Mitochondrial Transfer: From Bench to Bedside

Abstract

Intercellular mitochondrial transfer has emerged as a fundamental mechanism of tissue adaptation and repair in the cardiovascular system, with major implications for cardiovascular, neurologic, metabolic, and inflammatory diseases. Once thought to be static, mitochondria are now recognized as mobile organelles that move between cells via tunneling nanotubes (TNTs), extracellular vesicles (EVs), and free mitochondria. These pathways support two complementary axes of mitochondrial communication: “Rescue by Replenish,” in which healthy mitochondria or mitochondrial components restore bioenergetics and stress resistance in recipient cells, and “Relief by Release,” in which damaged mitochondria are exported for degradation to preserve homeostasis and limit inflammation. We summarize the molecular machinery governing TNT formation, mitochondria-derived vesicle (MDV) biogenesis, EV sorting, and free mitochondrial release and uptake, and discuss how these processes shape organ function. Building on these mechanistic insights, we outline four translational strategies: (i) cell-based therapies that donate healthy mitochondria or scavenge damaged ones; cell-free approaches using (ii) mitochondria-containing EVs or (iii) purified mitochondria; (iv) pharmacologic, nutritional, and lifestyle interventions that augment endogenous mitochondrial turnover and intercellular exchange. Finally, we discuss key barriers to clinical translation, including inflammatory and oncogenic risks, mitonuclear incompatibility, incomplete understanding of the fate and durability of transferred mitochondria, and the lack of standardized manufacturing, potency assays, and long-term storage methods. Continued integration of mechanistic biology with bioengineering and regulatory science will be essential to safely move mitochondrial transfer–based therapies from bench to bedside in cardiovascular medicine.

Introduction

Cardiovascular disease is increasingly recognized as a disorder of impaired cellular energetics rather than solely hemodynamic or neurohormonal dysregulation.^1–3 Despite substantial advances in guideline-directed medical therapy (GDMT), which have improved survival primarily by reducing myocardial workload and neurohormonal stress, the global burden of heart failure continues to rise, reaching epidemic proportions worldwide.⁴ This paradox underscores a fundamental limitation of current therapies: while GDMT alleviates energetic demand, it does not directly restore the compromised bioenergetic capacity of diseased cardiomyocytes. As a result, progressive metabolic insufficiency persists at the cellular level, driving ongoing myocardial dysfunction, adverse remodeling, and clinical deterioration.^5,6 This unmet therapeutic gap has renewed interest in strategies that target mitochondrial health, positioning mitochondrial function not as a passive downstream consequence of injury, but as a central and therapeutically actionable determinant of cardiovascular repair and resilience.

Mitochondria are central regulators of cellular homeostasis, integrating oxidative phosphorylation, calcium buffering, redox signaling, and apoptotic control, while also governing fundamental biological processes such as stem cell maintenance, differentiation, and cell-cycle progression. Consequently, mitochondrial dysfunction or inappropriate mitochondrial quality control mechanisms underlie a wide spectrum of pathologies, most prominently cardiovascular disease, including ischemia-reperfusion injury and heart failure, as well as pulmonary injury, neurodegeneration, metabolic disorders, cancer progression and metastasis, and chronic inflammation.^7–11 Recent discoveries have revealed that mitochondria are not static organelles confined within individual cells but rather dynamic entities capable of intercellular exchange under both physiological and pathological conditions. This bidirectional transfer of mitochondria, including both functional and damaged organelles, enables the modulation of bioenergetic capacity, redox balance, and cell fate in participating cells, including cardiomyocytes, endothelial cells (ECs), and immune cells. A seminal study first revealed that mitochondria can be horizontally transferred between mammalian cells.¹² Since that discovery, several distinct conduits of mitochondrial exchange have been elucidated, most prominently tunneling nanotubes (TNTs), extracellular vesicles (EVs), and freely released mitochondria, which together constitute a multifaceted network of intercellular mitochondrial communication.

Transferred healthy mitochondria augment the bioenergetic capacity of recipient cells, reinforce mitochondrial quality control and proteostasis, and fine-tune innate immune and inflammatory signaling. Conversely, stressed cells can actively extrude dysfunctional mitochondria via a mitochondrial clearance pathway that is functionally complementary to mitophagy. This process not only attenuates activation of proinflammatory cascades but also serves as a cytoprotective mechanism to preserve homeostasis in donor cells.^13,14 Together, these bidirectional processes define two mechanistic axes of mitochondrial communication: metabolic replenishment (“Rescue by Replenish”) and cytoprotective detoxification (“Relief by Release”) (Figure 1).

Figure 1. Maintenance of cardiovascular homeostasis through two distinct mechanisms of mitochondrial transfer: “Rescue by Replenish” and “Relief by Release.”.

In the “Rescue by Replenish” mechanism, functional mitochondria and mitochondrial-derived vesicles (MDVs) are generated within donor cells. Tunneling nanotubes (TNTs) are dynamic, actin-based membranous bridges formed through filopodia-like protrusions that facilitate direct intercellular exchange of organelles, including both intact mitochondria and MDVs. MDVs can also be incorporated into multivesicular bodies (MVBs) and released into the extracellular space, while both MDVs and intact mitochondria may be secreted via extracellular vesicles (EVs) or ectosomes formed through plasma membrane budding. The molecular mechanisms underlying the release of free mitochondria remains incompletely understood. Once internalized by recipient cells, transferred mitochondria either fuse with the native mitochondrial network or undergo lysosomal degradation, resulting in restoration of mitochondrial membrane potential, increased ATP production, and reduced apoptosis in cardiovascular system. Conversely, in the “Relief by Release” mechanism, damaged mitochondria generate MDVs and fragmented mitochondria that are either released as free mitochondria or encapsulated within EVs. Extracellular mitochondria are subsequently engulfed and degraded by recipient cells, thereby maintaining mitochondrial quality control in donor cells and mitigating excessive tissue inflammation in cardiovascular system. Illustration credit: Sceyence Studios

Despite promising preclinical results, key challenges remain for clinical translation. In particular, the mechanisms underlying the release of functional mitochondria into the extracellular space, their intracellular processes after uptake, and the duration of mitochondrial survival after transfer remain poorly understood. Furthermore, therapeutic mitochondrial transfer carries potential risks such as immune activation, tumorigenic transformation, and mtDNA incompatibility, raising major safety and reproducibility concerns. From a translational standpoint, the absence of standardized manufacturing protocols, scalable purification processes, validated potency assays, and stable long-term storage further limits clinical feasibility. Taken together, these challenges highlight the need for mechanistic elucidation, bioengineering innovation, and regulatory advancement to fully realize the therapeutic potential of mitochondrial transfer therapy.

In this review, we summarize recent advances in understanding of the molecular mechanisms, biological functions, and therapeutic implications of intercellular mitochondrial transfer. We highlight how mitochondrial exchange contributes to cellular bioenergetic recovery and tissue homeostasis through both replenishment and clearance pathways. In addition, we discuss emerging therapeutic strategies based on mitochondrial transfer, along with the current technical, biological, and translational challenges that must be addressed to achieve clinical translation in cardiovascular disease.

Intercellular Mitochondria Transfer

The first experimental evidence of intercellular mitochondrial transfer was provided by Spees et al. (2006), who demonstrated that direct mitochondrial transfer restored oxidative metabolism and exponential proliferation capacity in mitochondrial DNA (mtDNA)-depleted (ρπ) cells.¹² In this work, human bone marrow-derived mesenchymal stem cells (MSCs) were shown to donate functional mitochondria to adjacent A549 ρπ lung carcinoma cells, thereby rescuing their defect in oxidative phosphorylation after four days of co-culture. Fluorescence microscopy visualized the intercellular trafficking of mitochondria through cytoplasmic extensions and vesicular structures. In contrast, no mtDNA transmission was observed when mitochondria isolated from human MSC homogenates or from human platelets were used as donors, suggesting that mitochondrial transfer requires viable donor cells and may depend on cell type-specific, contact-dependent mechanisms. Later studies, however, revealed that mitochondrial transfer can occur under both contact and non-contact culture systems, suggesting that direct physical interaction is not strictly required for organelle exchange. The transfer process can be partially suppressed by pharmacological agents targeting endocytosis, gap junctions, or cytoskeletal/TNT formation pathways.^15–19 Based on these inhibition assays, together with ultrastructural and live-cell imaging analyses, multiple structural pathways have been proposed to mediate intercellular mitochondrial transfer, including TNTs, EVs, free mitochondria,^20–22 connexin 43 (Cx43)-mediated gap junctions,^23,24 and cell fusion.²⁵ This review focuses on TNT- and EV-mediated mitochondrial transfer, as well as the release and uptake of free mitochondria, which represent the most extensively characterized mechanisms to date, particularly in the context of cardiovascular biology and disease. Other potential routes, including gap junction- or cell fusion-mediated transfer, have been comprehensively reviewed elsewhere.^26,27

