Calnexin promotes glioblastoma progression by inducing protective mitophagy through the MEK/ERK/BNIP3 pathway

Abstract

Rationale: Glioblastoma multiforme (GBM), one of the most malignant tumors of the central nervous system, has a poor prognosis, mainly because of its high recurrence caused by the rapid development of drug resistance to postoperative chemotherapy. Although macroautophagy/autophagy is believed to be a fundamental factor in tumor survival during chemotherapy, there is still a lack of autophagy biomarkers for predicting patient prognosis and chemotherapeutic efficacy in clinical practice. Methods: We combined transcriptomic and single-cell sequencing data to identify differentially expressed autophagy-related genes in gliomas. Overexpression of calnexin (CANX), a key gene related to protein folding, and its secretion in the endoplasmic reticulum (ER) was identified, suggesting poor prognosis in GBM patients. The autophagy flow related to CANX was detected by transmission electron microscopy (TEM), Western blotting, and immunofluorescence. Flow cytometry, cell proliferation, activity assays, and the GBM intracranial xenograft mouse model were employed to validate CANX's role in GBM progression. Results: CANX knockdown inhibited proliferation and autophagosome formation in GBM cells. On the other hand, CANX overexpression increased mitogen-activated protein kinase (MAPK) activity, leading to the accumulation of BNIP3 (CL2/adenovirus E1B 19 kDa interacting protein 3, a critical factor regulating mitophagy) and protective mitophagy. Notably, when combined with temozolomide (TMZ), CANX knockdown extended the lifespan of GBM-bearing mice. Additionally, our studies revealed that the classic calcium inhibitor nimodipine (ND) decreased CANX expression and thus enhanced the sensitivity to TMZ. Conclusions: Our findings indicate that CANX functions as an oncogene in GBM. We also characterize the CANX/MEK/ERK/BNIP3 mitophagy pathway, provide new insights into the molecular mechanism of GBM drug resistance, and identify a therapeutic target.

Keywords: calnexin; glioblastoma; mitophagy; nimodipine; temozolomide.

Figure 1

CANX is an Autophagy-Related Gene that is Differentially Expressed in Gliomas and is Associated with Glioma Prognosis. (A) A volcano plot showing the DEGs in gliomas identified by data from the TCGA and GTEx databases. (B) A Venn diagram showing the intersection of DEGs with ARGs. (C) Clustering heatmap of DE-ARGs. (D) Prognostic model for glioma patients encompassing DE-ARGs on the basis of TCGA data. (E) Multivariate Cox regression analysis revealed the hazard ratios for each independent prognostic indicator gene. (F) Autophagy scores were calculated from a single-cell database. (G) Correlation analysis between CANX expression and autophagy. (H) Kaplan‒Meier survival curves were plotted on the basis of TCGA data for survival analysis of the high and low CANX expression groups. (I) Analysis of CANX expression in normal brain tissue (NBT) and tissues from patients with gliomas of different grades according to data from the TCGA and CGGA databases. (J) Representative images and statistical analysis of IHC staining results for CANX according to data from the HPA database. (K) Images of IHC staining for CANX in gliomas of different grades (scale bar: 100 μm). (L) Western blot showing the difference in CANX expression between NBT and glioma tissues. (M) Immunoblotting results showing CANX expression in normal human astrocytes and various glioma cells. (N) qPCR results showing the RNA expression levels of CANX in NHAs and different glioma cell lines. (O) Statistical analysis of normalized CANX protein levels on the basis of the Western blot results. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001, ****P < 0.0001 between the two indicated treatments.

