P11.17 Splicing dysregulation drives glioblastoma malignancy: SRSF3 as a potential therapeutic target to impair glioblastoma progression

Abstract

Abstract Glioblastomas (GBMs) remain the deadliest human brain tumors, with poor prognosis despite years of research. Currently, standard therapeutic strategies to treat GBM are not efficient and common survival from diagnosis is ~12–16 months. Thus, identification of new diagnostic/prognostic/therapeutic tools to tackle GBMs is crucial. Emerging evidence indicates that the cellular machinery controlling alternative splicing is altered in tumor pathologies, leading to oncogenic splicing events linked to tumor progression. Accordingly, we aimed to determine the expression pattern of the spliceosome components (SCs) and splicing factors (SFs) in high-grade astrocytomas (HGAs), mostly GBMs, and to ascertain the potential consequences of its dysregulation on GBM development. To this end, expression levels of SCs core and selected SFs were measured using a customized-microfluidic qPCR array in a well-characterized cohort of HGAs (n=33). Our results unveiled a profound alteration in the expression of multiple SCs and SFs in HGAs compared to healthy brain control-samples, wherein levels of particular elements (SRSF3/RBM22/PTBP1/RBM3) enabled perfect discrimination between non-pathological vs. tumor human-tissues, and between proneural and mesenchymal-like GBMs vs. control samples in mouse-models. Results were confirmed in an independent validation-cohort (n=49) and available Microarray dataset (Murat), which revealed that the expression of these splicing elements was correlated with relevant tumor markers and with survival. Remarkably, SRSF3/RBM22/PTBP1/RBM3 silencing (using specific siRNAs) decreased several aggressiveness parameters in vitro (e.g. proliferation, migration, tumorsphere formation, VEGFA secretion, etc.) and induced apoptosis, being SRSF3 the most relevant element affecting these parameters. Hence, a preclinical mouse model (U87MG-xenografts) with SRSF3 silencing drastically decreased in vivo tumor development/progression (i.e. tumor size, %MKI67, mitosis number, etc.) likely through a molecular/cellular mechanism involving the regulation of PDGFRB expression and its associated oncogenic signaling pathways. Overall, our results demonstrate that there is a profound dysregulation of the splicing machinery (spliceosome core and SFs) in HGAs/GBMs, which is directly associated to the development/progression of GBMs. Furthermore, this study reveals that SRSF3 can be a novel biomarker of malignancy and a potential therapeutic target to impair GBMs progression.