Cardiovascular constraints and molecular regulation of nanotube-mediated mitochondrial transfer

The most widely reported mechanism of intercellular mitochondrial transfer involves the formation of TNTs. TNTs were first described in 2004 by Rustom et al., who demonstrated that mammalian cells establish direct intercellular connections through nanotubular structures, enabling bidirectional transfer of cytoplasmic components, including organelles.²⁸ TNTs are generally classified into thin and thick subtypes based on diameter; thin TNTs contain only F-actin filaments, whereas thick TNTs (> 0.7 μm) contain both F-actin and microtubules.²⁹ Mitochondrial transport occurs predominantly through thick TNTs and may be unidirectional or bidirectional depending on cellular and environmental context.^29,30 Importantly, bidirectional mitochondrial exchange via TNTs between cardiomyocytes and fibroblasts has been demonstrated under physiological conditions.³¹ Under hypoxic stress, this exchange becomes unidirectional, with mitochondria transferred from fibroblasts to cardiomyocytes, suggesting a stress-adaptive mechanism that supports cardiomyocyte bioenergetics.³² These observations establish TNTs as a physiologically relevant and context-sensitive conduit for mitochondrial exchange in the heart. In cardiovascular tissues, particularly in the adult heart, nanotube-mediated mitochondrial transfer occurs within a uniquely constrained mechanical and metabolic environment characterized by continuous contraction, high energy demand, and tightly packed cellular architecture. These features distinguish the myocardium from many other tissues and are likely to influence both the feasibility and functional relevance of TNTs as conduits for intercellular mitochondrial exchange. However, direct evidence for TNT formation between fully differentiated cardiomyocytes in the normal adult myocardium is limited, as the majority of observations to date have been reported either during cardiac development,^33,34 or, as detailed below, under pathological or stress-associated conditions.

TNT formation is a tightly coordinated process involving cytoskeletal remodeling, membrane dynamics, and stress-responsive signaling. Biogenesis begins with filopodia-like protrusions that extend toward neighboring cells and detach from the substrate to form suspended nanotubular bridges.³⁵ The actin cytoskeleton provides the primary scaffold, with Rho-family GTPases (Cdc42, Rac1, RhoA) orchestrating early protrusion formation.^28,36 Membrane remodeling is further regulated by the p53–caspase-3–S100A4 axis, which governs TNT directionality through S100A4 concentration gradients between connected cells.³⁷ The M-Sec–Ral–exocyst complex mediates the emergence of long plasma-membrane extensions,³⁸ while Eps8 stabilizes actin bundles and myosin Va/X mediate vesicular and organelle transport along TNTs.^39,40 Cell-type-specific proteins, including connexin43 (Cx43) and GAP43 further stabilize TNTs and facilitate electrical or metabolic coupling between connected cells.^41–44

Under conditions relevant to cardiovascular disease, inflammatory and ischemic signaling pathways enhance TNT formation and mitochondrial transfer. NF-κB/TNFαIP2 signaling, Miro1-mediated mitochondrial trafficking, and KIF5B-driven microtubule transport enhance TNT-dependent mitochondrial exchange during inflammation.^32,45–47 Hypoxia activates HIF-1α–dependent pathways that further facilitate TNT-mediated communication,⁴⁸ while phosphatidylserine exposure on injured cells recruits stem cells via phosphatidylserine receptors to initiate targeted TNT formation and mitochondrial exchange.⁴⁹ Collectively, these pathways provide a mechanistic framework by which inflammatory and ischemic stress may permissively enable TNT-mediated mitochondrial transfer in the myocardium.

Functional Roles of Nanotube-Mediated Mitochondrial Transfer in Cardiovascular Repair

One of the earliest reports of TNT-mediated mitochondrial transfer in the cardiovascular system appeared in 2005, when nanotube-like connections between endothelial progenitor cells and cardiomyocytes were shown to enable mitochondrial transfer in approximately 2.6% of co-cultured cells after 24 hours.⁵⁰ Building on this initial observation, subsequent in vitro studies demonstrated that MSCs transfer functional mitochondria to diverse cardiovascular cell types via TNTs, thereby promoting cardiac repair and vascular homeostasis. Bidirectional mitochondrial transfer between MSCs and hypoxia-injured cardiomyocytes (H9C2) has been observed via TNTs.⁵¹ In models of anthracycline-induced cardiotoxicity, mitochondrial exchange between MSCs and injured cardiomyocytes attenuates apoptosis and restores cardiomyocyte bioenergetic function both in vitro and in vivo.⁴⁶ Several studies further report predominantly unidirectional mitochondrial delivery from MSCs to cardiomyocytes under stress conditions, serving as a “Rescue by Replenish” mechanism. Under hypoxia, MSC-derived mitochondria restored mitochondrial membrane potential and prevented cardiomyocyte apoptosis.^52,53 Even under physiological conditions, direct co-culture induces fine TNTs that support unidirectional mitochondrial transfer.⁵⁴ Beyond cardiomyocytes, TNT-mediated mitochondrial transfer from MSCs to ECs under physiological, hypoxic, or chemotherapeutic stress conditions restores endothelial respiration, enhances motility and angiogenic capacity, and suppresses apoptosis in vitro and in vivo.^19,30,49 Conversely, MSCs may also function as mitochondrial recipients; co-culture with vascular smooth muscle cells promotes MSC proliferation through TNT-dependent mitochondrial exchange.⁵⁵ Collectively, these findings indicate that TNT-mediated mitochondrial transfer enables dynamic, bidirectional metabolic coupling among cardiovascular cell populations and contributes to myocardial repair and tissue homeostasis under pathological conditions.

Cardiovascular Context and Molecular Regulation of Extracellular Vesicle-Mediated Mitochondrial Transfer

In cardiovascular tissues, mitochondria are subject to high metabolic demand, oxidative stress, and inflammatory pressure, necessitating highly adaptable mitochondrial quality-control systems. In this context, EV-mediated mitochondrial transfer has emerged as a critical mechanism that integrates mitochondrial turnover, intercellular communication, and tissue homeostasis. Vesicle-mediated mitochondrial release is an evolutionarily conserved process. In protists such as Tetrahymena, elevation in intracellular Ca²+ trigger the extrusion of mitochondrial fragments into the extracellular milieu, accompanied by relocalization of Hsp60 to the plasma membrane, thereby facilitating stress adaptation.⁵⁶ In mammalian cells, this ancestral mechanism has evolved into a regulated EV-mediated pathway in which intact mitochondria or mitochondria-derived vesicles (MDVs) are selectively packaged into EVs and released. Once secreted, these mitochondria-containing EVs (EV-mito) serve dual functions that are highly relevant to cardiovascular biology: they enable metabolic support of recipient cells (“Rescue by Replenish”) while simultaneously facilitating the disposal of damaged mitochondria from donor cells (“Relief by Release”). In contrast to TNT-mediated transfer, EV-mediated trafficking of mitochondrial components has been documented in vivo in differentiated cardiomyocytes under physiological conditions,^14,57 supporting the notion that EVs mediate a basal and continuous form of mitochondrial communication in the adult myocardium.

EV-associated mitochondrial cargo can be broadly classified into intact mitochondria and MDVs. MDVs are small vesicles (≈60–150 nm) that bud directly from mitochondria. Electron microscopy has identified both single-membraned MDVs, derived from the outer mitochondrial membrane, and double-membraned MDVs, containing both inner and outer membrane elements.^58,59 Importantly, MDV biogenesis is mechanistically distinct from canonical mitochondrial fission and mitophagy. Although DRP1 and its adaptors (MFF, MID49, MID51) mediate the final scission step,⁵⁹ MDVs selectively incorporate specific mitochondrial proteins while excluding others, reflecting an active, regulated cargo-sorting mechanism.^60–62 Notably, MDVs biogenesis occurs independently of the autophagy machinery and can proceed even in mitochondria maintaining a normal membrane potential,^61–63 which indicates that this process represents a dedicated and regulated quality-control pathway rather than a terminal degradation event.

MDV formation is governed by multiple conserved molecular pathways (Figure 2). The retromer complex, particularly VPS35, plays a central role in selective cargo sorting and trafficking of mitochondrial proteins.^64,65 VPS35 interacts with the outer mitochondrial membrane protein MAPL/MUL1 and is required for its transport from mitochondria to peroxisomes.^64,66 Disruption of VPS35 function alters mitochondrial morphology and impairs protein recycling, underscoring its importance in mitochondrial homeostasis.^67,68 MDV biogenesis has been most extensively characterized in PINK1/Parkin-dependent pathways, where mild oxidative stress selectively removes oxidized mitochondrial components prior to activation of mitophagy.^61,69,70 Following mild mitochondrial stress, such as that induced by the complex III inhibitor antimycin A, MDV formation is detected within 2–6 hours, whereas mitophagy becomes evident only after 12–24 hours.⁶⁹ This temporal distinction indicates that MDV biogenesis represents an early, selective quality-control process that precedes mitophagy and maintains mitochondrial function by eliminating localized damage, while mitophagy serves as a terminal degradation pathway under severe or widespread dysfunction. Particularly in high-energy-demanding cells such as cardiomyocytes, MDV formation likely constitutes a preferred adaptive response for maintaining mitochondrial homeostasis.⁷⁰

Figure 2. Mitochondria-derived veisicle formation and divergent fates.

