Universal catalytic endothelial lineage nanovesicles for targeting and rescuing ischemic endothelium

Hongyi Jiang; Yumeng Wang; Haosheng Huang; Liang Zhu; Hanting Shen; Liting Jiang; Chihao Lin; Huangyi Fang; Gang Zheng; Junfeng Shi; Xinghe Xue; Jilong Wang; Jian Lin; Junjie Deng; Liang Chen; Xiaoyun Pan

doi:10.1016/j.xcrm.2026.102972

. 2026 Aug 5;7(8):102972. doi: 10.1016/j.xcrm.2026.102972

Universal catalytic endothelial lineage nanovesicles for targeting and rescuing ischemic endothelium

Hongyi Jiang ^1,^2,^3,^4,⁸, Yumeng Wang ^1,^2,^3,^5,⁸, Haosheng Huang ^1,^2,³, Liang Zhu ^1,^2,³, Hanting Shen ^1,^2,³, Liting Jiang ^1,^2,³, Chihao Lin ^1,^2,³, Huangyi Fang ⁶, Gang Zheng ^1,^2,³, Junfeng Shi ⁷, Xinghe Xue ^1,^2,³, Jilong Wang ⁴, Jian Lin ^1,^2,^3,^∗, Junjie Deng ^4,^∗∗, Liang Chen ^1,^2,^3,^∗∗∗, Xiaoyun Pan ^1,^2,^3,^9,^∗∗∗∗

PMCID: PMC13522786 PMID: 42556339

Summary

Vascular endothelial dysfunction drives major ischemic diseases, yet current nanovesicle therapies lack precise targeting and fail to counteract severe oxidative stress. Here, we engineer a universal, dual-targeting therapeutic platform by functionalizing nanovesicles from induced pluripotent stem cell (iPSC)-derived endothelial cells with antioxidant ceria nanozymes (CeO₂-NMU_iECs). These nanovesicles home to ischemic tissues via the CXCR4/SDF-1 axis and preferentially enter local endothelial cells through homotypic affinity. At the target site, CeO₂-NMU_iECs act synergistically: the nanozymes catalytically neutralize reactive oxygen species, while the vesicles deliver endogenous pro-angiogenic proteins to rescue endothelial function. Systemic administration successfully promoted angiogenesis, mitigated tissue damage, and restored organ function across diverse rodent models of ischemia, including osteonecrosis, skin flap injury, myocardial infarction, and stroke. This nanozyme-arming strategy provides a broadly applicable platform, advancing cell-derived nanovesicles from passive carriers into active microenvironment-modulating therapeutics.

Keywords: ischemic diseases, endothelial dysfunction, biomimetic nanovesicles, ceria nanozymes, dual-targeting

Graphical abstract

Highlights

•
iPSC-derived endothelial lineage nanovesicles carry pro-angiogenic cargo
•
Nanozyme arming enables catalytic ROS scavenging in ischemic endothelium
•
CXCR4/SDF-1 homing and endothelial affinity drive dual targeting
•
CeO₂-NMU_iECs provide a universal strategy to treat diverse ischemic diseases

Jiang et al. present a dual-targeting nanovesicle platform armed with antioxidant nanozymes to treat ischemic diseases. By homing directly to injured endothelium, these engineered nanovesicles neutralize harmful reactive oxygen species and promote blood vessel regeneration, effectively repairing tissue damage across diverse rodent and large animal models.

Introduction

Ischemic diseases, including myocardial infarction, ischemic stroke, and peripheral vascular disorders, share a convergent pathological origin in vascular endothelial dysfunction, which remains a leading cause of morbidity and mortality worldwide.¹ The vascular endothelium, a dynamic and organ-spanning interface, regulates barrier integrity, blood flow, inflammation, and tissue repair.²^,³ When this system collapses under ischemic stress, barrier disruption, inflammatory cascades, and failed vascular regeneration rapidly ensue, driving irreversible organ injury.⁴^,⁵ This shared mechanism implies that therapeutics capable of directly repairing endothelial function could serve as a unifying strategy for diverse ischemic syndromes, yet no clinically viable platform has achieved this breadth of application.⁶

Cell-free nanovesicle therapies, including natural exosomes and biomimetic nanovesicles, offer the bioactivity of donor cells without the safety liabilities of live-cell transplantation.⁷^,⁸ However, their efficacy is constrained by the choice of cellular origin, which dictates molecular cargo and targeting behavior.⁹^,¹⁰ While numerous nanovesicle strategies have been deployed for ischemic diseases, most rely on conventional sources like mesenchymal stem cells (MSCs).¹¹^,¹²^,¹³ Although MSC-derived nanovesicles show general regenerative benefits, they lack sufficient specificity for direct vascular endothelial repair, encounter inherent scalability bottlenecks, and exhibit limited intrinsic pro-angiogenic efficacy.¹⁴ Induced pluripotent stem cells (iPSCs) provide an unlimited, highly scalable cellular source, yet their direct application in deriving endothelial-specific nanovesicles for the treatment of ischemic diseases remains unexplored. We hypothesize that matching the nanovesicle origin to the target tissue by specifically utilizing the endothelial lineage can overcome current therapeutic bottlenecks. By systematically comparing nanovesicle-mimicking units (NMUs) from isogenic iPSCs (NMU_iPSCs), endothelial cells (NMU_iECs), MSCs (NMU_iMSCs), and neural progenitor cells (NMU_iNPCs), we identified NMU_iECs as intrinsically enriched in angiogenesis-related proteins and uniquely capable of restoring endothelial networks. Specifically, proteomic analysis revealed that NMU_iECs inherit an abundance of critical pro-angiogenic receptors and regulatory proteins from their parent endothelial cells, including vascular endothelial growth factor receptor 1 (VEGFR1), fibroblast growth factor receptor 1 (FGFR1), neuropilin-2 (NRP2), and Ras-interacting protein 1 (RASIP1). This lineage-specific molecular cargo endows NMU_iECs with a robust capacity to stimulate endothelial cell proliferation, migration, and tube formation, directly orchestrating the regeneration of damaged vascular networks.

Ischemia triggers an excessive generation of reactive oxygen species (ROS), creating an oxidative microenvironment that directly destroys the surviving endothelium and exacerbates tissue damage.¹⁵ Consequently, even the lineage-optimized NMU_iECs falter under these oxidative conditions, as the ROS surge incapacitates their therapeutic cargo.¹⁶^,¹⁷^,¹⁸ Overcoming this microenvironmental barrier requires a dual-action construct capable of simultaneously blunting oxidative injury and delivering pro-regenerative cues. While traditional small-molecule antioxidants and natural enzymes often show limited clinical success due to poor stability and short in vivo half-lives,¹⁹ nanozymes have emerged as an alternative. Despite their superior stability and tunable catalytic activities, nanozymes lack the inherent biological signals required for comprehensive tissue regeneration, necessitating their integration with biological nanovesicles to bridge this gap. Compared with other nanozymes (e.g., MnO₂ or Pt) that often exhibit limited enzyme-mimetic range or large physical dimensions, cerium oxide (CeO₂) possesses unique advantages for this integration. Its ultra-small size allows it to be anchored onto nanovesicles without disrupting the membrane structure. Furthermore, its mixed surface valence (Ce³⁺ and Ce⁴⁺) enables the continuous, self-regenerating neutralization of broad-spectrum ROS.²⁰^,²¹ Crucially, the catalytic CeO₂ shield neutralizes oxidative stress to protect nanovesicle integrity, ensuring the sustained delivery of NMU_iECs-derived pro-angiogenic proteins to rescue the ischemic endothelium.

Here, we present a lineage-optimized, nanozyme-armed nanovesicle system built from human iPSC-derived endothelial cells (iPSC-ECs). By anchoring antioxidant ceria nanozymes (CeO₂) onto NMU_iECs surfaces, we generated CeO₂-NMU_iECs capable of catalytic ROS scavenging and sustained pro-angiogenic protein delivery. This hybrid design exploits an intrinsic dual-targeting mechanism, in which CXCR4 drives homing to ischemic regions via the SDF-1 gradient, followed by homotypic endothelial uptake, enabling precise localization without synthetic ligand modification. We validated this platform across in vitro models of oxidative (hydrogen peroxide, H₂O₂), pharmacological (methylprednisolone [MPS]), and metabolic (oxygen-glucose deprivation [OGD]) endothelial injury, and across primary endothelial cells from multiple tissue origins. Crucially, to validate the broad applicability of our platform, we demonstrated its potent therapeutic efficacy in vivo across four distinct models of organ ischemia affecting the bone (glucocorticoid-induced osteonecrosis of the femoral head [GIONFH]), skin (a skin flap model), heart (myocardial infarction [MI]), and brain (ischemic stroke). Collectively, our findings establish a therapeutic paradigm in which an engineered, lineage-optimized nanovesicle system overcomes key biological barriers within the disease microenvironment, providing a potent and universal platform for treating a wide spectrum of pathologies rooted in endothelial dysfunction.

Results

EC-derived nanovesicles exhibit superior pro-angiogenic potential

To identify an optimal cell source for nanovesicles aimed at vascular repair, we first differentiated human iPSCs into three distinct lineages: endothelial cells (iPSC-ECs),²²^,²³ MSCs (iPSC-MSCs),²⁴^,²⁵ and neural progenitor cells (iPSC-NPCs)²⁶ (Figure 1A). Differentiation success was confirmed by multi-modal characterization, establishing the high purity and lineage identity of iPSC-ECs (Figure S1), iPSC-MSCs (Figure S2), and iPSC-NPCs (Figure S3). Following a 7-day induction protocol (Figures S1A and S1B), the iPSC-ECs transitioned from pluripotent colonies to a characteristic cobblestone monolayer morphology (Figures S1C and S1D) and demonstrated angiogenic capacity in tube formation assays (Figures S1F). Flow cytometry confirmed a high purity of the resulting population, with over 94% of cells co-expressing the endothelial markers CD31 and CD144 (Figure S1E). Immunofluorescence staining further verified the expression of CD31, von Willebrand factor (vWF), and VE-cadherin at the protein level (Figure S1G). Transcriptomic analysis revealed a distinct gene expression profile for iPSC-ECs compared with undifferentiated iPSCs, as shown by principal-component analysis (PCA) (Figure S1H). We identified thousands of differentially expressed genes (Figures S1I–S1K), with Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes pathway analyses confirming the enrichment of pathways related to vascular development and angiogenesis (Figures S1J–S1L). Notably, key endothelial markers like PECAM1, vWF, CXCR4, and CDH5 were significantly upregulated, while pluripotency markers like POU5F1, NANOG, and SOX2 were downregulated (Figure S1M). Similarly, the successful generation of iPSC-MSCs was confirmed. These cells displayed a classic spindle-shaped morphology (Figure S2D) and possessed robust trilineage differentiation potential, as evidenced by their capacity to form mineralized nodules (alizarin red S), lipid droplets (oil red O), and cartilaginous pellets (Alcian blue) (Figure S2E). The iPSC-MSC population was defined by high surface expression of the canonical mesenchymal markers CD73, CD90, and CD105 (Figure S2F) and a unique transcriptome distinct from that of iPSCs (Figures S2G–S2L). Finally, the differentiation into iPSC-NPCs was also validated. These cells formed typical neural rosette-like structures (Figures S3C) and uniformly expressed the canonical neural progenitor markers Nestin, SOX2, and PAX6 (Figure S3F). Again, transcriptomic profiling confirmed a clear and significant divergence from the parent iPSC state (Figures S3E, S3G–S3K).

Using a previously established serial extrusion method,²⁷^,²⁸ we then generated NMUs from these three differentiated cell types and the undifferentiated iPSCs, yielding NMU_iECs, NMU_iMSCs, NMU_iNPCs, and NMU_iPSCs, respectively (Figure 1A). After serial extrusion and debris removal, the NMU-containing supernatant was purified and concentrated using a 100-kDa centrifugal filter, and the retentate was collected as the final NMU fraction for subsequent experiments (Figure S4A). Dynamic light scattering (DLS) analysis showed that the retentate contained a stable vesicle-sized population centered at approximately 100 nm, whereas the filtrate exhibited only weak and irregular scattering signals without a comparable vesicle-sized peak (Figure S4B). Cryo-transmission electron microscopy (TEM) further revealed that native NMU_iECs retained an intact vesicular morphology with a discernible bilayer-like membrane boundary (Figure S4C). Consistently, physicochemical characterization revealed that all four NMU types presented as cup-shaped vesicles with similar size distributions, peaking at approximately 100 nm in diameter, and demonstrated good colloidal stability in both PBS and serum-containing media over 48 h (Figures S4D and S4E). We next performed a comparison of their pro-angiogenic potential on human umbilical vein endothelial cells (HUVECs). NMU_iECs demonstrated a markedly superior capacity to promote both tube formation and cell migration compared with the other three NMU types (Figures 1E–1G). NMU_iMSCs also showed a significant, albeit lesser, pro-angiogenic effect, consistent with previous reports on the bioactivity of MSC-derived vesicles.²⁸^,²⁹ To uncover the molecular basis for the enhanced efficacy of NMU_iECs, we conducted a comparative proteomic analysis against NMU_iPSCs (Figure 1R). This revealed that NMU_iECs were intrinsically enriched with a distinct protein signature. A heatmap comparing the expression of key angiogenesis-related proteins illustrated this distinction, showing a clear and consistent upregulation of a broad panel of factors in NMU_iECs versus NMU_iPSCs (Figure S4F). Volcano plot and heatmap analyses highlighted a significant upregulation of proteins associated with pro-angiogenic signaling pathways in NMU_iECs (Figures 1S and 1T). Accordingly, GO enrichment analysis confirmed that pathways such as angiogenesis, cell adhesion, and collagen binding were among the most significantly enriched in the NMU_iEC proteome (Figure 1U). Western blot analysis further validated the higher abundance of key pro-angiogenic proteins, including FLT1, NRP2, SRPX2, and RASIP1,³⁰^,³¹^,³² in NMU_iECs compared with the other NMU types (Figures S4H and S4I; uncropped gels in Figure S24).

Having identified NMU_iECs as the most potent pro-angiogenic nanovesicles, we sought to evaluate their efficacy under conditions mimicking the ischemic microenvironment. We subjected HUVECs to oxidative stress induced by hydrogen peroxide (H₂O₂) and found that the therapeutic effects of NMU_iECs were severely blunted. While they retained some residual activity, their ability to promote tube formation and migration was substantially diminished compared with their performance under normal condition (Figures 1H, 1I, and 1K). To investigate the underlying cause of this functional attenuation, we assessed their impact on intracellular ROS. Confocal microscopy and flow cytometry analysis using the 2′,7′-dichlorofluorescin diacetate (DCFH-DA) probe revealed that NMU_iECs failed to reduce the elevated ROS levels in H₂O₂-treated HUVECs (Figures 1J–1L). Furthermore, by exposing HUVECs to a broader range of ischemic and oxidative insults (H₂O₂, MPS, and OGD), we observed through JC-1 and MitoSOX staining that NMU_iECs were unable to rescue the associated mitochondrial dysfunction, including the loss of membrane potential and the surge in mitochondrial ROS (Figures 1M–1P). Taken together, these findings demonstrate that while NMU_iECs, selected through systematic comparison, are a potent source of pro-angiogenic signals under normal physiological conditions, their efficacy is hampered in oxidative environments due to a lack of intrinsic antioxidant capacity. This highlights a fundamental barrier to their direct application for treating ischemic diseases (Figure 1Q).

Engineering of nanozyme-armed nanovesicles for synergistic therapy

Given the failure of unmodified NMU_iECs in oxidative environments (Figure 1), we engineered a nanozyme-armed nanovesicle platform, termed CeO₂-NMU_iECs, designed to synergistically neutralize ROS while promoting angiogenesis. This was achieved by covalently conjugating maleimide-functionalized ultrasmall, antioxidant ceria nanozymes (Mal-CeO₂) onto the surface of thiol-functionalized NMU_iECs (Thiol-NMU_iECs) via a maleimide-thiol click chemistry reaction²¹ (Figure 2P). First, we comprehensively characterized the synthesized CeO₂ nanozymes.³³ The CeO₂ nanoparticles exhibited a uniform spherical morphology with an average diameter of approximately 3–5 nm (Figure 2A). Energy-dispersive X-ray spectroscopy confirmed the elemental composition of cerium and oxygen, and a selected area electron diffraction pattern indicated their polycrystalline nature (Figure S5A). The crystalline structure was further verified by X-ray diffraction, which displayed characteristic diffraction peaks corresponding to the fluorite cubic structure of cerianite (Figure S5B). High-resolution analysis of the Ce 3d spectrum was crucial, revealing a deconvolution of peaks corresponding to mixed Ce (III) and Ce (IV) valence states, which is the basis for their catalytic ROS-scavenging activity³⁴ (Figure S5D). Functionally, the CeO₂ nanoparticles demonstrated potent, dose-dependent (0–200 μg/mL) scavenging capabilities against a wide spectrum of radicals, including hydroxyl radicals (⋅OH), superoxide anions (O₂⋅^-), ABTS, and DPPH, alongside catalase (CAT)- and superoxide dismutase (SOD)-mimetic activities (Figures S5E–S5K). For subsequent conjugation, these nanoparticles were functionalized with maleimide groups to yield Mal-CeO₂, whose morphology was confirmed by TEM (Figure S5L).

Engineering of a nanozyme-armed nanovesicle for synergistic therapy

(A) Representative transmission electron microscopy (TEM) images of native NMU_iECs, ultrasmall CeO₂ nanoparticles, and the assembled CeO₂-NMU_iEC nanohybrids.

(B) Super-resolution confocal microscopy images of CeO₂-NMU_iECs constructed from DiD-labeled NMU_iECs (red) and FITC-labeled CeO₂ (green). The ortho-display image (right) shows a three-dimensional co-localization view.

(C) Hydrodynamic size distribution of NMU_iECs (left) and CeO₂-NMU_iECs (middle) measured by dynamic light scattering (DLS). The bar chart (right) shows the zeta potential measurements for NMU_iECs, CeO₂, and CeO₂-NMU_iECs (n = 3, mean ± SD).

(D) Fluorescence intensity profiles along the selected line in the merged image of (B), showing the signal overlap for the DiD and FITC channels.

(E) High-resolution TEM (HRTEM) images showing the crystalline lattice fringes of CeO₂ nanoparticles (left) and CeO₂ nanoparticles on the surface of CeO₂-NMU_iEC (right).

(F–I) Assays for radical scavenging. (F) Scavenging of ABTS radicals, showing representative UV-visible absorbance curves (right) and corresponding scavenging rates (left) (n = 3, mean ± SD). (G) Scavenging of DPPH radicals (n = 3, mean ± SD). (H) Scavenging of hydroxyl radicals (⋅OH) (n = 3, mean ± SD). (I) Scavenging of superoxide anions (O₂⋅⁻) (n = 3, mean ± SD).

(J and K) Assays for enzyme-mimetic activities. (J) Superoxide dismutase (SOD)-mimetic activity (n = 3, mean ± SD). (K) Catalase (CAT)-mimetic activity (n = 3, mean ± SD).

(L) Measurement of H₂O₂ scavenging by CeO₂, NMU_iECs, and CeO₂-NMU_iECs (n = 3, mean ± SD).

(M) Colloidal stability of CeO₂-NMU_iECs in PBS and 50% FBS over 48 h, presented as diameter and polydispersity index (PDI) (n = 3, mean ± SD).

(N) SDS-PAGE analysis of total protein from iPSC-ECs, NMU_iECs, thiol-NMU_iECs, and CeO₂-NMUi_ECs.

(O) Western blot analysis of pro-angiogenic proteins (FLT1, FGFR1, FGFR2, NRP2, AMOTL1, RASIP1, SRPX2, and CXCR4) in iPSC-ECs, NMU_iECs, thiol-NMU_iECs, and CeO₂-NMU_iECs. Uncropped gels are provided in Figure S25.

(P) Schematic illustrating the synthesis and designed dual function of CeO₂-NMU_iECs. (Top) Covalent conjugation of maleimide-functionalized CeO₂ (Mal-CeO₂) to thiol-functionalized NMU_iECs (Thiol-NMU_iECs) via maleimide-thiol chemistry. (Bottom) Representation of the CeO₂-NMU_iECs concurrently scavenging ROS and promoting angiogenesis.

n represents the number of biologically independent samples. Statistical significance was determined by a two-tailed Student’s t test (for two groups) or one-way ANOVA (for multiple groups). p values are indicated in graphs. Scale bars: 30 nm for CeO₂ in (A), 100 nm for NMU_iECs and CeO2-NMU_iECs in (A), and 500 nm in (B).

We then confirmed the successful fabrication of the CeO₂-NMU_iEC nanohybrids. TEM images revealed that CeO₂-NMU_iEC retained the vesicular structure, with their surfaces uniformly decorated by the ultrasmall CeO₂ nanoparticles (Figure 2A). The covalent linkage of CeO₂ onto the NMU_iEC surface was definitively validated by super-resolution confocal microscopy of fluorescently labeled components (fluorescein isothiocyanate [FITC]-CeO₂ and DiD-NMU_iECs). Ortho-display images provided a three-dimensional view, and co-localization analysis confirmed a high degree of spatial overlap between the green (CeO₂) and red (NMU_iEC) signals, indicating a stable linkage (Figures 2B–2D). Consistent with successful surface conjugation, DLS analysis showed an increase in the average hydrodynamic diameter from ∼100 nm for NMU_iECs to ∼120 nm for CeO₂-NMU_iECs. Furthermore, zeta potential measurements demonstrated a significant shift from approximately −25 mV for the parent NMU_iECs to a less negative value of approximately −15 mV for CeO₂-NMU_iECs, reflecting the change in surface chemistry (Figure 2C). Finally, high-resolution TEM (HRTEM) imaging of CeO₂-NMU_iEC verified that the nanoparticles on the vesicle surface retained the characteristic crystalline lattice fringes of CeO₂, confirming their identity post-conjugation (Figure 2E).

Functionally, the assembled CeO₂-NMU_iECs inherited the potent antioxidant properties of the CeO₂ nanozymes. They demonstrated broad-spectrum scavenging of ROS, including ABTS, DPPH, ⋅OH, and O₂⋅⁻ (Figures 2F–2I), and exhibited superoxide dismutase (SOD)- and catalase (CAT)-mimetic activities (Figures 2J and 2K). Furthermore, the final CeO₂-NMU_iEC maintained excellent colloidal stability in both PBS and 50% FBS for 48 h, similar to that of the unmodified NMU_iECs (Figure 2M; Figure S5N). We assessed whether the chemical modification process compromised the intrinsic biological cargo of the NMU_iECs. SDS-PAGE analysis showed that the total protein profiles were well-preserved throughout the thiolation and conjugation steps (Figure 2N). More importantly, western blot analysis confirmed that the abundance of key pro-angiogenic proteins, such as FLT1, NRP2, SRPX2, and CXCR4, was not diminished in CeO₂-NMU_iEC compared with NMU_iECs (Figure 2O).

Finally, we characterized the production yield and cellular uptake of the platform. We estimated that approximately 22.6 ± 2.4 μg of protein and (11.0 ± 1.4) × 10⁹ particles could be produced from 10⁶ iPSC-ECs (Figure S5M). The conjugation efficiency of CeO₂ to NMU_iECs was optimized; the reaction reached a plateau when 100 μg/mL of CeO₂ was reacted with 50 μg/mL of NMU_iECs (Figure S5S). Cellular uptake assays in HUVECs showed that surface modification with CeO₂ did not impede the internalization of the nanovesicles, with both NMU_iECs and CeO₂-NMU_iECs showing comparable, dose-dependent uptake that appeared to saturate around 50 μg/mL after 6 h (Figures S5O and S5P). To further determine the intracellular fate of the internalized NMU_iECs, we performed a co-localization analysis with lysosomes. After 6 h of incubation, both NMU_iECs and CeO₂-NMU_iECs exhibited only partial co-localization with lysosomes, with a substantial fraction of the fluorescent signal being widely dispersed throughout the cytoplasm (Figures S5Q–S5R), indicating that the nanovesicles successfully evade lysosomal degradation. Cytotoxicity assays confirmed that NMU_iECs exhibited negligible toxicity toward HUVECs across a range of concentrations (0–50 μg/mL) at 24, 48, and 72 h (Figure S5T). Furthermore, at a fixed concentration (50 μg/mL for NMU_iEC components), the final CeO₂-NMU_iECs product, as well as its individual components (CeO₂ and NMU_iECs) and a non-covalent physical mixture (Mix), all demonstrated good biosafety (Figure S5U).

