BMSC-derived EVs alleviate doxorubicin-induced cardiotoxicity via miR-210-3p involving the ACVR1B-SMAD3-PGC-1α signaling axis

Abstract

Background

Doxorubicin-induced cardiotoxicity (DIC) limits the clinical use of doxorubicin, yet no specific therapeutic interventions are available. Although the cardioprotective role of miR-210 and the therapeutic potential of stem cell-derived extracellular vesicles have been reported, the specific downstream effectors and pathways mediating the cardioprotective effects of BMSC-EVs against DIC remain poorly understood.

Methods

DIC models were established in HL-1s and C57BL/6 mice. High-throughput RNA sequencing combined with EVs uptake assay and miRNA FISH staining was performed to screen functional miRNAs delivered by BMSC-EVs. Gain- and loss-of-function approaches were performed using miR-210-3p mimics, inhibitors, and ACVR1B siRNA. Cell viability and death, ROS and mitochondrial ROS, mitochondrial membrane potential, and mitochondrial function were assessed by fluorescence staining and western blotting. Bioinformatic analysis and dual-luciferase reporter assays were further applied to identify the downstream target genes of BMSC-EV-derived miRNAs. Molecular mechanisms involving the ACVR1B/SMAD3/PGC-1α signaling pathway were further investigated by RT–qPCR and Western blot.

Results

This study verified that miR-210-3p acts as one of the critical effector molecules mediating the cardioprotective effects of bone marrow mesenchymal stem cell-derived EVs. Both miR-210-3p overexpression and BMSC-EV treatment markedly alleviated cardiomyocyte death and mitochondrial dysfunction by DIC, whereas inhibition of miR-210-3p aggravated cellular injury. In vivo, miR-210-3p overexpression significantly improved cardiac dysfunction, attenuated myocardial fibrosis, and mitigated mitochondrial dysfunction. Mechanistically, miR-210-3p directly targeted the 1573-nt site within the ACVR1B 3’UTR, thereby suppressing ACVR1B expression. Downregulation of ACVR1B subsequently inhibited SMAD3 phosphorylation, relieved repression of PGC-1α, and restored mitochondrial biogenesis and respiratory function. Moreover, ACVR1B knockdown effectively rescued the aggravated DIC induced by miR-210-3p inhibition.

Conclusions

BMSC-EVs-derived miR-210-3p attenuates doxorubicin-induced mitochondrial damage, an effect that was accompanied by suppression of the ACVR1B–SMAD3 cascade and restoration of PGC-1α. This study proposes a potential “EVs–miR-210-3p–ACVR1B” regulatory axis, highlighting candidate targets for further therapeutic exploration in DIC.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13287-026-05261-5.

Background

Doxorubicin (DOX), a prototypical anthracycline antibiotic, is widely used as a first-line chemotherapeutic agent for various solid tumors and hematological malignancies due to its broad-spectrum antitumor activity [1, 2]. However, its clinical utility is severely limited by multiple adverse effects, including myelosuppression, neurocognitive impairment, and multi-organ toxicity [3]. Among these, cardiotoxicity is the most common and severe complication [4, 5], which not only significantly impairs the quality of life of cancer survivors but also adversely affects long-term prognosis. Three phase III clinical trials have confirmed that 26% of patients develop DOX-related congestive heart failure (CHF) compared with the placebo group [6]. Although the pathogenic mechanisms of DOX-induced cardiotoxicity (DIC) are complex, excessive reactive oxygen species (ROS) generation and subsequent mitochondrial dysfunction are universally recognized as core drivers [7, 8]. Unfortunately, effective therapeutic agents for this condition remain lacking, highlighting an urgent need for novel precision treatment strategies.

Extracellular vesicles (EVs) are bilayer-enclosed vesicles secreted via fusion of intracellular multivesicular bodies with plasma membrane [9]. They carry abundant nucleic acids, proteins and lipids to mediate intercellular molecular delivery and communication [10]. Accumulating evidence has demonstrated that EVs derived from various stem cell types exert cardioprotective effects against DOX-induced cardiomyopathy, but their key effector molecules and specific regulatory mechanisms remain incompletely defined [11–13]. For instance, exosomal miR-144 inhibits hypoxia-induced cardiomyocyte apoptosis via modulating the PTEN/AKT pathway [14], and human umbilical cord mesenchymal stem cell-derived exosomal miR-21 also alleviates cellular damage through this signaling cascade [15]. Various studies have advanced our understanding of the critical role of BMSC-EVs-derived microRNAs (miRNAs) in the treatment of diseases [16–18]. Nevertheless, whether the cardioprotective effects of BMSC-EVs against doxorubicin-induced cardiotoxicity are mediated by functional miRNAs remains largely unexplored. Emerging evidence has highlighted the promising potential of miRNAs in the diagnosis and therapeutic intervention of DIC [19, 20]. However, the functional miRNAs contained within BMSC-EVs and their underlying mechanisms remain to be elucidated.

In the present study, we identified differentially expressed miRNAs in BMSC-derived EVs via small RNA-seq and EVs uptake combined with miRNA FISH assay, and confirmed the direct targeting relationship between miR-210 and the 3’ untranslated region (3’UTR) of ACVR1B using miRDB, miRTarBase, and miRWalk database prediction and dual-luciferase reporter assay. In DOX-treated HL-1s, the regulatory effect of miR-210 on the ACVR1B-SMAD3 pathway was verified by RT-qPCR and Western blot. Its mitochondrial protective effect was further confirmed using JC-1 staining, mitoSOX staining, and Western blot (for mitochondrial function-related proteins). Finally, in a mouse model of doxorubicin-induced cardiotoxicity, in vivo experiments confirmed that miR-210-3p from BMSC-derived EVs attenuates myocardial mitochondrial injury and cardiac damage, an effect that was associated with suppression of the ACVR1B-SMAD3 signaling cascade. This study identifies miR-210-3p as an important effector molecule in BMSC-derived EVs in the DIC context and suggests a potential ACVR1B-SMAD3-mediated mechanism underlying mitochondrial damage, thereby highlighting promising therapeutic targets for future investigation.

