Three-dimensional model to predict cell‐mediated nanoparticles drug delivery for Glioblastoma treatment

Glioblastoma (GBM) is the most frequent and most malignant brain tumour, with a dismal five-year survival rate lower than 5%. treatment is complex due to the presence of the Blood Brain Barrier (BBB), the high genomic variability, and the heterogeneity of the tumour microenvironment (TME). Moreover, the presence of a cancer stem cell (CSCs) niche drives the development of treatment resistance. Identifying appropriate therapeutic agents and delivery mechanisms to grant effective targeting with reduced side effects is key in improving GBM management. Unfortunately, drug screening mainly relies on ethically debated animal models or on in vitro models, which do not fully replicate the TME and other relevant phenomena such as angiogenesis or immune response. To fill this gap, this study aims to develop a more reliable three-dimensional GBM model to investigate the transport and efficacy of nanoparticles (NPs)-based drug delivery systems. To achieve this, multicellular tumour spheroids (MTSs) modelling GBM histological heterogeneity were prepared. Different ratios of primary tumour cells (U87), microglia (HMC-3) and CSCs (GBM8) were used to prepare the MTSs and test the antitumor effect of a proteasome inhibitor, Bortezomib (BTZ), at different concentrations. BTZ is an extremely potent drug against GBM and GBM-CSC, as our in vitro results confirmed, but cannot be delivered to the brain due to its inability to cross the BBB. Hence, polymeric nanocarriers for BTZ encapsulation were developed and characterized to enhance drug accumulation at the target site. BTZ was successfully encapsulated in NPs with high efficiency (11±2%), achieving a sustained release over 5-7 days. The efficacy of BTZ-NPs against GBM MTSs was compared with that of free BTZ, showing a slightly delayed response, compatible with the slower drug release rate from the NPs. To further extend NPs uptake by tumour cells in the MTSs, we exploited microglia as cellular transporters for NPs by virtue of their capacity to penetrate the tumour mass. The efficacy of BTZ-NPs delivered through microglia was also assessed. To explore NPs transport, we increased the GBM model complexity, by including angiogenesis through a microfluidic platform (OrganoPlate® Graft, MIMETAS). The device has a central chamber to house the MTS within an extracellular matrix (ECM) gel and two lateral perfusion channels, which were seeded with human cerebral microvascular endothelial cells (HBEC-5i) to mimic blood vessels. The administration of angiogenic signals induces vessel sprouting towards the ECM gel and the MTS. Once the vascularization procedure was optimized, the morphology and integrity of the microvasculature was assessed through immunostaining and perfusion assays. Extravasation of free NPs and microglia was monitored in the vascularized model with fluorescent markers and cellular tracers to identify the most effective carrier. CD31-staining confirmed the homogeneous presence of endothelial cells forming tight junctions (observed by ZO-1 staining). The microvasculature replicates the retention of 120 nm-diameter NPs observed in previous in vivo studies. These promising results point to the possibility of updating the model to a higher level of complexity, for instance by including other TME components, like pericytes and astrocytes. Moreover, the model can be used to investigate nanocarrier- and cell-mediated transport through the BBB to ensure targeted, effective drug accumulation with minimal adverse effects.