CSRP2 promotes the glioblastoma mesenchymal phenotype via p130Cas-mediated NF-κB and MAPK pathways

Abstract

Background: Cysteine-rich protein 2 (CSRP2) plays a role in a variety of biological processes including cell proliferation and differentiation. However, whether and how CSRP2 participates in the malignancy of glioblastoma multiforme (GBM), including its proneural-to-mesenchymal transition (PMT), remains unclear.

Methods: CSRP2 expression in low-grade and high-grade gliomas was analyzed, and survival analyses were performed in patients with gliomas with high and low CSRP2 expression in various tumor databases. Quantitative real-time PCR (qRT-PCR) and western blotting (WB) were used to detect the expression of CSRP2 in GBM and control brain tissues. CSRP2 function in GBM was determined by a series of functional tests in vitro and in vivo. WB, co-immunoprecipitation (co-IP) and immunofluorescence were used to determine the relation between CSRP2 and p130Cas. Mechanisms of CSRP2 involvement in GBM progression were analyzed with gene set enrichment analysis and KEGG enrichment analysis in available databases. WB was used to determine the relation between CSRP2 and PMT markers, NF-κB and MAPK signaling-related proteins, and apoptosis-related proteins. Microscale thermophoresis assay was used to analyze whether mitoxantrone (MTO) and CSRP2 could bind. MTO function was determined by a series of functional tests in vitro, while the relation between MTO and PMT markers, NF-κB and MAPK signaling-related proteins, and apoptosis-related proteins was analyzed by WB in GBM cell lines stably overexpressing CSRP2.

Results: We found that CSRP2 expression significantly increased in GBM, especially mesenchymal GBM, and that glioma patients with high CSRP2 expression possibly had poor prognosis. CSRP2 overexpression in GBM cells promoted proliferation, colony formation, migration, invasion, temozolomide resistance, and PMT in vitro and tumor formation in vivo. While knockdown of CSRP2 had the opposite effects. Mechanistically, we revealed that CSRP2 interacted with p130Cas, thereby regulating the NF-κB and the MAPK signaling pathways. CSRP2 overexpression and knockdown increased and decreased p130Cas levels and NF-κB and MAPK activities, respectively. Both p130Cas downregulation and NF-κB inhibition reversed the elevated PMT and NF-κB and MAPK activities resulted from CSRP2 overexpression. Finally, we identified that MTO bound CSRP2 and inhibited the malignant effects of CSRP2 overexpression on GBM cells.

Conclusions: Our findings demonstrate that CSRP2 promotes GBM malignancy including PMT and temozolomide resistance through activating p130Cas-mediated NF-κB and MAPK signaling pathways. Inhibiting CSRP2 function, including using MTO, may become a novel therapeutic approach for GBM.

Keywords: Cysteine-rich protein 2; Glioblastoma multiforme; Mitoxantrone; NF-κB; p130Cas.

Fig. 1

CSRP2 is upregulated in GBM and correlated with advanced progression and poor prognosis of GBM. (A) Comparison of CSRP2 expression between GBM and normal controls and between LGG and normal controls, using data from the GEPIA database. (B) Comparison of CSRP2 expression between GBM proneural and mesenchymal subtypes using the CGGA database. Unpaired t test, n = 157 for proneural and n = 116 for mesenchymal. (C) Comparison of CSRP2 expression between GBM proneural and mesenchymal subtypes using the TCGA database. Unpaired t test, n = 138 for proneural and n = 157 for mesenchymal. (D) Comparison of CSRP2 expression between different chemotherapy status (0 and 1) of GBM using the CGGA database. Unpaired t test, n = 275 for chemotherapy status 0 and n = 635 for chemotherapy status 1. (E, F) Kaplan‒Meier survival analysis of patients with low or high CSRP2 expression using data from the Gravendeel database (E) and the GEPIA database (F). Log-rank test. (G) Receiver Operating Characteristic curve (ROC curve) of CSRP2 predicting survival or death in GBM patients using the TCGA database. (H, I) Equal amounts of cell lysates of NHA and GBM cell lines (U87-MG and U251) were subjected to western blotting (H) and quantitative comparison (I) for CSRP2. GAPDH was used as an internal loading control. One-way ANOVA with Tukey’s post hoc test. n = 4 per group. (J) CSRP2 mRNA levels in 4 non-tumor brain biopsies from traumatic brain injury or epilepsy patients (N1-N4) and 4 GBM samples (grades IV) were compared. Unpaired t test. (K, L) Equal amounts of protein lysates from the 4 GBM samples and the 4 non-tumor controls were subjected to western blotting (K) and quantitative comparison (L) for CSRP2. Unpaired t test. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 2

