Skip to main content
BMC is moving to Springer Nature Link. Visit this journal in its new home.
Download PDF

THBS1high macrophages promote aortic aneurysm formation by orchestrating an inflammatory and smooth muscle cell phenotypic transition program in vascular Behçet’s disease

BMC Medicine volume 23, Article number: 631 (2025) Cite this article

Abstract

Background

Vascular Behçet’s disease (VBD), a representative autoimmune inflammatory disorder, is a leading cause of mortality in BD, primarily due to the formation of inflamed aneurysms. However, both the histopathological features and the underlying mechanism of VBD remain elusive. Identifying specific VBD-associated cell populations and unravel their participation within the aorta microenvironment is crucial for deciphering the pathogenesis of VBD.

Methods

We conducted immunohistochemistry and single-cell RNA sequencing to comprehensively characterize the ascending aorta in VBD patients. Immunofluorescence staining, bulk RNA sequencing, and functional co-culture system were employed to investigate the phenotypic characteristics of pathological subsets and potential cell–cell interactions. A CaCl2-induced humanized VBD model was established to validate the transcriptomic and cellular results.

Results

We identified an expansion of THBS1high macrophages in VBD, particular in active VBD. These THBS1high macrophages exhibited a proinflammatory profile and promoted the phenotypic transition of smooth muscle cells. Mechanistically, transcription factor ETS2 mediated the proinflammatory development of this macrophage subset, while tumor necrosis factor-α (TNF-α) upregulated THBS1 expression in macrophages. The adoptive transfer of THBS1high macrophages exacerbated vascular inflammation and aneurysm formation in CaCl2-induced humanized VBD mice, which could be mitigated by TNF inhibitor. Clinically, plasma THBS1 levels were positively correlated with disease activity and inversely associated with the clinical benefits of TNF inhibitor, both observed at the systemic level.

Conclusions

Overall, our findings underscore the pivotal role of THBS1high macrophages in vascular degeneration, highlighting the therapeutic potential of anti-TNF therapy in VBD and the THBS1’s potential as a biomarker for clinical evaluation.

Peer Review reports

Background

Behçet’s disease (BD) is an autoimmune inflammatory disorder characterized by recurrent oral and genital ulcers, uveitis, and skin lesions, with the potential involvement of the cardiovascular, neurological, and gastrointestinal systems [1]. Arterial involvement is a hallmark of vascular BD (VBD), presenting as aneurysms, aortic valvular insufficiency, and arterial thrombotic occlusion [2, 3]. These complications can be fatal, particularly when large arterial aneurysms rupture, with a 7-year mortality rate of approximately 25% [4, 5]. Despite receiving advanced medical treatment with corticosteroids and immunosuppressants, approximately one-third of patients experience relapses or suffer from adverse effects [6]. Postoperative complications such as anastomotic leakage, false aneurysm recurrence, and thrombosis are also frequently observed after vascular surgery, underscoring the critical need to elucidate the underlying mechanisms and develop novel therapeutic strategies for VBD [3, 7, 8].

Rapidly evolving high-throughput transcriptomic and epigenomic technologies over recent decades have immensely helped in uncovering the immune landscape in peripheral blood mononuclear cells (PBMCs) and inflamed skin lesions of patients with BD [9, 10]. However, due to the challenges in aortic tissue accessibility, the application of these technologies to VBD remains in its infancy. A comprehensive profiling of cellular heterogeneity and intercellular communication in the aortic microenvironment of BD remain poorly understood.

Current knowledge suggests VBD is one kind of chronic inflammatory disorder, highlighting the dynamic interplay between immune and non-immune cells. Thrombogenesis in VBD is induced by activated leukocytes recruited to the sites of endothelial injury, inducing platelet activation, endothelial dysfunction, and impaired fibrinolysis, ultimately leading to thrombo-inflammation [3, 11]. Nevertheless, the pathogenetic mechanisms underlying the initiation and progression of the inflamed aortic aneurysms (AA) in VBD remain to be elucidated.

Studies on non-inflamed AA have shown that macrophage-smooth muscle cell (SMC) interactions are critical for modulating inflammation and tissue remodeling. M1-polarized macrophages could upregulate the inflammatory response and matrix metalloproteinase (MMP) expression in SMCs [12, 13]. In addition, TNF signaling governs the phenotypic transition of SMCs, contributing to the disruption of aortic tissue homeostasis [14]. Consistently, recent studies demonstrated that macrophages in patients with BD tend to polarize toward the M1 phenotype, secreting high levels of proinflammatory cytokines [15, 16]. Therefore, further investigation into the distinct inflammatory macrophage subsets and their effects on SMCs in VBD is essential.

The current study, for the first time, applies single-cell RNA sequencing (scRNA-seq) and performs pathological validation of patient specimens in an attempt to delineate the landscape of the arterial lesions in VBD. We identify an expansion of pro-inflammatory THBS1high macrophage subset, which promotes the phenotypic transition of SMCs and the formation of AA in VBD. Our findings contribute to a broader understanding of proinflammatory mechanisms in autoimmune diseases and highlight the innovative clinical implication of TNF inhibitors in the treatment of VBD and potentially other inflamed aneurysms.

Methods

Patients and samples

Peripheral blood and aortic tissues were obtained from patients with BD who were admitted into Zhongshan Hospital Affiliated to Fudan University and fulfilled the 2014 International Criteria for Behçet’s Disease [17]. VBD was diagnosed according to the clinical presentation and imaging findings (ultrasound and computed tomography) of the patients. The overall BD activity was assessed using the Behçet’s Disease Current Activity Form (BDCAF) based on patient’s symptoms over the past 4 weeks [18]. Active BD was identified when the BDCAF score was ≥ 1 or increased levels of inflammatory parameters were identified. Demographic characteristics, clinical manifestation, laboratory data, imaging data, associated-analytical parameters, and treatments of patients were collected.

Animals

Six to eight weeks male NOD/ShiLtJGpt-Prkdcem26Cd52Il2rgem26Cd22/Gpt (NCG) mice (GemPharmatech, Nanjing, China) were used to generate PBMC humanized mice models. Our study exclusively examined male mice for the character of higher prevalence of vascular involvement in BD and less sex hormone variations. The generalizability of these findings to female mice remains to be determined in the future studies. All mice were housed at 24 ± 2 °C and 40 ± 5% humidity under a 12-h light/dark cycle. Animal studies were performed in compliance with the National Institutes of Health Guidelines on the Care and Use of Laboratory Animals (National Institutes of Health Publication, 8th edition, 2011). To establish a CaCl2-induced AAA model, the mice were anesthetized using 1–2% isoflurane. After isolating the abdominal aorta between the renal artery and iliac artery bifurcation, cotton gauze containing 0.5 mol/L of CaCl2 was placed directly on the external surface of the aorta for 15 min. To establish an angiotensin II (AngII)-induced AAA model, an Alzet osmotic minipumps (Model 2004, Durect Corporation, USA) containing AngII (1000 ng/kg/min, Sigma, USA) was subcutaneously implanted into the flank of each mouse for 28 days. Thirteen days after the surgery, 10 μg of lipopolysaccharide was administered subcutaneously to the mice. To establish human immune system reconstituted mice models, the mice were injected with PBMCs (10 million per mouse) from patients with BD and autologous monocyte-derived macrophage (1 million per mouse) intravenously through the tail vein on the next day (D14). Human CD45 + cells in peripheral blood and human macrophage infiltration in the artery were assessed by flow cytometry. Mice with over 25% hCD45 + cells in their peripheral blood were classified as humanized. Mice receiving PBMCs and macrophages from the same patient were randomly allocated into groups.

For in vivo testing the role of THBS1high macrophages, the macrophages were transfected with the control lentivirus or THBS1 overexpression lentivirus before adoptive transfer. For the ADA group, adalimumab (3 mg/kg, A2010, Selleck, USA) was applied through intraperitoneal injection every 3 days simultaneously starting at D15 to evaluate the in vivo efficacy of TNF inhibitor. The dose of adalimumab was selected based on previous studies involving murine models of rheumatoid arthritis, uveitis, and retinal degeneration. No significant adverse effects, including opportunistic infections, were observed in the animals treated with adalimumab [19,20,21,22]. Four weeks after the operation, the abdominal aortae were harvested for morphological or histological analyses. Vascular ultrasound and micro-PET/CT were performed before the mice were sacrificed.

Cell preparation, culture, and treatment

The human acute monocytic leukemia cell line THP-1 (Cat. QC-h213, Qincheng Bio, China) was cultured in the RPMI-1640 medium containing 10% fetal bovine serum (FBS, Gibco). The THP-1 cells differentiated into M0 macrophages after 48 h incubation in 100 ng/mL of phorbol 12-O-tetradecanoylphorbol-13-acetate (PMA, Sigma-Aldrich, USA).

The PBMCs were separated from the EDTA-anticoagulated blood samples through density gradient centrifugation using Ficoll-Paque (Cytiva, Marlborough, MA). The primary monocytes were positively selected with anti-CD14 microbeads (Miltenyi Biotech, USA) from PBMCs and then cultured in the complete RPMI-1640 medium containing 40 ng/mL of macrophage colony-stimulating factor (M-CSF, R&D Systems, USA) for 6 days. The differentiation medium was refreshed every 3 days. Mϕ were detached from 6-well plates using StemPro™ Accutase™ Cell Dissociation Reagent (Gibco, USA).