Materials and methods

Materials

Mouse bone marrow mesenchymal stem cells (BMSCs) (Cat#: BMMC-C106) and HL-1 cells (Cat#: TCM-C783) were purchased from Suzhou Haixing Biotechnology. Dulbecco’s Modified Eagle Medium (DMEM) (Cat#:PM150210) and Minimum Essential Medium α(MEMα) (Cat#:PM150421) were purchased from Wuhan Pricella Biotechnology. Phosphate buffered saline (PBS) (Cat#: G0002-2 L) was purchased from Wuhan Servicebio Technology. Trypsin (0.25%, containing EDTA) (Cat#: MA0232)、protease inhibitor mix (Cat#: MB2678) and penicillin/streptomycin solution (Cat#: MA0110) were purchased from Dalian Meilun Biotechnology. Methylthiazolyldiphenyl-tetrazolium bromide (MTT) Kit (Cat#: M8180) and radioimmunoprecipitation (RIPA) lysis buffer (Cat#:R0010) were purchased from Beijing Solarbio Science & Technology. Fetal Bovine Serum (FBS) (Cat#: UT82901) was purchased from TOCYTO. TRnaZol Reagent Kit (Cat#:M5101) was purchased from NCM Biotech. Doxorubicin hydrochloride (DOX) (Cat#: T1020) was purchased from TargetMol. HighGene transfection reagent (Cat#:RM09014P) was purchased from ABclonal Technology. BCA protein assay kit (Cat#:KTD3001), CheKine™ Reactive Oxygen Species (ROS) Detection Fluorometric Assay Kit (Cat#: KTB1910), and Mitochondrial Membrane Potential Assay Kit (Cat#: KTA4001) were purchased from Abbkine. MitoSOX Red (Cat#: HY-D1055) was purchased from MedChemExpress (MCE).

Methods

Cell culture

The BMSCs and the HL-1 cells were both purchased from Hysigen Bioscience. BMSCs were routinely maintained in MEMα (PM150421, Pricell, China) supplemented with 10% FBS and 1% penicillin‑streptomycin. HL‑1 cells were cultured in high‑glucose DMEM (PM150210, Pricella, China) containing 10% fetal bovine serum (FBS, UT82901, TOCYTO, USA) and 1% penicillin/streptomycin solution (MA0110, Meilunbio, China). All cells were incubated at 37 °C in a humidified atmosphere with 5% CO₂ according to standard protocols.

Isolation of BMSC-EVs by ultracentrifugation

When BMSCs reached 90% confluence, they were cultured in serum-free basal MEMα for 48 h, and the supernatant was collected. The supernatant was sequentially filtered through a 0.8 μm membrane and centrifuged at 4 °C at 300 × g for 10 min, 2,000 × g for 15 min, and 10,000 × g for 30 min to remove cells, debris, and large vesicles. Subsequently, the supernatant was sterile-filtered through a 0.22 μm membrane and ultracentrifuged at 120,000 × g for 120 min at 4 °C. The supernatant was discarded, and the pellet was resuspended in 100–200 µL of sterile PBS. After aliquoting, the samples were stored at − 80 °C. The extracted EVs were characterized by transmission electron microscopy (TEM) for morphological observation, and western blot analysis for detection of specific marker proteins of EVs.

Animal studies

Wild-type male C57BL/6 mice (6–8 weeks old, weighing 20–25 g) were obtained from Beijing HFK Biotechnology Co., Ltd. All animals were housed in a specific pathogen‑free (SPF) barrier facility at the Laboratory Animal Center of Qingdao University (temperature: 22 ± 2 °C; 12 h light/dark cycle) with free access to food and water. All animal experimental procedures were approved by the Laboratory Animal Welfare and Ethics Committee of Qingdao University and was in line with the ARRIVE Guidelines 2.0. The number of animals used per experiment is shown in the figure legend. During the experimental procedures, the mice were anesthetized by inhalation of 2% isoflurane, and euthanized by exposure to CO₂ at the end of the study. Mice were randomly assigned to each experimental group. The acute myocardial injury model was established by intraperitoneal injection of DOX (10 mg/kg) on day 0 and day 7, resulting in a cumulative dose of 20 mg/kg. One day prior to the initial DOX administration, mice were administered BMSC-EVs (50 µg) or normal saline via the tail vein, with injections repeated every 3 days for a total duration of 14 days. For AAV9 experiment, before received DOX treated, AAV9 was administered by tail vein for 4 weeks. At the experimental endpoint, mice were anesthetized and subjected to echocardiography. Subsequently, heart tissues were harvested.

Echocardiography

Mice were anesthetized with 1.5% isoflurane and secured on a temperature-controlled platform. The VINNO 6 LAB (VINNO, China) ultrasound imaging system was used to acquire M-mode images at the left ventricular short-axis view. Based on end-diastolic and end-systolic internal diameters, left ventricular ejection fraction (LVEF) and fractional shortening (LVFS) were calculated to assess cardiac function.

Cell transfection

The microRNA-210-3p mimic (CTGTGCGTGTGACAGCGGCTGA), microRNA-210-3p inhibitor, and their respective negative controls were synthesized by Shanghai GenePharma Company. Acvr1b siRNA and adAcvr1b were synthesized by Hunan fenghbio Company. All were dissolved in sterile, RNase-free DEPC water to prepare a 20 µmol/L stock solution, aliquoted, and stored at -80 °C for subsequent use. HL-1 cells were seeded into 6-well plates at a density of 5.0 × 10⁴ cells per well and cultured in complete DMEM supplemented with 10% FBS at 37 °C in a humidified atmosphere containing 5% CO₂. Transfection was performed using the HighGene Transfection Reagent (RM09014P, Abclonal, China) when cell confluence reached 70–80%, strictly according to the manufacturer’s protocol. To establish a DOX-induced cellular injury model, the successfully transfected cells were divided into corresponding treatment groups. The culture medium was then replaced with fresh basal DMEM containing 2 µM DOX, and the cells were further incubated for 24 h. Following this treatment, the cells were harvested for subsequent experimental analyses.

Quantitative real-time polymerase chain reaction (RT-qPCR)

Total RNA was extracted from HL‑1 cells using the TRnaZol Reagent Kit (M5101, Ncmbio, China). Total RNA was reverse transcribed into complementary DNA (cDNA) using strand-specific kits according to the manufacturer’s instructions. For microRNA (miRNA) analysis, first-strand cDNA was synthesized from equal amounts of total RNA using the miRNA 1st Strand cDNA Synthesis Kit (by tailing A) (MR201-01, Vazyme, China). For messenger RNA (mRNA) analysis, cDNA was synthesized using the HiScript III RT SuperMix for qPCR kit (RK20429, Abclonal, China). Quantitative PCR was performed using the 2× Universal SYBR Green Fast qPCR Mix (RK21203, Abclonal, China) on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, USA). The thermal cycling conditions were as follows: initial denaturation at 95 °C for 3 min, followed by 40 cycles of denaturation at 95 °C for 3s and combined annealing/extension at 60 °C for 10 s. A melting curve analysis (65 °C to 95 °C, increment 0.5 °C per 5 s) was performed at the end of each run to confirm the specificity of amplification. The expression levels of target genes were normalized to the endogenous reference gene GAPDH (for mRNA) or U6 snRNA (for miRNA). The relative expression was calculated using the comparative 2^−ΔΔCt method. All the primer sequences involved in this study are listed in Supplementary Tables 2–3.

