Gut microbiota dysfunction mediates stress-exacerbated aortic dissection via the bacteroides vulgatus-outer membrane vesicles-stearic acid-JNK/MAPK axis

Abstract

Background

Aortic dissection (AD) is a lethal condition involving vascular smooth muscle cell (VSMC) transformation and extracellular matrix degradation. While gut microbiota dysbiosis is implicated in cardiovascular diseases, its role in stress-exacerbated AD pathogenesis is unknown. This study investigates the mechanism linking chronic restraint stress (CRS) to AD progression via gut microbiota modulation.

Methods

A β-aminopropionitrile (BAPN)-induced AD mouse model combined with CRS was utilized. Aortic dilation, mortality, and VSMC phenotype shift (assessed via α-SMA/SM22α and OPN/MMP2 expression) were evaluated. Gut microbiota composition was analyzed using 16 S rRNA sequencing. Microbiota depletion was achieved via antibiotics, and fecal microbiota transplantation (FMT) from CRS-exposed mice was performed. Serum metabolomics analysis, incorporating liquid chromatography-mass spectrometry (LC-MS), has demonstrated that outer membrane vesicles (OMVs) derived from Bacteroides vulgatus (B. vulgatus) contain high levels of the key metabolite stearic acid (SA). In vitro effects of stearic acid (SA) on AngII-induced JNK phosphorylation in VSMCs were tested, with validation using the JNK agonist anisomycin. Statistical analyses included correlation tests and appropriate comparisons (e.g., t-tests, ANOVA).

Results

CRS significantly accelerated aortic dilation, increased mortality, and promoted a synthetic VSMC phenotype (decreased α-SMA/SM22α, increased OPN/MMP2) in BAPN-treated mice. 16 S sequencing revealed CRS reduced gut microbiota diversity, particularly depleting B. vulgatus, which correlated negatively with AD severity. Antibiotic-mediated microbiota ablation mitigated CRS-aggravated AD, while FMT from CRS mice exacerbated it. Metabolomics identified stearic acid (SA), a metabolite derived from OMVs of B. vulgatus, as negatively correlated with aortic diameter. SA supplementation inhibited VSMC synthetic transformation, reduced AD incidence, and suppressed JNK/MAPK pathway activation in vivo. Mechanistically, SA attenuated AngII-induced JNK phosphorylation in VSMCs in vitro, an effect reversed by the JNK agonist anisomycin.

Conclusions

CRS exacerbates the pathogenesis of AD by disrupting the gut microbiota, particularly by reducing the abundance of B. vulgatus and the levels of SA, which is a metabolite encapsulated in the OMVs of B. vulgatus. This leads to unchecked JNK/MAPK signaling, driving detrimental VSMC transformation. Restoration of SA inhibits this pathway and mitigates AD progression. Targeting the gut microbiota-B. vulgatus-SA axis presents a novel therapeutic strategy for stress-aggravated AD.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04369-3.

Introduction

Aortic dissection (AD) represents a life-threatening cardiovascular emergency characterized by clinically significant mortality rates, primarily attributable to structural failure of the aortic wall manifested through an intimal tear and associated acute clinical presentations ranging from sudden-onset severe pain to fulminant cardiogenic collapse [1]. Epidemiologic data indicate a predominant occurrence between the 65–75 years of life, with population-based studies reporting annual incidence rates of 3–5 cases per 100,000 individuals [2]. While surgical intervention and endovascular repair remain the primary therapeutic modalities for confirmed AD cases, emerging clinical challenges highlight the critical need for adjunctive strategies to mitigate disease progression. Histopathological analyses reveal three hallmark pathological alterations in AD-affected aortas: (1) catastrophic depletion of vascular smooth muscle cells (VSMCs); (2) progressive extracellular matrix (ECM) degradation; and (3) elastic lamellae fragmentation within the medial layer [3]. VSMCs constitute fundamental structural and functional components of the vascular tunica media, orchestrating hemodynamic regulation through tonic modulation of vascular resistance and maintaining structural homeostasis. Emerging evidence positions VSMC dysregulation as a central pathomechanism driving maladaptive vascular remodeling following onset of medial degeneration in AD pathogenesis [4–7]. Accumulating studies demonstrate that phenotypic switching of VSMCs, characterized by dynamic alterations in proliferative capacity, apoptotic susceptibility, migratory potential, and contractile marker expression, which constitutes a critical transitional phase in AD progression [8]. These findings collectively suggest that targeted modulation of VSMC plasticity may represent a viable therapeutic paradigm for AD management.

The gut microbiota encompasses a complex ecological consortium of commensal bacteria, eukaryotic microorganisms, archaeal species, and viral particles, among which bacteria predominate as primary colonizers [9]. This microbial community plays a crucial role in maintaining intestinal homeostasis by metabolizing nutrients, regulating host immune functions, and participating in host defense mechanisms. For example, Iida et al. found significant differences in the component of the intestinal flora between non-cancer and cancer samples, indicating that intestinal flora had a significant effect on the progression of cancer [10]. Su et al. Indicated that bacteroides fragilis can generate short-chain fatty acids as well as indole compounds, which had anti-inflammatory influences via modulating the function and differentiation of Treg and Th17 cells [11]. Yan et al. suggested that intestinal flora regulated blood pressure through modulating the synthesis of intestinal-derived corticosterone in high salt-induced hypertension [12]. However, due to the presence of the mucus layer, communication between the intestinal microbiota and the host does not occur via direct cell-to-cell contact. Recent studies have demonstrated that bacteria can deliver bioactive molecules to host cells via OMVs; these OMVs play diverse physiological and pathological roles in interactions between bacteria and the host. Bacterial OMVs are spherical nanoparticles, ranging in size from 20 to 400 nm, naturally secreted by bacteria. They play a crucial role in bacterial communication with the external environment [13, 14]. A previous study showed that there was a significant difference in the composition of the gut flora in Ang II-induced abdominal aortic aneurysm mice [15]. However, the association and mechanism between AD and gut microbes is still to be explored.

