Tumor protease-activated theranostic nanoparticles for MRI-guided glioblastoma therapy

Abstract

Rationale: As a cancer, Glioblastoma (GBM) is a highly lethal and difficult-to-treat. With the aim of improving therapies to GBM, we developed novel and target-specific theranostic nanoparticles (TNPs) that can be selectively cleaved by cathepsin B (Cat B) to release the potent toxin monomethyl auristatin E (MMAE). Methods: We synthesized TNPs composed of a ferumoxytol-based nanoparticle carrier and a peptide prodrug with a Cat-B-responsive linker and the tubulin inhibitor MMAE. We hypothesized that intratumoral Cat B can cleave our TNPs and release MMAE to kill GBM cells. The ferumoxytol core enables in vivo drug tracking with magnetic resonance imaging (MRI). We incubated U87-MG GBM cells with TNPs or ferumoxytol and evaluated the TNP content in the cells with transmission electron microscopy and Prussian blue staining. In addition, we stereotaxically implanted 6- to 8-week-old nude mice with U87-MG with U87-MG GBM cells that express a fusion protein of Green Fluorescence Protein and firefly Luciferase (U87-MG/GFP-fLuc). We then treated the animals with an intravenous dose of TNPs (25 mg/kg of ferumoxytol, 0.3 mg/kg of MMAE) or control. We also evaluated the combination of TNP treatment with radiation therapy. We performed MRI before and after TNP injection. We compared the results for tumor and normal brain tissue between the TNP and control groups. We also monitored tumor growth for a period of 21 days. Results: We successfully synthesized TNPs with a hydrodynamic size of 41 ± 5 nm and a zeta potential of 6 ± 3 mV. TNP-treated cells demonstrated a significantly higher iron content than ferumoxytol-treated cells (98 ± 1% vs. 3 ± 1% of cells were iron-positive, respectively). We also found significantly fewer live attached cells in the TNP-treated group (3.8 ± 2.0 px²) than in the ferumoxytol-treated group (80.0 ± 14.5 px², p < 0001). In vivo MRI studies demonstrated a decline in the tumor signal after TNP (T₂= 28 ms) but not control (T₂= 32 ms) injections. When TNP injection was combined with radiation therapy, the tumor signals dropped further (T₂ = 24 ms). The combination therapy of radiation therapy and TNPs extended the median survival from 14.5 days for the control group to 45 days for the combination therapy group. Conclusion: The new cleavable TNPs reported in this work accumulate in GBM, cause tumor cell death, and have synergistic effects with radiation therapy.

Keywords: MMAE; MRI; glioblastoma; nanoparticles; theranostic.

Figure 1

TNP design. (A) Synthesis of TNPs. (B) TNPs enter cancer cells and Cat B cleaves drug MMAE off the TNPs. (C) Lysosomal Cat B cleaves TNPs and releases MMAE to inhibit tubulin polymerization and cause tumor cell death.

Figure 2

TNP characterization. (A) Chromatography of free MMAE released from the TNPs. (B) Product ion of free MMAE. (C) Representative TEM image of the original ferumoxytol (left) and TNPs (right). (D) Activation of TNP is cathepsin-dependent. (E) MMAE releasing profile of TNP. (F) T1-relaxivity and (G) T2-relaxivity were measured from changes in relaxation time at different concentrations of ferumoxytol or TNP. (H) MR images of increasing concentrations of TNP nanoparticles in Ficoll solution on T2-weighted MSME sequences and color-coded T2-maps using the FIRE scheme. Scale bars represent 20 nm.

Figure 3

Representative TEM images of GBM cells exposed to TNPs. U87-MG cells were incubated with TNPs at a concentration of 4 nM and observed under TEM at (A) 0, (B) 30, and (C) 120 min post-incubation. Secondary lysosomes are indicated by yellow arrowheads. Nanoparticles are indicated by red arrowheads. White scale bars are 2 μm. Inset shows zoomed-in views of nanoparticles around the cell surface or in the lysosomes. Black scale bars in the insets are 400 nm.

Figure 4

TNPs are internalized into GBM cells and kill tumor cells. U87-MG cells were exposed to (A) PBS, peptide prodrug (C-MMAE), free drug (MMAE), TNP and unconjugated ferumoxytol. (B) Internalized drugs or nanoparticles were visualized under a microscope. Internalizations are indicated by red arrowheads. (C) Internalized nanoparticles were measured by UV-Vis. Cell viability was measured by (D) crystal violet staining and (E) the cell titer blue assay. * p < 0.05, ** p < 0.005, one-way ANOVA.

Figure 5

TNPs inhibit GBM tumor growth. (A) Treatment scheme of the TNPs. (B) Representative coronal T2-weighted MR images of MMAE-, ferumoxytol-, TNP1×-, TNP2×-, and PBS-injected (control) orthotopic GMB mouse brains (n = 4). Top panel shows FSE images, middle panel shows T2 maps, and bottom panel shows fused images. Green arrows indicate darkened tumors. The regions of interest (ROIs) were defined on the T2-weighted images to identify the tumors. (C) Quantification of the bioluminescent signals in control and treated U87-MG tumors on day 0 and 7. Results are represented as mean ± SEM from three independent experiments. (D) Weight changes of mice on day 0 vs. day 7. Fold change in body weights represents the ratio of body weight after treatment on day 7/body weight before treatment on day 0. TNP1× indicates a total dose of 0.3 mg/Kg and TNP2× indicates a total dose of 0.6 mg/Kg. * p < 0.05, ** p < 0.005, one-way ANOVA.

Figure 6

Combination therapy of TNP and radiation therapy inhibited tumor growth and improved survival. (A) Cat B expression of U87-MG GBM cells after exposure to TNP and/or radiation therapy (RT). (B) U87 GBM cell viability after receiving TNP and/or RT treatment. (C) The amount of MMAE (M) in culture media of U87-MG cells treated with TNP with or without prior treatment of RT. (D) In vivo combination treatment schedule. (E) Representative coronal T2-weighted MR images fused with measured T2 values within the tumor ROIs of TNP-, RT-, TNP+RT- and PBS-injected (control) orthotopic GMB mouse brains. Top two panels show pre-treatment images and corresponding T2 measurements, and bottom two panels show 24 h post-treatment results for comparison. (F) BLI images of animals at day 0, 7, 14, and 21 after first treatment. (G-H) Comparison of T2 relaxation times in tumor and normal brain areas. Empty bars indicate pre-treatment; filled bars indicate post-treatment. (I) Quantification of BLI signals. (J) Kaplan-Meyer survival curves of control and treated mice demonstrate a significant survival benefit of TNP+RT as compared to vehicle, log-rank Mantel-Cox test.