Mesenchymal stem cell-derived extracellular vesicles for disease therapy by regulating ferroptosis: focus on diabetes mellitus and diabetic complications

Abstract

Ferroptosis is a novel form of programmed cell death, which has been demonstrated to play a pivotal role in various pathological processes due to its association with iron overload, lipid peroxidation, and dysregulation of the antioxidant system. In recent years, mesenchymal stem cell-derived extracellular vesicles (MSC-EVs) have garnered significant attention as a promising cell-free therapeutic strategy for modulating ferroptosis. This article elucidates the biological characteristics of MSC-EVs and the molecular mechanisms underlying ferroptosis, with a focus on how MSC-EVs regulate ferroptosis through three key pathways: iron metabolism, lipid metabolism, and the antioxidant defense system. Additionally, the therapeutic potential of both natural and engineered MSC-EVs in treating ferroptosis-related diseases is discussed, particularly highlighting their efficacy in diabetes mellitus and diabetic complications. Finally, this article evaluates the challenges and opportunities in translating MSC-EVs-based ferroptosis modulation therapies into clinical applications, providing valuable insights for future research and therapeutic development.

Introduction

Mesenchymal stem cells (MSCs) are a type of multipotent stem cells derived from the mesoderm, with common sources including umbilical cord, bone marrow, adipose tissue, placenta, and dental pulp [1]. MSCs possess multiple characteristics, which lay a solid foundation for their therapeutic applications [2]. For instance, they exhibit the potential to differentiate into various mesodermal cells and, under certain conditions, can transdifferentiate into neural-like or epithelial-like cells across germ layers [3–5]. Additionally, they modulate immune responses through dual pathways: secretion of soluble factors and direct cell-to-cell contact. Moreover, they can sense chemokine gradients at injury or inflammatory sites via surface receptors and migrate directionally to diseased tissues to enhance therapeutic efficacy [6]. Currently, more studies have focused on mesenchymal stem cell-derived extracellular vesicles (MSC-EVs), which are nanoscale membranous vesicles secreted by MSCs, containing bioactive components such as proteins, nucleic acids, lipids, and metabolites. Their functions have been demonstrated in various disease models: (1) alleviating inflammation by regulating macrophage polarization (e.g., sepsis) [7]; (2) accelerating skin wound repair [8]; (3) promoting vascular network reconstruction in ischemic myocardium and regulating vascular homeostasis [9]; (4) rescuing neuronal apoptosis [10]; and (5) mitigating mitochondrial damage to improve acute kidney injury [11]. Compared to the limitations of MSCs, such as low differentiation efficiency and potential tumorigenic risks, these advantages of MSC-EVs—their high penetration and targeting capabilities, low immunogenicity and stability, and cell-free therapeutic safety—are particularly valuable [12]. According to various studies, the drug delivery strategies of MSC-EVs are also more diverse and better adapted to diseases, including: (1) targeted peptide/antibody modification [13, 14]; (2) exogenous drug/nucleic acid loading [15, 16]; and (3) integration with biomaterials or nanotechnology [17, 18].

Ferroptosis, an iron-dependent, lipid peroxidation-driven form of regulated cell death, plays a significant role in various physiological and pathological processes. Its core mechanism involves intracellular iron accumulation and oxidative cascades in membrane lipids rich in polyunsaturated fatty acids, ultimately leading to loss of plasma membrane integrity. Compared to other forms of cell death such as apoptosis, necroptosis, pyroptosis, and autophagy, ferroptosis is distinct in its molecular mechanisms, morphological features, and functional implications [19–21]. Current strategies for regulating ferroptosis primarily encompass two aspects: one involves inhibiting ferroptosis by modulating iron metabolism, suppressing lipid peroxidation, or delivering antioxidants; the other promotes ferroptosis through opposite means, such as increasing the LIP or weakening antioxidant defenses [22]. In this context, MSC-EVs demonstrate a unique capacity for bidirectional regulation of ferroptosis (Tables 1 and 2). This duality primarily stems from differences in MSCs sources, the metabolic characteristics of target cells, and the heterogeneity of their microenvironment. Studies indicate that MSC-EVs can either promote or inhibit ferroptosis in liver disease models. For instance, in acute liver injury models, BMSC-EVs inhibit hepatocyte ferroptosis by delivering SLC7A11, thereby exerting protective effects [23]. Conversely, in the context of liver fibrosis, hucMSC-EVs can transport BECN1 protein, promoting ferroptosis in hepatic stellate cells and subsequently alleviating fibrosis progression [24]. The underlying mechanism for this phenotypic switch may be closely related to the redox status of the target cells themselves. In low to moderate oxidative stress environments (e.g., normal or mildly injured hepatocytes), where intracellular reactive oxygen species (ROS) levels are low, MSC-EVs tend to exert anti-ferroptotic effects by supplementing antioxidant components, maintaining intracellular redox homeostasis and protecting cells from damage. In contrast, in high oxidative stress microenvironments (e.g., fibrotic tissues), where target cells (such as activated hepatic stellate cells) are already at a critical state of oxidative stress, MSC-EVs may amplify lipid peroxidation signaling, thereby promoting ferroptosis in pathological cells and enabling selective clearance. However, current related research cases remain concentrated in limited fields with a narrow scope; therefore, understanding of the dual regulatory nature of MSC-EVs requires caution and should not be overgeneralized. Future studies should also dialectically consider the direction of MSC-EVs’ effects, avoiding simplistic or absolute judgments regarding their dual regulatory capacity.

Table 1.

MSC-EVs promote ferroptosis in liver fibrosis

EVs type	Source	Cargo	Diseases	Cell mode	Target	References
Native	HucMSCs	BECN1	Liver fibrosis	LX-2	SLC7A11/GPX4	[24]
Native	HucMSCs	miR-499a-5p	Liver fibrosis	HSCs/LX-2	ETS1	[25]
Native	BMSCs	miR-26a	Liver fibrosis	LX-2	SLC7A11	[26]
Native	BMSCs	miR-144-3p	Liver fibrosis	LX-2	SLC7A11	[27]

Table 2.