Western blot

Total protein was extracted from animal heart and HL-1 cells of each group using RIPA lysis buffer, and protein concentrations were determined using the BCA method. Samples containing equal amounts of protein were loaded per lane, separated by SDS-PAGE, and transferred onto PVDF membranes via the wet transfer method. The membranes were blocked with 5% skim milk for 1 h at room temperature, followed by overnight incubation at 4 °C with the corresponding primary antibodies. After washing with TBST, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. and then incubated overnight at 4 °C with the following primary antibodies: ACVR1B (1:1000), PGC-1α (1:1000), Smad3 (1:1000), pSmad3 (1:1000), SDHB (1:1000), NDUFS1 (1:2000), ATP5A1 (1:1000), COX IV (1:2000), and internal reference β-actin (1:10000) or GAPDH (1:10000). After washing with TBST, the membranes were incubated with an HRP-conjugated secondary antibody (Jackson ImmunoResearch, 1:10000) diluted in TBST with 5% non-fat milk for 1 h at room temperature. Finally, signals were visualized using an ECL detection reagent and analyzed by band density quantification with ImageJ software. For quantitative analysis, band intensities were measured using ImageJ software (NIH). The intensity value of each target protein band was first normalized to that of the corresponding loading control (β-Tubulin, GAPDH or β-actin) from the same sample. Subsequently, the relative protein expression was calculated as the ratio of the normalized target protein value in each experimental group to that of the designated control group, which was set to 1. All experiments were repeated at least three independent times, and the normalized relative expression levels were used for statistical comparisons.

Cell viability

Cell viability was assessed using the Methylthiazolyldiphenyl-tetrazolium bromide (MTT, M8180, Solarbio, China) assay according to the manufacturer’s protocol. MTT powder was dissolved in deionized water to prepare a 5 mg/mL stock solution. HL-1 cells were seeded in a 96-well plate at a density of 5 × 10³ cells per well. After 24 h of incubation for adherence, the cells were treated with 2 µM DOX for 24 h. Subsequently, medium containing MTT at a final concentration of 0.5 mg/mL was added to each well, and the plate was incubated at 37 °C for 4 h. The medium was then aspirated, and 150 µL of DMSO was added to each well to dissolve the formazan crystals. Absorbance was measured at 490 nm using a microplate reader, and the relative cell viability was calculated.

JC-1, ROS, mitoSOX immunofluorescent staining

Mitochondrial membrane potential was assessed using the JC‑1 fluorescent probe. HL‑1 cells were seeded into 6‑well plates at a density of 1 × 10⁴ cells/well and treated accordingly. After treatment, the medium was removed, and cells were washed with PBS. Each well was then incubated with 1 mL of working solution containing JC‑1 (final concentration 5 µg/mL) at 37 °C in the dark for 60 min. Following incubation, cells were washed twice with pre‑warmed JC‑1 buffer. Images were acquired under a fluorescence microscope using 488 nm excitation, with emission collected at 530 nm (monomeric form, green fluorescence) and 590 nm (aggregate form, red fluorescence). The red/green fluorescence intensity ratio was calculated to quantitatively evaluate mitochondrial membrane potential levels. To determine the reactive oxygen species (ROS) levels of the cells and sections, DCFH-DA and MitoSOX Red were added to stain the cells and sections. Sections were imaged using a fluorescence microscope.

Histological staining

Heart samples were embedded in OCT and cut cross-sectionally into 5-µm-thick sections. Hematoxylin and Eosin (H&E) staining was performed for histological examination. Sirius Red Stain Kit and Masson’s Trichrome Stain Kit were used to identify collagen fibers. For H&E staining, sections were fixed using 4% paraformaldehyde. For Sirius red staining, sections were fixed and sequentially incubated in picrosirius red solution and 0.5% acetic acid solution. For Masson’s trichrome staining, sections were fixed and sequentially incubated in ponceau-acid fuchsin solution and phosphomolybdic acid solution. Then, the sections were dehydrated with an ethanol gradient and sealed with neutral gum. After staining, images were taken with the Pannoramic MIDI (3DHISTECH, Hungary) and analyzed in the infarct area and myocardial fibrosis area using ImageJ (NIH, Bethesda, MD, USA). For each heart, three to five different sections were taken. The infarcted area and myocardial fibrosis value were determined by the percentage of positive staining area in the whole left ventricle area in each heart section.

Dual-luciferase reporter assay

The sequence of miR-210-3p and the corresponding Acvr1b mutants were cloned and inserted into the 3’UTR of psiCHECK-2 plasmid (Searching Biotechnology). In six-well plates, 293 T cells were cultured to approximately 70% confluence and then co-transfected with either wild type or mutant luciferase reporter vector (2 µg) and either mimic miRNAs or negative control (NC) (2 µg). After 48 h, luciferase activity was measured and normalized to the activity of Rluc.

miR-210-3p-deficient BMSC-EVs obtaining

To knock down endogenous miR-210-3p expression in BMSCs, lentiviral vectors carrying miR-210-3p short hairpin RNA (shRNA) or negative control shRNA were transfected into BMSCs. shRNA was purchased from fenghbio. Upon successful construction of miR-210-3p-knockdown BMSCs, the cells were cultured in serum-free basal MEMα. Subsequently, cell culture supernatants were collected, and BMSC-EVs were isolated and purified following the EV extraction procedure described in detail above. Total miRNAs were extracted from ultracentrifugation-isolated BMSC-EVs using EVs-specific extraction kit. The extracted miRNAs were performed to detect and verify the expression level of miR-210-3p in isolated EVs. The obtained EVs derived from miR-210-3p-deficient BMSCs were used for subsequent functional experiments.

EVs uptake and miR-210-3p fish colocalization staining

For EVs uptake and miR-210-3p colocalization staining, BMSC-EVs were first fluorescently labeled with PKH26 (C2071, Beyotime, China) dye according to standard protocols. After termination of the labeling reaction, PKH26-labeled BMSC-EVs were co-cultured with serum-starved HL-1 cardiomyocytes for 24 h to allow cellular internalization. Subsequently, RNA fluorescence in situ hybridization (FISH) was performed using an RNA FISH kit (Bes1001, Bersinbio, China) to detect the intracellular distribution of miR-210-3p. Cells were fixed, permeabilized, and dehydrated after EV uptake. Following pre-hybridization, cells were incubated with a specific miR-210-3p probe for overnight hybridization. After that, cell nuclei were counterstained with DAPI. Finally, the samples were mounted and observed via fluorescence microscope to assess the cellular uptake of BMSC-EVs and the intracellular colocalization of miR-210-3p.

