Therapeutic transplantation of mitochondria and Extracellular Vesicles: Mechanistic insights into mitochondria bioenergetics, redox signaling, and organelle dynamics in preclinical models

Abstract

Mitochondrial and extracellular vesicles (EV) transplantation have emerged as promising therapeutic strategies targeting mitochondrial dysfunction, a central feature of numerous pathologies. This review synthesizes preclinical data on artificial mitochondrial and EV transfer, emphasizing their therapeutic potential and underlying mechanisms. A systematic analysis of 123 animal studies revealed consistent benefits across diverse models, including ischemia-reperfusion injury (IRI), neurological disorders, drug-induced toxicities, and sepsis. Mitochondrial transfer improved organ function, reduced inflammation and apoptosis, and enhanced survival. Mechanistic insights revealed restored bioenergetics, increased oxidative phosphorylation, redox balance through activation of specific pathways, and modulation of mitochondrial dynamics via fusion/fission proteins. Mitochondrial homeostasis was supported through elevated mitophagy and biogenesis, alongside the preservation of mitochondrial-associated membranes. EV demonstrated similar effects, offering a potentially more targeted therapeutic alternative. Although pre-clinical studies have demonstrated safety and feasibility, broader application is limited by variability in isolation methods, lack of mechanistic clarity, and minimal human data. Standardization and mechanistic validation are critical to advance clinical translation. This review underscores the therapeutic promise of mitochondrial and EV transfer while highlighting the need for continued research to refine these interventions and unlock their full potential in regenerative medicine.

1. Introduction: mitochondrial and extracellular vesicles (EV) transfer mechanisms and therapeutic potential

Mitochondria play a central role in maintaining cellular homeostasis by regulating energy production, redox signaling, apoptosis, and immunomodulation. As the primary producers of adenosine tri-phosphate (ATP) through oxidative phosphorylation (OXPHOS), they are essential for cell viability and function. Beyond their role in energy metabolism, mitochondria interact closely with various cellular processes, including metabolic regulation, stress response, and inflammation modulation [1]. Their involvement in a wide range of pathologies, from neurodegenerative and cardiovascular diseases to cancer and metabolic disorders, underscores the importance of maintaining optimal mitochondrial function.

Recent studies have revealed that mitochondrial transfer between cells constitutes a natural biological mechanism that helps preserve cellular homeostasis and promote survival in response to environmental stress [2]. This transfer can occur through various mechanisms, including tunneling nanotubes (TNTs) [3]. For instance, mesenchymal stem cells (MSCs) have been shown to transfer functional mitochondria to damaged pulmonary epithelial cells, thereby enhancing their metabolic function and resistance to oxidative damage [4,5]. On the other hand, transfer of non-functional mitochondria from macrophages to cancer cells promoted their proliferation [6]. These observations suggest that inter-cellular mitochondrial transfer is an adaptive process aimed at restoring bioenergetic function and mitigating the deleterious effects of cellular injury.

Inspired by these endogenous biological mechanisms, therapeutic approaches based on artificial mitochondrial transfer have been investigated [7]. The concept of mitochondrial transplantation has emerged as an innovative strategy to restore mitochondrial function and improve cell viability in various pathological contexts. This approach is based on the hypothesis that healthy mitochondria, isolated from donor cells, can be introduced into recipient cells to enhance their bioenergetic capacity and stress resistance. Several studies suggest that transplanted mitochondria can be internalized by target cells, integrated into the endogenous mitochondrial network, and restore deficient metabolism [8–10]. Two main mitochondrial transplantation strategies are currently being explored: (i) direct administration of isolated mitochondria and (ii) the use of extracellular vesicles (EV), which represent a more physiological and potentially more targeted approach.

EV are lipid bound vesicles released by all cell types and play a crucial role in tissue crosstalk by transporting and delivering bioactive molecules such as proteins and RNA that contribute to intracellular signaling pathways and regulate functions in recipient cells [11,12]. They encompass various subtypes based on size and biogenesis pathway. In the last decade according to the different versions of the Minimal Information for Studies of Extracellular Vesicles (MISEV) [13,14] the EV can generally be divided into 2 classes based on size: small EV, described as <200 nm in diameter and large EV, with a diameter >200 nm. Moreover, there are also other classifications based on the biogenesis pathway: exosomes (30–150 nm) formed through the endosomal pathway, within multivesicular bodies (MVBs), which are then released into the extracellular space upon fusion of MVBs with the plasma membrane, microvesicles (100–1000 nm), formed by the outward budding and fission of the plasma membrane, apoptotic bodies (>1000 nm), formed during programmed cell death (apoptosis) through cell fragmentation, where the dying cell undergoes membrane blebbing and breaks into membrane-bound vesicles containing cytoplasmic contents, organelles, and nuclear fragments [14].

Recently another 2 classes of EV (not yet discussed in the MISEV 2023) were described: mitochondria derived vesicles (MDV) and the mitochondria-EV (mito-EV). MDV are small (~70–150 nm) vesicles that shuttle mitochondrial constituents to other organelles, through the sorting of mitochondrial components via two different pathways. The first pathway involves the delivery of mitochondrial material to EV through sorting nexin 9 (SNX9)-dependent MDVs. In the second pathway, oxidative stress induces the production of MDVs containing damaged mitochondrial components, targeted to lysosomes for degradation [15,16]. Mito-EV are a broader and functionally distinct category encompassing all EV subtypes—small-EV, large EV, exosomes and microvesicles —that carry mitochondrial cargo, including mitochondrial proteins, mtDNA, or even whole mitochondria [17]. Mito-EV originate from various biogenesis routes, including MDV–endolysosomal fusion, direct plasma membrane budding, and exocytosis of mito-lysosomes [17]. They are actively secreted and taken up by recipient cells, where they can modulate metabolism, promote survival, and activate signaling pathways. These EV have been isolated from many cell types and bodily fluids, reflecting great heterogeneity in size (tens of nm to μm) and composition [18]. Mito-EV can carry either intact mitochondria or selected mitochondrial components [19]. Despite these 2 new EV categories, in literature the main papers investigating the effects of EV on mitochondrial function didn’t explicitly isolate and characterize the EV as MDV or mito-EV as there is still not precise guidance. Moreover, the literature has shown that all EV subtypes can transfer intact mitochondria and other mitochondrial material between cells in response to stress conditions [20]. In particular, MSCs and macrophages can release EV containing functional mitochondria. These EV are then taken up by recipient cells, promoting their survival and metabolic function. Furthermore, electron microscopy and flow cytometry analyses have confirmed the presence of whole mitochondria or mitochondrial components (mtDNA, respiratory chain proteins) in EV [21–23]. It is likely that in the coming years we will be able to distinguish between MDVs and mito-EV among the many different types of EV. However, as the current literature still lacks the tools and criteria necessary to make this distinction reliably, in this review we refer more broadly to EV that carry mitochondrial cargo or exert mitochondrial-related effects.

These findings have paved the way for exploring EV as natural vectors for mitochondrial and mitochondrial related material transfer, with potential applications in cell therapy and regenerative medicine. The employment of EV represents a complementary and innovative approach. EV offers several advantages, including greater stability, more controlled biodistribution, and the potential for tissue-specific targeting due to their molecular signature. Furthermore, they can induce regenerative and adaptive responses in recipient cells by activating pathways involved in mitochondrial biogenesis and detoxification [17].

However, despite these advances, several challenges must be addressed before widespread clinical applications can be considered. Standardization of mitochondrial and EV isolation and characterization methods, understanding the mechanisms underlying their uptake and integration, and evaluating their immunogenicity and long-term safety remain critical areas for further investigation. Moreover, regulatory and ethical considerations must be considered to ensure the clinical translation of these emerging approaches.

The objective of this review is to report the therapeutic effects of mitochondrial and EV transfer in different disease models in preclinical studies and identify the related mechanisms (bioenergetics, redox signaling, organelle dynamics).

2. Materials & methods

Medline research was conducted via PubMed using the following search terms for mitochondria: [((Mitochondria OR mitochondrial) AND (Therapeutic OR supplementation OR transplantation OR enrichment OR treatment) AND (preclinical)) OR (mitotherapy) OR (mitoception) NOT biomarker]. For EV- the search term is [(“extracellular vesicles” OR microvesicles OR “extracellular vesicle” OR microvesicle) AND (mitochondria OR mitochondrial)) AND (treatment OR therapeutic OR transplantation OR supplementation)]. The search was conducted on October 10, 2024.

Only peer-reviewed papers in English investigating the effects in preclinical models were included. Studies examining the effects on cells, ex vivo organs, oocytes, and humans were excluded. For each study included in the analysis, the following parameters were recorded: first author, country of the first author, year of publication, animal model used, type of disease, source of mitochondria (cells or organs, and species), quality check of mitochondria after isolation (e.g., imaging, mitochondrial protein enrichment …), type of transplantation (xenogeneic, allogenic, autologous), the transfer parameters (dose of mitochondria, route of administration, site of administration), proof of mitochondrial transfer into host cells/organs (yes/no), and the study’s conclusions. For EV-mitochondria analysis, we also exclude paper where the EV isolation and characterization didn’t follow the MISEV statement [14]. A network visualization of author collaborations to identify the degree of independence of studies was also conducted. To identify the author networks, a Venn diagram was created using VOSviewer 1.6.20, and all authors appearing in at least two articles were included.

To evaluate potential bias in the studies included, we applied the SYRCLE risk of bias tool [24], specifically designed for animal research and adapted from the Cochrane Collaboration’s framework for assessing bias in randomized controlled trials. The assessment involved 10 criteria addressing six categories of bias: selection, performance, detection, attrition, reporting, and other potential sources of bias (incomplete report of the transfer parameters).

3. Results

3.1. Search results

Following a Medline search, 123 studies met the research criteria, with the first evidence of mitochondrial transplantation in animals published in 2013 and EV in 2017 (Fig. 1). The predominant experimental model used were rodents (52 %, n = 64 in mice and 41 %, n = 50 in rats), along with larger animals such as pigs (6 %, n = 7). The mitochondria were primarily sourced from organs (66 %, n = 69), whereas the EV were sourced exclusively from cells, leading predominantly to allogeneic (64 %, n = 93) or xenogeneic (29 %, n = 42) transplantations. Notably, 13 % (n = 13) of mitochondrial studies did not provide quality control data following mitochondrial isolation. Most studies used local administration (60 %, n = 68), although 13 % (n = 14) did not specify the method of delivery. Additionally, 22 % (n = 27) of studies did not report whether mitochondrial internalization into tissues or cells was confirmed (Fig. 2). Geographically, research on mitochondrial transplantation was predominantly led by teams in China (n = 50), followed by the USA (n = 23), then Iran (n = 11) and Taiwan (n = 10) (to note, 79 % of EV studies come from China, Fig. 2). This distribution was reflected in two key author clusters for mitochondria: one around Fu, Zang, and Su; and the other around McCully and Guarien (Fig. S1) with a high interconnection between the teams involved in the field; and one cluster for EV: one around Cai, Yang and Chen (Fig. S2) with poor interconnection between the teams involved in the field. Finally, the risk of bias of the studies was reported as supplementary data (Tables S1–7). It is worth noting that the methodological descriptions often lacked key details related to potential bias, including sample size calculations, random housing of animals, allocation concealment, and information on how incomplete data were handled during the study.

Fig. 1.

Prisma flow diagram.

Fig. 2.

General parameters of mitochondrial and EV transplantation-related studies

A. Animal models use for in vivo studies. B. Mitochondrial and EV sources. C. Type of transplantation. D. Administration root of mitochondria and EV. E. Proof of mitochondria and EV isolation. F. Proof of mitochondria and EV transfer. G. Geographical repartition of teams involved in mitochondrial transplantation.

3.2. Efficacy and safety of mitochondrial and EV transplantation on disease pathophysiology

3.2.1. Ischemia-reperfusion injury (IRI)

A total of 30 studies [25–54] explored the effects of mitochondrial transplantation on IRI (Table 1). Of these, 19 studies reported data on infarct size, with 84 % (16/19) demonstrating a reduction in infarct size (ranging from 12.6 % to 34.1 %) following transplantation. Among the three studies reporting negative outcomes, two used the lowest dose of mitochondria (10⁴). The third study was not designed to assess the effect of mitochondrial transplantation alone, but rather to demonstrate the added benefit of combining mitochondria with CoQ10—a benefit that was successfully shown. Furthermore, the reduction in infarct size was correlated with improvements in organ function (e.g., cardiac, brain, muscle) and behavior. The reported clinical improvement also aligned with biological observations: decreased inflammatory markers (mainly TNFα, IL1β and IL6) and decreased apoptosis.

Table 1.

Mitochondrial effect on ischemia-reperfusion injury.

