UC-MSCs-derived exosomal circFOXP1 alleviates allergic rhinitis through EZH2/EGR1/Blimp-1 axis

Background

Allergic rhinitis (AR) is a frequently diagnosed allergic disease with high prevalence worldwide. However, the understanding of AR pathogenesis remains limited.

Methods

qRT-PCR, RNA fluorescence in situ hybridization (FISH), IHC or western blot were used to detect the expression of circFOXP1, EZH2, EGR1, Blimp-1 and other molecules. Flow cytometry and ELISA assay were employed to Th1/Th2 cell ratio and cytokine secretion, respectively. Umbilical cord-derived mesenchymal stem cells (UC-MSCs)-derived exosomes were isolated and characterized by transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA) and western blot. Bioinformatics analysis, RNA pull-down, RIP, ChIP and luciferase assays were used to examine the direct association among circFOXP1, EZH2, EGR1 and Blimp-1. The in vitro findings were validated in ovalbumin (OVA)-induced AR mouse model.

Results

circFOXP1 was downregulated, while EGR1 and Blimp-1 were upregulated in the peripheral blood and nasal mucosa of AR patients. This was accompanied with Th1/Th2 imbalance. hnRNPA2B1 mediated the packaging of circFOXP1 into exosomes, and UC-MSCs-derived exosomal circFOXP1 was internalized by CD4+ T cells. Silencing of circFOXP1 counteracted UC-MSCs-derived exosomes-suppressed Th2 differentiation of CD4+ T cells. Mechanistic experiments showed that EGR1 was a transcriptional activator of Blimp-1. circFOXP1 triggered epigenetic repression of EGR1 to transcriptionally suppress Blimp-1 through recruiting EZH2. In AR mice, UC-MSCs-derived exosomal circFOXP1 protected against the impairments, possibly via EZH2/EGR1/Blimp-1 axis.

Conclusion

UC-MSC-derived exosomal circFOXP1 alleviated allergic rhinitis, suggesting a role for the EZH2/EGR1/Blimp-1 axis in regulating Th2 differentiation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-026-03604-7.

Highlights

CircFOXP1 was downregulated, while EGR1 and Blimp-1 were upregulated in the nasal mucosa of AR patients.

UC-MSCs-derived exosomal circFOXP1 was internalized by CD4⁺ T cells and suppressed Th2 cell differentiation.

EGR1 was a transcriptional activator of Blimp-1.

CircFOXP1 triggered epigenetic repression of EGR1 via recruiting EZH2, leading to transcriptional repression of Blimp-1.

UC-MSCs-derived exosomal circFOXP1 protected against the impairments in AR mice, possibly via EZH2/EGR1/Blimp-1 axis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-026-03604-7.

Introduction

Allergic rhinitis (AR), a type I allergic disease, is triggered by immunoglobulin E (IgE) responses to inhaled allergens [1]. The global incidence and prevalence of AR are progressively increasing over recent decades [2, 3]. AR patients experience symptoms, such as nasal congestion, itching, nasal discharge and sneezing; the condition is also frequently associated with asthma [1]. AR impairs AR patients’ life quality, and represents a significant global health burden [1, 4]. However, due to an incomplete understanding of AR pathogenesis, current AR treatments are symptomatic relief, such as allergen avoidance, intranasal corticosteroids and allergen-specific immunotherapy [5, 6]. In-depth investigation of AR pathogenesis and the development of novel therapy with long-term efficacy are urgently required to improve the clinical outcomes in AR.

It is well-known that T helper type 2 (Th2) cells are activated upon allergen exposure, thus triggering allergic responses through the release of cytokines such as IL-4, IL-5 and IL-13 etc [6, 7]. Th1/Th2 imbalance is a known etiology of AR, and inhibition of Th2 differentiation is recognized as a promising therapeutic approach for AR [8]. Mesenchymal stem cells (MSCs) are pluripotent stem cells with self-renewal and differentiation potentials [9]. MSCs are well-established to regulate immune responses and inflammation by releasing biological active substances [10, 11]. In recent years, MSC therapy has been emerged as novel therapeutic strategy for AR [12, 13]. Previous study has reported that adipose tissue-derived stem cells suppress eosinophilic inflammation in the nasal mucosa through modulating a Th2 to Th1 shift [14]. Intravenous injection of bone marrow (BM)-MSCs markedly alleviates allergic symptoms, OVA-specific IgE production, eosinophilic inflammation and the secretion of Th2 signature cytokines IL-4, IL-5 and IL-13, but increases Th1 cytokine IFN-γ in AR mice [15]. More importantly, a recent study has demonstrated that UC-MSCs-derived extracellular vesicles (EVs) successfully re-establish Th1/Th2 balance in PBMCs from AR patients [16]. Despite these advances, the mechanism underlying UC-MSCs-derived exosomes-regulated Th1/Th2 balance in AR remains largely undefined.

Circular RNAs (circRNAs) are a class of continuous loop RNAs generated from the back-splicing of pre-mRNA [17]. circRNA forkhead box protein P1 (circFOXP1) is generated from FOXP1 gene on chromosome 3q13 [18]. Recent studies have highlighted the diagnostic and therapeutic potentials of circFOXP1 in various diseases, such as cancers, osteoporosis and pulmonary fibrosis [19–21]. Intriguingly, circFOXP1 has been identified as a crucial regulator in maintaining MSC identity and differentiation [22]. Our preliminary results revealed a downregulation of circFOXP1 in the nasal mucosa of AR patients, and UC-MSCs-derived exosomal circFOXP1 was internalized by CD4⁺ T cells. In addition, catRAPID predicted the association between circFOXP1 and EZH2, and AnimalTFDB also predicted that EZH2 may serve as a transcription factor of EGR1. Furthermore, bioinformatics analysis based on JASPAR predicted the putative binding sites between EGR1 and Blimp-1 promoter. Notably, Blimp-1 is upregulated in hypo-responsive Th2 cells, and it is also essential for Th2 cell development in the lung [23, 24]. It is of great clinical significance to delineate the underlying mechanism of UC-MSCs-derived exosomal circFOXP1 in AR.

The central scientific question driving this study was to decipher the mechanism by which UC-MSC exosomes regulate Th2 differentiation in AR. We hypothesized that UC-MSCs-derived exosomal circFOXP1 was transferred to CD4⁺ T cells. CircFOXP1 triggered epigenetic repression of EGR1 via recruiting EZH2, leading to Blimp-1 downregulation, ultimately inhibiting Th2 differentiation to alleviate AR. These findings are expected to delineate a novel circRNA-dependent mechanism of MSC action and offer fresh translational implications for exosome-based therapies in AR. More specifically, this study on circFOXP1 overexpression paves the way for its delivery via engineered exosomes as a future therapy for AR.