MSC-EVs deliver non-coding RNAs to inhibit ferroptosis in various diseases

EVs type	Source	Cargo	Diseases	Cell mode	Target	References
Native	USCs	lncTUG1	Kidney IRI	HK-2	SRSF1	[28]
Native	MenSCs	miR-let-7	IPF	MLE-12	Sp3	[29]
Native	HucMSCs	miR-Let-7a-5p	CVB3-induced VMC	CMCs	SMAD2	[30]
Native	HucMSCs	miR-214-3p	Neuronal damage caused by ICH	HT22	/	[31]
Native	HucMSCs	miR-17–92	Skin lesions	HUVECs	/	[32]
Native	HucMSCs	miR-138-5p	IDD	Mouse NPCs	MALAT1	[33]
Native	HucMSCs	miR-129-5p	IBD	HCoEpiC/HEK293T	ACSL4	[34]
Native	HucMSCs	miR-1275	Attenuating aged HIRI	Primary hepatocytes/HLOs	SLC39A14	[35]
Native	HucMSCs	miR-125b-5p	DR	HRMECs	P53	[36]
Native	HucMSCs	lncTUBB6	TBI	Mouse cortical neurons	Nrf2	[37]
Native	HucMSCs	circ-BBS2	Ischemic stroke	SH-SY5Y	SLC7A11	[38]
Native	HUCB-MSCs	miR-23a-3p	Myocardial injury after AMI	Myocardial cells	DMT1	[39]
Native	BMSCs	miR-367-3p	EAE	BV2	EZH2	[40]
Native	BMSCs	miR-330-3p	Myocardial IRI	HL-1	BAP1	[41]
Native	BMSCs	miR-223-3p	HBx-induced ferroptosis in podocytes	HPCs	HDAC2	[42]
Native	BMSCs	miR-219-5p	SCI	PC −12	UBE2Z	[43]
Native	BMSCs	miR-194	OGD/R-induced neuronal injury	HBMEC	Bach1	[44]
Native	BMSCs	miR-16-5p	Liver IRI	BRL3A	SLC39A14	[45]
Native	BMSCs	miR-150-3p	SONFH	Primary osteoblasts	BTRC	[46]
Native	BMSCs	lncSNHG7	OA	Chondrocytes	FSP1	[47]
Native	BMSCs	lncMir9-3hg	I/R-induced cardiac injury	HL-1	Pum2	[48]
Native	BMSCs	lncGm36569	ASCI	HT-22/HEK-293 T HEK-293 T	FSP1	[49]
Native	BMSCs	lncGAS5	HF	H9C2	UL3	[50]
Native	BMSCs	circ-Snhg11	DM wounds	EPCs	SLC7A11	[51]
Native	BMSCs	circ-ITCH	DFU	HUVECs	TAF15	[52]
Native	BMSCs	circ-0072464	IDD	NPCs	NRF2	[53]
Native	ADSCs	miR-125b-5p	Sepsis lung injury	PMVEC	Keap1	[54]
Native	ADSCs	miR-760-3p	Cerebral IRI	N2a	CHAC1	[55]
Engineered (HO-1)	BMSCs	miR-124-3p	IRI	IAR20	STEAP3	[56]
Engineered (HO-1)	BMSCs	miR-214-3p	Ischemia-reperfusion injury after steatotic liver transplantation	S-IAR20	COX2	[57]
Engineered (HO-1)	BMSCs	miR-29a-3p	Steatotic liver IRI	SHPs	Ireb2	[58]
Engineered (HO-1)	BMSCs	miR-204-5p	Ischemia-reperfusion injury after steatotic liver transplantation	HIBCs	ACSL4	[59]
Engineered (miR-19b-3p)	ADSCs	miR-19b-3p	ICH	Primary cortical neurons	IRP2	[60]
Engineered (HO-1)	ADSCs	circ-Ash1l	UVB-induced skin lesions	EPCs	GPX4	[61]
Engineered (Hypoxic)	ADSCs	circ-Stt3b	Myocardial injury after myocardial infarction	HL-1	GPX4	[62]
Engineered (Hypoxic)	ADSCs	circ-Wdfy3	Neuronal injury after SCI	HT-22	GPX4	[63]
Engineered (Q10)	HucMSCs	miR-548ai/ miR-606	Diabetic cutaneous wound	HaCaT	ACSL4	[64]
Engineered (SARS-CoV-2-S-RBD)	HucMSCs	miR-486-5p	RILI and RIPF	MLE-12	SMAD2	[65]
Engineered (Strontium)	SMSCs	miR-143-3p	TMJOA	Primary condylar chondrocytes	Mfsd8	[66]

This review outlines the pivotal role of MSC-EVs in the regulatory network of ferroptosis, providing an in-depth analysis of the molecular basis by which MSC-EVs intervene in ferroptosis signaling pathways through the delivery of functional molecules. In particular, we focus on the connection between ferroptosis and diabetes mellitus and diabetic complications, systematically reviewing the potential applications of MSC-EVs-based targeted therapeutic strategies in improving pancreatic β-cell dysfunction, alleviating diabetic nephropathy, and promoting wound healing. This not only expands the understanding of the biological functions of MSC-EVs but also opens new research perspectives for the precise regulation of ferroptosis mechanisms in diabetes mellitus and diabetic complications.

Biogenesis and characteristics of MSC-EVs

The biogenesis of EVs is a highly regulated process, and EVs from different sources generally involve the formation and release of two major vesicle types: exosomes and ectosomes (Fig. 1). The mechanisms can be categorized as follows: (1) Exosomes originate from the endosomal system within cells. Initially, the cell membrane invaginates to form early endosomes, followed by the inward budding of the endosomal membrane to generate intraluminal vesicles (ILVs), which then form multivesicular bodies (MVBs) [67]. This process is regulated by the endosomal sorting complex required for transport (ESCRT): ESCRT-0 recognizes ubiquitinated membrane proteins and clusters them on the endosomal membrane; ESCRT-I/II promotes membrane curvature; and ESCRT-III completes the release of ILVs by severing the membrane neck. Additionally, tetraspanins (e.g., CD63, CD81) and lipids (e.g., ceramide) can facilitate ILV formation through ESCRT-independent pathways. Mature MVBs are transported to the cell membrane via microtubules and, upon fusion with the plasma membrane, release ILVs into the extracellular space, forming exosomes with a diameter of 50–150 nm [68, 69]. (2) Ectosomes are generated through direct budding and shedding of the cell membrane. This process is triggered by the disruption of phospholipid asymmetry in the plasma membrane: calcium influx activates scramblase, causing phosphatidylserine (PS) to flip to the outer membrane leaflet, while the cytoskeleton (e.g., actin) reorganizes under the regulation of Rho/ROCK or ARF6 signaling pathways, leading to local membrane protrusion and shedding, forming ectosomes with a diameter of 100–1000 nm [70]. Exosomes and ectosomes are both nanoscale vesicles secreted by cells, featuring a lipid bilayer membrane structure and carrying bioactive molecules such as proteins, nucleic acids, and lipids from their parent cells. They collectively participate in intercellular communication. However, due to their distinct biogenesis mechanisms, they exhibit functional specificity: exosomes primarily dominate tissue repair and immune regulation, while ectosomes are more focused on microenvironment remodeling.

Fig. 1.