Mitochondria-derived vesicles (MDVs) are generated through multiple pathways to maintain mitochondrial and cellular homeostasis. Under mild oxidative or metabolic stress, localized mitochondrial damage triggers PINK1/Parkin-dependent MDV formation, leading to the selective sequestration of oxidized or depolarized mitochondrial components. These MDVs are trafficked to late endosomes and lysosomes for degradation. In parallel, under basal or adaptive conditions, OPA1- and SNX9-dependent MDVs containing functional mitochondrial components are incorporated into multivesicular bodies (MVBs) and released extracellularly for the Rescue by Replenish. Together, these distinct routes represent a dual mitochondrial quality-control system: lysosomal degradation of damaged cargo and extracellular release of functional mitochondria for intercellular communication. MDVs derived from damaged mitochondria are preferentially targeted to lysosomes for degradation. Lysosomal inhibition enhances the secretion of EV-mito, which originate from MVBs and are released independently of autophagy. Deletion of Rab7, a key endolysosomal GTPase, markedly increases the secretion of ubiquitin-rich EVs containing mitochondrial cargo in cultured cells and mouse hearts. Illustration credit: Sceyence Studios

Beyond this canonical route, PINK1/Parkin-independent pathways are activated under inflammatory or metabolic stress conditions commonly encountered in cardiovascular disease. In macrophages, LPS stimulation induces MDVs containing 2-oxoglutarate dehydrogenase via SNX9 and RAB9.^71–73 SNX9 facilitates vesicle budding, while RAB9 mediates late endosomal trafficking. Moreover, microtubule-dependent pulling of mitochondrial tubules, mediated by MIRO1/2, CENP-F, and ARMCX proteins, drives membrane protrusions and MDV scission.^58,74–76 Finally, the fusion GTPase OPA1 contributes to the incorporation of inner-membrane and matrix cargo into double-membraned MDVs.⁷⁷ While direct in vivo evidence for MDV biogenesis in the cardiovascular system is still emerging, multiple lines of indirect evidence from inflammatory and heart failure models support the notion that MDV formation is upregulated under cardiac stress conditions.

Divergent Fates of Mitochondria-Derived Vesicles in Cardiovascular tissues: Degradation or Extracellular Secretion

In cardiomyocytes, MDVs are central components of mitochondrial quality control, operating at the intersection of degradation and intercellular communication. MDVs are typically trafficked to degradative compartments such as late endosomes and lysosomes, where their contents are degraded or recycled (Figure 2). This transport is mediated by SNARE-dependent fusion machinery composed of STX17, SNAP29 and VAMP7. STX17, localized to the outer membrane of MDVs, interacts with SNAP29 in the cytosol and VAMP7 on the target endolysosomal membrane, mediating the vesicle docking and membrane fusion.^78,79 In parallel, the small GTPase Rab7 and the HOPS tethering complex coordinate MDV transport along microtubules and ensure specific recognition between MDVs and late endosomes or lysosomes. Rab7 recruits HOPS components to promote tethering, while the HOPS complex stabilizes SNARE assembly and facilitates membrane merger.⁷⁹

Importantly, not all MDVs are destined for lysosomal degradation. A subset of MDVs can evade lysosomal routing and instead be incorporated into multivesicular bodies (MVBs) or budding ectosomes, subsequently released as EV-mito. Mechanistic insights, derived from a study using mouse embryonic fibroblasts, indicate that under basal, non-stressed conditions, healthy mitochondria generate MDVs through an OPA1- and SNX9-dependent process, in which OPA1 mediates inner mitochondrial membrane remodeling and SNX9 facilitates membrane curvature and vesicle scission.⁷⁷ These MDVs are subsequently incorporated into EVs and secreted extracellularly. In contrast, MDVs derived from damaged mitochondria are preferentially targeted to lysosomes for degradation through the PINK1/Parkin-dependent mitophagy pathway.⁷⁷ These distinct routes constitute a dual mitochondrial quality-control system: damaged components are selectively sequestered and degraded, while functional elements are exported via EVs. However, whether this dual routing system operates similarly in cardiomyocytes and other cardiovascular cell types in vivo remains incompletely defined.

Emerging evidence indicates that extracellular release of MDVs represents a regulated form of intercellular communication rather than a passive byproduct of mitochondrial turnover. One proposed role is to support neighboring or distant cells through a “Rescue by Replenish” mechanism, in which MDVs retaining functional integrity are taken up by recipient cells, exert cytoprotective effects, and contribute to the maintenance of organ-level homeostasis.⁸⁰Second, in stem cells, this process may contribute to the maintenance of quiescence and stemness.^81,82 In metabolically dormant cells, the extrusion of highly active mitochondria could serve to limit mitochondrial reactive oxygen species (mtROS) generation,^83,84 thereby fine-tuning intracellular redox balance and preserving genomic stability. Furthermore, the extracellular release of functional mitochondria serves as a paracrine signaling mechanism involved in immune modulation and inflammatory regulation. Circulating mitochondria and their components have been reported to influence macrophage polarization, phagocytic capacity, and systemic stress responses,⁸⁵ raising the possibility that similar pathways contribute to immune regulation within the myocardium.

MDV release through EVs constitutes a parallel quality-control pathway that complements mitophagy in maintaining mitochondrial homeostasis. Disruption of both Parkin-dependent and -independent mitophagy pathways increases the proportion of depolarized mitochondria within EV fractions.⁸⁶ Lysosomal inhibition enhances the secretion of EV-mito, which originate from MVBs and are released independently of autophagy. Deletion of Rab7, a key endolysosomal GTPase, markedly increases the secretion of ubiquitin-rich EVs containing mitochondrial cargo in cultured cells and mouse hearts.⁸⁷ These vesicles are engulfed by cardiac macrophages and degraded, exerting anti-inflammatory effects. Together, these findings position EV-mediated MDV release as an active and regulated arm of mitochondrial quality control in the cardiovascular system, operating in parallel with mitophagy in vivo to balance intracellular proteostasis, intercellular communication, and immune homeostasis.

Ectosome-Mediated Release and Uptake of Functional Mitochondria in the Cardiovascular System

In contrast to MDVs, entire and structurally intact mitochondria are rarely incorporated into MVBs but are instead packaged into ectosomes, the large plasma membrane-derived vesicles ranging from 0.3 to 3 μm in diameter, and released into the extracellular space through Ca²+-dependent enzymatic machinery.⁸⁸ A single ectosome can contain multiple mitochondria, contributing to the considerable heterogeneity of these structures, and cells can simultaneously release more than one subtype of EV-mito.⁸⁰ Notably, when EVs larger than 0.2 μm are removed by filtration, the transfer of mitochondria retaining membrane potential to recipient cells is significantly reduced,^{13, 89, 90} suggesting that functionally competent mitochondria are primarily enclosed within ectosomes. The molecular mechanisms governing the incorporation of mitochondria or MDVs into ectosomes remain incompletely defined. Current evidence suggests that mitochondrial fission is a prerequisite for their extracellular release.⁹¹ Mitochondrial fragmentation induced by Fis1 overexpression or OPA1 knockdown promotes Ca²+-dependent extrusion. Emerging evidence suggests that two Ca²+-mediated signaling pathways have been implicated: (i) CD38/ cADPR signaling,^80,91 and (ii) the CD38/IP₃R/Ca²+ axis,⁹² both facilitating stress-induced vesicular release. These pathways are particularly relevant in cardiovascular tissues, where Ca²+ flux, mechanical stress, and metabolic demand are tightly coupled.

Following their release, EVs interact with recipient cells primarily via integrin-mediated adhesion, which facilitates selective docking onto target cell surfaces.⁹³ Upon binding, EVs can either fuse directly with the plasma membrane or undergo endocytic internalization through multiple pathways, including clathrin-dependent,^94,95 caveolin-mediated,⁹⁶ and lipid-raft–associated endocytosis,⁹⁷ as well as phagocytosis and micropinocytosis.⁹⁸ However, which of these mechanisms predominantly mediates mitochondrial delivery to cardiomyocytes in vivo remains largely unknown.

