CircEif3c/miR-96–5p/PHF20L1/MEOX2 axis in perivascular preadipocyte exosomes mediates fibroblast dysfunction and vascular remodeling

Yixuan Liu; Peiqing Tian; Shuaiyong Zhang; Jiayu Wang; Ye Zhang; Shuxin Ge; Qinghai Wang; Peng Wang; Juan Zhang; Ping Liu

doi:10.1016/j.ncrna.2026.01.006

. 2026 Mar 3;19:1–17. doi: 10.1016/j.ncrna.2026.01.006

CircEif3c/miR-96–5p/PHF20L1/MEOX2 axis in perivascular preadipocyte exosomes mediates fibroblast dysfunction and vascular remodeling

Yixuan Liu ^a,^b,¹, Peiqing Tian ^b,¹, Shuaiyong Zhang ^b, Jiayu Wang ^c, Ye Zhang ^b, Shuxin Ge ^b, Qinghai Wang ^b, Peng Wang ^d, Juan Zhang ^b, Ping Liu ^b,^⁎

PMCID: PMC12969010 PMID: 41809968

Abstract

Background

Diabetes mellitus is a major modifiable risk factor for atherosclerotic cardiovascular disease and pathological vascular remodeling. During chronic hyperglycemia, exosomes serve as essential nanocarriers that coordinate intercellular communication and induce structural and functional alterations in the vascular wall. Perivascular pre-adipocytes (PVPACs) exhibit high exosome secretory activity, while adventitial fibroblasts (AFs) are key effector cells in vascular remodeling. Despite their anatomical proximity, the potential bidirectional crosstalk between PVPAC-derived exosomes and AFs, and its roles for vascular remodeling, remains largely unexplored.

Methods

A co-culture system of PVPACs and AFs, along with a mouse model of perivascular proliferation, was established under sustained hyperglycemic conditions. Exosomes were isolated via sequential ultracentrifugation and characterized based on morphology, size distribution, and exosomal markers. High-throughput RNA microarray was employed to profile PVPAC-derived exosomal RNAs, with qRT-PCR and in situ hybridization used to validate the expression of circEif3c, miR-96–5p, PHF20L1, and MEOX2. Mechanistic studies integrated bioinformatic predictions, CRISPR-Cas9 editing, with bio-functional assays, including RNA pull-down, RIP, dual-luciferase reporter, CO-IP, and protein interaction analyses, to elucidate the circEif3c/miR-96–5p/PHF20L1/MEOX2 axis. In vitro and in vivo rescue experiments evaluated the role of exosomal circEif3c in AF proliferation, migration, apoptosis, and vascular remodeling.

Results

PVPAC-derived exosomes were enriched with circEif3c, which acted as a competitive endogenous RNA by sequestering miR-96–5p, thus alleviating its suppression of PHF20L1 and inhibiting MEOX2 signaling. This axis enhanced AF proliferation and migration, reduced apoptosis, and exacerbated vascular remodeling. Conversely, inhibition of exosomal circEif3c or overexpression of MEOX2 attenuated AF activation, promoted apoptosis, and markedly improved vascular remodeling in diabetic mice.

Conclusion

Our findings establish the PVPAC-derived exosomal circEif3c/miR-96–5p/PHF20L1/MEOX2 axis as a critical driver of hyperglycemia-induced vascular remodeling. Targeting this pathway presents a promising therapeutic strategy to combat diabetes-associated vascular pathology and its complications.

Keywords: Perivascular preadipocyte, Exosome, Vascular remodeling

1. Introduction

Atherosclerotic cardiovascular disease (ASCVD) is a leading cause of disability and mortality in cardiovascular disorders. Obesity and hyperglycemia, as globally prevalent metabolic diseases, are two major risk factors driving the onset and progression of ASCVD. Diabetes mellitus is regarded as a “coronary heart disease equivalent”, not only is it an independent risk factor for ASCVD, but it can also directly cause diabetes-related target organ damage [1].

Obese patients often exhibit abnormal accumulation of visceral fat, with particular attention increasingly paid to perivascular adipose tissue (PVAT) as a high-risk factor for ASCVD [2,3]. Traditionally, PVAT and the vascular adventitia were viewed as protective cushions and energy reservoirs for blood vessels [3]. However, recent studies have revealed that PVAT acts as a highly active endocrine/paracrine organ, directly regulating vascular structure and function through the secretion of various adipokines and signaling molecules [3,4]. Under physiological conditions, PVAT releases vasoactive factors to modulate vascular tone and maintain homeostasis [3,4]. Under pathological conditions such as obesity and diabetes, however, PVAT undergoes adverse changes: brown adipose tissue can transdifferentiate into white adipose tissue and secrete large quantities of pathogenic adipokines, significantly promoting vascular remodeling and the progression of cardiovascular disease (the “Outside-In” theory) [4,5]. The main cellular components of PVAT are perivascular preadipocytes (PVPACs) and mature adipocytes (PVACs), with the proliferation and differentiation of PVPACs determining adipocyte quantity and function [6].

The vascular adventitia is anatomically adjacent to PVAT without a fascial barrier, facilitating intercellular communication [5,7]. Adventitial fibroblasts (AFs), the dominant residential cells of the adventitia, sit at the hub of vascular homeostasis. Operating in an “outside-in” fashion, they coordinate matrix turnover, inflammatory signaling, and structural reprogramming, rendering adventitial fibroblasts (AFs) prime drivers of vascular remodeling [5,8,9]. This process is first heralded in the adventitia, where AFs mount a rapid proliferative response. Although AFs are recognized as prime movers of this process, the molecular lexicon through which they converse with PVPACs/perivascular progenitor/dipocyte-ike cells remains largely undeciphered. The potential role of the AF-PVPAC interplay in determining the outcome of vascular remodeling by regulating critical cellular behaviors, such as proliferation, migration, autophagy, and apoptosis, in the context of obesity and diabetes remains an open question. As the body's largest endocrine organ, adipose tissue secretes a diverse array of signaling molecules including exosomes, adipokines, and non-coding RNAs that collectively constitute a systemic communication network [10]. Within this repertoire, exosomes act as privileged couriers, shuttling proteins, lipids and nucleic acids between cells and organs to preserve homeostasis or, when dysregulated, to drive disease [11,12]. In obesity, adipocyte-derived exosomes are massively released and their lipid-enriched cargo converts a physiological signaling system into a pathogenic driver that amplifies local inflammation and accelerates atherogenesis [11,12].

Non-coding RNAs (ncRNAs) have evolved from genomic “noise” to master regulators of cardiovascular homeostasis and disease, furnishing tractable biomarkers and therapeutic entry points for atherosclerosis [13,14]. While the miRNA and lncRNA landscapes are already charted in considerable detail, the circRNA continent remains largely unexplored [15,16]. CircRNAs are widely distributed, highly stable, and tissue-specific, giving them considerable value in disease pathogenesis, diagnosis, and therapy [15,16]. Their expression is especially prone to dysregulation under metabolic abnormalities such as diabetes [15,16]. The most well-characterized mechanism of circRNAs is their “molecular sponge” function in sequestering miRNAs [15,16]. However, it remains unreported whether and how circRNAs regulate PVPAC function.

We previously identified mesenchyme homeobox 2 (MEOX2, also known as Gax) as a homeobox gene that acts as a transcriptional brake to govern cardiovascular remodeling by restraining vascular cell proliferation [8,17]. Although pro-angiogenic activities have been ascribed to MEOX2 in neural and reproductive tissues, its dominant vascular portfolio comprises suppression of proliferation, limitation of fibrosis, and facilitation of apoptosis [8,18,19]. Emerging data further suggest that the MEOX2/Tcf15 heterodimer licenses endothelial fatty-acid uptake and trafficking, yet the molecular underpinnings of this metabolic axis remain unresolved [20].

Here we present the first transcriptomic atlas of the previously uncharted PVAT, uncovering the concerted and dysregulated expression of circEif3c, miR-96–5p, the homeobox transcription factor MEOX2, and the zinc-finger reader PHF20L1. Gain- and loss-of-function studies in vitro and in vivo demonstrate that PVPAC-derived exosomes ferry a circEif3c/miR –96–5p/PHF20L1/MEOX2 regulatory circuit that selectively programs AFs fate and thereby dictates the trajectory of vascular remodeling. These findings establish MEOX2 as a tractable gene-therapy tool for ASCVD and provide the mechanistic underpinning for exosome-based interception of adverse vascular remodeling.

2. Material and methods

2.1. Primary culture of dedifferentiated PVPAC and Oil Red staining experiment

Mature adipocytes were isolated using a modified ceiling culture method as previously described [21,22]. Adipocytes (5 × 10⁴) were plated in 25-cm² flasks completely filled with Dulbecco's Modified Eagle Medium (#C3113, Darthill Shanghai, China), supplemented with 15 % fetal bovine serum, and incubated at 37 °C in 5 % CO₂. The cells floated upward and adhered to the inner ceiling surface of the flask. After 6–7 days, the medium was removed and the flasks were inverted. The medium was changed every 3 days until the cells reached passage 4–6. Relevant antigen such as CD34+Sca1+Pref-1+ identification was respectively performed using antibodies, anti-CD34 (1:200, #50-0341-82, eFluor™ 660, Invitrogen, USA), anti-Sca-1(1:100, #bs-3752R, BIOSS, China), anti-pref-1 (1/100, #LS-C137763, LSBio, USA), for Cell identification. After correct identification, cells from passages 3–6 were used as experimental materials. Oil Red “O” staining was also performed for primary dedifferentiated adipocyte identification. After correct identification, cells from passages 3–6 were used as experimental materials. Detection of changes in factors secreted by PVPAC before and after ncRNAs stimulation; assessment of the time required for PVPAC induction; and observation of the ultrastructure, morphology, and function of PVPAC or PVAC.

2.2. Primary culture of AFs

AFs were isolated and cultured using the tissue explant method. The common carotid arteries of C57BL/6 mice were isolated, and the adventitia was separated. The tissue was then cut into small pieces approximately 1 mm × 1 mm in size and placed in culture dishes. DMEM medium supplemented with 10 % fetal bovine serum was added for tissue explant culture. Once the cells reached a subconfluent state, they were digested with 0.125 % trypsin-EDTA and passaged. Cells from passages 3–6 were used as experimental materials. Antigen identification was performed using CD31⁻, CD34⁺, and CD146⁻ markers, and differentiation from vascular smooth muscle cells was confirmed using monoclonal antibodies such as vimentin. Prior to Ad-RNAs transfection into target cells in vitro, AF cells were stimulated with Ad-ncRNAs and Ad-siR-ncRNAs at the optimal titer for 48 h. Changes in factors secreted by AF before and after ncRNAs stimulation were detected, and the morphology, function, and ultrastructure of AF were observed.

2.3. Plasmid and siRNA transfection and adenovirus transduction

The circEif3c overexpression plasmid, pcDNA3.1-CMV-circEif3c, was constructed. siRNA mimics and inhibitors targeting circEif3c and miR-96–5p were designed [Table 1, Table 2] and synthesized by Invitrogen (Shanghai, China) using Thermo Scientific technologies. Plasmid and miRNA transfections were performed using Lipofectamine 3000 (Invitrogen), while siRNA transfection was carried out with DharmaFECT 4 (Dharmacon, IL, USA). Adenovirus vectors (GFP-tagged) expressing MEOX2 supplied by Shanghai Genechem Co., Ltd.

Table 1.

The sequence of primer (F: Forward primer; R: Reverse primer).

Gene name	Sequence (5′ → 3′)
circEif3c (mmu_circ_0001628, NM_146200) -Divergent primer-F	TACTATGGAGTGTGTCTA
circEif3c -Divergent primer-R	GGACGATCACCATAGTAC
circEif3c -Convergent primer-F	GCAGGAGCGAAATCAGGAAC
circEif3c -Convergent primer-R	GTGGAACTGCTTGCTGATCA
Eif3c-F	CTCCGGGAACTACGGCAAAC
Eif3c-R	GCGCTTCGGACAACTCTCTT
PHF20L1-F	CGTCCTGGAATCACTTTTGAGA
PHF20L1-R	TTGCTGTCCCAGTAAATCCAC
miR-96–5p (mouse: NR_029750.1) (ploy A)-F miR-96-5p-R	TCGGCAGGTTTGGCACTAG CTCAACTGGTGTCGTGGA
MEOX2 (NM_017149.1)-F	GGAAGCTTTCATAAGTGTGCGTGCTCAG
MEOX2-R	AAGGTACCATGGAACACCCCCTCTTTGGC
GAPDH-F	AGCTTAGGTTCATCAGGTAAAC
GAPDH-R	TTACTCCTTGGAGGCCATGTAG

Open in a new tab

Table 2.

The sequence of probe.

Gene name	Pobe (5′ → 3′)
circEif3c -ISH probe	TGGGCAGCCATGTAGGGGATCTCCAGGAGC
circEif3c-biotin probe	TCCAGGTTGATGTGCAGGTGAAAGGGCACC
miR-96-5p-ISH probe	AAAATGTGCTAGTGCCAAAATCGGCCAAGCAG
miR-96–5p -biotin probe	GGCACTACACATGATTGCTCACAGCGGAGA

Open in a new tab

2.4. Extraction, identification, and tracking of exosomes

2.4.1 Isolation and extraction of exosomes: exosomes were isolated and purified from the cell culture supernatant or plasma using differential centrifugation. The supernatant (10 ml) from PVPAC culture was centrifuged at 2000×g for 30 min at room temperature to remove dead cells, cellular debris, and large particles. The supernatant was then transferred to a disinfected tube. The cell culture supernatant (10 mL) was mixed with 5 mL of Total Exosome Isolation Reagent (#4478359, TEIR kit; Invitrogen; Thermo Fisher Scientific, Inc.) from the cell culture medium, and the mixture was thoroughly vortexed until a homogeneous solution was formed. Subsequently, the sample was incubated and then centrifuged at 10000×g for 1 h at 4 °C. To further purify, wash the precipitate with PBS and centrifuge again at 100000×g for 1 h at 4 °C to obtain purified exosomes.