The biogenesis and intercellular communication of EVs. EVs are nanoparticles composed of a phospholipid bilayer, capable of carrying diverse bioactive components such as DNA, RNA, lipids, metabolites, and proteins. Their biogenesis involves the formation mechanisms of exosomes and the direct plasma membrane budding of ectosomes. The uptake of EVs by target cells can be achieved through pathways including endocytosis, receptor-ligand interactions, and membrane fusion

The critical steps in the therapeutic application of MSC-EVs include the isolation, purification, and preservation of EVs [71]. The International Society for Extracellular Vesicles (ISEV) states in MISEV2023 that EV isolation techniques are developed based on their biophysical properties (such as size, density, charge, and surface molecular markers) [72]. These methods exhibit significant trade-offs between specificity and yield, necessitating the selection of an optimal protocol according to research objectives. It is noteworthy that the guidelines particularly emphasize the detailed documentation of technical parameters during isolation, as co-isolated substances (e.g., EVs corona or contaminants adsorbed on the EVs surface) may influence the functional expression of EVs [73, 74]. Furthermore, improper preservation conditions may induce vesicle aggregation, degradation, or leakage of bioactive components, thereby compromising their therapeutic potential [75]. Since MSC-EVs function by directly or indirectly modulating ferroptosis in target cells through the delivery of signaling molecules, the accuracy of these techniques directly determines the reliability of mechanistic studies.

Mechanisms and applications of MSC-EVs in regulating ferroptosis

The molecular mechanisms of ferroptosis involve a complex interactive network encompassing iron homeostasis, lipid metabolism, antioxidant defense, and mitochondrial function. Existing studies indicate that MSC-EVs can deliver bioactive molecules such as functional proteins and non-coding RNAs to target key nodes of ferroptosis, thereby exerting regulatory roles in various diseases. Therefore, in-depth elucidation of the specific molecular mechanisms by which MSC-EVs mediate ferroptosis will not only contribute to uncovering the pathogenesis of related diseases but also provide a critical theoretical foundation for developing diagnostic and therapeutic strategies based on MSC-EVs (Fig. 2).

Fig. 2.

The mechanisms of ferroptosis occurrence. The molecular mechanisms underlying the regulation of ferroptosis through the modulation of iron homeostasis, lipid metabolism, antioxidant defense system, and mitochondrial function

MSC-EVs regulate ferroptosis through iron metabolism pathways and their applications in disease therapy

The role of iron metabolism in ferroptosis

Cellular iron overload serves as a fundamental catalyst for ferroptosis. In the systemic circulation, cells uptake transferrin-bound Fe³⁺ through transferrin (TF) receptor protein 1 (TFR1), which is reduced to Fe²⁺ by six-transmembrane epithelial antigen of prostate 3(STEAP3) and subsequently enters the labile iron pool (LIP) via divalent metal transporter 1 (DMT1). When TFR1-mediated iron uptake increases or iron export proteins (e.g., ferroportin 1 (FPN1)) are inhibited [76], excess Fe²⁺ catalyzes the conversion of hydrogen peroxide (H₂O₂) into highly reactive hydroxyl radicals (·OH) through the Fenton reaction, exacerbating lipid peroxidation and thereby inducing ferroptosis [77]. Studies have reported that the regulation of iron homeostasis also relies on “ferritinophagy,“ [78]. Nuclear receptor coactivator 4 (NCOA4) is the core receptor for ferritinophagy, and its expression level directly determines the rate of ferritin degradation [79–81].

Application of EVs in regulating ferroptosis via iron metabolism pathways for the treatment of various diseases

Recent studies have revealed that MSC-EVs target key molecules involved in iron metabolism (such as FTH, TFRC, DMT1, SLC39A14, etc.), emerging as a critical strategy for modulating ferroptosis. These molecules constitute a cascade pathway encompassing iron uptake (TFRC/DMT1/SLC39A14), storage (FTH), and oxidative stress response (IREB2/STEAP3).

In cardiovascular diseases (e.g., myocardial infarction), MSC-EVs coordinate iron homeostasis through dual mechanisms, suggesting that MSCs from different tissue origins possess distinct molecular expression profiles and may exert divergent effects in analogous pathologies: on one hand, pericardial adipose tissue-derived EVs directly intervene in the IRP2-FTH/TFRC axis via Adipsin protein, upregulating the iron storage protein FTH while suppressing the expression of the iron uptake receptor TFRC [82]; on the other hand, human umbilical cord blood mesenchymal stem cell (HUCB-MSC)-EVs deliver miR-23a-3p, which targets and inhibits the divalent metal transporter DMT1, thereby blocking pathological iron influx [39]. These two mechanisms operate at the levels of iron storage and uptake, respectively. Notably, Adipsin-mediated regulation may be limited by the local microenvironment of pericardial adipose tissue, whereas the broad-spectrum activity of miR-23a-3p renders it more therapeutically promising in AMI. In neurological disorders, miR-19b-3p-modified adipose stem cell (ADSC)-EVs further expand the applicability of this pathway (Fig. 3), achieving neuroprotection by suppressing IRP2 (which post-transcriptionally regulates TFRC and DMT1 expression) [60]. However, since IRP2 is an iron-sensitive protein, its inhibition may disrupt physiological iron metabolism, underscoring the need for precise control of EVs dosage and duration in clinical applications.

Fig. 3.

Engineering Strategies for EVs. The engineering strategies for EVs include changing the nurture environment, combining with physical materials or compounds, 3D microenvironments, and genetic engineering

Liver disease models have unveiled the unique regulatory capacity of MSC-EVs over the non-transferrin-bound iron (NTBI) pathway. SLC39A14, a major NTBI transporter, serves as a common target for miR-16-5p (bone marrow mesenchymal stem cell (BMSC)-EVs) [45] and miR-1275 (urine-derived stem cell (USC)-EVs) [35]. This phenomenon of multiple miRNA co-regulation highlights significant differences in the cargo of MSC-EVs from different sources, and the source-dependent cargo profile may be a key factor determining their direction of promoting or inhibiting ferroptosis. Research on heme oxygenase-1 (HO-1)-enriched BMSC-EVs (HM-EVs) further deepens mechanistic understanding: their cargo, miR-124-3p, directly blocks the execution phase of ferroptosis by inhibiting STEAP3 (a ferrireductase responsible for reducing Fe³⁺ to toxic Fe²⁺) [56]. Complementarily, miR-29a-3p suppresses IREB2 (which regulates TFR1 and DMT1 transcription), redistributing iron via upregulation of FTH1 and downregulation of TFR1 in a fatty liver ischemia-reperfusion injury (IRI) model [58]. Nevertheless, these studies also reveal potential limitations: concurrent inhibition of SLC39A14 and STEAP3 may induce compensatory iron metabolic disturbances, and while HO-1 modification enhances EVs functionality [83] (Fig. 3), its antioxidant effects may obscure the specific regulation of ferroptosis.

In summary, the core advantage of MSC-EVs lies in their ability to selectively activate or suppress iron metabolic nodes according to disease microenvironments (e.g., focusing on IRP2/TFRC in cardiovascular and cerebrovascular diseases, and SLC39A14/STEAP3 in liver diseases). However, current research faces three major challenges: (1) Most mechanisms are based on single miRNA-target pairs, lacking comprehensive validation of pathway cross-talk; (2) The potential interference of EVs with physiological iron metabolism has not been systematically evaluated; (3) Genetic modifications such as HO-1, while enhancing therapeutic efficacy, may introduce uncontrollable off-target effects. Future studies should optimize the precision of MSC-EVs regulation by dynamically tracking panoramic changes in the iron metabolic network post-delivery.