Emerging clinical evidence increasingly highlights the association between gut microbiota dysbiosis and AD. For example, a recent clinical study by Liu et al. identified the gut microbiota and their metabolites have a profound impact on the onset and progression of AD [16]. Another clinical investigation reported a correlation between circulating levels of gut microbiota-derived metabolites, such as trimethylamine N-oxide (TMAO), and both the prevalence and severity of AD [17]. These clinical observations are supported by a small but growing body of preclinical research. For example, studies utilizing β-aminopropionitrile (BAPN)-induced AD models, similar to the model employed in our research, have demonstrated that gut microbiota-derived tryptophan metabolite indole-3-carboxaldehyde ameliorates AD [18], while the microbial metabolite butyrate exerts protective effects against AD via GPR41 signaling [19]. However, the specific bacterial taxa responsible for these effects, as well as the precise mechanistic pathways connecting gut microbes to AD pathogenesis, particularly under chronic stress conditions, remain largely unelucidated. Our study was designed to address this critical knowledge gap.

Contemporary public health challenges highlight chronic psychological stress as a pervasive risk modulator for multisystem pathology, including elevated susceptibility to neuropsychiatric disorders (e.g., major depressive disorder) [20, 21] and functional gastrointestinal conditions (e.g., irritable bowel syndrome) [22–24]. Stress-associated diseases involve in complex and multifactorial etiology and trajectory. Recently, increasing attention was attached to the underlying effect of microbiota–gut–brain axis signaling in the aetiologies as well as therapy of stress-correlated disorders [25, 26]. For example, Preclinical models establish that CRS exposure induces significant gut microbiome dysbiosis [27–29]. What’s more, evidence suggested that a lot of microbiota-targeted interventions can alleviate CRS-related deficiency in mice, including the application of live bacterial strains [29, 30], and the supplement of dietary fibre indigestible to the host, which are fermented by the intestinal microbiota [27]. Notwithstanding these advances, the potential interplay between stress-induced microbiome alterations and AD progression remains uncharted territory in cardiovascular research. In this investigation, our objective was to assess the pivotal bacterial phenomena linked to the effects of CRS on AD. Our results demonstrate that CRS diminishes the population of B. vulgatus within the gut and reduces the concentration of its beneficial metabolite, SA. This disruption leads to the attenuation of SA-induced inhibition of the MAPK pathway, thereby facilitating the progression of AD. These findings furnish a theoretical foundation for the advancement of microbiome-based therapeutic strategies for early-stage AD.

Materials and methods

Animal

All animal experiments received approval from the Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. The housing and management of mice were strictly in accordance with the guidelines established by the Institutional Animal Care and Use Committee (IACUC) of Huazhong University of Science and Technology. Male C57BL/6J mice, aged three weeks, were procured from Charles River (Beijing, China). Male C57BL/6J mice (n = 10 per group) were used in all experiments unless otherwise specified. Prior to the experimental procedures, the mice were acclimatized in cages for a duration of three days under controlled environmental conditions: a temperature of 20 °C, humidity at 50%, and a 12-hour light/12-hour dark cycle. AD was induced in three-week-old male mice using BAPN (MedChemExpress, New Jersey, USA), following a previously established protocol [31]. The mice were maintained on a standard diet and were provided with freshly prepared BAPN dissolved in drinking water at a dosage of 0.5 g/kg/day for a period of 50 days. Vevo2100 ultrasound system (VisualSonics, FUJIFILM, Toronto, Canada) was used to evaluate aorta morphology of mice. Determination of the maximum internal diameter of the thoracic aorta was performed by using longitudinal images of the aortic arch in mice. For any mice that succumbed prior to the scheduled experimental endpoint, an immediate necropsy was conducted. Blood clots were noted within the thoracic cavity of these mice, and their aortas were promptly excised to ensure the freshness of the tissue. Mice that survived until the experimental endpoint were euthanized through an overdose of pentobarbital sodium, and aortic tissue samples were subsequently collected for further analysis.

Chronic restraint stress (CRS)

Mice were subjected to CRS by being confined in a 50 mL ventilated conical tube for 3 to 4 h each day. In contrast, control mice that were not subjected to stress had unrestricted access to food and water within their home cages. The assessment of depression-like behaviors was conducted using the Sucrose Preference Test (SPT) and the Forced Swim Test (FST). For the SPT, individual mice were housed separately and provided with two identical bottles: one filled with plain drinking water and the other containing a 2% sucrose solution. After a duration of 6 h, the positions of the two bottles were exchanged to negate any side preference. The total consumption of both solutions was recorded over a 12-hour timeframe, with sucrose preference determined as the ratio of sucrose solution intake to plain water intake. In the FST, each mouse was placed into a cylindrical container filled with water at a temperature of 25 °C for a period of 6 min. The duration of immobility—characterized by the absence of struggling or passive floating—was documented during the final 4 min utilizing the EthoVision XT video tracking system (Noldus, Netherlands).

Fecal microbiota transplantation (FMT)

Recipient mice underwent pretreatment with an antibiotic regimen consisting of ampicillin (1 g/L), metronidazole (1 g/L), vancomycin (0.5 g/L), and neomycin (0.5 g/L), administered daily via oral gavage for a duration of seven consecutive days to eliminate their native gut microbiota. Following this, AD models were developed in both donor and recipient mice. Fresh fecal specimens from donor mice were collected each day, homogenized in sterile phosphate-buffered saline (PBS) at a concentration of 100 mg/mL, and vortexed for ten seconds. The resultant mixture was subjected to centrifugation at 800 × g for three minutes, after which the supernatant was retrieved as the fecal microbiota transplant. Recipient mice received 200 µL of the fecal transplant derived from donors through daily oral gavage.

