An integrative analysis of Cell Adhesion Molecules (CAMs) in Glioma: The Essential role of PTGFRN and ASTN1 in Glioma and Glioma Stem-like Cells
The central nervous system tissue contains neural stem cells which differentiate and give rise to the neurons and glial cells. The glial cells maintain neurons by providing nutrients, physical support and protection. Unlike neurons, glial cells retain their capacity to divide and proliferate throughout the life span of an individual. During proliferation of glial cells, dysregulation in cellular processes that control growth leads to formation of glial neoplasms or gliomas. The glial tumors are classified based on the predominant cell type; astrocytomas that originate mainly from astrocytes, oligodendrogliomas from oligodendrocytes and oligoastrocytomas both astrocytes and oligodendrocytes. Among all gliomas, astrocytomas are the most common central nervous system neoplasms which makes up to 60% of all the primary brain tumors. Being the most prevalent type, the world health organizations (WHO) classifies them into grades ranging from I to IV based on their intensity of malignancy. Grade-IV astrocytoma or Glioblastoma (GBM) is the most malignant form with a median survival of less than 15 months, despite all therapeutic modalities. CAMs are involved in cell-cell and cell-extra cellular matrix (ECM) communications and physical attachment between cells and ECM. CAMs are also involved in ‘out-in and in-out’ signalling which helps in cell survival, migration and stemness in both normal and cancer cells. Cell adhesion system works in various mechanisms by which it translates basic genetic information into complex three-dimensional pattern of cells in tissues. Dysregulation of CAMs is involved in different neoplastic transformation, survival, self-renewal, proliferation, migration, EMT, and metastasis. The aim of this study was to investigate the role of CAMs in glioma development and progression. Part 1: Elucidation of Genetic and Epigenetic landscape alterations in Cell Adhesion Molecules (CAMs) in Glioblastoma CAMs are important group of molecules expressed on the surface of the cell and often found dysregulated in multiple cancers including GBM. Though several reports emphasize the importance of a number of individual CAMs in glioma initiation and progression, there is no single report which shows the comprehensive view of their importance in glioma development and progression. In this study, we have curated all known CAMs from the literature and analysed using various bioinformatics tools to show their genetic alterations, probable mechanisms for regulation, association with tumor grade/nature, and survival correlation in GBM. A comprehensive bioinformatics analysis of CAMs (n=518) in multiple available data sets revealed genetic and epigenetic alterations among CAMs in GBM. We identified several genetically altered CAMs (n=293, 56%) in GBM. The mutations in the LRFN5 and PCDHGC5 predicted poor prognosis. The differential regulation of CAMs (n=181, 35%) contributed by copy number variation (CNV), DNA methylation and miRNA targeting. We identified two sets of differentially regulated CAMs that may be implicated in initial astrocytic transformation and glioma progression. Further, we have also identified a unique set of differentially regulated CAMs (n=22) specific to glioma stem-like cells (GSC) as compared to both neural stem cells (NSC) and differentiated glioma cells (DGC). Part 2: The essential role of Prostaglandin F2 receptor inhibitor (PTGFRN) in glioblastoma PTGFRN is an IgSF CAM, surface expressed single pass transmembrane protein and serves as a scaffolding protein. It is widely found to be associated with multiple tetraspanin members such as CD9, CD81, CD82 and CD151. It was reported as a highly expressed antigen in metastatic cancer cells and induces cell migration in multiple cancer types. In our study, integrative analysis of CAMs revealed that PTGFRN is one of the upregulated genes in GSCs compared to both DGCs and NSCs. In this section, we identified PTGFRN to be highly expressed in GSCs vs DGCs, GBM samples compared to controls as analysed in publicly available multiple datasets such as TCGA (Agilent and RNA-seq), REMBRANDT, GSE22866 and GSE7696. Further, when we explored the expression of PTGFRN in GBM cell lines, transcript and protein levels were found to be more in GBM cell lines as compared to immortalized astrocyte derived control cell lines. The survival analysis revealed that PTGFRN high expression is correlated with poor patient survival in GBM. The functional relevance of PTGFRN in GBM development and progression was investigated by shRNA mediated silencing followed by functional assays. In multiple GBM cell lines, silencing of PTGFRN reduced cell proliferation, colony formation and anchorage-independent growth in soft agar as compared to silencing of non-targeting (shNT) control. Silencing of PTGFRN in GSCs reduced number of neurospheres formed as compared to shNT control as assessed by neurosphere