CSRP2 overexpression promotes the malignancy of GBM cells. (A) U87-MG and U251 cells were stably transduced with lentiviruses expressing CSRP2 or control (NC) and the mRNA levels of CSRP2 were compared. Unpaired t test, n = 3 per group. (B, C) CSRP2 protein levels in U87-MG and U251 cells stably overexpressing CSRP2 and controls were analyzed by western blotting (B) and densitometry quantification for comparison (C). Unpaired t test, n = 3 per group. (D, E) U87-MG and U251 cells with stable CSRP2 overexpression and controls were assayed for colony formation (D) and the colony numbers were compared (E). Unpaired t test, n = 6 per group. (F, G) U87-MG (F) and U251 (G) cells with stable CSRP2 overexpression and controls were assayed for their proliferation. Two-way ANOVA with Sidak’s post hoc test, n = 6 per group for U87-MG and U251. (H-K) Cells with stable CSRP2 overexpression and controls were studied for their migration ability (H, I) and invasion ability (J, K). Unpaired t test, n = 3 per group. (L, M) U87-MG cells with stable CSRP2 expression and controls were xenografted into the mouse brain. Tumor formation was observed by MRI brain images (L) and tumor weights were compared (M). Red cycles indicate tumor areas. Unpaired t test, n = 6 per group. (N) Kaplan-Meier survival curves of xenografted mice. Log-rank test, n = 6 per group. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 3

CSRP2 knockdown reduces the malignancy of GBM cells. (A) U87-MG and U251 cells were stably transduced with two CSRP2 shRNA lentiviruses (shCSRP2-1 and shCSRP2-2) or control shRNA lentiviruses (shNC), and the mRNA levels of CSRP2 were compared. One-way ANOVA with Tukey’s post hoc test, n = 3 per group. (B, C) CSRP2 protein levels in U87-MG and U251 cells stably transduced with CSRP2 shRNA and control shRNA lentiviruses were analyzed by western blotting (B) and densitometry quantification for comparison (C). One-way ANOVA with Tukey’s post hoc test, n = 3 per group. (D, E) U87-MG and U251 cells with CSRP2 knockdown and controls were assayed for colony formation (D) and the colony numbers were compared (E). One-way ANOVA with Tukey’s post hoc test, n = 6 per group. (F, G) U87-MG (F) and U251 (G) cells with CSRP2 knockdown and controls were assayed for their proliferation. Two-way ANOVA with Sidak’s post hoc test, n = 6 per group for U87-MG and U251. (H-K) U87-MG and U251 cells with CSRP2 knockdown were studied for their migration ability (H, I) and invasion ability (J, K). One-way ANOVA with Tukey’s post hoc test, n = 3 per group. (L, M) U87-MG cells with CSRP2 knockdown (shCSRP2) and controls (shNC) were xenografted into the mouse brain. Tumor formation was observed by MRI brain images (L) and tumor weights were compared (M). Red cycles indicate tumor areas. Unpaired t test, n = 6 per group. (N) Kaplan-Meier survival curves of xenografted mice. Log-rank test, n = 6 per group. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 4

CSRP2 promotes PMT and the NF-κB and the MAPK signaling activities in GBM. (A) GSEA analysis of the correlation between CSRP2 expression and the NF-κB signaling pathway based on the CGGA data. (B) KEGG enrichment analysis of the correlation between CSRP2 expression and the Ras/MAPK signaling pathway based on the TCGA data. (C-E) Levels of PMT markers and NF-κB and MAPK signaling-related proteins in U87-MG and U251 cells with stable CSRP2 overexpression (CSRP2) and controls were analyzed by western blotting (C) and densitometry quantification comparison (D, E). Unpaired t test, n = 3 per group. (F-H) Levels of PMT markers and NF-κB and MAPK signaling-related proteins in U87-MG and U251 cells with stable CSRP2 knockdown and controls were analyzed by western blotting (F) and densitometry quantification comparison (G, H). One-way ANOVA with Tukey’s post hoc test, n = 3 per group. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant

Fig. 5

Inhibition of the NF-κB activity reduces the malignancy of GBM cells with CSRP2 overexpression and suppresses PMT and the MAPK activity. (A, B) U87-MG (A) and U251 (B) cells with stable CSRP2 expression (CSRP2) were treated with or without the NF-κB inhibitor JSH 23 (80 µM) for indicated time periods and cell proliferation was analyzed. Cells transfected with control vector (NC) were used as a negative control. Two-way ANOVA with Sidak’s post hoc test, n = 6 per group for U87-MG and U251. (C, D) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 80 µM JSH 23 and analyzed for their colony formation ability (C) and colony numbers (D). Unpaired t test, n = 6 per group. (E-H) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 80 µM JSH 23 and studied for their migration (E, F) and invasion (G, H) abilities. Unpaired t test, n = 4 per group. (I, J) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 80 µM JSH 23 for 48 h and cell apoptosis was measured by flow cytometry (I) for apoptosis ratio comparison (J). Unpaired t test, n = 4 per group. (K-M) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 80 µM JSH 23 for 48 h and equal amounts of cell lysates were subjected to western blotting (K) and quantification analysis (L, M) for PMT markers, NF-κB and MAPK signaling-related proteins, and apoptosis-related proteins. Unpaired t test, n = 3 per group. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 6