EV-Mediated Mitochondrial Exchange and Homeostasis in the Cardiovascular System

In the cardiovascular system, EV-mediated mitochondrial transfer contributes to maintaining myocardial and vascular homeostasis through both metabolic replenishment and mitochondrial clearance. Induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) release EVs containing functional mitochondria, and delivery of these mitochondria-rich EVs to hypoxia-injured cardiomyocytes restores oxidative metabolism, improves contractility, and reduces cell death.⁸⁹ In vivo, local administration of mitochondria-rich EVs in murine myocardial infarction models enhances myocardial ATP content, improves left ventricular function, and attenuates adverse remodeling.⁸⁹ Similar protective effects have been observed in models of doxorubicin-induced cardiotoxicity, where MSC-derived EVs transfer mitochondria to injured cardiomyocytes, restoring bioenergetic capacity and contractile performance.⁹⁰

EV-mediated mitochondrial exchange in the heart appears to be bidirectional. Injured cardiomyocytes can export mitochondrial material to MSCs, where these mitochondria-associated signals act as danger-associated molecular patterns (DAMPs) that activate cytoprotective programs.⁹⁹ Although it remains unclear whether the exported entities represent entire mitochondria or sub-mitochondrial components, their uptake by MSCs triggers mitochondrial degradation–dependent signaling pathways, including induction of heme oxygenase-1, stimulation of PGC-1α– and TFAM-dependent mitochondrial biogenesis, and enhancement of the mitochondrial donor capacity of MSCs.⁹⁹ These primed MSCs confer robust cardioprotection in models of myocardial infarction and chemotherapy-induced injury.

Recent evidence suggests that EV-mediated mitochondrial clearance is essential for immune homeostasis in the healthy myocardium. Resident cardiac macrophages internalize mitochondria released from cardiomyocytes via membrane-bound particles.¹⁴ These mitochondria are functionally compromised and are extruded through an autophagy-dependent process. Disruption of macrophage-mediated clearance, such as loss of the phagocytic receptor Mertk, leads to inflammasome activation, impaired autophagy, accumulation of dysfunctional mitochondria, metabolic reprogramming, and deterioration of ventricular function.¹⁴ These cardiomyocyte-derived fragments differ from classical MDV-containing ectosomes in that they are micron-sized, contain large organelles, and are formed through an autophagy-dependent rather than endosome-derived pathway. Together, these findings position EV-mediated mitochondrial transfer as a fundamental mechanism by which cardiovascular tissues coordinate mitochondrial quality control, metabolic resilience, and intercellular communication under both physiological and pathological conditions.

Release and uptake of free mitochondria

The capacity of exogenous mitochondria to enter recipient cells and restore mitochondrial function was first demonstrated more than four decades ago. In 1979, isolated mitochondria were shown to be taken up by respiration-deficient (ρπ) yeast protoplasts, leading to functional recovery of oxidative metabolism when co-incubated in the presence of polyethylene glycol and CaCl₂.^100,101 These findings were subsequently extended to mammalian systems, where free mitochondria derived from drug-resistant mouse tumor cells were internalized by drug-sensitive cells, conferring antibiotic resistance and indicating preservation of mitochondrial genetic and functional integrity after uptake.¹⁰² Together, these studies established the principle that free mitochondria can be internalized and functionally incorporated into recipient cells under experimental conditions.

Despite these observations, the physiological relevance and mechanisms governing the release of free mitochondria into the extracellular space remain incompletely understood. Mitochondrial fission machinery has been implicated in facilitating mitochondrial extrusion,²² but direct evidence defining how cardiomyocytes or other cardiovascular cells release intact mitochondria in vivo is lacking. Free mitochondria have been identified in the circulation, with reported diameters ranging from approximately 0.2 to 1 μm, and have been shown to contain full-length mtDNA.^20,103 Some studies report preservation of mitochondrial membrane potential and oxygen consumption capacity,^21,103, whereas others, using high-resolution respirometry, failed to detect functional electron transport activity, raising questions about the bioenergetic competence of circulating free mitochondria in vivo.¹⁰⁴ In contrast, free mitochondria identified in the cerebrospinal fluid are considerably smaller (≈100–200 nm).^58,105 Electron microscopy revealed that these structures possess both inner and outer mitochondrial membranes but lack defined cristae architecture,⁵⁸ suggesting that these free mitochondria originate from MDVs. Approximately half of these free mitochondria retain membrane potential and are capable of generating ATP. Whether analogous populations of MDV-derived free mitochondria exist in the cardiovascular system, and whether they contribute meaningfully to myocardial homeostasis or repair, remains unknown.

Free mitochondria can be internalized by recipient cells through endocytic processes, but the dominant uptake mechanisms are not well defined.^106–111 Several studies have reported inefficient or absent mitochondrial internalization under simple co-culture conditions,^12,112 highlighting the context dependence of this process. Recently, a heparan sulfate-dependent endocytic pathway has been proposed as a potential route mediating mitochondrial uptake.¹¹³ Collectively, these observations underscore that, despite compelling experimental evidence for mitochondrial uptake, the physiological relevance, cellular sources, release mechanisms, and functional consequences of free mitochondria in vivo remain incompletely defined.

Fate of Transferred Mitochondria in Recipient Cells

The fate of exogenous mitochondria after cellular uptake remains unclear, but multiple lines of evidence indicates that functionally intact mitochondria can integrate into the host mitochondrial network, including within cardiovascular tissues. It has been reported that intact mitochondria introduced via various mechanisms, including cell fusion,^114–116 centrifugation-assisted delivery,¹¹⁷ spontaneous uptake of isolated mitochondria,^118,119 TNT-mediated transfer,^45,120 and EV-mediated delivery,^89,90,121 can integrate into the endogenous mitochondrial network of recipient cells. This integration involves fusion of both outer and inner mitochondrial membranes, although evidence suggests that donor and host mtDNA populations may remain spatially segregated within the recipient cell.¹¹⁶ Notably, mitochondrial fusion can occur even under conditions of impaired respiratory chain function, disrupted microtubule organization, or reduced intracellular ATP levels, whereas complete dissipation of mitochondrial membrane potential abolishes fusion and its restoration rescues this process.¹¹⁴ Overexpression of MFN1 further enhances mitochondrial merging, leading to improved bioenergetics, enhanced mitophagy-mediated quality control, and increased cell survival.¹¹⁴

Recently, a new hypothesis has been proposed regarding the cytoprotective effects of mitochondrial transfer.³⁰ During TNT-mediated mitochondrial transfer from MSCs to ECs, transferred mitochondria are reportedly not incorporated into the endogenous mitochondrial network but instead degraded via mitophagy.³⁰ This degradation triggers endogenous mitochondrial biogenesis, enhancing EC metabolic activity, apoptosis resistance, and migratory capacity. Thus, the degradation of exogenous mitochondria itself functions as a “priming effect” that reactivates cellular metabolism. This effect occurs regardless of whether the mitochondria are intact, depolarized, or mtDNA-deficient, indicating that functional incorporation of donor mitochondria into recipient cells is not required.³⁰

In post-mitotic cells, such as cardiomyocytes and neurons, mitochondrial transfer also plays a critical role in mitochondrial quality control through the “Relief by Release” mechanism. Damaged mitochondria extruded from stressed cardiomyocytes can be engulfed and degraded by cardiac macrophages, thereby limiting the intracellular accumulation of dysfunctional organelles and suppressing inflammation.^14,122 However, excessive or dysregulated mitochondrial release can have deleterious consequences. In particular, mitochondrial components released into the extracellular space—including mtDNA, cardiolipin, and N-formyl peptides—function as DAMPs. Although many studies have primarily examined these individual mitochondrial components rather than intact extracellular mitochondria, accumulating evidence indicates that such DAMP signals are sufficient to activate pattern recognition receptors, including Toll-like receptor 9, the NLRP3 inflammasome, and the cGAS–STING pathway, thereby promoting macrophage activation and dendritic cell maturation. The resulting paracrine and systemic cytokine cascades amplify inflammatory signaling, linking mitochondrial stress to innate immune activation and tissue injury.^123,124 Recent evidence further demonstrates that microvesicles released from LPS-primed macrophages can transfer mitochondrial material capable of inducing a similar inflammatory cascade.⁸⁵ These mitochondria-associated microvesicles reprogram recipient macrophages toward a pro-inflammatory phenotype, impair phagocytic capacity, and disrupt mitochondrial homeostasis, ultimately leading to organ injury in vivo. Notably, although this study does not directly establish the presence of intact, respiration-competent mitochondria within these vesicles, the transferred mitochondrial components are sufficient to activate inflammatory signaling pathways. Importantly, pharmacological inhibition of this pathway by metformin suppresses cytokine production, highlighting the mtROS/cGAS–STING–IFN-β axis as a potential therapeutic target for inflammatory cardiovascular conditions.⁸⁵ Collectively, these findings indicate that the fate of transferred mitochondria in cardiovascular tissues is highly context dependent. Importantly, future studies directly comparing intact extracellular mitochondria, mitochondria-containing vesicles, and isolated mitochondrial components within the same experimental systems will be required to resolve their relative stability, signaling capacity, and immunological impact in vivo. Functional mitochondria can integrate into recipient cells to support bioenergetic recovery and repair, whereas damaged mitochondria are preferentially eliminated through autophagy or immune-mediated clearance. In contrast, uncontrolled mitochondrial release promotes innate immune activation and inflammation, underscoring the need for tight regulation of mitochondrial transfer to preserve cardiovascular homeostasis.

