Mitochondrion-based organellar therapies for central nervous system diseases

Abstract

As most traditional drugs used to treat central nervous system (CNS) diseases have a single therapeutic target, many of them cannot treat complex diseases or diseases whose mechanism is unknown and cannot effectively reverse the root changes underlying CNS diseases. This raises the question of whether multiple functional components are involved in the complex pathological processes of CNS diseases. Organelles are the core functional units of cells, and the replacement of damaged organelles with healthy organelles allows the multitargeted and integrated modulation of cellular functions. The development of therapies that target independent functional units in the cell, specifically, organelle-based therapies, is rapidly progressing. This article comprehensively discusses the pathogenesis of mitochondrial homeostasis disorders, which involve mitochondria, one of the most important organelles in CNS diseases, and the machanisms of mitochondrion-based therapies, as well as current preclinical and clinical studies on the efficacy of therapies targeting mitochondrial to treat CNS diseases, to provide evidence for use of organelle-based treatment strategies in the future.

Graphical Abstract

Introduction

CNS diseases are among the leading causes of death and disability worldwide and are the third most common disease after cancer and cardiovascular diseases. Most CNS diseases are believed to be related to structural dysfunction of mitochondria. Mitochondria are organelles composed of a bilayer membrane-wrapped matrix that exist in most cells. The respiratory chain enzymes and adenosine triphosphatase (ATPase) complexes in mitochondria collaboratively participate in the formation of adenosine triphosphate (ATP) [1]. As energy factories of cells, mitochondria not only generate ATP for the brain and nervous system through oxidative phosphorylation, but also participate in various physiological processes, such as the maintenance of calcium ion balance and redox homeostasis. The CNS is highly dependent on energy metabolism to sustain the release of neurotransmitters and maintain ion gradients. The stable function of mitochondria is crucial for the homeostasis of the nervous system [2]. Mitochondrial dysfunction leads to decreased electron transport chain (ETC) enzyme activity, disrupted intracellular calcium homeostasis, reduced brain energy metabolism, increased mitochondrial reactive oxygen species (mtROS), and oxidative stress (OS) that damages lipids, proteins, and DNA, altering mitochondrial permeability. Oxidative damage to mitochondrial DNA (mtDNA) further exacerbates dysfunction, ultimately causing cell apoptosis or death [3,4,5,6]. Therefore, mitochondrial dysfunction is crucial in the pathogenesis of CNS diseases, involving a series of interconnected damage processes, from energy metabolism disorders to alterations in cellular signaling.

However, the current treatment regimens of CNS diseases, including conservative treatment [7] and supportive care, mainly involve the use of single-target drugs. For example, edaravone exerts its neuroprotective effects by inhibiting cellular oxidative damage, dopamine agonists relieve symptoms of Parkinson’s disease (PD) by replenishing dopamine levels in the brain, and antiepileptic drugs enhance GABA-mediated inhibition of excessive neuronal discharges. Although exogenous drugs can repair damaged cells and relieve symptoms, the response of brain cells may weaken over time, and normally functioning cells may become damaged and cause adverse effects. Although drug can directly or indirectly affect the behavior of cells such as neurons, restoring a damaged or dysfunctional cell population to a normal functional state is extremely difficult. Therefore, cell therapy, which replaces or eliminates damaged cells with new and normally functioning cells, is a promising treatment modality for CNS diseases [8]. In contrast to traditional drug therapy, cell therapy can replace damaged cells with normal cells, regulate multiple networks, and affect multiple targets, which has been proven to be clinically effective and safe. Cells be implanted into living organisms, function normally, circumvent the side effects of drugs to some extent, and secrete cytokines or other active substances, which can cure disease and results in favorable long-term outcomes. However, as cells are living functional units, transplantation has potential adverse effects, including immune rejection, tumorigenicity, and other unpredictable risks, limiting the application of cell therapy in clinical practice [9, 10]. Thus, relatively well-studied and multitargeted therapeutic strategies that are safe and effective are needed. Organelles, which are substructural units with specific functions, can be precisely targeted and are not living, and there are multiple types of organelles; therefore, organelle-based treatments that are more aggressive than cell therapy can be used. The use of organelle-based therapies not only avoids the obstacles of using living cells from the perspectives of quality control and clinical treatment but also maintain the specialized functions of cells.

Current organelle-based therapies target different types of organelles, such as mitochondria, nuclei, chloroplasts, lysosomes, and the endoplasmic reticulum (ER) [11]. Mitochondria are the most important functional targets for therapies affecting organelle transport [12]. Therefore, this review focuses on mitochondrial therapy, which involves intervening in diseases by transmitting healthy mitochondria. As early as 1982, Clark and Shay first proposed transplanting mitochondria of heterologous origin into another cell, thus opening the way for mitochondrial transplantation therapy to treat various diseases [13]. Healthy mitochondria are transferred from donor cells to recipient cells with mitochondrial dysfunction. The carrier containing these mitochondria is then disrupted, releasing intact mitochondria with their original structure, function, and mtDNA. These mitochondria participate in the fusion and fission of endogenous mitochondria and ATP generation [14, 15]. The dual membrane structure of mitochondria is crucial in protecting cells from harmful products like free radicals during these processes [16, 17]. Transferred mitochondria can bind to the recipient cell nucleus or other organelles, forming the mitochondrial information processing system (MIPS) for intercellular communication [18]. This increases mitochondrial-related biosynthesis in recipient cells, thereby affecting their biological functions. Practical theory has proven the feasibility of using organelle-level therapies, such as those involving mitochondria, to treat diseases. This provides potential therapeutic targets for CNS diseases, including cerebral ischemia/reperfusion injury (IRI), PD, Alzheimer’s disease (AD), and multiple sclerosis (MS) [19]. For example, the restoration of cellular function through the transfer of mitochondria from astrocytes and microglia to damaged neurons has also been studied [20]. Alternatively, reptide-1-enriched wild-type mitochondria isolated from parental heterozygous cells, have been transferred to heterozygous (MitoB2) and empty (Mito°) cells [21] to facilitate the treatment of myoclonic epilepsy in the context of ragged red fibers (MERRF) syndrome [20]. Thus, restoring the biological function of damaged mitochondria via mitochondrial transfer is a key research focus in treating neurological diseases.

This article reviews the advantages and limitations of using mitochondria as novel therapeutic targets for CNS diseases. This article summarizes the effects of mitochondrion-based therapies and related quality control requirements and discusses preclinical and clinical studies of mitochondrion-based treatments for cerebrovascular, degenerative, and other CNS diseases.

Problems caused by mitochondrial dysfunction in CNS diseases

Mitochondrial dysfunction mainly refers to energy metabolism disorders caused by mitochondrial ETC dysfunction leading to a decrease in ATP levels, triggering an increase in the production of mtROS, activation of related cell apoptosis and pyroptosis pathways, occurrence of the inflammatory reactions, and ATP dependent mitochondrial dynamics disorders. Mitochondrial dysfunction plays a crucial role in the development and negative consequences of CNS diseases. At present, various mitochondrial dysfunction has been found in CNS diseases including ischemic stroke (IS), AD, PD, MS, amyotrophic lateral sclerosis (ALS), and Huntington's disease (HD). Therefore, understanding the possible mechanisms of mitochondrial dysfunction in CNS diseases will provide a basis and theoretical support for the sue of mitochondrion-based therapies (Table 1).

Table 1 Problems caused by mitochondrial dysfunction in CNS diseases

Mitochondrial ETC dysfunction

Mitochondrial ETC dysfunction is a hallmark of CNS diseases. The five complexes in the ETC collaborate to maintain its normal function and ensure ATP production. Functional defects in any of these complexes can impair mitochondrial function. Notably, complex I (CI), the largest complex, has the most intricate subunit composition and is more prone to mutations. Severe ETC inhibition and a decrease in the mitochondrial membrane potential (Ψm) have been observed in patient samples or animal models of multiple CNS disease patients. The specific mechanism of ETC dysfunction in CNS diseases is still under investigation.

In AD, the hyperphosphorylated form of tau selectively damages CI, resulting in reduced ATP levels [22]. In postmortem analyses of AD patient brain specimens, decreased expression of mitochondrial complexes I-V (CI-V), was observed in the piriform and insular cortices [23]. Furthermore, as different mitochondrial complexes are inhibited in AD, the deposition of amyloid and tau proteins may also be a result of mitochondrial dysfunction. Mitochondrial CI dysfunction plays a major role in PD, and studies have shown that the CI inhibitor rotenone affects the function of cholinergic-like neurons to induce a neuropathological phenotype resembling Parkinson's disease dementia (PDD). Alpha-synuclein (α-Syn) is involved in many pathological processes of PD, such as suppression of mitochondrial permeability transport pore (mPTP) opening, inhibition of mitochondrial CI function, and a reduction in the Ψm [25]. Reduced respiratory chain activity has been reported in the spinal cord tissue of patients who died from ALS [31, 32]. Research suggests that this is related to the toxic effects of TAR DNA binding protein-43 (TDP-43) in ALS [33]. When the ETC is damaged and the oxygen and glucose supply is disrupted, cells fail to maintain a normal ion gradient [36], similar to what occurs as a result of inadequate perfusion in IS, which leads to impairment of mitochondrial oxidative phosphorylation and ATP synthesis; failure of sodium pumps, calcium pumps, and other ATP-dependent ion transporters; and inhibition of Ca²⁺ release, causing calcium overload and extracellular glutamate accumulation [37]. This leads to dysfunction of cells, including neurons, and other changes.

Specifically, when Ca²⁺ homeostasis is disturbed, TDP-43 disrupts the exchange of stored Ca²⁺ between the ER and mitochondria while reducing mitochondrial Ca²⁺ levels, which is one of the features of ALS [34]. Deficiency of mitochondrial Na⁺/Ca²⁺ exchangers (NCLXs) leads to chronic mitochondrial Ca²⁺ overload [20]. Disturbances in the shuttling of Ca²⁺ between the ER and mitochondria in AD lead to protein misfolding [26, 35], manifesting as memory loss and increased pathological changes in amyloid and tau protein levels [4]. In addition, Aβ proteins, including Aβ40 and Aβ42, form permeable pores on the Ca²⁺ membrane, which promotes the influx of Ca²⁺ and leads to excessive calcium uptake by mitochondria, followed by the production of free radicals and mitochondrial dysfunction [93], accelerating the progression of AD [5]. Impairment of Ca²⁺ uptake has also been observed in HD [24], whereas the loss of CI, is caused by depolarization and an imbalance in mitochondrial calcium homeostasis after inhibition of Ca²⁺ activity. This also occurs during the loss of substantia nigra neuronal in PD [27]. α-Syn aggregates bind to the sarcoendoplasmic reticulum calcium ATPase (SERCA) pump and activate it, leading to Ca²⁺ efflux from the cytoplasm to the ER and inducing mitochondrial degradation [28, 29]. In PD, mitochondrial function is also reflected by the expression of CIII, which reduces cytochrome C (CytC) levels through ubiquinone oxidation and induces the release of protons from the mitochondrial membrane to assist in the maintenance of the proton gradient. When electron transfer is inhibited, CIII participates in the production of superoxide anions [30]. An increase in glutamate levels in the intercellular space leads to excessive excitation of glutamate receptors on the postsynaptic membrane, promoting the activation of nucleotide-binding oligomeric domain-like receptor protein 3 (NLRP3) [38]. K⁺ efflux is a common trigger of NLPR3 activation, which induces excitotoxicity after ischemia and leads to cell death [39]. The main pathogenic changes in MS are inflammation and demyelination, which have significant effects on axonal mitochondria. Mitochondria reduce ATP production and increase reactive oxygen species (ROS) generation, whereas demyelinated axons increase energy demand. Thus, reversing manipulation of the Na⁺/Ca²⁺ pump further damages demyelinated axons [6]. Mahad hypothesized that demyelination leads to the transfer of dysfunctional mitochondria from the cell body to axons, promoting oxidative damage, increasing energy expenditure, and altering the calcium balance [40].

