MSC-small extracellular vesicles alleviated Th2-airway inflammation by regulating the metabolism of DCs in mice

Abstract

Background

Allergic airway inflammation is one of the chronic inflammatory diseases and is generally dominated by T helper 2 cells (Th2). Dendritic cells (DCs) are essential to mounting the Th2-mediated airway inflammation by presenting inhaled antigens to prime CD4⁺ T cells. Small extracellular vesicles (sEV) derived from mesenchymal stem cells (MSCs) exhibited great interest in intractable diseases. However, the effects and mechanisms of MSC-sEV on DCs in airway inflammation is still unclear.

Methods

We isolated MSC-sEV using anion-exchange chromatography. Mouse bone marrow-derived DCs (BMDCs) and human monocyte-derived DCs (moDCs) were used to study the effects of MSC-sEV on dendritic cell surface molecules and their cytokine release. Mice were sensitized with house dust mites (HDM) to induce airway inflammation, and treated with MSC-sEV, the effects of sEV on murine DCs were identified. Extracellular flux analysis techniques were used to study the effects of MSC-sEV on the metabolic state of dendritic cells. RNA sequencing to study altered gene expression in BMDCs after MSC-sEV treatment.

Results

MSC-sEV mitigated the accumulation of Th2-associated cDC2s and moDCs in mouse lung in response to HDM. In vitro, MSC-sEV treatment significantly attenuated the activation of BMDCs. Furthermore, we identified that DCs were able to take MSC-sEV in vitro and in vivo. Mechanistically, MSC-sEV exerted regulatory effects on the metabolic pathways of murine DCs, specifically enhancing the reliance on oxidative phosphorylation of BMDCs. Importantly, MSC-sEV displayed similar effects on human moDCs.

Conclusions

MSC-sEV are able to alter the metabolic state of DCs, favoring DCs to maintain OXPHOS (oxidative phosphorylation) rather than glycolysis, thereby reducing DCs-initiated inflammatory responses and attenuating Th2 lung inflammation, suggesting MSC-sEV can be a potential clinical therapy for airway inflammation.

Graphical abstract

Graphical Abstract.

Significance statement.

This study explores the mechanism by which mesenchymal stem cell-derived small extracellular vesicles (MSC-sEV) alleviate airway inflammation by influencing dendritic cells (DCs). The Seahorse assay was used to evaluate the impact of MSC-sEV on the metabolism of DCs. Our data indicate that MSC-sEV can alter the metabolic state of DCs, prompting DCs to maintain oxidative phosphorylation, thereby reducing the inflammatory response initiated by DCs and alleviating Th2-type pulmonary inflammation. This suggests that MSC-sEV has the potential to be a potential clinical treatment for airway inflammation.

Introduction

Allergic airway inflammation is one of the most common chronic respiratory diseases globally. Allergic airway inflammation exhibits various degrees of airway inflammation, hyperresponsiveness, airway remodelling, and mucus production depending on its phenotype, and these pathological changes are mainly mediated by a number of airway cells and cells involved in the immune response.¹ Current treatments for allergic airway inflammation mainly use steroids, surgery, and immunotherapy. However, more effective treatments need to be investigated due to their treatment side effects.²

Generally, allergic airway inflammation is considered as a T helper 2 (Th2) cell-driven pathology in response to inhalation of aeroallergen. Dendritic cells (DCs) are antigen-presenting cells and play important roles in allergic airway inflammation by taking allergens to mediastinal draining lymph nodes (mLNs) and priming T cell immune response.^3–5

In recent studies, lung DCs were divided into conventional DCs (cDCs) and plasmacytoid DCs (pDCs). According to their functions and distinct surface markers, cDCs are divided into CD103⁺ DCs (cDC1s) and CD11b⁺ DCs (cDC2s).^6,7 Particularly, CD103⁺ DCs are able to present exogenous antigen cross-priming CD8⁺ T cells,^8,9 while CD11b⁺ DCs are specialized in presenting antigens to stimulate CD4⁺ T cells.¹⁰ In response to antigens in airway inflammation, lung DCs can migrate into mLNs to prime Th2-mediated immunity in allergic airway inflammation.^11,12 Therefore, regulating the activation and function of DCs may be a therapeutic mechanism for allergic airway inflammation.

Mesenchymal stem/stromal cells (MSCs) have been applied in many diseases as their significant immunomodulatory property, and their use in allergic airway inflammation is extensively studied.^13,14 Moreover, researchers have turned their attention to MSC derived small extracellular vesicles (sEVs) in recent years as they inherit the functions of MSCs and exhibit benefits of low immunogenicity as well as easier storage and application. Several studies have identified the immunoregulatory effects of MSC-sEVs on various immune cells involved in airway inflammation, including DCs, macrophages and type 2 innate lymphoid cells (ILC2).^15,16 Our previous study found that MSC-sEV-treated human dendritic cells exhibited impaired ability to prime Th2 and ILC2 responses in allergic rhinitis,¹⁷ but the mechanisms of how MSC-sEV acts on DCs and Th2-mediated airway inflammation are still lacking.

Here, we investigated the role of DCs in the process of MSC-sEV suppressing allergic airway inflammation and the underlying mechanisms. We found that MSC-sEVs attenuated the HDM-induced airway inflammation by regulating the activation and metabolism of DCs.

