Simvastatin Induces Unfolded Protein Response and Enhances Temozolomide-Induced Cell Death in Glioblastoma Cells

Abstract

Glioblastoma (GBM) is the most prevalent malignant primary brain tumor with a very poor survival rate. Temozolomide (TMZ) is the common chemotherapeutic agent used for GBM treatment. We recently demonstrated that simvastatin (Simva) increases TMZ-induced apoptosis via the inhibition of autophagic flux in GBM cells. Considering the role of the unfolded protein response (UPR) pathway in the regulation of autophagy, we investigated the involvement of UPR in Simva-TMZ-induced cell death by utilizing highly selective IRE1 RNase activity inhibitor MKC8866, PERK inhibitor GSK-2606414 (PERKi), and eIF2α inhibitor salubrinal. Simva-TMZ treatment decreased the viability of GBM cells and significantly increased apoptotic cell death when compared to TMZ or Simva alone. Simva-TMZ induced both UPR, as determined by an increase in GRP78, XBP splicing, eukaryote initiation factor 2α (eIF2α) phosphorylation, and inhibited autophagic flux (accumulation of LC3β-II and inhibition of p62 degradation). IRE1 RNase inhibition did not affect Simva-TMZ-induced cell death, but it significantly induced p62 degradation and increased the microtubule-associated proteins light chain 3 (LC3)β-II/LC3β-I ratio in U87 cells, while salubrinal did not affect the Simva-TMZ induced cytotoxicity of GBM cells. In contrast, protein kinase RNA-like endoplasmic reticulum kinase (PERK) inhibition significantly increased Simva-TMZ-induced cell death in U87 cells. Interestingly, whereas PERK inhibition induced p62 accumulation in both GBM cell lines, it differentially affected the LC3β-II/LC3β-I ratio in U87 (decrease) and U251 (increase) cells. Simvastatin sensitizes GBM cells to TMZ-induced cell death via a mechanism that involves autophagy and UPR pathways. More specifically, our results imply that the IRE1 and PERK signaling arms of the UPR regulate Simva-TMZ-mediated autophagy flux inhibition in U251 and U87 GBM cells.

Keywords: ER stress; autophagy; autophagy flux; glioblastoma; mevalonate cascade; statin.

Figure 1

Simvastatin–temozolomide (Simva–TMZ) treatment induces unfolded protein response (UPR) concomitant with cell death and DNA damage in GBM cells. U87 and U251 cells were treated with TMZ, Simva, and Simva–TMZ for 72 h as described in the material and methods section. Cells were collected and lysed. The expression of protein markers of UPR (GRP-78, IRE-1, XBP-1s, ATF6, eIF2α, and p-eIF2α) was determined by immunoblotting. glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the loading control. Simva–TMZ increased the protein amount of GRP-78, IRE-1, and XBP-1s, and it decreased the ratio of p-eIF2α/eIF2α as compared to the time-matched control. Immunoblots are representative of three different biological replicates.

Figure 2

Inositol-requiring enzyme 1 (IRE1α) RNase inhibition does not affect Simva–TMZ-induced cell death. (A,B) U87 and U251 cells were treated with different concentrations of MKC8866 (IRE-1α RNAse inhibitor; 10, 20, 30, 40, and 80 µM) for 48 and 72 h, and cell viability was evaluated by 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide) (MTT) assay. Control cells were treated with the solvent (DMSO) for each time point. MKC8866 significantly reduced cell viability in U87 (30 µM; 48 h, p < 0.05, 72 h, p < 0.001), (40 µM; 48 h, p < 0.05, 72 h, p < 0.0001), (80 µM; 48 h, p < 0.001, 72 h, p < 0.0001) and U251 (80 µM; 48 h, p < 0.05, 72 h, p < 0.0001) cells. (C,D) U87 and U251 cells were pretreated with MKC8866 (30 µM) for 4 h and subsequently co-treated with TMZ, Simva, or Simva–TMZ for 72 h. Cell viability was measured by MTT assay. MKC8866 did not significantly change Simva-, TMZ-, or Simva–TMZ-induced cell death in glioblastoma (GBM) cells. (E,F) U87 and U251 cells were pretreated with MKC8866 (30 µM) for 4 h and subsequently co-treated with TMZ, Simva, or Simva–TMZ for 72 h. Apoptosis was measured by Nicoletti assay. MKC8866 did not significantly change Simva-, TMZ-, or Simva-TMZ-induced apoptosis in GBM cells. Findings are expressed as the mean ± SD of three independent experiments for MTT assay (15 replicates) and three independent replicates for Nicoletti assay **** p < 0.0001). NS; not significant.