2.4.2 Identification of exosomes: TEM method, FCS method and Western blotting method can be used respectively (to detect the expression of extracellular vesicle markers CD63 and TSG101). Finally, the separated exosomes were resuspended in PBS buffer for subsequent nanoparticle tracking analysis (NTA), 1 × RIPA buffer (Thermo Fisher Scientific, Inc.) supplemented with protease inhibitors (Roche Diagnostics) and collected in sterile PBS for subsequent studies (stored in a −80 °C freezer).

2.4.3 Tracking and localization of exosomes: after labeling the exosomal membrane of correctly identified exosomes with PKH-26 red fluorescence, they were added to PAVAC and AF cell culture media and incubated for 24 h. Concurrently, target cells were labeled with GFP green fluorescence, exosomes were marked with RFP-CD36 red fluorescent plasmid, and cell nuclei were stained with Hoechst 33258 blue fluorescent dye. Additionally, circEif3c was tagged with Cy5 (red) and miRNA with Cy3 (orange-yellow). Cellular dynamics were monitored using a BioNavigator, followed by washing steps. The relative spatial relationships among ncRNAs, exosomes, and cells were then analyzed via fluorescent confocal microscopy to determine exosome internalization by cells.

2.5. Screening, validation, amplification, and localization of ncRNAs

2.5.1. Construction of ncRNAs expression profiles and screening of differentially expressed ncRNAs

ncRNA expression profiles in exosomes derived from PAVAC and AF cells were analyzed using microarray technology. The accuracy of the microarray data was validated by qRT-PCR. Based on the background expression levels and fold-change values, two target ncRNAs, circEif3c and miR-96–5p, were identified as differentially expressed.

2.5.2. Validation of circRNA identity and circularization characteristics

Back-to-back (divergent) PCR primers and linear (convergent) primers were designed to target the back-splicing junction of circEif3c. Amplification was performed using both cDNA and genomic DNA (gDNA), followed by agarose gel electrophoresis to confirm circularization. Sequencing analysis validated the circular structure and identified precise junction sites. Additionally, RNase R resistance assays and actinomycin D-mediated transcription inhibition experiments were conducted to confirm circEif3c's stability as a circular RNA.

2.5.3. RNA extraction, subcellular localization, and qRT-PCR semi-quantitative analysis

To investigate the subcellular localization and expression dynamics of circEif3c and miR-96–5p, a combined approach of cytoplasmic/nuclear fractionation, qRT-PCR, and fluorescence in situ hybridization (FISH) was employed. Cytoplasmic and nuclear RNAs were isolated from normal control and experimental groups using a specialized RNA isolation kit. After measuring RNA concentration and purity with a spectrophotometer, cDNA was synthesized via reverse transcription and subjected to semi-quantitative analysis by qRT-PCR using specific primers listed in Table 3. In parallel, for FISH analysis, specific Cy5-and Cy3-labeled probes targeting circEif3c and miR-96–5p, respectively, were designed (sequences in Table 3). Cells at 80 %–90 % confluence were fixed, permeabilized, and pretreated, followed by overnight hybridization with the probes. After washing, nuclei were counterstained with DAPI, and signals were detected using a commercial FISH kit according to the manufacturer's protocol. All images were captured using a confocal microscopy system.

Table 3.

The sequence of siRNA.

Gene name	Target sequence (5′ → 3′)
siR-circEif3c-01	TATAACCGTACTATGGTGCAACT
siR-circEif3c-02	TAACCGTACTATGGTGCAACTGG
siR-miR-96-5p-01	AAGCAAAUGAUGCUAGUGCAA
siR-miR-96-5p-02	AGCUGUUGCUGUAAAGCAAA

Open in a new tab

2.5.4. Tissue FISH

Frozen tissue sections (8 μm) underwent sequential treatments: 0.2 M HCl (15 min, RT), 0.25 % pepsin (37 °C, 30 min), and 4 % PFA fixation (20 min). Post-hybridization buffer incubation (55 °C, 2 h), ncRNA probes (1:200) were denatured (85 °C, 2 min), equilibrated (37 °C, 2 min), and hybridized overnight (37 °C, dark environment). Sections received SSC washes, anti-digoxigenin (38 °C, 1–2 h) and Cy3-streptavidin (37 °C, 1 h), followed by DAPI nuclear staining and mounting. Fluorescence imaging identified cytoplasmic red signals as positives. Cluster and Treeview were used to analyz fluorescence intensity for heatmap generation.

2.6. Co-immunoprecipitation (Co-IP) and Western blotting

Total proteins were extracted from cells and clinical tissues using RIPA lysis buffer (P0013B, Beyotime, China) according to the manufacturer's protocol. Subsequently, the lysates were incubated with specified antibodies at 4 °C for 12 h and then mixed with Protein A/G magnetic beads (#ab286842, Abcam, UK) for an additional 4 h. After three washes with PBST (Phosphate-Buffered Saline containing Tween-20), the eluates were boiled for 8 min and analyzed by Western blot. Protein expression levels were assessed following a previously described protocol [6]. Antibodies against MEOX2 (1:1500, #ab262916, Abcam, UK), PHF20L1 (1:1500, #ab118190, Abcam, UK), β-actin (#AC004, 1:5000, ABclone, Wuhan), Bcl-2 (1:2000, #ab182858, Abcam, UK), N-cadherin (1:1500, Abcam, UK), vimentin (1:2000, #ab92547, Abcam, UK), Anti-CD63 (1:1000, #ab315108, Abcam, UK), and TSG101 (1:2000, #28283-1-AP, Proteintech, USA) were purchased. A horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:5000, UK) was also obtained. Finally, protein bands were visualized using an electrochemiluminescence (ECL) system, and the grey values were measured by Image J software to evaluate relative protein levels, and normalized to β-actin.

2.7. Dual-luciferase reporter assay

StarBase software were used to predict miR-96–5p binding sites in wild-type circEif3c (WT-circEif3c) and MEOX2 3′UTR (WT-MEOX2). Mutant vectors (Mut-circEif3c/Mut- MEOX2) were constructed. BioSune (Shanghai, China) synthesized firefly luciferase reporters containing WT or mutated sequences cloned into pGL3. 293T cells were co-transfected with 0.1 μM vectors (WT/Mut) using Lipofectamine 3000, cultured at 37 °C/5 % CO₂ for 48 h, then harvested via centrifugation (2000g, 5 min). Dual-luciferase activity (Synergy LX; BioTek, USA) was measured and normalized to Renilla luciferase activity.

2.8. Cell type identification and apoptosis detection by flow cytometry (FCM)

FCM was used to detect specific markers on the cell surface or inside to identify whether primary cultured cells are PVPAC or AF. The antibodies used are: CD34 Rabbit mAb (#R380824, Zenbio, Sichuan, China); Secondary antibody Goat Anti-Rabbit IgG (H + L) (Alexa Fluor® 488, #AB0141, Abways, Wuhan, China). Pref1 Goat pAb (#AF8277, Bio-techne, Minnesota, USA); Secondary antibody Donkey Anti-Goat IgG (H + L) (Alexa Fluor 594, #AB150132, Abways, Wuhan, China); Rabbit anti-Mouse Sca-1 mAb, ABflo 647 tag (#A22787, Abclonal, Wuhan, China).

The Annexin V-FITC/Propidium iodide (PI) double staining kit (#F6012M, Uelandy, Suzhou, China) was used to inspect cell apoptosis according to the manufacturer's protocol. Briefly, after harvesting the cells, they were separately stained with Annexin V-FITC/PI (#F6012M, UElandy, Suzhou, China), and then the flow cytometer (Beckman Coulter, Miami, FL, USA) was employed to detect the cell apoptosis rate. Fluorescein isothiocyanate (FITC) was observed at 530 nm, while Propidium iodide (PI) was detected at 575 nm.

2.9. Real-time quantitative PCR (RT-qPCR)

Total RNA was extracted from cells and tissues using the Trizol reagent kit obtained from Invitrogen (Carlsbad, CA, USA) according to the manufacturer's protocol. Real-time quantitative PCR was conducted to examine the expression levels of circEif3c and miR-96–5p, as well as the mRNA levels of MEOX2, and GAPDH, following the procedures provided in a previous study [8]. For the quantification of circEif3c, circular RNA was enriched and pretreated with RNase R enzyme (3 U/μg) for 20 min at 37 °C to eliminate linear circEif3c. The primer sequences for the aforementioned genes are listed in Table 1.

2.10. Wound healing assay

Cells were treated with stimulus (such as exosomes, high-sugar, and left untreated) for 24 h. Subsequently, the cells were harvested and seeded in a 6-well plate at a density of 2 × 10⁵ cells per well. A scratch wound was created using a sterile 200 μL pipette tip, and the suspended cells were removed by washing with 1 × PBS. Images of the scratches were captured using an inverted microscope (Olympus Inverted Research Microscope, Japan) at 40 × magnification at 0 h and 24 h after scratching.

2.11. Cell proliferation assay

Cell proliferation was assessed using CCK-8, EdU, flow cytometry (FCM), and crystal violet staining. For CCK-8 (#C6005M, Uelandy, Suzhou, China), 5 × 10³ cells/well (96-well plate) were incubated for 48 h, treated with 10 μL CCK-8 reagent (100 μL final volume), and absorbance measured at 450 nm. EdU assays (Epizyme, Shanghai, China) involved seeding 5 × 10⁴ cells/well (12-well plate), 48 h incubation, 2 h EdU labeling, fixation (4 % PFA/0.1 % Triton X-100), DAPI staining, and calculating EdU+/DAPI+ ratios. Cell cycle analysis required PBS-washed cells stained with DNA dye/permeabilization solution (MultiSciences Biotech, Co., Ltd. Hangzhou, China), analyzed via FlowJo V10. Cells were fixed in 4 % PFA, stained with crystal violet (Sigma, St. Louis, MO, USA), dissolved in 10 % glacial acetic acid, and measured OD590 value.

2.12. Glucose uptake experiment

Referring to our previous experimental protocol regarding cellular uptake of 2-[3H] deoxyglucose (2-DOG) [23], PVPAC cells underwent a 3-h fasting period. Subsequently, they were stimulated at 37 °C for 10 min by adding 2-[3H]DOG (0.1 Ci, with a final concentration of 0.1 mM) to Krebs-Ringer phosphate HEPES (KRPH) buffer. The cells were then washed, specifically by rinsing the monolayer four times with 500–1000 μL of ice-cold KRPH buffer. After terminating the uptake process, 300 μL of 0.05 M NaOH was added to each well, followed by mixing with a plate shaker and incubation at 37 °C for 2 h to facilitate cell lysis. The radioactivity associated with the cells was quantified using a liquid scintillation counter.

2.13. RNA immunoprecipitation (RIP)

RIP experiment was performed using anti-Ago2 and anti-IgG antibodies. The experiment utilized the Magna RIP® RNA-binding protein immunoprecipitation kit (#17–700, Merck KGaA, Darmstadt, Germany) and strictly followed the manufacturer's instructions. Enrichment values were normalized against the background RIP levels detected using the IgG isotype control.

2.14. Biotin-labeled probe RNA pull-down experiment

Biotin-labeled miR-96–5p and its scrambled negative control miRNA (#AM4621, invitrogen, USA) were synthesized. Biotin-labeled miRNAs were transfected into cells using Lipofectamine 3000 (#L3000-001, Invitrogen, Thermo Fisher Scientific, USA), and cell lysates were subsequently collected. After 24 h of transfection, the cells were fixed with 3 % paraformaldehyde for 30 min, followed by incubation with 1.25 M glycine for 5 min at room temperature, then centrifuged at 1500g–2000g for 5 min, the supernatant was discarded, and PBS was added to the suspended cells for counting. Cell lysates were collected 48 h after transfection and incubated with streptavidin magnetic beads (#65306, Invitrogen, USA) for 2 h. After centrifugation to wash the beads, the pulled-down miRNAs were extracted using Trizol (#15596018, Invitrogen, USA) and subjected to qRT-PCR. The primer sequences for qRT-PCR are detailed in Table 2.

2.15. Animal experiments

Male wild-type C57BL mice (3–4 weeks old) were purchased from Jinan Pengyue Experimental Animal Breeding Co., Ltd., while circEif3c(−/−) and MEOX2(±) conditional knockout mice were generated by Cyagen Biosciences Inc. (Suzhou, China).

Eighty-four male C57BL mice were allocated to 14 experimental groups (n = 6 each) on the basis of genotype and planned intervention. The groups were: (1) C57BL control (standard chow, no gene deletion); (2) ob/ob diabetic model; (3) circEif3c (−/−); (4) MEOX2 (±); (5) ob/ob + PVPAC-Exo; (6) ob/ob + circEif3c mimic plasmid; (7) ob/ob + Exo-circEif3c (over-expression); (8) ob/ob + Exo-siR-circEif3c (knock-down); (9) ob/ob + siR-miR-96–5p; (10) ob/ob + Exo-miR-96–5p (over-expression); (11) Ad-MEOX2 alone; (12) MEOX2(±) + Ad-MEOX2; (13) Ad-MEOX2+ PVPAC-Exo; (14) Ad-MEOX2+Exo-circEif3c. Except for the normal control group fed with standard diet before modeling and throughout the intervention period, all other groups were fed with high-fat diet (containing 3 % cholesterol, 10 % lard, 0.2 % methimazole, and 86.8 % basal feed). To exclude potential interference from viral vectors, all control and experimental group mice (except those receiving Ad-MEOX2 treatment) were concurrently administered Ad-GFP control virus.

