res. 2011;71(23):7155-
67.
29. Wang J, Wakeman TP, Lathia JD, Hjelmeland AB, Wang XF, White RR,
Rich JN, Sullenger BA. Notch promotes radioresistance of glioma stem cells.
Stem Cells. 2010;28(1):17-28.
30. Clark MJ, Homer N, O'Connor BD, Chen Z, Eskin A, Lee H, Merriman B,
Nelson SF. U87MG decoded: the genomic sequence of a cytogenetically aberrant human cancer cell line. PLoS Genet. 2010;6(1):e1000832.
31. Ishii N, Maier D, Merlo A, Tada M, Sawamura Y, Diserens AC, Van Meir
EG. Frequent co-alterations of TP53, p16/CDKN2A, p14ARF, PTEN tumor suppressor genes in human glioma cell lines. Brain Pathol. 1999;9(3):469-79.
152 32. Breznik B, Motaln H, Vittori M, Rotter A, Lah Turnsek T. Mesenchymal stem cells differentially affect the invasion of distinct glioblastoma cell lines.
Oncotarget. 2017;8(15):25482-99.
33. Hai L, Zhang C, Li T, Zhou X, Liu B, Li S, Zhu M, Lin Y, Yu S, Zhang K,
Ren B, Ming H, Huang Y, Chen L, Zhao P, Zhou H, Jiang T, Yang X. Notch1 is a prognostic factor that is distinctly activated in the classical and proneural subtype of glioblastoma and that promotes glioma cell survival via the NF-kappaB(p65) pathway. Cell Death Dis. 2018;9(2):158.
34. Hong X, Chedid K, Kalkanis SN. Glioblastoma cell line-derived spheres in serumcontaining medium versus serum-free medium: a comparison of cancer stem cell properties. Int J Oncol. 2012;41(5):1693-700.
35. Lane R, Simon T, Vintu M, Solkin B, Koch B, Stewart N, Benstead-Hume
G, Pearl FMG, Critchley G, Stebbing J, Giamas G. Cell-derived extracellular vesicles can be used as a biomarker reservoir for glioblastoma tumor subtyping.
Commun Biol. 2019;2:315.
36. McGillicuddy LT, Fromm JA, Hollstein PE, Kubek S, Beroukhim R, De
Raedt T, Johnson BW, Williams SM, Nghiemphu P, Liau LM, Cloughesy TF,
Mischel PS, Parret A, Seiler J, Moldenhauer G, Scheffzek K, Stemmer-
Rachamimov AO, Sawyers CL, Brennan C, Messiaen L, Mellinghoff IK,
Cichowski K. Proteasomal and genetic inactivation of the NF1 tumor suppressor in gliomagenesis. Cancer Cell. 2009;16(1):44-54.
37. Brett-Morris A, Wright BM, Seo Y, Pasupuleti V, Zhang J, Lu J, Spina R,
Bar EE, Gujrati M, Schur R, Lu ZR, Welford SM. The polyamine catabolic
153 enzyme SAT1 modulates tumorigenesis and radiation response in GBM. Cancer
Res. 2014;74(23):6925-34.
38. Hemler ME. Tetraspanin functions and associated microdomains. Nat Rev
Mol Cell Biol. 2005;6(10):801-11.
39. Yanez-Mo M, Barreiro O, Gordon-Alonso M, Sala-Valdes M, Sanchez-
Madrid F. Tetraspanin-enriched microdomains: a functional unit in cell plasma membranes. Trends Cell Biol. 2009;19(9):434-46.
40. Charrin S, Manie S, Billard M, Ashman L, Gerlier D, Boucheix C,
Rubinstein E. Multiple levels of interactions within the tetraspanin web. Biochem
Biophys Res Commun. 2003;304(1):107-12.
41. Le Naour F, Andre M, Boucheix C, Rubinstein E. Membrane microdomains and proteomics: lessons from tetraspanin microdomains and comparison with lipid rafts. Proteomics. 2006;6(24):6447-54.
42. Richardson MM, Jennings LK, Zhang XA. Tetraspanins and tumor progression. Clin Exp Metastasis. 2011;28(3):261-70.
43. Charrin S, le Naour F, Silvie O, Milhiet PE, Boucheix C, Rubinstein E.
Lateral organization of membrane proteins: tetraspanins spin their web. Biochem
J. 2009;420(2):133-54.
44. Hemler ME. Tetraspanin proteins mediate cellular penetration, invasion, and fusion events and define a novel type of membrane microdomain. Annual review of cell and developmental biology. 2003;19:397-422.
