Molecular profiling of Primary and Recurrent GBM reveal novel therapeutic opportunities
Presented at BNOS June 2016
Combiz Khozoie, Richard Perryman, Fernando Abaitua, Babar Vaqas, Kevin O’Neill, Nelofer Syed
Glioblastoma multiforme (GBM) is the most common and aggressive primary brain tumour. High patient mortality arising from a high propensity for tumour recurrence has driven a need to characterize the genetic changes associated with recurrent GBM. Here, we present a comprehensive landscape of transcriptome profiles of 10 primary and recurrent GBM cancers using RNA-seq, revealing extensive recurrent-specific gene expression changes and somatic mutations. Perturbation of the extracellular matrix, and enhanced ECM degradation, emerge as important changes accompanying GBM recurrence; this perturbation is multimodal and implicates diverse functions including enhanced matrix metalloproteinase (MMP) mediated ECM degradation, silencing of collagen encoding genes (e.g. COL1A1, COL1A2, COL3A1, COL6A3) or their mutation (e.g. COL7A1), and enhanced ECM degradation via dysregulation of SAPK/JNK signaling. Mutational signature analysis revealed striking enrichment of C to T transitions, accounting for 40 to 50% of all base-substitution mutations, implicating CT transition enhancing mutagenic processes in the etiology of GBM. Aberrant epigenetic regulation consistent with dysfunctional polycomb-mediated gene repression manifests in inappropriate activation of putative tumour promoter genes in recurrent tumours. Of 188 genes identified as differentially expressed in our GBM cohort, 38 genes were shared in a separate validation cohort of tumours from the TCGA GBM database. This includes seven genes (ADM, HPCAL4, SNCB, SV2B, TMEM130, H19, VSNL1) with correlations between gene expression and patient survival according to Kaplan-Meier plots generated from TCGA data. This study has highlighted new insights into GBM tumour recurrence and revealed potential novel candidates as prognostic biomarkers and rational therapeutic targets.