Figure 2

Knockdown of CANX Inhibits GBM Cell Migration and Promotes GBM Cell Apoptosis. (A) Volcano plot showed the DEGs identified based on the TCGA data. (B) KEGG enrichment analysis of the DEGs. (C) Analysis of the knockdown efficiency of three siRNAs targeting CANX in T98G, LN229, and GBM#P3 cells via Western blotting. (D) Statistic results of the knockdown efficiency of three siRNAs via Western blotting. (E) Analysis of the knockdown efficiency of three siRNAs targeting CANX in T98G, LN229, and GBM#P3 cells via qRT‒PCR. (F) EdU staining assay of T98G, LN229, and GBM#P3 cells treated with siCANX and siNC (scale bar: 100 μm). (G) Statistic results of the EdU staining assay of T98G, LN229, and GBM#P3 cells treated with siCANX and siNC. (H) A CCK-8 assay was used to analyze cell viability of T98G, LN229, and GBM#P3 cells with CANX knockdown. (I, J) Flow cytometry analysis of propidium iodide (PI) and Annexin V-FITC staining in T98G, LN229, and GBM#P3 cells transfected with siNC, siCANX-2, and siCANX-3 for the analysis of apoptosis and the statistic results of the proportion of apoptosis cells and necrotic apoptosis. (K, L) Validation of the efficiency of the CANX overexpression lentivirus via Western blotting in U118-MG cells and the statistic results of the relative protein level of CANX in U118-MG cells. (M) Validation of the efficiency of the CANX overexpression lentivirus via qRT‒PCR. (N, O) Analysis of apoptosis in CANX-overexpressing cells under normal and starvation conditions in U118-MG cells and the statistic results of the proportion of apoptosis cells and necrotic apoptosis cells. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 3

Knockdown of CANX Inhibits Autophagosome Formation. (A-C) TEM images showing the changes in the number of autophagosomes in the GBM#P3, T98G and LN229 cell lines following CANX knockdown under normal conditions and after bafilomycin A1 (Baf) treatment (scale bar: 3 μm). The arrows indicate autophagosomes. (D) Western blot analysis confirmed the changes in the expression of the autophagy markers LC3B-II and SQSTM1 after CANX knockdown. (E) The relative changes in LC3B levels were quantified via Western blot analysis. (F) Immunofluorescence staining and confocal microscopy revealed alterations in LC3B expression following CANX knockdown (scale bar: 50 μm). The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 4

Validation of the Role of CANX in Regulating Autophagic Flux in GBM. (A, B) Western blot analysis demonstrated the effect of CANX knockdown on LC3B levels in the presence of the autophagy inhibitors Baf and 3-MA in GBM#P3, T98G an LN229 cells. (C, D) The Western blot results validated the impact of CANX knockdown on MAPLC3B-II levels when BECN1 and ATG7 in GBM#P3, T98G an LN229 cells with siCANX and siNC. (E, F) Immunofluorescence assays of the LN229 cell line revealed the effect of CANX knockdown on autophagosome formation after treatment with 3-MA and Baf (scale bar: 25 μm). (G, H) Confocal microscopy analysis of immunofluorescence in the GBM#P3 cell line was used to assess the impact of CANX knockdown on autophagic flux following the knockdown of ATG7 and BECN1 (scale bar: 25 μm).

Figure 5

CANX Overexpression Promotes Protective Autophagy in GBM Cells In Vitro. (A, B) Assessment of the viability of NC-overexpressing/CANX-overexpressing U118-MG cells and shNC/shCANX-treated GBM#P3 cells after incubation with 3-MA or siBECN1 via the CCK-8 assay. (C, D) Representative images of the EdU staining assay of NC-overexpressing/CANX-overexpressing U118-MG cells and shNC-treated/shCANX-treated GBM#P3 cells after 3-MA or siBECN1 incubation (scale bar: 100 μm). (E, F) Statistical analysis of the EdU staining assay of NC-overexpressing/CANX-overexpressing U118-MG cells and shNC-treated/shCANX-treated GBM#P3 cells after 3-MA or siBECN1 incubation. (G, H) Flow cytometry analysis of Annexin V-FITC and PI staining in NC-overexpressing/CANX-overexpressing U118-MG cells and shNC-treated/shCANX-treated GBM#P3 cells after 3-MA or siBECN1 incubation for the analysis of apoptosis. (I, J) Statistic results of the proportion of apoptosis U118-MG and GBM#P3 cells and necrotic apoptosis after 3-MA or siBECN1 incubation. *P < 0.05, **P < 0.01, and ***P < 0.001 between the two indicated treatments.

