Horizontal mitochondrial transfer as a novel bioenergetic tool for mesenchymal stromal/stem cells: molecular mechanisms and therapeutic potential in a variety of diseases

In neurons, the extracellular protein S100A4 and its receptor RAGE (Receptor for Advanced Glycation End Product) play crucial role in directing TNTs biogenesis [107]. Under oxidative stress, p53-caspase-3 axis determines the cleavage and depletion of S100A4, thereby leading to a concentration gradient between neurons and astrocytes that drives the extension direction of TNTs from neurons (with low concentration of S100A4) to astrocytes (with high concentration of S100A4).

Miro 1 and 2 proteins are embedded in the outer mitochondrial membrane and involved in the structure of the mitochondrial contact-site and cristae organizing system complex (that connects outer and inner-mitochondrial membrane), Ca2+ metabolism, mitochondria-ER communication, and mitophagy. Besides these functions, Miro proteins have critical role in regulating the mitochondrial spatial positioning and transport acting as adaptors that link mitochondria to cytoskeleton-associated motor proteins [108, 109]. Generally, microtubules and actin microfilament system mediate long- and short-distance transport of mitochondria, respectively. In this regard, movement of mitochondria based on microtubules involves the interaction of Miro/TRAK (Trafficking Kinesin Protein) complex with different motor proteins, such as Kinesin, dynein, and kinesin superfamily KIF5 [110, 111], while the actin-based mitochondrial movement is mediated by myosin family members, including myosin II, V, VI, and XIX [112, 113].

Recent studies demonstrate the essential role of Miro proteins in TNTs‐mediated mitochondrial [38] transport [114–118]. Specifically, overexpression of Miro1 promotes MT from donor MSCs to epithelial injured cells, leading to the rescue of epithelial functions. On the contrary, Miro1 knockdown inhibits MSC-mediated MT [114]. Consistent with this, upregulation of Miro1 in MSCs increases the metabolic/bioenergetic benefits of MT following neuronal oxidative stress and mitochondrial damage, while decreasing Miro1 expression reduces these effects [115, 117]. Also, MT via TNTs from astrocytes to neurons rescues neurons from cisplatin-induced damage. Furthermore, siRNA-mediated knockdown of Miro1 in astrocytes reduces MT, thus preventing the normalization of neuronal calcium dynamics [116].

MT can also occur in a Cx43-dependent manner since the deletion of Cx43 in donor/acceptor cells usually negatively affects the process [119–121]. Canonically, the formation of GJs or hemichannels across the plasma membrane by six Cx43 monomers allows uni- and bidirectional transfer of ions, small molecules (e.g., glucose, prostaglandins, microRNAs, and secondary messengers), and organelles between cells, thereby regulating intracellular mechanisms of signaling and several cellular functions [50, 122]. In line with this context, the MT from BM-MSCs to injured alveolar epithelium requires the Ca2+ exchanges between the two cells via Cx43-GJs, in the lipopolysaccharide (LPS)-induced acute lung injury mouse model [119]. In addition, MT from hematopoietic progenitors to BM-MSCs is required to induce the metabolic recovery of recipient BM-MSCs following irradiation and is cell-contact dependent and mediated by Cx43 [121].

Besides this classical function, Cx43-GJs may play further physiological roles, such as the regulation of mitochondrial functions and mediation of MT through the formation of TNTs and GJs internalization. In particular, Cx43 has also been detected within mitochondria [123] and seems to play a crucial role in mitochondrial calcium homeostasis and cell survival [124]. Cx43 is also implicated in forming connections between TNTs [125, 126] and marked reduction of TNTs is observed following knocking out the Cx43 genes in human trabecular meshwork [127]. Also, Cx43-mediated TNTs formation is crucial for MT from human induced pluripotent stem cell (iPSC)-derived MSCs to the injured bronchial epithelial cells and results in the inhibition of asthma inflammation [120]. Importantly, a novel mechanism for MT mediated by Cx43-GJs has been reported [128]. Specifically, this process requires a distinct form of GJs turnover involving the engulfment of GJs (a form of trogocytosis). Thus, following GJs internalization whole mitochondria and endosomes are incorporated into vesicles (named connexosomes/annular gap junctions) and transferred between neighbouring granulosa cells.

