Suppression of glioblastoma by targeting the overactivated protein neddylation pathway

Abstract

Background: The neddylation pathway has been recently identified as an attractive anticancer target, and MLN4924, a specific NEDD8-activating enzyme inhibitor, has been developed as a first-in-class anticancer agent. However, neither the status of the neddylation pathway in glioblastoma (GBM) nor the effect of MLN4924 against GBM has been systematically investigated yet.

Methods: To measure the activation state of the neddylation pathway in GBM, expression of the NEDD8-activating enzyme (E1), NEDD8-conjugating enzyme (E2), and global protein neddylation in GBM tumor tissues versus adjacent tissues were examined by immunoblotting analysis and immunohistochemistry staining. To assess the therapeutic efficacy of neddylation inhibition in GBM, cell proliferation in vitro and tumor growth in vivo were determined upon neddylation inhibition by MLN4924, an investigational NEDD8-activating enzyme inhibitor.

Results: The neddylation pathway was overactivated in a majority of GBM tumor tissues when compared with adjacent normal tissues. The upregulation of this pathway in GBM tissues was positively correlated with high-grade disease and postoperative recurrence but was negatively associated with patient overall survival. MLN4924 treatment inhibited cullin neddylation, inactivated cullin-RING E3 ligase, and led to the accumulation of tumor-suppressive cullin-RING E3 ligase substrates to trigger cell-cycle arrest and senescence or apoptosis in a cell-line dependent manner. Moreover, inhibition of neddylation by MLN4924 significantly suppressed tumor growth in an orthotopic xenograft model of human GBM.

Conclusion: Our study indicates that an overactivated neddylation pathway may be involved in GBM progression and that inhibition of this oncogenic pathway is a potentially new therapeutic approach for GBM.

Keywords: MLN4924; NEDD8; cullin-RING E3 ligase; glioblastoma; neddylation.

Fig. 1.

The neddylation pathway is overactivated and correlated with the progression of gliomas. (A) Immunoblotting (IB) analysis to determine the expression of NAE1, UBA3, and UBC12 in glioma tissues and adjacent tissues. Quantification relative to GAPDH by densitometric analysis using Image J software. Representative results from 4 of 18 pairs of tissues are shown. (Abbreviations: P, patient; T, tumor tissues; A, adjacent tissues.) (B) Immunohistochemical staining of different grade human glioma tissues using NEDD8-specific antibodies. (C) Immunoreactive score (IRS) of NEDD8 in different grade gliomas. High-grade gliomas (WHO grades III and IV) have a higher expression of NEDD8 than low-grade gliomas (WHO grades I and II). P < .0001. (D) IRS of NEDD8 in paired GBM samples. NEDD8 expression in recurrent GBM was significantly higher than in primary GBM. P = .0043. (E) Kaplan–Meier curves for overall survival (OS) rate of patients with GBM according to the expression of NEDD8. Log-rank P = .0273.

Fig. 2.

MLN4924 specifically inhibits neddylation and suppresses the growth of GBM cells. (A) Specificity of the NAE inhibitor MLN4924 in inhibiting the neddylation pathway and suppressing CRL activity when compared with pan-proteasome inhibitors MG132 and bortezomib. U251 and A172 cells were treated with DMSO, MLN4924 (1 μM), MG132 (20 μM), and bortezomib (1 μM) for 1hour, followed by immunoblot analysis to determine the change of global neddylation (top panel) and cullin1 neddylation with β-actin as a loading control (bottom panel). (B and C) U251 and A172 cells seeded into 96-well plates were cultured overnight and treated with MLN4924 at various concentrations for 72 hours to determine its therapeutic efficacy on cell proliferation followed by ATPlite and CCK8 cell viability assays, respectively (n = 3) (B). U251 and A172 cells seeded into 24-well plates were cultured overnight and treated with DMSO or MLN4924 (0.3 μM) for the indicated time points, followed by cell counting (n = 3) (C). (D and E) U251 and A172 cells were seeded into 6-well plates in duplicate and then grown in the presence or absence of MLN4924 for 12 days. The colonies with more than 50 cells were counted, following crystal violet staining (D). Transwell migration assay was performed to analyze the efficacy of MLN4924 on cell migration as described in the Methods section (E). *, P < .05; **, P < .01; ***, P < .001.

Fig. 3.

MLN4924 induces the accumulation of CRL substrates and induces G2 cell-cycle arrest in GBM cells. (A) MLN4924 inhibited cullin neddylation and induced the accumulation of CRL substrates. U251 and A172 cells were treated with MLN4924 at different concentrations for 48 hours and subjected to immunoblot (IB) analysis using antibodies against p21, p27, p-IκB-α, t-IκB-α, CDT1, ORC1, p-H2AX, t-H2AX, WEE1, p-H3, and cullin1 with β-actin as a loading control. (B) MLN4924 induced G₂ cell-cycle arrest in glioma cells. U251 and A172 cells were treated with DMSO and MLN4924 (0.3 μM) for 48 hours and subjected to propidium iodide staining and fluorescence-activated cell sorting analysis to determine cell-cycle profile. (C) U251 and A172 cells were treated with DMSO and MLN4924 at different concentrations for 48 hours and subjected to IB analysis to determine the induction of WEE1 and p-H3 using β-actin as a loading control.

Fig. 4.

MLN4924 induces senescence or apoptosis in a cell line-dependent manner. (A–C) U251 and A172 cells were treated with DMSO and MLN4924 (0.3 μM) for 96 hours and then photographed (A top panel and (B) or subjected to the expression analysis of senescence-associated β–galactosidase by β–Gal staining (A, bottom panel), or collected, stained with AnnexinV and propidium iodide (C), followed by flow cytometric analysis. (D) U251 and A172 cells were treated with DMSO and MLN4924 at different concentrations for 48 hours and subjected to immunoblot analysis to determine the induction of cleaved caspase 3 and cleaved PARP.

Fig. 5.

MLN4924 suppresses the growth of orthotopic xenograft of human glioma. (A and B) MLN4924 inhibited tumor growth measured by fluorescence imaging system. Nude mice bearing glioma xenografts with U251-RFP cells were administered MLN4924 at 60 mg/kg s.c. twice daily on a 7-days-on/2-days-off schedule for 4 cycles. Tumor size was monitored twice a week with fluorescence imaging system. Five out of 10 pictures are shown (A). Abbreviation: D, day). The data were converted to tumor growth curves by ModFit LT software, and the duration of treatment was visualized with arrows in Fig. 5B. *, P < .05; **, P < .01. (C) MLN4924 significantly reduced tumor volume. Mice were euthanized at day 35 after treatment (the end of study, n = 10). Tumor tissues of mice were collected, photographed, and weighed. P <.01. (D) ki67 and p21 staining of the glioma sections. (E) Immunoblotting analysis to determine the expression of NAE1, UBA3, UBC12 and the global protein neddylation in the treated and control tumors. (F) A working model. Inhibition of overactivated neddylation pathway with the NAE inhibitor MLN4924 blocks protein neddylation (especially cullin neddylation), inactivates CRL, results in the accumulation of tumor-suppressive CRL substrates, and induces cell-cycle arrest, senescence, or apoptosis to inhibit tumor cell growth.