Animals

All in vivo studies were performed at the School of Medicine Animal Facilities in Changsha Medical University. The experimental procedures were approved by Changsha Medical University Animal Care and Use Committee and were conducted in full accordance with the Guide for the Care and Use of Laboratory Animals. All experimenters were blind to group assignment and outcome assessment.

All samples were randomised and all criteria used for exclusion of any data, samples or animals have been included. All experiments were conducted starting with male mice aged 4–6 weeks. Mice (C57BL/6J; The Jackson Laboratory, USA; https://www.jax.org/strain/000664) were fed a normal chow diet (NCD) ad libitum. The high-fat diet (HFD) stress model was established in mice by feeding an HFD (percentage of total energy provided by protein, fat and carbohydrate: protein 18.1%; fat 61.6%; carbohydrates 20.3%; D12492 Research Diets, NJ, USA) continuously for 18 weeks. The mice were bred in a standard environment with a 12 h light–dark cycle. Mice were fasted for 6 h and circulating triglyceride (TG) levels were measured in tail blood using reactive strips in an Accutrend GCT Glucometer (Roche, Germany). Circulating NEFA levels were determined in plasma from mice fasted for 6 h using a NEFA-HR (2) Kit (Wako, Germany) according to the manufacturer’s instructions. Total cholesterol (TC) was measured using an ELISA kit (F002-1; Nanjing Jiancheng, China). Plasma obtained from tail blood was centrifuged at 1500 g for 20 min (4°C).

Differentiation of adipocytes

The 3T3-L1 mouse embryonic fibroblast line was identified to be free from mycoplasma contamination. To obtain adipocytes, differentiation of 3T3-L1 cells was induced in the induction medium after 7–10 days. The DMEM/F12 medium contained 4.5 g/l glucose, 10% FBS (vol./vol.), penicillin–streptomycin and glutamine. The differentiation cocktail consisted of 0.5 mmol/l 3-isobutyl-l-methylxanthine, 1 μmol/l dexamethasone, 10 μg/ml insulin, 0.2 mmol/l indometacin and 1 μmol/l rosiglitazone in DMEM supplemented with 10% FBS, 1% l-glutamine and 1% penicillin/streptomycin (vol./vol.).

Isolation of primary hepatocytes

Primary hepatocytes were isolated as previously described [7]. Mice were infused through the vena cava with a calcium-free HEPES-phosphate buffer A (calcium- and magnesium-free PBS containing 0.2 μmol/l EGTA, 10 mmol/l HEPES, 1 mmol/l glucose and 0.2% BSA, pH 7.4) for 3–5 min followed by a collagenase-containing buffer B (PBS with 1 mmol/l magnesium, 1 mmol/l calcium, 0.2% BSA, 30 mmol/l HEPES and 0.5 mg/ml collagenase H) for 10 min. After the appearance of cracking on the surface of the liver, perfusion was stopped immediately, the digested liver was excised and hepatocytes were collected. Hepatocytes were suspended in buffer A, filtered through a 100 μm cell strainer and centrifuged at 50 g for 7 min at 4°C. The pellet was washed twice with collagenase-free buffer B and then mixed with Percoll (adjusted to physiological ionic strength with 10×PBS) to a final concentration of 36% and centrifuged at 100 g at 4°C for 10 min. After the supernatant fraction was removed, the hepatocyte pellet was washed once with collagenase-free buffer B and then cultured in Williams Medium E with 10% FBS on collagen-coated plates (GIBCO, Life Technologies, USA) and 1% penicillin/streptomycin. After overnight incubation, culture medium was replaced with fresh medium.

Mitochondrial morphology by confocal microscopy

Mitochondrial morphology images were acquired by confocal microscopy as previously described [13]. Cells were grown on poly-d-lysine-coated glass coverslips. Mitochondria were labelled with Mitotracker Red (50 nmol/l; Invitrogen, USA) at 37°C for 30 min and fixed with 4% paraformaldehyde according to the manufacturer’s instructions. Coverslips were subsequently mounted with Prolong Gold Antifade Reagent with DAPI (Invitrogen). Fluorescence images were captured at room temperature using a camera mounted to an inverted epifluorescence microscope. The overall contrast and brightness of the acquired images were adjusted using Adobe Photoshop CS6 (Adobe, USA).

IPGTT and IPITT

For GTTs, mice received one dose of dextrose (1 g/kg body weight) via i.p. injection after 12 h fasting. Blood glucose was determined at 0, 15, 30, 60, 90 and 120 min after injection. For ITTs, mice were fasted for 6 h and then injected intraperitoneally with insulin (0.70 U/kg body weight). Blood glucose was determined at 0, 15, 30, 60 and 90 min after injection.

