Sorafenib selectively depletes human glioblastoma tumor-initiating cells from primary cultures

Abstract

Glioblastomas are grade IV brain tumors characterized by high aggressiveness and invasiveness, giving patients a poor prognosis. We investigated the effects of the multi-kinase inhibitor sorafenib on six cultures isolated from human glioblastomas and maintained in tumor initiating cells-enriching conditions. These cell subpopulations are thought to be responsible for tumor recurrence and radio- and chemo-resistance, representing the perfect target for glioblastoma therapy. Sorafenib reduces proliferation of glioblastoma cultures, and this effect depends, at least in part, on the inhibition of PI3K/Akt and MAPK pathways, both involved in gliomagenesis. Sorafenib significantly induces apoptosis/cell death via downregulation of the survival factor Mcl-1. We provide evidence that sorafenib has a selective action on glioblastoma stem cells, causing enrichment of cultures in differentiated cells, downregulation of the expression of stemness markers required to maintain malignancy (nestin, Olig2 and Sox2) and reducing cell clonogenic ability in vitro and tumorigenic potential in vivo. The selectivity of sorafenib effects on glioblastoma stem cells is confirmed by the lower sensitivity of glioblastoma cultures after differentiation as compared with the undifferentiated counterpart. Since current GBM therapy enriches the tumor in cancer stem cells, the evidence of a selective action of sorafenib on these cells is therapeutically relevant, even if, so far, results from first phase II clinical trials did not demonstrate its efficacy.

Keywords: Mcl-1; glioblastoma; sorafenib; stemness; therapy; tumor initiating cells.

Figure 1. Inhibitory effect of sorafenib on proliferation and DNA synthesis of human GBM TICs. (A) TICs from six GBMs were plated in 96-well microplates and then treated with different concentrations of sorafenib 5 µM (left), 10 µM (center) and 20 µM (right) for 48 and 72 h. Results are expressed as means of percentage of cells vs. vehicle treated controls (measured by MTT assay) from four independent experiments, each performed in quadruplicate. All SE are lower than 5% of each mean value. (B) Sorafenib effects on DNA synthesis of GBM TIC cultures evaluated by BrdU incorporation in control (vehicle-treated) and in 24 h sorafenib-treated cells (IC₅₀). Results are expressed as means, and error bars represent standard deviations from three different experiments. * p < 0.05; ** p < 0.01.

Figure 2. Sorafenib induces apoptosis in GBM TICs. Apoptosis/cell death induction in GBM TICs after 24 and 48 h of treatment with sorafenib IC₅₀ (S) or control (C; vehicle-treated). Bars represent overall cell death and white boxes the apoptosis contribution (early and late apoptosis). Results are expressed as means, and error bars represent standard deviations from four different experiments. * p < 0.05; ** p < 0.01.

Figure 3. Effects of sorafenib on STAT-3, MAPK pathway, Akt and Mcl-1. Summary of western blots that evaluate the levels of phosphorylated/active forms of STAT-3, MEK, ERK1/2, Akt and Mcl-1 content after 24 h treatment with three concentration of sorafenib (0.5X IC₅₀, IC₅₀ and 2x IC₅₀) are showed. Results are representative blots from three independent experiments. Tubulin expression was evaluated to normalize immunoreactive band to total protein load. Densitometric evaluations of the data are reported in the Figure S1. Sorafenib treatment reduces the phosphorylation of MEK and ERK1/2 as well as of Akt in the more responsive cultures (GBM 3, 23, 24 and 17); in the less responsive culture (GBM 19), only a reduction of Akt activity at the higher dose is observed, and no changes are detected in GBM 6 culture. The STAT-3 phosphorylation levels differ among the six cultures tested without any correlation to the anti-proliferative or pro-apoptotic effects of sorafenib. The reduction of Mcl-1 correlates to the pro-apoptotic effects.

Figure 4. Sorafenib selectivity for cancer stem cells in GBM TIC cultures: Modulation of marker expression and impairment of clonogenic and tumorigenic potential. (A) Effects of sorafenib on in vitro clonogenic potential of GBM 3, 19 and 6. Cells were treated for 48 h with sorafenib (IC₅₀) or vehicle (control), and clone formation was estimated after two weeks from clonal seeding. Results are reported as mean percentage of clones from three independent experiments, and error bars represent standard deviations. (B) Expression of differentiation (GFAP and MAP2, upper panels) and stemness (Nestin and Olig2, middle panels and Sox2, lower panel) markers in GBM TIC cultures after 48 h of treatment with sorafenib (IC₅₀) or vehicle. Results are reported as mean percentage from three different experiments and error bars represent standard deviations. All determinations were performed by FACS, except that of Sox2 that was performed by western blot. (C) Reduced sensitivity to sorafenib of differentiated GBM cultures, as compared with undifferentiated TICs. Undifferentiated or differentiated cells were treated for 48 h with sorafenib (20 µM), and cell number was determined by MTT assay. Results are expressed as mean percentage of cells vs. control and error bars represent standard deviations of two experiments. (D) Reduction of GBM TIC tumorigenic potential after sorafenib treatment. Results are represented as tumor take rate in mice injected with untreated or in vitro sorafenib treated (48 h) cells from GBM 3. *p < 0.05; **p < 0.01 vs. control.