Bacterial culture and administration

Bacteroides vulgatus (ATCC 8482) was obtained from the American Type Culture Collection (Manassas, VA, USA) and cultured anaerobically in Bacteroides BHI broth at 37 °C for 48–72 h under an atmosphere of 85% N₂, 10% H₂, and 5% CO₂. Bacterial cells were harvested in the mid-logarithmic growth phase by centrifugation at 5,000×g for 15 min at 4 °C. The pellet was washed twice and resuspended in sterile, reduced phosphate-buffered saline (PBS) containing 15% glycerol. The bacterial concentration was adjusted to 1 × 10^9 colony-forming units (CFU) per mL based on optical density measurements (OD₆₀₀) and verified by plating serial dilutions on agar plates. For oral gavage, fresh bacterial suspensions were prepared daily from frozen stocks. Mice in the designated treatment groups (BAPN and BAPN + CRS) received a daily oral gavage of 200µL of the bacterial suspension (approximately 2 × 10^8 CFU per mouse) throughout the 50-day experimental period. Control groups received an equal volume of the sterile PBS-glycerol vehicle.

Cell culture and treatment

Primary vascular smooth muscle cells (VSMCs) were isolated from six-week-old male Sprague-Dawley (SD) rats, following previously established protocols [32]. The cells were cultured in DMEM/F12 medium supplemented with with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin, and maintained at 37 °C in a humidified environment containing 5% CO₂. To mimic the pathological alterations associated with AD, VSMCs were stimulated with angiotensin II (Ang II) for subsequent experiments.

Extraction and identification of OMVs

B. vulgatus was cultured anaerobically at 37 °C for 72 h. The bacterial suspension was centrifuged at 4 °C and 5,000 × g for 15 min. The resulting supernatant was then centrifuged at 4 °C and 15,000 × g for 20 min. The collected supernatant was sequentially filtered through 0.45 μm and 0.22 μm membranes to remove bacterial debris and other impurities. The filtrate was subjected to ultracentrifugation (Beckman Coulter, Fullerton, USA) at 4 °C and 150,000 × g for 2 h to pellet the OMVs. The OMVs were resuspended in sterile PBS and stored at − 80 °C until further use. The protein concentration of the OMVs was determined using a bicinchoninic acid (BCA) protein assay kit. Morphological examination of OMVs was performed using transmission electron microscopy (Thermo, Waltham, USA). The size distribution and particle concentration of OMVs were analyzed using a NanoSight NS300 system (NanoSight Technology, Malvern, UK).

In vitro effect of corticosterone on OMVs production by B. vulgatus

B. vulgatuswas cultured under anaerobic conditions at 37 °C. At the mid‑logarithmic growth phase, bacterial cultures were treated with 1 µM corticosterone (Sigma-Aldrich) or an equal volume of vehicle control. Following treatment, the cultures were further incubated anaerobically for 24 h. OMVs were harvested from the culture supernatant by sequential centrifugation. Nanoparticle tracking analysis was used to determine the particle concentration and size distribution of the OMVs.

Hematoxylin and eosin (H&E) staining

Fresh aortic tissue specimens were preserved in 4% paraformaldehyde for more than 24 h. Selected segments of the aorta were subsequently excised, subjected to a graded series of alcohol dehydration, and embedded in paraffin. The paraffin-embedded samples were sliced into sections measuring 4 μm in thickness. These sections were then deparaffinized using xylene and rehydrated through a descending alcohol gradient. For staining purposes, the sections were initially immersed in a hematoxylin solution for nuclear visualization, followed by rinsing in running tap water to eliminate any excess dye. The sections were then counterstained with eosin Y to visualize the cytoplasm and extracellular matrix. Post-staining, the sections underwent dehydration via an ascending alcohol gradient, were cleared in xylene twice for 5 min each, and were finally mounted using a neutral gum mounting medium for subsequent microscopic analysis.

Elastic-van gieson (EVG) staining

Following the processes of deparaffinization and rehydration, tissue sections were submerged in Verhoeff’s hematoxylin solution for the staining of elastic fibers, followed by a thorough rinse. The sections were then treated with a ferric chloride differentiation solution, which selectively stained the elastic fibers until they manifested as black against a gray backdrop. Thereafter, the sections underwent counterstaining with Van Gieson’s solution, a blend of picric acid and acid fuchsin, to render the collagen fibers visible. Upon completion of the staining process, the sections were washed with absolute ethanol to dehydrate and eliminate any residual dyes. Finally, the sections were cleared in xylene and mounted using a neutral gum mounting medium in preparation for microscopic examination.

Western blot analysis

Aortic tissues and VSMCs were extracted and lysed with the radio-immunoprecipitation assay (RIPA) buffer (Beyotime, China). Protein concentrations were then detected using the BCA protein assay kit (Thermo Fisher, Scientific Inc USA). Following this, protein (about 30 µg) was separated via sodium dodecyl sulfate (SDS)–PAGE, followed by electrotransfered onto a 0.45-µm polyvinylidene difluoride (PVDF) membranes. Then the membranes were blocked with 5% fat-free milk for 1 h, incubated by using primary (ERK, ab201015, abcam; p-ERK, ab32537, abcam; JNK, 51151-1-AP, Proteintech; p-JNK, 80024-1-RR, Proteintech; P38, 14064-1-AP, Proteintech; p-P38, 28796-1-AP, Proteintech; SM22α, 10493-1-AP, Proteintech; α-SMA, 14395-1-AP, Proteintech; OPN, 25715-1-AP, Proteintech; MMP2, 10373-2-AP, Proteintech; GAPDH, 10494-1-AP, Proteintech) for overnight at 4 °C, Next, the membrane was incubated with secondary antibodies at room temperature for 1 h. Finally, protein expression was analyzed through the ECL kit (Beyotime). And the protein grayscale of the protein bands was assessed by the ImageJ software. The raw, uncropped data of WB results is supplemented in the Fig. S9.