Human aortic smooth muscle cells (HASMCs) were purchased from Cellverse Co., Ltd. (Cat. HUM-iCell-c010, China) and cultured in a complete SMCM containing SMC growth supplement, 10% FBS, and 1% penicillin/streptomycin (Cellverse Co., Ltd., Cat. iCell-c010-002 h, China). An indirect macrophage–HASMC co-culture system was established using 0.4-µm transwell chambers (Corning, USA). The upper chamber contained THP-1-derived macrophages with THBS1 overexpression or knockdown, while the lower chamber was plated with the HASMCs. After 48 h of co-culture, the SMCs were harvested for immunoblotting or immunofluorescence assays.

To generate single-cell suspension of aortic tissues, freshly collected arteries were cut into 1–2-mm small pieces and digested with 1 mg/mL of collagenase type I (C8140, Solarbio, China), 1 mg/mL of collagenase type IV (C8160, Solarbio, China), 0.744 U/mL of elastase type (E1250, Sigma, USA), and 20 U/mL of DNase I (10608ES25, Yeasen, China) in Hanks’ balanced salt solution (14,025,092, Thermo Fisher Scientific, USA) for 50–60 min at 37 °C with gentle shaking. The digestion solution was then filtered through a 40-µm cell strainer to remove tissue debris. All cultures were maintained in a 37 °C incubator with 5% CO2.

Pathological examination

For the pathological examination of the aortic specimens, all slides were scanned using Pannoramic 250FLASH (3DHISTECH, Hungary) and were reviewed by two pathologists blinded to clinical information. The disruption of the medial layer was graded semi-quantitatively as follows under Verhoeff Van Gieson staining: non-disrupted, mild (< 1/3), medium (1/3 to 2/3), and high (> 2/3). The extent of overall inflammatory infiltration was estimated semi-quantitatively under hematoxylin and eosin (H&E) staining: absent ( −), rare ( +), medium (+ +), and numerous (+ + +) [23]. The number of macrophages in the three layers was counted from 10 randomly selected areas (approximately 30 mm2) with inflammatory infiltration using Fiji (Version 2.0.0-rc-69, National Institutes of Health) under × 400 magnification fields.

Single-cell RNA sequencing (sample and library preparation)

Aortic tissues for scRNA-seq were obtained from six patients with VBD and three heart transplant donors identified as controls. The harvested tissues were cut into small pieces and then placed in an enzyme cocktail for 40 min at 37 °C. After filtering through a 40-μm cell screen and centrifugation at 300 × g for 5 min, we collected and resuspended the precipitate. The resulting single-cell suspension was subjected to 10 × Genomics Chromium Next GEM Single Cell 3ʹ Reagent Kits v3.1 (10 × Genomics, USA), according to the manufacturer’s instruction for computer and library preparation. The sequencing was performed on an Illumina Nova 6000 platform with 150-bp paired-end reads.

Statistical analysis

Statistical analyses were conducted using GraphPad Prism 10.2.3 (GraphPad Software, Inc., USA) and R (V4.3.2). The data were summarized as mean ± standard deviation (SD). We employed the Shapiro–Wilk normality test to assess the normality of data distribution. For comparisons between two groups with normally distributed data and equal variances, an independent samples t-test was applied. A paired t-test was used to analyze the plasma THBS1 levels in patients with VBD at different time points. One-way analysis of variance (ANOVA) was performed for multiple group comparisons. The non-parametric Wilcoxon test or Mann–Whitney U test was employed to analyze non-normally distributed continuous data. Chi-squared tests or Fisher exact tests were applied to categorical variables. A two-tailed P < 0.05 was considered as statistically significant. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

For additional methodological details regarding sequencing data analysis and molecular biology experiments, please refer to Additional file 1: Supplementary methods.

Results

Histopathological characterization of aortic lesions in BD

We first conducted a systematic histopathological analysis of ascending aortic specimens collected from 18 VBD patients with aortic dilation and aneurysm and 16 control subjects (8 noninflammatory aneurysm cases, due to atherosclerosis and inherited aortopathy, and 8 healthy donor controls; Additional file 2: Table S1). While both VBD and noninflammatory aneurysm groups exhibited characteristic medial degeneration patterns including elastic fiber fragmentation, collagenous matrix deposition, and mucoid matrix accumulation (Additional file 3: Fig. S1A-C), quantitative analysis revealed significantly greater medial layer destruction scores in VBD specimens (P = 0.02; Additional file 2: Table S1). Distinctive inflammatory features were observed in VBD aortae, demonstrating increased infiltration by mixed leukocyte populations (CD4+ T cells, CD20+ B cells, CD68+ macrophages, CD66b+ neutrophils; P < 0.001; Additional file 2: Table S1, Additional file 3: Fig. S1D, S2A-D). Notably, CD68+ macrophage infiltration demonstrated significantly increased macrophage infiltration across all vascular layers (intima/media/adventitia) in BD patients compared to non-inflamed AA (which showed predominant accumulation in the media and adventitia) and HC (Additional file 3: Fig. S2E).

Single-cell landscape of the ascending aortic wall in BD

To mechanistically investigate immune-mediated vascular injury in VBD, we performed scRNA-seq on the ascending aortic tissues collected from three patients with active VBD (BD_Act, defined as BDCAF ≥ 1 or with increased levels of inflammatory parameters), three patients with VBD in remission (BD_Rem, BDCAF = 0 and with normal level of inflammatory parameters), and three age-/sex-matched non-BD controls (HC) from heart transplant donors (Fig. 1A). The clinical characteristics of the subjects enrolled for scRNA-seq are available in Additional file 2: Table S2. Apart from systemic activity, to further link our subsequent findings to aortic pathology, we also incorporated detailed aortic clinical data, including diameter, z-score, imaging changes, and histological lymphocytic infiltration in Additional file 2: Table S2. Notably, while discrepancy was observed between systemic activity (BDCAF ≥ 1) and aortic diameter/z-score/histology, radiological progression perfectly stratified active VBD cases (all demonstrating aneurysm enlargement/newly onset) from remission (all stable). These findings suggested that imaging dynamics may better reflect active vascular injury, complementing the correlation between scRNA-seq findings with systemic activity. After stringent quality control, we established an ascending aorta single-cell atlas for BD (Fig. 1B, Additional file 3: Fig. S3A-B). Twelve major cell types were identified based on their expression of classic cell markers (Fig. 1C,D, Additional file 3: Fig. S3C). Compared with HC, we noted an increased infiltration of T cells, macrophages, neutrophils, and B/plasma cells, corroborating histopathological findings (Fig. 1E, Additional file 3: Fig. S3D). Cross-group differential expression analysis (|log2FC|> 0.25, adjusted P < 0.05) identified SMCs (835 differentially expressed genes, DEGs) and macrophages (630 DEGs) as the most transcriptionally perturbed populations in VBD (Fig. 1F, Additional file 3: Fig. S3E), indicating the pivotal role of macrophages and SMCs in the pathogenesis of VBD.

Fig. 1
figure 1

Single-cell profiling of ascending aortic tissues obtained from BD patients and controls. A Overview of the experimental workflow. B UMAP visualization of the 12 major cell types, including T cells, natural killer cells, macrophages, dendritic cells, neutrophils, B cells, plasma cells, mast cells, fibroblasts, smooth muscle cells, mesenchymal cells, and endothelial cells in 68,878 cells from six patients with BD (42,806 cells) and three controls (26,072 cells). C Stacked violin plot showing the expression of marker gene for each cluster identified in aortic tissues. D UMAP plot depicting the expression of the canonical markers of the main cell lineages. E Bar plot showing the proportions of 12 cell types among total cells in each group and individual patients. F UMAP plot showing the number of differentially expressed genes between BD and HC across major clusters. UMAP, Uniform manifold approximation and projection; BD, Behçet’s disease; HC, healthy control; HASMC, human aortic smooth muscle cell

Full size image

To decipher the heterogeneity of enriched macrophages and their role in medial destruction in patients with BD, we sub-clustered macrophages/dendritic cells (DCs) and confirmed two conventional DCs clusters (cDC1 and cDC2) and eight macrophage subtypes (Monocyte-like (LYZ+S100A8+); THBS1high macrophage; classically activated CCL4high macrophage; tissue-resident homeostatic LYVE1high macrophage; C1Qhigh phagocytic macrophage; SPP1high matrix-remodeling macrophage; ISGhigh Interferon-responsive macrophage and proliferating macrophage) (Fig. 2A,B, Additional file 3: Fig. S4A-C). As indicated, THBS1 exhibited predominant expression in THBS1high macrophages and SPP1 was mainly expressed in SPP1high macrophages, validating the accuracy of macrophage subcluster identification.