Statistical analysis

Experimental data are presented as mean ± standard deviation. Statistical analyses were performed using GraphPad Prism 9.0 software. Comparisons among multiple groups were conducted using two-way analysis of variance (ANOVA), followed by Tukey’s test for post hoc pairwise comparisons. Comparisons between two groups were performed using unpaired Student’s t-test. *P-value < 0.05 was considered statistically significant.

Results

BMSC-derived EVs attenuate doxorubicin-induced cardiotoxicity

To explore the protective effect of Bone marrow mesenchymal stem cell (BMSC) - derived EVs against doxorubicin-induced cardiotoxicity, we isolated EVs from BMSC culture supernatant via ultracentrifugation. Subsequent identification verified their typical morphology (Fig. 1A). and high expression of EV-specific markers CD9 and CD81 (Fig. 1B). To investigate the protective effects of BMSC-derived EVs against DOX-induced cardiotoxicity, we established mouse cardiac injury models via intraperitoneal DOX injection at 10 mg/kg on day 0 and day 7. Mice received tail vein injection of 50 µg EVs or saline one day ahead of initial DOX treatment, with repeated administration every 3 days for 14 days (Fig. 1C). As shown in Fig. 1D–F, BMSC-derived EV treatment effectively alleviated DOX-triggered cardiac dysfunction and restored core cardiac functional indicators including LVEF and LVFS in mice. Compared with saline-treated mice, DOX-treated mice exhibited disorganized myocardium and increased perivascular/interstitial fibrosis, which were significantly alleviated by BMSC-EVs treatment (Fig. 1G–I). In vitro, MTT assay results demonstrated that BMSC-EVs treatment significantly enhanced the viability of DOX - treated cardiomyocytes (Fig. 1J). Propidium iodide (PI) staining indicated that BMSC-EVs treatment markedly reduced DOX - induced cardiomyocyte death (Fig. 1K). Additionally, ROS-specific probe DCHF-DA staining revealed that BMSC-EVs treatment significantly attenuated the elevated oxidative stress levels in DOX - treated cardiomyocytes (Fig. 1L). Collectively, these findings confirm that BMSC - derived EVs alleviate DOX - induced cardiac injury both in vivo and in vitro.

Fig. 1.

BMSC-derived EVs attenuate DIC. A Representative TEM image of BMSC-EVs showing typical cup‑shaped morphology; scale bar = 500 nm. B Representative western blots of the EVs markers CD81, CD9, and β-Tubulin in BMSC and BMSC-derived EVs. C experimental timeline and design of the DOX mice treated with BMSC-EVs. D–F Echocardiographic analysis of LV ejection fraction (LVEF) and left ventricular fractional shortening (LVFS) of mice (n = 5). G Hematoxylin and eosin-stained heart of mice. H Masson trichrome staining, and I Quantification of fibrosis area (n = 5, scale bar = 50 μm). J MTT assay for cell viability in BMSC-EVs vs. Vehicle treated DOX HL-1s (n = 7). K Representative PI-stained images and quantification of cell death levels of HL-1s (n = 6, scale bar = 100 μm). L, representative images of DCFH-DA-stained HL-1s and quantification of oxidative stress levels (n = 6, scale bar = 100 μm). Data are shown as mean ± SD. D through L, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

miR-210-3p serves as a key molecule mediating BMSC-EVs-regulated mitochondrial function in DIC

To identify key effectors underlying the therapeutic efficacy of BMSC-EVs against DIC, we performed RNA sequencing on cardiomyocytes from DOX-exposed and BMSC-EVs-treated groups. Sequencing analysis identified 25 upregulated and 25 downregulated miRNAs in the BMSC-EVs-treated group compared with the DOX group (Fig. 2A, Supplementary Table 1). We next explored whether the miRNAs within BMSC-EVs are responsible for their protective functions. Fourteen upregulated miRNAs with relatively high abundance were subjected to RT-qPCR verification. RT-qPCR results revealed that miR-210-3p was markedly upregulated in BMSC-EV-treated cardiomyocytes relative to the DOX group, consistent with the RNA-seq data (Fig. 2B). We further generated miR-210-3p knockdown BMSC-EVs (Supplementary Fig. 1). As shown in Fig. 2C, BMSC-EVs-WT markedly increased miR-210-3p expression in cardiomyocytes, whereas this effect was abolished in miR-210-3p-depleted BMSC-EVs. EVs uptake assay and FISH colocalization confirmed that intracellular miR-210-3p in cardiomyocytes was not endogenous but derived from BMSC-EVs (Fig. 2C). Gene Ontology (GO) enrichment analysis revealed that the differentially expressed miRNAs (DEMs) upregulated by BMSC-EVs were primarily enriched in the regulation of metabolism and mitochondrial biological processes (Fig. 2D). Accordingly, we further evaluated the level of mitochondrial dysfunction in both HL-1 cardiomyocytes and mouse cardiac tissues. Mitochondrial reactive oxygen species (mitoROS) were significantly increased in DOX-treated cells, which was markedly attenuated by BMSC-EVs treatment (Supplementary Fig. 2A). JC-1 staining showed the elevated red/green fluorescence ratio, an indicator of mitochondrial membrane potential (ΔΨm), indicating recovered mitochondrial membrane potential by BMSC-EVs (Supplementary Fig. 2B). Western blot results revealed that BMSC-EVs treatment notably upregulated the expression of mitochondrial electron transport chain (ETC) subunits NDUFS1, SDHB, ATP5A1 and COX IV both in vitro and in vivo (Supplementary Fig. 2C and Supplementary Fig. 3). Collectively, these results suggest that BMSC-EVs attenuate DIC, at least in part, by modulating mitochondrial damage, with EV-loaded miR-210-3p functioning as an important contributor to this protective effect.

Fig. 2.

Identification of miR-210-3p is a potential target of BMSC-EVs in DIC. A Heatmap of differentially expressed microRNA in Vehicle (n = 3) or BMSC-EVs (n = 3)-treated in DIC. B microRNA levels of in Ctrl, DOX and BMSC-EVs treated HL-1s by RT-qPCR (n = 4). C Representative images of FISH staining for miR-210-3p combined with BMSC-EVs labeling-PKH26. Green: miR-210-3p; Red: PKH26; Blue: DAPI. Scale bar = 50 μm. D Enrichment analysis of biological process (BP) of upregulated microRNA in BMSC-EVs vs. Vehicle treated DOX HL-1s by RNA-seq. Data are shown as mean ± SD. A, unpaired Student t-test. B, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01