CeO₂-NMU_iECs possess innate dual-targeting capacity to ischemic endothelium

Previous studies have established that ischemic tissues upregulate the chemokine stromal cell-derived factor-1 (SDF-1),³⁵^,³⁶^,³⁷^,³⁸ creating a chemical gradient that guides the homing of cells and nanoparticles expressing its cognate receptor, CXCR4. While strategies to enhance CXCR4 expression on therapeutic vesicles, such as viral transduction or genetic modification, have been explored,³⁵^,³⁶^,³⁷^,³⁸ the development of a delivery system with intrinsically high levels of surface CXCR4 is highly desirable. We hypothesized that NMU_iECs, derived from endothelial cells, which naturally express high levels of CXCR4, would inherit this critical homing receptor. To test this, we first compared the surface abundance of CXCR4 across the four NMU types. Immunogold labeling followed by TEM analysis revealed a significantly higher density of 6-nm gold particles on the surface of NMU_iECs compared with NMU_iPSCs, NMU_iMSCs, and NMU_iNPCs (Figure 3A). However, as the 6-nm gold particles are similar in size to the ultrasmall CeO₂ nanozymes used in this study, we developed an independent, non-particulate-based method to avoid potential interference in subsequent analyses. This magnetic bead-based capture assay utilized streptavidin-coated magnetic beads conjugated with biotinylated anti-CXCR4 antibodies to specifically isolate NMUs based on their surface CXCR4 expression (Figure 3C). When incubated with FITC-CD63-labeled NMUs, both confocal microscopy and flow cytometry analysis of the captured beads demonstrated that the quantity of captured NMU_iECs was markedly greater than that of the other three NMU types, reaffirming their superior surface CXCR4 expression (Figure 3B).

To validate the prerequisite for CXCR4/SDF-1-mediated homing in vivo, we first confirmed that SDF-1 was upregulated across a spectrum of ischemic tissues. For this, we established four distinct models to represent ischemia in different organ systems: GIONFH for bone, an ischemic flap model for skin, MI for the heart, and middle cerebral artery occlusion (MCAO) for the brain. In all four models, quantitative reverse-transcription PCR (RT-PCR) and ELISA analyses showed that both SDF-1 mRNA and protein levels were significantly elevated in the ischemic regions compared with their corresponding normal tissues, thus providing a chemotactic gradient for NMU_iECs (Figures S6A–S6D). The ischemic flap model, which involves a local administration route, was reserved for in vivo distribution evaluation in a later section. Prior to in vivo distribution studies, we established the pharmacokinetic profile of NMU_iECs. After a single intravenous injection via the tail vein (200 μg DiD-labeled NMU_iECs in 200 μL PBS), ex vivo imaging of major organs was performed at multiple time points (1, 2, 4, 6, 8, 12, 16, 24, 36, 48, and 72 h). The analysis revealed accumulation in the liver and spleen peaking at approximately 4 h, followed by peak accumulation in the bone and lungs at 6 h, and in the kidneys at 8 h, with a circulating half-life of roughly 24 h (Figures S6E and S6F). Based on this profile, we selected the 4-h time point for cross-model comparisons. Because major organ accumulation peaks at this time, 4 h provides a practical window of high systemic availability to capture SDF-1-guided homing to ischemic tissues before substantial systemic clearance. To verify if this time point is also optimal in pathological settings, we quantified the local time course of NMU_iECs within ischemic target tissues. NMU_iECs accumulation in ischemic bone, infarcted heart, and the ischemic brain increased rapidly, peaked at 4 h post-injection, and then gradually declined over the next 24 h (Figure S6G). We, therefore, compared the biodistribution of NMU_iECs in normal rats and in rats with GIONFH, MI, or MCAO at 4 h post-injection. Ex vivo imaging revealed a pronounced accumulation of NMU_iECs in the respective ischemic tissues (Figures 3D–3G). Specifically, the fluorescence intensity in the bone of GIONFH rats was significantly higher than that in normal rats. Similarly, a markedly elevated signal was detected in the hearts of rats with MI and in the brains of rats with MCAO compared with their counterparts in the normal group (Figures 3D–3G). Interestingly, this accumulation was highly localized to the specific regions of injury: near the femoral head in the GIONFH model, in the myocardial infarct zone for MI, and within the ischemic hemisphere for MCAO. To further visualize this enrichment at the microscopic level, we analyzed cryosections of these tissues. This confirmed a substantial accumulation of NMU_iECs within the ischemic regions, with fluorescence signals being higher in the bone from the GIONFH group, the heart from the MI group, and the brain from the MCAO group compared with controls (Figures 3E and 3F).

To confirm that this ischemia-homing was mediated by CXCR4 and retained after CeO₂ conjugation, we performed CXCR4 blocking experiments. First, an in vitro magnetic bead assay showed that CeO₂-NMU_iECs were captured as efficiently as NMU_iECs and that pre-incubation with a CXCR4 antibody (CXCR4 block) reduced the capture of both vesicle types, confirming the successful blockade of the receptor (Figure S7A). Next, we compared the in vivo biodistribution of four formulations—NMU_iECs, CeO₂-NMU_iECs, and their respective CXCR4-blocked counterparts—across normal, GIONFH, MI, and MCAO rats. The results demonstrated that CeO₂-NMU_iECs exhibited a similar biodistribution profile identical to that of unmodified NMU_iECs and crucially, retained the ability to target ischemic tissues (Figures S7B–S7I). Pre-blocking of CXCR4, however, significantly altered this distribution. Specifically, in the GIONFH model, CXCR4 blockade substantially reduced the accumulation of both vesicle types in the ischemic bone. Similarly, the enhanced accumulation in the heart of MI rats and in the ischemic hemisphere of MCAO rats was significantly diminished upon CXCR4 blockade (Figures S7B–S7I). Interestingly, the blockade of CXCR4 not only prevented homing to ischemic sites but also reduced the basal accumulation in bone, which may be due to the enrichment of SDF-1 in bone tissue. In these CXCR4-blocked groups, a concomitant increase in signal was observed in the liver and spleen, indicating a biodistribution shift when CXCR4 was blocked (Figures S7B–S7I).

Next, we sought to identify which cell type preferentially internalizes NMU_iECs within the ischemic region. Building on the observation that vesicles often exhibit homotypic affinity,⁸^,³⁹^,⁴⁰ we hypothesized that endothelial-derived nanovesicles would be preferentially taken up by endothelial cells. To investigate this in vivo, we co-stained for DiD-labeled NMU_iECs (red) and the endothelial marker CD31 (green). Confocal microscopy of tissue sections from GIONFH bone, MI heart, and MCAO brain revealed a high degree of co-localization between the DiD and CD31 signals, indicating that after homing to the ischemic tissue, NMU_iECs were indeed preferentially taken up by local endothelial cells (Figure 3I). To quantitatively validate this endothelial preference, we performed in vitro assays with a mixed population of five cell types representative of the bone microenvironment (bone marrow endothelial cells [BMECs], bone marrow-derived MSCs (BMSCs), chondrocytes, nucleus pulposus [NP] cells, and macrophages). Confocal microscopy qualitatively demonstrated that after 6 h of incubation, the uptake of DiD-labeled NMU_iECs was substantially higher in BMECs compared with the other four cell types (Figure 3J, top). This observation was then confirmed by flow cytometry, which showed that BMECs internalized a higher percentage of NMU_iECs (92.7%) compared with macrophages (43.5%), BMSCs (30.3%), chondrocytes (27.2%), and NP cells (25.2%) (Figure 3J, bottom; Figures S8A and S8B). Time-course confocal microscopy and 3D fluorescence intensity plots further visualized this rapid and preferential internalization over 6 h (Figures S8C and S8D). Finally, to demonstrate that this endothelial affinity was a universal property, we repeated the uptake assay with a panel of five different endothelial cell types from diverse tissue origins (HUVECs, BMECs, human dermal microvascular endothelial cells [HDMECs], rat brain microvascular endothelial cells [RBMVECs], and mouse coronary artery endothelial cells [MCAECs]). Qualitative confocal imaging confirmed efficient internalization of NMU_iECs across all five endothelial cell lines (Figure 3K, top; Figures S9C and S9D). This was further quantified by flow cytometry, which confirmed a consistently high level of uptake, with proportions of 95.7% for HUVECs, 91.4% for BMECs, 86.6% for HDMECs, 82.8% for RBMVECs, and 67.5% for MCAECs, underscoring the universal endothelial affinity of NMU_iECs (Figure 3K, bottom; Figures S9A and S9B). Collectively, these results support a dual-targeting mechanism whereby NMU_iECs first home to ischemic tissues via the CXCR4/SDF-1 axis and are then preferentially internalized by local endothelial cells (Figure 3H).

CeO₂-NMU_iECs rescue endothelial dysfunction under diverse ischemic insults in vitro

Given the successful fabrication of CeO₂-NMU_iECs, we next sought to validate its therapeutic efficacy in rescuing endothelial dysfunction under various conditions that mimic key aspects of ischemic pathologies (Figure 4A). We employed three distinct in vitro injury models on HUVECs. First, H₂O₂ was used to induce direct and acute oxidative stress, a hallmark of the initial phase of ischemia injury.⁴¹ Second, we used MPS, a potent glucocorticoid, to replicate the specific cellular damage that underlies the pathogenesis of GIONFH.⁴² Third, an OGD model was established to simulate the metabolic and hypoxic stress characteristic of ischemic stroke.⁴³ By testing our platform across these diverse insults, we aimed to demonstrate its broad applicability. We began by assessing the effects in the H₂O₂-induced injury model. Treatment with H₂O₂ (200 μM) markedly suppressed the mRNA expression of key angiogenic genes, including VEGF, ANG1, bFGF, and PDGF (Figure 4B). While treatment with CeO₂ alone had no restorative effect, and NMU_iECs only partially increased ANG1 and bFGF levels, both the physical mixture of CeO₂ and NMU_iECs (Mix) and the conjugated CeO₂-NMU_iECs significantly upregulated all four genes. Notably, the CeO₂-NMU_iEC group showed a significantly greater increase in VEGF and ANG1 expression compared with the Mix group, suggesting a synergistic benefit of covalent conjugation (Figure 4B). This functional superiority was confirmed in tube formation and Transwell migration assays. H₂O₂ treatment severely impaired the ability of HUVECs to form capillary-like networks and to migrate. While CeO₂ or NMU_iECs alone provided only marginal protection, both the Mix and CeO₂-NMU_iECs substantially restored these functions, with the covalently conjugated CeO₂-NMU_iECs consistently demonstrating the most potent pro-angiogenic and pro-migratory effects (Figures 4C–4E). This enhanced efficacy of the CeO₂-NMU_iECs over the physical mixture may be attributed to the stable linkage ensuring the co-delivery and synergistic action of the antioxidant and pro-angiogenic components at the cellular level.

Next, we investigated the underlying cytoprotective mechanisms. Consistent with our findings in Figure 1, NMU_iECs alone failed to mitigate the surge in intracellular ROS induced by H₂O₂. In contrast, both CeO₂ and CeO₂-NMU_iECs effectively scavenged ROS, as shown by DCFH-DA staining. Again, the conjugated CeO₂-NMU_iECs were significantly more effective than both CeO₂ alone and the Mix, suggesting that the nanozyme’s antioxidant activity is enhanced when targeted to the cell via the NMU_iECs vehicle (Figures 4F–4H). This superior antioxidant protection extended to mitochondrial function. H₂O₂ treatment led to mitochondrial dysfunction, characterized by a collapse in mitochondrial membrane potential (JC-1 staining) and an elevation in mitochondrial ROS (MitoSOX staining). CeO₂-NMU_iECs were the most effective formulation at preserving membrane potential and suppressing mitochondrial ROS, outperforming both individual components and their physical mixture (Figures 4G–4I). Consequently, CeO₂-NMU_iECs provided the most potent protection against apoptosis, as evidenced by a marked reduction in TUNEL-positive cells and a lower percentage of Annexin V-positive cells compared with all other treatment groups (Figures 4J and 4K).

To explore the molecular basis of this protection, we analyzed key antioxidant-related proteins in HUVECs. H₂O₂ exposure reduced the expression of sirtuin 1 (SIRT1) and nuclear respiratory factor 1 (NRF1), whereas CeO₂-NMU_iECs increased the levels of SIRT1, NRF1, and heme oxygenase 1 (HMOX1) under oxidative stress, indicating enhanced cellular antioxidant capacity (Figure 4L). To gain a global view of the cellular response, we performed transcriptomic analysis on HUVECs treated with H₂O₂ alone or with H₂O₂ and CeO₂-NMU_iECs. PCA showed a clear separation between the two groups, indicating a profound transcriptomic shift induced by the treatment (Figure 4M). Compared with the H₂O₂ group, cells treated with CeO₂-NMU_iECs showed 2,511 significantly upregulated genes (including anti-oxidant and pro-angiogenic factors like SIRT2, HMOX1, VEGFA, and VWF) and 842 significantly downregulated genes (including pro-inflammatory factors like COX2 and CXCL6) (Figures 4N and 4O). GO analysis confirmed that the differentially expressed genes were highly enriched in pathways such as angiogenesis, positive regulation of cell proliferation, and cellular response to hydrogen peroxide (Figure 4P). Furthermore, GSEA revealed a significant positive enrichment for both cell migration involved in sprouting angiogenesis and cellular response to hydrogen peroxide in the CeO₂-NMU_iECs-treated group, providing comprehensive evidence for the dual pro-angiogenic and antioxidant action of our platform at the transcriptomic level (Figure 4Q).

To validate the broad utility of our platform, we repeated the functional assays in the MPS- and OGD-induced injury models. In both models, treatment with CeO₂-NMU_iECs consistently and significantly restored the impaired tube formation and migration capabilities of HUVECs (Figures S10A–S10C; Figures S11A–S11C). Similarly, CeO₂-NMU_iECs were highly effective at mitigating intracellular ROS production and rescuing mitochondrial dysfunction in both the MPS and OGD contexts (Figures S10D–S10G; Figures S11D–S11G). In conclusion, these in vitro results collectively demonstrate that the engineered CeO₂-NMU_iECs platform can effectively rescue endothelial cell dysfunction across a range of clinically relevant ischemic and oxidative insults. The covalent conjugation of the CeO₂ nanozyme to the NMU_iECs vehicle creates a synergistic effect that is superior to the individual components or their physical mixture, simultaneously restoring angiogenic potential and protecting against cellular damage.

CeO₂-NMU_iECs exert a conserved protective effect across endothelial cells of diverse tissue origins

The vascular endothelium, once considered a homogenous monolayer, is now recognized as a vast, heterogeneous organ system with remarkable structural and functional diversity across different tissues.¹ Endothelial cells in the brain form the tight blood-brain barrier,⁴⁴ those in the bone marrow regulate hematopoietic stem cell niches,⁴⁵ and those in the heart withstand unique hemodynamic forces.⁴⁶ This organ-specific specialization presents a challenge for developing therapies for systemic endothelial dysfunction, as a successful strategy should prove effective across this spectrum of cellular diversity.³^,⁴⁷ To validate that the therapeutic efficacy of our CeO₂-NMU_iECs platform transcends this heterogeneity and represents a universal strategy, we next tested its effects on a panel of primary endothelial cells isolated from four distinct tissue origins, corresponding to our in vivo ischemia models: bone (rat BMECs), skin (HDMECs), heart (MCAECs), and brain (RBMVECs) (Figure 5A).

We began with the validation in BMECs. Primary rat BMECs were successfully isolated from bone marrow, purified using CD31 microbeads, and subsequently characterized.⁸ These cells displayed a typical cobblestone morphology (Figure 5B) and were confirmed to be a pure endothelial population by flow cytometry, showing high expression of CD31 and CD144, while being negative for the hematopoietic marker CD45 (Figure 5C). We then assessed the therapeutic effects in the H₂O₂-induced injury model. As observed with HUVECs, H₂O₂ treatment severely impaired the angiogenic potential of BMECs, drastically reducing their ability to form capillary-like networks and to migrate (Figures 5D–5F). Treatment with CeO₂ or NMU_iECs alone offered only minimal rescue of these functions. While the physical mixture (Mix) substantially restored tube formation and migration, the covalently conjugated CeO₂-NMU_iECs consistently demonstrated the most potent restorative effects, significantly outperforming all other groups (Figures 5D–5F). This functional superiority was mirrored by superior cytoprotective capabilities. CeO₂-NMU_iECs were the most effective formulation at mitigating cellular damage; they most effectively scavenged intracellular ROS, outperforming both CeO₂ alone and the Mix (Figures 5G–5I). This enhanced antioxidant defense translated to better preservation of mitochondrial integrity, where CeO₂-NMU_iECs were again superior in maintaining mitochondrial membrane potential and suppressing mitochondrial ROS production (Figures 5H–5J). Consequently, CeO₂-NMU_iECs provided the most effective protection against apoptosis, as evidenced by a significantly lower percentage of TUNEL-positive and Annexin V-positive cells compared with all other treatment groups (Figures 5K and 5L). To investigate the molecular response in BMECs, we analyzed the expression of antioxidant-related proteins. Similar to our observations in HUVECs, treatment with CeO₂-NMU_iECs increased the levels of antioxidant-related proteins, including NAD(P)H quinone dehydrogenase 1 (NQO1), NRF1, and SIRT1, in oxidatively stressed BMECs (Figure 5M). A comprehensive transcriptomic analysis further solidified these findings. PCA showed a clear separation between the H₂O₂ and H₂O₂ + CeO₂-NMU_iECs treatment groups, indicating a profound transcriptomic reprogramming induced by the treatment (Figure 5N). Compared with the H₂O₂ group, cells treated with CeO₂-NMU_iECs showed 576 significantly upregulated genes (including potent pro-angiogenic and antioxidant factors such as HMOX1, NQO1, ADM, FGF7, and VEGFA) and 499 downregulated genes (including pro-inflammatory chemokines such as CCL5 and CXCL10) (Figure 5O). A heatmap of representative differentially expressed genes visualized this distinct transcriptomic shift (Figure 5Q). GO analysis of the upregulated genes revealed a significant enrichment of pathways related to both cellular regeneration (e.g., positive regulation of angiogenesis, positive regulation of cell population proliferation, and regulation of cell migration) and stress response (e.g., endothelial cell chemotaxis, and response to hydrogen peroxide), highlighting a dual pro-regenerative and cytoprotective program (Figure 5R). This was further corroborated by GSEA, which confirmed a strong positive enrichment for multiple key pathways, including positive regulation of angiogenesis, positive regulation of endothelial cell migration, response to ischemia, and cellular oxidant detoxification in the CeO₂-NMU_iECs-treated group. Collectively, these data demonstrate a conserved dual-action mechanism in primary BMECs (Figure 5P).

To further establish the broad applicability of our platform, we extended our validation to endothelial cells from other critical organs. We first confirmed the identity of the commercially sourced HDMECs, which showed typical endothelial morphology and expressed high levels of VE-cadherin and CD31 (Figures S12A–S12C). In these cells, CeO₂-NMU_iECs again proved to be the most effective formulation for restoring angiogenic functions and protecting against oxidative stress-induced ROS production and mitochondrial dysfunction (Figures S12D–S12J). Next, we validated the therapeutic effects on MCAECs, which were also confirmed to be of endothelial identity via immunofluorescence staining for CD31 and VE-cadherin (Figures S13A–S13C). In this cardiac endothelial model, while CeO₂ alone did not promote tube formation or migration and NMU_iECs showed only partial effects, both the Mix and CeO₂-NMU_iECs potently rescued these angiogenic functions, with the covalently CeO₂-NMU_iECs demonstrating the highest efficacy (Figures S13D–S13F). As with the other endothelial cell types, CeO₂-NMU_iECs also provided the most effective protection against intracellular ROS production and mitochondrial dysfunction in H₂O₂-treated MCAECs (Figures S13G–S13J). Finally, we validated these effects in RBMVECs isolated from rat cortical gray matter.⁴⁸ These cells were characterized by their morphology, CD31 expression, and a CD31⁺/CD144⁺/CD105⁺/CD45⁻ flow cytometry profile (Figures S14A–S14C). Once again, CeO₂-NMU_iECs demonstrated a superior ability to restore angiogenic potential and mitigate oxidative damage in these brain-derived endothelial cells (Figures S14D–S14J). Collectively, these results demonstrate that the protective and regenerative effects of CeO₂-NMU_iECs are not cell-type specific but are conserved across endothelial cells of diverse species and tissue origins. This consistent and enhanced performance over individual components or their physical mixture validates our nanozyme-arming nanovesicle-mimicking platform as a potent and broadly applicable platform for treating endothelial dysfunction.

Effect of CeO₂-NMU_iECs in bone and skin ischemia

Having established the potent and universal cytoprotective effects of CeO₂-NMU_iECs in vitro, we next sought to translate these findings into therapeutic outcomes in vivo. We selected two distinct and clinically relevant models of peripheral ischemia: GIONFH, a common form of ischemic bone disease, and a random-pattern skin flap model, which represents acute dermal ischemia.

We first investigated the therapeutic efficacy of CeO₂-NMU_iECs in a rat model of GIONFH. The disease was induced by intramuscular injection of MPS, and rats were treated with PBS, CeO₂, NMU_iECs, a physical mixture (Mix), or CeO₂-NMU_iECs via the tail vein injection (Figure 6A). Based on our pharmacokinetic studies, a multi-dose regimen was employed. Three-dimensional micro-computed tomography (micro-CT) reconstructions of the femoral heads at 6 weeks revealed severe trabecular bone destruction, characterized by architectural collapse and bone loss, in the PBS group (Figure 6B). Treatment with CeO₂ alone showed negligible improvement, likely due to its rapid renal clearance and lack of specific targeting to the bone. While NMU_iECs and the Mix group showed partial preservation of bone structure, the CeO₂-NMU_iECs-treated group exhibited the most substantial therapeutic effect, with markedly preserved trabecular architecture (Figure 6B). Quantitative analysis of bone morphometric parameters confirmed these observations: CeO₂-NMU_iECs significantly improved bone volume/total volume, trabecular number, and trabecular thickness, while reducing trabecular separation to levels approaching the control group, and were significantly more effective than both NMU_iECs alone and the Mix (Figures 6C–6F). These results highlight both the critical role of arming the NMU_iECs with the CeO₂ nanozyme to overcome the oxidative stress that limits the efficacy of the vesicles alone and the superiority of the covalent conjugation strategy over a physical mixture in ensuring synergistic action at the target site. Histological analyses provided further evidence of this therapeutic effect. H&E staining of femoral head sections from the PBS group revealed severe pathological changes, including extensive trabecular necrosis, a high number of empty lacunae, and massive adipocyte infiltration in the bone marrow space. Treatment with CeO₂ alone showed negligible improvement over the PBS group. While the NMU_iECs and Mix groups exhibited partial preservation of bone structure and a reduction in necrotic areas, the CeO₂-NMU_iECs group displayed the most profound therapeutic effect, with a well-preserved bone structure and a significant reduction in both necrosis and adiposity (Figures 6G and 6H). Masson staining and Goldner staining further confirmed that CeO₂-NMU_iECs treatment effectively preserved the collagenous bone matrix and promoted the formation of mature, mineralized bone (Figures S15A and S25B). To assess angiogenesis and osteogenesis, we performed immunohistochemistry. The CeO₂-NMU_iECs group showed a significant increase in the density of CD31-positive blood vessels and the expression of the osteogenic markers collagen I and osteocalcin, as well as the pro-angiogenic factor ANG1, compared with all other treatment groups (Figures 6I and 6J; Figures S15C and S15D). Finally, TUNEL staining revealed widespread apoptosis in the femoral heads of the PBS group (∼60% TUNEL-positive cells), which was not alleviated by CeO₂ alone. In contrast, CeO₂-NMU_iECs provided the most potent anti-apoptotic effect, significantly reducing the percentage of TUNEL-positive cells (Figures S15E and S15F).

