Circulating exosomal mir-16-2-3p is associated with coronary microvascular dysfunction in diabetes through regulating the fatty acid degradation of endothelial cells

Patients were enrolled from the Cardiovascular Department of Nanjing Drum Hospital between May 2021 and May 2022. The study protocol was approved by the Ethics Committee of Nanjing Drum Hospital (2023-416-02) and adhered to the Declaration of Helsinki. The participants were divided into 3 groups according to the inclusion/exclusion criteria: the DM group, the DM-CMD group and the DM-CAD group (n = 3 each group). The diagnosis of diabetes followed the 1999 WHO diagnostic criteria [13] and included the following typical diabetic symptoms: (1) fasting blood glucose ≥ 7.0 mmol/L, (2) HbA1c ≥ 6.5%, and (3) random blood glucose ≥ 11.1 mmol/L. The diagnosis of CMD included the presence of labor-induced chest pain or tightness and objective evidence of myocardial ischemia (meeting at least one of the following criteria): (a) dynamic changes in ST-T during symptom onset; (b) positive results on the treadmill exercise test; (c) reversible ischemic changes observed in myocardial nuclide perfusion imaging during exercise/drug loading; (d) no obstruction in the coronary arteries from angiography. CAD diagnosis was established through coronary angiography or coronary CTA, which indicated stenosis in the lumen of the main coronary artery or branches with a diameter greater than 2 mm but less than 50%. All patients were excluded if they had (1) heart failure (ejection fraction, EF < 50%), (2) renal failure (estimated glomerular filtration rate, eGFR < 60 ml/min*1*73 m2), or (3) malignant tumors. Clinical characteristics, including age, sex, BMI, smoking status, hypertension status and the use of antidiabetic drugs, were collected. Peripheral blood samples were collected to analyze fasting glucose levels, HbA1c levels, eGFRs, total cholesterol levels, and triglyceride levels.

Serum exosomes were isolated using ultracentrifugation as previously described with modifications [14]. Briefly, the serum was obtained by centrifugation at 3000 × g for 15 min to remove the cell debris and platelets. Microvesicles were pelleted at 10,000 × g for 30 min, after which the exosomes were further purified from the supernatant by ultracentrifugation at 100,000 × g for 60 min. Following isolation, the exosomes were diluted in 100 µL of filtered PBS and stored at − 80 °C.

Exosomes were dissolved in lysis buffer and quantified using a BCA analysis kit (Thermo Fisher Scientific, USA). Western blotting was used to detect exosomal markers, including TSG101, HSP70, CD63 and CD9. Exosomes were fixed in 2.5% glutaraldehyde at 4 °C, dehydrated with gradient alcohol and embedded in epoxy resin. Sections were stained with uranyl acetate and citrate acid lead. The images were captured under a transmission electron microscope (JEM-1010, Japan). For nanoparticle tracking analysis, samples were loaded into the sample chamber of an NS500 unit (NanoSight, UK), and five 1-min videos of each sample were recorded. The data analysis was performed with NTA 2.3 software, and the size and concentration of the particles were calculated.

The raw reads from small RNA sequencing were generated in fastq format, and the linker sequence was removed using Cutadapt. Fastx_toolkit (version 0.0.13) software was used to perform Q20 quality control on the sequences, and sequences with a Q20 of 80 or above were retained. The NGSQC toolkit (version 2.3.2) was subsequently used to filter out reads containing N-bases. The resulting high-quality clean reads were used for subsequent analysis. Statistics were performed on the length distribution of the clean reads to initially assess the distribution of small RNAs in the sample. Using blastn software, the clean reads were compared with those in the Rfam (version 10.0) database to annotate sequences such as rRNA, snRNA, snoRNA, and tRNA. Using RepeatMasker software, the filtered sequences were compared to those in the repeat database to identify possible repeats. Small RNA sequences on unannotated devices were predicted using miRDeep2 software for new miRNAs, and secondary structures of miRNAs were predicted using RNAfold software. The miRNA expression calculation used TPM (transcript per million) to calculate the metric, where TPM = number of reads per miRNA/total sample alignment read number ×106. The miRNA array data have been deposited in the Gene Expression Omnibus under accession number GSE234464.

The miR-16-2-3p sequence was subcloned and inserted into an IRES-containing GFP-expressing plasmid, which was subsequently packaged in lentiviral vectors at a concentration of 5 × 10⁸ TU/ml. The intramyocardial injections were performed via a 30-gauge needle using a 20 µl micromanipulator (Hamilton, USA). A volume of 20 µl was injected into three distinct areas in the anterior wall of the left ventricle.