Fig. 2
figure 2

Identification of macrophage subpopulation involved in the activation of VBD A UMAP visualization of macrophage/DC subclusters, color-coded according to cell type. B Dot plot depicting the expression of marker genes in each cluster. C Volcano plot showing the DEGs between THBS1high macrophages and monocytes. D Boxplot comparing the proportions of macrophage subtypes between BD patients (n = 6) and healthy controls (n = 6). E Boxplot comparing the proportions of THBS1high macrophage and SPP1high macrophages in the aortas of patients with active BD, patients with remission BD, and controls. F Representative immunofluorescence images of CD68 (green), THBS1 (red), and DAPI (blue) in the aortas of healthy controls and patients with BD, and the corresponding quantification of the percentages of THBS1+ macrophages from all macrophages (n = 5 per group). Scale bars, 25 μM. G Transcriptional levels of THBS1 in the aortas of controls and BD patients (n = 6 per group). H,I Immunoblot analysis and quantification of THBS1 expression in primary monocyte-derived macrophages from healthy controls (n = 6) and BD patients (n = 7) (H), and in primary monocyte-derived macrophages from healthy controls cultured with serum from healthy controls or BD patients (n = 6 per group, I). J THBS1 levels in the plasma of active vascular BD patients (n = 74) and healthy controls (n = 39). K THBS1 levels in the plasma of patients with vascular BD (n = 80) with different BDCAF. P-values were calculated by using Wilcoxon test in D and E, by using Mann–Whitney U test in G and J, by using two-tailed Student’s t test in F, H, and I, by using Kruskal–Wallis test with Dunn’s multiple comparison test in K. UMAP, Uniform manifold approximation and projection; DEGs, differentially expressed genes; BD, Behçet’s disease; DAPI, 4′,6-diamidino-2-phenylindole; BDCAF, Behçet’s Disease Current Activity Form. ns, not significant, **P < 0.01, and ***P < 0.001.

Full size image

The THBS1high subset demonstrated a transitional phenotype between circulating monocytes and tissue macrophages, characterized by downregulated monocytic markers (LYZ: 1.6-fold decrease vs monocyte-like, adjusted P < 0.001; S100A8: 2.6-fold decrease vs monocyte-like, adjusted P < 0.001) and upregulated effector molecules: inflammasome components (NLRP3: 2.5-fold, adjusted P < 0.001), chemotactic signals (CXCL8: 2.5-fold, adjusted P < 0.001), matrix regulators (MMP19: 2.0-fold, adjusted P < 0.001), and angiogenic factors (VEGFA: 2.8-fold, adjusted P < 0.001) (Fig. 2C). The expression level of THBS1 in macrophages was significantly elevated in BD, with a fold change of 2.7 and adjusted P < 0.001 (Additional file 3: Fig. S4D). THBS1high macrophages also exhibited remarkably increased enrichment in BD (Additional file 3: Fig. S4E). We further integrated three control human samples of the ascending aorta from a public resource (GSE155468) into our data (Additional file 3: Fig. S5A-C). Consistently, the proportions of THBS1high macrophages were significantly elevated in patients with BD (Fig. 2D). Moreover, the percentages of THBS1high macrophages were significantly high in patients with active BD, while those of SPP1high macrophages, which were preferentially enriched in remission BD, showed a significant decrease in patients with active BD (Fig. 2E).

The enrichment of THBS1high macrophages in the aortas of patients with BD was further verified by immunofluorescence staining. Quantitative analysis revealed that the proportions of THBS1+ macrophages were significantly higher in the aortas of patients with BD compared with that of matched healthy controls (Fig. 2F). Furthermore, the mRNA levels of THBS1 were notably higher in the aortae of patients with VBD (Fig. 2G). Primary monocyte-derived macrophages (MDMs) of patients with VBD exhibited a higher THBS1 expression level (Fig. 2H). Accordingly, stimulating MDMs with the serum of patients with VBD increased the THBS1 expression level too (Fig. 2I).

To assess the clinical relevance of THBS1, we investigated the plasma level of THBS1 in 39 healthy controls and 80 VBD patients (Additional file 2: Table S3). The VBD cohort included 6 remission VBD and 74 active VBD (those with BDCAF ≥ 1). Among the active VBD patients, we further stratified them by disease activity: BDCAF 1–2 (n = 21), 3–4 (n = 42), and ≥ 5 (n = 11). The results demonstrated that patients with active VBD exhibited higher THBS1 levels than those in age-matched healthy controls (Fig. 2J). Elevated plasma THBS1 levels were found in patients with VBD who exhibited higher BDCAF scores (Fig. 2K). The positive correlation between the plasma THBS1 level and disease activity suggested that circulating THBS1 could serve as a potential indicator to reveal the active systemic inflammatory status of VBD, though its specific association with aortic involvement requires further investigation.

THBS1high macrophages possessed a pro-inflammatory and hypermetabolic phenotype

To further explore the distinct functions of these macrophage subtypes, we scored pathway activities between different subtypes. “TNFα signaling via NFκB signaling,” “IL6 JAK STAT3 signaling,” and “Inflammatory response” were significantly activated in THBS1high macrophages (Fig. 3A). We became increasingly interested in THBS1high macrophages because of its high frequency and inflammatory property in BD.

Fig. 3
figure 3

THBS1high macrophages demonstrated pro-inflammatory characteristics. A Heatmap illustrating the pathway activity scored by gene set variation analysis between macrophage subtypes. B Bar plot showing the gene ontology terms enriched in THBS1high macrophages. C Pseudotime reconstruction and developmental trajectory of macrophages, colored by macrophage subtype (upper) and pseudotime (lower). D DEGs (rows) are hierarchically clustered into three distinct profiles along the pseudotime (columns). Heatmap depicting the expression of pseudotime-dependent genes and representative identified genes of each cluster are labelled. The arrow indicates a possible differentiation direction (left: early state; right: late state). E Schematic of the bulk-RNA sequencing in control and THBS1-overexpressing macrophages. F GSEA of the TNFα signaling pathway between THBS1-overexpression macrophages and control macrophages. G Heatmap showing the expression of leading-edge genes in the TNFα signaling. H, J Immunoblot analysis and quantification of THBS1, TNF-α, and IL-1β expression in control or THBS1-overexpressing macrophages (H), and in control or THBS1-knockdown macrophages (J). I Quantification of TNF-α and IL-6 protein levels in the supernatant of control or THBS1-overexpressing macrophages by enzyme-linked immunosorbent assay. P-values in I, J, and K were calculated by using two-tailed Student’s t test. DEG, differentially expressed genes. *P < 0.05, **P < 0.01, and ***P < 0.001.

Full size image

GO enrichment analysis and Gene Set Enrichment Analysis (GSEA) further demonstrated the enrichment of several immune response-activating pathways in THBS1high macrophages, including response to lipopolysaccharide, positive regulation of cytokine production, canonical NF-κB signal transduction, and inflammatory response signaling, among others (Fig. 3B, Additional file 3: Fig. S6A). To evaluate metabolic activities across macrophage subtypes, we mapped the metabolic pathways of macrophage subsets using scMetabolism, a computational pipeline for quantifying single-cell metabolism [24], and calculated a comprehensive metabolic score based on the top 10 activated pathways. Metabolic analysis showed that THBS1high macrophages exhibited the highest metabolic scores among all the macrophage subsets, as well as increased metabolic activity in patients with active BD (Additional file 3: Fig. S6B-C). Pro-inflammatory fatty acid biosynthesis and pentose phosphate pathway were enriched in THBS1high macrophages (Additional file 3: Fig. S6D) [25]. These results suggest that THBS1high macrophages from patients with active BD maintained a high metabolic activation status, which could be associated with vascular inflammation in patients with BD. Pseudotime analysis indicated that THBS1high macrophages could give rise to TNF-secreting CCL4high macrophages, which then convert to homeostatic macrophages (Fig. 3C). THBS1 and EREG were highly expressed at the beginning of the trajectory, while inflammatory cytokine-related genes, such as TNFRSF9, TNF, and CCL3, were gradually upregulated at the subsequent stage (Fig. 3D).

To mechanistically characterize the inflammatory phenotype of THBS1high macrophages, we performed bulk-RNA sequencing between control and THBS1-overexpressing MDMs (Fig. 3E, Additional file 3: Fig. S7A). Transcriptomic profiling revealed that THBS1-overexpressing macrophages displayed marked upregulation of pro-inflammatory mediators (IL1B, IL6) and immune-modulating chemokines (TNFSF9) compared to control cells (Fig. S7B). Notably, pathway enrichment analysis demonstrated coordinated activation of two key inflammatory axes: (1) the TNFα-NFκB signaling cascade (adjusted P < 0.001) and (2) the pan-inflammatory response pathway (adjusted P < 0.05), suggesting THBS1 may function as a master regulator of macrophage polarization (Fig. 3F,G and Additional file 3: Fig. S7C-D).

These transcriptional changes translated to functional consequences at the protein level. Western blot quantification in THP-1-derived macrophages showed THBS1 overexpression induced 1.93-fold (p = 0.005) and 1.68-fold (p = 0.01) increases in tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) protein expression, respectively (Fig. 3H). Consistently, cytokine profiling of culture supernatants revealed significantly elevated TNF-α (185,318 ± 21,416 vs 45,422 ± 4039 pg/mL) and interleukin-6 (IL-6) (1510 ± 296 vs 643 ± 72 pg/mL) concentrations in THBS1-overexpressing macrophages compared to controls (Fig. 3I). Importantly, reciprocal experiments using siRNA-mediated THBS1 knockdown in primary macrophages resulted in 31% (p = 0.008) and 32% (p = 0.008) reductions in TNF-α and IL-1β expression, establishing a dose-dependent relationship between THBS1 levels and inflammatory output (Fig. 3J). Collectively, these multimodal in vitro analyses (transcriptomic, proteomic, and functional) demonstrate that THBS1high macrophages acquire a distinct pro-inflammatory phenotype through activation of canonical inflammatory pathways.

