CSIG-10. PLATELET-DERIVED GROWTH FACTOR RECEPTOR ALPHA ONCOGENE DEPENDENCY IN GLIOBLASTOMA. (12th November 2021)

• Record Type: Journal Article
• Authors: Berezovsky, Artem; Datta, Indrani; She, Ruicong; Transou, Andrea; Irtenkauf, Susan; Hasselbach, Laura; Poisson, Laila; deCarvalho, Ana
• Is Part Of: Neuro-oncology. Volume 23: Supplement 6(2021)
• Page Start: vi35
• Page End: vi35
• Publication Date: 2021-11-12
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noab196.136
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: PDGFRA is the second most frequently amplified gene encoding receptor tyrosine kinase in adult glioblastoma (GBM), oftentimes as extrachromosomal elements (ecDNA). Our overall objective is to elucidate mechanisms underlying PDGFRα dependency in GBM tumor maintenance. We have isolated distinct subpopulations from a GBM model (HF3253), harboring two alterations in PDGFRA: constitutively active genomic rearrangement and extrachromosomal amplification, that differ in the frequency of PDGFRA ecDNA. HF3253 tumor growth rate correlates with the initial proportion of ecDNA+ population implanted. Furthermore, slower tumor growth is due to selection for initially low-frequency PDGFRA ecDNA amplified clones based on histology and TaqMan Copy Number assay. Further exploiting intra-tumoral heterogeneity, we have isolated single cell clones from bulk cells. Compared to bulk cells, single cell clones do not express PDGFRα, PDGFRA mRNA and exhibit diploid PDGFRA copy number. Tumor growth was reduced in 4 ecDNA(-) clones compared to parental ecDNA(+) (log-rank test p= 0.00772, 0.00379, 0.00076, 0.00379). In contrast to parental HF3253, ecDNA(-) tumors demonstrated diffuse tumor morphology and weak PDGFRα activation. HF3253 ecDNA(-) PDX tumors lack detectable PDGFRα. Correspondingly, HF3253 ecDNA(-) cell populations do not exhibit de novo PDGFRA copy number gains post-implant. We conducted paired, whole RNA-sequencing on 20 HF3253 populations (ecDNA+/-: 6 clones from 3 biological replicates PDXs and 4 clones from 4 in vitro technical replicates). Employing a false discovery rate of 0.05, we identified 785 differentially expressed genes. Platelet-derived growth factor binding (GO:0048407) and central carbon metabolism were down-regulated in ecDNA(-) while genes significantly associated with astrocytic processes were upregulated. We demonstrated the dependency on PDGFRα signaling in a patient-derived GBM model carrying ecDNA PDGFRA amplification. Our data validates PDGFRɑ as a therapeutic target in a subset of GBM patients and demonstrates that detection of ecDNA-amplified PDGFRA has the potential to be a predictive biomarker of future PDGFRɑ targeted therapies.