Activation of Nrf2 Signaling Augments Vesicular Stomatitis Virus Oncolysis via Autophagy-Driven Suppression of Antiviral Immunity

Abstract

Oncolytic viruses (OVs) offer a promising therapeutic approach to treat multiple types of cancer. In this study, we show that the manipulation of the antioxidant network via transcription factor Nrf2 augments vesicular stomatitis virus Δ51 (VSVΔ51) replication and sensitizes cancer cells to viral oncolysis. Activation of Nrf2 signaling by the antioxidant compound sulforaphane (SFN) leads to enhanced VSVΔ51 spread in OV-resistant cancer cells and improves the therapeutic outcome in different murine syngeneic and xenograft tumor models. Chemoresistant A549 lung cancer cells that display constitutive dominant hyperactivation of Nrf2 signaling are particularly vulnerable to VSVΔ51 oncolysis. Mechanistically, enhanced Nrf2 signaling stimulated viral replication in cancer cells and disrupted the type I IFN response via increased autophagy. This study reveals a previously unappreciated role for Nrf2 in the regulation of autophagy and the innate antiviral response that complements the therapeutic potential of VSV-directed oncolysis against multiple types of OV-resistant or chemoresistant cancer.

Keywords: Nrf2; VSV; autophagy; cancer; innate antiviral response; interferon; oncolysis.

Graphical abstract

Figure 1

Sulforaphane Enhances VSVΔ51-Mediated Oncolytic Activity in Resistant PC-3 Cells PC-3 cells were pretreated for 24 hr with increasing concentrations of sulforaphane (SFN) and were subsequently infected with VSVΔ51-GFP (MOI 1). (A and B) Infectivity was determined by fluorescent microscopy (A) or quantified by flow cytometry (B) at the indicated times. (C) Viral replication was assessed by plaque assay. (D) Cell survival was imaged by light microscopy 72 hr following VSVΔ51 challenge. (E and F) The percentage of viable (E) and apoptotic cells (F) was assessed using an annexin-V/7AAD staining by flow cytometry at the indicated times. (A–F) Data are the means ± SEM from two experiments performed in triplicate. (G) The correlation between the percentage of VSVΔ51-GFP⁺ cells at 24 hr and the percentage of apoptotic annexin-V⁺ cells at 48 hr was calculated in PC-3 cells using a Spearman test (n = 33). (H) The same statistical test was used to calculate the correlation between the percentage of apoptotic annexin-V⁺ cells and the percentage of MitoSOX⁺ cells at 48 hr after infection (n = 36). (I) PC-3 cells were pretreated for 24 hr with SFN (20 μM) and challenged with VSVΔ51-GFP (MOI 1) for 48 hr. Whole-cell extracts (WCEs) were analyzed by immunoblotting for cleaved caspase-3, cleaved caspase-7 (high and low exposure), cleaved PARP, VSVΔ51-GFP, and β-actin. Results are from a representative experiment; all immunoblots are from the same samples.

Figure 2

VSVΔ51 and Sulforaphane Combinatorial Treatment Reduces Tumor Progression and Prolongs Survival in Mouse Syngeneic and Xenograft Tumor Models (A–C) A murine TS/A (3 × 10⁵ cells) subcutaneous model was established in immunocompetent BALB/c mice (n = 7). After 7 days, when tumors became palpable, they were injected with 2 × 10⁷ PFU of the VSVΔ51-GFP at day 0 and day 4. Sulforaphane (SFN) at 10 mg/kg was administered i.p. 1 day before the first VSVΔ51 injection and then days 1, 4, and 6 following the first VSVΔ51 administration. (A) Tumor growth was monitored using caliper measurements at the indicated times, and average tumor volumes (n = 7) are shown. To account for variable tumor size at the beginning of the experiment, tumor volume was normalized to its initial volume on day 0 and presented as percentage. Error bars correspond to the SEM. ***p < 0.001 by two-way ANOVA. (B) Cumulative survival rate (n = 7) was monitored over time. The log rank test indicates that the combined treatment is significantly prolonged over virus alone (**p = 0.0064) or SFN alone (***p = 0.0002). (C) The cytotoxicity of each treatment was assessed by monitoring mouse body weight. Error bars correspond to the SEM (n = 7). (D–F) Human PC-3 cells were established in athymic male nude mice. Four weeks later, animals were treated by intratumoral injection with VSVΔ51 (2 × 10⁷ PFU/dose) on days 0, 3, 6, 9, and 12 and then treated intraperitoneally with SFN (10 mg/kg) every day from day −1 to day 10. (D) Tumor volume was monitored for each group, and average tumor volumes were normalized as shown (n = 7). To account for variable tumor size at the beginning of the experiment, tumor volume was normalized to its initial volume on day 0 and presented as percentage. Error bars correspond to the SEM. ***p < 0.001 by two-way ANOVA. (E) Survival rate (n = 7) was monitored over time. The log rank test indicates that the combined treatment is significantly prolonged over non-treated (***p = 0.0005) and SFN alone (**p = 0.0036). (F) The cytotoxicity of each treatment was assessed by monitoring mouse body weight (n = 7).