p130Cas interacts with CSRP2 and mediates CSRP2 functions. (A) U87-MG and U251 cells were immunostained with antibodies against CSRP2 (in green) and p130Cas (in red). The Nuclei were stained by DAPI (in blue). (B, C) Equal amounts of U87-MG and U251 cell lysates (B) or GBM tissue lysates (C) were immunoprecipitated (IP) with an anti-CSRP2 antibody and IgG and then immunoblotted (WB) with anti-CSRP2 and anti-p130Cas antibodies. (D) HEK 293T cells were co-transfected with CSRP2-Flag and p130Cas-HA vectors. Equal amounts of cell lysates were immunoprecipitated with anti-Flag and anti-HA antibodies and IgG and then immunoblotted with anti-Flag and anti-HA antibodies. (E-H) The p130Cas protein levels in U87-MG and U251 cells with CSRP2 overexpression (E, F) or CSRP2 knockdown (G, H) were analyzed by western blotting (E, G) and quantified for comparison (F, H). Unpaired t test for (F) and One-way ANOVA with Tukey’s post hoc test for (H), n = 3 per group. (I-K) U87-MG and U251 cells with stable CSRP2 overexpression and controls (NC) were transfected with p130Cas shRNA (shp130Cas) or control shRNA (shNC) and equal amounts of protein lysates were subjected to western blotting (I) and quantification comparison (J, K) for PMT-, apoptosis-, and NF-κB and MAPK signaling-related proteins. Unpaired t test, n = 4 per group. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant

Fig. 7

CSRP2 promotes glioma cell resistance to TMZ. (A-D) U87-MG (A, C) and U251 (B, D) cells with CSRP2 overexpression (A, B) or with CSRP2 knockdown (C, D) and respective control cells were treated with indicated amounts of TMZ for 48 h, and then analyzed for cell proliferation. Two-way ANOVA with Sidak’s post hoc test, n = 5 per group. (E-H) U87-MG (E, F) and U251 (G, H) cells with stable CSRP2 overexpression and control cells were treated with 200 µM TMZ for 48 h. Cell apoptosis was measured by flow cytometry (E, G) for apoptosis ratio comparison (F, H). Unpaired t test, n = 4 per group. (I-L) U87-MG (I, J) and U251 (K, L) cells with stable CSRP2 knockdown and control cells were treated with 200 µM TMZ for 48 h. Cell apoptosis was measured by flow cytometry (I, K) for apoptosis ratio comparison (J, L). One-way ANOVA with Tukey’s post hoc test, n = 4 per group. Data represent mean ± SEM, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant

Fig. 8

Mitoxantrone binds CSRP2 and inhibits the malignant function of CSRP2 in GBM. (A) Fluorescence-labeled CSRP2 recombinant protein was incubated with different concentrations of mitoxantrone (MTO) and measured for their binding using the microscale thermophoresis assay. (B, C) U87-MG (B) and U251 (C) cells with stable CSRP2 overexpression (CSRP2) were treated with or without 0.5 µM MTO for indicated time periods and the cell proliferation was analyzed. Cells transfected with control vector (NC) were used as a negative control. Two-way ANOVA with Sidak’s post hoc test, n = 6 per group for U87-MG and U251. (D, E) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 0.5 µM MTO and analyzed for their colony formation ability (D) and colony numbers (E). Unpaired t test, n = 6 per group. (F-I) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 0.5 µM MTO and studied for their migration (F, G) and invasion (H, I) abilities. Unpaired t test, n = 4 per group. (J, K) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 0.5 µM MTO for 48 h and cell apoptosis was measured by flow cytometry (J) for apoptosis ratio comparison (K). Unpaired t test, n = 4 per group. (L) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 0.5 µM MTO for 48 h. Equal amounts of cell lysates were immunoprecipitated (IP) with anti-CSRP2 and anti-p130Cas antibodies and IgG and then immunoblotted with anti-CSRP2 and anti-p130Cas antibodies. (M-O) U87-MG and U251 cells with stable CSRP2 overexpression were treated with 0.5 µM MTO for 48 h and equal amounts of cell lysates were subjected to western blotting (M) and quantification analysis (N, O) for PMT markers, NF-κB and MAPK signaling-related proteins, and apoptosis-related proteins. Unpaired t test, n = 3 per group. Data represent mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001