SENP3 Drives Abdominal Aortic Aneurysm Development by Regulating Ferroptosis via De-SUMOylation of CTH

Abstract

Abdominal aortic aneurysm (AAA) is a high-risk inflammatory disorder. SENP3, a SUMO2/3-specific protease, is closely involved in the development of cancer. In this study, the aim is to explore the role of SENP3 in macrophages in AAA. It is found that the protein expression of SENP3 is significantly upregulated in both human and murine AAA specimens. SENP3 expression is negatively regulated by the E3 ubiquitin ligase STUB1/CHIP. Furthermore, myeloid-specific SENP3 knockout inhibited AAA formation in both AngII- and CaCl₂-induced mouse models. SENP3 deficiency repressed AAA lesion macrophage infiltration and inflammatory response. Mechanistic studies identified Cystathionine Gamma-Lyase (CTH), a critical enzyme involved in hydrogen sulfide production, as a target protein of SENP3 that mediated the exacerbating effects of SENP3 on ferroptosis and inflammatory programs in macrophages. SUMO-3 modification at Lysine 361 promoted CTH protein stability, whereas de-SUMOylation by SENP3 facilitated its proteasome-dependent degradation. Most importantly, it is found that CTH inhibitor counteracted the protective effect of SENP3 deficiency on AAA. Additionally, supplementation with ATB346, a novel H₂S-donating naproxen derivative, prevented AAA development in mice. These studies suggest that SENP3-mediated CTH deSUMOylation regulates macrophage ferroptosis and AAA development. The SENP3/CTH axis is therefore an important therapeutic target for aortic aneurysmal diseases.

Keywords: CTH; SENP3; SUMOylation; abdominal aortic aneurysm; ferroptosis.

Figure 1

The protein expression of SENP3 is significantly upregulated in human and murine AAA tissues. A–C) SENP3 protein and mRNA levels were determined by western blot (A,B) and real‐time quantitative polymerase chain reaction (RT‐qPCR; C) in human abdominal aortic aneurysm (AAA) samples and adjacent nonaneurysmal control samples (non‐AAA) (n = 6 per group). D–F) SENP3 protein and mRNA levels were determined by western blot (D,E) and RT‐qPCR (F) in abdominal aortic samples of ApoE^−/− mice infused with angiotensin II (AngII) or saline for 28 days (n = 6 per group). G) Representative images of dual immunofluorescence staining of SENP3 (green) and CD68 (red) in human AAA samples and adjacent nonaneurysmal control samples. Where indicated, nuclei were counterstained with 4′,6‐diamidino‐2‐phenylindole (DAPI; blue). Scale bar: 100 µm. H) Representative images of dual immunofluorescence staining for SENP3 (green) and CD68 (red) in abdominal aortic samples of ApoE^−/− mice infused with AngII or saline for 28 days. Scale bar: 100 µm. I,J) SENP3 protein levels were determined by western blot in BMDMs after stimulation with AngII (1 µmol L⁻¹) for the indicated time periods (n = 4 per group). K,L) SENP3 protein levels were determined by western blot in bone marrow‐derived macrophages (BMDMs) after stimulation with indicated concentrations of AngII for 180 min (n = 4 per group). M,N) Representative images of dual immunofluorescence staining for SENP3 (green) and CD68 (red) in BMDMs stimulated with AngII (1 µmol L⁻¹) for 180 min (M). Where indicated, nuclei were counterstained with DAPI (blue). Scale bar: 25 µm. Quantification of fluorescence intensity of SENP3 (N; n = 4 for each group). Data represent mean ± SEM. p values were determined using the student's t‐test (B,C,E,F,N) and Welch ANOVA test (J,L). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant (p > 0.05).

Figure 2

Myeloid‐specific SENP3 deficiency represses AngII‐induced AAA formation. A) Schematic protocol: ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice were subcutaneously injected with saline or AngII via a mini osmotic pump for 28 days (n = 30 per group). B) Survival curves in the indicated groups (n = 30 per group). Survival data were analyzed by the Kaplan–Meier method and compared using the log‐rank test. C,D) Representative images and quantification of the maximal abdominal aortic diameter monitored by micro‐ultrasound imaging (MUI) in the indicated groups (n = 20–30 per group). Scale bar: 5 mm. E) Representative images of macroscopic features of AAA formation in the indicated groups. Scale bar: 5 mm. F,G) Quantification of the maximal abdominal aortic diameter measured by a digital vernier caliper and ratio of aortic weight to body weight (BW) in the indicated groups (n = 20–30 per group). H,I) The incidences of AngII‐induced AAA (H) and aortic rupture (I) in the indicated groups (n = 30 per group). J) Representative images of suprarenal aortic sections stained with hematoxylin and eosin (H&E), Masson Trichrome, and Van Gieson after saline or Ang II infusion (n = 6 per group). Scale bar: 100 µm. K,L) Quantification of relative collagen content (K) and numbers of elastin breaks per vessel (L) in each group. Data represent mean ± SEM. p values were determined using two‐way ANOVA followed by Bonferroni post‐hoc test (D,F,G,K,L) and Fisher exact test (H,I). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3

