Bone marrow mesenchymal stem cell-derived exosomal microRNA regulates microglial polarization

Cell culture

BV-2 cells and BMSCs from C57BL/6 mice were purchased from Procell (#CL-0493A; Procell, Wuhan, China) and Cyagen Biosciences, Inc. (#MUBMX-01001; Cyagen Biosciences, Inc., Guangzhou, China), respectively. The cells were cultivated in DMEM (Gibco, Billings, MT, USA) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37 °C with 5% CO₂. Non-adherent cells were discarded after 3 days, and the medium was changed every 2–3 days. The cells were sub-cultured after detachment with trypsin-EDTA (Gibco) upon reaching 80–90% confluence.

Exosome isolation

Exosomes were extracted from the BMSCs using the ExoQuick-TC Exosome Precipitation Solution (System Biosciences, Palo Alto, CA, USA). To remove cells and debris, BMSCs were collected and centrifuged at 2,000 g for 10 min. The ExoQuick-TC solution was then combined with the supernatant in a fresh tube. The mixture was incubated overnight at 4 °C, followed by centrifugation at 1,500 g for 30 min. The exosomes were purified by centrifuging the resultant pellet at 100,000 g for 1 h while resuspending it in phosphate-buffered saline (PBS).

Characterization of exosomes

The morphology of the exosomes was examined using transmission electron microscopy (TEM). A drop of exosome suspension was dropped on a copper grid that had been coated with carbon and adsorbed for 5 min. Next, distilled water was utilized to clean the grid before it was negatively stained for 1 min with 2% uranyl acetate. Images were captured using a TEM (JEM-1400 Plus; JEOL Ltd, Tokyo, Japan).

Particle size analysis was performed using a Zetasizer Nano ZS (Malvern Panalytical Ltd, Worcestershire, UK). Exosome suspension was diluted with PBS and measured at room temperature.

The search for exosome markers was carried out using Western blot analysis. Briefly, exosome suspension was lysed with RIPA buffer (Beyotime, Shanghai, China). A BCA Protein Assay Kit (Beyotime) was employed to calculate the protein concentration. SDS-PAGE was performed for separating equal amounts of protein, which was then deposited onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). After being blocked with 5% skim milk for 1 h at room temperature, The membranes were then incubated with primary antibodies TSG101 (1:1,000; ab125011, Abcam, Cambridge, UK), CD9 (1:1,000; ab236630; Abcam), and Calnexin (1:1,000; #2679; Cell Signaling Technology, Danvers, MA, USA) overnight at 4 °C. After washing with TBST, the membranes were treated with secondary antibodies (1:2,000; #7076; CST) for 60 min at room temperature. Protein bands were detected using ECL.

PKH26 staining assay was performed to defined exosomes presence. Briefly, 20 µL PKH26 dye (Sigma-Aldrich) was combined with 200 µL exosome suspension, incubating for 10 min. Then, process was then halted by adding 800 µL PBS to the mixture. After being collected by ultracentrifugation at 100,000 g for 70 min, the tagged exosomes were resuspended in 200 µL PBS. The presence of PKH26-labeled exosomes was examined through fluorescence microscope (Nikon, Japan).

Oxygen-glucose deprivation (OGD) modeling and exosome treatment

Six-well plates with a seeding density of 4 × 10⁵ BV-2 cells each were used for the 20 h of culturation. After that, the media was changed to glucose-free DMEM containing 2 mM sodium pyruvate and 1% FBS. The cells were then incubated for 4 h at 37 °C in a humidified environment that included 1% O₂, 5% CO₂, and 94% N₂. After OGD, the glucose-free medium switched out for normal DMEM, and cells were developed for 4 h under normoxic conditions. After modelling, cells were treated with 15 µg/mL exosome. Cells without any treatment were set as control group.

Flow cytometry for apoptosis

To detect cell apoptosis using flow cytometry, cells were gathered by centrifugation and then homogenized in 1 × Binding buffer at a density of 2  × 10⁵ cells/mL. Annexin V-FITC (Sangon Biotech, Shanghai, China) was added and the mixture was incubated at 4 °C for 10 min. After washing and resuspending cells in 1 × Binding buffer, propidium iodide was added. Cells were immediately examined through flow cytometry for double staining detection of apoptosis.