Swelling of the mitochondrial matrix may also occur because of an imbalance in Na⁺ and K⁺ concentrations. As swelling progresses, mitochondrial ridges decreases in number or disappear, and mitochondria may fail to function normally. Therefore, mitochondria transport normal proteins encoding ETC components and respiration into recipient cells, restoring electron transport in damaged mitochondria and the function of damaged nerve cells.

Production of mtros

ROS are natural byproducts of normal oxygen metabolism that are involved in cell differentiation, cell proliferation, and apoptosis, and ROS are maintained at normal levels in the form of superoxide radical anions and hydrogen peroxide. Homeostasis is associated with the pathological progression of many diseases. Excessive downstream effectors produced by ROS are triggered when ROS levels exceed the body’s scavenging capacity, leading to varying degrees of OS [94, 95]. For example, abnormal ROS signaling and excessive ROS production significantly promote muscle atrophy in ALS models mice [41]. Similarly, OS -induces the hyperphosphorylation of tau proteins through the inactivation of protein phosphatases 1 and 2A, leading to the abnormal aggregation of tau proteins [42]. Increased OS reportedly precedes the formation of Aβ plaques and neurofibrillary tangles [43]. MtROS, which are produced by mitochondria, differ from nicotinamide adenine dinucleotide phosphate (NADPH) oxidase-derived ROS, which cause tissue damage [96, 97], and are responsible for the production of mtROS through direct activation of the NLRP3 inflammasome, changes in mitochondrial membrane lipid peroxidation, and changes in the activity of enzymes related to ATP production. During IRI, accumulated succinic acid is rapidly oxidized by succinate dehydrogenase (SDH), while ATP synthesis has not yet returned to normal levels, leading to the accumulation of electrons and protons. This promotes CI reversal, generating mtROS, which cause mitochondrial dysfunction and further exacerbate brain tissue damage [5]. In the pathological processes of PD and AD, a deficiency in mitochondrial CI occurs due to the formation of calcium-permeable pores or the activation of voltage-gated channels (VGCC) on the plasma membrane by certain pathogenic markers [26, 98]. This results in an imbalance of mitochondrial Ca²⁺ homeostasis and promotes the further accumulation of mtROS [44, 45, 76, 99]. Not only that, increased iron levels and decreased ferritin levels in the brain lead to excessive mtROS production, which not only affects CI but also leads to the accumulation of misfolded proteins in the ER, further causing ER stress and triggering mitochondrial OS [25, 100]. Mitochondrial bioenergetics are affected by mtROS, and the generation of ROS can damage the substantia nigra, causing the oxidation of DNA, proteins, and lipids [46, 47]. Additionally, mitochondrial stress-induced protein oxidation is more severe than lipid oxidation at 24 h after spinal cord injury (SCI) [48]. In addition to primary injury caused by hematoma, secondary injury involves intracranial hemorrhage (ICH) and causes mitochondrial OS, inflammation, and excitotoxicity [49, 50]. The level of mtROS increases with decreasing levels of antioxidant enzymes [101]. With the opening of the mPTP and the dysfunction of ion channels, the release of mtROS is further induced. The elevated level of mtROS continuously induces mPTP opening through positive feedback, further aggravating mitochondrial damage. Additionally, matrix metalloproteinases (MMPs), which degrade the extracellular matrix, can be activated by ROS, disrupting endothelial tight junctions and altering the expression of adherens junction proteins and blood–brain barrier (BBB) transporters by degrading the basal BBB layer. Consequently, the integrity of the BBB is destroyed, leading to brain edema and severe injury [102, 103].

Mitochondria are potential central regulators of cell apoptosis and pyroptosis

Under physiological conditions, free Ca²⁺ and bound calcium concentrations in mitochondria are in dynamic equilibrium, and Ca²⁺ can reduce ROS leakage from mitochondrial CI and CIII. However, under pathological conditions, this dynamic balance is disrupted. Excessive Ca²⁺ transfer from the ER to the mitochondria leads to mitochondrial Ca²⁺ overload and increased mtROS production. Together, this high-calcium and OS environment ensures the sustained opening of the mPTP, leading to a decrease in the Ψm, mitochondrial depolarization, and the activation of a series of apoptotic pathways [104, 105]. For example, as cells emit apoptotic signals, mitochondrial outer membrane permeabilization (MOMP), which is associated with oligomers formed by B-cell lymphoma-2 (BCL-2) antagonist/killer factor 1 (BAK1) and BCL-2-associated X protein (BAX) occurs. This leads to the formation of many small pores in the mitochondrial membrane and induces the release of BCL-2, BAX, p53, and apoptosis-promoting proteins such as CytC or mtDNA into the cytoplasm, which participate in mPTP-driven apoptosis and necrosis through voltage-dependent anion channels (VDACs), resulting in mitochondrial dysfunction, such as increased OS and respiratory chain damage [106, 107]. One study reported that in PD, α-Syn regulates the release of substances such as CytC and BCL2 family proteins by increasing the expression of the mitochondrial encoded CytC oxidase II (MT-CO2) gene, thereby impairing the bioenergetics of dopaminergic neuron [29]. In ALS, a significant increase in ROS levels leads to alterations in mitochondrial axonal transport, structure, and dynamics, ultimately inducing cell apoptosis [51]. Apoptotic bodies, which comprise apoptotic peptidase activator 1 (APAF1) and caspase 9, affect mitochondrial function [108]. Therefore, inner mitochondrial membrane (IMM) and mPTP dysfunction results in the release of mtDNA [109]. Moreover, during programmed cell death caused by inflammasomes, inflammatory caspase-1 and caspase-4/5/11 are activated by pathogenic invasion via the classic pathway. Classic inflammasome pathways are then activated, aggravating brain injury in patients with stroke [52]. After ICH, the NLRP3-pyroptosis signaling pathway in the ER is activated, leading to brain injury [53]. These inflammatory processes specifically disrupt intermediate junctions between the amino terminus and carboxy terminus of gasdermin D (GSDMD) (gsdmin-n and gsdmin-c). GSDMD aggregates on the cell surface through the release of the N-terminal domain, leading to the formation of pores through which extracellular water molecules flow, resulting in swelling, rupture, cytolysis, and the release of intracellular inflammatory factors, such as IL-6 and TNF, ultimately leading to inflammatory cell death or pyroptosis [54, 110, 111]. Thus, the NLRP3 inflammasome may be a molecular target for the activation of pathways upstream of pyroptosis in IS. In AD, Aβ fibrils induce the formation of the NLRP1 inflammasome in neurons, thereby inducing caspase-1-dependent cell death [55]. Caspase-1 can also be detected in microglia in demyelinating lesions and normal white matter in experimental autoimmune encephalomyelitis (EAE) animal models [56]. Apoptosis is involved in other CNS diseases such as infections.

Mitochondrion-mediated regulation of the inflammatory response

The pathogenesis of several CNS diseases involves inflammatory responses caused by damage-associated molecular patterns (DAMPs). Owing to the separation of mitochondrial membranes from outer membranes in tissues, DAMPs are normally not recognized by pattern recognition receptors (PRRs) [112]. However, DAMPs can be activated by PRRs under pathological conditions, triggering an inflammatory response and increased permeability to various cellular components [113]. When inflammation is severe and prolonged, or when the inflammatory response is beyond the homeostatic control of the adaptive immune response, severe CNS diseases may develop. Some markers, including NLRP and glycogen synthase kinase 3 (GSK3), mediate the inflammatory response mentioned above. The NLRP inflammasome is the main initiator of neuroinflammatory responses in CNS diseases. Currently, there are five recognized types of inflammasomes (the NLRP1, NLRP3, NLRP4, IPAF, and absent in melanoma 2 (AIM2) inflammasome), among which the NLRP3 inflammasome has been studied extensively. NLRP3 is an intracellular receptor that activates the NF-κβ pathway to promote the expression of inflammasome-related proteins and to recruit caspase-1 by triggering the aggregation of inflammasome-sensing proteins and inflammasome-linker proteins.

Following mitochondrial dysfunction [114], hyperactivated inflammasomes regulate inflammation, and the inflammatory activation of microglial NLRP3 is a key factor in neuroinflammation in individuals with ALS [51]. The NLRP3 inflammasome converts precursor caspase to caspase-1 and releases IL-1β, IL-6, and other substances into the extracellular environment, leading to cell death [115]. The NLRP3 inflammasome is the most common inflammasome involved in neurodegenerative diseases, especially AD [58]. The Aβ or tau protein can regulate the cleavage and maturation of proinflammatory cytokines, including IL-1β and IL18, which can exacerbate inflammation in AD [57]. Multiple studies have reported high levels of inflammatory cytokine expression in the cerebrospinal fluid, cortex, and postmortem brain tissue of AD patients [59, 60, 116].

Similarly, in PD, dying neurons release synucleins that activate the NLRP3 inflammasome in human blood mononuclear cells [63, 64]. A similar phenomenon occurs in IS, as middle cerebral artery thrombosis activates the NLRP1 inflammasome [67]. This inflammasome may also be activated by the induction of harmful signaling pathways or mtROS production due to the accumulation of intracellular Ca²⁺, which promotes the maturation of pro-caspase-1. Caspase-1 cleaves pro-IL-1β and pro-IL-18, leading to inflammatory cell death [68]. Research has shown that the NLRP3 inflammasome is activated in a GSK3-dependent manner. As mentioned earlier, upregulation of IL-18 by NLRP3 inflammasome activation can increase the level of GSK3, thereby contributing to tau protein hyperphosphorylation and exacerbating neuroinflammation in AD [61, 62].

