Identification of differentially expressed microRNAs (DEmiRNAs)
The GSE101639 dataset included three normal controls, three patients with sepsis, and three patients with septic shock and was used to analyze DEmiRNAs associated with SAE. The expression matrix was normalized using the quantile method through the limma package. A |log2FoldChange| > 1 and adjusted P < 0.05 were used to identify DEmiRNAs. TargetScan (http://www.targetscan.org/mamm_31/) was used to predict the miRNAs targeting the combination of HMGB1. The intersection was used to obtain 15 prediction target miRNAs for subsequent research.
Cell culture and intervention
The methods used to extract and identify MSCs and MSCs-exo, and other protocols are shown in Supplementary Methods. Human microglia HMC3 cells (AW-CNH003, Abiowell, Changsha, China) were cultured in minimum essential medium (MEM) supplemented with 10% fetal bovine serum, 100 IU/mL penicillin, and 10 μg/mL streptomycin at 37 °C, 5% CO2. The cells were randomly divided into the normal group, LPS group, MSCs-exo group, and MSCs-exo + miR-140-3p mimics group. Then, 100 ng/mL lipopolysaccharide (LPS) was added and incubated for 4 h under standard culture conditions29. MSCs-derived exosome (MSCs-exo) (1 μg/mL)30 or S-lactoylglutathione (10 mM)31 were with the LPS. The MSCs-exo + miR-140-3p mimics group consisted of MSCs-exo transfected with miR-140-3p mimics (1 μg/mL).
Dual-luciferase reporter assay
TargetScan (http://www.targetscan.org/mamm_31/) was used to predict the binding sites of miR-140-3p and Hmgb1. Mutant (mut-Hmgb1) and wild-type (wt-Hmgb1) sequences were designed, synthesized according to the predicted results, cloned, and inserted into the pmirGLO reporter vector (Synthgene Biotech, Nanjing, China). The vector was transfected into HEK293T cells (AW-CNH086, Abiowell) with miR-140-3p mimics or mimics-NC (HonorGene, Changsha, China). Then, luciferase activity was determined by a dual-luciferase reporter gene assay kit (Promega, Shanghai, China). Renilla luciferase activity was used as an internal control.
Animal experiment
C57BL/6J male mice, which were aged 6–8 weeks and weighed 25–30 g, were purchased from Hunan Slack Jingda Laboratory Animal Co., Ltd. The IRB of Third Xiangya Hospital, Central South University, approved all the experiments in this study (2023-S633). The experiments complied with all relevant ethical regulations for animal use. The mice were randomly divided into five groups (n = 10): the sham group, model group, MSCs-exo + miR-140-3p mimic (Exo) group, antibiotic (ABX) group, and ABX + Exo group. The cecal ligation puncture (CLP) model was constructed12. In brief, animals were anesthetized with 2% sevoflurane in a well-ventilated room. The distal cecum was ligated 40% from the base and pierced once with a 20-G needle. The sham group underwent the same laparotomy but not CLP. The remaining groups were used to construct CLP models. Bupivacaine (3 mg/kg) and buprenorphine (0.1 mg/kg) were injected subcutaneously to avoid postoperative pain. On the 8th day after surgery, 200 μg of exosomes were injected through the tail vein32. For five consecutive days after CLP, 25 mg/kg/day of imipenem/cilastatin and 0.9% NaCl were injected intraperitoneally once per day33. The mice were continuously observed for 25 days and killed on the 31st day. The whole brain was dissected, rinsed with precooled saline, and fixed in 4% paraformaldehyde for paraffin-embedded sections or further dissected for the hippocampus tissue. Then, the dissected hippocampus tissue was snap-frozen by liquid nitrogen and stored at −80 °C for subsequent molecular biological detection (n = 3) and metabolites analysis (n = 3).
Morris water maze
As previously described12, the mice were trained four times daily for 5 days. After pretraining, the mice were tested at intervals of 5 s, 20 min, and 2 h (the interval between trials 1 and 2). Each test session lasted for 3 days. The time/path length was calculated by subtracting Trial 2 time/path length from that of Trial 1. A larger time/path length difference indicated better performance.