EVs are heterogeneous, phospholipid-enclosed structures that play a pivotal role in cell-to-cell communication at a longer range [129]. They are released by any cells into the extracellular space under different physiological and pathological conditions and encompass small EVs or exosomes (diameter of 30–200 nm), microvesicles (MVs or ectosomes; diameter of 100–1000 nm), and apoptotic bodies or apoptosomes (> 1 μm in size) [130, 131] Very recently, a specific subset of small EVs of mitochondrial origin has been identified in mouse and human brains named “mitovesicles” [132–134]. Mitochondria encapsulated into EVs can be transferred between cells to maintain the survival of metabolically compromised cells [58], to regulate immune responses [135], or to maintain tissue homeostasis [136]. In a mouse model of focal cerebral ischemia, MT by EVs between astrocytes and neurons acts as a survival mechanism protecting neurons from glucose deprivation and hypoxia [60]. EVs-mediated MT from renal scattered tubular cells (STC-like cells) to injured tubular epithelial cells (TEC) exerts protective effects, resulting in attenuating renal stenosis and recovering mitochondrial respiration [137]. Very recently, it has been demonstrated that mitochondria containing MVs (diameter of about 185 nm) derived from a human brain microvascular endothelial cell line significantly increase ATP levels, mitochondrial respiration, and glycolytic capacities of the ischemic primary human brain microvascular endothelial cells [138]. Interestingly, EVs-based interorgan transport of mitochondria may protect the heart through hormetic responses [139].

Specifically, circulating EVs derived from energetically stressed adipocytes, and containing oxidatively damaged mitochondria are taken up by cardiomyocytes, where they promote oxidative stress, resulting in a metabolic and redox adaptation of the heart that may confer protection against a future lethal stress.

In in vitro models of acute respiratory distress syndrome (ARDS), EVs-mediated MT from MSCs to human macrophages promotes phagocytosis and abolishes proinflammatory cytokine secretion [e.g., TNF-α] by macrophages, thereby mitigating lung injury [140]. Furthermore, in a model of allergic airway disease, myeloid-derived regulatory cells (MDRCs) encapsulate and transfer mitochondria to peripheral T cells via EVs, thereby affecting their bioenergetic and/or redox profile [141, 142]. Active mitochondria (in EVs and isolated) released by platelets stimulates the pro-angiogenic activity of MSCs via metabolic remodeling, including increased de novo fatty acid synthesis [143]. In this context, the clathrin-dependent endocytosis is the mechanism by which mitochondria are internalized by acceptor cells. Accordingly, EVs from MSCs are emerging as a novel nano-strategy approach to attenuate mitochondrial damage and improving TFAM-mtDNA complex stability, essential for regenerative capability of different tissues [144].

Although most data to date suggests that MT can occur via TNTs, Cx43-GJs, and EVs, various other routes have been proposed, including cell fusion, mitochondrial extrusion, and the migrasome-mediated mitocytosis [21, 145].

Cell fusion involves the physical merging of two or more cellular membranes, which would hypothetically allow the exchange of multiple protein complexes and even organelles, such as mitochondria [146]. Although cell fusion is infrequent under normal conditions, it may result in a marked mitochondrial delivery into recipient cells following injury and inflammation [147], hypoxia-induced apoptosis [148], or irradiation [149]. Cell fusion between human MSCs and cardiomyocytes results in MT into cardiomyocytes, and the resulting hybrid cells are reprogrammed toward a progenitor-like state [150]. Stem cells can fuse with other cellular models, including hepatocytes [151] and neurons [152], generating hybrid phenotypes which summarize distinct characteristics of both cells [153].