Author/Year	Model/Disease	Mitochondria source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Bafadam 2024 [25]	Wistar rat IRI cardiac	Pectoralis major Rat Allogenic	Dose: 6*106 mito Administration: intraventricular Site: cardiac	No effect on infarct size No effect on cardiac pressure and function	↓ MMP (JC-1) ↓ cellular ROS (DCFHDA)
Boutonnet 2024 [26]	SWISS mice IRI limb	Rectus muscle Mice Allogenic	Dose: 106 mito Administration: IM Site: gastrocnemius	No clinical data reported	↑ mito quantity (VDAC) ↑ Ca2+ retention capacity ↑ anti-ox (GPX) ↓ fusion (Mfn2)
Salman 2024 [27]	C57BL/6J mice IRI stroke	Liver Mice Allogenic	Dose: 100 μg mito proteins Administration: intranasal	↓ infarct size (≈33 % reduction vs. vehicle) ↑ motor coordination and cerebral oedema ↓ microglia and astrocyte activationα↑ inflammation (IL1β, NLRP3, NFκB, IgG)	↑ mito biogenesis (SIRT1, PGC1α) No effect on autophagy (pAMPKα-Thr172) ↑ OXPHOS proteins (NDUF8, SDHB, UQCRC2, MITCO1) ↓ oxidative stress (NO, HNE) ↑ anti-ox (eNOS, SULT4A1)
Sun 2024 [28]	C57BL/6J mice IRI stroke	Liver Mice Allogenic	Dose: 107 mito Administration: IV Site: tail	↓ infarct size (≈33 % reduction vs. vehicle) and mortality ↑ neurogenesis and inhibit pyroptosis ↓ anxious behavior, ↑ memory	↑ ATP level ↓ cellular ROS (DCFHDA) ↑ fusion (Mfn2)
Wu 2024 [29]	Sprague Dawley rat IRI cardiac	MSC cells Human Xenogeneic	Dose: 5*109 mito Administration: oral	↓ infarct size (≈5 % vs. 40 %) ↑ ejection fraction and fractional shortening ↓ apoptosis, ↓ gene neutrophil chemotaxis	↑ ATP level ↓ mito swelling, disruption, cristae loss
Xu 2024 [30]	Sprague Dawley rat IRI global	Gastrocnemius Rat Allogenic	Dose: 5*108 mito Administration: IV Site: femoral	↑ spatial memory ↓ cerebral oedema and S100β/NSE (cerebral injury marker) ↓ apoptosis	↑ mito quantity (mitotracker green), ↓ oxidative stress marker (MDA) ↑ anti-ox (SOD) ↑ mitophagy (PINK1, Parkin) ↓ PTP opening
Liang 2023 [31]	C57B/6J mice IRI cardiac	BM-MSC cells Human Xenogeneic	Dose: 7.5*105 mito Administration: IM Site: cardiac	↑ ejection fraction and fractional shortening ↓ fibrosis and apoptosis ↑ number of mature vessels	↑ ATP level ↑ mito quantity (mitotracker green)
Mokhtari 2023 [32]	Wistar rat IRI cardiac	Pectoralis major Rat Allogenic	Dose: 2*105 mito Administration: intramyocardial Sites: 3	↑ ejection fraction and fractional shortening ↓ troponin I ↓ inflammation (TNFα, IL1β, IL6)	↓LC3-II/LC3-I ratio
Norat 2023 [33]	C57BL/6J mice IRI stroke	Gastrocnemius Mice Allogenic	Dose: 200 μL mito preparation Administration: intra-arterial Site: cerebral	↓ infarct size (≈34 % reduction vs. vehicle) ↑ cellular viability & ↓ apoptosis	↑ ATP level
Sun 2023 [34]	C57BL/6J mice IRI cardiac	Heart Mice Allogenic	Dose: 7.5–10*104 mito Administration: IV Site: tail	No effect on infarct size No effect on cardiac function (ejection/fractional shortening) No effect on apoptosis	Not investigated
Zeng 2023 [35]	C57BL/6J mice IRI limb	MSC cells Human Xenogeneic	Doses: 107 mito Administration: IM Site: gastrocnemius	↓ fibrosis ↓ apoptosis ↓ neurophil infiltrate	↑ ATP level
Chen 2022 [36]	C57BL/6 mice IRI stroke	Liver Mice Allogenic	Dose: 2*107 mito Administration: injection Site: ischemic site	↑ locomotion ↑ myelinization (FABP5/7) ↓ apoptosis	↑ ATP level ↑ complex activity (I, II, III) No effect on complex activity (IV, V)
Doulamis 2022 [37]	Zucker diabetic fatty rat IRI cardiac	Pectoralis major Rat (diabetes or not) Allogenic	Dose: 106 mito Administration: intraventricular Sites: 6, cardiac	↓ infarct size (7.0–5.3 % vs. 33.6 %) ↑ fractional shortening (more important if non-diabetes mitochondria) ↓ inflammation (IL1α/β, IL4, IL6, IL13, TNFα, IFNγ, CXCL1/2/10, CX3CL-1)	↑ ATP level (more important if non-diabetes mitochondria)
Kubat 2021 [38]	Sprague Dawley rat IRI renal	MSC cells Rat Allogenic	Dose: 4*106 mito Administration: injection Site: renal cortex	↓ apoptosis ↑ cellular proliferation No effect on tissue histology, creatinine and urea level	↓ fission (DRP1)
Sun 2021 [39]	C56BL/6 mice IRI cardiac	Cardiomyocyte cells/>2021Mice Allogenic	Dose: 5*104 mito Administration: intraventricular Site: cardiac	No effect on infarct size No effect on cardiac function (ejection/ fractional shortening) No effect on apoptosis	↑ fusion (OPA1, Mfn1/2) ↑ mito biogenesis (TFAM)
Xie 2021 [40]	Sprague Dawley rat IRI stroke	N2a cells Mice Xenogeneic	Dose: 107 mito Administration: intra-arterial Site: carotid	↓ infarct size (13.4 % vs. 26.0 %) ↑ neurological behavior	↑ fusion (Mfn1, OPA1)
Blitzer 2020 [41]	Yorkshire pig IRI cardiac	Pectoralis major Pig Autologous	Dose: 109 mito Administration: intra-arterial Site: coronary	↓ infarct size (7.4 % vs. 38.0 %) ↑ left ventricular pressure, ejection fraction	Not investigated
Doulamis 2020 [42]	Yorkshire pig IRI renal	Sternocleidomastoid Pig Autologous	Dose: 109 mito Administration: intra-arterial Site: kidney	↓ injury in the cortex, no effect on medulla ↓ creatinine and urea level No effect on ionogram ↓ inflammation (IL6)	Not investigated
Guariento 2020 [43]	Yorkshire pig IRI cardiac	Pectoralis major Pig Autologous	Dose: 109 mito Administration: intra-arterial Site: coronary	↓ infarct size (3.8–4.2 % vs. 37.9 %) ↑ ejection fraction, left anterior descending artery flow	Not investigated
Jabbari 2020 [44]	Wistar rat IRI renal	Pectoralis major Rat Allogenic	Dose: 3*106 mito Administration: intra-arterial Site: kidney	↓ apoptosis ↓ tubular necrosis ↓ creatinine and urea level	Not investigated
Moskowitzova 2020 [45]	C57BL/6J mice IRI lung	Gastrocnemius Mice Allogenic	Doses: 108 mito injection in left pulmonary artery or 3*108 mito if ultrasonic nebulization	Same effect regarding doses/routes of administration ↓ histological tissue injury ↑ inspiration capacity, dynamic ↓ neutrophil infiltrate, apoptosis No effect on cytokines levels	Not investigated
Nakamura 2020 [46]	C57BL/6 mice IRI stroke	Placenta Mice Allogenic	Dose: 100 μg mito proteins Administration: IV Site: not specified	↓ infarct size (≈20 % reduction vs. vehicle)	Not investigated
Orfany 2020 [47]	C57BL/6J mice IRI limb	Muscles Mice Allogenic	Dose: 106–9/g of muscle Administration: IM Sites: 4 muscles in the limb	↓ infarct size (69 % vs. 87 %, dose effect) ↑ percentage shared stance time ↓ stance factor ↓ apoptosis	↓ ATP level (dose: 109)
Pourmohammadi-Bejarpasi 2020 [48]	Wistar rat IRI stroke	UC-MSC cells Human Xenogeneic	Dose: from 3*107 cells Administration: intraventricular Site: cerebral	↓ infarct size (12 % vs. 31 %) Restore motor and balance neurobehavioral function Neuroprotective effect on cyto-architecture ↓ apoptosis, CPK, microglial activity, astrogliosis	Not investigated
Shin 2019 [49]	Yorkshire pig IRI cardiac	Pectoralis major Pig Autologous	Dose: 109 mito Administration: intra-arterial Site: coronary	↓ infarct size (7.3 % vs. 38.6 %) Restore contractile measures ↑ coronary blood flow	↑ ATP level
Zhang 2019 [50]	Sprague Dawley rat IRI stroke	Pectoralis major Rat Autologous	Dose: 5*106 mito Administration: intraventricular Site: cerebral	↓ infarct size (≈23 % vs. 35 %) ↓ oedema ↓ apoptosis ↑ neurogenesis	↑ ATP level ↑ mito quantity (COX4) ↓ oxidative stress (MDA, 8-OHDG, NT) ↑ anti-ox (SOD, GSH)
Kaza 2017 [51]	Yorkshire pig IRI cardiac	Pectoralis major Pig Autologous	Dose: 9.9*106 mito Administration: sub endocardial Site: area at risk	↓ infarct size (5.2 % vs. 13.0 %) ↓ creatine kinase and troponin level No effect on electrocardiogram parameters No effect on cytokines levels	Not investigated
Huang 2016 [52]	Sprague Dawley rat IRI stroke	BHK-21 cells Hamster Xenogeneic	Doses: 75 μg intracerebral OR 750 μg in femoral artery expression in mito proteins	Better recovery if injection intracerebral than femoral artery ↓ infarct size (7.1 % cerebral vs. 12.6 arterial vs. 20.7 %) ↓ apoptosis	↑ ALAMAR blue signal ↑ MTT assay signal
Lin 2013 [53]	Wistar rat IRI liver	Liver Rat Allogenic	Dose: 7.7*105 mito Administration: injection Site: spleen	↓ histological tissue injuries ↓ apoptosis ↓ ALAT	↓ oxidative stress (HNE)
Masuzawa 2013 [54]	New Zeeland White rabbit IRI cardiac	Pectoralis major Rabbit Autologous	Dose: 1.2*106 mito Administration: intracardiac Sites: 8	↓ infarct size (9.8 % vs. 36.6 %) ↓ apoptosis ↓ creatine kinase, troponin I No arrythmia or immune reaction after injection ↓ inflammation (TNFa, IL6, CRP)	↑ cellular respiration

8-OHDG: 8-hydroxyl-2′-deoxyguanosine, ALAT: alanine aminotransferase, anti-ox: anti-oxidant; ATP: adenosine tri-phosphate, BHK21: baby hamster kidney cells-clone 21, BM-MSC: bone-marrow MSC, COX: cytochrome C oxidase, CPK: creatine phosphokinase, CRP: C reactive protein, CXCL: Chemokine (CXC motif) ligand, DCFHDA: 2′,7′-dichlorodihydrofluorescein diacetate, DRP1: dynamin-related protein 1, eNOS: endogenous endothelial nitric oxide synthase, FABP: fatty acid-binding protein, GPX: glutathione peroxidase, GSH: reduced glutathione, HNE: 4-hydroxy-2-nonenal, IFN: inter-feron, IgG: immunoglobulin G, IL: interleukin, IM: intramuscular, IRI: ischemia-reperfusion injury, IV: intra-veinous, JC1: 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide, LC3: microtubule-associated protein light chain 3, MDA: malondialdehyde, Mfn: mitofusin, MITCO1: protein complex IV, mito: mitochondria, MMP: Mitochondrial membrane potential, MSC: mesenchymal stromal cell, MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromure, N2A: Neuro-2A, NDUF8: protein complex I, NFκB: nuclear factor-kappa B, NLRP3: NOD-, LRR- and pyrin domain-containing protein 3, NO: nitric oxide, NSE: neuron specific enolase, NT: nitrotyrosine, OPA1: opticatrophy 1, OXPHOS: oxidative phosphorylation, PAMPK-α-Thr172: Phospho-AMPK-alpha-Threonine 172, PGC1α: Pparg coactivator 1 alpha, PINK1: PTEN induced kinase 1, PTP: permeability transition pore, ROS: relative oxygen species, S1000β: S100-beta protein, SDHB: protein complex II, SIRT1: sirtuin 1, SULT4A1: sulfotransferase 4A1, SOD: superoxide dismutase, TFAM: mitochondrial transcription factor A, TNFα: tumor necrosis factor α, UC-MSC: Umbilical-cord MSC, UQCRC2: protein complex III, VDAC: voltage-dependent anion channel, ↑: increase/improve, ↓: decrease, ≈: data have been estimated from the graphs.

A total of six studies [55–60] explored the effects of EV on IRI (Table 7). They all reported improvement of organ function (brain, heart, liver) but only one reported significant reduction of infarct size (54 %).

Table 7.

EV treatment paper.