Figure 3

VSVΔ51 Replication Relies on Nrf2 and HO-1 (A) Intracellular levels of phosphorylated Nrf2 were detected by Phosflow in HEK293T stimulated for 18 hr with increasing doses of SFN. (B) HEK293T cells were pretreated for 24 hr with increasing doses of SFN, and the ARE promoter activity was assessed using a luciferase assay. (C) High-throughput analysis of gene expression was evaluated by qPCR BioMark analysis on PC-3 cells pretreated with SFN (20 μM) for 24 hr and subsequently infected with VSVΔ51-GFP (MOI 1) for 24 hr. Gene expression levels were calculated using the ΔΔCt method, and the gene-wise standardized expression (Z score) was generated for each gene. The scale represents Z score values, with red showing upregulation and blue showing downregulation in gene expression. Data are representative of three independent experiments. Each box of the heatmap represents one experiment. (D) WT, Nrf2^−/−, and Keap1^−/− MEF cells were pretreated for 18 hr with SFN (10 μM) and subsequently challenged with VSVΔ51-GFP (MOI 0.1). Infectivity was quantified by flow cytometry 24 hr post-infection. Data are the means ± SEM of one representative experiment performed in triplicate. The experiment was repeated twice with the same trend. (E–H) A549 cells were transfected with control, Nrf2, or HO-1 siRNA and 48 hr later were infected with VSVΔ51-GFP (MOI 0.01) for an extra 24 or 48 hr. (E) Knockdown efficiency was assessed by immunoblotting. Viral infectivity and replication were determined by microscopy (F) and plaque assay (G) at 24 hr post-infection, respectively. (H) Apoptosis was determined by flow cytometry 48 hr post-infection. Data are the means ± SEM of one representative experiment performed in triplicate. Experiment was repeated twice with the same trend. (I) A549 cells were transfected with control or Nrf2 siRNA and 48 hr later were infected with VSVΔ51-GFP (MOI 0.01) in the presence or absence of SFN (7.5 μM) for an extra 24 hr. Viral infectivity was determined by flow cytometry. Data are the means ± SEM of two independent experiments performed in duplicate. (J) U-2 OS cells were transfected with control or Nrf2 siRNA and 48 hr later were infected with VSVΔ51-GFP (MOI 0.001). Whole-cell extracts were also analyzed by immunoblotting for Nrf2, VSV proteins, and β-actin. (K) Viral infectivity was determined by flow cytometry 24 hr post-infection. Data are the means ± SEM of three independent experiments.

Figure 4

Sulforaphane-Enhanced VSVΔ51 Replication and Oncolysis Is Dependent on Nrf2-Dependent Autophagy (A) Immunofluorescence analysis of U-2 OS cells was conducted to detect changes in the levels of p62 and LC3-II in response to a 24 hr treatment with SFN (20 μM). (B) PC-3 cells were pretreated with SFN (20 μM) for 24 hr and were subsequently infected with VSVΔ51-GFP (MOI 1). Whole-cell extracts were analyzed by immunoblotting for LC3B, VSV proteins, ATG5, and β-actin. Results are from a representative experiment; all immunoblots are from the same samples. (C) PC-3 cells were pretreated with SFN (20 μM) in the presence or absence of the PI3K/autophagy inhibitor 3-MeA (10 mM) for 24 hr. Cells were subsequently infected with VSVΔ51-GFP (MOI 1). Infectivity was quantified by flow cytometry at 24 hr post-infection. Data are the means ± SEM from three independent experiments. (D–F) WT Atg5 and Atg5^−/− MEFs were stimulated overnight with SFN (7.5 μM) before challenge with VSVΔ51 (MOI 0.01). Viral infectivity was determined by flow cytometry (D) and viral replication by plaque assay (E). (F) Whole-cell extracts were analyzed by immunoblotting for LC3B, VSV proteins, ATG5, and β-actin. (G and H) A549 cells were transfected with control, Nrf2, or HO-1 siRNA and 48 hr later were treated with SFN (15 μM) for an extra 24 hr. Whole-cell extracts were analyzed by immunoblotting for Nrf2, HO-1, LC3, and β-actin. (I) Immunofluorescence analysis of U-2 OS cells was conducted to detect the changes in the levels of p62 and LC3-II in response to a 24 hr treatment with SFN (20 μM) in the presence of Nrf2 (siCtrl) or absence of Nrf2 (siNrf2).