Myeloid‐specific SENP3 deficiency represses AAA Lesion macrophage infiltration and inflammatory response. A) The heatmap shows differentially expressed genes (DEGs) identified by RNA‐seq in aortic tissues from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days. Each column represents an individual replicate, and each row represents an individual gene. Upregulated genes are shown in red, and downregulated genes are displayed in blue. N = 4 for each group. Genes with an adjusted p‐value < 0.01 and Fold Change ≥ 2 found by DESeq2 were assigned as differentially expressed. B) Volcano plot reveals the magnitude and significance of DEGs in aortic tissues from AngII‐infused ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice. C) Gene Ontology (GO) pathway enrichment analysis of DEGs in aortic tissues from AngII‐infused ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice. D) Representative images of immunofluorescence staining for CD68 (marker for macrophage) in abdominal aortic samples from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days. Scale bar: 100 µm. Quantification of fluorescence intensity of CD68 (n  =  4 for each group). E) Serum levels of IL‐6, TNF‐α, and MCP‐1 in ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days were assessed by ELISA. F) Immunohistochemistry analysis of IL‐6, TNF‐α, and MCP‐1 in abdominal aortas from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days. Scale bar: 100 µm. G) Representative images of transwell migration assay and quantification of migrated cells (number/field) in BMDMs isolated from Senp3^flox/flox and Senp3^△Mø mice. Scale bar: 50 µm. H) Representative images of dual immunofluorescence staining of MMP‐9 (green) and CD68 (red) in abdominal aortic samples from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days. Where indicated, nuclei were counterstained with DAPI (blue). Scale bar: 100 µm. I) MMP9 protein levels were determined by western blot in abdominal aortic samples of ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII or saline for 28 days. J) In situ zymography for gelatinase activity in abdominal aortic samples from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days. Scale bar: 100 µm. Data represent mean ± SEM. p values were determined using the student's t‐test (D‐F,H‐J) and two‐way ANOVA followed by the Bonferroni post hoc test (G). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 4

SENP3 deficiency mitigates ferroptosis in macrophage. A) The protein levels of GPX4, ACSL4, SLC7A11, and FTH1 were determined by western blot in human AAA samples and adjacent nonaneurysmal control samples (non‐AAA) (n = 6 per group). B) Representative images and quantitative analysis of Berlin blue staining of iron in human non‐AAA and AAA samples (scale bar, 100 µm, n = 4). C,D) The mRNA levels and protein levels of GPX4, ACSL4, SLC7A11, and FTH1 were determined by RT‐qPCR (C) and western blot (D) in abdominal aortic samples from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII or saline for 28 days. E) Representative images and quantitative analysis of berlin blue staining of iron in abdominal aortic samples from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII or saline for 28 days (scale bar, 100 µm, n = 4). F) Representative images and quantitative analysis of dihydroethidium (DHE) staining in abdominal aortic samples from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days (scale bar, 100 µm, n = 4). G) The ratio of reduced glutathione (GSH) to oxidized glutathione (GSSG) (GSH/GSSG) and levels of superoxide dismutase (SOD) and malondialdehyde (MDA) in aortic tissues from ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII or saline for 28 days were measured. H) THP‐1 were transfected with control siRNA (si‐con) and Senp3 siRNA (si‐Senp3) for 48 h. The protein levels of GPX4, ACSL4, SLC7A11, and FTH1 were determined by western blot. I,J) The levels of cellular ferrous iron (Fe²⁺) and mitochondrial Fe²⁺ were assessed by FerroOrange probes and Mito‐FerroGreen probes in BMDMs which were isolated from Senp3^flox/flox and Senp3^△Mø mice and stimulated with AngII (1 µmol L⁻¹) or PBS for 24 h (scale bar, 100 µm, n = 4). K) Representative images of DHE staining in BMDMs isolated from Senp3^flox/flox and Senp3^△Mø mice and stimulated with AngII (1 µmol L⁻¹) for 24 h (scale bar, 100 µm, n = 4). L) Representative images of Mito tracker (above) and electron micrograph of mitochondria (below) in BMDMs isolated from Senp3^flox/flox and Senp3^△Mø mice and stimulated with AngII (1 µmol L⁻¹) or PBS for 24 h (scale bar, 20 or 0.5 µm, n = 4). Data represent mean ± SEM. p values were determined using the student's t‐test (A, B, F) and two‐way ANOVA followed by Bonferroni post‐hoc test (C–E,G–J). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant (p > 0.05).

