Extracellular CIRP Promotes GPX4-Mediated Ferroptosis in Sepsis

Abstract

Sepsis is characterized by life-threatening organ dysfunction caused by a dysregulated host response to infection. Extracellular cold-inducible RNA-binding protein (eCIRP) is a damage-associated molecular pattern (DAMP) that promotes inflammation and induces cell death via apoptosis, NETosis, and/or pyroptosis. Ferroptosis is a form of regulated cell death characterized by the accumulation of lipid peroxide on cellular membranes. We hypothesize that eCIRP induces ferroptosis in macrophages and lung tissue during sepsis. RAW 264.7 cells stimulated with recombinant murine (rm) CIRP significantly decreased the expression of glutathione peroxidase 4 (GPX4), a negative regulator of ferroptosis, and increased lipid reactive oxygen species (ROS) in a TLR4 dependent manner. In TLR4^-/- peritoneal macrophages, depression of GPX4 expression and increase in lipid ROS levels were attenuated after rmCIRP-treatment compared to WT macrophages. rmCIRP also induced cell death in RAW 264.7 cells which was corrected by the ferroptosis inhibitor, ferrostatin-1 (Fer-1). Intraperitoneal injection of rmCIRP decreased GPX4 expression and increased lipid ROS in lung tissue, whereas the increase of lipid ROS was reduced by Fer-1 treatment. GPX4 expression was significantly decreased, while malondialdehyde (MDA), iron levels, and injury scores were significantly increased in lungs of WT mice after cecal ligation and puncture (CLP)-induced sepsis compared to CIRP^-/- mice. Treatment with C23, a specific eCIRP inhibitor, in CLP mice alleviated the decrease in GPX4 and increase in MDA levels of lung tissue. These findings suggest that eCIRP induces ferroptosis in septic lungs by decreasing GPX4 and increasing lipid ROS. Therefore, regulation of ferroptosis by targeting eCIRP may provide a new therapeutic approach in sepsis and other inflammatory diseases.

Keywords: GPX4; acute lung injury; eCIRP; ferroptosis; lung; macrophage; sepsis.

Figure 1

eCIRP downregulates GPX4 expression in macrophages. RAW 264.7 cells were stimulated with (A) 1 μg/mL of rmCIRP for 1, 4 and 20 h and (B) 0.01, 0.1 and 1 μg/mL of rmCIRP for 20 h. Cells were collected and GPX4 expression was determined by Western blotting. Representative western blot showing GPX4 expression was obtained from a single experiment. The experiments were performed at least 2-3 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as mean ± SEM (n = 4-6 sample/group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 vs. PBS).

Figure 2

eCIRP increases lipid ROS, and cell death which are attenuated by Fer-1 treatment in macrophages. RAW 264.7 cells were treated with PBS, rmCIRP or rmCIRP + Fer-1 for 20 h. Cells were collected and stained with C11 BODIPY 581/591 and assessed lipid ROS by flow cytometry. (A) Dot blots and histograms representing lipid ROS in RAW 264.7 cells are shown. (B) A total of 0.1 × 10 ⁶ cells were suspended in 1 mL of PBS and stained with 1 μL C11-BODIPY. Unstained cells were used as a negative control to establish the flow cytometer voltage setting. The FSC-A/FSC-H gating was used to remove doublets. Lipid peroxidation was quantified by the fluorescence intensities in FITC channel. We expressed the data in terms of percentages (%) of cells that were positive for C11-BODIPY staining as represented by the dotted line in the histogram. Quantitative bar diagrammatic presentation of lipid ROS in RAW 264.7 cells in PBS control, rmCIRP stimulation, rmCIRP stimulation with Fer-1 treatment are shown. The experiments were performed 2 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as mean ± SEM (n = 6 sample/group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 vs. PBS; ^#p < 0.05 vs. rmCIRP). (C, D) RAW 264.7 cells were treated with PBS, rmCIRP or rmCIRP + Fer-1 for 20 h. A total of 0.1 × 10 ⁶ cells were suspended in 1 mL of PBS and stained with 1 μL LIVE/DEAD Fixable Violet Dead Cell Stain kit. Unstained cells were used as a negative control to establish the flow cytometer voltage setting. The FSC-A/FSC-H gating was used to remove doublets. Cell death was quantified by the fluorescence intensities in Pacific Blue channel. We expressed the data in terms of percentages (%) of cells that were positive for Live/DEAD Fixable Violet Dead Cell staining as represented by the dotted line in the histogram. (C) Dot blots and histograms representing cell death in RAW 264.7 cells are shown. (D) Quantitative bar diagrammatic presentation of cell death in PBS control, rmCIRP stimulation, rmCIRP stimulation with Fer-1 treatment are shown. The experiments were performed 2 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as mean ± SEM (n = 8 sample/group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 vs. PBS; ^#p < 0.05 vs. rmCIRP).