Inhibiting NLRP3 inflammation downregulates GSK3, indicating that NLRP3 and GSK3 are closely related [117]. There are two isoforms of GSK3, GSK-3α and GSK-3β, which share a highly conserved catalytic domain but are encoded by different genes. Among them, GSK-3β is a serine/threonine protein kinase highly expressed in the brain that is involved in the pathogenesis of various CNS diseases. For example, it has been confirmed as one of several kinases related to abnormal posttranslational modifications (PTMs) of key proteins known to cause and be affected in PD [65]. Along with the activation of GSK-3β, proteins in the typical wingless/integrated (Wnt) signaling pathway upregulate death signaling proteins such as caspase3 and CytC through dynamic interactions, promoting dopaminergic neuron degeneration and the activation of glial cells [66]. Similarly, activation of the Wnt/GSK3β/β-catenin signaling pathway enhances inflammation, OS, and neuronal apoptosis in IS model rats [69,70,71]. In epilepsy, an increase or decrease in GSK-3β activity can exacerbate cell death induced by epileptic seizures, which may be due to the differential expression of downstream targets of activated GSK-3β [72, 73]. In addition to triggering the Wnt signaling pathway, as mentioned above, GSK-3β also triggers the protein kinase B (Akt) pathway. Traumatic brain injury (TBI) reduces p-Akt levels and increases GSK-3β, thus increasing the release of caspase3 and other substances, leading to neuronal death [74]. SCI reduces the ratio of phosphorylated GSK-3β to total GSK-3β (p-GSK-3β/t-GSK-3β ratio) and increases the number of apoptotic cells in the spinal dorsal horn [75].

Owing to the difference in pore size between the outer membrane and mPTP and the difference in the number of BAX-BAK1 oligomers, mtDNA is more likely to be released through the mPTP than through the outer membrane to activate cyclic GMP AMP synthase (cGAS), which responds to stimulator of interferon response cGAMP interactor 1 (STING1) and worsens the inflammatory response [118, 119]. MtDNA can increase inflammasome activation and initiate a positive feedback loop to regulate mitochondrial function [120]. However, the specific interactions between mtDNA and inflammasomes are not fully understood. A recent review reported that mtDNA and other mitochondrial components, such as ROS, can trigger PRR-mediated inflammation through different cGAS pathways and inflammasomes. The naked and protein-bound mtDNA sensors include toll-like receptor 9 (TLR9) and receptor for advanced glycation end products (RAGE) [121, 122]. For instance, activated platelets release respiratory mitochondria containing mtDNA, both within membrane-encapsulated particles and as free organelles [123]. Moreover, researchers have proposed that when mtDNA replication cannot continue, it leads to the accumulation of mitochondrial and nuclear ribosomes, which are then excreted by cells [124]. However, when excess ribosomes are present, they can also cause the release of mtDNA into cells, which can trigger a strong inflammatory response [125], potentially leading to irreversible opening of the mPTP and disruption of MOM integrity [126]. These views suggest that the release of mtDNA associated with tissue damage appears to be pro-inflammatory, thereby regulating the activation state of the immune system Fig. 1.

Fig. 1

Mitochondrial dysfunction in ETC dysfunction, production of mtROS, cell apoptosis and pyroptosis, inflammatory response. a When oxygen and glucose levels are low, the normal ion gradients cannot be maintained, mitochondrial oxidative phosphorylation is impaired, ATP synthesis is reduced, ATP-dependent ion transporters such as sodium pumps and calcium pumps are disabled, and Ca²⁺ excretion is blocked, leading to calcium overload and excessive extracellular glutamate levels. Moreover, owing to dysfunction of the sodium and potassium pumps, mitochondria cannot function properly. b When ROS levels exceed their scavenging capacity, downstream effectors generated by excess ROS are triggered, resulting in different degrees of OS, especially mtROS generation. With the opening of the mitochondrial mPTP and the disturbance of ion channels, the release of mtROS is further induced. The high level of mtROS continues to activate mPTP opening through positive feedback, further aggravating mitochondrial damage. c Elevated levels of mtROS trigger cellular stress, disrupting mitochondrial homeostasis. The proapoptotic proteins BAX and BAK binds each other after activation, leading to MOMP. This results in the formation of numerous small pores in the mitochondrial membrane, releasing the pro-apoptotic factor CytC or mtDNA into the cytoplasm. CytC binds to Apaf1, forming “apoptotic bodies” that activate caspase-9 and initiate a cascade of events. MtDNA is recognized by cGAS, which activates STING to induce an inflammatory response. d cGAS-STING signaling triggers an inflammatory response, which subsequently induces inflammasome and pyroptosis

Mitochondrial dynamics

Mitochondrial quality control is a prerequisite for ensuring cellular metabolism. Mitochondrial dynamics, including biogenesis, continuous fusion, and division, are key steps in cell quality control [127].

Within hours of CNS injury, mitochondrial biogenesis increases mtDNA levels in response to repair mechanisms [128]. A series of transcription factors and transcriptional coactivators, including peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α), which is a potent activator involved in mitochondrial biosynthesis, are involved in this process [129]. It plays a crucial role in connecting mitochondria and the nucleus by activating downstream nuclear respiratory factor 1 (NRF1) and mitochondrial transcription factor A (TFAM) and is considered a potential new therapeutic target for various CNS diseases [130]. The mRNA level of the nuclear-encoded gene RC, which responds to PGC-1α, is reduced in the dopaminergic neurons of PD patients, resulting in decreased PGC-1α protein expression and disruption of mitochondrial ultrastructure in dopaminergic neurons [76, 131]. A decrease in PGC-1α expression is accompanied by a decrease in mitochondrial antioxidant and uncoupling protein (UCP) expression, which can lead to functional damage to mitochondria in gray matter in MS [79]. After IS, PGC-1α and transcription factor estrogen receptor alpha (ERR α) in microglia synergistically regulate unc-51 like autophagy activated kinase 1 (ULK1) to promote autophagy and mitosis. Phosphatase and tensin homologs (PTEN) as well as PTEN-induced putative kinase 1 (PINK1) also participate in this process [132], modulating the mitochondrial autophagy process in microglia, thereby decreasing ROS production and inflammasome activation, inhibiting microglia-mediated neuroinflammatory responses, and ultimately mitigating ischemic brain injury [80]. PGC-1α acts as a transcriptional activator to activates a substrate protein (sirtuin3, SIRT3), which encodes a mitochondrial protein and is responsible for posttranscriptional protein modifications [133]. In AD, ApoE-4 reduces PGC-1α levels, SIRT3 damages mitochondria by impairing mitochondrial biogenesis or directly or indirectly reducing the levels of fusion and mitogenic proteins [83], and tau inhibits parkin translocation to mitochondria, consequently impairing mitochondrial phagocytosis [84, 85]. In addition, Scadden et al. reported that malic enzyme (ME2) can link fumaric acid, an intermediate of the tricarboxylic acid cycle, to mitochondrial biogenesis. Deoxyuracil nucleoside 5'-triphosphate nucleotide hydrolase (deoxyuridine triphosphatase, DUT) is activated by the ME2 dimer, which promotes an increase in mtDNA levels, and then fumarate dimers induce the binding of ME2 monomers to the mitochondrial ribosomal protein L45, releasing mtDNA-encoded proteins and jointly activating mitochondrial production of energy [134]. Mitochondrial dynamics, including biogenesis and continuous fusion and division, is key for controlling cellular quality. To maintain normal morphology and function, mitochondrial fusion and fission occur continuously and simultaneously. Dysfunctional mitochondria fuse with normally functioning mitochondria to form healthy mitochondria, and mitofusions 1 and 2 (Mfn1 and Mfn2, respectively) in the MOM and optic atrophy 1 (Opa1) participate in this process [86]. Research has shown that 72 h after SAH, the expression of Mfn1/2 and Opa1 decreases [90, 91]. When fusion occurs but normal function cannot be maintained, mitochondria undergo division to split dysfunctional mitochondria. Mitochondrial fission is regulated by dynamic-related protein 1 (Drp1). By recruiting cytokinesis factors, Drp1 induces mitochondrial degradation and damage and plays a role in stabilizing mitochondria. Under OS, Drp1 is destroyed, leading to further aggravation of toxicity [135]. The expression of Drp1 and Fission 1 (Fis1) is increased in ALS mice models, whereas the expression of Mfn1 and Opa1 gradually decreases [92]; a similar phenomenon has been observed in the posterior hippocampus of patients with AD [136]. A decrease in Drp1 expression is also observed in the fibroblasts of patients with sporadic AD [137]. The tau protein binds and stabilizes actin [87], thereby blocking the transport of DRP1 to mitochondria, which is crucial for subsequent mitochondrial fission [88, 89]. This imbalance in mitochondrial fusion and fission leads to increased mitochondrial fragmentation in the brains of patients with AD. Overexpression of Drp1 and loss of Mfn2 induced by excess α-Syn oligomers and fibroblasts in PD patients lead to impaired mitochondrial transport in the substantia nigra and striatal dopaminergic neurons, resulting in severe movement disorders [77, 78]. Together, they determine whether mitochondria survive or die by complementary salvage (in which mitochondria remain useful after division) or autophagic clearance (in which mitochondria lose their function owing to a decreased membrane potential) [138]. Mitophagy is one of the major processes for controlling mitochondrial quality, as it selectively removes dysfunctional mitochondria from the cytoplasm, thereby maintaining mitochondrial function [139]. Studies have shown that the reestablishment of blood flow after cerebral ischemia can activate mitochondrial autophagy and reduce neuronal damage [140]. Through various experimental methods, it was discovered that axonal mitochondria in ischemic neurons are transported back to the neuronal cell body and undergo autophagy rather than being cleared by autophagy in situ in daughter axons [81]. The specific promotion of mitochondrial reverse transport can increase the mitochondrial mass in ischemic neurons, reduce apoptosis, and exert a neuroprotective effect against cerebral ischemia by activating mitochondrial autophagy. Liu et al. reported that there is a synergistic effect between mitochondrial biogenesis and mitophagy and that abnormal mitophagy can negatively affect mitochondrial biogenesis. Specifically, the PGC-1α–NRFG1–FUNDC1 pathway plays a crucial role in maintaining the stability of mitochondrial quality and quantity [82] Fig. 2.

Fig. 2

Mitochondrial dynamics in CNS diseases. Mitochondrial dynamics, including biogenesis and continuous fusion and fission, is a key step in cell quality control. To maintain normal morphology and function, mitochondrial fusion and fission occur continuously and simultaneously in a dynamic process. Mitochondrial fusion is regulated by Mfn2 and OPA1. Dysfunctional mitochondria fuse with functional mitochondria to form healthy mitochondria. However, when fusion occurs but normal function cannot be maintained, mitochondria undergo fission to split dysfunctional mitochondria. Mitochondrial fission is regulated by Drp1. Abnormal mitochondrial autophagy can provide negative feedback and inhibit mitochondrial biogenesis

Technical quality control and characteristics of mitochondrion-based therapies

Organelle-based therapies are effective in cells, animals, and humans, and their effects largely depend on the success of mitochondrial delivery. This section introduces technical quality control considerations and the characteristics of mitochondria from different sources, mitochondrial quality, mitochondrial pathways, biological mechanisms, mitochondrial tags, and methods for monitoring mitochondrial localization, which will help us better understand the mechanisms underlying the effects of organelle-based therapy to aid future research.