Akt phosphorylation assay

Liver, skeletal muscle and epididymal visceral adipose tissue (eVAT) were evaluated by measuring insulin-stimulated Akt phosphorylation as previously described [7]. Briefly, after being fasted for 8 h, mice were anaesthetised and samples of these insulin target tissues were collected to measure basal levels of Akt phosphorylation. After a dose of insulin (0.75 U/kg body weight) injected via the vena cava, samples of liver, skeletal muscle and eVAT were collected at 3 min, 7 min and 10 min, respectively. The phosphorylation of Akt was measured by western blot analysis with antibodies against phosphorylated Akt (s473) or pan-Akt.

Cell culture

Bone-marrow-derived macrophages (BMDMs) were generated in the presence of L-929 conditional medium and granulocyte–macrophage colony-stimulating factor as described previously [13]. The RAW264.7 mouse monocyte/macrophage cell line (American Type Culture Collection, Manassas, VA, USA) was identified to be free of mycoplasma contamination. RAW264.7 cells were cultured in DMEM (Hyclone, UT, USA). RAW264.7 cells were maintained in 10% (vol./vol.) FBS, 1% penicillin–streptomycin at 5% CO2 and 37°C. 3T3-L1 cells (American Type Culture Collection) were maintained in DMEM (Hyclone). HEK293 cells (American Type Culture Collection) were cultured in DMEM (Nacalai Tesque, Japan) supplemented with 10% (vol./vol.) FBS (Sigma-Aldrich, USA). HEK293 cells were authenticated using the PowerPlex16 STR system (Promega, USA). L6 myocytes (American Type Culture Collection) were cultured in minimum essential medium (MEM) with 4% (vol./vol.) FBS for 8–10 days.

MicroRNA, siRNA, plasmid transfection and adenovirus infection

Cy3-labelled miR-27-3p mimic and antagomir and relevant negative controls (NC) were purchased from RIBOBIO (Guangzhou, China). Compared with common miR-27-3p inhibitors  (5’-GCAGAACTTAGCCACTGTGAA-3’; GenePharma, Shanghai, China), antagomir has higher stability and inhibition in animal experiments and cell experiments and is more easily enriched in target cells through cell membranes and tissue gaps. The sequences of the miRNA mimic and antagomir are listed in electronic supplementary material (ESM) Table 1. The GFP-tagged microtubule-associated protein 1 light chain 3 (LC3) plasmid was provided by Addgene (USA). Lipofectamine 2000 reagent (Invitrogen) was used to perform transient transfection in cells according to the producer’s protocol for 24 or 48 h before harvesting. Adenovirus harbouring mCherry-GFP-LC3 (Ad-mCherry-GFP-LC3) was purchased from Beyotime Biotechnology (Shanghai, China).

Palmitic acid solutions

Palmitic acid (PA) (C 16:0) was purchased from Sigma (USA). PA solutions were prepared according to the method of Martino et al [14]. Briefly, PA was dissolved at 70°C in 0.1 mol/l NaOH to obtain a 100 mmol/l stock solution. A 5% (wt/vol.) solution of PA-free BSA was prepared in serum-free RPMI medium. Then, a 5 mmol/l PA–BSA mixture was prepared by combination of these two solutions. Finally, the PA stock solutions were diluted in RPMI supplemented with 1% FBS to obtain 0, 0.25, 0.5 and 1 mmol/l final concentrations at a fixed concentration of 0.5% BSA. INS-1 (823/13) cells (Sigma-Aldrich, USA) were transfected with construct using Lipofectamine Plus (Life Technologies).

Lipopolysaccharide+PA treatment of BMDMs

To prepare lipopolysaccharide (LPS)+PA-treated BMDMs, about 2.0×106 macrophage cells were seeded in 100 mm culture dishes. The cells were cultured in LPS (100 ng/ml) + PA (0.05 mmol/l) and 1000 U/ml IFN-γ for 8 h [15].

Characterisation of Exos from BMDMs

The Exos from BMDM culture medium were extracted as previously described [7, 13]. The characterisation of Exos was confirmed by measuring expression of the Exo-specific markers arginase-1 (Arg-1) and inducible nitric oxide synthase (iNOS), and EV-associated protein markers tumour suppressor gene 101 (TSG101), ALG-2-interacting protein X (Alix), CD63 and Syntenin1 by western blot analysis, and by NanoSight analysis of particle size (Malvern Instruments, UK).