ETS2 mediated the proinflammatory development of THBS1 high macrophages

We next attempted to identify the candidate transcriptional regulators involved in the developmental trajectory from THBS1high macrophages to TNF-secreting macrophages. Through systematic screening, we identified 324 regulons encompassing 13,788 target genes in the macrophage population. These regulons were classified into 18 regulon modules based on the similarity between their activities (Fig. 4A). Among these, Module M6 exhibited a significantly enhanced activity in BD (Fig. 4B). Variance analysis showed that ETS2 regulon was prominently ranked within M6 with its transcriptional activity specifically increased in activated THBS1high macrophages and CCL4high macrophage populations (Fig. 4C,D, Additional file 3: Fig. S8). Functional annotation of ETS2 target genes showed significant enrichment in the TNFα–NFκB signaling pathway (Fig. 4E), suggesting a potential regulatory role in inflammatory responses. Consistent with these findings, both the transcriptional activity and expression levels of ETS2 were markedly elevated in BD patients, particularly during active disease phases (Fig. 4F,G). Building on recent evidence identifying ETS2 as a central regulator of inflammatory macrophages, we hypothesized that ETS2 mediates the pro-inflammatory development of THBS1high macrophages [26].

Fig. 4
figure 4

ETS2 mediated the proinflammatory development of THBS1high macrophages. A Heatmap displaying the 18 regulon modules identified based on the connection Specificity Index matrix in the macrophage population. B UMAP plot of BD and control macrophages colored according to the activity of module M6 (left). Violin plot comparing the module M6 activity between BD and HC. C Regulons (blue dots) in module M6 ranked by fraction of variance across group. ETS2 regulon was highlighted with red dots. D THBS1high macrophages (red dots) and binarized regulon activity scores (RAS) of the EST2 regulon (green dots) are highlighted on the UMAP plot for all macrophages. E Dot plot showing the regulons enriched in “HALLMARK_TNFA_SIGNALING_VIA_NFKB” signaling. F,G Violin plot comparing the expression (left) and regulon activity of ETS2 (right) between BD patients and controls (F), and among controls, patients with remission BD, and patients with active BD (G). H Boxplot comparing the expression of ETS2 between control and THBS1-overexpressing macrophages. I Immunofluorescence staining of the subcellular localization of p-ETS2 in control and THBS1-overexpression macrophages. Scale bar, 50 μm. J Immunoblot analysis of ETS2 phosphorylation and protein levels in cytoplasmic and nuclear fractions extracted from control and THBS1-overexpressing macrophages. P-value in B, F, G was calculated by Wilcoxon test and in H was calculated by two-tailed Student’s t test. DAPI, 4′,6-diamidino-2-phenylindole; TNF-α, tumor necrosis factor-α; UMAP, Uniform manifold approximation and projection. ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001

Full size image

To validate this hypothesis, we conducted in vitro experiments using THBS1-overexpressing MDMs. These studies demonstrated increased ETS2 expression in THBS1-overexpressing MDMs (Fig. 4H). Furthermore, in THP-1-derived macrophages, THBS1 overexpression led to increased total ETS2 (t-ETS2) expression and enhanced nuclear translocation of phosphorylated ETS2 (p-ETS2) (Fig. 4I,J). These findings suggest that elevated THBS1 expression in macrophages exerts downstream transcriptional regulatory effects by facilitating ETS2 nuclear translocation, thereby potentially driving the pro-inflammatory differentiation program. To further explore whether ETS2 was essential to drive TNF-α and THBS1 expression in this population, we performed ETS2-knockdown experiments in the macrophages. Results demonstrated that ETS2 knockdown in THBS1-overexpressing macrophages significantly attenuated TNF-α upregulation. However, ETS2 knockdown did not significantly affect TNF-α-induced THBS1 expression, confirming ETS2’s essential role in mediating TNF-α induction but not THBS1 expression (Additional file 3: Fig. S9A-B).

Phenotypic transition of SMCs in VBD patients

To elucidate cellular crosstalk in VBD, we performed CellChat analysis. While fibroblast-SMC interactions showed the most pronounced increase in VBD, we also observed a marked elevation in both the quantity and intensity of macrophage-SMC interactions in VBD patients compared to HC (Additional file 3: Fig. S10A–C). Given the pathological hallmark of medial layer degeneration in VBD, coupled with the identification of SMCs as the most transcriptionally dysregulated cell type (highest number of DEGs between VBD and HC) and their enhanced cross-talk with macrophages, we further investigated SMC heterogeneity and functional reprogramming in VBD.

After sub-clustering, six SMC subtypes were identified according to their specific marker gene expression and GO terms enrichment. SMC_SFRP1 expressed high levels of WNT signaling-associated genes (SFRP1 and CCN5) suggesting that they were involved in cell growth and migration. SMC_PLAC9 was notably enriched in metabolic pathways, which are essential for maintaining normal functions of SMCs [27]. SMC_MT2A was characterized by its elevated expression of proliferation-associated genes (EGR1, ATF3, and NR4A3) and inflammatory gene [28,29,30]. SMC_COL1A1 enriched in the extracellular matrix (ECM) organization, SMC_stressed enriched in stress-response signaling and SMC_ISG expressing ISGs were also identified. In terms of distribution, SMC_PLAC9 was the predominant SMC subtype in the HC group, while the other SMC subtypes were more enriched in the BD group (Fig. 5A–D). Differential analysis showed that SMCs from patients with BD exhibited higher expression of ECM-related genes, inflammatory factors, stress-related genes, and lower expression of the tissue inhibitor of matrix metalloproteinases (Fig. 5E). GSEA analysis revealed that the TNF-α-NFκb pathway, apoptosis, and ECM receptor interaction were activated in the SMCs obtained from patients with BD, while oxidative phosphorylation and muscle contraction were inhibited (Fig. 5F). Collectively, these results suggested that SMCs transitioned from a contractile phenotype to synthetic and inflammatory phenotypes in BD. The phenotypic transition of SMCs in BD was additionally validated through immunofluorescence analysis of aortic tissues, which revealed upregulation of synthetic (FN) and inflammatory markers (IL-1β, TNF-α) in BD specimens compared to healthy controls (Fig. 5G, Additional file 3: Fig. S11A-B). Further exploration of cell–cell communication between macrophage and SMC subsets revealed that the THBS1high macrophages exhibited active communication with multiple SMC subgroups, with the upregulation of the THBS1-CD47 signaling pathway (Fig. 5H, Additional file 3: Fig. S10D), indicating that THBS1high macrophages might promote the phenotypic transition of SMCs partially via THBS1–CD47 interaction.

Fig. 5
figure 5

Characterization of the gene expression and functional profiles of SMC subpopulations. A UMAP visualization of SMC subclusters, color-coded according to cell type. B UMAP visualization of SMC subclusters, split by group. C Dot plot showing the expression of marker genes in SMC subtypes. D Bubble diagram showing the representative Gene Ontology terms enriched in SMC subtypes. E Volcano plot showing the DEGs of SMC, including extracellular matrix-related genes (COL1A1, COL3A1, and COL5A2), inflammatory factors (CXCL8, CCL3, and CCL3L1), stress-related genes (JUND and JUNB) and the tissue inhibitor of matrix metalloproteinases TIMP1 and TIMP3, between controls and BD patients. F GSEA analysis depicting the pathways enriched in SMCs of BD patients in comparison to controls based on the hallmark (upper) and Kyoto Encyclopedia of Genes and Genomes (lower) gene sets. G Representative immunofluorescence images of α-SMA (green), FN (red), and DAPI (blue) in the aortas of healthy controls and patients with BD. Scale bars, 100 μM. H Dot plot comparing the significant increased ligand-receptor pairs between BD and HC, which are involved in the signaling from macrophage subtypes to SMC subtypes. UMAP, Uniform manifold approximation and projection; SMC, smooth muscle cell; DEGs, differentially expressed genes; BD, Behçet’s disease; GSEA, Gene Set Enrichment Analysis; HC, healthy control

Full size image

THBS1 high macrophages drove the phenotypic transition and apoptosis of SMCs

To mechanistically interrogate the role of THBS1high macrophages in SMC dysfunction, we established a macrophage–SMC co-culture system combining THBS1-overexpressing/control macrophages with HASMCs, followed by bulk RNA sequencing of HASMCs to delineate transcriptomic alterations (Fig. 6A). HASMCs co-cultured with THBS1-overexpressing macrophages exhibited a pronounced upregulation of ECM remodeling genes, matrix metalloproteinases (MMP2, MMP3), and pro-inflammatory cytokines (IL6, CXCL1), alongside downregulation of tissue inhibitor of metalloproteinases (TIMP1, TIMP3) and contractile markers (CNN1) (Fig. 6B). ECM remodeling and inflammatory signaling pathways were enriched in HASMCs co-cultured with THBS1-overexpressing macrophages (Fig. 6C). THBS1-overexpressing macrophages could also inhibit oxidative phosphorylation in HASMCs, while promoting epithelial–mesenchymal transition (Fig. 6D). These findings are consistent with the pathological changes observed in SMCs during medial degeneration, suggesting that THBS1high macrophages facilitate the phenotypic transition of HASMCs.