Source of transplanted mitochondria

Transplanted mitochondria can be obtained from various cell lines. Classify the origin of mitochondria based on whether they are transplanted endogenously or exogenously. Endogenous transplantation involves the transfer of mitochondria via other carriers, such as cells. Exogenous transplantation involves the direct transfer of mitochondria, including various cell lines and stem cells like mesenchymal stem cells (MSCs) and induced pluripotent stem cells (iPSCs), as well as tissues such as muscle, placenta, brain, and liver.

When neurons, astrocytes, or exosomes derived from astrocytes are cocultured with other nerve cells, these mitochondria-containing cell carriers release mitochondria for transfer to recipient nerve cells, including damaged or healthy mitochondria. For example, endogenous neurons release damaged mitochondria and transfer them to astrocytes, which engulf and degrade them; this process may also play an important role in the response to ischemic injury [141]. Additionally, astrocytes can release functional mitochondria to neurons through CD38-mediated mechanisms to increase ATP levels and restore cell viability after IS [142]. Chen et al. demonstrated that the transfer of mitochondria from astrocytes to neurons protects neurons from TBI-like injury in vitro [143]. Brain endothelial cells (BECs)-derived vesicles spontaneously release healthy mitochondria to treat BEC damage, increasing BEC metabolic function and tight junction integrity [144]. These mitochondrial sources are carried by cells as carriers and exert their effects through endogenous transfer mechanisms. Mitochondria separated from cells or tissues of different origins through exogenous transfer pathways are then transported to damaged cells, internalized, and integrated into the mitochondrial network in the recipient cell to restore aerobic respiration and energy metabolism, including increasing the membrane potential and oxygen consumption during injury repair [145,146,147,148]. Mitochondria derived from HeLa cells [149, 150] and human hepatocellular carcinoma (HepG2) cells [110], were used for transplantation therapy in AD and PD model mice, and improvements in bioenergetics and behavior were observed in these mice. When mitochondria are transferred to mtDNA-deficient cells, the expression of mtDNA-encoded proteins and mitochondrial function are restored [151]. The transfer of donor mitochondria can decrease apoptosis in recipient cells by regulating the imbalance between BAX and BCL-2 levels and decreasing the expression of caspase-3, preserving cellular bioenergetics [152]. Additionally, the placenta is considered a source for cell therapy even after cryopreservation, and healthy mitochondria with a normal membrane potential and ATP levels extracted from frozen mouse embryos that were transplanted into middle cerebral artery occlusion (MCAO) model mice were found to significantly reduce the ischemic area [153]. Transplantation of mitochondria isolated from the brain tissues into PC12 cells subjected to oxygen–glucose deprivation (OGD) or TBI mice models was shown to improve the biological function of the recipient cells and promote angiogenesis [154]. Mitochondria isolated from the gastrocnemius muscle were transplanted into ischemic mice, and these mitochondria were found to penetrate the damaged BBB and enter neural cells in the brain, increasing ATP synthesis [155]. These studies suggest that mitochondria from various sources can operate via endogenous transfer between neighboring cells and exogenous transplantation following separation from the parent cell. This finding also indicates, to some extent, that there is no consensus on the ideal source of mitochondria for transplantation, which will lead to a lack of unified standards for evaluating the comprehensive effects of mitochondrial transplantation in future research. Regarding the possibility of future clinical application of mitochondrial transplantation, homologous endothelium-derived extracellular vesicles (EVs) can transfer more mitochondria to recipient cells than can heterotype macrophage-derived EVs [156]. Thus, autologous transplantation can reduce inflammatory responses and prevent immune rejection; however, this is not to the case in diseases caused by congenital mitochondrial defects. Therefore, immune responses induced by allogeneic mitochondrial transplantation require further investigation.

Quality of isolated mitochondria

The key to mitochondrial transfer is to obtain functional mitochondria, while maintaining the integrity of the outer and inner membranes during mitochondrial extraction to ensure mitochondrial quality. Moreover, ideal storage strategies are especially important for maintaining the optimal function of the mitochondria. Isolated mitochondria may be affected by the freezing rate, thawing rate, storage temperature, and number of freeze–thaw cycles, which making mitochondrial storage challenging.

Functional identification of isolated mitochondria

Ensuring the functional integrity of isolated mitochondria is the key to successful transplantation; therefore, functional identification of isolated mitochondria is necessary. The biological properties of mitochondria need to be comprehensively evaluated by assessing their number, function, purity, and localization through different analyses [157, 158], including analyses of morphology, the Ψm, ATP levels, the mtDNA copy number, respiratory activity, and protein expression [159,160,161,162,163], to achieve maximum effective control of mitochondrial quality. For morphological evaluation, transmission electron microscopy (TEM) is commonly used to assess mitochondrial structure and purity. TEM can also be used to measure the Ψm, ATP production, permeability, the respiration rate, the extracellular acidification rate, and mitochondrial enzyme activity to determine mitochondrial function and quality [145, 164], but this method is not suitable for rapid evaluation. Considering the low biological activity of isolated mitochondria, although studies have shown that the lifespan of mitochondria in the cytoplasm is 2–4 w [165], it is necessary to explore the duration of mitochondrial functional activity in the host. To make mitochondrion-based therapies more widely applicable, there is an urgent need to establish standardized methods for mitochondrial extraction and purification. JC-1 staining is commonly used to measure the Ψm to evaluate the integrity of isolated mitochondria, and changes in the color of fluorescence, which are measured as the red/green fluorescence intensity ratio, reflect changes in the Ψm or the degree of mitochondrial depolarization [163, 166]. In addition, tetramethylrhodamine ethyl ester (TMRE) and tetramethylrhodamine methyl ester (TMRM) are widely used for flow cytometry analysis of mPTP opening [167]. Multiple probes can be used to detect mPTP opening, which can be used to accurately evaluate the membrane integrity of mitochondria. Moreover, fluorescence can be generated by the catalysis of luciferin by luciferase to assess the level of ATP production [153, 168]. MtDNA abundance can be evaluated by measuring the mtDNA/nuclear DNA ratio (as a proxy for mitochondrial copy number and function) [163], and qPCR can also be used to quantify mtDNA levels. A standard curve can be established on the basis of dilutions of an ND1 (mitochondrial gene)-expressing plasmid, and the copy number of mtDNA can be calculated [165]. To measure activity, the function of extracellular mitochondria can be determined by measuring oxygen consumption [168], and mitochondrial energy metabolism can be evaluated by assessing mitochondrial respiratory chain complex activity [169]. Furthermore, the presence of extracellular mitochondria can be confirmed by measuring the expression of the outer mitochondrial membrane 40 (TOMM40) protein by western blotting (WB) [168]. These methods can be used to systematically evaluate the integrity of transplanted mitochondria from multiple aspects to ensure successful transplantation.

Storage of isolated mitochondria

Isolated mitochondria can be stored for up to 72 h at 25 °C [170], whereas the outer and inner membranes of mitochondria may be damaged and normal mitochondrial function may be lost when they are stored on ice for 1–2 h [171]. Therefore, to ensure the quality of transplanted mitochondria, isolated methods and delivery time windows should be strictly controlled. To meet this requirement, differential centrifugation can be used to efficiently and quickly isolate mitochondria, but the optimal parameters depend on the differential settling velocity between mitochondria and the nucleus in a particular medium. Zhang isolated autologous mitochondria in just 40 min and achieved favorable therapeutic results in rats with MCAO [169]. The method that was used is simple but not very effective, and it cannot be used to obtain large quantities of purified mitochondria at one time. For procedures requiring many mitochondria, further improvements in long-term storage strategies are needed to maximize clinical benefits. To extend the storage time of mitochondria, permeable cryoprotectants such as dimethyl sulfoxide (DMSO) and glycerol or the impermeable cryoprotectant trehalose are added to isolated mitochondria [172], as they can maintain mitochondrial integrity after freeze‒thaw cycles [173]. These cryoprotectants form hydrogen bonds with water molecules, which bind water and lower its freezing point, effectively preventing the formation of ice crystals and providing protection. Nukala et al. added 10% DMSO to isolated and purified rat cortical mitochondria to demonstrate successful cryopreservation and mitochondrial recovery. The mitochondria were cooled at a stable rate of approximately 1 °C/min and stored for long periods at − 80 °C [174]. Studies have shown that storing frozen mitochondria in buffer solutions containing sugar and trehalose maintains the integrity of the MOM and preserves their biological functions, including their ultrastructure, ATP-synthesizing ability, and transmembrane potential [175]. However, recent research has shown the opposite trend. Nakamura successfully isolated healthy mitochondria from mouse embryonic tissue and stored them at − 80 °C for one week [153]. Huang transplanted frozen mitochondria into ischemia model rats and found that mitochondrial activity was not restored even on the 7th day after reperfusion but was restored in the 2nd week [159]. Chou et al. reported that the JC-1 ratio of mitochondria after a freeze‒thaw cycle was close to the ratio before freezing‒thawing [176]. However, Wechsler et al. reported that mitochondrial function decreased to 10%–15% of normal levels after freezing [177]. In summary, these results provide a certain degree of experimental and theoretical evidence for the proper long-term storage of mitochondria, and it is necessary to explore better methods for ensuring mitochondrial quality in the future.

Pathway of transplanted mitochondria

The selection of carriers for mitochondrial transplantation is crucial to easily transport mitochondria to damaged sites and effectively reduce the likelihood of mitochondrial inactivation [36, 178]. Delivery methods for in vitro mitochondrial transplantation include cocultivation, magnetic transfer, photothermal nanoblade, and fluidic force microscopy (FluidFM), whereas in vivo methods include direct injection (intravenous, arterial, or stereotactic) and nasal administration.

Mitochondrial transplantation in vitro

Cocultivation

The cocultivation method involves the spontaneous transfer of mitochondria between cells via coculture of healthy mitochondria with recipient cells. Babenko et al. reported that when multipotent mesenchymal stromal cells (MMSCs) were cocultured with neural cells (including neurons and astrocytes), mitochondria were spontaneously transferred from the MMSCs to the neural cells and performed regulatory functions [179], and it was found that this transfer process was unidirectional. In 2018, researchers found that when ROS levels in damaged cells were elevated, astrocytes exhibited intracellular transfer in a similar but more efficient manner, and they also demonstrated that this finely regulated transfer process was mediated by mitochondrial Rho GTPase 1 (MIRO1) rather than ordinary cell diffusion [180]. Similarly, when bone marrow mesenchymal stem cells (BMSCs) or human brain microvascular endothelial cells (hCMEC/D3EV) were cocultured with neurons or BECs subjected to OGD, mitochondrial transfer from the BMSCs to the neurons was observed, resulting in improved mitochondrial bioenergetics [156, 181]. The cocultivation method is the most common and simplest technique among all in vitro mitochondrial delivery methods; however, considering that the main pathway by which mitochondria enter cells is actin-dependent endocytosis, the efficiency of mitochondrial transfer during coculture is relatively low because of the limitations of this cellular endocytosis.