MSC-EVs regulate ferroptosis through lipid metabolism pathways and their applications in disease therapy

The role of lipid metabolism in ferroptosis

Ferroptosis activates a unique lipid metabolic pathway [84]. The hallmark events include: cell membranes are rich in phospholipids containing polyunsaturated fatty acids (PUFA-PLs), which serve as specific substrates for lipid peroxidation, and their abundance enhances susceptibility to ferroptosis [85]. First, long-chain acyl-CoA synthetase 4 (ACSL4) catalyzes the conjugation of free PUFAs with coenzyme A to form acyl-CoA, which is then esterified by lysophosphatidylcholine acyltransferase 3 (LPCAT3) into the sn-2 position of membrane phospholipids (e.g., phosphatidylethanolamine, PE), yielding oxidation-sensitive PUFA-PLs (e.g., PE-AA or PE-AdA) [86]. Second, lipoxygenases (LOXs), particularly 15-LOX, oxidize the bis-allylic regions of PUFAs to generate lipid hydroperoxides (PL-OOH), while cytochrome P450 oxidoreductase (POR) consumes NADPH to produce superoxide anions (O₂⁻), which are further converted to H₂O₂. Together with Fe²⁺, these reactive species amplify free radical reactions [87]. Ultimately, lipid peroxidation products disrupt membrane fluidity and plasma membrane integrity by crosslinking membrane proteins and lipids, leading to cellular disintegration [88].

Application of EVs in regulating ferroptosis via lipid metabolism pathways for the treatment of various diseases

ACSL4 is a key regulator of ferroptosis, which triggers ferroptosis by catalyzing the esterification of PUFAs into phospholipids, thereby promoting the accumulation of lipid peroxides. The aberrant activation of this metabolic pathway is closely associated with the pathological progression of various diseases, including osteoarthritis (OA), inflammatory bowel disease (IBD), and organ IRI.

In OA, BMSC-EVs reverse chondrocyte oxidative stress and ferroptosis by inhibiting METTL3-mediated m6A modification of ACSL4 mRNA, thereby reducing its stability and protein expression [89]. This epigenetic regulatory mechanism highlights the critical role of RNA modification in ferroptosis. Similarly, human umbilical cord mesenchymal stem cell (hucMSC)-EVs directly suppress ACSL4 expression by delivering miR-129-5p in IBD, reducing lipid peroxidation and restoring intestinal barrier function (e.g., upregulation of Occludin and Claudin-1) [34]. Notably, although both EVs target ACSL4, their regulatory mechanisms differ: BMSC-EVs act at the post-transcriptional modification level, whereas hucMSC-EVs directly inhibits translation via miRNA, suggesting that MSC-EVs may precisely regulate ACSL4 activity through a multi-level network.

In the field of organ IRI, the regulation of ACSL4 exhibits greater organ specificity. For instance, USC-EVs alleviate renal IRI by delivering lncRNA TUG1, which dynamically modulates ACSL4 mRNA stability through interaction with the splicing factor SRSF1 [28]. In contrast, HM-EVs inhibit ACSL4 via miR-204-5p, simultaneously ameliorating inflammation and fibrosis in biliary IRI [59]. These studies not only confirm ACSL4 as a common hub for ferroptosis but also reveal that EVs can achieve tissue-specific delivery by carrying non-coding RNAs (e.g., lncRNAs, miRNAs). However, this organ-specific regulation further underscores the dual potential of MSC-EVs functions: MSC-EVs from the same source may dynamically switch between promoting or inhibiting ACSL4 patterns through differential delivery of RNA molecules in distinct organ microenvironments. This necessitates researchers to incorporate both tissue microenvironments and MSCs subtype characteristics into a unified analytical framework. Furthermore, engineered modification techniques have expanded the therapeutic potential of MSC-EVs. For example, FNDC5-modified BMSC-EVs exert neuroprotective effects in ischemic stroke models by targeting the YAP/EGR1/ACSL4 axis via irisin [90], underscoring the advantages of EVs as delivery vehicles (Fig. 3).

Although existing studies demonstrate the broad therapeutic potential of MSC-EVs in regulating ferroptosis via ACSL4, there remains significant room for improvement. First, while effector molecules (e.g., miR-129-5p and miR-204-5p) from different EVs sources all target ACSL4, their tissue tropism and delivery efficiency may vary substantially. Researchers could further optimize carrier design to enhance efficacy. Second, the regulatory mechanisms of ACSL4 are complex, and interactions such as those between m6A modification and RNA-binding proteins remain incompletely elucidated, representing a promising direction for future exploration.

MSC-EVs regulate ferroptosis through the antioxidant system and their applications in disease therapy

MSC-EVs regulate ferroptosis through the GPX4-xCT system and their applications in disease therapy

The role of the GPX4-xCT system in ferroptosis

Glutathione Peroxidase 4(GPX4), a member of the glutathione peroxidase family, relies on glutathione (GSH) as a cofactor to catalyze the reduction of lipid peroxides (LPO) into non-toxic lipid alcohols, thereby inhibiting the chain reaction of lipid peroxidation [91]. The synthesis of GSH depends on cystine uptake mediated by the cystine/glutamate antiporter (System Xc-), and the function of system Xc- is facilitated by the heterodimer composed of SLC7A11 and SLC3A2 [92]. Notably, GPX4 exists in two isoforms: mitochondrial GPX4 (mGPX4) and cytosolic GPX4 (cGPX4), which differ in spatial distribution and function. cGPX4 primarily protects the plasma membrane and organelle membranes, while mGPX4 specifically maintains mitochondrial membrane stability [93–95].

Application of EVs in regulating ferroptosis via the GPX4-xCT system for the treatment of various diseases

The SLC7A11/GPX4 pathway serves as the core metabolic axis regulating ferroptosis, whose function relies on SLC7A11-mediated cystine uptake and GPX4-catalyzed reduction of lipid peroxides, collectively maintaining intracellular redox homeostasis. The activity of this pathway is regulated at multiple levels, including transcription factors, epigenetic modifications, protein interactions, and metabolic reprogramming. Dysregulation of this pathway can lead to lipid peroxide accumulation and ferroptosis. MSC-EVs precisely target key nodes of this pathway, thereby exerting therapeutic effects in various fibrotic and IRI.

The core mechanism of liver fibrosis is closely associated with the aberrant activation of hepatic stellate cells (HSCs). In liver fibrosis, MSC-EVs regulate the SLC7A11/GPX4 axis through a dual mechanism: on one hand, miR-499a-5p in hucMSC-EVs downregulates GPX4 expression by inhibiting the transcription factor ETS1, selectively inducing ferroptosis in HSCs [25]; on the other hand, BMSC-EVs directly suppress SLC7A11 transcription via miR-26a and miR-144-3p [26, 27], synergistically blocking the cystine-glutathione-GPX4 antioxidant axis. Notably, the heterogeneity of EVs cargo significantly influences therapeutic efficacy. For instance, EVs derived from spindle-shaped urine-derived stem cells (SS-USCs) outperform those from rice-shaped subtypes (RS-USCs) in fatty liver ischemia-reperfusion injury due to their high GPX4 protein expression [96], highlighting the importance of cell source selection for therapeutic optimization.

