Targeting CH25H/25HC-ferroptosis axis: a novel mechanism of MSC-EVs mediated renoprotection in ischemic AKI

Abstract

Background

Acute kidney injury (AKI) remains a major clinical challenge due to the lack of effective interventions. While mesenchymal stem cell-derived extracellular vesicles (MSC-EVs) show therapeutic promise for AKI, their exact mechanisms are largely to be understood.

Methods

Human umbilical cord-derived MSC-EVs were isolated, characterized, and tested in a murine bilateral renal ischemia reperfusion injury (bIRI) model and in hypoxia/reoxygenation (H/R) treated tubular epithelial cells in vitro. Integrated transcriptomic, miRNA, and biochemical analyses were performed to elucidate the metabolic pathways and molecular mechanisms underlying the renoprotective effects of MSC-EVs.

Results

MSC-EVs preferentially targeted injured kidneys and significantly improved renal function, ameliorated tubular injury, and suppressed inflammation in IRI-AKI. RNA sequencing and targeted metabolomics revealed substantial dysregulation of steroid metabolism after IRI, marked by activation of the cholesterol 25-hydroxylase (CH25H)/25-hydroxycholesterol (25HC) axis. Importantly, accumulated 25HC induced lipid peroxidation and ferroptosis in tubular epithelial cells. MSC-EVs treatment reversed these pathological changes by downregulating CH25H, lowering 25HC levels, and restoring redox homeostasis. miRNA profiling further identified miR-26b-5p as a key MSC-EVs cargo that directly targets the 3′UTR of CH25H mRNA to repress its expression. Notably, inhibiting miR-26b-5p within EVs abrogated their ability to suppress CH25H/25HC-driven ferroptosis, thereby demonstrating its essential role in the metabolic and cytoprotective actions of MSC-EVs.

Conclusions

Our findings unveil the CH25H/25HC axis as a key metabolic checkpoint governing tubular ferroptosis in ischemic AKI. MSC-EVs deliver miR-26b-5p to suppress this axis, thereby rectifying oxysterol metabolism and preventing ferroptosis.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04339-9.

Introduction

Acute kidney injury (AKI) represents a serious clinical condition marked by rapid renal dysfunction, affecting 10–15% of hospitalized patients and over half of critically ill individuals [1, 2]. Despite its high morbidity and mortality, and its established role as a major risk factor for chronic kidney disease, no effective pharmacological therapy currently exists to prevent AKI or enhance recovery [3–5]. This urgent unmet need underscores the importance of developing novel mechanistic and therapeutic insights.

Ischemia-reperfusion injury (IRI) is a leading cause of AKI, frequently occurring in settings such as hypovolemia, cardiac surgery, and kidney transplantation [2, 6]. During IRI, metabolically active tubular epithelial cells, particularly proximal tubules in the corticomedullary junction, are highly susceptible to hypoxia and oxidative stress, leading to energy collapse, mitochondrial failure, and eventual cell death [7–9]. Recent studies have identified ferroptosis, an iron-dependent form of regulated necrosis driven by lipid peroxidation and redox imbalance, as a central mechanism underlying tubular destruction in ischemic AKI, offering new potential targets for intervention [10–12].

Mesenchymal stem cell-derived extracellular vesicles (MSC-EVs) have emerged as promising cell-free therapeutics, delivering bioactive cargo to injured cells and exerting anti-apoptotic, anti-inflammatory, and pro-regenerative effects [13–15]. Compared with whole-cell therapies, MSC-EVs present advantages in safety, stability, and scalability [13]. While their benefits have been demonstrated in various AKI models, including modulation of apoptosis, cell cycle arrest, and autophagy, whether MSC-EVs can regulate the newly recognized ferroptotic death pathway remains unknown [16–18].

Here, using a xeno-free human platelet lysate (HPL) culture system to generate clinical-grade MSC-EVs, we investigated their role in a murine renal IRI model. We identified the cholesterol 25-hydroxylase (CH25H)/25-hydroxycholesterol (25HC) axis as a critical metabolic driver of tubular ferroptosis during ischemic injury. Mechanistically, MSC-EVs deliver miR-26b-5p to suppress CH25H expression, thereby restoring oxysterol homeostasis and protecting tubular cells from ferroptosis. Our findings reveal a previously unrecognized link between MSC-EVs, sterol metabolism, and ferroptosis, providing a molecular basis for the development of EV-based therapies against ischemic kidney injury.

Methods

Isolation, culture, and characterization of MSCs

Primary mesenchymal stem cells were isolated from human umbilical cords. All procedures were approved by the Ethics Committee for Clinical Research of Zhongda Hospital, Southeast University (Approval No. 2023ZDSYLL118-P01). Umbilical cords were rinsed with ice-cold phosphate-buffered saline (PBS) containing 1% penicillin-streptomycin-amphotericin B (PSA, BL141A, Biosharp, China) until residual blood was completely removed. The amniotic membrane, two arteries, and one vein were carefully excised. Wharton’s jelly was minced into 1–3 mm³ fragments, washed twice with PBS, and expanded in MSC NutriStem XF basal medium (05-200-1 A, Sartorius, Germany) supplemented with 5% HPL (A300-100, Sartorius, Germany) at 37 °C in a humidified atmosphere with 5% CO₂. Once visible MSCs migrated out from the explants, the medium was replaced every 2–3 days. MSCs from passages 3–6 were used for subsequent experiments.

To ensure the phenotypic authenticity and functional integrity of MSCs, we performed immunophenotypic characterization and trilineage differentiation assays. MSCs at passage 3 were harvested and resuspended (1 × 10⁶ cells in 100 µL), and incubated for 30 min at room temperature in the dark with fluorochrome-conjugated antibodies against CD29, CD44, CD73, CD90, and CD105 (positive markers, 555443, 555479, 550257, 555596, 560839, BD Biosciences, USA), as well as CD11b, CD34, and CD45 (negative markers, 557321, 555822, 555483, BD Biosciences). After two PBS washes, cells were analyzed using a BD FACSCalibur™ flow cytometer (BD Biosciences, USA). The trilineage differentiation potential of MSCs was evaluated using commercial differentiation media according to the manufacturers’ instructions: osteogenic medium (7531, Sciencell, USA), adipogenic medium (7541, Sciencell, USA), and chondrogenic medium (05455, Stemcell Technologies, Canada). Passage 3 MSCs were induced for 28 days (osteogenesis), 10 days (adipogenesis), or 21 days (chondrogenesis). Differentiation was verified by Alizarin Red, Oil Red O, and Alcian Blue staining, respectively.