Effect of CeO₂-NMU_iECs in bone and skin ischemia

(A–J) Treatment of glucocorticoid-induced osteonecrosis of the femoral head (GIONFH) in rats. (A) Schematic of the experimental design. GIONFH was induced in rats by intramuscular injection of methylprednisolone ([MPS], 20 mg/kg/day) for the first 3 days of each week over a 3-week period (indicated by yellow arrows); rats in the Control group received corresponding injections of normal saline. Six groups of rats were included: Control, GIONFH (treated with PBS), and GIONFH treated with CeO₂, NMU_iECs, Mix, or CeO₂-NMU_iECs. Therapeutic formulations were administered via the tail vein injection at a dose equivalent to 200 μg of the NMU_iECs component in 200 μL PBS. The injections were administered three times a week for the first 3 weeks, followed by weekly injections for the subsequent 3 weeks (indicated by red arrows). Femoral heads were harvested for analysis after 6 weeks for micro-CT, H&E staining, Masson staining, Goldner staining, immunohistochemistry (IHC), and TUNEL staining assays. (B) Representative three-dimensional micro-CT reconstruction images of the femoral heads from each group. (C–F) Quantitative micro-CT analysis of bone morphometric parameters: bone volume to total volume ratio (BV/TV) (C), trabecular number (Tb.N) (D), trabecular thickness (Tb.Th) (E), and trabecular separation (Tb.Sp) (F) (n = 8, mean ± SD). (G) Quantification of the necrotic area based on H&E staining (n = 8, mean ± SD). (H) Representative H&E-stained sections of the femoral heads. (I) Representative IHC images for the endothelial marker CD31 and the osteogenic marker collagen I (COL1). (J) Quantification of the CD31-positive vessel number and the COL1-positive area from IHC images (n = 8, mean ± SD).

(K–Q) Treatment of ischemia in a mouse random skin flap model. (K) Schematic of the experimental design. A random-pattern skin flap (1.5 × 4.5 cm) was elevated on the mouse dorsum, and the sacral arteries supporting the blood supply to the flap were resected to induce ischemia. The five treatment formulations (PBS, CeO₂, NMU_iECs, Mix, or CeO₂-NMU_iECs) were then subcutaneously injected via a microinjection needle at six sites along the longitudinal axis of the flap (100 μg/mL, 20 μL per site). Flaps were harvested on day 7 for analysis, including macroscopic photography, infrared thermal imaging, laser Doppler blood flow (LDBF) imaging, H&E staining, immunohistochemistry, DHE staining, western blot analysis, TUNEL staining assay, and immunofluorescence. (L) Representative images of the skin flaps on day 7 post-surgery, including macroscopic photographs, infrared thermal images, and LDBF images. (M) Quantification of the flap survival area, temperature, and blood flux from the images in (L) (n = 8, mean ± SD). (N) Representative images of H&E staining, CD31 IHC staining, and DHE staining (ROS, red; DAPI, blue) of flap tissues on day 7. (O) Quantification of the CD31-positive vessel density and the DHE fluorescence intensity (n = 8, mean ± SD). (P and Q) Western blot analysis of skin flap tissue lysates. (P) Representative blots for the endothelial markers CD31 and VE-cadherin. (Q) Densitometric quantification of the protein bands in (P) (n = 5, mean ± SD).

n represents the number of biologically independent samples. Statistical significance was determined by a two-tailed Student’s t test (for two groups) or one-way ANOVA (for multiple groups). p values are indicated in graphs. Scale bars: 200 μm in (H), 100 μm in (I), 100 μm in (N, CD31 IHC and H&E), and 80 μm in (N, DHE).

Next, we evaluated the therapeutic potential of CeO₂-NMU_iECs in a mouse ischemic skin flap model. A random-pattern flap was created on the dorsum of the mice, and treatments were administered locally, a route more suitable for this type of localized dermal injury.⁴⁹ Specifically, the formulations were delivered via six subcutaneous injections along the longitudinal axis of the flap (Figure 6K; Figure S16A). We first investigated whether this local administration route would still leverage the CXCR4-dependent homing mechanism. As previously confirmed, the ischemic flap tissue highly expresses SDF-1 (Figure S6B). After subcutaneous injection of DiD-labeled NMU_iECs, we observed significantly higher retention and broader distribution of the nanovesicles within the ischemic flap compared with normal skin, where the signal remained more localized to the injection sites (Figure S16G). Furthermore, this enhanced local retention in the ischemic flap was significantly reduced by pre-blocking with a CXCR4 antibody, confirming that even with local delivery, the CXCR4/SDF-1 axis plays a key role in directing the NMU_iECs toward the injured tissue (Figure S16H). We then optimized the therapeutic dose for local administration. A clear dose-dependent improvement in flap viability was observed with CeO₂-NMU_iECs. While the 25 and 50 μg/mL doses offered partial protection, treatment with 100 μg/mL resulted in the most substantial preservation of viable tissue, the most uniform temperature distribution across the flap surface, and the most widespread blood perfusion signal and was thus selected as the optimal dose for subsequent experiments (Figures S16B and S16C). Using this optimal dose, we compared the five treatment groups (PBS, CeO₂, NMU_iECs, Mix, and CeO2-NMU_iECs). On day 7 post-surgery, the PBS-treated group exhibited severe distal necrosis, characterized by a dark, desiccated appearance and a sharp demarcation line (Figures 6L and 6M). This was accompanied by a pronounced temperature drop in the ischemic region and a near-total absence of blood flow as measured by LDBF imaging (Figures 6L and 6M). Treatment with CeO₂, NMU_iECs, or the Mix all resulted in varying degrees of improved flap survival. In contrast, the CeO₂-NMU_iECs group exhibited the largest viable area, with healthy-appearing tissue extending to the distal tip, a restored and homogenous temperature profile, and strong, widespread blood perfusion signals across the entire flap (Figures 6L and 6M). Histological analysis confirmed these macroscopic findings. H&E staining of the CeO₂-NMU_iECs-treated flaps revealed well-preserved dermal structures with a greater number of visible vascular lumens, and significantly reduced edema and inflammatory cell infiltration compared with the other groups (Figure 6N). Crucially, immunohistochemistry for CD31, an endothelial cell marker, showed a significantly higher density of well-formed, lumen-containing microvessels in CeO₂-NMU_iECs group (Figures 6N and 6O). This was consistent with the widespread and intense immunofluorescence signal for the pro-angiogenic factor VEGFA observed in this group (Figures S16E and S16F). Furthermore, dihydroethidium (DHE) staining revealed extensive ROS production in the necrotic regions of PBS-treated flaps, which was most effectively mitigated by CeO₂-NMU_iECs (Figures 6N and 6O). Similarly, TUNEL staining showed widespread apoptosis in the PBS-treated flaps, whereas the CeO₂-NMU_iECs-treated flaps exhibited a dramatic reduction in TUNEL-positive cells (Figures S16D and S16F). To further assess endothelial cytoprotection in vivo, we performed CD31/TUNEL co-immunofluorescence staining of ischemic flap tissues on day 3. PBS-treated flaps showed abundant TUNEL signals around CD31-positive endothelial structures, whereas CeO₂-NMU_iECs treatment reduced endothelial-associated TUNEL staining, supporting early protection of ischemic endothelial cells from apoptosis (Figures S16I and S16J). Finally, western blot analysis of skin tissue lysates confirmed these pro-survival and pro-angiogenic effects at the molecular level, showing that CeO₂-NMU_iECs treatment led to the highest expression of the endothelial markers CD31 and VE-cadherin (Figures 6P and 6Q). Taken together, these results from two distinct models of peripheral ischemia (a chronic, systemic GIONFH model and an acute, localized skin flap model) provide strong evidence for the in vivo therapeutic potential of the CeO₂-NMU_iECs platform. Despite the differences in the pathology and administration route, the nanozyme-armed nanovesicles consistently promoted angiogenesis, combated oxidative stress, and restored ischemic tissue in both bone and skin.

Effect of CeO₂-NMU_iECs in myocardial and cerebral ischemia

The therapeutic success of our platform in peripheral tissues prompted us to test its efficacy in MI and ischemic stroke. These pathologies, the leading causes of death and disability, are fundamentally diseases of the vasculature, precipitated by acute endothelial dysfunction and thrombotic occlusion in the coronary or cerebral arteries, respectively.⁵⁰ The subsequent ischemia-reperfusion injury unleashes a torrent of oxidative stress and inflammation, with the endothelium at its epicenter, leading to widespread cardiomyocyte death or neuronal loss.⁴ We, therefore, hypothesized that our dual-targeting, synergistic CeO₂-NMU_iECs platform, designed to rescue the endothelium from this pathological cascade, could provide substantial therapeutic benefits in these life-threatening conditions.

We first established a rat model of MI by permanent ligation of the left anterior descending artery. Rats were randomized into six groups (sham, MI + PBS, MI + CeO₂, MI + NMU_iECs, MI + Mix, and MI + CeO₂-NMU_iECs) and received intravenous treatments based on our pharmacokinetic data (Figure 7A). At 28 days post-MI, echocardiography was performed to assess cardiac function. M-mode images revealed that the MI group treated with PBS exhibited severe left ventricular dilation and systolic dysfunction (Figure 7B; Figures S17A). Treatment with CeO₂ alone provided negligible improvement in cardiac function. While both the NMU_iECs and Mix groups provided partial improvement in these outcomes, the CeO₂-NMU_iECs group exhibited the most significant preservation of cardiac function (Figure 7B; Figure S17A). Quantification confirmed that CeO₂-NMU_iECs group had the highest ejection fraction and fractional shortening, the smallest left ventricular internal dimension at systole, and the best-preserved LV wall thickness compared with all other groups (Figure 7C; Figures S17E). Histological analysis corroborated these functional improvements. Masson staining at day 28 showed extensive transmural fibrosis and thinning of the ventricular wall in the MI group treated with PBS. The CeO₂ and NMU_iECs groups showed only a modest reduction in fibrosis, while the Mix group showed a more noticeable improvement. The CeO₂-NMU_iECs group, however, displayed the smallest fibrotic scar area and the most preserved myocardial tissue, indicating the most effective prevention of adverse cardiac remodeling (Figures 7D and 7E). H&E staining further confirmed the greater preservation of viable myocardium and reduced inflammatory infiltration in the CeO₂-NMU_iECs group (Figure S17B). To investigate the underlying mechanisms of cardiac repair, we performed further histological analyses. Immunofluorescence staining at day 28 revealed a significantly higher density of CD31-positive microvessels in the infarct border zone of the CeO₂-NMU_iECs group, indicating enhanced angiogenesis (Figures 7F and 7G). Co-staining for the endothelial marker vWF and the smooth muscle marker α-SMA suggested a greater number of mature, stabilized vessels in the CeO₂-NMU_iECs-treated hearts (Figure S17C). Furthermore, the expression of Connexin 43 (Cx43), a key gap junction protein essential for cardiomyocyte electrical coupling, was better preserved at the intercalated discs of cardiomyocytes in the CeO₂-NMU_iECs group, suggesting improved tissue viability and electrical stability (Figure 7G; Figure S17F). Analysis of acute injury at day 7 confirmed that the massive burst of ROS in the ischemic myocardium, measured by DHE staining, was most effectively suppressed by the CeO₂-NMU_iECs treatment (Figures 7F and 7G). Consistent with this, TUNEL staining at day 7 revealed that CeO₂-NMU_iECs provided the most potent protection against apoptosis (Figures S17D and S17G).

Effect of CeO₂-NMU_iECs in myocardial and cerebral ischemia

(A–G) Treatment of myocardial infarction (MI) in rats. (A) Schematic of the experimental design. MI was induced in male Sprague-Dawley rats by ligation of the left anterior descending artery (LAD); sham-operated rats underwent the same procedure without ligation. Thirty minutes after MI induction, the five treatment groups (PBS, CeO₂, NMU_iECs, Mix, or CeO₂-NMU_iECs) received the first intravenous injection via the tail vein at a dose equivalent to 200 μg of the NMU_iECs component. Treatments were administered three times a week for the first 2 weeks, followed by twice-weekly injections for the subsequent 2 weeks. Analyses were performed at day 7 (for acute *in vivo* ROS detection and TUNEL staining assay) and day 28 (for cardiac function and histology). (B) Representative M-mode echocardiography images of the left ventricle from each group at day 28. (C) Quantification of cardiac function parameters from echocardiography, including ejection fraction (EF), fractional shortening (FS), and left ventricular internal dimension at systole (LVIDs) (n = 8, mean ± SD). (D) Representative images of heart sections from the six groups (Sham, MI, MI + CeO₂, MI + NMU_iECs, MI + Mix, and MI + CeO₂-NMU_iECs) at day 28, stained with Masson’s trichrome to delineate the fibrotic scar area (blue) from viable myocardium (red). (E) Quantification of the infarct size as a percentage of the left ventricle area (n = 8, mean ± SD). (F) Quantification of CD31-positive vessel density (left) and DHE fluorescence intensity (right) (n = 8, mean ± SD). (G) Representative fluorescence microscopy images of the infarct border zone. (Top) immunofluorescence for the endothelial marker CD31 (green) at day 28. (Middle) DHE staining for superoxide (red) at day 7. (Bottom) Co-immunofluorescence for the gap junction protein connexin 43 (Cx43, green) and the cardiomyocyte marker α-actinin (red) at day 28. Nuclei were counterstained with DAPI (blue).

(H–M) Treatment of middle cerebral artery occlusion (MCAO) in rats. (H) Schematic of the experimental design. The MCAO model was established by the intraluminal filament method to transiently occlude the left MCA for 1.5 h, followed by reperfusion; sham-operated rats underwent the same surgical procedure without filament insertion. Thirty minutes after reperfusion, treatments were administered intravenously to six groups of rats: Sham, MCAO (treated with PBS), and MCAO treated with CeO₂, NMU_iECs, Mix, or CeO₂-NMU_iECs, followed by injections thrice weekly. Neurological function was assessed at days 0, 1, 3, 5, and 7. Brain tissues were harvested at day 1 (for TTC staining and cerebral edema evaluation) or day 7 (for Nissl staining, H&E staining, *in vivo* ROS detection, TUNEL staining assay, immunohistochemistry, and immunofluorescence). (I) Representative images of 2,3,5-triphenyltetrazolium chloride (TTC)-stained coronal brain sections at day 1 post-MCAO. (J) Quantification of the infarct volume, calculated from the TTC-stained sections in (I) (left). Brain water content (edema) was determined using the wet-dry weight method (right) (n = 8, mean ± SD). (K) Representative images of Nissl staining at day 7 for the six experimental groups (Sham, MCAO, and MCAO treated with CeO₂, NMU_iECs, Mix, or CeO₂-NMU_iECs), showing surviving neurons. (M) Representative images of CD31 immunohistochemistry (top), DHE staining (middle; red), and TUNEL staining (bottom; green) in the brain. Nuclei were counterstained with DAPI (blue). (L) Quantification of Nissl-positive neurons, CD31-positive vessel density, DHE fluorescence intensity, and the percentage of TUNEL-positive cells (n = 8, mean ± SD).

n represents the number of biologically independent samples. Statistical significance was determined by a two-tailed Student’s t test (for two groups) or one-way ANOVA (for multiple groups). p values are indicated in graphs. Scale bars: 500 μm in (D), 130 μm in (G, CD31 and Cx43/α-actinin), 65 μm in (G, DHE), 400 μm in (K), 100 μm in (M, CD31), 65 μm in (M, DHE), and 2,700 μm in (M, TUNEL).

Next, we evaluated the therapeutic efficacy of CeO₂-NMU_iECs in a rat model of transient MCAO, which mimics ischemic stroke. The MCAO model was established by occluding the left MCA for 1.5 h, followed by reperfusion. Treatments were administered intravenously 30 min after reperfusion and every other day thereafter (Figure 7H). At day 1 post-MCAO, 2,3,5-triphenyltetrazolium chloride (TTC) staining revealed a large, pale infarct area in the PBS-treated group. While NMU_iECs and the Mix offered partial protection, CeO₂-NMU_iECs most significantly reduced the infarct volume and attenuated the development of brain edema, as measured by the wet-dry weight method (Figures 7I and 7J). This potent neuroprotective effect translated into a marked improvement in functional recovery. Over a 7-day observation period, rats treated with CeO₂-NMU_iECs showed a rapid and sustained improvement in neurological function, as evidenced by significantly lower Longa scores, reduced turning bias in the elevated body swing test, and improved motor coordination in the rotarod and adhesive removal tests, compared with all other MCAO groups (Figure S18A). Histological analysis at day 7 provided further evidence of neuroprotection. H&E and Nissl staining of the ischemic penumbra showed that the CeO₂-NMU_iECs group had the most preserved tissue structure and the highest number of surviving, healthy-appearing neurons (Figure 7K; Figure S18B). This was accompanied by a significant increase in the density of CD31-positive microvessels, indicating enhanced post-stroke angiogenesis (Figures 7L and 7M). Furthermore, CeO₂-NMU_iECs treatment effectively mitigated key pathological hallmarks of stroke-induced brain injury. Immunostaining revealed a significant reduction in GFAP-positive reactive astrocytes, indicating an attenuation of detrimental neuroinflammation and glial scarring (Figures S18C and S18D). Consistent with the neuroprotective effect observed by Nissl staining, immunofluorescence for the neuronal marker NeuN confirmed a higher number of surviving neurons in the CeO₂-NMU_iECs-treated group (Figures S18E and S18F). DHE and TUNEL staining confirmed that these protective effects were associated with a significant reduction in ROS production and neuronal apoptosis at day 7 in the ischemic penumbra (Figures 7L and 7M).

Collectively, the therapeutic success of CeO₂-NMU_iECs across four distinct and challenging in vivo models (GIONFH, ischemic flap, MI, and MCAO) demonstrates the broad potential and universal applicability of our nanozyme-armed endothelial-derived nanovesicle platform. These pathologies, though affecting disparate organs, all share a common root in the endothelial dysfunction that is precipitated by ischemic and oxidative insults. Our in vivo results confirm that by simultaneously combating oxidative stress and delivering potent pro-angiogenic signals, the CeO₂-NMU_iECs can effectively intervene in this core pathological process to promote tissue regeneration and functional recovery.

Preclinical evaluation and optimization of CeO₂-NMU_iECs for clinically relevant applications

A critical prerequisite for the clinical translation of nanomedicine is a comprehensive evaluation of its biocompatibility and potential to elicit an immune response.⁵¹ Vesicles derived from allogeneic sources, such as iPSCs, while offering scalability, necessitate a thorough assessment of their immunogenic potential. We, therefore, conducted a multi-tiered investigation into the biosafety and immunogenicity of our CeO₂-NMU_iECs platform, encompassing different species, dose regimens, and administration routes, to provide robust support for its future clinical application. First, we performed long-term, repeated-dose toxicity studies in healthy rodents. In C57BL/6 mice, intravenous injection of CeO₂-NMU_iECs via the tail vein (50 μg/mouse) three times a week for 6 weeks did not induce any noticeable adverse effects (Figure S19A). Histological analysis of major organs, including the heart, liver, spleen, lungs, kidneys, brain, and bone marrow, revealed no signs of pathological damage, inflammation, or lesions compared with the PBS-treated control group (Figure S19B). Furthermore, a comprehensive panel of hematological and serum biochemical markers remained within the normal physiological range, indicating no damage to hematopoietic, hepatic, or renal functions (Figure S19C). We then extended this long-term safety evaluation to a second species, Sprague-Dawley rats. Following a similar 6-week injection schedule (200 μg/rat) (Figure S19D), we again observed no overt toxicity. Histological sections of all major organs were unremarkable (Figure S19E). Blood analysis revealed a minor but statistically significant increase in red blood cell count and a slight decrease in ALT levels, both of which remained well within the normal reference range for healthy rats and were not considered indicative of toxicity (Figure S19F). To further probe the safety profile in a large animal model, we administered CeO₂-NMU_iECs (5 mg/pig) to healthy Bama minipigs via ear margin vein injection three times a week for 2 weeks. This treatment was well tolerated, with no pathological abnormalities observed in any major organs and no significant alterations in hematological parameters, apart from a minor, clinically insignificant increase in platelet count (Figures S20A–S20C). Finally, to determine the safety margin, we conducted a dose-escalation study in mice for 4 weeks. Even at doses up to five times the therapeutic dose (5X, 250 μg/mouse), CeO₂-NMU_iECs did not cause any organ damage or significant changes in blood biochemistry, demonstrating a wide therapeutic window (Figures S20D–S20F).

Next, we systematically assessed the immunogenicity of the platform. To evaluate the potential for a systemic inflammatory response, we administered a single intravenous dose of CeO₂-NMU_iECs (50 μg) to healthy mice (Figure S21A). Analysis of serum at 24, 48, and 72 h showed no significant elevation in the levels of key pro-inflammatory cytokines, including IL-1β, IL-6, TNF-α, and IFN-γ, compared with the PBS control group (Figure S21B). To determine whether the platform triggered an adaptive immune response, we analyzed lymphocyte populations in primary and secondary immune organs at 72 h post-injection. Flow cytometric analysis revealed no significant changes in the proportions of T cells (CD3⁺), B cells (CD19⁺), or helper (CD4⁺) and cytotoxic (CD8⁺) T cell subsets in the bone marrow and thymus. A slight but statistically significant increase in the percentage of CD8⁺ T cells was noted in the spleen, a common observation for nanoparticle administration that did not translate to a systemic inflammatory state (Figures S21C–S21E; Figure S22). Finally, we evaluated the local inflammatory response to subcutaneous and intramuscular injections, routes relevant for certain therapeutic applications (Figure S21F). Histological analysis revealed a mild, transient inflammatory infiltrate at the injection sites at 24 h, characterized by an increase in F4/80-positive macrophages. This response was largely resolved by 72 h in both skin and muscle tissues, consistent with a normal, self-limiting foreign body response (Figures S21G and S21H). This was mirrored by a transient, localized increase in the pro-inflammatory cytokines IL-1β and TNF-α at 24 h, which returned to baseline levels by 72 h (Figures S21I and S21J).

Beyond safety, clinical translation requires standardized manufacturing. Because source-cell state directly affects nanovesicle quality, we tested whether iPSC-EC passage number influences NMU_iECs efficacy. NMU_iECs were generated from iPSC-ECs at passages 1, 2, 3, 5, and 7 (P1, P2, P3, P5, and P7, respectively) and evaluated for pro-angiogenic activity. NMU_iECs from early-passage cells (P1–P3) markedly promoted endothelial tube formation and migration, whereas efficacy declined at P5 and was largely lost at P7 (Figures S23A–S23C). These results show that source-cell physiological status determines NMU_iECs bioactivity and support the use of well-controlled early-passage iPSC-ECs for manufacturing. To assess translational potential, we extended therapeutic evaluation to a Bama minipig ischemic skin flap model, which more closely resembles human skin anatomy and physiology (Figure S23D). Consistent with the rodent data, gross inspection and infrared thermography on postoperative day 7 showed pronounced distal necrosis and reduced surface temperature in PBS-treated flaps. Local subcutaneous delivery of NMU_iECs partially improved tissue viability, whereas CeO₂-NMU_iECs produced the best outcome, with the largest flap survival area and near-normalization of surface temperature (Figures S23E–S23H). Histological analyses supported these findings. CD31 immunohistochemistry and CD34 immunofluorescence showed a denser microvascular network in the CeO₂-NMU_iECs group than in the NMU_iECs and PBS groups (Figures S23I and S23J, S23L, and S23M), and in situ DHE staining indicated more effective suppression of local ROS within the ischemic tissue (Figures S23K–S23N). Moreover, H&E and picrosirius red staining under polarized light revealed marked dermal disruption and loss of mature type I collagen in PBS controls, while CeO₂-NMU_iECs preserved dermal architecture and maintained a more organized, collagen-rich matrix compared with unmodified NMU_iECs (Figures S23O–S23Q).

In summary, the comprehensive preclinical evaluation demonstrates that the CeO₂-NMU_iECs platform possesses a highly favorable biosafety and immunogenicity profile across multiple species. Crucially, by defining the optimal manufacturing window for the source iPSC-ECs and confirming therapeutic efficacy in large animal models, this integrated platform presents a highly viable and standardizable strategy for the clinical treatment of ischemic diseases.