Materials and methods

Clinical specimens

Whole blood and nasal mucosal specimens were collected from AR patients (n = 23), disease controls (DCs) who were diagnosed with non-atopic obstructive snoring (n = 23) and healthy controls (HCs, n = 23) from May, 2023 to May, 2024 in The Second Xiangya Hospital, Central South University. AR patients were recruited to this study according to the inclusion criteria as previously described [25]. Inferior turbinate mucosal tissues from 23 AR patients who underwent septoplasty. The control tissues were harvested under endoscopic guidance with minimal mechanical trauma during the same period [26]. The Total Nasal Symptom Score (TNSS) was employed as a separate metric of AR severity [27]. Scores were assigned to each nasal symptom and subsequently summed to yield a total score for each AR case. A matched design was used for cohort recruitment. The DC and HC groups were specifically selected to match the AR group for age (± 5 years) and sex ratio. All written contents were obtained and this study was approved by The Second Xiangya Hospital, Central South University. It was performed in accordance with the Declaration of Helsinki.

Animal and ovalbumin (OVA)-induced AR mouse model

All animal studies were approved by The Second Xiangya Hospital, Central South University and were performed in accordance with the ARRIVE guidelines. Female BALB/c mice (6-week-old, body weight 18–20 g, n = 6 per group) were from Hunan SJA Laboratory Animal Co., LTD (Changsha, China), and housed under standard condition (22 ± 1 °C, humidity of 50 ± 5%) with a 12 h light/12 h dark cycle. The experiments were performed in a blinded manner. Mice were intraperitoneally injected with 50 µg OVA (Sigma-Aldrich, St. Louis, MO, USA) in PBS containing 1 mg of aluminum hydroxide on day 1, 7, 14 and 21. Mice were then nostril challenged with 400 µg OVA from day 21 to day 27 (every other day, total of 4 times). For the treatment of exosomes, mice were randomly divided into four groups (n = 6 per group) using a computer-generated random number sequence: Control, AR + PBS, AR + UC-MSC-Exo^NC and UC-MSC-Exo^circFOXP1. Exosomes (100 µg) were administered to the mice from day 21 to day 27 (every day) via tail vein injection [28]. The frequencies of rubbing and sneezing behavior were counted on day 27, and the mice were sacrificed at day 28 after anesthesia with alfaxan (Jurox Pty Ltd, Rutherford, NSW, Australia).

Evaluation of sneezing and rubbing

The nasal symptoms of mice were evaluated blindly. Following the last OVA treatment and a period of one-hour rest, the counts of sneezing and rubbing were recorded for 15 min.

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

Total RNA was extract from blood sample or nasal mucosa using Trizol Plus reagent (Invitrogen, Carlsbad, CA, USA), and reverse transcribed using PrimeScript First Strand Kit (TaKaRa, Dalian, China). qRT-PCR was conducted using SYBR Green Master Mix (ABI, Foster City, CA, USA), and the relative expression of target gene was calculated using 2^−ΔΔCT method. GAPDH or U6 was used as an internal control. Sequence information was listed in Table S1.

Hematoxylin and eosin (H&E) staining, sinus red staining and immunohistochemistry (IHC) analysis

The anterior portion of mouse head was dissected and fixed with 4% paraformaldehyde (PFA) and decalcified with 0.5 M EDTA for 10 days, followed by paraffin-embedding. The sections of nasal mucosa (4 μm-thickness) were stained with H&E solution (Beyotime, Shanghai, China). The H&E-stained nasal mucosa sections were analyzed as described with modifications [29]. The scoring system ranged from 0 (normal tissue) to 5 (severe, diffuse inflammation) based on the combined assessment of eosinophil infiltration, tissue edema, and epithelial damage. Eosinophils infiltration was monitored by Sirius Red staining. Briefly, the nasal mucosal sections were stained with Sirius Red solution (Sigma-Aldrich). In a blinded manner, eosinophils were counted in five randomly selected high-power fields (400× magnification) of the nasal septal mucosa by two independent observers using a light microscope (Nikon, Tokyo, Japan) [30]. For IHC analysis, the rehydrated sections were subjected to antigen retrieval. After blocking, the slides were then incubated with anti-EGR1 (1:500, ab300449, Abcam, Cambridge, UK) or anti-Blimp-1 antibody (1:100, ab241568, Abcam) at 4 ℃ overnight. This was followed by the incubation with HRP-conjugated secondary antibody, and signal was detected using DAB substrate (Beyotime). Images were acquired under a microscope (Nikon). The IHC results were quantified by calculating the percentage of positive-staining cells as previous described [30, 31].

RNA fluorescence in situ hybridization (FISH)/immunofluorescence (IF) staining

The sections of nasal mucosa or UC-MSCs were fixed with 4% PFA and prehybridized with 0.5% Triton X-100. UC-MSCs were then incubated with Cy3-labeled circFOXP1 probe in hybridization buffer at 37 ℃ overnight. For RNA FISH/IF staining, the slides were then stained with ant-EZH2 antibody (1:250, ab191080, Abcam) at 4 ℃ overnight, followed by the incubation of goat anti-rabbit antibody, Alexa Fluor 488 (1:2000, A11008, Invitrogen). Nuclei were visualized with DAPI (Invitrogen). Images were acquired using a confocal microscope (Nikon). The RNA FISH results were quantified by calculating the percentage of circFOXP1-postivie cells.

RNA pull-down assay

RNA pull-down assay was conducted using Pierce RNA Pull-Down Kit (20164, Pierce, Rockford, IL, USA). In brief, UC-MSCs were lysed using M-PER Protein Extraction Reagent (78501, Pierce). Biotinylated circFOXP1 or control probe (GenePharma, Suzhou, China) was conjugated to Streptavidin beads, followed by the incubation with cell lysates. The protein levels of hnRNPA2B1 and EZH2 in the elutes were detected by western blot.

RNA immunoprecipitation (RIP) assay

Magna RIP Kit (Millipore, Billerica, USA) was used to perform RIP assay. UC-MSCs were lysed with RIP Lysis Buffer. Anti-hnRNPA2B1 antibody (2 µg, ab259894, Abcam) or normal rabbit IgG (10500 C, Invitrogen) was conjugated to magnetic beads, and RNA-protein complexes were then enriched by these beads. RNA purification was conducted, followed by qRT-PCR analysis.

Western blot

Total protein lysates were prepared using RIPA lysis buffer (Beyotime), and quantified using BCA protein assay kit (Pierce). After gel electrophoresis, proteins were transferred onto PVDF membrane and blocked with 5% non-fat milk. The blots were then incubated with primary antibody at 4 ◦C overnight, followed by secondary antibody incubation. Signal was detected by BeyoECL Plus kit (Beyotime). Primary antibodies used in western blot were from Abcam: anti-hnRNPA2B1 (1:1000, ab259894), anti-TSG101 (1:1000, ab125011), anti-CD63 (1:1000, ab315108), anti-CD81 (1:1000, ab79559), anti-GRP94 (1:1000, ab3674), anti-GATA3 (1:1000, ab199428), anti-IL-4 (1:1000, ab62351), anti-EGR1 (1:1000, ab300449), anti-Blimp-1 (1:1000, ab241568), anti-EZH2 (1:500, ab191080), anti-GAPDH (1:2000, ab8245) antibodies.

