Hypoxia preconditioned MSC exosomes attenuate high-altitude cerebral edema via the miR-125a-5p/RTEF-1 axis to protect vascular endothelial cells

Abstract

Vasogenic edema, caused by the disruption of the blood-brain barrier (BBB), is a significant pathological factor in high-altitude cerebral edema (HACE). Due to the rapid progression and high mortality rate of HACE, prophylactic treatment is important. Mesenchymal stem cell exosomes (MSC-EXO) are increasingly being used in tissue injury repair, and research suggests that appropriate conditioning can enhance the targeted efficacy of exosome therapy. Our in vitro experiments revealed that hypoxia preconditioned MSC-EXO (H-EXO) significantly outperformed normoxic MSC-EXO (N-EXO) in multiple protective aspects. Specifically, H-EXO demonstrated enhanced capacity to mitigate hypoxia-induced aberrant angiogenesis, maintain vascular endothelial cell viability, and suppress ROS accumulation and apoptotic signaling under hypoxic stress. Mechanistic investigation identified miR-125a-5p cargo in H-EXO as a key mediator of RTEF-1 targeted inhibition during hypoxic exposure. In corresponding in vivo studies, H-EXO administration effectively attenuated HACE-induced pathological angiogenesis while maintaining crucial vascular homeostasis markers. The therapeutic effects manifested through three principal aspects: 1) downregulation of RTEF-1/VEGF hyperexpression, 2) modulation of VE-cadherin, SMA, and PDGFRα + β expression to preserve BBB integrity, and 3) concurrent protection of neurovascular functions against HACE-induced damage. This investigation elucidates the miR-125a-5p/RTEF-1 axis as the central mechanism through which hypoxic preconditioning enhances MSC-EXO's endothelial protective properties. Our findings establish H-EXO's multimodal therapeutic potential, demonstrating its capacity to simultaneously inhibit pathological angiogenesis, restore BBB function, and protect neural tissue under hypoxic stress conditions. The study elucidates key mechanisms underlying clinical prevention and management of HACE by delineating H-EXO's preventive mechanisms against hypoxia-induced cerebrovascular injury.

Keywords: High-altitude cerebral edema; Hypoxia preconditioned exosomes; Neurological recovery; Vascular endothelial cells; miR-125a-5p.

Graphical abstract

Fig. 1

Identification and characterization of H-EXO. (A) Extraction schematic of N-EXO and H-EXO (created with biorender.com). (B) Morphological characteristics of N-EXO and H-EXO under transmission electron microscope. (C), (D) NTA technique was used to detect N-EXO and H-EXO. (E) Dil labels phagocytic exosomes in vascular endothelial cells. (F) To investigate the expression of exosome-specific protein markers CD9, CD63, CD81, Calnexin and GM130 by WB. (G) Protein concentrations of N-EXO and H-EXO. All data are presented as mean ± SD; Scale bars: 1 μm (B left), 500 nm (B right), 573 nm (D), 25 μm (E left), 3.125 μm (E right).

Fig. 2

Effect of H-EXO on functional recovery of vascular endothelial cells after hypoxia. (A) Schematic diagram of transwell cell permeability assay (created with biorender.com). (B) Lower chamber fluorescence over time. (C), (D) Representative pictures and statistical analysis of in vitro wound healing test results. (E), (F) DCFH-DA staining was used to detect intracellular ROS production and statistical analysis. (G), (H) Expression and statistical analysis of RTEF-1, VEGF, P2Y12, and INOS were investigated by WB. All data are presented as mean ± SD, statistical analysis using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus Hypo group, # versus N-EXO-Hypo group. Scale bar: 200 μm (C).

Fig. 3

Effect of H-EXO on vascular function after hypoxia. (A) Representative plot of vessel formation. (B) Statistical graph of length of vessel formation and number of nodes. (D), (E) and (F) Representative flow charts of cell cycles, statistical graphs of % G1 phase and statistical graphs of % S phase. (H) Statistical plot of cell viability detected by ATP. (I), (J) Caspase-3 immunofluorescence representative maps and caspase-3 positive statistics. All data are presented as mean ± SD, statistical analysis using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus Hypo group, # versus N-EXO-Hypo group. Scale bars: (A) 100 μm, (I) 50 μm.