Discussion

We report the development of a universal, dual-targeting nanovesicle platform, CeO₂-NMU_iECs, that homes to and rescues endothelial dysfunction across diverse ischemic pathologies involving the bone, skin, heart, and brain. Engineered from iPSC-derived endothelial cells, these vesicles combine two innate targeting mechanisms. CXCR4/SDF-1 chemotaxis directs them to the ischemic tissue, and homotypic endothelial affinity ensures selective uptake by the local endothelium. Upon arrival, ceria nanozyme conjugation provides potent ROS-scavenging activity, which works in concert with the retained pro-angiogenic cargo to restore endothelial function in oxidative microenvironments that normally suppress vascular repair. This integrated approach addresses a long-standing challenge in ischemic therapy by combining microenvironmental correction with lineage-specific regenerative signaling.

Our work also builds on recent progress in hiPSC-derived extracellular vesicle and nanovesicle therapeutics. Previous studies have shown that hiPSC-derived MSC exosomes can attenuate limb ischemia,⁵² accelerate cutaneous wound repair,⁵³ and enhance post-stroke angiogenesis,⁵⁴ largely by promoting vascular regeneration and tissue repair. Endothelial-lineage hiPSC vesicles have also begun to emerge as a biologically matched strategy for vascular repair. For example, hiPSC-EC-derived exosomes were reported to promote postnatal angiogenesis in ischemic limbs,⁵⁵ while iPSC-EC-derived bioinspired nanovesicles generated by serial extrusion were shown to convert the skeletal endothelium-associated secretory phenotype to treat osteoporosis.⁸ Compared with these studies, our platform differs in three main aspects. First, instead of selecting a single iPSC-derived cell source a priori, we systematically compared nanovesicles from multiple isogenic iPSC-derived lineages and identified the endothelial lineage as the most pro-angiogenic source for endothelial repair. Second, rather than relying on unmodified vesicle cargo alone, we armed NMU_iECs with ultrasmall CeO₂ nanozymes to overcome the oxidative microenvironment that limits vesicle efficacy in ischemic tissues. Third, we validated the resulting platform across multiple endothelial injury models and four anatomically distinct ischemic disease models, supporting a broader endothelial-repair framework rather than a disease-specific vesicle therapy.

In comparison with existing strategies such as antibody- or peptide-modified nanoparticles,⁵⁶ exogenous growth factor delivery,⁵¹^,⁵⁷ or gene therapy,³⁸^,⁵⁸ our platform avoids the need for synthetic targeting ligands or viral vectors. It achieves systemic delivery through intravenous administration, displays intrinsic specificity for injured vasculature, and unites redox modulation with angiogenic stimulation in a single therapeutic entity. This design integrates advances from nanotechnology, regenerative medicine, and redox biology into a clinically translatable format.

The use of iPSC-derived endothelial cells also opens a path toward precision medicine.⁵⁹^,⁶⁰ Patient-specific iPSCs could be differentiated into endothelial cells and used to generate autologous CeO₂-NMU_iECs, potentially reducing immune rejection and enabling repeated administrations. For acute clinical scenarios, universal HLA-edited iPSC banks or cryopreserved intermediates may allow rapid, off-the-shelf production.⁶¹^,⁶² Clinical translation will require GMP-compliant manufacturing, robust batch-to-batch quality control based on defined functional metrics such as ROS-scavenging capacity and CXCR4 expression, and rigorous safety testing in large-animal models.⁶³ For future large-animal MI and stroke studies, intravenous administration would be the preferred initial route because it preserves the intended SDF-1/CXCR4-mediated ischemic homing and endothelial-lineage homotypic uptake of CeO₂-NMU_iECs. Local or regional delivery may be suitable for surgically accessible ischemic tissues, whereas catheter-based intracoronary or intra-arterial delivery could be compared with systemic administration in future MI or stroke models. In parallel, scalable translation will require closed, xeno-free expansion of banked allogeneic iPSCs and iPSC-derived endothelial cells, scalable nanovesicle production and purification, and release criteria that capture vesicle yield, CeO₂ loading, endothelial identity, potency, safety, and batch consistency.

Several important questions remain. First, while we demonstrated that endothelial-lineage nanovesicles possess stronger pro-angiogenic activity than those derived from MSCs and NPCs, pericytes and macrophages are also critical orchestrators of angiogenesis and tissue repair. Future studies expanding the iPSC differentiation repertoire to include these isogenic nanovesicles for comparative analysis will further optimize this lineage-guided engineering strategy. Furthermore, while we confirmed SDF-1-mediated homing, the spatial and temporal dynamics of SDF-1 expression in different ischemic tissues were not examined, yet this information could guide the optimal timing and frequency of administration. In this study, CeO₂-NMU_iECs were administered shortly after ischemia induction to test their proof-of-concept efficacy during the early phase of endothelial oxidative injury. However, delayed treatment is highly relevant in clinical scenarios such as MI, ischemic stroke, traumatic flap ischemia, and postoperative tissue ischemia, where treatment may be initiated after delayed diagnosis, transfer, reperfusion, or surgical decision-making. We have not systematically defined the latest post-injury time point at which CeO₂-NMU_iECs remain beneficial. Future studies should determine the therapeutic window of CeO₂-NMU_iECs by evaluating delayed administration at clinically relevant time points after ischemic injury. Mechanistic insights in this study were primarily obtained from proteomics, which confirmed enrichment of angiogenesis-related proteins, but extracellular vesicles carry a diverse array of cargo including microRNAs (miRNAs), lipids, and phosphorylated proteins.⁶⁴ The contribution of these components should be elucidated through targeted functional studies. Furthermore, although the therapeutic benefit to endothelial cells was evident, the influence of CeO₂-NMU_iECs on endothelial communication with other cell types such as macrophages, osteogenic cells, or astrocytes has not been explored. Such interactions may underlie broader tissue repair processes beyond vascular restoration.⁸^,⁶⁵

Beyond preserving the structural integrity of the nanovesicles, the continuous ROS scavenging mediated by CeO₂ nanozymes likely exerts a direct synergistic enhancement on the functional activity of this diverse cargo. In the oxidative stress of the ischemic microenvironment, unprotected therapeutic proteins and nucleic acids are susceptible to rapid oxidative modification and premature degradation. The local catalytic shield provided by CeO₂ actively neutralizes these reactive species during extracellular transit and cellular uptake. This protection potentially enables the lineage-specific pro-angiogenic proteins and regulatory miRNAs to retain their bioactivity upon intracellular delivery. Therefore, the extent to which CeO₂-mediated ROS scavenging exerts a direct synergistic enhancement on the functional activity of specific vesicular cargo, including proteins and miRNAs, warrants thorough exploration in future studies.

In summary, CeO₂-NMU_iECs combine lineage-optimized vesicle targeting with nanozyme-based modulation of the ischemic microenvironment, providing a versatile and non-invasive platform for the treatment of ischemic diseases across multiple organ systems. Future research should focus on establishing the translational framework including imaging-guided delivery, dosing window optimization, and pharmacokinetic–pharmacodynamic modeling, while also enhancing therapeutic potency by incorporating additional regenerative or metabolic cargo without compromising safety or immunogenicity.

Limitations of the study

Although our study establishes CeO₂-NMU_iECs as a dual-targeting endothelial lineage nanovesicle platform with therapeutic efficacy across multiple ischemia models, several limitations should be acknowledged. First, the findings remain primarily based on experimental models, and validation using human ischemic tissues or clinically relevant patient-derived systems will be needed to further assess endothelial targeting and therapeutic relevance in heterogeneous human disease settings. Second, large-animal evaluation was performed in an ischemic skin flap model; whether similar biodistribution and efficacy can be achieved in large-animal models of osteonecrosis, MI, or ischemic stroke remains to be determined. Third, although the CXCR4/SDF-1 axis and endothelial homotypic affinity support the proposed dual-targeting mechanism, the molecular determinants underlying endothelial-selective uptake and the relative contribution of individual vesicle cargo proteins require further investigation. Finally, only male animals were used in this study, and potential sex-dependent differences in efficacy, biodistribution, and immune compatibility were not directly assessed.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Xiaoyun Pan (xiaoyunpan@wmu.edu.cn).

Materials availability

All unique/stable reagents generated in this study are available from the lead contact with a completed materials transfer agreement.

Data and code availability

•
Data: The RNA-seq data generated in this study have been deposited in the Genome Sequence Archive (GSA) under accession numbers HRA019213, HRA019214, and HRA019215. The proteomics data generated in this study have been deposited to the ProteomeXchange Consortium via the iProX partner repository with the dataset identifier PXD080200.
•
Code: This paper does not report original code.
•
General statement: Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.

Acknowledgments

This work was funded by the National Natural Science Foundation of China (nos. 32101127, 31971260, and 82272564), the Medical and Health Technology Program of Zhejiang Province (nos. 2020KY187 and 2023KY148), the Natural Science Foundation of Zhejiang Province (no. LQN25H060012), and the Wenzhou Municipal Science and Technology Bureau (no. Y20210400). We thank Scientific Research Center of Wenzhou Medical University for providing excellent consultation and instrumental supports.

Author contributions

J.D., L.C., and X.P. conceived the project. H.J., Y.W., J.D., L.C., and X.P. designed the studies. H.J. and Y.W. performed most of the cell culture and molecular, biochemical, and animal experiments and analyzed the proteomics and RNA-seq data. L.Z. and H.S. assisted with the iPSC differentiation and nanovesicle preparation and characterization. L.J. and C.L. assisted with the in vitro functional assays. H.H. assisted with the ischemic skin flap animal model. G.Z. assisted with the GIONFH model. H.F., J.L., and J.S. provided assistance in the MI and MCAO animal experiments. X.X. and J.W. provided valuable suggestions for the project. H.J., Y.W., J.D., and X.P. wrote and revised the manuscript with input from all authors.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies

Cyanine7 anti-human CD31 Antibody	BioLegend	Cat#303117; RRID: AB_2114314
APC anti-human CD144 (VE-Cadherin) Antibody	BioLegend	Cat#348508; RRID: AB_10639949
PE anti-human CD73 (Ecto-5′-nucleotidase) Antibody	BioLegend	Cat#344004; RRID: AB_2298698
APC anti-human CD90 (Thy1) Antibody	BioLegend	Cat#328114; RRID: AB_893431
FITC anti-human CD105 Antibody	BioLegend	Cat#323203; RRID: AB_755955
PE anti-rat CD45 Antibody	BioLegend	Cat#202207; RRID: AB_314006
Mouse/Rat CD31/PECAM-1 APC-conjugated Antibody	R&D Systems	Cat#FAB3628A-100
Rabbit monoclonal [UMB2] to CXCR4	Abcam	Cat#ab124824
6 nm Colloidal Gold AffiniPure® Goat Anti-Rabbit IgG (H + L) (EM Grade)	Jackson ImmunoResearch	Cat#111-195-144
Biotin anti-human CXCR4	BioLegend	Cat#306504; RRID: AB_314610
Alexa Fluor® 488 Mouse monoclonal [KILL150A] to CD63	Abcam	Cat#ab318246
Mouse monoclonal [JC/70A] to CD31	Abcam	Cat#ab9498; RRID: AB_307284
Rabbit polyclonal to Von Willebrand Factor	Abcam	Cat#ab6994; RRID: AB_305689
Rabbit polyclonal to VE Cadherin - Intercellular Junction Marker	Abcam	Cat#ab33168; RRID: AB_870662
Rabbit monoclonal [SP103] to Nestin	Abcam	Cat#ab105389; RRID: AB_10859398
Mouse monoclonal [20G5] to SOX2	Abcam	Cat#ab171380; RRID: AB_2732072
Mouse monoclonal [AD2.38] to PAX6	Abcam	Cat#ab78545; RRID: AB_1566562
VEGF Receptor 1 (E7T9H) Rabbit mAb	Cell Signaling Technology	Cat#64094; RRID: AB_3697182
FGF Receptor 1 (D8E4) XP® Rabbit mAb	Cell Signaling Technology	Cat#9740; RRID: AB_11178519
Rabbit monoclonal [EPR24075-418] to FGFR2	Abcam	Cat#ab289968; RRID: AB_3668756
Rabbit monoclonal [EPR23808-72] to NRP2	Abcam	Cat#ab273584
Rabbit monoclonal [EPR11803-97] to Angiomotin like 1	Abcam	Cat#ab171976
Rabbit monoclonal [EPR7130(B)] to TSG101	Abcam	Cat#ab125011; RRID: AB_10974262
RASIP1 Polyclonal antibody	Proteintech	Cat#17971-1-AP; RRID: AB_2878474
SRPX2 Polyclonal antibody	Proteintech	Cat#11845-1-AP; RRID: AB_2195006
Rabbit monoclonal [EPR18239] to SIRT1	Abcam	Cat#ab189494; RRID: AB_2864311
Rabbit monoclonal [EPR5554(N)] to NRF1 - ChIP Grade	Abcam	Cat#ab175932; RRID: AB_2629496
Rabbit polyclonal to beta Actin	Abcam	Cat#ab8227; RRID: AB_2305186
NRF2 (E5F1A) Rabbit mAb	Cell Signaling Technology	Cat#20733; RRID: AB_2934224
HO-1 (E3F4S) Rabbit mAb	Cell Signaling Technology	Cat#43966; RRID: AB_2799254
NQO1 (D6H3A) Rabbit mAb	Cell Signaling Technology	Cat#62262; RRID: AB_2799623
PE/Cyanine7 anti-mouse CD45 Antibody	BioLegend	Cat#157206; RRID: AB_2860726
FITC anti-mouse CD3 Antibody	BioLegend	Cat#100204; RRID: AB_312661
Brilliant Violet 421™ anti-mouse CD4 Antibody	BioLegend	Cat#116023; RRID: AB_2800579
APC anti-mouse CD8a Antibody	BioLegend	Cat#100712; RRID: AB_312751
APC anti-mouse CD19 Antibody	BioLegend	Cat#115512; RRID: AB_313647
Rabbit polyclonal anti- COL1 antibody	Proteintech	Cat#14695-1-AP; RRID: AB_2082037
Rabbit polyclonal anti-ANG1 antibody	Proteintech	Cat#23302-1-AP; RRID: AB_2879250
Rabbit polyclonal anti-CD31 antibody	Servicebio	Cat#GB11063; RRID: AB_2922436
Rabbit polyclonal anti-osteocalcin antibody	Servicebio	Cat#GB11233; RRID: AB_3676307
Rabbit monoclonal [EP1176Y] to VEGFA - C-terminal	Abcam	Cat#ab52917; RRID: AB_883427
Anti-Sarcomeric Alpha Actinin Rabbit pAb	Servicebio	Cat#GB11555; RRID: AB_2814686
Anti-Connexin 43/GJA1 Mouse mAb	Servicebio	Cat#GB12234
Anti-alpha smooth muscle Actin Rabbit pAb	Servicebio	Cat#GB111364; RRID: AB_2910228
Anti-Von Willebrand Factor Mouse mAb	Servicebio	Cat#GB120019
Anti-GFAP Rabbit pAb	Servicebio	Cat#GB11096; RRID: AB_2904015
Anti-NeuN Rabbit pAb	Servicebio	Cat#GB11138; RRID: AB_2868432
Goat Anti-Rabbit IgG H&L (Alexa Fluor® 488)	Abcam	Cat#ab150077; RRID: AB_2630356
Goat Anti-Mouse IgG H&L (Alexa Fluor® 488)	Abcam	Cat#ab150113; RRID: AB_2576208
Goat Anti-Rabbit IgG H&L (Alexa Fluor® 594)	Abcam	Cat#ab150080; RRID: AB_2650602
Goat Anti-Mouse IgG H&L (Alexa Fluor® 594)	Abcam	Cat#ab150116; RRID: AB_2650601
HRP-conjugated Goat Anti-Rabbit IgG(H + L)	Proteintech	Cat#SA00001-2; RRID: AB_2722564
HRP-conjugated Goat Anti-Mouse IgG(H + L)	Proteintech	Cat#SA00001-1; RRID: AB_2722565

Chemicals, peptides, and recombinant proteins

Collagenase type II	Sigma-Aldrich	Cat# C6885
Cerium(III) acetate	Aladdin	Cat# C106112
Oleylamine	Aladdin	Cat# O431504
DSPE-PEG(2000)-methoxy	Avanti Polar Lipids	Cat# 880120P
DSPE-PEG(2000)-maleimide	Avanti Polar Lipids	Cat# 880126P
Traut’s reagent	Thermo Fisher Scientific	Cat# 26101

Critical commercial assays

STEMdiff Endothelial Differentiation Kit	STEMCELL Technologies	Cat# 08005
STEMdiff Mesenchymal Progenitor Kit	STEMCELL Technologies	Cat# 05240
STEMdiff Neural System	STEMCELL Technologies	Cat# 08581
STEMdiff Neural Progenitor Medium	STEMCELL Technologies	Cat# 05833
Osteogenic differentiation kit	OriCell	Cat# HUXXC-90021
Adipogenic differentiation kit	OriCell	Cat# HUXXC-90031
Chondrogenic differentiation kit	OriCell	Cat# HUXXC-90041
TUNEL staining kit for cultured cells	Yeasen Biotechnology	Cat# 40306ES20
Annexin V-FITC and PI apoptosis detection kit	Yeasen Biotechnology	Cat# 40302ES20
Masson’s trichrome staining kit	Servicebio	Cat# G1006
Goldner’s trichrome staining kit	Servicebio	Cat# G1064
Nissl staining kit	Servicebio	Cat# G1036

Deposited data

RNA-seq data generated in this study	This paper	GSA: HRA019213; HRA019214; HRA019215
Proteomics data generated in this study	This paper	iProX: PXD080200

Experimental models: Cell lines

Human: induced pluripotent stem cell line	Hunan Fenghui Biotechnology Co., Ltd.	N/A
Human: iPSC-derived endothelial cells (iPSC-ECs)	This paper	N/A
Human: iPSC-derived mesenchymal stem cells (iPSC-MSCs)	This paper	N/A
Human: iPSC-derived neural progenitor cells (iPSC-NPCs)	This paper	N/A
Human: HUVECs	Shanghai Cell Bank	N/A
Human: HDMECs	ATCC	Cat#PCS-110-010
Rat: primary bone marrow endothelial cells (BMECs)	This paper	N/A
Rat: primary brain microvascular endothelial cells (RBMVECs)	This paper	N/A
Mouse: coronary artery endothelial cells (MCAECs)	Gift from Prof. Chang Jiang	N/A
Rat: primary bone marrow-derived mesenchymal stem cells (BMSCs)	This paper	N/A
Rat: primary chondrocytes	This paper	N/A
Rat: primary nucleus pulposus cells	This paper	N/A
Mouse: RAW 264.7 macrophages	ATCC	ATCC: TIB-71

Experimental models: Organisms/strains

Sprague-Dawley rat	Animal Center of the Chinese Academy of Sciences (Shanghai, China)	N/A
C57BL/6 mice	Animal Center of the Chinese Academy of Sciences (Shanghai, China)	N/A
Bama minipig	Jiangsu Changzhou Beiwang Biological Technology Co., Ltd.	N/A

Oligonucleotides

Primers for quantitative PCR, see Table S1	Sangon Biotech	N/A

Software and algorithms

BioRender	BioRender	https://www.biorender.com/
GraphPad Prism	GraphPad	https://www.graphpad.com/
FlowJo	BD Life Sciences	https://www.flowjo.com/
Zetasizer	Malvern Instruments	https://www.malvernpanalytical.com/en/
ImageJ	ImageJ	https://imagej.net/ij/
LivingImage	PerkinElmer	https://www.perkinelmer.com.cn/
Image Lab	Bio-Rad	https://www.bio-rad.com/
AngioTool	National Cancer Institute	https://ccrod.cancer.gov/confluence/display/ROB2/Home
NRecon	Bruker microCT	https://www.bruker.com/
CTAn	Bruker microCT	https://www.bruker.com/
CTVol	Bruker microCT	https://www.bruker.com/
fastp	OpenGene	https://github.com/OpenGene/fastp
HISAT2	Johns Hopkins University, Center for Computational Biology	http://daehwankimlab.github.io/hisat2/
DESeq2	Bioconductor	https://bioconductor.org/packages/release/bioc/html/DESeq2.html
Gene Set Enrichment Analysis software	Broad Institute/UC San Diego	https://www.gsea-msigdb.org/gsea/index.jsp

Open in a new tab

Experimental model and study participant details

Human induced pluripotent stem cells (iPSCs)

A commercially available human induced pluripotent stem cell (iPSC) line derived from a male donor was purchased from Hunan Fenghui Biotechnology Co., Ltd. and cultured under feeder-free conditions. Culture plates were pre-coated with Matrigel basement membrane matrix (Corning; 1:80 dilution in DMEM/F-12) for 2 h at 37°C before use. Cells were seeded on the coated plates and maintained in mTeSR1 medium (STEMCELL Technologies), which was replaced daily. Cultures were maintained in a humidified incubator at 37°C with 5% CO₂. When cell colonies reached approximately 70–80% confluency, they were passaged as small aggregates using ReLeSR, an enzyme-free gentle cell dissociation reagent (STEMCELL Technologies), at a ratio of 1:6. Undifferentiated parental iPSCs were used to generate NMU_iPSCs and served as the isogenic source for iPSC-derived lineages used in this study. No additional independent STR authentication or karyotyping was performed in this study; cell identity was supported by supplier documentation and by the lineage-specific characterization described below. The iPSC cultures were tested and confirmed negative for mycoplasma contamination before experimental use.

iPSC-derived lineages

The induction of iPSC-ECs followed a seven-day stepwise differentiation protocol using STEMdiff Endothelial Differentiation Kit (STEMCELL Technologies). In brief, iPSCs were first induced to form mesoderm, then guided toward the endothelial lineage at the meso-endothelial progenitor stage, and finally matured into functional endothelial cells. The success of the differentiation process was validated through a multi-dimensional approach. Cell morphology was monitored by phase-contrast microscopy for the transition from a pluripotent colony form to a characteristic cobblestone monolayer. Angiogenic potential was assessed by tube formation on Matrigel (Corning). The purity of the resulting endothelial population was evaluated by flow cytometry for CD31 and CD144 co-expression. Endothelial protein expression was further confirmed by immunofluorescence staining for CD31, von Willebrand factor (vWF), and VE-cadherin. Transcriptomic analysis by RNA sequencing (RNA-seq) was performed to compare the gene expression profile of iPSC-ECs with that of undifferentiated iPSCs (OE Biotech Co., Ltd., Shanghai, China).

The induction and expansion of iPSC-MSCs followed an established stepwise differentiation protocol using STEMdiff Mesenchymal Progenitor Kit (STEMCELL Technologies). Successful differentiation was validated by assessing cell morphology, trilineage differentiation potential, surface marker expression, and transcriptomic profile. Morphological characteristics were observed for a typical spindle-shaped, fibroblast-like appearance. Trilineage differentiation potential was evaluated by inducing osteogenesis (OriCell), adipogenesis (OriCell), and chondrogenesis (OriCell), followed by staining with Alizarin Red S, Oil Red O, and Alcian blue, respectively. Surface marker expression of CD73, CD90, and CD105 was analyzed by flow cytometry. The transcriptomic profile was compared to that of the parent iPSCs using RNA-seq data (OE Biotech Co., Ltd., Shanghai, China).

The induction and culture of iPSC-NPCs followed a stepwise differentiation protocol using STEMdiff Neural System (STEMCELL Technologies) and STEMdiff Neural Progenitor Medium (STEMCELL Technologies). The success of the differentiation was validated by assessing cell morphology, marker protein expression, and transcriptomic profile. Morphological changes, including the formation of neural rosette-like structures, were monitored during the differentiation process. Neural progenitor identity was confirmed by immunofluorescence staining for Nestin, SOX2, and PAX6. Transcriptomic profiling was used to compare the gene expression pattern of iPSC-NPCs with that of the parental iPSC state (OE Biotech Co., Ltd., Shanghai, China). All iPSC-derived lineages were generated from the same parental male iPSC line. Therefore, sex was defined by the parental iPSC donor. All iPSC-derived cultures used for experiments were tested and confirmed negative for mycoplasma contamination.

