DPSC culture
The use of teeth and the study protocols were approved by the Institutional Ethics Committee Board of the Guanghua School of Stomatology, Sun Yat-sen University (KQEC-2019-06). Informed consent was obtained from the donor patient family. We obtained exfoliated third molars from healthy donors 18-24 years in age who provided informed consent. DPSCs were isolated as previously described. Briefly, the pulp in the chamber and canals was gently removed using various instruments and cut into small fragments. The dental pulp tissue was added to a solution of collagenase type I (4 mg·mL−1; Sigma-Aldrich, MO, USA) and dispase (4 mg·mL−1; Sigma-Aldrich). The solution was placed at 37 °C for 30 min, with the tube inverted at 10-min intervals. The single-cell suspension was cultured in low-glucose Dulbecco’s modified Eagle’s medium (Gibco, Grand Island, NY, USA) containing 20% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Sigma-Aldrich) in a 60 mm culture flask (Corning, Cambridge, MA, USA) or 60 mm ultra-low-attachment culture dish (Corning) at 37 °C in a 5% CO2 humidified atmosphere.
Isolation of DPSC-exos
To prepare exosome-depleted fetal bovine serum (FBS), FBS was ultracentrifuged at 4 °C at 120 000 × g for 18 h. The supernatant was filtered using a 0.22-μm syringe filter and stored at 4 °C. For 2D-exo collection, as described in a previously published protocol,25 cells were cultivated in Dulbecco’s modified Eagle medium containing 10% exosome-depleted FBS for 2 days. The same amount of MSCs used for 2D culture was used for inoculation in an ultra-low-attachment culture dish (Corning) for 3D culture with the same complete medium used for 2D culture for 2 days. After the culture supernatants were collected, exosomes from the 2D and 3D cultures were isolated through multistep centrifugation as previously described.25 To erase debris and dead cells, we first centrifuged the supernatant at 300 × g for 10 min, 2 000 × g for 20 min, and 10 000 × g for 30 min. Then, we ultracentrifuged the supernatant at 100 000 × g for 90 min and washed the pellet with PBS before centrifugation at 100 000 × g for 90 min (Optima-90 K, Beckman Coulter). At last, we resuspended the pellets in PBS.
Exosomes were characterized by TEM, NTA, and western blot analysis of exosome markers. The morphology and ultrastructure of the exosomes were analyzed using TEM (JEOL, Tokyo, Japan). The yields of exosomes from 1 × 107 MSCs cultivated in the 2D and 3D culture systems were quantified by a Micro Bicinchoninic Acid Protein Assay Kit (CWBio, Beijing, China) and NTA according to the manufacturer’s recommended protocol. CD63 and TSG101 protein levels were determined using Western blot analysis. The therapeutic effects of 3D-exos and 2D-exos were examined in a mice model of periodontitis and DSS-induced colitis in vivo and in vitro.
To understand the distribution of the exosomes after injection, exosomes were stained with DiO (Invitrogen, USA), a fluorescent dye that can label the plasma membrane. At 24 h after the first injection of exosomes into the palatal gingiva near maxillary left second molar of DSS-P mice, gingivae from the maxilla and colons were collected. They were either fixed in 4% PFA and cut into frozen sections for confocal fluorescence analysis or treated with enzymatic digestion (RPMI-1640 medium containing 4 mg/mL dispase and 3 mg·mL−1 collagenase type I) for flow cytometric analysis. The digest solution was then filtered through a 70-μm cell strainer (Biologix Research Company, USA) to obtain a single-cell suspension. The ex vivo fluorescence intensity was analyzed by flow cytometry.
Animals
Six- to eight-week-old male C57BL/6J mice were purchased from the National Resource Center of Model Mice (Nanjing, China). With the importance of IBD alleviation accounted for in our study, the sample size was estimated by using one-way analysis of variance F-tests with the mean difference of the DAI of the colon representing the severity of colitis among the PBS-treated group, 2D-exo-treated group and 3D-exo-treated group. We estimated at least 6 mice are required to reach a statistical power of 90% with type I error at 5%. The sample size was calculated using PASS software. All mice were maintained under specific pathogen-free conditions in an environmentally controlled clean room at the Center for Experimental Medicine. All experiments were performed with the approval of the Animal Care and Use Committee of Sun Yat-sen University (SYSU-IACUC-2019-000096). A total of 5 experimental groups are described and listed in Table S3. Each group was comprised of 6 mice.
Ligature-induced periodontitis model
The mouse was anaesthetized with 4% isoflurane flow (RWD, Shenzhen, Guangdong, China). Then, a ligature (5-0 silk) was placed around the maxillary left second molar from day 0 to day 14, as described previously.25 After 14 days, PBS, 2D-exos, 3D-exos, NCI-3D-exos or miR1246I-3D-exos (50 μg per mouse) was injected into the palatal gingiva of the experimental mice over a period of 14 days (once every 7 days). The mice were sacrificed and analyzed 14 days after PBS or exosome treatment.