Photothermal nanoblade method

By rapidly heating a miniature straw coated with peptides with laser pulses, Ting Hsiang Wu developed a new method using a photothermal nanoblade to transfer mitochondria to recipient cells and found that it alleviated pyrimidine deficiency-related phenotypes and improved respiratory function in P⁰ cells lacking mtDNA. Unlike cell fusion, this method does not require endocytosis and can reduce the transfer of other cytoplasmic biomolecules, such as miRNAs and metabolites, that affect cellular functions. The photothermal nanoblade method optimizes laser energy, improves the membrane opening efficiency and increases the viability of different cell types, making it a highly important mitochondrial transfer technology [182].

MitoPunch

MitoPunch is a simple, high-throughput device for transferring isolated mitochondria and mtDNA to recipient cells and inducing the transcription of specific mitochondrial and nuclear genes in composite cells [183, 184]. This method can achieve consistent transfer of many mitochondria isolated from different donor cell types to multiple recipient cell types, which is consistent with the need for mitochondria from diverse sources. Unlike the photothermal nanoblade method, MitoPunch does not require complex lasers and optical systems. Instead, pressure is used to push an isolated mitochondrial suspension through a porous membrane covering the cells; through this pressure gradient, mitochondria can directly enter the receptor cells. This is a major improvement over existing mitochondrial transfer technology, as it enables the simultaneous transfer of mitochondria to 100,000 or more receptor cells and the control of cellular functions at the gene level [185]. This is undoubtedly the fastest and most efficient method for transferring and isolating mitochondria in vitro.

Magnetic transfer

Magnetic transfer is an efficient way for direct mitochondria transfer into cells that utilizes anti outer mitochondrial membrane 22 (TOM22) magnetic beads to label TOM22 fragments and bind to them, greatly improving the purity of isolated mitochondria [14, 186]. However, owing to magnetic attraction, the structure of the outer mitochondrial membrane (OMM) may be damaged, making it difficult for isolated mitochondria to enter cells. Moreover, not only healthy functional mitochondria, but also dysfunctional mitochondria labeled with anti-TO22 magnetic beads may be transferred. Therefore, using this method to extract and purify mitochondria may result in loss of function. This technology still needs further development to determine how to target mitochondria while avoiding the negative effects of magnets.

FluidFM

Vorholt et al. extracted intact mitochondria from the plasma membrane via the multifunctional probe FluidFM and single-cell microscopy and transplanted them into recipient cells, allowing them to fuse with the mitochondrial network in the recipient cells for more than 16 h; this resulted in the inhibition of replication of donor mtDNA in the recipient cells [187]. This relatively novel method for delivering mitochondria, which can achieve rapid, accurate, and effective removal and delivery of organelles without compromising cell viability, is more efficient than the photothermal nanoblade method. However, the throughput of this method is reduced to a certain extent, and it has strict technical requirements.

Mitochondrial transplantation in vivo

Direct injection

As mitochondria are small and do not interfere with oxygen transport, they can be transferred directly through injection. The routes include the intravenous injection (IV), arterial injection (IA), and intra-cerebroventricular injection (ICV). Compared with ICV, IV and IA are more effective in treating multiple mitochondrial disorders, and cause fewer negative effects; thus, they are more conducive to future clinical application. IV transplantation of mitochondria is a noninvasive transplantation method in which isolated mitochondria or cells containing healthy mitochondria are injected via the tail vein. The quantity of mitochondria cannot be accurately controlled when IV techniques are used, but the number of mitochondria administered depends on whether rats or mice are used. In general, 3 × 10^{^6} cells (including MMSCs containing mitochondria) [179, 180] or 100 µg isolated mitochondria are administered intravenously to rats [188]. However, 10^{^7} cells or 100–200 µg mitochondria can be administered intravenously to mice [144, 149, 150, 153, 163, 189], or mitochondria can be administered to mice by IV at a dose of 0.5–1 mg/kg [110, 166, 190]. This transplantation method is relatively simple, but mitochondria injected intravenously into the body are not widely distributed. Blood circulation may cause the number of mitochondria remaining in tissues other than the brain, such as lung tissue, to far exceed the number entering the brain. This, to some extent, weakens the therapeutic effect of mitochondrial transplantation. Compared with IV, IA of mitochondria requires greater technical skill by the operator, as it requires intubation. According to empirical evidence, 750 µg mitochondria should be administered arterially to rats [159] while 2 × 10^{^7} mitochondria should be administered to mice by IA [165]. Owing to the specificity of this technique, it is possible to deliver mitochondria accurately to the target site to a certain extent and avoid mitochondrial waste. However, this method is not very safe, as it can cause mitochondrial aggregation in arteries, which can cause embolism and cerebral ischemia. ICV involves the direct injection of mitochondria into the lateral ventricle after stereotaxic localization. Through CSF, it spreads to the lesion site or multiple regions of the CNS. According to empirical evidence, 5 × 10^{^6}–3 × 10^{^7} or 50 µg mitochondria should be delivered to rats via the lateral ventricle [152, 169, 191], and 1 × 10^{^6} mitochondria should be delivered to mice [192]. This transplantation method can directly achieve a high concentration of mitochondria in the injury site, but it is an invasive technique that not only requires particular equipment and a highly skilled operator but also is limited by infection risk and low reproducibility.

Nasal delivery

Nasal delivery is a relatively safe and noninvasive technique that involves the application of mitochondria into the nasal cavity, utilizing the distribution of nasal nerves to allow mitochondria to be rapidly delivered to the CNS and migrate to the site of injury. This method has obvious advantages, including easy control of dosage, ease of repeated administration, reduced risk of embolism, and most importantly, the ability to bypass the BBB. There are currently two studies on the treatment of mitochondrial diseases via nasal administration of mitochondria, one on PD and the other on MCAO. The results of the first study revealed that mitochondria isolated from the rat liver were delivered to different brain regions in PD model rats through the olfactory bulb and migratory flow neurons on both sides of the midline of the brain and expressed in the striatum. Although the number of dopaminergic neurons in the substantia nigra was not altered in the PD model rats, rotational behavior was alleviated and motor function was improved in the rats after treatment [193]. In another study, mitochondria isolated from BMMSCs and purified were used to treat mice subjected to cerebral infarction. Before nasal delivery, each mouse was also given hyaluronidase through the nasal cavity to enhance nasal mucosal permeability. The results revealed an increase in synaptic marker expression and improved cognitive function in the stroke model mice [194]. Nasal delivery is an effective pathway for mitochondrial delivery, but some issues, including how to ensure the quantity of mitochondria delivered, the specific pathway through which mitochondria travel, and the distribution of mitochondria that enter the brain, urgently need to be clarified.

Biological mechanisms of mitochondrial transplantation

Regardless of the method of mitochondrial transplantation, the ultimate goal is to ensure that healthy mitochondria from the donor enter the recipient’s body. Presently, mitochondria can be transferred through biological molecules such as microtubules and motor proteins (including dynein, mitochondrial adaptors, and mitochondrial receptors) [195].

Tunnel nanotubes (tnts)

TNTs are the most common structures involved in the in vitro delivery of mitochondria, and their existence was first proposed in 2004 [196]. TNTs are dynamic structures with a diameter of approximately 50–1,000 nm and are membranous tubes containing F-actin, tubulin, and microfilaments. TNTs participate in cellular communication and biological development, promoting the intercellular exchange of signals and components [197]. Mitochondrial TNTs are slender tubular structures extended by mitochondria that come into contact with neighboring mitochondria to form connections and allow transport, thereby improving the biological activity of the recipient mitochondria. The mechanism by which mitochondria are transported through TNTs, including whether similar to cellular TNTs [197, 198], to promote the transfer of healthy mitochondria from pluripotent MSCs to neural cells after cerebral ischemia [180], has been explored. Damaged cells release a “help me” signal. Yorgov et al. reported that this signal may be due to mtDNA acting as a DAMP to increase the expression of the cytoprotective enzyme heme oxygenase-1 (HO-1) [147] to guide mitochondrial transport in stem cells. TNTs can also form protrusions in the cell envelope. Liu et al. reported that damaged cells express phosphatidylserine, which activates MSCs to form TNTs. When an organ undergoes ischemia or reperfusion injury, MSC membrane protrusions create TNTs between damaged cells. MIRO1 is more likely than other proteins to form protrusions and thus promotes the transmission of healthy mitochondria to damaged cells and increases cell survival [146, 148]. Yorgov et al. reported that MSCs engulf mtDNA from mitochondria, resulting in impaired function. As the expression of the cytoprotective enzyme HO-1 increases, the ability of MSCs to release healthy mitochondria for transport into damaged cells also increases [147]. The connections between these cells are formed and regulated by various factors and processes, such as proinflammatory cytokines and OS-induced apoptosis, which can promote the formation of TNTs [147, 199]. There are different opinions on whether TNTs are bidirectional or unidirectional, with some studies suggesting that the mitochondrial transfer of TNTs between MSCs and human umbilical vein endothelial cells (HUVECs) is bidirectional. Other studies have also suggested that under certain stressful conditions, such as cytarabine toxicity, mitochondria can only undergo one-way transfer from the donor cell to the recipient cell [200]. However, when dissociated F-agonist drugs, cytochalasin B, LatA, or Annexin V nanotube blockers are administered, the number of mitochondrial transferred is reduced due to the inhibition of TNT formation [201, 202]. Further clarification is needed to understand how different process of TNT formation are interconnected and their role in mitochondrial transfer.

Evs

In addition to TNTs, mitochondria can also be transferred through EVs. EVs are capable of carrying microRNA and genetic material for transfer to receptor cells [203]. MSC-derived EVs carry mitochondria and promote the oxidative phosphorylation of macrophages to regulate polarization and exert a protective effect [204]. The transfer function of EVs is determined by their diameter, as larger vesicles (100–1,000 nm) transfer intact mitochondria, whereas smaller EVs (30–150 nm) transfer mitochondrial components [205]. Dave et al. reported that when mitochondria are carried by brain endothelial cells (EVs with large or medium diameters rather than small diameters), the survival and connectivity of dead endothelial cells around the BBB after stroke increase [144]. Compared with studies on TNTs, relatively little research has been conducted on the transport of mitochondria by EVs, and further exploration is needed.

Cell fusion

Cell fusion is also one of the mechanisms by which mitochondria are transferred. Cells fuse the membranes of other cells, allowing them to share organelles, including mitochondria. Two dyes were used to label mitochondria isolated from Neuro-2a (N2a) cells (MitoTracker Green FM) and the N2a culture medium (MitoTracker Red CMXRos). N2a mitochondria were then added to the N2a culture medium, resulting in the complete fusion of the red and green dyes. The mitochondrial fusion protein Mfn1 may facilitate this process, suggesting that exogenous mitochondria can fuse with endogenous mitochondria within the host cell [163]. However, it is unclear whether this process relies on partial cell fusion, similar to TNT formation, or complete cell fusion without the need for external factors. Compared with research on the mechanisms discussed earlier, research on cell fusion is relatively lacking.

