Adipose-derived MSCs (ADSCs) were obtained from C57BL/6J mice following the procedure previously reported (26). The isolated cells were resuspended in Dulbecco’s modified eagle medium (DMEM)/F12 (Gibco) containing 10% exosome-depleted fetal bovine serum (FBS, Gibco) (160,000 ×g at 4 °C, overnight) (27), and 100 U/mL penicillin/streptomycin (Gibco). The immunophenotype of MSCs were detected using flow cytometry. And their multilineage differentiation was confirmed by osteogenic and adipogenic differentiation. Cells at 3^rd to 5^th passages were used for further experiments. Preconditioning with LPS (Sigma) was performed once the cells reached 70–80% confluence. MSCs were cultured in fresh medium supplemented with LPS (1 μg/mL) or PBS for 48 h before supernatant collection.

Exosomes were isolated from the supernatants of MSCs and LPS-treated MSCs using differential ultracentrifugation as previously described (28). Cell supernatants were centrifuged at 350 ×g for 10 min to remove dead cells, 2,000 ×g for 10 min to remove cell debris, and 10,000 ×g for 30 min at 4°C to remove microvesicles. Next, the supernatant was filtered through a 0.22 μm filter and centrifuged at 120,000 ×g for 70 min at 4 °C. The pallets were re-suspended in PBS and centrifuged at 120,000 ×g for 70 min at 4 °C again. The final pellets were re-suspended with PBS and stored at -80 °C for further experiment.

The morphology of the isolated exosomes was observed with transmission electron microscopy (TEM) (Hitachi, HT7700, Japan). The particle size distribution and concentration of exosomes were analyzed with nanoparticle tracking analysis (NTA) (Malvern, NS300, UK). The specific markers CD9, CD63, and TSG101 were detected using western blot. The protein concentration was quantified using a BCA protein assay (Applygen).

C57BL/6J male mice, 6–8 weeks old, were purchased from Vital River Laboratory Animal Technology (Beijing, China). All animal experiments carried out in accordance with the National Institutes of Health guide for the care and use of Laboratory animals, and approved by the Animal Care and Use Committee of the General Hospital, Tianjin Medical University.

The mice were randomly divided into groups and then intraperitoneally injected with LPS (12.5 mg/kg), and the mice in the control group were given the same dose of PBS. Two hours after the LPS injection, mice were treated with LPS-pretreated exosomes (LPS-Exo, 200 μg/mouse), normal exosomes (Exo, 200 μg/mouse), or PBS through the tail vein injection. The same dose of PBS was administered to the mice in the control group. Status of survival of the mice were monitored every 6 h. Serum samples were collected 12 h after LPS injection. Mice were euthanized 48 h after LPS injection, and organ damage was examined. Mice were anesthetized with isopentane and sacrificed by cervical dislocation.

BMDMs were obtained from C57BL/6J mice as previously described (29). The isolated cells were cultured in DMEM with 10% FBS, 100 U/ml penicillin/streptomycin, and 10 ng/mL M-CSF (Pepro Tech) in 5% CO₂ at 37 °C for 7 days. The BMDMs were stimulated with LPS (1 μg/mL) to establish an in-vitro sepsis model and then treated with Exo, LPS-Exo, or PBS.

To trace the internalization of exosomes by BMDMs, exosomes were labeled with DiI (MedChemExpress), washed with PBS, ultracentrifuged and resuspended in PBS. After being incubated with DiI-labeled exosomes for 12 h, the BMDMs were rinsed with PBS, fixed for 30 min in 4% paraformaldehyde, and stained with DAPI. Finally, the cellular uptake of exosomes was detected using a laser-scanning confocal microscope (Zeiss LSM 800, Germany).

To investigate the molecular cargo of LPS-Exo which improved sepsis. LPS-Exo was treated with proteinase K (100 μM/mL, Sigma) for 30 min at 37 °C or with RNase A (100 μM/mL, Sigma) for 15 min at 37 °C. Exosome-free proteins or RNA was used for further experiments.