45. Termini CM, Gillette JM. Tetraspanins Function as Regulators of Cellular
Signaling. Front Cell Dev Biol. 2017;5:34.
154 46. Odintsova E, Sugiura T, Berditchevski F. Attenuation of EGF receptor signaling by a metastasis suppressor, the tetraspanin CD82/KAI-1. Curr Biol.
2000;10(16):1009-12.
47. Stipp CS, Orlicky D, Hemler ME. FPRP, a major, highly stoichiometric, highly specific CD81- and CD9-associated protein. The Journal of biological chemistry. 2001;276(7):4853-62.
48. Orlicky DJ, Nordeen SK. Cloning, sequencing and proposed structure for a prostaglandin F2 alpha receptor regulatory protein. Prostaglandins Leukot
Essent Fatty Acids. 1996;55(4):261-8.
49. Orlicky DJ. Negative regulatory activity of a prostaglandin F2 alpha receptor associated protein (FPRP). Prostaglandins Leukot Essent Fatty Acids.
1996;54(4):247-59.
50. Charrin S, Le Naour F, Oualid M, Billard M, Faure G, Hanash SM,
Boucheix C, Rubinstein E. The major CD9 and CD81 molecular partner.
Identification and characterization of the complexes. The Journal of biological chemistry. 2001;276(17):14329-37.
51. Colin S, Guilmain W, Creoff E, Schneider C, Steverlynck C, Bongaerts M,
Legrand E, Vannier JP, Muraine M, Vasse M, Al-Mahmood S. A truncated form of CD9-partner 1 (CD9P-1), GS-168AT2, potently inhibits in vivo tumour-induced angiogenesis and tumour growth. Br J Cancer. 2011;105(7):1002-11.
52. Guilmain W, Colin S, Legrand E, Vannier JP, Steverlynck C, Bongaerts M,
Vasse M, Al-Mahmood S. CD9P-1 expression correlates with the metastatic
155 status of lung cancer, and a truncated form of CD9P-1, GS-168AT2, inhibits in vivo tumour growth. Br J Cancer. 2011;104(3):496-504.
53. Orlicky DJ, Lieber JG, Morin CL, Evans RM. Synthesis and accumulation of a receptor regulatory protein associated with lipid droplet accumulation in 3T3-
L1 cells. J Lipid Res. 1998;39(6):1152-61.
54. Sala-Valdes M, Ursa A, Charrin S, Rubinstein E, Hemler ME, Sanchez-
Madrid F, Yanez-Mo M. EWI-2 and EWI-F link the tetraspanin web to the actin cytoskeleton through their direct association with ezrin-radixin-moesin proteins.
The Journal of biological chemistry. 2006;281(28):19665-75.
55. Kitadokoro K, Bordo D, Galli G, Petracca R, Falugi F, Abrignani S, Grandi
G, Bolognesi M. CD81 extracellular domain 3D structure: insight into the tetraspanin superfamily structural motifs. The EMBO journal. 2001;20(1-2):12-8.
56. Seigneuret M, Delaguillaumie A, Lagaudriere-Gesbert C, Conjeaud H.
Structure of the tetraspanin main extracellular domain. A partially conserved fold with a structurally variable domain insertion. The Journal of biological chemistry.
2001;276(43):40055-64.
57. Maecker HT, Todd SC, Levy S. The tetraspanin superfamily: molecular facilitators. FASEB journal : official publication of the Federation of American
Societies for Experimental Biology. 1997;11(6):428-42.
58. Hemler ME. Tetraspanin proteins promote multiple cancer stages. Nature
Reviews Cancer. 2013;14(1):49-60.
59. Romanska HM, Berditchevski F. Tetraspanins in human epithelial malignancies. J Pathol. 2011;223(1):4-14.
156 60. Hori H, Yano S, Koufuji K, Takeda J, Shirouzu K. CD9 expression in gastric cancer and its significance. J Surg Res. 2004;117(2):208-15.