Overexpression of miR-210-3p attenuates DOX-induced mitochondrial damage

To further elucidate the regulatory mechanism and correlation of miR-210-3p overexpression with DIC and mitochondrial dysfunction, we specifically upregulated miR-210-3p in cardiomyocytes via transfection with miR-210-3p mimics. PI staining revealed that the rate of cardiomyocyte death was markedly reduced in the miR-210-3p overexpression group compared with the DOX group (Fig. 3A). Consistently, MTT assays confirmed that cardiomyocyte viability was markedly improved in the miR-210-3p overexpression group relative to the DOX group (Fig. 3B), demonstrating that miR-210-3p alleviates DOX-induced cardiomyocyte injury. DCHF-DA staining showed a remarkable increase in intracellular ROS in DOX-treated cells, while miR-210-3p overexpression effectively reversed this elevation (Fig. 3C). Mitochondrial ROS levels, assessed using the MitoSOX Red probe, were substantially higher in the DOX group, and this increase was greatly diminished upon miR-210-3p overexpression (Fig. 3D). These results indicate that miR-210-3p helps maintain mitochondrial redox homeostasis in cardiomyocytes. JC-1 staining further revealed that the red/green fluorescence ratio was significantly higher in the miR-210-3p overexpression group than in the DOX group (Fig. 3E), reflecting restored ΔΨm. Western blot analysis demonstrated that the protein levels of ETC subunits, including NDUFS1, ATP5A1, SDHB and COX IV, were markedly upregulated following miR-210-3p overexpression (Fig. 3F). Taken together, miR-210-3p overexpression protects cardiomyocytes against DOX-induced damage by repressing oxidative stress, restoring mitochondrial membrane potential, and upregulating the expression of respiratory chain complexes.

Fig. 3.

miR-210-3p overexpression improves DOX induced cell death and mitochondrial dysfunction. A Representative PI-stained images and quantification of cell death levels in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 6, scale bar = 100 μm). B MTT assay for cell viability in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 8). C Representative images of DCFH-DA-stained HL-1s and quantification of oxidative stress levels in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 6, scale bar = 100 μm). D Representative images of mitoSOX-stained HL-1s and quantification of oxidative stress levels in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 6, scale bar = 100 μm). E Representative JC-1-stained HL-1s and quantification of mitochondrial membrane potential levels in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 5, scale bar = 100 μm). F Representative western blotting images, and quantitative analysis of ETC-related proteins in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 5). Data are shown as mean ± SD. A through F, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

miR-210-3p deficiency exacerbates DIC by aggravating mitochondrial dysfunction

To clarify the causal relationship between miR-210-3p and DIC, miR-210-3p expression was specifically knocked down in HL-1 cardiomyocytes using a targeted inhibitor, and the regulatory effect of miR-210-3p silencing on DOX-triggered myocardial injury was systematically explored. PI staining demonstrated that miR-210-3p deficiency induced cardiomyocyte death and further exacerbated DOX-mediated myocardial cell injury (Fig. 4A). Consistent with these findings, MTT assays showed that miR-210-3p inhibition markedly reduced cardiomyocyte viability. Moreover, combined treatment with miR-210-3p inhibitor and DOX exerted a more severe inhibitory effect on cell viability than DOX treatment alone (Fig. 4B). miR-210-3p deficiency significantly increased cellular and mitochondrial ROS accumulation, reduced ΔΨm, and aggravated oxidative stress and mitochondrial dysfunction in cardiomyocytes. Furthermore, miR-210-3p deficiency further amplified DOX-induced oxidative stress disorder and ΔΨm (Fig. 4C and E). Western blot analysis indicated that the downregulation of miR-210-3p prominently suppressed the expression of ETC subunits, including NDUFS1, ATP5A1, SDHB, and COX IV. In addition, the combination of miR-210-3p inhibitor and DOX treatment further aggravated the DOX-induced reduction in ETC protein levels (Fig. 4F). These results suggest that miR-210-3p deficiency triggers mitochondrial oxidative respiratory dysfunction and ultimately deteriorates DOX-induced myocardial injury.

Fig. 4.

miR-210-3p deficiency aggravates DOX induced cell death and mitochondrial dysfunction. A Representative PI-stained images and quantification of cell death levels of NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 6, scale bar = 100 μm). B MTT assay for cell viability in NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 7). C Representative images of DCFH-DA-stained HL-1s and quantification of oxidative stress levels in NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 5, scale bar = 100 μm). D Representative images of mitoSOX-stained HL-1s and quantification of oxidative stress levels in NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 6, scale bar = 100 μm). E Representative JC-1-stained HL-1s and quantification of mitochondrial membrane potential levels in NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 6, scale bar = 100 μm). F Representative western blotting images, and quantitative analysis of ETC-related proteins in NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 5). Data are shown as mean ± SD. A through F, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

miR-210-3p alleviates DIC by targeting the 3’UTR of ACVR1B to regulate mitochondrial dysfunction

To elucidate the downstream molecular mechanism by which miR-210-3p regulates doxorubicin (DOX)-induced cardiotoxicity, we first predicted potential target genes of miR-210-3p using three independent bioinformatics databases, namely miRDB, miRTarBase, and miRWalk. A total of 6 candidate target genes were identified through cross-analysis of the prediction results (Fig. 5A). Literature review revealed that ACVR1B is closely associated with mitochondrial biogenesis and energy metabolism, which is consistent with the regulatory role of miR-210-3p in mitochondrial function observed in our previous experiments. Thus, ACVR1B was selected as a candidate target gene for subsequent verification. To confirm the direct binding relationship between miR-210-3p and ACVR1B, we further predicted the potential binding sites in the 3’ untranslated region (3’UTR) of ACVR1B using starBase and miRDB databases (Fig. 5B). A mutant ACVR1B 3’UTR construct was generated by site-directed mutagenesis of the predicted binding site. Dual-luciferase reporter assay results demonstrated that co-transfection of miR-210-3p mimics significantly reduced the luciferase activity of the wild-type ACVR1B 3’UTR reporter plasmid, whereas no significant change was observed in the luciferase activity of the mutant reporter plasmid. These findings confirmed that miR-210-3p directly binds to the 1573 site within the 3’UTR of ACVR1B (Fig. 5C).

