Ischemic diseases, encompassing conditions such as peripheral artery disease and coronary artery disease, represent a leading cause of global morbidity and mortality[1]. These pathologies are defined by a critical imbalance between oxygen supply and metabolic demand within affected tissues[2, 3]. The therapeutic induction of robust and functional angiogenesis remains a paramount, yet persistently elusive, goal in vascular medicine. The ischemic microenvironment is notoriously pathological, characterized by profound hypoxia, elevated levels of reactive oxygen species (ROS), nutrient deprivation and sustained inflammation[4]. These factors collectively impose severe bioenergetic stress on resident vascular endothelial cells, compromising their capacity for repair and regeneration.

Central to cellular survival and function, mitochondria are indispensable for energy production, redox homeostasis and critical signaling processes that govern angiogenesis, including proliferation, migration and tube formation [5–7]. Compelling evidence now indicates that mitochondrial dysfunction is a fundamental hallmark of ischemic tissues, directly contributing to cellular apoptosis and the failure of re-vascularization[5–7]. Consequently, innovative strategies designed to directly ameliorate mitochondrial bioenergetics within ischemic cells present a compelling and targeted therapy.

Adipose-derived stem cells (ADSCs) have emerged as a potent source for regenerative therapy, prized for their abundance, accessibility and pro-angiogenic secretome[8–10]. However, a groundbreaking paradigm shift has occurred with the recognition of horizontal mitochondrial transfer as a fundamental mechanism of ADSC-mediated cytoprotection and repair[11]. Intercellular mitochondrial trafficking can occur via several routes, including the formation of tunneling nanotubes (TNTs), gap junctions and cell fusion. While physiologically relevant, these direct cell-contact mechanisms are inherently limited in a therapeutic context by their low efficiency and spatial constraints[11, 12]. In contrast, the transfer of mitochondria via extracellular vesicles (EVs) represents a sophisticated, distance-independent delivery system[13]. These mitochondrial-enriched extracellular vesicles (mitoEVs) are naturally packaged by donor cells and can be efficiently internalized by recipient cells, conferring a full complement of functional proteins, lipids and mitochondrial DNA (mtDNA) to resuscitate bioenergetics[13]. This vesicle-mediated mechanism offers distinct advantages for therapy, including enhanced biological stability in circulation, a reduced risk of immunogenicity compared to whole cells, and the potential for targeted delivery to sites of injury, thereby bypassing the limitations of direct cell-cell contact.

In this study, we isolated and characterized mitoEVs from ADSC-conditioned media through Nano Flow Cytometry and analyzed their protein components. We then investigated their efficacy in promoting angiogenesis in a murine model of critical hindlimb ischemia. Utilizing integrated proteomic profiling and real-time analysis of cellular metabolism via Seahorse extracellular flux technology, we further elucidate the core mechanism through which ADSC-mitoEVs reprogram cellular energy metabolism and activate potent angiogenic pathways in ischemic VECs.

Both mitoEVs and mito-free EVs from ADSCs exhibited characteristic cup-shaped morphology under transmission electron microscopy (Fig. 1B). MitoEVs shown increased mean particle size of 218.6 ± 32.5 nm compared with that of mito-free EVs (105.3 ± 24.1 nm) as determined by nanoparticle tracking analysis (Fig. 1C). Western blot analysis, as shown in Fig. 1D, confirmed positive expression of EV markers (CD63 and TSG101) in both mitoEVs and mito-free EVs, while positive expression Mitochondrial outer membrane protein (TOM20) only observed in mitoEVs.

We simulated in vivo ischemia by constructing a cell culture microenvironment with hypoxia and serum deprivation. Colocalization analysis indicated that internalized mitoEVs preferentially bind the endogenous mitochondrial network within 12 hours of co-culture, as visualized via confocal microscopy using miRFP670-labeled mitoEVs and DsRed-labeled mitochondria in HUVECs. Increased uptake rate and mitochondrial binding efficiency of mitoEVs was observed in HUVECs cultured under ischemic condition, compared to Normal condition (Fig. 1E-F).

Apoptosis analysis by flow cytometry indicated that ADSC-mitoEVs significantly increased the viability of ischemic HUVECs (under hypoxia and serum deprivation) in a dose-dependent manner up to 40 µg/ml (Fig. 2A-C). Similarly, scratch wound healing and tube formation assays demonstrated markedly enhanced migration and angiogenic capacity in mitoEV-treated groups compared to controls under ischemic conditions, with efficacy plateauing beyond 40 µg/ml (Fig. 2D-G). Doses exceeding 40 µg/ml provided minimal additional benefit to either cell survival or angiogenesis. Therefore, 40 µg/ml was selected as the optimal concentration of ADSC-mitoEVs for all subsequent experiments.

