Abstract 2372: Kinome profiling of non-germline, genetically engineered mouse models of glioblastoma driven by Cdkn2a , Egfr , and/or Pten mutations reveals genotype-dependent kinase targets

• Published in: Cancer research (Chicago, Ill.) Vol. 78; no. 13_Supplement; p. 2372
• Main Authors: Smithberger, Erin, Flores, Alex R., Butler, Madison K., Dhruv, Harshil D., Johnson, Gary L., Berens, Michael E., Furnari, Frank B., Miller, C. Ryan
• Format: Journal Article
• Language: English
• Published: 01-07-2018
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2018-2372

The most attractive target for personalized medicine of glioblastoma (GBM) remains epidermal growth factor receptor (EGFR), due to its frequency in and specificity for the disease and the number of drugs that target its tyrosine kinase domain. Yet several EGFR tyrosine kinase inhibitors (TKI) have failed clinically, in part due to multiple molecular mechanisms of resistance. The most common oncogenic mutation in GBM is EGFRvIII, while the most common tumor suppressors lost are CDKN2A and PTEN. Mutations in these 3 genes frequently co-occur. To identify potentially attractive kinase targets for use in combination regiments with an EGFR TKI, we used multiplex inhibitor bead/mass spectrometry (MIB-MS) and RNA-seq to examine the baseline kinomes and transcriptomes of non-germline genetically engineered mouse (nGEM) models of GBM, specifically cultured Cdkn2a-null murine astrocytes (C) engineered to harbor human EGFRvIII (CEv3), Pten deletion (CP), or both (CEv3P). Among these 4 lines, 5.2-9.7% of the transcriptome was differentially expressed using DESeq2 at Q<0.001. Gene set analysis showed progressive loss of astrocyte and gain of stemness signatures in more heavily mutated cells (CEv3P>CEv3> CP>C), suggesting that EGFRvIII and Pten deletions cooperate to induce astrocyte de-differentiation into glioma stem cells (GSC). Principal components analysis showed a significant influence of both EGFRvIII (component 1, 44-48% variance) and Pten (component 2, ~33% variance) status on the transcriptome and kinome. Of the 228 expressed kinases detected using MIB-MS, 86 (38%) were differentially expressed in 1 or more genotypes. Integrated transcriptome and kinome analysis showed that Egfr was significantly over-expressed and hyperactive in EGFRvIII-mutated cells. Akt1 showed a non-expression driven increase in kinase activity in Pten-deleted cells, consistent with known effects of Pten on PI3K signaling. Pairwise genotype comparisons revealed 5-20 additional kinases that were differentially activated. Some, including Pdgfrb, Fgfr2, Lyn, Ddr1, and several members of the Ephrin family, represent potential targets for dual therapy with EGFR TKI. Clinically-curated human GBM patient-derived xenograft (PDX) models matched to CEv3P, CEv3, CP, and C nGEM models will afford comparisons in a patient-based preclinical setting for translational support. Functional kinome analysis using targeted EGFR TKI and MIB-MS in nGEM and PDX will help define the kinase networks required for EGFRvIII-driven GBM pathogenesis and may aid in the identification of novel treatment combinations.
Citation Format: Erin Smithberger, Alex R. Flores, Madison K. Butler, Harshil D. Dhruv, Gary L. Johnson, Michael E. Berens, Frank B. Furnari, C. Ryan Miller. Kinome profiling of non-germline, genetically engineered mouse models of glioblastoma driven by Cdkn2a, Egfr, and/or Pten mutations reveals genotype-dependent kinase targets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2372.