In neurological and cardiovascular diseases, MSC-EVs indirectly activate the SLC7A11/GPX4 pathway via circRNA-miRNA regulatory networks. For instance, circBBS2 in hucMSC-EVs relieves miR-494-mediated suppression of SLC7A11 [38], while BMSC-EVs target EZH2 via miR-367-3p to reverse its epigenetic silencing of GPX4 [40]. These mechanisms reveal a universal strategy whereby MSC-EVs inhibit ferroptosis by enhancing cellular antioxidant defenses in neuro- and cardio-protective contexts, which stands in sharp contrast to their pro-ferroptotic role in liver fibrosis. This further confirms that the directionality of their function depends on the specific pathological milieu and target cell types. Pretreatment strategies significantly enhance EVs efficacy: hypoxia-pretreated hucMSC-EVs promote SLC7A11 expression through the Trx1-mTORC1 axis [97], whereas IL-1β-pretreated BMSC-EVs stabilize GPX4 protein activity via HSPA5 [98]. However, the translational potential of these studies is constrained by the yield of EVs. Emerging 3D culture technology addresses this limitation by enabling the production of high-yield, functionally enhanced EVs, which are enriched with functional active components. These 3D-EVs demonstrate superior therapeutic potential over conventional 2D-EVs in applications such as tissue repair and anti-cellular senescence [99, 100], thereby offering novel strategies for advancing their scalable application (Fig. 3).

In summary, current research on the mechanisms by which MSC-EVs regulate the SLC7A11/GPX4 pathway still has limitations. First, most studies focus on the role of single molecules (e.g., miRNA or circRNA), overlooking the global dynamic regulation of the pathway. For example, post-translational modifications of GPX4 (e.g., ubiquitination) or membrane localization regulation of SLC7A11 remain underexplored. Second, the heterogeneity of EVs cargo may lead to efficacy fluctuations. For instance, BMSC-EVs modified by the deubiquitinase ubiquitin-specific peptidase 10 (USP10) stabilize SLC7A11 through deubiquitination [101], but batch-to-batch variations in USP10 loading may affect reproducibility. Furthermore, engineered EVs (e.g., three-dimensional (3D) printed GelMA (methacrylated gelatin) hydrogel composite system (GelMA-EVs) [102]) enable programmed release, but their long-term safety and immunogenicity require further validation (Fig. 3). Future studies should integrate single-cell sequencing and proteomics to systematically analyze EVs-recipient cell interaction networks and develop standardized production protocols to facilitate clinical translation.

MSC-EVs regulate ferroptosis through the FSP1-CoQ10 system and their applications in disease therapy

The role of the FSP1-CoQ10 system in ferroptosis

Ferroptosis suppressor protein 1 (FSP1) is another key ferroptosis defense factor independent of GPX4. It reduces coenzyme Q10 (CoQ10) to generate CoQ10H₂, which acts as a lipophilic antioxidant to directly capture lipid radicals, thereby blocking the chain reaction of lipid peroxidation [103]. The reducing capacity of CoQ10 relies on the dynamic balance of the intracellular NADH/NAD + ratio, and metabolic reprogramming (an active, adaptive, and precisely regulated transition of metabolic states, such as the enhancement of glycolysis) may influence ferroptosis sensitivity by modulating this ratio [104]. Additionally, FSP1 can reduce vitamin K (VK) to generate VKH₂, activating the non-canonical vitamin K cycle and further expanding its antioxidant functions [105–107].

Application of EVs in regulating ferroptosis via the FSP1-CoQ10 system for the treatment of various diseases

The dual defense system against ferroptosis is constituted by two parallel pathways, FSP1/CoQ10 and DHODH/CoQ10, which maintain plasma membrane redox homeostasis through their mechanisms in ferroptosis.

In the OA model, BMSC-EVs deliver lncRNA SNHG7, which competitively binds to miR-485-5p via a “molecular sponge” mechanism, thereby relieving its transcriptional repression on the ferroptosis suppressor protein FSP1. This leads to reduced levels of malondialdehyde (MDA) and ROS in chondrocytes while inhibiting the release of pro-inflammatory factors [47]. This regulatory axis exhibits similar neuroprotective effects in acute spinal cord injury (ASCI) EVs lncGm36569 maintains FSP1 expression by adsorbing miR-5627-5p, reducing abnormal Fe²⁺ accumulation and lipid peroxidation in neurons [49]. Notably, as a GPX4-independent ferroptosis suppressor, the upregulation of FSP1 not only scavenges free radicals by reducing CoQ10 but also forms a positive feedback loop with vitamin K reductase. However, this mechanism was not thoroughly explored in either study, suggesting that BMSC-EVs may regulate a broader antioxidant network.

MSC-EVs regulate ferroptosis through the GCH1-BH4 system and their applications in disease therapy

The role of the GCH1-BH4 system in ferroptosis

GTP cyclohydrolase 1 (GCH1), as the key rate-limiting enzyme in the synthesis of tetrahydrobiopterin (BH4), effectively reduces intracellular ROS levels by accelerating BH4 production, thereby alleviating oxidative stress and ultimately enhancing cell survival [108]. Its primary mechanism relies on the unique molecular structure of BH4. As an electron donor with a conjugated diene structure, BH4 efficiently captures lipid radicals (LO·/LOO·) through a single-electron transfer mechanism, thereby blocking the chain reaction of lipid peroxidation [109]. Additionally, BH4 can enhance the antioxidant function of CoQ10 by promoting its regeneration [110, 111].

Application of EVs in regulating ferroptosis via the GCH1-BH4 system for the treatment of various diseases

The transcription factor Nrf2, as the core regulatory component in the Nrf2/GCH1/BH4 signaling axis, promotes the synthesis of BH4 by activating its downstream target gene GCH1. BH4 is not only an essential cofactor for the synthesis of monoamine neurotransmitters but also inhibits ferroptosis by stabilizing antioxidant enzymes, thereby maintaining neuronal survival. Yinxin Chen et al. employed Nrf2 knockout experiments to inversely validate the necessity of this pathway—its inactivation led to impaired BH4 synthesis, elevated levels of key ferroptosis markers, and exacerbated neuronal death after SCI [112]. Following intervention with BMSC-EVs, the motor function of SCI rats (as indicated by improved BBB scores) was significantly enhanced, accompanied by mitigation of pathological phenotypes such as syringomyelia formation and inflammatory infiltration.

MSC-EVs regulate ferroptosis through NRF2 (nuclear factor erythroid 2-like 2)-related pathways and their applications in disease therapy>

The role of NRF2-Related pathways in ferroptosis

Under steady-state conditions, NRF2 is maintained at low expression levels by binding to Kelch-like ECH-associated protein 1(KEAP1) and undergoing continuous ubiquitination and degradation [113]. When cells encounter oxidative stress or electrophilic stimuli, the sensor cysteine residues of KEAP1 are modified, leading to the dissociation of NRF2 and its translocation into the nucleus, where it binds to antioxidant response elements (ARE) to initiate the transcription of downstream target genes [114]. Notably, the majority of genes discovered to date that are associated with ferroptosis are involved in the transcriptional regulation of the NRF2 signaling pathway [115–117].