Methods

Generation of MSC-sEV

The MSCs derived from induced pluripotent stem cells (iPSCs) used in this study were cultured and identified according to the methods reported in our previous reports.¹⁸ In short, frozen iPSC-MSCs were resuscitated in two 15 cm dishes and cultured in complete culture medium with fetal serum for 3-4 days with a cell density of 70-80%. Then, the cells were collected and planted in twenty 15 cm petri dishes and cultured for another 3-4 days. When the cell density reached 70-80%, the chemically defined and protein-free medium was replaced, discarded after 6 h of incubation and replaced with fresh above medium, and the culture was continued for 42 h.^17,19 MSC-sEVs were then isolated from exosome-rich culture medium supernatant by anion exchange chromatography as our previous study.²⁰ The concentration and particle size of MSC-sEV were determined by Bradford Protein Assay and Nanoparticle Tracking Analysis (NTA), and stored at −80° for future use.

Mice

Mice C57BL/6 were purchased from GemPharmatech company. All mice were bred and maintained in specific pathogen-free (SPF) facilities. Here, 7-12 weeks of age and sex-matched mice were used according to the guidelines approved by the Animal Ethics Committee of Sun Yat-sen University. The method of execution for experimental animals is anaesthesia followed by neck breaking. Anaesthesia was administered by direct inhalation of isoflurane with an anaesthetic pump at a concentration dose of 1% and a gas flow rate of 300 mL/min. All animal experiments follow the National Research Council’s Guide for the Care and Use of Laboratory Animals.

HDM-induced allergic airway inflammation model

The HDM-induced allergic airway inflammation model was introduced as previously.¹¹ Mice were sensitized with intratracheal (i.t.) administration of 100 μg HDM (Dermatophagoidespteronyssinus extracts, Greer Laboratories) on day 0 and were subsequently challenged with 10 μg HDM i.t. on day 7-11. Mice were treated with MSC-sEV (1.5 × 10¹⁰) on day 6, 7, 8, and 10 by i.v. The mice were sacrificed on day 12 and bronchoalveolar lavage fluid (BALF), lung and mediastinal lymph nodes (mLNs) were collected. DCs frequency in the lung were evaluated using flow cytometry.

Lung tissue samples were stained with hematoxylin and eosin (H&E). Histopathological assessment included the evaluation of inflammatory cell infiltration and alveolar wall thickness. Inflammation was scored based on the extent of inflammatory cell infiltration (Supplementary Table S1—see online supplementary material). Alveolar wall thickness was measured by examining three to six morphologically regular, nonoverlapping, and nondeformed alveoli per high-power field in each sample. Specifically, the midline of each alveolar wall was measured using the software’s line tool. The arithmetic mean of all measurements from a single sample was calculated and recorded as its alveolar wall thickness.

Preparation of human and murine DCs

As described previously, for BMDCs, 3 × 10⁶ bone marrow cells per well were cultured in 6 well plates in 3 mL complete medium (RPMI 1640 supplemented with 10% Fetal Bovine Serum [FBS], glutamine, penicillin, 2-mercaptoethanol, streptomycin and HEPES buffer; ThermoFisher) added with 20 ng/mL GM-CSF and 5 ng/mL IL-4 (Peprotech).⁶ For human monocyte-derived DCs,¹⁷ CD14⁺monocytes from human peripheral blood mononuclear cells (PBMCs) from human blood buffy coats provided by Guangzhou Blood Center were positively isolated using MACS microbeads (Miltenyi Biotec, Germany). The study about using human buffy coat was approved by the Ethics Committee of The First Affiliated Hospital, Sun Yat-sen University, China, and exemption of written informed consent was approved. 1 × 10⁶ cells were cultured in 1 mL complete medium supplemented with 100 ng/mL GM-CSF and 40 ng/mL IL-4. For murine and human DCs activation, 1 μg/mL lipopolysaccharide (LPS; Sigma, USA) was used on day 5 for 2 h, and then, BFA (Brefeldin A) was added for 8-12 h. DCs were finally collected, identified by flow cytometry using antibodies against DCs-specific markers, and then, utilized in subsequent experiments.

Extracellular flux analysis

Mito stress test (Mito stress Test kit, Agilent) and glycolytic rate assay kit (Seahorse XF Glycolytic Rate Assay, Agilent) were performed as manufacturers’ instructions. Briefly, 8 × 10⁴ DCs per well were plated in seahorse XF96 tissue culture microplates coated with Cell-Tak (Corning, USA) in DMEM medium supplemented with 2 mM glutamine, 1 mM pyruvate, and 10 mM glucose at PH 7.4. After placing for 30-60 min at 37 °C without CO₂, the extracellular flux of mito stress test was performed with 1.5 μM Oligomycin, 1.0 μM FCCP, 0.5 μM ROT/AA. The extracellular flux of the glycolytic rate assay was measured in response to 0.5 μM ROT/AA and 50 mM 2-DG using Agilent Seahorse XFe 96 Analyzer.

sEV uptake assay

mCherry-labeled MSC-sEVs were prepared as described in our previous research.²⁰ Briefly, iPSC-MSCs were transfected with lentivirus vectors carrying the mCherry-CD63 fusion gene, and then, MSC-sEVs with mCherry expression were isolated and identified.²⁰ For sEV uptake in vitro assay, 10 μg/mL or 20 μg/mL mCherry-labeled sEV was added and cultured at 37 °C and 5% CO₂ for 12 or 24 h on day 5 of BMDCs culture. Then, mCherry MFI was detected by flow cytometry.

