Human studies

The studies involving participants were reviewed and approved by the Institutional Review Board of Xijing Hospital. Written informed consent was obtained from all participants. Serum samples were collected from healthy individuals and type 2 diabetic patients. According to the manufacturer's instructions, serum Adipsin levels were quantified using an ELISA kit (Elabscience Biotechnology Co., Ltd, E-EL-H6007).

Animal studies

All experimental animal procedures were approved by the Animal Care and Use Committee of the Fourth Military Medical University and followed the Animal Research Advisory Committee of the National Institutes of Health guidelines. AdipsinLSL/+ mice and Adipoq-Cre (Cre) mice were purchased from Shanghai Southern Model Biotechnology Co., Ltd. Adipose tissue-specific Adipsin overexpression mice (AdipsinLSL/LSL-Cre) were constructed using Cre recombinase-mediated excision of a STOP cassette (Additional file 1: Fig. S1). Genotypes were confirmed by genomic PCR according to the manufacturer's procedure (Bimake, B40015). Male mice were used for all experiments.

High-fat diet (Research Diets, D12492) and low-dose injections of streptozotocin (STZ, 50 mg/kg, Sigma-Aldrich, S0130) dissolved in 0.1 mol/l citrate buffer (pH 4.2–4.5) for 5 consecutive days were used to induce an animal model of type 2 diabetes mellitus (DM) as described with further optimization [17,18,19]. Furthermore, the non-diabetic group (Non-DM) received a standard chow diet and an intraperitoneal injection with an equal volume of citrate buffer. One week after the final STZ injection, mice that failed to meet the criteria (fasting blood glucose levels, 16.6 mmol/L) of diabetes diagnosis were given an additional injection of STZ. Four groups were assigned as follows: Non-DM + AdipsinLSL/LSL, Non-DM + AdipsinLSL/LSL-Cre, DM + AdipsinLSL/LSL, and DM + AdipsinLSL/LSL-Cre. The glucose tolerance test was performed using intraperitoneal injection of glucose following 16h-fasting, and the insulin resistance tolerance test was performed using intraperitoneal injection after 2-h fasting in mice. Blood glucose levels were measured using blood samples from the tail vein at the indicated time points.

Csk knockdown was achieved by intramyocardial injection of Adeno-associated virus 9 (Tsingke Biotechnology Co., Ltd) vector carrying Csk shRNA in vivo as described previously [20, 21]. Detailed information was provided in the Additional file 1: Fig. S2. In brief, mice were anesthetized with 2% isoflurane inhalation and secured on the shelf. Anesthetized mice were mechanically ventilated before aseptic left thoracotomy. The AAV9-Csk-shRNA or scramble (2.0+E10.0 viral genome particles per mouse) was injected into the left ventricular wall of transgenic mice at 3 different sites using the Hamilton micro-syringes. Exosomes were delivered via tail vein injections as follows. A total of 200 µL exosomes (1.0+E10.0 particles/kg) was administered once a week after 8 weeks of high-fat diet feeding until the termination of the experiment [22]. Equal volumes of PBS were injected as the vehicle control. At the end of the experiment, mice were euthanized using 30% VD/minof 100% carbon dioxide in adherence to the 2020 AVMA Guidelines on Euthanasia.

Echocardiography

Mice were anesthetized with 2% isoflurane inhalation and were placed in supine position on a heated plate at 37 °C. The chest and upper abdomen were debrided to expose the skin thoroughly, and cardiac function was evaluated by transthoracic echocardiography (VisualSonics, Vevo2100). The left ventricular diastolic function was assessed by the measurement of early peak flow velocity (E) to late peak flow velocity (A) wave ratio. The left ventricular systolic function was indicated by measurement of the left ventricular internal diameter at end-diastole (LVIDd) and left ventricular internal diameter at end-systole (LVIDs). Each measurement was obtained by calculating the average result of five consecutive heartbeats.

Enzyme-linked immunosorbent assay (ELISA)

Serum Adipsin levels were detected using the Mouse Adipsin ELISA Kit according to the manufacturer's procedure (Elabscience Biotechnology Co., Ltd, E-EL-M0335c). The optical density (OD value) was determined at the wavelength of 450 nm using a spectrophotometer.

Isolation of mouse primary cardiac microvascular endothelial cells

CMECs were isolated as described with appropriate optimization methods [23, 24]. Briefly, left ventricles were isolated from mice under general anesthesia (2% isoflurane inhalation). The heart was washed with ice-cold PBS and cut into 1 mm3 size, digested with 0.2% collagenase II at 37°C for 10 min until tissue pieces became gelatinous. Then digestion was continued with 0.25% trypsin at 37°C for 10 min, filtered through a 70-μm filter, and centrifuged at 1000g for 5 min. After removal of the supernatant, the pellet was resuspended in culture flasks with a complete medium containing 20% fetal bovine serum. CMECs were differentiated using the DiI-Ac-LDL marker, and cell proportions more significant than 90% were used for subsequent experiments. The diabetic group (HG + PA) was exposed to 25 mM glucose and palmitic acid (PA, 300 μM) to mimic diabetes. The control group (NG + vehicle) was given 5 mM glucose and solvent control. Mannitol at the same concentration was used as osmolarity control.