61. Soyuer S, Soyuer I, Unal D, Ucar K, Yildiz OG, Orhan O. Prognostic significance of CD9 expression in locally advanced gastric cancer treated with surgery and adjuvant chemoradiotherapy. Pathol Res Pract. 2010;206(9):607-10.
62. Yamazaki H, Xu CW, Naito M, Nishida H, Okamoto T, Ghani FI, Iwata S,
Inukai T, Sugita K, Morimoto C. Regulation of cancer stem cell properties by CD9 in human B-acute lymphoblastic leukemia. Biochem Biophys Res Commun.
2011;409(1):14-21.
63. Hwang JR, Jo K, Lee Y, Sung BJ, Park YW, Lee JH. Upregulation of CD9 in ovarian cancer is related to the induction of TNF-alpha gene expression and constitutive NF-kappaB activation. Carcinogenesis. 2012;33(1):77-83.
64. Podergajs N, Motaln H, Rajcevic U, Verbovsek U, Korsic M, Obad N,
Espedal H, Vittori M, Herold-Mende C, Miletic H, Bjerkvig R, Turnsek TL.
Transmembrane protein CD9 is glioblastoma biomarker, relevant for maintenance of glioblastoma stem cells. Oncotarget. 2016;7(1):593-609.
65. Shi Y, Zhou W, Cheng L, Chen C, Huang Z, Fang X, Wu Q, He Z, Xu S,
Lathia JD, Ping Y, Rich JN, Bian XW, Bao S. Tetraspanin CD9 stabilizes gp130 by preventing its ubiquitin-dependent lysosomal degradation to promote STAT3 activation in glioma stem cells. Cell Death Differ. 2017;24(1):167-80.
66. Li X, Wu C, Chen N, Gu H, Yen A, Cao L, Wang E, Wang L.
PI3K/Akt/mTOR signaling pathway and targeted therapy for glioblastoma.
Oncotarget. 2016;7(22):33440-50.
157 67. Zhao HF, Wang J, Shao W, Wu CP, Chen ZP, To ST, Li WP. Recent advances in the use of PI3K inhibitors for glioblastoma multiforme: current preclinical and clinical development. Mol Cancer. 2017;16(1):100.
68. Cancer Genome Atlas Research N. Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature.
2008;455(7216):1061-8.
69. Franke TF. PI3K/Akt: getting it right matters. Oncogene.
2008;27(50):6473-88.
70. Li HF, Kim JS, Waldman T. Radiation-induced Akt activation modulates radioresistance in human glioblastoma cells. Radiat Oncol. 2009;4:43.
71. Chautard E, Loubeau G, Tchirkov A, Chassagne J, Vermot-Desroches C,
Morel L, Verrelle P. Akt signaling pathway: a target for radiosensitizing human malignant glioma. Neuro Oncol. 2010;12(5):434-43.
72. Golding SE, Morgan RN, Adams BR, Hawkins AJ, Povirk LF, Valerie K.
Pro-survival AKT and ERK signaling from EGFR and mutant EGFRvIII enhances
DNA double-strand break repair in human glioma cells. Cancer Biol Ther.
2009;8(8):730-8.
73. Heimberger AB, Suki D, Yang D, Shi W, Aldape K. The natural history of
EGFR and EGFRvIII in glioblastoma patients. J Transl Med. 2005;3:38.
74. Mukherjee B, McEllin B, Camacho CV, Tomimatsu N, Sirasanagandala S,
Nannepaga S, Hatanpaa KJ, Mickey B, Madden C, Maher E, Boothman DA,
Furnari F, Cavenee WK, Bachoo RM, Burma S. EGFRvIII and DNA double-
158 strand break repair: a molecular mechanism for radioresistance in glioblastoma.
Cancer Res. 2009;69(10):4252-9.
75. Netland IA, Forde HE, Sleire L, Leiss L, Rahman MA, Skeie BS, Miletic H,
Enger PO, Goplen D. Treatment with the PI3K inhibitor buparlisib (NVP-BKM120) suppresses the growth of established patient-derived GBM xenografts and prolongs survival in nude rats. J Neurooncol. 2016;129(1):57-66.