For another experiments, after co-culturing with mCherry-labeled sEV for 4 or 12 h, 1 × 10⁵ BMDCs were placed on Poly-L-Lysine treated glass coverslips. After 30 min, cells were washed with PBS and were fixed with 4% paraformaldehyde for 60 min. 0.2% Trizon X-100 was used to permeabilize cells for 30 min. After blocking with 5% goat serum, cells were mounted using 2.5 μg/mL DAPI for 5 min. Finally, fluorescence was visualized with Nikon C2 confocal microscope. The images were processed by Imaris 8.4 software and ImageJ was used to calculate mean fluorescence intensity (MFI).

For sEV uptake in vivo assay, mice were injected intravenously with 100 μg sEV at day-1 and sensitized with 100 μg HDM i.t. after sEV injection 2 h on day 0. To maintain sEV concentration in airway inflammation, mCherry-labeled sEV were also injected subsequently on day 1 and 2. On day 3, mice were sacrificed and MFI of lung sEV was obtained and analyzed by flow cytometry.

Flow cytometry

Flow cytometry was performed using Beckman Coulter Cytoflex, and FlowJo 10.4 software (Tree star) was used to analyze. Staining was performed at 4 °C in FACS buffer (PBS, 0.5%BSA, 2 mM EDTA, 0.1% Sodium Azide) for 20 min. All antibodies purchased from Invitrogen and were as followed: anti-mouse CD11c (N418), MHC II (M5/114.15.2), CD80 (16-10A1), CD86 (GL1), CD40 (IC40), TNFα (MP6-XT22), IL-12 (C17.8), FcεR1 alpha (MAR-1), CD64 (X54-5/7.1), siglecF (1RNM44N), ly6G (RB6-8C5), CD11b (M1/70), CD103 (2E7), CCR6 (29-2L17), CCR2 (SA203G11), CCR7 (4B12); anti-human CD11c (3.9), HLA-DR (G46-6, BD), CD80 (2D10.4), CD86 (IT2.2), CD40 (5C3), TNFα (MAb11), IL-12 (C8.6); ROS (abcam, ab113851), Mitotracker Green (Invitrogen, M7514).

RNA extraction and qPCR

Total RNA was extracted from BMDCs using RNAiso Plus reagent (Takara, Japan) according to the manufactures’ protocol. Complementary DNA (cDNA) was obtained with PrimeScript RT Master Mix (Takara, Japan). Quantitative PCR analysis was performed with SYBR Green Master (Roche, Switzerland) on ABI 7500 Real-Time PCR system (Applied Biosystems, USA). Final data were calculated with 2^−ΔΔCt method and levels of mRNA were normalized to actin. Forward primers of iNOS and actin were, CACCAAGCTGAACTTGAGCG, ATGCCGGAGCCGTTGTC, respectively. Reverse primers of iNOS and actin were CGTGGCTTTGGGCTCCTC, GCGAGCACAGCTTCTTTG, respectively.

Western blot analysis

The protein levels of sEV and cell lysates were quantified using the Pierce™ BCA protein assay (Beyotime, China). Firstly, 50 µg of protein was denatured in SDS loading buffer at 95 °C for 5 min. Then, 10 µg of the samples were loaded, separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes (Merck millipore, Germany). The membranes were blocked with 5% skimmed milk at RT for 1 h and further incubated with rabbit antibodies against CD9 (1:2000), CD63 (1:2000), CD81 (1:2000), Alix (1:5000), TSG101 (1:1000), and Calnexin (1:1000) at 4 °C overnight. After three washes, the membranes were incubated with horseradish peroxidase-conjugated anti-rabbit antibodies for 1 h at RT. The membranes were washed, incubated with ECL Western Blotting Substrate (Merck millipore, Germany), and photographed using an imaging analysis system (ChemiDoc XRS⁺ System, bio-rad, USA). Primary antibodies were purchased from Abcam, Cambridge, UK and the secondary antibody was purchased from CoWin Biosciences, China.

Nanoparticle tracking analysis

The particle size of sEVs was determined using the NanoSight NS300 (Malvern Panalytical, Netherlands). The sEVs were diluted 50-400 times with sterile PBS and carefully loaded into the sample cell. Once the measurement concluded, particle size and number were obtained.

RNA-seq sequencing

BMDCs samples cultured in vitro (in either steady state or LPS-activated state) and treated with sEV or PBS were collected. RNA sequencing was performed using the DNBSEQ platform. After sequencing, raw data were processed by filtering out low-quality reads, adapter contaminants, and reads with an excessive content of unknown bases (N). The resulting clean reads were aligned to the reference genome sequence GRCm38 to obtain the raw count matrix for subsequent analysis. Differential expression analysis: The raw count matrix was processed using DESeq2 (v1.42.1) for size factor normalization, variance stabilization, and negative binomial model fitting. Differential testing was performed using the Wald test, with multiple testing correction applied via the Benjamini–Hochberg method. Significantly differentially expressed genes were filtered using the thresholds: P.adjust <.05 and |log2FoldChange| > 1. Gene Set Enrichment Analysis (GSEA): Using the log2FoldChange output from DESeq2 as the ranking metric, GSEA was performed using the built-in GSEA function in the clusterProfiler package (v4.10.1), with a significance threshold of P.adjust <.05. KEGG pathway enrichment analysis: Downregulated genes common to both comparison groups were selected, and enrichment analysis was conducted using clusterProfiler via the hypergeometric test, with the Mus musculus whole genome as the background; P-values were adjusted using the Benjamini–Hochberg method, with a significance threshold of P.value <.05. Gene Set Variation Analysis (GSVA): Using the GSVA package (v1.50.5), the activity scores of the OXPHOS pathway (gene set: “WP_OXIDATIVE_PHOSPHORYLATION.v2025.1.Mm.gmt”) were assessed across the four sample groups. Significant differences between groups were evaluated using the T-test.