Author/Year	Model/Disease	EV source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Gai 2024 [55]	C57BL/6J Mice IRI Stroke	MSC cells Mice Allogenic	Dose: 100 μg of EV proteins Administration: intra-nasal	↓ hypoxia-ischemia-induced injury ↓ neuronal damage ↓ infarct size (≈40 % reduction HI-EV vs. HI, ≈20 % reduction H2S-EV vs HI-EV)	↑ anti-ox (Nrf2, PARK7, PRDX1, PRDX2) ↑ mito quantity (mtDNA) ↑ fusion (mfn1 and mfn2 mRNA)
Wu 2023 [56]	Balb/c Mice IRI stroke	Hypoxic neurons cells Mice Allogenic	Dose: 108 EV Administration: stereotactic microinjection Sites: cortex or nasal drops	↓ cerebral infarction volume (≈45 % vs. vehicle) ↑ neuronal neurite survival ↑ neuronal integrity ↑ neuroprotection	↓ mitochondria-associated apoptosis
Ikeda 2021 [57]	C57BL/6 Mice IRI cardiac	Induced cardiomyocyte cells Human Xenogeneic	Dose: 108 EV Administration: Injection Site: peri-infarct region	↑ ejection fraction (≈20 % vs. vehicle) ↑ restoration of intracellular bioenergetics and contractile property ↑ post-MI cardiac function improvement	↑ mito biogenesis (PGC-1α)
Lu 2021 [58]	C57BL/6 Mice IRI liver	UC-MSC cells Human Xenogeneic	Dose: 109 EV Administration: IV Site: tail	↓ liver IRI ↑ hepato-protection & ↓ liver damage ↓ NETs formation	↑ fusion (Mfn1) ↓ oxidative stress (mtROS)
Zhao 2021 [59]	C57BL/6 Mouse IRI renal	MSC cells Human and mouse Allogenic/Xenogeneic	Dose: 2*107 EV Administration: IV Site: tail	↓ kidney inflammation ↓ renal lesion formation	↑ OXPHOS ↓ mtDNA and mitochondrial damage ↑ mito biogenesis (restoring TFAM protein and stabilizing the TFAM-mtDNA complex)
Yao 2019 [60]	Sprague-Dawley rats IRI liver	UC-MSC cells Human Xenogeneic	Dose: 10 mg/kg EV Administration: IV Site: tail	↑ hepatoprotection ↓ liver damage ↓ mRNA levels of IL-1β, IL-6, TNF-α, C-C motif ligand 12, IFN-κ, and TLR4	↑anti-ox (MnSOD) ↓ respiratory burst ↓ ROS (CellRox Deep Red, mitosox)
Bao 2022 [108]	C57BL/6J Mice Sepsis associated coagulopathy	Primary neutrophils cells Mice Allogenic	Dose: 2*105 EV Administration: IV Site: tail	↓ mortality 40 % with EV treatment vs ↓ mortality 20 % with vehicle treatment at 70 h ↑ neutrophil-mediated prevention of DIC ↑ antithrombotic function of neutrophils ↓ endothelial dysfunction ↓ DIC severity	↑anti-ox (SOD2) mediated endothelial protection ↓ endothelial ROS accumulation
Zheng 2021 [109]	Sprague-Dawley Rat Sepsis	MSC cells Rat Allogeneic	Dose: 2*107 EV Administration: IV Site: tail	↓ mortality (from 56 % to 25 %) ↓ intestinal barrier dysfunction	↑ fusion (mfn2) ↑ mito biogenesis (PGC-1α)
Yao 2024 [117]	C57BL/6J Mice Wound healing	ADSCs cells Human Xenogeneic	Dose: 3*108 EV Administration: hydrogel MNP Site: skin	↓ wound healing time (≈5 day less vs. vehicle) ↑ macrophage polarization (M2 subtype)	↑ MMP ↑ ATP levels ↓ ROS (Cell Rox probe)
Zhuang 2024 [118]	C57BL/6 Mice Aging	Metformin treated MSC cells Human Xenogeneic	Dose: 1010 EV Administration: hydrogel wound beds Site: skin	↑ aged skin repair ↓ cellular senescence	↑ ATP production ↑ OXPHOS ↑ mitophagy (LC3 II)
Chen 2024 [119]	C57BL/6 Mice Aging	Plasma Mice Allogenic	Dose 36 μg of EV proteins Administration: IV Site: not specified	↓ age-associated functional decline ↓ senescence ↑ tissue regeneration ↑ lifespan	↑ mito biogenesis (PGC-1α)
Liang 2024 [140]	Sprague-Dawley Mice Erectile disfunction	PC-12 cells Rat Xenogeneic	Dose: 108 MVs Administration: IC Site: Corpus cavernosum	↑ erectile function ↓ ferroptosis ↓ apoptosis	↓ oxidative stress marker (MDA)
Shen 2024 [141]	C57BL/6 Mice Autoimmune Hepatitis	MSC cells Human Xenogeneic	Dose: 3*1010 MVs Administration: IP	↓ liver injury & ↑ liver protection ↑ glycolysis inhibition ↓ CD4+ T-cell activation ↓ mRNA IFN-γ, TNF-α, and IL-2	↓ OXPHOS
Tolomeo 2024 [142]	Pig Graft quality transplant for IRI	MSC cells Pig Allogenic	Dose: 1011 MVs Administration: into the solution of heart perfusion machine	↑ myocardial viability ↓ apoptosis ↓ IL-1ra, IL-2 and IL-6	↓ anti-ox (SOD, CAT, GPX) ↓ oxidative stress (carbonylated proteins) ↓ mitochondrial cristae loss ↓ mitochondrial swelling
Cao 2022 [143]	BALB/c Mice Solid tumors	Dendritic cells Mice Allogenic	Dose: 80 μg EV proteins Administration: IV Site: tail	↓ tumor size (≈80 % reduction vs. vehicle) ↑ T cell activation ↑ immunogenic cell death in tumor cells ↑ immune responses against primary, distant tumors and metastases tumors ↑ efficient eradication of primary, distant and metastases tumors ↓ cancer stem cells	Not investigated
Lu 2022 [144]	C57BL/6 Mice Cardiac hypertrophy	Hypoxic MSC cells Human Xenogeneic	Dose: 109 EV Administration: IV Site: tail	↑ hepatoprotective effects ↓ NETs formation (suppressed by hUC-MSC-EV) ↓ liver IRI severity	↓ mitochondrial ROS (mitosox) ↑ mitochondrial fusion in neutrophils (Mfn2) ↓ cytokine release from neutrophils (IFN-γ, IL-6 and TNF-α)
Li 2022 [145]	C57BL/6 Mice Liver Fibrosis	Plasma Mice Allogenic	Dose: 4*108 EV Administration: IV Site: tail	↓ liver fibrosis progression ↑ collagen synthesis and degradation regulation ↓ HSC activation	↑ anti-ox (GSH levels) ↓ ROS (CMXROS)
Dutra Silva 2021 [146]	C57BL/6 Mice ARDS	BM-MSC cells Human Xenogeneic	Dose: 98.3 μg of EV proteins Administration: IV Site: tail	↑ barrier integrity ↑ barrier integrity with ARDS plasma ↓ LPS-induced inflammation	↑ in vivo mitochondrial respiration (OCR)
Morrison 2017 [147]	C57BL/6 Mouse ARDS	MDMs cells Human Xenogeneic	Dose: not specified Administration: intra-nasal	↑ expression of the M2 phenotype marker CD206 ↑ phagocytic capacity of macrophages ↑ protection in lung injury ↓ production of cytokines (TNF-α and IL-8)	↓ OXPHOS

ADSCS: adipose derived stem cells, anti-ox: anti-oxidant, ARDS: acute respiratory distress syndrome, BM-MSC: bone marrow MSC, CAT: catalase, DIC: disseminated intravascular coagulation, EV: extracellular vesicles, GPX: glutathione peroxidase, GSH: reduced glutathione, HI: hypoxia-ischemia, HypEV: EV release by cells under hypoxia, IC: intracavernous, IFN: interferon, IL: interleukin, IP: intraperitoneal, IRI: ischemia-reperfusion injury, IV: intravenous, LC3: microtubule-associated protein light chain 3, MDA: malondialdehyde, MDMs: monocyte derived macrophages, Mfn: mitofusin, mito: mithcondria, M2 phenotype: alternatively activated macrophages, MMP: mitochondrial membrane potential, MNP: microneedle patch, MSC: mesenchymal stromal cells, MV: microvesicles, NRF2: nuclear factor erythroid 2-related factor 2, OCR: oxygen consumption rate, OXPHOS: oxidative phosphorylation, PGC1-α: Pparg coactivator 1-alpha, PRDX: peroxiredoxin, ROS: reactive oxygen species, SOD: superoxide dismutase, TFAM: mitochondrial transcription factor A, TNFα: tumor necrosis factor alpha, UC-MSC: umbilical cord MSC, ↑: increase/ improve, ↓: decrease, ≈: data have been estimated from the graphs.

3.2.2. Neurological disorders

Mitochondrial transplantation was examined in 29 studies [61–89] on various neurological disorders, including spinal cord injury (n = 5), Parkinson’s disease (n = 4), stress/depression (n = 4), aging (n = 3), traumatic brain injury (n = 2), and other conditions (n = 11, Table 2). Notably, most of studies (22/23) reported positive clinical outcomes (motor and cognitive) aligning with biological observation: improvement of neurogenesis, inflammation decrease, regulation of microglial and astrocyte proliferation, and apoptosis decrease. However, six studies focused exclusively on biological effects (e.g., apoptosis, cell differentiation & axonal flow) without evaluating clinical outcomes, and one study found no clinical benefit in spinal cord injury, which was potentially linked to the source of mitochondria (PC12 cells and soleus muscle). No EV studies were found in the search.

Table 2.

Mitochondrial transplantation effects on neurologic disorders.