Endothelial and uptake assay cell models

Primary rat bone marrow endothelial cells (BMECs) were isolated from the femurs and tibias of 4-week-old male Sprague-Dawley (SD) rats as described before.⁸ Briefly, bone marrow was flushed out and cultured overnight in DMEM (Gibco) supplemented with 20% FBS (Gibco). The adherent cells were then harvested, and BMECs were positively selected using CD31 MicroBeads (Miltenyi Biotec). Purified BMECs were cultured on flasks pre-coated with rat-tail collagen type I (Corning) in Endothelial Cell Medium (ECM, Sciencell). Characterization was performed by morphology and flow cytometry analysis for a CD31⁺/CD144⁺/CD45^- phenotype. Primary rat brain microvascular endothelial cells (RBMVECs) were isolated from the cerebral cortices of neonatal male SD rats.⁴⁸ Briefly, minced gray matter was digested with 0.05% trypsin and filtered. The cell pellet was then subjected to a 20% BSA density gradient centrifugation to remove myelin. The resulting microvessel pellet was further digested with 0.1% collagenase type II (Sigma-Aldrich). The isolated RBMVECs were cultured in Endothelial Cell Medium. Characterization was performed by immunofluorescence for Factor-VIII and flow cytometry for a CD31⁺/CD144⁺/CD105⁺/CD45^- profile. Human umbilical vein endothelial cells (HUVECs) were obtained from the Shanghai Cell Bank and cultured in ECM. Human dermal microvascular endothelial cells (HDMECs; ATCC, PCS-110-010) were purchased from ATCC and cultured in ECM. Mouse coronary artery endothelial cells (MCAECs) were provided by Prof. Chang Jiang and cultured in ECM. The endothelial identity of these cells was confirmed by morphology and marker expression, including CD31, CD144, CD45, CD105, Factor VIII, and/or VE-cadherin, depending on the cell type. Commercially obtained or gifted endothelial cells were used according to the corresponding supplier or provider information. Donor-level information, including sex when available, was not independently determined in this study. Additional primary cells and cell lines used for comparative uptake studies were prepared as follows. Bone marrow-derived mesenchymal stem cells (BMSCs) were isolated by flushing the femoral and tibial cavities of 4-week-old male SD rats and culturing the adherent cell fraction.⁶⁶ Primary chondrocytes were isolated from the knee cartilage of neonatal male SD rats by digestion with 2 mg/mL collagenase type II.⁶⁷ Rat nucleus pulposus (NP) cells were isolated from the intervertebral discs of 4-week-old male SD rats using a similar collagenase digestion method.⁶⁸ RAW 264.7 macrophages were purchased from ATCC. These four cell types were cultured in DMEM/F12 (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin-streptomycin (Gibco) in a humidified incubator at 37°C with 5% CO₂. Commercial cell lines were authenticated by the original suppliers or cell banks, and no additional independent STR authentication was performed after receipt. For primary cultures and gifted endothelial cells, cellular identity was confirmed by morphology and lineage marker expression as described above. All cell lines and primary cultures used in this study were tested and confirmed negative for mycoplasma contamination before experimental use.

Animals

All animal experiments were conducted in accordance with the Regulations of the People’s Republic of China on the Administration of Experimental Animals and were approved by the Animal Care and Use Committee of Wenzhou Medical University under approval numbers wydw2024-0465, wydw2025-0427, wydw2025-0434, and wydw2025-0442. Wild-type male Sprague-Dawley rats and C57BL/6 mice were obtained from the Animal Center of the Chinese Academy of Sciences (Shanghai, China). Animals were housed in a specific-pathogen-free facility under controlled temperature (18°C–23°C), humidity (40–60%), and a 12-h light/dark cycle, with food and water available ad libitum. Animals were drug- and test-naive before model induction unless otherwise specified. Male Sprague-Dawley rats were used for the glucocorticoid-induced osteonecrosis of the femoral head model, myocardial infarction model, middle cerebral artery occlusion model, pharmacokinetic studies, biodistribution studies, and systemic biosafety analyses. Specifically, 8-week-old male Sprague-Dawley rats were used for the glucocorticoid-induced osteonecrosis model, and 6-week-old male Sprague-Dawley rats were used for the myocardial infarction and middle cerebral artery occlusion models. Male C57BL/6 mice were used for the ischemic random-pattern skin flap model, dose-escalation study, systemic immunogenicity assessment, local immunogenicity assessment, and mouse biosafety evaluation. Six-week-old male C57BL/6 mice were used for the ischemic skin flap model. For large-animal preclinical evaluation, male Bama minipigs aged approximately 8 months and weighing 20–30 kg were obtained from Jiangsu Changzhou Beiwang Biological Technology Co., Ltd. Male Bama minipigs were used to establish a random-pattern ischemic skin flap model for local therapeutic evaluation. Bama minipigs were housed under standard conditions and managed according to the approved animal protocol. Detailed procedures for model establishment, treatment administration, imaging, tissue collection, and endpoint analyses are provided in METHOD DETAILS. Only male animals were used in this study to reduce variability across ischemic injury models and to maintain consistency with the preclinical disease models used. Therefore, the influence of sex on therapeutic efficacy and safety was not directly assessed, which should be considered when generalizing the findings.

Method details

Synthesis and functionalization of ceria (CeO₂) nanozymes

Ultrasmall CeO₂ nanoparticles were synthesized via a thermal decomposition-hydrolysis method.³³ Briefly, cerium (III) acetate (0.43 g, 1 mmol; Aladdin), oleylamine (3.25 g, 12 mmol; Aladdin) and xylene (15 mL; Aladdin) were stirred vigorously for 12 h at room temperature. The solution was then heated to 90°C under a nitrogen atmosphere. Following this, 1 mL of deionized water was rapidly injected. The reaction was maintained at 90°C for 3 h, then cooled to room temperature. The resulting CeO₂ nanoparticles were precipitated by the addition of acetone, collected by centrifugation (12,000g for 10 min), and redispersed in chloroform for storage. For subsequent conjugation and in vivo stabilization, the nanoparticles were functionalized with maleimide groups. This was achieved by encapsulating the hydrophobic CeO₂ nanoparticles with an amphiphilic phospholipid-polyethylene glycol (PEG) layer. This amphiphilic DSPE-PEG coating facilitates the phase transfer of hydrophobic CeO₂ into a stable aqueous suspension, introduces reactive maleimide handles for covalent nanovesicle anchoring, and provides a hydrophilic steric corona designed to minimize in vivo non-specific protein adsorption and prolong systemic circulation. A solution containing 20 mg of DSPE-PEG(2000)-methoxy (Avanti Polar Lipids) and 10 mg of DSPE-PEG(2000)-maleimide (Avanti Polar Lipids) in 10 mL of chloroform was mixed with 2 mL of the CeO₂ nanoparticle suspension in chloroform (10 mg/mL) and sonicated in an ice bath for 5 min. The chloroform was then removed by rotary evaporation to form a thin lipid-nanoparticle film. The film was further dried in a vacuum oven at 60°C for 3 h. Finally, 4 mL of deionized water was added, and the mixture was sonicated to obtain a transparent colloidal suspension of maleimide-functionalized CeO₂ (Mal-CeO₂). The Mal-CeO₂ was purified by centrifugation and filtration through a 0.22 μm syringe filter to remove any non-encapsulated lipids or aggregates. The synthesized nanozymes were comprehensively characterized. The size and morphology were observed by transmission electron microscopy (TEM; Hitachi, HT-7700). The crystal structure was determined by X-ray diffraction (XRD; Shimadzu, XRD-6100). Elemental composition and crystalline nature were analyzed by energy-dispersive X-ray spectroscopy (EDS) and selected area electron diffraction (SAED). The chemical valence states of cerium were analyzed by high-resolution X-ray photoelectron spectroscopy (XPS; Thermo Scientific, Nexsa).

Preparation of nanovesicle mimicking units (NMUs)

NMUs were prepared from the four cell types (iPSCs, iPSC-ECs, iPSC-MSCs, and iPSC-NPCs) using an extrusion technique.⁷^,²⁸ Briefly, 1 × 10⁶ cells were resuspended in PBS and subjected to sequential extrusion through polycarbonate membrane filters (Whatman) with pore sizes of 5 μm and 1 μm, respectively, using a mini-extruder (Avanti Polar Lipids). Following extrusion, the samples were centrifuged at 10,000g for 10 min to remove cell debris and larger vesicles. The NMUs were then purified and concentrated through three rounds of centrifugation at 1000g for 15 min using a 100 kDa centrifugal filter (Millipore, USA). The final product was stored at −80°C for future use. The total protein content of the NMUs was quantified using a bicinchoninic acid (BCA) assay kit (Beyotime, China). For subsequent conjugation, the NMU_iECs were functionalized with thiol groups. The NMU_iECs pellet was resuspended in pH 8 PBS and incubated with Traut’s reagent (Thermo Fisher Scientific) with gentle stirring for 30 min at room temperature, according to the manufacturer’s protocol. The resulting thiol-functionalized NMU_iECs (Thiol-NMU_iECs) were then purified by washing three times with Tyrode’s buffer via ultracentrifugation to remove unreacted reagent and stored at −80°C until use.

Covalent conjugation of CeO₂ nanozymes to NMU_iECs

The purified Thiol-NMU_iECs were resuspended in Tyrode’s buffer and mixed with the Mal-CeO₂ suspension to initiate the thiol-maleimide reaction.²¹ For the conjugation, 50 μg/mL of Thiol-NMU_iECs were combined with 100 μg/mL of Mal-CeO₂ and the mixture was gently stirred for 3 h at room temperature. To purify the final CeO₂-NMU_iECs, the reaction mixture was washed several times with PBS via ultracentrifugation to remove any unreacted Mal-CeO₂. The final product was resuspended in PBS for subsequent experiments.

Morphology, size, and surface potential analysis

The morphology of NMUs, CeO₂ and CeO₂-NMU_iECs was observed by TEM (HitachiHT-7700, Japan). High-resolution TEM (HRTEM) was used to examine the crystal lattice fringes of the conjugated CeO₂ nanoparticles. To confirm the stable conjugation of CeO₂ onto the vesicle surface, NMU_iECs were labeled with DiD (Thermo Fisher Scientific) and CeO₂ was labeled with FITC (Thermo Fisher Scientific). The co-localization of the two fluorescent signals was assessed using a super-resolution confocal microscope (A1R-SIM-STORM, Nikon), and the spatial overlap was analyzed using three-dimensional ortho-display images and fluorescence intensity line profiles (A1R-SIM-STORM, Nikon). The hydrodynamic diameter, polydispersity index (PDI), and zeta potential of the nanovesicles were measured by dynamic light scattering (DLS) using a Zetasizer Nano ZS instrument (Malvern). For colloidal stability assessment, NMUs and CeO₂-NMU_iECs were suspended in PBS or PBS containing 50% fetal bovine serum (FBS), and the hydrodynamic diameter and PDI were monitored over a 48-h period at 37°C using DLS.

Proteomics and western blot analysis

Comparative proteomic analysis of NMU_iECs and NMU_iPSCs was performed by OE Biotech Co., Ltd. (Shanghai, China). Briefly, total protein was extracted from samples using a phenol extraction buffer, precipitated with ammonium acetate-methanol, and quantified by bicinchoninic acid assay. Equal quantities of protein from each sample were digested for analysis by liquid chromatography-tandem mass spectrometry (LC-MS/MS). For western blot analysis, protein concentration was determined using a BCA Protein Assay Kit (Beytime). For each sample, 30 μg of protein was combined with 5x SDS-PAGE loading buffer, boiled for 10 min, and separated on a 10% SDS-PAGE gel. Proteins were then transferred onto a 0.22 μm PVDF membrane (Merck Millipore). The membrane was blocked with 5% skim milk for 2 h at room temperature before an overnight incubation at 4°C with primary antibodies against pro-angiogenic proteins (FLT1, FGFR1, FGFR2, NRP2, AMOTL1, RASIP1, SRPX2, and CXCR4). After washing three times with TBST, membranes were incubated with HRP-conjugated secondary antibody for 2 h at room temperature. Protein bands were detected using chemiluminescence and imaged with the ChemiDoc XRS+ system with Image Lab V3.0 software (Bio-Rad). To assess the integrity of the total protein profile throughout the chemical modification process, samples from each step (iPSC-ECs, NMU_iECs, Thiol-NMU_iECs, and CeO₂-NMU_iECs) were also analyzed by SDS-PAGE followed by Coomassie Brilliant Blue staining (Beytime).

In vitro ROS scavenging and enzyme-mimetic activity assays

The antioxidant properties of the CeO₂, NMU_iECs and CeO₂-NMU_iECs were evaluated through a series of radical scavenging and enzyme-mimetic assays. The scavenging of 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radicals was assessed by monitoring the reduction in absorbance at 734 nm after incubating the samples with a pre-formed ABTS⁺⋅ solution (Macklin). Similarly, the scavenging of 2,2-diphenyl-1-picrylhydrazyl (DPPH) radicals was determined by measuring the decrease in absorbance at 517 nm after incubation with a 0.1 mM ethanolic DPPH solution (Macklin). The hydroxyl radical (⋅OH) scavenging capacity was quantified using a hydroxyl radical detection kit (Solarbio), where the inhibition of a colorimetric reaction was measured at 550 nm. The superoxide anion (O₂⋅^-) scavenging ability was evaluated using a WST-1 based assay kit (Solarbio), where the inhibition of formazan dye formation by xanthine oxidase was measured at 450 nm. The catalase (CAT)-mimetic activity was determined using a Catalase Assay Kit (Solarbio) by monitoring the decomposition of H₂O₂ via the decrease in absorbance at 240 nm. The superoxide dismutase (SOD)-mimetic activity was measured using a WST-1 based SOD assay kit (Solarbio), which quantifies the inhibition of the xanthine oxidase-catalyzed reaction at 450 nm. For dose-dependent studies on CeO₂ nanoparticles, concentrations ranging from 0 to 200 μg/mL were used. For comparative studies, CeO₂, NMU_iECs and CeO₂-NMU_iECs were used at equivalent concentrations as specified in the experiments.

Surface CXCR4 expression assessment

For qualitative analysis by immunogold labeling,⁶⁹ 10 μL of each NMU suspension was deposited onto a carbon-coated copper grid and incubated for 20 min at room temperature. After washing with PBS and 50 mM glycine, grids were blocked with 1% BSA for 15 min. The samples were then incubated with a primary anti-CXCR4 antibody for 1 h at room temperature, followed by incubation with a 6 nm Colloidal Gold AffiniPure Goat Anti-Rabbit IgG (H + L) secondary antibody (Jackson, 111-195-144) for 1 h. After fixation with 1% glutaraldehyde and extensive washing, the grids were visualized by TEM (HitachiHT-7700). For quantitative analysis, a magnetic bead capture assay was employed. First, the four NMU types were fluorescently labeled by incubation with a recombinant Alexa Fluor 488 Anti-CD63 antibody. Separately, Dynabeads MyOne Streptavidin C1 magnetic beads (Thermo Fisher Scientific) were washed and then incubated with a Biotin anti-human CXCR4 Antibody to prepare CXCR4-capture beads. The antibody-coated beads were then incubated with the Alexa Fluor 488-labeled NMUs. After magnetic separation and washing to remove unbound vesicles, the fluorescence intensity of the bead-captured NMUs was analyzed by flow cytometry (Beckman Coulter) and confocal microscopy (Nikon).

Cellular uptake and cytotoxicity analysis

Cellular uptake and cytotoxicity were evaluated in HUVECs. For the cellular uptake assay, HUVECs were seeded in 24-well plates (for microscopy) or 12-well plates (for flow cytometry) and allowed to adhere overnight. The medium was then replaced with fresh medium containing DiD-labeled NMU_iECs or CeO₂-NMU_iECs at concentrations of 0, 6.25, 12.5, 25, 50, and 100 μg/mL. The concentration of CeO₂-NMU_iECs was normalized to the protein content of the NMU_iECs component. After a 6-h incubation, cells were washed with PBS to remove non-internalized nanovesicles. For confocal microscopy, cells were fixed, counterstained with DAPI (Thermo Fisher Scientific), and imaged. For flow cytometry, cells were trypsinized, resuspended in PBS, and the mean fluorescence intensity (MFI) of the DiD signal was quantified. For cytotoxicity analysis, HUVECs (4 × 10³ cells per well) were seeded in a 96-well plate and cultured overnight. The medium was then replaced with fresh medium containing the corresponding treatments. Cell viability was assessed at 24, 48, and 72 h using a Cell Counting Kit-8 (CCK-8; Meilunbio) according to the manufacturer’s protocol. The absorbance at 450 nm was measured using a microplate reader. First, the cytotoxicity of NMU_iECs was evaluated across a concentration range of 0, 6.25, 12.5, 25, and 50 μg/mL. Subsequently, the cytotoxicity of different components was compared using the following treatments: PBS, CeO₂, NMU_iECs, a physical mixture (Mix), and CeO₂-NMU_iECs, all at a concentration equivalent to 50 μg/mL of the NMU_iECs component.

In vitro preferential uptake analysis

The preferential uptake of NMU_iECs by endothelial cells was assessed both quantitatively by flow cytometry (Beckman Coulter) and qualitatively by confocal microscopy (Nikon). For quantitative flow cytometry analysis, equal amounts of five different pre-labelled cell populations were cultured on Petri dishes and incubated overnight. DiD-labelled NMU_iECs were then added at a concentration of 50 μg/mL and incubated with five different cell populations for 6 h. Two separate mixed-cell models were established. For the bone microenvironment model, the five cell types were uniquely labeled as follows: BMECs (Hoechst 33342-labelled nuclei and Dil-labelled membranes; Hoechst⁺/Dil⁺/DiO⁻), BMSCs (Hoechst 33342-labelled nuclei and DiO-labelled membranes; Hoechst⁺/DiO⁺/Dil⁻), chondrocytes (Hoechst 33342-labelled nuclei; Hoechst⁺/Dil⁻/DiO⁻), NP cells (Dil-labelled membranes; Hoechst⁻/Dil⁺/DiO⁻), and macrophages (no pre-labelling; Hoechst⁻/Dil⁻/DiO⁻). For the diverse endothelial model, the five endothelial cell types were uniquely labeled using a similar strategy. The detailed gating strategies used to distinguish these populations and quantify the percentage of DiD-positive cells within each are provided in Figures S8A and S9A. For qualitative and time-course analysis by confocal microscopy, the five cell types for each model were seeded separately onto glass-bottom dishes. The cells were incubated with DiD-labelled NMU_iECs (deep red, 50 μg/mL), and imaging of fixed cells was performed at 1, 2, 4, and 6 h. For visualization of cellular morphology and NMU_iECs internalization, cells were fixed, permeabilized, and counterstained for F-actin with Phalloidin (green) and nuclei with DAPI (blue). Three-dimensional (3D) surface plots of DiD fluorescence intensity were generated from the confocal images using ImageJ software.

In vitro ischemic injury models

Three models were used to simulate key pathological features of ischemic injury in HUVECs or other primary endothelial cells (BMECs, HDMECs, MCAECs and RBMVECs). For the acute oxidative stress model, cells were co-incubated with 200 μM hydrogen peroxide (H₂O₂) with or without the indicated treatments. For the drug-induced injury model, cells were co-incubated with 160 μM of the glucocorticoid methylprednisolone (MPS) with or without the indicated treatments, simulating the pathology of glucocorticoid-induced osteonecrosis of the femoral head. For the oxygen-glucose deprivation (OGD) model, cells were first subjected to oxygen-glucose deprivation (OGD) by culturing in glucose-free medium for 8 h at 37°C in a hypoxic incubator (1% O₂, 5% CO₂). Following this ischemic insult, the medium was replaced with normal glucose-containing medium, and the cells were returned to normoxic conditions and treated with the indicated treatments. For initial comparative studies of different NMU types, cells were treated with NMU_iPSCs, NMU_iECs, NMU_iMSCs, or NMU_iNPCs (50 μg/mL). For subsequent experiments evaluating the engineered platform, the treatment groups included PBS, CeO₂, NMU_iECs, Mix, and CeO₂-NMU_iECs, with all concentrations normalized to an NMU_iECs component of 50 μg/mL.

Tube formation assay

The tube formation assay was utilized to assess angiogenic capacity. Matrigel matrix (Corning) was thawed on ice and 150 μL was added to each well of a pre-chilled 48-well plate. The plate was then incubated at 37°C for 1 h to allow for Matrigel polymerization. Approximately 6×10⁴ endothelial cells (including HUVECs, BMECs, HDMECs, MCAECs, and RBMVECs) were seeded onto the gelled Matrigel in each well. The cells were then treated with the different formulations under either normal or injury-inducing conditions (H₂O₂, MPS, or OGD). After an 8 h incubation period, the formation of tubular structures was visualized by staining with Calcein-AM (Thermo Fisher Scientific). Images were captured using a fluorescence microscope (Keyence, BZ-X800), and various parameters, including vessels percentage area, total number of junctions and total vessels length, were quantified using AngioTool.

Migration assay

A Transwell migration assay was employed to evaluate cell migration ability. Approximately 3×10⁴ endothelial cells were seeded into the upper compartments of 24-well inserts with an 8.0 μm pore size polycarbonate membrane (Corning). The cells were treated with the different therapeutic formulations under either normal or injury-inducing conditions. After an 18 h incubation period, cells that had migrated to the lower surface of the membrane were fixed in 4% paraformaldehyde for 15 min and then stained with 0.2% (w/v) crystal violet (Beyotime, China) for 15 min. Non-migrated cells on the upper surface of the membrane were carefully removed with a cotton swab. The stained cells were visualized and imaged using a bright-field microscope (Keyence, BZ-X800), and the number of migrated cells was counted for quantification by ImageJ software.

Intracellular and mitochondrial ROS analysis

To measure total intracellular ROS levels, endothelial cells were seeded in 24-well plates (for microscopy) or 12-well plates (for flow cytometry). Following the specified treatments under normal or injury-inducing conditions, the cells were washed with PBS and then incubated with 10 μM 2′,7′-dichlorofluorescin diacetate (DCFH-DA; Beyotime) in serum-free medium for 30 min at 37°C in the dark. After incubation, cells were washed again with PBS to remove the excess probe. The intracellular fluorescence of the oxidized product (DCF) was immediately visualized by confocal microscopy (Nikon) or quantified by flow cytometry (Beckman Coulter). To specifically measure mitochondrial superoxide, cells were similarly treated and then incubated with 5 μM MitoSOX Red Mitochondrial Superoxide Indicator (Thermo Fisher Scientific) in HBSS (Gibco) for 10 min at 37°C, protected from light. After incubation, cells were washed, and the resulting fluorescence was visualized by confocal microscopy (Nikon). Fluorescence intensity was quantified from the captured images using ImageJ software.

Mitochondrial membrane potential and apoptosis assessment

After the specified treatments under normal or injury-inducing conditions, mitochondrial integrity and apoptosis were evaluated. To assess mitochondrial membrane potential (ΔΨm), cells were incubated with the JC-1 probe (Beyotime) for 20 min at 37°C. After washing, the fluorescence was immediately imaged by confocal microscopy (Nikon). Healthy mitochondria with high ΔΨm exhibit red fluorescent J-aggregates, while apoptotic or damaged mitochondria with low ΔΨm contain green fluorescent JC-1 monomers. Red fluorescence intensity was calculated using ImageJ software to quantify changes in ΔΨm. Apoptosis was analyzed using two independent methods. First, DNA fragmentation in apoptotic cells was detected in situ using a TUNEL staining kit (Yeasen Biotechnology) according to the manufacturer’s instructions. The percentage of TUNEL-positive cells was quantified from fluorescence microscopy images. Second, apoptosis was quantified by flow cytometry using an Annexin V-FITC and propidium iodide (PI) apoptosis detection kit (Yeasen Biotechnology). Briefly, cells were harvested, washed, and resuspended in binding buffer, then co-stained with Annexin V-FITC and PI. The populations of live (Annexin V⁻/PI⁻), early apoptotic (Annexin V⁺/PI⁻), late apoptotic (Annexin V⁺/PI⁺), and necrotic (Annexin V⁻/PI⁺) cells were then quantified by flow cytometry (Beckman Coulter).

Quantitative PCR

Total cellular RNA was extracted using TRIzol reagent (Thermo Fisher Scientifi), and its concentration was measured with a Nanodrop 2000 spectrophotometer. The conversion of RNA to cDNA was facilitated using the RevertAid First Strand cDNA Synthesis Kit (Takara). For the quantification of key angiogenic genes (VEGF, ANG1, bFGF, PDGF), the LightCycler 96 Real-Time PCR System (Roche, IN, USA) was employed in conjunction with SYBR Green assay reagent (Takara). The thermal cycling conditions were as follows: an initial denaturation at 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. Gene expression levels were normalized to the housekeeping gene GAPDH, and relative expression was calculated using the 2−ΔΔCq method. The primers used for qPCR were synthesized by Sangon Biotech (Shanghai, China), and their sequences are listed in Table S1.