Before and after intervention vector transfection, echocardiography (transthoracic + intravascular ultrasound) was used to measure the structural morphology of common carotid arteries in each group. An animal model was established using the Exo-circEif3c gel method applied to the adventitia of mouse common carotid arteries: experimental mice were randomly selected for modeling. First, CCA transthoracic ultrasonography was performed and parameters were recorded. After anesthesia with pentobarbital sodium (30 mg/kg, intravenous injection), the left common carotid artery (CCA) was exposed under sterile conditions. A gel containing 0.5 ml of optimally concentrated Exo-circRNA Eif3c and 20 % (3–5 mL) Pluronic F-68 was applied to the perivascular tissue. The same procedure was performed on the right common carotid artery with application of an equal volume of Exo-circEif3c gel. After suturing the incision for 4 weeks, the animal model was established by stimulating perivascular tissue and adventitial proliferation with Exo-circEif3c plasmid, and the modeling effects were compared between exosome-treated and untreated groups. The control group underwent the same procedures but with gel containing only Exo-Ad-GFP.

Based on the above model, in vivo intervention of carotid artery PVAT was performed: the PVAT of CCA in modeled mice was re-exposed. After dissolving the previously applied gel with ice water, new Pluronic F-68 gels containing equal volumes of (Exo-)Ad-MEOX2 or Exo-(siR)-ncRNAs were applied to the perivascular tissue of the left/right common carotid arteries, respectively. Changes in relevant indicators were observed in each group after 2 weeks. To euthanize the mice for tissue harvest, the animals were first anaesthetized with pentobarbital sodium (30 mg/kg, intravenous injection) and then euthanized by cervical dislocation.

2.16. In vivo imaging of mice with DIR-labeled exosomes

Exosomes were fluorescently labeled via DiR dye (MCE, USA) by incubating 2 μL dye with 100 μL exosome suspension (1 μg/μL) in PBS for 1 h. Unbound dye was removed by adding equal-volume serum, followed by ultracentrifugation (100,000×g, 4 °C, 70 min). Pelleted exosomes were resuspended in 100 μL PBS, filtered (0.22 μm), and adjusted to 1 μg/μL using BCA assay. For in vivo application, 100 μg DiR-exosomes were topically applied to mouse carotid sheaths and controls received PBS. Imaging was conducted at 12 h post-application.

2.17. Immunohistochemistry (IHC) and immunofluorescence

Cells (1 × 10⁵/well, 6-well plates) on coverslips were fixed in 4 % paraformaldehyde (40 min, 4 °C), permeabilized with 0.1 % Triton X-100 (20 min), and blocked with 5 % BSA (30 min). For IHC, sections were incubated with primary antibodies (anti-MEOX2 1:500 #12449-1-AP, Proteintech) and (anti-Bcl-2, 1:100, ab194583; anti-PHF20L1, 1:100, ab118190; anti-Bax, 1:250, ab32503; anti-Bak, 1:200, ab104124), all Abcam at 37 °C (90 min), followed by HRP-secondary antibody (1:3000, AS003, ABclonal; 30 min), DAB staining (15 min, dark environment), and hematoxylin counterstaining (10 min). For IF, sections underwent primary antibody incubation (4 °C overnight), FITC-conjugated secondary antibody (37 °C, 1 h), and DAPI staining. Imaging used Leica MZ125 (brightfield) or Nikon ECLIPSE Ti2-E (fluorescence).

2.18. Transthoracic vascular Doppler ultrasound in mice

After establishment of the carotid artery proliferation model, a VEVO 3100 echocardiography system (VisualSonics, Toronto, ON, Canada) was used for transthoracic vascular Doppler ultrasound of the common carotid artery to assess vascular remodeling. Mice were removed from the induction chamber, and chest hair was removed with depilatory cream. Anaesthetized animals were placed on a heating pad equipped with built-in ECG leads to maintain body temperature. A nose cone connected to the anesthesia system was used to deliver continuous sedation with 1.0–1.5 % isoflurane in 100 % O₂ at 0.5 L min⁻¹. Anesthesia depth was adjusted to maintain a target heart rate of 450 ± 50 beats min⁻¹. All four paws were coupled to ECG electrodes with conductive gel. During scanning, the probe was gently placed on the chest to localize the left ventricle. Three consecutive cardiac cycles were recorded and averaged for each mouse. All data were analyzed post-experiment using the system's built-in software. To euthanize the mice, the animals were first anaesthetized with pentobarbital sodium (30 mg/kg, intravenous injection) and then euthanized by cervical dislocation.

2.19. Statistical analysis

All statistical analyses were conducted using IBM SPSS 24.0 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism9.5 (GraphPad Software Inc., San Diego, USA). The sample size was calculated using the G∗Power 3.1 tool [24]. The values are derived from at least three independent experimental samples, each of which is repeatedly validated using the same technique, unless otherwise specified. The Kolmogorov-Smirnov test was used to assess the normality of the data distribution. The comparison of mean values between two sets of quantitative data is conducted using a student t-test. Analysis of Variance (ANOVA) was used to analyze the differences in sample means among multiple sets of measurement data, and Tukey's post hoc test was used for multiple comparisons. The comparison of rates between count data uses chi square test.

3. Results

3.1. Effects of different concentrations of glucose on the biological functions of PVPACs

Primary cultures of PVPAC and AF cells were established. As mentioned earlier [8,21], primary PVPAC and AF cells were established separately by the ceiling method [21] from isolated perivascular adipose tissue and by the explant tissue-block method [8] from isolated vascular outer-membrane tissues. Immunocytofluorescence staining confirmed that primary cells expressing vimentin and α-smooth muscle actin corresponded to the desired AF phenotype, whereas cells positive for CD34, Sca-1, and Pref-1 were identified as PVPAC. Both cell populations have been authenticated 100 % purity [Fig. 1A-D]. To explore the effects of hyperglycemia on the biological functions of AF and PVPAC cells, different concentrations of glucose were added to the complete culture medium, and mannitol was added to maintain isotonic conditions. Compared with normal glucose (NG, 5 mmol/L), high glucose (HG, 30 mmol/L) significantly enhanced the proliferation, migration and differentiation of PVPAC [Fig. 1E-J].

Fig. 1 — **Characterization of cultured cells and the impact of high glucose on PVPAC function.** (A and D) Immunofluorescence and (B and C) flow cytometry (FCM) identification of PVPACs/AFs. Characteristic surface markers of PVPACs and AFs were analyzed by FCM, with blue curves indicating isotype controls and red curves representing test samples. PVPACs and AFs exhibited positive expression of CD34, Pref1, Sca1, vimentin, and α-actin, respectively, scale bar = 50 μm. (E–J) Normal glucose (NG, 5.5 nmol/L) and high glucose (HG, 30 nmol/L) had different effects on the biological functions of PVPACs. (E) Glucose uptake assay. Quantitative determination of 2-[3H] deoxyglucose (2-DOG) uptake by PVPAC cells under stimulation of glucose at different concentrations. HG effectively stimulated glucose uptake in PVPAC, while it had no obvious change in NG group. n (the number of experiments) = 6 (F) EDU experiment, scale bar = 100 μm. (G) CCK8 assays. n (the number of experiments) = 3 (H) Crystal violet staining; Scale bar = 200 μm. (I) Scratch migration assays. Scale bar = 200 μm. (J) Adipogenic induction and differentiation assessed by Oil Red O staining, scale bar = 200 μm. Data are shown as mean ± SD from three independent experiments. n (the number of experiments) = 3. Statistical significance was determined by one-way ANOVA with Dunnett's post-hoc test; vs. the control (NG) group, ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; n.s, no significance.

3.2. AF and PVPAC interact with each other through exosomes

Exosomes in the culture medium of PVPACs and AFs cultured alone or co-cultured were isolated by differential centrifugation [Fig. 2A]. To investigate whether exosomes act as mediators between PVPACs and AFs under HG conditions and to clarify their role in modulating cellular biology, we first established PVPACs/AFs co-culture to determine the relative capacity of each cell type to deliver exosomes and to quantify the impact of these exosomes on the proliferative function of both PVPACs and AFs, the isolation and identification of exosomes were first conducted [Fig. 2B]. Apparently, the concentration and diameter of exosomes detected in the AF (upper cell)/PVPAC (lower cell) culture way with HG stimulating are the highest, followed by PVPAC/AF with HG, then PVPAC with HG, and the smallest is AF with NG [Fig. 2B2, 2B3 and 2B5]. TEM analysis revealed that these vesicles in PVPACs and AFs were spherical and had a typical cup-shaped appearance [Fig. 2B2 and 2E1]. The NTA results showed that the average size of PVPAC-Exo was also higher than that of AF-Exo [Fig. 2B3 and 2B6]. The co-culture experiments described above indicated that PVPAC, located in the lower compartment of the Transwell, likely functioned as the signal-sending cell, whereas AF in the upper chamber served as the signal-receiving cell [Fig. 2B]. Cell immunofluorescence co-localization analysis revealed a significant presence of exosomes in both PVPAC and AF [Fig. 2C]. PKH67-labeled exosomes from PVPAC-Exo cells were added to PVPACs and AFs respectively and clearly expressed in the cytoplasm [Fig. 2C1 and 2C2]. After AFs were co-cultured with PKH67-labeled PVPAC-derived exosomes for 4 h, abundant fluorescent exosomes were observed throughout the AF cytoplasm, demonstrating the efficient uptake of PVPAC-Exo by AFs [Fig. 2C3]. It is evident that the fluorescence intensity of PVPAC-derived exosomes in the cytoplasm of AFs was no less than that of AF-derived exosomes or the exosomal fluorescence intensity in the cytoplasm of PVPACs [Fig. 2C]. To verify the effects of PVPAC-Exo, we incubated AFs with 1 × 10⁶ of PVPAC-derived exosomes. After 24 h, the scratch test and crystal violet assay demonstrated that, compared with PBS treatment (control), PVPAC-Exo significantly enhanced the migration and proliferation of AFs [Fig. 2D]. Conversely, pretreating AFs with the exosome inhibitor GW4869 markedly attenuated their proliferative and migratory capacity [Fig. 2D]. To assess the impact of HG stimulation on PVPAC-derived exosomes, PVPACs were exposed to normal glucose (NG, 5.5 mM), high-glucose (HG, 30 mM), or GW4869 for 24 h, after which exosome morphology and the expression of exosomal marker proteins were evaluated across all groups [Fig. 2E]. Compared with the NG group, the HG group exhibited significantly enhanced cell proliferation capacity, increased exosome particle size, and elevated expression of exosomal marker proteins. However, pretreatment with GW4869 to inhibit exosome activity markedly attenuated the stimulatory effects on AF proliferation, migration, and vimentin expression [Fig. 1D]. This indicated that PVPAC-derived exosomes had a significant impact on the biological functions of AFs [Fig. 1D]. To investigate the effects of different glucose concentrations, PVPAC-derived exosomes (PVPAC-Exo) and culture modalities (single-cell vs. dual-cell co-culture) on exosomal marker proteins, PVPACs and AFs were pretreated with NG, HG, and PVPAC-Exo under the indicated culture modes [Fig. 2F-J]. Biomarkers (CD9, CD63, TSG-101, and HSP70) of exosomes were detected by Western blotting in uncleaved PVPAC-Exo and AF-Exo [Fig. 2F-J]. Analysis of exosome-specific proteins in the culture medium revealed that CD9, CD63, TSG101, and HSP70 levels were highest in the PVPAC/AF co-culture model and in PVPAC-Exo-pretreated groups under high-glucose conditions, intermediate in the high-glucose PVPAC-alone group, and lowest in the low-glucose AF group [Fig. 2F-J]. The above results suggested that exosomes secreted by PVPAC might mediate cell-to-cell communication between PVPAC and AF, as well as the realization of AF's biological functions.

3.3. circEif3c was identified and enriched in PVPAC-Exos

Following 24 h of high-glucose (30 nM) stimulation, exosomes isolated from the co-culture medium of PVPACs and AFs were subjected to comprehensive analysis combined high-throughput microarray profiling with bioinformatics (such as circBase, TargetScan, and Circbank). Compared with the NG control, the HG group demonstrated significantly differential expression of both circRNAs and miRNAs [Fig. 3A-D]. After the initial screening through bioinformatics analysis and PVPAC exosomal microarray screening, three candidate PVPAC-Exo-circRNAs with highly specific differential expression were preliminarily selected, namely, exo_circ_0001628 (circEif3c), exo_circ_0001629, and exo_circ_0003881[Fig. 3C], circEif3c was subsequently validated by qRT-PCR as the optimal candidate, with Eif3c identified as its parental gene [Fig. 3D]. Leveraging bioinformatics databases and genomic technologies, we annotated the genomic physical structure of circEif3c [Fig. 3E]. CircEif3c was composed of one exon from chr7:126551975-12655233. Firstly, the ring structure of circEif3c was verified by divergent primer (DP) and convergent primer (CP), and further identified by Sanger sequence [Fig. 3E]. Subsequently, following amplification with divergent primers, the head-to-tail backsplice junction (BSJ) of circEif3c was confirmed by Sanger sequencing and was consistent with its annotation in the circBase database [Fig. 3E]. Divergent primers produced amplicons from cDNA but not from gDNA, the result consistent with a backsplicing origin rather than trans-splicing or genomic rearrangement [Fig. 3F]. qPCR analysis in the Actinomycin D assay displayed the expression of circEif3c and linear Eif3c in AFs at 0, 5, 10, and 15 h post-treatment [Fig. 3G], demonstrating that the circular form is more stable than its linear counterpart [Fig. 3G]. Moreover, its resistance to the effects of Actinomycin D also confirms that circEif3c forms a closed-loop structure [Fig. 3G]. Therefore, based on the aforementioned results, it is proven that CircEif3c possesses a stable circular structure [Fig. 3E-G]. Biotinylated probes were designed to target the back-splice junction (BSJ) of circEif3c [Table 2] and RNA pull-down assay verified their robust pull-down efficiency in AFs transfected with the circEif3c overexpression plasmid [Fig. 3H].

