NRF2 and glutathione are key resistance mediators to temozolomide in glioma and melanoma cells

Abstract

Cancer is a leading cause of death worldwide, and while great advances have been made particularly in chemotherapy, many types of cancer still present a dismal prognosis. In the case of glioma, temozolomide (TMZ) is the main option for treatment, but it has limited success due to drug resistance. While this resistance is usually associated to DNA repair mechanisms, in this work we demonstrate that oxidative stress plays an important role. We showed that upon TMZ treatment there is an induction of the nuclear factor erythroid 2-related factor 2 (NRF2), which is the main antioxidant transcription factor regulator in human cells. This is accompanied by an enhancement of glutathione (GSH) concentration in the tumor cells. The effectiveness of this pathway was proven by silencing NFR2, which greatly enhanced cell death upon TMZ treatment both in vitro and in vivo. Also, higher DNA damage and induced cell death was observed by combining BSO - a GSH inhibitor - with TMZ. Similar effects were also observed using in vitro and in vivo models of melanoma, thus possibly indicating that GSH has a decisive role in TMZ resistance in a wider range of tumors. Thus, a combined regimen of BSO and TMZ configures an interesting therapeutic alternative for fighting both glioma and melanoma.

Keywords: NRF2; glioma; melanoma; resistance; temozolomide.

Figure 1. Expression of NRF2 and its target genes in glioma cell lines

A-B. Representative image and quantification of NFR2, GCLM and GSTπ mRNA in U138MG and U87MG at basal level or 4 h after TMZ treatment (300 μM); C. NRF2 detection in glioma cells untreated or treated with 300 μM TMZ. Samples were collected 24 h after TMZ treatment and analyzed by western blot; D. Quantification of NRF2 protein expression in U87MG and U138MG submitted or not to TMZ treatment. Data were normalized by GAPDH expression followed by normalization by NRF2/GAPDH ratio verified on untreated U87MG cell line. Values are mean ± SEM of three independent experiments, *P< 0.05, **P< 0.01, ***P< 0.001.

Figure 2. Consequences of oxidative stress induction after TMZ treatment

A. Intracellular GSH quantification in glioma cells treated with TMZ (300 μM) for 24 h; B. Glioma cells were pre-incubated with BSO (100 μM) or NAC (1 mM) for 16 h and then treated with TMZ (300 μM) for 2 h. ROS was detected by DCFDA probe, and analyzed by flow cytometry; C. Quantification of mitochondrial O₂⁻ generation using MitoSOX Redin glioma cells treated withTMZ (300 μM) for 2 h; D-E. Quantification and representative image of alkaline comet assay of U138MG glioma cells treated with TMZ alone (300 μM for 2 h) or in combination with BSO (100 μM, pre-incubated for 16 h). Quantification was done by measuring tail length of cells nuclei incubated or not with FPG endonuclease. Values are mean ± SEM of three independent experiments, *P< 0.05, **P< 0.01, ***P< 0.001.

Figure 3. Cellular response of NRF2 silenced cells to TMZ treatment

A. NRF2 detection by western blot in U138MG cells transduced with shCTRL or shNRF2 lentivirus; B. A dose-response curve of U138MG shCTRL or U138MG shNRF2 cell lines treated with increasing concentrations of TMZ (10 to 500 μM) and analyzed 72 h after drug treatment measured by XTT assay; C-D. Representative histogram and quantification of sub-G1 population of glioma cells treated with TMZ (100 μM) for 72 h, respectively; E-F. Flow cytometry analysis of percentage of active caspase-3 or γH2AX positive staining in cells NRF2 silenced or transduced with shCTRL upon treatment with TMZ (100 μM) for 72 h, respectively. Values are mean ± SEM of three independent experiments, *P< 0.05, **P< 0.01, ***P< 0.001.

Figure 4. In vivo response of NRF2 silenced cells to TMZ treatment

A. Representative bioluminescent image of shCTRL or shNRF2 expressing luciferase cells on day 10 after beginning treatment with TMZ (30 mg/kg); B. Ex vivo shCTRL or shNRF2 tumor at day 10 after initial TMZ treatment; C. Time-course of shCTRL or shNRF2 tumor volume progression, as determined by caliper measurement; D. Quantification of GSH concentration on shCRTL or shNRF2 tumors. Values are mean ± SEM; 5 animals were used per group.

Figure 5. In vitro and in vivo response of melanoma cells to treatment with TMZ in combination with GSH modulators

A-C. Dose response curve of murine and 2 human melanoma cell lines, respectively, to treatment with TMZ alone or in combination with BSO or EZA. Importantly, cell viability, for any of the cell lines, was not affected by BSO (blue line) or EZA (red line) in the absence of TMZ. Cell viability was measured 72 h after drug treatment by XTT assay; D. Time-course of B16Luc tumor volume progression, as determined by caliper measurement; E. Representative image of C57Bl/6 mice bearing B16Luc tumor at day 9 after treatment with TMZ (30 mg/kg) and/or BSO (450 mg/kg). Values are mean ± SEM ; 5 animals were used per group.

Figure 6. Proposed model of NRF2 role on TMZ resistance

TMZ induces genomic and mitochondrial DNA methylation damage. Mitochondrial DNA damage could lead to malfunction of this organelle, increasing its ROS production, which in turn activates NRF2. This transcription factor induces expression of genes related to GSH synthesis and utilization. GSH could act as an antioxidant (neutralizing ROS induced upon TMZ treatment) or detoxification agent (by GSH binding to TMZ through GST activity, eliminating TMZ inside the cells). Increased NRF2 activity, leading to higher GSH levels, would be a key resistance mechanism to TMZ. Thus, we propose the use of the GSH inhibitor, BSO, in combination with TMZ to circumvent resistance to this drug in glioma and melanoma tumor.