Author/Year	Model/Disease	Mitochondria source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Eo 2024 [61]	C57BL/6 mice Parkinson’s disease	UC-MSC cells Human Xenogeneic	Doses: 0.5–10 μg mito proteins Administration: IV Site: not specified	↑ behavioral (dose effect) ↓ neurological damage (only 10 μg tested) ↓ neuro-inflammation (TNFα, IL6) (only 10 μg tested)	Not investigated
Jain 2024 [62]	C57BL/6 mice Parkinson’s disease	Liver Mice Allogenic	Dose: 0.5 mg/kg Administration: IV Site: not specified	↑ locomotor activity	↑ OXPHOS prot expression (NDUFB8, SDHB, UQCRC2, MTCO1, vATP5α) ↑ mito biogenesis (PGC1a, TFAM)
Xu 2024 [63]	C57BL/6 mice Spinal cord injury	Macrophages cells Mice Allogenic	Dose: from 107 cells, engineered mitochondria Administration: IV Site: tail	↑ functional recovery ↑ motor coordination ↑ phagocytosis of myelin debris ↑ tissue repair ↓ apoptosis	↑ ATP production ↑ mito quantity (mitotracker) ↑ basal & max intensity ↓ cellular ROS (DCFHDA)
Bamshad 2023 [64]	Wistar rat Traumatic brain injury	UC-MSC cells Human Xenogeneic	Dose: 3*107 mito Administration: intraventricular Site: cerebral	↑ sensitivity recovery ↓ brain tissue destruction ↑ neurogenesis ↓ apoptosis, CPK, microglial activation, astrogliosis	Not investigated
Ene 2023 [65]	Wistar rat Schizophrenia	Brain Rat Allogenic	Dose: 100 μg mito proteins Administration: intracerebral Site: prefrontal cortex	↓ schizophrenia after development in sick rats, neurological impairment in healthy rats No effect on inflammation (IL1 p)	↑ COX activity ↓ Cellular ROS (DCFHDA)
Hayashida 2023 [66]	Sprague Dawley rat Cardiac arrest neurotoxicity	Brain Rat Allogenic	Dose: 500 μL of mito preparation (fresh or fractioned) Administration: IV Site: not specified	Results obtained only for fresh mito ↑ survival at day 3 (90.9 % vs. 54.5 %) ↑ neurological function and microperfusion cerebral ↓ cerebral oedema and lactate	Brain: ↓ fusion (Mfn1/2), no effect on OPA1 No effect on fission (DRP1, Fis1) Spleen: ↑ fission (DRP1), no effect on Fis1 No effect on fusion (Mfn1/2, OPA1)
Huang 2023 [67]	Sprague Dawley rat Neuropathic pain	Soleus, pectoralis major, biceps, gastrocnemius, abdominal muscles Rat Allogenic	Dose: 100 μg mito proteins Administration: intracerebral Site: ganglion dorsal root	↓ pain hyper-sensibility (mechanical, thermal) ↓ activation of glial cells ↓ neuro-inflammation (TNFα, IL1β, IL6, NFκB) ↓ apoptosis	Not investigated
Javani 2023 [68]	Wistar rat Chronic stress	Brain Rat Allogenic	Dose: 100 μg mito proteins Administration: intraventricular Site: cerebral	↓ apoptosis	↑ ATP level ↑ MMP (JC-1)
Javani 2023 [69]	Wistar rat Chronic stress	Brain Rat Allogenic	Dose: 100 μg mito proteins Administration: intraventricular Site: cerebral	↓ stress-induced spatial learning and memory losses ↑ neuronal dendritic morphology	↓ oxidative stress (MDA)
Jia 2023 [70]	C57BL/6J mice Epilepsia	Hippocampus Mice Allogenic	Dose: 1 mg/kg Administration: IV Site: tail	Improve cognitive impairment, epileptic depression and anxiety state ↓ microglial and astrocytes proliferation ↑ neurodegeneration	Not investigated
Yang 2023 [71]	C57BL/6 mice Alzheimer’s disease	Brain Mice Allogenic	Dose: 3*106 mito Administration IV Site: caudal	↑ cognitive ability	↑ ATP/ADP ratio ↓ oxidative stress (MDA) ↑ anti-ox (GSH, MAPK and FOXO) ↑ complex activity (I, IV) ↑ autophagy (FOXO3, BNIP3, LC3II/LC3I) ↑ mito biogenesis (ERK, SIRT1)
Zhu 2023 [72]	Sprague Dawley rat Spinal cord injury	Platelet cells Rat Allogenic	Dose: 3*105 mito Administration: intracerebral	No clinical data reported ↑ neuron quantity and myelin ↓ apoptosis	↑ ATP level ↑ proportion of elongated mito ↑ complex activity (I, II, III, IV, V) ↑ anti-ox (Nrf2, HO-1, NQO1, SOD) ↓ oxidative stress markers (Nox2, MDA, NT, 8-OHDG)
Adlimoghaddam 2022 [73]	C57BL/6 mice Aging and hippocampal function	Liver Mice Allogenic	Dose: 10–20 mg/kg Administration: IV Site: not specified	No dose effect No clinical data reported	↑ OXPHOS (CII-SDHB)
Hosseini 2022 [74]	C57BL/6 mice Cognitive impairment post stroke	BM-MSC cells Mice Allogenic	Doses: 85/170/340 μg mito proteins Administration: intranasal	↑ social interaction (only 340 μg) ↓ cognitive impairment ↓ loss of synaptic protein	↑ ATP level ↑ MMP (JC-1) ↓ cellular ROS (DCFHDA)
Javani 2022 [75]	Wistar rat Depression	Brain Rat Allogenic	Dose: 100 μg mito proteins Administration: intraventricular Site: cerebral	↓ anxiety ↓ depression-like behavior	↑ ATP level ↑ MMP (JC-1)
Lin 2022 [76]	Sprague Dawley rat Spinal cord injury	Soleus Rat Allogenic	Dose: 50 μg mito proteins Administration: injection Site: spinal cord	↑ sensitivity and motor function ↓ demyelinization ↓ inflammation (TNFα, IL6)	↓ fission (DRP1) ↓ oxidative stress markers ↓NO & 3NT
Zhang 2022 [77]	TOPGAL mice Age-related cognitive decline	Liver Mice Allogenic	Dose: 1.5–2*106 mito Administration: injection Site: hippocamp	↑ cognitive function	↑ mito quantity (TOM20) ↑OXPHOS (complexes protein activity (I, II, IV))
Fang 2021 [78]	Sprague Dawley rat Spinal cord injury	Soleus Rat Allogenic	Dose: 100 μg mito proteins Administration: IV Site: jugular	↑ locomotor function and recovery ↓ apoptosis ↓ neuro-inflammation (TNFα, IL6)	Not investigated
Sheu 2021 [79]	Sprague Dawley rat Sciatic crush injury	BHK-21 cells Hamster Xenogeneic	Dose: 195 μg mito proteins Administration: IM Site: gastrocnemius	↑ nerve regeneration ↑ speed axonal flow Restore nerve and muscle morphology	Not investigated
Zhao 2021 [80]	C57BL mice Traumatic brain injury	Liver Mice Allogenic	Dose: 1.2–1.4*106 mito Administration: intracerebral	↓ anxiety ↑ cognitive function ↑ spatial memory ↓ apoptosis	Not investigated
Ma 2020 [81]	db/db mice DT2 cognitive impairment	Platelet cells Rat Xenogeneic	Dose: 105 mito Administration: intraventricular Site: cerebral	↑ cognitive function ↓ apoptosis ↓ β amyloid deposit No effect on glucose level and weight	↑ ATP level ↑ mito quantity (mitotracker) ↓ damage mito ↑ CI, III, IV, V activity, no effect on CII ↓ oxidative stress (H2O2, MDA, NT, 8-OHDG) ↑ anti-ox (total antioxidant capacity expressed in Trolox equivalent)
Nascimento-Dos-Santos 2020 [82]	Lister Hooded rat Optic nerve injury	Liver Rat Allogenic	Dose: 1.25 μg mito proteins (fresh or fractioned) Administration: injection Site: intravitreal	Results obtained for both (fresh, fractioned mito) Promote neuroprotection ↑ number of axons ↑ retinal electrophysiological	Not investigated
Zhao 2020 [83]	BALB/c mice Cognitive motor impairment in aging	Liver Mice Allogenic	Dose: 5 mg/kg Administration: IV Site: tail	↑ age-related behaviors ↑ motor performance ↑ macrophages phagocytosis	↑ ATP level ↓ oxidative stress (MDA) ↑ anti-ox (GSH) ↑ pyruvate DH, α-ketoglutarate DH, NADH
Nitzan 2019 [84]	C57BL/6 mice Alzheimer’s disease	HeLa cells Human Xenogeneic	Dose: 200 μg mito proteins Administration: IV Site: tail	↑ cognitive function ↑ neuronal disarrangement	↑ citrate synthase activity ↑ COX activity
Wang 2019 [85]	ICR mice Depression	Hippocampus Mice Allogenic	Dose: 1 mg/kg Administration: IV Site: not specified	↑ depression-like behavior ↑ neurogenesis ↓ neuro-inflammation (IL1β, TNFα, COX-2)	↑ ATP level J oxidative stress (DCFHDA, MDA) ↑ anti-ox (SOD)
Gollihue 2018 [86]	Sprague Dawley rat Spinal cord injury	PC12Adh/soleus Rat Allogenic	Dose: 100 μg mito proteins Administration: injection Site: spinal cord	No comparison between cell sources No effect on locomotor and sensory threshold ↓ number of neurons	No effect on RCR
Robicsek 2018 [87]	Wistar rat Schizophrenia	EBV transform lymphocyte/ brain Human/Rat Xenogeneic/Allogenic	Dose: 100 μg mito proteins Administration: injection Sites: 3, pre-frontal cortex	No comparison between cell sources ↑ neuronal differentiation Restore neuronal function Restore mitochondrial membrane potential	↑ MMP (JC-1)
Shi 2017 [88]	C57BL/6J mice Parkinson’s disease	HepG2 cells Human Xenogeneic	Dose: 0.5 mg/kg Administration: IV Site: tail	↑ motor activity	↑ ATP level ↑ OXPHOS (complex I activity) ↓ oxidative stress (ROS) ↑ anti-ox (GSH)
Chang 2016 [89]	Sprague Dawley rat Parkinson’s disease	PC12 cells Rat Allogenic	Dose: 1.05 μg mito proteins Administration: injection Site: medial forebrain bundle	↑ mobility ↓ dopaminergic neurons deterioration	↓ OXPHOS (CI, CIII and IV activity activity) ↑ fusion (Mfn2) ↑ fission (Drp1) ↑ mitophagy (Parkin) ↓ oxidative stress (8-OHDG)

8-OHDG: 8-hydroxyl-2′-deoxyguanosine, anti-ox: anti-oxidant; ATP: adenosine tri-phosphate, BM-MSC: bone-marrow mesenchymal stromal cell, BNIP3: Bcl-2/ adenovirus E1B-interacting protein 3, CII-SDHB: complex II - succinate dehydrogenase subunit B, COX: cytochrome C oxidase, CPK: creatine phosphokinase, DCFHDA: 2′,7′-dichlorodihydrofluorescein diacetate, DRP1: dynamin-related protein 1, EBV: Epstein-Barr virus, ERK: extracellular signal-regulated kinase, Fis1: mitochondrial fission protein 1, FOXO: forkhead box O, GSH: reduced glutathione, HO-1: heme oxygenase-1, IL: interleukin, IM: intra-muscular, IV: intra-veinous, JC1: 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide, LC3: microtubule-associated protein light chain 3, MAPK: mitogen-activated protein kinase, MDA: malondialdehyde, Mfn: mitofusin, mito: mitochondria, MMP: Mitochondrial membrane potential, NDUF8: protein complex I, MTCO1: protein complex IV, NFκB: nuclear factor kappa B, NO: nitric oxide, NOX2: NADPH oxidase 2, NRF2: nuclear factor erythroid 2, NQO1: NAD(P)H:quinone oxidoreductase 1, NT: nitrotyrosine, OPA1: optic atrophy 1, OXPHOS: oxidative phosphorylation, PGC1α: Pparg coactivator 1 alpha, RCR: Respiratory control ratio, ROS: relative oxygen species, SDHB: protein complex II, SIRT1: sirtuine 1, SOD: superoxide dismutase, TFAM: mitochondrial transcription factor A, TOM: translocase of the outer membrane, TNFα: tumor necrosis factor α, Trolox: synthetic form of vitamin E, UC-MSC: Umbilical-cord MSC, UQCRC2: protein complex III, vATP5a: protein complex V, ↑: increase/ improve, ↓: decrease, ≈: data have been estimated from the graphs.

3.2.3. Drug-related toxicity

Eleven studies [90–100] investigated the impact of mitochondrial transplantation on drug-related toxicity (Table 3), eight of which focused on chemotherapy agents (doxorubicin and cisplatin) known for their cardiotoxicity, neurotoxicity, and nephrotoxicity via oxidative stress. Mitochondrial transplantation showed significant potential in mitigating these drug-related toxicities by reducing inflammation and cell apoptosis. Additionally, it improved histological and biochemical markers of acetaminophen-induced toxicity. No EV studies were found in the search.

Table 3.

Mitochondrial transplantation effect to prevent drug-related toxicity.

Author/Year	Model/Disease	Mitochondria source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Jin 2024 [90]	C57BL/6N mice Doxorubicin cardiotoxicity	MSC cells Not specified	Dose: 3 mg/kg Administration: IV Site: not specified	Restore ejection fraction and fractional shortening ↓ fibrosis ↓ apoptosis	↑ ATP level ↑ mito quantity (image) ↑ anti-ox (mRNA Sdhaf2/4) ↓ autophagy (mRNA LC3BII, p62) ↓ autophagosomes (inhibit AMPK/mTOR pathway)
Kim 2024 [91]	Sprague Dawley rat Dexamethasone related amyotrophy	UC-MSC cells Human Xenogeneic	Doses: 0.5–5 μg/limb Administration: IM Site: soleus	Positive effect with 5 μg Recover muscle mass, fiber area, desmin, myogenin ↓ lactate	Act AMPK/AKT/Foxo pathway (↑Akt, ↓ FOXO3, ↓ MuRF-1, ↓ MAFbx) ↑ OXPHOS (CI, II, IV, V activity) ↑ mito biosynthesis (PGC-1, TFAM)
Kubat 2024 [92]	Sprague Dawley rat Doxorubicin muscular toxicity	Femoris Rat Allogenic	Dose: 6.5 μg mito proteins Administration: IM Sites: 3 tibialis	No protective effect No effect on apoptosis No effect on inflammation	↑ anti-ox (succinate DH, SOD)
Maia 2023 [93]	SWISS mice Oxaliplatin neurotoxicity	Liver Mice Allogenic	Dose: 0.5 mg/kg Administration: IV Site: retro-orbital	↓ cold mechanical and sensitive alteration ↓ neuro-inflammation (TNFα, TLR4)	Not investigated
Maleki 2023 [94]	Wistar rat Doxorubicin cardiotoxicity	Liver Rat Allogenic	Dose: 8*109 mito Administration: IV Site: tail	Restore ejection fraction and fractional shortening ↓ apoptosis ↓ inflammation (TNFα, IL1β, IL6)	↑ ATP level ↓ oxidative stress markers (MDA) ↑ anti-ox (SOD, CAT, GPX, GSH)
Sun 2023 [95]	C57BL/6 mice Doxorubicin cardiotoxicity	Heart Mice Allogenic	Dose: 105 mito Administration: intraventricular Site: cardiac	↑ ejection fraction and fractional shortening ↓ apoptosis	↓ Superoxide (mitosox probe) ↑ anti-ox (SOD2, OGG1) ↑ mito function (IDH2, citrate synthase) ↑ mitogenesis (PGC1-b) ↑ fusion (Opa1) ↑ OCR/ECAR ratio Act glutamine & glutamate metabolic pathway
Zhang 2023 [96]	C57BL/6 mice Doxorubicin cardiotoxicity	Heart Mice Allogenic	Dose: 5*104 mito Administration: intraventricular Site: cardiac	↑ ejection fraction and fractional shortening	↑ mito quantity (hs & ms mtDNA) ↓ intracellular ROS, ↑ mito ROS ↑ MMP ↑ baseline and max OCR
Alexander 2021 [97]	C57BL/6J mice Cisplatin neurotoxicity	MSC cells Human Xenogeneic	Dose: 170 μg mito proteins Administration: intra-nasal	↑ executive functioning Restore spatial/working memory Restore synaptic damage	↓ atypical mito morphology ↓ synaptosomal mito damage ↑ anti-ox (Nrf2)
Kubat 2021 [98]	Sprague Dawley rat Doxorubicin nephrotoxicity	MSC cells Rat Allogenic	Dose: 4*106 mito Administration: injection Site: below kidney capsule	↑ cellular regeneration ↓ apoptosis ↓ proteinuria	↑ anti-ox (SOD, GPX)
Ulger 2021 [99]	Sprague Dawley rat Acetaminophen hepatotoxicity	MSC cells Rat Allogenic	Dose: 8.2*106 mito Administration: injection Site: subscapular spleen	↓ histological alteration ↓ apoptosis and ALAT	↓ anti-ox (GSH)
Shi 2018 [100]	Sprague Dawley rat Acetaminophen hepatotoxicity	HepG2 cells Human Xenogeneic	Dose: 10 mg/kg (fresh or fractioned) Administration: IV Site: tail	Results obtained only for fresh mito ↓ histological alteration ↓ ALAT and ASAT	↑ ATP level, only for intact mito ↑ anti-ox (GSH), only for intact mito ↓ ROS, only for intact mito

Akt: protein kinase B, ALAT: alanine aminotransferase, AMPK: AMP-activated protein kinase, anti-ox: anti-oxidant; ASAT: aspartate aminotransferase, ATP: adenosine tri-phosphate, CAT: catalase, ECAR: extracellular acidification rate, FOXO: forkhead box O, GPX: glutathione peroxidase, GSH: Reduced glutathione, IDH2: isocitrate dehydrogenase 2, IL: interleukin, IM: intra-muscular, IV: intra-veinous, LC3: microtubule-associated protein light chain 3, mito: mitochondria, MAFbx: muscle atrophy F-box, MDA: malondialdehyde, MMP: Mitochondrial membrane potential, MSC: mesenchymal stromal cell, mTOR: mammalian target of rapamycin, MuRF-1: muscle RING finger protein 1, NRF2: nuclear factor erythroid 2, OCR: oxygen consumption rate, OGG1: 8-oxoguanine DNA glycosylase 1, OPA1: optic atrophy 1, OXPHOS: oxidative phosphorylation, PGC1b: Pparg coactivator 1-b, ROS: relative oxygen species, Sdhaf: Succinate dehydrogenase assembly factor 1, SOD: superoxide dismutase, TFAM: mitochondrial transcription factor A, TLR4: Toll-like receptor 4, TNFα: tumor necrosis factor α, UC-MSC: Umbilical-cord MSC, ↑: increase/improve, ↓: decrease.