Isolation and purification of MSC-EVs

Passage 6 MSCs were cultured to 80–85% confluence, washed three times with PBS, and then incubated in EV-depleted basal medium (DMEM/F12, C11330, Gibco, USA) for 48 h to collect conditioned medium for MSC-EV isolation. MSC-EVs were purified by differential ultracentrifugation. Briefly, the medium was sequentially centrifuged at 2,000 × g for 30 min to remove cells and debris, 15,000 × g for 30 min to eliminate large vesicles, and 100,000 × g for 2 h at 4 °C (Type 45 Ti rotor, Optima XPN-100, Beckman Coulter, USA) to pellet EVs. The pellet was resuspended in PBS and further purified using a 10 mL size-exclusion chromatography (SEC) column (Echo9103A, Echo Biotech, China). According to the manufacturer’s instructions, fractions 4–7 were collected for downstream analyses. Freshly isolated MSC-EVs were stored at −80 °C and used within one month.

Characterization of MSC-EVs

MSC-EVs samples were diluted to a concentration range of 10⁷-10⁹ particles/mL in sterile PBS, and their size distribution was measured using the ZetaView PMX nanoparticle tracking analysis (NTA) system (ZetaView PMX 110, Particle Metrix, Germany). Data acquisition and analysis were performed using the accompanying ZetaView software (version 8.04.02). The morphology of EVs was examined by transmission electron microscope (TEM). EV samples were mixed with glutaraldehyde at a 1:1 volume ratio, and 10 µL of the mixture was applied onto 200-mesh copper grids and incubated for 20 min at room temperature for fixation. Excess liquid was removed with filter paper, followed by negative staining with 2% uranyl acetate for 10 s. Grids were air-dried at room temperature and imaged using a TEM (H-7650, Hitachi, Japan) operated at 80 kV to assess vesicle morphology. The protein concentration of the isolated MSC-EVs was quantified using a bicinchoninic acid (BCA) protein assay kit (KGB2101, KeyGEN BioTECH, China) according to the manufacturer’s protocol. Surface markers including EV-associated (TSG101, ab125011, Abcam, USA; CD63, sc-5275, Santa Cruz, USA; CD81, ab155760, Abcam, USA), endoplasmic reticulum-associated protein (calnexin, ab22595, Abcam, USA), Golgi-associated protein (GM130, ab52649, Abcam, USA) and lipid-associated protein (ApoA1, sc-68755, Santa Cruz, USA) were detected by Western blotting.

Labeling of MSC-EVs

Purified MSC-EVs were resuspended and labeled with DiD or DiO fluorescent dyes (V22889, Invitrogen, USA) at a ratio of 5 µL dye per 1 mL EV suspension (5 × 10¹¹ particles/mL). The mixture was vortexed and incubated for 2 h at room temperature on an orbital shaker under light-protected conditions. Labeled EVs were subsequently purified using SEC columns. Eluted fractions were collected for downstream characterization. NTA was performed to determine particle concentration, whereas nano-flow cytometry was used to assess fluorescent labeling efficiency of MSC-EVs.

Ex vivo distribution of MSC-EVs

To assess the biodistribution of EVs, bilateral renal IRI and sham-operated mice were intravenously injected with DiD-labeled MSC-EVs (1 × 10¹⁰ particles) immediately after reperfusion. Animals were euthanized 6 h post-injection and the heart, lungs, liver, spleen, and kidneys were collected for imaging using the IVIS Spectrum imaging system (PerkinElmer).

Cellular uptake of MSC-EVs in vitro

For uptake studies, DiO-labeled MSC-EVs were incubated with HK-2 cells under normoxia or hypoxia/reoxygenation (H/R) conditions for 12 h. Internalization of EVs was visualized using a super-resolution confocal microscope (Zeiss Elyra 7, Zeiss, Germany) and quantified by counting DiO⁺ puncta per cell. Cytoskeletal organization was assessed with F-actin staining, and nuclei were counterstained with DAPI.

Animals

All animal experiments were conducted in accordance with standard guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee of Southeast University (No. 20220224039). Male C57BL/6 mice (8 to 10 weeks old, weighing 20 to 22 g, Beijing Vital River Laboratory Animal Technology Co., Ltd., China) were anesthetized with pentobarbital sodium (50 mg/kg, Sigma-Aldrich, USA) by intraperitoneal injection and then bilateral renal IRI was established. Bilateral renal pedicles were occluded for 26 min using microaneurysm clamps. Following clamp removal, renal reperfusion was confirmed by visual assessment of color restoration prior to wound closure. Sham-operations underwent identical surgical exposure of kidneys without pedicle occlusion. Core body temperature was maintained at 36.6–37.2 °C using continuous rectal monitoring. At 0, 24, and 48 h post-reperfusion, mice received intravenous injections of MSC-EVs (2 × 10⁹ or 1 × 10¹⁰ particles in 200 µL PBS) or an equal volume of vehicle. For 25HC treatment, 25HC (HY-113134, MCE, USA) dissolved in 20% 2-Hydroxypropyl-β-cyclodextrin (HP-β-CD, HY-101103, MCE, USA) was administered intravenously at 0.05 mg/kg at the same time points. For miRNA-deficient EVs treatment, mice received EVs (1 × 10¹⁰ particles in 200 µL PBS) intravenously at 0, 24, and 48 h after reperfusion. All mice were euthanized three days after reperfusion for tissue and serum collection.

HK-2 cell culture and treatment

Human renal proximal tubular epithelial cells (HK-2) were purchased from American Type Culture Collection and cultured in DMEM/F12 (C11330, Gibco, USA) supplemented with 10% fetal bovine serum (10099–141 C, Gibco, ThermoFisher Scientific, USA) and 1% PSA (PSA, BL141A, Biosharp, China) in a humidified atmosphere of 5% CO₂ at 37 °C. For hypoxia/reoxygenation (H/R) experiments, cells were incubated in glucose- and serum-free medium under hypoxia (1% O₂, 5% CO₂, 94% N₂) for 6, 12, 24, or 48 h, followed by reoxygenation in complete medium under normoxic conditions for 2 h.

Renal function assessment

Blood samples were collected from mice at 72 h after the induction of renal IRI. Serum creatinine and blood urea nitrogen (BUN) levels were quantified using commercial assay kits (C011-2-1, C013-2-1, Nanjing Jiancheng Bioengineering Institute, China) according to the manufacturer’s protocol.