Transthoracic echocardiography was performed before myocardial injection and one month later using an In Vivo Microimaging System (VisualSonics, Canada) under 2.0% isoflurane inhalation. Briefly, the mice were abdominally shaved, anesthetized and placed on a heating pad. The left ventricular internal systolic dimension (LVIDs) and left ventricular internal diastolic dimension (LVIDd) were measured and averaged from three consecutive cardiac cycles. The left ventricular ejection fraction (EF) and fractional shortening (FS) were calculated. EF was calculated with the Simpson method [15], and FS was calculated as [(LVIDd-LVIDs)/LVIDs] x100%.

The mice were euthanized with 5% isoflurane, and the hearts were removed after 4 weeks. The hearts were perfused with PBS and fixed with 4% paraformaldehyde. After fixation for 24 h, the hearts were embedded in paraffin and sliced into 5 μm thick sections. Hematoxylin and eosin (H&E) staining and Picrosirius Red staining were used to evaluate cardiac morphology and collagen deposition. For immunofluorescence staining, the sections were incubated with the miR-16-2-3p probe (Servicebio, China) or with antibodies against CD31, cTnT and αSMA (all from Abcam, UK). To evaluate the microcirculation, 100 µl of FITC-lectin (Sigma, USA) at a concentration of 1 mg/ml was injected into the mice via the caudal vein to label the perfused vessels. Ten minutes after injection, heart samples were harvested to prepare frozen sections, which were subsequently immunoblotted for CD31. The histological areas of interest were all semiquantified by ImageJ software (NIH, USA).

Human aortic endothelial cells (HAECs) and human cardiac microvascular endothelial cells (HCMECs) were purchased from ATCC. The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin/streptomycin (P/S; Invitrogen, USA). These two cell lines were treated with high glucose (40 µM) for 48 h to mimic CAD or CMD, respectively.

A wild-type luciferase reporter was constructed with the 3’-UTR of ACADM containing miR-16-2-3p binding sites, while the mutant reporter had no binding sites. The wild-type reporter, the mutant reporter, the let-7i mimic or the negative controls were transfected into HEK293 cells using Lipo 2000 (RiboBio; China). After 24 h, the cell medium was collected for luciferase activity measurement. Firefly luciferase activity was calculated and normalized to Renilla luciferase activity.

Total RNA was extracted from cardiac tissues and cells. Subsequently, 1 µg of RNA was reverse transcribed into cDNA using HiScriptII Q RT SuperMix (Vazyme; China), and quantitative RT‒PCR was carried out with ChamQ SYBR qPCR Master Mix (Vazyme; China) following the provided protocol. All the results were normalized against glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression.

Cells and exosomes were lysed with lysis buffer containing protease and phosphatase inhibitors, and the concentrations of the proteins were determined using a BCA kit (Thermo Fisher Scientific, USA). Total protein lysates were separated by SDS‒PAGE and transferred to PVDF membranes (Millipore, USA). The membranes were incubated with the corresponding primary antibodies overnight at 4 °C and then exposed to secondary antibodies for 2 h. Western blot bands were visualized using an enhanced chemiluminescence (ECL) kit (Keygene; China). The protein expression levels were normalized to the β-actin levels. The following antibodies from Abcam, USA, were utilized: Acadm, TSG101, HSP70, CD63, and CD9. The relative protein expression was quantified using ImageJ.

HCMECs were seeded in 6-well plates to reach 90% confluence. The cells were scraped with a pipe and washed twice with PBS to remove floating cells. After 24 h, the wound was photographed again, and the scratch width was recorded. The migration distance was subsequently calculated.

HCMECs were seeded in 12-well plates precoated with 100 µl/well of Matrigel (Corning, USA). After 8 h of incubation at 37 °C, morphological images of the tubes were captured using a computer-assisted microscope. The number of junctions and vessel length and area were analyzed using ImageJ.

Lipids were extracted according to the modified method of Bligh and Dyer. Briefly, 50 mg heart tissue samples were homogenized in 500 µl cold phosphate-buffered saline (PBS). Then, lipids were extracted by adding chloroform:methanol (2:1, v/v). Glass tubes (5 mL) were used to avoid polymer contamination. The samples were vortexed for 2 min, followed by 5 min of centrifugation at 1000 rpm after they had allowed to sit still for 20 min. The lower chloroform layer was transferred to a glass syringe and subsequently dried under nitrogen. Lipid samples were subjected to liquid chromatography‒electrospray ionization tandem mass spectrometry (Thermo, CA).