3.2.4. Sepsis-related outcomes

Seven studies [101–107] investigated the role of mitochondrial transplantation in sepsis (Table 4). Five of these studies reported survival rates with an improvement ranging from 21 % to 40 %, accompanied by reduced biological markers of inflammation and enhanced bacterial clearance. Two studies [108,109] investigated the role of EV in sepsis (Table 7), one showing improvement of survival rate and the other one a decrease in intestinal barrier dysfunction, both associated with improvement in inflammatory markers.

Table 4.

Mitochondrial transplantation effect on sepsis.

Author/Year	Model/Disease	Mitochondria source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Kim 2023 [101]	Sprague Dawley rat Sepsis	L6 cells Rat Allogenic	Dose: 200 μg mito proteins Administration: IV Site: tail	No clinical data reported ↑ expression of RT1-m2 and Cbln2	Not investigated
Kim 2023 [102]	Sprague Dawley rat Sepsis	L6, clone 9, UC-MSC cells Rat Allogenic	Dose: 200 μg mito proteins Administration: IV Site: tail	↑ survival rate at day 12 (40 % vs. 0 % only L6 mitochondria) ↓ creatinine and ALAT levels	Not investigated in vivo In vitro: ↑ basal & max respiration ↑ ATP production ↑ proton leak ↑ non mito oxygen consumption
Mokhtari 2023 [103]	Wistar rat Sepsis	Pectoralis major Rat Allogenic	Dose: 3*106 mito Administration: IV Site: tail	↑ survival rate at day 3 (50 % vs. 10 %) ↓ inflammation (TNFα, IL1β, IL6), LDH	↓ cellular ROS (DCFHDA) ↑ mito biogenesis (SIRT1, PCG-1α) ↑ fusion: (Mfn2) ↓ fission (DRP1)
de Carvalho 2021 [104]	C57BL/6 mice Sepsis	MSC cells Not specified	Dose: from 3*106 cells Administration: IV Site: jugular	↑ survival rate at day 2 (57 % vs. 36 %) ↓ bacteria load ↓ inflammation (IL1β), ↑IL10	Vascular endothelial cells: ↑basal & max respiration, Iuncoupled respiration Alveolar epithelial cells: ↑ max respiration, no effect on basal and uncoupled respiration
Hwang 2021 [105]	Sprague Dawley rat Sepsis	UC-MSC cells Rat Allogenic	Dose: 50 μg mito proteins Administration: IV Site: tail	↑ survival rate at day 14 (50 % vs. 10 %) ↑ bacterial clearance and ↓ lactate	↑ ATP level ↓ anti-ox (SOD2) ↑ ROS (DCFHDA) Restore O2 consumption spleen, muscle, no effect on liver, kidney, heart
Zhang 2021 [106]	C57BL/6 mice Sepsis	Pectoralis major Mice Allogenic	Dose: 2.5–3*107 mito Administration: IV Site: tail	↑ survival rate at day 7 (75 % vs. 36.8 %) ↑ bacterial clearance, nitrate + nitrite ↓ organ injury ↓ inflammation (IL1β and IL6) ↓ apoptosis	↓ OCR lung, liver, kidney, no effect on brain
Yan 2020 [107]	C57BL/6 mice Sepsis	Pectoralis major Mice Allogenic	Dose: 4*106 mito Administration: intraventricular Site: cerebral	No effect on mortality, mortice and explorative function ↑ memory retention performance ↑ cognitive impairment ↓ inflammation (TNFα, IL6, IL1β, TGFβ, IL4), no effect on IL10 ↓ apoptosis	↓ oxidative stress (iNOS)

ALAT: alanine aminotransferase, ATP: adenosine tri-phosphate, Cbln-2: cerebellin 2, DCFHDA: 2′,7′-dichlorodihydrofluorescein diacetate, DRP1: dynamin-related protein 1, IL: interleukin, iNOS: inducible nitric oxide synthase, IV: intra-veinous, LDH: lactate dehydrogenase, Mfn: mitofusin, mito: mitochondria, MSC: mesenchymal stromal cell, OCR: oxygen consumption rate, PGC1-α: Pparg coactivator 1-alpha, ROS: relative oxygen species, RT1-m2: major histocompatibility complex gene, SOD: superoxide dismutase, SIRT1: sirtuine 1, TGFβ: transforming growth factor beta, TNFα: tumor necrosis factor α, UC-MSC: Umbilical-cord MSC, ↑: increase/ improve, ↓: decrease.

3.2.5. Wound healing and aging

Five studies [110–116] reported the role of mitochondrial transplantation and three [117–119] the role of EV in the aging and healing process (Tables 5 and 7). They show a consistent increase in cell proliferation associated with a decrease in apoptosis, and improvement in related clinical investigation (i.e., healing, hair growth, muscular function).

Table 5.

Mitochondrial transplantation effect on wound healing and aging.

Author/Year	Model/Disease	Mitochondria source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Arroum 2024 [110]	C57BL/6J mice Aging and muscular function	Hind limbs Mice Allogenic	Dose: 109 mito Administration: IM Sites: 3	↑ maximal speed and distance	↑ ATP level ↑ COX activity Plantaris: ↑parkin, TFAM, BNIP3, Drp1 Soleus: ↑parkin, DRP1, no effect TFAM BNIP3 Femoris: No effect
Arroum 2024 [111]	Rat Aging and muscular function	Femoris Rat Allogenic	Dose: 109 mito Administration: IM Sites: 3	No effect on speed and distance	↑ ATP level ↑ COX activity Plantaris and soleus: ↑ fusion (OPA1), ↑ mitochondrial homeostasis (Parkin, TFAM)
Li 2024 [112]	Kunming mice Burn injury	Liver Mice Allogenic	Dose: 10 μg/g Administration: SC Site: not specified	↑ healing regeneration, ↓ scarring ↓ apoptosis ↓ inflammation (TNFα, IL1β, IL6) ↑ IL10	Not investigated
Raz 2024 [113]	C57BL/6 mice Corneal wound healing	Liver Mice Allogenic	Dose: 25 μL mito preparation (fresh or fractioned) Administration: drop Site: eyes	↑ re-epithelialization and better cyto-architecture ↓ cellular infiltration Only the fresh ↓ wound area	↓ ATP level
Kshersagar 2023 [114]	Wistar rat Endometrium Injury	Placenta Human Xenogeneic	Dose: from 106 cells Administration: injection Site: intrauterine	↑ regeneration ↑ endometrial protein junction	↑ OXPHOS (activity CI, CII, CIII, CIV, CytB5, CytP450, NADPH-CytC red, NADH-CytC red)
Wu 2022 [115]	RCS/Kyo rat Retinal degeneration	Liver Rat Allogenic	Dose: 50–100 μg mito proteins Administration: injection Site: intra-retinal	Restore retinal layer thickness ↑ number of cells (only 100 μg) Anti-degeneration activity	Not investigated
Wu 2020 [116]	C57B/6 mice Hair aging	Liver Mice Allogenic	Dose: 200 μg mito proteins Administration: SC Site: not specified	↑ hair length, number of hair follicles No effect on derma thickness ↑ collagen	↑ ATP synthase ↑ mito quantity (mtDNA)

ATP: adenosine tri-phosphate, BNIP3: Bcl-2/adenovirus E1B-interacting protein 3, COX: cytochrome C oxidase, DRP1: dynamin-related protein 1, IL: interleukin, IM: intra-muscular, mito: mitochondria, OPA1: optic atrophy 1, SC: subcutaneous, TFAM: mitochondrial transcription factor A, TNFα: tumor necrosis factor α, ↑: increase/ improve, ↓: decrease.

3.2.6. Other diseases and conditions

A total of 28 studies [120–147] explored other potential applications of mitochondrial transplantation (Table 6). Six studies focused on lung disorders, with beneficial effects on pulmonary hypertension [127,139], hyper-reactivity [138], endotoxin-induced injury [129] and acute respiratory distress syndrome [146,147]. Four studies [124,130,133,143] explored the impact on tumors, with mixed results, and four other studies reported benefit on osteo-muscular disorders (myo-dystrophy [120,122], osteoarthritis [128], tendinopathy [131]). Additionally, there were positive outcomes reported in cardiological conditions (cold ischemia [135], right heart failure [132], hypertrophy [144] and graft quality [142]), liver conditions (CCl₄ injury [134], fatty liver [137], fibrosis [145] and hepatitis [141]), and diabetes [125,126] (improved glycemic control). Two studies also demonstrated efficacy in treating erectile dysfunction [121,140]. Lastly, a study investigating the safety of mitochondrial transplantation in a large animal model [123] (horse) found no evidence of inflammation or adverse effects. This was further supported by a study examining the immunogenicity of different mitochondrial sources (allogeneic, syngeneic), with no reported immune or inflammatory reactions [136].

Table 6.

Other mitochondrial transplantation effects.