Kidney histology and quantification

Kidneys were fixed in 4% formaldehyde, embedded in paraffin, and sectioned at a thickness of 4 μm. Sections were stained with periodic acid-Schiff (PAS) to assess histopathological changes. Tubular injury was semi-quantitatively scored in 10 randomly selected fields per mouse using the following criteria: 0, no injury (0%); 1, mild injury (< 25%); 2, moderate injury (25–50%); 3, severe injury (50–75%); 4, very severe injury (>75%).

RNA sequencing library construction and data processing

Total RNA was extracted using Trizol reagent (Invitrogen, USA), and RNA integrity was assessed with the Agilent 2200 system. Samples with an RNA integrity number (RIN) ≥ 7.0 were used for library preparation. Strand-specific cDNA libraries were constructed using the TruSeq Stranded mRNA Library Prep Kit (Illumina, USA) following the manufacturer’s protocol. Briefly, poly(A)⁺ mRNA was enriched using oligo(dT) magnetic beads, fragmented, and reverse-transcribed to generate first- and second-strand cDNA. After end repair, A-tailing, and adapter ligation, libraries were amplified and quality-checked before sequencing. Paired-end sequencing (150 bp) was performed on the DNBSEQ-T7 platform. Raw sequencing reads were processed to remove adapters and low-quality sequences, and high-quality clean reads were aligned to the mouse reference genome (mm10) using Hisat2 [19]. Gene-level counts were obtained with HTSeq, and expression values were calculated as FPKM [20]. Differentially expressed genes (DEGs) were identified using the DESeq2 algorithm [21]. Significant DEGs were defined by |fold change| ≥ 2, p < 0.05, and FDR < 0.05 [22]. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were subsequently performed to determine the biological pathways associated with the identified DEGs.

Exosome miRNA library construction and sequencing

miRNA library preparation and sequencing were performed by Echo Biotech. miRNAs were extracted from MSC-EVs using the exoEasy Maxi Kit (QIAGEN) and quantified with a Qubit 2.0 Fluorometer (Life Technologies, USA). RNA integrity was assessed using the Agilent 2100 TapeStation (Agilent Technologies). For library construction, 50 ng of exosomal RNA per sample was utilized to generate small RNA libraries following standard protocols. Sequencing was carried out on the Illumina HiSeq 2500 platform, producing reads ranging from 15 to 41 nucleotides in length.

Dual-luciferase reporter assay

The 3′-untranslated region (3′-UTR) of human CH25H containing the miR-26b-5p target site or the mutation in the seed region of the miR-26b-5p binding site was inserted into the pGL3 luciferase reporter vector (Promega, Madison, WI, USA). HEK293T cells were co-transfected with the reporter plasmid (pGL3-CH25H-3′-UTR or mutant CH25H-3′-UTR vector) and miR-control or miR-26b-5p inhibitors or mimics by Lipofectamine™ 3000 (L3000015, Invitrogen, ThermoFisher, USA) according to the manufacturer’s protocol. Finally, luciferase activity was measured at 48 h post-transfection using a Dual-Luciferase Reporter Assay System Kit (HB-DLR-100, HNIBIO REAGENT, China).

MSCs and HK-2 cells transfection with miRNA

P6 MSCs and HK-2 cells were transfected with miR-26b-5p mimics, miR-26b-5p inhibitor and their corresponding negative controls (HNIBIO REAGENT, China) with Lipofectamine™ 3000. The sequences are as follows: miR-26b-5p inhibitor (5′-ACCUAUCCUGAAUUACUUGAA-3′); NC inhibitor (5′-UCUACUCUUUCUAGGAGGUUGUGA-FAM-3′). miR-26b-5p mimic: Sense: 5′-UUCAAGUAAUUCAGGAUAGGU-3′; Antisense: 5′-ACCUAUCCUGAAUUACUUGAA-3′. NC mimic: Sense: 5′-UCACAACCUCCUAGAAAGAGUAGA-FAM-3′; Antisense: 5′-UCUACUCUUUCUAGGAGGUUGUGA-FAM-3′. All oligonucleotides were synthesized with appropriate chemical modifications according to manufacturer specifications.

25HC-ELISA

The levels of 25HC in HK-2 cell supernatants and kidney tissues were measured using a commercial human/mouse 25HC ELISA kit (YFXEH00741, YFXEM00214, YIFEIXU Biotechnology, China). For HK-2 cells, culture supernatants were collected after H/R treatment and centrifuged at 3,000 rpm for 30 min at 4 °C. The clarified supernatants were then subjected to ELISA analysis. For renal tissues, samples were homogenized in PBS and centrifuged under the same conditions, and the supernatants were used for analysis. 25HC concentrations were normalized to total protein content, as determined by the BCA assay and expressed as ng/g protein.

MDA

The levels of Malondialdehyde (MDA) in the kidney tissues and cells were detected by a lipid peroxidation MDA assay kit (Beyotime; S0131, China) according to the manufacturer’s instructions.

Transmission electron microscopy

Kidney tissues were processed for ultrastructural examination to assess mitochondrial damage characteristic of ferroptosis, including condensed mitochondria with reduced cristae, using standard TEM procedures. The freshly-dissected renal tissues were sectioned into small pieces (< 1 mm³) and fixed in 2.5% glutaraldehyde solution at 4 °C for 2–4 h. Sequential dehydration was performed in a graded ethanol series. Dehydrated specimens were infiltrated overnight with a 1:1 mixture of acetone and epoxy resin (Epoxy 812, GN340, Ted Pella, USA), followed by pure resin infiltration overnight and polymerization at 60 °C for 48 h. Ultrathin sect. (80 nm) were obtained using an ultramicrotome (Leica UC7, Leica Microsystems, Germany) and stained with 2% uranyl acetate in 50% acetone and 0.2% lead citrate. After staining, sections were air-dried overnight at room temperature and examined with a transmission electron microscope (HT7700, Hitachi, Japan).

Cell counting kit-8 (CCK8)

HK-2 cells were cultured in 96-well plates overnight, followed by 25HC treatment with increasing concentrations of 25HC (HY-113134, MCE, USA) at 0–40 µM for 24 h. Cell viability was measured with a CCK8 assay kit (BS350B, Biosharp, China) according to the manufacturer’s protocol.

ROS and BODIPY staining

Reactive oxygen species (ROS) staining was performed using a commercial kit (S0033S, Beyotime, China) according to the manufacturer’s protocols. For kidney tissue lipid peroxidation analysis, BODIPY (D2191, Thermo Fisher Scientific, USA) staining was performed, and for cellular lipid peroxidation, BODIPY 581/591 C11 (S0043S, Beyotime, China) was performed following the manufacturer’s instructions.

