Isolation and identification of BMSCs-Exos

Human BMSCs (SNP-H096, SUNNCELL, Wuhan, China) were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and 100 μg/mL penicillin/streptomycin at 37 °C with 5% CO2. Exos were isolated from BMSCs by ultracentrifugation and labelled with PKH67 (green fluorescence). The morphology of the BMSC-Exos was observed under a transmission electron microscope (JEM-1400 Flash, JEOL, Japan), and the particle size was measured using a Flow NanoAnalyzer (NanoFCM, Xiamen, China). In addition, western blotting was performed to detect the expression of surface markers (CD63 and CD81) in the BMSC-Exos (details are described later).

Establishment of a rat model of HF and treatments

Male Sprague–Dawley (SD) rats (1–3 days old, 180–220 g) purchased from HFK Bioscience (Beijing, China) were acclimatised for a week in the laboratory at 22 ℃ and 50–60% humidity with access to food and water ad libitum. Transverse aortic coarctation (TAC) is commonly used for HF model in previous publications [35, 36], so TAC model was established for the following exploration in our study as previous publications. Specifically, rats were first anaesthetised using pentobarbital sodium (60 mg/kg), and normal breathing was maintained by tracheal cannula (tidal volume = 2–3 ml/100 g; frequency = 60–80 times/minute; respiration ratio = 1:1). Then, the chest of rat was opened, the aorta was separated, and a silk suture was used to ligate the arterial segment between the truncus brachiocephalicus and aortic root. The constriction maintained ~ 70% of the original diameter of the aorta (note: less than 70% of the diameter easily leads to sudden cardiac arrest). The animal experiments were approved by the ethical committee of Xiamen University in accordance with the Guide for the Care and Use of Laboratory Animals (XMULAC20220034-18).

All the rats were divided into following groups: sham, Model + PBS, Model + Exos (BMSC-Exos), Model + ferrostatin-1 (Fer-1, ferroptosis inhibitor), Model + Exos + Fer-1, and Model + Exos + GAS5 (lenti-oe-GAS5), with three replications in each group. Rats that underwent surgery without TAC were assigned to the sham group; the model rats were intra -myocardially injected with 50 μL PBS; the rats in Model + Exos group were injected 50 μL PBS containing BMSCs-Exo; the rats in Model + Fer-1 group were injected 50 μL PBS containing Fer-1 (2 mg/kg, ABclonal RM02804, Wuhan, China); the rats in Model + Exos + Fer-1 group were injected 50 μL PBS containing BMSCs-Exo and Fer-1. The concentration of Fer-1 referenced a publication [37]. After three weeks, the animals were killed for the subsequent experiments.

Evaluation of cardiac function

After modelling for 21 days, cardiac function parameters, including left ventricular end-diastolic diameter (LVEDD), end-systolic diameter (LVESD), ejection fraction (EF), and fractional shortening (FS), were measured by echocardiography using an animal ultrasound imaging system (VisualSonics Vevo 2100, Toronto, Canada).

Histopathological evaluation

After measuring cardiac function, the rats were anaesthetised and killed by cervical dislocation. Myocardial tissues were resected, fixed in 10% formaldehyde, embedded in paraffin, and sliced into 5 μm sections. After dewaxing and rehydration, the sections were stained with haematoxylin–eosin (HE) (Beyotime, Beijing, China) to determine pathological damage, MASSON (Haematoxylin, Masson's ponceau acid fuchsin solution, and aniline blue; Solarbio, Beijing, China) to determine fibrosis, Prussian blue (Solarbio) to determine iron deposition, or terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL; Beyotime) to determine apoptosis. A part of each section was used to detect UL3 expression by immunohistochemistry (IHC). Stained sections were observed under a microscope (Olympus, Tokyo, Japan).

Establishment of a hypoxia model in myocardial cells and treatments

H9C2 cells, a rat myocardial cell line (American Type Culture Collection, Manassas, VA, USA), is used for HF cell model construction [38, 39], and were cultured in DMEM supplemented with 10% FBS and 100 μg/mL penicillin/streptomycin at 37 °C with 5% CO2. The hypoxia model was established by culturing the cells for 6 h in an atmospheric environment of 95% N2 and 5% CO2 (model cells). H9C2 cells cultured under normal oxygen were used as the normal control (NC) group. The model cells were transfected with different agents according to the following groups: Model + PBS, Model + Exos (BMSCs-Exos), Model + Fer-1, Model + Exos + Fer-1, Model + Exos + GAS5 (lenti-oe-GAS5), Model + Exos + oe-NC (lenti-oe-NC), Model + Exos + Ver (Verteporfin; Adooq, Nanjing, China), Model + Exos + shNC, Model + Exos + shUL3, Model + Exos + shUL3 + oe-NC, and Model + Exos + shUL3 + GAS5. Overexpression vectors carrying GAS5 (oe-GAS5), empty overexpression vector (oe-NC), shRNA-UL3 (sh-UL3), and shRNA-NC (sh-NC) were packaged in lentivirus (GenePharma, Shanghai, China) and transfected into cells using Highgene transfection reagent (ABclonal). The concentration of Fer-1 used in vitro is 1 mmol/L. There were three replications in each group.

