To our knowledge, this review is one of the first comprehensive summaries focusing on mitochondrial application of mesenchymal stem cells in ophthalmology. As a novel emerging field, mitochondrial transplantation opens up a new direction for stem cell therapy, addressing the urgent need for innovative therapeutic strategies in ophthalmic diseases. It advances the field by systematically organizing the latest research findings, clarifying the potential of mesenchymal stem cell mitochondria in treating ocular disorders. The review imparts new perspectives on the clinical transformation value of this novel technology, guiding subsequent basic and clinical research, and promoting the development of innovative stem cell-based therapeutic approaches for refractory ophthalmic diseases.

Mitochondria, the “powerhouses of the cell,” generate ATP through oxidative phosphorylation and are essential for calcium signaling, reactive oxygen species (ROS) regulation, apoptosis, and inflammation [1, 2]. Balanced mitochondrial fusion, fission, biogenesis, and mitophagy preserve mitochondrial integrity and function [2]. Mutations in mitochondrial or nuclear DNA disrupt these processes, leading to structural and functional abnormalities collectively termed mitochondrial diseases [3].

The first inherited mitochondrial optic neuropathy described was Leber hereditary optic neuropathy (LHON), caused by three primary mtDNA mutations (ND1 G3460A, ND4 G11778A, and ND6 T14484C). These mutations impair complex I, reduce ATP production, elevate ROS, and selectively damage retinal ganglion cells (RGCs) and the optic nerve [4]. Mitochondrial dysfunction is also a key pathogenic driver in secondary ophthalmic disorders, such as glaucoma, diabetic retinopathy (DR), and age-related macular degeneration (AMD) [5–8].

Conventional therapies—including pharmacological agents, laser treatments, and surgery—slow progression but fail to restore function in severely damaged corneal, retinal, or optic nerve tissues [9]. Mesenchymal stem cells (MSCs) have thus gained attention for their regenerative potential in ocular diseases [10]. While MSCs exert effects through differentiation, paracrine signaling, and immune regulation, growing evidence highlights mitochondrial transfer (MT) as an additional mechanism of repair [11].

This review provides a systematic overview of MT in ocular cells, with emphasis on MSC-mediated MT (Table 1). We discuss its therapeutic applications across diverse ophthalmic diseases, analyze key regulatory factors, and propose perspectives for clinical translation.

Corneal epithelial cells (CECs) are highly susceptible to oxidative stress and mitochondrial dysfunction, contributing to conditions such as dry eye disease and corneal chemical burns [15, 34–37]. It has been observed that intercellular MT occurs among CECs and corneal endothelial cells. Stress conditions further facilitate this process (Figure 1B), implying a damage-induced “SOS” signaling mechanism [21, 38, 39].

In alkali-burned corneal models, MSCs transplantation supported wound healing and mitochondrial repair (Figure 1B). Notably, when tunneling nanotubes (TNTs)-mediated contact was blocked, MSCs lost much of their reparative effect, confirming the essential role of MT [15]. Thus, MT complements paracrine signaling to protect corneal cells from mitochondrial dysfunction. Interestingly, human corneal epithelial cells could internalize exogenous mitochondria isolated from MSCs (Figure 1B). A single subconjunctival injection of isolated mitochondria reduced inflammatory cell infiltration induced by acid burn and accelerated the corneal repair process [16].

Most studies focus on acute injury models, and the long-term stability of transferred mitochondria in chronic corneal diseases remains unclear. Future work should evaluate chronic models, clarify the relative role of MT versus paracrine signaling, and optimize safe delivery methods for clinical translation.

Müller glia (MG) play a central role in retinal homeostasis. Studies in degenerative models show that MG internalize MSCs-derived mitochondria, restoring mitochondrial structure and suppressing gliosis [20]. After ILM disruption, transplanting human embryonic stem cell derived RGCs into the mouse retina resulted in the transfer of donor RGCs substances, including mitochondrial antigens, to recipient MG (Figure 1F) [46]. Moreover, RGCs and cones can donate damaged mitochondria to glial cells for degradation, a process termed transmitophagy (Figure 1G) [47, 48]. This highlights glial cells participation in mitochondrial quality control and neuroprotection.

