Immunomodulatory effect of PLGA-encapsulated mesenchymal stem cells-derived exosomes for the treatment of allergic rhinitis

Six to eight weeks old male Balb/c mice were purchased from Vital River Animal Technology Co., Ltd (Beijing, China) and housed in specific pathogen free (SPF) Laboratory Animal House of Tongji University (Shanghai, China). All animal experiments and procedures were conducted following the ethical guidelines of the School of Medicine, Tongji University (Permit no. T3-HB-LAL-2023-23) and recommendations of Guide for the Care and Use of Laboratory Animals of Ministry of Science and Technology of the People's Republic of China.

Human nasal epithelial cells (HNEpCs) were purchased from Procell Life Science and Technology Co., Ltd (Wuhan, China) (Cat NO.: CP-H252) and THP-1 cell line was obtained from Chinese Academy of Sciences Typical Culture Preservation Committee Cell bank (ATCC, Shanghai, China) and maintained in MEM (Shanghai BasalMedia Technologies, Shanghai, China) supplemented with 10% fetal bovine serum (FBS) (Gibco, Shanghai, China) and 1% penicillin-streptomycin (Shanghai BasalMedia Technologies, Shanghai, China) in humidified conditions with 5% CO₂ and 95% air at 37 °C. The RPMI2650 cell line was provided by Jennio Biotech Co., Ltd (Guangzhou, China) and grown in RPMI1640 with L-glutamine (Corning, Manassas, USA) supplemented with 10% FBS. Human umbilical cord derived mesenchymal stem cells (HUC-MSCs) were purchased from Cytoniche Biotechnology Co., Ltd (Beijing, China) and maintained in Dulbecco's modified eagle medium (DMEM) (Biosharp, Beijing, China) supplemented with 10% FBS.

HUC-MSCs-derived exosomes (MSC-Exos) were isolated from the collected culture medium by ultracentrifugation as described previously (23, 24). Briefly, HUC-MSCs were cultured with DMEM containing exosome-free FBS (A27208-03, Gibco, Shanghai, China). When the cell confluency reached 80-90%, the culture medium was collected after 48 hr and centrifuged at 10,000 g for 30 min to remove the dead cells and cell debris. The culture medium was then filtered through 0.22 µm pore sized sterile filters (Millipore, Billerica, MA, USA) to ensure the depletion of large particles. The filtered culture medium was then centrifuged at 100,000 g for 90 min to obtain the pellet of exosomes. The pellet was resuspended in sterile PBS, centrifuged again at 100,000 g for 90 min, and stored at -80 °C until future use.

The morphology and size distribution of isolated MSCs-Exos were analyzed using transmission electron microscopy (TEM) (FEI F20, Columbus, USA) and nanoparticle tracking analysis (NTA) (NS300, NanoSight, Malvern, UK), respectively (24). Western blotting was performed to determine the expression of CD81, CD63, HSP70 and calnexin (Proteintech, USA) protein markers in the extracted MSCs-Exos (25).

