Evaluation of Proteomic and N-Glycomic Alterations in Glioblastoma Multiforme

Abstract

Glioblastoma multiforme (GBM) is a tumor of glial origin. It is the most malignant, aggressive, and prevalent with the highest mortality rate in adult brain cancer. Surgical resection of tumor followed by Temozolomide (TMZ) therapy is currently available however, the development of resistance to TMZ is the limiting factor in effective treatment. GBM shows a high resistance to radiotherapy as well. Over the past few years, the mechanisms underlying radioresistance have been carefully investigated. However, it is still unclear how the interplay of biological components like proteins and various signaling pathways contributes to the establishment of the radioresistant phenotype in GBM. Proteomics approaches provide an unprecedented opportunity for the discovery of biomarkers and tumor markers for early detection and diagnosis by providing valuable information on the identity, expression levels, and modification of proteins. In the present work high throughput approaches based on mass spectrometric (MS) analysis i.e., nanoLC MS/MS was utilized to profile N-glycans isolated from radiosensitive (BT112) and radioresistant (NCH644) GBM cell lines. Differential proteome profiling was also performed for isocitrate dehydrogenase (IDH)-mutant and IDH-wild type GBM using Formalin-Fixed-paraffin-Embedded (FFPE) GBM tissue sections. The N-glycan profiling identified 183 and 185 N-glycans in BT112 and NCH644 cells, respectively. A noteworthy observation was the related N-glycan landscapes with the dominance of complex N-glycans in both the cell lines, which suggests that the N-glycan have been processed completely. About 41 N-glycans were differentially regulated in NCH644 compared to BT112 cells, consisting of 16 upregulated and 25 downregulated compositions. A greater percentage of complex highly branched, hyper-fucosylated and hyper-sialylated N-glycans was significantly upregulated, while a lower percentage of paucimannose, high-mannose, low-antennary complex, Abstract xxii hypo-fucosylated and hypo-sialylated N-glycans was significantly downregulated in NCH644 cells compared to BT112 cells. These outcomes suggest an initial point for the development of potential N-glycan targets to enhance the effectiveness of radiotherapy. These outcomes suggest an initial point for the development of potential N-glycan targets to enhance the effectiveness of radiotherapy. Moreover, isocitrate dehydrogenase (IDH) which is considered as the most significant genetic biomarker of GBM; its mutation is estimated in about 10% of all GBMs. Most notably, mutant GBMs are characterized by better survival than wild-type GBMs (31 months vs. 15 months). The present study explored the differential protein expression between the IDH-mutant and IDH-wild type GBM, via nano LC MS/MS, utilizing FFPE GBM tissue sections and identified 2251 proteins. Based on the statistical analysis and considering the log fold change, 66 differentially regulated proteins were identified between the two groups. About 28 and 38 proteins were upregulated and downregulated, respectively in IDH-mutant GBM compared to IDH-wild type GBM. Gene ontology (GO) enrichment analysis revealed molecular function, biological process, cellular content, and protein class for both the upregulated and downregulated proteins. Pathway enrichment analysis identified 12 upregulated and 16 downregulated proteins as significantly enriched in various molecular pathways while protein−protein interaction network analysis unveiled 21 upregulated proteins with two distinct interconnected networks and 24 downregulated proteins prominently enriched in two networks. Also, significant survival difference between IDH mutant GBM and IDH-wild type GBM was observed. Implementation of negligibly invasive tests as potential diagnostic technique and biofluid-based monitoring of GBM emphasizes on discovering biomarkers in CSF and blood. Therefore, a comprehensive in silico analysis was also performed to identify potential circulating biomarkers Abstract xxiii for GBM. A total of 2145 brain-specific proteins were screened through six gene and protein databases i.e., BioGPS database, C-It database, TiGER (Tissue-specific Gene Expression and Regulation) database, Human protein atlas (HPA), TiSGeD and CellFinder database. The screened proteins were filtered using a channel of five tools i.e., UniProtKB, SignalP 4.1 server, SecretomeP 2.0 server, ExoCarta database and TargetP 1.1 to predict the secretory proteins. The expression profile of the secreted proteins was verified and searched in literature for their relationship with GBM, keeping special emphasis on secretome proteome. The study identified nine secretory proteins i.e., Growth/differentiation factor 1 (GDF1), Slit homolog 1 protein (SLIT1), Neuronal pentraxin 1 (NPTX1), cellular repressor of E1A-stimulated genes 2 (CREG2), Serpin peptidase inhibitor-clade I (SERPINI), Leucine-rich glioma inactivated-1 (LGI1), Contactin-2 (CNTN2), Lymphocyte antigen 6H (LY6H) and opioid binding protein/cell adhesion molecule-like (OPCML). These predicted secretory proteins with brain-specific expression are associated with several cellular and molecular pathways that might be linked with GBM pathology. The study further examined the possible interactions of TMZ with these nine substantial secretory proteins through molecular docking and molecular dynamics (MD) simulation approach. Automated docking was carried out using Autodock 4.2 which indicated a significant binding affinity of TMZ with all the targeted proteins. The strongest interaction and binding affinity of TMZ was observed for GDF1 and SLIT1 followed by NPTX1, CREG2 and SERPINI. MD simulation of protein-ligand complexes was performed via CABS-flex V2.0 and iMOD server to assess Root Mean Square (RMSF) and to measure the stability of the proteins, respectively. The results demonstrated that the docked models were significantly flexible and stable with TMZ, proposing that it might be able to target putative proteins involved in gliomagenesis. Nevertheless, Abstract xxiv further in vitro and in vivo studies can discover the potential of the selected proteins to serve as novel targets for TMZ in the treatment of GBM. In conclusion, the present study encompassing high throughput proteomic approaches provide novel information on differential proteome and N-glycan signatures existing between IDH- mutant and wild type GBM and between BT112 and NCH644 cells, respectively, that may prove to be useful molecular indicators of diagnostic or prognostic value for GBM. In addition, it is proposed that the secretory proteins identified through in silico approach can be further investigated in vivo and in vitro to explore their potential role as circulating biomarkers that will further facilitate the development of minimally invasive diagnostic methods and novel therapeutic interventions for GBM providing insights for targeting heterogeneity and overcoming therapy resistance.