Although mitochondrial transfer has been implicated in photoreceptor rescue, RPE repair, and Müller glia support, several important questions remain unanswered. The relative contribution of MT versus other forms of material exchange—such as cytoplasmic or nuclear transfer—remains controversial, particularly in photoreceptor transplantation studies. Moreover, while MSCs-derived mitochondria have shown benefits in AMD and retinitis pigmentosa models, differences in therapeutic outcomes between tissue sources (e.g., Wharton’s Jelly vs. endometrium MSCs) highlight the need for systematic comparative studies. The persistence of transferred mitochondria and their functional integration into highly specialized retinal cells also remain poorly defined. Furthermore, transmitophagy appears to play a role in Müller glia support and neuroprotection, but its regulation and therapeutic implications in human disease are not well understood. Future directions should include longitudinal in vivo imaging, single-cell transcriptomic approaches to map donor–recipient interactions, and controlled studies comparing MT to direct mitochondrial transplantation. Ultimately, clarifying these mechanisms will be essential to translate MT into durable therapies for retinal degenerative diseases.

Neurovascular coupling in the retina depends on pericyte communication through interpericyte TNTs (IP-TNTs) [49]. In diabetic retinopathy, pericyte loss and disrupted TNTs contribute to vascular dysfunction [50, 51]. Similarly, glaucoma is associated with impaired IP-TNTs, leading to reduced capillary diameter and blood flow [52, 53]. Whether mitochondrial transfer through IP-TNTs contributes to neurovascular regulation remains an open question.

Previous studies have shown that MSCs exert their immunomodulatory effects by releasing various mediators, including immunosuppressive molecules, growth factors, EVs, chemokines, cytokines, complement components, and various metabolites [54, 55]. MSCs-mediated MT also modulates immune responses (Figure 1H). By transferring mitochondria to T cells, MSCs promote regulatory T cell differentiation and suppress retinal inflammation in ischemia-reperfusion models (Figure 1H) [23, 56–61]. In stroke models, MSCs restored mitochondrial morphology and reduced RGCs loss [24]. These findings suggest MT as a dual mechanism: supporting vascular integrity and regulating retinal immune homeostasis.

The exact role of MT in neurovascular coupling and visual restoration is still uncertain, and current evidence is mainly from rodent models. Future efforts should employ large-animal models and real-time tracking to clarify mechanisms and assess long-term safety.

Under physiological conditions, MT occurs occasionally at a low frequency among ocular cells. Live-cell imaging of ^PhNT-connected photoreceptors revealed merely two definitive MT events across 50 cell pairs during a 60-min observation period [45]. Fluorescence assays performed after 24 h of co-culture showed that the basal transfer rate was below 5% in untreated ARPE19 cells and 4% in CECs, with an even lower rate observed in corneal endothelial cells [21, 39]. The efficiency of MT is shaped by a combination of donor cell characteristics, transfer routes, recipient cell conditions, and the surrounding microenvironment. Among donor-related factors, the source of MSCs, the functional integrity of their mitochondria, and the expression of trafficking proteins such as Miro1 play decisive roles in determining transfer capacity [14, 15, 23, 62, 63]. As superior mitochondrial donors, MSCs achieve much higher transfer efficiency than native ocular cells. In co-culture, the MT rate from MSCs to corneal endothelial cells was approximately 29.42%, greatly surpassing the low spontaneous transfer level among common ocular cells [21]. With respect to transfer routes, TNTs represent the predominant mechanism, and their formation is tightly regulated by cytoskeletal remodeling and Rho GTPase signaling [64]. Pharmacological modulation, such as the use of ROCK inhibitors (e.g., Y-27632), has been shown to enhance TNTs formation and consequently increase MT efficiency (Figure 1I) [39, 65, 66]. At optimal concentrations, Y-27632 raised MT efficiency above 20% in ARPE-19 cells and increased the rate from approximately 4% to 7%–10% in CECs [39, 66]. Besides conventional TNTs, Y-27632 induced Y-NTs, an elongated subtype of intercellular nanotubes, in ARPE-19 cells and facilitated nanotube-dependent mitochondrial transport via cytoskeletal remodeling [66]. Mechanistically, it upregulated Miro1 to regulate mitochondrial dynamics; Miro1 controls mitochondrial movement by interacting with actin filaments and microtubules, thereby boosting intercellular mitochondrial transfer [2]. In addition, alternative pathways—including EVs, gap junctions, and transient cell fusion—also contribute under specific conditions [20, 67–72]. Wang, Y., et al discovered that the release of mitochondria-containing EVs from MSCs is regulated by the CD38/IP3R/Ca²⁺ pathway. Using a non-viral gene delivery vector to overexpress CD38 in MSCs, they generated “super donor” cells that produce “Super-EV-Mitochondria” with three times the mitochondrial yield and improved quality compared to normal MSC-derived EVs [13]. Recipient cell status is another critical determinant; cells experiencing mitochondrial dysfunction or oxidative stress tend to emit “rescue” signals that increase their susceptibility to accepting functional mitochondria [21, 73]. However, this improvement is limited and cannot substantially elevate the overall transfer efficiency. After hyperosmotic injury to CECs, MT increased slightly but remained markedly lower than in the Y27632 group, rendering large-scale cell repair unfeasible [39]. Additionally, recipient ocular cell type affected mitochondrial uptake efficiency; the photoreceptor cell line 661W exhibited a greater capacity to accept mitochondria from co-cultured MSCs than ARPE-19 cells [21]. Finally, the microenvironment exerts a substantial influence, as oxidative stress, hypoxia, and inflammatory signaling pathways, such as TNFα/TNFαip2, NFκB/TNFαip2, promote TNTs formation and intercellular transfer (Figure 1H) [15, 22, 63, 74, 75]. Collectively, these regulatory factors highlight potential intervention points to optimize MSC-mediated MT for therapeutic application in ophthalmic diseases.