Application of EVs in regulating ferroptosis via NRF2-Related pathways for the treatment of various diseases

During ferroptosis, Nrf2 as a central regulator of the antioxidant response, plays a pivotal role in suppressing ferroptosis by modulating downstream target genes. The activity of Nrf2 is controlled by its negative regulator Keap1, which limits Nrf2 stability through ubiquitin-mediated degradation. Deacetylases such as SIRT1 can further expand the regulatory dimensions of this pathway by modulating the Keap1-Nrf2 interaction or directly enhancing Nrf2 transcriptional activity. Recent studies have revealed that MSC-EVs can target key nodes of the Nrf2 pathway to achieve effective intervention in ferroptosis.

In the field of neural injury repair, MSC-EVs exert protective effects through multi-level regulation of the Nrf2 signaling network. For instance, lncRNA TUBB6 in hucMSC-EVs inhibits post-traumatic brain injury(TBI) inflammation and ferroptosis by stabilizing Nrf2 expression [37], while BMSC-EVs target the ubiquitin ligase UBE2Z via miR-219-5p to reduce Nrf2 degradation, thereby reversing iron accumulation and lipid peroxidation in SCI [43]. Notably, muscle-derived stem cell (MDSC)-EVs further elucidate the dual function of the Keap1-Nrf2-HO-1 axis in peripheral nerve injury—suppressing ferroptosis while enhancing neurotrophic support via upregulated BDNF (brain-derived neurotrophic factor) [118]. This functional duality suggests that the regulatory logic of MSC-EVs should be understood dynamically: their ultimate effects depend on the molecular dialogue network within the microenvironment. Under conditions of excessive oxidative stress in the tissue microenvironment, MSC-EVs may tend to suppress the Nrf2 pathway to promote ferroptosis; whereas in contexts dominated by the demand for neural regeneration, they may achieve comprehensive repair through coordinated activation of multiple pathways.

In respiratory diseases, hucMSC-EVs and ADSC-EVs alleviate lung injury via the Nrf2/HO-1 and SIRT1/NRF2 pathways, respectively [119, 120]. The latter also enhances the antioxidant capacity of pulmonary microvascular endothelial cells by targeting Keap1 through miR-125b-5p [54]. Additionally, menstrual blood-derived stem cell (MenSC)-EVs relieve Nrf2 transcriptional repression via the miR-let-7/Sp3/HDAC2 axis, revealing the role of epigenetic regulation in pulmonary fibrosis [29]. It is worth in-depth exploration that the regulatory characteristic differences exhibited by these MSC-EVs from various tissue sources essentially reflect the adaptive evolution of their parental cells within specific physiological microenvironments—for instance, the enrichment of the miR-let-7 family in MenSC-EVs may be associated with their capacity for epigenetic remodeling required for cyclic endometrial regeneration. This provides a crucial perspective for understanding how MSCs source heterogeneity influences the functions of their EVs.

Research in osteoarticular diseases further extends the application boundaries of the Nrf2 pathway. BMSC-EVs inhibit Nrf2 degradation via the miR-150-3p/BTRC axis in steroid-induced osteonecrosis of the femoral head (SONFH) [46], while in intervertebral disc degeneration (IDD), they rely on the circ_0072464/miR-431/NRF2 axis to delay matrix degradation [53]. This has also been validated in models of chronic kidney disease (CKD) and chemotherapy-induced premature ovarian insufficiency (POI). For example, dental pulp stem cell (DPSC)-EVs alleviate renal fibrosis via the Nrf2/GPX4 pathway [121], while hucMSC-EVs restore ovarian function through the same pathway [122].

Although existing studies have fully demonstrated the universality of MSC-EVs in inhibiting ferroptosis via the Nrf2 pathway, several limitations remain. For instance, excessive activation of Nrf2 may promote tumor cell drug resistance, necessitating a balance between antioxidant and pro-survival effects in therapeutic EVs design. Future research could further explore the interplay between Nrf2 and other cell death pathways (e.g., apoptosis, necroptosis) to advance clinical translation. More importantly, future research needs to establish a systematic framework to explain the functional duality of MSC-EVs: it is proposed to analyze their regulatory direction from three dimensions—(1) the tissue origin and epigenetic memory of the parent cells; (2) the metabolic state and signaling network context of the recipient cells; (3) the interactive relationships among cellular communities within the microenvironment.

MSC-EVs regulate ferroptosis through mitochondrial function and their applications in disease therapy

The role of mitochondrial function in ferroptosis

Mitochondria, serving as the central hubs for cellular energy and signaling, play a critically important role in functional regulation [123, 124]. During ferroptosis, mitochondria exhibit characteristic pathological alterations, including overall volume reduction, increased membrane density, degeneration or disappearance of cristae structures, and outer membrane rupture [125, 126]. The underlying mechanisms involve: Dysregulation of the iron homeostasis feedback network mediated by mitochondrial ferritin, iron regulatory protein 1 (IRP1), and iron-sulfur cluster (Fe-S) formation [127, 128]; Superoxide anion (O₂⁻·) generated via electron leakage from mitochondrial respiratory chain complex I, which is converted to H₂O₂ by superoxide dismutase (SOD) [129], forming a redox microenvironment that—along with free Fe²⁺—initiates lipid peroxidation; Impaired dihydroorotate dehydrogenase (DHODH)-catalyzed dihydroorotate (DHO) oxidation and the concomitant reduction of coenzyme Q (CoQ) to CoQH₂, a process localized on the outer surface of the mitochondrial inner membrane [130, 131]. Additionally, distinct from the above pathways, mitophagy can delay ferroptosis by clearing damaged mitochondria [132].

Application of EVs in regulating ferroptosis via mitochondrial function for the treatment of various diseases

The core mechanism underlying delayed wound healing in diabetes mellitus involves ferroptosis-mediated endothelial dysfunction, while excessive formation of neutrophil extracellular traps (NETs) induces ferroptosis in endothelial cells by releasing pro-inflammatory mediators such as free DNA and myeloperoxidase. Recent studies have revealed that hucMSC-EVs significantly suppress NETs generation by delivering functional mitochondria to neutrophils, promoting mitochondrial fusion, and restoring mitochondrial membrane potential. Mechanistically, hucMSC-EVs activate the PI3K/AKT pathway, reduce the accumulation of ferroptosis markers, and restore endothelial cell proliferation and angiogenesis capacity, thereby achieving dual intervention. In animal models, hucMSC-EVs treatment accelerates wound closure and enhances vascular density and tissue perfusion. Meanwhile, impaired mitochondrial function (e.g., with Rho-EVs treatment) diminishes their therapeutic efficacy, further confirming that mitochondrial transfer constitutes the central mechanism of this process [133].