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA was extracted from the aortic tissues with TRIzol reagent (Beyotime, China) based on the manufacturer’s instructions. NanoDrop ND-2000 spectrophotometer (Thermo Fisher, USA) was used to evaluate the concentration as well as quality of the purified RNA. The reverse transcription was performed by applying HiScript III All-in-one RT SuperMix (Vazyme, R333-01). AceQ SYBR Green Master Mix (Vazyme, Q141-02) was employed to carried out the real-time qRT-PCR analyses. The relative expression of targeted genes was assessed with the 2 − ΔΔCt method and β-actin was employed as an internal standard. The primer sequences were shown in the Supplementary Table 1.

Immunofluorescence staining

The frozen aortic tissues of the mouse were fixed by 4% paraformaldehyde for 30 min, permeabilized with freshly 0.5% TritonX-100 for 30 min as well as blocked for 1 h in 5% normal goat serum (NGS). Following this, the tissues were incubated overnight at 4 °C and the following primary antibodies were employed: α-SMA (1:2000, 14395-1-AP, Proteintech, USA), SM22α (1:500, GB11366, Servicebio, China), OPN (1:200, GB11500, Servicebio, China), MMP2 (1:200, GB11130, Servicebio, China), then the following secondary antibody Goat Anti-Rabbit IgG (H + L) Alexa Fluor^® 488(1:200, PN0108, Pinofi, China) were incubated for 1 h at 37 °C. Nuclei were counterstained by using DAPI. Finally, the immunofluorescent signals were analyzed with a laser scanning confocal microscope (Zeiss LSM 780, Germany).

16S rRNA Sequencing and analysis

Fecal specimens from mice were gathered and preserved at -80 °C until subsequent analysis. Genomic DNA extraction from the fecal specimens was performed utilizing the QIAamp Rapid DNA Kit (Qiagen, USA). The hypervariable regions V3-V4 of the 16 S rRNA gene were amplified through polymerase chain reaction (PCR). Library preparation was executed using the MetaVx™ Library Preparation Kit (GENEWIZ, USA) in accordance with the manufacturer’s guidelines. Sequencing was accomplished on the Illumina MiSeq platform (Illumina, USA). Bioinformatic evaluation of the 16 S rRNA sequencing data was carried out employing the QIIME 2 software suite (version 2023.2). The sequences underwent quality filtering, were clustered into operational taxonomic units (OTUs) at a 97% similarity threshold, and were aligned with the SILVA reference database (v138) for taxonomic identification.

Analysis of the metabolomic profile via UPLC Q-TOF/MS

Fecal/plasma samples were processed using a two-phase extraction protocol with methanol: water (4:1 v/v) and chloroform. After centrifugation (15,000 × g, 15 min, 4 °C), supernatants were concentrated by vacuum centrifugation and reconstituted in 80% methanol. Quality control (QC) samples were prepared by pooling equal volumes from all samples.

LC-MS analysis

Metabolite separation was performed on a UHPLC system (Agilent 1290) equipped with a HILIC column (2.1 × 100 mm, 1.7 μm) using gradient elution (0.1% formic acid in water/acetonitrile). Mass spectrometry was conducted using a Q-TOF mass spectrometer (Sciex X500R) in both positive and negative ionization modes (m/z 50-1000). Instrument parameters included: drying gas 10 L/min, nebulizer 35 psig, capillary voltage 3500 V, fragmentor 120 V, and skimmer 65 V.

Data processing

Raw data were processed using MS-DIAL or XCMS Online (v3.0) with parameters: mass tolerance 10 ppm, minimum peak height 1000, retention time tolerance 0.2 min. Metabolites were identified by matching to HMDB (v4.0) and METLIN databases with MS/MS confirmation when available. Multivariate analysis (PLS-DA, OPLS-DA) was performed using SIMCA-P (v15.0) after Pareto scaling.

Integration with microbiome data

Spearman correlations between significantly altered metabolites (VIP > 1.0, p < 0.05) and microbial taxa (relative abundance > 0.1%) were calculated using R (v4.1). Metabolic pathways were reconstructed via KEGG Mapper using both identified metabolites and predicted metagenomic functions (PICRUSt2/KEGG).

Statistical analysis

Statistical evaluations were conducted utilizing GraphPad Prism v9.0.0 (GraphPad Software, USA). The data are expressed as mean ± standard deviation (SD). Between-group statistical significance was determined using two-tailed Student’s t-tests for comparisons of two groups. For analyses involving three or more groups, one-way ANOVA was employed. Survival rates were assessed using the Kaplan-Meier method, and relationships between variables were examined with Spearman’s rank correlation coefficient. A p-value of less than 0.05 was deemed statistically significant.

Result

CRS accelerates the development of BAPN-induced AD in murine models

To explore the ramifications of CRS on the progression of AD, we executed in vivo studies employing a BAPN-induced AD mouse model. Both the control cohort and BAPN-treated mice were subjected to approximately 50 days of CRS to assess its influence on AD advancement (Fig. 1A). Our findings indicated that, in comparison to their counterparts who were not subjected to CRS, both the control and BAPN-treated groups demonstrated a marked increase in immobility duration during the forced swim test (FST) and a notable reduction in the preference for sucrose water in the sucrose preference test (SPT) following CRS exposure (Fig. 1B-C). Furthermore, we noted that during the 50-day CRS exposure period, 20% (n = 2) of the BAPN-treated mice succumbed to AD and rupture, while 30% (n = 3) developed AD. Notably, whereas CRS alone did not result in fatalities, 60% (n = 6) of the mice in the BAPN + CRS co-treatment group perished due to AD and rupture, with 40% (n = 4) manifesting AD (Fig. 1D-F). Although AD was identified in both BAPN-treated groups, it was evident that CRS significantly facilitated its progression. Vascular ultrasound imaging and assessment of the maximum aortic diameter on the 30th day post-modeling further corroborated that CRS exacerbated BAPN-induced aortic dilation in comparison to the BAPN-only treatment group (Fig. 1G-H). Histological examinations utilizing hematoxylin, eosin, and elastic Van Gieson staining of the ascending aorta, aortic arch, and proximal descending aortic segments indicated that the development of dissecting aneurysms and elastin disorganization was markedly more severe in the BAPN + CRS co-treatment group than in the BAPN-only treatment group (Fig. 1I). In contrast, no significant differences were observed regarding maximum aortic diameter and AD incidence between the non-BAPN-treated control and CRS-treated groups (Fig. 1D-I). In summary, these findings bolster the proposition that CRS intensifies the progression of AD.

