Vasculogenic Mimicry in Glioblastoma: Investigating the impact of natural and synthetic compounds and the role of tumor microenvironment

Glioblastoma (GBM) is the most common and aggressive malignant brain tumor in adults. Standard therapy, consisting of maximal surgical resection followed by radiotherapy and chemotherapy with agents such as temozolomide (TMZ) and cisplatin, provides only modest benefits, and recurrence is almost inevitable. A major limitation of current chemotherapeutics lies in their lack of selectivity, as they damage normal cells in addition to tumor cells, and their prolonged use is associated with toxicity and the development of chemoresistance. Furthermore, GBM progression is accompanied by neuronal loss (Rawal et al., 2025), which further contributes to the devastating impact of the disease. The remarkable resilience of GBM derives not only from intrinsic features—such as high migratory and invasive potential and cellular plasticity—but also from the supportive role of the tumor microenvironment (TME). In particular, astrocytes secrete factors that promote GBM migration, invasion, stemness, and therapy resistance. Moreover, GBM cells are capable of engaging in vasculogenic mimicry (VM), forming vessel-like networks independently of endothelial cells. VM sustains tumor perfusion, therapeutic resistance, and recurrence, thereby representing a crucial adaptive mechanism. Targeting both tumor-intrinsic adaptations and microenvironmental support thus emerges as an essential therapeutic challenge. Natural products are valuable sources of novel compounds with therapeutic potential in several human diseases, including cancer. In this context, previous studies from our laboratory investigated the activity of natural extracts from Ruta graveolens (RGWE), which were shown to selectively induce the death of glioblastoma cells and neural progenitors while sparing mature neurons. Interestingly, upon differentiation, A1 cells became resistant to rue’s toxic effects but not to those of TMZ and cisplatin, the two alkylating agents commonly used in glioblastoma therapy. These findings demonstrated that rue extracts induce glioma cell death by discriminating between proliferating/undifferentiated and non-proliferating/differentiated neuronal cells, representing a promising tool both for isolating novel therapeutic compounds and for identifying potential targets for intervention (Gentile et al., 2015). More recently, RGWE was also shown to exert neuroprotective effects in vivo, ameliorating ischemic damage and improving neurological deficits in a rat model of transient focal brain ischemia (Campanile et al., 2022). Building on these observations, our first aim was to assess whether RGWE could exert differential effects on primary glial cells. To this end, we analyzed its activity in primary neurons, oligodendrocytes, astrocytes, and microglia, comparing the results to those obtained in rat C6 glioma cells. Dose–response experiments revealed that RGWE reduced proliferation in glioma cells without detectable effects on healthy neurons or glial cells. By contrast, TMZ at comparable doses impaired the viability of both glioma and normal glial cells, underscoring its lack of selectivity. Having established the selective action of RGWE, we next sought to broaden our investigation beyond its effects. Since our main focus was to further explore the adaptive properties of GBM and strategies to counteract them, we extended our analysis to additional classes of compounds. In particular, we reasoned that applying natural (e.g., RGWE), synthetic (e.g., uPAcyclin and its derivatives), and clinically approved chemotherapeutics (e.g., TMZ and cisplatin) at sub-lethal concentrations could help minimize toxicity while unveiling novel therapeutic windows. This approach allowed us not only to confirm the discriminatory activity of RGWE, but also to comparatively assess other agents with distinct mechanisms of action. These studies revealed that RGWE significantly impaired GBM migration, invasion, and VM. Importantly, RGWE also proved effective against patient-derived cancer stem cells (CSCs), which are recognized as the main drivers of tumor aggressiveness and therapy resistance. Similarly, uPAcyclin and its derivatives consistently reduced GBM invasiveness by interfering with αV-integrin–dependent pathways. TMZ, at sub-lethal concentrations, attenuated migration, stemness, and VM, indicating a potential therapeutic window with lower cytotoxicity; however, its poor selectivity remained a concern. In contrast, sub-lethal doses of cisplatin paradoxically enhanced VM formation, raising caution regarding its use in this context. To better capture the complexity of GBM, we further employed advanced 3D co-culture models to evaluate the effects of our compounds in a more physiologically relevant setting. These experiments confirmed the pivotal role of astrocytes in driving GBM aggressiveness. Astrocyte-conditioned media and direct co-cultures enhanced migration, invasion, and spheroid expansion. Interestingly, TMZ partially counteracted these astrocyte-driven effects, highlighting its dual capacity to act not only on tumor cells but also on TME-derived support. In conclusion, this thesis identifies RGWE and uPAcyclin derivatives as promising anti-GBM candidates capable of impairing invasive and adaptive features while sparing healthy brain cells. It also emphasizes the dual, yet limited, action of TMZ—effective at reducing aggressiveness but lacking selectivity—and the potential risks of cisplatin in promoting pro-tumor adaptations. By integrating both tumor plasticity and astrocyte-mediated support, this work provides a rationale for the development of therapeutic strategies that are at once more effective and less toxic in the treatment of glioblastoma.