Biosafety evaluation of MSC-sEV in mice

Healthy mice were administered either iPSC-MSC-sEV (1.5 × 10¹⁰) in 200 µL PBS or an equal volume of PBS alone (control group) via the tail vein on day 0, 1, 2, and 4. Subsequently, mice were euthanized on day 5. Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine kinase (CK), lactate dehydrogenase (LDH), urea, and creatinine (CREA) were measured. Tissue sections from the heart, liver, spleen, lungs, and kidneys, embedded in paraffin, were subjected to routine hematoxylin and eosin (H&E) staining.

Statistical analysis

Statistical analysis was evaluated using GraphPad Prism 7 software and presented as the mean ± SEM. Comparisons between the two groups were performed by a two-way Student’s t-test. P-values of .05 or less were considered significant.

Results

MSC-sEV generation and identification

MSC-sEVs were generated using the anion-exchange chromatography method as our previous study,²⁰ which simply means that the negatively charged property of sEV is used to adsorb sEV through an anion exchange chromatography column, remove impurities, and then, elute sEV down, collect it and perform steps such as purification and concentration, followed by analysis and experiments (Figure 1A). According to the recommendations in MISEV2023 (Minimal information for studies of extracellular vesicles 2023) we quantified the particle number and size of the extracted MSC-sEV by NTA detection²¹ (Figure 1B). The NTA results showed that the particle number of our prepared MSC-sEV was 30 × 10¹⁰ particles/mL, where the particle size was mainly distributed at 103 nm, which is consistent with the exosome size of 30-150 nm range. Then, we performed transmission electron microscopy to directly observe the morphology of the small extracellular vesicles, and the results showed that our extracted sEV possessed a typical bilayer membrane structure (Figure 1C). Then, at least two positive protein markers (including at least one transmembrane/lipid binding protein or cytoplasmic protein) as well as one negative protein were detected by WB as recommended by MISEV2023,²¹ and the results showed that MSC-sEV expressed CD63, CD9, CD81, TSG101, and Alix but not Calnexin, in line with the exosome marker protein profile (Figure 1D and Supplementary Figure S1—see online supplementary material for a color version of this figure). The results indicated that we successfully prepared MSC-sEV.

Figure 1.

Successful preparation and characterization of MSC-sEV. (A) Schematic diagram of MSC-sEV preparation by anion exchange chromatography. (B) Nanosight detection of MSC-sEV particle size distribution and particle number. The particle number is 30 × 10¹⁰/mL, and the particle size is around 103 nm. (C) Projection electron microscopy shows that MSC-sEV have a typical bilayer membrane structure. (D) Western blot validation of MSC-sEV marker proteins. The original blots are presented in Supplementary Figure S1 (see online supplementary material for a color version of this figure).

sEV alleviated HDM-induced allergic airway inflammation through downregulating cDC2s and moDCs accumulation in the lung

DCs were vital to mediate Th2- mediated immunity in the context of inhaled HDM. Our previous studies revealed that MSC-sEV ameliorated ovalbumin-induced Th2-mediated allergic airway inflammation.¹⁹ To understand if MSC-sEV also played a role in regulating the process of DCs-induced Th2-mediated immunity in response to inhaled antigen, we established HDM model by sensitizing mice with 100 μg HDM i.t. on day 0 and challenging mice with 10 μg HDM i.t. on day 7 to 11 while mice were treated with MSC-sEV on day 6, 7, 8, and 10 by i.v. (Figure 2A). The reason for this administration is to give a sufficient amount of sEV at the beginning of the excitation, and then, again during the excitation to maintain the effect. We found that sEV significantly reduced the frequency and the number of eosinophils in the lung but not neutrophils (Figure 2B). It suggested that sEV played an important role in decreasing eosinophils accumulation in the lung after exposure to HDM.

Figure 2.

sEV alleviated HDM-induced allergic airway inflammation and downregulated cDC2s and moDCs accumulation in the lung. (A) Schematic of HDM model. Mice were continuously sensitized with HDM i.t. from day 7 to day 11 and PBS or sEV was used for treatment at day 6, 7, 8, and 10. Mice were sacraficed on day 12. Naïve mice with non-HDM sensitization were used as control mice. (B) Summary graphs showed the proportion and the number of eosinophils and neutrophils in the lung. (C) Flow plots and graph of total CD11⁺MHC II⁺ DCs frequency in the lung. (D) Summary data of the percentage of CD103⁺CD11b^- cDC1s and CD103^-CD11b⁺ cDC2s in the lung. (E) Flow plots and graph show the frequency of lung Mar1⁺ CD64⁺ moDCs. (F) Histopathology of lungs on day 12. Lung tissues were stained with hematoxylin and eosin (HE) for histopathology assessment. (G) Summary graph of inflammation score and alveolar wall thickness. Data were represented as two independent experiments with n = 4 to 7 mice in each group. Error bars represented the SEM. (*P < .05; **P < .01; ***P < .001; ****P < .0001).