Figure 3

Simva–TMZ treatment induces IRE1 activation in GBM cells. (A) U87 and U251 cells were pretreated with MKC8866 (30 µM, 4 h), followed by co-treatment with TMZ, Simva, or Simva–TMZ for 72 h. XBP-1s protein levels were determined by Western blot analysis; GAPDH was used as the loading control. Simva–TMZ induced XBP-1 splicing in GBM cells, which was inhibited by MKC8866. (B,C) Densitometric analysis of the Western blot bands confirmed that Simva–TMZ significantly (p < 0.0001) induced XBP-splicing in an IRE1α-dependent manner. (D,E). U87 and U251 cells were pretreated with MKC8866 (30 µM, 4 h) and then co-treated with TMZ, Simva, or Simva–TMZ for 72 h, after which and XBP-1s mRNA expression was measured by real-time PCR. In line with the effects on protein amount, Simva–TMZ significantly (p < 0.0001) induced XBP splicing, which was inhibited by MKC8866. The data are expressed as the mean ± SD of three independent experiments ** p < 0.01, *** p < 0.001; **** p < 0.0001).

Figure 4

Simva–TMZ modulates the autophagy machinery via the IRE-1 pathway. (A) After pretreatment with MKC8866 (30 µM, 4 h), U87 and U251 cells were co-treated with TMZ, Simva, or Simva–TMZ for 72 h. The protein levels of Beclin-1, p62, LC3β-II, and LCβ-I were determined by immunoblotting. Simva–TMZ induced an inhibition of autophagy flux (accumulation of p62 and LC3β-II) in GBM cells. In Simva–TMZ-treated cells, MKC8866 increased p62 and Beclin-1 degradation, while it differentially affected the LC3β-II/LC3β-I ratio; GAPDH was used as loading control. Densitometric analysis of the Western blot bands confirmed that Simva–TMZ significantly induced Beclin-1 and p62 accumulation in both U87 and U251 cells (p < 0.0001), which was markedly prevented in the presence of MKC8866 (B,C,E,F). In addition, MKC8866 increased the LC3β-II/LC3β-I ratio in Simva–TMZ-treated U251 cells (p < 0.0001) (G), whereas it did not change LC3β-II/LC3β-I in U87 cells (p < 0.0001) (D). The data are shown as the mean ± SD from three independent experiments (* p < 0.05; ** p < 0.01, *** p < 0.001; **** p < 0.0001).

Figure 5

Protein kinase RNA-like endoplasmic reticulum kinase (PERK) inhibition further reduces cell viability in Simva–TMZ-treated U87 cells but not in U251 cells. (A,B) GBM cells (U87 and U251) were treated with different concentrations of PERK inhibitor GSK-2606414 (PERKi) (1, 5, 10, and 20 µM) for 48 and 72 h, after which cell viability was assessed by MTT assay. Controls were treated with the solvent (DMSO). The PERKi induced significant cell death at all concentrations in both cell lines, except for 1 µM in U87 cells. (C,D) U87 and U251 were pretreated with 5 µM PERKi for 30 min and then co-treated with TMZ, Simva, and Simva–TMZ for 72 h; cell viability was measured by MTT assay. GSK-2606414 (PERKi) significantly increased Simva–TMZ-induced cell death in U87 cells (p < 0.0001), but was without effect in Simva–TMZ-treated U251 cells. (E,F) U87 and U251 were pretreated with 5 µM PERKi for 30 min and then co-treated with TMZ, Simva, and Simva–TMZ for 72 h; apoptosis was measured by Nicoletti assay. GSEK–PERK inhibition did not significantly change Simva–TMZ-induced apoptosis in both cell lines. Data are expressed as the means ± SD of 15 replicates from three independent experiments for MTT assay and three independent replicates for Nicoletti assay (* p < 0.05; ** p< 0.01, *** p < 0.001, **** p < 0.0001).