Umbilical cord-derived mesenchymal stem cells (UC-MSCs) isolation, culture, treatment and transfection

Human umbilical cords were collected with the written consents of the donors in The Second Xiangya Hospital, Central South University. This study was approved by The Second Xiangya Hospital, Central South University and was performed in accordance with the Declaration of Helsinki. After removal of umbilical cord vessels, the cords were cut into 1 cm³-pieces. The tissues were rinsed with DMEM (Gibco, Grand Island, NY, USA) and digested with collagenase II (Gibco) at 37 ℃ for 1 h, followed by the incubation with trypsin (Gibco) at 37 ℃ for 30 min [32]. UC-MSCs were grown in F12/DMEM with 10% FBS (Gibco) at 37 ℃/5% CO₂. UC-MSCs were then characterized by flow cytometry usingFITC-labeled anti-CD90, CD73, CD105 and CD34 antibodies (Invitrogen) as described [25]. To detect the exosomal circFOXP1 level, UC-MSCs were treated with 1 µg/mL RNase A or combined with 0.1% Triton X-100 for 30 min. Scramble siRNA or si-hnRNPA2B1 were synthesized by the siRNA-mediated silencing system (JTSBIO Co., Ltd. Wuhan, China). The full-length of hnRNPA2B1, EGR1 or EZH2 was cloned into pcDNA3.1 vector (Invitrogen). UC-MSCs or CD4⁺ T cells were transfected with overexpression construct using Lipofectamine 3000 (Invitrogen). Cells were harvested for subsequent analysis at 48 h post-transfection. To block EZH2, CD4⁺ T cells were treated with EZH2 inhibitor GSK126 (5 µM, Sigma-Aldrich) for 24 h.

Lentiviral packaging and transduction

Lentiviral construct containing shRNA against circFOXP1 and scramble shRNA were from GenePharma (Shanghai, China). In brief, HEK293T cells were co-transfected with pLP-VSVG and psPAX2 using PEI40K (Polysciences, Warrington, PA, USA) for 48 h. The supernatant was collected and filtered. After centrifugation, the lentivirus particles were transduced into UC-MSCs. The transduced cells were selected by 2 µg/mL puromycin (Sigma-Aldrich).

Isolation of CD4⁺ T cells and Th2 differentiation

CD4⁺ T cells were isolated from human peripheral blood mononuclear cells (PBMCs) of AR patients using Dynabeads FlowComp Human CD4 Kit (11361D, Invitrogen). Briefly, PBMCs were incubated with antibody on ice for 10 min. After rinsing with isolation buffer, PBMCs were incubated with Dynabeads at room temperature for 15 min, and the enriched CD4⁺ T cells were then released from magnetic beads. For Th2 differentiation, CD4⁺ T cells were stimulated with anti-CD3 (3 µg/mL, 16-0037-81, Invitrogen), anti-CD28 (5 µg/mL, 16-0289-81, Invitrogen), anti-IFN-γ (20 µg/mL,16-7318-81, Invitrogen) antibody, IL-4 (20 ng/mL, 200-04-5UG, Lot# 030814ST, PeproTech, Rocky Hill, NJ, USA) and IL-2 (2 ng/mL, 200-02-10UG, Lot# 012412, PeproTech), and these cells were co-incubated with UC-MSCs-derived exosomes for 7 days.

Isolation and characterization of UC-MSCs-derived exosomes

Exosomes were isolated from the culture medium of UC-MSCs as described [33]. Briefly, cell culture medium was centrifuged at 3000 × g for 30 min, followed by a filtration with 0.22-µm filter (Millipore). Exosomes were then isolated using VEX exosome isolation reagent (Vazyme, Nanjing, China). Exosomes were characterized by transmission electron microscopy (TEM). In brief, exosomes were placed on a carbon-coated copper grid, followed by 2% phosphotungstic acid staining. Images were acquired using FEI Tecnai 12 TEM (Thermo Fisher Scientific, Waltham, MA, USA). Nanoparticle tracking analysis (NTA) was employed to assess the size distribution and particle concentration of exosomes by using Zetasizer (Malvern Panalytical, Malvern, UK).

Exosome uptake assay

PKH26 Green Fluorescent Cell Linker Kit (Sigma-Aldrich) was employed to monitor exosome uptake. In brief, CD4⁺ T cells were incubated with UC-MSCs-derived exosomes (50 µg) for 24 h. CD4⁺ T cells were then subjected to IF analysis. Images were acquired by a confocal microscope (Nikon).

ELISA assay

Human or mouse whole blood was collected and allowed to clot for 30 min. After centrifugation, serum was harvested for ELISA assay. Human serum IFN-γ (88-7316-22, Invitrogen), IL-4 (88-7046-22), IL-5 (88-7056-22) and IL-13 (88-7439-22) levels, as well as mouse serum IFN-γ (88-7314-22), IL-4 (88-7044-22) and IL-5 (88-7054-22) levels, were detected using commercial ELISA kits (Invitrogen) according to the manufacturer’s instructions. The mouse serum level of IgE was measured using mouse IgE ELISA kit (88-50460-22, Invitrogen), and mouse OVA specific IgE (sIgE) was assessed using OVA sIgE ELISA Kit (CSB-E08914m, CUSABIO, Wuhan, China). A450 was measured using a microplate reader (Thermo Fisher Scientific).

Flow cytometry

The percentages of Th1 or Th2 in human PBMCs were determined by flow cytometry. In brief, PBMCs were treated and analyzed using CellVivo Human Th1 or Th2 Differentiation Kit (R&D System, Minneapolis, MN, USA). For the analysis of CD4⁺ T cells, splenocytes were first gated on the lymphocyte population based on forward scatter (FSC) and side scatter (SSC). Single cells were selected, and CD3-APC⁺ T cells (17-0032-82, Invitrogen) were identified. Subsequently, the frequencies of IFN-γ-PE⁺ (61-7311-82, Invitrogen) and IL-4-PE⁺ (12-7041-82, Invitrogen) cells within the CD4-FITC⁺ T cell (11-0042-86) population were determined following intracellular staining. The stained cells were analyzed using BD FACSAria II Flow cytometer (BD Biosciences, St. Jose, CA, USA).

Subcellular fractionation

Subcellular fractionation was conducted using PARIS kit (AM1921, Invitrogen). Briefly, CD4⁺ T cells were resuspended in Cell Fractionation Buffer, and centrifugated at 500 × g for 5 min. The supernatant (cytoplasmic fraction) was collected, and the pellets (nuclear fraction) were then lysed with Cell Disruption Buffer. After RNA purification, the cytoplasmic and nuclear circFOXP1 levels were detected by qRT-PCR. U6 or GAPDH was used as a nuclear or cytoplasmic marker, respectively.

Chromatin immunoprecipitation (ChIP) assay

ChIP assay was conducted using Pierce Magnetic ChIP Kit (26157, Pierce). In brief, CD4⁺ T cells were crosslinked using 1% formaldehyde and lysed. The chromatin fractions were prepared by MNase digestion and incubated with anti-EZH2 (2 µg, ab191080, Abcam), anti-H3K27me3 (3 µg, MA5-11198, Invitrogen) antibody or normal IgG at 4 ◦C overnight. The DNA-protein complexes were enriched by Protein A/G magnetic beads, and DNA was purified and analyzed by qRT-PCR.