PLGA particles of different sizes (1 µm, 800 nm, 400 nm and 200 nm) with or without encapsulation of indocyanine green (ICG) (Macklin, Shanghai, China) or exosomes were prepared using double emulsion method as described previously (26, 27) with minor modifications. In brief, 100 mg of 0.67 dL/g carboxy-terminated 50:50 PLGA polymer (Daigang Co, Jinan, China) was dissolved completely in dichloromethane (Titan Scientific, Shanghai, China), and 50 µL ICG solution (20 mg/mL, in DMSO) or MSCs-Exos (3 µg, 5 µg, 10 µg or 50 µg/mg PLGA) were added and sonicated using an ultrasonic homogenizer (JY92-IIN, Ningbo Xinzhi Biotech., Ningbo, China) at 90% for 6 min (10 sec sonication and 15 sec break). The mixture was then added in 50 mL 1% poly vinyl alcohol (PVA) (Macklin, Shanghai, China) and sonicated again at 90% for 6 min in case of 200 nm, 4 min for 400 nm, 2 min for 800 nm, and 1.5 min for 1µm sized PLGA particles (10 sec sonication and 15 sec break). Finally, 100 mL dd H₂O was added in the solution and placed overnight on a magnetic stir bar to evaporate the dichloromethane. Finally, the solution was centrifuged (at 3000 rpm for 1 µm, 6000 rpm for 800 nm, 9000 rpm for 400 nm, and 12000 rpm for 200 nm particles) to collect the PLGA micro/sub-micron particles (PLGA-MPs/SMPs). All sizes of particles with ICG encapsulated PLGA-MPs/SMPs (ICG-PLGA MPs/SMPs) were centrifuged at 20,000 rpm for 10 min to remove the unencapsulated ICG. The resulting PLGA particles were stored at -20 °C. ICG-labelled exosomes (ICG-Exos) for in vivo tracking were prepared by incubating 10 µg/mL ICG in 50 µg exosomes for 12 hr at 4 °C (28, 29). The exosomes were centrifuged at 100,000 g for 90 min by adding sterile PBS to remove the unattached ICG.

Scanning electron microscopy (SEM) (Hitachi, S-4800, Japan) was performed to visualize the surface morphology and shape of prepared PLGA-MPs/SMPs. The size distribution and zeta potential were analyzed using dynamic light scattering (DLS) (OPT2301140, Opptronix, Shanghai, China).

In order to evaluate the sinonasal retention of ICG-Exos and different sized (1 µm, 800 nm, 400 nm and 200 nm) ICG-PLGA MPs/SMPs, 50 µg of ICG-Exos or 5 mg of ICG-PLGA-MPs/SMPs were suspended in 30 µL PBS and administered intranasally (15 µL in each nasal cavity) into every Balb/c mouse. The mice were placed under anesthesia by isoflurane inhalation for in vivo imaging of whole head using InVivo Smart-LF system (VISQUE, B12BAA002, Korea) at different time points (30 min, 2 hr, 6 hr, 12 hr, 24 hr, 48 hr, 72 hr, 96 hr, 120 hr and 168 hr). The exposure time was 1 min, whereas excitation and emission wavelengths were 740-790 nm and 810-860 nm, respectively. Three mice for each group were imaged for statistical analysis.

The in vitro release profile of exosomes from 800 nm sized PLGA SMPs encapsulated exosomes (PLGA-Exos) with different concentrations of exosomes (3 µg, 5 µg, 10 µg or 50 µg/mg PLGA) over time was measured for 11 days. A BCA assay kit (Beyotime, Shanghai, China) was used to assess the release profile. Based on the values, a cumulative release curve was plotted (n=3). In brief, 5 mg of PLGA-Exos with different concentrations were added in sterile 1.5 mL Eppendorf tube with 1 mL PBS and PLGA-Exos were suspended thoroughly. The tubes were placed in an incubator at 37 °C and 20 µL supernatant was replaced with fresh PBS by centrifugation at 12,000 g for 10 min on days 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 11. The exosomes release was determined by the released protein content.

RPMI2650 cell line was used as nasal epithelial cells, whereas THP-1 cells stimulated by phorbol-12-myristate-13-acetate (PMA) (Sigma-Aldrich, MO, USA) were used as macrophages in the in vitro models. 100 ng/mL of PMA was added in 1×10⁶ THP-1 cells for 24 hr in an incubator at 37 °C with 95% air and 5% CO₂ to obtain macrophages. PLGA-Exos labelled with PHK26 fluorescent dye (Solarbio, Beijing, China) were co-cultured with RPMI2650 and THP-1 cells. Briefly, RPMI2650 and THP-1 cells were cultured in 15 mm polystyrene cell culture dish (NEST, USA) for 24 hr until 80-90% confluency was attained. The cells were co-cultured with 50 µL of 800 nm PLGA-Exos (5 mg PLGA, 10 µg exosomes/mg PLGA) after adding 500 µL culture medium for 4 hr, 12 hr, 24 hr and 48 hr. Likewise, THP-1 cells were also co-cultured with 50 µg PHK26 labelled exosomes. Cells were fixed with 4% paraformaldehyde (PFA) (Titan, Shanghai, China) for 30min at room temperature (RT), incubated with β-actin (Abcam, Cambridge, UK) primary antibody overnight at 4 °C, and stained with Alexa Fluor^® 488 (Abcam, Cambridge, UK) secondary antibody for 1 hr at RT. DAPI was used to stain the nuclei of the cells. Finally, the cells were imaged using a laser scanning confocal microscope (LSCM) (ECLIPSE Ti2, Nikon, Japan).