Fig. 3 — **Verification of the circEif3c and miR-96-5p interaction. (A**–**C) Identification of differentially expressed exosomal circRNAs (Exo-circRNAs).** All the exosomal circRNAs in the culture medium following a co-culture model of AFs/PVPACs cells stimulated by HG (30 nM) for 24 h. The differential Exo-circRNA was analyzed by RNA microarray and displayed by volcano plot (A) and Heatmap (B). (C) A flowchart depicting the screening strategy. By integrating bioinformatics data from PVPAC-Exo, AF-Exo, and circBase, the top three highly expressed circRNAs were selected. **(D) Selection of the optimal circRNA candidate.** After the initial screening, the three candidate circRNAs were further narrowed down to the optimal one. Following 24 h of HG stimulation, qRT-PCR in PAPVCs revealed that circEif3c was expressed at significantly higher levels than the other two candidates. **(E and F) Validation of circular structure of circEif3c**. (E) Divergent primer design and Sanger sequencing validation for circEif3c amplification. (F)The primer of exosomal circEif3c was identified and verified by RT-PCR. Agarose gel analysis of circEif3c PCR products amplified with divergent versus convergent primers. **(G) Stability assessment of circEif3c.** Actinomycin D assay for circEif3c and Eif3c expression. **(H) Efficiency of the circEif3c probe**. A circRNA pull-down assay in AFs transfected with OE-circEif3c or OE-control plasmid confirmed the specific enrichment efficiency of the circEif3c probe, with a control probe as a negative control. (I–**K) Screening of candidate miRNAs.** Differential miRNA expression in exosomes was analyzed by microarray and displayed as a volcano plot (I) and a heatmap (J). (K) A flowchart showing the selection of miR-96–5p, miR-15a-5p, and miR-322–5p from the intersection of miRNA microarray data, TargetScan, and Circbank. **(L) circRNA pull-down assay for miRNA interaction.** In AFs transfected with OE-miR-96–5p mimic or control plasmid, the circEif3c probe exhibited significant enrichment of miR-96–5p, confirming their interaction. **(M) AGO2-RIP assay.** RIP analysis using an anti-Ago2 antibody in AF exosomes from cells overexpressing circEif3c or miR-96–5p demonstrated significant co-enrichment of both circEif3c and miR-96–5p with Ago2, indicating their incorporation into the RISC complex. **(N) Luciferase reporter assay.** Bioinformatic prediction identified putative miR-96–5p binding sites on circEif3c. Luciferase assay confirmed a direct interaction, as the miR-96–5p mimic suppressed the activity of the wild-type (WT) but not the mutant (Mut) circEif3c reporter. **(O) miR-96-5p enrichment by circEif3c probe.** The circRNA pull-down assay further confirmed the direct binding between the circEif3c probe and miR-96–5p. **(P) Immunofluorescence colocalization.** Cy5-labeled circEif3c and Cy3-labeled miR-96–5p plasmids showed clear colocalization in AFs 24 h post-transfection, providing visual evidence of their interaction, scale bars = 30 μm. n (the number of experiments) = 3, ∗P < 0.05; ∗∗P < 0.01.

3.4. circEif3c functions as a micro-sponge for miR-96–5p

We hypothesized that circEif3c acts as a miRNA sponge, potentially targeting miR-96–5p or other miRNAs. Under experimental conditions consistent with those in Fig. 3A-C, we integrated microarray data with TargetScan and circBank databases to preliminarily identify the top three miRNAs exhibiting both strong binding potential to circEif3c and significant differential expression [Fig. 3I–K]. We then designed biotinylated probes targeting the backsplice junction of circEif3c and confirmed their high pull-down efficiency in AFs transfected with OE-circEif3c or OE-control plasmids [Fig. 3L]. RNA pull-down assays revealed that the circEif3c probe specifically enriched miR-96–5p from AF lysates, while no enrichment was observed for miR-15a-5p or miR-322–5p. This indicated that miR-96–5p is the predominant miRNA bound by circEif3c [Fig. 3L]. Subsequent AGO2-RIP assays confirmed that circEif3c and miR-96–5p interact with the AGO2 protein, supporting circEif3c's role in miRNA binding [Fig. 3M]. These results provide a molecular basis for circEif3c highly likely functioning as a “molecular sponge” for miR-96–5p and suggest that its regulatory effects in target cells are likely mediated through miR-96-5p-induced translational repression or degradation of downstream targets (such as PHF20L1 or MEOX2), rather than through other transcriptional or post-transcriptional mechanisms [Fig. 3M]. Integrative bioinformatic analysis further identified two specific binding sites between circEif3c and miR-96–5p, leading to the construction of a full-site mutant circEif3c plasmid [Fig. 3N]. When AFs were transfected with the circEif3c overexpression plasmid, the circEif3c probe pulled down significantly more miR-96–5p compared to the control probe [Fig. 3O]. Confocal microscopy revealed cytoplasmic colocalization of Cy5-labeled circEif3c and Cy3-labeled miR-96–5p in transfected AFs, suggesting their mutual enrichment and potential interaction in the cytoplasm [Fig. 3P]. Together, these findings demonstrate that circEif3c selectively sponges miR-96–5p in both PVPACs and AFs, thereby modulating the biological functions of target cells.

3.5. Exosomal circEif3c promoted biological functions of AFs

To explore the biological function of circEif3c derived from PVPAC exosomes (Exo-circEif3c) in AFs, the exosomal circEif3c could be taken up by AFs. PAPVC-Exo was isolated and incubated with AFs. It was found that, compared with the group without exosome treatment (control, added PBS buffer), the expression of circEif3c was upregulated in the PVPAC-Exo group approximately 6 h and 12 h after treatment [Fig. 4A]. These results confirmed that PVPAC-Exo-circEif3c might enter AFs by being assembled into exosomes. Next, we constructed a stable cell line with circEif3c knockdown in AFs. The results displayed that, compared with the siR-control group, circEif3c was significantly downregulated in the siR-circEif3c group, while its parental gene remained unchanged [Fig. 4B]. Loss of function and gain of function analyses were conducted in vitro to explore the effect of Exo-circEif3c on the biological functions of AFs [Fig. 4C and D]. To evaluate the biological effects of PVPAC-Exo-mediated circEif3c knockdown on AFs, exosomes were isolated from stable cell lines. Scratch wound healing and EdU assays revealed that exosomes overexpressing Exo-siR-circEif3c significantly impaired the migratory and proliferative capacities of AFs [Fig. 4C and D]. Specifically, both Exo-siR-circEif3c-1 and -2 groups exhibited markedly delayed wound closure and reduced cell migration compared to the control, Exo-siR-control, and Exo-circEif3c groups. Conversely, the Exo-circEif3c group displayed accelerated migration and wound healing relative to all other groups [Fig. 4C and D]. Flow cytometry (FCM) analysis revealed that Exo-circEif3c, Exo-siR-miR-96–5p, and Exo-siR-MEOX2 significantly promoted AF proliferation and suppressed apoptosis. In contrast, Exo-siR-circEif3c, Exo-miR-96–5p, and Exo-MEOX2 markedly inhibited AF proliferation and promoted apoptosis. Additionally, Exo-MEOX2 was able to reverse the pro-proliferative effect induced by Exo-circEif3c [Fig. 4E and F]. Additionally, Western blot analysis revealed that siR-circEif3c could cause the downregulation of vimentin and PHF20L1 expression, while, conversely, it upregulated MEOX2 expression [Fig. 4G]. Compared with the non-exosome control, exosome-delivered circEif3c significantly enhanced AF bioactivity, PHF20L1 expression and autophagy while down-regulating MEOX2; its absence produced the opposite effects [Fig. 4H].

Fig. 4 — **Role of PVPAC-Exo-circEif3c in regulating AF biological functions and its potential mechanism.** PVPAC-derived exosomal circEif3c (Exo-circEif3c) promoted AFs migration and proliferation, whereas silencing exosomal circEif3c suppresses these processes. (A) Time-course analysis of circEif3c expression in AFs after Exo-circEif3c treatment (0, 6, and 12 h; 0 h as control). (B) Stable silencing efficiency and specificity of circEif3c in AFs; Exo-siR-control served as the control. (C and D) Effects of PVPAC-Exo-siR- circEif3c-1 and -2 on AF migration and proliferation assessed by wound healing and proliferation assays. Scratch closure percentage and migrated cell numbers were quantified using ImageJ and GraphPad Prism 9.5, scale bar = 150 μm. (E) and (F) FCM analysis of AF proliferation and apoptosis following treatment with PVPAC-Exo-circEif3c, Exo-miR-96–5p, and Ad-MEOX2 interaction. (G) Western blot analysis of vimentin, PHF20L1, and MEOX2 expression in AFs under high glucose and circEif3c modulation. (H) Effects of Exo-circEif3c on the expression of vimentin, PHF20L1, MEOX2, and LC3 in AFs. GAPDH was used as a loading control. All data above are presented as mean ± SD from three independent experiments. vs. the control group, ∗P < 0.05, ∗∗P < 0.01(one-way ANOVA with Dunnett's post-hoc test), n (the number of experiments) = 3.

3.6. miR-96–5p regulated the biological functions of AFs through the PHF20L1/MEOX2 signaling pathway

To validate the interplay between miR-96–5p and PVPAC-Exo, AFs were pretreated with PVPAC-Exo for varying durations, and the expression level of miR-96–5p in AFs was measured accordingly. The results demonstrated gradual enhancement in miR-96–5p expression with prolonged incubation time [Fig. 5A]. miR-96–5p markedly suppressed both the migration and proliferation of AFs, whereas its knockdown significantly enhances these capacities [Fig. 5B and C]. Correlation analysis between extracellular ncRNA in cell culture supernatant and intracellular expression of PHF20L1 and MEOX2 was determined by RT-qPCR [Fig. 5D]. Accordingly, in exosomes, circEif3c was positively correlated with PHF20L1 but significantly negatively correlated with MEOX2. Conversely, miR-96–5p was negatively correlated with PHF20L1 and significantly positively correlated with MEOX2. [Fig. 5D]. In preliminary experiments, tissue microarrays of arterial PVAT proliferation–associated vascular remodeling in diabetic mice were examined in conjunction with bioinformatics analyses that integrated the TargetScan, ENCORI, and miRDB databases, revealing PHF20L1 and MEOX2 as the primary target proteins of miR-96–5p [Fig. 5E]. Bioinformatics analysis identified miR-96–5p binding sites in the 3′UTR of PHF20L1 [Fig. 5F], and luciferase reporter assays confirmed that the miR-96–5p mimic markedly reduced luciferase activity in the PHF20L1-WT group [Fig. 5G]. Although bioinformatics analysis indicated the presence of miR-96–5p binding sites in MEOX2 [Fig. 5H], the miR-96–5p mimic significantly reduced PHF20L1 protein expression in AFs [Fig. 5G], while emerging no notable effect on the differential expression of MEOX2 protein [Fig. 5I]. Western blot analysis of PHF20L1 and MEOX2 expression in AFs incubated for 24 h with control exosomes (Exo-control), PVPAC-Exo overexpressing (OE-exosomes), Exo-miR- control, Exo-miR-96–5p mimic, and the PHF20L1-specific inhibitor UNC1215, respectively [Fig. 5I]. Relative to control exosomes or Exo-miR-control, OE-exosomes elevated the expression of vimentin and PHF20L1 while suppressing MEOX2, whereas Exo-miR-96–5p mimic produced the opposite profile-down-regulating vimentin and PHF20L1 and up-regulating MEOX2. These data indicated that exosomal miR-96–5p modulates AF biology via the PHF20L1/MEOX2 axis [Fig. 5I]. By integrating GEPIA, ENCORI, miRNet, NDEx, Cytoscape, and BioGRID, a complex network interconnecting the ncRNAs (circ-Eif3c and miR-96–5p) with PHF20L1 and MEOX2 was graphically mapped [Fig. 5J]. Western blot analysis revealed differential the expression of vimentin, PHF20L1 and MEOX2 in AFs treated with exosome, Exo-miR-96–5p, and UNC1215 [Fig. 5K]. OE-exosomes significantly increased the expression of vimentin and PHF20L1 while decreasing MEOX2 levels compared to controls (control-exosome or miR-control). Conversely, the Exo-miR-96–5p mimic and UNC1215 significantly reduced vimentin and PHF20L1 expression and increased that of MEOX2. This phenotypic reversal suggests that exosome-derived miR-96–5p regulates AF biology potentially through the PHF20L1/MEOX2 pathway [Fig. 5K]. Furthermore, ZDOCK 3.0.2 and PyMOL 2.5.5 revealed a specific interface between PHF20L1 and MEOX2: ASP30 of PHF20L1 forms a hydrogen bond with TRP92 of MEOX2 [Fig. 5K]. SwissDock and PyMol clearly revealed hydrogen bonding sites between PHF20L1 and MEOX2 [Fig. 5L]. Their interaction was further confirmed by CO-IP [Fig. 5M].