Immunohistochemistry staining and immunofluorescence staining

Kidney tissues were first fixed in 4% paraformaldehyde, embedded in paraffin and then sectioned into 4-µm slices using a microtome. Before blocking, the sections were deparaffinized, rehydrated, and subjected to antigen retrieval, after which they were incubated with 3% hydrogen peroxide for blocking. For immunohistochemistry, kidney sections were incubated overnight at 4 °C with primary antibodies against CD63 (ab16669, Abcam, USA), CD68 (ab955, Abcam, USA), Ly6g (PA5-16672, Invitrogen, USA), GPX4 (Catalog No. 67763-1-Ig, Proteintech, China), and ACSL4 (SC-271800, Santa Cruz, USA). After PBS washes, sections were processed using a streptavidin-peroxidase detection kit (KIT-9921, Maixin, China) according to the manufacturer’s instructions, and visualized with diaminobenzidine (DAB-0031/1031, Maixin, China) as chromogen. Slides were digitized using a slide scanner (SLIDEVIEW VS200, Olympus, Japan), and image quantification was performed with ImageJ software (NIH). For immunofluorescence, sections were incubated overnight at 4 °C with primary antibodies against KIM-1 (MA5-28211, Invitrogen, USA), AQP1 (SC-25287, Santa Cruz, USA) and CH25H (ab214295, Abcam, USA), followed by PBS washes and incubation with appropriate fluorescent secondary antibodies. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI), and samples were imaged using a confocal microscope (FV3000, Olympus, Japan).

Western blot analysis

Proteins from EVs, cultured cells, and renal tissues were extracted using a commercial kit (P0013B, Beyotime, China), and concentrations were determined with a BCA assay (KGB2101, KeyGEN BioTECH, China). Equal amounts of protein were separated on 4–20% SDS-PAGE gels and transferred to polyvinylidene difluoride membranes (PVDF, Merck Millipore IPFL00010, USA). After blocking with rapid blocking buffer (KGP902, KeyGEN BioTECH, China) for 15 min at room temperature, membranes were incubated with primary antibodies overnight at 4 °C. Primary antibodies were against KIM-1 (AF1817, R&D Systems, USA), GPX4 (Catalog No. 67763-1-Ig, Proteintech, China), ACSL4 (SC-271800, Santa Cruz, USA), HMOX-1 (Catalog No. 66743-1-Ig, Proteintech, China), CH25H (YN4959, Immunoway, China), Bax (Catalog No. 50599-2-Ig, Proteintech, China), Bcl-2 (Catalog No. 12789-1-AP, Proteintech, China) and GSDMD (PU224937, Abmart, China). Following three washes with TBST, membranes were incubated with HRP-conjugated secondary antibodies (1:10000) for 2 h at room temperature. Protein bands were detected with enhanced chemiluminescence (ECL) reagents and quantified using ImageJ software, with β-actin serving as the internal control.

Real-time fluorescence quantitative PCR (RT-qPCR)

Total RNA from cultured cells and renal tissues was extracted with FreeZol Reagent (R711-01, Vazyme, China) according to the manufacturer’s protocol. cDNA was synthesized from total RNA using HiScript III RT SuperMix for qPCR (R323, Vazyme, China). Quantitative PCR was performed on a QuantStudio 3 system (Applied Biosystems, Thermo Fisher Scientific, USA) with ChamQ SYBR Color qPCR Master Mix (Low ROX Premixed, Q331, Vazyme, China). miRNA reverse transcription and qPCR were conducted with the All-in-One miRNA qRT-PCR Detection Kit 2.0 (Catalog No. QP115, GeneCopoeia, USA) following the manufacturer’s instructions. Relative expression of target mRNAs and miRNAs was calculated using β-actin and U6 as internal references, respectively. Primer sequences are listed in Supplemental Table 1.

Statistical analysis

All statistical analyses and data visualization were performed using GraphPad Prism 10.0 (GraphPad Software, USA). Data are presented as means ± standard deviation (SD). Comparisons between two groups were conducted using the Mann-Whitney test for non-normally distributed data or the unpaired two-tailed Student’s t-test for normally distributed data. For comparisons among multiple groups, one-way analysis of variance (ANOVA) was applied, followed by Dunnett’s test or Bonferroni correction for post hoc pairwise comparisons. P < 0.05 was considered statistically significant. Different assays were performed with sample sizes appropriate to their quantitative or qualitative nature, in accordance with established experimental standards.

Results

Enhanced renal uptake of MSC-EVs following IRI-AKI

MSCs were successfully isolated and expanded under xeno-free, HPL-based conditions. P3 MSCs differentiated into osteogenic, adipogenic, and chondrogenic lineages under lineage-specific induction, verified by Alizarin Red S, Oil Red O, and Alcian Blue staining (Fig. S1A). Flow cytometry confirmed the expected surface marker profile (CD29, CD90, CD73, CD44, CD105 positive; CD19, CD11b, CD34, HLA-DR, CD45, CD14 negative) (Fig. S1B). Moreover, EVs derived from these MSCs displayed canonical structural and molecular features. TEM revealed bilayered vesicles (Fig. 1A), and NTA showed a median diameter of ~144.6 nm (Fig. 1B). Western blotting confirmed enrichment of EV markers CD63, CD81, TSG101, while excluding endoplasmic reticulum (Calnexin), Golgi (GM130), and lipoprotein (ApoA1) contaminants (Fig. 1C). These results collectively confirm that xeno-free MSCs and their EVs meet established standards [23, 24], providing a reliable basis for downstream functional studies.

Fig. 1.

Characterization and biodistribution of MSC-EVs. (A) Representative TEM images of MSC-EVs (Scale bars: 500 and 100 nm). (B) NTA assay showing the size distribution of MSC-EVs. (C) Western blot analysis of EV-associated markers (CD63, CD81 and TSG101) and negative markers for endoplasmic reticulum (Calnexin), Golgi apparatus (GM130), and lipoprotein (ApoA1). (D) Biodistribution of DiD-labeled MSC-EVs in mice (n = 3). (E) Imaging of kidneys from sham and bIRI (the kidneys were isolated from the mice shown in (D) and re-imaged separately, n = 3). (F) Representative fluorescent images show the uptake of DiO-labeled MSC-EVs by hypoxia-treated HK-2 cells (For each sample, four random fields were analyzed, n = 3). Data represent the means ± SD. **p < 0.01, ****p < 0.0001, two-tailed t-test

To evaluate in vivo biodistribution, DiD-labeled MSC-EVs were intravenously administered to sham and IRI mice. MSC-EVs predominantly accumulated in the liver (Fig. 1D), consistent with the well-established hepatic clearance of circulating nanoparticles [25]. Notably, compared with sham controls, IRI kidneys exhibited markedly increased DiD fluorescence (Fig. 1E), indicating preferential uptake by damaged renal tissue. In parallel, hypoxia-exposed HK-2 cells internalized substantially more DiO-labeled MSC-EVs than normoxic cells (Fig. 1F). Together, these results demonstrate that MSC-EVs exhibit preferential tropism toward injured kidneys.

