A Novel Patient-Personalized Nanovector Based on Homotypic Recognition and Magnetic Hyperthermia for an Efficient Treatment of Glioblastoma Multiforme

Abstract

Glioblastoma multiforme (GBM) is the deadliest brain tumor, characterized by an extreme genotypic and phenotypic variability, besides a high infiltrative nature in healthy tissues. Apart from very invasive surgical procedures, to date, there are no effective treatments, and life expectancy is very limited. In this work, an innovative therapeutic approach based on lipid-based magnetic nanovectors is proposed, owning a dual therapeutic function: chemotherapy, thanks to an antineoplastic drug (regorafenib) loaded in the core, and localized magnetic hyperthermia, thanks to the presence of iron oxide nanoparticles, remotely activated by an alternating magnetic field. The drug is selected based on ad hoc patient-specific screenings; moreover, the nanovector is decorated with cell membranes derived from patients' cells, aiming at increasing homotypic and personalized targeting. It is demonstrated that this functionalization not only enhances the selectivity of the nanovectors toward patient-derived GBM cells, but also their blood-brain barrier in vitro crossing ability. The localized magnetic hyperthermia induces both thermal and oxidative intracellular stress that lead to lysosomal membrane permeabilization and to the release of proteolytic enzymes into the cytosol. Collected results show that hyperthermia and chemotherapy work in synergy to reduce GBM cell invasion properties, to induce intracellular damage and, eventually, to prompt cellular death.

Keywords: glioblastoma; homotypic targeting; magnetic hyperthermia; personalized medicine.

Figure 1

A) Intensity distribution (%) as a function of the hydrodynamic diameter (nm) for MNVs (black) and CDMNVs (red). B) ζ‐potential (mV) distribution of MNVs (black) and CDMNVs (red). C,D) Representative BF‐TEM images of MNVs and CDMNVs.

Figure 2

A) Scheme of CDMNV coating, with an optical microscopy image of a typical patient‐derived GBM cell culture. B) FTIR spectra of MNVs (black), CM extracts (blue), and CDMNVs (red). C–E) High‐resolution XPS scans for MNVs (black) and CDMNVs (red) for Fe 2p, C1s, and N1s. F) Table summarizing the feature of the CM coating in CDMNVs. G) Western blotting of typical proteins involved in the homotypic adhesion mechanism (CD44, N‐cadherin, neuroplastin SDR‐1, beta‐catenin) in CM extracts of U87‐MG cells and of patient‐derived GBM cells and in CDMNVs coated with the CM extract obtained from the same patient's cells; the molecular weight of the proteins is reported on the right.

Figure 3

A) VSM measurements for IONPs (black) and CDMNVs (blue) powders acquired at room temperature. Inset: a zoom of the low magnetization and low field area. B) Temperature profile of CDMNVs (3 mg mL⁻¹ in terms of iron oxide) during the stimulation with an AMF (AMF ON) and thereafter (AMF OFF).

Figure 4

Cumulative release (%) of regorafenib from CDMNVs in different conditions: A) pH 4.5 (black squares), pH 4.5 + AMF (red circles), pH 7.4 (blue triangles), pH 7.4 + AMF (green upside down triangles), cell lysate (dark green diamond), cell lysate + AMF (dark blue left triangles). B) pH 4.5 + 100 × 10⁻⁶ m H₂O₂ (black squares), pH 4.5 + 100 × 10⁻⁶ m H₂O₂ + AMF (red circles), pH 7.4 + 100 × 10⁻⁶ m H₂O₂ (blue triangles), pH 7.4 + 100 × 10⁻⁶ m H₂O₂ + AMF (green upside down triangles). In the insets in (A) and (B), a zoom on the release profiles during the last time points to better highlight differences in regorafenib release in the different conditions.

Figure 5

A) Representative confocal images depicting the uptake of CD*MNVs and CDMNVs (in green) by different cell cultures (hCMEC/D3, human astrocytes, neuron‐like cells, patient‐derived GBM cells) after 6 h of treatment in dynamic conditions. B) Quantitative analysis showing nanovectors/cells co‐localization (*p < 0.05). C) Quantitative analysis showing the amount of CD*MNVs and CDMNVs in the lower chamber of the BBB in vitro model at different experimental points (*p < 0.05); the red dashed line indicates the end of the nanovector perfusion. D) Representative confocal images of CD*MNVs and CDMNVs (in white) internalized by patient‐derived GBM cells upon BBB model crossing after 48 h, and E) quantitative analysis showing the nanovectors / cells co‐localization (*p < 0.05).

Figure 6

Confocal analysis of the internalization of CDMNVs in patient‐derived GBM cells mediated by A) clathrin‐coated vesicles or B) caveolae, at 6 and 24 h. C) Quantitative evaluation of colocalization (through Mader's overlap analysis) at 24 and 72 h (*p < 0.05). D) Confocal analysis of the co‐localization of CDMNVs (in white) with lysosomes (in red) in patient‐derived GBM cells after 6, 24, and 72 h of treatment. E) Quantitative evaluation of co‐localization through Mader's overlap analysis (*p < 0.05).

Figure 7

Metabolic activity of patient‐derived GBM spheroids treated for A) 24 h and B) 72 h with increasing concentrations of CDMNVs (green), Reg‐CDMNVs (red), and free regorafenib (yellow), the latter at concentrations corresponding to the respective loading in Reg‐CDMNVs. All data were normalized with respect to control group (gray) (*p < 0.05).

Figure 8

A) Representative cathepsin D confocal imaging (in green) in patient‐derived GBM cells after different treatment protocols. B) Quantitative analysis of the cathepsin D signal intensity (*p < 0.05).

Figure 9

A) Representative HSP70 confocal imaging (in green) in patient‐derived GBM cells after different treatment protocols. B) Quantitative analysis of HSP70 signal intensity (*p < 0.05). C) Representative TRAP1 confocal imaging (in green) in patient‐derived GBM cells after different treatment protocols. D) Quantitative analysis of TRAP1 signal intensity (*p < 0.05).

Figure 10

A) Representative optical images of migration tests performed on patient‐derived GBM spheroids treated with different experimental protocol. B) Quantitative analysis of the migration length of cells from patient‐derived GBM spheroids represented as box chart. The box represents the interquartile range, the bold line is the median, the circle is the average, the bar is the 5%–95% percentile, and the crosses are the minimum and the maximum value (*p < 0.05).

Figure 11

A) Metabolic activity of patient‐derived GBM cells upon different treatment protocols. B) Investigation of the expression of caspase‐8 in patient‐derived GBM spheroids in analogous experimental groups (*p < 0.05). C) Representative confocal images of the live/dead assay on patient‐derived GBM cells treated with regorafenib, CDMNVs, or Reg‐CDMNVs, with or without AMF stimulation. D) Dead and live cells (%) quantified from the analysis of the images in B (*p < 0.05).