ADSC-mitoEVs significantly improved the angiogenic potential of HUVECs in vivo. Subcutaneous grafts co-implanted with HUVECs and ADSC-mitoEVs exhibited extensive networks of perfused microvessels by day 7, in contrast to control grafts, which showed poor vascularization (Fig. 3A-H). Furthermore, ADSC-mitoEVs administration promoted functional recovery in ischemic hindlimbs, as evidenced by markedly improved perfusion over a three-week period and prevention of foot necrosis (Fig. 3I-N). Histological analysis of muscle tissues retrieved from the ADSC-mitoEV-treated group revealed increased capillary density (CD31⁺ structures) and higher skeletal muscle density (α-SMA⁺) (Fig. 3O-Q). These findings demonstrate that ADSC-mitoEVs not only stimulate robust vascular network formation but also enhance tissue perfusion and protect against ischemic injury, highlighting their therapeutic potential in promoting vascular regeneration and functional recovery.

Using JC-1 staining, we observed that ADSC-mitoEVs markedly enhanced the mitochondrial membrane potential (MMP) and increased total ATP production in ischemic HUVECs (Fig. 5A-C). Furthermore, ADSC-mitoEV treatment significantly reduced mitochondrial ROS levels (Fig. 5D-E). Mitochondrial stress assays demonstrated that ADSC-mitoEVs improved basal oxygen consumption rate (OCR), mitochondrial ATP production, spare respiratory capacity, and non-mitochondrial ATP generation (Fig. 5F-G). These results indicate that ADSC-mitoEVs restore mitochondrial bioenergetics, enhance redox balance, and boost metabolic flexibility in hypoxic endothelial cells, underscoring their therapeutic potential in ameliorating ischemia-induced mitochondrial dysfunction.

To investigate the impact of ADSC-mitoEVs on mitochondrial quality control under ischemic conditions, we first employed mt-Keima lentiviral transduction to monitor mitophagy flux. The results indicated that ADSC-mitoEVs significantly enhanced mitophagy, as evidenced by an increased ratio of mitochondria within lysosomes to those in the neutral cytosol (Fig. 6A-B). Moreover, ADSC-mitoEV treatment improved mitochondrial morphology under ischemic stress, promoting increased mitochondrial network length and size (Fig. 6C-D).

We next evaluated the persistence and integration of transferred mitochondria. Using droplet digital PCR (ddPCR) with primers specific to mtDNA and ADSCs and HUVECs (Supplementary Data 1), we detected approximately 32,000 copies of ADSC-derived mtDNA in HUVECs after 12 hours of co-culture with ADSC-mitoEVs. This copy number declined to 22,400 at 24 hours and 12,800 at 48 hours, becoming undetectable by 72 hours (Supplementary Fig. 2 and Fig. 6E).

To elucidate the mechanisms underlying enhanced mitophagy, we analyzed markers of autophagic flux and mitophagy pathways (Fig. 6F). ADSC-mitoEVs treatment increased the LC3-II/I ratio and decreased p62 expression, confirming enhanced autophagic activity. Western blot analysis further revealed elevated levels of PINK1, phospho-ubiquitin, and phospho-Parkin, indicating activation of the PINK1/Parkin-mediated mitophagy pathway. In contrast, no significant activation was observed in the BNIP3/NIX pathway.

In this study, we demonstrate that mitochondrial-enriched extracellular vesicles derived from adipose-derived stem cells serve as an efficient bioenergetic cargo system capable of rescuing ischemic endothelial cells through enhanced mitophagy and mitochondrial protein transfer. These vesicles promote angiogenesis, improve mitochondrial function, and facilitate functional recovery in models of hindlimb ischemia, thereby presenting a novel cell-free strategy for ischemic vascular diseases.

Our experimental strategy was designed to systematically elucidate the functional and mechanistic contributions of ADSC-mitoEVs. Beginning with proteomic characterization, we identified a significant enrichment of mitochondrial proteins within ADSC-mitoEVs, which contribute to oxidative phosphorylation, redox homeostasis, and anti-apoptotic processes. This initial profiling suggested that mitoEVs might directly supplement dysfunctional mitochondria in recipient cells. However, internalized ADSC-mitoEVs did not integrate into the mitochondrial network of recipient endothelial cells. Instead, they rapidly activated the PINK1/Parkin-mediated mitophagy pathway within 12 hours and were completely degraded within 72 hours. This activation promotes mitochondrial turnover and quality control, served as a prerequisite for initiating angiogenesis in ischemic endothelial cells.

