The PI3K/Akt/mTOR signalling network is activated in almost 90% of all glioblastoma, the most common primary brain tumour, which is almost invariably lethal within 15 months of diagnosis. Despite intensive research, modulation of this signalling cascade has so far yielded little therapeutic benefit, suggesting that the role of the PI3K network as a pro-survival factor in glioblastoma and therefore a potential target in combination therapy should be re-evaluated. Therefore, we used two distinct pharmacological inhibitors that block signalling at different points of the cascade, namely, GDC-0941 (Pictilisib), a direct inhibitor of the near apical PI3K, and Rapamycin which blocks the side arm of the network that is regulated by mTOR complex 1. While both substances, at concentrations where they inhibit their primary target, have similar effects on proliferation and sensitisation for temozolomide-induced apoptosis, GDC-0941 appears to have a stronger effect on cellular motility than Rapamycin. In vivo GDC-0941 effectively retards growth of orthotopic transplanted human tumours in murine brains and significantly prolongs mouse survival. However, when looking at genetically identical cell populations that are in alternative states of differentiation, i.e. stem cell-like cells and their differentiated progeny, a more complex picture regarding the PI3K/Akt/mTOR pathway emerges. The pathway is differently regulated in the alternative cell populations and, while it contributes to the increased chemo-resistance of stem cell-like cells compared to differentiated cells, it only contributes to the motility of the latter. Our findings are the first to suggest that within a glioblastoma tumour the PI3K network can have distinct, cell-specific functions. These have to be carefully considered when incorporating inhibition of PI3K-mediated signals into complex combination therapies.

Temozolomide (TMZ) currently remains the only chemotherapeutic component in the approved treatment scheme for Glioblastoma (GB), the most common primary brain tumour with a dismal patient’s survival prognosis of only ~15 months. While frequently described as an alkylating agent that causes DNA damage and thus—ultimately—cell death, a recent debate has been initiated to re-evaluate the therapeutic role of TMZ in GB. Here, we discuss the experimental use of TMZ and highlight how it differs from its clinical role. Four areas could be identified in which the experimental data is particularly limited in its translational potential: 1. transferring clinical dosing and scheduling to an experimental system and vice versa; 2. the different use of (non-inert) solvent in clinic and laboratory; 3. the limitations of established GB cell lines which only poorly mimic GB tumours; and 4. the limitations of animal models lacking an immune response. Discussing these limitations in a broader biomedical context, we offer suggestions as to how to improve transferability of data. Finally, we highlight an underexplored function of TMZ in modulating the immune system, as an example of where the aforementioned limitations impede the progression of our knowledge.

Glioblastoma is the most common brain tumor in adults and among the deadliest malignancies per se with a highly invasive phenotype upon presentation. To achieve rapid colonialization throughout the central nervous system, glioblastoma cells have to be equipped with a high resistance to several forms of programmed cell death, such as apoptosis. All this occurs in the absence of any tumor-initiated signature mutations. Using a comprehensive comparative analysis combining expression profiles and functional analysis of normal brain glia cells, primary tumors and tumor-derived organoids with distinct differentiation subtypes we investigated the underlying features associated with the high resistance to cell death-induced by conventional treatment. To break this resistance, we looked into the possibility to add the small molecule inhibitor Venetoclax, that targets the Bcl-2 family, to conventional therapy. The Bcl-2 family are a number of evolutionarily-conserved proteins that share Bcl-2 homology (BH) domains and are most notable for their regulation of apoptosis at the mitochondrion. Interestingly, although stem cell-like cells (SCs) express more pro-apoptotic Bcl-2 family proteins than their differentiated progeny they remain more resistant to apoptosis, suggesting that the Bcl-2 family is not the main mediator of apoptosis resistance. In stark contrast to leukemia cells, inhibition of Bcl-2 alone has no effect on glioblastoma cells, but combining this with either Temozolomide (TMZ), the standard chemotherapeutic option, or serum starvation leads to synergistic effects. These are, however, weak hence we investigated whether compensatory mechanisms are activated. Indeed, in both SCs and DCs we found an upregulation of Mcl-1, a molecule known to compensate for inhibited Bcl-2. However, additionally blocking Mcl-1 with several different small molecule inhibitors did not further sensitize primary Glioblastoma cells. This suggests that while mechanical upregulation of Mcl-1 occurs in Glioblastoma, it is of little functional consequence. The reduced importance of the Bcl-2 family is also reflected in the non-transformed astrocytes which are precursor of glioblastoma. They already display an intrinsic high resistance to apoptotic cell death. Treatment with TMZ and modulators of apoptosis does not significantly affect their viability. Our findings suggest that the Bcl-2 family has a reduced role in mediating the survival of brain cells, at least when compared to haemopoietic cells. While a certain sensitizing effect can be achieved in Glioblastoma cells by combining cellular stressors with inhibitors of the Bcl-2 family, this is unlikely to be a sufficient to overcome therapy resistance. Citation Format: Helene von Bandemer, Felix Seyfried, Anne Fleischmann, Georg Karpel-Massler, Aurelia Peraud, Hannah Strobel, Marcin Lyszkiewicz, Klaus Michael Debatin, Mike Andrew Westhoff. Mechanisms of cell death escape in glioblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 2 (Clinical Trials and Late-Breaking Research); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(8_Suppl):Abstract nr LB304.

No abstract