Towards Emx2 therapy of Glioblastoma multiforme

The homeodomain-containing transcription factor Emx2 encodes for a homeobox protein essential for territorial specification of rostral CNS as well as for proper spatio-temporal tuning of neural cell growth and differentiation (Gangemi et al. 2006). Previous experiments done in my laboratory demonstrated its role in inhibiting cortico-cerebral astrogenesis, limiting proliferation of astrocytes-committed progenitors upon its overexpression in neural stem cells. This control takes place via a functional cascade, which includes stimulation of Bmp signaling and Sox2 repression, through the downregulation of Egfr and Fgf9 (Falcone et al., 2015). Meanwhile, studies by other labs also reported an inverse correlation between Emx2 expression levels and aggressiveness of several human cancers, including lung, endometrial and gastric tumours (Okamoto et al. 2010) (Li et al. 2012). Inspired by these findings, we activated a research program aimed at exploring the possibility to use Emx2 in therapy of glioblastoma multiforme. This neoplasm is the most common and aggressive malignant primary tumor of the CNS, responsible of 4% of all tumor death in humans. Conventional therapeutic options for it are unfortunately limited: after surgical resection, GBM-affected patients routinely undergo radiotherapy and adjuvant chemotherapy (temozolomide, TMZ); nevertheless, their median survival is no longer than 14 months. In 2016, Carmen Falcone demonstrated that Emx2 overexpression suppresses a number of different glioblastomas in vitro, within 7-10 days, by inducing cell death and inhibiting cell proliferation. Molecular mechanisms underlying this phenomenon resulted to be highly pleiotropic, indeed Emx2 affects different pathways and genes, including RTK cascades, cell cycle control circuitries and other malignancy-related processes (Falcone et al. 2016). Then, given limits of conventional therapies, we decided to score the actual benefit of experimental Emx2 gene therapy, evaluate its possible interaction with standard chemo- and radiotherapy, and explore novel routes for its delivery. These issues were investigated under the framework of this thesis. Results were as follows. First, Emx2 displayed a therapeutically appealing, anti-oncogenic activity in vivo. Indeed the median survival time of mice transplanted with Emx2-GOF tumour cells was twenty days longer compared to the control group (i.e. 35 days), outperforming TMZ. Next, these results were confirmed by in vitro kinetic assays. Here we observed at least an additive effect between TMZ and Emx2-GOF treatments; specifically, in case of the U87 line, Emx2 sensitized GBM cells to chemotherapy. As for X-rays, we found that their association to Emx2 overexpression resulted in an enhanced anti-oncogenic effect. By means of ad hoc calibrated, in vitro kinetic assays, we discovered that Emx2 sensitized GBM cells to radiation. This was predominantly due to an inhibition of homologous recombination-based DNA-repair, likely leading to GBM cell suicide. A pronounced downregulation of SOX2 and FOXG1, two key drivers of GBM malignity, was instrumental to that. Finally, interested in developing new effective and biosafe strategies for in vivo delivery of the therapeutic Emx2 transgene, we collected preliminary proofs of principle, supporting the feasibility of a novel design based on combined use of HSV-1-derived, oncolytic viruses and amplicon vectors.