MT can also involve the extrusion or internalization of free mitochondria or mitochondrial components without membranous carriers. Under stress conditions, these events are crucial for regulating mitochondrial turnover and homeostasis [145, 154, 155]. Uncouplers of oxidative phosphorylation strongly stimulate the complete release of fragmented mitochondria from highly glycolytic HeLa cells via a mitoptosis process [154, 156, 157]. An analogous mechanism is also confirmed in in vivo models, including cardiomyocytes [158] and neurons [159, 160] suggesting a possible role as “waste removal” [145]. In Caenorhabditis elegans, under proteotoxic stress, neurons release dysfunctional mitochondria and protein aggregates via large membrane‐bound vesicles called “exophers” to support cellular homeostasis and functionality [160]. Murine cardiomyocytes can remove mitochondria and a significant portion of subcellular components via exopher‐like structures that are finally taken up and eliminated by cardiac macrophages. This process would support heart homeostasis by preventing extracellular accumulation of waste material, autophagic block, and inflammasome activation [158].

In hepatocytes and fibroblasts, the extrusion of mitochondria is also promoted by LPS through a process resembling secretory autophagy [161]. TNF-α induces the release of naked mitochondria into the extracellular spaces [155]. Similarly, massive mitochondrial extrusion has been shown in cell models of TNF‐α-induced necroptosis [162]. In this regard, free mitochondria would act as a specific danger signal to trigger inflammatory processes. Therefore, activated platelets can also release active mitochondria, both within membranous carriers and as free compartments, able to induce inflammation [163].

Importantly, MT may also occur via migrasomes, vesicular structures that grow on the tips and intersections of retraction fibers of migrating cells via tetraspanin microdomains [164, 165]. Therefore, to maintain cellular homeostasis, damaged mitochondria would be transported to the cell periphery before disposal by mitocytosis, an important mitochondrial quality-control process [166]. This process would deliver damaged mitochondria to surrounding cells [167].

MT-mediated by MSCs may be used in restoring the bioenergetic metabolism and cell functionality and may be a useful tool for several diseases’ treatment. The MSCs potential therapeutic role has reported in several papers with a proposed mechanism which suggests that MSCs may enhance tissue repair after injury by MT and shedding of membrane vesicles [168]. MT is a strategy investigated in several kinds of cells such as pulmonary, cardiac, renal, corneal epithelium, and brain cortical [37]. Rustom et al. reported mitochondrial donation via a new form of cell-to-cell interaction based on TNTs [169]. Moreover, Jang et al. described intercellular mitochondrial transportation from MSCs to corneal endothelial cells, photoreceptors, and retinal pigment cells [170] and that recipient cells exhibited increased mitochondrial respiratory abilities. In the same paper Jang et al. reported that direct contact is a prerequisite for TNT formation and identified F-actin-based TNTs bridging MSCs and recipient cells. Several authors reported that mitochondrial dysfunction is associated also with several neurological diseases such as stroke, spinal cord injury and Alzheimer diseases (AD) [171, 172]. Moreover, there are evidence that MT may be a new approach to restoring mitochondrial functions and that the MT can be used to correct a range of problems caused by mitochondrial dysfunction [173, 174].

In several retinal and corneal diseases has been shown that mitochondrial dysfunction is a critical phenomenon [92, 175]. For example, it has observed that the increase of mitochondrial fission and mitochondrial DNA damage in retinal vasculature precede apoptosis of retinal endothelial cells in diabetic retinopathy [176, 177]. Therefore, several authors thought that targeting mitochondrial dysfunction may be an approach to prevent the development and progression of both retinal and corneal degeneration. Jiang et al. showed that intercellular mitochondrial transport is a vital mechanism for regeneration of corneal epithelial cells and retinal ganglion cells [51, 178]. Moreover, Jiang et al. describe intercellular mitochondrial transportation from MSCs to corneal endothelial cells, photoreceptors, and retinal pigment cells. In this paper they showed that the cell that receive the mitochondrion increased respiratory abilities and elevated expression of mitochondrial structure and function related gene [170]. Mitochondrial dysfunction has observed in Leber’s hereditary optic neuropathy (LHON). Recent evidence showed that LHON derives from a genetic mutation in mitochondria, which evolves to optic atrophy, which gives rise to visual acuity and blindness. Thus, the replacement of mitochondria may be a possible solution, which can be achieved by MT utilizing mesenchymal stem cells or their conditioned media derivate [179]. Several authors investigated a possible role of MT in AD.