MSC-EVs attenuate renal dysfunction and tubular injury after IRI

To evaluate the therapeutic potential of MSC-EVs, a bilateral IRI model was established, and mice received PBS or MSC-EVs once daily at 0, 24, and 48 h after reperfusion (Fig. 2A). Bilateral IRI markedly increased serum creatinine and BUN levels (Fig. 2B) and induced severe tubular injury, characterized by TEC swelling, vacuolar degeneration, necrosis, and luminal cast formation (Fig. 2C-D). MSC-EVs treatment significantly improved renal function and attenuated these histopathological damage. Consistently, LTL staining showed preserved proximal tubular integrity in MSC-EVs-treated mice compared with IRI controls (Fig. 2E-F). At the molecular level, MSC-EVs reduced the expression of tubular injury markers KIM-1, NGAL, and Socs3, while upregulating the protective factor Hnf1 (Fig. S2). Western blot analysis further confirmed a marked decrease in KIM-1 protein levels following MSC-EVs treatment (Fig. 2G). Moreover, MSC-EVs reduced renal inflammatory responses, as demonstrated by reduced expression of IL-18, CXCL10, CCL2, and TNF-α (Fig. 2H), along with diminished infiltration of CD3⁺ T cells, CD68⁺ macrophages, and Ly6g⁺ neutrophils (Fig. 2I-J). Together, these findings demonstrate that MSC-EVs significantly protect against renal dysfunction, tubular injury, and inflammation in IRI-induced AKI.

Fig. 2.

MSC-EVs attenuate renal injury following IRI-AKI. (A) Schematic diagram of the experimental design. Briefly, MSC-EVs were administered at 0, 24, and 48 h after bilateral renal ischemia. (B) Serum creatinine and blood urea nitrogen (BUN) levels in different groups (sham: n = 9; bIRI: n = 8; L-MSC-EVs: n = 7; H-MSC-EVs: n = 9). (C-D) Representative images of PAS staining and quantification of tubular injury score (scale bar: 50 μm; sham: n = 9; bIRI: n = 8; L-MSC-EVs: n = 7; H-MSC-EVs: n = 9). (E–F) Representative images of Lotus tetragonolobus lectin (LTL) staining and semiquantification of LTL⁺ proximal tubules (scale bar: 50 μm; n = 6). (G) Western blot and quantification of KIM-1 in renal tissues (n = 3). (H) RT-qPCR analysis of inflammatory cytokines and chemokines, including IL-18, CXCL10, CCL2, and TNF-α (n = 6). (I-J) Representative immunohistochemical staining and quantification showing renal infiltration of inflammatory cells: T cells (CD3), macrophages (CD68), and neutrophils (Ly6g) (scale bar: 50 μm; n = 3). Data represent the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA. (L-MSC-EVs: 2 × 10⁹ particles in 200 μL PBS; H-MSC-EVs: 1 × 10¹⁰ particles in 200 μL PBS)

MSC-EVs regulate steroid metabolic disturbances via CH25H/25HC axis in IRI-AKI

RNA-seq analysis of kidney tissues was performed to identify the potential molecular mechanisms underlying MSC-EVs-mediated renoprotection. Sham, bIRI, and MSC-EVs-treated kidneys exhibited distinct transcriptional profiles (Fig. S3A-B), and a subset of genes altered by IRI was reversed after MSC-EVs treatment (Fig. S3C-E), suggesting partial restoration of IRI-induced transcriptional alterations. GO enrichment analysis revealed marked alterations in metabolic pathways, with lipid metabolism and steroid metabolism being most prominently affected (Fig. 3A). Cholesterol, the core precursor for steroid biosynthesis, is metabolized into multiple steroid metabolites through enzyme-driven pathways [26] (Fig. 3B). Heatmap analysis showed that IRI-induced alterations in steroidogenic genes, including CH25H and Cyp46a1, were partially mitigated by MSC-EVs (Fig. 3C). Among these, CH25H, the rate-limiting enzyme catalyzing 25HC production, was strongly upregulated after IRI and downregulated following MSC-EVs treatment (Fig. 3D). In contrast, other cholesterol metabolism-related enzymes shown in Fig. 3B, including Cyp11a1, Cyp27a1, and Cyp46a1, did not show statistically significant differences among the experimental groups (Fig. S4). Western blot and immunofluorescence further confirmed elevated CH25H expression in injured proximal tubular cells after IRI, which was attenuated by MSC-EVs treatment (Fig. 3E-G), accompanied by corresponding changes in renal 25HC levels (Fig. 3H). In vitro, 25HC accumulation was observed in HK-2 cells during early H/R (Fig. 3I). Collectively, these findings indicate that MSC-EVs partially reverse IRI-induced transcriptional and metabolic reprogramming in the kidney, with the CH25H/25HC axis representing one of the affected pathways.

Fig. 3.

MSC-EVs modulate renal steroid metabolism after IRI. (A) Gene ontology (GO) enrichment of differentially expressed genes (DEGs) highlighted metabolic pathway changes by MSC-EVs. (B) Schematic diagram of cholesterol metabolism according to previous literature. (C) Heatmap showing the expression of key metabolic enzymes according to RNA-seq data. (D) Renal CH25H mRNA expression detected by RT-qPCR (n = 7). (E–F) Western blot analysis and densitometric quantification of CH25H protein in renal tissue (n = 4). (G) Immunofluorescence staining showing CH25H expression in AQP1⁺ proximal tubules. (CH25H: red; AQP1: green; cell nuclei marker DAPI: blue; scale bar: 25 μm). (H) Renal 25-hydroxycholesterol (25HC) levels in different treatment groups determined by ELISA (sham: n = 5; bIRI: n = 5; L-MSC-EVs: n = 6; H-MSC-EVs: n = 9). (I) 25HC concentrations in HK-2 supernatants after hypoxia/reoxygenation (H/R) at various time points, determined by ELISA (n=3). Data represent the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA. (L-MSC-EVs: 2 × 10⁹ particles in 200 μL PBS; H-MSC-EVs: 1 × 10¹⁰ particles in 200 μL PBS)