Fig. 5.

miR-210-3p target gene screening and regulatory mechanism validation. A Venn diagram of three independent miRNA differential analysis databases reveals 6 possible mRNAs directly targeting miR-210-3p. B Sequence of ACVR1B 3’ UTR with miR-210 targeting sites. C Luciferase reporter assay to evaluate the direct binding of miR-210 with the 3’ UTR of ACVR1B (n = 3). D Acvr1b mRNA expression levels in NC mimics or miR-210-3p mimics-transfected HL-1s (n = 4). E Acvr1b mRNA expression levels in NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 4). F Acvr1b mRNA expression levels in BMSC-EVs vs. Vehicle treated DOX HL-1s (n = 4). G Representative western blotting images and quantitative analysis of NC mimics or miR-210-3p mimics-transfected HL-1s (n = 3). H Representative western blotting images and quantitative analysis of NC inhibitor or miR-210-3p inhibitor-transfected HL-1s (n = 3). Data are shown as mean ± SD. A through H, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

We then investigated the regulatory effect of miR-210-3p on ACVR1B expression at both mRNA and protein levels. At the mRNA level, overexpression of miR-210-3p significantly inhibited the DOX-induced upregulation of ACVR1B expression (Fig. 5D), whereas inhibition of miR-210-3p further exacerbated the DOX-induced increase in ACVR1B expression (Fig. 5E). Of note, BMSC-EVs recapitulated the inhibitory effect of miR-210-3p overexpression and efficiently blunted the DOX-mediated increase in ACVR1B mRNA levels. Consistently, Western blot analysis confirmed that miR-210-3p overexpression significantly reduced the DOX-induced elevation of ACVR1B protein levels (Fig. 5G), while inhibition of miR-210-3p further enhanced the DOX-induced upregulation of ACVR1B protein expression (Fig. 5H).

miR-210-3p attenuates DOX-induced mitochondrial dysfunction through a potential ACVR1B–SMAD3–PGC-1α axis

To clarify the functional impact of ACVR1B on HL-1 cardiomyocytes, we achieved overexpression of ACVR1B in cardiomyocytes using adenoviral vectors. MTT assay demonstrated that overexpression of ACVR1B significantly reduced cardiomyocyte viability (Supplementary Fig. 4A). Consistently, PI staining revealed that ACVR1B overexpression markedly increased cardiomyocyte death (Supplementary Fig. 4B). Notably, ACVR1B overexpression led to a significant elevation in mitochondrial reactive oxygen species (ROS) levels (Supplementary Fig. 4C) and a marked reduction in ΔΨm (Supplementary Figs. 4D-4E). Western blot analysis showed that overexpression of ACVR1B resulted in significant downregulation of key subunit proteins of ETC, including NDUFS1, ATP5A1, SDHB and COX IV (Supplementary Figs. 4G-4 K). Since ACVR1B typically exerts its biological functions by mediating SMAD3 phosphorylation to inhibit PGC-1α, we further examined the changes in SMAD3 phosphorylation and PGC-1α expression. Western blot results confirmed that ACVR1B overexpression significantly enhanced SMAD3 phosphorylation and suppressed PGC-1α expression (Supplementary Figs. 4G and 4 L-4M).

To further explore the regulatory relationship between miR-210-3p and its target gene ACVR1B in DIC, we treated cardiomyocytes with miRNA inhibitor and small interfering RNA (siRNA) and divided the cells into four groups: siNC + NC inhibitor, siNC + miR-210-3p inhibitor, siACVR1B + NC inhibitor, and siACVR1B + miR-210-3p inhibitor. PI staining and MTT assay results showed that targeted knockdown of ACVR1B significantly reversed the cardiomyocyte injury induced by doxorubicin (DOX) and miR-210-3p inhibition (Fig. 6A and B). Specifically, ACVR1B knockdown notably reduced the elevated mitochondrial ROS levels induced by DOX and miR-210-3p inhibition (Fig. 6C and D) and prevented the decrease of ΔΨm (Fig. 6E and F). Western blot analysis revealed that ACVR1B knockdown significantly attenuated the DOX- and miR-210-3p inhibition-induced downregulation of ETC subunit proteins (NDUFS1, ATP5A1, and COX IV) (Fig. 6H and L). Additionally, ACVR1B knockdown significantly reversed the increased SMAD3 phosphorylation and decreased PGC-1α expression caused by DOX and miR-210-3p inhibition (Fig. 6H, M and N). These results support that ACVR1B functions as a downstream effector of miR-210-3p. The protective effects of miR-210-3p against DOX-induced cardiomyocyte injury were associated with suppression of ACVR1B and SMAD3 phosphorylation, along with restoration of PGC-1α expression and mitochondrial molecular markers. These observations suggest a potential miR-210-3p–ACVR1B–SMAD3–PGC-1α signaling axis that may contribute to the regulation of mitochondrial damage in this model.

Fig. 6.

miR-210-3p targeting ACVR1B inhibits mitochondrial damage in DIC. A Representative PI-stained images and B MTT assay for cell viability in NC inhibitor, miR-210-3p inhibitor, siNC+ siACVR1B or miR-210-3p inhibitor+siACVR1B-transfected HL-1s (n = 9). C Representative images of mitoSOX-stained and D Quantification of oxidative stress levels in NC inhibitor, miR-210-3p inhibitor, siNC+ siACVR1B or miR-210-3p inhibitor+siACVR1B-transfected HL-1s (n = 5, scale bar = 100 μm). E Representative JC-1-stained HL-1s and F quantification of mitochondrial membrane potential levels (n = 5, scale bar = 100 μm). G Representative western blotting images, H Quantitative analysis of ACVR1B, I–L quantitative analysis of ETC-related proteins, and M, N quantitative analysis of PGC-1α and phosphorylated SMAD3 in HL-1s (n = 3). Data are shown as mean ± SD. A through N, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Overexpression of mir-210-3p ameliorates doxorubicin-induced cardiac dysfunction and mitochondrial impairment

To investigate the regulatory role of miR-210-3p in the pathological progression of doxorubicin (DOX)-induced myocardial injury in vivo, 7-week-old male C57BL/6 mice were used as experimental subjects. Mice were intravenously injected with adeno-associated virus serotype 9 carrying the miR-210-3p gene under the control of the cardiac troponin T promoter to achieve cardiac-specific overexpression of miR-210-3p. An acute myocardial injury model was established by intraperitoneal injection of DOX (10 mg/kg) on day 0 and day 7 (Fig. 7A). Subsequent assessments included cardiac function evaluation, histopathological examination and mitochondrial function-related indicators. Echocardiographic results demonstrated that overexpression of miR-210-3p significantly ameliorated DOX-induced cardiac dysfunction in mice: left ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS) were effectively preserved, and systolic ventricular dilation was notably attenuated (Fig. 7B and C). Consistently, histopathological analyses revealed that the myocardial injury area (H&E staining, Fig. 7E) and fibrotic area (Masson staining, Fig. 7F and G) in the AAV9-miR-210-3p + DOX group were significantly smaller than those in the AAV9-NC + DOX group.