Western blotting

Protein extraction was conducted using RIPA lysis buffer (Thermo Fisher Scientific), which includes inhibitors for both protease and phosphatase (Sigma-Aldrich). The concentration of protein was determined through the BCA Protein Assay Kit (Beytime, China). For denaturation, 30 μg of protein was combined with a 5x concentration of SDS-PAGE loading buffer and subjected to boiling for 10 min. The protein underwent separation via a 10% SDS-PAGE gel and was then transferred onto a 0.22 μm polyvinylidene difluoride (PVDF) membrane (Merck Millipore). The membrane was blocked with 5% non-fat milk in Tris-buffered saline with 0.1% Tween 20 (TBST) for 2 h at room temperature. Following blocking, the membrane was incubated with the appropriate primary antibodies overnight at 4°C. After washing three times with TBST, the membrane was incubated with the corresponding HRP-conjugated secondary antibody for 2 h at room temperature. Protein bands were detected using an enhanced chemiluminescence (ECL) kit (Meilunbio) and imaged with the ChemiDoc XRS+ system with Image Lab V3.0 software (Bio-Rad). The specific primary antibodies and HRP-conjugated secondary antibody utilized are itemized in the key resources table.

Transcriptomics study

For the transcriptomics study, total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific). RNA purity and concentration were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific), and RNA integrity was evaluated with an Agilent 2100 Bioanalyzer (Agilent Technologies). Sequencing libraries were generated using a VAHTS Universal V6 RNA-seq Library Prep kit (Vazyme) following the manufacturer’s recommendations. The libraries were sequenced on an Illumina Novaseq 6000 platform, generating 150 bp paired-end reads. Raw reads were processed using fastp software to obtain clean reads, which were then aligned to the reference genome using HISAT2. Gene expression levels were calculated, and DESeq2 was used to identify differentially expressed genes. Genes with a p-value <0.05 and an absolute log2 (fold change) ≥ 1 were considered differentially expressed. Gene Ontology (GO) enrichment analysis and Gene Set Enrichment Analysis (GSEA) were performed to identify enriched biological processes and metabolic pathways. Transcriptome sequencing and analysis were conducted by OE Biotech Co., Ltd. (Shanghai, China).

Animal models of ischemia and therapeutic interventions

Male SD rats and C57BL/6 mice were obtained from the Animal Center of the Chinese Academy of Sciences (Shanghai, China) and housed in a specific-pathogen-free facility with a controlled temperature (18–23°C), humidity (40–60%) and a 12-h light–dark cycle, with food and water freely available. A total of four different animal models (GIONFH, ischemic flap, MI, MCAO) were used to simulate ischemic diseases in various organs.

For glucocorticoid-induced osteonecrosis of the femoral head (GIONFH), the model was established in 8-week-old male SD rats by intramuscular injection of methylprednisolone (MPS; 20 mg/kg/day) for the first 3 days of each week for three consecutive weeks⁷^,⁷⁰; control rats received saline. Rats were randomly divided into six groups: Control, GIONFH+PBS, GIONFH+CeO₂, GIONFH+NMU_iECs, GIONFH+Mix, and GIONFH+CeO₂-NMU_iECs. Therapeutic formulations were administered via the tail vein injection at a dose equivalent to 200 μg of the NMU_iECs component in 200 μL PBS, three times a week for the first three weeks, followed by weekly injections for the subsequent three weeks. After 6 weeks, femoral heads were harvested for analysis by micro-CT (for BV/TV, Tb.N, Tb.Th, Tb.Sp), histology (H&E, Masson, Goldner), IHC (for CD31, COL1, ANG1, OCN), and TUNEL staining.

For the ischemic skin flap model, a 1.5 × 4.5 cm random-pattern flap was created on the dorsum of 6-week-old C57BL/6 mice, and the sacral arteries were resected.⁴⁹ Immediately post-surgery, mice were randomly divided into five groups (PBS, CeO₂, NMU_iECs, Mix, CeO₂-NMU_iECs) and treated with a single subcutaneous injection of the respective formulation (100 μg/mL, 20 μL per site at six sites). On day 7, flap viability was assessed by photography, thermal imaging, and laser Doppler blood flow imaging. Harvested tissues were analyzed by H&E, IHC (CD31), DHE staining, Western blot (CD31, VE-cadherin), TUNEL staining, and immunofluorescence (VEGFA).

For the myocardial infarction (MI) model, 6-week-old male SD rats were anesthetized with isoflurane, intubated, and mechanically ventilated. A left thoracotomy was performed in the fourth intercostal space to expose the heart. The left anterior descending (LAD) artery was permanently ligated 1–2 mm distal to the left atrial appendage using a 6-0 suture.⁷¹ Successful infarction was confirmed by the immediate blanching of the anterior ventricular wall. Sham-operated rats underwent the same procedure without LAD ligation. Rats were then randomly assigned to six groups (Sham, MI + PBS, MI + CeO₂, MI + NMU_iECs, MI + Mix, MI + CeO₂-NMU_iECs) and treated intravenously with formulations equivalent to 200 μg of NMU_iECs. The first injection was given 30 min post-MI, followed by injections three times a week for two weeks, then twice weekly for two weeks. At day 28, cardiac function was assessed again by echocardiography (EF, FS, LVIDs), and hearts were harvested for histology (Masson’s trichrome for infarct size, H&E) and immunofluorescence (CD31, Cx43/α-actinin). At day 7, a separate cohort of hearts was analyzed for oxidative stress (DHE staining) and apoptosis (TUNEL assay) in the border zone. The survival rate of the animals was 80% throughout the experiment.

For the middle cerebral artery occlusion (MCAO) model,⁷² 6-week-old male SD rats were anesthetized with isoflurane, and body temperature was maintained at 37°C. A midline cervical incision was made, and the left common carotid artery (CCA), external carotid artery (ECA), and internal carotid artery (ICA) were carefully isolated. The distal ECA was ligated, and a temporary clip was placed on the ICA. A 4-0 monofilament nylon suture with a silicone-coated tip was introduced through an incision in the ECA stump and advanced into the ICA for approximately 18–20 mm from the carotid bifurcation until a mild resistance was felt, indicating the occlusion of the origin of the MCA. After 1.5 h of occlusion, the filament was withdrawn to allow for reperfusion. Sham-operated rats underwent the same procedure, but the filament was not advanced to occlude the MCA. Rats were randomly divided into six groups (Sham, MCAO+PBS, MCAO+CeO₂, MCAO+NMU_iECs, MCAO+Mix, MCAO+CeO₂-NMU_iECs) and treated intravenously with formulations equivalent to 200 μg of NMU_iECs, starting 30 min after reperfusion and continuing thrice weekly. Neurological function was assessed on days 1, 3, 5, and 7. At day 1, infarct volume was measured by TTC staining and cerebral edema by the wet-dry weight method. At day 7, brains were harvested for histology (Nissl, H&E), IHC (CD31, GFAP), immunofluorescence (NeuN), DHE staining, and TUNEL assay.

For the large animal pre-clinical evaluation, a Bama minipig random-pattern skin flap model was established. Two male Bama minipigs aged approximately 8 months and weighing 20–30 kg were obtained from Jiangsu Changzhou Beiwang Biological Technology Co., Ltd. The animals were housed under standard conditions, fasted for 24 h, and deprived of water for 6 h prior to surgery. Anesthesia was induced via intramuscular injection of Zoletil 50. After standard surgical preparation, three random-pattern skin flaps (each measuring 14 cm × 4 cm) were designed on each side of the dorsum, separated by a 6 cm interval, resulting in a total of 12 flaps across the two pigs. Incisions were made along the design lines, and the subcutaneous tissue was elevated down to the deep fascia. Crucially, axial perforating vessels emerging from the deep tissue were identified, sectioned, and ligated to ensure strict reliance on the random microvascular network for blood supply, thus inducing ischemia. 12 ischemic flaps were randomly assigned to three treatment groups (n = 4 flaps per group): PBS, NMU_iECs, and CeO₂-NMU_iECs. Based on the therapeutic dose validated in the mouse model and adjusted for the expanded tissue volume, formulations were administered via subcutaneous injection at 10 equidistant sites, uniformly distributed in two parallel longitudinal columns (five sites per column) across the flap. A total volume of 3 mL per flap (300 μL per injection site) was delivered. After surgery, the wounds were properly dressed with a custom-made abdominal binder to prevent mechanical injury. On day 7 post-surgery, therapeutic efficacy was evaluated via macroscopic photography and infrared thermal imaging to calculate the survival area and temperature gradients. The animals were then euthanized, and the flap tissues were harvested for comprehensive histological and molecular assessments, including H&E staining, Picrosirius Red staining (to distinguish type I and type III collagen under polarized light), immunohistochemistry (CD31), immunofluorescence (CD34), and in situ ROS detection (DHE staining).

Biodistribution, pharmacokinetics, and targeting studies

The in vivo behavior of the nanovesicles was investigated through a series of pharmacokinetic, biodistribution, and targeting studies. For all fluorescence imaging, NMU_iECs and CeO₂-NMU_iECs were labeled with DiD lipophilic dye (Thermo Fisher Scientific). For pharmacokinetic studies, a single intravenous injection of DiD-labeled NMU_iECs via the tail vein (200 μg in 200 μL PBS) was administered to healthy rats, and major organs (heart, liver, spleen, lung, kidney, bone, and brain) were collected at multiple time points from 1 to 72 h for ex vivo fluorescence imaging using an IVIS Spectrum system (PerkinElmer) to determine the biodistribution over time. For biodistribution and homing studies, the accumulation of DiD-labeled NMU_iECs was compared between normal rats and ischemic model rats (GIONFH, MI, MCAO) at 4 h post-injection. At this time point, major organs (heart, liver, spleen, lung, kidney, bone, and brain) were harvested for ex vivo fluorescence imaging. Additionally, tissues from the ischemic regions (bone, heart, and brain) were embedded in OCT compound, sectioned using a Leica CM1950 cryostat, and imaged by confocal microscopy (Nikon) to visualize vesicle distribution at the tissue level. For the skin flap model, the biodistribution of DiD-labeled NMU_iECs was assessed at 24 h after local subcutaneous injection (100 μg/mL, 20 μL per site at six sites). To investigate the homing mechanism, a CXCR4 blocking experiment was conducted.⁸ DiD-labeled NMU_iECs and CeO₂-NMU_iECs were pre-incubated with an anti-CXCR4 antibody (Abcam) before intravenous injection via the tail vein into normal or ischemic model animals. The biodistribution was then compared to that of unblocked nanovesicles at 4 h post-injection using ex vivo IVIS imaging (PerkinElmer). To identify the target cells within the ischemic tissue, a cellular targeting analysis was performed. Ischemic tissues from animals injected with DiD-labeled NMU_iECs were harvested, cryosectioned, and co-stained with an anti-CD31 antibody (green) to label endothelial cells. The co-localization of DiD-labeled NMU_iECs (red) with CD31-positive cells was then examined by confocal microscopy (Nikon).

Micro-computed tomography (Micro-CT) analysis

At the end of the treatment period for the GIONFH model, rats were euthanized, and the femoral heads were harvested and fixed in 4% paraformaldehyde. The samples were then scanned using a micro-CT system (Skyscan, Bruker). Three-dimensional images were reconstructed, and the trabecular bone region of interest was analyzed to determine bone morphometric parameters, including bone volume/total volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp).

Skin flap viability assessment

On day 7 post-surgery in the skin flap model, flap viability was assessed using multiple methods. The flap survival area was determined by digital photography, and the percentage of the necrotic area relative to the total flap area was calculated using ImageJ software. Infrared thermal imaging was performed using a thermal camera (FLIR One Pro) to assess tissue temperature as an indicator of perfusion. Blood flow was measured using a laser Doppler blood flow (LDBF) imaging system (Moor Instruments), which provided a quantitative map of blood perfusion across the flap surface.

Echocardiography

In the MI model, cardiac function was evaluated by transthoracic echocardiography at day 28. Rats were anesthetized with isoflurane (3% for induction, 1.5–2% for maintenance) and placed in the left lateral decubitus position. Body temperature was maintained at 37°C, and heart rate was monitored via ECG to ensure stability (approximately 350–400 bpm). Two-dimensional M-mode images were acquired from the parasternal long-axis view using a high-frequency ultrasound system equipped with a linear array transducer. From these images, left ventricular internal dimensions at end-diastole (LVIDd) and end-systole (LVIDs) were measured. These measurements were used to calculate the ejection fraction (EF) and fractional shortening (FS) to assess global systolic function.

TTC staining

At 24 h post-MCAO, rats were euthanized and the brains were collected. The brains were sectioned into 2-mm-thick coronal slices using a brain matrix (Servicebio). The slices were then immersed in a 2% solution of 2,3,5-triphenyltetrazolium chloride (TTC, Servicebio) in saline at 37°C for 30 min in the dark. After staining, the slices were fixed in 4% paraformaldehyde. The infarct area (pale tissue) was quantified from digital images of the stained sections using ImageJ software, and the total infarct volume was calculated.

Histology, immunohistochemistry, and immunofluorescence

Harvested tissues from all models were fixed in 4% PFA, processed, and embedded in either paraffin or OCT compound. Paraffin-embedded tissues were sectioned at 4–5 μm using a microtome, while OCT-embedded tissues were sectioned at 10 μm using a cryostat (Leica CM1950). For histological analysis, sections were stained with hematoxylin and eosin (H&E) for general morphology, Masson’s trichrome for fibrosis, Goldner’s trichrome for bone mineralization, and Nissl stain for neuronal survival. All staining kits and solutions were obtained from Servicebio (China) and used according to the manufacturer’s instructions. For immunohistochemistry (IHC) and immunofluorescence (IF), sections underwent antigen retrieval followed by blocking with goat serum. Sections were then incubated with primary antibodies overnight at 4°C, followed by incubation with either HRP-conjugated secondary antibodies and DAB for IHC or fluorescently-labeled secondary antibodies for IF. Key markers analyzed included CD31 (vessel density), Collagen I (osteogenesis), GFAP (astrogliosis), and NeuN (neurons). The specific primary antibodies and HRP-conjugated secondary antibody utilized are itemized in the key resources table. For in situ detection of ROS and apoptosis, frozen sections were incubated with dihydroethidium (DHE) or processed using a TUNEL assay kit (Servicebio), respectively, according to the manufacturer’s instructions. Quantification was performed by measuring fluorescence intensity or counting positive cells using ImageJ software.

In vivo biosafety and immunogenicity

Long-term systemic toxicity was evaluated in multiple animal models. In healthy rodent models, C57BL/6 mice (50 μg/mouse) and SD rats (200 μg/rat) received intravenous injections of PBS or CeO2-NMU_iECs via the tail vein three times a week for six weeks. Systemic toxicity was assessed through hematological analysis, serum biochemical tests, and H&E histopathological examination of major organs (heart, liver, spleen, lungs, kidneys, brain, and bone marrow). Hematological analysis was performed using an automated hematology analyzer to measure a complete blood count, including white blood cell (WBC), red blood cell (RBC), hemoglobin (HGB), hematocrit (HCT), and platelet (PLT) counts. Serum biochemical tests were conducted using an automated biochemistry analyzer (LW C400) to assess liver function (aspartate aminotransferase, AST; alanine aminotransferase, ALT), kidney function (creatinine, CR; blood urea nitrogen, BUN; uric acid, UA), and metabolic status (triglycerides, TG; total cholesterol, TC; albumin, ALB; globulin, GLB). In a large animal model, a preliminary two-week safety study was conducted in Bama minipigs (5 mg/pig, ear margin vein injection, three times a week), monitoring tolerance, pathological changes, and hematological parameters. Additionally, a four-week dose-escalation study was performed in mice, testing doses up to five times the therapeutic dose (250 μg/mouse) to determine the therapeutic window. The immunogenic potential of CeO₂-NMU_iECs was assessed at both systemic and local levels in healthy C57BL/6 mice. For systemic immunogenicity, mice received a single intravenous injection of either PBS or CeO₂-NMU_iECs via the tail vein (50 μg in 50 μL PBS). Blood was collected at 24, 48, and 72 h post-injection for cytokine analysis. Serum levels of key pro-inflammatory cytokines (IL-1β, IL-6, TNF-α, IFN-γ) were measured using commercial ELISA kits (Peprotech) according to the manufacturer’s instructions. To analyze the adaptive immune response, immune organs (bone marrow, thymus, and spleen) were harvested at 72 h post-injection. Single-cell suspensions were prepared for flow cytometry. Spleen and thymus tissues were mechanically dissociated and passed through a 70-μm cell strainer. Bone marrow cells were flushed from the femurs and tibias. Red blood cells in spleen and bone marrow samples were lysed using an RBC lysis buffer (Sigma-Aldrich). For T cell analysis, cells were stained with antibodies against CD45, CD3, CD4, and CD8. For B cell analysis, a separate panel of antibodies against CD45, CD3, and CD19 was used. Flow cytometry was performed on a cytometer (Beckman Coulter CytoFlex), and the data were analyzed using FlowJo software. The detailed gating strategies for identifying T cell subsets (CD3⁺ within CD45⁺; CD4⁺ and CD8⁺ within CD3⁺) and B cells (CD19⁺ within CD3⁻) are provided in Figure S22. For local immunogenicity, healthy mice received a single injection of PBS or CeO₂-NMU_iECs (100 μg/mL in 20 μL PBS) via either subcutaneous or intramuscular routes. At 24 and 72 h post-injection, the skin or muscle tissue at the injection site was harvested. One portion of the tissue was fixed in 4% paraformaldehyde for histological analysis, including H&E staining and immunohistochemistry (IHC) for the macrophage marker F4/80. Another portion of the tissue was homogenized in lysis buffer, and the supernatant was collected for the measurement of local cytokine levels (IL-1β, IL-6, TNF-α, IFN-γ) by ELISA (Peprotech).

Quantification and statistical analysis

All quantitative data are expressed as the mean ± standard deviation (s.d.). Statistical analysis was performed using GraphPad Prism (version 9.0). Comparisons between two groups were made using an unpaired, two-tailed Student’s t test. Comparisons among three or more groups were performed using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test for multiple comparisons. A p-value of less than 0.05 was considered statistically significant. In the figures, p values are provided as exact values whenever available; when p values are smaller than 0.0001, p < 0.0001 is shown. The exact sample size (n), representing the number of biologically independent samples for each experiment, is indicated in the corresponding figure legends.

Published: August 5, 2026

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.xcrm.2026.102972.

Contributor Information

Jian Lin, Email: linj0208@163.com.

Junjie Deng, Email: j.deng@ucas.ac.cn.

Liang Chen, Email: breakingsunshine@wmu.edu.cn.

Xiaoyun Pan, Email: xiaoyunpan@wmu.edu.cn.