Dual luciferase reporter assay

The promoter region of Blimp-1 containing the putative EGR1 binding sites was cloned into pGL-3 vector (Promega, Madison, WI, USA). The luciferase construct and EGR1 overexpression construct were co-transfected into HEK293 cells using Lipofectamine 3000 (Invitrogen). At 48 h post-transfection, the luciferase activity was quantified using Dual Luciferase Reporter System (Promega). Renilla luciferase activity was employed as an internal control.

Statistical analysis

Data were presented as mean ± SD from at least 3 independent experiments. Statistical analyses were conducted using SPSS 22.0. The differences between two groups or among multiple groups were analyzed by Student’s t test or one-way ANOVA, respectively. P < 0.05 was considered to be statistically significant.

Results

Downregulation of circFOXP1 is correlated with Th1/Th2 dysregulation in AR patients

To study the biological role of circFOXP1 in AR, we examined the circFOXP1 level in the peripheral blood and nasal mucosa of AR patients by qRT-PCR. As presented in Fig. 1A-B, circFOXP1 was downregulated in the peripheral blood and nasal mucosa of both DCs and AR patients, and a more prominent decrease of circFOXP1 was observed in AR patients. The clinical characteristics of the participants in this study were shown in TableS2. The mean TNSS of all AR cases was 8.2 (Table S2). In addition, the analysis of receiver operating characteristic (ROC) curve revealed that circFOXP1 distinguished AR patients from DCs or HCs (Fig. 1B-C). Spearman’s correlation analysis showed that circFOXP1 was positively correlated to the percentage of Th1 cells and Th1/Th2 axis, but negatively correlated with Th2 cell proportion in AR (Fig. 1D-F). Furthermore, an remarkable positive correlation was found between circFOXP1 and the Th1 signature cytokine IFN-γ level, while a negative correlation between circFOXP1 and the Th2 signature cytokine IL-4 was observed in AR patients (Fig. 1G-H). There was a negative correlation between circFOXP1 and TNSS in AR patients (Fig. 1I). These findings indicate that circFOXP1 is downregulated in AR patients, and it is associated with Th1/Th2 dysregulation in AR.

Fig. 1.

Downregulation of circFOXP1 is correlated with Th1/Th2 dysregulation in AR patients. The peripheral blood and nasal mucosal samples were collected from healthy controls (HCs), disease controls (DCs) and AR patients (n=23). A The level of circFOXP1 in the peripheral blood and nasal mucosa were detected by qRT-PCR. B-C ROC curves of circFOXP1 in distinguishing AR patients from DCs or HCs. D-I The correlations of circFOXP1 with Th1 cells, Th2 cells, Th1/Th2 axis, IFN-γ and IL-4, TNSS in AR patients were evaluated by Spearman’s correlation analysis. *P<0.05, **P<0.01, ***P<0.001

The comparison of circFOXP1 distribution in nasal mucosa between HCs and AR patients

The histological changes in the nasal mucosa of AR patients were evaluated by H&E staining. As shown in Fig. 2A, thickened basement membrane, abnormal arrangement of epithelial cells and infiltration of inflammatory cells were observed in the nasal mucosa of AR patients, compared with HCs. Moreover, RNA FISH showed that the expression of circFOXP1 was remarkably decreased in nasal mucosa of AR patients, in comparison with that of HCs (Fig. 2B). These data suggest that circFOXP1 is decreased in the nasal mucosa of AR patients.

Fig. 2.

The comparison of circFOXP1 distribution in nasal mucosa between HCs and AR patients. A The histological changes of nasal mucosa (n=23) in HCs and AR patients were assessed by H&E staining. Scale bar, 100 μm for 100X and 25 μm for 400X. B The localization and expression of circFOXP1 in the nasal mucosa of HCs and AR patients were detected by RNA FISH with quantitative analysis. Scale bar, 25 μm for 400X. Red, circFOXP1; Blue, DAPI. Results were presented as mean ± SD from 3 independent experiments. A representative RNA FISH result from three independent experiments is shown. **P<0.01

hnRNPA2B1 mediates the exosomal delivery of circFOXP1 to CD4⁺ T Cells

Our preliminary data revealed that extracellular circFOXP1 could be incorporated into exosomes. As shown in Fig.S1A, RNase A alone was unable to be accessible in the exosomes in which no change of circFOXP1 was detected by qRT-PCR, while circFOXP1 was greatly decreased when UC-MSCs were treated with RNase A and Triton X-100, indicating that circFOXP1 was contained inside the exosomes. In addition, circFOXP1 was detected in the whole culture medium and exosomes derived from UC-MSCs (Fig.S1B). CD4⁺ T cells were then incubated with PBS or UC-MSCs-derived exosomes (UC-MSC-Exo). qRT-PCR showed that the level of circFOXP1 was remarkably elevated in UC-MSC-Exo-treated CD4⁺ T cells (Fig.S1C). Moreover, RNA FISH confirmed the induction of circFOXP1 in CD4⁺ T cells upon UC-MSC-Exo treatment (Fig.S1D). Previous studies have demonstrated that hnRNPA2B1 acts as a key player in sorting and recruiting RNAs into exosomes [34, 35]. We next investigated if hnRNPA2B1 was implicated in the recruitment of circFOXP1 into exosomes. RNA pull-down assay showed the direct interaction between circFOXP1 and hnRNPA2B1 in UC-MSCs (Fig. 3A). Additionally, this association was further validated by RIP assay in which antibody against hnRNPA2B1 successfully enriched circFOXP1 in UC-MSCs (Fig. 3B). Gain- and loss-of-function experiments were conducted. As anticipated, transfection of si-hnRNPA2B1 or hnRNPA2B1 overexpression plasmid markedly reduced or induced hnRNPA2B1 mRNA and protein level in UC-MSCs, respectively (Fig. 3C-D). qRT-PCR further revealed that the downregulation of hnRNPA2B1 was accompanied with the reduction of exosomal circFOXP1, and the upregulation of exosomal circFOXP1 was observed in hnRNPA2B1-overexpressing UC-MSCs (Fig. 3E). These findings suggest that hnRNPA2B1 regulates exosomal circFOXP1 packaging. We next sought to investigate if exosomal circFOXP1 was internalized by CD4⁺ T cells. As illustrated in Fig. 3F, UC-MSCs were negative for CD34 and positive for CD90/CD73/CD105. In addition, UC-MSCs exhibit trilineage differentiation potential into adipocytes, osteocytes, and chondrocytes. These implied the phenotypic characterization of UC-MSCs, which was prepared for the transfection with lentiviral scramble shRNA (KD-NC) or sh-circFOXP1 (KD-circFOXP1). UC-MSCs-derived exosomes were purified and characterized by TEM and NTA. As presented in Fig. 3G, exosomes exhibited double-membrane bound morphology as detected by TEM. NTA also showed the particle size range from 30 to 150 nm, and the particle concentration of UC-MSC-Exo^KD−NC and UC-MSC-Exo^{KD−circFOXP1} was approximately 8 × 10⁹/10⁸ cells (Fig. 3H). As expected, UC-MSCs-derived exosomes were positive for exosomal markers TSG101, CD63 and CD81 whereas these markers can’t be detected in cell lysates. As expected, the exosomes were negative for GRP94 (Fig. 3I). The successful knockdown of circFOXP1 in UC-MSCs and UC-MSCs-derived exosomes was confirmed by qRT-PCR (Fig. 3J). Exosome uptake assay revealed the intercellular trafficking of PKH26-labeled exosomes in CD4⁺ T cells (Fig. 3K). Moreover, the circFOXP1 level was dramatically increased in CD4⁺ T cells which were co-cultured with UC-MSCs-derived exosomes (UC-MSC-Exo^KD−NC). By contrast, co-culturing of exosomes derived from circ-FOXP1-knockdown UC-MSCs (UC-MSC-Exo^{KD−circFOXP1}) led to a reduction of circFOXP1 in CD4⁺ T cells, compared with corresponding control (Fig. 3L). Collectively, these data indicate the uptake of exosomal circFOXP1 by CD4⁺ T cells.