Author/Year	Model/Disease	Mitochondria source	Transfer parameters	Clinical & biological evidences	Mitochondrial mechanisms modulation
Dubinin 2024 [120]	C57BL/10 mice Muscular dystrophy	Skeletal muscle Mice Allogenic	Dose: 1 μg/g body weight Administration: IM Sites: gastrocnemius/ quadriceps	↑ muscle strength, grips strength ↑ number of muscular fibers ↓ calcification ↓ creatine kinase No effect on inflammation (TNFα, IL6)	↓ MAM surface area ↓ RCR ↓ Ca2+ retention capacity ↓ oxidative stress (TBARs)
Zhai 2024 [121]	Sprague Dawley rat Erectile dysfunction	Adipose MSC cells Rat Allogenic	Dose: 300 μg mito proteins Administration: injection Site: intra-cavernosum	Restore erectile function Restore smooth muscle content ↓ apoptosis	↑ ATP level ↓ oxidative stress (mitosox) ↑ anti-ox (SOD)
Always 2023 [122]	C57BL/6 mice Injured skeletal muscle	Liver Mice Allogenic	Dose: 50 μg mito proteins Administration: IV Site: not specified	↓ collagen (day 7 not day 14) No effect on inflammation, muscular mass, maximal plantar force	Not investigated
Cassano 2023 [123]	Horse Safety	Platelet cells Horse Autologous	Dose: 1 mL of mito preparation Administration: injection Site: intra-synovial	No inflammation No adverse effect	Not investigated
Celik 2023 [124]	Athymic nude mice Ovarian cancer	Cardiac fibroblast cells Human Xenogeneic	Dose: 107 mito Administration: IV Site: not specified	No effect on tumor size	Not investigated in vivo In vitro: ↑ mtDNA, ↑ ATP level
Mudgal 2023 [125]	Wistar rat Diabetes	Pectoralis major Rat Allogenic	Dose: 0.5 mg/kg Administration: IV Site: not specified	↓ blood glucose ↓ inflammation (NFκB, IL6)	↑ ATP level ↑ mito quantity (mtDNA) ↑ complex activity (I, II, III) ↑ anti-ox (SOD, CAT, GSH) ↓ oxidative stress (NO, TBARs) ↑ mito biogenesis (PGC1α)
Paliwal 2023 [126]	Wistar rat Diabetes	Pectoralis major Rat Allogenic	Dose: 100 μg mito proteins Administration: IV Site: not specified	↓ diastolic/systolic pressure ↓ abdominal fat, cholesterol, BMI, ALAT, ASAT ↓ glycaemia	↑ ATP level ↑ complex activity (I, II, III, IV, V) ↑ anti-ox (SOD, CAT, GSH) ↓ oxidative stress (NO, TBARs) ↓ mito biogenesis (PGC1α)
Hsu 2022 [127]	Sprague Dawley rat Pulmonary hypertension	Soleus Rat Allogenic	Dose: 100 μg mito proteins Administration: IV Site: jugular	No effect on mortality ↑ right ventricular function, ↓ BNP Restore contractile phenotype of pulmonary artery	↑ ATP level
Lee 2022 [128]	Wistar rat Osteoarthritis	L6 cells Rat Allogenic	Dose: 10 μg mito proteins Administration: injection Site: intra-articular	↓ osteoarthritis progression, suppressing pain ↓ bone damage ↑ bone surface and volume ↓ inflammation (TNFα, IL1β, IL6, NFκB)	↑ mito quantity (mitotracker) ↓ mito diameter ↑ MMP (JC-1) ↓ ROS (mitosox) ↓ autophagy (LC3–2/LC3–1 ratio, p62)
Pang 2022 [129]	Sprague Dawley rat Lung injury endotoxins	Soleus Rat Allogenic	Dose: 100 μg mito proteins Administration: IV Site: jugular	↑ gas exchange and bicarbonate levels ↓ lung parenchyma damage	↑ ATP level
Zhou 2022 [130]	BALB/c mice Hepatocellular carcinoma	Liver Mice Allogenic	Dose: 6*107 mito Administration: IV Site: tail	↓ tumor volume (≈ decrease of 30 %) ↓ apoptosis and lactate ↓ tumor metabolism (hexokinase, lactic acid, ATP), ↑ pyruvate DH, ↑ succinate DH, ↑ isocitrate DH	↓ cellular ROS (DCFHDA) ↑ anti-ox (SOD, CAR, GSH)
Lee 2021 [131]	Sprague Dawley rat Tendinopathy	L6 cells Rat Allogenic	Doses: 10–50 μg mito proteins Administration: injection Site: tendon	↓ tendon thickness (both) ↓ apoptosis ↓ inflammation (TNFα, IL6, NFκB: both; IL1β: only 50 μg)	↑ ATP level ↑ mito quantity (mitotracker) ↓ Superoxide and cellular ROS (mitosox, DCFHDA) ↑ OXPHOS prot level ↓ fission (Fis1, DRp1) ↑ fusion (Mfn2)
Weixler 2021 [132]	Yorkshire pig Right heart failure	Gastrocnemius, soleus Rat Xenogeneic	Dose: 107 mito Administration: intraventricular Site: cardiac	No effect on mortality Restore fractional area change ↑ contractile function ↓ fibrosis and apoptosis	Not investigated in vivo In vitro: ↑ ATP level
Yu 2021 [133]	BALB/c mice Melanoma	Liver Mice Allogenic	Dose: 106 mito Administration: IV Site: not specified	↓ tumor volume (≈ decrease of 50 %) ↓ lactate	↑ ATP level ↑ autophagy (LC3), ↑ mitophagy (Parkin)
Zhao 2021 [134]	Kunming mice CCl4 hepatic lesion	Liver Mice Allogenic	Dose: 0.2–0.4 mg/kg Administration: IV Site: tail	Trend for a dose effect ↓ lesion area and fibrosis ↓ ALAT and ASAT	↑ ATP level ↑ MMP (JC-1) ↑ anti-ox (GSH, SOD, NADH DH, pyruvate DH) ↓ oxidative stress (MDA)
Moskowitzova 2019 [135]	C57BL/6J mice Cold ischemia time and cardiac function	Gastrocnemius Mice Allogenic	Dose: 108 mito Administration: IV Site: coronary	↑ ejection fraction and fractional shortening ↓ necrosis ↓ neutrophil infiltrate No effect on apoptosis	Not investigated
Ramirez-Barbieri 2019 [136]	BALB/c mice Safety and immune reaction	Gastrocnemius Mice Allogenic and syngenic	Doses: 105–7 mito Administration: intraperitoneal	No dose effect, no type of transplantation effect No rejection of arterial graft No immune reaction (TNFα, IFNγ, IL2, IL4, IgM)	Not investigated
Fu 2017 [137]	C57BL/6J mice Fatty liver	HepG2 cells Human Xenogeneic	Dose: 0.5 mg/kg Administration: IV Site: tail	↓ cholesterol and hepatic lipid infiltrate ↓ ALAT and ASAT	↑ ATP level ↑ CCO ↓ oxidative stress (Fenton reaction, MDA) ↑ anti-ox (SOD, GSH)
Su 2016 [138]	Sprague Dawley rat Hyper-reactivity bronchial	Airway epithelial cells Rat Allogenic	Dose: 3.75*107 mito Administration: Intratracheal	↓ airway hyper-reactivity	↑ MMP (JC-1)
Zhu 2016 [139]	Sprague Dawley rat Hypoxia and pulmonary hypertension	Pulmonary artery/lung Rat Allogenic	Dose: 2.25*108 mito (fresh or fractioned) Administration: IV Site: tail	Results obtained for both (fresh and fractioned mito) Inhibit hypoxia pulmonary vasoconstriction Restore pulmonary artery pressure ↓ apoptosis ↓ vascular remodeling	↑ mitochondrial quantity (ROS increase)

ALAT: alanine aminotransferase, anti-ox: anti-oxidant, ASAT: aspartate aminotransferase, ATP: adenosine tri-phosphate, BMI: body mass index, BNP: B-type natriuretic peptide, CAT: catalase, DCFHDA: 2′,7′-dichlorodihydrofluorescein diacetate, DRP1: dynamin-related protein 1, Fis1: mitochondrial fission protein 1, GSH: reduced glutathione, IgM: immunoglobulin M, IL: interleukin, IM: intra-muscular, IV: intra-veinous, JC1: 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazo-lylcarbocyanine iodide, LC3: microtubule-associated protein light chain 3, MAM: Mitochondria-associated endoplasmic reticulum membranes, MDA: malondialdehyde, Mfn: mitofusin, mito: mitochondria, MMP: Mitochondrial membrane potential, MSC: Mesenchymal stromal cell, NFκB: nuclear factor kappa B, NO: nitric oxide, RCR: respiration control rate, PGC1-α: Pparg coactivator 1-alpha, ROS: reactive oxygen species, SOD: superoxide dismutase, TBARS: Thiobarbituric Acid Reactive Substances, TNFα: tumor necrosis factor α ↑: increase/improve, ↓: decrease, ≈: data have been estimated from the graphs.

3.3. Safety of mitochondrial transplantation

Across all studies, the majority were conducted under significant stress conditions (IRI, sepsis, etc.), where the transplantation of EV led to a reduction in pro-inflammatory cytokines (IL-1β, TNF-α, etc.) in 87 % (27/31) of cases.

However, in several studies, the administration of mitochondria under non-stress conditions has also been investigated to assess potential reactions. First, the study by Cassano et al. [123] showed that there was no inflammatory reaction at the injection site of the mitochondria (joint), and no systemic inflammatory response was observed (no change in blood cell counts).

However, with 94 % (115/123) of the studies using a transplantation type other than autologous, immunogenic reactions would theoretically be possible. The safety results observed in the autologous model have been confirmed in various studies using rat or mouse models with allogeneic [76,94,107,138] or xenogeneic [81] mitochondria transplantation, which did not lead to changes in the number of immune cells (leukocytes, T lymphocytes), oxidative stress, or pro-inflammatory cytokines (IL-6, TNFα, IL-1β).

Finally, these results were confirmed biologically by Ramirez-Barbieri et al., [136] showing an absence of acute direct or indirect alloreactivity. This team further investigated by verifying that mitochondrial administration did not lead to chronic immunization in the mice. They sequentially injected syngeneic or allogeneic mitochondria into mice (3 injections), and after a few days, they performed an allogeneic skin graft. The graft rejection time was comparable between the groups, suggesting an absence of immunization of these mice by the previously administered mitochondria.

3.4. Unraveling mitochondrial complexity: functional and structural mechanisms underlying mitochondrial-based therapies

Although the positive therapeutic outcomes reported across various diseases are promising, the underlying biological mechanisms remain incompletely understood. As recently emphasized by Monzel, Enríquez, and Picard (2023), mitochondria are far more than cellular powerhouses. They participate in a wide array of critical functions, including membrane potential generation, calcium uptake and release, lipid metabolism, mtDNA maintenance and expression, sodium ion exchange, OXPHOS, permeability transition, respiration, redox homeostasis, and reactive oxygen species (ROS) production. Importantly, the authors highlighted not only the diverse functions of mitochondria but also their dynamic behaviors—such as cristae remodeling, mitochondrial fusion and fission, and overall mitochondrial homeostasis—underscoring a more nuanced and integrative view of mitochondrial biology [148]. In this review, we examine both well-established and underexplored mitochondrial functions modulated by mitochondrial and EV transplantation, aiming to clarify how these alterations may contribute to the observed therapeutic benefits across disease models (Fig. 3). However, it is important to acknowledge that 17 % of the selected studies did not assess mitochondrial activity or behavior, representing a limitation in fully understanding the mechanistic basis of these interventions.

Fig. 3. Mechanisms of action and clinical outcomes of mitochondrial and extracellular vesicle (EV) transplantation.

Mitochondria and EV can be isolated from various autologous or allogeneic sources, including immune cells, cardiac and neuronal tissues, organs and biological fluids. These mitochondrial therapies have been associated with multiple clinical outcomes such as infarct size reduction, decreased inflammation, rescue of cognitive functions, vascular remodeling, and regeneration of cells and organs. Mechanistically, transplanted mitochondria and EV exert their effects through: (1) enhancement of bioenergetic and redox processes by increasing oxidative phosphorylation (OXPHOS) and ATP production via upregulation of NDUFB8, SDHB, UQCRC, MTCO1, and ATP5A; (2) redox homeostasis maintenance through antioxidant pathways involving Nrf2, SOD, CAT, GSH, MAPK, and FOXO signaling; (3) modulation of mitochondrial dynamics via regulation of fission (Drp1) and fusion (Mfn1, Mfn2, Opa1) processes; and (4) preservation of mitochondrial homeostasis through enhanced mitophagy (increased LC3-II/LC3-I ratio) and mitochondrial biogenesis (upregulation of PGC1α and TFAM). These mechanisms underpin the therapeutic potential of mitochondrial and EV-based interventions across a range of diseases.

3.5. Bioenergetic and redox mechanisms underlying the therapeutic effects of mitochondrial and EV transplantation

Mitochondria are essential organelles responsible for producing ATP through OXPHOS, a process that converts energy from nutrients into a useable form for cellular functions [149,150]. This energy production is vital for maintaining cellular activities and overall organismal health. Additionally, mitochondria play a crucial role in regulating redox homeostasis, balancing the production and neutralization of ROS to prevent oxidative stress and cellular damage [151,152]. In this section, we explore the impact of mitochondrial and EV transplantation on enhancing OXPHOS efficiency, increasing ATP production, and maintaining redox balance. We also examine how these interventions contribute to restoring mitochondrial function and improving clinical outcomes across various disease models.

3.5.1. OXPHOS and ATP increase

The OXPHOS system, composed of five multi-protein complexes, includes the electron transport chain (ETC), which transfers electrons to oxygen, generating an electrochemical gradient. This gradient powers key mitochondrial functions, including ATP synthesis, ion/metabolite transport, protein import, and maintaining mitochondrial dynamics [153]. In many studies utilizing mitochondrial or EV transplantation, researchers observed increased expression of OXPHOS proteins—such as NDUFB8, SDHB, UQCRC2, MTCO1, and ATP5A—alongside enhanced activities of complexes I through V, indicating improved mitochondrial function [154]. In a majority of publications, the authors showed an increase in ATP production in IRI disease models [27,33,36,37], in neurological disease models (increase in ATP level and increase in pyruvate DH, α-ketoglutarate DH, NADH) [62,71–73,77,81,83,88,89], in the prevention of drug-related toxicity [91], in sepsis [106], in wound healing [114], in diabetes [125,126], and in osteoarthritis [128]. In all these papers, where an increase in OXPHOS protein expression/activity, and in particular ATP level increase was observed, the authors also described clinical benefits. In the IRI disease model publications, the authors showed an increase in ATP level concomitant with a reduction of infarct size [28,29,31,33,35–37,47,49,50]. However, it is important to note that not all improvements in OXPHOS protein expression and ATP levels translated into clinical benefits. In studies focused on neurological disorders, drug-related toxicity, aging/wound healing, and tumors, some papers reported increases in ATP levels and OXPHOS protein expression without corresponding clinical effects [72,73,86,110, 124].

Based on the selected literature, we can formulate three hypotheses to explain how mitochondria and EV transplantation may enhance OXPHOS system activity and increase ATP level. First, the increase in ATP levels can be explained by the integration of healthy mitochondria into the recipient cell’s network, where they actively generate ATP through OXPHOS [155]. Second, transplanted mitochondria may regulate intracellular Ca²⁺ levels, which subsequently increase ATP synthesis [156]. Indeed, in one of the selected papers in IRI disease [26], the authors showed an increase in Ca²⁺ retention capacity (CRC), although in another article in a muscular dystrophy model, the CRC decreased after mitochondrial transplantation into muscles [157]. However, it is important to consider that even though CRC is crucial for efficient ATP production, excessive Ca²⁺ can lead to mitochondrial dysfunction and cell death [158]. Third, mitochondrial morphology and dynamics (particularly fusion and fission) may also play a role, although this process will be discussed further in later paragraphs.