Fig. 6
figure 6

THBS1high macrophages mediate the phenotypic transition of contractile SMCs to synthetic SMCs. A, Schematic of bulk-RNA sequencing in SMCs co-cultured with control or THBS1-overexpressing macrophages. B, Volcano plot showing the DEGs of SMCs cocultured with control or THBS1-overexpressing macrophages. C, Lollipop diagram displaying enriched pathways of the significantly upregulated genes in SMCs cocultured with THBS1-overexpressing macrophages. D, Gene Set Enrichment Analysis of the oxidative phosphorylation and epithelial mesenchymal transition signaling pathway in SMCs cocultured with THBS1-overexpressing macrophages versus control macrophages. Genes were ranked by foldchange in expression between two groups. NES, normalized enrichment score. E, Immunoblot analysis and quantification of the contractile/synthetic markers and apoptotic protein expression in SMCs cocultured with THBS1-overexpressing macrophages (left) and THBS1-knockdown macrophages (right). P-value was calculated by two-tailed Student’s t-test. F, Immunofluorescence staining of α-SMA (red), phalloidin (green) (upper), and CNN1 (red, lower) in SMCs cocultured with control or THBS1-overexpressing macrophages. The nuclei were stained with DAPI. Scale bar, 50 μM. G, Representative fluorescence images of mitochondria in SMCs cocultured with control or THBS1-overexpressing macrophages. Scale bar, 50 μM. SMCs, smooth muscle cells; DEGs, differentially expressed genes; DAPI, 4′,6-diamidino-2-phenylindole. *P < 0.05 and **P < 0.01

Full size image

We further confirmed the ability of THBS1high macrophages to induce phenotypic transition in HASMCs at the protein level using western blot and immunofluorescence. The results showed a significant reduction in the expression of contraction-related proteins α-smooth muscle actin (α-SMA) and calponin 1 (CNN1) in HASMCs co-cultured with THBS1-overexpressing macrophages, accompanied by a notable increase in the expression of collagen type I (COL1), fibronectin (FN), and MMP2. Additionally, THBS1-overexpressing macrophages promoted apoptosis in HASMCs, as evidenced by the upregulated expression of apoptosis-associated protein BAX. Conversely, co-culturing with THBS1-knockdown macrophages led to an increase in the expression of α-SMA and CNN1 and a significant reduction in the expression of COL1, FN, MMP2, and BAX (Fig. 6E,F). Given that THBS1-overexpressing macrophages might inhibit oxidative phosphorylation in HASMCs, we utilized Mitotracker probes to label the mitochondria of HASMCs. The mitochondria in HASMCs co-cultured with THBS1-overexpressing macrophages transformed from the typical elongated shape to a punctate morphology, suggesting THBS1-overexpressing macrophage-induced mitochondrial damage in HASMCs (Fig. 6G).

Collectively, these findings demonstrate that THBS1high macrophages orchestrate a phenotypic switch in SMCs, from a contractile, quiescent state to a synthetic, inflammatory, and apoptosis-prone phenotype via regulation of ECM dynamics, inflammatory signaling, and metabolic reprogramming. This mechanism likely contributes to the medial layer disintegration characteristic of VBD.

THBS1 high macrophages promoted vascular inflammation in humanized VBD mice

To further investigate the pathological role of THBS1high macrophages in vascular inflammation and degeneration in vivo, we established a humanized VBD mouse model. First, a mini osmotic pump loaded with AngII was subcutaneously implanted in the NCG mouse model to induce the formation of aneurysm. On day 14, these NCG mice were then immuno-reconstituted by the adoptive transfer of BD PBMCs along with autologous MDMs, generating a humanized VBD model. To assess the role of THBS1high macrophages in vivo, macrophages in the control group were transfected with control lentivirus, while those in the THBS1-OE experimental group received THBS1-overexpression lentivirus (Additional file 3: Fig. S12A). Immunohistochemistry analysis confirmed the presence of human CD68 + macrophages in the aorta of mouse (Additional file 3: Fig. S12B). Although neither group developed overt aneurysms—likely due to the use of NCG mice instead of the classical apolipoprotein E-deficient model for AngII-induced abdominal aortic aneurysm (AAA) model, THBS1high macrophages still partially promoted SMC phenotypic switch in the experimental group (Additional file 3: Fig. S12C). Moreover, abdominal aortas in the THBS1-OE group displayed an increased expression of inflammatory cytokines IL-1β and TNF-α (Additional file 3: Fig. S12D). Exacerbated vascular inflammation was further validated by PET-CT imaging, showing significantly higher 18F-FDG uptake in the abdominal aortas of the experimental group (Additional file 3: Fig. S12E-F). These data indicated that THBS1high macrophages contributed to the phenotypic switching of contractile SMCs and aggravated vascular inflammation in AngII-infused mouse aorta.

TNF-α upregulated THBS1 expression in macrophages

Having observed the pathological role of THBS1high macrophages, we next explored the factors involved in the expression of THBS1 in macrophages. Sequencing results revealed that THBS1high macrophages consistently displayed upregulated expression of TNF-α receptors, TNFRSF1A and TNFRSF1B (Fig. 7A), indicating heightened sensitivity to TNF-α signaling. To further explore this relationship, we stimulated macrophages with TNF-α and observed a significant increase in THBS1 expression (Fig. 7B,C). These results suggest the existence of a potential positive feedback loop between THBS1 and TNF-α, which may contribute to the sustained activation and expansion of pathological THBS1high macrophages.

Fig. 7
figure 7

TNF-α upregulated THBS1 expression in macrophages. A Dot plot showing the expression of TNF receptors in each cluster. B Immunofluorescence staining of THBS1, phalloidin and DAPI in macrophages treated with TNF-α (20 ng/mL). Scale bars, 50 μM. C Immunoblot analysis and quantification of THBS1 expression in macrophages treated with TNF-α (20 ng/mL). D Immunoblot analysis and quantification of THBS1 levels in macrophages treated with TNF-α (20 ng/mL) in the presence or absence of adalimumab (ADA), thalidomide (THAL) or tofacitinib (TOFA). E THBS1 levels in the plasma of VBD patients (n = 15) before and after 6 months of adalimumab treatment. P-value in C was calculated by two-tailed Student’s t-test, in D was calculated by one-way ANOVA with Tukey’s multiple comparison test and in E was calculated by paired t-tests. DAPI, 4′,6-diamidino-2-phenylindole; TNF-α, tumor necrosis factor-α; UMAP, Uniform manifold approximation and projection. ns, not significant, *P < 0.05, **P < 0.01, and ***P < 0.001

Full size image

After observing the pro-inflammatory characteristics of THBS1high macrophages and the facilitative role of TNF-α in their expansion, we next evaluated whether THBS1high macrophages respond to adalimumab (a TNF inhibitor). In addition, since immunosuppressive agent emerged as effective medication in VBD, and JAK inhibitors are considered as alternative therapeutic options for VBD and other vasculitis [31,32,33,34], we also explored the role of thalidomide and tofacitinib respectively. It was found that adalimumab treatment greatly decreased the TNF-α-stimulated THBS1 expression in macrophages (Fig. 7D). Further analysis showed that adalimumab treatment could significantly decrease the plasma THBS1 levels in VBD patients undergoing remission, suggesting that circulating THBS1 levels could respond to adalimumab treatment in patients with VBD and function as a potential indicator for assessing therapeutic efficacy in the context of systemic disease activity (Fig. 7E).

Adalimumab alleviated THBS1 high macrophages-induced vascular inflammation and degeneration in humanized VBD mice

Given that THBS1high macrophages contributed to the phenotypic switch of SMCs and that the TNF inhibitor adalimumab suppressed the expansion of THBS1high macrophages in vitro, we hypothesized that adalimumab might alleviate vascular injury in BD caused by THBS1high macrophages. After initial exploration of the AngII-induced model revealed limited aortic dilation aneurysm development due to the absence of apolipoprotein E deficiency, we established an optimized CaCl₂-induced humanized VBD model to investigate this hypothesis (Fig. 8A). At day 0, the mice were treated with CaCl2. After administering the mice with lipopolysaccharide on day 13, the mice were randomly assigned to three groups with one control group immuno-reconstituted by BD PBMC plus control macrophages and the other two experimental groups immuno-reconstituted with BD PBMC plus THBS1-overexpressing macrophages on the next day. To test the potential of adalimumab to attenuate aortic dilation and vascular inflammation, humanized mice in one experimental group were administered 3 mg/kg of adalimumab subcutaneously every 3 days, while the other two groups were administered saline for 2 weeks.

Fig. 8
figure 8

TNF inhibitor alleviated THBS1high macrophages-induced vascular degeneration and inflammation in VBD mice. A Schematic diagram showing the establishment of CaCl2-induced humanized VBD model and the timeline for adalimumab treatment. B Representative images of the macroscopic view of abdominal aorta. C Representative longitudinal ultrasound images of infrarenal aortas detected with B model. D Quantification of the maximal diameter in indicated groups (n = 8 per group). E Elastin degradation was reflected by Verhoeff-Van Gieson staining of infrarenal aorta. Scale bar, 100 µm. F Immunofluorescence staining of fibronectin (FN, red) and α-SMA (green) in aorta. The nuclei were stained with 4′,6-diamidino-2-phenylindole. Scale bar, 100 µm. G Quantification of FN and α-SMA fluorescence intensity in indicated groups (n = 6 per group). H Immunofluorescence staining of TUNEL (red) and α-SMA (green) in aorta. The nuclei were stained with 4′,6-diamidino-2-phenylindole. Scale bar, 100 µm. I Quantification of apoptotic cells in indicated groups (n = 6 per group). J Immunohistochemical staining of TNF-α (upper) and IL-1β (lower) in infrarenal aorta. Scale bar, 25 µm. K Representative PET-CT scanning images of mice on day 28 after CaCl2 treatment. L Quantification of mean 18F-FDG SUV of abdominal aorta (n = 3 per group). P-value in D, G, I, and L was calculated by one-way ANOVA with Tukey’s multiple comparison test. VBD, vascular Behçet’s disease; PET-CT, positron emission tomography computed tomography; SUV, standardized uptake value. **P < 0.01; and ***P < 0.001

Full size image

The results demonstrated that THBS1high macrophages significantly facilitated pathological vascular remodeling in the THBS1-OE group (by adopting BD PBMCs and THBS1-overexpressing macrophages), which was characterized by aortic dilation, elastin degradation, SMC apoptosis, and medial degeneration (Fig. 8B–I). An increased expression of inflammatory markers and an elevated 18F-FDG uptake were observed in the THBS1-OE group compared to that with the vehicle group (adopting BD PBMC and control macrophages) (Fig. 8J–L). Hence, these data suggested that THBS1high macrophages may lower the threshold required to induce vascular inflammation, SMC phenotypic switch, SMC apoptosis, and aortic dilation in the CaCl2-induced humanized VBD mouse model. The adalimumab treatment mitigated THBS1high macrophage-accelerated aortic dilation with a decreased maximal aortic diameter and reduced elastin disarray. Adalimumab was found to elevate the expression of contractile markers (α-SMA), simultaneously lower the expression of synthetic marker (FN) in SMC, and reduce SMC apoptosis. In addition, adalimumab treatment exhibited beneficial effects on vascular inflammation characterized by reduced IL-1β and TNF-α expression as well as decreased 18F-FDG uptake, as observed in PET-CT imaging (Fig. 8B–L).

