Ncf1 knockout in smooth muscle cells exacerbates angiotensin II–induced aortic aneurysm and dissection by activating the STING pathway

Abstract

Aims

Aortic aneurysm and dissection (AAD) is caused by the progressive loss of aortic smooth muscle cells (SMCs) and is associated with a high mortality rate. Identifying the mechanisms underlying SMC apoptosis is crucial for preventing AAD. Neutrophil cytoplasmic factor 1 (Ncf1) is essential in reactive oxygen species production and SMC apoptosis; Ncf1 absence leads to autoimmune diseases and chronic inflammation. Here, the role of Ncf1 in angiotensin II (Ang II)–induced AAD was investigated.

Methods and results

Ncf1 expression increased in injured SMCs. Bioinformatic analysis identified Ncf1 as a mediator of AAD-associated SMC damage. Ncf1 expression is positively correlated with DNA replication and repair in SMCs of AAD aortas. AAD incidence increased in Ang II–challenged Sm22^CreNcf1^fl mice. Transcriptomics showed that Ncf1 knockout activated the stimulator of interferon genes (STING) and cell death pathways. The effects of Ncf1 on SMC death and the STING pathway in vitro were examined. Ncf1 regulated the hydrogen peroxide–mediated activation of the STING pathway and inhibited SMC apoptosis. Mechanistically, Ncf1 knockout promoted the ubiquitination of nuclear factor erythroid 2-related factor 2 (NRF2), thereby inhibiting the negative regulatory effect of NRF2 on the stability of STING mRNA and ultimately promoting STING expression. Additionally, the pharmacological inhibition of STING activation prevented AAD progression.

Conclusion

Ncf1 deficiency in SMCs exacerbated Ang II–induced AAD by promoting NRF2 ubiquitination and degradation and activating the STING pathway. These data suggest that Ncf1 may be a potential therapeutic target for AAD treatment.

Graphical Abstract

Graphical Abstract.

Time of primary review: 14 days

1. Introduction

Aortic aneurysm and dissection (AAD) is a potentially lethal cardiovascular disease that carries a high risk of morbidity and mortality¹ and leads to >150 000200 000 deaths worldwide each year.^2,3 AAD is often asymptomatic with slow expansion until the onset of complications or rupture. The risk factors for AAD include advanced age, male sex, smoking, and family history.² Currently, no drugs have been clinically proven effective in preventing AAD; open surgery and endovascular aortic repair remain the main treatment methods.⁴

Many cell types, including vascular smooth muscle cells (VSMCs),⁴ endothelial cells,⁵ neutrophils,⁶ monocytes/macrophages,⁷ lymphocytes,⁷ adipocytes,⁸ and mast cells,⁷ contribute to AAD development. AAD pathological changes are characterized by extracellular matrix (ECM) degradation, thinning of the aortic wall, immune cell infiltration, increased oxidative stress, and loss of VSMCs in the media.⁴ Specifically, smooth muscle cell (SMC) apoptosis has long been considered as a hallmark of AAD pathology.^3,4

Reactive oxygen species (ROS), including hydrogen peroxide (H₂O₂), superoxide (O₂⁻), and hydroxyl radical (•OH), are highly reactive oxygen-derived chemical molecules.⁹ AAD tissues are characterized by abnormal nicotinamide adenine dinucleotide phosphate oxidase (NOX) activity and ROS production.¹⁰ Evidence from prior studies shows that ROS are intricately implicated in AAD pathogenesis.¹¹ ROS cause SMC DNA damage in sporadic ascending thoracic AAD (ATAAD) tissues and promote SMC apoptosis via the stimulator of interferon genes (STING) pathway.¹²

Recent studies have shown that some diseases are associated with ROS deficiency, such as chronic granulomatous disease and systemic lupus erythematosus (SLE),¹³ which are related to missense mutations or low copy numbers of NCF1,¹⁴ a key regulatory gene of ROS. Additionally, NCF1-derived ROS protect animal models from arthritis, psoriasis, and lupus.¹⁵ Moreover, low ROS production and NCF1 mutation are accompanied by an exaggerated type I interferon (IFN) response, leading to overexpression of IFN-stimulated genes (ISGs).^15,16 Although these data indicate the protective effect of ROS in diseases, the molecular mechanism remains largely unknown.

Owing to the dual role of ROS in many diseases, the role of Ncf1 in AAD was investigated. This research indicates the protective effect of ROS in AAD, suggesting Ncf1 as a potential therapeutic target for AAD treatment.

2. Methods

2.1. Ethics statement of human studies

Patients’ aortic tissues were from the Department of Cardiovascular Surgery, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. Written informed consent, according to the Declaration of Helsinki, was obtained from patients and organ donors. The study protocol was approved by the Union Hospital Independent Ethics Committee (Ethical Review No. 2023-0385) of Huazhong University of Science and Technology.

2.2. Human study

Details of this section are shown in Supplementary material online, Table S1.

2.3. Animal studies

ApoE^−/− (C57BL/6J-KO) mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Ncf1^fl (C57BL/6JSmoc-Ncf1em1^(flox)Smoc) mice were purchased from Shanghai Model Organisms Center, Inc. (Shanghai, China; online Figure IV). The genotyping primers are listed in Supplementary material online, Table S2. Sm22-Cre (B6.Cg-Tg(Tagln-cre)1Her/J) mice were purchased from Jackson Laboratory (Maine, USA). All mice were aged 8–12 weeks, weighted 20–30 g, and were housed in a specific pathogen-free laboratory with a standard chow diet at Huazhong University of Science and Technology (Wuhan, China). The mice used in our study were male mice, except where otherwise indicated. Mice were infused with angiotensin II (Ang II; 1000 ng/min/kg, 4 weeks; Sigma-Aldrich) or saline with subcutaneously implanted Alzet osmotic minipumps (Model 2004, Alzet, USA) implanted under isoflurane anaesthesia (2%). Blood pressure was measured by tail-cuff plethysmography at Days 0, 7, 14, 21, and 28 during the Ang II infusion. Measurements were performed at the same time of the day, and at least five individual observations were taken for each animal and averaged.