Flow cytometry for cell phenotype

For flow cytometry analysis of the BV-2 cell phenotype, cells were harvested and resuspended in 100 µL of PBS. Anti-mouse CD40-FITC (#124607; Biolegend, San Diego, CA, USA), CD86-PE (#105007, Biolegend), and CD206-FITC (#141703, BioLegend) antibodies were incubated with cells at 4 °C for 30 min. Cells were resuspended in 500 µL of PBS after being washed twice with PBS. The samples were then evaluated using a flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

miRNA identification and analysis

RNA was extracted from BV-2 cells in OGD or OGD + exo groups. Using a NanoDrop spectrophotometer, concentration and purity of the RNA were assessed, and RNA integrity was confirmed through gel electrophoresis. Using Illumina HiSeq platform, six small RNA libraries were built and sequenced. By eliminating adaptor sequences, poor-quality reads, and reads smaller than 18 nt or bigger than 30 nt from the raw sequencing data, clean data was obtained. Bowtie was used to map the clean reads to the reference genome (Mus musculus), and miRDeep2 was utilized to determine the miRNA expression levels.

Differentially expressed miRNA analysis

The edgeR package in R was used to carry out a differential expression analysis. Absolute P < 0.05 and —log2FC— ≥ 1 were applied to define which miRNAs were differentially expressed (Yang et al., 2022). Volcano plots were generated to visualize the distribution of differentially expressed miRNAs. To investigate the link between various samples, a hierarchical clustering analysis was carried out using the R heatmap package.

Functional annotation and enrichment analysis of miRNA targets

We used TargetScan, miRanda, and DIANA-microT algorithms to search for differentially expressed miRNA-related genes. Using the DAVID tool, functional annotation and enrichment analysis were performed on the anticipated targets. The predicted targets with score = 1 and validated target genes using qRT-PCR were subjected to enrichment analyses of Gene Ontology (GO, http://www.geneontology.org/), Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg/), disease ontology (DO, http://www.disease-ontology.org), and Reactome (http://reactome.org/). Enrichment results were visualized using bubble plots. Adjust p value < 0.05 was used as the cutoff criterion.

Cell transfection

BV-2 cells were cultured in 6-well plates and transfected with miRNA mimics, inhibitors, or plasmids using Lipofectamine 3000. The miRNA mimics and inhibitors sequences were supplied by GenePharma (Shanghai, China). Plasmids used in this study were psiCHECK-2-Traf6 (wild-type 3′UTR), psiCHECK-2-Traf6-mut (mutated 3′UTR), oe-Traf6 (overexpression of Traf6), and sh-Traf6-1 (shRNA against Traf6-1), sh-Traf6-2, and sh-Traf6-3.

Dual luciferase reporter gene assay

BV-2 cells were collected after 48 h of co-transfection and washed with pre-chilled PBS. Careful washing was performed to prevent cell detachment and loss. 5 × PLB was diluted to 1 × PLB with deionized water. 100 µL 1 × PLB was added to each well of a white, opaque 96-well plate for cell lysis, and shaken on a shaker for 15 min at room temperature. After 20 µL cell lysate was added to each well, pre-mixed Stop&Glo Reagent (100 µL) was added, and luminescence data were measured after a 2 s delay. Luciferase activity detection was performed under light-avoiding conditions.

RT-qPCR analysis

Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA). The RNA concentration was determined using a NanoDrop spectrophotometer, and cDNA was synthesized using a reverse transcription kit (Thermo Fisher Scientific). The expression levels of mmu-miR-6238, mmu-miR-3102-3p, mmu-miR-6984-5p, mmu-miR-495-3p, mmu-miR-3095-3p, mmu-miR-6975-5p, mmu-miR-7013-5p, mmu-miR-7652-5p, and mmu-miR-146a-5p were detected by quantitative real-time PCR (qPCR) using the PowerUp SYBR Green Master Mix (Thermo Fisher Scientific) and specific primers (Table S1). The qPCR conditions were as follows: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 20 s. U6 snRNA was used as an internal control. The expression levels were calculated using the 2^−ΔΔCt method.