Figure 5

SENP3 regulates AAA by targeting CTH to regulate ferroptosis signals. A) By utilizing IP‐MS in conjunction with the STRING database based on PI3K/AKT signaling pathways, as well as macrophage gene set, two overlapping targets, CTH and Cox4i1, were identified. In the IP‐MS analysis results, the enrichment score of CTH was higher than that of Cox4i1. B) CTH protein levels were determined by western blot in human AAA samples and adjacent nonaneurysmal control samples (non‐AAA) (n = 6 per group). C) CTH protein levels were determined by western blot in abdominal aortic samples of ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII or saline for 28 days. D) CTH protein levels were determined by western blot in BMDMs which were isolated from Senp3^flox/flox and Senp3^△Mø mice and stimulated with AngII (1 µmol/L) for 180 min. E,F) Representative images of dual immunofluorescence staining for CTH (green) and CD68 (red) in abdominal aortic samples of ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with AngII for 28 days. Where indicated, nuclei were counterstained with DAPI (blue). G,H) Representative images of dual immunofluorescence staining for CTH (green) and CD68 (red) in BMDMs which were isolated from Senp3^flox/flox and Senp3^△Mø mice and stimulated with AngII (1 µmol L⁻¹) for 180 min. Where indicated, nuclei were counterstained with DAPI (blue). I) The plasma levels of H₂S in ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice infused with saline or AngII for 28 days were determined (n = 6 for each group). J) H₂S concentrations were determined for the cellular homogenate of BMDMs which were isolated from Senp3^flox/flox and Senp3^△Mø mice and stimulated with AngII (1 µmol L⁻¹) for 24 h. K) BMDMs were transfected with control siRNA (si‐NC) and CTH siRNA (si‐CTH) for 48 h. The knockout efficiency of CTH was verified by western blot. L) The protein levels of GPX4, ACSL4, SLC7A11, and FTH1 were determined by western blot in BMDMs which were isolated from Senp3^flox/flox and Senp3^△Mø mice and were further transfected with si‐NC and si‐CTH for 48 h before stimulation with AngII (1 µmol L⁻¹) for 24 h. M) BMDMs isolated from Senp3^flox/flox and Senp3^△Mø mice were transfected with si‐NC and si‐CTH for 48 h before stimulation with AngII (1 µmol L⁻¹) for 24 h. The mRNA levels of IL‐1β, TNF‐α, CCL2, and IL‐6 were determined by RT‐qPCR. p values were determined using student's t‐test (B, F, H, and K) and two‐way ANOVA followed by Bonferroni post‐hoc test (C,D,I,J,L,M). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant (p > 0.05).

