A New Nrf2 Inhibitor Enhances Chemotherapeutic Effects in Glioblastoma Cells Carrying p53 Mutations

Abstract

TP53 tumor suppressor gene is a commonly mutated gene in cancer. p53 mediated senescence is critical in preventing oncogenesis in normal cells. Since p53 is a transcription factor, mutations in its DNA binding domain result in the functional loss of p53-mediated cellular pathways. Similarly, nuclear factor erythroid 2-related factor 2 (Nrf2) is another transcription factor that maintains cellular homeostasis by regulating redox and detoxification mechanisms. In glioblastoma (GBM), Nrf2-mediated antioxidant activity is upregulated while p53-mediated senescence is lost, both rendering GBM cells resistant to treatment. To address this, we identified novel Nrf2 inhibitors from bioactive compounds using a molecular imaging biosensor-based screening approach. We further evaluated the identified compounds for their in vitro and in vivo chemotherapy enhancement capabilities in GBM cells carrying different p53 mutations. We thus identified an Nrf2 inhibitor that is effective in GBM cells carrying the p53 (R175H) mutation, a frequent clinically observed hotspot structural mutation responsible for chemotherapeutic resistance in GBM. Combining this drug with low-dose chemotherapies can potentially reduce their toxicity and increase their efficacy by transiently suppressing Nrf2-mediated detoxification function in GBM cells carrying this important p53 missense mutation.

Keywords: Nrf2; chemotherapy; glioblastoma; p53; small molecule compounds.

Figure 1

Schematic of study workflow. (A) A molecular imaging biosensor identifies Nrf2 inhibitors and activators from a set of small molecule compounds. A2780 cells were transduced by a construct that carries a Firefly luciferase reporter gene under an Nrf2 responsive target gene promoter, the NQO1. Upon activation of Nrf2, the FLuc signal is enhanced, while inhibition reduces the FLuc signal intensity; (B) evaluation of Nrf2-mediated mechanism in p53 wild-type and mutant cell lines in response to Nrf2 modulators. LN308 cell line with null-p53 were transduced with five different constructs containing mutant p53 and GFP. In each construct, the mutation was placed in the hotspot structural position in GBM. These cell lines as well as U87-MG with wild-type p53 were then evaluated for apoptotic populations and growth inhibition when co-treated with Nrf2 inhibitor compounds and chemotherapeutic drugs.

Figure 2

Evaluation of Nrf2 pathway after treatment with six different compounds (CET-CH-1 to CET-CH-6) tested on stably biosensor A2780 ovarian cancer cells. (A) Bioluminescence imaging of Firefly Luciferase (FLuc) showing the alteration in Nrf2 pathway in A2780 after treatment with compounds CET-CH-1 to CET-CH-6 at concentrations of 10, 20 and 30 μM, with (B) a quantitative plot in which samples are tested for differential expression of FLuc signal in comparison to DMSO as a solvent control sample used for the study. *, **, ***, **** represent p-value < 0.05, <0.01, <0.001, <0.0001, respectively. (C) FLuc imaging of A2780 cells after treatment with CET-CH-1 to CET-CH-6 at concentrations of 1.25, 2.5, 5, and 10 μM, with (D) quantitative plot. In each respective experiment, DMSO and TBHQ images show the minimum (background) and maximum (ROS) bioluminescence signal, respectively. Experiments were performed in triplicate wells. Error bars represent mean ± SD.

Figure 3

Mechanism of inhibition of Nrf2 in response to treatment with CET-CH-6 in combination with TMZ. (A) Confocal microscopy of MDA-MB-231 cells treated with CET-CH-6 and stained with Nrf2 and Keap1 (zoom = 10×). Zoomed in images of the white boxes are shown below each image; (B) quantitative analyses of immunofluorescence images shown in (A), signals are normalized to the total amount of proteins expressed; (C) Western blotting images of MDA-MB-231 cells treated with CET-CH-6, TMZ or their combination for 24 h with quantitative plot of (D) Keap1 and (E) Nrf2. Protein expressions measured in cytoplasm (Cy) and nucleus (Nu) are normalized to GAPDH and Lamin B1, respectively. All the Western blot images are shown in Figure S2 in the Supplementary Material. In (D,E) the plots are normalized to control Cy and control Nu, respectively. Error bars represent mean ± SD. * represent p-value < 0.05.