Isolation and identification of adipose tissue-derived exosomes

Subcutaneous and epididymal adipose tissues were obtained from AdipsinLSL/LSL and AdipsinLSL/LSL-Cre mice under general anesthesia (2% isoflurane inhalation). The tissue blocks were cut to approximately 1×1×1 mm size following rinsing in PBS on a sterile operating table, and the supernatant was collected after 24h of incubation in the exosomes-free medium. Exosomes were isolated from supernatant according to the conventional exosome isolation protocol [25]. Briefly, cells and impurity debris were removed from supernatants by centrifugation at 500g for 10min, 2000g for 10min, and 10,000g for 30min. The supernatant was filtered with a sterile 0.22-μm filter, and exosomes were isolated by ultracentrifugation at 100,000g for 70min. Exosomes in plasma were isolated using the commercial kit (Invitrogen, 4484450) according to the manufacturer’s protocol. The harvested exosomes were identified by transmission electron microscopy (JEM-2000EX) and Nanoparticle tracking analysis (Nanosight-NS300).

Scanning electron microscopy

Cardiac microvascular casting was employed to assess the integrity of microvascular endothelial cells as previously described [23, 24]. Low-viscosity resin mixed with benzoyl peroxide was perfused through the aorta and was then immersed in a 5% sodium hydroxide solution at room temperature. Connective tissues around the vessels were removed after 3–4 days. Isolated cardiac microvessels were dehydrated, dried, plated, and observed with a scanning electron microscope (HITACHI, S4800).

Transmission electron microscopy

The cardiac microvascular structure was visualized using standard transmission electron microscopy. Myocardial tissues were fixed with 2.5% glutaraldehyde overnight, washed, dehydrated, penetrated, and polymerized following lanthanum nitrate perfusion. The resin-embedded samples were imaged and captured using transmission electron microscopy (HITACHI, FE-2000).

Permeability assay in vitro

As previously reported [26], endothelial permeability was evaluated using 6.5-mm Transwell® 0.4-µm pore size polycarbonate membrane chambers (Corning, 3413). Briefly, CMECs were inoculated and changed to 70 kDa FITC-labeled dextran (Sigma-Aldrich, 46945) after growing into the monolayer. Fluorescence intensity was measured using a plate reader with filters for 485 nm excitation and 535 nm emission.

Trans-endothelial electrical resistance (TEER)

Trans-endothelial electrical resistance (TEER) analysis was performed to monitor the monolayer integrity as described previously [26]. In brief, CMECs were cultured as monolayers on semi-permeable filters (Corning, 3413) and exposed to NG + vehicle or HG + PA treatment. TEER was measured using EVOM Volt-Ohm Meter (World Precision Instruments Inc) at the indicated points. The percentage of TEER change was calculated for statistical analyses.

siRNA transfection

CMECs were cultured to a 60-80% confluence in a 6-well plate. Then cells were transfected with a siRNA against Csk or a negative control siRNA (Tsingke Biotechnology Co., Ltd). According to the manufacturer’s instruction, siRNA diluted in Lipofectamine 2000 was premixed with Opti-MEM media (50nM) at room temperature for 15min and was incubated with cells for 24h or 48 h for further study.

Co-immunoprecipitation (Co-IP)

Cell lysates were obtained following digestion with a RIPA buffer (Beyotime, P0013B). Then an immune complex was formed by binding the target protein with the specific antibody and co-incubation with the Pierce Protein A/G Magnetic Beads. Finally, the antibody/protein complex was separated from other cellular proteins by magnetic or centrifugal methods. Samples are electrophoresed on SDS-PAGE gels.

Liquid chromatography-mass spectrometry analysis (LC-MS)

The experiment was performed with the support of Applied Protein Technology. Briefly described as follows: samples were lysate to extract proteins before quantification using the BCA method. Each sample was separated by NanoElute, an HPLC liquid phase system with a nanoliter flow rate. Buffer A was 0.1% formic acid aqueous solution, and B was 0.1% formic acid acetonitrile aqueous solution. The chromatographic column was equilibrated with 95% of liquid A. Samples were loaded onto the loading column by the autosampler and separated by the analytical column at a flow rate of 300 nL/min. Samples were separated and analyzed using a timsTOF Pro mass spectrometer. Protein identification and quantitative analysis were performed using mass spectrometry assisted by the MaxQuant software.