DSS-induced colitis model
After the ligature had remained in place for 14 days, mice received 1.5% DSS (molecular mass ~40 000 kD; Sigma-Aldrich, UK) in their drinking water (tap water) for 14 days, after which they received normal tap water for a 2-day recovery period. The DAI was applied to measure the severity of colitis, which was scored as follows, as previously described: body weight loss (0, none; 1, 1%–5%; 2, 5%–10%; 3, 10%–20%; 4, >20%), stool consistency (2, loose stools; 4, diarrhea), and bleeding (2, positive haemoccult; 4, gross bleeding).85 The scores of individual measurements were added to calculate the DAI. Mice were sacrificed on day 28. Colons were collected, and their lengths were measured.
Histological staining and histopathological evaluation
After collection, the maxillae and colons were fixed in 4% PFA. The colons were then paraffin-embedded and sectioned, followed by H&E staining. At the same time, the maxillae were decalcified in 0.5 mol·L–1 EDTA for 3 weeks, dehydrated in a 30% sucrose solution and embedded in Tissue-Tek optimum cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA, USA). They were cut with a freezing microtome (Leica CM1900, Germany) and stained with H&E (Servicebio, Wuhan, China) as well as tartrate-resistant acid phosphatase (TRAP, 387A-1KT, Sigma-Aldrich). The distance between the cementoenamel junction and alveolar bone crest (CEJ-ABC distance) of the sections stained by H&E was measured to evaluate bone loss. TRAP-positive multinucleate cells were considered osteoclasts and a sign of bone resorption.
For assessment of colon inflammation, histological scores were determined by the following criteria: the severity of epithelial/crypt loss (score, 0-4) and the extent of inflammatory cell infiltration in the lamina propria (score, 0-4). Each score was multiplied by a factor representing the percentage of the colon involved (1, 0%–25%; 2, 26%–50%; 3, 51%–75%; 4, 76%–100%), and the scores on the individual measures were then added to calculate the overall histological score for each sample, as previously described.41
Micro-CT
We collected the mouse maxillary bones from mice under each experimental condition and used them for three-dimensional high-resolution micro-CT analysis (Scano Micro-CT, μCT50, Switzerland). The key parameters were set as follows: 70 kV, 110 mA, and 7-μm increments. Three-dimensional microstructural image data were reconstructed and analyzed by using image analysis software (Mimics Research 21.0, Materialize, Belgium). The CEJ-ABC distance was measured at six sites, including mesial, middle, and distal points of both the buccal and palatal sides, and the mean CEJ-ABC distance was then calculated.
T lymphocyte culture and differentiation
The mice were subjected to cervical dislocation following anesthesia with isoflurane to reduce their pain. They were then sterilized in 70% ethanol. The spleen was dissected by aseptic operation and then physically homogenized into pieces. The crushed spleen tissue was dissolved in a small amount of PBS solution and filtered with a 0.45-μm sieve. The filtered solution was added to 3 times the volume of RBC lysis buffer (CWBio), incubated on ice for 15 min, and vortexed slightly in the tube at 5-min intervals. After centrifugation at 450 × g for 5 min, the precipitated cells were suspended in RPMI-1640 medium (Gibco).
Naive CD4+ T cells were harvested from the spleen via the MojoSort™ Mouse CD4 Naive T-Cell Isolation Kit (BioLegend, San Diego, CA, USA) and seeded into 48-well plates (3 × 106 per well) precoated with anti-CD3 (5 mg·mL−1) and anti-CD28 (2 mg·mL−1) solutions. To investigate the roles of miR-1246 in the expression of Nfat5 in CD4+ T cells, CD4+ T cells were incubated with miR-1246 mimics (RiboBio, Guangzhou, China) or inhibitors (RiboBio) for 24 h.
To examine the differentiation propensity, CD4+ T cells were cultured in Th17 differentiation media or Treg differentiation media. The Th17 differentiation medium contained 1.0 ng·mL–1 TGF-β, 30 ng·mL–1 IL-6, 20 ng·mL–1 IL-1β, 20 ng·mL–1 IL-23, 10 ng·mL–1 anti-IL-4 and 10 mg·mL–1 anti-IFN-γ. The Treg differentiation medium contained 50 ng·mL–1 IL-2 and 5 ng·mL–1 TGF-β. All cytokines were purchased from R&D Systems. T cells were further cultured with PBS, 2D-exos (10 μg of exosomes per 105 cells), 3D-exos (10 μg of exosomes per 105 cells), or 3D-exos with the miRNA inhibitor (RiboBio) in differentiation medium for 72 h and restimulated with PMA (Sigma-Aldrich) and ionomycin (Sigma-Aldrich) in the presence of brefeldin A (BD Biosciences, CA, USA) for 5 h before further analysis of intracellular cytokines. These cells were maintained in a standard 37 °C CO2 (5%) incubator. miRNA mimics and inhibitors were transfected with Lipofectamine 2000 (Invitrogen, CA, USA).