In vivo and in vitro Exo treatment

For in vitro assays, 1×108 Exos based on NanoSight analysis were added to 0.1×106 cells for 36 h. For in vivo treatment, recipient mice were treated by i.v. injection of 1×109 extracellular vesicles twice per week. In the control groups, blank liposomes (FormuMax, USA) were used.

Isolation and identification of adipose tissue macrophages

Isolation and identification of adipose tissue macrophages (ATMs) were performed as previously described [16]. To isolate macrophages, mouse adipose tissues were mechanically chopped and then digested with 1 mg/ml collagenase II (Sigma-Aldrich) for 30 min at 37°C with shaking. The digested tissues were filtered through a 100 μm filter and centrifuged at 1000 g for 10 min. Primary adipocytes in the supernatant fraction were separated and erythrocyte buffer (Roche, Germany) was added to the pellet containing stromal vascular cells. Cells were resuspended and incubated with fluorescence-tagged antibodies against CD11b or F4/80. CD11b+F4/80+ macrophages were purified using a BD FACS Aria II flow cytometer (BD Biosciences, USA). The presence of macrophage contamination in isolated primary adipocytes was detected by measuring the cell surface markers CD11b and F4/80.

Exosomal RNA preparation and miRNA sequencing

Exosomal RNA preparation and miRNA sequencing were performed as previously described [17]. Total exosomal RNA isolation and extraction were performed using an exosomal RNA isolation kit (Norgen Biotek, Canada) according to the manufacturer’s protocol. RNA purity was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, USA). A small RNA library was prepared using the TruSeq Small RNA Sample Prep Kit (Illumina, USA) and an Illumina HiSeq 2500 instrument was used to sequence the constructed library. Resulting data were processed using the ACGT101-miR program v4.2 (LC Sciences, Houston, TX, USA) to remove adaptor dimers, junk sequences, low complexity and repeats. Finally, a BLAST search (https://blast.ncbi.nlm.nih.gov/blast/Blast.cgi) was performed to map the 18–26 nucleotide long distinctive sequences to species-specific precursors in miRBase 22.0 and identify known miRNAs and novel 3p and 5p miRNAs.

miRNA quantitative PCR assay

miRNA and mRNA levels were assessed by quantitative real-time PCR (qPCR) using gene-specific primers (ESM Table 2). Data were normalised to levels of U6 snRNA or total Exo protein levels.

Flow cytometry analysis

Macrophage subtypes were detected with antibodies against CD206 (141706, BioLegend, USA) and CD11c (12-0114-83, Invitrogen) according to the manufacturers’ instructions. Data were acquired with a FACSCalibur flow cytometer.

Glucose uptake assay and output assay

The assays of glucose uptake and output were performed as previously described [7]. For the glucose uptake assay, after 8 h of serum starvation, cells were stimulated with 100 nmol/l insulin for 30 min in KRH buffer (137 nmol/l NaCl, 4.8 mmol/l KCl, 1.2 mmol/l KH2PO4, 1.2 mmol/l MgSO4, 2.5 mmol/l CaCl2, 0.2% BSA, 16 mmol/l HEPES) at 37°C. Then, 2-deoxy-d-[3H]glucose (0.1 mmol/l, 3.7 MBq) was added to cells. After 10 min incubation at 37°C, cells were washed twice with ice-cold PBS. NaOH (1 mmol/l) was then added for 20 min to efficiently dissolve the cells. An aliquot was used for measurement of protein concentration. After neutralisation of NaOH with HCl (1 mmol/l), the extracts were transferred to scintillation vials. Scintillation fluid was added and the radioactivity was counted. Results were normalised to protein concentrations of cell lysates.

For the glucose output assay, after 6 h of serum starvation, primary hepatocytes were washed twice and then exposed to glucose-free buffer (10 mmol/l HEPES, 4 mmol/l KCl, 125 mmol/l NaCl, 0.85 mmol/l KH2PO4, 1.25 mmol/l Na2HPO4, 1 mmol/l CaCl2 and 15 mmol/l NaHCO3) containing glucagon (200 ng/ml), insulin (10 nmol/l) or a combination of glucagon and insulin for 4 h at 37°C. Glucose production was determined by measurement of glucose in the media. The primary hepatocytes attached to the culture plate were dissolved by adding NaOH (1 mmol/l) and protein content was determined. The glucose results were normalised to protein concentrations of cell lysates.

Electron microscopy analysis

Cells cultured on the Aclar embedding film (2 mil [51 μm] thickness; Electron Microscopy Sciences, USA) were fixed in 2.5% glutaraldehyde and 4% sucrose in a 0.05 mol/l phosphate buffer, pH 7.4, and examined with a JEOL 1200EX electron microscope (Tokyo, Japan).

