Abstract
Recurrence of therapy resistant Glioblastoma multiforme (GBM) is responsible for patient mortality. Not all patients qualify for surgical resection or for chemotherapy, but radiation is almost a universally tolerated therapy. We are modeling acquired radiation resistance using 8 radiation-sensitive GBM patient-derived xenolines (PDX) made resistant by six irradiation series (6x2Gy each) in vivo. In 4 resistance-induced PDX, MGMT (O6-methylguanine–DNA methyltransferase) protein expression increased over original isogenic PDX. Paradoxically, temozolomide screening of orthotopic radiation-resistant PDX with increased MGMT expression revealed increased, not decreased chemo-sensitivity. This suggests an unanticipated mechanism associating acquired radiation resistance and alkylating chemotherapy sensitivity. RNA-seq data of long non-coding RNAs (lncRNAs) has revealed associations with patient overall survival and age at diagnosis. Out of 24,076 lncRNAs, 5 are significantly differentially expressed with regard to patient overall survival and age at diagnosis. The functions of lnc-ZNF117-1, lnc-DCUN1D4-1, and LINC01397 are unknown, but they are enriched in brain over other tissues. Tissue specific expression and function are unknown for lnc-TBL1XR1-5, but it is highly expressed in HepG2 (hepatocellular carcinoma) as well as in non-functioning pituitary adenomas (NFPAs). The lnc-CDH17-1 transcript is also highly expressed in NFPAs. All 5 of these lncRNAs have complex secondary and tertiary structures, but their physiologic or pathologic functions are unknown. LncRNAs most likely contribute to acquired resistance through epigenetic regulation of transcription and chromatin state. Deep sequencing of total RNA isolated from intracranial radiation–sensitive/–resistant PDX to relate transcriptional programs and therapy resistance is underway. These data will be paired with kinomic profiling of matched tumors to elucidate basal signaling modality changes between radiation–sensitive/–resistant tumors. Analysis of differentially expressed lncRNA will be used to predict lncRNA structure/function, elucidating druggable mechanisms mediated by lncRNAs.