As researchers have identified CD11b⁺cDC2s, specifically moDCs, expressed several chemokines with activity on eosinophils,¹² we, then, investigated the role of sEV on orchestrating DCs subsets in the lung after HDM exposure. The gating strategy is shown in a Supplementary Figure S2 (see online supplementary material for a color version of this figure). Interestingly, the proportion of total CD11c⁺MHC II⁺DCs in the lung was downregulated after sEV treatment (Figure 2C). Particularly, sEV efficiently decreased the frequency of CD11b⁺CD103^-cDC2s (Figure 2D). Given that mainly CD11b⁺cDC2s but not cDC1s induced Th2-mediated airway inflammation in HDM model, not surprisingly, the percentage of cDC1s was oppositely upregulated in naïve mice and sEV-treated group compared with PBS treatment (Figure 2D). In addition, cDC2s expressing MAR-1 and CD64 was recognized as moDCs which played a crucial role in initiating Th2 immunity to inhale allergen.^22,23 moDCs was decreased in sEV-treated lung tissue (Figure 2E).

Consistently, histopathology of alveolar wall thickness and inflammation scores in lung tissues revealed that inflammatory levels of sEV-treated mice were mitigatory (Figure 2F and G). Furthermore, our safety assessment results confirmed the absence of any detectable side effects or toxicity in the organs of hearts, livers, spleens, lungs, and kidneys following MSC-sEV administration (Supplementary Figure S3—see online supplementary material for a color version of this figure).

In short, sEV efficiently alleviated Th2-mediated immunity to HDM allergen by orchestrating the generation of HDM-specific cDC2s and moDCs.

DCs were able to take MSC-sEV in vitro and in vivo

To validate the effect of sEV on DCs, the uptake effects of DCs on mCherry-sEV in vitro and in vivo were evaluated. First, we generated mouse bone marrow-derived DCs and co-cultured BMDCs with sEV. To validate if DCs function was promoted by direct uptake of sEV, we treated BMDCs with mCherry-sEV at different concentrations (10 μg/mL or 20 μg/mL) for 12 or 24 h as our previous study.²⁰ We found that BMDCs were able to take up sEV in a dose-dependent with the peak at least 12 h (Figure 3A). Using confocal immunofluorescence, we found that BMDCs could directly take up sEV. (Figure 3B). The uptake of sEV by murine DCs was evaluated in HDM-induced airway inflammation model in mice (Figure 3C), and DCs in murine lungs were able to uptake sEV (Figure 3D). To understand the effect of sEV on human DCs, we purified CD14⁺ monocytes from human PBMC and cultured them in persistence with GM-CSF and IL-4 to generate human moDCs. Not surprisingly, human moDCs improved sEV-uptake capacity with increasing doses of sEV treatment (Figure 3E). Overall, these results indicated that DCs were able to take in MSC-sEV in allergic airway inflammation.

Figure 3.

Uptake of sEV by BMDCs and mo-DCs in vitro and in vivo. (A) mCherry-labelled sEV (10 μg/mL or 20 μg/mL) was added to BMDCs 12 or 24 h before the flow-through assay on day 6 of BMDCs incubation. (B) mCherry-labelled sEV (10 μg/mL per well) was added to the BMDCs on day 6. Then, the cell was photographed by confocal microscopy. (C, D) A mouse model of HDM-induced airway inflammation in mice was established (see methods section for details) and administered intravenously at 100 μg sEV on per day of the model. Mouse lung was taken on the third day for analysis. The absorption of sEV in mouse lung DCs was shown in the FACS analyses. (E) PBMCs were extracted from human peripheral blood, CD14⁺ cells were sorted by magnetic beads, then, induced into iDCs with human IL-4 with human GM-CSF, and mCherry-labelled sEV (10 μg/mL or 20 μg/mL per well) was added 12 h before the flow assay on day 6. Displays representative FACS analyses with numbers representing the proportion of mCherry positive cells. (n = 3, *P < .05, **P < .01, ***P < . 001).

sEV relieved the activation of murine BMDCs in vitro

To understand sEV function on DCs activation, sEV was added one day before the LPS-induced activation of BMDCs. It has been shown that co-stimulatory molecules and cytokines such as IL-12, were essential to antigen-presenting capacity to prime T cell response.⁶ Interestingly, sEV decreased the expression of co-stimulatory molecules, such as CD80 and CD86 (Figure 4A and B). In addition, the productions of TNFα, IL-6, and IL-12 cytokines, were downregulated after sEV treatment (Figure 4C).

Figure 4.

sEV relieved LPS-induced activation of murine BMDCs in vitro. Murine BMDCs were treated with sEV one day before LPS activation. (A, B) Histograms and summary graphs of the MFI of BMDCs CD80 (A) and CD86 (B) expression. (C) Graphs of BMDCs TNFα, IL6, and IL12 production. (D) RNA-sequencing was performed on CD11⁺MHC II⁺ BMDCs subsets on day 6. Heatmap showed the relative expression of selected differentially expressed genes (DEG). Summary data (mean ± SEM in A-C) were represented as two independent experiments. (*P < .05; **P < .01; ***P < .001; ****P < .0001).