The STING inhibitor C-176 (MedChemExpress, USA)¹⁷ was dissolved with dimethyl sulfoxide (DMSO). Mice were given 750 nmol C-176 (in 200 μL corn oil) or 200 μL corn oil with 0.54% DMSO (control group) daily through intraperitoneal injection.

Finally, animals were euthanized by CO₂ or isoflurane (2%) narcosis. Ethics approval for the animal experiments in this study was obtained from the Institutional Animal Care and Use Committee of Huazhong University of Science and Technology. Our study conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication, 8th edition, 2011).

2.4. Ultrasound imaging

The mice were anaesthetized with isoflurane (2%) and fixed on the control board. The Vevo 1100 ultrasound system (VisualSonics, Australia) was used to monitor the overall changes of the aorta through B-mode ultrasound imaging. The data analysis was performed on the accompanying Vevo LAB software (VisualSonics, Australia).

2.5. Aortic diameter measurement

As previously described,¹² the aorta was exposed and rinsed with cold phosphate-buffered saline (PBS). Images of the excised aorta were obtained using a microscope at a magnification of ×1.0 (scale bar = 1 mm), and the diameter of different aortic segments was measured with the use of DP2-BSW software (Olympus Life Science Solutions, Center Valley, PA, USA) by two independent, blinded observers.

2.6. Definition of aortic dilatation, aortic aneurysm, and aortic dissection

As previously described,¹² for different aortic segments of Ncf1^fl or Sm22^CreNcf1^fl mice, dilatation and aneurysm represented an aortic diameter ≥1.25 but <1.5 and ≥1.5 times the mean aortic diameter of the same segment in unchallenged mice with an identical genetic background, respectively. Aortic dissection was identified by the presence of haematoma in the aortic wall upon gross examination or by the presence of layer separation, accompanied by false lumen haematoma, within the aortic media or at the medial–adventitial boundary observed upon histologic examination of the aorta. Rupture and premature death were recorded.

2.7. AAD severity classification

Severe AAD was defined as the presence of dissection or rupture.¹² The aneurysm tissue was evaluated by three independent observers who were blinded and agreed on the differences.

2.8. Histology

In euthanized mice, the heart and the aorta were exposed, and the left ventricle was then flushed with cold PBS. Approximately 1–2 mL of yellow latex (Ward’s Natural Science) was injected directly into the left ventricle. Mice were then kept moist at 4°C for 3–4 h to facilitate the setting of the injected latex.¹⁸ Aortic segments were fixed and embedded in paraffin or an OTC compound for staining. Haematoxylin and eosin (H&E; Sigma-Aldrich) and Verhoeff–Van Gieson (Sigma-Aldrich) staining were performed according to the manufacturer’s instructions. The extent of elastic fibre fragmentation was scored on a scale of 1–4 (1, none; 2, minimal; 3, moderate; and 4, severe).

2.9. Western blot

Total protein was lysed in RIPA lysis buffer with protease and phosphatase inhibitor. Membrane and cytosol proteins were isolated with Membrane and Cytosol Protein Extraction Kit (Beyotime, Shanghai, China). Protein concentrations were qualified with the Enhanced BCA Protein Assay Kit (Beyotime, Shanghai, China). Equal amounts of total protein were loaded onto SDS PAGE gels and blotted on polyvinyl difluoride membranes. The membranes were blocked in 5% nonfat milk for 1 h at room temperature and then incubated overnight at 4°C with primary antibodies. Antibodies are listed in Supplementary material online, Table S3. Membranes were probed with species-appropriate HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were detected by enhanced chemiluminescence (GE Healthcare Life Sciences, Germany) and analysed using ImageJ software. Protein expression levels were normalized to β-actin.

2.10. Immunofluorescence staining

The aortic sections were blocked with 10% serum and incubated at 37°C for 30 min to reduce nonspecific staining. Then, the sections were incubated with primary antibody overnight at 4°C, washed with PBS, and incubated with secondary antibody. Antibodies are listed in Supplementary material online, Table S3. Stained slices were examined using a fluorescence microscope (BX51, Olympus, Tokyo, Japan).

2.11. Immunohistochemistry

The aortic sections were placed in a pH 6.0 Citrate Antigen Retrieval Solution (Beyotime, Shanghai, China) and heated for 3 min for antigen repair. After cooling and washing with PBS, the samples were placed in 3% H₂O₂ for 20 min to block endogenous peroxide activity. Then, the samples were further proceeded as described previously. Antibodies are listed in Supplementary material online, Table S3. Images were observed and captured under an optical microscope (ECLIPSE 80i, Nikon, Tokyo, Japan).

2.12. Quantitative real-time polymerase chain reaction

Total mRNA was extracted using a TRIzol reagent. The cDNA was prepared using the HiScript III RT SuperMix for quantitative real-time polymerase chain reaction (qPCR; Vazyme, China). Real-time PCR was performed using the StepOne Plus Real-time PCR System (Applied Biosystems, Foster City, CA, USA) using AceQ qPCR SYBR Green Master Mix (Vazyme, China). The qPCR protocol consisted of 95°C for 30 s, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. The expression levels of genes were quantified using the comparative cycle threshold (ΔΔCt) method and normalized against the Ct of β-actin. The primers used for qPCR are listed in Supplementary material online, Table S4.

2.13. Cells

Human aortic smooth muscle cells (HASMCs, ScienCell, USA) were cultured in smooth muscle cell medium (SMCM, ScienCell, USA) with 10% foetal bovine serum (FBS, Gibco, USA) and 1% penicillin/streptomycin (Gibco, USA). HASMCs were treated with 500 μM H₂O₂ for 24 h to cause DNA damage. Intracellular delivery of 2′3′-cGAMP (InvivoGen, USA) was achieved using Lipofectamine 2000 (Thermo Fisher) diluted in serum-free medium with a ratio of 1:1. The final concentration for cGAMP was 4 μg/mL.¹⁹

HEK293T cells (ScienCell, USA) were cultured in Dulbecco’s modified Eagle’s medium (Life Technology Gibco, USA) supplemented with 10% FBS, 1% L-glutamine, and 1% penicillin/streptomycin.