76. Wen PY, Yung WKA, Mellinghoff IK, Ramkissoon S, Alexander BM, Rinne
ML, Colman H, Omuro AMP, DeAngelis LM, Gilbert MR, Groot JFD, Cloughesy
TF, Chi AS, Lee EQ, Nayak L, Batchelor T, Chang SM, Prados M, Reardon DA,
Ligon KL. Phase II trial of the phosphatidyinositol-3 kinase (PI3K) inhibitor buparlisib (BKM120) in recurrent glioblastoma. Journal of Clinical Oncology.
2014;32(15_suppl):2019-.
77. Cloughesy TF, Mischel PS, Omuro AMP, Prados M, Wen PY, Wu B,
Rockich K, Xu Y, Lager JJ, Mellinghoff IK. Tumor pharmacokinetics (PK) and pharmacodynamics (PD) of SAR245409 (XL765) and SAR245408 (XL147) administered as single agents to patients with recurrent glioblastoma (GBM): An
Ivy Foundation early-phase clinical trials consortium study. Journal of Clinical
Oncology. 2013;31(15_suppl):2012-.
78. Foster P, Yamaguchi K, Hsu PP, Qian F, Du X, Wu J, Won KA, Yu P,
Jaeger CT, Zhang W, Marlowe CK, Keast P, Abulafia W, Chen J, Young J,
Plonowski A, Yakes FM, Chu F, Engell K, Bentzien F, Lam ST, Dale S, Yturralde
O, Matthews DJ, Lamb P, Laird AD. The Selective PI3K Inhibitor XL147
(SAR245408) Inhibits Tumor Growth and Survival and Potentiates the Activity of
159 Chemotherapeutic Agents in Preclinical Tumor Models. Mol Cancer Ther.
2015;14(4):931-40.
79. Koul D, Shen R, Kim YW, Kondo Y, Lu Y, Bankson J, Ronen SM,
Kirkpatrick DL, Powis G, Yung WK. Cellular and in vivo activity of a novel PI3K inhibitor, PX-866, against human glioblastoma. Neuro Oncol. 2010;12(6):559-69.
80. Salphati L, Shahidi-Latham S, Quiason C, Barck K, Nishimura M, Alicke B,
Pang J, Carano RA, Olivero AG, Phillips HS. Distribution of the phosphatidylinositol 3-kinase inhibitors Pictilisib (GDC-0941) and GNE-317 in
U87 and GS2 intracranial glioblastoma models-assessment by matrix-assisted laser desorption ionization imaging. Drug Metab Dispos. 2014;42(7):1110-6.
81. Nonnenmacher L, Westhoff MA, Fulda S, Karpel-Massler G, Halatsch ME,
Engelke J, Simmet T, Corbacioglu S, Debatin KM. RIST: a potent new combination therapy for glioblastoma. Int J Cancer. 2015;136(4):E173-87.
82. Kaley TJ, Panageas KS, Mellinghoff IK, Nolan C, Gavrilovic IT, DeAngelis
LM, Abrey LE, Holland EC, Lassman AB. Phase II trial of an AKT inhibitor
(perifosine) for recurrent glioblastoma. J Neurooncol. 2019;144(2):403-7.
83. Fruman DA, Rommel C. PI3K and cancer: lessons, challenges and opportunities. Nat Rev Drug Discov. 2014;13(2):140-56.
84. Pegg AE. Spermidine/spermine-N(1)-acetyltransferase: a key metabolic regulator. Am J Physiol Endocrinol Metab. 2008;294(6):E995-1010.
85. Thakur VS, Aguila B, Brett-Morris A, Creighton CJ, Welford SM.
Spermidine/spermine N1-acetyltransferase 1 is a gene-specific transcriptional
160 regulator that drives brain tumor aggressiveness. Oncogene. 2019;38(41):6794-
800.
86. Veeravalli KK, Ponnala S, Chetty C, Tsung AJ, Gujrati M, Rao JS. Integrin alpha9beta1-mediated cell migration in glioblastoma via SSAT and Kir4.2 potassium channel pathway. Cell Signal. 2012;24(1):272-81.
87. Bao S, Wu Q, McLendon RE, Hao Y, Shi Q, Hjelmeland AB, Dewhirst
MW, Bigner DD, Rich JN. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature.
2006;444(7120):756-60.
88. Chakravarti A, Zhai GG, Zhang M, Malhotra R, Latham DE, Delaney MA,
Robe P, Nestler U, Song Q, Loeffler J. Survivin enhances radiation resistance in primary human glioblastoma cells via caspase-independent mechanisms.