Gap junction channels (gjcs)

Gap junctions are connections established by coupling of the cytoplasm of cells to mediate intercellular signals. Adjacent cells form channels comprising gap junction (GJ) proteins called GJCs [206]. In this context, the term “gap” has two meanings. First, it refers to the gap of 2–3 nm between adjacent cytoplasmic membranes at the GJ; second, it refers to the fact that at the junction point of the GJ, the double lipid layer is connected not directly but by two linkers to form a channel, allowing small molecules to flow directly from one cell to another [207]. Some studies have shown that the membrane protein contactin associated protein 1 regulates the aggregation of GJ proteins by regulating cytoskeletal microfilaments [208], which helps recipient cells recognize and integrate foreign mitochondria. Although GJCs play a beneficial role in mitochondrial transplantation, they have disadvantages such as low selectivity and low transport efficiency and allow transfer only over a limited distance, and further research and technological improvements are needed.

Methods for labeling and monitoring the localization of transplanted mitochondria

Labeling and monitoring the localization of free mitochondria after transplantation aids research on the mechanisms of mitochondrial transplantation. Currently, the labeling methods for transplanted mitochondria include MitoTracker dye, JC1 dye (to measure the Ψm), fluorescent proteins, BrdU (to label mtDNA), the pDsRed2 Mito vector, the Mito-DsRed2 vector, and DsRed Cox8-GFP [209]. Methods for assessing the localization of transplanted mitochondria in vitro include real-time fluorescence imaging, including the use of a fluorescent probe affected by the membrane potential, the use of nuclear-encoded fluorescent proteins that target mitochondria, and immunoelectron microscopy combined with three-dimensional reconstruction or superresolution structured illumination microscopy [210].

Malekpour K et al. used MitoTracker Deep Red to label mitochondria in MSCs and measured the difference in the fluorescence intensity of the mitochondrial dye between the donor and recipient cells via fluorescence microscopy [211]. Huang performed labeled BHK-21 cells with MitoTracker before extracting mitochondria and measured the total number of stained particles via a flow analyzer [159]. Hayakawa also transferred MitoTracker-labeled functional mitochondria from rat cortical astrocytes into neurons [142]. In other related studies [110, 162, 168, 169, 176], MitoTracker was used to label mitochondria. Robicsek evaluated the distribution and connections of JC1-stained cells via confocal microscopy [212]. Ramirez Barbieri G and Joshi AU also used JC-1 to evaluate the Ψm [162, 213]. Chang labeled mitochondria with green fluorescent protein by transiently transfecting cells with a plasmid harboring GFP, under the control of a mitochondrial matrix-specific promoter and then isolated the mitochondria [178]. Golihue also labeling the mitochondria of PC12 cells with a tGFP transgene to isolate mitochondria [214]. While most studies used only MitoTracker or a fluorescent fusion protein to label mitochondria, Ramirez Barbieri et al. used both techniques to label mitochondria and reported that mitochondria labeled via different methods were also separated [162], which, to some extent, proves that different conclusions can be drawn on the basis of the labeling methods. Kitani et al. used PCR, real-time PCR, real-time fluorescence imaging, three-dimensional reconstruction, continuous delayed microscopic observation, flow cytometry, and EM to determine the intracellular localization of transplanted mitochondria in recipient cells [15]. Vorholt et al. observed dynamic changes in mitochondria after they were injected into cells in real time and reported that transplanted mitochondria fused with the mitochondria of recipient cells 20 min after transplantation and for more than 16 h [187]. Different phenomena were observed by monitoring transplanted mitochondria. Mitochondria are injected into mice subjected to cerebral ischemia through the carotid artery, and as the endothelial barrier opened during acute ischemia, mitochondria were able to penetrate the damaged BBB [155]. Other related studies revealed that exogenous mitochondria can reach the brains of mice [153], and can even reach multiple organs, such as the heart, kidney, and liver [110]. Different monitoring methods can be used to comprehensively evaluate the morphology, localization, activity, and dysfunction of transplanted mitochondria. Each of these methods has its own advantages and disadvantages, such as time- or metabolism-related factors, so the choice of the method depends on the experimental requirements Fig. 3.

Fig. 3

Technical quality control and characteristics of mitochondrion-based therapies. a After IS, astrocytes release functional mitochondria to restore the functions of neurons. In addition to astrocytes, mitochondria also originate from MSCs (b), endothelial progenitor cells (c), induced pluripotent stem cells (d), skeletal muscles (e), and hepatocytes (f). Isolation of mitochondria for transfer: Exogenous mitochondria from histiocytes rich in mitochondria can be obtained via different methods, including identity gradient separation (g), differential separation (h), and immunological beam separation. i Mitochondria can be obtained through differential density and speed centrifugation, as shown in (g) and (h), respectively. As shown in (i), antibody-bound magnetic beads are added, and cells connected to the magnetic field are adsorbed by the magnetic beads. After the supernatant is discarded, the labeled cells are analyzed to obtain target mitochondria. Function of transplanted mitochondria: To obtain functional mitochondria, ensuring the integrity of the outer and inner membranes during the mitochondrial extraction process is necessary. For the identification of isolated mitochondria, EM (j), Ψm-specific probes (k), flow cytometry (l), hippocampal potential measurement (m), WB (n), and other methods are used. If many mitochondria are needed, further improvement of long-term mitochondrial storage strategies is needed (o). Mitochondria are cooled at a rate of approximately 1 °C/minute and stored at − 80 °C. Pathway of transplanted mitochondria: The methods used to deliver mitochondria to recipient cells in vitro include coculturing (p), the photothermal nanoblade method (q), MitoPunch (r), magnetic transfer (s), FluidFM (t), direct injection (u), and nasal administration (v). Biological mechanisms of transplanted mitochondria: Various components of cells, such as mitochondria, other organelles, proteins, lysosomes, and plasma membranes, can be transferred through TNTs (w), EVs (x), cell structures (y), and GJC (z)

The application of organelle-based therapies involving mitochondrial transfer in the treatment of CNS diseases

Mitochondrial transplantation has shown great potential in replacing dysfunctional mitochondria and alleviating CNS diseases by delivering healthy functional mitochondria to damaged cells and tissues and has also received widespread attention. The viability of recipient cells increases as the number of exogenous mitochondria increases. However, when a certain number of mitochondria is reached, the therapeutic effect does not increase, as excess mitochondria are removed by autophagy [215]. Both preclinical and clinical studies on mitochondrial transplantation have achieved good results. Herein, we present a summary of the studies that have been and are currently being conducted on different CNS diseases Fig. 4.

Fig. 4

Application of organelle-based therapies involving mitochondrial transfer in the treatment of CNS diseases. To date, the ability of mitochondrion-based therapies to treat CNS diseases such as IS, SAH, AD, PD, Epilepsy, TBI, MS, and SCI in preclinical studies and Pearson’s disease, stroke, refractory polymyositis and dermatomyositis in clinical studies has been evaluated. The goal of mitochondrial transplantation is to restore mitochondrial function by replacing damaged mitochondria with healthy mitochondria

Preclinical studies

Preclinical research has indicated that mitochondrion-based therapies can effectively treat CNS diseases (Tables 2 and 3), such as IS, SAH, AD, PD, TBI, MS, Epilepsy, SCI. Among CNS diseases, ischemic disease is the most extensively studied. Owing to the high levels of metabolism in the CNS, many mitochondria are required to provide sufficient energy. Considering that OS and impairment of ion transfer ultimately lead to cell death during ischemia and hypoxia, these processes cannot be separated from mitochondrial dysfunction. Therefore, restoring healthy and functional mitochondria is required to maintain proper cellular function. Huang et al. transplanted allogeneic mitochondria into mice via intradermal injection. They reported that the transplanted mitochondria were distributed mainly in cells around the ischemic infarct area, such as in astrocytes and microglia [159], and may have even been transferred to other remote tissues. Additionally, mitochondria increased overall energy production, reduced OS, decreased cell apoptosis, and promoted the repair of damaged nerves. Therefore, mitochondrial transplantation has great potential as a clinical treatment.

Table 2 Mitochondrial transplantation therapy preClinical trials for CNS disorders

Table 3 Mitochondrial transplantation therapy clinical trials for CNS disorders

IS

In IS, brain tissue death is caused by reduced cerebral blood flow or insufficient cerebral oxygen supply, and IS is characterized by the rapid development of functional deficits in local or distant brain regions. IS is a major disease that seriously endangers human health, and during ischemia, mitochondria, as energy factories that regulate cell viability and homeostasis, exhibit functional impairment, resulting in sustained changes in ion-related, biochemical, and cellular processes. Mitochondria are the main factors that initiate the cellular cascade that leads to neuronal damage after cerebral ischemia. Therefore, protecting mitochondrial function through exogenous mitochondrial supplementation is a powerful strategy for treating IS. In the past decade, the ability of mitochondrial transplantation to treat IS has been widely studied. The clinical application of mitochondrial transplantation requires strict control of the source and quality of mitochondria, and the ideal mitochondrial source should be easily obtainable and easily amplified in large quantities. Therefore, since the discovery of mitochondrial transfer in 2004, intercellular communication through mitochondria has been studied in different types of stem cells and damaged cells, including hUC-MSCs [152], neural stem cells (NSCs) [163], and MMSCs [179, 180, 202], and the unidirectional transfer of mitochondria from MMSCs to damaged cells has also been confirmed. Research has shown that mitochondria derived from stem cells can effectively repair brain tissue damaged by IRI in several ways, such as by reducing ROS levels, inhibiting apoptosis activating microglia, increasing the expression of synaptic markers and increasing cell viability. However, MMSCs tend to disappear rapidly after transplantation, so their regenerative potential is often limited by time [179]. Therefore, genetic modification of stem cells, such as increased miro1 expression, can be used to aid mitochondrial transfer and repair damaged brain tissues [180]. In addition to stem cells, other types of mitochondria can be successfully obtained from various sources, including rapidly frozen placentas [153], EVs from hCMECs/D3 EVs [156], BHK-21 cells [159], autologous muscle cells [169], astrocytes [194], EVs from BECs [144], and endothelial progenitor cells [168]. Mitochondria derived from currently available and well-characterized immortalized human brain capillary endothelial cell lines have also been extensively studied. This is because BECs have greater metabolic activity than other nonbrain endothelial cells do and that their mitochondria produce large amounts of energy to aid in neurovascular recovery after stroke. However, as the earliest cells of the BBB to perceive hypoxia/reoxygenation signals in ischemic diseases, their metabolism and survival depend more strongly on the function of mitochondria [216]. Some mitochondria originate from cellular vesicles because EVs contain abundant intrinsic biological molecules that transmit information between cells, resulting in additional synergistic therapeutic effects of their cargo. To achieve higher transfer efficiency, EVs can be further altered. For example, engineered hCMECs/D3 EVs carrying the Luc-pDNA plasmid can transfer mitochondria to recipient BECs to a greater extent to exert therapeutic effects [144]. Moreover, studies have shown that EV/HSP27 mixtures further increase BEC metabolism in ischemia [144]. Overall, the injection of mitochondria from differences sources via different routes (intravenous or intraventricular) has beneficial effect in IS.