Total RNA of the exosomes was extracted using the miRNeasy Serum/Plasma Advanced Kit (Qiagen). The final ligation PCR products were sequenced using the Illumina NovaSeq 6000 platform (Echo Biotech, Beijing, China). miRNAs in Exo and LPS-Exo were profiled in three biological replicates. The expression differences between Exo and LPS-Exo were analyzed with a t-test. Those with a fold change > 1.5 and a p-value < 0.05 were regarded as significantly different expressed.

miRanda and RNAhybrid were used to predict the target genes of the differentially expressed miRNAs. GO categories of the predict target genes were performed using the DAVID bioinformatics database. Pathway analysis was performed using the KEGG database and the 20 most enriched signaling pathways were listed to identify the most relevant signaling pathways.

Total RNA of the exosomes was extracted using the miRNeasy Serum/Plasma Advanced Kit. Total RNA of the cells was extracted using the TRIzol reagent. The isolated RNA was reverse transcribed using a cDNA synthesis kit (Qiagen). PCR was performed in a CFX Connect (BIO-RAD) system using SYBR Green PCR Master Mix (Qiagen), using GAPDH or U6 as an internal control. The primer and probe sequences are listed in Table 1 .

miR-150–5p inhibitor and negative control (NC) were purchased from GenePharma and diluted with DEPC water. For miR-150–5p inhibition, BMDMs were transfected with miR-150–5p inhibitor or NC at a concentration of 50 nM using Lipofectamine3000 (ThermoFisher) before treated with Exo or LPS-Exo. MSCs were transfected with miR-150–5p inhibitor or NC at a concentration of 50 nM using Lipofectamine3000 for 6 h and then cultured in fresh medium with LPS (1μm/ml) for further 48 h (30). The exosomes were extracted after transfection as previously described. The transfection efficiency was evaluated by qRT-PCR.

The concentration of TNF-a, IL-6, and IL-10 in serum and the supernatant of cells was determined using commercial ELISA kits (Biolegend). The levels of NO in the supernatant of the cells were determined using NO kits (Beyotime).

Lung, liver, and kidney tissues were obtained and fixed with 10% paraformaldehyde, embedded in paraffin, and then cut into 5-μm sections. After that, the sections were stained with hematoxylin and eosin for evaluation. Injury areas were analyzed as previously described (31).

Cells and exosomes were isolated and lysed with ice-cold radioimmunoprecipitation assay lysis buffer containing protease and phosphatase inhibitors (Applygen). Protein concentration was tested by a BCA assay. Equal amounts of proteins from each sample were used for immunoblot analysis as previously described (32). Data were analyzed using the ImageJ software. The primary antibodies employed for western blotting are listed in Table 2 .

HEK293T cells (1×10⁵ cells/well) were seeded into a 12-well plate 24 h before transfection. Then, 1 μg pmir-GLO vectors containing a wild-type (WT) or mutant (MUT) fragment of the IRS1 3’UTR, 50 nM miR-150–5p mimic or NC were co-transfected using Lipofectamine3000. Luciferase activity was detected using a Dual-Glo luciferase assay system 24 h after transfection.

Data were analyzed using GraphPad Prism 9.0 and presented as mean ± standard deviation. A t-test was used to compare the significance between two groups. Survival rates between the groups were compared using the log-rank test. A p-value < 0.05 was considered to be significant.

We first obtained MSCs from the subcutaneous adipose tissue of C57BL/6 mice. The MSCs were exhibited a spindle-shaped morphology under the light microscope ( Supplementary Figure S1A ). Additionally, these cells were capable of differentiating into osteoblasts and adipocytes ( Supplementary Figures S1B, C ). FACS analysis demonstrated that these cells were positive for CD29 and CD44, and negative for CD45 ( Supplementary Figure S1D ). These data indicated that we successfully obtained ADSCs.

We then collected the supernatants of the LPS-pretreated or non-pretreated MSCs and extracted the exosomes using differential ultracentrifugation. TEM, NTA, and western blotting were performed to verify the characteristics of exosomes. The TEM images showed that LPS-Exo and Exo possessed a typical cup-like appearance with double-membrane structures ( Figure 1A ). NTA results showed that the diameters of Exo and LPS-Exo were between 50 and 200 nm ( Figure 1B ). Western blotting indicated that the specific exosomal markers CD9, CD63, and TSG101 were expressed in Exo and LPS-Exo ( Figure 1C ). These results revealed that LPS-Exo and Exo share identical exosomal characteristics, including shape, size, and biomarkers. However, using BCA and NTA assays, we found significantly higher levels of proteins and particles in LPS-Exo than in Exo, confirming that LPS preconditioning promoted the secretion of exosomes from MSCs ( Figures 1D, E ).

