LAUSR

Somatic oncogene amplification happens frequently in glioblastoma (GBM). The second most frequently amplified gene encoding receptor tyrosine kinases in GBMs is platelet derived growth factor alpha (PDGFRA) (15%). In contrast, MYC and MYCN amplification occurs in 1.6% and 2.9%, respectively. Our goal was to characterize the role of PDGFRɑ and Myc in GBM. Neurosphere cultures were implanted in cohorts of 10–15 nude mice. 5 PDX lines, presenting median survival of 29–59 days were classified as short survivors, and 5 lines with median survival between 104–134 days classified as long survivors. Total RNA was extracted from PDX terminal tumors (3 biological replicates) and sequenced in a paired-end read format. Mouse reads were filtered out using Xenome. MYC and PDGFRA expression patterns were analyzed in tissue microarrays representing duplicated samples from 40 glioma neurosphere-derived PDX lines by IHC (1 anaplastic oligodendroglioma, 8 recurrent GBM with 2 newly diagnosed/recurrent pairs). Normalized staining intensity (MI) and area (A) were quantified using Fiji/ImageJ. PDGFRA, MYC, MYCN gene amplifications were represented in a molecularly diverse panel of GBM patient-derived cancer stem-like cells (CSC) and orthotopic mouse xenografts (PDX). Transforming to a normal distribution (log10), 4/13 of cell lines had a PDGFRA mRNA expression (RPKM) higher than 1.5. Similarly, one PDX line had a staining index of greater than 10, 11 (27.5%) had an index between 5–10. The range of intra-tumoral variance, represented by standard deviation, was 0.09–24.25 highlighting the heterogeneity of PDGFRɑ expression. PDGFRɑ phosphorylation (Y754) did not differ between 8 cell lines cultured in NMGF, but deviated in alternate medias without growth factors, supplemented with FBS. In comparison, MYC(N) mRNA expression is only elevated in the context of a known amplification. Furthermore, a a MYC activity signature consisting of 18 target genes was only evident in the 5 amplified CSC lines. Taking advantage of genomic heterogeneity, we have isolated subclones lacking PDGFRA amplification from a PDGFRA amplified GBM CSC. The absence of PDGFRA amplification reduced the self-renewal potential to 37% of the PDGFRA amplified cell population (p=0.001) in clone 1 and 57% in clone 2 (p=0.013). Pertaining to determinants of in vivo survival, MYC was altered in 80% of short survivors (2/5 MYC, 2/5 MYCN amplification) and in 0% of long survivors. Myc signature was highly correlated with in vivo survival (Pearsons’ corr. = -0.77) and MYC gene expression was correlated with in vivo TMZ resistance (corr. = 0.7). These results suggest that PDGFRɑ expression and activity can occur in the absence of gene amplification, while Myc activity is dependent on gene amplification. Both oncogenes drive oncogenic pathways that should be explored as therapeutic targets.