Fig. 5 — **The regulatory mchanisms of miR-96-5p in AF biology.** (A) Time-course analysis of miR-96–5p expression in AFs treated with PVPAC-Exo at 0, 6, 12, and 24 h. (B) and (C) AFs were transfected with miR-96–5p mimic and NC mimic for 24 h. Then, the migration ability of AFs through migration experiments (B) and EDU assay (C) were evaluated using Image J and GraphPad Prism 9. vs. control mimic, ∗P < 0.05, ∗∗P < 0.01, n (the number of experiments) = 3, scale bar = 150 μm. (D) Correlation analysis between extracellular ncRNA levels in culture supernatant and intracellular expression of PHF20L1 and MEOX2 by RT-qPCR. (E) Bioinformatic prediction identifying PHF20L1 and MEOX2 as potential targets of miR-96–5p. (F) Predicted miR-96–5p binding sites in the 3′UTR of PHF20L1. (G) Luciferase reporter assay displayed that miR-96–5p mimic significantly reduced luciferase activity in the PHF20L1-WT group, but not in the PHF20L1-Mut group (vs. PHF20L1-Mut, ∗∗P < 0.01), n (the number of experiments) = 3. (H) Bioinformatics analysis indicated the miR-96–5p binding sites in the 3′UTR of MEOX2. (I) Western blot exhibited no significant change in MEOX2 protein levels upon miR-96–5p overexpression. (J) Interaction network among miR-96–5p, circEif3c, PHF20L1, and MEOX2, constructed using GEPIA, ENCORI, miRNet, NDEx, and Cytoscape. (K) Western blot analysis of PHF20L1 and MEOX2 expression in AFs transfected with control-exosome, OE-exosomes, miR-comtrol mimic, miR-96–5p mimic, and UNC1215, respectively. vs. control-exosome, miR-comtrol mimic, ∗P < 0.05, ∗∗P < 0.01, n (the number of experiments) = 3. (L) Predicted protein–protein interaction interface between PHF20L1 and MEOX2 using Zdock 3.0.2 and PyMOL 2.5.5. (M) Co-IP experiments confirmed an interaction between PHF20L1 and MEOX2.

3.7. PVPAC-(Exo-)circEif3c promoted the migration and proliferation of AFs through the miR-96–5p/PHF20L1/MEOX2 signaling

In vitro, loss- and gain-of-function assays confirmed that overexpression of circEif3c and PHF20L1, combined with underexpression of miR-96–5p and MEOX2, accelerated the biological functions of AFs. Conversely, knockdown of PVPAC-derived circEif3c and PHF20L1, along with overexpression of miR-96–5p and MEOX2, restrained the biological functions of AFs [Fig. 6A-C]. Moreover, circEif3c was able to counteract the inhibitory effects of miR-96–5p or MEOX2 on AFs. Functionally, these results supported a molecular mechanism whereby circEif3c acts as a sponge for miR-96–5p, ultimately repressing the inhibitory transcription factor MEOX2 and thereby promoting the biological functions of AFs [Fig. 6A-C]. At the in vitro level, MEOX2 exhibited a significantly stronger regulatory effect on AFs at the transcriptional level compared to ncRNAs. This mechanistic difference stems from MEOX2 mediating direct transcriptional regulation of AFs, which is more potent than the indirect regulatory effects exerted by ncRNAs (e.g., circEif3c and miR-96–5p) through mechanisms like miRNA sequestration [Fig. 6A-C]. Meanwhile, PVPAC-Exo was co-incubated with AFs to examine how exosomal ncRNAs and PHF20L1/MEOX2 affect AF function. We found that circEif3c in PVPAC exosomes positively correlated with PHF20L1 expression but negatively with MEOX2 in AFs. In contrast, exosomal miR-96–5p was negatively associated with PHF20L1 and positively correlated with MEOX2 [Fig. 6E]. These trends align closely with those previously observed for non-exosomal ncRNAs in relation to PHF20L1 and MEOX2 [Fig. 6C]. After 48 h of co-incubation with AFs using control mimic, siR-control, Exo-(siR-)circEif3c mimic, Exo-(siR-)miR-96–5p mimic, Exosome (PVPAC-Exo), GW4689, and Exo-(siR-) pAd-MEOX2, respectively, the results exhibited that PVPAC-Exo-circEif3c mimic, siR-miR-96–5p mimic, and siR-pAd-MEOX2 significantly stimulated the migration and proliferation of AFs. In contrast, GW4689, siR-circEif3c mimic, and pAd-MEOX2 markedly suppressed the migration and proliferation of AFs. These effects were more pronounced than those induced by their non-exosomal counterparts [Fig. 6D-F]. This result supported the molecular mechanism of the exosomes derived from PVPAC mediate the circEif3c/miR-96–5p/PHF20L1/MEOX2 signaling axis that drives the biological functions of AFs. Immunofluorescence co-localization clearly revealed the subcellular positions of circEif3c, miR-96–5p and MEOX2 in AFs, indicating a potential interaction among these molecules [Fig. 6G].

Fig. 6 — **CircEif3c modulates AF proliferation and migration via the miR-96-5p/PHF20L**1/MEOX2 axis. (A–C) Cell migration and proliferation assays. AFs were transfected for 24 h with Ad-GFP, siR-circEif3c, miR-96–5p mimic, or siR-MEOX2. Migration (A) and proliferation (B) were quantified (C). (D–F) AFs were co-incubated for 48 h with control mimic, Exo-(siR-)circEif3c mimic, Exo-(siR-)miR-96–5p mimic, PVPAC-exosome (Exo-control), GW4869, or Exo-siR-pAd-MEOX2. Migration (D) and proliferation (E) were assessed (F), scale bar = 150 μm. (G) Cellular fluorescence immunolocalization. nuclei (DAPI, blue), circEif3c (Cy5, red), miR-96–5p (Cy3, orange-yellow), MEOX2 (GFP, green).Scale bar = 30 μm. The above data were presented as mean ± SD. vs. Ad-GFP group, ∗P < 0.05, ∗∗P < 0.01, n (the number of experiments) = 3.

3.8. In vivo validation of the effects of the circEif3c/miR-96–5p/PHF20L1/MEOX2 signaling molecules on vascular remodeling

To further validate the effects and mechanisms of the aforementioned ncRNAs, PHF20L1, and MEOX2 on vascular remodeling in vivo, C57BL/6 mice (control) and ob/ob mice (DM model) were fed a high-fat diet to establish models [Fig. 7A]. A 4-week perivascular application of (Exo-Ad-)GFP, Exo-circEif3c mimic, Exo-miR-96–5p or Ad-MEOX2 gel onto the PVAT of the CCA was performed, followed by local transfection of Cy5-labeled circEif3c into PVAT at the lesion site for 24 h [Fig. 7B]. The mice were then sacrificed and the expression levels of the target molecules were compared between the control and lesion segments of the vessel wall. As anticipated, Exo-circEif3c, Exo-siR-miR-96–5p, and Exo-siR-MEOX2 significantly increased intima-media thickness (IMT), total vascular wall thickness, and lumen stenosis ratio, while decreasing lumen area and diastolic lumen diameter compared to control, circEif3c, and siR-miR-96–5p groups. Additionally, local inflammatory cell and adipocyte infiltration increased with disrupted tissue architecture [HE staining] [Fig. 7B]. Concurrently, the modeling site exhibited elevated peak systolic velocity (PSV) and resistance index (RI) [Fig. 7B]. Western blot analysis revealed that Exo-circEif3c and Exo-siR-miR-96–5p significantly upregulated the expression of PHL1, vimentin, LC3(II/I), and CD63, while downregulating MEOX2 [Fig. 7C-E]. Conversely, Exo-siR-circEif3c, Exo-miR-96–5p, and Exo-MEOX2 reduced IMT, total vascular wall thickness, and lumen stenosis ratio, increased lumen area and diastolic lumen diameter, decreased local inflammatory cell and adipocyte accumulation, and improved tissue organization. Furthermore, they downregulated the expression of PHL1, vimentin, LC3(II/I), and CD63, while upregulating MEOX2. Exo-MEOX2 and Exo-miR-96–5p were able to strikingly attenuate the proliferative vascular remodeling induced by Exo-circEif3c [Fig. 7C-E]. As described above, immunohistochemical analysis revealed Exo-circEif3c and Exo-siR-miR-96–5p could promote autophagy in PVAT, enhance vascular cell proliferation, and exacerbate proliferative vascular remodeling. In contrast, MEOX2 could reduce the expression of circEif3c, autophagy-related markers, and exosome production, while upregulating miR-96–5p expression. This ultimately attenuates vascular cell proliferation and thereby ameliorates proliferative vascular remodeling [Fig. 7C-E]. In vivo imaging revealed that, after topical application of DiR-labeled exosomes to the common carotid artery, mice in the experimental groups exhibited variable degrees of fluorescence at the treated site. The pattern mirrored the results obtained from Western blot and immunofluorescence analyses of tissue sections [Fig. 7F-G]: compared with the control group, fluorescence intensity at the locally injured CCA was markedly elevated in ob/ob (DM), MEOX2(±), PVPAC-Exo, Exo-circEif3c mimic and Exo-siR-miR-96–5p groups, indicating robust local accumulation of exosomes. Conversely, treatment with Exo-siR-circEif3c, Exo-miR-96–5p mimic, or Ad-MEOX2 significantly reduced exosome density at the carotid artery. Moreover, Ad-MEOX2 effectively abolished the pro-accumulation effects of both PVPAC-Exo and Exo-circEif3c on carotid exosome enrichment [Fig. 7C-I]. Therefore, in vivo experiments strongly support the conclusions drawn from in vitro studies: dysregulation of exosomal circEif3c and miR-96–5p constitutes a crucial mechanism underlying cellular dysfunction of AFs and vascular remodeling. Specifically, PVAT-Exo facilitates the circ-Eif3c/miR-96–5p/PHF20L1/MEOX2 signaling axis to accelerate proliferation in AFs and PVAT, ultimately driving vascular remodeling. Oppositely, MEOX2 over-expression efficiently suppresses circEif3c and PHF20L1, up-regulates miR-96–5p, and thereby inhibits AFs activity and ameliorates adverse vascular remodeling, positioning MEOX2 as a promising therapeutic gene for ASCVD.

Fig. 7 — **Exosomal circEif3c/miR-96-5p/PHF20L1/MEOX2 axis drives vascular remodeling in vivo.** (A) Workflow: a stable PVPAC line over-expressing circEif3c supplied exosomes (Exo-Ad-circEif3c, 10 μg/mouse) that were micro-injected into perivascular adipose tissue (PVAT) surrounding the left carotid artery for 4 weeks to initiate remodeling. Subsequently, after the model was established, treatments with (Exo)-Ad-GFP, (Exo)-Ad- circEif3c, (Exo)-Ad-miR-96–5p, and (Exo)-Ad-Meox2 were administered continuously for 2 weeks, respectively. Normal saline (NS) was used as a negative control. (B) Representative H&E-stained cross-sections and concomitant ultrasonography of the common carotid artery. Black scale bars = 50 μm, yellow scale bars = 1 mm, and white scale bars = 0.1 s. (C) Immunohistochemistry. Scale bars = 20 μm. (D) Western blotting. (E) Quantification of protein levels. (F) Tissue localization of Cy5-labeled circEif3c by immunofluorescence, scale bar = 100 μm. (G) Fluorescence intensity quantification. (H) Comparative fluorescence imaging of vascular sections: (H1) Bright-field H&E vs. dark-field GFP before and after Ad-MEOX2 transfection; Scale bars = 50 μm; (H2) DM-remodeling vs MEOX2-intervention groups. Scale bars = 30 μm. (I) Whole-animal in vivo imaging of Cy5 signal. All quantitative data above are presented as mean ± SD. vs. control, ∗P < 0.01.∗∗P < 0.01. n (the number of animals) = 6 in each group.

4. Discussion

Here, we demonstrate that PVPAC-Exo are rapidly internalized by AFs, triggering their migration and proliferation. CircEif3c was identified as the most highly enriched exosomal circRNA (Exo-circRNA) in PVPAC-Exo. Mechanistically, functional analyses confirmed that exosomal circEif3c promotes AF migration and proliferation by sequestering miR-96–5p to activate the PHF20L1/MEOX2 axis, hence drive AF migration and proliferation and ultimately fuels vascular remodeling [Fig. 8].

Fig. 8 — Schematic illustration of the PVPAC-Exo mediated circEif3c/miR-96–5p/PHF20L1/MEOX2 axis regulating vascular remodeling.