Mitochondrial transfer represents a paradigm shift in the understanding of intercellular communication and tissue repair within ophthalmology. Accumulating preclinical evidence demonstrates that MSC-mediated MT can restore cellular bioenergetics, enhance resistance to oxidative stress, and protect both retinal and corneal cells from degeneration. Despite this promise, several critical challenges must be overcome before clinical application can be realized.

Unlike systemic mitochondrial therapy, intraocular mitochondrial transplantation is restricted by the blood-ocular barrier, vitreoretinal barrier and the eye’s immune-privileged microenvironment, which greatly hinder the delivery, survival and functional integration of donor mitochondria. Common routes including intravitreal, subretinal and intravenous injection also have technical defects and potential complications. In addition, non-standard isolation and storage methods reduce mitochondrial activity [41]. The long-term efficacy, biosafety and molecular mechanisms of mitochondrial transplantation in retinal repair remain unclear.

The MitoCatch system, a cell-specific mitochondrial delivery platform, effectively addressed the inadequate targeting capability and low delivery efficiency of conventional mitochondrial transplantation. Its reliability was validated in cellular models, ex vivo human tissues and in vivo animal models. Relevant studies on LHON neurons and mouse optic nerve injury models further confirmed that targeted mitochondrial transplantation reversed mitochondrial dysfunction and prevented cellular degeneration [76]. This technology provided a new direction for the treatment of neurodegenerative diseases, optic atrophy and other intractable mitochondrial disorders.

Current researches mainly verify the short-term protective effects of MSC-mediated MT via in vitro cell models and animal models. These simplified models cannot fully simulate the complex in vivo ocular microenvironment. Moreover, there is a lack of mature quantitative methods to assess MT efficiency and functional recovery in vivo. Existing studies have only observed phenotypic improvements such as improved energy metabolism and relieved oxidative stress, while the regulatory roles of MT in ROS signaling, mitochondrial quality control and cell apoptosis remain poorly understood.

MSCs have been proven superior to ocular cells as mitochondrial donors. Nevertheless, the field still lacks unified criteria for donor screening and tissue-specific delivery schemes. Inconsistent donor sources and immature delivery systems may trigger abnormal cell proliferation and oncogenic risks. Comparative studies between MSC-mediated indirect mitochondrial transplantation and direct mitochondrial transplantation are required to evaluate their operability, efficacy and safety. Systematic assessments of long-term mitochondrial stability, immune responses and therapeutic outcomes in classic models such as AMD and LHON are also urgently needed.

Future studies should first establish standardized quantitative assays to dynamically monitor the in vivo behavior of mitochondria and evaluate therapeutic outcomes. Second, in-depth investigations are required to clarify the regulatory mechanisms of core signaling pathways including PTEN/PI3K/AKT during treatment [77]. Furthermore, given the substantial pathological heterogeneity among various retinal degenerative diseases, individualized transplantation protocols should be formulated according to disease categories and patient-specific conditions.

In summary, MT is a promising therapeutic strategy for retinal degenerative diseases. By correcting mitochondrial dysfunction, it restores cellular energy balance and mitigates oxidative damage. With continuous technological optimization, mechanistic research and safety verification, MT is expected to evolve into a mature clinical therapy for irreversible visual impairment.