The involvement of the Stat1/Nrf2 pathway in exacerbating Crizotinib-induced liver injury: implications for ferroptosis

Abstract

Crizotinib carries an FDA hepatotoxicity warning, yet analysis of the FAERS database suggests that the severity of its hepatotoxicity risks, including progression to hepatitis and liver failure, might be underreported. However, the underlying mechanism remains poorly understood, and effective intervention strategies are lacking. Here, mRNA-sequencing analysis, along with KEGG and GO analyses, revealed that DEGs linked to Crizotinib-induced hepatotoxicity predominantly associate with the ferroptosis pathway which was identified as the principal mechanism behind Crizotinib-induced hepatocyte death. Furthermore, we found that ferroptosis inhibitors, namely Ferrostatin-1 and Deferoxamine mesylate, significantly reduced Crizotinib-induced hepatotoxicity and ferroptosis in both in vivo and in vitro settings. We have also discovered that overexpression of AAV8-mediated Nrf2 could mitigate Crizotinib-induced hepatotoxicity and ferroptosis in vivo by restoring the imbalance in glutathione metabolism, iron homeostasis, and lipid peroxidation. Additionally, both Stat1 deficiency and the Stat1 inhibitor NSC118218 were found to reduce Crizotinib-induced ferroptosis. Mechanistically, Crizotinib induces the phosphorylation of Stat1 at Ser727 but not Tyr701, promoting the transcriptional inhibition of Nrf2 expression after its entry into the nucleus to promote ferroptosis. Meanwhile, we found that MgIG and GA protected against hepatotoxicity to counteract ferroptosis without affecting or compromising the anti-cancer activity of Crizotinib, with a mechanism potentially related to the Stat1/Nrf2 pathway. Overall, our findings identify that the phosphorylation activation of Stat1 Ser727, rather than Tyr701, promotes ferroptosis through transcriptional inhibition of Nrf2, and highlight MgIG and GA as potential therapeutic approaches to enhance the safety of Crizotinib-based cancer therapy.

Fig. 1. Signal association plots of hepatotoxicity and selected key hepatotoxicity PTs were analyzed for various TKIs in the FAERS database.

A Association map of different TKIs with reported hepatotoxicity signals. B Heatmap of the association between different TKIs and selected key hepatotoxicity PTs signals. ALK anaplastic lymphoma kinase, ROS1 ROS proto-oncogene 1, cMET cellular-mesenchymal to epithelial transition factor, EGFR epithelial growth factor receptor, RET proto-oncogene tyrosine-protein kinase receptor ret, HER2 human epidermal growth factor receptor-2, VEGFR vascular endothelial growth factor receptor, CSF-1R colony-stimulating factor 1 receptor, KIT KIT proto-oncogene receptor tyrosine kinase, FLT3 FMS-like tyrosine kinase 3, RAF1 Raf-1 proto-oncogene, serine/threonine kinase, PDGFR platelet-derived growth factor receptors, ROR reporting odds ratio, 95%CI 95% confidence interval.

Fig. 2. Crizotinib induced ferroptosis in AML12 cells (n = 3).

A Signification enriched top 20 KEGG pathway for the DEGs between the control and Crizotinib group. B and C The changes of ACSL4, SLC7A11, GPX4 protein expression (B), and PTGS2 mRNA levels (C) in AML12 cells treated with Crizotinib in different concentrations for 48 h were observed. D and E Iron content (D) and MDA (E) were measured by reagent kit following treatment with increasing concentrations of Crizotinib for 48 h. F The induction of lipid ROS was determined by BODIPY™ 581/591 C₁₁ and flow cytometry for 48 h. G Representative TEM images showed mitochondrial damages treated with 13 μM Crizotinib for 48 h in the AML12 cell. Scale bars, 2 μm (upper panel) and 500 nm (lower panel). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group. Crizo Crizotinib.

Fig. 3. Fer1 and DFO alleviate Crizotinib-triggered hepatotoxicity and ferroptosis.

A and B The survival rate of AML12 cells (A, n = 6) and HL-7702 cells (B, n = 3) treated with Crizotinib and/or DFO in different concentrations for 48 h. C57 male mice were treated with 120 mg/kg/day Crizotinib and/or 1 mg/kg/day Fer1 or 30 mg/kg/day DFO for 3 weeks. C and D The levels of serum ALT (C) and AST (D) were analyzed (n = 7–8). E Representative images of H&E staining in liver tissues (H&E staining, ×20). F The expression of SLC7A11 and GPX4 protein were measured by western blot (n = 6). G The GPX4 mRNA level was determined by RT-qPCR (n = 8). H–J GSH, the reduced GSH/HSSG (H, n = 8), total Iron concentration measured by reagent box and ICP-MS (I, n = 8), MDA and LPO (J, n = 8) were measured. K and L Representative pictures of 4-HNE expression in each group measured by immunohistochemical staining (K, ×20). And the staining outcomes were quantitatively assessed using the H-score system (L, n = 5). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 vs. Crizotinib group^. Crizo Crizotinib.

Fig. 4. Nrf2 overexpression attenuated Crizotinib-induced hepatocellular ferroptosis in AML12 cells.

A The protein level of total Nrf2 and nuclear Nrf2 in liver tissues was detected by western blot (n = 6). B The mRNA levels of Nrf2 and its downstream gene in liver tissues (n = 8) were determined by RT-qPCR. Vector or Nrf2 OE transfected AML12 cells were treated with or without 13 μM Crizotinib for 48 h. C Protein expression of Nrf2, HO1, SLC7A11, GPX4 and ACSL4 was measured by western blot (n = 3). D Fold change in MDA level (n = 3). E Fold change in Iron content (n = 3). F Fe²⁺ was detected by FerroOrange probe (×20). G The induction of lipid ROS was determined by BODIPY™ 581/591 C11 and flow cytometry (n = 3). H The survival rate of AML12 cells was treated with 13 μM Crizotinib and/or tBHQ at different concentrations for 48 h (n = 6). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 vs. Crizotinib group^. Crizo Crizotinib.