Figure 6

PERK inhibition does not change the p-eIF2α/eIF2α ratio in Simva–TMZ-treated cells. (A) U87 and U251 were pretreated with PERKi (5 µM, 30 min) and then co-treated with Simva–TMZ for 72 h. The protein levels of eIF2α and p-eIF2α were determined using immunoblotting; GAPDH was used as a loading control. (B,C) Densitometric analysis of the immunoblots showed that Simva–TMZ by itself significantly reduced the p-eIF2α/eIF2α ratio, which was not further decreased by the PERKi in either cell line. Of note, control levels of p-eIF2α were significantly decreased by the PERKi as well. The data are expressed as the means ± SD of three independent experiments ** p < 0.01; **** p < 0.0001).

Figure 7

PERK inhibition differentially affects autophagy flux in U87 and U251 cells treated with Simva–TMZ. (A) U87 and U251 cells were pretreated GSK PERK inhibitor (5 µM, 30 min) and then co-treated with Simva–TMZ as described for 72 h. The protein levels of p62, LC3β-II, and LCβ-I were determined by immunoblotting. Simva–TMZ induced an inhibition of autophagy flux (accumulation of p62 and LC3β-II) in GBM cells. The PERKi decreased p62 degradation (autophagosome degradation) in both U87 and U251 cells, while it increased the LC3β-II/LC3β-I ratio in U251 cells and decreased it in U87 cells. GAPDH was used as a loading control. (B–E) Densitometric analysis of the Western blot bands to quantify p62 and LC3β-II/LC3β-I protein amount. Data are expressed as the mean ± SD of three independent experiments (** p < 0.01; **** p < 0.0001).

Figure 8

p-eIF2α phosphatase inhibition does not affect the cytotoxicity of Simva–TMZ in GBM cells. (A,B) U87 and U251 cells were treated with different concentrations of p-eIF2α phosphatase inhibitor (salubrinal; 1, 5, 10, and 20 µM) for 48 and 72 h. Controls were treated with the solvent (DMSO). Salubrinal induced significant cell death in both cell lines, except for 1 µM in U87 cells in 48 h. (C,D) U87 and U251 cells were pretreated with salubrinal (15 µM, 30 min) and then co-treated with Simva, TMZ, or Simva–TMZ for 72 h; cell viability was measured by MTT assay. Salubrinal had no significant effects on Simva-, TMZ-, or Simva–TMZ-induced cell death in either cell line. (E,F) U87 and U251 cells were pretreated with salubrinal (15 µM, 30 min) and then co-treated with Simva, TMZ, or Simva–TMZ for 72 h; apoptosis was measured by Nicoletti assay. Salubrinal had no significant effects on Simva-, TMZ-, or Simva–TMZ-induced apoptosis in either cell line. Data are expressed as the means ± SD of 15 replicates from three independent experiments for MTT assay and three independent replicates for Nicoletti assay (* p < 0.05; ** p < 0.01, *** p < 0.001, **** p < 0.0001).

Figure 9

p-eIF2α phosphatase inhibition increases the p-eIF2α/eIF2α ratio in Simva–TMZ treated in GBM cells. (A) U87 and U251 cells were pretreated with salubrinal (15 µM, 30 min) followed by co-treatment with Simva–TMZ for 72 h. Cell lysates were collected, and the p-eIF2α/eIF2α protein amount ratios were determined using immunoblotting; GAPDH was used as a loading control. (B,C) Densitometric analysis of the Western blot bands shows that salubrinal significantly (p < 0.0001) increased the p-eIF2α/eIF2αratio with Simva–TMZ treatment. Data are expressed as the means ± SD of three independent experiments (** p < 0.01, **** p < 0.0001).

Scheme 1

Simva–TMZ co-treatment induces UPR in GBM cells. It initiated IRE1α RNase activity and PERK activation. Activation of these arms of UPR induces Simva–TMZ inhibition of autophagy flux in GBM cells.