Reagents and antibodies

The PKH26 mini kit (MINI26) and DAPI, a blue-fluorescent dye (D9542), were obtained from Sigma-Aldrich. The following primary antibodies were used for western blotting: anti-IFN-γ (11276905001, Roche), anti-Alix (ab186429), anti-iNOS (ab178945), mouse anti-Flag (1:1000; GSA00170-40), rabbit anti-Myc (1:2000; ab9106) and anti-CD63 (ab216130) (Abcam, USA); anti-Arg-1 (93668T), anti-NLRP3 (ab4207) and anti-caspase-3 (9662) (Cell Signaling Technology, USA); anti-dynamin-related protein 1 (DRP1) (8570S), anti-mitochondrial fission factor (MFF) (84580S), anti-mitofusin 2 (MFN2) (9482S), anti-optic atrophy protein 1 (OPA1) (80471S), anti-COX IV (4844S) and anti-GAPDH (8884S) (Cell Signaling Technology); rabbit anti-LC3 (L8918; Sigma), rat anti-lysosome-associated membrane protein (LAMP2) (ab13524; Abcam), anti-β actin antibody (ab3280; Abcam), rabbit polyclonal anti-autophagic protein 7 (Atg7) (2631; Cell Signaling Technology), anti-mouse F4/80 antigen (AB_469653; Abcam), anti-mouse CD11c (AB_469346; Abcam), PE anti-mouse CD206 (AB_10895754; Abcam) and guinea pig polyclonal anti-p62 (GP62-N; ProGen, Germany). Alexa Fluor-conjugated antibodies (A21057, A21076 and A21096; Molecular Probes) were used as secondary antibodies. Chemiluminescence was detected using ECL plus reagents (GE Healthcare Bio-Sciences) in an LAS4000 Lumi-Imager (Fuji Photo Film, Japan).

Western blot analysis

Mice were euthanised at 12 weeks of age by cervical dislocation. Tissues were excised and frozen in liquid nitrogen. Cell and tissue lysates were separated by SDS-PAGE prior to transfer to PVDF membranes and immunoblotting. Data were normalised relative to levels of actin or GAPDH.

Autophagic flux assay

Detection of autophagic flux, the specific processing of autophagy protein LC3, with or without a lysosomal protease inhibitor, was determined by western blot analysis. Cells were treated with or without both pepstatin A (PepA) and E46d (10 μmol/l for each) or siAtg7, and autophagic protein levels were detected by western blotting.

Preparation of encapsulated small nucleic acids with Invivofectamine

Cy3-labelled miR-27-3p mimics were encapsulated using Invivofectamine, a cationic liposome-based formulation according to the manufacturer’s instructions (Thermo Fisher Scientific, USA).

H&E staining and immunofluorescence

Adipose tissue was fixed in 10% formalin and embedded in paraffin using standard histological protocols. Subsequently, tissue was sectioned and stained with H&E and immunofluorescence was measured using a digital image analysis system (Image-Pro Plus, version 6.0, https://www.bioimager.com/product/image-pro-image-pro-plus-analyzer-software/).

Luciferase reporter assay

HEK293 cells (1×105) were transfected with 3′ untranslated region (3′ UTR) luciferase reporter constructs (3′UTR-NC, 3′UTR-Miro1, 3′UTR-Miro1-mutant), miRNA (miRNA-NC or miR-27-3p) and Renilla luciferase using Lipofectamine 2000 (Genechem, Shanghai, China), according to the manufacturer’s instructions. After 24 h of transfection, the luciferase activity of cells was measured using the Dual Luciferase Assay Kit (E1910; Promega, Germany) and microplate reader (BioTeK, USA).

Quantification and statistical analysis

All values are expressed as means ± SEM. Comparison of differentially expressed mouse miRNAs was carried out by single-channel chip analysis and the data were fitted by locally weighted scatterplot smoothing (Lowess) using the PROC LOESS procedure in SAS/STAT software version 9.2 110 (https://www.sas.com/en_us/software/stat.html). For normally distributed data with equal variance, significant differences were determined using Student’s t test (when two groups were compared) or one-way or two-way ANOVA to test the effect of group (when more than two groups were compared). For non-normally distributed data or data with unequal variances, significant differences were determined using the non-parametric Mann–Whitney U test (when two groups were compared) or Kruskal–Wallis test followed by a post hoc Bonferroni test (when more than two groups were compared). p<0.05 was considered statistically significant.