Intercellular communication is indispensable to cellular function. Cell-to-cell communication requires the transmission and reception of biological signals [25]. Each cell has the potential to both receive and send signals; however, under different ecological conditions, distinct cells may assume different communication roles. Identifying which cells act as the commanding secretory cells (sending cells) that issue signals and which serve as the responsive worker cells (receiving cells) that receive them is the primary issue that needs to be clarified in this study. This study, utilizing co-culture systems and exosome tracing techniques, demonstrated that PVPACs act as the signal sender, whereas AFs function as the recipient [Fig. 2]. Exosomes are a key node in intercellular communication between PVPACs and AFs in driving vascular remodeling; targeting their cargo can markedly influence the remodeling process. The arterial wall is composed of multiple cell types, including endothelial cells (ECs), macrophages, vascular smooth muscle cells (VSMCs), PVPACs and AFs. Their highly intricate intercellular communication is essential for maintaining normal vascular structure and physiological function [14]. Previous studies have predominantly focused on signaling between ECs and SMCs, while overlooking the crosstalk between the outermost PVPACs and AFs. To the best of our knowledge, this study is the first to demonstrate that PVPACs and AFs communicate via exosomes. Growing evidence suggests that exosomes serve as crucial mediators in intercellular communication and inter-organ crosstalk by delivering their cargo to target cells or through the interaction of exosomal signaling molecules with cell surface receptors [26]. Exosomes possess a topological structure similar to that of cells and contain substances such as DNA, RNA, proteins, lipids, small-molecule metabolites, and cell surface proteins [14]. Wang S. et al. found that under high-glucose conditions, exosomes secreted by mouse aortic endothelial cells (MAECs) promote high-glucose-induced VSMC proliferation and inhibit VSMC apoptosis. MAEC exosomes exposed to high glucose can deliver enriched circHIPK3 to VSMCs [27]. Exosomes secreted by adipose tissue play a pivotal role in coordinating communication between adipocytes and other key metabolic organs [26]. Exosomes can be released by various cells in response to different biological or chemical stimuli such as oxidative stress, low pH, and hypoxia [14]. Both PAPVC and AF are highly bioactive vascular cells that secrete and load abundant exosomes [Fig. 2]. This study demonstrated that PVPACs possess a markedly superior capacity for exosome secretion compared with AFs, functioning as the active “sender” that releases signals, whereas AFs act as the “receiver” that integrates these exosomal cues [Fig. 2]. Both in vivo and in vitro, exosome-mediated molecular effects consistently outweigh those observed with non-exosome delivery, underscoring the superior potency of the exosomal route [Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7]. Our experiments consistently identified that, at every critical step, exosomes serve as the delivery platform that enables ncRNAs to engage the PHF20L1/MEOX2 axis, thereby driving AF biology and initiating and propelling vascular remodeling [Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7].

Non-coding RNA (ncRNA) stands as a pivotal functional molecule within exosomes, playing a crucial role in promoting angiogenesis and driving the onset and progression of cardiovascular diseases [28,29]. This study confirms that high glucose stimulation significantly boosts the secretion of exosomes by PVPAC, elevates circEif3c expression, and facilitates AFs proliferation as well as the progression of vascular remodeling, as evidenced in Fig. 2, Fig. 3, Fig. 4, 6, and 7. Exosomes have been shown to selectively interact with specific target cells based on their cargo and origin, efficiently releasing their contents and thereby inducing phenotypic alterations in these cells. Our study further demonstrates that PVPAC-Exos not only secrete ncRNAs but also harness them to induce functional changes in target AF cells, as illustrated in Fig. 4, Fig. 5, Fig. 6. Imbalanced expression of circRNAs and miRNAs within exosomes is a key mechanism underlying cellular dysfunction. Ultra-sensitive RNA-seq has demonstrated that exosomal RNA composition is cell-type-specific, with oncogenic circRNAs/miRNAs enriched >10-fold in tumor-derived exosomes [30]. The dysregulation of exosomal circRNA functioning as a competitive endogenous RNA (ceRNA) is a core mechanism of cellular dysfunction [31]. This study uncovered that under high-glucose stimulation, the expression of circEif3c derived from PVPAC and AF was elevated by approximately 4-fold and 2-fold, respectively, compared with the control group. Notably, PVPAC-derived circEif3c expression was markedly higher than that from AF under high-glucose conditions [Fig. 2F]. This result further supports the notion that PVPAC is the active signal sender, while AF is the receiver.

Circular RNA not only possesses a more stable covalently closed-loop structure compared to other non-coding RNAs but also exhibits diverse biological regulatory functions [28]. CircEif3c is an exonic circRNA derived from the back-splicing of exon 1 of the EIF3C gene (chr7:126551975-126552338), establishing a “host gene-circRNA” relationship with EIF3C [32]. Recently, Zhong XM et al. reported that circEIF3C promotes proliferation, progression and immune evasion of human intrahepatic cholangiocarcinoma via the miR-34a-5p/B7-H4 axis [32]. This study not only confirmed that PVPAC-Exo-circEif3c could be taken up by AFs, but also demonstrated that the expression of circEif3c gradually increased with the prolongation of incubation time between PVPAC-Exos and AFs [Fig. 4A]. As expected, circRNA-RIP, pull-down, and luciferase reporter assays (using wild-type/mutant circEIF3C or circEIF3C knockdown) all confirmed the specific interaction between circEIF3C and miR-miR-96–5p [Fig. 3H and 3L-3O]. Rescue experiments employing either Exo-circEif3c knockdown or overexpression confirmed that, compared to the normal control, Exo-circEif3c markedly enhanced the bioavailability of circEif3c, significantly promoted AF migration and proliferation, and substantially reduced AF apoptosis [Fig. 4C-F]. Under combined high-glucose and circEif3c stimulation, AFs exhibited elevated vimentin and PHF20L1 expression, whereas MEOX2 levels were markedly reduced. Conversely, siR-circEif3c reversed the HG–induced up-regulation of PHF20L1 and restored MEOX2 expression [Fig. 4G], indicating that circEif3c promotes AF activation and proliferation via the PHF20L1/MEOX2 signaling axis. Moreover, compared to the exosome-free control, PVPAC-Exos pre-treated with siR-circEif3c significantly reduced PHF20L1 expression and enhanced MEOX2 expression in AFs. Conversely, PVPAC-Exos pre-treated with circEif3c markedly increased PHF20L1 expression while decreasing MEOX2 expression in AFs [Fig. 4H]. These findings demonstrate that the robust capacity of exosomes to package and deliver functional genetic cargo [Fig. 4].

The most prevalent and fundamental way for circRNAs to exert biological functions is by interacting with downstream miRNAs. Consistent with the classic mechanism of ncRNA interactions, this study confirmed that PVPAC-Exo-circEif3c acts as a molecular sponge to sequester and inhibit miR-96–5p, thereby regulating the biological functions of AFs. miR-96–5p is a transcriptional target of p21CIP1 and ZEB1 that suppresses cellular proliferation and induces fibroblast senescence [33]. However, no studies have reported the regulation of the transcription factor MEOX2 and zinc finger protein PHF20L1 by miR-96–5p. Although some studies have reported that miR-96–5p acts as a competing endogenous RNA (ceRNA) to circRNA and thereby accelerates the bio-functions of ovarian cancer cell [34], our findings and the majority of evidence indicate that miR-96–5p restrains target cell biofuntions [33]. Whether miR-96–5p promotes or suppresses the proliferation of target cells requires a comprehensive assessment within specific tissue/cellular contexts, considering its target genes and signaling pathways. miR-96–5p significantly inhibited the migration and proliferation capabilities of Afs [Fig. 4, Fig. 5C]. Mechanistically, through bioinformatic analysis and dual-luciferase reporter assays, we identified that miR-96–5p has direct binding sites and exerts a regulatory effect on PHF20L1 [Fig. 5D-F]. Although bioinformatics predicted interaction sites between miR-96–5p and MEOX2, dual-luciferase reporter assays demonstrated that miR-96–5p might not exert a direct regulatory effect on MEOX2 [Fig. 5G and H]. To dissect the exosomal miR-96–5p/PHF20L1/MEOX2 signaling axis, AFs were pre-incubated with PVPAC-Exo, miR-96–5p mimic, or UNC1215 (PHF20L1 inhibitor). Consequently, exosomes boosted vimentin/PHF20L1 and depressed MEOX2, whereas miR-96–5p or UNC1215 generated the opposite profile and completely reversed these effects [Fig. 5I], indicating that miR-96–5p tunes AF behaviour through the PHF20L1/MEOX2 pathway. As miR-96–5p targets PHF20L1 but not MEOX2, we postulated PHF20L1 to be upstream of MEOX2. In silico protein-protein interaction (PPI) screening predicted a hydrogen bond between PHF20L1-Asp30 and MEOX2-Trp92 [Fig. 5J and K], and Co-IP verified their direct binding [Fig. 5L]. qPCR of exosomes from HG cultures showed a strong positive correlation between circ-Eif3c and PHF20L1 and a negative correlation with MEOX2 [Fig. 5M].

In vitro and in vivo, circEif3c, miR-96–5p and MEOX2 were systematically manipulated to evaluate their impact on AF behaviour. Overexpression of circ-Eif3c, knock-down of miR-96–5p or silencing of MEOX2 markedly enhanced AF proliferation and migration, whereas miR-96–5p mimic or MEOX2 overexpression strongly suppressed both responses [Fig. 6A-C]. Packaging these interventions into PVPAC exosomes further amplified the effects: exosomal circ-Eif3c or si-MEOX2 produced the most pronounced changes in AF growth and motility [Fig. 6D-F]. Immunofluorescence confirmed cytoplasmic localization of ncRNAs (circEif3c and miR-96–5p) and nuclear presence of MEOX2, suggesting spatial interaction [Fig. 6G].

To further verify the in vivo effects and mechanisms, we established a perivascular hyperplasia model in MEOX2⁺/⁻ ob/ob and C57BL/6 mice using exosome-packed Ad-circEif3c. Local gel-based adventitial delivery of (Exo)-(siR)-circ-Eif3c, (Exo)- (siR)-miR-96–5p, and (Exo)-Ad-MEOX2 was performed [Fig. 7A]. After two weeks, (Exo)-circEif3c, (Exo)-siR-miR-96–5p, and (Exo)-Ad-MEOX2 were found to promote PVAT proliferation, increase IMT and wall thickness, elevate resistance index, and reduce lumen area, thereby accelerating remodeling [Fig. 7B]. Conversely, (Exo)-siR-circEif3c, (Exo)-miR-96–5p, and (Exo)-Ad-MEOX2 attenuated these effects and alleviated remodeling [Fig. 7C]. Notably, MEOX2^−/− was lethal, but MEOX2⁺/⁻ mice still exhibited significant perivascular hyperplasia, confirming model validity [Fig. 7D]. High-resolution ultrasound and in vivo fluorescence imaging further validated that ncRNAs/MEOX2 modulate vascular remodeling. Immunohistochemistry verified the circ-Eif3c/miR-96–5p/PHF20L1/MEOX2 axis as a key pathway driving hyperglycemia-induced vascular remodeling in atherosclerosis [Fig. 8].

5. Conclusion

To summarize, our findings demonstrated that under high glucose conditions, PVPAC-derived exosomes could load and deliver circEif3c to AFs, mediating intercellular communication, inducing proliferation and migration of AFs, inhibiting apoptosis, and thereby promoting the occurrence and progression of vascular remodeling. Mechanistically, under high glucose conditions, circEif3c was found in AFs to target and regulate miR-96–5p, triggering the PHF20L1/MEOX2 signaling pathway, which leads to vascular cell proliferation and the initiation and development of vascular remodeling. This study revealed a novel mechanism of hyperglycemia-induced proliferative vascular remodeling and might provide effective strategies for preventing and treating pathological vascular remodeling [Fig. 8].

6. Limitation

Firstly, the precise mechanisms by which PVPAC-secreted exosomes are recognized, captured, and utilized by adjacent AF cells remain to be thoroughly investigated. Secondly, between ncRNAs and between ncRNAs and their target proteins lies a complex interaction network; however, our study focuses on only two ncRNAs and does not address interactions involving lncRNAs or other target proteins. Thirdly, the study did not employ the latest or most advanced exosome technologies, such as exosome barcoding (EXO-BC), to track exosome trajectories. Our tracking methods, while demonstrating uptake, cannot discriminate between administered and endogenous exosomes. Fourthly, the mechanisms by which exosomes are specifically recognized, internalized, and transported were not explored. Fifthly, intravascular ultrasound (IVUS) holds an irreplaceable, central position as one of the “gold-standard” tools for evaluating vascular remodeling in both clinical research and interventional procedures. In the present study, however, the small diameter of mouse arteries renders IVUS technically impractical for assessing carotid atherosclerotic plaques and vascular remodeling. Additionally, due to the spectral similarity or overlap among the fluorescent labels used for the four target molecules, this study was unable to perform simultaneous immunofluorescence co-localization of circEif3c, miR-96–5p, PHF20L1, and MEOX2. Finally, due to the limited volume of venous blood obtainable from mice and the consequent instability in ncRNA measurements, which could introduce significant bias into the statistical analysis, the ncRNA results from the animal models were excluded from this study.

CRediT authorship contribution statement

Yixuan Liu: Writing – original draft, Methodology, Investigation, Data curation. Peiqing Tian: Writing – review & editing, Investigation, Data curation. Shuaiyong Zhang: Investigation, Data curation. Jiayu Wang: Investigation, Data curation. Ye Zhang: Investigation. Shuxin Ge: Investigation. Qinghai Wang: Resources. Peng Wang: Investigation. Juan Zhang: Investigation. Ping Liu: Writing – review & editing, Writing – original draft, Supervision, Resources, Methodology, Funding acquisition, Data curation, Conceptualization.

Ethics approval and consent to participate

All animal care and experimental protocols were approved by the Research Ethics Committee of the Qilu Second Hospital of Shandong University [No. KYLL-2021(KJ)A-0027] and complied with the national standards established by the National Institutes of Health.

Availability of data and material

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

Funding

This work was supported by the National Natural Science Foundation of China (81170274, 82170462), the Major Research & Development Program of Shandong Province (ZR2020MH041), the Central Government Guidance Fund for Local Science and Technology Development (2021Szvup073), the Science and Technology Innovation Program of Jinan (201602153, 202019193), the “20 Universities Regulations” Innovation Team Grant of Jinan (2021GXRC107), and the Natural Science Foundation of Shandong Province (GG201703080074).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Footnotes

Peer review under the responsibility of Editorial Board of Non-coding RNA Research.

