STING/ACSL4 axis-dependent ferroptosis and inflammation promote hypertension-associated chronic kidney disease

Abstract

Hypertension is a primary modifiable risk factor for cardiovascular diseases, which often induces renal end-organ damage and complicates chronic kidney disease (CKD). In the present study, histological analysis of human kidney samples revealed that hypertension induced mtDNA leakage and promoted the expression of stimulator of interferon genes (STING) in renal epithelial cells. We used angiotensin II (AngII)- and 2K1C-treated mouse kidneys to elucidate the underlying mechanisms. Abnormal renal mtDNA packing caused by AngII promoted STING-dependent production of inflammatory cytokines, macrophage infiltration, and a fibrogenic response. STING knockout significantly decreased nuclear factor-κB activation and immune cell infiltration, attenuating tubule atrophy and extracellular matrix accumulation in vivo and in vitro. These effects delayed CKD progression. Immunoprecipitation assays and liquid chromatography-tandem mass spectrometry showed that STING and ACSL4 were directly combined at the D53 and K412 amino acids of ACSL4. Furthermore, STING induced renal inflammatory response and fibrosis through ACSL4-dependent ferroptosis. Last, inhibition of ACSL4 using small interfering RNA, rosiglitazone, or Fer-1 downregulated AngII-induced mtDNA-STING-dependent renal inflammation. These results suggest that targeting the STING/ACSL4 axis might represent a potential strategy for treating hypertension-associated CKD.

Keywords: angiotensin II; chronic kidney disease; hypertension; inflammation; renal end-organ damage; stimulator of interferon genes.

Graphical abstract

Figure 1

Increased STING expression in patients with hypertensive nephropathy, AngII- treated HK-2 cells, and mouse kidneys (A) IF for STING and dsDNA in human kidney biopsies. Scale bar, red, 100 μm; white, 20 μm (n = 12). (B) IF for STING and LTL in renal epithelial cells. LTL was used to stain renal epithelial cells. Scale bar, red, 100 μm; white, 20 μm (n = 12). (C) Correlation between renal tubular STING expression and inflammation (CD68, n = 12) or estimated fibrosis (Masson’s trichrome). (D) t-Distributed stochastic neighbor embedding maps of the single-cell RNA sequencing dataset GSE174219. (E) Western blot analysis and quantitative assessment of STING expression in HK-2 cells treated with 0, 10⁻⁷, or 10⁻⁶ M AngII (n = 3). (F) Western blot analysis and quantification of STING expression in HK-2 cells treated with 10⁻⁶ Mol AngII for 0, 12, 24, 36, or 48 h (n = 3). (G) Representative transmission electron micrographs of HK-2 cells treated with AngII. Scale bar, white, 1 μm (n = 3). (H) IF for STING and dsDNA in HK-2 cells. Scale bar, white, 20 μm (n = 3). (I) Workflow for generating an AngII-induced hypertension model. Mice were infused with vehicle (PBS) or 1,000 ng/kg/min AngII using Alzet minipumps for 4 weeks. (J) Representative transmission electron micrographs of renal epithelial cells in control and AngII-treated mouse groups. Scale bar, white, 1 μm (n = 6). (K and L) IF for STING/dsDNA and STING/LTL in control and AngII-treated mouse kidneys. Scale bar, red, 100 μm; white, 20 μm (n = 6). (M) IHC for F4/80 in control and AngII-treated mouse kidneys. Scale bar, black, 100 μm (n = 6). (N) Workflow for generating a 2K1C mouse hypertension model. The right renal artery was clipped with a U-shaped silver clip (internal diameter, 0.12 mm) for 4 weeks. (O) Representative transmission electron micrographs of renal epithelial cells in control and 2K1C groups. Scale bar, white, 1 μm (n = 6). (P and Q) IF for STING/dsDNA and STING/LTL in control and 2K1C-induced mouse kidneys. Scale bar, red, 100 μm; white, 20 μm (n = 6). (R) IHC for F4/80⁺ reveals the 2K1C-induced renal inflammatory response. Scale bar, black, 100 μm (n = 6); ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 compared with the control. HSC, hematopoietic stem cells.