AD is a chronic neurodegenerative disease and manifests symptoms such as: short-term memory loss, visual-spatial perception disorders and impaired language and executive functions [180]. Several papers showed that MSCs could inhibit amyloid β-peptide (Aβ) generation and promote its effective clearance, alter amyloid precursor protein (APP) processing, decrease tau phosphorylation, and increase proteasomal activity resulting in reduced accumulation of ubiquitin-conjugated proteins [181–183]. In AD has been observed the presence of Aβ in the mitochondria. This presence of Aβ generates hyperphosphorylated tau proteins by interacting with mitochondrial Drp1 protein, which, in turn, disrupts microtubule function and induces neural toxicity [184, 185]. Moreover, mitochondrial Aβ can also interact with Aβ-binding alcohol dehydrogenase (ABAD), leading to mitochondrial dysfunction and the production of ROS [186]. The main consequence of mitochondrial dysfunction, ROS accumulation, and increased oxidative stress are main factors involved in AD pathogenesis. Several studies showed that MSCs may promote microglia and autophagy-mediated clearance of protein aggregates as Aβ [187–189]. MSCs can protect neurons from cell death by secretion of some neuroprotective factors or by MT [52, 190, 191]. Zhan et al. showed that UC-MSC-CM significantly decreased tau phosphorylated at the Thr181 level, which increased in AD-alleviated intracellular and mitochondrial oxidative stress of okadaic acid (OA)-treated SH-SY5Y cells. In addition, UC-MSC-CM suppressed apoptosis and improved mitochondrial function in OA-treated SH-SY5Y cells. In this paper, Zhan et al. showed that UC-MSC-CM exerted protective effects relying on or partly extracellular vesicle (EV) MT from UC-MSCs to OA-treated SH-SY5Y cells [191].

Moreover, it observed that UC-MSC-CM decreased the level of p181-tau in the AD cell model, improved cell viability, and suppressed apoptosis in OA-treated Sh-SY5Y cells. In addition, they showed that UC-MSC-CM improved mitochondrial functions in OA-treated SH-SY5Y cells [192].

Spinal cord injury (SCI) is a destructive neurological disease that causes major motor, sensory and autonomic dysfunction that in some cases can lead blood vessels rupture and vasoconstriction reflexive led to an oxygen reduction and consequent mitochondrial damage. In addition, it is possible to observe a series of damages such as mitochondrial permeability, calcium overload, excitatory toxicity, oxidative stress, and increased ROS production [193]. These phenomenon results in impaired capability to maintain mitochondrial homeostasis with less energy available [194]. In this context, several authors are evolving new strategies to improve secondary injuries such as repairing or replacing damaged mitochondria, the use of antioxidants, and restoring mitochondrial permeability [195]. Li et al. showed that either MSCs or MSC-derived mitochondria injected into the injured spinal cord of a rat contusion SCI model significantly improved locomotor functions 6 weeks after injury [50]. Other authors reported that MSCs are able to improve the secondary injury caused by inflammation, myelin insulation, and assist the angiogenesis process [9, 196–199]. Sykova et al. showed the safety of the use of intravenous and intraarterial delivery of MSCs in SCI patients [200], while Deng et al. used MSCs coupled with collagen in SCI patients and compared with the control group (only collagen) [201]. After 12 months the study showed the treatment group vs control group showed significantly improved American Spinal Injury Association scores and better bowel and urinary functions [44].The ability of cells to interact with other cells via mitochondria has been demonstrated also in other tissues and organs where different modes of MT from MSCs to injured or damaged cells in order to restore or support nonfunctional mitochondria has been discovered [168].