SAH

SAH is a common hemorrhagic cerebrovascular disease that refers to sudden bleeding in the subarachnoid space and is characterized by sudden severe headache. The sharp increase in intracranial pressure after SAH severely affects cerebral perfusion. After SAH, bleeding and the release of toxic substances can cause the impairment of mitochondrial integrity and function, leading various types of damage, such as OS and cellular apoptosis, causing ischemia and hypoxia in related tissues and resulting in tissue damage and functional impairment. The aforementioned studies on the use of mitochondrial transplantation for the treatment of IS, revealing that damaged neurons use mitochondria as a “help me” signal and transmit the signal to neighboring nerve cells. It has been reported that extracellular mitochondria can mediate intercellular communication, but their specific effect depends on the health of the mitochondria [164]. Chou et al. reported that both model rats and patients with SAH presented an increase in mitochondrial concentrations in the cerebrospinal fluid, which is closely related to glial cells [176], indicating that mitochondria may be released into the extracellular space. Furthermore, to investigate the potential cellular origin of these mitochondria in the CSF, they labeled all functional mitochondria with MitoTracker, and they ultimately determined that the mitochondria were derived from astrocytes. However, the therapeutic effect of mitochondrial transplantation in SAH has not yet been studied, and related research may provide new targets for the diagnosis and treatment of SAH in the future.

AD

AD is a progressive degenerative disease of the CNS with insidious onset that is characterized by memory impairment, cognitive impairment, and loss of mental function. Its pathological features include the deposition of extracellular Aβ plaques, excessive phosphorylation of tau, and mitochondrial dysfunction. Changes in mitochondria with age are believed to play a key role in the pathological development of AD, such as through OS, apoptosis, iron metabolism dysfunction, energy imbalance, and interactions with Aβ and phosphorylated tau proteins. During the pathogenesis of AD, Aβ is localized in the mitochondria and interacts with mitochondrial components to exert toxic effects, leading to a decrease in mitochondrial energy production. In addition, excessive phosphorylation of the tau protein affects mitochondrial morphology and function. As mentioned earlier, mitochondrial dysfunction may be the main cause of excessive phosphorylation of Aβ and tau. However, regardless of the type of mitochondria, the delivery of healthy mitochondria is an effective option for treating AD and replacing damaged mitochondria. We found only two relevant articles on the use of mitochondrial transplantation for the treatment of AD, both from the same team and published in 2019 [149] and 2023 [150].

The researchers isolated mitochondria from HeLa cells and assessed the effects of intravenous injection of these mitochondria on short-term pharmacological AD mouse models and long-term chronic AD mouse models. The results showed that mitochondrial transplantation reduced glial cell proliferation, reduced neuronal damage and the amyloid burden, improved cognitive function, and demonstrated high safety in the AD model mice. Moreover, the researchers contended the previous statement that engineering isolated mitochondria can further improve transplantation efficiency. Instead, they believe that unmodified mitochondria have strong therapeutic effects because the original composition of the organelles is not altered, thus avoiding abnormal effects. However, notably, after the injection of mitochondria into AD model mice, strong fluorescence signals were detected in the liver but not the brain, and these signals persisted for 15 days, at which point weak signals were observed in the brain. The authors speculated that the mitochondrial response in the liver may be stronger than the response in the brain. To further investigate the mechanism of the interaction between the mitochondrial response in the liver and that in the brain, they conducted experiments on the effects of long-term mitochondrial transplantation on a chronic AD model and performed metabolomics and proteomics analyses. The results confirmed that intravenously injected mitochondria enter the liver through the blood circulation and subsequently induce extensive metabolism in the liver. Neuroprotective metabolites produced by these reactions enter the brain via the liver-blood‒brain axis, affecting the pathogenesis of diseases. However, currently, the only source of mitochondria for research on the use of mitochondrial transplantation for the treatment of AD is HeLa cells, and the only practical transplantation method is intravenous injection. There is still much work that can be done in the future.

PD

PD is a progressive neurodegenerative disease that is common in middle-aged and elderly individuals that and is characterized by tremors, muscle rigidity, and bradykinesia. One of the pathological mechanisms of PD is the abnormal aggregation of toxic α-synuclein oligomers, which impairs the activity of mitochondrial CI and increases ROS generation, further accelerating α-synuclein oligomerization and leading to mitochondrial dysfunction. Mitochondrial dysfunction is a key factor in the progression of PD and manifests as increased susceptibility of dopaminergic neurons to excitatory neurotoxin death. Therefore, increasing ETC activity and reducing ROS levels through mitochondrial transfer can effectively treat PD. The construction of PD models often involves the use of drugs such as the neurotoxin MPP⁺ [110, 217], the neurotransmitter dopamine antagonist 6-OHDA [178, 193, 217], and rotenone [190, 217, 218], which exert their effects by inducing mitochondrial dysfunction. As we mentioned earlier, mitochondrial sources should be easily obtainable and easily amplified in large quantities, and both tumors and stem cells meet these conditions. Shi et al. innovatively transferred mitochondria isolated from HepG2 cells to SH-SY5Y human neuroblastoma cells and injected them into multiple sites in mice, and the researchers found that the treatment increased ETC activity, reduced ROS levels and apoptosis, and preventing the progression of PD [110]. In addition, research has also been conducted on the ability of stem cell (iPSC and UC-MSC)-derived mitochondria to treat PD [190, 217]. Research has confirmed that p38 is spontaneously phosphorylated in iPSC-derived astrocytes to induce the transfer of mitochondria to damaged dopaminergic neurons, significantly reversing their degeneration and axonal pruning [188]. UC-MSCs protect dopaminergic neurons and improve the behavior of PD model mice by reducing the expression of inflammatory factors and the activation of microglia through in the nuclear factor kappa B (NF-κB) signaling pathway [217]. For the modification and isolation of mitochondria, considering that the cell-penetrating peptide Pep-1 can promote the uptake of heterologous mitochondria by cells, Chang et al. labeled mitochondria via a Pep-1-expressing vector. During mitochondrial transfer, there was a difference between the transfer of Pep-1-labeled mitochondria and unlabeled mitochondria. The Pep-1-labeled mitochondria significantly increased mitochondrial CI protein levels in dopaminergic neurons in the substantia nigra, restored mitochondrial function, and reduced mitochondrial oxidative damage [178]. However, there are some concerns about primary mitochondrial diseases (PMDs), such as the possibility of purified mitochondria losing their function and the inability of DA neurons to effectively internalize the transplanted mitochondria. Some studies have suggested that mitochondria from exercised mice can avoid these concerns and, to some extent, solve the issues of mitochondrial OXPHOS. They compared the effects of the allogeneic transplantation of normal mitochondria and mitochondria from exercised mice in PD model mice and reported that the mitochondria from exercised mice significantly increased the respiratory capacity of the PD model mice by increasing the internalization rate through F-actin [190]. These findings suggest that increasing the transplantation efficiency of functional mitochondria can further promote the recovery of mitochondrial function. In addition to the widely studied delivery routes of intravenous injection and intracranial injection, it is also feasible to treat PD via the intranasal delivery of mitochondria [193]. Researchers labeled allogeneic mitochondria isolated from the rat liver with Pep-1 and administered them to PD model rats, and significant alleviation of rotational behavior and improvements in motor function were observed. The authors suggested that these improvements were related to the restoration of mitochondrial function and a reduction in oxidative damage to the substantia nigra. However, the efficacy of intranasally administered mitochondria was similar to that of locally injection mitochondria. Mitochondrial CI levels in dopaminergic neurons were significantly lower following intranasal administration of mitochondria than that following local injection. At present, research on the use of mitochondrial transplantation for the treatment of PD involves mitochondria from multiple sources and different transplantation methods, all of which are effective to some extent. In the future, further detailed exploration of methods for modifying and transplanting isolated mitochondria to achieve effective neural uptake in the brain will be particularly important.

Epilepsy

Epilepsy is a brain disorder characterized by temporary functional impairment caused by excessive excitation and abnormal discharge of electrical activity in the brain, causing symptoms such as convulsions, seizures, and loss of consciousness. As the main producer of ATP for normal neuronal activity and synaptic transmission, mitochondrial dysfunction is not only the main cause of epileptic seizures but also may further lead to secondary brain damage after seizures. Therefore, mitochondrial supplementation is a promising treatment strategy for epilepsy-related mental and cognitive dysfunction. For example, in vitro, coculture of human osteosarcoma 143B cells harboring the A8344G mutation in mtDNA with skin fibroblasts from MERRF syndrome patients revealed that mitochondrial internalization promoted the recovery of cellular function [14]. In vivo, mitochondria isolated from the hippocampus of normal mice were intravenously injected into mice with status epilepticus (SE) induced by pilocarpine hydrochloride, and the results showed that the exogenous mitochondria reduced ROS production, activated glial cells, and altered several metabolic pathways, such as the methylmalonic acid pathway, thereby alleviating cognitive impairment and depressive and anxiety-like behavior [166, 219]. Although it is unclear how mitochondria are internalized into recipient cells, this research provides the first direct experimental evidence that exogenous mitochondria ameliorate hippocampal damage after SE. At present, there is relatively little research on the therapeutic efficacy of mitochondrial transplantation for epilepsy, but there is no doubt that it may be a new strategy for treating epilepsy and related mental and cognitive disorders; however, further exploration is still needed.

TBI

TBI usually results from a primary injury due to mechanical force, which causes loss of consciousness disorder, brain dysfunction, or death. After TBI, mitochondrial dysfunction occurs in brain tissue, leading to secondary responses such as excessive ROS generation and calcium overload to cause neuronal damage, which is the key mechanism of secondary injury. Therefore, protecting mitochondrial function can improve the prognosis of TBI. Currently, the sources of mitochondria for treating TBI in preclinical studies vary and include astrocytes [143], healthy neurons [167], fresh brain tissue [154], allogeneic liver and autologous muscle [220], and mitochondria from all of these sources have shown significant therapeutic effects. The long-term prognosis of TBI depends on the subsequent regeneration of damaged neurons, and Chien et al. reported significant recovery of neuronal transmission and the neuronal membrane potential after the transfer of mitochondria derived from astrocytes to damaged hippocampal neurons [143]. Chien et al. observed significant neuronal burst regeneration and recovery of damaged neuronal membrane potential by providing mitochondria derived from healthy mice cortical neurons to damaged hippocampal neurons [167]. In addition, the transfer of mitochondria from allogeneic liver and autologous muscle upregulated the expression of neurotrophic factors in reactive astrocytes and improved spatial memory ability in TBI model rats [220]. In addition, mitochondrial activity is particularly important during the transplantation process. A comparison of mitochondria extracted from frozen/thawed brain tissue and fresh brain tissue revealed that mitochondria extracted from fresh tissue had greater biological activity [154]. These healthy fresh brain tissue-derived mitochondria were subsequently transferred to cells with OGD-induced injury and transplanted into cortical impact-induced TBI models, and the results revealed that the exogenous mitochondria improved the function of damaged mitochondria and reduced cell apoptosis, brain edema, and BBB leakage. These studies explored the effects of exogenous mitochondria on neuronal and endothelial cell function after TBI, but further research is needed to explore new mitochondrion-related strategies for TBI treatment.