Wound healing scratch assay

The migration ability of cells was assayed using the wound-healing scratch assay. In brief, CMECs were inoculated in a 6-well plate. When the cells were fully plated, a straight line was drawn at the bottom of the well. Then, the culture was continued with a serum-free or low-serum medium, and photographs of cell scratches were taken at selected appropriate time points. The migration distance of cells was determined by the ratio of the 12-hour scratch width to the 0-hour scratch width.

Cell migration assay

Cell migration ability was evaluated using 6.5-mm Transwell 8.0-µm pore-sized polycarbonate membrane chambers (Corning, 3422). First, 200 µl of serum-free medium cell suspension was inoculated in the upper chamber, and 600 µl of medium containing 20% FBS was added in the lower chamber. CMECs were then incubated for 24 h, fixed with 4% paraformaldehyde, and stained with 0.1% crystal violet (Beyotime, C0121). Finally, non-migrated cells in the upper chamber were gently wiped with a cotton swab, and migrated cells were observed with an Olympus microscope.

Tube formation assay

Endothelial cell tube formation experiments were performed in 250-µl Matrigel (Corning, 354248) coated 24-well plates. Inoculate 200 µl of cell suspension at a density of 2.0+E5 into a 24-well plate covered with Matrigel, and then continue to incubate for 12 h. After fixing with 4% paraformaldehyde at an appropriate time, the tube formation was visualized using microscopy.

Flow cytometry

The Annexin V-FITC Apoptosis Assay Kit (Beyotime, C1062) was used to detect apoptosis according to the manufacturer’s manual. Briefly, CMECs were digested and centrifuged. Then 5-μl Annexin V-FITC and 10-μl propidium iodide staining solution were sequentially added. After incubation for 20 min at room temperature, apoptosis was detected using flow cytometry (Beckman Coulter, CytoFLEX S).

TUNEL staining

TUNEL staining was performed according to the instructions of the Beyotime one-step TUNEL Apoptosis Detection Kit (Beyotime, C1088). Finally, nuclei were stained with DAPI (Beyotime, C1006), and images were collected by confocal fluorescence microscopy.

CCK-8 proliferation assay

To assess cell viability, CMECs were plated and given indicated drug treatment in 96-well plates. Then cell viability was determined via the Cell Counting Kit-8 (Sigma, 96992) according to the manufacturer’s instructions. The absorbance at 450 nm was read using the microplate reader (Molecular Devices, SpectraMax M5).

Immunofluorescence staining

CMECs or 10-μm frozen sections were fixed with 4% paraformaldehyde for 15min before incubation with an immunostaining permeabilization solution Triton X-100 (Beyotime, P0096) for 20min at room temperature and blocking with goat serum (Beyotime, C0265) for 1 h at room temperature. Then cells were incubated with primary antibody overnight at 4°C and incubated with fluorescent secondary antibody for 1 h at room temperature, and nuclei were stained with DAPI (Beyotime, C1006). Images were acquired by observing under a confocal fluorescence microscope. Antibody dilution information was shown in the Supplementary Materials (Additional file 1: Table S1).

Quantification of blood vessel density

Anti-CD31 immunofluorescence intensity was used to determine blood vessel density in heart. The blood vessel density was analyzed by using the VesselJ plugin of ImageJ software as described [27].

Quantitative real-time PCR

Total RNA was extracted using TRIzol™ reagent (Invitrogen, 15596018). RNA concentration was measured with the Nano Drop2000 spectrophotometer (Thermo Scientific). The cDNA was synthesized, and genomic DNA was removed using the PrimeScript™ RT reagent Kit with gDNA Eraser (Takara, RR047A). Quantitative analysis was performed using TB Green® Premix Ex Taq™ II (Takara, RR820A) and StepOnePlusTM Real-Time PCR (Thermo Fisher Scientific, 4376600). The primer sequence information was shown in the Supplementary Material (Additional file 1: Table S2).

Western blot

Serum protein was extracted using the Serum Protein Extraction Kit (Bestbio, BB-31954). Tissue or cells were lysed with RIPA lysis buffer (Beyotime, P0013E) with protease inhibitor and phosphatase inhibitor. Protein concentration was measured using the BCA protein concentration assay kit (Beyotime, P0010). Then protein samples were separated in SDS-PAGE gels and transferred to PVDF membranes with 0.22-μm pore size. Then membranes were closed for 20 min by fast closure solution and incubated with primary antibody at 4°C overnight, followed by horseradish peroxidase (HRP)-labeled secondary antibody at room temperature for 1 h, using the Image Lab system for analysis. Antibody dilution information was shown in the Supplementary Material (Additional file 1: Table S1).

Statistical analysis

Data analysis was performed using GraphPad Prism 9.0 software. Results were expressed as mean ± the standard error of the mean, and n indicates sample size of each group. Statistical analyses between groups were done by unpaired Student’s t‐test and one-way ANOVA followed by a Fisher’s post hoc comparison test. A p-value less than 0.05 was accepted to be a significant difference.