Fig. 1.

Chronic restraint stress accelerates β-aminopropionitrile (BAPN)-induced aortic dissection (AD) progression in mice. (A–I) Experimental analyses of mice in Control, CRS, BAPN, and BAPN + CRS groups. (A) Schematic overview of the experimental design. (B–C) Immobility time in the forced swim test (FST) and sucrose preference rate in the sucrose preference test (SPT). (D) Survival rates analysis. (E) Incidence of AD. (F) Representative macroscopic images of aortas. (G) Representative ultrasound images of thoracic aortas. Scale bar: 1 mm. (H) Maximum aortic diameter measurements. (I) Representative hematoxylin and eosin (H&E) and Elastic-Van Gieson (EVG)-stained aortic sections. Scale bars: 100 μm (overview), 50 μm (inset). **p < 0.01, ***p < 0.001

CRS facilitates the phenotypic shift of VSMCs from a contractile to a synthetic phenotype within the aortic tissues of BAPN-treated mice

To evaluate the influence of CRS on the phenotypic alteration of VSMCs in the aortic tissues of BAPN-treated mice, the research team utilized Western blot and quantitative real-time polymerase chain reaction (qRT-PCR) methodologies for detection. Through the application of Western blot and qRT-PCR techniques, we quantified the expression levels of contractile markers (α-SMA), SM22α) and synthetic markers (OPN, MMP2) in the aortas of BAPN-treated mice. The findings revealed that, in comparison to the non-CRS-exposed control group, there were no notable differences in the protein and RNA levels of both contractile and synthetic markers in the aortas of the CRS-only group. However, in the BAPN-only group, the expression levels of contractile markers were significantly diminished, while those of synthetic markers experienced a substantial increase in the aortas of these mice. These alterations were even more pronounced in the group subjected to co-treatment with BAPN and CRS (Fig.S1A-B). Consistent results were also obtained from immunofluorescence staining experiments examining the protein levels of α-SMA and SM22α (Fig.S1C-D). Collectively, these findings corroborate that CRS promotes the phenotypic transition of VSMCs from a contractile to a synthetic phenotype in the aortic tissues of BAPN-treated mice.

CRS facilitates the onset of AD by perturbing the gut microbiota

Recognizing the pivotal function of the gut microbiota in both cancer advancement and suppression, we performed 16 S rRNA gene sequencing analysis on the gut microbiota present in the fecal samples of mice across control, CRS, BAPN, and BAPN + CRS groups to investigate the underlying mechanisms of CRS-induced AD progression. The results regarding intersection size indicated notable differences in the composition of gut microbiota among the various mouse groups (Fig.S2A). Furthermore, although differences in alpha diversity were observed, comparison between the non-CRS-exposed control group (Control) and the CRS-only group revealed no statistically significant differences in gut microbiota abundance. In contrast, the gut microbiota abundance was markedly diminished in the BAPN and CRS co-treatment group (BAPN + CRS) relative to the BAPN-only treatment group (Fig.S2B-C), implying the potential involvement of gut microbiota in the CRS-induced AD development process. To substantiate the validity of the sequencing outcomes, we administered an antibiotic mixture intervention to the mice in both the BAPN-treated and BAPN + CRS co-treatment groups (Fig.S2D). Quantitative PCR analysis of the 16 S rRNA gene revealed a reduction in total bacterial load following Abx treatment (Fig. S2E), confirming the efficacy of the antibiotic regimen in depleting the microbiota. The antibiotic mixture effectively depleted the gut microbiota, resulting in comparable microbiota abundance in both groups of mice. As anticipated, the antibiotic intervention successfully inhibited CRS-induced AD progression (Fig. S2F-I). Hematoxylin-eosin (HE) staining and Victoria blue-elastic van Gieson (EVG) staining results of aortic sections further corroborated the crucial role of the antibiotic intervention in obstructing CRS-induced AD progression (Fig. S2J). Moreover, no significant differences were noted in the protein and RNA levels of both contractile and synthetic markers of vascular smooth muscle cells (VSMCs) in the aortas of mice between the two groups (Fig. S2K-L). The immunofluorescence staining results based on the protein levels of α-SMA and SM22α were consistent with these findings (Fig. S2M-N). To investigate the role of stress modified gut microbiota in modulating the progression of AD, fecal microbiota transplantation (FMT) was performed using donor microbiota obtained from either BAPN treated mice or BAPN + CRS treated mice. After transplantation, 16 S rRNA gene sequencing demonstrated that the microbial communities of FMT recipients closely resembled those of their respective donors (Fig. S3A). Moreover, principal coordinate analysis (PCoA) confirmed no significant separation between donor and corresponding FMT recipient groups (Fig. S3B), indicating successful microbial engraftment. Intriguingly, fecal samples from the BAPN + CRS group were found to induce AD progression in recipient mice (Fig. S3C-G). In summary, these findings suggest that CRS disrupts the equilibrium of the gut microbiota in mice, thereby facilitating the development of AD.