It is worth mentioning that in recent years, engineered MSC-EVs have demonstrated great potential for synergistic effects in disease treatment by loading exogenous drugs or functional nanoparticles (Fig. 3). They not only serve as delivery vehicles to address issues of targeting and biocompatibility but also exert synergistic effects through their intrinsic biological activity and cargo, enabling multi-level regulation of complex disease pathways. Importantly, these engineering strategies confer novel and precisely designable therapeutic functions, which should be distinguished from the inherent biological functions of EVs [134]. For instance, in the treatment of acute liver injury, Yiwei Tian et al. constructed an MSC-EXO/MnO₂@Dexamethasone (DEX) multifunctional nanoreactor [135]. The EVs utilize their natural homing ability to target the liver, while the loaded MnO2 nanozymes catalytically decompose H₂O₂ to alleviate hypoxia and scavenge ROS. Meanwhile, DEX suppresses the NF-κB pathway to modulate immunity. In osteosarcoma therapy, Wenkai Chen’s team genetically engineered MSC-EVs by modifying their surface with the bone-targeting peptide (BT) SDSSD, constructing a BT-EXO-Capreomycin (CAP) system [136]. This modification enables specific accumulation in bone tissue and subsequent internalization by tumor cells. The loaded anti-tuberculosis drug CAP induces ferroptosis by activating the Keap1/Nrf2/GPX4 pathway. Similarly, another study utilized ANG peptide-modified EVs to target the blood–brain barrier, allowing loaded nanoparticles to accumulate in glioblastoma regions. By delivering drugs or siGPX4, they disrupted the antioxidant defense system and cooperatively induced ferroptosis in tumor cells [137]. Conversely, in a Parkinson’ s disease model, engineered MSC-EVs loaded with dihydrotanshinone I inhibited microglial ferroptosis via activation of the Nrf2–GPX4 pathway [138]. Additionally, membrane proteins competitively bound inflammatory cytokines, thereby blocking neuroinflammation. In summary, MSC-EVs not only function as “active therapeutic agents” but also serve as “intelligent delivery platforms.” Critically distinguishing their inherent functions from their engineered carrier roles, while avoiding overestimation of their capabilities, is essential for realizing their potential in precision medicine.

Application of MSC-EVs in ferroptosis associated with diabetes mellitus and diabetic complications

Recent studies have elucidated the molecular logic by which MSC-EVs modulate ferroptosis in diabetes mellitus and diabetic complications. HucMSC-EVs deliver miR-125b-5p to directly suppress p53 transcriptional activity, thereby relieving its repression on SLC7A11 and maintaining GSH synthesis to protect the retinal barrier [36]. Concurrently, they inhibit JNK phosphorylation to block KEAP1-mediated ubiquitin degradation of NRF2, promoting NRF2 nuclear translocation and activating downstream antioxidant gene expression, thereby ameliorating mitochondrial dysfunction in diabetic nephropathy (DN) [139]. This finding aligns with the mechanism of BMSC-EVs in a diabetes mellitus-sepsis model, where they activate the NRF2/HO-1 pathway while suppressing inflammatory cytokine release and lipid ROS accumulation, improving pulmonary microvascular endothelial cell function [140]. Notably, distinct tissue microenvironments elicit differential regulatory strategies—in diabetic wound studies, BMSC-EVs stabilize Nrf2 mRNA and prolong its half-life via the circ-ITCH/TAF15 complex [52], whereas circ-Snhg11 acts as a molecular sponge for miR-144-3p to relieve its inhibition of SLC7A11, establishing a GPX4-dependent protective mechanism [51]. This “dual-track” regulation highlights the precise adaptation of EVs to tissue-specific oxidative stress.

Further analysis reveals that EVs modification significantly enhances targeting and efficacy: hucMSC-EVs pre-treated with coenzyme Q10 (Q10-EV) enrich miR-548ai/miR-606 to directly silence ACSL4, blocking PUFA phospholipid synthesis [64]. This metabolic reprogramming-based strategy proves more effective than mere antioxidant intervention. Additionally, hucMSC-EVs transfer functional mitochondria to activate the PI3K/AKT pathway, inhibiting NET formation and ameliorating endothelial ferroptosis, while mitochondrial dysfunction attenuates this effect [133], indicating crosstalk between energy metabolism reprogramming and ferroptosis regulation. In diabetic osteogenic impairment, H₂O₂-preconditioned adipose stem cell-EVs (H-EVs) reverse BMSCs senescence and restore osteogenic differentiation via the NRF2/HO-1 pathway [141], further expanding the link between ferroptosis and stem cell aging.

Critical issues emerge from current research: First, some experiments rely solely on high-glucose-induced cell models, failing to fully replicate the metabolically dysregulated microenvironment of diabetic patients. Second, EVs cargo heterogeneity may lead to therapeutic variability—while Q10-EVs significantly enriches specific miRNAs, its in vivo delivery efficiency remains limited by microvascular lesions in target tissues. Notably, Longqing Xia’s team developed a targeted delivery strategy using polyethylene glycol (PEG) and β-cell aptamer-modified hucMSC-EVs, which not only prolongs circulation half-life but also enhances homing to pancreatic β-cells, enabling efficient delivery of AKT/ERK pathway proteins to activate NRF2 and improve insulin secretion [142]. This approach provides a novel solution to overcome targeting bottlenecks in diabetes mellitus therapy, though long-term safety validation is still required (Fig. 4).

Fig. 4.

MSC-EVs improve diabetes complications. EVs derived from MSCs with different sources or engineered modifications can regulate the initiation and progression of ferroptosis through their cargo of functional components, thereby ameliorating diabetes complications

Conclusions and perspectives

The therapeutic potential of MSC-EVs is increasingly evident. However, MSCs exhibit tissue source heterogeneity, and different microenvironments endow MSCs with distinct physiological functions and molecular phenotypes. This “source imprinting” profoundly shapes the cargo composition of their EVs, ultimately influencing their role in ferroptosis regulation. As the primary choice in most studies, BMSC-EVs are enriched with molecules related to basic cellular homeostasis [143], primarily providing robust antioxidant defense and acting as “ferroptosis suppressors.” However, their efficacy is significantly influenced by donor age; BMSC-EVs from elderly donors may exhibit diminished protective cargo, and donor heterogeneity can sometimes exert a greater impact on function than inter-source differences, leading to considerable variability in their effectiveness [144]. In contrast, EVs derived from “younger” and more proliferative cell sources, such as hucMSCs and MenSCs, carry a richer repertoire of regeneration-associated cargo and can potently target ferroptosis pathways. ADSC-EVs present a unique dimension, being enriched with lipid metabolism-related proteins that may more effectively clear lipid ROS and block ferroptosis execution by inhibiting mechanisms such as ACSL4, demonstrating significant advantages in metabolic diseases [145]. USCs, originating from the urinary system, may confer upon their EVs a natural affinity and reparative capacity for renal or bladder cellular injury. Nevertheless, the core appeal of MSC-EVs lies in their role convertibility, i.e., their bidirectional regulatory potential. This bidirectional regulation is highly dependent on the interaction between EVs and the microenvironment: although ADSC-EVs show promise in inhibiting ferroptosis, their potential pro-inflammatory effects (e.g., increased IL-6) in hypoxic microenvironments warrant caution [146, 147]. During early inflammation, while ADSC-EVs suppress ferroptosis, their parental cell characteristics might lead to the release of pro-inflammatory signals into the microenvironment, potentially exacerbating local inflammation and counteracting some protective effects. This underscores the necessity of considering microenvironmental characteristics when selecting EVs sources for research. Future efforts must delve into the interplay among “source-cargo-mechanism-microenvironment”: deciphering the molecular signatures of MSC-EVs subpopulations to accurately predict their functional orientation; investigating the real-time impact of dynamic microenvironmental changes on EVs function; and comprehensively elucidating the precise mechanisms by which MSC-EVs regulate ferroptosis by integrating single-cell sequencing, conditional gene knockout, and standardized microenvironment simulation. Only through these approaches can we ultimately achieve personalized precision therapy based on MSC-EVs.