For example, the active transfer from adult stem cells and somatic cells can rescue aerobic respiration in mammalian cells with non-functional mitochondria [36, 168].

Another pathology that may benefit from studies on stem cell-mediated MT is stroke and ischemia–reperfusion injury. A stroke occurs when something blocks the blood supply to part of the brain or when a blood vessel in the brain bursts. Therefore, we can identify two types of strokes: ischemic or hemorrhagic. The blockage of one or more arteries is the main characteristic of acute ischemic stroke, with blood flow reduction and cellular dysfunction, damage, and/or death. The revascularization process is necessary for stroke treatment, but the oxygen and nutrient trasport to the damaged tissues may lead to the activation of the innate and adaptive immune responses that may cause secondary damage to the remaining cells [202, 203]. The hallmark of ischemia/reperfusion process is a mitochondrial dysfunction, ATP production decrease, ROS increase and cellular death [204]. During this process, cell switchs to anaerobic metabolism with a consequent indirect increase of Ca2+ [205]. This Ca2+ overload and oxidative stress lead to the opening of mitochondrial permeability transition pore in the inner mitochondrial membrane and an increase in ROS production [206, 207]. Therefore, MT from other cells could represent a useful tool in management of pathological damage caused by mitochondrial dysfunction in ischemia–reperfusion injury. It has been shown that several brain cells as neurons, astrocytes, endothelial cells, and MSCs are able to transfer mitochondria [171]. In preclinical studies, Liu et al. showed that human bone marrow MSCs could save endothelial cells during hypoxia and nutrients deprivation-induced stress. MSCs could abolish apoptosis in the endothelial cells induced by dysfunctional mitochondria during hypoxia by shifting the functional mitochondria from MSCs through TNTs-like cell protrusions [204]. Babenko et al.

showed that BM-MSCs are able to save astrocytes and PC12 cells during hypoxia and glucose deprivation induced by oxidative stress and mitochondrial damage. Moreover, the same author reported that in the same in vivo study BM-MSCs improve the neurological impairments of cerebral ischemia rats and that this phenomenon is mediated by MT from BM-MSCs via TNTs. Moreover, it has been reported that BM-MSCs with overexpression of Miro1 improve the recovery of ischemic rats [115, 208]. Transplantation of healthy mitochondria has been studied as a solution in rescuing injured cells and tissue for treatment of others different pathologies connected to ischemia vascular diseases. The first studies have been conducted in animal models of ischemia–reperfusion [209]. For example, Mc Cully et al. have been demonstrated that administering mitochondria isolated from the left ventricle of the rabbit to the site of partial ischemia–reperfusion allowed a significant reduction of myocardial infarction and apoptosis markers with a consequent recovery of myocardial infarction [210]. Then, Masuzawa et al. demonstrated in their work the internalization of the transplanted mitochondria and enhanced myocardial energetics despite the lack of demonstration of internalization (by tunneling nanotubes) mechanism [211]. Similar results have been obtained in rat models of ischemia reperfusion. Kaza et al., administering mitochondria in in vivo rat models prior to reperfusion, obtained a decrease of Infarct Size and Area at Risk (IS/AAR) index and altogether enhanced myocardia and cell viability [212]. Similarly, Guariento and coworkers and Blitzer et al. investigated on the therapeutic use single or multiple intracoronary doses of isolated mitochondria before ischemia–reperfusion episodes [213, 214].