Adipose–Cardiac–Immune Axis of interorgan mitochondrial exchange

Recent evidence indicates that EVs carrying mitochondrial components or free mitochondria circulate systemically and mediate interorgan mitochondrial communication, with potential implications for cardiometabolic adaptation. In white adipose tissue (WAT), adipocytes transfer mitochondria to resident macrophages via a heparan sulfate-dependent mechanism.¹¹³ Conditional deletion of the heparan sulfate biosynthetic gene Ext1 in myeloid cells disrupts mitochondrial uptake, leading to reduced energy expenditure, adipose expansion, and exacerbated diet-induced obesity. These findings indicate that intercellular mitochondrial transfer between adipocytes and macrophages is essential in maintaining immunometabolic homeostasis under physiological conditions. Under metabolic or oxidative stress, however, this local communication appears to be altered. Adipocytes instead release mitochondria encapsulated in small EVs (50–300 nm) into the circulation (Figure 3).⁵⁷ These small EV-mito, which retain respiratory activity despite oxidative damage, can be taken up by cardiomyocytes, where they induce a transient burst of ROS. This ROS surge activates compensatory antioxidant pathways and has been shown, in experimental models, to precondition the heart against ischemia-reperfusion injury.⁵⁷ Consistent with a functional role for this pathway, Parkin deficiency, which reduces mitochondrial content within small EVs, abolishes this cardioprotective effect.

Figure 3. Context-dependent and complementary model of mitochondrial transfer in the cardiovascular tissue.

Multiple modes of mitochondrial transfer appear to operate in parallel to support myocardial homeostasis. In the normal heart, EV-mediated transfer represents a constitutive pathway in which cardiomyocytes export damaged or dysfunctional mitochondria to cardiac macrophages, linking mitochondrial disposal to immune surveillance and tissue homeostasis (Relief by Release). EVs may also mediate mitochondrial delivery from extracardiac sources via the circulation, including white adipose tissue (WAT), potentially facilitating cardiac preconditioning against ischemia-reperfusion injury (Rescue by Replenish). In contrast, TNT-mediated mitochondrial transfer is constrained in the healthy, actively contracting myocardium due to the mechanical instability of the cytoskeletal scaffolds required for TNT formation, and may preferentially occur among non-cardiomyocytes or in cardiomyocytes with suppressed contractile activity under injury or disease conditions. Free mitochondrial transfer has been predominantly described in therapeutic contexts, where direct or intracoronary delivery augments myocardial bioenergetics; whether this mechanism represents an endogenous physiological process in the heart remains unclear. Illustration credit: Sceyence Studios

Complementary evidence from a murine study further demonstrates that dietary long-chain fatty acids selectively suppress mitochondrial capture by macrophages in WAT, diverting adipocyte-derived mitochondria into systemic circulation for redistribution to distal organs such as the heart.²⁰ Macrophage depletion increases circulating adipocyte-derived mitochondria, supporting their role as a local sink under physiological conditions. Together, these findings suggest the existence of a dynamic adipose-cardiac-immune axis, in which the fate of adipocyte-derived mitochondria, whether captured locally by macrophages or redistributed to distant organs, is determined by nutritional and metabolic context. This system functions as a regulatory switch that balances tissue-specific mitochondrial recycling and systemic metabolic preconditioning, providing a potential mechanistic framework for the adaptive cardiometabolic resilience observed in obesity, sometimes referred to as the “obesity paradox.”

Complementary model of mitochondrial transfer in the heart

At present, no studies have directly compared TNT-mediated, EV-mediated, and free mitochondrial transfer in the heart under identical experimental conditions, precluding conclusions regarding pathway dominance. Rather than a single prevailing mechanism, available evidence supports a context-dependent and complementary model in which mitochondrial transfer contributes to cellular and tissue homeostasis through multiple parallel pathways (Figure 3). Post-mitotic cells such as cardiomyocytes and neurons, and metabolically active cells such as adipocytes, appear capable of exporting these organelles to neighboring cells, particularly macrophages, thereby limiting intracellular accumulation of bioenergetically inefficient and pro-inflammatory mitochondria.^14,113,122 EV-mediated delivery of cardiomyocyte-derived mitochondria to cardiac macrophages represents one such potential pathway, linking mitochondrial disposal by cardiomyocytes to immune surveillance and myocardial homeostasis.¹⁴ Importantly, EV-mediated transfer may also enable mitochondrial delivery from extracardiac sources via the circulation,^57,87 indicating that mitochondrial exchange in the heart is not restricted to locally interacting cells. Conversely, cardiomyocytes or neurons experiencing insufficient energy supply may acquire mitochondria or mitochondrial components from surrounding cells, independent of the specific transport modality.^32,80 TNT-mediated mitochondrial transfer in vivo may be constrained in the healthy, actively contracting myocardium, where formation of stable cytoskeletal scaffolds required for TNTs is unfavorable. Accordingly, TNT-mediated transfer may preferentially occur among non-cardiomyocytes or within cardiomyocytes whose contractile activity is impaired by injury or disease. Free mitochondrial transfer has been primarily described in therapeutic contexts, where direct or intracoronary delivery augments myocardial bioenergetics. However, whether free mitochondrial transfer represents a physiologically occurring endogenous process in the heart remains unclear.

Cell-type specificity and disease context of mitochondrial transfer in cardiovascular system

Beyond transport modality, cell-type specificity and disease context critically shape the efficiency and functional consequences of mitochondrial transfer. Although direct head-to-head comparisons across donor and recipient cell types remain limited, available evidence supports a model in which mitochondrial transfer efficacy is highly cell-type– and context-dependent rather than universal (Table).

Table.

Cell-type specificity and disease context of mitochondrial transfer in cardiovascular system