Oncogene. 2004;23(45):7494-506.
89. Hatanpaa KJ, Burma S, Zhao D, Habib AA. Epidermal growth factor receptor in glioma: signal transduction, neuropathology, imaging, and radioresistance. Neoplasia. 2010;12(9):675-84.
90. Chambrion C, Le Naour F. The tetraspanins CD9 and CD81 regulate
CD9P1-induced effects on cell migration. PloS one. 2010;5(6):e11219.
91. Andre M, Chambrion C, Charrin S, Soave S, Chaker J, Boucheix C,
Rubinstein E, Le Naour F. In situ chemical cross-linking on living cells reveals
CD9P-1 cis-oligomer at cell surface. J Proteomics. 2009;73(1):93-102.
92. Vivanco I, Sawyers CL. The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nature reviews Cancer. 2002;2(7):489-501.
161 93. Chakravarti A, Zhai G, Suzuki Y, Sarkesh S, Black PM, Muzikansky A,
Loeffler JS. The prognostic significance of phosphatidylinositol 3-kinase pathway activation in human gliomas. J Clin Oncol. 2004;22(10):1926-33.
94. Nakamura JL, Karlsson A, Arvold ND, Gottschalk AR, Pieper RO, Stokoe
D, Haas-Kogan DA. PKB/Akt mediates radiosensitization by the signaling inhibitor LY294002 in human malignant gliomas. J Neurooncol. 2005;71(3):215-
22.
95. Saito Y, Tachibana I, Takeda Y, Yamane H, He P, Suzuki M, Minami S,
Kijima T, Yoshida M, Kumagai T, Osaki T, Kawase I. Absence of CD9 enhances adhesion-dependent morphologic differentiation, survival, and matrix metalloproteinase-2 production in small cell lung cancer cells. Cancer Res.
2006;66(19):9557-65.
96. Sugiura T, Berditchevski F. Function of alpha3beta1-tetraspanin protein complexes in tumor cell invasion. Evidence for the role of the complexes in production of matrix metalloproteinase 2 (MMP-2). J Cell Biol. 1999;146(6):1375-
89.
97. Rhodes DR, Kalyana-Sundaram S, Mahavisno V, Varambally R, Yu J,
Briggs BB, Barrette TR, Anstet MJ, Kincead-Beal C, Kulkarni P, Varambally S,
Ghosh D, Chinnaiyan AM. Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles. Neoplasia. 2007;9(2):166-
80.
162 98. Mizuno H, Kitada K, Nakai K, Sarai A. PrognoScan: a new database for meta-analysis of the prognostic value of genes. BMC medical genomics.
2009;2:18.
99. Behnan J, Finocchiaro G, Hanna G. The landscape of the mesenchymal signature in brain tumours. Brain : a journal of neurology. 2019;142(4):847-66.
100. Galli R, Binda E, Orfanelli U, Cipelletti B, Gritti A, De Vitis S, Fiocco R,
Foroni C, Dimeco F, Vescovi A. Isolation and characterization of tumorigenic, stem-like neural precursors from human glioblastoma. Cancer Res.
2004;64(19):7011-21.
101. Hemmati HD, Nakano I, Lazareff JA, Masterman-Smith M, Geschwind DH,
Bronner-Fraser M, Kornblum HI. Cancerous stem cells can arise from pediatric brain tumors. Proceedings of the National Academy of Sciences of the United
States of America. 2003;100(25):15178-83.
102. Ignatova TN, Kukekov VG, Laywell ED, Suslov ON, Vrionis FD, Steindler
DA. Human cortical glial tumors contain neural stem-like cells expressing astroglial and neuronal markers in vitro. Glia. 2002;39(3):193-206.
103. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, Henkelman
RM, Cusimano MD, Dirks PB. Identification of human brain tumour initiating cells.
Nature. 2004;432(7015):396-401.
104. Pastrana E, Silva-Vargas V, Doetsch F. Eyes wide open: a critical review of sphere-formation as an assay for stem cells. Cell Stem Cell. 2011;8(5):486-98.