HNEpCs were cultured in 12-well plate at a density of 1×10⁵ cells/well in 500 µL culture media. The culture media were removed completely after 12 hr of incubation when 90% confluency was obtained. Medium with 0.1 µg/mL lipopolysaccharide (LPS) was added in the wells of positive control and treatment group, whereas unstimulated cells were considered as negative control. After 24 hr of stimulation, PLGA-Exos (5 mg/well) were added in the treatment group and incubated for further 24 hr. The culture media was aseptically collected from wells of each group and centrifuged at 3000 rpm for 5 min to remove the cell debris and supernatant, and then preserved at -80 °C until assayed. Enzyme linked immunosorbent assay (ELISA) and real time quantitative PCR (RT-qPCR) was performed to analyze the effect of PLGA-Exos treatment at protein and transcriptomic levels, respectively.

The AR model was established as described previously (30, 31) with minor modifications. Briefly, Balb/c mice were randomly divided into 5 groups, i.e., negative control (NC, normal mice), positive control (OVA, AR mice), blank PLGA (PLGA), exosomes (Exos) and exosomes-encapsulated in PLGA sub-micron particles (PLGA-Exos) groups (n = 12/group). Mice from the last 4 groups were first subjected to basic sensitization by intraperitoneal injection of 200 µL of Grade V ovalbumin (OVA) and aluminum hydroxide (Al(OH)₃) (Sigma-Aldrich, GmbH Germany) (1 mg/mL OVA+100 mg/mL Al(OH)₃) on day 0, 7 and 14. These mice were then challenged for 7 consecutive days with intranasal administration of 20 µL OVA solution (40 mg/mL) (10 μL in each nostril) on a daily basis up to day 21. The control group (NC) was sensitized and challenged with the same amount of PBS at a same frequency.

On day 22 (day 0 for treatment), all 5 groups were intranasally administered with individual regimen (NC and OVA with PBS, PLGA with Blank PLGA, Exos with only exosomes and PLGA-Exos with PLGA encapsulated exosomes) with a total volume of 20 µL after every five days for six sessions. The AR symptoms and signs, i.e., nasal scratching, sneezing and number of eosinophils in nasal lavage fluid, of 6 mice from each group were observed, recorded and scored after the last treatment for a period of 20min by a blinded observer (). The nasal tissues, spleens, blood samples from each treatment group were collected at 2 (after 3 sessions of treatment) and 4 weeks (after 6 sessions of treatment). 1

The concentrations of IFN-γ (Servicebio®, Wuhan, China), IL-2, IL-4, and IL-10 (EK102-01, EK104-03, EK110/2-02; Multi Sciences, Hangzhou, China) in the HNEpCs cell culture supernatant were measured using ELISA as per manufacturer's protocol. Blood sample collected after eyeball removal from mice of each treatment group was placed for 3 hr at 4 °C and then centrifuged at 3000 rpm for 15 min to obtain serum. The isolated serum was stored at -80 °C until further use. The serum was analyzed for IFN-γ, IL-4, (GEM0006, GEM0006; Servicebio®, Wuhan, China), IL-10 (88-7105; Invitrogen, Vienna, Austria), IL-17 (EK217/2-01; Multi Sciences, Hangzhou, China) and sIgE (Konodi Bio, Xiamen, China) concentrations using ELISA according to given protocols.