3.5.2. Redox homeostasis maintenance

Another essential mitochondrial function is redox homeostasis as maintenance of redox reactions are central to the existence of life [159]. Indeed, reactive oxygen species (ROS), nitrogen and sulfur species, mediate redox control of a wide range of essential cellular processes [160]. Moreover, excessive levels of these species are associated with many diseases. Maintaining a finely tuned balance between ROS production and scavenging is essential for cellular homeostasis [160]. Several studies have shown that mitochondrial transplantation can effectively reduce oxidative stress across various conditions. In IRI, mitochondrial transplantation has been associated with decreased ROS levels and lower oxidative stress markers, such as MDA, TBARS, lipid peroxidation, 8-OHDG, and nitrotyrosine [25,27,28,30,50,53] Similarly, in neurological diseases, reductions in both ROS content and oxidative stress markers following mitochondrial transplantation have been reported [63,65,68,71,74,81,83,85,88]. Studies on drug-related toxicity suggest that mitochondrial transplantation helps prevent oxidative stress, as evidenced by decreased ROS levels and oxidative stress markers such as MDA and 8-OHDG [94–96,99,100]. In sepsis, mitochondrial transplantation has also been linked to a reduction in ROS content and oxidative stress [103,105,107]. Furthermore, studies on erectile dysfunction, diabetes, osteoarthritis, hepatocellular carcinoma, and tendinopathy have reported a decrease in oxidative stress following mitochondrial transplantation [121,125,126,134,137]. In muscular dystrophy, diabetes, osteoarthritis, hepatic lesions, and fatty liver, the reduction in oxidative stress markers such as MDA, TBARS, lipid peroxidation, 8-OHDG, and nitro-tyrosine further supports the potential therapeutic role of mitochondrial transplantation in mitigating oxidative stress-related damage [120,125,126,134,137]. Regarding the transplantation of EV, the selected papers reported a decrease in ROS levels and respiratory burst in various conditions, including IRI [58,60], sepsis [108], wound healing [117] and liver fibrosis [145]. Only one of the selected articles which conducted EV transplantation explicitly investigated the reduction of oxidative stress markers, demonstrating that EV decrease levels of carbonylated proteins, thereby enhancing graft quality [142].

The decrease in ROS, oxidative stress markers, and the overall oxidative stress environment, essential for maintaining redox homeostasis and restoring the cellular steady state, is primarily mediated by endogenous antioxidant enzymes, such as superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), glutathione reductase (GSR), and peroxiredoxins (PRxs), along with antioxidant molecules like glutathione (GSH). The expression of many antioxidant enzymes is orchestrated by the stress-responsive transcription factor Nuclear factor erythroid 2-related factor 2 (Nrf2), which also regulates stress-response proteins, such as heat shock proteins (HSPs), to maintain cellular redox homeostasis [161]. It is well known that mitochondrial respiration produces ROS essential for signaling and the mitochondria themselves possess antioxidant enzyme systems for the maintenance of Redox balance [162]. In the papers selected for this review, the majority of studies indicate that a decrease in ROS is associated with an increase in antioxidant enzymes content, particularly manganese superoxide dismutase (SOD2, the mitochondrial form), glutathione peroxidase (GPX, which utilizes glutathione (GSH) as a substrate), heme oxygenase-1 (HO-1), NAD(P)H-quinone oxidoreductase 1 (NQO1), 8-oxoguanine DNA glycosylase 1 (OGG1), catalase (CAT), and total antioxidant capacity. This increase is also associated with the translocation of transcription factors such as Nrf2, following the inhibition of its co-factor, Kelch-like ECH-associated protein 1 (Keap1). This inhibition allows Nrf2 to translocate into the nucleus, promoting the transcription of antioxidant enzymes such as CAT and SOD [163]. Additionally, there is an upregulation of Succinate Dehydrogenase Complex Assembly Factor (Sdhaf2/4) mRNA, along with the activation of other signaling pathways, including mitogen-activated protein kinase (MAPK) and Forkhead box class O (FOXO) [164]. These effects have been observed in IRI [26, 30,50], neurological disorders [71,72,81,83,85,88], drug-related toxicity prevention [90,92,94,95,97,98,100], erectile dysfunction [121], diabetes [125,126], hepatocellular carcinoma [130], hepatic lesions [134] and fatty liver disease [137]. Additionally, EV can trigger an antioxidant enzyme response in IRI [55,60], sepsis [108] and liver fibrosis [145]. However, the activation of the antioxidant response and reduction in ROS levels do not always translate into clinical benefits. In the selected papers on IRI, some studies reported no effect on clinical outcomes [25,26], as well as in studies on neurological diseases [72] and drug-related toxicity prevention [92].

Based on the review of the selected articles, it is reasonable to suggest that the observed decrease in ROS levels is due to an increase of the antioxidant system response after the mitochondria and EV transplantation. Several hypotheses can be proposed to explain this phenomenon. First, Mitochondria and EV can increase intracellular ROS levels, leading to a transient activation of the Nrf2 pathway through disruption of Keap1-mediated repression. This allows Nrf2 to translocate into the nucleus, where it binds to antioxidant response elements (AREs) and promotes the expression of key antioxidant enzymes such as SOD, CAT, GPx1, and glutamate–cysteine ligase (GCL), thereby restoring the GSH:GSSG ratio critical for redox buffering and detoxification of lipid peroxides [163,165,166]. Additionally, ROS generated by mitochondrial respiration can induce epigenetic modifications—such as histone acetylation (e.g., H3K27ac) and DNA demethylation—at antioxidant gene loci, enhancing chromatin accessibility to Nrf2 and other transcription factors, and contributing to sustained antioxidant enzyme expression [167]. Second, mitochondrial-derived signals modulate FOXO transcription factor activity through phosphorylation by AKT and JNK resulting in nuclear translocation or cytoplasmic retention of FOXO1/3a, thereby regulating expression of genes involved in oxidative stress resistance (e.g., catalase, SOD2, and GADD45) [165]. Third, the exposure to mitochondrial material activates redox-sensitive MAPK cascades (e.g., p38, ERK1/2), which in turn phosphorylate key transcription factors or co-activators, enhancing transcription of stress-responsive genes (e.g., HO-1, NQO1) [168].

3.5.3. Mitochondrial dynamics and structural remodeling in mitochondrial and EV transplantation

Mitochondria are enclosed by two membranes with distinct structures and functions: the outer membrane and the inner membrane, which are further organized into specialized regions. The outer membrane separates mitochondria from the cytosol and is the site where fission and fusion processes occur, mediated by key mitochondrial-shaping proteins such as Mitofusin 1 and 2 (Mfn1 and Mfn2) and Fission 1 protein (Fis1) [169]. The inner membrane is subdivided into the inner boundary membrane (IBM) and the cristae [169]. The IBM contains the translocase of the inner membrane, which imports proteins into the matrix. Cristae are invaginated, bag-like structures separated from the intermembrane space by narrow tubular junctions, creating specialized compartments that limit the diffusion of molecules critical for the OXPHOS system [170]. A variety of proteins, many of which remain incompletely characterized, regulate cristae biogenesis and structure. Among these, optic atrophy protein (OPA1), another essential protein involved in mitochondrial fusion and the mitochondrial contact site and cristae organizing system (MICOS complex) are the primary regulators of cristae dynamics [171].

3.5.4. Fission and fusion

Mitochondrial fission (fragmentation) and fusion (elongation) are fundamental processes that regulate mitochondrial morphology and function. Their dynamic balance is essential for maintaining cellular homeostasis and preventing disease. Together with cristae remodeling, these interconnected mechanisms continuously shape the mitochondrial network, adapting it to cellular needs and stress conditions. This structural plasticity plays a critical role in supporting mitochondrial quality control, energy metabolism, and overall cellular health [172]. Indeed, mitochondrial fusion allows the transfer of gene products between mitochondria for optimal functioning, especially under metabolic and environmental stress. On the other hand, fission is crucial for mitochondrial division and quality control [173].

In the literature, fission is described as a faster process than fusion. In most cases, fission is used for the removal of damaged mitochondria by mitophagy or the homogeneous distribution of mitochondria in dividing cells. Fusion is a slower process that allows mitochondria to share their contents (proteins, lipids, mitochondrial DNA), facilitating repair and optimizing ATP production essential for mitochondria that are unable to perform their respiratory function optimally because of a disease condition [174]. Moreover, fusion can trigger fission because fused mitochondria form elongated, interconnected networks that enhance cellular energy distribution. However, these extended structures cannot be maintained indefinitely and require fission to reshape and preserve mitochondrial dynamics.

This delicate balance between fusion and fission is tightly regulated by key proteins. DRP1, a major driver of fission, can be recruited even after a fusion event to break down mitochondrial networks when needed. Meanwhile, MFN1, MFN2, and OPA1 orchestrate fusion but also influence subsequent fission, ensuring mitochondria remain adaptable to cellular demands. Indeed, where fusion and fission were studied, a majority of publications found an increase in fusion processes, rather than fission. In 3 studies, fission was also observed to coincide with an increase in ATP levels [28,40,72]. An increase in fission was also found to coincide with a reduction in infract size in IRI [28,40], improvement of cognitive ability [71] and improve in tendon thickness in Tendinopathy condition [131].

Many of the selected papers showed that mitochondrial transplantation led to modulation of mitochondrial dynamics, affecting both fission and fusion processes. In IRI, this involved a decrease in the fission protein DRP1 [38], and an increase in fusion proteins such as OPA1 [39] and Mfn1/2 [39,40]. In neurological diseases, there was a decrease in fusion proteins Mfn1/2 [66] and a decrease in the fission protein DRP1 [76], along with an increase in both fusion (Mfn2) and fission (DRP1) [89] as well as a publication that showed an increase in the proportion of elongated mitochondria [72]. Additionally, there was a reduction in atypical mitochondrial morphology and synaptosome mitochondrial damage [97]. In drug-related toxicity, mitochondrial fusion increased (OPA1) [95], while in sepsis, fusion increased (Mfn2) and fission decreased (DRP1) [103]. In healing and aging, there was an increase in mitochondrial fission (DRP1) [110], accompanied by reduced mitochondrial swelling, disruption, and cristae loss, along with upregulation of genes related to mitochondrial structure in IRI [26,28,29].

Regarding EV transplantation several studies have reported modulation of mitochondrial dynamics, particularly fusion and fission processes, across various disease models. In IRI, increased of MFN1 and MFN2 mRNA levels was observed following EV treatment [55]. One study demonstrated that human-derived mitochondria delivered via EV successfully fused with native mitochondrial networks in mouse cardiomyocytes, enhancing mitochondrial function and decreasing the dysfunctional mitochondrial fission [57]. In an IRI liver model, EV promoted mitochondrial fusion by increasing MFN1 [58] while in sepsis this resulted in an increase of MFN2 [109]. Similarly, in cardiac hypertrophy, EV transplantation led to increased mitochondrial fusion and elongation in neutrophils [144]. Additionally, EV were shown to reduce mitochondrial cristae loss and swelling, improving mitochondrial structural integrity in graft quality for transplant after IRI disease in pig [142].

Among the 15 publications that examined mitochondrial fusion and fission processes, 11 reported an increase in fusion. One study observed no effect on either process, while two studies noted an increase in fission. The increase of fusion process with the mitochondrial transplantation can be explained with different hypotheses. First, fusion of transplanted mitochondria when they enter in the recipient cells, share protein, lipids and mitochondrial DNA and thereby improve complex mitochondrial cell function which could also explain the increase in ATP levels [175]. Second, the fusion of transplanted mitochondria could be a mechanism of metabolic adaptation to efficiently distribute respiratory chain enzymes and optimize ATP production [175]. Third, the mitochondria of the recipient cells could be damaged and in a hyper-fragmented status, and the healthy mitochondria transplant may promote fusion to dilute damage and restore mitochondrial function, reducing oxidative stress and decreasing mitochondrial dysfunction [176].

Alternative hypotheses can be formulated to explain the observed increase in fission. First, following transplantation, the transplanted mitochondria may need to be efficiently distributed and reorganized within the recipient cell. Fission would allow the transplanted mitochondria to be split into smaller pieces, facilitating their distribution and integration into the cell’s pre-existing mitochondrial networks [177]. Second, balance between fusion and fission: If mitochondrial transplantation leads to initial fusion to restore mitochondrial function, fission could subsequently be activated to maintain mitochondrial balance and turnover, preventing the formation of overly long mitochondrial networks that would be unsustainable [177]. Third, stress response: Under conditions of cellular or metabolic stress (e.g., following mitochondrial transplantation), the cell may increase fission to remove damaged mitochondria or to better adapt to new energy demands. Fission also allows the separation of malfunctioning mitochondria, reducing oxidative stress and improving energy efficiency [176].