Fig. 5. Overexpression of Nrf2-attenuated Crizotinib-induced hepatocytes ferroptosis and hepatotoxicity in vivo.

AAV8-TBG- Vector or AAV8-TBG- Nrf2 were injected into C57 male mice through the tail vein. 4 weeks later, mice were treated with 0.5% CMC-Na or 120 mg/kg/day Crizotinib (n = 8). A and B The levels of serum ALT (A) and AST (B) were measured (n = 8). C Representative images of H&E staining in liver tissues (20×). D The protein levels of total Nrf2, SLC7A11, and GPX4 of liver lysate were analyzed by western blot (n = 6). E The mRNA levels of Nrf2, GCLC, GCLM, GPX4, FPN, and FTH1 of liver lysate were analyzed by RT-qPCR (n = 8). F–H Cys (F, n = 8), GSH, the reduced GSH/HSSG (G, n = 8), MDA, and LPO (H, n = 8) were measured by the reagent kit. I total Iron concentration was measured by reagent box and ICP–MS (n = 8). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. the WT group. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 vs. WT+ Crizotinib group. ^&P < 0.05, ^&&P < 0.01 and ^&&&P < 0.001 vs. Nrf2^+/+ group. Crizo Crizotinib.

Fig. 6. Silence of Stat1 attenuated Crizotinib-induced hepatocellular ferroptosis in AML12 cells.

SiNC or Stat1 siRNA transfected AML12 cells were treated with or without 13 μM Crizotinib for 48 h. A Protein expression of Stat1, pStat1 Ser727, SLC7A11, GPX4 and ACSL4 was measured by western blot (n = 3). B Fold change in MDA level (n = 3). C Fold change in Iron content (n = 3). D The induction of lipid ROS was determined by BODIPY™ 581/591 C11 and flow cytometry (n = 3). E The survival rate of AML12 cells was treated with 13 μM Crizotinib and/or NSC118218 at different concentrations for 48 h (n = 6). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 vs. Crizotinib group. Crizo Crizotinib.

Fig. 7. STAT1 is a transcription factor of the Nrf2 gene.

SiNC or Stat1 siRNA transfected AML12 cells were treated with or without 13 μM Crizotinib for 48 h. A protein expression of total Nrf2 and HO1 was measured by western blot (n = 3). AML12 cells were treated with 13 μM Crizotinib and/or NSC118218 at different concentrations for 48 h. B protein expression of total Nrf2 and HO1 was measured by western blot (n = 3). C Representative images of immunofluorescence for p-Stat1 (Ser727) staining and Nrf2 staining in liver tissues (×20). D Dual-luciferase reporter assay of Nrf2 and STAT1 with or without Crizotinib treatment (n = 3). E Stat1 binding motif provided by the JASPAR database. F STAT1 binds to its binding sites in the Nrf2 promoter region as shown by the ChIP assay in AML12 cells with or without the transfection of Stat1 overexpression plasmid. *P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 vs. Crizotinib group.

Fig. 8. Phosphorylation of Stat1 Ser727 induction is associated with Nrf2 inhibition and ferroptosis (n = 3).

A Transfection efficiency of Stat1 WT and pStat1 Ser727A plasmid was evaluated by western blot. B Protein expression of pStat1 Ser727, total Nrf2, and GPX4 was measured by western blot. C Fold change in MDA level. D The induction of lipid ROS was determined by BODIPY™ 581/591 C11 and flow cytometry. E Fold change in Iron content. F Fe²⁺ was detected by FerroOrange probe (×20). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group. ^#P < 0.05 and ^##P < 0.01 vs. Crizotinib group.

Fig. 9. GA and MgIG attenuate Crizotinib-induced hepatocyte ferroptosis via the Stat1/Nrf2 pathway in vivo.

A and B the cell viability of HL7702 and AML12 cells treated with 13 μM Crizotinib and/or GA (n = 3) or MgIG (n = 6) at different concentrations for 48 h. C57 male mice were treated with 120 mg/kg/day Crizotinib and/or 5 mg/kg/day GA or 10 mg/kg/day MgIG for 3 weeks. C and D The levels of serum ALT (C) and AST (D) were analyzed (n = 8). E Representative images of H&E staining in liver tissues (×20). F The protein levels of pStat1 Ser727, Stat1, total Nrf2, SLC7A11, and GPX4 of liver lysate were analyzed by western blot (n = 6). G The mRNA levels of Nrf2, GCLC, GCLM, GPX4, FPN, and FTH1 of liver lysate were analyzed by RT-qPCR (n = 8). H, J Cys, GSH, the reduced GSH/HSSG (H, n = 8), MDA, and LPO (J, n = 8) were measured by reagent kit. I Total Iron concentration was measured by reagent box and ICP-MS (n = 8). K Representative pictures of 4-HNE expression in each group measured by immunohistochemical staining (×20). *P < 0.05, **P < 0.01 and ***P < 0.001 vs. the control group. ^#P < 0.05, ^##P < 0.01 and ^###P < 0.001 vs. Crizotinib group.

Fig. 10. Schematic diagram showing the mechanism of Crizotinib-induced hepatotoxicity.

Crizotinib triggers the phosphorylation of Stat1 at Ser727, which facilitates the transcriptional suppression of Nrf2 expression upon nuclear entry, promoting ferroptosis, and leading to liver toxicity.