References

1.Sattar N., McMurray J., Borén J., Rawshani A., Omerovic E., Berg N., Halminen J., Skoglund K., Eliasson B., Gerstein H.C., McGuire D.K., Bhatt D., Rawshani A. Twenty years of cardiovascular complications and risk factors in patients with type 2 diabetes: a nationwide Swedish cohort study. Circulation. 2023;147(25):1872–1886. doi: 10.1161/CIRCULATIONAHA.122.063374. [DOI] [PubMed] [Google Scholar]
2.Sigdel S., Udoh G., Albalawy R., Wang J. Perivascular adipose tissue and perivascular adipose tissue-derived extracellular vesicles: new insights in vascular disease. Cells. 2024;13(16):1309. doi: 10.3390/cells13161309. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Antoniades C., Tousoulis D., Vavlukis M., Fleming I., Duncker D.J., Eringa E., Manfrini O., Antonopoulos A.S., Oikonomou E., Padró T., Trifunovic-Zamaklar D., De Luca G., Guzik T., Cenko E., Djordjevic-Dikic A., Crea F. Perivascular adipose tissue as a source of therapeutic targets and clinical biomarkers. Eur. Heart J. 2023;44(38):3827–3844. doi: 10.1093/eurheartj/ehad484. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Wang Z., Mu W., Xu R., Zhong J., Xiong W., Zhao X., Liang X., Guo Y., Zhang J., Jiang Z.S., Yang B., Chen Y.E., Chang L. Perivascular adipose tissue dysfunction contributes to thoracic aortic aneurysm development. Cardiovasc. Diabetol. 2025;24(1):223. doi: 10.1186/s12933-025-02765-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Kim H.W., Shi H., Winkler M.A., Lee R., Weintraub N.L. Perivascular adipose tissue and vascular perturbation/atherosclerosis. Arterioscler. Thromb. Vasc. Biol. 2020;40(11):2569–2576. doi: 10.1161/ATVBAHA.120.312470. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Jiang Y., Liu P., Jiao W., Meng J., Feng J. Gax suppresses chemerin/CMKLR1-induced preadipocyte biofunctions through the inhibition of Akt/mTOR and ERK signaling pathways. J. Cell. Physiol. 2018;233(1):572–586. doi: 10.1002/jcp.25918. [DOI] [PubMed] [Google Scholar]
7.Angueira A.R., Sakers A.P., Holman C.D., Cheng L., Arbocco M.N., Shamsi F., Lynes M.D., Shrestha R., Okada C., Batmanov K., Susztak K., Tseng Y.H., Liaw L., Seale P. Defining the lineage of thermogenic perivascular adipose tissue. Nat. Metab. 2021;3(4):469–484. doi: 10.1038/s42255-021-00380-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Liu P., Zhang C., Feng J.B., Zhao Y.X., Wang X.P., Yang J.M., Zhang M.X., Wang X.L., Zhang Y. Cross talk among smad, MAPK, and integrin signaling pathways enhances adventitial fibroblast functions activated by transforming growth factor-beta1 and inhibited by gax. Arterioscler. Thromb. Vasc. Biol. 2008;28(4):725–731. doi: 10.1161/ATVBAHA.107.159889. [DOI] [PubMed] [Google Scholar]
9.Liu P., Zhang C., Zhao Y.X., Feng J.B., Liu C.X., Chen W.Q., Yao G.H., Zhang M., Wang X.L., Zhang Y. Gax gene transfer inhibits vascular remodeling induced by adventitial inflammation in rabbits. Atherosclerosis. 2010;212(2):398–405. doi: 10.1016/j.atherosclerosis.2010.06.001. [DOI] [PubMed] [Google Scholar]
10.Scheja L., Heeren J. The endocrine function of adipose tissues in health and cardiometabolic disease. Nat. Rev. Endocrinol. 2019;15(9):507–524. doi: 10.1038/s41574-019-0230-6. [DOI] [PubMed] [Google Scholar]
11.Kalluri R., LeBleu V.S. The biology, function, and biomedical applications of exosomes. Science. 2020;367(6478) doi: 10.1126/science.aau6977. eaau6977. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Crewe C., Joffin N., Rutkowski J.M., Kim M., Zhang F., Towler D.A., Gordillo R., Scherer P.E. An endothelial-to-adipocyte extracellular vesicle axis governed by metabolic state. Cell. 2018;175(3):695–708.e13. doi: 10.1016/j.cell.2018.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Singh D.D., Kim Y., Choi S.A., Han I., Yadav D.K. Clinical significance of MicroRNAs, long non-coding RNAs, and CircRNAs in cardiovascular diseases. Cells. 2023;12(12):1629. doi: 10.3390/cells12121629. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Zhang X., Sun S., Ren G., Liu W., Chen H. Advances in intercellular communication mediated by exosomal ncRNAs in cardiovascular disease. Int. J. Mol. Sci. 2023;24(22) doi: 10.3390/ijms242216197. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Wang Y., Liu B. Circular RNA in diseased heart. Cells. 2020;9(5):1240. doi: 10.3390/cells9051240. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Nemeth K., Bayraktar R., Ferracin M., Calin G.A. Non-coding RNAs in disease: from mechanisms to therapeutics. Nat. Rev. Genet. 2024;25(3):211–232. doi: 10.1038/s41576-023-00662-1. [DOI] [PubMed] [Google Scholar]
17.Wang J., Chen Y., Wang Q., Xu H., Wu C., Jiang Q., Wu G., Zhou H., Xiao Z., Chen Y., Zhang T., Lan Q. MEOX2-mediated regulation of cathepsin S promotes cell proliferation and motility in glioma. Cell Death Dis. 2022;13(4):360. doi: 10.1038/s41419-022-04845-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Wu Z., Guo H., Chow N., Sallstrom J., Bell R.D., Deane R., Brooks A.I., Kanagala S., Rubio A., Sagare A., Liu D., Li F., Armstrong D., Gasiewicz T., Zidovetzki R., Song X., Hofman F., Zlokovic B.V. Nat. Med. 2005;11(9):959–965. doi: 10.1038/nm1287. [DOI] [PubMed] [Google Scholar]
19.Chen Y., Gorski D.H. Regulation of angiogenesis through a microRNA (miR-130a) that down-regulates antiangiogenic homeobox genes GAX and HOXA5. Blood. 2008;111(3):1217–1226. doi: 10.1182/blood-2007-07-104133. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Coppiello G., Collantes M., Sirerol-Piquer M.S., Vandenwijngaert S., Schoors S., Swinnen M., Vandersmissen I., Herijgers P., Topal B., van Loon J., Goffin J., Prósper F., Carmeliet P., García-Verdugo J.M., Janssens S., Peñuelas I., Aranguren X.L., Luttun A. Meox2/Tcf15 heterodimers program the heart capillary endothelium for cardiac fatty acid uptake. Circulation. 2015;131(9):815–826. doi: 10.1161/CIRCULATIONAHA.114.013721. [DOI] [PubMed] [Google Scholar]
21.Liu P., Feng J., Kong F., Lu Q., Xu H., Meng J., Jiang Y. Gax inhibits perivascular preadipocyte biofunction mediated by IGF-1 induced FAK/Pyk2 and ERK2 cooperative pathways. Cell. Signal. 2014;26(12):3036–3045. doi: 10.1016/j.cellsig.2014.09.017. [DOI] [PubMed] [Google Scholar]
22.Liu P., Kong F., Wang J., Lu Q., Xu H., Qi T., Meng J. Involvement of IGF-1 and MEOX2 in PI3K/Akt1/2 and ERK1/2 pathways mediated proliferation and differentiation of perivascular adipocytes. Exp. Cell Res. 2015;331(1):82–96. doi: 10.1016/j.yexcr.2014.09.011. [DOI] [PubMed] [Google Scholar]
23.Liu Z., Wang J., Tian P., Liu Y., Xing L., Fu C., Huang X., Liu P. Sodium-glucose cotransporter 1 promotes the biofunctions of perivascular preadipocytes mediated by Akt/mTOR/p70S6K signaling pathway. Am. J. Physiol. Cell Physiol. 2024;326(6):C1611–C1624. doi: 10.1152/ajpcell.00606.2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Faul F., Erdfelder E., Buchner A., Lang A.G. Statistical power analyses using G∗Power 3.1: tests for correlation and regression analyses. Behav. Res. Methods. 2009 Nov;41(4):1149–1160. doi: 10.3758/BRM.41.4.1149. [DOI] [PubMed] [Google Scholar]
25.Hu K., Le Vo K.L., Wang F., Zhang X., Gu C., Fang N., Phan N.T.N., Ewing A.G. Single exosome amperometric measurements reveal encapsulation of chemical messengers for intercellular communication. J. Am. Chem. Soc. 2023;145(21):11499–11503. doi: 10.1021/jacs.3c02844. [DOI] [PubMed] [Google Scholar]
26.Tang Y., Yang L.J., Liu H., Song Y.J., Yang Q.Q., Liu Y., Qian S.W., Tang Q.Q. Exosomal miR-27b-3p secreted by visceral adipocytes contributes to endothelial inflammation and atherogenesis. Cell Rep. 2023;42(1) doi: 10.1016/j.celrep.2022.111948. [DOI] [PubMed] [Google Scholar]
27.Wang S., Shi M., Li J., Zhang Y., Wang W., Xu P., Li Y. Endothelial cell-derived exosomal circHIPK3 promotes the proliferation of vascular smooth muscle cells induced by high glucose via the miR-106a-5p/Foxo1/Vcam1 pathway. Aging (Albany NY) 2021;13(23):25241–25255. doi: 10.18632/aging.203742. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Mao L.W., Yue Z., Zhu D.S., Lu C., Cheng K. Exosomes: bridge metabolic regulation in cardiac repair. npj Biomed. Innov. 2025;2(1):21–30. [Google Scholar]
29.Li C., Ni Y.Q., Xu H., Xiang Q.Y., Zhao Y., Zhan J.K., He J.Y., Li S., Liu Y.S. Roles and mechanisms of exosomal non-coding RNAs in human health and diseases. Signal Transduct. Targeted Ther. 2021;6(1):383. doi: 10.1038/s41392-021-00779-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Jeppesen D.K., Fenix A.M., Franklin J.L., Higginbotham J.N., Zhang Q., Zimmerman L.J., Liebler D.C., Ping J., Liu Q., Evans R., Fissell W.H., Patton J.G., Rome L.H., Burnette D.T., Coffey R.J. Reassessment of exosome composition. Cell. 2019;177(2):428–445.e18. doi: 10.1016/j.cell.2019.02.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Kristensen L.S., Andersen M.S., Stagsted L.V.W., Ebbesen K.K., Hansen T.B., Kjems J. The biogenesis, biology and characterization of circular RNAs. Nat. Rev. Genet. 2019;20(11):675–691. doi: 10.1038/s41576-019-0158-7. [DOI] [PubMed] [Google Scholar]
32.Zhong X., Ji C., Ren D., Ke A., Yang Z. Circular RNA circEIF3C promotes intrahepatic cholangiocarcinoma progression and immune evasion via the miR-34a-5p/B7-H4 axis. Genes Dis. 2022;10(2):370–372. doi: 10.1016/j.gendis.2022.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Santiago F.E., Adige T., Mahmud S., Dong X., Niedernhofer L.J., Robbins P.D. miR-96-5p expression is sufficient to induce and maintain the senescent cell fate in the absence of stress. Proc. Natl. Acad. Sci. U. S. A. 2024;121(40) doi: 10.1073/pnas.2321182121. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Wu M., Qiu Q., Zhou Q., Li J., Yang J., Zheng C., Luo A., Li X., Zhang H., Cheng X., Lu W., Liu P., Lu B., Lu Y. circFBXO7/miR-96-5p/MTSS1 axis is an important regulator in the Wnt signaling pathway in ovarian cancer. Mol. Cancer. 2022;21(1):137. doi: 10.1186/s12943-022-01611-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

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

[bib1] 1.Sattar N., McMurray J., Borén J., Rawshani A., Omerovic E., Berg N., Halminen J., Skoglund K., Eliasson B., Gerstein H.C., McGuire D.K., Bhatt D., Rawshani A. Twenty years of cardiovascular complications and risk factors in patients with type 2 diabetes: a nationwide Swedish cohort study. Circulation. 2023;147(25):1872–1886. doi: 10.1161/CIRCULATIONAHA.122.063374. [DOI] [PubMed] [Google Scholar]

[bib2] 2.Sigdel S., Udoh G., Albalawy R., Wang J. Perivascular adipose tissue and perivascular adipose tissue-derived extracellular vesicles: new insights in vascular disease. Cells. 2024;13(16):1309. doi: 10.3390/cells13161309. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] 3.Antoniades C., Tousoulis D., Vavlukis M., Fleming I., Duncker D.J., Eringa E., Manfrini O., Antonopoulos A.S., Oikonomou E., Padró T., Trifunovic-Zamaklar D., De Luca G., Guzik T., Cenko E., Djordjevic-Dikic A., Crea F. Perivascular adipose tissue as a source of therapeutic targets and clinical biomarkers. Eur. Heart J. 2023;44(38):3827–3844. doi: 10.1093/eurheartj/ehad484. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Wang Z., Mu W., Xu R., Zhong J., Xiong W., Zhao X., Liang X., Guo Y., Zhang J., Jiang Z.S., Yang B., Chen Y.E., Chang L. Perivascular adipose tissue dysfunction contributes to thoracic aortic aneurysm development. Cardiovasc. Diabetol. 2025;24(1):223. doi: 10.1186/s12933-025-02765-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Kim H.W., Shi H., Winkler M.A., Lee R., Weintraub N.L. Perivascular adipose tissue and vascular perturbation/atherosclerosis. Arterioscler. Thromb. Vasc. Biol. 2020;40(11):2569–2576. doi: 10.1161/ATVBAHA.120.312470. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 6.Jiang Y., Liu P., Jiao W., Meng J., Feng J. Gax suppresses chemerin/CMKLR1-induced preadipocyte biofunctions through the inhibition of Akt/mTOR and ERK signaling pathways. J. Cell. Physiol. 2018;233(1):572–586. doi: 10.1002/jcp.25918. [DOI] [PubMed] [Google Scholar]