Transfer	Model systems	Donor	Recipient	Functional outcomes	Ref
TNT	In vitro/co-culture (Normal)	MSCs	CMs (Neonatal rat)	Bidirectional cytoplasmic exchange; induction of β-myosin heavy chain in MSCs	54
	In vitro/co-culture (Hypoxia/reoxygenation)	MSCs	CMs (Neonatal mouse)	Reduction of CM apoptosis; effects dependent on intact microtubules and functional donor mitochondria	53
	In vitro/co-culture (Hypoxia/reoxygenation)	MSCs	H9C2	Reduction of apoptosis; recovery of mitochondrial membrane potential; effects attenuated by TNT inhibition	52
	In vitro/co-culture (Oxygen glucose deprivation)	MSCs*	H9C2	Reduction of cell death; increased number of viable cells; effects dependent on direct cell-to-cell contact	51
	In vitro/co-culture (Dox-induced injury) In vivo mouse (Anthracucline-induced cardiomyopathy model)	MSCs	CMs (Neonatal mouse)	Improved mitochondrial respiration; reduced cell death; MIRO1- and TNF-α/TNFaIP2/NF-κB–dependent TNT formation; in vivo improvement of LVEF and myocardial ATP level	46
	In vitro/co-culture (Normal) Ex vivo mouse heart (Normal)	Fibroblasts*	CMs (Neonatal rat)	Bidirectional transfer of vesicles/organelles within nanotubes; propagation of Ca2+; similar thin nanotuble in adult heart tissue	31
	In vitro/co-culture (Hypoxia/reoxygenation)	Fibroblasts	CMs (Neonatal rat)	Reduction of cell death; effects abolished by transwell separation, microtubule depolymerization, or KIF5B knockdown	32
	In vitro/co-culture (Oxygen glucose deprivation)	MSCs	ECs (HUVECs)	Reduction of cell death; recovery of mitochondrial respiration; effects attenuated by actin inhibition, or mtDNA-depleted MSCs; detection of MSC-derived mtDNA	49
	In vitro/co-culture (Normal) In vivo mouse (hindlimb ischaemia)	MSCs	ECs	Increased ATP and reduced apoptosis; transfer enhanced by stress and reduced by TNFAIP2 or MIRO1 knockdown; transient presence of donor mitochondria in ECs in vivo; increased EC survival and microvessel density; improve blood flow in ischaemic hindlimb; donor mitochondria do not stably fuse with recipient mitochondria and are cleared; EC engraftment requires PINK1–Parkin-dependent mitophagy	30
	In vitro/co-culture (Normal)	VSMCs*	MSCs	Increased MSC proliferation in direct co-culture; abolished by actin inhibitors; effect lost with mtDNA-depleted VSMCs despite preserved TNT formation	19
	In vitro/co-culture (Normal)	CMs	Endothelial progenitor cells	Mitochondrial transfer via nanotubes stable cell or nuclear fusion	50
EV	In vitro (H2O2) In vivo mouse (myovardial infarction)	MSCs*	CMs (RL-14) ECs (HUVECs)	Increased mitochondrial respiration and ATP levels; reduced apoptosis in vitro; in vivo detection of EV-derived mitochondria in CMs within infarcted myocardium; reduction of CM apoptosis	99
	In vitro/co-incubation (Hypoxia/reoxygenation) In vivo mouse (myovardial infarction)	CMs (iPSC-derived)	CMs (iPSC-derived)	Increased mitochondrial respiration and ATP levels; reduced apoptosis in vitro; in vivo detection of EV-derived mitochondria in CMs; Improvement of LVEF	89
	In vitro/co-culture DOX	MSCs	CMs (iPSC-derived)	Improved CM contractility and cell viability: reduced apoptosis; reduced ROS; increased ATP	90
	In vivo mouse (Isoproterenol)	CMs	Macrophage	Macrophage depletion reduced mitochondrial fitness, decreased ATP production, altered　18F-FDG uptake, and impaired cardiac function. Isoproterenol increases exopher formation and macrophage uptake of CM-derived mitochondria	14
	In vivo mouse (obese/myocardial ischemia-reperfusion injury)	Adipocytes	CMs	Adipocytes under mitochondrial stress release EV-mito; these EVs are taken up by CMs in vivo; increased cardiac ROS and induction of antioxidant signaling; a single injection of EVs reduced infarct size	57
Free mitochondria	In vitro/co-incubation (mtDNA-depletion)	MSCs	H9C2	Internalization of isolated mitochondria into recipient cells; transient perinuclear localization; restoration of cellular viability and mitochondrial respiration in mtDNA-depleted cells; effects attenuated by macropinocytosis inhibition	108
	In vivo rabbit (myocardial ischemia–reperfusion)	Skeletal muscle	CMs	Direct myocardial injection results in uptake of transplanted mitochondria by CMs with preserved membrane potential; increased myocardial ATP and mitochondrial respiration; reduction of CM apoptosis and infarct size; improved LVEF	125
	In vivo porcine (myocardial ischemia–reperfusion)	Skeletal muscle	CMs	Direct intramyocardial injection reduced infarct size; preservation of myocardial ultrastructure; absence of immune activation; transplanted mitochondria detected in myocardium for up to 4 weeks	126
	In vitro/co-incubation (Normal)	Skeletal muscle	H9C2	Transient perinuclear localization; short-term enhancement of mitochondrial respiration and ATP production; bioenergetic indices returned to baseline by 28 days without increased mitochondrial ROS	127
	In vitro/co-incubation (mtDNA-depletion)	CMs (Neonatal rat)	CMs (Neonatal rat) HeLa	Internalization via actin-dependent endocytosis; increased ATP and mitochondrial respiration; replacement of mtDNA in mtDNA-depleted cells	109

^*:Bidirectional exchange of mitochondria

Among donor cell types, MSCs are the most extensively studied source of mitochondria transferred to cardiomyocytes. MSC-mediated transfer via TNTs has been primarily characterized in in vitro coculture systems; however, accumulating evidence suggests that MSCs can unidirectionally donate mitochondria to cardiomyocytes both at baseline and under stress conditions, consistent with a “Rescue by Replenishment” paradigm.^46,51–54 In contrast, EV-mediated mitochondrial exchange between MSCs and cardiomyocytes appears bidirectional, and in vivo mouse models of myocardial infarction indicate that this reciprocal exchange contributes to cardioprotection and functional recovery.⁹⁹ In addition, EV-mito derived from MSCs or iPS-CMs, as well as isolated free mitochondria, can be taken up by cardiomyocytes following intramyocardial injections, improving myocardial bioenergetics and function in in vivo myocardial infarction models.^89,125,126 Cardiac fibroblasts also exhibit context-dependent mitochondrial transfer behavior, showing bidirectional exchange with cardiomyocytes under physiological conditions, but predominantly unidirectional transfer toward cardiomyocytes during acute stress,^31,32 supporting a compensatory bioenergetic mechanism. Vascular ECs also participate in mitochondrial transfer within the cardiovascular system, with evidence suggesting that mitochondrial donation from MSCs via TNTs contributes to endothelial metabolic reprogramming, angiogenic capacity, and maintenance of vascular homeostasis under stress conditions.^30,49 In contrast, terminally differentiated cardiomyocytes preferentially engage EV-mediated pathways to export damaged mitochondria to resident macrophages as part of mitochondrial quality control and immune surveillance, representing a complementary “Relief by Release” mechanism.^14,87 Beyond the myocardium, adipocyte-derived mitochondria primarily participate in systemic EV-mediated signaling that influences cardiac stress adaptation within a broader metabolic context.⁵⁷ These comparative observations suggest that MSC- or cardiomyocyte-derived mitochondrial transfer may be particularly suited for acute myocardial injury, whereas cardiomyocyte–macrophage mitochondrial exchange appears central to mitochondrial quality control and immune homeostasis in the adult heart. Collectively, these observations argue against a single dominant transfer mechanism and instead support a disease-, cell-type-, and cell-state-specific framework in which mitochondrial exchange fulfills distinct yet complementary roles in cardiovascular homeostasis.

Therapeutic Implications for mitochondrial transfer and potential risks

Conceptually, therapeutic strategies leveraging mitochondrial transfer can be categorized into at least four layers of intervention (Figure 4): (i) Cell-based therapies designed to transfer healthy mitochondria or to scavenge damaged ones, cell-free approaches using (ii) EV-mito or (iii) purified mitochondria, (iv) pharmacological or lifestyle interventions that enhance endogenous mitochondrial exchange, and improve the clearance of damaged mitochondria to limit inflammation. Regardless of the mode of transfer, mitochondrial transfer carries three major potential risks that must be overcome: inflammation, oncogenic transformation, and mtDNA incompatibility.

Figure 4. Therapeutic implication for mitochondria transfer.

Mitochondrial transfer offers therapeutic potential through three major approaches: cell-based, cell-free (vesicle or free mitochondria), and indirect interventions. In cell-based therapies, donor cells transfer functional mitochondria via tunneling nanotubes (TNTs) or extracellular vesicles (EVs), improving bioenergetics and tissue repair via paracrine effect. Cell-free approaches use mitochondria-containing extracellular vesicles (EV-mito) or isolated mitochondria as therapeutic agents. EV-mediated delivery provides low immunogenicity and protection of mitochondria from extracellular environment. Free mitochondria offer rapid metabolic rescue but face stability and immune-activation challenges. Indirect enhancement through exercise, nutrition, or activation of PGC-1α/AMPK/SIRT1 pathways promotes mitochondrial biogenesis and may enhance transfer efficiency and integration. Although the field of mitochondrial transfer therapy is still in its infancy, rapid progress in bioengineering, delivery systems, and mechanistic understanding holds promise for overcoming current challenges and realizing its clinical potential. Illustration credit: Sceyence Studios

Inflammation and Immunological Considerations

At present, direct evidence that administration of autologous or allogeneic mitochondria elicits overt or clinically meaningful immune responses remains limited. Notably, in a rabbit ischemia–reperfusion model, intracoronary delivery of autologous free mitochondria did not induce systemic inflammation or autoimmune responses.¹²⁵ Inflammatory markers were reduced compared with vehicle-treated controls, and antimitochondrial antibodies were not detected up to 28 days after transplantation, supporting the immunological tolerability of autologous mitochondrial transfer in vivo. Consistent findings were reported in a porcine model from the same group, in which intramyocardial injection of autologous mitochondria did not alter circulating inflammatory cytokine levels at 30 days post-treatment.¹²⁶

However, from a mechanistic perspective, mitochondria are intrinsically immunogenic organelles. Even when derived from self, mitochondrial components can function as DAMPs and activate innate immune pathways. In line with this concept, accumulating evidence indicates that donor-derived circulating extracellular mitochondria contribute to immune activation and allograft rejection following organ transplantation.^128–130 Experimental infusion of isolated mitochondria into heart donors significantly exacerbated allograft rejection in a murine heterotopic heart transplantation model.¹²⁸ Moreover, co-incubation of human peripheral blood mononuclear cells with mitochondria-exposed ECs increased the frequency of effector CD8+ T cells producing IFN-γ and TNF-α, indicating potentiation of adaptive immune responses.¹²⁸ These observations highlight an important safety consideration for the systemic administration of mitochondria, particularly free mitochondria, in which mitochondrial DAMPs are directly exposed to the extracellular immune surveillance system. Accordingly, comprehensive preclinical evaluation of safe dosage ranges, biodistribution, clearance kinetics, and tissue-specific immune activation is essential prior to clinical translation. Beyond these considerations, an additional unresolved question concerns whether concomitant immunosuppression will be required during mitochondrial transfer–based therapies, particularly in non-autologous or systemic administration settings. To address the immunological challenges, innovative strategies to mitigate mitochondrial immunogenicity, including surface modification, encapsulation within vesicular carriers, controlled release systems, or transient immunosuppression, should be explored to enhance the safety profile and translational feasibility of mitochondrial transfer-based therapeutic approaches.^131–134