Figure 4

LN308 cell line with no p53 expression engineered to stably express structural p53 variants at different locations. (A) the vector designed for the transduction of LN308 cell line to stably express p53 wild-type as well as five different structural mutations in the frequently mutated sites at the DNA binding regions at amino acid residue 175, 220, 245, and 282. These six cell lines were used throughout the study to test the Nrf2 inhibitors and activators; (B) confocal microscopy of different clones of LN308 cell lines that stably express green fluorescence protein (GFP) after being transduced with a lentivirus vector constructed for expression of p53 and GFP; (C) flow cytometry analysis of different LN308 cell lines stably expressing p53 mutations and GFP; (D) Western blotting of LN308 cell lines stably expressing different p53 variants; (E) changes in Keap1 expression after treatment of different LN308 cell lines with CET-CH-6 (2.5 μM), with (F) quantitative plot normalized to GAPDH and control of each clone. Error bars represent mean ± SD. C: control and C6: CET-CH-6. For whole Western blot membranes, please refer to Figure S2. Keap1 in each cell line is compared to the control signal, and * represents p-value < 0.05.

Figure 5

U87-MG cells with wt-p53 expression tested for therapeutic outcome in response to the treatment of Nrf2 activators and inhibitors in the presence and absence of DOX. (A) fixed cell population assessed by PI staining based FACS after treatment of U87-MG-p53^wt by DOX (0.05 μM) in the presence of two Nrf2 activators (CET-CH-1 and CET-CH-2) and two Nrf2 inhibitors (CET-CH-5 and CET-CH-6), the gating results are shown in Figure S5; (B) fixed cell population assessed by PI staining based FACS after treatment of LN308 cells stably expressing different p53 mutations with DOX (0.05 μM) in the presence of CET-CH-6 (concentrations in μM), the gating results are shown in Figure S6. Experiments were performed three times. Cells with fragmented DNA, late apoptotic population, show diffused PI staining compared to non-apoptotic cells. Error bars represent mean ± SD. *, **, ***, **** markers above the columns represent p-value < 0.05, <0.01, <0.001, <0.0001, respectively. Each sample is compared to non-treated control sample of the same cell line.

Figure 6

MTT assay result shows that co-treatment of CET-CH-6 with DOX and TMZ inhibits cell growth in LN308 cell lines with different p53 variants. Differential therapeutic responses of LN308 cells stably expressing different p53 variants as assessed using MTT in response to co-treatment with (A) CET-CH-6 and DOX, and (B) CET-CH-6 and TMZ. Each sample was normalized to the untreated control. Experiments were performed three times, and error bars represent mean ± SD. *, **, ***, **** represent p-value < 0.05, <0.01, <0.001, <0.0001, respectively. Samples were assayed 72 h post treatment. Each sample was tested against the untreated control.

Figure 7

Genes in downstream pathway of p53 and Nrf2 are altered in response to treatment with CET-CH-6 and DOX. Treatment of LN308 cells stably expressing p53 variants (p53^wt, p53¹⁷⁵, p53²⁴⁵, p53²⁸²) activated expression of p53 responsive pathway proteins and the Nrf2 pathway. For representative whole Western blot membranes, please refer to Figure S10.

Figure 8

In vivo evaluation of CET-CH-6 in combination with TMZ on NSG mice bearing LN308-p53¹⁷⁵, LN308-p53^wt, and U87-MG cells. (A) Schematic workflow of in vivo treatment. Mice (N = 4–6/group) were implanted subcutaneously into the left and right flank, respectively. When the tumors reached 3 mm/diameter, the mice were randomized into four groups: control, TMZ (12.5 mg/kg body weight), CET-CH-6 (2.5 mg/kg body weight), CET-CH-6 + TMZ (co-treated with 2.5 and 12.5 mg/kg body weight of CET-CH-6 and TMZ, respectively). The mice received i.p. doses of drugs every four days for ten cycles; (B) survival curve of mice with LN308-p53¹⁷⁵ tumors. Tumor volume measurements for each cell lines over time: (C) LN308-p53¹⁷⁵; (D) U87-MG; (E) LN308-p53^wt; (F) TUNEL assay to show the apoptotic population in tumor tissue; and (G) quantification of tunnel assay, each row shows one cell line. Error bars represent mean ± SD. * markers above the columns represent p-value < 0.05.