Figure 5

Sulforaphane Dampens the Innate Antiviral Response (A) HEK293T cells were pretreated for 24 hr with increasing doses of SFN (1–20 μM), and the ISRE promoter activity was assessed using a luciferase assay in the presence or absence of VSVΔ51 (MOI 0.01). (B) PC-3 cells were pretreated with sulforaphane (SFN) (20 μM) for 24 hr and were subsequently infected with VSVΔ51-GFP (MOI 1). Antiviral and inflammatory gene expression levels were assessed by high-throughput qPCR. (C) PC-3 cells were pretreated for 24 hr with SFN (20 μM) and subsequently challenged with VSVΔ51-GFP (MOI 1) for 24 hr. Whole-cell extracts were analyzed for P-IRF3, IRF3, P-STAT1, STAT1, STING, HO-1, LC3, VSV proteins, and β-actin by immunoblotting. (D) PC-3 cells were pretreated for 24 hr with SFN (20 μM) and subsequently challenged with VSVΔ51-GFP (MOI 1) for 8 hr. Translocation of IRF3 and Nrf2 from cytoplasm to nucleus was investigated by immunoblotting in cytoplasmic and nuclear fractions at the same time. The experiment was repeated twice.

Figure 6

Nrf2-Induced Autophagy Curtails Antiviral Innate Immunity (A) 6 hr following transfection with increasing doses of an Nrf2 overexpression plasmid, HEK293T cells were infected with Sendai virus (SeV) (40 hemagglutinin [HA]/mL) for 18 hr and ISRE promoter activity was assessed by a luciferase reporter assay. (B) A549 cells were transfected with control or Nrf2 siRNA and 48 hr later were challenged with VSVΔ51-GFP (MOI 0.01) for an extra day. Whole-cell extracts were analyzed for Nrf2, VSV-GFP, LC3, P-STAT1, STAT1, RIG-I, and β-actin by immunoblotting. (C) 6 hr following transfection with increasing doses of an HO-1 overexpression plasmid, HEK293T cells were infected with SeV (40 HA/mL) for 18 hr and ISRE promoter activity was assessed by a luciferase reporter assay. (D) A549 cells were transfected with control or HO-1 siRNA and 48 hr later were challenged with VSVΔ51-GFP (MOI 0.01) for an extra day. Whole-cell extracts were analyzed for HO-1, VSV-GFP, LC3, P-STAT1, STAT1, RIG-I, and β-actin by immunoblotting. (E and F) A549 cells were transfected with control or Atg7 siRNA and 48 hr later were challenged with VSVΔ51-GFP (MOI 0.01) for an extra day. (E) VSVΔ51 infection was determined by flow cytometry. Data are the means ± SEM of three independent experiments. (F) Whole-cell extracts were analyzed for VSV proteins, RIG-I, P-STAT1, STAT1, and β-actin by immunoblotting. (G and H) U-2 OS cells were transfected with control or Nrf2 siRNA and 48 hr later were infected with VSVΔ51-GFP (MOI 0.001). (G) Viral infectivity was determined by flow cytometry 24 hr post-infection and observed using a fluorescent microscope. Data are the means ± SEM of three independent experiments. (H) Whole-cell extracts were also analyzed by immunoblotting for Nrf2, VSV proteins, LC3, P-STAT1, and β-actin.

Figure 7

Nrf2-Dependent Autophagy Decreases the Innate Antiviral Response to Augment VSVΔ51 Replication and Oncolytic Activity in Cancer Cells (A) Under homeostatic conditions, Nrf2 is sequestered in the cytoplasm by Keap1, a substrate adaptor protein for a E3 ubiquitin ligase complex that targets Nrf2 for ubiquitination and degradation by the proteasome. (B) Upon stimulation with sulforaphane (SFN), cysteine residues in Keap1 are modified, thereby inactivating its substrate adaptor function and disrupting the cycle of Nrf2 degradation. This results in Nrf2 stabilization, its nuclear translocation, and the transcriptional upregulation of a multitude of genes bearing antioxidant response element (ARE) in their promoter, including HO-1. (C and D) Increased Nrf2 signaling induces autophagy (C), which negatively regulates VSVΔ51-induced immune response (D) by disrupting RIG-I-MAVS interactions. (E and F) Suppression of the antiviral response facilitates VSVΔ51 replication (E) and oncolysis (F) in cancer cells.