CRS facilitates the onset of AD by diminishing the prevalence of Bacteroides vulgatus

To elucidate the bacterial species that significantly contribute to the emergence of CRS-induced AD, we employed 16 S rRNA gene sequencing at the genus level. Our findings indicated that, in comparison to the control group, there was no notable difference in the gut microbiota composition within the CRS-only group of mice. Conversely, a significant disparity was observed in the relative abundance of the Bacteroides genus between the BAPN-only treatment group and the BAPN + CRS co-treatment group (Fig. 2A-C, Fig. S4A-E). To pinpoint the specific Bacteroides species exerting a crucial influence, we performed qRT-PCR analysis on prevalent Bacteroides species. The data revealed that, relative to the BAPN-only treatment group, the population of B. vulgatus was markedly reduced in the BAPN + CRS co-treatment group, while no significant variations were noted among the other four identified Bacteroides species across the different cohorts (Fig. 2D). B. vulgatus is a pivotal constituent of the gut microbiota, playing a role in metabolic regulation and the progression of diseases. Research has demonstrated that B. vulgatus can mitigate atherosclerosis by decreasing the synthesis of gut microbial lipopolysaccharides [33], and it is also implicated in the development of insulin resistance and modifications in bile acid metabolism [34]. Furthermore, it has been shown to predict immune-related adverse events associated with immune checkpoint blockade therapy for metastatic melanoma [35]. These investigations underscore its dual function in metabolic disorders and cancer treatment. Consequently, we propose that the reduction in B. vulgatus abundance may intensify the negative impacts of CRS on the advancement of AD. To test this hypothesis, we administered B. vulgatus daily to mice within both the BAPN-treated and BAPN + CRS co-treatment groups. The results illustrated that B. vulgatus supplementation could mitigate CRS-induced damage to arterial structural integrity and inhibit the progression of CRS-induced AD (Fig. 2E-J; Fig. S5A-D). In conclusion, B. vulgatus may play a protective role against the exacerbation of AD due to CRS in murine models.

Fig. 2.

Reduced abundance of Bacteroides vulgatus under chronic restraint stress exacerbates AD progression. (A) Relative genus abundance in indicated four groups. (B) Relative genus abundance in each single sample. (C) A difference analysis was conducted on the five genera with the highest abundance among the four groups. (D) Bacteroides subspecies levels in stressed vs. non-stressed CRC mice (qRT-PCR). (E–J) BAPN and BAPN + CRS mice treated with PBS or B. vulgatus for 50 days. (E) Experimental timeline. (F) Survival rates analysis. (G) AD incidence. (H) Macroscopic aortic images. (I) Maximum aortic diameter. (J) H&E and EVG-stained aortic sections. Scale bars: 100 μm (overview), 50 μm (inset). *p < 0.05, **p < 0.01

Outer membrane vesicles (OMVs) derived from bacteroides vulgatusinhibit the progression of AD exacerbated by CRS

Based on the findings, we investigated how B. vulgatus inhibits the progression of AD under CRS. To assess the impact of bacterial metabolites on VSMCs, we cultured VSMCs in a basal medium containing supernatant from B. vulgatus. The results from Western blot analysis revealed a significant increase in contractile markers and a notable decrease in synthetic markers in VSMCs treated with the bacterial supernatant compared to those maintained solely in the basal medium (Fig.S6A). This suggests that the supernatant from B. vulgatus contains unidentified factors that can prevent VSMCs from switching their phenotype from a contractile to a synthetic state. Furthermore, there is growing evidence that bacteria can release bioactive molecules to host cells via OMVs, which can influence various physiological processes in the host [36, 37]. Consequently, we hypothesized that OMVs from B. vulgatus play a crucial role in reducing the progression of stress-related AD. We isolated OMVs from the culture supernatant of B. vulgatus using ultracentrifugation (Fig. S6B). Transmission electron microscopy confirmed that the isolated vesicles displayed a typical bilamellar membrane structure (Fig. S6C). Additionally, nanoparticle tracking analysis showed that the diameters of the OMVs ranged from 20 to 270 nm (Fig. S6D). Western blot analysis further confirmed the presence of the Gram-negative outer membrane proteins OmpA and OmpC in the vesicle preparations (Fig. S6E). Collectively, these findings validate the successful isolation of OMVs from B. vulgatus. We next assessed the stability of OMVs under gastrointestinal conditions. OMVs were incubated at 37 °C for 2 h in simulated salivary, gastric, and intestinal fluids, which contained α-amylase, pepsin, and pancreatin with bile acids, respectively (Fig. S6F). The experiments showed that none of these digestive conditions altered the morphology or particle size of the OMVs, and there was no significant membrane fragmentation observed (Fig. S6G-H). These results suggest that the OMVs from B. vulgatus maintain their structural integrity while passing through the digestive tract. To investigate whether these OMVs could be taken up by recipient cells, we co-cultured OMVs isolated from B. vulgatus (OmpA-Cre) with vascular smooth muscle cells (VSMCs) that were genetically modified to express LoxP-DsRed-Stop-LoxP-GFP. Fluorescence microscopy revealed that some VSMCs internalized the OMVs, evidenced by a shift from red to green fluorescence (Fig. S6I-J).

To determine whether the host stress environment can directly modulate OMVs production by B. vulgatus, bacterial cultures were treated in vitro with corticosterone (1 µM), the predominant glucocorticoid in mice. Notably, the concentration of secreted OMVs, as measured by nanoparticle tracking analysis (Fig. 3A), showed no significant difference between the corticosterone-treated and vehicle control groups. These results indicate that the stress hormone corticosterone does not directly suppress the ability of B. vulgatus to generate OMVs. To assess the impact of B. vulgatus OMVs on CRS-induced AD progression, mice in both the BAPN group and the BAPN + CRS group received daily doses of B. vulgatus OMVs. The findings indicated that these OMVs significantly reduced the progression of AD induced by CRS (Fig. 3B-G). Additionally, there were no notable differences in the protein or mRNA levels of contractile and synthetic markers in aortic VSMCs between the two groups (Fig. 3H-I). Immunofluorescence staining results corroborated these findings (Fig. 3J-K). Collectively, these data demonstrate that B. vulgatus OMVs can inhibit the worsening of AD caused by CRS.

Fig. 3.