Oncogenic Risks of Mitochondrial Transfer Therapy

Mitochondrial function is indispensable for tumor growth, invasiveness, and metastatic potential. Accumulating evidence indicates that intercellular mitochondrial transfer between stromal cells and cancer cells appears to occurs predominantly via TNTs.¹³⁵ TNTs are frequently formed by cancer cells themselves, and most studies identify cancer cells as the primary recipients of transferred mitochondria. The biological consequences of mitochondrial transfer to cancer cells remain controversial. While some studies report tumor-suppressive effects of mitochondrial transfer,¹³⁶ a larger body of evidence suggests tumor-promoting outcomes, including enhanced proliferation, invasiveness, and metastatic capacity.^137–139 In particular, several studies have demonstrated that MSCs can transfer mitochondria to cancer cells, leading to increased oxidative phosphorylation, ATP production, and metabolic fitness, thereby promoting tumor cell growth, migration, invasion, and resistance to chemotherapy.^140–142 From a clinical perspective, the potential for tumor exacerbation represents a critical safety concern for mitochondrial transfer-based therapies. To date, no animal studies or clinical trials of mitochondrial transfer therapies or MSC-based treatments have reported increased tumor growth or de novo tumorigenesis attributable to these interventions. Existing safety data are largely derived from small-scale studies with short- to mid-term follow-up, and therefore remain insufficient to definitively exclude long-term oncogenic risk. Although no established bioengineering strategies currently exist to selectively prevent mitochondrial transfer to malignant cells, these considerations underscore the importance of rigorous pre-treatment screening for occult malignancies, as well as long-term post-treatment surveillance in future clinical trials. Such safeguards will be essential to ensure the safe translation of mitochondrial transfer-based therapeutic approaches, particularly in patient populations at elevated risk for cancer.

Genetic Incompatibility and Mitonuclear Conflict

To date, clinical studies have primarily utilized autologous mitochondrial transplantation (NCT02851758; NCT03639506; NCT05669144; NCT04998357),¹⁴³ thereby minimizing the risk of heteroplasmy and reducing both immunological and genetic incompatibility. Future efforts to expand the universality of mitochondrial transplantation, through allogeneic or xenogeneic sources, may introduce additional biological and immunogenetic challenges. In particular, the coexistence of mitochondria carrying distinct mtDNA haplotypes within the same recipient cell may give rise to heteroplasmic imbalance, defined as a progressive shift in the relative abundance of coexisting mtDNA populations that can cross pathogenic thresholds and impair mitochondrial and cellular function. While mtDNA carryover and reversion driven by genetic drift are well-recognized limitations of mitochondrial replacement therapy,^144,145 such phenomena are generally considered less likely in somatic mitochondrial transplantation, where limited cell division and constrained mitochondrial replication reduce opportunities for genetic drift, especially in post-mitotic cells such as cardiomyocytes. Nevertheless, under conditions involving non-autologous transplantation or long-term persistence of exogenous mitochondria, heteroplasmic imbalance cannot be entirely excluded and warrants continued mechanistic and longitudinal investigation.

Furthermore, the transplantation of xenogeneic mitochondria harboring genetically divergent mtDNA may disrupt the finely coordinated interaction between mitochondrial and nuclear genomes, resulting in mitonuclear conflict, a state of genomic incompatibility that compromises cellular function, bioenergetic efficiency, and metabolic integrity.¹⁴⁶ Collectively, these considerations highlight the importance of donor–recipient genetic compatibility and long-term safety assessment as mitochondrial transplantation strategies move toward broader clinical application.

Cell-based therapies for mitochondrial delivery

Extensive preclinical evidence across both small and large animal models has demonstrated that administration of MSCs and other progenitor cell populations improves cardiac structure and function following acute and chronic myocardial infarction.^147,148 Although the quantitative contribution of mitochondrial transfer—whether mediated through TNTs or EVs—to the overall paracrine effects of MSCs remains to be fully elucidated, it is generally accepted that the therapeutic efficacy of MSCs primarily stems from their paracrine activity. Despite promising preclinical evidence, clinical outcomes of MSC-based therapies in cardiovascular diseases have remained limited.^149,150 These translational discrepancies may highlight several intrinsic limitations of cell-based approaches, including immune incompatibility and host immune responses, loss of stemness and metabolic fitness during ex vivo expansion, donor-to-donor variability, cellular heterogeneity, and limited tissue homing, retention and engraftment.^151–153

Importantly, emerging insights into mitochondrial transfer biology suggest that cell-based therapies may influence myocardial homeostasis not only by supplying functional mitochondria (“rescue by replenish”), but also by indirectly facilitating mitochondrial quality control through immune-mediated disposal mechanisms (“relief by release”). In this context, MSCs may shape mitochondrial quality control within the injured myocardium by modulating the immune microenvironment. In particular, extensive evidence indicates that MSCs promote a reparative polarization of cardiac-resident macrophages through paracrine signaling.^154,155 These cardiac-resident macrophages appear to play a central role in clearing damaged cardiomyocyte-derived mitochondria and limiting the accumulation of bioenergetically inefficient or pro-inflammatory organelles. From a translational perspective, this framework reframes mitochondrial transfer as a coordinated, immune-coupled process, highlighting that effective therapies may need to modulate the myocardial immune–metabolic microenvironment in addition to enhancing mitochondrial delivery itself.

To overcome the translational barriers associated with conventional cell-based therapies, the development of next-generation bioengineering strategies that enhance mitochondrial transfer and its downstream handling is warranted. These may include reprogramming chemokine receptor expression to improve tissue homing,¹⁵⁶ employing localized or scaffold-based delivery systems to increase retention,¹⁵⁷ and identifying or genetically engineering donor cells with superior mitochondrial donation capacity.^92,158–160 Collectively, these approaches highlight the need to integrate mitochondrial delivery strategies with immune modulation to maximize therapeutic efficacy.

Cell-free Extracellular Vesicle-Mediated Delivery

EV-mediated mitochondrial delivery represents a promising cell-free strategy for organelle replacement therapy. EVs possess inherently low immunogenicity and can be administered repeatedly without the oncogenic risks inherent to live-cell transplantation. Their lipid bilayer effectively shields encapsulated mitochondria from extracellular degradation, particularly under high-calcium conditions, and from direct extracellular immune surveillance, thereby maintaining mitochondrial integrity and facilitating efficient internalization by recipient cells. Moreover, surface engineering of EVs with targeting peptides,¹⁶¹ fusogenic peptides,¹⁶² antibodies,¹⁶³ or anti-phagocytic molecules¹⁶⁴ can further enhance organ- or cell-type–specific homing.

Despite these advantages, EV-mediated mitochondrial transfer remains largely at the preclinical stage and faces several key challenges. A major bottleneck is the intrinsic heterogeneity of EV populations. Current preparations contain diverse subtypes, including exosomes, ectosomes, and MDVs, of which only a small fraction harbor intact and functional mitochondria. The precise isolation of EV-mito from dysfunctional vesicles remains technically difficult. It also remains unclear under what cellular conditions or culture environments mitochondria with preserved respiratory competence are selectively packaged into EVs, and which donor cell types are optimal for producing EVs enriched with functional mitochondria. Additionally, the intracellular fate of transferred mitochondria is not fully understood—specifically, how they evade lysosomal degradation and integrate into the host mitochondrial network. Elucidating these mechanisms will be crucial for achieving sustained bioenergetic recovery in target tissues. Technical challenges also persist in maintaining mitochondrial structural integrity and membrane potential during EV isolation, purification, and storage. Furthermore, the absence of standardized potency assays to quantitatively assess mitochondrial content, quality, and transfer efficiency hampers inter-study comparisons and regulatory evaluation. Addressing these limitations will be essential for the rational development of scalable, reproducible, and clinically translatable EV-based mitochondrial delivery platforms.