Fig. 7.

miR-210-3p suppresses mitochondrial Dysfunction and cardiac dysfunction in DIC. A Experimental timeline and design of the DOX mice treated with AAV9-NC and AAV9-miR-210-3p. B–D Echocardiographic analysis of LV ejection fraction (LVEF) and left ventricular fractional shortening (LVFS) of mice (n = 5). E Representative images of hematoxylin and eosin-stained mouse heart tissue (scale bar = 50 μm). F Masson trichrome staining, and G Quantification of fibrosis area (n = 5, scale bar = 50 μm). H Representative western blotting images, and I quantitative analysis of ETC-related proteins in mouse heart tissue (n = 3). J Representative western blotting images, and K quantitative analysis of ACVR1B, PGC-1α and phosphorylated SMAD3 in mouse heart tissue (n = 3). Data are shown as mean ± SD. A through J, two-way ANOVA followed by the Bonferroni multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Further assessments of molecular phenotypes linked to mitochondrial injury verified that miR-210-3p regulates doxorubicin-triggered mitochondrial damage. Western blot analysis of myocardial tissues revealed that miR-210-3p overexpression reversed the DOX-induced downregulation of mitochondrial electron transport chain complex proteins (NDUFS1, ATP5A1, SDHB and COX IV), indicating improved molecular signatures associated with mitochondrial respiration (Fig. 7H and I). In addition, miR-210-3p overexpression suppressed DOX-mediated upregulation of ACVR1B and phosphorylation of SMAD3. This was accompanied by a reversal of PGC-1α suppression and a restoration of markers related to impaired mitochondrial biogenesis, suggesting a potential link between the ACVR1B-SMAD3 cascade and mitochondrial injury (Fig. 7J and K). Collectively, in vivo experiments demonstrated that miR-210-3p overexpression alleviates doxorubicin-induced cardiac dysfunction and mitochondrial damage in mice, an effect that correlated with improved molecular signatures of mitochondrial biogenesis and electron transport chain proteins. These data, coupled with the validated targeting of ACVR1B, propose a potential miR-210-3p–ACVR1B–SMAD3–PGC-1α signaling axis that may critically govern mitochondrial injury under doxorubicin-induced cardiotoxicity.

Discussion

Despite its potent antitumor activity, DOX is constrained by dose-dependent cardiac toxicity in clinical practice. The lack of specific preventive or therapeutic strategies of DIC underscores the clinical significance of exploring therapeutic approaches [4, 5]. BMSC-EVs, as key carriers for intercellular signaling, have been confirmed to play important roles in the pathological regulation of DIC [13], but the core effector molecules and their mechanisms remain incompletely elucidated. This study demonstrated through multi-dimensional experiments that BMSC-EVs can deliver miR-210-3p to target the 3’UTR of ACVR1B, inhibit SMAD3 phosphorylation, relieve its inhibition on PGC-1α, thereby improving mitochondrial dysfunction and ultimately alleviating DIC. This study proposes a potential “BMSC-EVs–miR-210-3p–ACVR1B–SMAD3–PGC-1α–mitochondrial function” regulatory axis, providing candidate molecular targets and therapeutic strategies for further exploration in DIC.

Current studies on BMSC-EVs-based therapy for DIC have predominantly focused on the global protective effects of EVs at the phenotypic level, such as inhibiting cardiomyocyte apoptosis and alleviating oxidative stress [11, 21]. However, the elucidation of key effector miRNAs and their downstream molecular mechanisms underlying these protective effects remains inadequately addressed, which precludes the clinical translation of BMSC-EVs therapy due to the lack of precise therapeutic targets. To address this knowledge gap, the present study systematically screened miRNAs significantly upregulated following BMSC-EVs treatment via small RNA transcriptome sequencing. Combined with Gene Ontology (GO) enrichment analysis, these miRNAs were indicated to be primarily involved in metabolic regulation and mitochondrial biosynthesis. This is consistent with the proposal by Wu et al. that “mitochondrial quality control mechanisms serve as therapeutic targets for doxorubicin-induced cardiotoxicity.” Further verification using quantitative RT-qPCR identified six core miRNAs that were downregulated under DOX treatment and upregulated following EV intervention. Ultimately, functional correlation analysis confirmed miR-210-3p as the key effector molecule. This research design addresses the deficiency of “prioritizing phenotype over mechanism” in existing EVs-based DIC studies. It provides experimental evidence for the targeted optimization of BMSC-EVs therapy, thereby highlighting the innovation of this study in research strategy.

Existing studies have reported that miR-210 exerts a protective role in diseases such as myocardial infarction and heart failure primarily through regulating cardiomyocyte proliferation, inhibiting apoptosis, and improving mitochondrial function [22, 23]. However, in-depth studies on the role and mechanism of miR-210 in DIC are lacking, and whether it acts as a key effector molecule in BMSC-EVs to mediate the protective effect against DIC remains unproven. Notably, mature miR-210 gives rise to two functional strands, miR-210-3p and miR-210-5p, which can exert distinct biological functions in a context‑dependent manner. Our preliminary sequencing and quantitative RT–qPCR data revealed that miR-210-5p was expressed at extremely low levels in bone marrow mesenchymal stem cells, BMSC‑derived EVs, and recipient cardiomyocytes, falling well below the threshold needed to drive meaningful biological regulation. In contrast, miR-210-3p was consistently enriched across all samples examined, supporting its role as a key regulatory mediator. Based on this, we focused all subsequent mechanistic studies exclusively on miR‑210‑3p. While the cardioprotective role of miR-210 has been previously documented, this work extends its functional relevance to the DIC context and identifies miR-210-3p as an important functional cargo of BMSC-EVs. More importantly, we identify the ACVR1B/SMAD3/PGC-1α signaling axis as a potential downstream mechanism linking miR-210-3p to mitochondrial protection. Furthermore, EV-mediated delivery may offer a potential strategy for harnessing miR-210-3p in DIC treatment, circumventing the inherent instability of direct miRNA administration and thus representing a promising avenue for future translational exploration.

In this study, ACVR1B was identified as a target gene of miR-210-3p through cross-prediction using four authoritative databases (miRDB, miRTarBase, miRWalk, and starBase). As a member of the transforming growth factor-β (TGF-β) superfamily, ACVR1B primarily regulates downstream signaling pathways by mediating the phosphorylation of SMAD family proteins [24, 25]. Aberrant activation of the ACVR1B-SMAD3 signaling pathway is closely associated with the pathogenesis and progression of cardiovascular diseases, such as promoting myocardial fibrosis and exacerbating myocardial remodeling [25–28]. PGC-1α, a core regulator of mitochondrial biogenesis, exhibits downregulated expression that induces mitochondrial respiratory chain dysfunction, thereby aggravating cardiomyocyte injury [29–32]. Previous studies have confirmed that phosphorylation of the TGF-β/Smad3 pathway can inhibit PGC-1α and impair mitochondrial biogenesis during diabetes progression [33]. However, the regulatory crosstalk between the ACVR1B-SMAD3 pathway and PGC-1α in DIC, as well as the potential involvement of upstream miRNA-mediated targeted regulation, remains unreported. Utilizing bioinformatics prediction combined with dual-luciferase reporter assay, this study is the first to demonstrate that miR-210-3p directly targets the 3’UTR (position 1573) of ACVR1B to suppress its expression. This discovery supplements a novel upstream molecule to the regulatory network of ACVR1B. Further functional validation revealed that ACVR1B overexpression exacerbates DOX-induced cardiomyocyte injury and mitochondrial dysfunction by promoting SMAD3 phosphorylation and inhibiting PGC-1α expression. In contrast, ACVR1B knockdown reverses the aggravated pathological progression of DIC induced by miR-210-3p inhibition. Collectively, these data propose a potential “miR-210-3p–ACVR1B–SMAD3–PGC-1α” signaling axis in DIC, suggesting that suppression of the ACVR1B–SMAD3 cascade and restoration of PGC-1α are associated with the attenuation of mitochondrial damage by miR-210-3p. These findings highlight a candidate molecular axis that warrants further investigation as a potential therapeutic strategy for DIC.