Western blotting

Total proteins were extracted from BMSC-Exos, myocardial tissues, and myocardial cells by lysis in RIPA buffer (Beyotime). The protein concentration was measured using a BCA kit (Beyotime), and the proteins were separated by 10% SDS polyacrylamide gel electrophoresis and transferred onto PVDF membranes (Beyotime). Subsequently, the membranes were blocked with 5% non-fat milk and incubated with specific primary antibodies (anti-CD63, -CD81, -UL3, -YAP, -TAZ, -ACSL4, -GPX4, and -GAPDH; 1:2,000, Abcam, Cambridge, UK) overnight at 4 °C, followed by incubation with HRP-conjugated secondary antibody (goat anti-rabbit IgG, 1:5000, Abcam) for 1 h at 25 °C. Protein bands were visualised using an ECL reagent (Thermo Fisher Scientific, Waltham, MA, USA) and observed using a gel imaging system (Tanon 3500, China).

Quantitative real time-polymerase chain reaction (qRT-PCR)

Total RNAs were extracted from BMSC-Exos, myocardial tissues, or myocardial cells using TRIzol reagent (Thermo Fisher Scientific) and immediately reverse-transcribed into cDNAs using a cDNA Synthesis kit (Tiangen, Beijing, China). qRT-PCR was performed on an Mx3000P instrument (Stratagene, Carlsbad, CA, USA) at 95 °C for 3 min, followed by 40 cycles at 95 °C for 12 s and 62 °C for 40 s. Relative expression levels were quantified according to the 2−∆∆Ct method using GAPDH as an internal control. The primers used in qRT-PCR included GAS5 (human)-F, 5′-TTC TGC GTT AGG AAG CCT GG-3′, GAS5 (human)-R, 5′-CAA GCC GAC TCT CCA TAC CC-3′; GAPDH (human)-F, 5′-TGT GGG CAT CAA TGG ATT TGG-3′, GAPDH (human)-R, 5′-ACA CCA TGT ATT CCG GGT CAA T-3′; GAS5 (rat)-F, 5′-AAC TGA CTT TAT GCT TGC CC-3′, GAS5 (rat)-R, 5′-CCA TCT TCC ACC TGT AGG GT-3′; UL3 (rat)-F, 5′-GGT GAC CAG TCG TTG GCA TA-3′, UL3 (rat)-R, 5′-TGC GAT CTT TCT TGA GCG GT-3′; GAPDH (rat)-F, 5′-GCG AGA TCC CGC TAA CAT CA-3′, GAPDH (rat)-R, 5′-CTC GTG GTT CAC ACC CAT CA-3′.

Measurement of oxidative stress parameters, Fe2+, and ATP levels

In the serum and supernatants of model rats and cells, respectively, the levels of oxidative stress parameters (malondialdehyde [MDA], reactive oxygen species [ROS], superoxide dismutase [SOD], and glutathione [GSH]), Fe2+, and ATP levels were measured using commercial kits (MDA/ROS/SOD, Solarbio; GSH, Elabscience, Wuhan, China; Fe2+, BioVision, Milpitas, CA, USA; ATP, Solarbio) as per the manufacturer’s instructions.

Cell counting kit-8 (CCK-8) assay

H9C2 cell viability was determined using the CCK-8 kit (Beyotime). Briefly, H9C2 cells were seeded in 96-well plates and co-cultured with BMSC-Exos (with or without other treatments) for 12 and 24 h, respectively. After incubation with CCK-8 for 2 h, the optical density at 450 nm was measured using a microplate reader (DR-3518G, Hiwell Diatek, Wuxi, China).

Cell cycle assay

The cell cycle of H9C2 cells was detected by flow cytometry. Briefly, the cells were digested with trypsin and fixed in 70% ethanol for 6 h at 4 °C. After incubation with RNase A for 30 min at 37 °C, cells were stained with PI (Beyotime) for 30 min in the dark. Cells in the G1 phase were monitored using a flow cytometer (CytoFLEX S, Beckman), and the relative percentage was analysed using Cell Quest software (BD Biosciences, NJ, USA).

RNA immunoprecipitation (RIP) assay

The target relationship between GAS5 and UL3 was identified using RIP assay. Briefly, cells were lysed in RIPA buffer and incubated with beads conjugated with anti-Ago2 or IgG (Geneseed, Guangzhou, China) for 12 h at 4 °C. After centrifugation and 30 min of incubation with proteinase K at 55 °C, the precipitates were collected, and immunoprecipitated RNAs were extracted using TRIzol reagent (Thermo Fisher Scientific). As mentioned earlier, the relative expression of GAS5 and UL3 was detected by qRT-PCR.

Fluorescence in situ hybridisation (FISH)

The subcellular colocalisation of GAS5 and UL3, as well as UL3 and YAP was detected using a FISH kit (RiboBio, Guangzhou, China). Briefly, cells were fixed in 4% paraformaldehyde for 15 min, permeabilised with 0.1% Triton X-100 for 15 min, soaked in 2 × SSC solution for 30 min, and dehydrated in graded ethanol. After hybridisation with a 1 μg/mL probe for 12 h at 37 °C, the samples were washed in 0.4 × SSC solution containing 0.3% Triton X-100 for 2 min at 65 °C and then in 2 × SSC solution containing 0.1% Triton X-100 for another 2 min at 25 °C. The cells were finally stained with DAPI for 5 min in the dark and observed under a microscope (Olympus).

Statistical analysis

Data are presented as mean ± standard deviation and were statistically analysed using GraphPad Prism 7.0 (GraphPad, San Diego, CA, USA). Comparisons between two and among multiple groups were analysed by t-test and one-way ANOVA followed by Tukey’s test, respectively. Statistical significance was set at P < 0.05.