MS

MS is an immune-mediated inflammatory demyelinating disease of the CNS characterized by multiple temporal and spatial occurrences, mainly manifested as limb numbness, limb weakness, unstable walking, and even paralysis. The pathogenesis of MS involves mitochondrial dysfunction, resulting in inflammation, OS, and bioenergetic disruption, which can lead to an imbalance of neurotrophic substances in neurons and oligodendrocytes, resulting in increased axonal demyelination. Therefore, restoring normal mitochondrial function in neurons and glial cells, such as via mitochondrial transplantation, may be an effective method for treating MS. Functional mitochondria were isolated from NSC-EVs and transferred to L929 Rho⁰ cells with mtDNA deficiency and EAE animal models (models of MS), and the results revealed that exogenous mitochondria rescued cellular mitochondrial function and alleviated clinical symptoms in EAE model mice [192]. The researchers provided a detailed description of how extracellular mitochondria enter mononuclear phagocytes, and they speculated that the internalization of exogenous mitochondria is mainly mediated not by phagocytosis but by an endocytosis process similar to microphagocytosis. These findings provide crucial information for exploring the transfer of healthy mitochondria from stem cells to immune cells in the future.

SCI

SCI is a CNS lesion caused by various factors, such as trauma, inflammation, and tumors, and mainly manifests as impairment of motor, sensory, and autonomic nervous system functions below the level of injury to varying degrees. SCI includes acute, subacute, and chronic phases. The loss of ATP synthesis and glutamate excitotoxicity in the acute phase, neuroinflammation in the subacute phase, apoptosis and necrosis in the chronic phase, and axonal demyelination all result from to mitochondrial dysfunction. Therefore, the goal of mitochondrial transplantation is to restore mitochondrial function in the injured spinal cord, providing more energy, alleviating neuronal apoptosis and promoting functional recovery, to treat SCI [181, 188, 191, 214]. Numerous studies have confirmed that mitochondrial transfer from stem cells, especially MSCs, to adjacent damaged cells restores the biological functions of the cells. In 2018, mitochondrial transplantation was first studied as a method for treating SCI. Researchers used homologous soleus muscle tissue as a source of mitochondria. This type of muscle, made up of slow muscle fibers, contains large amounts of mitochondria and myoglobin and have a high capillary density, making it a good source of healthy mitochondria that can produce large amounts of energy [214].The following year, another research team explored the protective effect of mitochondria by coculturing BMSCs with OGD-exposed neurons and injecting BMSC-derived mitochondria into SCI model rats through electrode microneedles. The results showed that mitochondrial internalization improved neuronal function and motor function after SCI [181]. Afterward, a team isolated mitochondria from the soleus muscle again and transplanted them to the ischemic spinal cord through the internal jugular vein. The results showed that mitochondrial transplantation significantly improved the lower limb motor function of rats with spinal cord ischemia [188, 191]. Mechanistically, mitochondrial transplantation decreased the expression of CCAAT enhancer binding C/EBP homologous protein (CHOP), restored immunoglobulin levels, and inhibited ER stress in ischemic areas, thereby reducing neuronal inflammation and apoptosis. In addition to mitochondria derived from the soleus muscle, mitochondria isolated from BMSCs could also be internalized by neurons, thereby improving bioenergetics, reducing apoptosis of damaged neurons, and enhancing motor function in SCI model rats. At 28 days after treatment, the transplanted mitochondria were still detectable in the damaged spinal cord [191]. Although Gollihue reported that exogenous mitochondria maintained mitochondrial bioenergetics and improved the motor function of rats to some extent, they did not exert long-term neuroprotective effects [214]. This may have been due to the low internalization rate of the transplanted mitochondria, which did not enter host neurons in significant numbers; however, the impact of the extracellular mitochondria on cells was not elucidated. Thus, it is crucial to focus on improving the efficiency of mitochondrial integration into neurons and other nerve cells in future research.

Clinical studies

According to preclinical animal studies, the transfer of healthy mitochondria into damaged cells is effective in treating diseases. Therefore, corresponding clinical research has emerged. However, to date, few clinical trials involving mitochondrial transfer have been registered. We used the keyword “mitochondrial transplantation” to search for mitochondrial transplantation-related clinical studies on CNS diseases and analyzed the results.

The first clinical study was jointly completed by Sheba Medical Center Hospital and Minovia Therapeutics Ltd. In this study, seven patients with Wilson’s disease (WD) (aged 3–18 years) were recruited to evaluate the safety and efficacy of MNV-BM-BLD (autologous CD34 + cells rich in blood-derived mitochondria) for the treatment of WD. The main indicators included the number of participants with treatment-related adverse events graded according to the CTCAE Version 5.0 and changes in mitochondrial disease scores. This study, which, to our knowledge, is the first clinical study on the transplantation of mitochondria from autologous hematopoietic stem cells (auto-HSCTs) into patients with primary mitochondrial disease, has been completed and published. The results revealed that the treatment of six children with mtDNA deficiency with hematopoietic stem cells cultured with healthy mitochondria from the mother exerted favorable therapeutic effects, improving muscle strength and endurance and allowing patients to walk and stand for the first time [221]. Adverse events related to treatment include anemia, hypocalcemia, and alkalosis. However, given the significant benefits of mitochondrial therapy in enhancing patient metabolism, aerobic capacity, and quality of life, these adverse events do not compromise the safety of transplantation.

The second clinical study was the first related study on the human brain and was performed at the University of Washington in 2021 on patients with cerebral ischemia. The researchers recruited 20 patients with cerebral ischemia aged 18–85 years, aiming to confirm the safety of autologous mitochondrial transplantation for the treatment of cerebral ischemia. Autologous muscle tissues were used as the source of mitochondria. Autologous mitochondria were extracted from the patient and injected into the cerebral artery during reperfusion via microcatheters. At this time, no corresponding articles have been published.

The third clinical study was jointly completed at Seoul University Hospital and Paean Biotechnology, Inc. This project was a phase 1/2a trial, and the researchers recruited 18 patients with refractory polymyositis or dermatomyositis aged ≥ 19 years. Mitochondria (PN-101) were isolated from allogeneic umbilical cord-derived mesenchymal stem cells and administered intravenously to confirm safety and tolerance and to explore the efficacy of this treatment method. The main evaluation indicators included dose-limited toxicity (DLT) and the International Myositis and Clinical Study Group Total Improvement Score (IMACS-TIS). The study is currently being registered through an invitation, and no relevant articles have been published.

Currently, there are limited clinical studies on treating neurological diseases with mitochondrial organelles. However, some clinical registration projects indicate that disease-related tissues, such as muscles and the brain, are highly dependent on mitochondrial energy supply. Treatment plans for CNS diseases using organelles may select appropriate indications based on the biological functions of mitochondria and their associated organelles in the future. Additionally, more preclinical research data on mitochondrial therapy may be needed for future exploration.

Limitations and challenges of mitochondrion-based treatment of CNS diseases

Although mitochondrion-based therapies have been development, there are still limitations and challenges that need to be addressed in the process of mitochondrial transplantation, including the source, method, pathway, efficiency, and tracking, in order to better facilitate clinical translation: (1) Limitations of mitochondrial source: As mitochondrial transplantation involves the transfer of genetic material, the need for special supervision of mitochondrial sources and other ethical issues have sparked discussion [222,223,224]. During the aging process, mutations in mtDNA encoding core subunits of different respiratory chain complexes accumulate. When the accumulation of these mutations reaches a certain threshold, OXPHOS is impaired. Therefore, mitochondria from younger individuals have better therapeutic effects [225]. Additionally, individualized mitochondrion-targeted modifications based on gene polymorphisms may lead to improved therapeutic effects. (2) Challenges of mitochondrial transplantation methods: Since the mitochondrial matrix has a highly negative charge (approximately -240 mV), it may be easier for them to bind to cells in ischemic tissue, which is highly acidic, after transplantation. Therefore, it is necessary to determine the ideal method, dose and number of treatments for mitochondrial transfer. Thus, further investigation and clarification of the underlying mechanisms are needed [153]. (3) Issues of mitochondrial transplantation pathway: Mitochondria may be degraded in cell culture medium and lose their normal function during transplantation, leading to the formation of DAMPs that activate the innate immune system and trigger the production of inflammatory cytokines and chemokines [226,227,228]. Therefore, selectively regulating the packaging of mitochondrial proteins in EVs can prevent the free release of damaged components and exert pro-inflammatory DAMP effects [229]. (4) Increasing the efficiency of mitochondrial transplantation: The permeability of the plasma membrane of recipient cells in different individuals may lead to the uptake of different quantities of mitochondria and of mitochondria of different qualities, and the transfer efficiency of exogenous mitochondria cannot be fully guaranteed, leading to different therapeutic effects. Therefore, it is necessary to better understand the mechanisms of mitochondrial transmission [230]. (5) Tracking and monitoring mitochondrial transplantation: Owing to the lack of an effective labeling strategy, long-term tracking and monitoring of mitochondria after transplantation cannot be achieved. When used for membrane potential-dependent labeling, existing fluorescent dyes are associated with hazardous effects such as mitochondrial toxicity and the possibility of mitochondrial leakage; thus, their use is not recommended. Therefore, there is an urgent need to develop effective probes for mitochondrial labeling.

Conclusions

Organelle-based therapies, including mitochondrial transplantation, represent new avenues in biomedical research. They involve and the transfer of functional mitochondria into recipient cells with damaged mitochondria and affect metabolism and energy production in damaged or stressed cells to improve the energy balance, thus exerting therapeutic effects. Although these strategies are still in their infancy, revolutionary advances have attracted widespread attention and provided ideas for future research, as well as insights into alternative treatments for CNS disorders. Despite the aforementioned issues, mitochondrial therapy is an emerging treatment method for various diseases. We believe that overcoming the limitations discussed above will allow the broader application of this novel treatment method in humans.

In conclusion, this paper provides a comprehensive review of organelle-based therapies, especially mitochondrion-based therapies. We emphasize the importance of organelle-based therapeutic strategies for the precision treatment of CNS diseases and believe that mitochondrial transplantation represents a revolution in clinical treatment.