Sulforaphane induces cell morphology change and cell apoptosis by activating endoplasmic reticulum stress in glioblastoma

Abstract

Background: Sulforaphane (SFN), a naturally occurring isothiocyanate derived from cruciferous vegetables, has shown promise as a multitargeted therapeutic agent in glioblastoma (GBM). This study aimed to elucidate the role and underlying molecular mechanisms of SFN in regulating GBM progression, particularly through the endoplasmic reticulum stress (ERS) and unfolded protein response (UPR) pathways.

Methods: Primary human glioma cells and established GBM cell lines were treated with various concentrations of SFN. RNA sequencing and qPCR analyses were conducted to identify transcriptional changes associated with the UPR pathway. Western blot and immunofluorescence were used to assess the expression and subcellular localization of key ER stress-related proteins. A CHOP knockdown model was employed to examine the functional role of CHOP in SFN-induced apoptosis. Additionally, normal human astrocytes (HA) were used to evaluate the selectivity of SFN's cytotoxicity. In vivo validation was performed using an intracranial glioma xenograft mouse model.

Results: SFN significantly induced apoptotic cell death in GBM cells. Mechanistically, SFN activated multiple branches of the UPR, notably increasing the expression and nuclear translocation of ATF4 and CHOP. CHOP knockdown markedly attenuated SFN-induced apoptosis. RNA-seq and KEGG enrichment analyses confirmed the involvement of the ER stress pathway. Treatment with 4-phenylbutyrate (4-PBA) suppressed SFN-induced cytotoxicity, further supporting ER stress-mediated apoptosis. In vivo, SFN reduced tumor burden and upregulated ER stress markers in intracranial tumor tissues. Importantly, SFN had minimal cytotoxic effects on normal astrocytes, suggesting a favorable therapeutic window.

Conclusions: This study demonstrates that SFN induces GBM cell apoptosis via activation of the UPR pathway, particularly through the ATF4-CHOP axis. These findings support the potential of SFN as a promising therapeutic agent for glioblastoma.

Keywords: Cleaved Caspase-3; Endoplasmic reticulum stress; Glioblastoma; In vivo model; Sulforaphane; Unfolded protein response.

Fig. 1

SFN Inhibits Cell Viability and Promotes Cell Apoptosis in Glioma Cells. U87 and U251 cells were incubated with SFN at the indicated concentrations for 24 h (A) or 48 h (B). Cell viability was determined by CCK8 assay. Cell morphological changes following the indicated treatment for 24 h were examined and photographed, bar: 200 μm (C). TUNEL staining of U87 cells treated with SFN at the indicated concentrations for 24 h (D). Quantitative analysis of TUNEL-positive U87 cells in D (E). Statistical values are expressed as the mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 vs. the control group

Fig. 2

SFN Activates ER Stress Pathways in Glioma Cells. Workflow for RNA-sequencing in two groups (A). Volcano plot exhibiting 555 up-regulated and 1557 down-regulated SFN-related genes (B). KEGG analysis of different expression gene sets between control and SFN-treated groups (C). Representative immunoblots against ER stress-related proteins from U87 and U251 cells treated with SFN (20, 40, and 60 µM) for 24 h (D). Quantitative analysis of protein levels in D left (E) and D right (F). Values are expressed as the mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 vs. the control group

Fig. 3

Enhanced ER Stress Response with Increased Stimulation Duration of SFN in Glioma Cells. Representative immunoblots against ER stress-related proteins from U87 cells and U251 cells treated with SFN (20, 40, and 60 µM) for 48 h (A). Quantitative analysis of protein levels of GRP78 (B), p-eIf2α (C), ATF6 (D), xBP1s (E), ATF4 (F) and CHOP (G) in A left U87 cells. U87 (H) and U251 (I) cells were treated with SFN (20, 40, and 60 µM) for 48 h, and the mRNA levels of GRP78, eIf2α, xBP1s, ATF4, PERK and CHOP were assessed by qRT-PCR, with GAPDH serving as a reference gene. Data are presented as the mean ± SEM from three replicate experiments. *P < 0.05, **P < 0.01, ***P < 0.001 vs. the control group

Fig. 4

SFN Changes Morphology and Activates ER Stress Pathways in Glioma Cells in Time-dependent Manners. Cell morphological changes following the indicated concentration of 40 µM SFN treated for 6, 12, 24 and 48 h in U87 cells (A) and U251 (D) cells were examined and photographed. Bar: 100 μm. Representative immunoblots against ER stress-related proteins from U87 cells (B) and U251 cells (E) treated with SFN 40 µM for 6, 12, 24 and 48 h. Quantitative analysis of protein levels in B, E (C, F). Data are presented as the mean ± SEM from three separate experiments.*P < 0.05, **P < 0.01, ***P < 0.001 vs. control

Fig. 5

SFN Induces the Nuclear Translocation of CHOP and ATF4 in Glioma Cells. Representative immunofluorescence images depict nuclear translocation of CHOP (A) and ATF4 (B) in U87 cells following treatment with 40 µM SFN or the ER stress inducer tunicamycin (TM). Scale bar = 100 μm

Fig. 6

ER Stress Expression in Clinical Samples. U87 and U251 cells were incubated with SFN (40 µM) for 6, 12, 24 and 48 h. The CCK8 assay was used to detect cell viability (A). Immunohistochemistry for detecting GRP78 expression in glioma samples (B). Western Blot for detecting expression of ER-related markers GRP78, eIf2α, p-eIf2α, ATF6, xBP1s, ATF4 and CHOP in Samples (C). Primary glioma cells pretreated with the ER stress inhibitor 4-PBA (7 mM) for 4 h, followed by incubation with SFN (40 µM) or TM (2.5 µM) for 24 h. The CCK8 assay was used to detect cell viability (D). The expression levels of the ER stress-related proteins GRP78, ATF6, XBP1s, ATF4, CHOP, and cleaved-caspase3 were determined by WB (E). Data are presented as the mean ± SEM from three separate experiments. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control

Fig. 7

Overview of SFN’s therapeutic potential in glioblastoma