MSC-EVs protect tubular cells by attenuating 25HC-induced ferroptosis

25HC is a multifunctional oxysterol capable of reshaping lipid metabolic pathways and exerting diverse immunomodulatory effects, which can be anti- or pro-inflammatory depending on the biological context [27–29]. To investigate how tubular cells respond to 25HC accumulation, HK-2 cells were treated with increasing concentrations of 25HC. CCK-8 assays demonstrated a dose-dependent reduction in cell viability (Fig. S5A), indicating 25HC-induced cytotoxicity. However, the underlying mode of cell death remained unclear. Western blot analysis revealed no significant changes in apoptosis- or pyroptosis-related proteins, whereas the pro-ferroptotic enzyme ACSL4 was markedly upregulated (Fig. S5B-C), suggesting that 25HC primarily induces ferroptosis. This was further supported by increased lipid peroxidation (Fig. S6A-B), elevated ROS levels (Fig. S6A), and higher MDA content (Fig. S6C), along with upregulation of ACSL4 and HMOX-1, and downregulation of GPX4 (Fig. S6D-E). Furthermore, TEM revealed shrunken mitochondria with condensed membranes, confirming ferroptotic ultrastructural changes (Fig. S6F). Together, these results indicate that pathological 25HC accumulation drives tubular injury predominantly by sensitizing cells to ferroptosis.

Next, we explored whether MSC-EVs confer renoprotection by suppressing the CH25H/25HC metabolic axis. Consistent with our earlier findings, MSC-EVs treatment markedly improved survival (Fig. 4A), lowered serum creatinine and BUN levels (Fig. 4B-C), and attenuated tubular injury in IRI mice (Fig. 4D-E). However, all these benefits were completely abolished when 25HC was co-administered, and 25HC alone further exacerbated renal dysfunction (Fig. 4A-E). Regarding ferroptosis, MSC-EVs reversed IRI-induced changes in ACSL4, HMOX-1, and GPX4 expression (Fig. 4F-I) and suppressed lipid peroxidation, as shown by decreased MDA levels and BODIPY-C11 oxidation (Fig. 4J-K). Importantly, all of these protective effects were fully negated by 25HC. Collectively, these results establish that 25HC accumulation drives ferroptosis and kidney injury, while MSC-EVs provide potent anti-ferroptotic effects by directly targeting the CH25H/25HC axis.

Fig. 4.

MSC-EVs target the CH25H/25HC axis to mitigate ferroptosis following IRI. (A) Survival rate of mice. (initial group size: sham: n = 6; bIRI: n = 9; MSC-EVs: n = 8; MSC-EVs+25HC: n = 9, 25HC: n = 10. Survival data shown are based on the initial number of animals in each group). (B-C) Serum creatinine and BUN levels in different groups (sham: n = 6; bIRI: n = 7; MSC-EVs: n = 7; MSC-EVs+25HC: n = 7, 25HC: n = 5). (D-E) Representative images of PAS staining and semiquantification for tubular injury score (scale bar: 50 μm; sham: n = 6; bIRI: n = 7; MSC-EVs: n = 7; MSC-EVs+25HC: n = 7, 25HC: n = 5). (F–G) Representative immunohistochemical staining and quantification of ACSL4 and GPX4 expression (n = 3). (H-I) Western blot analysis and quantification of ACSL4, GPX4, and HMOX-1 in renal tissues (n = 3). (J) Malondialdehyde (MDA) levels measured by commercial assay kit (sham: n=6; bIRI: n = 7; MSC-EVs: n = 7; MSC-EVs+25HC: n = 7, 25HC: n = 5). (K) Representative images of Boron-dipyrromethene C11 (BODIPY C11) staining (scale bar: 50 μm; 10 μm; n = 3). Data represent the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA

MSC-EV enriched miR-26b-5p directly targets CH25H

Previous studies have shown that miRNAs are key functional components of MSC-EVs [30, 31] and that CH25H expression is markedly reduced following MSC-EVs treatment. We hypothesized that MSC-EVs deliver miRNAs to suppress CH25H. Small RNA sequencing identified 400 miRNAs in MSC-EVs (Supplementary Table 2), with the 30 most abundant shown in Fig. 5A. TargetScan analysis predicted several of these, including miR-221-3p, miR-26b-5p, and miR-93-5p, as potential CH25H regulators (Fig. 5B). To functionally assess these candidates, miRNA mimics and inhibitors were transfected into HK-2 cells, followed by quantification of CH25H expression. Notably, miR-26b-5p markedly reduced CH25H mRNA and protein levels (Fig. 5C-D, Fig. S7), whereas miR-221-3p and miR-93-5p had no significant effect. Further supporting this specificity, dual-luciferase reporter assays demonstrated that miR-26b-5p directly binds to the predicted site within the CH25H 3′UTR (Fig. 5E-F). In vivo, MSC-EVs administration increased renal miR-26b-5p abundance (Fig. 5G), indicating that MSC-EVs deliver miR-26b-5p to tubular cells, thereby suppressing the CH25H/25HC axis.

Fig. 5.

MSC-EVs enriched miR-26b-5p directly targets CH25H. (A) Heatmap showing the top 30 miRNAs enriched in MSC-EVs. (B) Bioinformatic prediction identifying miRNAs potentially regulating CH25H via TargetScan databases. (C) RT-qPCR analysis of CH25H mRNA levels in HK-2 cells transfected with miRNA mimics or inhibitors (n = 6). (D) Western blot and quantification of CH25H protein expression in HK-2 cells (n = 6). (E) Predicted binding sites between miR-26b-5p and the CH25H 3′ untranslated region (3′UTR). (F) Dual-luciferase reporter assay showing direct inhibition of CH25H by miR-26b-5p (n = 3). (G) In vivo validation of increased renal miR-26b-5p expression after MSC-EVs treatment (n = 6). Data represent the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns (not significant), one-way ANOVA(C, D, F), two-tailed t-test (G)

Exosomal miR-26b-5p mediates ferroptosis suppression via CH25H/25HC axis

To assess the role of miR-26b-5p, MSCs were transfected with a miR-26b-5p inhibitor to generate miR-26b-5p-deficient EVs (inhi-EVs) (Fig. S8). In vitro, unlike control MSC-EVs (NC-EVs), inhi-EVs failed to suppress H/R-induced CH25H upregulation and 25HC accumulation (Fig. 6A-C), and consequently did not alleviate ferroptotic injury, as evidenced by elevated ROS levels, increased lipid peroxidation (Fig. 6D-E), and disrupted ferroptosis-related protein expression (Fig. 6F). These findings indicate the critical role of miR-26b-5p in MSC-EV-mediated inhibition of the CH25H/25HC/ferroptosis axis.