Total RNA was isolated from the cultured cells and spleen tissues of each treatment group using an RNA-Quick Purification Kit (ES Science, Shanghai, China). The concentration of RNA isolated from each sample was quantified using a NanoDrop1000 Spectrometer (ThermoScientific, Waltham, MA, USA). 1 µg of RNA from each sample was used to synthesize cDNA by reverse transcription using a HiScript 1^st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). The RT-qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China). The primer sequences for HNEpCs IFN-γ, IL-2, IL-4, IL-10, and GAPDH and mice IFN-γ, IL-2, IL-4, IL-10, IL-17 and GAPDH are presented in Supplementary Tables S1, S2, respectively. The expression of mRNA was evaluated using RT-qPCR on an ABI 7500 Real-Time PCR System (Applied Biosystems, California, USA). The amplification conditions for PCR were 95 °C for 30 sec, followed by 40 cycles of denaturation at 95 °C for 10 sec and annealing at 60 °C for 20 sec. Housekeeping gene (GAPDH) was used to calculate the targeted gene expression levels and data was analyzed using the comparative ΔΔCT method.

Nasal tissues collected from mice of different treatment groups (2 and 4 weeks) were fixed in 4% PFA for 5 days at 4 °C and then decalcified by EDTA decalcification buffer (Cat#G1105, Servicebio, Wuhan, China) for 2-3 days at RT in a shaking incubator. The tissues were then dehydrated and embedded in paraffin, cut into sections of 4 µm thickness using a microtome, and observed under light microscope (Nikon) using routine H&E and PAS staining. The mean numbers of eosinophils and goblet cells were calculated by counting 10 different fields at 400× magnification, using Image-Pro Plus software (Media Cybernetics, Rockville, MD).

Spleens collected at 2 and 4 weeks from mice of different treatment groups (n=6) were processed to isolate the lymphocytes using a lymphocyte separation buffer (Cat#562574, BD Biosciences, San Diego, USA) as per manufacturer's protocol. The isolated lymphocytes were stained by surface staining with anti-mouse APC-CD4 (Cat#116014, Biolegend), FITC-CD19 (Cat#152404, Biolegend) and PE/Cyanine7-CD25 (Cat#102016, Biolegend) for 30min at 4°C. After fixation for 45 min at 4 °C using a fixation buffer (Cat#420801, Biolegend), the intracellular staining was performed with anti-mouse PE-IFN-γ (Cat#505808, Biolegend), Brilliant Violet 605-IL-4 (Cat#504126, Biolegend) and Brilliant Violet 421-Foxp3 (Cat#126419, Biolegend) for 1 hr at 4 °C. Finally, flow cytometric analysis was performed using a LSRFortessa™ Cell Analyzer (Cat#647794L6, BD Biosciences, San Jose, USA) and data was analyzed using the v7.6.1 FlowJo software.

The total RNA was extracted using a Trizol reagent kit (Invitrogen, USA). Extracted mRNA was enriched by Oligo(dT) beads fragmented into short fragments and reversely transcribed into cDNA using a NEB Next Ultra RNA Library Prep Kit for Illumina (New England Biolabs, USA). PCR and size selection by agarose gel electrophoresis were applied to construct the cDNA library. The library was sequenced using Illumina Novaseq6000 by Shanghai Wancheng Biotechnology Co., Ltd. The reads obtained from the sequencing machines were aligned to the reference genome using HISAT2. 2.4.

Principal component analysis (PCA) was performed using 'gmodels' package. DESeq2(v1.4.5) was used for differential expression analysis. The genes with false discovery rate (FDR) below 0.05 and fold change ≥ 2 was identified as differentially expressed genes (DEGs). The 'ggplot2' and 'ImageGP' were used to illustrate these DEGs when comparing different groups. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed by Phyper to uncover the underlying molecular mechanism.