Figure 2

Global STING knockout protects against AngII-mediated renal dysfunction, inflammation, and fibrosis (A) Schematic representation illustrating the genetic approach used to generate global STING knockout mice and determine STING deficiency via genomic DNA assessment. (B) IF for STING and dsDNA in AngII-treated TmemFF and CMVKO mouse kidneys. Scale bar, white, 20 μm. (C) Level of 24-h urine albumin in AngII-treated TmemFF mice and global knockout mice. (D) Level of GFR in AngII-treated TmemFF mice and global knockout mice. (E) Representative images of H&E- and PAS-stained kidneys of TmemFF and CMVKO mice treated with AngII. Scale bar, black, 100 μm. (F) Western blotting of P-TBK1(Ser172), TBK1, P-P65 (Ser536), and P65 in STING-deficient mouse kidneys. (G) Real-time PCR analysis of renal pro-inflammatory cytokine genes (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6). (H) IHC staining of renal F4/80⁺ macrophages in mice with AngII-induced nephropathy with or without knockdown of STING. Scale bar, red, 1,000 μm; black, 100 μm. (I) Masson staining and IHC images of α-SMA expression in the kidneys. Scale bar, black, 100 μm. Data are presented as the mean ± SEM of 6–7 independent experiments; ∗∗∗p < 0.001 compared with the control; ^##p < 0.01 and ^###p < 0.001 compared with the TmemFF-AngII group.

Figure 3

Conditional STING knockout protects against AngII-induced renal dysfunction, inflammation, and fibrosis (A) Schematic representation illustrating the genetic approach used to generate STING conditional knockout mice and determine of STING deficiency via genomic DNA assessment. (B) IF for STING and dsDNA in AngII-treated TmemFF and CDH16KO mouse kidneys. Scale bar, white, 20 μm. (C) Level of 24-h urine albumin in the mice. (D) GFR levels in AngII-treated TmemFF mice and CDH16KO mice. (E) Representative images of H&E- and PAS-stained kidneys of TmemFF and CDH16KO mice treated with or without AngII. Scale bar, black, 100 μm. (F) Western blotting of P-TBK1 (Ser172), TBK1, P-P65 (Ser536), and P65 in AngII-treated TmemFF and CDH16KO mouse kidneys. (G) Real-time PCR analysis of renal pro-inflammatory cytokine genes (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6). (H) Immunohistochemical staining of renal F4/80⁺ macrophages in TmemFF and CDH16KO mice with AngII-induced nephropathy. Scale bar, black, 1,000 μm; black, 100 μm. (I) Representative Masson staining and immunohistochemical images of renal α-SMA expression in TmemFF and CDH16KO mice treated with or without AngII. Scale bar, black, 100 μm. Data are presented as the mean ± SEM of 6–7 independent experiments; ∗∗∗p < 0.001 compared with the control; ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 compared with the TmemFF-AngII group. CDH16KO, STING conditional knockout in mouse kidney epithelial cells.

Figure 4

Global STING knockout protects against 2K1C-induced renal dysfunction, inflammation, and fibrosis (A) IF for STING and dsDNA in 2K1C-treated TmemFF and CMVKO mouse kidneys. Scale bar, white, 20 μm. (B) Level of GFR in 2K1C-treated TmemFF mice and global knockout mice. (C) Level of 24-h urine albumin in 2K1C-treated TmemFF mice and global knockout mice. (D) Serum blood urea and nitrogen levels in mice. (E) Representative images of H&E- and PAS-stained kidneys of TmemFF and CMVKO 2K1C mice. Scale bar, black, 100 μm. (F) Real-time PCR analysis of renal pro-inflammatory cytokine genes (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6). (G) IHC staining of renal F4/80⁺ macrophages in mice with 2K1C-induced nephropathy with or without knockdown of STING. Scale bar, black, 1000 μm; black, 100 μm. (H) Representative Masson staining and immunohistochemical images of α-SMA expression in TmemFF and CMVKO 2K1C mouse kidneys. Scale bar, black, 100 μm. Data are presented as the mean ± SEM of 6–7 independent experiments; ∗∗∗p < 0.001 compared with the control; ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 compared with the TmemFF-2K1C group. dsDNA, double-stranded DNA; CMVCre⁺/TmemFF.