Supplemental information

Document S1. Figures S1–S27 and Table S1

mmc1.pdf^{(71.2MB, pdf)}

Document S2. Article plus supplemental information

mmc2.pdf^{(174.2MB, pdf)}

References

1.Augustin H.G., Koh G.Y. A systems view of the vascular endothelium in health and disease. Cell. 2024;187:4833–4858. doi: 10.1016/j.cell.2024.07.012. [DOI] [PubMed] [Google Scholar]
2.Barnett S.N., Cujba A.M., Yang L., Maceiras A.R., Li S., Kedlian V.R., Pett J.P., Polanski K., Miranda A.M.A., Xu C., et al. An organotypic atlas of human vascular cells. Nat. Med. 2024;30:3468–3481. doi: 10.1038/s41591-024-03376-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Trimm E., Red-Horse K. Vascular endothelial cell development and diversity. Nat. Rev. Cardiol. 2023;20:197–210. doi: 10.1038/s41569-022-00770-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Zhang M., Liu Q., Meng H., Duan H., Liu X., Wu J., Gao F., Wang S., Tan R., Yuan J. Ischemia-reperfusion injury: molecular mechanisms and therapeutic targets. Signal Transduct. Target. Ther. 2024;9:12. doi: 10.1038/s41392-023-01688-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Fantini D.A., Yang G., Khanna A., Subramanian D., Phillippi J.A., Huang N.F. Overcoming big bottlenecks in vascular regeneration. Commun. Biol. 2024;7:876. doi: 10.1038/s42003-024-06567-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Pasut A., Lama E., Van Craenenbroeck A.H., Kroon J., Carmeliet P. Endothelial cell metabolism in cardiovascular physiology and disease. Nat. Rev. Cardiol. 2025;22:923–943. doi: 10.1038/s41569-025-01162-x. [DOI] [PubMed] [Google Scholar]
7.Jiang H., Zhu X., Yu J., Wang W., Mao Y., Jiang L., Zhu L., Shen H., Lou C., Lin C., et al. Biomimetic Extracellular Vesicles Based on Composite Bioactive Ions for the Treatment of Ischemic Bone Disease. ACS Nano. 2024;18:34924–34948. doi: 10.1021/acsnano.4c13028. [DOI] [PubMed] [Google Scholar]
8.Cui Y., Li Z., Guo Y., Qi X., Yang Y., Jia X., Li R., Shi J., Gao W., Ren Z., et al. Bioinspired Nanovesicles Convert the Skeletal Endothelium-Associated Secretory Phenotype to Treat Osteoporosis. ACS Nano. 2022;16:11076–11091. doi: 10.1021/acsnano.2c03781. [DOI] [PubMed] [Google Scholar]
9.Choi W., Park D.J., Eliceiri B.P. Defining tropism and activity of natural and engineered extracellular vesicles. Front. Immunol. 2024;15 doi: 10.3389/fimmu.2024.1363185. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Adnani L., Kassouf J., Meehan B., Spinelli C., Tawil N., Nakano I., Rak J. Angiocrine extracellular vesicles impose mesenchymal reprogramming upon proneural glioma stem cells. Nat. Commun. 2022;13:5494. doi: 10.1038/s41467-022-33235-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Babaei M., Rezaie J. Application of stem cell-derived exosomes in ischemic diseases: opportunity and limitations. J. Transl. Med. 2021;19:196. doi: 10.1186/s12967-021-02863-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Xie Y., Deng T., Xie L., Xie Y., Ma J., Zhong D., Huang X., Li Y. Effects of extracellular vesicles for ischemic stroke: A meta-analysis of preclinical studies. Exp. Ther. Med. 2024;28:287. doi: 10.3892/etm.2024.12575. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Palanisamy C.P., Pei J., Alugoju P., Anthikapalli N.V.A., Jayaraman S., Veeraraghavan V.P., Gopathy S., Roy J.R., Janaki C.S., Thalamati D., et al. New strategies of neurodegenerative disease treatment with extracellular vesicles (EVs) derived from mesenchymal stem cells (MSCs) Theranostics. 2023;13:4138–4165. doi: 10.7150/thno.83066. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Alpeeva E.V., Ryabchenko A.S., Vorotelyak E.A. Characterization, Preconditioning, Safety, and Other Issues of MSC-Derived EVs and Secretome. Int. J. Mol. Sci. 2026;27:1688. doi: 10.3390/ijms27041688. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Hou Z., Brenner J.S. Developing targeted antioxidant nanomedicines for ischemic penumbra: Novel strategies in treating brain ischemia-reperfusion injury. Redox Biol. 2024;73 doi: 10.1016/j.redox.2024.103185. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Mu J., Li C., Shi Y., Liu G., Zou J., Zhang D.Y., Jiang C., Wang X., He L., Huang P., et al. Protective effect of platinum nano-antioxidant and nitric oxide against hepatic ischemia-reperfusion injury. Nat. Commun. 2022;13:2513. doi: 10.1038/s41467-022-29772-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Kumar M.A., Baba S.K., Sadida H.Q., Marzooqi S.A., Jerobin J., Altemani F.H., Algehainy N., Alanazi M.A., Abou-Samra A.B., Kumar R., et al. Extracellular vesicles as tools and targets in therapy for diseases. Signal Transduct. Target. Ther. 2024;9:27. doi: 10.1038/s41392-024-01735-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.He L., Huang G., Liu H., Sang C., Liu X., Chen T. Highly bioactive zeolitic imidazolate framework-8-capped nanotherapeutics for efficient reversal of reperfusion-induced injury in ischemic stroke. Sci. Adv. 2020;6 doi: 10.1126/sciadv.aay9751. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Briones-Valdivieso C., Briones F., Orellana-Urzúa S., Chichiarelli S., Saso L., Rodrigo R. Novel Multi-Antioxidant Approach for Ischemic Stroke Therapy Targeting the Role of Oxidative Stress. Biomedicines. 2024;12 doi: 10.3390/biomedicines12030501. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Bai Y., Li Y., Li Y., Tian L. Advanced Biological Applications of Cerium Oxide Nanozymes in Disease Related to Oxidative Damage. ACS Omega. 2024;9:8601–8614. doi: 10.1021/acsomega.3c03661. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Koo S., Sohn H.S., Kim T.H., Yang S., Jang S.Y., Ye S., Choi B., Kim S.H., Park K.S., Shin H.M., et al. Ceria-vesicle nanohybrid therapeutic for modulation of innate and adaptive immunity in a collagen-induced arthritis model. Nat. Nanotechnol. 2023;18:1502–1514. doi: 10.1038/s41565-023-01523-y. [DOI] [PubMed] [Google Scholar]
22.Orlova V.V., van den Hil F.E., Petrus-Reurer S., Drabsch Y., Ten Dijke P., Mummery C.L. Generation, expansion and functional analysis of endothelial cells and pericytes derived from human pluripotent stem cells. Nat. Protoc. 2014;9:1514–1531. doi: 10.1038/nprot.2014.102. [DOI] [PubMed] [Google Scholar]
23.Lian X., Bao X., Al-Ahmad A., Liu J., Wu Y., Dong W., Dunn K.K., Shusta E.V., Palecek S.P. Efficient differentiation of human pluripotent stem cells to endothelial progenitors via small-molecule activation of WNT signaling. Stem Cell Rep. 2014;3:804–816. doi: 10.1016/j.stemcr.2014.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Lian Q., Zhang Y., Zhang J., Zhang H.K., Wu X., Zhang Y., Lam F.F.Y., Kang S., Xia J.C., Lai W.H., et al. Functional mesenchymal stem cells derived from human induced pluripotent stem cells attenuate limb ischemia in mice. Circulation. 2010;121:1113–1123. doi: 10.1161/circulationaha.109.898312. [DOI] [PubMed] [Google Scholar]
25.Zhao Q., Gregory C.A., Lee R.H., Reger R.L., Qin L., Hai B., Park M.S., Yoon N., Clough B., McNeill E., et al. MSCs derived from iPSCs with a modified protocol are tumor-tropic but have much less potential to promote tumors than bone marrow MSCs. Proc. Natl. Acad. Sci. USA. 2015;112:530–535. doi: 10.1073/pnas.1423008112. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Chambers S.M., Fasano C.A., Papapetrou E.P., Tomishima M., Sadelain M., Studer L. Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat. Biotechnol. 2009;27:275–280. doi: 10.1038/nbt.1529. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Fan J., Lee C.S., Kim S., Chen C., Aghaloo T., Lee M. Generation of Small RNA-Modulated Exosome Mimetics for Bone Regeneration. ACS Nano. 2020;14:11973–11984. doi: 10.1021/acsnano.0c05122. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Jiang H., Lou C., Jiang L., Lin C., Wang W., Yan Z., Yu J., Cai T., Lin S., Wang J., et al. Simvastatin-enhanced bioinspired exosome mimetics regulate osteogenesis and angiogenesis for the treatment of glucocorticoid-induced osteonecrosis of the femoral head. Chem. Eng. J. 2023;472 doi: 10.1016/j.cej.2023.144729. [DOI] [Google Scholar]
29.Kou M., Huang L., Yang J., Chiang Z., Chen S., Liu J., Guo L., Zhang X., Zhou X., Xu X., et al. Mesenchymal stem cell-derived extracellular vesicles for immunomodulation and regeneration: a next generation therapeutic tool? Cell Death Dis. 2022;13:580. doi: 10.1038/s41419-022-05034-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Liu Z.-L., Chen H.-H., Zheng L.-L., Sun L.-P., Shi L. Angiogenic signaling pathways and anti-angiogenic therapy for cancer. Signal Transduct. Target. Ther. 2023;8:198. doi: 10.1038/s41392-023-01460-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.de Kreuk B.J., Gingras A.R., Knight J.D., Liu J.J., Gingras A.C., Ginsberg M.H. Heart of glass anchors Rasip1 at endothelial cell-cell junctions to support vascular integrity. eLife. 2016;5 doi: 10.7554/eLife.11394. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Benwell C.J., Johnson R.T., Taylor J.A.G.E., Lambert J., Robinson S.D. A proteomics approach to isolating neuropilin-dependent α5 integrin trafficking pathways: neuropilin 1 and 2 co-traffic α5 integrin through endosomal p120RasGAP to promote polarised fibronectin fibrillogenesis in endothelial cells. Commun. Biol. 2024;7:629. doi: 10.1038/s42003-024-06320-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Ji W., Li Y., Peng H., Zhao R., Shen J., Wu Y., Wang J., Hao Q., Lu Z., Yang J., Zhang X. Self-Catalytic Small Interfering RNA Nanocarriers for Synergistic Treatment of Neurodegenerative Diseases. Adv. Mater. 2022;34 doi: 10.1002/adma.202105711. [DOI] [PubMed] [Google Scholar]
34.Kim Y.G., Lee Y., Lee N., Soh M., Kim D., Hyeon T. Ceria-Based Therapeutic Antioxidants for Biomedical Applications. Adv. Mater. 2024;36 doi: 10.1002/adma.202210819. [DOI] [PubMed] [Google Scholar]
35.Jiang H., Xia W., Xia T., Jiang L., Yu J., Zhu X., Lin C., Lou C., Wang W., Chai Y., et al. Chemotactic recruitment of genetically engineered cell membrane-camouflaged metal-organic framework nanoparticles for ischemic osteonecrosis treatment. Acta Biomater. 2024;185:410–428. doi: 10.1016/j.actbio.2024.07.024. [DOI] [PubMed] [Google Scholar]
36.Ma J., Zhang S., Liu J., Liu F., Du F., Li M., Chen A.T., Bao Y., Suh H.W., Avery J., et al. Targeted Drug Delivery to Stroke via Chemotactic Recruitment of Nanoparticles Coated with Membrane of Engineered Neural Stem Cells. Small. 2019;15 doi: 10.1002/smll.201902011. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Bose R.J., Kim B.J., Arai Y., Han I.B., Moon J.J., Paulmurugan R., Park H., Lee S.H. Bioengineered stem cell membrane functionalized nanocarriers for therapeutic targeting of severe hindlimb ischemia. Biomaterials. 2018;185:360–370. doi: 10.1016/j.biomaterials.2018.08.018. [DOI] [PubMed] [Google Scholar]
38.Luo L., Zang G., Liu B., Qin X., Zhang Y., Chen Y., Zhang H., Wu W., Wang G. Bioengineering CXCR4-overexpressing cell membrane functionalized ROS-responsive nanotherapeutics for targeting cerebral ischemia-reperfusion injury. Theranostics. 2021;11:8043–8056. doi: 10.7150/thno.60785. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Qiao L., Hu S., Huang K., Su T., Li Z., Vandergriff A., Cores J., Dinh P.U., Allen T., Shen D., et al. Tumor cell-derived exosomes home to their cells of origin and can be used as Trojan horses to deliver cancer drugs. Theranostics. 2020;10:3474–3487. doi: 10.7150/thno.39434. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Chen P., Liu X., Gu C., Zhong P., Song N., Li M., Dai Z., Fang X., Liu Z., Zhang J., et al. A plant-derived natural photosynthetic system for improving cell anabolism. Nature. 2022;612:546–554. doi: 10.1038/s41586-022-05499-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Zheng Z., Wang M., Cheng C., Liu D., Wu L., Zhu J., Qian X. Ginsenoside Rb1 reduces H2O2-induced HUVEC dysfunction by stimulating the sirtuin-1/AMP-activated protein kinase pathway. Mol. Med. Rep. 2020;22:247–256. doi: 10.3892/mmr.2020.11096. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Chen C.Y., Rao S.S., Yue T., Tan Y.J., Yin H., Chen L.J., Luo M.J., Wang Z., Wang Y.Y., Hong C.G., et al. Glucocorticoid-induced loss of beneficial gut bacterial extracellular vesicles is associated with the pathogenesis of osteonecrosis. Sci. Adv. 2022;8 doi: 10.1126/sciadv.abg8335. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Yu B., Li H., Zhang Z., Chen P., Wang L., Fan X., Ning X., Pan Y., Zhou F., Hu X., et al. Extracellular vesicles engineering by silicates-activated endothelial progenitor cells for myocardial infarction treatment in male mice. Nat. Commun. 2023;14 doi: 10.1038/s41467-023-37832-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Martin M., Vermeiren S., Bostaille N., Eubelen M., Spitzer D., Vermeersch M., Profaci C.P., Pozuelo E., Toussay X., Raman-Nair J., et al. Engineered Wnt ligands enable blood-brain barrier repair in neurological disorders. Science (New York, N.Y.) 2022;375 doi: 10.1126/science.abm4459. [DOI] [PubMed] [Google Scholar]
45.Rohde D., Vandoorne K., Lee I.H., Grune J., Zhang S., McAlpine C.S., Schloss M.J., Nayar R., Courties G., Frodermann V., et al. Bone marrow endothelial dysfunction promotes myeloid cell expansion in cardiovascular disease. Nat. Cardiovasc. Res. 2022;1:28–44. doi: 10.1038/s44161-021-00002-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Mehta V., Pang K.L., Givens C.S., Chen Z., Huang J., Sweet D.T., Jo H., Reader J.S., Tzima E. Mechanical forces regulate endothelial-to-mesenchymal transition and atherosclerosis via an Alk5-Shc mechanotransduction pathway. Sci. Adv. 2021;7 doi: 10.1126/sciadv.abg5060. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Augustin H.G., Koh G.Y. Organotypic vasculature: From descriptive heterogeneity to functional pathophysiology. Science. 2017;357 doi: 10.1126/science.aal2379. [DOI] [PubMed] [Google Scholar]
48.Ji B.S., Cen J., He L., Liu M., Liu Y.Q., Liu L. Modulation of P-glycoprotein in rat brain microvessel endothelial cells under oxygen glucose deprivation. J. Pharm. Pharmacol. 2013;65:1508–1517. doi: 10.1111/jphp.12122. [DOI] [PubMed] [Google Scholar]
49.Yu G., Chen Y., Yang N., Zhang H., Zhang X., Geng Y., Zhao J., Chen Z., Dong C., Lin L., et al. Apoptotic Bodies Derived from Fibroblast-Like Cells in Subcutaneous Connective Tissue Inhibit Ferroptosis in Ischaemic Flaps via the miR-339-5p/KEAP1/Nrf2 Axis. Adv. Sci. 2024;11 doi: 10.1002/advs.202307238. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Baaten C.C.F.M.J., Vondenhoff S., Noels H. Endothelial Cell Dysfunction and Increased Cardiovascular Risk in Patients With Chronic Kidney Disease. Circ. Res. 2023;132:970–992. doi: 10.1161/circresaha.123.321752. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.You Y., Tian Y., Guo R., Shi J., Kwak K.J., Tong Y., Estania A.P., Hsu W.H., Liu Y., Hu S., et al. Extracellular vesicle-mediated VEGF-A mRNA delivery rescues ischaemic injury with low immunogenicity. Eur. Heart J. 2025;46:1662–1676. doi: 10.1093/eurheartj/ehae883. [DOI] [PubMed] [Google Scholar]
52.Hu G.W., Li Q., Niu X., Hu B., Liu J., Zhou S.m., Guo S.c., Lang H.l., Zhang C.q., Wang Y., Deng Z.f. Exosomes secreted by human-induced pluripotent stem cell-derived mesenchymal stem cells attenuate limb ischemia by promoting angiogenesis in mice. Stem Cell Res. Ther. 2015;6:10. doi: 10.1186/scrt546. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Zhang J., Guan J., Niu X., Hu G., Guo S., Li Q., Xie Z., Zhang C., Wang Y. Exosomes released from human induced pluripotent stem cells-derived MSCs facilitate cutaneous wound healing by promoting collagen synthesis and angiogenesis. J. Transl. Med. 2015;13:49. doi: 10.1186/s12967-015-0417-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Xia Y., Ling X., Hu G., Zhu Q., Zhang J., Li Q., Zhao B., Wang Y., Deng Z. Small extracellular vesicles secreted by human iPSC-derived MSC enhance angiogenesis through inhibiting STAT3-dependent autophagy in ischemic stroke. Stem Cell Res. Ther. 2020;11:313. doi: 10.1186/s13287-020-01834-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Ye M., Ni Q., Qi H., Qian X., Chen J., Guo X., Li M., Zhao Y., Xue G., Deng H., Zhang L. Exosomes Derived from Human Induced Pluripotent Stem Cells-Endothelia Cells Promotes Postnatal Angiogenesis in Mice Bearing Ischemic Limbs. Int. J. Biol. Sci. 2019;15:158–168. doi: 10.7150/ijbs.28392. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Wang X., Chen Y., Zhao Z., Meng Q., Yu Y., Sun J., Yang Z., Chen Y., Li J., Ma T., et al. Engineered Exosomes With Ischemic Myocardium-Targeting Peptide for Targeted Therapy in Myocardial Infarction. J. Am. Heart Assoc. 2018;7 doi: 10.1161/jaha.118.008737. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Chen W.C.W., Lee B.G., Park D.W., Kim K., Chu H., Kim K., Huard J., Wang Y. Controlled dual delivery of fibroblast growth factor-2 and Interleukin-10 by heparin-based coacervate synergistically enhances ischemic heart repair. Biomaterials. 2015;72:138–151. doi: 10.1016/j.biomaterials.2015.08.050. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Ruozi G., Bortolotti F., Mura A., Tomczyk M., Falcione A., Martinelli V., Vodret S., Braga L., Dal Ferro M., Cannatà A., et al. Cardioprotective factors against myocardial infarction selected in vivo from an AAV secretome library. Sci. Transl. Med. 2022;14 doi: 10.1126/scitranslmed.abo0699. [DOI] [PubMed] [Google Scholar]
59.Mandai M., Watanabe A., Kurimoto Y., Hirami Y., Morinaga C., Daimon T., Fujihara M., Akimaru H., Sakai N., Shibata Y., et al. Autologous Induced Stem-Cell-Derived Retinal Cells for Macular Degeneration. N. Engl. J. Med. 2017;376:1038–1046. doi: 10.1056/NEJMoa1608368. [DOI] [PubMed] [Google Scholar]
60.Schweitzer J.S., Song B., Herrington T.M., Park T.Y., Lee N., Ko S., Jeon J., Cha Y., Kim K., Li Q., et al. Personalized iPSC-Derived Dopamine Progenitor Cells for Parkinson's Disease. N. Engl. J. Med. 2020;382:1926–1932. doi: 10.1056/NEJMoa1915872. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Deuse T., Hu X., Gravina A., Wang D., Tediashvili G., De C., Thayer W.O., Wahl A., Garcia J.V., Reichenspurner H., et al. Hypoimmunogenic derivatives of induced pluripotent stem cells evade immune rejection in fully immunocompetent allogeneic recipients. Nat. Biotechnol. 2019;37:252–258. doi: 10.1038/s41587-019-0016-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Hu X., White K., Olroyd A.G., DeJesus R., Dominguez A.A., Dowdle W.E., Friera A.M., Young C., Wells F., Chu E.Y., et al. Hypoimmune induced pluripotent stem cells survive long term in fully immunocompetent, allogeneic rhesus macaques. Nat. Biotechnol. 2024;42:413–423. doi: 10.1038/s41587-023-01784-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Thakur A., Rai D. Global requirements for manufacturing and validation of clinical grade extracellular vesicles. J. Liq. Biopsy. 2024;6 doi: 10.1016/j.jlb.2024.100278. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Valadi H., Ekström K., Bossios A., Sjöstrand M., Lee J.J., Lötvall J.O. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat. Cell Biol. 2007;9:654–659. doi: 10.1038/ncb1596. [DOI] [PubMed] [Google Scholar]
65.Lin R.Z., Im G.B., Luo A.C., Zhu Y., Hong X., Neumeyer J., Tang H.W., Perrimon N., Melero-Martin J.M. Mitochondrial transfer mediates endothelial cell engraftment through mitophagy. Nature. 2024;629:660–668. doi: 10.1038/s41586-024-07340-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Jiang H., Wang W., Mao Y., Jiang L., Yu J., Zhu X., Fu H., Lin Z., Shen H., Pan X., Xue X. Morroniside-mediated mitigation of stem cell and endothelial cell dysfunction for the therapy of glucocorticoid-induced osteonecrosis of the femoral head. Int. Immunopharmacol. 2024;127 doi: 10.1016/j.intimp.2023.111421. [DOI] [PubMed] [Google Scholar]
67.Jiang H., Yu J., Yan Z., Lin Z., Lin M., Mao Y., Hong Z., Lin J., Xue X., Pan X. Pharmacological activation of the Nrf2 pathway by Taxifolin remodels articular cartilage microenvironment for the therapy of Osteoarthritis. Int. Immunopharmacol. 2023;122 doi: 10.1016/j.intimp.2023.110587. [DOI] [PubMed] [Google Scholar]
68.Wu O., Jin Y., Zhang Z., Zhou H., Xu W., Chen L., Jones M., Kwan K.Y.H., Gao J., Zhang K., et al. KMT2A regulates the autophagy-GATA4 axis through METTL3-mediated m(6)A modification of ATG4a to promote NPCs senescence and IVDD progression. Bone Res. 2024;12:67. doi: 10.1038/s41413-024-00373-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Deng R., Zhao R., Zhang Z., Chen Y., Yang M., Lin Y., Ye J., Li N., Qin H., Yan X., et al. Chondrocyte membrane-coated nanoparticles promote drug retention and halt cartilage damage in rat and canine osteoarthritis. Sci. Transl. Med. 2024;16 doi: 10.1126/scitranslmed.adh9751. [DOI] [PubMed] [Google Scholar]
70.Jiang H., Lin C., Cai T., Jiang L., Lou C., Lin S., Wang W., Yan Z., Pan X., Xue X. Taxifolin-mediated Nrf2 activation ameliorates oxidative stress and apoptosis for the treatment of glucocorticoid-induced osteonecrosis of the femoral head. Phytother Res. 2024;38:156–173. doi: 10.1002/ptr.8031. [DOI] [PubMed] [Google Scholar]
71.Liu W., Zhao N., Yin Q., Zhao X., Guo K., Xian Y., Li S., Wang C., Zhu M., Du Y., et al. Injectable Hydrogels Encapsulating Dual-Functional Au@Pt Core-Shell Nanoparticles Regulate Infarcted Microenvironments and Enhance the Therapeutic Efficacy of Stem Cells through Antioxidant and Electrical Integration. ACS Nano. 2023;17:2053–2066. doi: 10.1021/acsnano.2c07436. [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Tang L., Yin Y., Liu H., Zhu M., Cao Y., Feng J., Fu C., Li Z., Shu W., Gao J., et al. Blood-Brain Barrier-Penetrating and Lesion-Targeting Nanoplatforms Inspired by the Pathophysiological Features for Synergistic Ischemic Stroke Therapy. Adv. Mater. 2024;36 doi: 10.1002/adma.202312897. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S27 and Table S1

mmc1.pdf^{(71.2MB, pdf)}

Document S2. Article plus supplemental information

mmc2.pdf^{(174.2MB, pdf)}

Data Availability Statement

•
Data: The RNA-seq data generated in this study have been deposited in the Genome Sequence Archive (GSA) under accession numbers HRA019213, HRA019214, and HRA019215. The proteomics data generated in this study have been deposited to the ProteomeXchange Consortium via the iProX partner repository with the dataset identifier PXD080200.
•
Code: This paper does not report original code.
•
General statement: Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.

[bib1] 1.Augustin H.G., Koh G.Y. A systems view of the vascular endothelium in health and disease. Cell. 2024;187:4833–4858. doi: 10.1016/j.cell.2024.07.012. [DOI] [PubMed] [Google Scholar]

[bib2] 2.Barnett S.N., Cujba A.M., Yang L., Maceiras A.R., Li S., Kedlian V.R., Pett J.P., Polanski K., Miranda A.M.A., Xu C., et al. An organotypic atlas of human vascular cells. Nat. Med. 2024;30:3468–3481. doi: 10.1038/s41591-024-03376-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] 3.Trimm E., Red-Horse K. Vascular endothelial cell development and diversity. Nat. Rev. Cardiol. 2023;20:197–210. doi: 10.1038/s41569-022-00770-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Zhang M., Liu Q., Meng H., Duan H., Liu X., Wu J., Gao F., Wang S., Tan R., Yuan J. Ischemia-reperfusion injury: molecular mechanisms and therapeutic targets. Signal Transduct. Target. Ther. 2024;9:12. doi: 10.1038/s41392-023-01688-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Fantini D.A., Yang G., Khanna A., Subramanian D., Phillippi J.A., Huang N.F. Overcoming big bottlenecks in vascular regeneration. Commun. Biol. 2024;7:876. doi: 10.1038/s42003-024-06567-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 6.Pasut A., Lama E., Van Craenenbroeck A.H., Kroon J., Carmeliet P. Endothelial cell metabolism in cardiovascular physiology and disease. Nat. Rev. Cardiol. 2025;22:923–943. doi: 10.1038/s41569-025-01162-x. [DOI] [PubMed] [Google Scholar]

[bib7] 7.Jiang H., Zhu X., Yu J., Wang W., Mao Y., Jiang L., Zhu L., Shen H., Lou C., Lin C., et al. Biomimetic Extracellular Vesicles Based on Composite Bioactive Ions for the Treatment of Ischemic Bone Disease. ACS Nano. 2024;18:34924–34948. doi: 10.1021/acsnano.4c13028. [DOI] [PubMed] [Google Scholar]

[bib8] 8.Cui Y., Li Z., Guo Y., Qi X., Yang Y., Jia X., Li R., Shi J., Gao W., Ren Z., et al. Bioinspired Nanovesicles Convert the Skeletal Endothelium-Associated Secretory Phenotype to Treat Osteoporosis. ACS Nano. 2022;16:11076–11091. doi: 10.1021/acsnano.2c03781. [DOI] [PubMed] [Google Scholar]