105. Awan MJ, Dorth J, Mani A, Kim H, Zheng Y, Mislmani M, Welford S, Yuan
J, Wessels BW, Lo SS, Letterio J, Machtay M, Sloan A, Sohn JW. Development
163 and Validation of a Small Animal Immobilizer and Positioning System for the
Study of Delivery of Intracranial and Extracranial Radiotherapy Using the Gamma
Knife System. Technol Cancer Res Treat. 2016.
106. Chu G. Double strand break repair. The Journal of biological chemistry.
1997;272(39):24097-100.
107. Kumar A, Fernandez-Capetillo O, Carrera AC. Nuclear phosphoinositide
3-kinase beta controls double-strand break DNA repair. Proceedings of the
National Academy of Sciences of the United States of America.
2010;107(16):7491-6.
108. Pridham KJ, Le L, Guo S, Varghese RT, Algino S, Liang Y, Fajardin R,
Rodgers CM, Simonds GR, Kelly DF, Sheng Z. PIK3CB/p110beta is a selective survival factor for glioblastoma. Neuro Oncol. 2018;20(4):494-505.
109. Pu P, Kang C, Zhang Z, Liu X, Jiang H. Downregulation of PIK3CB by siRNA suppresses malignant glioma cell growth in vitro and in vivo. Technol
Cancer Res Treat. 2006;5(3):271-80.
110. Yu J, Zhang Y, McIlroy J, Rordorf-Nikolic T, Orr GA, Backer JM.
Regulation of the p85/p110 phosphatidylinositol 3'-kinase: stabilization and inhibition of the p110alpha catalytic subunit by the p85 regulatory subunit. Mol
Cell Biol. 1998;18(3):1379-87.
111. Kumar A, Redondo-Munoz J, Perez-Garcia V, Cortes I, Chagoyen M,
Carrera AC. Nuclear but not cytosolic phosphoinositide 3-kinase beta has an essential function in cell survival. Mol Cell Biol. 2011;31(10):2122-33.
164 112. Shen H, Hau E, Joshi S, Dilda PJ, McDonald KL. Sensitization of
Glioblastoma Cells to Irradiation by Modulating the Glucose Metabolism. Mol
Cancer Ther. 2015;14(8):1794-804.
113. Wang GP, Han XF. CD9 modulates proliferation of human glioblastoma cells via epidermal growth factor receptor signaling. Molecular medicine reports.
2015;12(1):1381-6.
114. Pegg AE, Casero RA, Jr. Current status of the polyamine research field.
Methods Mol Biol. 2011;720:3-35.
115. Holst CM, Nevsten P, Johansson F, Carlemalm E, Oredsson SM.
Subcellular distribution of spermidine/spermine N1-acetyltransferase. Cell Biol
Int. 2008;32(1):39-47.
116. deHart GW, Jin T, McCloskey DE, Pegg AE, Sheppard D. The alpha9beta1 integrin enhances cell migration by polyamine-mediated modulation of an inward-rectifier potassium channel. Proceedings of the National Academy of Sciences of the United States of America. 2008;105(20):7188-93.
117. Uemura T, Yerushalmi HF, Tsaprailis G, Stringer DE, Pastorian KE, Hawel
L, 3rd, Byus CV, Gerner EW. Identification and characterization of a diamine exporter in colon epithelial cells. The Journal of biological chemistry.
2008;283(39):26428-35.
118. Baek JH, Liu YV, McDonald KR, Wesley JB, Zhang H, Semenza GL.
Spermidine/spermine N(1)-acetyltransferase-1 binds to hypoxia-inducible factor-
1alpha (HIF-1alpha) and RACK1 and promotes ubiquitination and degradation of
HIF-1alpha. The Journal of biological chemistry. 2007;282(46):33358-66.
165 119. Lee SB, Park JH, Folk JE, Deck JA, Pegg AE, Sokabe M, Fraser CS, Park
MH. Inactivation of eukaryotic initiation factor 5A (eIF5A) by specific acetylation of its hypusine residue by spermidine/spermine acetyltransferase 1 (SSAT1).
Biochem J. 2011;433(1):205-13.
120. Yoo J, Kim H, Aksimentiev A, Ha T. Direct evidence for sequence- dependent attraction between double-stranded DNA controlled by methylation.
Nat Commun. 2016;7:11045.