With deepening research, the application of MSC-EVs has gradually transitioned from experimental stages to clinical trials (Table 3). Currently registered MSC-EVs-related clinical studies not only focus on efficacy evaluation but also systematically assess safety and tolerability, laying a critical foundation for subsequent translation. In terms of administration routes, MSC-EVs exhibit high flexibility, including local injection (e.g., intravitreal injection for retinitis pigmentosa), dressing carriers (for burn wound repair), and topical formulations (e.g., psoriasis ointment). Moreover, the biological properties of MSC-EVs have highlighted their potential in disease diagnostics, with multiple studies confirming their utility as novel biomarkers for chronic kidney disease, diabetes mellitus, and other conditions, offering new approaches for early diagnosis and therapeutic monitoring. In summary, EVs possess dual potential as both therapeutic agents and diagnostic biomarkers, promising transformative roles in regenerative medicine and precision medicine.

Table 3.

Clinical studies on MSC-EVs

No	Study title*	Status	Condition	Type of EVs	Intervention	Clinical trial ID
1	Extracellular Vesicles as Biomarkers for Chronic Renal Failure	Completed	Chronic Kidney Diseases	USC-EVs	/	NCT04700631
2	Urine Extracellular Vesicles: Non-invasive Biomarkers of Β-cell Function and Novel Therapeutic Agents in Diabetes	Active, Not recruiting	Diabetes	USC-EVs	/	NCT06832215
3	Pilot Study of Human Adipose Tissue Derived Exosomes Promoting Wound Healing	Completed	Wounds	ADSC-EVs	Dressing	NCT05475418
4	Adipocyte-Derived Extracellular Vesicles, Weight Loss, and Endothelial Function	Recruiting	Obesity-Related Endothelial Dysfunction	ADSC-EVs	/	NCT06776081
5	Safety of Extracellular Vesicles for Burn Wounds	Completed	Burns	BMSC-EVs	Dressing	NCT05078385
6	Study of ExoFlo for the Treatment of Perianal Fistulas	Completed	Perianal Fistulizing Crohn’s Disease	BMSC-EVs	Local injection	NCT05836883
7	Safety and Efficacy of Stem Cell Small Extracellular Vesicles in Patients With Retinitis Pigmentosa	Recruiting, Phase 2	Retinitis Pigmentosa	BMSC-EVs	Intravitreal injection	NCT06242379
8	Safety and Efficacy of MSC-EVs in the Prevention of BPD in Extremely Preterm Infants	Recruiting, Phase 2	Bronchopulmonary Dysplasia	HucMSC-EVs	Endotracheal administrations	NCT06279741
9	Intra-articular Injection of UC-MSC Exosome in Knee Osteoarthritis	Recruiting, Early Phase 1	Osteo Arthritis Knee	HucMSC-EVs	Intra-articular injection	NCT06431152
10	Safety and Tolerability Study of MSC Exosome Ointment	Completed	Psoriasis	MSC-EVs	Ointment	NCT05523011

^*Study titles are indicated as listed on ClinicalTrials. https://clinicaltrials.gov/.Accessed 12 July 2025

However, current research still faces significant limitations: the lack of standardized dose-effect evaluations across studies undermines the reliability of cross-source efficacy comparisons. Nevertheless, we believe that through global collaborative efforts, the limitations of EVs in clinical applications will be overcome one by one, and their translational value as novel therapeutics and biomarkers will be more fully realized.

Abbreviations

HucMSCs

Human umbilical cord mesenchymal stem cells

BMSCs

Bone marrow mesenchymal stem cells

HSCs

Hepatic stellate cells

USCs

Human urine-derived stem cells

MenSCs

Menstrual blood-derived stem cells

HUCB-MSCs

Human umbilical cord blood mesenchymal stem cells

ADSCs

Adipose-derived stem cells

SMSCs

Synovial mesenchymal stem cells

IRI

Ischemic reperfusion injury

IPF

Idiopathic pulmonary fibrosis

CVB3

Coxsackievirus B3

VMC

Viral myocarditis

ICH

Intracranial hemorrhage

IDD

Intervertebral disc degeneration

IBD

Inflammatory bowel disease

HIRI

Hepatic ischemia reperfusion injury

DR

Diabetic retinopathy

TBI

Traumatic brain Injury

AMI

Acute myocardial infarction

EAE

Experimental autoimmune encephalomyelitis

SCI

Spinal cord injury

OGD/R

Oxygen-glucose deprivation/reoxygenation

SONFH

Steroid-induced osteonecrosis of the femoral head

OA

Osteoarthritis

I/R

Ischemia-reperfusion

ASCI

Acute spinal cord injury

HF

Heart failure

DM

Diabetes mellitus

DFU

Diabetic foot ulcer

UVB

Ultraviolet B-light

RILI

Radiation-induced lung injury

RIPF

Radiation-induced lung fibrosis

TMJOA

Temporomandibular joint osteoarthritis

CMCs

Cardiomyocytes

HUVECs

Human umbilical vein endothelial cells

NPCs

Nucleus pulposus cells

HCoEpiC

Numan colonic epithelial cells

HLOs

Human liver organoids

HRMECs

Human retina microvascular endothelial cells

HPCs

Human renal podocyte cells

HBMECs

Human brain microvascular endothelial cells

EPCs

Endothelial progenitor cells

PMVECs

Pulmonary microvascular endothelial cells

SHPs

Steatotic hepatocytes

HIBCs

Human intrahepatic bile duct epithelial cells

Author contributions

Jiayuan Wu, Zixuan Zhou and Hui Qian have contributed equally to this work.

Funding

This work was supported by the National Natural Science Foundation of China (82172102, 82472170), Zhenjiang Key Laboratory of High Technology Research on Exosomes Foundation and Transformation Application (SS2018003), and the Priority Academic Program Development of Jiangsu Higher Education Institutions (Phase IV, Clinical Medicine).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.