Discussion

VBD poses substantial clinical challenges due to its propensity for arterial complications, which is the leading cause of death [35, 36]. Despite decades of research, two critical knowledge gaps persist: (1) the absence of definitive histopathological markers for VBD and (2) incomplete understanding of the molecular cascades driving proinflammatory response and medial layer destruction. Addressing these limitations is paramount for developing targeted therapies and personalized monitoring strategies.

Our investigation bridges these gaps through an integrative approach combining scRNA-seq of human VBD aorta, functional in vitro coculture systems, and a humanized murine VBD model. Three key advances emerge from this multi-platform analysis: First, we establish the first single-cell atlas of VBD-affected aorta, revealing previously unrecognized cellular heterogeneity within the inflammatory infiltrate. Second, we identify and functionally validate a pathogenic THBS1high macrophage subpopulation with unique proinflammatory signature, which participates in a positive proinflammatory feedback loop between THBS1 and TNF-α. Third, we delineate a novel cellular interaction between THBS1high macrophages and the HASMCs. Moreover, the therapeutic implications of TNF inhibitor in VBD are underscored by our humanized model findings: Adalimumab administration attenuated vascular inflammation and aortic dilation progression (Fig. 9).

Fig. 9
figure 9

Schematic diagram illustrating the proposed mechanism of THBS1high macrophages in vascular inflammation and degeneration in vascular Behçet’s disease

Full size image

In this study, we observed an increase in the circulating THBS1 level in patients with active VBD and an association between the THBS1 level and systemic disease activity. Furthermore, our longitudinal analysis revealed that plasma THBS1 levels significantly reduced in anti-TNF responders. These clinical findings contextualize previous reports of THBS1 involvement in cardiovascular pathologies [37,38,39]. While Liu et al. identified increased myeloid-derived THBS1 in CaCl2-treated mouse noninflamed aortic aneurysm, where it promotes disease progression through TIMP1 suppression and enhanced macrophage infiltration [40], the role of THBS1high macrophages in BD and other vasculitis, such as giant cell arteritis or Takayasu arteritis remain unexplored. Importantly, our work advances the field by (1) demonstrating its functional role in HASMCs, (2) establishing circulating THBS1’s association with VBD disease stages, and (3) linking its systemic dynamics to treatment response. This tripartite significance—as a mechanistic driver, monitoring marker, and therapeutic target—positions THBS1high macrophages as central players in VBD pathogenesis. Critically, although our data primarily characterize THBS1 as a plasma biomarker reflecting systemic disease activity, its association with vascular lesions warrants further investigation. Radiological progression stratified active VBD cases from remission cases in our scRNA-seq cohort, complementing correlations between THBS1high macrophage enrichment and systemic disease activity, and suggesting THBS1high macrophage enrichment may mark aortopathy. Future studies incorporating larger cohorts and advanced imaging (e.g., MRA, PET-CT) will be essential to elucidate these important relationships and validate our findings.

THBS1 is widely recognized for its critical role in modulating inflammatory responses. Previous studies have demonstrated that recombinant mouse thrombospondin-1 protein can polarize macrophages toward an M1-like pro-inflammatory phenotype [38]. Another single-cell analysis data of mouse atherosclerotic plaques revealed that macrophage populations with high THBS1 expression exhibited elevated levels of inflammatory cytokines [40, 41]. Despite these findings, the specific characteristics of THBS1high macrophages in VBD and their pathological contributions to medial damage remain unexplored.

By integrating scRNA-seq data with in vitro experiments, we validated the pro-inflammatory phenotype of THBS1high macrophages. Importantly, we identified a novel positive feedback loop between THBS1 and TNF-α during macrophage activation which amplifies the inflammatory responses. Mechanistically, we demonstrated that ETS2 is essential for THBS1-induced TNF-α upregulation but dispensable for TNF-α-mediated THBS1 expression. This differential regulation suggested that other transcription factors may be involved in this process, revealing a complex transcriptional network underlying this pathway. These findings provide critical insight into the self-sustaining nature of this pathogenic feedback loop. Beyond its role in inflammation, THBS1 is known to regulate the proliferation and migration of vascular SMCs and has been implicated in the progression of AAAs by suppressing TIMPs [40, 42, 43]. Our findings extend these observations by revealing a previously unrecognized role for THBS1high macrophages in mediating phenotypic transition and apoptosis of HASMCs. These processes are widely regarded as key contributors to aortic instability and aneurysm formation [12]. Collectively, our findings highlighted the pathological significance of macrophage-specific THBS1 expression in VBD progression, rather than reflecting the broader biological effects of THBS1 that have been previously documented in other contexts. Furthermore, while recognizing the importance of fibroblast-SMC interactions revealed by CellChat analysis, we believe these findings open new avenues for future investigation into fibroblast-mediated mechanisms in VBD pathogenesis, and additional studies are needed to elucidate their critical role.

We further addressed the pathological role of THBS1high macrophages in vivo. Numerous scholars have made substantial efforts to establish the animal models of BD, including the HLA-B51 transgenic model, herpes simplex virus type 1-infected model, and heat shock protein model [44]. Nevertheless, because of the complex etiology and highly variable clinical manifestations of BD, none of these models can fully replicate the phenotypes of BD, particularly the critically severe vascular involvement associated with BD. Considering the systemic inflammatory state and high prevalence of AA in VBD, we generated a seminal VBD animal model by coupling the PBMC humanized model with the traditional AAA mouse model (Additional file 3: Fig. S13A). CaCl2 treatment permitted the migration of human immune cells to the inflamed tissues (Additional file 3: Fig. S13B-D) and the formation of AA, circumventing the limitations of AngII-induced model that requires apolipoprotein E-deficient backgrounds. While our preliminary AngII model demonstrated THBS1high macrophages’ pathogenic role in vascular dysfunction prior to significant aortic dilation and aneurysm formation [12], the CaCl₂ model provided a robust platform to evaluate their sustained contribution to vascular damage and therapeutic responses. The AngII model data were also retained in supplementary materials to document our exploratory process in developing humanized VBD model. Importantly, we also confirmed that reconstitution with control macrophages alone (without CaCl₂/LPS) did not induce significant vascular pathology, as evidenced by the absence of elastin degradation, SMC phenotypic switching, or apoptosis, thereby enabling deeper exploration of our humanized model (Additional file 3: Fig. S14A-D).

We adapted a previously published approach involving the adoptive transfer of phenotypically distinct macrophages within a humanized mouse model [45]. Consistent with our in vitro findings, we noted that THBS1high macrophages promoted vascular inflammation and medial degeneration, underscoring their critical role in the pathogenesis of arterial involvement in BD. Furthermore, we demonstrated that adalimumab, a TNF inhibitor, effectively attenuated THBS1high macrophage-driven vascular inflammation and AA formation. These findings not only supported the use of an anti-TNF therapy for the treatment of VBD, but also highlight the utility of our humanized model for studying disease mechanisms and evaluating potential therapeutic strategies for VBD.

Our study has several limitations. First, the number of aortic tissue samples available for single-cell RNA sequencing was limited due to the rarity and clinical challenges associated with obtaining such specimens from VBD patients. Nevertheless, the scarcity of these samples underscores the value of our findings. To mitigate this limitation, we validated our results using publicly available datasets, multiplexed immunohistochemistry, and a series of functional assays, thereby strengthening the robustness of our conclusions. Second, while the lack of commercially available THBS1 flow cytometry antibodies presented technical challenges, we successfully characterized the THBS1high macrophages through comprehensive single-cell transcriptomic analysis and a series of in vitro and in vivo assays. Future studies may identify surrogate surface markers to enable the isolation and further investigation of this macrophage subset. Third, the molecular mechanisms driving the TNF-induced expansion of THBS1high macrophages remain incompletely understood and warrant further exploration.