We injected septic mice via the tail vein with LPS-Exo (200 μg/mouse), Exo (200 μg/mouse), or PBS, and the effect on survival rate was evaluated. Septic mice treated with LPS-Exo and Exo exhibited significantly increased survival rates compared to mice treated with PBS (53.3% vs. 6.7% and 33.3% vs. 6.7%, respectively; p < 0.05), and the survival rate in the LPS-Exo group was the highest ( Figure 2A ). As expected, levels of IL-6 and TNF-α in the plasma were increased in the setting of sepsis. Both LPS-Exo and Exo inhibited the increase of these pro-inflammatory cytokines. After exosome treatment, IL-6 and TNF-α levels decreased by 91.4% and 76.8%, respectively, in the LPS-Exo group as well as by 43.9% and 32.1%, respectively, in the Exo group. The reduction of cytokines levels in the LPS-Exo group was significantly greater than that in the Exo group ( Figures 2B, C ).

Multiple organ dysfunction syndrome (MODS) is the most common serious complication of sepsis, associating with high mortality rates. The effects of LPS-Exo and Exo on sepsis-induced MODS were evaluated using histological examination. Indeed, the septic mice exhibited interalveolar septum thickening and alveolar inflammatory cell infiltration in the lungs, diffuse hemorrhage and inflammatory cell infiltration in the kidneys, and multiple hepatic cell vacuolations and steatosis in the liver, as opposed to the control mice ( Figure 2D ). Administration of LPS-Exo or Exo considerably attenuated these organ injury scores, whereas the efficacy of LPS-Exo was better than that of Exo ( Figures 2E-G ). These results indicated that LPS-Exo possessed potent anti-inflammatory activity and protected against sepsis with higher efficacy.

To explore the immunosuppressive mechanism of LPS-Exo, we established an in vitro model of sepsis using BMDMs. First, we demonstrated the uptake of LPS-Exo and Exo by BMDMs using confocal microscopy ( Figure 3A ). Then, BMDMs were stimulated with LPS and co-cultured with LPS-Exo, Exo, or PBS for 24 h. The expression levels of inflammatory cytokines in the supernatant of BMDMs were evaluated at 12 h using ELISA. The results indicated that the levels of pro-inflammatory cytokines (IL-6, TNF-α, and NO) were significantly decreased and the anti-inflammatory cytokine (IL-10) was remarkably increased in the LPS-Exo and Exo groups compared with those in the PBS groups, whereas the LPS-Exo had a stronger effect than Exo ( Figures 3B-E ). Western blotting and qRT-PCR were performed to detected the phenotype markers expression of BMDMs at 24 h. The results showed that after treatment with LPS-Exo and Exo, M1 phenotype marker (iNOS) expression was decreased, and the M2 phenotype marker (CD206 and Arg1) expression were increased. And LPS-Exo had a stronger effect than Exo ( Figures 3F-H ). These data indicated that LPS-Exo and Exo treatment promoted M2 polarization and inhibited M1 polarization and that LPS-Exo exhibited a stronger effect than Exo.

Given that LPS-Exo possessed a better efficacy than Exo in regulating macrophage polarization, which was primarily through delivering miRNA or protein to recipient cells, we proposed that LPS stimulation changes the composition of LPS-Exo cargo through which it exerts its activity. Thereby, LPS-Exo was processed with proteinase K or RNase to determine the molecular cargo responsible for its immunomodulatory activity. Western blot analysis of iNOS and CD206 showed that LPS-Exo was unaffected by proteinase K treatment, whereas LPS-Exo failed to exert protective effects following RNase treatment ( Figure 4A ). Similar results were obtained using qRT-PCR ( Figures 4B, C ). These results demonstrated that miRNA molecules in LPS-Exo responsibility for its superior immunomodulatory properties.