Statistical analysis

Statistical analysis was carried out using GraphPad Prism 8.0 software. Data are presented as mean ± standard deviation, and were analyzed using one-way ANOVA followed by Tukey’s test or unpaired Student’s t-test. P value of less than 0.05 was considered statistically significant.

Exosomes were successfully isolated from BMSCs in mice

We isolated exosomes from BMSCs and characterized them using various methods. TEM images of the isolated exosomes revealed their typical cup-shaped morphology, indicating successful isolation (Fig. 1A). Particle size analysis demonstrated a relatively homogeneous population of exosomes, with an average particle size of 135.7 nm and concentration of 6.0 × 10¹⁰ particles/mL (Fig. 1B). Western blotting detected presence of exosome markers TSG101 and CD9. Absence of Calnexin further indicated the purity of the exosome preparation (Fig. 1C). Additionally, a tracing assay using the PKH26 dye showed strong red fluorescence signals in the labeled exosomes, providing further evidence for the presence of exosomes in the isolated sample (Fig. 1D).

BMSCs-derived exosomes modulated microglial conversion

We then investigated the function of BMSCs-derived exosomes on BV-2 microglial cell polarization. Results revealed that BV-2 cells displayed a typical activated phenotype with an amoeboid shape and retracted processes (Fig. 2A). In comparison with the control cells, apoptosis in OGD-exposed cells were elevated, and there was a significant increase of apoptosis in exosome-treated OGD cells compared with untreated cells (P < 0.0001, Fig. 2B). Furthermore, the expression of CD40 and CD86 (M1-like microglia markers) and CD206 (M2-like microglia marker) was assessed. Positive ratio of CD40 and CD86 was increased while that of CD206 was decreased in the OGD-exposed cells (P < 0.05, Fig. 2C). Exosome treatment led to significant upregulation of CD40 and CD86 while downregulation of CD 206 (P < 0.05, Fig. 2C).

Biology information analyses of miRNA in microglia

miRNA identification

RNA was extracted from OGD-exposed microglial cells treated with or without exosomes and sequencing analyses was performed. Using the Illumina platform, the base quality scores were determined, and the base recognition accuracy rate was found to be greater than 90% (Table 1). The sequencing base quality in this study was primarily consistent with the Q30 criterion (Fig. 3A). We then employed miRDeep2 to identify miRNA expression in the samples. A total of six miRNAs were identified, including mmu-let-7c-2-3p, mmu-let-7a-1-3p, mmu-let-7a-5p (mmu-let-7a-1 and mmu-let-7a-2), mmu-let-7a-2-3p, and mmu-let-7b-3p. Subsequently, we performed hierarchical clustering analysis on the samples, and the dendrogram displayed the Euclidean distances between different samples, with a truncation height ranging from 10,000 to 80,000 (Fig. 3B). Principal component analysis (PCA) and heatmap visualization revealed substantial heterogeneity between the two different treatment groups (Fig. 3C and 3D).

Table 1:

Illumina sequencing quality value.

Phred score	Incorrect base identification ratio	Correct base identification ratio	Q.score
10	1/10	90%	Q10
20	1/100	99%	Q20
30	1/1,000	99.9%	Q30
40	1/10,000	99.99%	Q40

DOI: 10.7717/peerj.16359/table-1

Differentially expressed miRNA

We analyzed the length distribution of the identified expressed miRNAs, revealing a predominant length of approximately 21 bp (Fig. 4A). Differential expression analysis on these miRNAs was performed using a selection criterion of P < 0.05 and —log2FC— ≥ 1. The volcano plot showed that exosome treatment led to significant changes in miRNA expression in OGD-exposed microglial cells, with 10 upregulated miRNAs and 33 downregulated miRNAs (Fig. 4B). The differentially expressed miRNAs were visualized using a heatmap and a bar chart; The red color is upregulated miRNAs, while the blue color represents downregulated miRNAs in the heatmap (Fig. 4C). The bar chart displayed mature and precursor names of these differentially expressed miRNAs (Fig. 4D).