Conclusions

Our study provides a comprehensive single-cell atlas of immune and stromal cells in VBD lesions, revealing altered cellular composition and intercellular communication networks that contribute to disease progression. We identified THBS1high macrophages as key mediators of proinflammatory responses, SMC phenotypic switching and medial degeneration. By employing a humanized VBD mouse model, we further demonstrated the therapeutic potential of anti-TNF therapy in mitigating vascular inflammation and degeneration. We believe our findings might provide insights into the pathogenesis, clinical assessment, and effective therapeutic targets of VBD and potentially other inflamed aneurysms.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

AA:

Aortic aneurysm

AAA:

Abdominal aortic aneurysm

AngII:

Angiotensin II

CNN1:

Calponin 1

COL1:

Collagen type I

ECM:

Extracellular matrix

FN:

Fibronectin

GSEA:

Gene Set Enrichment Analysis

IL-1β:

Interleukin-1β

IL-6:

Interleukin-6

MDM:

Monocyte-derived macrophage

MMP:

Matrix metalloproteinase

PBMC:

Peripheral blood mononuclear cell

scRNA-seq:

Single-cell RNA sequencing

SMC:

Smooth muscle cell

TNF-α:

Tumor necrosis factor-α

VBD:

Vascular Behçet’s disease

α-SMA:

α-Smooth muscle actin

References

  1. Greco A, De Virgilio A, Ralli M, Ciofalo A, Mancini P, Attanasio G, et al. Behcet’s disease: New insights into pathophysiology, clinical features and treatment options. Autoimmun Rev. 2018;17(6):567–75.

    Article  CAS  PubMed  Google Scholar 

  2. Zou J, Luo JF, Shen Y, Cai JF, Guan JL. Cluster analysis of phenotypes of patients with Behcet’s syndrome: a large cohort study from a referral center in China. Arthritis Res Ther. 2021;23(1):45.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Bettiol A, Alibaz-Oner F, Direskeneli H, Hatemi G, Saadoun D, Seyahi E, et al. Vascular Behcet syndrome: from pathogenesis to treatment. Nat Rev Rheumatol. 2023;19(2):111–26.

    Article  CAS  PubMed  Google Scholar 

  4. Yazici H, Seyahi E, Hatemi G, Yazici Y. Behcet syndrome: a contemporary view. Nat Rev Rheumatol. 2018;14(2):107–19.

    Article  CAS  PubMed  Google Scholar 

  5. Alibaz-Oner F, Aldag B, Aldag M, Unal AU, Mutis A, Toptas T, et al. Post-thrombotic syndrome and venous disease-specific quality of life in patients with vascular Behcet’s disease. J Vasc Surg Venous Lymphat Disord. 2016;4(3):301–6.

    Article  PubMed  Google Scholar 

  6. Seyahi E, Melikoglu M, Akman C, Hamuryudan V, Ozer H, Hatemi G, et al. Pulmonary artery involvement and associated lung disease in Behcet disease: a series of 47 patients. Medicine (Baltimore). 2012;91(1):35–48.

    Article  PubMed  Google Scholar 

  7. Emmi G, Bettiol A, Silvestri E, Di Scala G, Becatti M, Fiorillo C, et al. Vascular Behcet’s syndrome: an update. Intern Emerg Med. 2019;14(5):645–52.

    Article  PubMed  Google Scholar 

  8. Gaudric J, Jayet J, Saadoun D, Couture T, Ferfar Y, Davaine JM, et al. Factors influencing the recurrence of arterial involvement after surgical repair in Behcet disease. J Vasc Surg. 2020;72(5):1761–9.

    Article  PubMed  Google Scholar 

  9. Zheng W, Wang X, Liu J, Yu X, Li L, Wang H, et al. Single-cell analyses highlight the proinflammatory contribution of C1q-high monocytes to Behcet’s disease. Proc Natl Acad Sci U S A. 2022;119(26):e2204289119.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Chang L, Zheng Z, Zou Q, Zhong B, Chen C, Cheng X, et al. Single cell Transcriptome and T cell Repertoire Mapping of the Mechanistic Signatures and T cell Trajectories Contributing to Vascular and Dermal Manifestations of Behcet’s Disease. bioRxiv. 2022:2022.03.22.485251.

  11. Ruf W, Ruggeri ZM. Neutrophils release brakes of coagulation. Nat Med. 2010;16(8):851–2.

    Article  CAS  PubMed  Google Scholar 

  12. Gao J, Cao H, Hu G, Wu Y, Xu Y, Cui H, et al. The mechanism and therapy of aortic aneurysms. Signal Transduct Target Ther. 2023;8(1):55.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Xiong W, MacTaggart J, Knispel R, Worth J, Persidsky Y, Baxter BT. Blocking TNF-alpha attenuates aneurysm formation in a murine model. J Immunol. 2009;183(4):2741–6.

    Article  CAS  PubMed  Google Scholar 

  14. Luo Y, Luo J, An P, Zhao Y, Zhao W, Fang Z, et al. The activator protein-1 complex governs a vascular degenerative transcriptional programme in smooth muscle cells to trigger aortic dissection and rupture. Eur Heart J. 2024;45(4):287–305.

    Article  PubMed  Google Scholar 

  15. Hirahara L, Takase-Minegishi K, Kirino Y, Iizuka-Iribe Y, Soejima Y, Yoshimi R, et al. The roles of monocytes and macrophages in Behcet’s disease with focus on M1 and M2 Polarization. Front Immunol. 2022;13:852297.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Wu X, Wang Z, Shi J, Yu X, Li C, Liu J, et al. Macrophage polarization toward M1 phenotype through NF-kappaB signaling in patients with Behcet’s disease. Arthritis Res Ther. 2022;24(1):249.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. International Team for the Revision of the International Criteria for Behcet’s D. The international criteria for Behcet’s disease (ICBD): a collaborative study of 27 countries on the sensitivity and specificity of the new criteria. J Eur Acad Dermatol Venereol. 2014;28(3):338–47.

    Article  Google Scholar 

  18. Lawton G, Bhakta BB, Chamberlain MA, Tennant A. The Behcet’s disease activity index. Rheumatology (Oxford). 2004;43(1):73–8.

    Article  CAS  PubMed  Google Scholar 

  19. Kanakaraj P, Puffer BA, Yao XT, Kankanala S, Boyd E, Shah RR, et al. Simultaneous targeting of TNF and Ang2 with a novel bispecific antibody enhances efficacy in an in vivo model of arthritis. MAbs. 2012;4(5):600–13.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Binder NB, Puchner A, Niederreiter B, Hayer S, Leiss H, Bluml S, et al. Tumor necrosis factor-inhibiting therapy preferentially targets bone destruction but not synovial inflammation in a tumor necrosis factor-driven model of rheumatoid arthritis. Arthritis Rheum. 2013;65(3):608–17.

    Article  CAS  PubMed  Google Scholar 

  21. Johnsen-Soriano S, Sancho-Tello M, Arnal E, Diaz-Llopis M, Navea A, Miranda M, et al. Comparison of the acute effects of anti-TNF-alpha drugs on a uveitis experimental model. Ocul Immunol Inflamm. 2010;18(3):208–15.

    Article  CAS  PubMed  Google Scholar 

  22. Martinez-Fernandez de la Camara C, Hernandez-Pinto AM, Olivares-Gonzalez L, Cuevas-Martin C, Sanchez-Arago M, Hervas D, et al. Adalimumab Reduces Photoreceptor Cell Death in A Mouse Model of Retinal Degeneration. Sci Rep. 2015;5:11764.

  23. Kong X, Xu M, Cui X, Ma L, Cheng H, Hou J, et al. Potential role of macrophage phenotypes and CCL2 in the pathogenesis of Takayasu arteritis. Front Immunol. 2021;12:646516.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Wu Y, Yang S, Ma J, Chen Z, Song G, Rao D, et al. Spatiotemporal immune landscape of colorectal cancer liver metastasis at single-cell level. Cancer Discov. 2022;12(1):134–53.

    Article  CAS  PubMed  Google Scholar 

  25. Van den Bossche J, O’Neill LA, Menon D. Macrophage Immunometabolism: Where Are We (Going)? Trends Immunol. 2017;38(6):395–406.

    Article  PubMed  Google Scholar 

  26. Stankey CT, Bourges C, Haag LM, Turner-Stokes T, Piedade AP, Palmer-Jones C, et al. A disease-associated gene desert directs macrophage inflammation through ETS2. Nature. 2024;630(8016):447–56.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Shi J, Yang Y, Cheng A, Xu G, He F. Metabolism of vascular smooth muscle cells in vascular diseases. Am J Physiol Heart Circ Physiol. 2020;319(3):H613–31.

    Article  CAS  PubMed  Google Scholar 

  28. Zhou H, Li N, Yuan Y, Jin YG, Guo H, Deng W, et al. Activating transcription factor 3 in cardiovascular diseases: a potential therapeutic target. Basic Res Cardiol. 2018;113(5):37.

    Article  PubMed  Google Scholar 

  29. Chakraborty A, Li Y, Zhang C, Li Y, Rebello KR, Li S, et al. Epigenetic induction of smooth muscle cell phenotypic alterations in aortic aneurysms and dissections. Circulation. 2023;148(12):959–77.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Wen Y, Liu Y, Li Q, Tan J, Fu X, Liang Y, et al. Spatiotemporal ATF3 expression determines VSMC fate in abdominal aortic aneurysm. Circ Res. 2024;134(11):1495–511.