RNA-seq
Total RNA was isolated from the gingivae of 3D-exo-treated, 2D-exo-treated, and PBS-treated mice with NucleoZOL reagent (Gene Company Limited, Hong Kong, China). RNA-seq libraries were prepared with the NEBNext® Ultra™ RNA Library Prep Kit (NEB, USA), followed by sequencing with Illumina Sequencing (HiSeq, Fasteris SA, Switzerland) at Novogene Co. Ltd. (Beijing, China). Small RNAs were isolated from the 3D-exos and 2D-exos for miRNA-seq. The miRNA-seq libraries were prepared and sequenced with an Illumina HiSeq platform at RiboBio Co. Ltd. (Guangzhou, China).
Feature counts were used to calculate read counts, and DESeq2 was used to analyse the differential expression of genes. Genes with a corrected p-value ≤ 0.05 and an absolute log2 (fold-change) > 2 were considered differentially expressed. GO enrichment analysis of the top 200 DEGs was performed with the Database for Annotation, Visualization and Integrated Discovery (DAVID).
RNA extraction, reverse transcription, and RT-qPCR
Total RNA was extracted from the colons, the gingivae and CD4+ T cells with NucleoZOL reagent (Gene Company Limited) and was then reverse-transcribed into cDNA using PrimeScript RT Master Mix (TaKaRa, Ltd, Osaka, Japan). Real-time polymerase chain reaction (RT-qPCR) was performed to measure gene expression levels in a Bio-Rad CFX96™ detection system (Roche, Sweden) with Hieff qPCR SYBR Green Master Mix (Yeasen, Shanghai, China). Small RNA was isolated with a miRNA isolation kit (Qiagen, Hilden, Germany), and cDNA was prepared with a miRNA reverse transcription kit (Shenggong, Shanghai, China). RT-qPCR was performed to measure the expression level of genes in a Bio-Rad CFX96™ Detection System (Roche) with Hieff qPCR SYBR Green Master Mix (Yeasen). U6 was applied as the internal reference. The primers used in the process are shown in Supplementary Table S1.
Western blot analysis
After lysis in RIPA buffer (Millipore, Billerica, MA, USA) on ice for 30 min and centrifugation for 15 min, proteins were extracted from CD4+ T cells and tissues. A BCA protein assay kit (CWBio) was used to detect the total protein concentration. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used to separate the proteins, which were then transferred onto poly (vinylidene fluoride) membranes (Millipore). The membranes were blocked in buffer containing 5% bovine serum albumin at room temperature for 30 min and then incubated with the primary antibodies at 4 °C overnight followed by horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. The antibodies used in the process are shown in Supplementary Table S2.
Tissue extraction and single-cell preparations
The gingival tissues were isolated and cut into small pieces, followed by enzymatic digestion with RPMI-1640 medium containing 4 mg/mL dispase and 3 mg/mL collagenase type I for 60 min at 37 °C. The digest solution was then filtered through a 70-μm cell strainer (Biologix Research Company, USA) to acquire a single-cell suspension.
Flow cytometry
For surface antigen staining, single-cell suspensions were stained with the indicated antibodies at 4 °C for 30 min. Dead cells were removed using Zombie viability dye (BioLegend). For intracellular antigen staining, cells were fixed in fixation buffer (0.5 mL per tube; BioLegend) for 20 min after staining with surface antigen antibody. Then, they were stained with predetermined intracellular antibodies at 4 °C for 30 min. The gating strategies are shown in Figs. S3, S4, and S6. Data were acquired using CytoFlex (Beckman CytoFlex, USA) and analyzed with FlowJo V10.0 (TreeStar, Ashland, OR, USA).
Luciferase activity assay
miR-1246-binding sites on the 3’-untranslated region (UTR) of Nfat5 were recognized by TargetScan online bioinformatics software (http://www.targetscan.org). Dual luciferase activity assay was then measured to confirm the relationship between miR-1246 and Nfat5. In brief, Nfat5 recombinant plasmids (Nfat5-WT and Nfat5-Mut) and mimic NC or miR-1246 were transfected into 293T cells. The luciferase activity was ultimately analyzed using a dual luciferase assay kit (Promega Corporation, USA) following the manufacturer’s instructions.
Statistics
All data are presented as the mean ± SEM of at least three independent experiments. After normality testing, all data were analyzed by 2-tailed unpaired Student’s t-test, the Kruskal–Wallis test (nonparametric samples) or 1-way ANOVA (parametric sample) followed by either Dunn’s test (nonparametric samples) or Tukey’s test (parametric samples) as the post hoc test. All statistical analyses were performed with GraphPad Prism software.