Isolated Mitochondrial Transplantation

Direct injection of free mitochondria into the myocardium has been shown to ameliorate IR injury in both small and large animal models,^125,126 and a small clinical study.¹⁴³ Mitochondrial transplantation has emerged as a promising therapeutic strategy; however, several scientific and technical challenges remain to be addressed. It has been reported that free mitochondria are unlikely to survive in circulation due to the high Ca²+ concentration in the extracellular milieu (above ~1 mM), which can induce irreversible rupture of the outer mitochondrial membrane and the subsequent release of pro-inflammatory mediators.^165,166 Conversely, other studies have demonstrated that free mitochondria with preserved membrane potential can be detected in circulation,^21,103 and cerebrospinal fluid,⁵⁸ although it remains unclear whether all such mitochondria exist as free organelles. Moreover, previous reports have shown the uptake and functional integration of exogenous mitochondria into recipient cells in media containing 1.8 mM of calcium.^167–169 At present, there is no clear consensus regarding the duration for which free mitochondria retain structural integrity and bioenergetic competence under physiological calcium conditions. Furthermore, the mechanisms governing cellular uptake, intracellular trafficking, and integration of free mitochondria into the native mitochondrial network remain incompletely understood. In addition, transplanted mitochondria have been shown to persist only transiently in vivo,¹²⁷ suggesting that further investigation is warranted to clarify the duration and sustainability of their therapeutic effects. From a technical standpoint, free mitochondria are highly fragile organelles that rapidly lose their structural and functional integrity after isolation. Although they can remain viable on ice for up to one hour, prolonged storage or cryopreservation at −80 °C disrupts the outer membrane, markedly reducing transplantation efficiency.¹⁷⁰

To address these challenges, recent advances have focused on encapsulating isolated mitochondria within hydrogels,¹⁷¹ liposomes,¹³² or polymeric carriers¹⁷² designed to preserve mitochondrial structural and functional integrity by mitigating oxidative insult, enzymatic degradation, calcium overload, and mechanical shear stress encountered during handling and transplantation. In parallel, bioengineering approaches functionalizing mitochondrial membranes with targeting ligands have been developed to enhance organ-specific delivery and cellular uptake efficiency.^173–175 Taken together, while isolated mitochondrial transplantation represents a promising therapeutic concept, critical questions regarding mitochondrial survival, trafficking, functional integration, and long-term impact in vivo remain largely unresolved.

Indirect Enhancement through Pharmacological, Nutritional, and Lifestyle Interventions

Currently, no interventions are known to directly promote mitochondrial transfer. However, several approaches may indirectly enhance intercellular mitochondrial exchange and/or facilitate the functional integration and persistence of transferred mitochondria. Nutritional and metabolic cues profoundly influence mitochondrial turnover and trafficking.¹⁷⁶ Physiological stimuli such as exercise¹⁷⁷ and intermittent fasting,¹⁷⁸ as well as pharmacological activation of PGC-1α, AMPK, and SIRT1,¹⁷⁹ have been shown to stimulate mitochondrial biogenesis and improve cellular bioenergetics. Modulators of mitochondrial fusion and fission may also optimize donor-cell mitochondrial fragmentation and recipient-cell mitochondrial fusion processes, thereby improving the efficiency of mitochondrial transfer. Furthermore, pharmacological interventions for metabolic disorders such as obesity, diabetes, and non-alcoholic fatty liver disease can modulate mitochondrial dynamics and function.¹⁷⁹ Although these strategies hold promise for indirectly enhancing mitochondrial exchange, rigorous mechanistic studies are still required before their deliberate application to promote mitochondrial transfer in clinical settings.

Unanswered questions and future research directions

Despite substantial progress, several key questions remain unresolved in the field of intercellular mitochondrial transfer, continuing to limit its mechanistic understanding and translational advancement. First, the molecular and cellular triggers that initiate mitochondrial release remain incompletely defined. While metabolic stress, inflammation, and tissue injury have been implicated, the upstream signaling pathways and contextual cues that determine whether mitochondria are transferred via TNTs, packaged within EV, or released as free mitochondria are poorly understood. Dissecting how donor cell identity, energetic state, and microenvironmental signals bias these distinct transfer routes represents an important unmet need. Second, the tissue and cellular origins of circulating cell-free mitochondria remain largely undefined. Determining whether EV-Mito arise predominantly from cardiomyocytes, ECs, immune cells, or other stromal populations will be critical for interpreting their biological significance. Future studies employing lineage-tracing strategies, surface protein signatures, and mitochondrial markers such as TOMM20 or cell-type-specific mitochondrial reporters will be essential to resolve this question. Third, the intracellular fate of transferred mitochondria within recipient cells remains incompletely understood. It is unclear which mitochondria functionally integrate into the endogenous mitochondrial network, which are selectively eliminated through mitophagy or endolysosomal pathways, and how these divergent outcomes contribute to cytoprotection, metabolic adaptation, immune modulation, or tissue repair. Clarifying the determinants of mitochondrial persistence versus degradation will be central to understanding functional benefit versus neutral or deleterious effects. Fourth, the trafficking dynamics of EV-Mito or free mitochondria in vivo remain poorly characterized. Fundamental questions remain regarding how EV-Mito enter the circulation, cross physiological barriers such as the blood–brain barrier or vascular endothelium, and undergo clearance. Defining organ biodistribution, cellular uptake pathways, and residence time will be necessary to assess both physiological relevance and therapeutic potential. Fifth, beyond the mitochondria themselves, the molecular cargo of EV-Mito warrants further investigation. Associated proteins, lipids, metabolites, and nucleic acids may critically shape immunologic, metabolic, or regenerative signaling in recipient cells, yet remain largely uncharacterized. Understanding how these co-transported components modulate biological responses may help explain context-dependent outcomes of mitochondrial transfer. Sixth, the physiological relevance and quantitative contribution of mitochondrial transfer under baseline conditions, compared with pathological or therapeutic settings such as ischemia-reperfusion injury or heart failure, have yet to be established. Whether mitochondrial transfer represents a rare stress-induced rescue mechanism or a constitutive homeostatic process remains an open question. Finally, it remains unclear to what extent mitochondrial transfer mechanisms and functional consequences differ across cell types, developmental stages, and disease contexts, including acute myocardial injury versus chronic cardiomyopathy. Addressing these issues will require robust in vivo tracking and quantification strategies, standardized functional and bioenergetic assays, and carefully designed translational studies that integrate mechanistic insight with clinical relevance. Together, addressing these unresolved questions will be essential to define the mechanistic principles governing mitochondrial transfer across physiological and pathological contexts and to enable rational translation of this biology into cardiovascular therapeutics.

Conclusion

Mitochondrial transfer and transplantation have emerged as conceptually novel therapeutic strategies aimed at restoring cellular bioenergetics and tissue resilience in diseases characterized by mitochondrial dysfunction. Accumulating evidence across cardiovascular and other organ systems indicates that mitochondria function not only as intracellular powerhouses but also as dynamic intercellular signaling entities that coordinate metabolic adaptation, quality control, and immune homeostasis. At the same time, this field remains at an early stage of development, and successful clinical translation will require overcoming substantial biological and technical challenges, including control of inflammation and immunogenicity, mitigation of oncogenic and genetic risks, and optimization of mitochondrial sourcing, manufacturing, preservation, and delivery. Importantly, emerging data suggest that therapeutic efficacy will depend not simply on mitochondrial supplementation per se, but on the precise engagement of context-dependent mechanisms that integrate mitochondrial transfer with immune regulation, metabolic remodeling, and cellular quality-control pathways. In cardiovascular biology, this paradigm challenges the traditional view of mitochondria as static, cell-autonomous organelles and instead positions them as intercellular signaling units that actively shape cardiac homeostasis and disease progression. From a cardiovascular perspective, these insights define several actionable research priorities. Future studies should (i) delineate cell-type–specific donor–recipient relationships among cardiomyocytes, ECs, fibroblasts, and cardiac-resident immune cells; (ii) determine how distinct mitochondrial transfer modalities—TNT-mediated, EV-mediated, or free mitochondrial transfer—are differentially engaged in acute ischemic injury versus chronic heart failure; and (iii) develop disease-relevant in vivo models that enable quantitative tracking of mitochondrial transfer and directly link transfer dynamics to cardiac metabolism, inflammation, remodeling, and functional outcomes. Ultimately, by reframing mitochondrial exchange as a coordinated, multicellular and context-dependent biological process rather than a single delivery modality, future research may uncover opportunities to complement existing cardiovascular therapies, refine patient selection, and redefine how mitochondrial dysfunction is therapeutically targeted in cardiovascular medicine.

Non-standard Abbreviations and Acronyms

Cx43

connexin 43

DAMPs

danger-associated molecular patterns

ECs

endothelial cells

EVs

extracellular vesicles

EV-mito

mitochondria-containing extracellular vesicles

GDMT

guideline-directed medical therapy

HUVECs

human umbilical vein endothelial cells

iPSC-CMs

induced pluripotent stem cell–derived cardiomyocytes

MDVs

mitochondria-derived vesicles

MSCs

mesenchymal stem cells

mtDNA

mitochondrial DNA

mtROS

mitochondrial reactive oxygen species

MVBs

multivesicular bodies

ρπ cells

mitochondrial DNA–depleted cells

TFAM

mitochondrial transcription factor A

TNTs

tunneling nanotubes

WAT

white adipose tissue