Human acute monocytic leukaemia cells (THP-1, ScienCell, USA) were cultured in RPMI 1640 (Thermo Fisher) supplemented with 10% FBS. THP-1 cells were treated with 100 ng/mL phorbol 12-myristate 13-acetate (PMA, Sigma-Aldrich) for 72 h to differentiate into macrophages.

All cell lines were assessed for mycoplasma contamination, and the results were negative.

2.14. Macrophage phagocytosis of HASMCs

HASMCs were added into the insert of a 24-well transwell (0.4 μm, Corning) and treated with 500 µM H₂O₂ for 24 h to induce cell injury. THP-1–derived macrophages were pre-plated onto the 24-well plate with the stimulation of PMA. Macrophages and HASMCs were co-cultured without direct cell contact. Untreated HASMCs were used as controls.

2.15. Plasmid and viral transfection

The NCF1 overexpression (NCF1-OE) plasmid was purchased from Sangon Biotech (Shanghai, China). The NCF1 siRNA sequence was listed in Supplementary material online, Table S4. HEK293T cells were transfected with siRNA or plasmid by using Lipofectamine 2000 (Thermo Fisher). NCF1-OE and NCF1 knockout (NCF1-KO) in HASMCs were established using a lentivirus (ObioTechnology, China). HASMCs were cultured in a 24-well culture plate, and polybrene (5 μg/mL) was added to help improve the transfection efficiency. Puromycin (2 μg/mL) was used for screening, and the remaining cells were further cultured and expanded for subsequent experiments.

2.16. ROS detection

Aortic sections were stained with dihydroethylene (DHE; 10 μM, Sigma) in dark conditions at 37°C for 30 min. Images were observed and collected through a fluorescence microscope (BX51, Olympus, Tokyo, Japan).

2.17. Transferase dUTP nick-end labelling staining

The transferase dUTP nick-end labelling (TUNEL) staining was performed using an in situ cell death detection kit (Roche Applied Science, Indianapolis, IN, USA) to investigate SMC apoptosis in mice. The images were examined using a fluorescence microscope (BX51, Olympus, Tokyo, Japan) and captured from three randomly selected views.

2.18. Transmission electron microscope

The suprarenal aortas from various conditions were removed and fixed at room temperature in the electron microscope fixative (Servicebio) for 2 h before being transferred to 4°C for storage and transportation. Sections were observed under a transmission electron microscope (H-7500, Hitachi, Brisbane, CA, USA).

2.19. Annexin V/propidium iodide staining

HASMCs were harvested, washed, and stained using a FITC Annexin V Apoptosis Detection Kit I (BD Biosciences, Franklin Lakes, NJ, USA) in the dark at room temperature for 15 min. The samples were examined immediately within 1 h on a flow cytometer (BD Biosciences), and the data were analysed with FlowJo software (V10).

2.20. RNA-seq and bioinformatic analysis

Total RNA was extracted from the entire aortas of Ncf1^fl and Sm22^CreNcf1^fl mice using a TRIzol reagent. Differentially expressed genes (DEGs) were identified using R package ‘limma’ (version 3.42.2), and DEGs with | log₂FC | at least > 1 and P adjusted <0.05 were considered to be significant. The main R packages and details of data processing are presented in Supplementary material online, Table S5.

2.21. Single-cell RNA sequencing and bioinformatic analysis

The entire excised aortas were digested in an enzyme cocktail composed of Hank’s buffered saline solution containing 2 U/mL Liberase™ (Roche) and 2 U/mL elastase (Worthington) for 1 h at 37°C to make a single-cell suspension. The suspensions from three mice were pooled together as one sample. The construction and data processing processes of a single-cell RNA-seq library are the same as previously described.¹² Subsequent analysis was performed in the R package ‘Seurat’ (version 3.1.4). Subsequently, the data were normalized, scaled, and regressed by mitochondrial gene percentage after filtering out low-quality cells (<200 or >5000 genes per cell and >25% mitochondrial genes in the cell). The principal component analysis was performed to further construct a shared nearest neighbour (SNN) modularity clustering algorithm for cell population deconvolution. The cell clusters were identified by using the ‘FindAllMarkers’ function in Seurat. The marker genes of each cluster are shown in Supplementary material online, Table S6. Within Sting^hi and Sting^low (Ncf1^hi and Ncf1^low) clusters, DEGs between two groups were identified by using a Wilcoxon rank sum test of the ‘FindMarkers’ function in Seurat (avg_logFC > 0.1 and P < 0.05). The main R packages and corresponding versions used for data analysis are listed in Supplementary material online, Table S5.

2.22. Immunoprecipitation

HASMCs were lysed with 500 μL of 150 mM immunoprecipitation (IP) lysis buffer. After centrifugation (12 000 g for 15 min), the supernatant was collected and taken 60 μL for input. The remaining supernatant was incubated with protein A/G agarose beads (Bestchorm Biosciences Co., Ltd, Shanghai) for 1 h and then incubated with antibodies overnight at 4°C. After centrifugation (3500 g for 5 min), the beads were washed thrice with 300 mM IP lysis buffer and twice with 150 mM IP lysis buffer. Finally, the beads were heated at 95°C in SDS loading buffer for 15 min and used for IP.

2.23. Ubiquitination assays

HEK293T cells were transfected with plasmids as described and treated with 10 μM MG132 for 3 h before collection. HEK293T cells were lysed in 90 μL of 150 mM IP lysis buffer and 10 μL of 10% SDS lysis buffer and then denatured by heating at 95°C for 15 min. After sonication, 900 μL of 150 mM IP lysis buffer was added to the lysates. The beads were obtained and used for western blotting as previously described.