Figure 6

Inhibition of CANX Enhances Sensitivity to TMZ Chemotherapy. (A) Comparison of the IC50 values of TMZ between cells with high and low CANX expression on the basis of RNA expression data from the TCGA database. (B) Correlation between CANX expression levels and the IC50 values of TMZ in glioma cells. (C) Western blot analysis showing the inhibition of TMZ-induced autophagy following CANX knockdown. (D) Confocal microscopy images of GFP-LC3B and RFP-mito illustrating mitophagy under TMZ treatment and CANX knockdown in T98G and GBM#P3 cells (scale bar, 40 µm). (E) Statistic results of GFP-LC3 puncta with RFP-Mito per cell under TMZ treatment and CANX knockdown in T98G and GBM#P3 cells. (F, G) EdU assay and statistic results showing the effect of reduced CANX expression on TMZ efficacy (scale bar: 100 μm). (H) CCK-8 assay showing the effect of CANX inhibition on sensitivity to TMZ. (I, J) Apoptosis assays demonstrating the impact of reduced CANX expression on TMZ-induced apoptosis with statistic results in GBM#P3 and T98G cells. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 7

CANX Knockdown Enhances the Efficacy of TMZ in an Orthotopic GBM Mouse Model. (A) GBM#P3 cells expressing luciferase were orthotopically implanted into immunosuppressed nude mice, and tumor growth was monitored via the IVIS-200 imaging system. Bioluminescent signals were measured on days 7, 14, 21, and 28 post-implantation. (B) HE staining showing the extent of tumor invasion within the brain. (C) Overall survival was determined via Kaplan‒Meier survival curves, with statistical significance assessed via the log-rank test. (D) Bioluminescence values for assessing tumor growth. (E-L) Immunohistochemical staining images and statistical analysis of MKi67, LC3B, CANX, and cleaved caspase 3 expression in tumors from each group (scale bar: 100 μm) (E-L). The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 8

CANX Regulates GBM Progression by Modulating BNIP3-Mediated Mitophagy. (A, B) A cluster heatmap and volcano plot of the PCR data showing significantly DE-ARGs between the shCANX and shNC groups. (C-G) PCR validation of the expression levels of these DEGs (IFNG, GAA, BNIP3, ARSA and MAP1LC3B) in the T98G cell line after the knockdown of CANX. (H-K) Correlation analysis of single-cell data was performed to determine the correlations between CANX expression and GAA, BNIP3, IFNG and ARSA expression. (L) Western blot showing the expression levels of LC3B, BNIP3 and CANX in shCANX-treated and shNC-treated T98G, GBM#P3 and LN229 cells overexpressing BNIP3. (M, N) Confocal microscopy images of GFP-LC3B and RFP-mito were used to reveal changes in mitophagy following CANX knockdown and BNIP3 overexpression in T98G and GBM#P3 cells (scale bar: 20 μm) . (O) EdU assay showing the effect of BNIP3 overexpression on shCANX-treated and shNC-treated T98G cells' proliferation (scale bar: 100 μm). (P) Flow cytometry analysis showing the impact of BNIP3 overexpression on apoptosis in T98G cells. (Q) The statistical analysis of the EdU assay with the treatment of BNIP3 overexpression on shCANX-treated and shNC-treated T98G cells. (R) Statistic results of the proportion of apoptosis and necrotic apoptosis T98G cells with oeBNIP3 and shCANX treatment. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 9