Fig. 3.

hnRNPA2B1 mediates the exosomal delivery of circFOXP1 to CD4⁺ T Cells. A The interaction between circFOXP1 and hnRNPA2B1 in UC-MSCs was assessed by RNA pull-down assay. B The association between circFOXP1 and hnRNPA2B1 in UC-MSCs was detected by RIP assay. Normal IgG served as a negative control. UC-MSCs were transfected with siRNA or overexpression construct. C-D The mRNA and protein levels of hnRNPA2B1 in UC-MSCs were detected by qRT-PCR and western blot. E The level of exosomal circFOXP1 was detected by qRT-PCR. UC-MSCs were transfected with lentiviral scramble shRNA (KD-NC) or sh-circFOXP1 (KD-circFOXP1). F Surface marker characterization and induced trilineage differentiation of UC-MSC were analyzed by flow cytometry and adipogenic (Oil Red O), osteogenic (Alizarin Red) and chondrogenic (Alcian Blue) staining. G The morphology of exosomes was examined by TEM. H Particle size distribution of exosomes was detected by NTA analysis. I The protein levels of TSG101, CD63, CD81 and GRP94 in exosomes and UC-MSCs were detected by western blot. J The level of circFOXP1 in UC-MSCs and exosomes were detected by qRT-PCR. K The uptake of PKH26-positive exosomes by CD4+ T cells was monitored by exosome uptake assay. Scale bar, 25 μm for 400X. Green, PKH26; Blue, DAPI. CD4+ T cells were co-cultured with PBS or UC-MSCs-derived exosomes. L The level of circFOXP1 in CD4+ T cells was detected by qRT-PCR. Results were presented as mean ± SD from 3 independent experiments. A representative western blot from three independent experiments is shown. *P<0.05, **P<0.01, ***P<0.001

Silencing of circFOXP1 counteracts UC-MSCs-derived exosomes-suppressed Th2 differentiation of CD4⁺T cells

To investigate the effects of UC-MSCs-derived exosomes on Th2 differentiation, CD4⁺ T cells were incubated with PBS or the designated UC-MSCs-derived exosomes. Flow cytometry revealed that UC-MSC-Exo^KD−NC decreased the percentage of Th2 cells, while UC-MSC-Exo^{KD−circFOXP1} caused a rebound of Th2 proportion (Fig. 4A). In line with this finding, qRT-PCR and ELISA assay unequivocally showed that Th2 cytokines IL-4, IL-5 and IL-13 expression and secretion were decreased by UC-MSC-Exo^KD−NC, whereas UC-MSC-Exo^{KD−circFOXP1} rescued the downregulation of these Th2 cytokines (Fig. 4B-C). In addition, western blot revealed that GATA3, IL-4, EGR1 and Blimp-1 were reduced in UC-MSC-Exo^KD−NC-treated CD4⁺ T cells. By contrast, circFOXP1-knockdown UC-MSCs-derived exosomes counteracted these effects in CD4⁺ T cells (Fig. 4D). These findings suggest that circFOXP1 acts as a key player in UC-MSCs-derived exosomes-suppressed Th2 differentiation of CD4⁺ T cells.

Fig. 4.

Silencing of circFOXP1 counteracts UC-MSCs-derived exosomes-suppressed Th2 differentiation of CD4+ T cells. A The percentage of Th2 cells were detected by flow cytometry. B The expression of IL-4, IL-5 and IL-13 in CD4+ T cells were detected by qRT-PCR. C The secretion of IL-4, IL-5 and IL-13 in CD4+ T cells were detected by ELISA assay. D The protein levels of GATA3, IL-4, EGR1 and Blimp-1 in CD4+ T cells were detected by western blot. Results were presented as mean ± SD from 3 independent experiments. A representative western blot from three independent experiments is shown. *P<0.05, **P<0.01, ***P<0.001

EGR1 and Blimp-1 are elevated in AR

We next tested the expression of EGR1 and Blimp-1 in clinical specimens. qRT-PCR showed that EGR1 and Blimp-1 were markedly elevated in the nasal mucosal samples of AR patients, compared with HCs and DCs (Fig. 5A-B). Moreover, ROC curve revealed that both EGR1 and Blimp-1 distinguished AR patients from DCs or HCs (Fig. 5C-D). Consistently, the immunoreactivities of EGR1 and Blimp-1 were also increased in the nasal mucosa of AR patients (Fig. 5E-F). These data indicate that EGR1 and Blimp-1 are upregulated in AR. Spearman’s correlation analysis further revealed that EGR1 or Blimp-1 was negatively correlated with circFOXP1 in AR patients, while no significant correlation between EGR1 Blimp-1 and circFOXP1 was found in DCs or HCs (Fig.S2A, 2F). In addition, a negative correlation between EGR1 or Blimp-1 and Th1 cells/IFN-γ was observed in AR patients, whereas EGR1 or Blimp-1 was positively correlated with Th2 cells/IL-4 in AR patients (Fig.S2B-E, 2G-J). These data indicate that EGR1 or Blimp-1 was associated with circFOXP1-mediated Th1/Th2 dysregulation.

Fig. 5.