BMSCs-derived exosomes mmu-miR-146a-5p modulated microglial conversion

Based on the sequencing results, we selected mmu-miR-6238, mmu-miR-3102-3p, mmu-miR-6984-5p, mmu-miR-495-3p, mmu-miR-3095-3p, mmu-miR-6975-5p, mmu-miR-7013-5p, mmu-miR-7652-5p, and mmu-miR-146a-5p for validation. The results demonstrated that in OGD-exposed microglial cells, only mmu-miR-6984-5p expression was upregulated after exosome treatment, but the change was not statistically significant (Fig. 5). Expression of mmu-miR-3102-3p, mmu-miR-3095-3p, mmu-miR-6975-5p, mmu-miR-7013-5p, mmu-miR-7652-5p, and mmu-miR-146a-5p was significantly decreased (P < 0.05), with mmu-miR-146a-5p showing the most prominent downregulation (P < 0.0001, Fig. 5). Therefore, we chose mmu-miR-146a-5p for further analysis.

To explore the role of mmu-miR-146a-5p on microglial conversion, OGD-exposed BV-2 cells were transfected with miR-146a-5p mimics or inhibitors. The cell morphology diagram shows that the treated cells are growing well (Fig. 6A). After upregulated miR-146a-5p, CD40 positive ratio was increased compared with mimics NC groups (P < 0.001, Fig. 6B). Oppositely, miR-146a-5p inhibitors reduced the number of CD40 and CD86 cells compared to inhibitor NC groups (P < 0.05, Fig. 6B). The number of CD206-positive cells showed no significant changes after transfection (Fig. 6B).

BMSCs-derived exosomes mmu-miR-146a-5p modulated microglial conversion by inhibiting Traf6

Subsequently, we validated the downstream targets of mmu-miR-146a-5p. mmu-miR-146a-5p mimics was transfected into OGD BV-2 cells. Compared to miR NC, mmu-miR-146a-5p expression significantly increased after transfection (P < 0.0001), indicating successful transfection of mmu-miR-146a-5p into cells (Fig. 7A). In contrast, the Traf6 mut group did not affect mmu-miR-146a-5p expression compared to psiCHEK-2-Traf6 (Fig. 7A). Dual-luciferase reporter assay revealed that mmu-miR-146a-5p reduced luciferase activity (P < 0.0001), suggesting that mmu-miR-146a-5p could target Traf6 3′UTR (Fig. 7B). We then verified the regulatory relationship between mmu-miR-146a-5p and Traf6 in BV-2. Results showed that transfection of mmu-miR-146a-5p mimics decreased Traf6 expression in BV-2, while mmu-miR-146a-5p inhibitors upregulated Traf6 expression (Fig. 7C), indicating a negative regulatory relationship between mmu-miR-146a-5p and Traf6.

We further explored role of the mmu-miR-146a-5p/Traf6 axis on microglial cell polarization. qPCR results demonstrated that sh-Traf6-1 and sh-Traf6-2 significantly downregulated Traf6 expression (P < 0.01, Fig. 7D) and oe-Traf6-1 significantly elevated the expression of Traf6 (P < 0.01, Fig. 7E). Therefore, BV-2 cells were then co-transfected with mmu-miR-146a-5p mimics and oe-Traf6-1 or mmu-miR-146a-5p inhibitor and sh-Traf6-1/2, and all cell groups showed good growth (Fig. 7F). No significant changes were found in CD40 positive ratio (Fig. 7G). miR-146a-5p mimics and oe-Traf6 increased the number of CD86-positive cells while decreasing CD206-positive cells (P < 0.0001, Fig. 7G). In contrast, mmu-miR-146a-5p inhibitor and sh-Traf6-1/2 reduced the number of CD86-positive cells while increasing CD206-positive cells (P < 0.0001, Fig. 7G).