Figure 6

SENP3 de‐SUMOylates CTH and regulates its protein stability. A) 293T cells were transfected with Flag‐CTH, HA‐SUMO3, GFP‐SENP3, and Ubc‐9 for 48 h. The SUMOylation of Flag‐CTH was determined by the immunoprecipitation (IP) assay using Flag beads and western blot using anti‐Flag, anti‐HA, and anti‐GAPDH antibodies. B) Exogenous interaction between SENP3 and CTH was determined by co‐immunoprecipitation (co‐IP) using GFP‐beads. C,D) The endogenous interaction between SENP3 and CTH was determined in BMDMs stimulated with AngII by coimmunoprecipitation assay using antibody against SENP3 or CTH. E,F) SUMOplotTM Analysis Program online tool (https://www.abcepta.com/sumoplot, accessed on 11 February 2023), JASSA, and GPS‐SUMO were employed to predict the SUMOylation sites of CTH. G) Structure of CTH protein and site of K361. H) 293T cells were transfected with Flag‐CTH/WT or Flag‐CTH/K361R, HA‐SUMO3, and Ubc‐9 for 48 h. The SUMOylation of Flag‐CTH was determined by the IP assay using Flag beads and western blot using anti‐Flag, anti‐HA, and anti‐GAPDH antibodies. I) 293T cells were transfected with Flag‐CTH and increasing amounts of GFP‐SENP3 for 48 h. The protein levels of Flag‐CTH, GFP‐SENP3, and GAPDH in whole cell lysates were determined by western blot with anti‐Flag, anti‐GFP, and anti‐GAPDH antibodies. J) BMDMs were infected with increasing amounts of adenovirus expressing HA‐SENP3 for 48 h. The levels of CTH and HA‐SENP3 in whole cell lysates were determined by western blot with anti‐CTH, anti‐HA, anti‐SENP3, and anti‐GAPDH antibodies. K) 293T cells were transfected with control siRNA (si‐NC) and SENP3 siRNA (si‐SENP3) for 48 h. Lysates were prepared and analyzed by western blot. L) 293T cells were transfected with Flag‐CTH and equal amounts of PCDNA or GFP‐SENP3 or GFP‐SENP3/C532A for 48 h. Lysates were prepared and analyzed by western blot. M) 293T cells were transfected with Flag‐CTH and increasing amounts of GFP‐SENP3 for 48 h, in the presence or absence of MG132 (10 µmol L⁻¹) for the last 10 h. Lysates were prepared and analyzed by western blot. N) 293T cells were transfected with Flag‐CTH and PCDNA or GFP‐SENP3 for 36 h and were subsequently exposed to the protein synthesis inhibitor cycloheximide (CHX) for the indicated time. Lysates were prepared and analyzed by western blot. The relative level of CTH was evaluated by densitometry and normalized to GAPDH. O) 293T cells were transfected with Flag‐CTH/WT or Flag‐CTH/K361R and increasing amounts of GFP‐SENP3 for 48 h. The levels of Flag‐CTH in whole‐cell lysates were determined by western blot with anti‐Flag, anti‐GFP, and anti‐GAPDH antibodies.

Figure 7

CTH inhibitor counteracts the protective effect of SENP3 deficiency on AAA. A) Schematic protocol: ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice were infused with saline or AngII (1.44 mg kg⁻¹ d⁻¹) and intraperitoneally injected with DL‐Propargylglycine (PAG) (50 mg kg⁻¹ d⁻¹) for 28 days (n = 20 per group). B) Survival curves in the indicated groups (n = 20 per group). Survival data were analyzed by the Kaplan–Meier method and compared using the log‐rank test. C,D) Representative images and quantification of the maximal abdominal aortic diameter monitored by micro‐ultrasound imaging in each group. Scale bar: 5 mm. E) Representative images of the abdominal aorta visualized by macroscopic examination in the indicated groups. Scale bar: 5 mm. F) Quantification of the maximal abdominal aortic diameter measured by a digital vernier caliper for each group. G,H) The incidences of AngII‐induced AAA (G) and aortic rupture (H) in the indicated groups. Data were analyzed by a Fisher exact test. I) Representative images of suprarenal aortic sections stained with hematoxylin and eosin (H&E), Masson Trichrome, and Van Gieson in each group. Scale bar: 100 µm. J,K) Quantification of relative collagen content (J) and numbers of elastin breaks per vessel (K) in each group. Data represents mean ± SEM. p values were determined using two‐way ANOVA followed by Bonferroni post‐hoc test (D,F,J,K) and Fisher exact test (G,H). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant (p > 0.05).

Figure 8

Supplementation with the H₂S donor ATB346 protects against AAA formation. A) Schematic protocol: ApoE^−/−;Senp3^flox/flox and ApoE^−/−;Senp3^△Mø mice were infused with saline or AngII (1.44 mg kg⁻¹ d⁻¹) and intragastrically injected with ATB346 (16 mg kg⁻¹) for 28 days (n = 20 per group). B) Survival curves in the indicated groups (n = 20 per group). Survival data were analyzed by the Kaplan‐Meier method and compared using the log‐rank test. C,D) Representative images and quantification of the maximal abdominal aortic diameter monitored by micro‐ultrasound imaging in each group. Scale bar: 5 mm. E) Representative images of the abdominal aorta visualized by macroscopic examination in the indicated groups. Scale bar: 5 mm. F) Quantification of the maximal abdominal aortic diameter by a digital vernier caliper for each group. Data were analyzed by two‐way ANOVA followed by the Bonferroni post hoc test. G,H) The incidences of AngII‐induced AAA (G) and aortic rupture (H) in the indicated groups. I) Representative images of suprarenal aortic sections stained with hematoxylin and eosin (H&E), Masson Trichrome, and Van Gieson in each group. Scale bar: 100 µm. J,K) Quantification of relative collagen content (J) and numbers of elastin breaks per vessel (K) in each group. Data represent mean ± SEM. p values were determined using two‐way ANOVA followed by Bonferroni post‐hoc test (D,F,J,K) and Fisher exact test (G.H). For all panels, *p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant (p > 0.05).