EGR1 and Blimp-1 are elevated in AR. The nasal mucosal samples were collected from healthy controls (HCs), disease controls (DCs) and AR patients (n=23). A-B The mRNA levels of EGR1 and Blimp-1 in the nasal mucosa were detected by qRT-PCR. C-D ROC curves of EGR1 and Blimp-1 in distinguishing AR patients from DCs or HCs. E-F The immunoreactivities of EGR1 and Blimps in the nasal mucosal samples were detected by IHC analysis. Scale bar, 100 μm for 100X and 25 μm for 400X. *P<0.05, **P<0.01, ***P<0.001

EGR1 acts as a transcriptional activator of Blimp-1

JASPAR database predicted the putative binding sites of EGR1 on the promoter region of Blimp-1, including Binding Site (BS)1 p(-1079/-1066) and BS2 p(-1479/-1466) (Fig. 6A-B). ChIP assay showed that antibody against EGR1 successfully immunoprecipitated both BS1 and BS2 fragments in HEK293 cells (Fig. 6C). Moreover, overexpression of EGR1 induced the promoter activity of Blimp-1 as detected by luciferase reporter assay in HEK293 cells (Fig. 6D). In UC-MSCs, EGR1 overexpression upregulated EGR1 and Blimp-1 mRNA and protein levels (Fig. 6E-G). These data indicate that EGR1 positively regulates Blimp-1 expression at the transcriptional level.

Fig. 6.

EGR1 acts as a transcriptional activator of Blimp-1. A-B The putative binding sites of EGR1 on the promoter region of Blimp-1 were predicted by JASPAR database. C The interaction between EGR1 and Blimp-1 promoter was detected by ChIP assay. Normal IgG served as a negative control. D The relative luciferase activity was measured by luciferase reporter assay. Renilla luciferase activity was used as an internal control. UC-MSCs were transfected with EGR1 overexpression plasmid. E-F The mRNA levels of EGR1 and Blimp-1 were detected by qRT-PCR. G The protein levels of EGR1 and Blimp-1 were detected by western blot. Results were presented as mean ± SD from 3 independent experiments. A representative western blot from three independent experiments is shown. **P<0.01, ***P<0.001

circFOXP1 triggers epigenetic repression of EGR1 to transcriptionally suppress Blimp-1 through recruiting EZH2

We next sought to delineate the mechanism underlying circFOXP1 regulated EGR1/Blimp-1 axis. Subcellular fractionation showed that circFOXP1 was expressed in the nucleus and cytoplasm of CD4⁺ T cells (Fig. 7A). A direct association between circFOXP1 and EZH2 was also detected by RNA pull-down and RIP assays (Fig. 7B-C). Additionally, RNA FISH/IF staining also confirmed the co-localization of circFOXP1 and EZH2 in CD4⁺ T cells (Fig. 7D). ChIP assay further showed that knockdown of circFOXP1 resulted in a decreased enrichment of EZH2 in EGR1 promoter, in comparison with control group (Fig. 7E). This was accompanied with the reduced association between H3K27me3 and EGR1 promoter (Fig. 7F). Furthermore, qRT-PCR and western blot revealed that silencing of circFOXP1 downregulated EZH2, but upregulated EGR1 and Blimp-1. Overexpression of EZH2 reversed circFOXP1-mediated changes of EZH2, EGR1 and Blimp-1 in CD4⁺ T cells (Fig. 7G-H). As presented in Fig.S3A, overexpression of circFOXP1-decreased Th2 ratio was reversed by GSK126 as detected by flow cytometry. qRT-PCR and ELISA assay showed that IL-4, IL-5 and IL-13 expression and secretion were attenuated in circFOXP1-overexpressing CD4⁺ T cells, while the reduction of these cytokines were counteracted by GSK126 (Fig.S3B-C). Furthermore, circFOXP1-upregulated EZH2, as well as circFOXP1-downregulated EGR1 and Blimp1, were reversed in CD4⁺ T cells upon GSK126 treatment (Fig.S3D). Collectively, these findings suggest that circFOXP1 triggers epigenetic repression of EGR1 to transcriptionally suppress Blimp-1 through recruiting EZH2. We next investigated if EGR1 acted as a downstream molecule of circFOXP1 in the regulation of Th2 cell differentiation. As anticipated, transfection of EGR1 overexpression construct successfully upregulated EGR1 mRNA and protein levels in CD4⁺ T cells (Fig.S4A-B). Flow cytometry revealed that circFOXP1 overexpression decreased the proportion of Th2 cells, while overexpression of EGR1 attenuated circFOXP1-decreased Th2% (Fig.S4C). In line with these findings, circFOXP1-reduced IL-4, IL-5 and IL-13 were abolished by EGR1 overexpression as detected by qRT-PCR and ELISA assay (Fig.S4D-E). In addition, western blot also showed that circFOXP1 downregulated EGR1 and Blimp-1 expression, whereas EGR1 overexpression successfully compensated the loss of EGR1. This was accompanied with the induction of Blimp-1 in OE-circFOXP1 + EGR1 group, in comparison with corresponding control (Fig.S4F). These data suggest that EGR1 overexpression reverses UC-MSCs-derived exosomal circFOXP1-suppressed Th2 cell differentiation of CD4⁺ T cells.

Fig. 7.

circFOXP1 triggers epigenetic repression of EGR1 to transcriptionally suppress Blimp-1 through recruiting EZH2. A The nuclear and cytoplasmic expression of circFOXP1 in CD4⁺ T cells was detected by qRT-PCR. The association between circFOXP1 and EZH2 in CD4⁺ T cells was assessed by RNA pull-down B and RIP assays C. Scramble RNA was used as a control probe for RNA pull-down assay. Normal IgG served as a negative control for RIP assay. D The subcellular co-localization of circFOXP1 and EZH2 in CD4+ T cells was detected by RNA FISH/IF staining. Scale bar, 25 μm for 400X. Red, circFOXP1; Green, EZH2; Blue, DAPI. E-F The enrichments of EZH2 and H3K27me3 in the promoter region of EGR1 were detected by ChIP assay. Normal IgG served as a negative control. G The mRNA levels of EGR1 and Blimp-1 in transfected CD4+ T cells were detected by qRT-PCR. H The protein levels of EZH2, EGR1 and Blimp-1 in transfected CD4+ T cells were detected by western blot. Results were presented as mean ± SD from 3 independent experiments. A representative western blot from three independent experiments is shown. *P<0.05, **P<0.01, ***P<0.001

Blimp-1 is a critical downstream effector of circFOXP1

To validate the role of Blimp-1, CD4⁺ T cells were transfected with circFOXP1 or/and BlimP1 overexpression construct in the presence of UC-MSC-Exo^KD-circFOXP1. Compared with circFOXP1-overexpressing CD4⁺ T cells, the Th2-differentiation phenotype was restored by Blimp1 overexpression (Fig. 8A). In line with this finding, overexpression of Blimp1 cause a rebound in the expression and secretion of IL-4, IL-5 and IL-13 in circFOXP1-overexpressing CD4⁺ T cells upon UC-MSC-Exo^KD-circFOXP1 treatment (Fig. 8B-C). Moreover, circFOXP1-downregulated GATA3, IL-4 and Blimp1 were counteracted by Blimp1 overexpression in UC-MSC-Exo^KD-circFOXP1-treated CD4⁺ T cells (Fig. 8D). These findings indicate that Blimp-1 serves as a critical downstream effector mediating circFOXP1’s function.

Fig. 8.