OMVs mitigates CRS-accelerated AD progression. (A) Particle concentration of OMVs in culture supernatants from corticosterone treated or vehicle control groups, as measured by nanoparticle tracking analysis (NTA). (B–K) Mice in the BAPN and BAPN + CRS groups were fed with OMVs for 50 days. (B) Experimental design. (C) Survival rates. (D) AD incidence. (E) Maximum aortic diameter. (F) Macroscopic aortic images. (G) H&E and EVG-stained aortic sections. Scale bars: 100 μm (overview), 50 μm (inset). (H) Western blot analysis of SM22α, α-SMA, OPN, and MMP2. (I) qRT-PCR analysis of mRNA levels of SM22α, α-SMA, OPN, and MMP2. (J) Immunofluorescence staining of SM22α and α-SMA. Scale bars: 100 μm (overview), 50 μm (inset). (K) Quantification of SM22α and α-SMA positive areas. *p < 0.05, **p < 0.01, ***p < 0.001

The outer membrane vesicles (OMVs) derived from bacteroides vulgatusare rich in stearic acid

It is widely recognized that the gut microbiota primarily exerts its biological effects through the production of metabolites. To explore the potential metabolites synthesized by B. vulgatus that may inhibit AD, we performed a metabolomic analysis on the fecal OMVs samples of mice. The 3D t − SNE discriminant analysis demonstrated significant variations in the fecal OMVs samples metabolic profiles between the BAPN-only treatment group and the BAPN and CRS co-exposure group of mice (Fig. 4A). The heatmap and volcanic plot indicated that among the top 10 metabolites identified, 4 metabolites were significantly elevated, whereas 6 metabolites were significantly lowered in the BAPN + CRS co-treatment group compared to the BAPN-only treatment group (Fig. 4B-C). Considering the reduced abundance of B. vulgatus in the intestinal tracts of mice subjected to CRS exposure, we proceeded with a Pearson correlation analysis between the aortic diameter and these 6 significantly altered metabolites. The results revealed that SA exhibited the most pronounced negative correlation with the aortic diameter (Fig. 4D). SA is a crucial saturated fatty acid that can lower low-density lipoprotein cholesterol and diminish the risk of atherosclerosis, in addition to promoting fatty acid β-oxidation to mitigate fat accumulation [38]. To confirm whether SA is produced as a metabolite by B. vulgatus, we employed liquid chromatography-mass spectrometry (LC-MS) to quantify the concentration of SA in the OMVs of B. vulgatus. The results revealed that the OMVs of B. vulgatus contain a high concentration of SA, with SA content varying depending on incubation time and bacterial concentration (Fig. 4E).

Fig. 4.

OMVs derived from Bacteroides vulgatus are notably abundant in stearic acid. (A) 3D plots of t-SNE discrimination analysis revealed clear separations of AD and AD + CRS mice. (B–C) Heatmap and volcano plot of differentially expressed metabolites between BAPN and BAPN + CRS group. (D) Correlation between SA levels and aortic diameter. (E) SA levels measured by LC-MS

The metabolite stearic acid derived from bacteroides vulgatus has been shown to impede the exacerbation of AD induced by CRS through the inhibition of the JNK/MAPK signaling pathway

To delve deeper into the mechanisms by which SA from B. vulgatus mitigates the progression of AD associated with CRS, we conducted RNA-seq and gene set enrichment analysis (GSEA) on aortic tissue specimens obtained from mice in the BAPN + CRS and BAPN + CRS+SA cohorts. KEGG pathway enrichment analysis revealed notable enrichment of the MAPK pathway in the BAPN + CRS samples (Fig. 5A-B). Subsequently, we examined the interplay between SA and the MAPK pathway via in vitro experiments. Western blot (WB) analysis was employed to assess the activation of the MAPK pathway (including ERK, JNK, and P38) in vascular smooth muscle cells (VSMCs) treated with SA and/or AngII. The findings indicated that the heightened JNK phosphorylation caused by AngII could be counteracted by SA, whereas the phosphorylation levels of ERK and P38 remained unaffected by either AngII or SA (Fig. 5C). Immunofluorescence assays further corroborated that the level of p-JNK was elevated in VSMCs exposed to AngII, and this effect was reversed by SA (Fig. 5D). To substantiate whether SA mediates its effects in vitro through the JNK/MAPK signaling cascade, we introduced the JNK-specific agonist anisomycin (ANI) into four groups of VSMCs: control, AngII, AngII + SA, and AngII + SA+ANI. The Western blot results demonstrated that, in comparison to the AngII + SA group, the AngII + SA+ANI group exhibited significantly elevated levels of p-JNK/JNK and synthetic markers, while the expression of contractile markers markedly diminished in VSMCs (Fig. 5E). Immunofluorescence analysis further confirmed that ANI treatment reversed the downregulation of p-JNK observed in the AngII + SA group (Fig. 5F). These results suggest that ANI negates the inhibitory effect of SA on JNK activation. To further assess whether the facilitation of AD by CRS is contingent upon the activation of the JNK/MAPK signaling pathway, we performed in vivo experiments with mice from the BAPN + CRS, BAPN + CRS+SA, and BAPN + CRS+SA + ANI groups. The results indicated that ANI reinstated the inhibitory influence of SA on CRS-induced AD progression (Fig. S7A-J). Collectively, these findings suggest that the OMV-packed metabolite SA from B. vulgatus effectively inhibits the promotion of AD driven by CRS through the suppression of the JNK/MAPK pathway (Fig. S8).

Fig. 5.

SA inhibits the JNK/MAPK pathway. (A) Top 10 enriched KEGG pathways based on gene set enrichment analysis (GSEA). (B) MAPK as the most significantly enriched pathway. (C) Western blot analysis of phosphorylated and total ERK, JNK, and p38 in SA- and/or Ang II-treated VSMCs. (D) Immunofluorescence images of p-JNK in VSMCs. Scale bars: 10 μm. (E) Western blot analysis of SM22α, α-SMA, OPN, MMP2, and JNK phosphorylation. (F) Immunofluorescence images of p-JNK in VSMCs. Scale bars: 10 μm

Discussion

The current research reveals an innovative mechanism through which CRS amplifies AD via gut microbiota dysbiosis and the subsequent activation of the JNK/MAPK signaling pathway. Our results emphasize the essential role of B. vulgatus and its metabolite SA, transported via OMVs, in mitigating stress-induced AD progression, presenting a microbiome-focused therapeutic strategy for early-stage AD intervention. While these findings enhance our comprehension of the gut-vascular axis, several limitations and broader implications merit discussion.