121. Ceccarelli M, Barthel FP, Malta TM, Sabedot TS, Salama SR, Murray BA,
Morozova O, Newton Y, Radenbaugh A, Pagnotta SM, Anjum S, Wang J,
Manyam G, Zoppoli P, Ling S, Rao AA, Grifford M, Cherniack AD, Zhang H,
Poisson L, Carlotti CG, Jr., Tirapelli DP, Rao A, Mikkelsen T, Lau CC, Yung WK,
Rabadan R, Huse J, Brat DJ, Lehman NL, Barnholtz-Sloan JS, Zheng S, Hess K,
Rao G, Meyerson M, Beroukhim R, Cooper L, Akbani R, Wrensch M, Haussler D,
Aldape KD, Laird PW, Gutmann DH, Network TR, Noushmehr H, Iavarone A,
Verhaak RG. Molecular Profiling Reveals Biologically Discrete Subsets and
Pathways of Progression in Diffuse Glioma. Cell. 2016;164(3):550-63.
122. Kim SH, Joshi K, Ezhilarasan R, Myers TR, Siu J, Gu C, Nakano-Okuno
M, Taylor D, Minata M, Sulman EP, Lee J, Bhat KP, Salcini AE, Nakano I. EZH2 protects glioma stem cells from radiation-induced cell death in a MELK/FOXM1- dependent manner. Stem cell reports. 2015;4(2):226-38.
123. Joshi K, Banasavadi-Siddegowda Y, Mo X, Kim SH, Mao P, Kig C, Nardini
D, Sobol RW, Chow LM, Kornblum HI, Waclaw R, Beullens M, Nakano I. MELK-
166 dependent FOXM1 phosphorylation is essential for proliferation of glioma stem cells. Stem Cells. 2013;31(6):1051-63.
124. Wang J, Cheng P, Pavlyukov MS, Yu H, Zhang Z, Kim SH, Minata M,
Mohyeldin A, Xie W, Chen D, Goidts V, Frett B, Hu W, Li H, Shin YJ, Lee Y, Nam
DH, Kornblum HI, Wang M, Nakano I. Targeting NEK2 attenuates glioblastoma growth and radioresistance by destabilizing histone methyltransferase EZH2. J
Clin Invest. 2017;127(8):3075-89.
125. Chen X, Muller GA, Quaas M, Fischer M, Han N, Stutchbury B, Sharrocks
AD, Engeland K. The forkhead transcription factor FOXM1 controls cell cycle- dependent gene expression through an atypical chromatin binding mechanism.
Mol Cell Biol. 2013;33(2):227-36.
126. Sharma V, Malgulwar PB, Purkait S, Patil V, Pathak P, Agrawal R,
Kulshreshtha R, Mallick S, Julka PK, Suri A, Sharma BS, Suri V, Sharma MC,
Sarkar C. Genome-wide ChIP-seq analysis of EZH2-mediated H3K27me3 target gene profile highlights differences between low- and high-grade astrocytic tumors. Carcinogenesis. 2017;38(2):152-61.
127. Coleman CS, Huang H, Pegg AE. Structure and critical residues at the active site of spermidine/spermine-N1-acetyltransferase. Biochem J. 1996;316 (
Pt 3):697-701.
128. Huang L, Chen K, Cai ZP, Chen FC, Shen HY, Zhao WH, Yang SJ, Chen
XB, Tang GX, Lin X. DEPDC1 promotes cell proliferation and tumor growth via activation of E2F signaling in prostate cancer. Biochem Biophys Res Commun.
2017;490(3):707-12.
167 129. Santos JS, Fonseca NA, Vieira CP, Vieira J, Casares F. Phylogeny of the teashirt-related zinc finger (tshz) gene family and analysis of the developmental expression of tshz2 and tshz3b in the zebrafish. Dev Dyn. 2010;239(3):1010-8.
130. Pridham KJ, Varghese RT, Sheng Z. The Role of Class IA
Phosphatidylinositol-4,5-Bisphosphate 3-Kinase Catalytic Subunits in
Glioblastoma. Front Oncol. 2017;7:312.
131. Sun SQ, Gu X, Gao XS, Li Y, Yu H, Xiong W, Yu H, Wang W, Li Y, Teng
Y, Zhou D. Overexpression of AKR1C3 significantly enhances human prostate cancer cells resistance to radiation. Oncotarget. 2016;7(30):48050-8.