Immunohistochemistry
Hippocampal tissue was fixed with 4% paraformaldehyde and cut into 2 μm slices. The slices were baked at 62 °C for 8 h, dewaxed, and rehydrated. The slices were soaked in 0.01 M citrate buffer at high temperature for 20 min and placed in a periodate solution to inactivate endogenous enzymes. The slices were incubated with primary antibodies against Caspase 1 (1:300, 22915-1-AP, Proteintech, USA) and NLRP3 (1:300, 19771-1-AP, Proteintech, USA) at 4 °C overnight. The slices were incubated with HRP-conjugated polyclonal goat anti-rabbit IgG (AWS0002, Abiowell, China). The slices were observed and analyzed under a microscope (BA210T, Motic, Singapore).
Immunofluorescence analysis
The slices were subjected to immunohistochemical staining. After the cells were fixed with 4% paraformaldehyde for 30 min, they were permeabilized with 0.3% Triton X-100 for 30 min. Then, the slices were blocked with 5% bovine serum albumin (BSA) at 37 °C for 90 min. The slices were incubated with primary antibodies against Caspase 1 (1:300, 22915-1-AP, Proteintech, USA), NLRP3 (1:300, 19771-1-AP, Proteintech, USA), HMGB1 (1:200, ab190377, Abcam, UK) and ionized calcium-binding adaptor molecule-1 (IBA-1) (1:50, 10904-1-AP, Proteintech, USA) at 4 °C overnight. The slices were incubated with secondary antibodies at 37 °C for 60 min. The slices were then preserved in glycerin buffer and observed under fluorescence microscopy.
Quantitative reverse transcription polymerase chain reaction (qRT-PCR)
Total RNA was extracted from cells and reverse-transcribed into complementary DNA (cDNA) using TRIzol reagent (15596026, Thermo Fisher Scientific, MA, USA) and a reverse transcription kit (CW2141, CWBIO, Beijing, China). PCR was performed using UltraSYBR mixture (CW2601, CWBIO). U6 was used as the internal control for miRNA expression. Relative miRNA expression was determined by the 2−ΔΔCt method. The primer sequences are shown in Supplementary Table 1.
Western blot analysis
Total proteins were extracted from brain tissue and neuronal cells with RIPA lysis buffer (AWB0136, Abiowell). The bicinchoninic acid (BCA) method was used to determine the protein concentration (AWB0104, Abiowell). The proteins were separated by 10% sodium dodecyl sulfate‒polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (AWB0231, Abiowell). The membranes were blocked in 5% skim milk at room temperature for 2 h and incubated with the primary antibodies shown in Supplementary Table 2. The membrane was incubated with secondary antibody HRP Goat anti-mouse IgG (SA00001-1, 1:5000, Proteintech, USA) or HRP Goat anti-rabbit IgG (SA00001-2, 1:6000, Proteintech, USA) at room temperature for 90 min. A ChemiScope6100 system (CLiNX) was used to capture the chemiluminescence signals from the membrane and produce a digital image of the protein bands. The band density in the image was then quantified using ImageJ software (ImageJ 1.5, USA). Uncropped blot images are shown in Supplementary Fig. 1.
Statistics and reproducibility
The data were analyzed and plotted using GraphPad Prism 9 software. The data are expressed as the mean ± standard deviation (SD). Each test was repeated independently three times from distinct samples. Kolmogorov–Smirnov test and exploratory descriptive statistics test were used to analyze whether the data conformed to a normal distribution and homogeneity of variance. The measurement data obeyed the normal distribution and homogeneity of variance. The data were analyzed by parametric test. The unpaired Student’s t-test was used to compare the data of two groups that were not one-to-one correspondence using a two-tailed approach. One-way ANOVA and Tukey’s post-hoc test were used to compare data among three groups. Data comparisons between groups at different time points were analyzed by two-way ANOVA with Bonferroni as a post hoc test. The significance level was set at P < 0.05.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.