Figure 3

eCIRP causes ferroptosis via TLR4 in macrophages. RAW 264.7 cells were treated with PBS, rmCIRP, rmCIRP + TLR4 Ab or rmCIRP + Isotype IgG for 20 h. (A) GPX4 expression were determined by Western blotting. (B) Dot blots and histograms representing lipid ROS in RAW 264.7 cells are shown. (C) Quantitative bar diagrammatic presentation of lipid ROS in PBS control, rmCIRP stimulation, rmCIRP stimulation with TLR4 Ab and rmCIRP stimulation with isotype IgG are shown. Representative western blot showing GPX4 expression was obtained from a single experiment. The experiments were performed at least 2 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as mean ± SEM (n = 5-6 sample/group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 vs. PBS; **p < 0.05 vs. rmCIRP; ^#p < 0.05 vs. rmCIRP + TLR4 Ab). (D–F) Peritoneal macrophages were isolated from both WT and TLR4^-/- mice and stimulated with 1 μg/mL rmCIRP for 20 h. (D) GPX4 expression was determined by Western blotting. Representative western blot showing GPX4 expression was obtained from a single experiment. (E) Dot blots and histograms representing lipid ROS in peritoneal macrophages are shown. (F) Quantitative bar diagrammatic presentation of lipid ROS in PBS control, rmCIRP stimulation of WT, and TLR4^-/- mice are shown. The experiments were performed at least 2 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as mean ± SEM (n = 4-8 sample/group). The groups were compared by one-way ANOVA and SNK method (*p < 0.05 vs. WT PBS; #p < 0.05 vs. WT rmCIRP).

Figure 4

Intraperitoneal injection of rmCIRP induces ferroptosis in lungs. (A) Lung tissues were collected after 4 h of intraperitoneal injection of rmCIRP and subjected to Western blotting to assess GPX4 expression. (B) Lung tissues were collected after 4 h of intraperitoneal injection of rmCIRP with or without Fer-1 treatment and subjected to MDA assay to assess lipid ROS. Representative western blot showing GPX4 expression was obtained from a single experiment. The experiments were performed at least 3 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as mean ± SEM (n = 9-12 mice/group). The groups were compared by Student’s t-test or one-way ANOVA and SNK method (*p < 0.05 vs. PBS; ^#p < 0.05 vs. rmCIRP).

Figure 5

GPX4 expression is reduced, and MDA levels are increased in the lungs after sepsis in WT mice, but not in CIRP^−/− mice. After 20 h of CLP or sham procedure, lung tissues were collected from WT and CIRP^−/− mice and subjected to (A) Western blotting to assess GPX4 expression, (B) MDA assay to assess lipid ROS, and (C) iron assay. Representative western blot showing GPX4 expression was obtained from a single experiment. The experiments were performed at least 2 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as means ± SE (n = 5-8 mice/group). The groups were compared by one-way ANOVA (*p < 0.05 vs. WT sham; ^#p < 0.05 vs. WT CLP mice). (D, E) After 20 h of CLP or sham procedure, lung tissues were collected from WT and CIRP^−/− mice. (D) Representative images of H&E-stained lung tissue at original magnification ×200. (E) Lung injury score calculated at original magnification ×400. n = 5 high-powered fields/group. Data were expressed as means ± SEM. The groups were compared by one-way ANOVA and SNK method (*p < 0.05 vs. WT sham; ^#p < 0.05 vs. WT CLP mice).

Figure 6

C23 protects mice from ferroptosis in the lungs in sepsis. After 20 h of CLP or sham procedure, lung tissues were collected from sham, vehicle (PBS), and C23 (8 mg/kg) treatment groups. (A) GPX4 expression was assessed by Western blotting and (B) lipid ROS was assessed by MDA assay. Representative western blot showing GPX4 expression was obtained from a single experiment. The experiments were performed 2 times, and all the data obtained were analyzed to create the bar diagrams. Data were expressed as means ± SE (n = 6 mice/group). The groups were compared by one-way ANOVA (*p < 0.05 vs. sham; ^#p < 0.05 vs. vehicle).

Figure 7

Summary schema.In sepsis, eCIRP is increased in the circulation and organs. eCIRP binds to TLR4 and decreases GPX4 expression, leading to the accumulation of lipid ROS to induce ferroptosis. Ferroptosis causes the release of DAMPs such as eCIRP and contributes to lung injury. C23 acts as an inhibitor of eCIRP through blocking the interaction between eCIRP and TLR4, thus decreasing lung ferroptosis in sepsis. The schema was prepared by BioRender.com.