Figure 5

Conditional STING knockout protects against 2K1C-induced renal dysfunction, inflammation, and fibrosis (A) IF for STING and dsDNA in 2K1C-treated TmemFF and CDH16KO mouse kidneys. Scale bar, white, 20 μm. (B) Level of GFR in 2K1C-treated TmemFF mice and STING-deficient mice. (C) Level of 24-h urine albumin in the mice. (D) Serum blood urea and nitrogen levels in 2K1C-treated TmemFF mice and CDH16KO mice. (E) Representative images of H&E− and PAS-stained kidneys of 2K1C-treated TmemFF and CDH16KO mice. Scale bar, black, 100 μm. (F) Real-time PCR analysis of renal pro-inflammatory cytokine genes (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6). (G) IHC staining of renal F4/80⁺ macrophages in TmemFF and CDH16KO mice with AngII-induced nephropathy. Scale bar, black, 1000 μm; black, 100 μm. (H) Representative Masson staining and immunohistochemical images of α-SMA expression in 2K1C-treated TmemFF and CDH16KO mouse kidneys. Scale bar, black, 100 μm. Data are expressed as the mean ± SEM of 6–7 independent experiments; ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 compared with the control; ^##p < 0.01 and ^###p < 0.001 compared with 2K1C-treated TmemFF group. dsDNA, double-stranded DNA; CDH16KO, STING conditional knockout in mouse kidney epithelial cells, CDH16-Cre⁺/TmemFF.

Figure 6

STING knockout alleviates inflammation and fibrosis in AngII-treated HK-2 cells (A) STING knockout in HK-2 cells. (B) IF for STING and dsDNA in AngII-treated HK-2 cells and STING knockout cells. (C and D) Western blotting and real-time PCR analysis of downstream phosphorylation of TBK1 and P65, as well as inflammatory cytokine production (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6). (E) Real-time PCR of TGF-β1, Collagen I, and α-SMA mRNA expression levels in AngII-treated HK-2 cells and STING knockout cells. (F) IF for Collagen I and α-SMA expression. Scale bar, white, 20 μm. Data are presented as the mean ± SEM of 3–4 independent experiments; ∗p < 0.05 and ∗∗∗p < 0.001 compared with the control; ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 compared with the CT-AngII group. CT, control; KO, STING knockout.