The MAPK Pathway Plays a Role in CANX-Mediated Regulation of Protective Mitophagy. (A) Volcano plot showing DEGs identified by whole-transcriptome sequencing of GBM#P3 cells in the shCANX group and shNC group. (B) KEGG enrichment analysis of these DEGs. (C) GO enrichment analysis of CANX-related DEGs according to data from the TCGA database. (D) Correlation analysis between MAPK and CANX expression. (E) Differences in MAPK expression between GBM cells with high and low CANX levels. (F) Western blot showing the protein levels of MEK, p-MEK, ERK, p-ERK, LC3B, BNIP3, CANX and GAPDH after treatment with PD98059 and C16-PAF in CANX-overexpressing U118-MG cells and shCANX-treated GBM#P3 cells. (G) Confocal microscopy images of GFP-LC3B and RFP-mito showing mitophagy changes after MAPK pathway agonist and inhibitor treatment in GBM#P3 cells (scale bar: 10 μm). (H) EdU assay demonstrating the impact of CANX and MAPK pathway alterations on GBM#P3 cells' proliferation (scale bar: 100 μm). (I) Flow cytometry analysis showing the impact of PD98059 and C16-PAF to assess the effects of changes in CANX expression on U118-MG and GBM#P3 cells' apoptosis level. (J) Statistic results of GFP-LC3 puncta with RFP-Mito per cell after PD98059 and C16-PAF treatment in GBM#P3 cells. (K) The statistical analysis of the EdU assay with the treatment of PD98059 and C16-PAF in GBM#P3 cells. (L) Statistic results of the proportion of apoptosis and necrotic apoptosis GBM#P3 cells with PD98059 and C16-PAF treatment. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 10

ND Inhibits CANX-Induced Protective Autophagy by Decreasing Intracellular Calcium Levels. (A) Fluorescence microscopy showing ER morphology and calcium ion distribution in GBM#P3 cells stained with Fluo-4 Ca²+ and ER-Tracker (scale bar: 80 μm). (B) Analysis of the calcium ion fluorescence intensity in GBM#P3 cells (shNC/shCANX group) treated with TMZ. (C) Quantitative analysis of the Ca²+ fluorescence intensity in GBM#P3 cells (treated with shNC/shCANX) before and after TMZ treatment via a microplate reader. (D) Molecular structure of ND. (E) IC50 curves of ND in the GBM#P3, T98G, and LN229 cell lines. (F) Western blot analysis showing the protein level of LC3B and CANX in GBM#P3, T98G, and LN229 cells after treatment with TMZ and ND. (G, H) EdU assay showing the proliferation of GBM#P3 and T98G cells after treatment with TMZ and ND with statistic results (scale bar: 40 μm). (I, J) Representative confocal images showing changes in GFP-labeled LC3 levels in GBM#P3 and T98G cells stained with MitoTracker Red after treatment with ND and TMZ with statistic results of GFP-LC3 puncta with RFP-Mito per cell (scale bar: 100 μm). (K, L) Flow cytometry analysis of apoptosis levels in GBM#P3 and T98G cells after treatment with TMZ and ND with statistic results. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.

Figure 11

In Vivo Experiments Confirm that ND Enhanced the Chemotherapeutic Effect of TMZ. (A) GBM#P3 cells expressing luciferase were implanted intracranially into immunosuppressed nude mice, followed by treatment with ND, TMZ, or a combination of both. Tumor growth was monitored via the IVIS-200 imaging system, and bioluminescence signals were measured on days 7, 14, 21, and 28 post-implantation. (B) Bioluminescence values for assessing tumor growth on day 28. (C) Overall survival was determined via Kaplan‒Meier survival curves, and statistical significance was assessed via the log-rank test. (D) HE staining showing the extent of tumor invasion in the brain. (E-H) Immunohistochemical staining images of MKi67, LC3B, CANX, and cleaved caspase 3 in tumors from each treatment group. (I) Study flowchart. The data are shown as the means ± SDs and are representative of three independent experiments. *P < 0.05; **P < 0.01, ***P < 0.001 between the two indicated treatments.