Blimp-1 is a critical downstream effector of circFOXP1. CD4+ T cells were transfected with vector, circFOXP1 or/and Blimp1 overexpression construct, and treated with UC-MSC-ExoKD-circFOXP1. A The percentage of Th2 cells were detected by flow cytometry with quantitative analysis. B The expression of IL-4, IL-5 and IL-13 in CD4+ T cells were detected by qRT-PCR. C The secretion of IL-4, IL-5 and IL-13 in CD4+ T cells were detected by ELISA assay. D The protein levels of GATA3, IL-4 and Blimp-1 in CD4+ T cells were detected by western blot. Results were presented as mean ± SD from 3 independent experiments. A representative western blot from three independent experiments is shown. *P<0.05, **P<0.01, ***P<0.001

UC-MSCs-derived exosomes protect against the impairments in AR mice

To validate the in vitro findings, AR mice were treated with UC-MSCs-derived exosomes. Mice were randomly divided into four groups: Control, AR + PBS, AR + UC-MSC-Exo^NC and UC-MSC-Exo^circFOXP1. As presented in Fig. 9A-B, the counts of rubbing and sneezing were greatly increased in AR group, while UC-MSC-Exo^NC slightly alleviated these nasal symptoms and UC-MSC-Exo^circFOXP1 exerted a more prominent rescue effect. Consistent with these results, the serum levels of IgE and sIgE were elevated in AR mice, whereas UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 reversed the induction of serum IgE and sIgE in AR mice (Fig. 9C-D). Moreover, AR mice also exhibited decreased IFN-γ and increased IL-4 level which were attenuated by UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 (Fig. 9E-F). H&E and Sirius staining unequivocally showed increased eosinophils infiltration in the nasal mucosa of AR mice, while UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 rescued this impairment. It was worth noting that the rescue effect of UC-MSC-Exo^circFOXP1 was more prominent (Fig. 9G). As anticipated, IHC analysis revealed that the elevations of EGR1 and Blimp-1 in AR mice were abrogated by UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 (Fig. 9H). More importantly, RNA FISH and qRT-PCR revealed that UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 led to a rebound of circFOXP1 in the nasal mucosa of AR mice (Fig. 9I-J). Western blot further showed that EZH2 was downregulated, but GATA3, EGR1 and Blimp-1 were upregulated in AR mice, whereas UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 reversed AR-mediated changes of these proteins. The effects of UC-MSC-Exo^circFOXP1 were more significant (Fig. 9K). Flow cytometry also illustrated that the Th1/Th2 ratio in the spleen were rescued by UC-MSC-Exo^NC, and the protective effect was further potentiated by UC-MSC-Exo^circFOXP1 (Fig. 9L). Together, these findings indicate that UC-MSCs-derived exosomes rescued the pathological impairments in AR, and circFOXP1-overexpressing UC-MSCs-derived exosomes exerted more remarkable protective effects.

Fig. 9.

UC-MSCs-derived exosomes protect against the impairments in AR mice. Mice were randomly divided into four groups (n=6 per group): Control, AR+PBS, AR+UC-MSC-ExoNC and UC-MSC-ExocircFOXP1. A-B The counts of rubbing and sneezing were evaluated. C-F The serum levels of IgE, sIgE, IFN-γ and IL-4 were measured by ELISA assay. G Eosinophils infiltration was monitored by H&E and Sirius Red staining with quantitative analysis. Scale bar, 25 μm for 400X. H The immunoreactivities of EGR1 and Blimp-1 in the nasal mucosa were assessed by IHC staining with quantitative analysis. Scale bar, 25 μm for 400X. I The localization and expression of circFOXP1 in the nasal mucosa were detected by RNA FISH. Scale bar, 25 μm for 400X. Red, circFOXP1; Blue, DAPI. J The level of circFOXP1 in the nasal mucosa was detected by qRT-PCR. K The protein levels of EZH2, GATA3, EGR1 and Blimp-1 in the nasal mucosa were detected by western blot. L Flow cytometry gating strategy for quantifying spleen-derived CD4⁺ T cells expressing IFN-γ and IL-4. *P<0.05, **P<0.01, ***P<0.001

Discussion

AR affects more than 400 million people globally [36], and the development of a long-term effective therapy remains a worldwide challenge. In the present study, we reported that UC-MSCs-derived exosomal circFOXP1 was internalized by CD4⁺ T cells, where it bound to EZH2 to mediate epigenetic repression of EGR1. Decreased EGR1 further downregulated Blimp-1 expression, thus re-establishing Th1/Th2 balance to ameliorate AR. These findings identified a novel circFOXP1/EZH2/EGR1/Blimp-1 axis in AR, and shed light on the MSC therapy for AR. Enforced circFOXP1 expression in UC-MSCs might enhance the clinical efficacy of MSC therapy for AR.

MSCs can be easily isolated from various sources, including UC, BM, adipose tissue and placenta [37]. Compelling evidence indicates that MSCs exert immunomodulatory effects primarily through the release of cytokines, chemokines, growth factors, metabolites and EVs, particularly exosomes [38, 39]. Recent studies have reported the vital role of exosomes in AR occurrence and progression [40]. For instance, exosomes derived from MSCs suppresses Th2 cell differentiation in AR via miR-146-5p/SERPINB2 signaling [41]. Similarly, our data revealed that UC-MSCs exosomes were internalized by CD4⁺ T cells and inhibited Th2 cell differentiation, while this suppressive effect was attenuated in UC-MSC-Exo^{KD−circFOXP1}-treated CD4⁺ T cells. By contrast, the in vivo data showed that UC-MSC-Exo^circFOXP1 potentiated UC-MSC-Exo^NC-mediated protection in AR mice. These findings suggest the crucial role of exosomal circFOXP1 in the regulation of Th2 cell differentiation in AR in vitro and in vivo.

Exosomes are a subtype of EVs which contain various bioactive molecules, such as proteins, DNA, lipids, microRNA (miRNA), long non-coding RNA (lncRNA) and circRNA [42]. More importantly, ncRNAs have been emerged as promising biomarkers and therapeutic targets in allergic disorders in recent years [43]. In AR, circ_0067835 modulates allergic inflammation through miR-155/GATA3 axis [44]. Additionally, AR epithelium-derived exosomal lncGAS5 results in Th1/Th2 imbalance in AR via suppressing EZH2 and T-bet [45]. In this study, we reported that circFOXP1 was downregulated in AR, and it was correlated with Th1/Th2 dysregulation. Moreover, circFOXP1 was packaged into exosomes in a hnRNPA2B1-dependent manner, and exhibited its effects on Th2 differentiation through exosome-mediated intercellular communication. Notably, both our study and the recent report identify exosomal ncRNAs as key regulators of Th1/Th2 balance in AR [45], there are fundamental distinctions. Firstly, our work focused on therapeutic, UC-MSC-derived exosomes that alleviated AR, whereas the cited study described AR patient nasal mucus-derived exosomes that exacerbated disease. Secondly, we demonstrated that circFOXP1 recruited EZH2 to epigenetically suppress the EGR1/Blimp-1 axis and inhibited Th2 differentiation. In contrast, lncGAS5 in pathogenic exosomes downregulates EZH2 and T-bet, impairing Th1 differentiation. This comparison highlights that exosomal regulation in AR is dual-faceted. BM-derived eosinophils marginate to peripheral blood and nasal mucosa in AR, thereby leading to eosinophilic inflammation [46]. H&E and Sirius Red staining revealed that UC-MSC-Exo^NC or UC-MSC-Exo^circFOXP1 rescued eosinophils infiltration in nasal mucosa of AR mice, suggesting that UC-MSCs-derived exosomes protected against eosinophilic inflammation in vivo. It is noteworthy that overexpression of circFOXP1 potentiated the rescue effects of UC-MSCs-derived exosomes in AR mice. A recent study has demonstrated that UC-MSCs alleviate pulmonary fibrosis via facilitating circFOXP1-mediated autophagy by HuR/EZH2/STAT1/FOXK1 axis [21]. Our findings showed a novel protective role of UC-MSCs-derived exosomal circFOXP1 in AR. Despite its emerging roles in cancers respiratory, skeletal and cardiovascular disorders [18], the function of circFOXP1 in allergic diseases beyond AR requires further exploration.

