333 Obtaining the Genetic Fingerprint of Resistance to Glioblastoma Through a Novel Multigenerational Xenograft Model. (August 2016)

• Record Type: Journal Article
• Authors: Jahangiri, Arman; Chen, William; Yagnik, Garima; De Lay, Michael; Wagner, Jeffrey; Sidorov, Maxim; Flanigan, Patrick Michael; Aghi, Manish Kumar
• Is Part Of: Clinical neurosurgery. Volume 63(2016)Supplement 1
• Publication Date: 2016-08
• Subjects: Nervous system -- Surgery -- Congresses; Neurosurgery; Nervous system -- Surgery; Neurologie; Congresses; Conference papers and proceedings; 617.48
• DOI: 10.1227/01.neu.0000489822.13595.ca
• Languages: English
• ISSNs: 0069-4827

Abstract

Abstract : INTRODUCTION: Despite positive preclinical/clinical trials, a major hurdle in the clinical application of bevacizumab, an antiangiogenic therapy, in glioblastoma is the development of resistance and progression following a transient response period. METHODS: We established a multigenerational glioblastoma xenograft model of acquired bevacizumab resistance through subcutaneous implantation of U87 cells, bevacizumab treatment, and selection and reimplantation of the fastest growing tumor in each generation to new mice. Whole human genome microarray (Illumina) was performed on 3 tumor samples from generations 1, 4, and 9, and bioinformatic analysis of gene expression data was performed in Matlab2014a. RESULTS: Using published statistical methods, we identified a set of genes exhibiting significant intergenerational variance. Protein-protein interaction (PPI) scores were extracted of String database (v10); subsequent spectral clustering revealed 13 gene subnetworks of closely interrelated genes. Gene set overrepresentation (GSO) analysis via ConsensusPathDB suggested biologically meaningful subnetworks mediating distinct functions, including inflammation, extracellular matrix remodeling, cell cycle, metabolism, and cytoskeletal dynamics. Gene set enrichment analysis revealed significant overexpression across generations of previously identified gene expression signatures of the mesenchymal subtype. Important markers, including putative tumor-stemness marker CD44 and critical epithelial-mesenchymal-transition transcription factors SNAI2 and ZEB2 were upregulated across generations. These results suggest tumor progression under bevacizumab to be accompanied by a gene expression shift toward the mesenchymal subtype, associated with enhanced invasiveness, resistance, and worse outcomes. Our analysis revealed expression changes in angiogenesis-related pathways. Genes identified via GSO suggested a tumor proangiogenic response to bevacizumab, composed of converging pathways involving inflammation, hypoxia, extracellular matrix remodeling, upregulation of alternative proangiogenic pathways, and downregulation of antiangiogenic factors. CONCLUSION: Using microarray analysis of a model of bevacizumab resistance in glioblastoma, we found development of resistance to be accompanied by a gene expression shift toward the mesenchymal subtype, as well as activation of alternative proangiogenic pathways. These findings shed light on the mechanisms of resistance to antiangiogenic therapy in glioblastoma.