The first clinical application of mitochondrial transplantation was carried out to treat myocardial ischemia–reperfusion injury in pediatric patients of Boston Children’s Hospital (United States) [215]. The role of MT has been demonstrated also in SENECA trials where induced patients derived cardiomyocytes (iCM) co-cultured with MSCs, thanks to MSCs EV release, improved their viability and physiology, reducing ROS production and preserving mitochondrial biogenesis. Moreover, some authors suggested a mechanism by which EV from MSCs resulted enriched in mitochondria that probably were transferred to iCM [216]). Also Liu and colleagued have been demonstrated a mechanism by which MSCs can be used as a novel treatment of ischemic vascular disease; they demonstrated a transfer of mitochondria via a tunnelling nanotube-like structure from stem cells to injured human umbilical vein endothelial cells [204].

The effect of dysfunctional mitochondria has been reported also in respiratory and pulmonary diseases. Among these, Chronic Obstructive Pulmonary Disease (COPD) induced by cigarette smoke (CS) is an example of lung diseases characterized by inflammation and damage of cells. In particular, cigarette smoke induces mitochondrial disfunctions in lung epithelial cells. MSCs and MSC-derived exosomes have been proposed as therapeutic intervention of COPD induced by cigarette smoke as reported by Maremanda et al. which demonstrated as BEAS2B-mMSC co-cultures showed protective response against the CSE-altered mitochondrial respiration parameters, confirming the beneficial effect of MSC towards human bronchial lung epithelial cells [217]. Morrison et al. studying the mechanisms of MSCs effects in ARDS demonstrated that MSCs are able to induce in this environment an anti-inflammatory and highly phagocytic macrophage phenotype through EV-mediated MT. In particular, alveolar macrophage treated with MSC-derived EVs ameliorate lung injury in vivo. Transfer of functional mitochondria via EVs determined an increase in Oxidative phosphorylation which led to enhanced phagocytosis and a decrease of TNF-a and IL-8 levels of secretion by macrophages in vitro and in vivo [140]. Similarly, Jackson et al. reported that human bone marrow derived MSCs transfer their mitochondria to macrophages both in vivo and in vitro via TNT and micro vesicle secretion. This leads to enhanced macrophage phagocytosis and improved bioenergetics. So mitochondrial donation represented a novel mechanism to explain the antimicrobial effect of MSCs in a condition determined by bacterial infection providing additional evidence about their therapeutic use in acute, inflammatory lung diseases [64].

Moreover, the effect of mitochondrial transplantation or donation has been reported in other cell pathologies including renal and liver diseases. For example, Kubat et al. observed good therapeutic effects of mitochondrial transplantation in a nephrotoxicity model [218]. Otherwise, Lu et al. demonstrated a nanotherapeutic effect of UC-MSC-EVs on inhibiting local NETs formation by transferring functional mitochondria to intrahepatic neutrophils and repairing their mitochondrial function [219]. In addition, Bi et al. reported a functional therapeutic effect of mitochondrial transfer to combat Non-alcoholic fatty liver disease [220]. By the way, MT has been observed in lung diseases also associated with bioenergetic impairment and dysfunctional mitochondria such as allergic airway inflammation, ARDS or asthma but what signalling events can trigger this mechanism remain elusive.

Mitochondrial transfer and the mechanisms by which mitochondria enter recipient cells are concepts that have gained much attention in recent years. In particular, intercellular communication and mitochondrial transfer in MSCs promise interesting therapeutic results in various pathological conditions. In general, the active transfer of mitochondria from mesenchymal cells to somatic cells could, for instance, restore aerobic respiration in cells with non-functional mitochondria. So, the possible transplantation of healthy mitochondria has been studied as a solution in rescuing injured cells and tissue for treatment of several pathologies such as pulmonary, cardiac, renal, corneal epithelium, and brain cortical. Therefore, understanding the molecular and cellular mechanisms of mitochondrial transfer/transplantation and demonstrating its efficacy could be an important milestone that lays the foundation for future clinical trials.

Conceptualization, I.R., V.M. and S.D.M.; writing—original draft preparation, S.P, I.R., S.D.M. and V.M.; writing—review and editing, S.P., I.R., V.M. and S.D.M.; supervision, I.R., V.M. and S.D.M. All authors have read and agreed to the published version of the manuscript.