To uncover sEV-affected genes relevant to DCs, we performed RNA sequencing of sorted CD11c⁺MHC II⁺BMDCs. We compared PBS and sEV-treated BMDCs in steady and activated states. As expected, sEV-treated BMDCs expressed lower several genes both in steady and activated states including genes encoding activated molecules Cd80, Cd40, Cd86, Il1a, and Il1b; cytokines known to be critical for T cell-priming such as Tnf and Il15; the chemokines Ccl3, Ccl4, and Ccl5, as well as genes involved to DCs differentiation such as Itgax, Id2. These findings collectively demonstrated that sEV treatment can efficiently mitigate DCs activation (Figure 4D).

Differential expression profiles and gene set enrichment signatures

To further identify how sEV regulated DCs activation, the analysis of global genes was applied to understand the mechanism. Under steady condition, the sEV treatment group exhibited a total of 847 differentially expressed genes (276 upregulated, 571 downregulated) compared to the PBS control group (Figure 5A). GSEA revealed significant enrichment (P.adjust <.05) of five immune signaling pathways after sEV treatment, including the Hematopoietic cell lineage, Cytosolic DNA-sensing pathway, NOD-like receptor signaling pathway, PI3K-Akt signaling pathway, and T cell receptor signaling pathway (Figure 5B). Under LPS stimulation conditions, the sEV_LPS group showed 699 differentially expressed genes (180 upregulated, 519 downregulated; Figure 5C) compared to the PBS_LPS group. The volcano plot indicated that downregulated genes were predominantly clustered within immune-related functional categories, suggesting that sEV primarily exerts immunosuppressive effects. GSEA results confirmed that this immunomodulatory effect of sEV persisted under LPS stimulation, demonstrating that sEV’s immunoregulatory effect is independent of the LPS background (Figure 5D). Venn analysis showed that the two comparisons shared 421 downregulated genes (Figure 5E). KEGG enrichment analysis indicated that these genes were significantly enriched in pathways including hematopoietic cell lineage, platelet activation, T cell receptor signaling, Th1 and Th2 cell differentiation, cytokine-cytokine receptor interaction, PI3K-Akt signaling, and JAK-STAT signaling (Figure 5F), further supporting the dual inhibitory role of sEV on both adaptive and innate immunity. Subsequent analysis of Th2-related pathways in the sEV_LPS and PBS_LPS groups showed a decreasing trend after sEV treatment, although it did not reach statistical significance, possibly due to insufficient sequencing sample size (Supplementary Figure S4A and B—see online supplementary material for a color version of this figure). Previous studies have shown that immature DCs rely on OXPHOS to maintain survival, whereas their activation requires glycolysis to support maturation, upregulation of co-stimulatory molecules, cytokine secretion, and T cell stimulatory capacity.²⁴ GSVA scores demonstrated that sEV significantly enhanced OXPHOS pathway activity both in the steady state and under LPS stimulation (P < .01, Figure 5G), suggesting that sEV reprograms immune cell metabolism by enhancing mitochondrial oxidative phosphorylation. Notably, the heatmap revealed significant differences in metabolism-related genes: OXPHOS-associated genes (such as Cox8a, Slc2a1, Ndufa7, Ndufa6, Atp5e) were upregulated in sEV-treated BMDCs, whereas glycolysis pathway genes (such as Pfn1, Hk3, Cd36, Pgk1) were more active in the PBS group (Figure 5H and Supplementary Figure S4C—see online supplementary material for a color version of this figure). In contrast, the expression levels of Nos2 and Ldha, proteins crucial for DCs activation during glycolysis,^25,26 were reduced in the sEV group following LPS treatment (Supplementary Figure S4C—see online supplementary material for a color version of this figure). In summary, these findings reveal a mechanism whereby sEV reshapes DCs metabolism by enhancing OXPHOS and inhibiting glycolysis, thereby maintaining DCs homeostasis and alleviating excessive activation.

Figure 5.

sEV orchestrated the metabolism of murine BMDCs both in steady and activated states. (A) Volcano plot of differentially expressed genes for sEV vs PBS. Red dots: 276 significantly upregulated genes (Padj <.05, log2FC > 1); Blue dots: 571 significantly downregulated genes (Padj <.05, log2FC <−1). The top 10 downregulated genes based on Padj are labeled. (B) GSEA enrichment heatmap for sEV vs PBS. (C) Volcano plot of differentially expressed genes for sEV_LPS vs PBS_LPS. Red: 180 upregulated genes; Blue: 519 downregulated genes. The top 10 downregulated genes are labeled. (D) GSEA enrichment heatmap for sEV_LPS vs PBS_LPS. (E) Venn diagram of downregulated genes. The number indicates the 421 downregulated genes shared between the sEV_vs_PBS and sEV_LPS_vs_PBS_LPS comparisons. (F) Bubble plot of KEGG enrichment analysis for the common downregulated genes. Y-axis: Pathway names; X-axis: GeneRatio (number of enriched genes/number of background genes in the pathway); Dot size: Number of enriched genes; Color: P.value. (G) Box plot of GSVA scores for the OXPHOS pathway across the four sample groups. (H) Heatmap of selected differentially metabolism-related genes among PBS, sEV and LPS-BMDCs treated with PBS or sEV).