Small RNAs from MSC-Exos were obtained using a QIAzol kit (QIAGEN, Shanghai, China). Sequencing libraries were generated from each sample using RNeasy MinElute spin column. The small RNAs were sequenced using next-generation sequencing technology and the raw sequencing data were aligned with the database to identify and analyze the small RNA sequence. RNAhybrid, miRanda and TargetScan were used to accurately predict the target genes of miRNAs. The functional analyses were conducted as described in section 2.12.

SDT buffer (bioWORLD, Inc., USA) was used to extract protein from MSC-Exos. The protein was digested by trypsin according to the filter-aided sample preparation (FASP) procedure as described by Burnner et al., 2022 (32). Liquid chromatography-mass spectrometry (LC-MS) was performed on a Q Exactive mass spectrometer (Thermo Scientific, USA). The LC-MS data was acquired using a data-dependent method. The most abundant precursor ions from the survey scan (300–1800 m/z) were acquired for fragmentation and the data of LC-MS were identified using the MaxQuant 1.5.3.17 software. In addition, the sequences of protein were searched in the InterProScan software to identify the domain signatures.

Immune microenvironment analysis was performed using the R software. The core function utilized CIBERSORT deconvolution algorithm to dig immune cells from a gene expression matrix in samples. Stacked bar plots were used to illustrate the relative abundance of different immune cell subsets in the NC, OVA and PLGA-Exos groups.

The data were statistically analyzed, and figures were generated using GraphPad Prism 6.0 software (GraphPad, La Jolla, CA, USA). All the data were compared using one-way analysis of variance (ANOVA) and then by Tukey's post-hoc test. All the data were expressed as mean ± SD. The p-value < 0.05 was considered statistically significant. All the experiments were repeated in triplicates unless otherwise mentioned.

Exosomes derived from HUC-MSCs (MSCs-Exos) were identified using NTA, TEM and western blot analysis (10). Firstly, the size distribution was determined using NTA, which indicated that more than 90% of the exosomes were between 30 nm to 150 nm in diameter (Figures 1A, B). Secondly, as expected, a double layered membrane structure with cup-shaped morphology was revealed using TEM (Figure 1C). Finally, western blotting exhibited that the exosomes were positive for CD63, CD81 and HSP70 but negative for calnexin (Figure 1D).

All four different sizes of blank PLGA MPs/SMPs, ICG-PLGA MPs/SMPs, and PLGA-Exos MPs/SMPs fabricated using double emulsion method displayed smooth surface morphology and spherical shape (Figure 1E). The average sizes and zeta potentials of blank PLGA MPs/SMPs were 1013.5 nm, 797.3 nm, 400.1 nm and 198.1 nm, and -13 ± 5.81 mV, -12.2 ± 6.13 mV, -7.2 ± 7.32 mV and -4.1 ± 7.01 mV, respectively (Supplementary Figures S2A, B; Table 1). In comparison to blank ones, ICG-PLGA MPs/SMPs had similar sizes but slightly lower zeta potential (-15.6 ± 2.81 mV, -14.0 ± 5.32 mV, -10.1 ± 1.92 mV and -7.2 ± 2.40 mV, respectively) (Supplementary Figures S3A, B), while successfully presenting green fluorescence under fluorescence microscopy (data not provided). Moreover, the PLGA-Exos MPs/SMPs were comparatively bigger in size and lower in zeta potential when compared with blank ones (1073.9 nm, 852.7 nm, 480.8 nm and 250.1 nm and -19.8 ± 3.53 mV, -15.6 ± 1.35 mV, -14.2 ± 2.97 mV and -9.4 ± 3.21 mV, respectively) (Supplementary Figures S4A, B).

