Targeting neuroblastoma stem cells with retinoic acid and proteasome inhibitor

Abstract

Background: Neuroblastma cell lines contain a side-population of cells which express stemness markers. These stem-like cells may represent the potential underlying mechanism for resistance to conventional therapy and recurrence of neuroblastoma in patients.

Methodology/principal findings: To develop novel strategies for targeting the side-population of neurobastomas, we analyzed the effects of 13-cis-retinoic acid (RA) combined with the proteasome inhibitor MG132. The short-term action of the treatment was compared with effects after a 5-day recovery period during which both chemicals were withdrawn. RA induced growth arrest and differentiation of SH-SY5Y and SK-N-BE(2) neuroblastoma cell lines. Inhibition of the proteasome caused apoptosis in both cell lines, thus, revealing the critical role of this pathway in the regulated degradation of proteins involved in neuroblastoma proliferation and survival. The combination of RA with MG132 induced apoptosis in a dose-dependent manner, in addition to promoting G2/M arrest in treated cultures. Interestingly, expression of stem cell markers such as Nestin, Sox2, and Oct4 were reduced after the recovery period of combined treatment as compared with untreated cells or treated cells with either compound alone. Consistent with this, neurosphere formation was significantly impaired by the combined treatment of RA and MG132.

Conclusions: Given that stem-like cells are associated with resistant to conventional therapy and are thought to be responsible for relapse, our results suggest that dual therapy of RA and proteasome inhibitor might be beneficial for targeting the side-population of cells associated residual disease in high-risk neuroblastoma.

Figure 1. Effects of the combined RA/MG132 treatment on apoptosis and cell cycle.

(A) The neuroblastoma cell line SK-N-BE(2) was treated with increasing doses of MG132 (100nM -1μM) for 3 days and analysed by flow cytometry using the fluorescent dye Yoyo1. Flow cytometry diagram and quantitative data. The percentage of cells is indicated in each quadrant. (B) Western blot analysis of the 4 treatment conditions after 3 days. (C) Quantitative data of the different phases of the cell cycle after 24h treatment. Notice that after 24h almost no apoptosis has been detected in MG132 and RA/MG132 cells. The increases of the percentage of cells in G2/M in response to combined treatment or to MG132 are statistically significant (asterisks). (D) Quantitative data of the different phases of the cell cycle after 72h of treatment.

Figure 2. Effects of the combined RA/MG132 treatment on cell cycle (A) Flow cytometry diagram of propidium iodide stained cells after 24h treatment with RA and MG132.

The percentage of cells is indicated in each quadrant. (B) Flow cytometry diagram of propidium iodide stained cells after 72h treatment with RA and MG132. The quantitative data of this experiment are shown in Figure 1C and D.

Figure 3. Combination of MG132 and RA induces p53-independent apoptosis and RelA inhibition.

(A, B) The apoptosis induced by MG132 and MG132/RA is p53 independent. (A) Luciferase assay performed on SK-N-BE(2) cells and (B) SH-SY5Y cells to assess transcriptional activity of p53 upon treatment. SK-N-BE(2) cells express mutant p53, SH-SY5Y neuroblastoma cells are wildtype for p53. LRU= Luciferase reference unit. (C) Representative western blot showing the accumulation of p53 protein in response to MG132 treatment. MG132 treatment does not induce up-regulation of the p53 responsive gene p21/Waf1 in SK-N-BE(2) cells, (D) in contrast to SH-SY5Y cells. (E, F) Effects of combined RA/MG132 treatment on NF-κB signalling. Both MG132 as well as RA cause significant decrease of nuclear expression of RelA (p65) protein. (E) Immunofluorescent detection of sub-cellular localization of RelA (p65) protein under the different treatment conditions. Arrows indicate RelA (p65) protein in the nucleus. Scale bar = 25μm. (F) Quantitative data. The reduction of the cell fraction with nuclear RelA expression in response to combined treatment is statistically significant (*) as compared to control (p< 0.037), to RA treated cells (p=0.037), but also to MG132 treated cells (p=0.016).

Figure 4. Effects of the combined RA/MG132 treatment on the expression stem cell- related markers.

(A) Western blot analysis at completion of a 3 day treatment with (+) or without (-) RA and/or MG132, as indicated, and (B) at completion of a 3 day treatment plus 5 additional days in the absence of compounds.

Figure 5. Combination of MG132 and RA reduces long-term expression of stemness markers.

Western blot analysis at completion of a 3 day treatment with (+) or without (-) RA and/or MG132, as indicated, plus 5 additional days in the absence of compounds. In the lane labelled with an asterisk, the recovery was performed with conditioned medium from untreated cells.

Figure 6. Effects of the combined RA/MG132 treatment on the expression of Nestin, Oct4 and IR-red viability at completion of a 3-day treatment.

(A) Cells were treated with (+) or without (-) RA and/or MG132, as indicated, for 3 days. Flow cytometry dot plots and graphical representation of viable Nestin labelled (+) or not labelled (-) cells. The percentage of cells is indicated in each quadrant. (B) Flow cytometry dot plots and graphical representation of viable Oct4 labelled (+) or not labelled (-) cells. The percentage of cells is indicated in each quadrant.

Figure 7. Effects of the combined RA/MG132 treatment on the expression of Nestin, Oct4 and IR-red viability at completion of a 3 day treatment plus 5 additional days of recovery without compounds.

(A) Flow cytometry dot plots and graphical representation of viable Nestin labelled (+) or not labelled (-) cells. The percentage of cells is indicated in each quadrant. (B) Flow cytometry dot plots and graphical representation of viable Oct4 labelled (+) or not labelled (-) cells. The percentage of cells is indicated in each quadrant.

Figure 8. Study of viability and expression of stem-cell markers of cells treated with MG132 and RA.

(A) Quantitative data of CD34-expressing cells (+) or non-expressing cells (-) that are dead or remain alive after the treatment at completion of a 3 day treatment. (B) Quantification of Nanog-expressing (+) or non-expressing cells (-) at completion of a 3 day treatment. (C) Confocal microscopy analysis of Oct4 stained cells at completion of a 3 day treatment. Scale bar = 15μm.

Figure 9. Effects of the combined RA/MG132 treatment on neuronal differentiation.

(A) Cells were treated with (+) or without (-) RA and/or MG132, as indicated. Confocal microscopy analysis of MAP2 stained cells at completion of a 3 day treatment. Scale bar = 20µm. (B) Western blot analysis of MAP2 at completion of a 3 day treatment and after a recovery period of 5 additional days.

Figure 10. Combined treatment with RA and MG132 impairs neurosphere formation.

(A-D) After completion of a 3-day treatment, SK-N-BE(2) cells were seeded on an ultra-low attachment 6-well plates and cultured in sphere medium (SM). After 9 days of culture, cells were examined by phase-contrast microscopy. Scale bars: 50µM. (E) Higher magnification of a sphere, scale bar: 50µM. (F) Neurospheres ≥50 µm were quantified after five days in SM culture and (G) the same cultures were quantified after 9 days in SM culture, defined as the percentage of spheres in each sample in relation to that in a control sample (without RA and MG132). The primary assay in (F) shows the percentage of stem/progenitor cells in the population, and the secondary assay in (G) shows self-renewal capability. Data represents means ± SE of three independent experiments.