Fig. 4

The miRNA expression profile in H-EXO and its mechanism of action in hypoxia-induced vascular endothelial cell dysfunction. (A) Expression of Hypoxia-Related miRNAs and Prediction of Upstream miRNAs of RTEF-1 Using miRTargetLink 2.0. (B) Volcano plot of differentially expressed miRNAs between H-EXO and N-EXO. (C) Scatter plot of differentially expressed miRNAs between H-EXO and N-EXO. (D) TOP10-20 enriched biological processes in GO enrichment analysis of all differentially expressed miRNA target genes. (E) TOP20-enriched signaling pathways analyzed for KEGG enrichment of miR-125a-5p. (F) Prediction of binding sites for RTEF-1 to miR-125a-5p using TargetScan. (G) Dual-luciferase reporter assay design wt-RTEF-1 versus mut-RTEF-1. (H) Dual-luciferase reporter assay results. (I) Protein expression of RTEF-1 in vascular endothelial cells after miR-125a-5p mimics inhibition. All data are presented as mean ± SD, Statistical analysis was performed using a t-test. ∗P < 0.05, ∗∗P < 0.01.

Fig. 5

Functional study of miR-125a-5p mimics/inhibitor. (A) Representative plot of vessel formation. (B) Statistical graph of length of vessel formation and number of nodes. (D), (E) and (F) Representative flow charts of cell cycles, statistical graphs of % G1 phase and statistical graphs of % S phase. (H) Statistical plot of cell viability detected by ATP. (I), (J) Caspase-3 immunofluorescence representative maps and caspase-3 positive statistics. All data were presented as mean ± SD and statistical analysis was performed using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus Hypo group, # versus H-EXO-in-Hypo group. Scale bars: (A) 100 μm, (I) 50 μm.

Fig. 6

RTEF-1 as a target gene of miR-125a-5p in H-EXO alleviating hypoxia-induced vascular endothelial cell dysfunction. (A) Representative images of in vitro wound healing assay results. (B) Statistical analysis of in vitro wound healing. (C) Lower chamber fluorescence over time in permeability experiments. (D), (E) DCFH-DA staining fluorescence was used to detect intracellular ROS production and statistical analysis. (F), (G) Expression and statistical analysis of RTEF-1, VEGF, P2Y12, and INOS were investigated by WB. (H) Correlation between RTEF-1 and VEGF expression in GSE122952 dataset. All data are presented as mean ± SD, statistical analysis using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus Hypo group, # versus Si-RTEF-1-Sham group. Scale bars: 200 μm (B), 50 μm (E).

Fig. 7

Potential benefits of H-EXO on neurological recovery in HACE mice in vivo. (A) Schematic representation of HACE mouse disease model construction (created with biorender.com). (B) Representative swimming paths in each group in Morris water maze test. (C) Escape latency of each group in Morris water maze test. (D) Target quadrant residence time of each group in Morris water maze test. (E) Comparison of the number of platform crossings in each group in Morris water maze test. (F) Time through balance beam. (G) Duration on rotarod. (H) Forelimb grip strength. All data are presented as mean ± SD, statistical analysis using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus HACE group, # versus N-EXO-HACE group.

Fig. 8

Effect of H-EXO on promoting tissue remodeling and alleviating neuronal injury in HACE mouse model. (A), (B) Representative pictures of HE staining results and statistical analysis of cavity area in each group. (C), (D), (E) Representative pictures of Map2/NeuN immunofluorescence results with statistical analysis of neuronal number and statistical analysis of Map2 fluorescence intensity. (F), (G) Representative pictures of NISSL staining results and statistical analysis of Nissl number. All data are presented as mean ± SD, statistical analysis using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus HACE group, # versus N-EXO-HACE group. Scale bars: 100 μm (top A), 33.3 μm (bottom A), 100 μm (left C), 20 μm (right C), 100 μm (top F), 33.3 μm (bottom F).

Fig. 9

Evaluation of the effect of H-EXO on cerebral vessels in HACE mice. (A), (B) Representative pictures of CD31 immunofluorescence staining results and statistical analysis of CD31 fluorescence intensity. (C), (D) Representative pictures of GFAP immunofluorescence staining results and statistical analysis of GFAP fluorescence intensity. (E), (F) Representative pictures of VE-Cadherin immunofluorescence staining results and statistical analysis of VE-Cadherin fluorescence intensity. (G) Expression and statistical analysis of RTEF-1 and VEGF were investigated by WB. All data are presented as mean ± SD, statistical analysis using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus HACE group, # versus N-EXO-HACE group. Scale bars: 100 μm (A left), 20 μm (A right).

Fig. 10

Assessment of the effect of H-EXO on other functions of cerebral vessels in HACE mice. (A), (B) Representative pictures of SMA immunofluorescence staining results and statistical analysis of CD31 fluorescence intensity. (C), (D) Representative pictures of PDGFRα + β immunofluorescence staining results and statistical analysis of PDGFRα + β fluorescence intensity. (E), (F) Representative pictures of P53 immunofluorescence staining results and statistical analysis of P53 fluorescence intensity. All data were presented as mean ± SD and statistical analysis was performed using one-way analysis of variance. ∗P < 0.05, ∗∗P < 0.01, ∗ versus HACE group, ^# versus N-EXO-HACE group. Scale bars: 100 μm (A left), 33.3 μm (A right).