2.24. Statistical analysis

Statistical analysis was conducted using the SPSS software (version 26.0) and the GraphPad Prism software (version 9.4.0). The normality of the data was examined using the Kolmogorov–Smirnov test. An independent t-test was used to make comparisons between two groups. For comparisons among three or more groups, one-way analysis of variance (ANOVA) with the Bonferroni post hoc test (with equal variances) or the Tamhane T2 post hoc test (with unequal variances) was used. A two-way ANOVA was used to analyse the differences in two independent variables between groups. The AAD incidence was analysed using the Fisher exact test. Kaplan–Meier survival curves with the log-rank (Mantel–Cox) test were used to investigate mouse survival rates. For all statistical analyses, two-tailed probability values were used. All data are presented as mean ± standard error of the mean. A probability value of P < 0.05 was considered significant. Raw data for statistical analysis are shown in the Supplementary material.

3. Results

3.1. Ncf1 expression was associated with the activation of the STING pathway

To investigate the pathological changes in SMCs in aneurysm tissue, SMC gene expression profiles were analysed using aortic medial SMCs (GSE140947). Gene Set Enrichment Analysis (GSEA) showed increased ROS activation, inflammation, DNA damage, and apoptosis in aneurysm aortic medium SMCs (Figure 1A). Heatmap analysis showed that genes involved in these pathways were highly upregulated (see Supplementary material online, Figure S1A).

Figure 1.

Ncf1 is expressed in injured SMCs and responds to the activation of the STING pathway. (A) DNA damage, ROS, inflammation, the STING, and cell death pathways were significantly activated in medial SMCs from patients with AAD. (B–H) Single-cell transcriptome analysis was performed in entire aortas isolated from ApoE^−/− mice challenged with saline or Ang II (1000 ng/kg/min) infusion for 4 weeks. The suspensions from three aortas were pooled as one sample. (B) Schematic diagram of the single-cell sequencing process (created with FigDraw). (C, D) Uniform Manifold Approximation and Projection (UMAP) plots of the aorta, SMCs, and the Sting^hi and Sting^low clusters. The mean and specific expression values of Sting in the Sting^hi and Sting^low clusters. (E) GSEA of DEGs from the Sting^hi cluster compared to that of the Sting^low cluster. (F, G) Heatmap analysis shows STING, cell death, and ROS-related pathways are upregulated in the Sting^hi cluster. (H) The ROS-related gene Ncf1 showed the greatest changes in the Sting^hi cluster (*). (I) NCF1 has enhanced expression in HASMCs treated with 4 μg/mL 2′3′-cGAMP for 3 h (n = 5). A two-tailed unpaired t-test was used for two-group comparisons (D, I). Data are presented as the mean ± standard error of the mean.

To further examine the mechanism of SMC death in AAD, ApoE^−/− mice were challenged with saline or Ang II (1000 ng/min/kg) for 28 days, and single-cell transcriptome analyses were performed (Figure 1B). The aortic wall comprised 12 cell clusters²⁰ (Figure 1C, see Supplementary material online, Figure S1B), including two SMC clusters (Myh11⁺, Myl9⁺, Tagln⁺, Acta2⁺), two clusters of fibroblasts (Dcn⁺, Gsn⁺, Col1a1⁺, Lum⁺), and two clusters of macrophages (Lyz2⁺, C1qa⁺, Pf4⁺, F13a1⁺; see Supplementary material online, Figure S1C and D). SMCs constituted the predominant cellular clusters in the normal aorta. Conversely, within the context of the aneurysmal aorta, the primary cellular clusters predominantly comprised inflammatory cells, including T cells, B cells, and macrophages (see Supplementary material online, Figure S1E and F).

The STING pathway plays a considerable role in SMC apoptosis and necroptosis.¹² Therefore, all SMCs (Clusters 0 and 4) were divided into Sting^hi and Sting^low clusters based on Sting expression levels (Figure 1D). The ROS, inflammation, DNA damage, and cell death pathways were markedly activated in the Sting^hi cluster (Figure 1E and F, see Supplementary material online, Figure S1G and H). In SMCs, DNA damage and STING pathway activation in aneurysms are mediated by ROS¹²; thus, the expression of ROS-related genes (Cyba, Cybb, Ncf1, and Ncf2) in the Sting^hi and Sting^low clusters was investigated. Ncf1 showed the greatest changes in the Sting^hi cluster (Figure 1G and H). Additionally, NCF1 was also upregulated in cGAMP-stimulated HASMCs (Figure 1I). These data suggest that Ncf1 responds to the activation of the STING pathway.

3.2. NCF1 was significantly expressed in AAD SMCs and might play a protective role

NCF1 was significantly upregulated in human and mouse AAD tissues (Figure 2A and B), especially in areas with severe lesions (see Supplementary material online, Figure S2A). Moreover, immunostaining showed that NCF1 and phosphorylation of NCF1 were highly activated in the aortic media (Figure 2C, see Supplementary material online, Figure S2B and C). Further analysis of NCF1 and a-SMA immunostaining indicated that NCF1 was mainly expressed in SMCs (Figure 2D and E, see Supplementary material online, Figure S2D and E).

Figure 2.

Ncf1 is mainly expressed in SMCs and may play a protective role in AAD. (A–C) NCF1 is highly upregulated in the human (n = 3 in A, n = 8 in B, and n = 4 in C) and mouse (n = 3 in A, n = 6 in B, and n = 4 in C) diseased aortas and expressed in SMCs (D, E). (F) Uniform Manifold Approximation and Projection (UMAP) plots of the Ncf1^hi and Ncf1^low clusters. The mean and specific expression values of Ncf1 in the Ncf1^hi and Ncf1^low clusters. (G–I) Ncf1 prevents apoptosis and DNA damage in the Ncf1^hi cluster but promotes DNA replication. Scale bars: 100 μm (C, D) and 50 μm (E). A two-tailed unpaired t-test was used for two-group comparisons (A, B, F). Data are presented as the mean ± standard error of the mean.