Fig. 6.

Exosomal miR-26b-5p attenuates 25HC-induced ferroptosis by targeting CH25H in vitro. (A-B) Western blot analysis and quantification of CH25H in HK-2 cells (n = 3). (C) Intracellular 25HC levels detected by ELISA. (D) Representative images of ROS staining (scale bar: 100 μm; n = 3). (E) Representative images of BODIPY C11 staining (scale bar: 20 μm; n = 3). (F) Western blot and quantification of ACSL4 and GPX4 in HK-2 cells (n = 3). Data represent the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA. NC-EVs: control MSC-EVs; inhi-EVs: miR-26b-5p-deficient EVs

In IRI mice, NC-EVs enriched with miR-26b-5p markedly improved renal function and alleviated tubular injury, whereas inhi-EVs showed minimal protection (Fig. 7A-C). Consistently, depletion of miR-26b-5p in MSC-EVs abolished their ability to reduce renal KIM-1 expression (Fig. 7D). Mechanistically, NC-EVs effectively suppressed CH25H expression and decreased renal 25HC accumulation, effects that were abolished upon miR-26b-5p inhibition (Fig. 7E-G). Consequently, miR-26b-5p-deficient EVs failed to influence ferroptosis, as reflected by elevated renal MDA levels (Fig. 7H), increased BODIPY-C11 oxidation (Fig. 7I), dysregulated ACSL4/GPX4 expression (Fig. 7J-L), and the presence of shrunken, electron-dense mitochondria (Fig. 7M). Taken together, these data identify miR-26b-5p as an essential functional cargo within MSC-EVs. By targeting CH25H and limiting 25HC accumulation, miR-26b-5p is required for MSC-EV-mediated protection against ferroptosis and renal IRI.

Fig. 7.

Exosomal miR-26b-5p protects against renal IRI-induced ferroptosis via CH25H/25HC modulation. (A) Serum Creatinine and BUN levels in different groups (sham: n = 5; IRI: n = 9; NC-EVs: n = 12; inhi-EVs: n = 7). (B) Quantification for kidney injury score (sham: n = 5; IRI: n = 9; NC-EVs: n = 12; inhi-EVs: n = 7) (C) Representative images of PAS staining (scale bar: 50 μm). (D) Immunofluorescence staining of KIM-1 (scale bar: 50 μm; n=3). (E) Western blot and quantification of KIM-1 expression in renal tissues (n = 6). (F) Western blot analysis and quantification of CH25H expression in renal tissues (n = 3). (G) Renal 25HC levels in different groups (n = 3). (H) MDA levels in different groups (sham: n = 5; IRI: n = 9; NC-EVs: n = 12; inhi-EVs: n = 7). (I) Representative BODIPY C11 staining (scale bar: 50 μm and 10 μm; n = 3). (J) Western blot and quantification of ACSL4 and GPX4 in renal tissues (n = 3). (K-L) Representative immunohistochemical staining and quantification showing ACSL4 and GPX4 expression in different groups (scale bar: 50 μm; n = 3). (M) Representative transmission electron microscope (TEM) images of renal tissues from sham, bIRI, NC-EVs- or inhi-EVs treated mice. Ferroptotic mitochondria are indicated by mitochondrial shrinkage and increased electron density (scale bar: 500 nm). Data represent the means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns (not significant), one-way ANOVA. NC-EVs: control MSC-EVs; inhi-EVs: miR-26b-5p-deficient EVs

Discussion

AKI remains a major clinical challenge, largely due to the absence of therapeutic strategies to prevent tubular cell death and facilitate renal recovery [2]. Tubular cell death in AKI is heterogeneous, encompassing apoptosis, ferroptosis, pyroptosis, and other regulated forms of cell death [32, 33]. Among these, ferroptosis has emerged as a key driver of tubular injury in IRI-induced AKI [34]. Increasing evidence highlights that metabolic dysregulation, particularly disturbances in lipid metabolism, contributes critically to tubular ferroptosis during AKI [35, 36]. Our study uncovers a novel mechanism by which MSC-EVs protect against IRI-AKI by suppressing tubular ferroptosis via the CH25H/25HC axis, with miR-26b-5p serving as a critical EV cargo that directly targets CH25H, thereby limiting 25HC accumulation and mitigating ferroptosis-driven tubular injury (Fig. 8). Collectively, these findings uncover a previously unrecognized link between oxysterol metabolism and ferroptosis in AKI, offering new mechanistic insight into MSC-EVs based renoprotection.

Fig. 8.

Schematic illustration showing MSC-EVs regulate tubular cells ferroptosis sensitivity through CH25H/25-HC metabolic axis. As a key component highly enriched within MSC-EVs, miR-26b-5p directly targets CH25H, limits aberrant 25HC accumulation, and consequently alleviates ferroptosis-mediated tubular injury

Previous studies have consistently reported that MSC-EVs exert renoprotective effects in AKI, primarily through modulation of TEC survival and repair pathways [37, 38]. Consistent with these findings, we confirmed that MSC-EVs significantly improve renal function and histological recovery following IRI. Importantly, our findings extend beyond these established paradigms by identifying cholesterol-derived oxysterol metabolism as a previously underappreciated pathological process that can be alleviated by MSC-EVs. TECs are highly susceptible to metabolic reprogramming during AKI, as their energy metabolism shifts from fatty acid β-oxidation toward glycolysis in response to hypoxia, mitochondrial dysfunction, and impaired nutrient sensing pathways [39]. This maladaptive metabolic shift impairs tubular recovery and promotes sustained inflammation, lipid accumulation, and fibrosis, thereby increasing TEC vulnerability to metabolic stress and cell death and ultimately contributing to the progression from AKI to CKD [40–42]. Therefore, targeting early metabolic reprogramming may represent a promising strategy to halt the transition from AKI to CKD [43]. Nevertheless, the role of MSC-EVs in modulating early metabolic alterations during AKI has not been fully elucidated. Our findings extend the mechanistic scope of MSC-EVs therapy, revealing that MSC-EVs can reprogram cholesterol metabolism via the CH25H/25HC axis, a previously unrecognized pathway in AKI.