3.5.5. Mitochondria homeostasis

Mitochondria play a critical role in maintaining cellular homeostasis, requiring regulated mitochondrial mass and function to protect mtDNA and meet the cell’s energy demands. This balance is preserved through a fine coordination between two opposing processes: mitochondrial biogenesis, which generates new mitochondria, and mitophagy/autophagy, which removes damaged ones [178,179]. Many of the selected papers showed that mitochondrial transplantation maintained mitochondrial homeostasis by increasing mitophagy in IRI, thanks to proteins that promote the clearance of damaged mitochondria (PINK1, Parkin) [30]. In IRI, one study also reported a decrease in the LC3-II/LC3-I ratio [32], which could reflect an increase in autophagic flux, where formed autophagosomes are rapidly degraded in the lysosomes. This is associated with more efficient clearance of cellular materials. Additionally, two studies in IRI demonstrated downregulation of mitochondrial apoptosis [37,53]. In addition to functional and quantitative alterations, structural changes in mitochondria can also impair homeostasis. For instance, in muscular dystrophy models, mitochondrial transplantation was associated with restoration of mitochondrial-associated membrane (MAM) surface area, which is typically reduced in the disease. In neurological diseases, there is evidence of increased mitochondrial biogenesis as showed by elevated expression of proteins such as PGC1α and TFAM [62], as well as ERK and SIRT1, accompanied by increased autophagy, indicated by the expression of FOXO3, BNIP3, and LC3-II/LC3-I [71]. In the prevention of drug related toxicity, there is a decrease in autophagy (mRNA LC3BII, p62 decrease) and autophagosomes (inhibit AMPK/mTOR pathway, decrease of p70S6K) [90] as well as an increase in mitochondrial biogenesis with an increase of PGC-1, mtTFAM [91], Ppargc1b [95], SIRT1, PGC-1α [103]. One study also observed an increase in mitochondrial content, determined by mitochondrial DNA quantification and mitochondrial-specific staining [96]. Another publication shows mitochondrial biogenesis in sepsis with an increase of PGC-1α expression followed by a decrease in mitochondria fragmentation, following EV transplantation [109]. In the context of healing and aging, one study on mitochondrial transplantation reported an increase in TFAM expression [110], while another study demonstrated an elevation in mitochondrial quantity, as indicated by increased mitochondrial DNA levels [116] while a publication with EV showed an increase in mitophagy, specifically with the increase of LC3 II [118]. In various disease models, studies have reported alterations in mitochondrial dynamics, including changes in mitochondrial biogenesis, content, autophagy, and mitophagy. For instance, in diabetes, an increase in mitochondrial biogenesis has been observed, indicated by elevated levels of markers such as PGC-1α, mtTFAM and increase in mitochondrial DNA [125]. Conversely, in a metabolic syndrome model study, authors reported a decrease in mitochondrial biogenesis, evidenced by reduced PGC-1α expression [126]. In osteoarthritis, a decrease in autophagy markers, such as the LC3-II/LC3-I ratio and p62 levels, has been noted [128]. In melanoma models, there is evidence of increased autophagy, indicated by elevated LC3 levels, and enhanced mitophagy activity, as shown by increased Parkin expression [133]. Additionally, in acute kidney injury (AKI) models induced by IRI, the transplantation of EV contributes to the restoration of TFAM levels and stabilization of the TFAM-mtDNA complex, leading to a reduction in mitochondrial DNA damage [59]. These findings highlight the complex and context-dependent nature of mitochondrial dynamics in various pathological conditions.

We can formulate several hypotheses regarding the impact of mitochondrial transplantation on mitochondria homeostasis. First, mitochondrial and EV transplantation work like a “quality control jump-start”, reactivating mitophagy via known pathways (PINK1/Parkin, LC3-II), which clears dysfunctional mitochondria, restores energy homeostasis, and reduces stress signals This is especially critical in injured or diseased tissues where these pathways are often suppressed [180]. Second, mitochondrial and EV transplantation stimulates mitochondrial biogenesis with the upregulation of transcriptional regulators such as PGC1α, TFAM, and SIRT1 in multiple disease models, indicating the activation of endogenous renewal programs that support energy recovery and mtDNA stability. This can be possible because of metabolic and signaling stimulus, pushing the cell to increase mitochondrial production to restore energy balance [181]. Third, another mechanism that can restore mitochondrial homeostasis is through the structural reorganization of the MAMs. Indeed, transplanted mitochondria may contribute directly by physically integrating into host mitochondrial networks and re-establishing Endoplasmic Reticulum contact sites [182]. This can be possible also with the delivery of essential MAM-regulatory proteins such as Mfn2 or VDAC1, or modulate signaling pathways (e.g., SIRT1 activation) that facilitate membrane tethering [183]. These structural effects improve ER-mitochondria calcium exchange and support efficient OXPHOS, as observed in muscular dystrophy models [182].

4. Discussion

4.1. Study heterogeneity and lack of standardized quality control

The field of mitochondrial and EV transplantation is beginning to expand into various applications, which makes the generalization of results complex. Importantly, the reliability of the current body of evidence is constrained by heterogeneous study quality as evidence by the presence of high risk of bias in the preclinical studies (determined by the SYRCLE guidelines). Many studies lacked critical design features such as randomization, or blinded outcome assessment, leading to an unclear or high risk of bias. This distinction is crucial for guiding translational priorities and identifying research gaps. Moreover, the lack of consistency between studies is evident, both in terms of post-isolation quality controls for mitochondria or EV, as well as their subsequent internalization. Furthermore, several instances of over-interpretation have been noted, such as attributing the observed increase in fusion proteins to the fusion between exogenous and endogenous mitochondria, although this conclusion is not definitively supported by the data. This raises a major question: are the observed effects directly related to the function of exogenous mitochondria? In other words, is it necessary to transplant functional and viable mitochondria to achieve clinical and biological effects, as well as an impact on redox homeostasis? All the studies have demonstrated (or assumed in the absence of quality control) the use of functional mitochondria, with only four comparing the efficacy of functional and non-functional mitochondria [66,82,100,139]. Two studies reported beneficial effects only with functional mitochondria, while the other two found that the effects were similar regardless of whether the mitochondria were functional or not. Moreover, studies rarely assessed respiratory competence directly (e.g., oxygen consumption rate or RCR), limiting mechanistic insight. It is still unclear whether mitochondrial structural preservation, biochemical cargo, or actual respiratory activity drives therapeutic effects. Further research should prioritize standardized functional assessments (e.g., ATP generation, membrane potential, electron transport chain activity) to clarify the precise determinants of efficacy in mitochondrial-based therapies. The question remains open for debate and has also been investigated in vitro, with diverging results. However, if the use of non-functional mitochondria proves effective, it will be possible to identify the factors responsible for the observed effects, thus developing a potentially more robust and consistent therapeutic solution. Moreover, future studies on EV transplantation should include comprehensive characterization of the different markers of EV subtypes to clarify their origin—particularly to distinguish them from already known EV subtypes (small-EV, large-EV, exosomes, microvesicles, and apoptotic bodies), as well as from the emerging classes of mitochondrial-derived extracellular vesicles (MDEV) and mito-EV.

4.2. Perspectives: from pre-clinical to clinical application

Although numerous in vitro and preclinical studies have demonstrated the benefits of mitochondrial transplantation (particularly in the context of IRI), clinical data remains limited.

The first study, published by Emani et al., in 2017 [184], described an improvement in systolic dysfunction and a decrease or disappearance of regional segmental hypokinesia in 5 children, with no acute adverse events related to the injection of 10⁷ mitochondria into the myocardium. In 4 out of 5 patients, this led to the cessation of ECMO cannulation, which had not been possible before. Although encouraging, the lack of a control group in this study prevents any conclusions about whether the improvement was directly linked to mitochondrial administration. The same team subsequently published data in 2021 [185] on 24 patients (14 without mitochondrial administration and 10 with), confirming the absence of acute adverse events (e.g., arrhythmia, inflammation), and better successful and durable separation from ECMO following mitochondrial transplantation, fewer cardiovascular events, and enhanced ventricular strain. More recently, in 2024, another team conducted a Phase 1, prospective, parallel-group, randomized study to investigate the safety and efficacy of the transplantation of 1.2 × 10⁸ mitochondria via coronary artery infusion in ischemic heart disease in 30 patients [186]. While no acute adverse events were reported, efficacy was limited to an improvement in exercise capacity compared to non-treated patients. No significant difference was observed in the improvement of ejection fraction between the groups. There is hope that the three ongoing clinical trials investigating the effects of mitochondria on cardiac ischemia (NCT02851758, NCT05669144) and cerebral ischemia (NCT04998357) will provide positive results to improve patient care and further investigate the potential benefits of mitochondrial transplantation.

5. Conclusion

Mitochondrial and EV transplantation represents innovative and versatile strategies to counteract mitochondrial dysfunction across a wide spectrum of preclinical disease models. By modulating mitochondrial bioenergetics, redox signaling, structural integrity, and organelle homeostasis, these therapies offer a multifactorial mechanism of action that aligns with the complex pathophysiology of many disorders.

While preclinical evidence supports the therapeutic potential of mitochondrial and EV transplantation (particularly in models of IRI, neurodegeneration, and drug-induced damage) translational readiness remains limited. Moreover, preclinical studies mostly do not adhere to the SYRCLE guidelines, which further limits the strength of the evidence they provide. The existing clinical data are preliminary, derived from small patient cohorts with limited control groups and heterogeneous methodologies. Furthermore, several critical barriers must be addressed before broader clinical application can be envisioned. These include the lack of standardized protocols for mitochondrial or EV isolation and characterization, incomplete understanding of the mechanisms of action, and unresolved questions surrounding immunogenicity and long-term safety. Regulatory challenges also remain considerable, especially regarding product classification, quality control, and clinical-grade manufacturing. As such, mitochondrial-based therapies should still be considered experimental, and ongoing efforts must prioritize mechanistic clarity, methodological harmonization, and rigorously designed clinical trials. A cautious yet optimistic outlook is warranted as the field progresses toward therapeutic maturity.

Supplementary Material

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.freeradbiomed.2025.06.040.

Abbreviations:

8-OHDG

8-hydroxyl-2′-deoxyguanosine

ADSCS

adipose derived stem cells

Akt

protein kinase B

ALAT

alanine aminotransferase

AMPK

AMP-activated protein kinase

anti-ox

anti-oxidant

ARDS

acute respiratory distress syndrome

ASAT

aspartate aminotransferase

ATP

adenosine tri-phosphate

BHK21

baby hamster kidney cells-clone 21

BM-MSC

bone marrow mesenchymal stromal cell

CAT

catalase

CII-SDHB

complex II – succinate dehydrogenase subunit B

COX

cytochrome C oxidase

CPK

creatine phosphokinase

CRP

C-reactive protein

CXCL

chemokine (CXC motif) ligand

DCFHDA

2′,7′-dichlorodihydrofluorescein diacetate

DIC

disseminated intravascular coagulation

DRP1

dynamin-related protein 1

ECAR

extracellular acidification rate

eNOS

endothelial nitric oxide synthase

ERK

extracellular signal-regulated kinase

EV

extracellular vesicles

FABP

fatty acid-binding protein

FOXO

forkhead box O

GPX

glutathione peroxidase

GSH

reduced glutathione

HI

hypoxia-ischemia

HNE

4-hydroxy-2-nonenal

HO-1

heme oxygenase-1

IC

intracavernous

IDH2

isocitrate dehydrogenase 2

IFN

interferon

IgG

immunoglobulin G

IgM

immunoglobulin M

IL

interleukin

IM

intra-muscular

IP

intraperitoneal

IRI

ischemia-reperfusion injury

iNOS

inducible nitric oxide synthase

IV

intra-veinous/intravenous

JC1

5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide

LC3

microtubule-associated protein light chain 3

MAFbx

muscle atrophy F-box

MDA

malondialdehyde

MDMs

monocyte derived macrophages

Mfn

mitofusin

MITCO1

protein complex IV

MMP

mitochondrial membrane potential

MNP

microneedle patch

MSC

mesenchymal stromal cell

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromure

mTOR

mammalian target of rapamycin

MuRF-1

muscle RING finger protein 1

N2A

Neuro-2A

NDUF8

protein complex I

NFκB

nuclear factor-kappa B

NLRP3

NOD-, LRR- and pyrin domain-containing protein 3

NO

nitric oxide

NOX2

NADPH oxidase 2

NQO1

NAD(P)H:quinone oxidoreductase 1

NRF2

nuclear factor erythroid 2–related factor 2

NSE

neuron specific enolase

NT

nitrotyrosine

OCR

oxygen consumption rate

OGG1

8-oxoguanine DNA glycosylase 1

OPA1

optic atrophy 1

OXPHOS

oxidative phosphorylation

PAMPK-α

Thr172 phospho-AMPK-alpha-Threonine 172

Parkin

E3 ubiquitin ligase involved in mitophagy

PGC1-α

Pparg coactivator 1-alpha

PINK1

PTEN induced kinase 1

PRDX

peroxiredoxin

PTP

permeability transition pore

RCR

respiration control ratio

ROS

reactive oxygen species

RT1-m2

major histocompatibility complex gene

S100β

S100-beta protein

SDHB

protein complex II

Sdhaf

succinate dehydrogenase assembly factor

SIRT1

sirtuin 1

SOD

superoxide dismutase

SULT4A1

sulfotransferase 4A1

TBARS

thiobarbituric acid reactive substances

TFAM

mitochondrial transcription factor A

TGFβ

transforming growth factor beta

TLR4

Toll-like receptor 4

TNFα

tumor necrosis factor α

TOM

translocase of the outer membrane

Trolox

synthetic vitamin E analogue

UC-MSC

umbilical cord mesenchymal stromal cell

UQCRC2

protein complex III

VDAC

voltage-dependent anion channel

vATP5a

protein complex V

Data sharing

The datasets generated during the current study are available from the corresponding authors on reasonable request.