formation and limiting dilution assays. We also found that the knockdown of PTGFRN reduces migration and invasion in GBM cell lines. Cell cycle analysis showed G1 or G2/M phase arrest after silencing PTGFRN in GBM cell lines. Further, we observed that in GBM cell lines silencing of PTGFRN leads to apoptotic cell death by 40 to 50% as compared to shNT control. In order to understand the functional role of PTGFRN in GBM, we first identified the significantly differentially regulated genes between PTGFRN-high and PTGFRN-low expressed groups of GBM samples in TCGA Agilent dataset. The gene sets were further used for KEGG pathway, Gene Ontology (GO) and GSEA analyses. We found that the upregulated genes showed significant enrichment for focal adhesion and CAMs pathways, whereas negatively regulated genes showed significant enrichment for drug and xenobiotic metabolism by cytochrome p450 pathways in KEGG pathway database. The GO and GSEA analyses showed significant enrichment for hallmarks: Epithelial-mesenchymal transition (EMT), IL6 JAK STAT3 signalling, TNF alpha signalling via NF-kB and Notch signalling and significantly depleted for the hallmark: KRAS signalling DN (genes down-regulated by KRAS activation). To decipher the molecular mechanism for PTGFRN function in GBM, we checked the status of the most dysregulated oncogenic pathways after its knockdown in GBM cell lines. Silencing of PTGFN reduced phospho-AKT, phospho-ERK, phospho-4EBP and phospho-P70S6 levels compared to control condition in GBM cell lines. In order to test the role of PTGFRN in EMT and Focal adhesion processes, we silenced PTGFRN in GBM cell lines and observed a reduction in vimentin and Focal Adhesion Kinase (FAK) protein levels compared to control. It indicates that PTGFRN might be playing a potential role in pro-survival and promigratory pathways in GBM. Furthermore, we also evaluated various mechanisms by which PTGFRN expression is regulated in GBM. Bioinformatics analysis revealed that DNA hypomethylation and down regulation of microRNAs such as miR-107, miR-133b and miR-137 correlated with PTGFRN high expression in GBM. In order to test the role of methylation, we treated LN18 cell line (PTGFRN-low) with 5-Aza 2’-cytidine and observed that no rescue of PTGFRN expression which indicates that methylation may not play a role in its regulation. However, overexpression of miR-137 in the PTGFRN-high glioma cell lines (U373, U251 and U87) reduced PTGFRN protein levels indicating microRNA-dependent regulation of PTGFRN. Moreover, a recent report suggests that TGF-β signalling pathway may also regulate PTGFRN in GBM. Inhibition of TGF-β signalling with ALK5 inhibitor also reduced PTGFRN protein levels relative to the control treatment. The high activation of TGF-β signalling and down regulation of miR-137 expression might be responsible for high expression of PTGFRN in GBM. Part 3: Astrotactin 1 (ASTN1), a neural CAM is essential for survival and growth of Glioma Stem-like Cells Astrotactins (ASTN1 and ASTN2) are neural CAMs which are known to be involved in the development of CNS. ASTN1 was also shown to be upregulated in premalignant stem-like or progenitor cells of glioblastoma compared to astrocytes and hESCs. In our study, integrated gene expression analysis revealed ASTN1 as a highly upregulated CAM in GSCs compared to both NSCs and DGCs. We analysed ASTN1 transcript and protein levels in various GSC cell lines and corresponding differentiated glioma cells. We found ASTN1 to be upregulated in GSCs compared to corresponding DGCs at both transcript and protein levels. Silencing of ASTN1 in GSCs formed reduced number of neurospheres in neurosphere and limiting dilution assays as compared to shNT control. We also found that the cell viability was significantly reduced after silencing of ASTN1 in GSCs as assessed by trypan blue exclusion assay. In addition to knockdown, blocking of ASTN1 with antibody also inhibited the neurosphere formation in GSCs. We further investigated the mechanisms by which ASTN1 exhibits its role in GSC survival and growth. Silencing of ASTN1 in GSCs increased the number of annexin-V positive cells which indicates increased apoptosis. To further validate ASTN1 role in GSC stemness maintenance, we checked the essential four-reprogramming factors; Sox2, Po3f2, Olig2, and Sall2 which are necessary and sufficient for maintenance of stemness in GSCs. In GSCs, silencing of ASTN1 reduced transcript levels of all four reprogramming factors as compared to vector control. In conclusion, we have elucidated the altered landscape of CAMs in GBM and given an insight into impact of such alterations on functional and molecular role taking PTGFRN and ASTN1 as examples. In first part we provided a panoramic view of the various alterations in CAMs encountered in GBM. In second part, we identified that PTGFRN is required for GBM cell growth, migration, and invasion and we also found its regulation. In the third part, we demonstrated that ASTN1 is required for GSC survival and growth.