[bib9] 9.Choi W., Park D.J., Eliceiri B.P. Defining tropism and activity of natural and engineered extracellular vesicles. Front. Immunol. 2024;15 doi: 10.3389/fimmu.2024.1363185. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] 10.Adnani L., Kassouf J., Meehan B., Spinelli C., Tawil N., Nakano I., Rak J. Angiocrine extracellular vesicles impose mesenchymal reprogramming upon proneural glioma stem cells. Nat. Commun. 2022;13:5494. doi: 10.1038/s41467-022-33235-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11.Babaei M., Rezaie J. Application of stem cell-derived exosomes in ischemic diseases: opportunity and limitations. J. Transl. Med. 2021;19:196. doi: 10.1186/s12967-021-02863-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Xie Y., Deng T., Xie L., Xie Y., Ma J., Zhong D., Huang X., Li Y. Effects of extracellular vesicles for ischemic stroke: A meta-analysis of preclinical studies. Exp. Ther. Med. 2024;28:287. doi: 10.3892/etm.2024.12575. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] 13.Palanisamy C.P., Pei J., Alugoju P., Anthikapalli N.V.A., Jayaraman S., Veeraraghavan V.P., Gopathy S., Roy J.R., Janaki C.S., Thalamati D., et al. New strategies of neurodegenerative disease treatment with extracellular vesicles (EVs) derived from mesenchymal stem cells (MSCs) Theranostics. 2023;13:4138–4165. doi: 10.7150/thno.83066. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Alpeeva E.V., Ryabchenko A.S., Vorotelyak E.A. Characterization, Preconditioning, Safety, and Other Issues of MSC-Derived EVs and Secretome. Int. J. Mol. Sci. 2026;27:1688. doi: 10.3390/ijms27041688. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] 15.Hou Z., Brenner J.S. Developing targeted antioxidant nanomedicines for ischemic penumbra: Novel strategies in treating brain ischemia-reperfusion injury. Redox Biol. 2024;73 doi: 10.1016/j.redox.2024.103185. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16.Mu J., Li C., Shi Y., Liu G., Zou J., Zhang D.Y., Jiang C., Wang X., He L., Huang P., et al. Protective effect of platinum nano-antioxidant and nitric oxide against hepatic ischemia-reperfusion injury. Nat. Commun. 2022;13:2513. doi: 10.1038/s41467-022-29772-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] 17.Kumar M.A., Baba S.K., Sadida H.Q., Marzooqi S.A., Jerobin J., Altemani F.H., Algehainy N., Alanazi M.A., Abou-Samra A.B., Kumar R., et al. Extracellular vesicles as tools and targets in therapy for diseases. Signal Transduct. Target. Ther. 2024;9:27. doi: 10.1038/s41392-024-01735-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.He L., Huang G., Liu H., Sang C., Liu X., Chen T. Highly bioactive zeolitic imidazolate framework-8-capped nanotherapeutics for efficient reversal of reperfusion-induced injury in ischemic stroke. Sci. Adv. 2020;6 doi: 10.1126/sciadv.aay9751. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] 19.Briones-Valdivieso C., Briones F., Orellana-Urzúa S., Chichiarelli S., Saso L., Rodrigo R. Novel Multi-Antioxidant Approach for Ischemic Stroke Therapy Targeting the Role of Oxidative Stress. Biomedicines. 2024;12 doi: 10.3390/biomedicines12030501. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20.Bai Y., Li Y., Li Y., Tian L. Advanced Biological Applications of Cerium Oxide Nanozymes in Disease Related to Oxidative Damage. ACS Omega. 2024;9:8601–8614. doi: 10.1021/acsomega.3c03661. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] 21.Koo S., Sohn H.S., Kim T.H., Yang S., Jang S.Y., Ye S., Choi B., Kim S.H., Park K.S., Shin H.M., et al. Ceria-vesicle nanohybrid therapeutic for modulation of innate and adaptive immunity in a collagen-induced arthritis model. Nat. Nanotechnol. 2023;18:1502–1514. doi: 10.1038/s41565-023-01523-y. [DOI] [PubMed] [Google Scholar]

[bib22] 22.Orlova V.V., van den Hil F.E., Petrus-Reurer S., Drabsch Y., Ten Dijke P., Mummery C.L. Generation, expansion and functional analysis of endothelial cells and pericytes derived from human pluripotent stem cells. Nat. Protoc. 2014;9:1514–1531. doi: 10.1038/nprot.2014.102. [DOI] [PubMed] [Google Scholar]

[bib23] 23.Lian X., Bao X., Al-Ahmad A., Liu J., Wu Y., Dong W., Dunn K.K., Shusta E.V., Palecek S.P. Efficient differentiation of human pluripotent stem cells to endothelial progenitors via small-molecule activation of WNT signaling. Stem Cell Rep. 2014;3:804–816. doi: 10.1016/j.stemcr.2014.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] 24.Lian Q., Zhang Y., Zhang J., Zhang H.K., Wu X., Zhang Y., Lam F.F.Y., Kang S., Xia J.C., Lai W.H., et al. Functional mesenchymal stem cells derived from human induced pluripotent stem cells attenuate limb ischemia in mice. Circulation. 2010;121:1113–1123. doi: 10.1161/circulationaha.109.898312. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Zhao Q., Gregory C.A., Lee R.H., Reger R.L., Qin L., Hai B., Park M.S., Yoon N., Clough B., McNeill E., et al. MSCs derived from iPSCs with a modified protocol are tumor-tropic but have much less potential to promote tumors than bone marrow MSCs. Proc. Natl. Acad. Sci. USA. 2015;112:530–535. doi: 10.1073/pnas.1423008112. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] 26.Chambers S.M., Fasano C.A., Papapetrou E.P., Tomishima M., Sadelain M., Studer L. Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat. Biotechnol. 2009;27:275–280. doi: 10.1038/nbt.1529. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Fan J., Lee C.S., Kim S., Chen C., Aghaloo T., Lee M. Generation of Small RNA-Modulated Exosome Mimetics for Bone Regeneration. ACS Nano. 2020;14:11973–11984. doi: 10.1021/acsnano.0c05122. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Jiang H., Lou C., Jiang L., Lin C., Wang W., Yan Z., Yu J., Cai T., Lin S., Wang J., et al. Simvastatin-enhanced bioinspired exosome mimetics regulate osteogenesis and angiogenesis for the treatment of glucocorticoid-induced osteonecrosis of the femoral head. Chem. Eng. J. 2023;472 doi: 10.1016/j.cej.2023.144729. [DOI] [Google Scholar]

[bib29] 29.Kou M., Huang L., Yang J., Chiang Z., Chen S., Liu J., Guo L., Zhang X., Zhou X., Xu X., et al. Mesenchymal stem cell-derived extracellular vesicles for immunomodulation and regeneration: a next generation therapeutic tool? Cell Death Dis. 2022;13:580. doi: 10.1038/s41419-022-05034-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30.Liu Z.-L., Chen H.-H., Zheng L.-L., Sun L.-P., Shi L. Angiogenic signaling pathways and anti-angiogenic therapy for cancer. Signal Transduct. Target. Ther. 2023;8:198. doi: 10.1038/s41392-023-01460-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31.de Kreuk B.J., Gingras A.R., Knight J.D., Liu J.J., Gingras A.C., Ginsberg M.H. Heart of glass anchors Rasip1 at endothelial cell-cell junctions to support vascular integrity. eLife. 2016;5 doi: 10.7554/eLife.11394. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib32] 32.Benwell C.J., Johnson R.T., Taylor J.A.G.E., Lambert J., Robinson S.D. A proteomics approach to isolating neuropilin-dependent α5 integrin trafficking pathways: neuropilin 1 and 2 co-traffic α5 integrin through endosomal p120RasGAP to promote polarised fibronectin fibrillogenesis in endothelial cells. Commun. Biol. 2024;7:629. doi: 10.1038/s42003-024-06320-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Ji W., Li Y., Peng H., Zhao R., Shen J., Wu Y., Wang J., Hao Q., Lu Z., Yang J., Zhang X. Self-Catalytic Small Interfering RNA Nanocarriers for Synergistic Treatment of Neurodegenerative Diseases. Adv. Mater. 2022;34 doi: 10.1002/adma.202105711. [DOI] [PubMed] [Google Scholar]

[bib34] 34.Kim Y.G., Lee Y., Lee N., Soh M., Kim D., Hyeon T. Ceria-Based Therapeutic Antioxidants for Biomedical Applications. Adv. Mater. 2024;36 doi: 10.1002/adma.202210819. [DOI] [PubMed] [Google Scholar]

[bib35] 35.Jiang H., Xia W., Xia T., Jiang L., Yu J., Zhu X., Lin C., Lou C., Wang W., Chai Y., et al. Chemotactic recruitment of genetically engineered cell membrane-camouflaged metal-organic framework nanoparticles for ischemic osteonecrosis treatment. Acta Biomater. 2024;185:410–428. doi: 10.1016/j.actbio.2024.07.024. [DOI] [PubMed] [Google Scholar]

[bib36] 36.Ma J., Zhang S., Liu J., Liu F., Du F., Li M., Chen A.T., Bao Y., Suh H.W., Avery J., et al. Targeted Drug Delivery to Stroke via Chemotactic Recruitment of Nanoparticles Coated with Membrane of Engineered Neural Stem Cells. Small. 2019;15 doi: 10.1002/smll.201902011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] 37.Bose R.J., Kim B.J., Arai Y., Han I.B., Moon J.J., Paulmurugan R., Park H., Lee S.H. Bioengineered stem cell membrane functionalized nanocarriers for therapeutic targeting of severe hindlimb ischemia. Biomaterials. 2018;185:360–370. doi: 10.1016/j.biomaterials.2018.08.018. [DOI] [PubMed] [Google Scholar]

[bib38] 38.Luo L., Zang G., Liu B., Qin X., Zhang Y., Chen Y., Zhang H., Wu W., Wang G. Bioengineering CXCR4-overexpressing cell membrane functionalized ROS-responsive nanotherapeutics for targeting cerebral ischemia-reperfusion injury. Theranostics. 2021;11:8043–8056. doi: 10.7150/thno.60785. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39.Qiao L., Hu S., Huang K., Su T., Li Z., Vandergriff A., Cores J., Dinh P.U., Allen T., Shen D., et al. Tumor cell-derived exosomes home to their cells of origin and can be used as Trojan horses to deliver cancer drugs. Theranostics. 2020;10:3474–3487. doi: 10.7150/thno.39434. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40.Chen P., Liu X., Gu C., Zhong P., Song N., Li M., Dai Z., Fang X., Liu Z., Zhang J., et al. A plant-derived natural photosynthetic system for improving cell anabolism. Nature. 2022;612:546–554. doi: 10.1038/s41586-022-05499-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Zheng Z., Wang M., Cheng C., Liu D., Wu L., Zhu J., Qian X. Ginsenoside Rb1 reduces H2O2-induced HUVEC dysfunction by stimulating the sirtuin-1/AMP-activated protein kinase pathway. Mol. Med. Rep. 2020;22:247–256. doi: 10.3892/mmr.2020.11096. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] 42.Chen C.Y., Rao S.S., Yue T., Tan Y.J., Yin H., Chen L.J., Luo M.J., Wang Z., Wang Y.Y., Hong C.G., et al. Glucocorticoid-induced loss of beneficial gut bacterial extracellular vesicles is associated with the pathogenesis of osteonecrosis. Sci. Adv. 2022;8 doi: 10.1126/sciadv.abg8335. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] 43.Yu B., Li H., Zhang Z., Chen P., Wang L., Fan X., Ning X., Pan Y., Zhou F., Hu X., et al. Extracellular vesicles engineering by silicates-activated endothelial progenitor cells for myocardial infarction treatment in male mice. Nat. Commun. 2023;14 doi: 10.1038/s41467-023-37832-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 44.Martin M., Vermeiren S., Bostaille N., Eubelen M., Spitzer D., Vermeersch M., Profaci C.P., Pozuelo E., Toussay X., Raman-Nair J., et al. Engineered Wnt ligands enable blood-brain barrier repair in neurological disorders. Science (New York, N.Y.) 2022;375 doi: 10.1126/science.abm4459. [DOI] [PubMed] [Google Scholar]

[bib45] 45.Rohde D., Vandoorne K., Lee I.H., Grune J., Zhang S., McAlpine C.S., Schloss M.J., Nayar R., Courties G., Frodermann V., et al. Bone marrow endothelial dysfunction promotes myeloid cell expansion in cardiovascular disease. Nat. Cardiovasc. Res. 2022;1:28–44. doi: 10.1038/s44161-021-00002-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] 46.Mehta V., Pang K.L., Givens C.S., Chen Z., Huang J., Sweet D.T., Jo H., Reader J.S., Tzima E. Mechanical forces regulate endothelial-to-mesenchymal transition and atherosclerosis via an Alk5-Shc mechanotransduction pathway. Sci. Adv. 2021;7 doi: 10.1126/sciadv.abg5060. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] 47.Augustin H.G., Koh G.Y. Organotypic vasculature: From descriptive heterogeneity to functional pathophysiology. Science. 2017;357 doi: 10.1126/science.aal2379. [DOI] [PubMed] [Google Scholar]

[bib48] 48.Ji B.S., Cen J., He L., Liu M., Liu Y.Q., Liu L. Modulation of P-glycoprotein in rat brain microvessel endothelial cells under oxygen glucose deprivation. J. Pharm. Pharmacol. 2013;65:1508–1517. doi: 10.1111/jphp.12122. [DOI] [PubMed] [Google Scholar]

[bib49] 49.Yu G., Chen Y., Yang N., Zhang H., Zhang X., Geng Y., Zhao J., Chen Z., Dong C., Lin L., et al. Apoptotic Bodies Derived from Fibroblast-Like Cells in Subcutaneous Connective Tissue Inhibit Ferroptosis in Ischaemic Flaps via the miR-339-5p/KEAP1/Nrf2 Axis. Adv. Sci. 2024;11 doi: 10.1002/advs.202307238. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] 50.Baaten C.C.F.M.J., Vondenhoff S., Noels H. Endothelial Cell Dysfunction and Increased Cardiovascular Risk in Patients With Chronic Kidney Disease. Circ. Res. 2023;132:970–992. doi: 10.1161/circresaha.123.321752. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] 51.You Y., Tian Y., Guo R., Shi J., Kwak K.J., Tong Y., Estania A.P., Hsu W.H., Liu Y., Hu S., et al. Extracellular vesicle-mediated VEGF-A mRNA delivery rescues ischaemic injury with low immunogenicity. Eur. Heart J. 2025;46:1662–1676. doi: 10.1093/eurheartj/ehae883. [DOI] [PubMed] [Google Scholar]

[bib52] 52.Hu G.W., Li Q., Niu X., Hu B., Liu J., Zhou S.m., Guo S.c., Lang H.l., Zhang C.q., Wang Y., Deng Z.f. Exosomes secreted by human-induced pluripotent stem cell-derived mesenchymal stem cells attenuate limb ischemia by promoting angiogenesis in mice. Stem Cell Res. Ther. 2015;6:10. doi: 10.1186/scrt546. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] 53.Zhang J., Guan J., Niu X., Hu G., Guo S., Li Q., Xie Z., Zhang C., Wang Y. Exosomes released from human induced pluripotent stem cells-derived MSCs facilitate cutaneous wound healing by promoting collagen synthesis and angiogenesis. J. Transl. Med. 2015;13:49. doi: 10.1186/s12967-015-0417-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib54] 54.Xia Y., Ling X., Hu G., Zhu Q., Zhang J., Li Q., Zhao B., Wang Y., Deng Z. Small extracellular vesicles secreted by human iPSC-derived MSC enhance angiogenesis through inhibiting STAT3-dependent autophagy in ischemic stroke. Stem Cell Res. Ther. 2020;11:313. doi: 10.1186/s13287-020-01834-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] 55.Ye M., Ni Q., Qi H., Qian X., Chen J., Guo X., Li M., Zhao Y., Xue G., Deng H., Zhang L. Exosomes Derived from Human Induced Pluripotent Stem Cells-Endothelia Cells Promotes Postnatal Angiogenesis in Mice Bearing Ischemic Limbs. Int. J. Biol. Sci. 2019;15:158–168. doi: 10.7150/ijbs.28392. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib56] 56.Wang X., Chen Y., Zhao Z., Meng Q., Yu Y., Sun J., Yang Z., Chen Y., Li J., Ma T., et al. Engineered Exosomes With Ischemic Myocardium-Targeting Peptide for Targeted Therapy in Myocardial Infarction. J. Am. Heart Assoc. 2018;7 doi: 10.1161/jaha.118.008737. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib57] 57.Chen W.C.W., Lee B.G., Park D.W., Kim K., Chu H., Kim K., Huard J., Wang Y. Controlled dual delivery of fibroblast growth factor-2 and Interleukin-10 by heparin-based coacervate synergistically enhances ischemic heart repair. Biomaterials. 2015;72:138–151. doi: 10.1016/j.biomaterials.2015.08.050. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib58] 58.Ruozi G., Bortolotti F., Mura A., Tomczyk M., Falcione A., Martinelli V., Vodret S., Braga L., Dal Ferro M., Cannatà A., et al. Cardioprotective factors against myocardial infarction selected in vivo from an AAV secretome library. Sci. Transl. Med. 2022;14 doi: 10.1126/scitranslmed.abo0699. [DOI] [PubMed] [Google Scholar]

[bib59] 59.Mandai M., Watanabe A., Kurimoto Y., Hirami Y., Morinaga C., Daimon T., Fujihara M., Akimaru H., Sakai N., Shibata Y., et al. Autologous Induced Stem-Cell-Derived Retinal Cells for Macular Degeneration. N. Engl. J. Med. 2017;376:1038–1046. doi: 10.1056/NEJMoa1608368. [DOI] [PubMed] [Google Scholar]

[bib60] 60.Schweitzer J.S., Song B., Herrington T.M., Park T.Y., Lee N., Ko S., Jeon J., Cha Y., Kim K., Li Q., et al. Personalized iPSC-Derived Dopamine Progenitor Cells for Parkinson's Disease. N. Engl. J. Med. 2020;382:1926–1932. doi: 10.1056/NEJMoa1915872. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib61] 61.Deuse T., Hu X., Gravina A., Wang D., Tediashvili G., De C., Thayer W.O., Wahl A., Garcia J.V., Reichenspurner H., et al. Hypoimmunogenic derivatives of induced pluripotent stem cells evade immune rejection in fully immunocompetent allogeneic recipients. Nat. Biotechnol. 2019;37:252–258. doi: 10.1038/s41587-019-0016-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib62] 62.Hu X., White K., Olroyd A.G., DeJesus R., Dominguez A.A., Dowdle W.E., Friera A.M., Young C., Wells F., Chu E.Y., et al. Hypoimmune induced pluripotent stem cells survive long term in fully immunocompetent, allogeneic rhesus macaques. Nat. Biotechnol. 2024;42:413–423. doi: 10.1038/s41587-023-01784-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib63] 63.Thakur A., Rai D. Global requirements for manufacturing and validation of clinical grade extracellular vesicles. J. Liq. Biopsy. 2024;6 doi: 10.1016/j.jlb.2024.100278. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib64] 64.Valadi H., Ekström K., Bossios A., Sjöstrand M., Lee J.J., Lötvall J.O. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat. Cell Biol. 2007;9:654–659. doi: 10.1038/ncb1596. [DOI] [PubMed] [Google Scholar]

[bib65] 65.Lin R.Z., Im G.B., Luo A.C., Zhu Y., Hong X., Neumeyer J., Tang H.W., Perrimon N., Melero-Martin J.M. Mitochondrial transfer mediates endothelial cell engraftment through mitophagy. Nature. 2024;629:660–668. doi: 10.1038/s41586-024-07340-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib66] 66.Jiang H., Wang W., Mao Y., Jiang L., Yu J., Zhu X., Fu H., Lin Z., Shen H., Pan X., Xue X. Morroniside-mediated mitigation of stem cell and endothelial cell dysfunction for the therapy of glucocorticoid-induced osteonecrosis of the femoral head. Int. Immunopharmacol. 2024;127 doi: 10.1016/j.intimp.2023.111421. [DOI] [PubMed] [Google Scholar]

[bib67] 67.Jiang H., Yu J., Yan Z., Lin Z., Lin M., Mao Y., Hong Z., Lin J., Xue X., Pan X. Pharmacological activation of the Nrf2 pathway by Taxifolin remodels articular cartilage microenvironment for the therapy of Osteoarthritis. Int. Immunopharmacol. 2023;122 doi: 10.1016/j.intimp.2023.110587. [DOI] [PubMed] [Google Scholar]

[bib68] 68.Wu O., Jin Y., Zhang Z., Zhou H., Xu W., Chen L., Jones M., Kwan K.Y.H., Gao J., Zhang K., et al. KMT2A regulates the autophagy-GATA4 axis through METTL3-mediated m(6)A modification of ATG4a to promote NPCs senescence and IVDD progression. Bone Res. 2024;12:67. doi: 10.1038/s41413-024-00373-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib69] 69.Deng R., Zhao R., Zhang Z., Chen Y., Yang M., Lin Y., Ye J., Li N., Qin H., Yan X., et al. Chondrocyte membrane-coated nanoparticles promote drug retention and halt cartilage damage in rat and canine osteoarthritis. Sci. Transl. Med. 2024;16 doi: 10.1126/scitranslmed.adh9751. [DOI] [PubMed] [Google Scholar]

[bib70] 70.Jiang H., Lin C., Cai T., Jiang L., Lou C., Lin S., Wang W., Yan Z., Pan X., Xue X. Taxifolin-mediated Nrf2 activation ameliorates oxidative stress and apoptosis for the treatment of glucocorticoid-induced osteonecrosis of the femoral head. Phytother Res. 2024;38:156–173. doi: 10.1002/ptr.8031. [DOI] [PubMed] [Google Scholar]

[bib71] 71.Liu W., Zhao N., Yin Q., Zhao X., Guo K., Xian Y., Li S., Wang C., Zhu M., Du Y., et al. Injectable Hydrogels Encapsulating Dual-Functional Au@Pt Core-Shell Nanoparticles Regulate Infarcted Microenvironments and Enhance the Therapeutic Efficacy of Stem Cells through Antioxidant and Electrical Integration. ACS Nano. 2023;17:2053–2066. doi: 10.1021/acsnano.2c07436. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib72] 72.Tang L., Yin Y., Liu H., Zhu M., Cao Y., Feng J., Fu C., Li Z., Shu W., Gao J., et al. Blood-Brain Barrier-Penetrating and Lesion-Targeting Nanoplatforms Inspired by the Pathophysiological Features for Synergistic Ischemic Stroke Therapy. Adv. Mater. 2024;36 doi: 10.1002/adma.202312897. [DOI] [PubMed] [Google Scholar]

PERMALINK

Universal catalytic endothelial lineage nanovesicles for targeting and rescuing ischemic endothelium

Hongyi Jiang

Yumeng Wang

Haosheng Huang

Liang Zhu

Hanting Shen

Liting Jiang

Chihao Lin

Huangyi Fang

Gang Zheng

Junfeng Shi

Xinghe Xue

Jilong Wang

Jian Lin

Junjie Deng

Liang Chen

Xiaoyun Pan

Summary

Graphical abstract

Highlights

Introduction

Results

EC-derived nanovesicles exhibit superior pro-angiogenic potential

Figure 1.

Engineering of nanozyme-armed nanovesicles for synergistic therapy

Figure 2.

CeO2-NMUiECs possess innate dual-targeting capacity to ischemic endothelium

Figure 3.

CeO2-NMUiECs rescue endothelial dysfunction under diverse ischemic insults in vitro

Figure 4.

CeO2-NMUiECs exert a conserved protective effect across endothelial cells of diverse tissue origins

Figure 5.

Effect of CeO2-NMUiECs in bone and skin ischemia

Figure 6.

Effect of CeO2-NMUiECs in myocardial and cerebral ischemia

Figure 7.

Preclinical evaluation and optimization of CeO2-NMUiECs for clinically relevant applications

Discussion

Limitations of the study

Resource availability

Lead contact

Materials availability

Data and code availability

Acknowledgments

Author contributions

Declaration of interests

STAR★Methods

Key resources table

Experimental model and study participant details

Human induced pluripotent stem cells (iPSCs)

iPSC-derived lineages

Endothelial and uptake assay cell models

Animals

Method details

Synthesis and functionalization of ceria (CeO2) nanozymes

Preparation of nanovesicle mimicking units (NMUs)

Covalent conjugation of CeO2 nanozymes to NMUiECs

Morphology, size, and surface potential analysis

Proteomics and western blot analysis

In vitro ROS scavenging and enzyme-mimetic activity assays

Surface CXCR4 expression assessment

Cellular uptake and cytotoxicity analysis

In vitro preferential uptake analysis

In vitro ischemic injury models

Tube formation assay

Migration assay

Intracellular and mitochondrial ROS analysis

Mitochondrial membrane potential and apoptosis assessment

Quantitative PCR

Western blotting

Transcriptomics study

Animal models of ischemia and therapeutic interventions

Biodistribution, pharmacokinetics, and targeting studies

Micro-computed tomography (Micro-CT) analysis

Skin flap viability assessment

Echocardiography

TTC staining

Histology, immunohistochemistry, and immunofluorescence

In vivo biosafety and immunogenicity

CeO₂-NMU_iECs possess innate dual-targeting capacity to ischemic endothelium

CeO₂-NMU_iECs rescue endothelial dysfunction under diverse ischemic insults in vitro

CeO₂-NMU_iECs exert a conserved protective effect across endothelial cells of diverse tissue origins

Effect of CeO₂-NMU_iECs in bone and skin ischemia

Effect of CeO₂-NMU_iECs in myocardial and cerebral ischemia

Preclinical evaluation and optimization of CeO₂-NMU_iECs for clinically relevant applications

Synthesis and functionalization of ceria (CeO₂) nanozymes

Covalent conjugation of CeO₂ nanozymes to NMU_iECs