The intranasal retention of ICG-Exos and different sized ICG-PLGA MPs/SMPs in Balb/c mice was evaluated using in vivo near-infrared imaging. As demonstrated in Figures 2A, B, strong fluorescence was observed in all sizes of ICG-PLGA MPs/SMPs up to 6 hr after intranasal administration. Compared with 1 µm and 200 nm sized ICG-PLGA MPs/SMPs, the ones in 800 nm and 400 nm sizes exhibited longer retention (120 hr vs. 48 hr). The quick disappearance of 1µm sized PLGA particles might be the larger size, the epithelial cells and macrophages cannot uptake MPs easily and can be cleared quickly from the nasal cavity by mucociliary clearance (33, 34), on the other hand, 200 nm sized PLGA particles were supposed to be uptaken by nasal epithelial cells and macrophages easily but degraded quickly because of very small diameter (35). Whereas, in case of ICG-Exos group strong fluorescence was observed up to 2 hr time point following complete disappearance after 24 hr.

In vitro release of exosomes from 800 nm PLGA-Exos SMPs with varying exosome concentrations (3 µg, 5 µg, 10 µg, and 50 µg/mg PLGA) was measured using the BCA quantification method (Figure 2C). In vitro exosomes release PLGA SMPs was completely concentration based as higher concentration (50 µg, and 10 µg) were released after 9 and 11 days respectively, whereas 5 µg, and 3 µg were released after 7 days.

Considering the above results in various aspects, i.e., in vivo retention time (800nm/400nm > 200 nm/1 µm), and in vitro exosome release (10 µg/50 µg > 3 µg/5 µg), PLGA-SMPs in 800 nm with 10 µg/mg exosomes concentration were determined to be most suitable for AR treatment and used thereafter.

PLGA-Exos labelled with PHK-26 fluorescent dye were co-cultured with RPMI2650 and THP-1 cells. Confocal imaging was performed to visualize the uptake of only exosomes and PLGA-Exos at 4, 12, 24 and 48 hr. The results demonstrated that PLGA-Exos were gradually internalized by RPMI2650 and THP-1 cells (Figure 3A; Supplementary Figure S5). It was observed that the uptake of PLGA-Exos was maximum at 24 hr followed by a decline of fluorescence at 48 hr in both nasal epithelial cells and macrophages (Figure 3A; Supplementary Figure S5). Whereas, maximum uptake of only exosomes was observed at 12 hr followed by a decline of fluorescence at 24 hr and 48 hr, respectively (Supplementary Figure S6). When dye-loaded PLGA particles were incubated with cells at 4 °C, where energy-dependent processes slow down, no intracellular fluorescence is expected. Any intracellular fluorescence obtained at 4 °C incubation would indicate the free dye that leaked out of the particles and diffused into the cells. This is an important issue (dye leakage) and is particularly relevant for in vivo biodistribution/clearance studies and needs to be addressed. In addition, our findings in uptake pattern are inconsistent with other studies (36, 37). Future work needed to be performed to clarify this issue.

LPS-stimulated HNEpCs were used as an in vitro model for the therapeutic effect of PLGA-Exos treatment. The Th1/Th2 balance was determined by the level of Th1-associated cytokines (e.g., IFN-γ and IL-2) and Th2- associated cytokines (e.g., IL-4). PLGA-Exos treatment increased the levels of IFN-γ, IL-2 and IL-10, and decreased those of IL-4 (Figure 3B). Moreover, the results of RT-qPCR also aligned with ELISA results (Figure 3C). These data suggested that PLGA-Exos treatment upregulated Th1 cytokines and downregulated Th2 cytokines in a cellular model.

The allergic response in the blood serum of OVA-challenged AR mice is characterized by elevated Th2 cytokines (e.g., IL-4, IL-13) and antigen-specific IgE, and decreased Th1 cytokines (e.g., IFN-γ, IL-2) (38). In our study, the levels of IL-4, IL-17 and sIgE increased whereas those of IFN-γ and IL-10 decreased in the OVA group, suggesting a successful establishment of AR model. As shown in Figure 4A and Supplementary Figure S7A, treatment with Exos alone and PLGA-Exos significantly decreased the levels of IL-4, IL-17 and sIgE but increased those of IFN-γ and IL-10 at both 2- and 4-weeks. Additionally, a significant increase in the levels of IFN-γ and IL-10 and decrease in IL-4 and sIgE were observed in PLGA-Exos treatment group when compared with Exos alone group at 4-weeks.