The miRNA expression profiles of LPS-Exo and Exo were sequenced and compared, and 31 significant differentially expressed miRNAs were found (fold change > 1.5, adjusted p-value < 0.05), shown in a hierarchical clustering plot ( Figure 4D ). GO analysis revealed the influence of these differentially expressed miRNAs on various biological functions ( Figure 4E ). KEGG analysis revealed the most enriched signaling pathways relate to these differentiated miRNAs, and our data indicated that the PI3K/Akt and mTOR signaling pathways were within the 20 most enriched pathways ( Figure 4F ).

Among the markedly up-regulated differentially expressed miRNAs, miR-150–5p ranked second, which was previously reported to regulate macrophages polarization. We then rectified that the expression of miR-150–5p in LPS-Exo was significantly higher than that in Exo using qRT-PCR ( Figure 4G ). Additionally, macrophages treated with LPS-Exo exhibited higher expression of miR-150–5p ( Figure 4H ).

To investigate the functional role of miR-150–5p in LPS-Exo in vivo, we transfected MSCs with miR-150–5p inhibitor or NC prior to LPS stimulation. The exosomes were extracted and named inhibitor LPS-Exo or NC LPS-Exo, and the transfection efficacy was validated by qRT-PCR ( Figure 5A ). We then established an LPS-induced sepsis model and treated via tail vein injection with inhibitor LPS-Exo, NC LPS-Exo, or PBS. The survival rate of septic mice treated with the inhibitor LPS-Exo was significantly lower than that of mice treated with the NC LPS-Exo ( Figure 5B ). ELISA also showed that the protective effect of LPS-Exo on the release of proinflammatory cytokines was blocked by miR-150–5p inhibition ( Figures 5C, D ). Notably, pathological examination showed that downregulation of miR-150–5p in LPS-Exo led to a notable aggravation of the histological damage of organs ( Figures 5E-H ). Collectively, these data demonstrated that LPS-Exo delivers miR-150–5p alleviates sepsis.

To elucidate the mechanism through which miR-150–5p mediates macrophage polarization, bioinformatics analysis was performed to investigate the putative target genes of miR-150–5p. As predicted by miRanda and RNAhybrid databases, miR-150–5p conserved the binding site in the 3’UTR of insulin receptor substance 1 (Irs1) that was a positive regulator of the PI3K/Akt/mTOR signaling pathway ( Figure 6A ). The dual-luciferase reporter assay was conducted to validate our bioinformatic predictions, and the results indicated that luciferase activity was significantly decreased by miR-150–5p overexpression in Irs1 WT 3’UTR group, but not in MUT 3’UTR group ( Figure 6B ). Furthermore, the protein expression of Irs1 was decreased in LPS-Exo-treated macrophages, which was reversed by transfection with miR-150–5p inhibitors ( Figures 6C , 7B ). Altogether, these data suggested that miR-150–5p could inhibit Irs1 by directly binding with the 3’UTR of Irs1.

Studies have revealed the PI3K/Akt/mTOR signaling pathway is strongly associated with macrophage polarization, which can be positive regulated by Irs1. We hypothesized that miR-150–5p modulates the PI3K/Akt/mTOR pathway by targeting Irs1, which correspondently regulating macrophage polarization. To confirm our hypothesis, we employed western blotting to examine the expression levels of key proteins in the PI3K/Akt/mTOR pathway and noticed that the phosphorylation levels of PI3K, Akt, and mTOR were decreased after LPS-Exo treatment ( Figures 6C-F ). In contrast, transfection with miR-150–5p inhibitor ( Figure 7A ) prior to culturing with LPS-Exo significantly increased the phosphorylation of PI3K, Akt, and mTOR ( Figures 7B-E ). In addition, the inhibition of miR-150–5p reversed the upregulation of M2 macrophage markers and the downregulation of M1 markers induced by LPS-Exo ( Figure 7F ). We demonstrated that LPS-Exo transferred miR-150–5p to macrophage, and modulated the polarization and inflammatory response of macrophage by down-regulating PI3K/Akt/mTOR pathway via targeting Irs1.