Several limitations of the present study warrant acknowledgment and refinement in future work. First, the in vitro experiments were performed exclusively using the HL-1 cardiomyocyte cell line. Although HL-1 cells are widely used as a cardiac model, validation in human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs), which more faithfully recapitulate human cardiac physiology, would strengthen the translational relevance of this study. Second, mitochondrial status was primarily assessed using surrogate markers, including mitochondrial ROS, mitochondrial membrane potential, and electron transport chain protein expression, without direct measurement of mitochondrial respiratory function (e.g., oxygen consumption rate). This somewhat limits the depth of mitochondrial functional characterization. Third, although BMSC-EVs were characterized by transmission electron microscopy and immunoblotting for markers such as CD9 and CD81. More robust size distribution data and a more comprehensive marker panel (e.g., including CD63) would further enhance the rigor of EV characterization. Fourth, the identified miR-210-3p–ACVR1B–SMAD3–PGC-1α signaling axis has not yet been validated in clinical specimens from patients with doxorubicin-induced cardiotoxicity, and its clinical translational value remains to be established. Fifth, although we focused on miR-210-3p because of its high abundance, the function of the minor strand miR-210-5p was not examined, and a potential contributory role cannot be entirely excluded. Gradual refinement of these limitations will facilitate a more accurate assessment of the centrality of this pathway in doxorubicin-induced cardiotoxicity.

Furthermore, from the perspectives of clinical translation and precision intervention, the following additional limitations should be noted. First, this study did not investigate the role of the BMSC-EVs–miR-210-3p–ACVR1B pathway in the tumor context. Given that doxorubicin is a commonly used chemotherapeutic agent, balancing its antitumor efficacy with myocardial protection is a critical prerequisite for future clinical application. Subsequent studies should employ tumor cell-cardiomyocyte co-culture systems and tumor-bearing animal models to verify whether this pathway interferes with the antitumor activity of doxorubicin, thereby ensuring its clinical safety. Second, this study focused on miR-210-3p within EVs, while the roles of other functional EV components (e.g., proteins, lipids, and other nucleic acids) in DIC were largely overlooked. As the overall function of EVs relies on the synergistic effects of multiple cargo molecules, future studies could utilize multi-omics integrative analyses to systematically delineate the key molecular networks within EVs that regulate DIC, thereby providing a more comprehensive experimental basis for multi-target optimization of BMSC-EV therapy. Finally, compared with existing studies on targeted BMSC-EV therapy, EV targeting modification was not performed in this work. In future studies, our group will adopt the strategy described by Mao et al. [34] to engineer EVs with a myocardial homing peptide (HHP) at the stem cell level for targeted treatment of doxorubicin-induced cardiotoxicity, which is expected to further enhance the precision and efficacy of the intervention.

Conclusion

This study firstly confirmed that BMSC-EVs-derived miR-210-3p mediates DIC protection by directly targeting ACVR1B 3’UTR, inhibiting ACVR1B-induced SMAD3 phosphorylation, relieving PGC-1α inhibition, improving mitochondrial dysfunction, and alleviating DOX-induced cardiotoxicity. It reveals a novel BMSC-EVs therapy mechanism for DIC, identifies miR-210-3p/ACVR1B as potential targets, and provides experimental basis for clinical translation.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Abbreviations

DIC

Doxorubicin-induced cardiotoxicity

DOX

Doxorubicin

BMSC

Bone marrow mesenchymal stem cell

EVs

Extracellular vesicles

ACVR1B

Activin a receptor, type 1b

PGC-1α

Peroxisome Proliferator-Activated Receptor Gamma Coactivator 1-alpha

SMAD3

Mothers against decapentaplegic homolog 3‌

FISH

Fluorescence in situ hybridization

EDTA

Ethylenediaminetetraacetic acid

AAV9

Adeno-associated virus serotype 9

ETC

Electron transport chain

SDHB

Succinate dehydrogenase complex subunit b

NDUFS1

NADH dehydrogenase (ubiquinone) Fe-S protein 1‌

ATP5A1

ATP synthase, H+ transporting, mitochondrial F1 complex, alpha subunit 1

COX IV

Cytochrome c oxidase subunit 4

ΔΨm

Mitochondrial membrane potential

DCFDA

2′,7′-dichlorodihydrofluorescein diacetate

Author contributions

X.M.C. designed and supervised the experiments. R.L.C. performed most of the cellular experiments. L.Q. performed the animal experiments and EVs uptake and miR-210-3p fish colocalization staining. R.L.C. wrote the manuscript. B.L. and X-M.C. provided the funding. R.L.C., L.Q., N.Z., X.W., Z.L., X.X., H.Z., C.Y., Y.Z., B.L. collected the samples.

Funding

This work was supported by the National Natural Science Foundation of China (no. 82172574 and 81871231), the Natural Science Foundation of Shandong Province (no. ZR2024MH083, ZR2025MS1420 and ZR2023MH082), the Shandong Taishan Scholars Young Experts Program (no. TSQN202103056), and the Qingdao Natural Science Foundation Key Project (no. 24-8-4-ZRJJ-8-JCH). 2024 Health and medical leading innovation team program of Shandong province (Xianming Chu).

Data availability

Data supporting the findings of this study are present in the paper and/or the Supplementary Materials. Additional data related to this paper can be requested from the corresponding authors.

Declarations

Ethics approval and consent to participate

“This study, entitled ‘BMSC-derived EVs alleviate doxorubicin-induced cardiotoxicity through miR-210-3p by targeting the ACVR1B–p-SMAD3–PGC-1α pathway,’ is conducted as a subproject of the National Natural Science Foundation of China (NSFC, No. 82172574), which focuses on the mechanism by which bone marrow mesenchymal stem cell-derived exosomal piRNA (DICAPIR) mitigates anthracycline-induced cardiotoxicity through the regulation of ferroptosis.” All the animal procedures were performed following the Declaration of Helsinki and were approved by the Laboratory Animal Management and Ethics Committee of Qingdao University (Approval No. 20240822C573220240907219). Date of approval was August 22, 2024.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.