sEV promoted aerobic respiration of murine BMDCs

To directly test the role of sEV in regulating metabolic capacity, we performed an extracellular flux assay of PBS or sEV-treated BMDCs in the activated state. Consistent with our RNA-sequencing analysis, sEV promoted oxygen consumption (OCR) capacity compared with PBS BMDCs (Figure 6A). Specifically, sEV BMDCs significantly exhibited higher basal, maximal, and spare respiratory levels (Figure 6B). Furthermore, ATP production, which was produced in the OXPHOS process was also increased in sEV BMDCs (Figure 6C) while sEV had no effect on mitochondrial contents in dendritic cells (Figure 6D). To verify if sEV decreased DCs activation via promoting OXPHOS, we tried to use OXPHOS inhibitor oligomycin to test DCs activation including surface markers and cytokine productions. However, it was limited to the use of oligomycin in the LPS activation environment because almost cells tended to be apoptosis without OXPHOS capacity (Figure 6E). Even in this way, sEV BMDCs exhibited better survival capacity, which suggested that sEV-treated BMDCs was more adaptable to environment changes compared with PBS BMDCs.

Figure 6.

sEV promoted aerobic respiration of murine BMDCs. (A) The kinetic curve of OCR in PBS or sEV-treated BMDCs after LPS activation for 10 h. (B, C) Summary graphs of respiratory capacity (B) and ATP production of PBS or sEV-treated BMDCs after LPS activation. (D) Histogram and graph showing mitochondrial content of PBS or sEV-treated BMDCs after LPS activation. (E) Graph of the frequency of Annexin V⁺ apoptosis BMDCs. Summary data (mean ± SEM in B-D) were represented as two independent experiments. (*P < .05; **P < .01; ***P < .001).

sEV downregulated the activation of human DCs in vitro via modulating metabolism

To understand the effect of sEV on human DCs, we purified CD14⁺ monocytes from human PBMCs and cultured them in persistence with GM-CSF and IL-4 to generate human moDCs. Consistent with our mouse experiments, sEV decreased the activation of human moDCs including TNFα and IL-12 production (Figure 7A) as well as the expression of CD80 and HLA-DR (Figure 7B and C). Mechanically, sEV promoted OXPHOS to mitigate human DCs activation while PBS moDCs exhibited a higher glycolysis process of OCR and ECAR (Extracellular acidification rate; Figure 7D and E). These results showed sEV alleviated DCs activation by promoting oxygen consumption to maintain homeostasis both in human and murine DCs.

Figure 7.

sEV downregulated the activation of human DCs in vitro via modulating metabolism. (A-C) Histogram and graphs of the proportion of TNFα and IL-12 cytokines production (A), and MFI of CD80 as well as HLA-DR (B, C). (D, E) Extracellular flux analysis of OCR and ECAR in human DCs after PBS or sEV treatment. Summary data (mean ± SEM in A-C) were represented as two independent experiments. (*P < .05; **P < .01; ***P < .001; ****P < .0001).

Discussion

Currently, the therapeutic strategy for allergic rhinitis and asthma has been established, ranging from first-line intranasal corticosteroids and antihistamines to leukotriene receptor antagonists, and extending to biological agents targeting specific immune pathways, and allergen immunotherapy (AIT) aimed at inducing immune tolerance. The evidence has confirmed that these therapies could effectively control symptoms, improve patients’ quality of life, and, in some cases, potentially affect the natural course of the disease. However, clinical practice still faces numerous challenges, such as less response to existing medications in some patients, the long duration of AIT with its associated risk of allergic reactions, and the limited capacity of current strategy to repair damaged airway mucosa.²⁷

Against this backdrop, mesenchymal stem cell-derived exosome (MSC-exosome) therapy, as an emerging cytokine delivery system, presents a treatment concept that shows significant differences and complementary potential compared to existing methods. Unlike conventional drugs that primarily focus on antagonizing specific inflammatory mediators (such as histamine, IgE, leukotrienes), exosomes carry a rich cargo of bioactive substances capable of acting at an upstream level of immune regulation. By modulating T-cell differentiation and balancing the Th1/Th2 immune response, they may intervene more fundamentally in the process of allergic inflammation.²⁸ This multi-targeted, network-based mechanism of action suggests potential not only for pure allergic rhinitis but also for providing a synergistic treatment strategy for “unified airway” diseases involving both the upper and lower respiratory tracts, particularly in complex cases comorbid with asthma. Furthermore, while existing therapies effectively suppress inflammation, their impact on airway remodeling and tissue repair is minimal. In contrast, exosomes have been demonstrated in experimental models to possess the capacity to promote epithelial repair and mitigate fibrosis, offering new hope for improving the long-term prognosis of the disease.

In this study, we demonstrated that MSC-sEV significantly mitigated Th2-mediated airway inflammation by decreasing cDC2s and moDCs accumulation in the lung. Mechanically, MSC-sEV promoted oxidative phosphorylation of dendritic cells to maintain the homeostasis in response to inflammatory environment. Notably, MSC-sEV not only modulated murine DCs but also exhibited great effects on alleviating human moDCs activation and decreasing glycolysis pathway against excessive activation.