Figure 7

STING interacts with ACSL4 and regulates inflammation and fibrosis via ACSL4-dependent ferroptosis in HK-2 cells (A) ACSL4 detection in mass fragmentation of amino acids following digestion of proteins by immunoprecipitation assay using an anti-STING antibody in AngII-treated HK-2 and STING-overexpressing cells. (B) ACSL4 and STING detection using an IP assay with an anti-STING antibody in AngII-treated HK-2 and STING-overexpressing cells. (C) STING detection in mass fragmentation of amino acids following digestion of proteins by IP assay using the anti-ACSL4 antibody in AngII-treated HK-2 cells. (D) ACSL4 and STING detection using an IP assay with an anti-ACSL4 antibody in AngII-treated HK-2 cells. (E) Surface plasmon resonance. (F) Molecular docking of STING and ACSL4. (G) IP assay using an anti-STING antibody in STING EV cells transfected with ACSL4 EV and STING OE cells transfected with ACSL4 EV or ACSL4 wild type (overexpression of ACSL4; ACSL4 WT) or mutant vectors (T20A, Y31A, N40A, K49A, D53A, K412A, and Y415A). (H) Real-time PCR analysis of pro-inflammatory cytokine genes (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6) in STING OE-ACSL4 WT, STING OE-ACSL4 D53A, and STING OE-ACSL4 K412A cells. (I) Real-time PCR of TGF-β1, Collagen I, and α-SMA mRNA expression levels in STING OE-ACSL4 WT, STING OE-ACSL4 D53A, and STING OE-ACSL4 K412A cells. (J) Heatmap of untargeted lipidomics LC-MS/MS analysis showing differentially PE species and MS spectra of AdA-containing PE (18:0/22:4) in AngII-treated CT and STING KO cells representing substrates for oxygenation during ferroptosis (n = 3). (K–N) GSH, MDA, lipid peroxide, and Fe²⁺ levels measured in AngII-treated CT and STING knockout cells (n = 3). (O) Real-time PCR of inflammatory genes (ICAM, GM-CSF, TNF-α, CCL2, and CXCL6) and fibrogenic genes (TGF-β1, Collagen I, and α-SMA) in AngII-treated CT or SiACSL4 cells. (P) Real-time PCR of inflammatory and fibrogenic genes in STING OE cells transfected with CT or SiACSL4. (Q) Real-time PCR of inflammatory and fibrogenic genes in AngII-treated HK-2 cells with or without 100 μM ROSI. (R) Real-time PCR of inflammatory and fibrogenic genes in STING OE cells with or without 100 μM ROSI. (S) Real-time PCR of inflammatory and fibrogenic genes in AngII-treated HK-2 cells treated with or without 1 μM Fer-1. (T) Real-time PCR of inflammatory and fibrogenic genes in STING OE cells treated with or without 1 μM Fer-1. Data are presented as the mean ± SEM of 3–4 independent experiments; ∗ p < 0.05, ∗∗ p < 0.01 and ∗∗∗ p < 0.001 compared with the control; ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 compared with the AngII-CT group or OE-CT group or STING-OE-ACSL4-WT group. and ALA, alanine; CT, control; D and ASP, aspartic acid; EV, empty vector; Fer-1, ferrostatin-1; IP, immunoprecipitation; K and LYS, lysine; N and ASN, asparagine; OE, overexpression; ROSI, rosiglitazone; T and THR, threonine; Y and TYR, tyrosine. .

Figure 8

ACSL4 disruption attenuates AngII-induced renal dysfunction, inflammation, and fibrosis in vivo (A) Schematic representation illustrating the genetic approach for ROSI (i.p., every other day) administered to mice. (B) Level of 24-h urine albumin in AngII-treated mice with or without ROSI. (C) GFR levels in AngII-treated mice with or without ROSI. (D) Real-time PCR of inflammatory genes and fibrogenic genes in AngII-treated mouse kidneys. (E) Representative images of H&E, PAS, immunohistochemical staining of F4/80⁺ macrophages, Masson and immunohistochemical staining for α-SMA expression in the kidney of AngII-treated mice with or without ROSI. Scale bar, black, 100 μm. (F) Schematic illustrating the genetic approach for Fer-1 (i.p., every other day) administered to mice. (G) Level of 24-h urine albumin in AngII-treated mice with or without Fer-1. (H) GFR levels in AngII-treated mice with or without Fer-1. (I) Real-time PCR of renal inflammatory genes and fibrogenic genes. (J) Representative images of H&E, PAS, immunohistochemical staining of F4/80⁺ macrophages, Masson and immunohistochemical staining for α-SMA expression in the kidneys of AngII-treated mice with or without Fer-1. Scale bar, black, 100 μm. Data are presented as the mean ± SEM of 6–7 independent experiments; ^#p < 0.05, ^##p < 0.01, and ^###p < 0.001 compared with the AngII-Control group. AngII, angiotensin II; Fer-1, ferrostatin-1; ROSI, rosiglitazone.