This work connects CRS, gut microbiota dysbiosis, OMVs, and vascular pathology—a set of interactions that has been previously underexplored in cardiovascular research. CRS is increasingly acknowledged as a modulator of systemic inflammation and metabolic dysfunction, yet its direct influence on aortic remodeling has remained unclear [39, 40]. By illustrating that stress-induced depletion of B. vulgatus and its OMVs-encapsulated SA exacerbates AD through JNK/MAPK activation, our study aligns with emerging frameworks that link microbial metabolites to vascular homeostasis. For example, recent research by Luo et al. underscored the role of lncRNA H19-encoded GMRSP in regulating VSMC metabolic reprogramming to alleviate AD, highlighting the therapeutic potential of targeting metabolic pathways in aortic disorders [41]. Likewise, the cardioprotective properties of SA, a saturated fatty acid known for its anti-atherogenic effects, resonate with studies that have identified microbial-derived metabolites as pivotal regulators of vascular inflammation and remodeling [42, 43]. Furthermore, our in vitro data demonstrate that the stress hormone corticosterone does not directly impair OMVs production by B. vulgatus. This finding suggests that the reduction in circulating Bacteroides derived OMVs observed in CRS mice is likely a consequence of stress induced decreases in B. vulgatus abundance, rather than a direct functional suppression of OMVs biogenesis. These results further underscore the central role of microbial population dynamics in the gut-vascular axis.

The identification of JNK/MAPK signaling as a pivotal mediator of SA’s protective effects further reinforces the significance of this pathway within cardiovascular pathophysiology. Previous investigations concerning myocardial infarction have shown that HIP-55-mediated suppression of MAPK/JNK signaling mitigates ferroptosis, illuminating the dual role of this pathway in cell survival and death [44]. Our findings extend this paradigm to AD, suggesting that JNK inhibition—whether through microbial metabolites or pharmacological agents—could represent a universal strategy for alleviating stress-exacerbated vascular conditions.

While our study elucidates the downstream consequences of B. vulgatus depletion, it raises a critical question regarding the upstream mechanism: how does CRS specifically lead to the loss of this bacterium? Although not the focus of the current investigation, emerging evidence suggests several plausible pathways. Chronic stress activates the hypothalamic-pituitary-adrenal (HPA) axis, resulting in elevated glucocorticoid levels that can substantially alter the gut microenvironment. First, stress-induced glucocorticoids are known to suppress the expression of mucin genes and reduce goblet cell numbers, thereby thinning the protective mucus layer [45, 46]. Bacteroides species, including B. vulgatus, are proficient at utilizing mucin glycans as a nutrient source [47]. A compromised mucus barrier, while potentially increasing bacterial access to mucins, may also intensify host-bacterial contact and immune surveillance. This could precipitate an inflammatory state that selectively affects certain symbiotic members [48]. Second, stress can alter gut motility and secretion, changing luminal physicochemical conditions that may become suboptimal for B. vulgatus proliferation [49]. Finally, catecholamines released during stress can directly influence bacterial growth; although this effect is broad, bacteria such as B. vulgatus may be particularly susceptible due to their specific receptor profiles or metabolic pathways [50]. Thus, it is plausible that CRS creates an intestinal milieu—via a combination of impaired barrier function, altered nutrient availability, and neuroendocrine signaling—that is particularly unfavorable for B. vulgatus, leading to its specific depletion.

Several limitations of this study should be acknowledged. First, our investigation focused primarily on B. vulgatus and SA, while the gut microbiota functions as a complex, interdependent community. Interactions with other species may modulate the observed effects. Second, the translational relevance from murine models to human AD requires validation, given the interindividual variability in human gut microbiota and stress responses. Finally, while our in vitro data support OMVs stability and uptake, direct in vivo tracking of administered B. vulgatus-OMVs to the aortic wall would further solidify the mechanistic link.

These limitations delineate clear avenues for future research. Multi-omics studies are needed to map the broader microbial and metabolic networks involved in AD progression. Prospective clinical cohorts with longitudinal biospecimen collection are essential to validate the B. vulgatus-SA axis in patients. Furthermore, elucidating the precise upstream mechanisms by which CRS impairs the gut microenvironment to selectively disadvantage B. vulgatus will be crucial. Finally, therapeutic strategies employing engineered OMVs or SA-based formulations warrant preclinical development for stress-exacerbated cardiovascular conditions.

Conclusions

In conclusion, this research elucidates a mechanistic connection between persistent stress, dysbiosis of gut microbiota, and the advancement of AD, facilitated by the B. vulgatus-OMVs-SA-JNK/MAPK pathway. Although challenges regarding the generalizability of the models and the complexity of the mechanisms remain, our findings lay the groundwork for microbiome-targeted interventions in cardiovascular medicine. Future investigations should emphasize translational validation and examine synergistic strategies that integrate microbial modulation with pathway-specific inhibitors to enhance the management of AD.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

JCC, SL, DSW, DLS and CL conceived this project. The data analysis was done by JCC, PFL, and DW. JCC, PFL, CL and DW participated in wet experiments. PFL, CL and DW participated in mouse experiments. CL, DLS and DSW gave some suggestions about the project. All authors read and approved the final manuscript.

Funding

This work was supported by the Joint Construction project of Henan Medical Science and Technology (LHGJ20230213) and Yourh Science Foundation of Henan Province (242300421488).

Data availability

All data generated or analyzed during this study are included in this published article and its additional files. The data used in the current study are available from the corresponding authors upon reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Ethics Committee and institutional Review Committee of Wuhan Union Hospital.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.