Emerging evidence indicates that cholesterol oxidation products, particularly oxysterols, contribute to renal injury. In early ischemic AKI, both 25HC and 27HC are markedly increased, with 27HC identified as a key driver of tubular ferroptosis, highlighting the importance of oxysterol dysregulation in ferroptosis [44]. Consistently, inhibition of 7-dehydrocholesterol reductase has also been shown to improve renal IRI by modulating cholesterol intermediate metabolism [45]. However, the specific contribution of CH25H-derived 25HC to tubular injury has remained unclear. Given the multiplicity of regulated cell death pathways in AKI, we sought to clarify the specific modality associated with CH25H-derived 25HC. In vitro, exposure to 25HC predominantly induced ferroptosis-related changes, whereas markers of other regulated cell death pathways remained largely unchanged under the same conditions. This observation justified prioritizing ferroptosis in our study. Our findings identify CH25H-driven 25HC accumulation as a previously unrecognized metabolic trigger of tubular ferroptosis following renal IRI.

CH25H catalyzes the conversion of cholesterol into 25HC, a potent oxysterol known to regulate lipid homeostasis, inflammation, and apoptosis [28, 46, 47]. In ischemic kidneys, CH25H was markedly upregulated, leading to excessive 25HC accumulation. Mechanistically, 25HC may promote ferroptosis through multiple interconnected processes. Consistently, we observed upregulation of ACSL4, an enzyme known to facilitate incorporation of polyunsaturated fatty acids into phospholipids and widely implicated in ferroptosis [48]. Excessive 25HC was also associated with elevated ROS levels and reduced GPX4 expression, suggesting potential impairment of the antioxidant defense system that normally constrains lipid peroxidation [49]. Furthermore, induction of HMOX-1 suggests alterations in redox balance and iron handling, which may contribute to enhanced oxidative lipid damage [50]. Collectively, these findings support a model in which CH25H-mediated accumulation of 25HC synergistically disrupts lipid metabolism and redox balance by promoting lipid peroxidation and impairing the antioxidant defenses, thereby increasing the susceptibility of tubular cells to ferroptosis. The specific mechanisms underlying these effects remain to be further investigated. Importantly, MSC-EVs administration effectively reversed these changes, restoring redox homeostasis and mitochondrial integrity; however, these protective effects were completely abrogated when 25HC was co-administered. This indicates that MSC-EVs alleviate ferroptosis at least in part by suppressing the CH25H/25HC axis. Overall, these results highlight the therapeutic potential of targeting cholesterol metabolism in the treatment of ferroptosis-associated kidney injury.

To further dissect the molecular mediators of this effect, we examined MSC-EV cargo composition. Extracellular vesicles are increasingly recognized as potent intercellular communicators, whose biological activities largely depend on their cargo, particularly miRNAs, which play a central role in post-transcriptional gene regulation and mediate the functional effects of MSC-EVs on recipient cells [51–53]. Here, miRNA sequencing revealed that miR-26b-5p is highly enriched in MSC-EVs and directly targets the 3′UTR of CH25H mRNA, suppressing its expression. Notably, inhibition of miR-26b-5p within MSC-EVs markedly reduced their ability to suppress the CH25H/25HC metabolic axis. These findings identify miR-26b-5p as a critical mediator linking MSC-EVs to the regulation of cholesterol-derived oxysterol metabolism and ferroptosis. Although miR-26b-5p has previously been implicated in the control of lipid metabolism and redox balance in cardiovascular disorders [54], its role in renal injury had not been characterized. Our study therefore provides evidence that miR-26b-5p can modulate the CH25H/25HC axis in the kidney, establishing a novel post-transcriptional regulatory mechanism through which MSC-EVs exert metabolic protection against ischemic injury. This discovery broadens the mechanistic scope of MSC-EV therapy from general effects (immunomodulation, anti-apoptotic and cell proliferation and so on.) toward targeted metabolic reprogramming, offering a promising strategy for ferroptosis-related renal diseases.

While this work provides impressive evidence that MSC-EVs reprogram the CH25H/25HC metabolic axis via miR-26b-5p to alleviate tubular ferroptosis in ischemic AKI, several limitations warrant further investigation. Although miR-26b-5p was identified as a key functional mediator, additional EV cargo-including other miRNAs, proteins, and lipids-may also contribute to redox regulation and lipid metabolism. Future multi-omics analyses are needed to define the relative contributions of these components. The downstream mechanisms linking CH25H/25HC perturbation to ferroptotic lipid damage also require further clarification. Moreover, evaluation of MSC-EVs in other clinically relevant AKI models will be important to determine the broader applicability of this mechanism. Finally, translational challenges such as scalable EV production, purification consistency, and renal targeting must be addressed to facilitate clinical translation.

In summary, our study identifies the CH25H/25HC axis as a key metabolic driver of tubular ferroptosis in ischemic AKI and demonstrates that MSC-EVs confer renoprotection by reprogramming cholesterol metabolism. We further show that EV-enriched miR-26b-5p directly targets CH25H, limiting 25HC accumulation, restoring redox balance, and alleviating ferroptotic stress in tubular cells. These findings broaden the mechanistic understanding of MSC-EV therapy and highlight metabolic modulation as a novel strategy for precision treatment of ischemic and ferroptosis-related kidney diseases.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Q. Y. and Y.-L.Z. conceptualized and designed the experiments, interpreted experimental results, and wrote the original draft, with contributions from all authors; Q.Y. and Y.-L.Z. conducted most of the experiments, with help from S.-J.C., J.S., Q.Y., Y.W. and H.-B.Y., Q.-L.W., Y.-Z., and Y.-Q.F. carried out experiments and analyzed the data; B.W., M.Y., S.-F.C., L.-L.L revised the manuscript. T.-T.T. and B.-C.L. contributed to the conceptualization, project administration, funding acquisition, and revision of the manuscript.

Funding

This work was supported by grants from the National Natural Science Foundation of China (82230022, 82030024, 82570873, 82200772), the National Key Research and Development Program of China (2022YFC2502500), the National Outstanding Physician (CZXM-RC-52), and the Jiangsu Province Innovation Centre of Nephrology (2023GSPKY05).

Data availability

The RNA sequencing dataset(s) supporting the conclusions of this article are available in the NCBI Sequence Read Archive (SRA) repository under BioProject accession number PRJNA1393078 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1393078). The data will be made publicly available upon publication.

Declarations

Ethics approval and consent to participate

All procedures involving human umbilical cords were approved by the Ethics Committee for Clinical Research of Zhongda Hospital, Southeast University (approval number: 2023ZDSYLL118-P01). All animal experiments were approved by the Institutional Animal Care and Use Committee of Southeast University (approval number: 20220224039) and were conducted in accordance with internationally accepted guidelines for the care and use of laboratory animals.

Competing interests

The authors declare no competing interests.