To investigate the specific role of Ncf1 in aortic aneurysm SMCs, all SMCs were divided into Ncf1^hi and Ncf1^low clusters based on their Ncf1 expression levels (Figure 2F). Notably, Ncf1 exhibited the highest increase in SMCs, as opposed to macrophages or neutrophils, following the Ang II infusion. Furthermore, consistent with these observations, Ncf1 expression in the AAD group was higher than that in the control group (see Supplementary material online, Figure S2F). Gene ontology analysis revealed that pathways involved in extracellular structure organization, muscle tissue development, and muscle contraction were remarkably upregulated in the Ncf1^hi cluster (see Supplementary material online, Figure S2G). Further analysis showed that the DNA damage and cell death pathways were significantly downregulated in Ncf1^hi SMCs (Figure 2G–I). Conversely, the DNA replication pathway was upregulated in Ncf1^hi SMCs (Figure 2H). Altogether, these data demonstrate that Ncf1 may play a role in promoting DNA repair in injured SMCs and protecting them from death during AAD progression.

3.3. Conditional NCF1-KO in SMCs promoted the development of Ang II–induced AAD in mice

The role of Ncf1 was evaluated by comparing disease development in Ncf1^fl (see Supplementary material online, Figure S3A and B) and Sm22^CreNcf1^fl mice in a model of AAD. AS was not observed in the murine model.

In challenged Sm22^CreNcf1^fl mice, marked aortic degeneration (Figure 3A), diameter enlargement (Figure 3B–D), lower survival rate (Figure 3E), dilatation (100%), AAD (including aortic aneurysm, dissection, and rupture; 54%), severe AAD (including dissection and rupture; 35%), rupture (23%, Figure 3F), and elastic fibre break (Figure 3H) were observed. Most challenged Ncf1^fl mice survived the experiments without rupture. Moreover, AAD and severe AAD incidences in challenged Sm22^CreNcf1^fl mice were increased in different aortic segments (Figure 3G). Similar results were observed in female mice (see Supplementary material online, Figure S3C–E). Ang II infusion increased blood pressure similarly in Ncf1^fl and Sm22^CreNcf1^fl mice (see Supplementary material online, Figure S3F). Notably, an aneurysm in the ascending aorta of an approximately 43-week-old male Sm22^CreNcf1^fl mouse without any treatment (see Supplementary material online, Figure S3G). Additionally, elastic fibre breakage increased in older Sm22^CreNcf1^fl mice (see Supplementary material online, Figure S3H). Taken together, these data suggest that Ncf1 deletion in SMCs exacerbates Ang II–induced AAD in mice.

Figure 3.

Ang II–induced AAD formation is exacerbated in Ncf1-deficient mice. (A, B) Representative images of excised aortas showing aggravated aortic damage in challenged Sm22^CreNcf1^fl mice. (C) Representative ultrasound images of aortas from challenged mice showing the presence of AAD in the thoracic and suprarenal aortic segments. (D) The mean aortic diameters of various aortic segments are increased in challenged Sm22^CreNcf1^fl mice. The measurements are based on the excised aortas (as shown in A and B). (E) Kaplan–Meier survival analysis showing decreased survival in challenged Sm22^CreNcf1^fl mice. In comparison to challenged Ncf1^fl mice, the incidence of AAD, severe AAD, and rupture in challenged Sm22^CreNcf1^fl mice is significantly increased in overall (F) and different aortic segments (G). (H) H&E and elastic Verhoeff–Van Gieson (EVG) staining of suprarenal abdominal aortic sections showing severely damaged aortic structures in challenged Sm22^CreNcf1^fl mice. Scale bars: 1 cm (A, B) and 50 μm (H). A two-way ANOVA with the Bonferroni post hoc test was used for pairwise comparisons in (D) and (H). The Fisher exact test was used in (F) and (G). Data are presented as the mean ± standard error of the mean. Asc, ascending aorta; Arch, aortic arch; Desc, descending aorta; IR, infrarenal aorta; SR, suprarenal aorta.

3.4. Ncf1 deficiency exacerbated DNA damage and activated the STING pathway during AAD development

To further study the potential mechanisms, bulk RNA sequencing of entire aortas isolated from Ncf1^fl and Sm22^CreNcf1^fl mice challenged with saline or Ang II (1000 ng/min/kg) infusions for 4 weeks was performed (Figure 4A). GSEA showed that Ang II infusion activated several biologic processes, including inflammatory response and cytosolic DNA sensing pathways in Sm22^CreNcf1^fl mice compared to Ncf1^fl mice (see Supplementary material online, Figure S4A). Ncf1 deletion promoted apoptosis and DNA damage; this effect was significantly exacerbated by Ang II (Figure 4B and C, see Supplementary material online, Figure S4B–E). Heatmap analysis showed that key genes involved in the inflammatory response, STING, and cell death–related pathways were highly upregulated in challenged Sm22^CreNcf1^fl mice (Figure 4D). The same results were observed in aortas isolated from Ncf1^fl and Sm22^CreNcf1^fl mice (Figure 4E and F, see Supplementary material online, Figure S4F). Together, these data indicate that Ncf1 deficiency in SMCs aggravated Ang II–induced AAD development via promoting SMC apoptosis and STING pathway activation.

Figure 4.

Ncf1 deficiency exacerbates the activation of the STING and apoptosis pathways. (A–D) Bulk RNA sequencing analysis was performed in aortas isolated from Ncf1^fl mice and Sm22^CreNcf1^fl mice challenged with saline or Ang II (1000 ng/kg/min) infusion for 4 weeks. (A) Schematic diagram of bulk RNA sequencing process (created with FigDraw). (B–D) Ncf1 deficiency promotes the activation of DNA damage, apoptosis, and the STING pathways. (E–G) The STING and cell death–related pathways are highly activated in the aortas of Ang II–challenged Sm22^CreNcf1^fl mice (n = 5 in E and n = 3 or 4 in F and G). Scale bars, 100 μm (H). A two-way ANOVA with the Bonferroni post hoc test was used for pairwise comparisons in (F). A two-tailed unpaired t-test was used for two-group comparisons (G). Data are presented as the mean ± standard error of the mean. FC, Ncf1^fl mice + saline; FA, Ncf1^fl mice + Ang II; SNC, Sm22^CreNcf1^fl mice + saline; SNA, Sm22^CreNcf1^fl mice + Ang II.