132. Wang S, Yang Q, Fung KM, Lin HK. AKR1C2 and AKR1C3 mediated prostaglandin D2 metabolism augments the PI3K/Akt proliferative signaling pathway in human prostate cancer cells. Mol Cell Endocrinol. 2008;289(1-2):60-
6.
133. Panagopoulos AT, Gomes RN, Almeida FG, da Costa Souza F, Veiga
JCE, Nicolaou A, Colquhoun A. The prostanoid pathway contains potential prognostic markers for glioblastoma. Prostaglandins Other Lipid Mediat.
2018;137:52-62.
134. Moreno M, Pedrosa L, Pare L, Pineda E, Bejarano L, Martinez J,
Balasubramaniyan V, Ezhilarasan R, Kallarackal N, Kim SH, Wang J, Audia A,
Conroy S, Marin M, Ribalta T, Pujol T, Herreros A, Tortosa A, Mira H, Alonso
MM, Gomez-Manzano C, Graus F, Sulman EP, Piao X, Nakano I, Prat A, Bhat
KP, de la Iglesia N. GPR56/ADGRG1 Inhibits Mesenchymal Differentiation and
Radioresistance in Glioblastoma. Cell Rep. 2017;21(8):2183-97.
168 135. Cahill KE, Morshed RA, Yamini B. Nuclear factor-kappaB in glioblastoma: insights into regulators and targeted therapy. Neuro Oncol. 2016;18(3):329-39.
136. Gan L, Yang Y, Li Q, Feng Y, Liu T, Guo W. Epigenetic regulation of cancer progression by EZH2: from biological insights to therapeutic potential.
Biomark Res. 2018;6:10.
137. Vastrad B, Vastrad C, Godavarthi A, Chandrashekar R. Molecular mechanisms underlying gliomas and glioblastoma pathogenesis revealed by bioinformatics analysis of microarray data. Med Oncol. 2017;34(11):182.
138. Gustafson-Wagner E, Stipp CS. The CD9/CD81 tetraspanin complex and tetraspanin CD151 regulate alpha3beta1 integrin-dependent tumor cell behaviors by overlapping but distinct mechanisms. PloS one. 2013;8(4):e61834.
139. Berditchevski F. Complexes of tetraspanins with integrins: more than meets the eye. J Cell Sci. 2001;114(Pt 23):4143-51.
140. Kotha J, Longhurst C, Appling W, Jennings LK. Tetraspanin CD9 regulates beta 1 integrin activation and enhances cell motility to fibronectin via a
PI-3 kinase-dependent pathway. Exp Cell Res. 2008;314(8):1811-22.
141. Spina R, Voss DM, Asnaghi L, Sloan A, Bar EE. Atracurium Besylate and other neuromuscular blocking agents promote astroglial differentiation and deplete glioblastoma stem cells. Oncotarget. 2016;7(1):459-72.
142. Voss DM, Spina R, Carter DL, Lim KS, Jeffery CJ, Bar EE. Disruption of the monocarboxylate transporter-4-basigin interaction inhibits the hypoxic response, proliferation, and tumor progression. Sci Rep. 2017;7(1):4292.
169 143. Yu Z, Huang, Z. and Lung, M. L. Subcellular Fractionation of Cultured
Human Cell Lines. Bio-protocol 2013;3(9):e754.
144. Kass EM, Helgadottir HR, Chen CC, Barbera M, Wang R, Westermark
UK, Ludwig T, Moynahan ME, Jasin M. Double-strand break repair by homologous recombination in primary mouse somatic cells requires BRCA1 but not the ATM kinase. Proceedings of the National Academy of Sciences of the
United States of America. 2013;110(14):5564-9.
145. Fleisig H, Wong J. Measuring cell cycle progression kinetics with metabolic labeling and flow cytometry. J Vis Exp. 2012(63):e4045.
146. Takeuchi H, Saito H, Noda T, Miyamoto T, Yoshinaga T, Terahara K, Ishii
H, Tsunetsugu-Yokota Y, Yamaoka S. Phosphorylation of the HIV-1 capsid by
MELK triggers uncoating to promote viral cDNA synthesis. PLoS Pathog.
2017;13(7):e1006441.
147. Krieg AJ, Hammond EM, Giaccia AJ. Functional analysis of p53 binding under differential stresses. Mol Cell Biol. 2006;26(19):7030-45.
170