Concurrently, mitophagy induction enhanced reactive oxygen species clearance, alleviating mitochondrial stress[14, 15]. These effects align with mechanisms reported for both free mitochondrial transfer from ADSCs and TNT-mediated mitochondrial delivery[11]. Although the internalization routes differ (endocytosis for EVs versus direct intercellular transfer for TNTs) both pathways converge on a shared functional outcome: activation of mitophagy and subsequent cytoprotection. We propose that this convergence is mediated by the delivery of mitochondrial components, such as some kinds of mitochondrial proteins and mtDNA, which are recognized as “non-self” motifs by the recipient cell’s surveillance system. These motifs trigger PINK1 stabilization on the outer mitochondrial membrane, Parkin recruitment, ubiquitination of mitochondrial substrates, and ultimately autophagosome formation around the exogenous material[16–18]. The PINK1/Parkin pathway, central to mitochondrial quality control via mitophagy, plays a critical and predominantly restrictive role in angiogenesis by clearing dysfunctional mitochondria in endothelial cells. This pathway maintains cellular fitness by limiting reactive oxygen species, preventing apoptosis, and supporting the necessary metabolic reprogramming for cell migration and proliferation[19–22]. Moderate activation of mitophagy is beneficial, as it facilitates the removal of dysfunctional organelles and stimulates mitochondrial biogenesis, leading to increased ATP production[19–22]. Indeed, following mitoEV treatment, we observed increased total mitochondrial content, well-developed mitochondrial network and enhanced ATP generation in endothelial cells, indicating improved bioenergetic capacity as well as enhanced cellular fitness under stress conditions.

Notably, unlike free mitochondrial transfer as reported previously[11], ADSC-mitoEVs retained partial functionality, such as antioxidant capacity and improvements in energy metabolism, even in PINK1-silenced endothelial cells. We attribute this PINK1-independent effect to the efficient vesicle-mediated delivery of functional mitochondrial enzymes, particularly superoxide dismutase 2 (SOD2). Owing to the protective bilayer phospholipid membrane of extracellular vesicles[23, 24], the structural and functional integrity of SOD2 and other oxidative phosphorylation-related proteins is preserved during intercellular transit, enabling them to remain catalytically active upon delivery. Mechanistically, once internalized, SOD2-enriched ADSC-mitoEVs preferentially localize to the mitochondrial compartments of recipient endothelial cells, where the encapsulated SOD2 is released and integrates into the endogenous mitochondrial matrix. SOD2 anions dismutates superoxide anions into hydrogen peroxide and oxygen, thereby markedly reducing mitochondrial reactive oxygen species levels and protecting electron transport chain complexes from oxidative damage, stabilizing mitochondrial membrane potential and enhancing ATP synthesis independently of PINK1/Parkin signaling[25–28]. Proteomic analyses further suggest that ADSC-mitoEVs carry additional mitochondrial enzymes, such as ubiquinone oxidoreductase subunits and ATP synthase, which may act synergistically with SOD2 to ameliorate bioenergetic deficits and enhance mitochondrial resilience under ischemic stress. This multi-protein delivery mechanism distinguishes ADSC-mitoEVs from conventional mitochondrial transplantation and underscores their potential as a comprehensive mitochondrial therapy for vascular diseases.

Mitochondrial transplantation has emerged as a promising treatment for conditions such as cancer[29], Parkinson’s disease[30], and myocardial infarction[31]. ADSC-mitoEVs offer a cell-free strategy that combines the advantages of mitochondrial transplantation and conventional EV therapies while overcoming following limitations. Unlike direct mitochondrial transplantation, which suffers from poor stability and short retention times, ADSC-mitoEVs provide a membrane-protected biological vector that enhances protein stability and delivery efficiency of multiple mitochondrial-derived enzymes. Moreover, compared with heterogeneous conventional MSC-EV preparations, ADSC-mitoEVs represent a defined subpopulation with elevated mitochondrial content, enabling more targeted bioenergetic modulation.

Despite these promising findings, our study has several limitations. First, advanced purification strategies, such as immunoaffinity-based sorting for mitochondrial markers, could yield more homogeneous mitoEV populations if the membrane protein profile of mitoEVs can be characterized. Additionally, while diabetic murine models of hindlimb ischemia offer useful insights, they hardly recapitulate the complex pathophysiology of human vascular diseases, particularly in aging or comorbid settings. Further validation in large animal models is essential before clinical translation. Finally, the long-term safety, immunogenicity, and fate of administered mitoEVs warrant systematic investigation.

In this study, we demonstrate that ADSC-mitoEVs, a novel cell-free therapeutic platform, ameliorates ischemic endothelial damage through dual mechanisms: activation of PINK1/Parkin-mediated mitophagy and direct transfer of functional mitochondrial proteins. This strategy enhances mitochondrial quality control, redox homeostasis and energy metabolism reprogramming, offering a promising approach for ischemic vascular diseases treatment.

To evaluate the persistence of transferred mitochondria in recipient HUVECs over time, we employed a mitochondrial DNA (mtDNA) genotyping strategy based on the detection of heteroplasmy. mtDNA was isolated from ADSCs and HUVECs using a Mitochondrial DNA Isolation Kit (MCE, NJ, USA). Regions with sequence divergence between the two cell types were identified by Sanger sequencing. Specific primers were designed to target these differential sites, and their amplification efficiency was validated by Droplet Digital PCR (ddPCR) on a QuantStudio Absolute Q system (Thermo Fisher Scientific, MA, USA). Using these primers, we quantified the copy number of ADSC-derived mtDNA in HUVECs co-cultured with ADSC-mitoEVs at 12, 24, 48, and 72 h post-treatment via ddPCR. Three independent experiments were performed.