3.5. Ncf1 deficiency promoted SMC death and macrophage MMP9 production in mice

To examine the mechanism by which Ncf1 deficiency activates the STING pathway and promotes SMC death, a series of experiments were performed using cultured HASMCs. H₂O₂ was used to trigger DNA damage. H₂O₂ activated NCF1 and promoted its transfer from the cytoplasm to the cell membrane (see Supplementary material online, Figure S5A and B). H₂O₂ markedly increased the phosphorylation of STING, TANK-binding kinase 1 (TBK1), and interferon regulatory factor 3 (IRF3) in HASMCs, indicating the activation of the STING pathway. Moreover, STING-mediated cell death pathways were highly activated, as indicated by increased cleaved caspase 3 and phosphorylation of mixed lineage kinase domain-like (p-MLKL; Figure 5A, see Supplementary material online, Figure S5C). Additionally, the STING and cell death pathways were more significantly activated in NCF1-KO HASMCs; however, this phenomenon was inhibited in NCF1-overexpressing HASMCs (Figure 5B–D, see Supplementary material online, Figure S5D–F). The same result was supported by flow cytometry (Figure 5D).

Figure 5.

Ncf1 deficiency exacerbates SMC apoptosis in vivo and in vitro. H₂O₂ activates STING and cell death pathways (A), which are exacerbated by silencing NCF1 (B) and prevented by overexpressing NCF1 (C). (D) Flow cytometry analysis showing that silencing NCF1 exacerbates H₂O₂-induced apoptosis (n = 4). (E, F) Representative immunostaining of dsDNA and mitochondria (Tomm20) showing that Ncf1 deficiency increases the amount of cytosolic DNA in the aortic media of Sm22^CreNcf1^fl mice. The asterisk indicates nuclear DNA, the arrowhead indicates mitochondrial DNA, and the arrow indicates cytosolic DNA. (G) Transmission electron microscopy images of suprarenal aortas showing increased apoptosis of SMCs (nuclear fragmentation, chromatin condensation, and mitochondrial vacuolation) and necrotic SMCs (swollen nuclei, mitochondria, and endoplasmic reticulum) in challenged Sm22^CreNcf1^fl mice. (H) Terminal deoxynucleotidyl TUNEL-stained images of suprarenal aortas showing increased SMC apoptosis in challenged Sm22^CreNcf1^fl mice. (I) MMP9 expression increases, and the STING pathway is activated in macrophages co-cultured with H₂O₂-treated SMCs. (J) MMP9 expression is increased in macrophages of Ang II–challenged Sm22^CreNcf1^fl mice. Scale bars: 200 μm (H), 50 μm (E, F, H), 20 μm (E, F), 10 μm (E, F), and 5 μm (G), 2 μm (G). A one-way ANOVA with the Bonferroni post hoc test was used in the first two of (D). A one-way ANOVA with the Tamhane T2 post hoc test was used in the last one of (D). A two-way ANOVA with the Bonferroni post hoc test was used for pairwise comparisons in (I). Data are presented as the mean ± standard error of the mean.

There was a marked increase in the amount of cytosolic DNA in the SMCs of challenged Sm22^CreNcf1^fl mouse aortas (Figure 5E). The same result was observed in an older Sm22^CreNcf1^fl mouse with an ascending aortic aneurysm (Figure 5F). The cytosolic DNA then activated the STING pathway and promoted SMC death in Ang II–challenged Sm22^CreNcf1^fl mice, which was supported by transmission electron microscopy and terminal deoxynucleotidyl TUNEL staining (Figure 5G and H). ROS level in the aorta of challenged Sm22^CreNcf1^fl mice did not decrease due to Ncf1 deficiency (see Supplementary material online, Figure S5G), likely because SMCs expressed other ROS-related genes, such as Nox2, Nox4, and Cyba (see Supplementary material online, Figure S5H and I), which compensated for the loss of Ncf1. Moreover, NCF1 was also expressed in endothelial cells and macrophages of the diseased aorta (see Supplementary material online, Figure S5J), and NCF1-KO in SMCs did not affect ROS production in endothelial cells and macrophages. Consequently, ROS levels were elevated in Ang II–challenged Sm22^CreNcf1^fl mice, leading to nuclear and mitochondrial DNA damage in SMCs and the subsequent DNA leakage into the cytosol that activated STING signalling, ultimately inducing SMC apoptosis and necroptosis. Significantly increased matrix metallopeptidase 9 (MMP9) production was observed in macrophages co-cultured with H₂O₂-treated SMCs compared with macrophages co-cultured with untreated SMCs; the MMP9 production was aggravated in macrophages co-cultured with NCF1-KO SMCs and reduced in macrophages co-cultured with NCF1-OE SMCs (Figure 5I). Compared with control aortic tissues, aortic tissues in challenged Sm22^CreNcf1^fl mice also contained significantly elevated MMP9 levels (Figure 5J). These data suggested that Ncf1 deficiency promoted SMC apoptosis, MMP9 production in macrophages, and AAD formation.

3.6. Ncf1 deficiency exacerbated the ubiquitination and degradation of NRF2 and promoted STING expression

NCF1 deletion increased STING expression, whereas NCF1-OE inhibited STING expression (Figure 5B and C). However, the specific reasons for this phenomenon remain unclear. Previous studies have shown that NCF1 physically binds to the NRF2/kelch-like ECH-associated protein 1 (KEAP1) complex, suppresses the ubiquitination of NRF2, and activates NRF2 function.²¹ Moreover, NRF2 negatively regulates STING expression by decreasing STING mRNA stability.¹⁹ Therefore, we speculated whether NCF1-KO would aggravate NRF2 ubiquitination, leading to increased stability and expression of STING. We first confirmed that NCF1 could bind to NRF2 in HASMCs (Figure 6A). Subsequently, whether the absence of NCF1 affects NRF2 ubiquitination by KEAP1 was examined. HEK 293T cells were transfected with plasmids encoding HA-NRF2, FLAG-KEAP1, Myc-Ubiquitin, or EGFP-NCF1 (see Supplementary material online, Figure S6A–D) in the presence or absence of the proteasome inhibitor MG132. HA-NRF2 was precipitated using an anti-HA antibody and analysed by immunoblotting for Myc-Ubiquitin to reveal NRF2 ubiquitination. NRF2 ubiquitination was evident in the presence of KEAP1 and MG132 (Lane 2), which was decreased by NCF1 (Lane 3) and exacerbated by the absence of NCF1 (Lane 4), suggesting that NCF1 interfered with NRF2 ubiquitination (Figure 6C). In addition, the lack of NCF1 enhanced NRF2 ubiquitination and promoted STING expression (Figure 6C). Consistently, the expression of Nrf2 and the Nrf2-dependent gene Ho-1 decreased in the aortas of challenged and unchallenged Sm22^CreNcf1^fl mice; however, STING expression increased (Figure 6C and D). Immunostaining further suggested that the Ncf1 deletion facilitated STING expression (Figure 6E). Increased STING expression further activated the STING pathway, promoting ISG15 expression and AAD formation²² (Figure 6E). Together, these data demonstrate that Ncf1 deficiency in SMCs aggravates Ang II–induced AAD formation by promoting NRF2 ubiquitination and degradation and inducing STING expression.

Figure 6.

Ncf1 deficiency promotes ubiquitination and degradation of nuclear factor erythroid 2-related factor 2 (NRF2) and induces STING expression. (A) IP and Co-IP indicate that NCF1 physically binds to NRF2 in HASMCs. (B) GSEA showing Ncf1 deficiency promotes the biological process associated with ubiquitin-mediated proteolysis. Co-IP assays of the ubiquitination of NRF2 after different treatments. NCF1 deletion promotes the ubiquitination of NRF2 (C) and increases STING expression in Sm22^CreNcf1^fl mice (D, E; n = 3 or 4). Scale bar, 100 μm (E). A two-way ANOVA with the Bonferroni post hoc test was used for pairwise comparisons in (D). A two-tailed unpaired t-test was used for two-group comparisons (E). Data are presented as the mean ± standard error of the mean.

3.7. The STING inhibitor C-176 prevented AAD development in Ang II–infused Sm22^CreNcf1^fl mice

Having established the importance of Ncf1 deficiency in facilitating AAD development through the upregulation of STING expression, we conducted an investigation to ascertain whether AAD progression could be prevented through pharmacological inhibition of STING activation in Sm22^CreNcf1^fl mice. C-176 is a strong and covalent mouse STING inhibitor.¹⁷Sm22^CreNcf1^fl mice were challenged with Ang II infusion and also received either C-176 (750 nmol per mouse in 200 μL corn oil) or corn oil with DMSO (control) during Ang II infusion. C-176 treatment showed a better preserved aortic structure (Figure 7A), reduced aortic enlargement (Figure 7B–D), improved survival rate (Figure 7E), significantly reduced incidences of AAD (6%, P = 0.0007; Figure 7F), better aortic structural integrity (Figure 7G), and less elastic fibre breakage (Figure 7G) in Sm22^CreNcf1^fl mice. Similar results were observed in female mice (see Supplementary material online, Figure S7A–D). Moreover, C-176 treatment significantly reduced the amount of cytosolic DNA in SMCs (see Supplementary material online, Figure S7E), the activation of the STING pathway (Figure 7H, see Supplementary material online, Figure S7F and G), the number of dead SMCs (Figure 7I and J), MMP9 production in macrophages (Figure 7K), and the accumulation of macrophages and neutrophils (see Supplementary material online, Figure S7H) in Ang II–challenged Sm22^CreNcf1^fl mouse aortas. These data clearly indicate a protective effect of C-176 against aortic degeneration and AAD formation.

Figure 7.

Prevention of AAD development and SMC death in Ang II–challenged Sm22^CreNcf1^fl mice treated with the STING inhibitor C-176. (A) Representative images of excised aortas and ultrasound images showing C-176 reduces aortic damage in Ang II–challenged Sm22^CreNcf1^fl mice. (B–D) C-176 treatment significantly reduces the mean diameter of different aortic segments in Ang II–challenged Sm22^CreNcf1^fl mice. The measurements are based on the excised aortas (as shown in A). (E) Kaplan–Meier survival analysis showing C-176 increases the survival of Ang II–challenged Sm22^CreNcf1^fl mice. (F) C-176 significantly reduces the overall incidences of AAD, serve AAD, and rupture in Ang II–challenged Sm22^CreNcf1^fl mice. (G) Representative H&E and elastic Verhoeff–Van Gieson (EVG) staining of suprarenal abdominal aortic sections showing better aortic structure in Ang II–challenged Sm22^CreNcf1^fl mice treated with C-176 (n = 5). C-176 reduces the activation of the STING and cell death pathways (H), SMC death (I, J), and MMP9 production of macrophages (K) in Ang II–challenged Sm22^CreNcf1^fl mice. Scale bars: 1 cm (A), 200 μm (I), 50 μm (G, I), and 5 and 2 μm (J). A two-tailed unpaired t-test was used in (G) and (I). A two-way ANOVA with the Bonferroni post hoc test was used for pairwise comparisons in (C) and (D). The Fisher exact test was used in (F). Data are presented as the mean ± standard error of the mean.

4. Discussion

AAD is a highly dangerous and fatal disease; thus, identifying therapeutic targets to prevent AAD progression is critical. Previous studies have shown the importance of ROS in AAD progression.¹¹ We found that Ncf1 deletion in SMCs aggravates AAD development and progression in a murine AAD model, and C-176 could rescue AAD in Ncf1-deficient mice. Our mechanistic studies revealed that Ncf1 deficiency activated the STING pathway by promoting NRF2 ubiquitination and degradation, which activated cell death pathways, causing SMC death. The DNA released from dead SMCs engulfed by macrophages increased MMP9 expression in macrophages, contributing to AAD development.

Previous studies have demonstrated that Ncf1 plays a pro-inflammatory and pro-apoptotic role and regulates ROS production against invading pathogens.¹⁴ Therefore, imbalanced Ncf1 expression is closely associated with several diseases, including AAD,²³ hypertension,²⁴ and autoimmune diseases.¹⁴ Notably, previous studies have found that Ncf1 deletion attenuates Ang II–induced AAD formation in ApoE^−/− mice.²³ Herein, for the first time, we report that NCF1-KO in SMCs promoted Ang II–induced AAD formation. We speculated that NCF1-KO in SMCs caused SMCs to lose protection and be attacked by ROS produced by endothelium and macrophages, leading to increased apoptosis. This may be an endogenous mechanism that exacerbates AAD progression.

Previous studies have shown that Ncf1 deficiency is associated with autoimmune diseases¹⁴ and accompanied by an ISG overexpression,¹⁶ which is related to AAD development.²² This is not the first time that ROS deficiency has been reported to aggravate AAD occurrences.²⁵ To obtain direct evidence supporting the protective effect of Ncf1 in AAD, single-cell transcriptome analysis was performed. Our data suggested that Ncf1 expression in SMCs increased in response to STING pathway activation. Recently, the STING pathway has been implicated in multiple inflammatory diseases, including autoimmune diseases,²⁶ vascular inflammation and degeneration,²⁷ and AAD.¹² Additionally, upon further analysis, it became evident that heightened Ncf1 expression in SMCs was conductive to self-DNA replication and repair processes, concomitant with the suppression of pathways associated with inflammatory responses, DNA damage, and apoptosis induction.

We also observed that Ncf1 deficiency in SMCs did not reduce ROS levels in the aortic wall after Ang II stimulation. To further investigate the underlying mechanism, bulk RNA sequencing was performed, showing that NCF1-KO promoted the activation of DNA damage- and apoptosis-related pathways. Consistently, the amount of cytosolic DNA was markedly increased in the aortas of Ang II–challenged Sm22^CreNcf1^fl mice, which subsequently activated the STING signalling pathway. The STING-IRF3 pathway reportedly promotes apoptosis and necroptosis through type I IFN signalling,^28,29 consistent with the in vivo changes caused by Ncf1 deficiency.¹⁶ Coherently, SMC apoptosis was significantly increased in the aortas of Ang II–challenged Sm22^CreNcf1^fl mice, as corroborated by a series of in vitro experiments conducted in HASMCs. These results deepen our knowledge of the role of Ncf1 in AAD pathology.

This study offers novel insights into the regulation of STING-mediated apoptosis by targeting Ncf1. Our data indicated that NCF1-KO promoted the ubiquitination and degradation of NRF2. Recent studies have shown that NRF2, a negative regulator of STING, mediates a decrease in the stability of STING mRNA.¹⁹ Furthermore, our study suggested that NCF1-KO promoted STING expression, subsequently activating the STING pathway and related apoptosis pathways. Previous studies have reported that SMC-derived DNA is engulfed by macrophages, activating the STING-TBK1-IRF3 pathway, which subsequently promotes MMP9 production and AAD occurrence.¹² This SMC–macrophage crosstalk has also been observed in Ang II–challenged Sm22^CreNcf1^fl mice. Finally, pharmacologically inhibiting STING activation with C-176 alleviated STING pathway activation, SMC death, MMP9 production in macrophages, and AAD formation.

We provided evidence that Ncf1 deficiency exacerbated AAD development in Ang II–challenged Sm22^CreNcf1^fl mice; however, we could not determine the reason AAD occurred in unchallenged older Sm22^CreNcf1^fl mice. We speculate it occurred through the STING pathway or after AAD formation. In addition, further exploration is needed on how Ncf1 deficiency promotes DNA damage and apoptosis without Ang II infusion. Finally, whether Ncf1 mutations exist in patients with AAD and whether the STING pathway inhibitors are safe and effective in patients with AAD warrant further investigation.

In summary, this study reveals a key role of Ncf1 in preventing SMC apoptosis and AAD development, suggesting a dual role for ROS in AAD. This study may help identify a potential target for pharmacologic therapy for AAD.

Translational perspective.

This work reveals the role of neutrophil cytoplasmic factor 1 (Ncf1) in aortic AAD, outlining the importance of Ncf1 in regulating apoptosis and the STING pathway and indicating the protective effect of ROS in AAD. This provides new ideas and directions for the clinical diagnosis and pharmacologic therapy of AAD.

Supplementary material

Supplementary material is available at Cardiovascular Research online.

Authors’ contributions

J.X. and Pi.Y. conceptualized and supervised the study. H.L., Pe.Y., and S.W. performed the main experiments and acquired data. H.L., X.C., L.J., and S.L. performed single-cell RNA sequencing and bulk RNA sequencing with the help of X.X., Shu.C., and Sha.C. H.L., Shu.C., and Pe.Y. conducted the animal experiments. L.J. and X.X. collected and processed clinical samples. H.L., Pe.Y., S.L., and Shu.C. analysed experimental data and bioinformatic data. H.L., S.W., and Shu.C. drafted the manuscript. X.C., Pi.Y., and J.X. revised the manuscript for important intellectual content. All authors have read and approved the final version of this manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (81730015, 81974048, 81800296, 82370492, and 82170504) and Fundamental Research Funds for the Central Universities (2021GCRC037).

Data availability

The scRNA-seq data are available in Gene Expression Omnibus under GSE239620 (ApoE^−/−). The RNA-seq data are under GSE237188 (Sm22^CreNcf1^fl). The RNA-seq data of human AAD tissues are available under GSE140947.