Given that MSCs displayed the ability to reduce lung inflammation in allergic airway inflammation,^29,30 here, our experimental studies revealed that MSC-sEV show similar function and MSC-sEV also has the potential to treat allergic airway inflammation. Our Th2 airway inflammation murine models showed that lung DCs and eosinophils were decreased after MSC-sEV treatment. In more detail, MSC-sEV downregulated cDC2s and inflammatory moDCs thereby reducing Th2 lung inflammation level. To determine how exosome effect on DCs functions, we used cytokine GM-CSF and IL-4 to culture moDCs. We found that DCs were able to uptake exosomes, which relieved inflammatory level of dendritic cells after MSC-sEV treatment, including the expression co-stimulatory molecules CD80 as well as CD86, and the production of pro-inflammatory cytokines, such as IL-6, TNF-α, and IL-12. What’s more, the results of RNA-sequencing showed that the expression of dendritic cells activated-related genes decreased, which is consistent with our previous data, including co-stimulatory molecules, cytokines, and migratory-related chemokines such as CCR7.

Interestingly, we found that MSC-sEV displayed great effect on the mechanism pathway on dendritic cells. DCs subsets display various adaptions of metabolism in different settings. Steady-stage DCs depend on a catabolic metabolism for cell maintenance, exhibiting a hallmark of active OXPHOS driven by the tricarboxylic acid (TCA) cycle. In this process, steady DCs use intracellular glycogen to provide substrate for mitochondrial respiration.³¹ In addition, active DCs apply an anabolic metabolism to generate substrate to support cell growth and activation, switching to the glycolysis pathway dependence on mTOR/HIF-1α signaling.³² Nitric oxide (NO) and ROS produced in the glycolysis process promoted the expression of co-stimulatory molecules including CD40, CD80, and CD86, and cytokine production in DCs such as IL-12, TNFα, and IL-6. Here, we revealed that MSC-sEV promoted the expression of OXPHOS pathway-related genes in DCs, such as Cox8a, Slc2a1, Ndufa7, Ndufa6, and Atp5e and mitochondrial respiration-related genes, such as Cox5a, Cox7c, and Ndufv3. Indeed, the results of extracellular flux analysis showed that the level of DCs oxygen consumption increased after MSC-sEV treatment even though the mitochondrial content showed similar level, suggesting that exosome inhibition of DCs activation may be achieved by altering their metabolic state.

Previous studies have shown that MSCs could inhibit the differentiation, maturation, and function of human DCs.³³ To examine whether our experimental results are equally applicable to human moDCs, we performed similar experiments on human moDCs. Similar to our murine data, human moDCs were able to uptake MSC-sEV. Besides, MSC-sEV treatment significantly reduced the production of cytokines and the expression of co-stimulatory molecules in human moDCs. Furthermore, the results of extracellular assay showed that oxygen consumption upregulated while glycolysis pathway decreased in human moDCs after MSC-sEV treatment, meaning the activation of human moDCs was suppressed. In short, MSC-sEV not only can be used to relieve the inflammatory levels of murine dendritic cells, but also showed a potential therapeutic capacity for allergy by improving the inflammation of human moDCs.

We acknowledgment that there are still some limitations for our study. Due to dendritic cell survival depending on their metabolic capacity, relevant trials using OXPHOS inhibitors on dendritic cells were not able to be completed. Besides, the issues of large-scale preparation and clinical ethics of MSC-sEV application remained to be advanced. Nevertheless, it was undeniable that MSC-sEV shows the promising potential of biological therapy in Th2-mediated airway inflammation.

Conclusion

To be concluded, this study suggested that MSC-sEV can exert therapeutic effect on allergic airway inflammation by regulating DCs, mainly mediated by metabolic reprogramming of DCs. These findings provide new insights into the mechanisms and contribute the clinical application of MSC-sEVs therapy in allergic airway inflammation.

Author contributions

Long-Xin Huang: Formal analysis, Investigation, Writing—original draft, Writing—review & editing. Lifen Wen: Conceptualization, Methodology, Validation, Writing—original draft, Investigation, Resources. Tian Tian: Investigation, Data curation, Formal analysis, Validation, Supplementary analysis & experiments. Zi-Cong Wu, Chan-Gu Li, Qi-Sun: Investigation, Data curation, Formal analysis, Validation. Xiao-Qing Liu, Zhi-Rou Zhou: Validation, Resources. Zhi-Bin Xu, Bi-Xin He, Ying-Chun Xie: Writing—review & editing. Qing-Ling Fu: Conceptualization, Funding acquisition, Project administration, Supervision, Writing—review & editing, Data curation.

Supplementary material

Supplementary material is available at Stem Cells Translational Medicine online.

Funding

This work was supported by grants from National Key R&D Program of China (No. 2022YFA1104900), National Natural Science Foundation of China (Nos. 82571286, 82271144, 82501380), Guangdong Basic and Applied Basic Research Foundation (No. 2021B1515120062) and Guangzhou Key R&D Program (Nos. 2023B03J1233 and 20220600003).

Conflicts of interest

The authors declare no conflict of interest.

Data Availability

The datasets used and analysed during the current study are available from the corresponding author on reasonable request. The RNA sequence data have been deposited at the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm. nih.gov/geo/) under accession number GSE268139.

Ethics approval and consent to participate

Human-derived dendritic cells experiment was approved by The Ethics Committee of The First Affiliated Hospital, Sun Yat-sen University on September 16, 2019 (No. 2019-331). The project title was “The effects of exosomes derived from stem cells on immune cells,” and exemption of written informed consent was approved. The animal experiment was approved by the Laboratory Animal Ethics Committee of Sun Yat-sen University on December 31, 2022 (No. 2022003376). The project title was “Immunomodulatory effects of stem cell-exosomes on airway allergic inflammation.”