Likewise, RT-qPCR was also performed at 2- and 4-weeks to detect the mRNA expression of IL-2, IL-4, IL-10, IL-17 and IFN-γ in the spleen of mice from different treatment groups. The RT-qPCR results were in accordance with the ELISA results as manifested by upregulated IL-2, IL-10, IFN-γ and downregulated IL-4 and IL-17 (Figure 4B; Supplementary Figure S7B). The reduction of IL-17 after PLGA-Exos treatment was more evident at 4 weeks (6 sessions of treatment) when compared to 2 weeks (3 sessions of treatment). After 6 treatments (4-weeks), significantly higher levels of mRNA expression in IFN-γ and IL-10 and lower IL-2 level were observed when PLGA-exos group was compared with Exos group.

The main histopathological manifestation in nasal mucosa of AR is the accumulation and infiltration of various immune cells (39). The number of eosinophils and goblet cells increased in the OVA group but reduced in the Exos alone and PLGA-Exos treatment groups (Figures 4C, D; Supplementary Figures S7C, D). Moreover, a significant decrease in goblet cells was observed in PLGA-Exos group when compared with Exos only group. The results indicated that treatment with Exos and PLGA-Exos significantly inhibited the recruitment of inflammatory cells (i.e., eosinophils and goblet cells).

The immune cells infiltration landscape was portrayed using sequencing results of mouse nasal mucosal tissues. Cibersort analysis of mRNA expression showed that compared with the negative control group (PBS group), there were more mast cells in the OVA group. Moreover, compared with OVA group, monocytes, mast cells and neutrophils were decreased in PLGA-Exos group. These results confirmed the successful establishment of study model and the anti-inflammatory effects of PLGA-Exos therapy. Naïve B cells were also increased in the PLGA-Exos group compared to the OVA group, which may imply more bone marrow mobilization (Figure 4E; Supplementary Figure S8A). In addition, as presented in Figure 4F and Supplementary Figure S8B when compared with OVA group, Th2 cells were decreased whereas Th1 and Tregs were increased significantly in PLGA-Exos group donating excellent therapeutic effect of exosomes.

The cellular changes after PLGA-Exos treatment were assessed by monitoring the counts of Th1/Th2/Tregs/Bregs in the spleen of mice from each treatment group. The results showed increased Th2 and decreased Th1 in OVA-sensitized AR mice. In addition, the percentage of IFN-γ⁺CD4⁺ cells in the spleen were 4.48 ± 0.42, 2.69 ± 0.58, 2.61 ± 0.58, 10.70 ± 0.94 and 14.17 ± 2.45 in PBS, OVA, PLGA, Exos and PLGA-Exos groups, respectively, at 4-weeks (Figure 5A). Similarly, Tregs counts were 4.85 ± 1.08, 2.63 ± 0.45, 3.61 ± 0.71, 7.28 ± 0.98 and 9.47 ± 1.25 in PBS, OVA, PLGA, Exos and PLGA-Exos groups, respectively, at 4-weeks (Figure 5B). Conversely, the population of IL-4⁺CD4⁺ cells increased to 5.28 ± 1.09 in the OVA group but reduced to 2.58 ± 0.77 and 1.99 ± 0.79 in the Exos alone and PLGA-Exos groups after 6 sessions of treatment (Figure 5C). The proportion of Bregs was not statistically different among all the groups at both 2- and 4-weeks (Figure 5D; Supplementary Figure S8D). However, the proportion of Tregs did increase at 2-weeks' time point (Supplementary Figure S9B). These results revealed that Th1/Th2 disbalance was corrected in Exos alone and PLGA-Exos treated AR mice after 6 doses (4 weeks). In addition, PLGA-Exos treatment exhibited better effect than Exos alone group in terms of reestablishing the Th1/Th2 balance (14.17% > 10.70% for IFN-γ, 1.99% < 2.58% for IL-4) and upregulation of Tregs (9.47% > 7.28% for CD4⁺CD25⁺Foxp3⁺). Being consistent with the flow cytometric results, transcriptomic analysis suggested that the proportion of Th2 cells increased in the OVA group and reduced in the PLGA-Exos group. In addition, the changes of γδT cells, which also secrete IL-17, were similar to the trend of decreased IL-17 as proven by ELISA and RT-qPCR (Figures 4C, D, F).