A recent study has reported that UC-MSCs protects against pulmonary fibrosis via modulating circFOXP1-mediated autophagy. Specifically, downregulation of circFOXP1 blocks HuR translocation and facilitates its degradation, thereby suppressing autophagy negative regulator EZH2 [21]. In the present study, a direct association between circFOXP1 and EZH2 was identified, indicating that circFOXP1 not only indirectly regulates EZH2 through HuR, but also directly interacts with EZH2 to trigger downstream signaling. It is reported that EZH2 is implicated in AR development by regulating Th2 differentiation [45, 47]. In addition, a number of studies have demonstrated the critical role of EZH2-dependent H3K27me3 in cancer, and EZH2 suppresses the expression of target proteins by modulating H3K27me3 [48, 49]. EZH2-mediated epigenetic silencing of EGR1 was observed in CD4⁺ T cells. Based on reported presence of H3K27me3 at the EGR1 promoter and our functional data, we focused on this regulatory pathway [50]. Mechanistically, loss of circFOXP1 reduced the enrichments of EZH2 and H3K27me3 at EGR1 promoter region, suggesting that EZH2-mediated H3K27me3 is required for circFOXP1-induced epigenetic repression of EGR1. However, further validation is required to confirm the direct role of EZH2-mediated H3K27me3 in the EGR1 promoter region. Moreover, the UCSC Genome Browser predicted that a landscape of putative histone modifications (e.g., H3K4me1, H3K4me3, H3K27ac) was also present at the EGR1 promoter region (chr5:138,463,479 − 138,465,578). Our future studies will aim to profile these histone modifications, elucidate their interplay with EZH2 and other relevant enzymes, and validate their respective contributions to the development of AR. In short, EZH2 was identified as a downstream effector of circFOXP1-regulated Th2 differentiation in AR in the present study.

EGR1 is a zinc finger transcription factor involved in regulating cell proliferation, differentiation or cell death through the modulation of target gene transcription [51]. Notably, EGR1 is predominantly expressed in Th2 cells, where it regulates the acute phase of IL-4 transcription upon T cell receptor stimulation [52]. Moreover, EGR1 is a target of miR-183 and miR-96 which is involved in T cell activation [53]. In the current study, EGR1 was identified as a transcriptional activator of Blimp-1. Blimp-1 has been recognized as a regulator in cytotoxic CD4⁺ T cell differentiation, and hypo-responsive Th2 cells elevates Blimp-1 [23, 54, 55], suggesting the significant role of Blimp-1 in T cells, especially in Th2 cells. We demonstrated that Blimp-1 acted as a key player in the regulation of Th2 differentiation in AR. GATA3 coordinates Th1 and Th2 cell differentiation via lineage-specific targeting of distal regulatory elements [56]. More importantly, previous study has demonstrated that Blimp-1-supressed Bcl-6 results in GATA3 repression, ultimately facilitating Th2 cell differentiation in response to allergens [24]. The current study reported that GATA3 potentially emerged as a key downstream target, whose expression in AR mice was upregulated but can be restored by UC-MSCs-derived exosomes. However, the detailed mechanism underlying Blimp-1-modulated Th1/Th2 balance still needs in-depth investigation in the future study.

In conclusion, these findings illustrated that UC-MSCs-derived exosomal circFOXP1 was internalized by CD4⁺ T cells, and it induced epigenetic silencing of EGR1 via recruiting EZH2, leading to the downregulation of Blimp-1. Dysregulated Blimp-1 further inhibited Th2 differentiation to mitigate AR. circFOXP1/EZH2/EGR1/Blimp-1 regulatory axis was identified in this study, and UC-MSCs-derived exosomal circFOXP1 exhibits promising therapeutic efficacy in AR. In addition, a notable limitation of this study is the translational relevance of the acute OVA-induced model, which does not fully replicate the chronic nature of most human AR cases. The therapeutic potential of this molecular mechanism warrants further validation in larger clinical cohorts. A remaining limitation is the absence of a direct genetic rescue in UC-MSCs; conducting this experiment will be an important objective for future work to conclusively validate the functional axis.

Abbreviations

AR

Allergic rhinitis

Blimp-1

B-lymphocyte induced maturation protein 1

ChIP

Chromatin immunoprecipitation

circRNAs

Circular RNAs

DCs

Disease controls

EVs

Extracellular vesicles

FISH

Fluorescence in situ hybridization

FOXP1

Forkhead box protein P1

FSC

Forward scatter

HCs

Healthy controls

H&E

Hematoxylin and eosin

IF

Immunofluorescence

IgE

Immunoglobulin E

IHC

Immunohistochemistry

KD

Knockdown

lncRNA

Long non-coding RNA

miRNA

MicroRNA

NTA

Nanoparticle tracking analysis

OVA

Ovalbumin

PBMCs

Peripheral blood mononuclear cells

RIP

RNA immunoprecipitation

ROC

Receiver operating characteristic

shRNA

Short hairpin RNA

SSC

Side scatter

TEM

Transmission electron microscopy

Th2

T helper type 2

TNSS

Total Nasal Symptom Score

UC-MSCs

Umbilical cord-derived mesenchymal stem cells

Authors’ contributions

Ping Li: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Writing - Review & Editing. Benquan Yu: Investigation, Data Curation, Writing - Original Draft, Writing - Review & Editing, Visualization. Junjie Zeng: Formal analysis, Resources, Writing - Review & Editing, Supervision, Funding acquisition.

Funding

The project was supported by the Natural Science Foundation of Hunan Province, China (Grant No.2024JJ6588).

Data availability

The data that supports the findings of this study are available in the supplementary material of this article.

Declarations

Ethics approval and consent to participate

The studies involving human subjects were approved by The Second Xiangya Hospital, Central South University and was performed in accordance with the Declaration of Helsinki. All participants provided written informed consent. All animal studies were approved by The Second Xiangya Hospital, Central South University and were performed in accordance with the ARRIVE guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.