[bib7] 7.Angueira A.R., Sakers A.P., Holman C.D., Cheng L., Arbocco M.N., Shamsi F., Lynes M.D., Shrestha R., Okada C., Batmanov K., Susztak K., Tseng Y.H., Liaw L., Seale P. Defining the lineage of thermogenic perivascular adipose tissue. Nat. Metab. 2021;3(4):469–484. doi: 10.1038/s42255-021-00380-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] 8.Liu P., Zhang C., Feng J.B., Zhao Y.X., Wang X.P., Yang J.M., Zhang M.X., Wang X.L., Zhang Y. Cross talk among smad, MAPK, and integrin signaling pathways enhances adventitial fibroblast functions activated by transforming growth factor-beta1 and inhibited by gax. Arterioscler. Thromb. Vasc. Biol. 2008;28(4):725–731. doi: 10.1161/ATVBAHA.107.159889. [DOI] [PubMed] [Google Scholar]

[bib9] 9.Liu P., Zhang C., Zhao Y.X., Feng J.B., Liu C.X., Chen W.Q., Yao G.H., Zhang M., Wang X.L., Zhang Y. Gax gene transfer inhibits vascular remodeling induced by adventitial inflammation in rabbits. Atherosclerosis. 2010;212(2):398–405. doi: 10.1016/j.atherosclerosis.2010.06.001. [DOI] [PubMed] [Google Scholar]

[bib10] 10.Scheja L., Heeren J. The endocrine function of adipose tissues in health and cardiometabolic disease. Nat. Rev. Endocrinol. 2019;15(9):507–524. doi: 10.1038/s41574-019-0230-6. [DOI] [PubMed] [Google Scholar]

[bib11] 11.Kalluri R., LeBleu V.S. The biology, function, and biomedical applications of exosomes. Science. 2020;367(6478) doi: 10.1126/science.aau6977. eaau6977. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Crewe C., Joffin N., Rutkowski J.M., Kim M., Zhang F., Towler D.A., Gordillo R., Scherer P.E. An endothelial-to-adipocyte extracellular vesicle axis governed by metabolic state. Cell. 2018;175(3):695–708.e13. doi: 10.1016/j.cell.2018.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] 13.Singh D.D., Kim Y., Choi S.A., Han I., Yadav D.K. Clinical significance of MicroRNAs, long non-coding RNAs, and CircRNAs in cardiovascular diseases. Cells. 2023;12(12):1629. doi: 10.3390/cells12121629. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Zhang X., Sun S., Ren G., Liu W., Chen H. Advances in intercellular communication mediated by exosomal ncRNAs in cardiovascular disease. Int. J. Mol. Sci. 2023;24(22) doi: 10.3390/ijms242216197. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] 15.Wang Y., Liu B. Circular RNA in diseased heart. Cells. 2020;9(5):1240. doi: 10.3390/cells9051240. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16.Nemeth K., Bayraktar R., Ferracin M., Calin G.A. Non-coding RNAs in disease: from mechanisms to therapeutics. Nat. Rev. Genet. 2024;25(3):211–232. doi: 10.1038/s41576-023-00662-1. [DOI] [PubMed] [Google Scholar]

[bib17] 17.Wang J., Chen Y., Wang Q., Xu H., Wu C., Jiang Q., Wu G., Zhou H., Xiao Z., Chen Y., Zhang T., Lan Q. MEOX2-mediated regulation of cathepsin S promotes cell proliferation and motility in glioma. Cell Death Dis. 2022;13(4):360. doi: 10.1038/s41419-022-04845-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.Wu Z., Guo H., Chow N., Sallstrom J., Bell R.D., Deane R., Brooks A.I., Kanagala S., Rubio A., Sagare A., Liu D., Li F., Armstrong D., Gasiewicz T., Zidovetzki R., Song X., Hofman F., Zlokovic B.V. Nat. Med. 2005;11(9):959–965. doi: 10.1038/nm1287. [DOI] [PubMed] [Google Scholar]

[bib19] 19.Chen Y., Gorski D.H. Regulation of angiogenesis through a microRNA (miR-130a) that down-regulates antiangiogenic homeobox genes GAX and HOXA5. Blood. 2008;111(3):1217–1226. doi: 10.1182/blood-2007-07-104133. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20.Coppiello G., Collantes M., Sirerol-Piquer M.S., Vandenwijngaert S., Schoors S., Swinnen M., Vandersmissen I., Herijgers P., Topal B., van Loon J., Goffin J., Prósper F., Carmeliet P., García-Verdugo J.M., Janssens S., Peñuelas I., Aranguren X.L., Luttun A. Meox2/Tcf15 heterodimers program the heart capillary endothelium for cardiac fatty acid uptake. Circulation. 2015;131(9):815–826. doi: 10.1161/CIRCULATIONAHA.114.013721. [DOI] [PubMed] [Google Scholar]

[bib21] 21.Liu P., Feng J., Kong F., Lu Q., Xu H., Meng J., Jiang Y. Gax inhibits perivascular preadipocyte biofunction mediated by IGF-1 induced FAK/Pyk2 and ERK2 cooperative pathways. Cell. Signal. 2014;26(12):3036–3045. doi: 10.1016/j.cellsig.2014.09.017. [DOI] [PubMed] [Google Scholar]

[bib22] 22.Liu P., Kong F., Wang J., Lu Q., Xu H., Qi T., Meng J. Involvement of IGF-1 and MEOX2 in PI3K/Akt1/2 and ERK1/2 pathways mediated proliferation and differentiation of perivascular adipocytes. Exp. Cell Res. 2015;331(1):82–96. doi: 10.1016/j.yexcr.2014.09.011. [DOI] [PubMed] [Google Scholar]

[bib23] 23.Liu Z., Wang J., Tian P., Liu Y., Xing L., Fu C., Huang X., Liu P. Sodium-glucose cotransporter 1 promotes the biofunctions of perivascular preadipocytes mediated by Akt/mTOR/p70S6K signaling pathway. Am. J. Physiol. Cell Physiol. 2024;326(6):C1611–C1624. doi: 10.1152/ajpcell.00606.2023. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] 24.Faul F., Erdfelder E., Buchner A., Lang A.G. Statistical power analyses using G∗Power 3.1: tests for correlation and regression analyses. Behav. Res. Methods. 2009 Nov;41(4):1149–1160. doi: 10.3758/BRM.41.4.1149. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Hu K., Le Vo K.L., Wang F., Zhang X., Gu C., Fang N., Phan N.T.N., Ewing A.G. Single exosome amperometric measurements reveal encapsulation of chemical messengers for intercellular communication. J. Am. Chem. Soc. 2023;145(21):11499–11503. doi: 10.1021/jacs.3c02844. [DOI] [PubMed] [Google Scholar]

[bib26] 26.Tang Y., Yang L.J., Liu H., Song Y.J., Yang Q.Q., Liu Y., Qian S.W., Tang Q.Q. Exosomal miR-27b-3p secreted by visceral adipocytes contributes to endothelial inflammation and atherogenesis. Cell Rep. 2023;42(1) doi: 10.1016/j.celrep.2022.111948. [DOI] [PubMed] [Google Scholar]

[bib27] 27.Wang S., Shi M., Li J., Zhang Y., Wang W., Xu P., Li Y. Endothelial cell-derived exosomal circHIPK3 promotes the proliferation of vascular smooth muscle cells induced by high glucose via the miR-106a-5p/Foxo1/Vcam1 pathway. Aging (Albany NY) 2021;13(23):25241–25255. doi: 10.18632/aging.203742. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Mao L.W., Yue Z., Zhu D.S., Lu C., Cheng K. Exosomes: bridge metabolic regulation in cardiac repair. npj Biomed. Innov. 2025;2(1):21–30. [Google Scholar]

[bib29] 29.Li C., Ni Y.Q., Xu H., Xiang Q.Y., Zhao Y., Zhan J.K., He J.Y., Li S., Liu Y.S. Roles and mechanisms of exosomal non-coding RNAs in human health and diseases. Signal Transduct. Targeted Ther. 2021;6(1):383. doi: 10.1038/s41392-021-00779-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30.Jeppesen D.K., Fenix A.M., Franklin J.L., Higginbotham J.N., Zhang Q., Zimmerman L.J., Liebler D.C., Ping J., Liu Q., Evans R., Fissell W.H., Patton J.G., Rome L.H., Burnette D.T., Coffey R.J. Reassessment of exosome composition. Cell. 2019;177(2):428–445.e18. doi: 10.1016/j.cell.2019.02.029. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31.Kristensen L.S., Andersen M.S., Stagsted L.V.W., Ebbesen K.K., Hansen T.B., Kjems J. The biogenesis, biology and characterization of circular RNAs. Nat. Rev. Genet. 2019;20(11):675–691. doi: 10.1038/s41576-019-0158-7. [DOI] [PubMed] [Google Scholar]

[bib32] 32.Zhong X., Ji C., Ren D., Ke A., Yang Z. Circular RNA circEIF3C promotes intrahepatic cholangiocarcinoma progression and immune evasion via the miR-34a-5p/B7-H4 axis. Genes Dis. 2022;10(2):370–372. doi: 10.1016/j.gendis.2022.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Santiago F.E., Adige T., Mahmud S., Dong X., Niedernhofer L.J., Robbins P.D. miR-96-5p expression is sufficient to induce and maintain the senescent cell fate in the absence of stress. Proc. Natl. Acad. Sci. U. S. A. 2024;121(40) doi: 10.1073/pnas.2321182121. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] 34.Wu M., Qiu Q., Zhou Q., Li J., Yang J., Zheng C., Luo A., Li X., Zhang H., Cheng X., Lu W., Liu P., Lu B., Lu Y. circFBXO7/miR-96-5p/MTSS1 axis is an important regulator in the Wnt signaling pathway in ovarian cancer. Mol. Cancer. 2022;21(1):137. doi: 10.1186/s12943-022-01611-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

CircEif3c/miR-96–5p/PHF20L1/MEOX2 axis in perivascular preadipocyte exosomes mediates fibroblast dysfunction and vascular remodeling

Yixuan Liu

Peiqing Tian

Shuaiyong Zhang

Jiayu Wang

Ye Zhang

Shuxin Ge

Qinghai Wang

Peng Wang

Juan Zhang

Ping Liu

Abstract

Background

Methods

Results

Conclusion

1. Introduction

2. Material and methods

2.1. Primary culture of dedifferentiated PVPAC and Oil Red staining experiment

2.2. Primary culture of AFs

2.3. Plasmid and siRNA transfection and adenovirus transduction

Table 1.

Table 2.

2.4. Extraction, identification, and tracking of exosomes

2.5. Screening, validation, amplification, and localization of ncRNAs

2.5.1. Construction of ncRNAs expression profiles and screening of differentially expressed ncRNAs

2.5.2. Validation of circRNA identity and circularization characteristics

2.5.3. RNA extraction, subcellular localization, and qRT-PCR semi-quantitative analysis

Table 3.

2.5.4. Tissue FISH

2.6. Co-immunoprecipitation (Co-IP) and Western blotting

2.7. Dual-luciferase reporter assay

2.8. Cell type identification and apoptosis detection by flow cytometry (FCM)

2.9. Real-time quantitative PCR (RT-qPCR)

2.10. Wound healing assay

2.11. Cell proliferation assay

2.12. Glucose uptake experiment

2.13. RNA immunoprecipitation (RIP)

2.14. Biotin-labeled probe RNA pull-down experiment

2.15. Animal experiments

2.16. In vivo imaging of mice with DIR-labeled exosomes

2.17. Immunohistochemistry (IHC) and immunofluorescence

2.18. Transthoracic vascular Doppler ultrasound in mice

2.19. Statistical analysis

3. Results

3.1. Effects of different concentrations of glucose on the biological functions of PVPACs

Fig. 1.

3.2. AF and PVPAC interact with each other through exosomes

Fig. 2.

3.3. circEif3c was identified and enriched in PVPAC-Exos

Fig. 3.

3.4. circEif3c functions as a micro-sponge for miR-96–5p

3.5. Exosomal circEif3c promoted biological functions of AFs

Fig. 4.

3.6. miR-96–5p regulated the biological functions of AFs through the PHF20L1/MEOX2 signaling pathway

Fig. 5.

3.7. PVPAC-(Exo-)circEif3c promoted the migration and proliferation of AFs through the miR-96–5p/PHF20L1/MEOX2 signaling

Fig. 6.

3.8. In vivo validation of the effects of the circEif3c/miR-96–5p/PHF20L1/MEOX2 signaling molecules on vascular remodeling

Fig. 7.

4. Discussion

Fig. 8.

5. Conclusion

6. Limitation

CRediT authorship contribution statement

Ethics approval and consent to participate

Availability of data and material

Funding

Declaration of competing interest

Footnotes

References

Associated Data

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases