MEOX2 as a transcription factor involved in glioblastoma stem cell key properties

Glioblastoma (GBM) is the most aggressive and malignant brain cancer occurring in adults, according to the World Health Organization (WHO) classification. It is characterized by genomic instability, hyperproliferation, resistance to apoptosis, necrosis, and angiogenesis (Zheng et al., 2008). Current therapy of maximal surgical resection followed by radiotherapy with concurrent chemotherapy offers only palliation, with median survival of less than two years (Stupp et al., 2009). Glioblastoma stem cells (GSCs) are considered the main tumor initiating cells; indeed, they represent a subpopulation of GBM cells with a self-renewal ability and an active proliferation, and they are believed to be at the origin of drug resistance and disease recurrence in GBM. For this reason, it is essential to better identify new genes marking out this disease and in particular its initiating cells (Bao et al., 2006). The starting point of my PhD project was an RNA-sequencing analysis (RNA-seq) performed on six patient-derived glioblastoma stem cells, on one human glioblastoma cell line (T98G), on cultured healthy astrocytes, and on healthy brain tissue. This transcriptome analysis showed that GSCs are specifically characterized by overexpression of the transcription factor Mesenchyme Homeobox 2 (MEOX2), whose coding gene is located at 7p21.2 locus. Interestingly, gain of chromosome 7, and in particular of its short arm, is part of the molecular signature of glioblastoma. MEOX2 belongs to the homeobox gene family, and, in a physiological context, it has been established as a cell cycle inhibitor in vascular smooth muscle and vascular endothelial cells, via activation of the cyclin dependent kinase inhibitors p21 and p16 (Douville et al., 2011). In contrast, in a tumoral context, both pro- and antitumoral roles have been described for MEOX2 (Ohshima et al., 2009; Zhou et al., 2012; ÁvilaMoreno et al., 2014; Tian et al., 2018). Intriguingly, in the specific context of glioblastoma, MEOX2 has been regarded as a member of a 4-transcription factors scoring system for prediction of GBM prognosis (Cheng et al., 2019), and, very recently, it has been shown to contribute to GSCs’ viability and proliferation (Tachon et al. 2021). On this basis, the first step of my work was to investigate the possible pro-tumoral effect of MEOX2 ectopic expression in non-stem cells of human glioblastoma, by analyzing the main hallmarks of tumor cells such as proliferation, migration, and anchorage-independent growth ability. This has been performed taking advantage of two glioblastoma cell lines, namely 2 LN229 and T98G cells, stably transfected with a MEOX2-expressing construct. Interestingly, MTS assay and viable cell count did not show an increase in proliferation of cells ectopically expressing MEOX2, compared to the control ones, transfected with an empty vector. In order to investigate if MEOX2 could affect the migration ability of these cell lines, we performed a transwell migration assay that highlighted a trend toward a lower migratory ability, especially in LN229 cells stably expressing MEOX2. On the contrary, when we evaluated by a soft agar assay, the anchorage-independent growth capacity of MEOX2 expressing cell lines, neither the LN229 cells nor the T98G ones showed any change in their capability to grow independently of a solid surface in vitro. Meanwhile, we analyzed by RNA-seq the cell lines stably overexpressing MEOX2 and we found a single mRNA consistently modulated in both cellular contexts. This mRNA, encoding for Periostin (POSTN), was clearly induced upon MEOX2 transfection in LN229 and T98G cells. Periostin is a secreted extracellular protein that is involved in tissue development and regeneration, playing an important role in extracellular matrix (ECM) structure and organization. Interestingly, in glioblastoma, it has been described that GSCs secrete Periostin to recruit tumor-associated macrophages (TAMs) belonging to the M2 tumor supportive subtype (Zhou et al., 2015). The second part of my PhD work was devoted to the depletion of MEOX2 in GSCs, where in fact we had firstly revealed its overexpression. To this end, we took advantage of two different shRNAs, directed to either 3’UTR region (shRNA53) or the coding sequence of MEOX2 mRNA (shRNA18). The shRNAs were harbored in lentiviral vectors by which we have transduced two patient-derived GSCs, BT273 and BT379 cells, thus obtaining cells stably depleted of MEOX2 expression. We subsequently evaluated whether MEOX2 depletion could affect the key aspect of GSCs, namely their self-renewal ability, and found that it was deeply reduced, as both BT273 and BT379 cells transduced with both shRNAs showed a great decrease in sphere forming ability compared to the same cells transduced with a control construct. A proliferation assay performed on Geltrex coated plates highlighted that also the viability of both GSC lines was negatively affected by MEOX2 knock down. Driven by curiosity to understand whether the reduced growth capacity, seen particularly in BT379 GSCs, was linked to an increased level of apoptosis, we evaluated by Western blot analysis the level of Caspase-3 cleavage in BT273 and BT379 cells transduced with shRNA18 or shRNA53. Both shRNAs induced a Caspase-3 processing, even if to different extent. In BT379 cells activation of Caspase-3 was much 3 stronger than in BT273 ones. In agreement with this result, a Propidium Iodide (PI) or 7-AAD (7-Aminoactinomycin D)-Annexin V staining flow cytometry analysis confirm a higher level of apoptosis in BT379 cells compared to BT273 ones, in which we found instead a higher level of necrosis. We then searched the transcriptome of the MEOX2 depleted GSCs to investigate the basis of the observed phenotypes. We found a large number of genes modulated by MEOX2 knock down, among which a “core set” of genes showed an overlapping differential expression, consequent to MEOX2 repression, in both BT273 and BT379 GSC lines. These genes, all down regulated by MEOX2 depletion, are involved in the response to hypoxic environment and in the glycolytic pathway. This very interesting finding suggest that the transcription factor MEOX2 could be involved in the regulation of the GSCs ability to respond to unfavorable conditions, and in their metabolic shift to glycolysis as the main energetic pathway. Overall, the data obtained during my PhD work suggest an important function for the MEOX2 transcription factor in glioblastoma stem cell biology, and its possible role as a therapeutic target for this stem cell population, which contributes to make this brain tumor still incurable.