    Article  CAS  PubMed  Google Scholar 

  31. Shek LP, Lim DL. Thalidomide in Behcet’s disease. Biomed Pharmacother. 2002;56(1):31–5.

    Article  CAS  PubMed  Google Scholar 

  32. Bursi R, Cafaro G, Perricone C, Riccucci I, Calvacchi S, Gerli R, et al. Contribution of Janus-kinase/signal transduction activator of transcription pathway in the pathogenesis of vasculitis: a possible treatment target in the upcoming future. Front Pharmacol. 2021;12:635663.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Schwartz DM, Bonelli M, Gadina M, O’Shea JJ. Type I/II cytokines, JAKs, and new strategies for treating autoimmune diseases. Nat Rev Rheumatol. 2016;12(1):25–36.

    Article  CAS  PubMed  Google Scholar 

  34. Liu J, Hou Y, Sun L, Li C, Li L, Zhao Y, et al. A pilot study of tofacitinib for refractory Behcet’s syndrome. Ann Rheum Dis. 2020;79(11):1517–20.

    Article  CAS  PubMed  Google Scholar 

  35. Kural-Seyahi E, Fresko I, Seyahi N, Ozyazgan Y, Mat C, Hamuryudan V, et al. The long-term mortality and morbidity of Behcet syndrome: a 2-decade outcome survey of 387 patients followed at a dedicated center. Medicine (Baltimore). 2003;82(1):60–76.

    Article  PubMed  Google Scholar 

  36. Saadoun D, Wechsler B, Desseaux K, Le Thi HD, Amoura Z, Resche-Rigon M, et al. Mortality in Behcet’s disease. Arthritis Rheum. 2010;62(9):2806–12.

    Article  CAS  PubMed  Google Scholar 

  37. Bosman M, Kruger D, Van Assche C, Boen H, Neutel C, Favere K, et al. Doxorubicin-induced cardiovascular toxicity: a longitudinal evaluation of functional and molecular markers. Cardiovasc Res. 2023;119(15):2579–90.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Zeng T, Yuan J, Gan J, Liu Y, Shi L, Lu Z, et al. Thrombospondin 1 is increased in the aorta and plasma of patients with acute aortic dissection. Can J Cardiol. 2019;35(1):42–50.

    PubMed  Google Scholar 

  39. Yamashiro Y, Thang BQ, Shin SJ, Lino CA, Nakamura T, Kim J, et al. Role of thrombospondin-1 in mechanotransduction and development of thoracic aortic aneurysm in mouse and humans. Circ Res. 2018;123(6):660–72.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Yang H, Zhou T, Sorenson CM, Sheibani N, Liu B. Myeloid-derived TSP1 (thrombospondin-1) contributes to abdominal aortic aneurysm through suppressing tissue inhibitor of metalloproteinases-1. Arterioscler Thromb Vasc Biol. 2020;40(12):e350–66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Cochain C, Vafadarnejad E, Arampatzi P, Pelisek J, Winkels H, Ley K, et al. Single-cell RNA-Seq reveals the transcriptional landscape and heterogeneity of aortic macrophages in murine atherosclerosis. Circ Res. 2018;122(12):1661–74.

    Article  CAS  PubMed  Google Scholar 

  42. Stein JJ, Iwuchukwu C, Maier KG, Gahtan V. Thrombospondin-1-induced vascular smooth muscle cell migration and proliferation are functionally dependent on microRNA-21. Surgery. 2014;155(2):228–33.

    Article  PubMed  Google Scholar 

  43. Majack RA, Goodman LV, Dixit VM. Cell surface thrombospondin is functionally essential for vascular smooth muscle cell proliferation. J Cell Biol. 1988;106(2):415–22.

    Article  CAS  PubMed  Google Scholar 

  44. Charles J, Castellino FJ, Ploplis VA. Past and Present Behcet’s disease animal models. Curr Drug Targets. 2020;21(16):1652–63.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Hu Z, Zhao TV, Huang T, Ohtsuki S, Jin K, Goronzy IN, et al. The transcription factor RFX5 coordinates antigen-presenting function and resistance to nutrient stress in synovial macrophages. Nat Metab. 2022;4(6):759–74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We would like to thank Huiyong Chen, Zongfei Ji, Lili Ma, Ying Sun, Lingying Ma, Xiaomin Dai, Yun Liu and Huijing Huang for clinical support. We would like to thank Rongyi Chen, Sifan Wu, Shuwai Chang and Mengdi Li for experimental help. We would like to thank all the patients and their families for their support.

Funding

This work was supported by the National Natural Science Foundation of China (grant numbers 82271834, 82201987, 82302014, 82201989), Shanghai of Science and Technology Commission (grant number 23Y11905000, 22YF1406900, 24ZR1410900), Shanghai Shenkang Hospital Development Center (SHDC22024245), and the Clinical Research Project of Zhongshan Hospital (grant number ZSLCYJ202307, ZSLCYJ202326).

Author information

Author notes

  1. Xianting Sun and Xiufang Kong contributed equally to this work and should be considered co-first authors.

Authors and Affiliations

  1. Department of Rheumatology, Zhongshan Hospital, Fudan University, Xuhui District, 180 Fenglin Road, Shanghai, 200032, China

    Xianting Sun, Xiufang Kong & Lindi Jiang

  2. Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai, China

    Jun Hou

  3. Department of Cardiac Surgery, Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, China

    Jun Li

  4. Department of Anesthesiology, Zhongshan Hospital, Fudan University, Shanghai, China

    Kefang Guo

  5. Shanghai Key Laboratory of Perioperative Stress and Protection, Shanghai, China

    Kefang Guo

  6. Department of Immunology, Shanghai Key Laboratory of Bioactive Small Molecules and State Key Laboratory of Medical Neurobiology, School of Basic Medical Sciences, Fudan University, Shanghai, China

    Yuanyuan Wei

  7. Department of Physiology and Pathophysiology, School of Basic Medical Sciences, Department of Rheumatology, Zhongshan Hospital, Fudan University, Shanghai, China

    Dan Meng

  8. Center of Clinical Epidemiology and Evidence-Based Medicine, Fudan University, Shanghai, China

    Lindi Jiang

Authors
  1. Xianting Sun
    View author publications

    Search author on:PubMed Google Scholar

  2. Xiufang Kong
    View author publications

    Search author on:PubMed Google Scholar

  3. Jun Hou
    View author publications

    Search author on:PubMed Google Scholar

  4. Jun Li
    View author publications

    Search author on:PubMed Google Scholar

  5. Kefang Guo
    View author publications

    Search author on:PubMed Google Scholar

  6. Yuanyuan Wei
    View author publications

    Search author on:PubMed Google Scholar

  7. Dan Meng
    View author publications

    Search author on:PubMed Google Scholar

  8. Lindi Jiang
    View author publications

    Search author on:PubMed Google Scholar

Contributions

LJ, XS, XK, YW and DM conceived and designed the study. XK recruited patients and performed clinical evaluation. XK, JH, JL and KG helped collected clinical samples and data. XS and XK performed bioinformatic analyses and experiments. LJ and XK supervised the study. LJ and XS wrote the original draft. All authors reviewed the manuscript and approved the submission. LJ is responsible for the overall content as guarantor. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Lindi Jiang.

Ethics declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Zhongshan Hospital Affiliated to Fudan University (B2013-115(3)) and complied with the 1975 Declaration of Helsinki. All study participants provided signed written informed consent. Animal studies were authorized by the Institutional Animal Care and Use Committee (IACUC) of Charles River Accelerator & Development Laboratories (Shanghai) Co., Ltd. (P202312250001).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1. Supplementary methods.

12916_2025_4434_MOESM2_ESM.docx

Additional file 2. Table S1 Clinical characteristics and pathological features of BD patients and control subjects enrolled for histopathological analysis. Table S2 Clinical characteristics of subjects enrolled for scRNA-sequencing. Table S3 Clinical characteristics of subjects enrolled for experimental cohort.

12916_2025_4434_MOESM3_ESM.docx

Additional file 3. Fig S1-[Pathological features of the ascending aortae in patients with BD and control subjects]. Fig S2-[Infiltration and distribution of immune cells in the ascending aorta of BD]. Fig S3-[Quality control and cell cluster identification in our scRNA-seq data]. Fig S4-[Characteristic features of gene expression and proportion of macrophage subsets]. Fig S5-[Integrated analysis of macrophages with public datasets]. Fig S6-[THBS1high macrophages demonstrated pro-inflammatory and metabolically activated characteristics]. Fig S7-[Characterization of THBS1-overexpressing macrophages compared to control macrophages]. Fig S8-[Activity of top-ranked regulons in Module M6 between BD patients and controls across different macrophage subtypes]. Fig S9-[ETS2 was essential for THBS1-induced TNF-α upregulation in macrophages]. Fig S10-[Altered intercellular communication in the aorta of BD]. Fig S11-[Phenotypic transition of SMCs in the aorta of BD]. Fig S12-[THBS1high macrophages contributed to the vascular inflammation in angiotensin II-induced humanized VBD mice]. Fig S13-[Validation of BD PBMC humanized mice model]. Fig S14-[Baseline vascular assessment in control macrophage-reconstituted mice without CaCl₂/LPS treatment].

Additional file 4. Original, uncropped gels/blots.

Additional file 5. ARRIVE guidelines.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sun, X., Kong, X., Hou, J. et al. THBS1high macrophages promote aortic aneurysm formation by orchestrating an inflammatory and smooth muscle cell phenotypic transition program in vascular Behçet’s disease. BMC Med 23, 631 (2025). https://doi.org/10.1186/s12916-025-04434-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • DOI: https://doi.org/10.1186/s12916-025-04434-y

Keywords

BMC Medicine

ISSN: 1741-7015

Contact us