Nucleic acids, especially miRNA, are one of the most important functional cargoes of exosomes. As Figure 6A; Supplementary Figure S9A demonstrated, the proportions of miRNA and length distributions of small RNA were similar among samples. In addition, the sequencing quality of small RNA was acceptable. Figure 6B illustrates the mechanism of microRNA involved in post-transcriptional silencing. pre-miRNA, the precursor of microRNA, is about 70-90 bases in length. After Dicer enzyme digestion, pre-miRNA became mature miRNA with a length of 20~24 nt. The mature miRNA binds to the target mRNA and causes its degradation. miRNAs in MSC-Exos might play multiple roles in the treatment of AR, such as inhibition of cytokine production, activation of T cells during immune response, and regulation of GM-CSF production (Figure 6C). Interestingly, MSC-Exos might not only inhibit Th17 cell lineage commitment and IL-17 mediated signaling pathway, but also attenuate the differentiation of Th0 to Th17 cells by suppressing cytokines that promote Th17 differentiation, such as IL-1, IL-21 and IL-23 (Figure 6C).

The boxplots in Supplementary Figures S10A, B indicated that the extracted RNA was of good quality and integrity. In addition, the distribution of FPKM of mouse nasal mucosal tissues were similar among different samples (Supplementary Figures S10C–E). The protein components of MSC-Exos were detected using non-target mass spectrometry, and a total of 905 quantified proteins were detected (Figure 7A). Firstly, our sequencing results validated the successful establishment of an animal model of AR. The bubble plot revealed multiple inflammatory and immune-related pathways in OVA group (e.g., IgA complex, IgM complex, eosinophil activation involved in immune response, and positive regulation of mast cell activation) (Figure 7B). The chord plot, showing the enriched pathway in different colors, also showed the total gene number of the pathway and the number of up-regulated and down-regulated genes in detail (Figure 7C). Secondly, the differentially expressed genes (DEGs) between the OVA group (AR model) and treatment group (PLGA-Exos) were illustrated by the volcano plots and heatmap (Figures 7D, E). Finally, functional enrichment analysis was performed based on the Gene Ontology (GO) annotation (section 3.11) (Figure 7F).

In order to analyze the potential mechanism of PLGA-Exos treatment of AR, the proteomics data of MSC-Exos, microRNA seq data of MSC-Exos, and transcriptomic data of PLGA-Exos treated mucosal tissue were cross-verified. The domain analysis based on LC-MS showed many immunoglobulin subunits (e.g., Ig V-set domain, Ig C1-set domain and Ig I-set domain) in MSC-Exos (Figure 8A). In addition, the miRNA target gene enrichment analysis also revealed multiple inhibited signaling pathways that are related to immune and inflammatory cytokines (e.g., T cell and B cell activation during immune response, activation of Mitogen-activated protein kinase (MAPK) activity, NK cell activation, mast cell cytokine production and innate immune response) (Figure 8B). These results supported the anti-inflammatory effect of PLGA-Exos in the treatment of AR. The KEGG analysis of PLGA-Exos treated nasal mucosal tissues demonstrated several potentially crucial signaling pathways (e.g., PPAR pathway, metabolic pathway, glycolysis pathway and AMPK pathway) (Figure 8C). In addition, the GO enrichment analysis suggested dozens of important processes such as the metabolism process, immune system process and ATP-dependent activity (Figure 8D). Finally, after intersecting three sets of omics data, we identified two important pathways that might explain the profile changes in immune cells and cytokines after PLGA-Exos treatment, i.e., the activation of PPAR signaling pathway and the disruption of glycolysis pathway (Supplementary Figures S11A, B).