Ectopic endometrial mesenchymal stem cell-derived exosomal miR-4466 promotes angiogenesis by targeting RUNX1 in adenomyosis

Jindan Wang: Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft. Yongyan Ni: Formal analysis, Investigation, Visualization, Writing – review & editing. Jiali Sun: Conceptualization, Data curation, Writing – review & editing. Yingying Qiu: Data curation, Formal analysis, Funding acquisition. Xinjun Wei: Investigation, Methodology. Bei Wang: Data curation, Project administration. Meihua Huang: Resources, Supervision, Writing – review & editing. Zhenli Li: Resources, Supervision, Writing – review & editing. Tao Gui: Conceptualization, Funding acquisition, Resources, Supervision, Writing – review & editing.

This work is financially supported by the 10.13039/501100001809 National Natural Science Foundation of China ( 81403321 ), the Science and Technology Project of Jiangsu Provincial Administration of Traditional Chinese Medicine (MS2023039, MS2023079, and ZD202419), the Development Fund Project of the Affiliated Hospital of 10.13039/501100012217 Xuzhou Medical University ( XYFY202322 ), and the Graduate Research and Practice Innovation Plan of 10.13039/501100002949 Jiangsu Province ( SJCX23 _0817).

The results of the cell phenotype analysis confirmed that both N-eMSCs and Ec-eMSCs isolated in this study exhibited characteristics typical of mesenchymal stem cells, as evidenced by their positive expression of CD73, CD90, and CD105, along with negative expression of HLA-DR ( Fig. S1 ). To analyse the paracrine pro-angiogenic effects of eMSCs derived from normal endometrium, eutopic endometrium and adenomyosis lesions, an in vitro co-culture system was utilised to assess changes in proliferation, invasion, migration, and tube formation of HUVECs ( Fig. 1 A). Results from the EdU, transwell, wound healing, and μ-slide angiogenesis assays demonstrated that Ec-eMSCs have a significantly greater capacity to enhance the proliferation, invasion, migration, and tube formation of HUVECs compared to N-eMSCs and Eu-eMSCs ( Fig. 1 B–E). These findings indicate that Ec-eMSCs exert potent effects on the angiogenesis process via paracrine signalling. Fig. 1 The pro-angiogenic effects of eMSCs from adenomyotic lesions on HUVECs. (A) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs co-cultured with Ec-eMSCs compared with Eu-eMSCs or N-eMSCs, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (B) Quantitative analysis of EdU positive cell proportion. (C) Statistical analysis of the number of invasive cells in per field in different groups. (D) Quantitative analysis of the wound healing assay results. (E) Quantification of tubule formation, defined as total vessels length. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 1 The pro-angiogenic effects of eMSCs from adenomyotic lesions on HUVECs. (A) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs co-cultured with Ec-eMSCs compared with Eu-eMSCs or N-eMSCs, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (B) Quantitative analysis of EdU positive cell proportion. (C) Statistical analysis of the number of invasive cells in per field in different groups. (D) Quantitative analysis of the wound healing assay results. (E) Quantification of tubule formation, defined as total vessels length. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Given that the pro-angiogenic paracrine activity of mesenchymal stem cells (MSCs) is primarily mediated by exosomes, we hypothesised that Ec-eMSCs may exert their pro-angiogenic effect through exosomes. To test this, Ec-eMSCs were pre-treated with GW4869, an inhibitor that impairs exosome synthesis, before co-culturing with HUVECs. The pro-angiogenic effects of Ec-eMSCs were then analysed as described above ( Fig. 2 A). Phenotypic analysis revealed that GW4869-treated Ec-eMSCs significantly reduced the proliferation, invasion/migration, and tube formation of HUVECs compared to DMSO-treated Ec-eMSCs ( Fig. 2 B–E). These observations suggest that the paracrine pro-angiogenic effect of Ec-eMSCs is primarily mediated by exosomes. Fig. 2 Effect of GW4869 on the pro-angiogenic capabilities of Ec-eMSCs. (A) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs co-cultured with Ec-eMSCs treated with or without GW4869, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (B) Quantitative analysis of EdU positive HUVECs. (C) Statistical analysis of the number of invasive cells in per field at different groups. (D) Quantitative analysis of the wound healing assay results. (E) Quantification of tubule formation, defined as total vessels length. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 2 Effect of GW4869 on the pro-angiogenic capabilities of Ec-eMSCs. (A) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs co-cultured with Ec-eMSCs treated with or without GW4869, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (B) Quantitative analysis of EdU positive HUVECs. (C) Statistical analysis of the number of invasive cells in per field at different groups. (D) Quantitative analysis of the wound healing assay results. (E) Quantification of tubule formation, defined as total vessels length. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Extracellular vesicle-like particles were isolated from the supernatant of Ec-eMSCs and characterised by NTA, Western blot, and TEM analyses. NTA analysis revealed that the particle size distribution ranged from 60 to 400 nm, with a peak at 158.5 nm ( Fig. 3 A). Western blot results showed that Ec-exo were positive for CD9, CD63, and TSG101, but negative for calnexin, compared to Ec-eMSCs ( Fig. 3 B). TEM analysis demonstrated that these vesicles were spherical and membrane-encapsulated ( Fig. 3 C), confirming the successful isolation of Ec-exo from Ec-eMSCs. Additionally, cellular uptake experiments revealed that PKH26-labelled Ec-exo were internalised and distributed in the perinuclear region of HUVECs. In contrast, PFA-treated HUVECs were unable to internalise the Ec-exo, displaying a uniform diffusion pattern ( Fig. 3 D). Fig. 3 Exosomes derived from Ec-eMSCs enhances angiogenesis of HUVECs. (A) The size and distribution of Ec-eMSCs-derived exosomes (Ec-exo) by NTA. (B) Western blot of the positive exosome markers CD9, CD63, and TSG101 and the negative exosome marker calnexin in lysates from Ec-eMSCs and Ec-exo (Full-length blot are presented in Supplementary Figure). (C) Representative TEM images of Ec-exo (Scale bar = 100 nm). (D) Representative immunofluorescence images showing Ec-exo internalisation by HUVECs (Scale bar = 20 μm). Red color indicated exosomes (PKH26) and blue color indicated nuclei (DAPI). n = 3. (E) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs co-cultured with exosomes derived from Ec-eMSCs, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (F) Quantitative analysis of EdU positive cells. (G) Statistical analysis of the number of invasive cells in per field at different groups. (H) Quantitative analysis of the wound healing assay results. (I) Quantification of tubule formation, defined as total vessels length. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 3 Exosomes derived from Ec-eMSCs enhances angiogenesis of HUVECs. (A) The size and distribution of Ec-eMSCs-derived exosomes (Ec-exo) by NTA. (B) Western blot of the positive exosome markers CD9, CD63, and TSG101 and the negative exosome marker calnexin in lysates from Ec-eMSCs and Ec-exo (Full-length blot are presented in Supplementary Figure). (C) Representative TEM images of Ec-exo (Scale bar = 100 nm). (D) Representative immunofluorescence images showing Ec-exo internalisation by HUVECs (Scale bar = 20 μm). Red color indicated exosomes (PKH26) and blue color indicated nuclei (DAPI). n = 3. (E) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs co-cultured with exosomes derived from Ec-eMSCs, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (F) Quantitative analysis of EdU positive cells. (G) Statistical analysis of the number of invasive cells in per field at different groups. (H) Quantitative analysis of the wound healing assay results. (I) Quantification of tubule formation, defined as total vessels length. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. To validate the pro-angiogenic effect of Ec-exo, we utilised an in vitro co-culture system followed by a series of assays, including EdU, transwell, wound healing, and angiogenesis assays. The representative images showed that Ec-exo significantly promoted the proliferation, invasion/migration, and tube formation of HUVECs compared to N-exo ( Fig. 3 E–I). Collectively, these data indicate that Ec-exo exert a significant pro-angiogenic effect in vitro. To elucidate how Ec-exo exert their pro-angiogenic activity on HUVECs, exosomal small RNAs from N-eMSCs and Ec-eMSCs were subjected to miRNA sequencing. The findings revealed the distribution of small RNAs (smRNAs) in Ec-exo and N-exo, including miRNA, ribosomal RNA (rRNA), transfer RNA (tRNA), piwi-interacting RNA (piRNA), small nuclear RNA (snRNA), and small nucleolar RNA (snoRNA), with piRNA constituting over half of the total (56.1 % and 56.44 %) ( Fig. 4 A). The results indicated 173 DE-miRNAs between N-exosomes and Ec-exo, comprising 81 upregulated and 92 downregulated miRNAs. These findings were visualized using a heatmap ( Fig. 4 B) and a volcano plot ( Fig. 4 C). Fig. 4 MiRNA sequencing for exosomes and validation by qRT-PCR analysis. (A) Rainbow plot of classification annotation of smRNAs in N-exo and Ec-exo. (B) Heatmap showing differentially expressed (DE)-miRNAs between Ec-exo and N-eMSCs-derived exosomes (N-exo). n = 3. (C) Volcano plot showing DE-miRNAs between Ec-exo and N-exo. n = 3. D, E GO biological process (D) and KEGG (E) enrichment analysis of DE-miRNAs between Ec-exo and N-exo. n = 3. (F) The top 11 upregulated miRNAs (UMIs >10) were further validated by qRT-PCR. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 4 MiRNA sequencing for exosomes and validation by qRT-PCR analysis. (A) Rainbow plot of classification annotation of smRNAs in N-exo and Ec-exo. (B) Heatmap showing differentially expressed (DE)-miRNAs between Ec-exo and N-eMSCs-derived exosomes (N-exo). n = 3. (C) Volcano plot showing DE-miRNAs between Ec-exo and N-exo. n = 3. D, E GO biological process (D) and KEGG (E) enrichment analysis of DE-miRNAs between Ec-exo and N-exo. n = 3. (F) The top 11 upregulated miRNAs (UMIs >10) were further validated by qRT-PCR. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. To explore the functional categories of exosomal DE-miRNAs, we utilised miRTarBase, an experimentally validated miRNA-target interaction database, to predict the target genes of DE-miRNAs. The predicted target genes were enriched in various biological processes according to Gene Ontology (GO) terms, including regulation of angiogenesis, cytokinesis, wound healing, vascular permeability, endothelial cell activation/migration/chemotaxis, and blood vessel remodelling ( Fig. 4 D). Additionally, KEGG pathway analysis revealed significant enrichment in pro-angiogenic pathways, such as the VEGF signalling pathway and the NF-κB signalling pathway ( Fig. 4 E). These bioinformatics analyses suggest that the pro-angiogenic effect of Ec-exo may be mediated by exosomal miRNAs. Among the 81 upregulated miRNAs, low-expression miRNAs with a mean of unique molecular identifiers (UMIs) ≤ 10 were excluded. Consequently, 11 upregulated miRNAs listed in Supplementary Table 6 were selected for validation by qRT-PCR. The qRT-PCR results showed that 9 miRNAs were upregulated, 1 miRNA was downregulated, and 1 miRNA showed no change in Ec-exo ( Fig. 4 F). Overall, the qRT-PCR results were largely consistent with the sequencing analysis. Notably, miR-4466 was the most upregulated miRNA (3.7-fold, p <  0.001), drawing significant attention. Given the pro-angiogenic effect of Ec-exo and the significant upregulation of miR-4466 in Ec-exo compared to N-exo, we hypothesised that exosomal miR-4466 plays a crucial role in this effect. To confirm this hypothesis, Cy3-labelled miR-4466 (miR-4466-Cy3) was successfully transfected into Ec-eMSCs, as confirmed by immunofluorescence staining and qRT-PCR ( Fig. S2A and B ). Ec-exo were isolated from miR-4466-Cy3 transfected Ec-eMSCs and subsequently co-cultured with recipient HUVECs. Fluorescence images revealed the co-localisation of miR-4466-Cy3 and PKH67-labelled exosomes within the cytoplasm and around the nucleus of HUVECs ( Fig. 5 A). As expected, naked miR-4466-Cy3 treatment showed no fluorescence signal. These results confirm that miR-4466 can be transferred from Ec-eMSCs to HUVECs via exosome secretion and internalisation. Cell phenotype assays of HUVECs transfected with miR-4466 mimics demonstrated increased proliferation, invasion, migration, and tube formation compared to those transfected with NC-mimics. Conversely, HUVECs transfected with the miR-4466 inhibitor exhibited decreased proliferation, invasion, migration, and tube-forming capabilities compared to those transfected with the NC-inhibitor ( Fig. 5 B–F). In the presence of miR-4466 mimics, there was a significant increase in both intracellular and extracellular VEGFA expression in HUVECs ( Fig. 5 G and H), suggesting a potential relationship between miR-4466 and VEGFA. Collectively, these results clearly indicate that exosomal miR-4466 plays a crucial role in the pro-angiogenic activity of Ec-exo. Fig. 5 miR-4466 promotes angiogenesis capability of HUVECs in vitro. (A) The up panels show the presence of Cy3 fluorescence and PKH67 lipid dye in HUVECs after addition of PKH67-labelled exosomes derived from Ec-eMSCs for 12 h. HUVECs incubated with naked-miR-4466-Cy3 served as negative controls (the down panels). Blue color indicated nuclei (DAPI). Scale bar = 10 μm. n = 3. (B) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs transfected with miR-4466 mimics/inhibitor (miR-mimics/inhibitor) or corresponding NC-mimics/inhibitor, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (C) Quantitative analysis of EdU positive cells. (D) Statistical analysis of the number of invasive cells in per field at different groups. (E) Quantitative analysis of the wound healing assay results. (F) Quantification of tubule formation, defined as total vessels length. (G) Relative expression of VEGFA in HUVECs transfected with miR-mimics/inhibitor and NC-mimics/inhibitor using qRT-PCR. n = 3. (H) Relative levels of extracellular protein VEGFA in the supernatants from the indicated groups. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 5 miR-4466 promotes angiogenesis capability of HUVECs in vitro. (A) The up panels show the presence of Cy3 fluorescence and PKH67 lipid dye in HUVECs after addition of PKH67-labelled exosomes derived from Ec-eMSCs for 12 h. HUVECs incubated with naked-miR-4466-Cy3 served as negative controls (the down panels). Blue color indicated nuclei (DAPI). Scale bar = 10 μm. n = 3. (B) Representative images on the proliferation, invasion, migration, and tube formation abilities of HUVECs transfected with miR-4466 mimics/inhibitor (miR-mimics/inhibitor) or corresponding NC-mimics/inhibitor, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (C) Quantitative analysis of EdU positive cells. (D) Statistical analysis of the number of invasive cells in per field at different groups. (E) Quantitative analysis of the wound healing assay results. (F) Quantification of tubule formation, defined as total vessels length. (G) Relative expression of VEGFA in HUVECs transfected with miR-mimics/inhibitor and NC-mimics/inhibitor using qRT-PCR. n = 3. (H) Relative levels of extracellular protein VEGFA in the supernatants from the indicated groups. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. To explore the mechanism underlying the pro-angiogenic effect of miR-4466 in HUVECs, we identified RUNX1, DOT1L, and CALN1 as potential target genes of miR-4466 through bioinformatics analysis, which involved intersecting predicted target genes from the miRDB, miRWalk, miRPathDB, and mirDIP databases ( Fig. 6 A). TargetScan and miRDB were used to identify potential binding sites between miR-4466 and its target genes. Both databases provided strong evidence indicating that the sequence of miR-4466 highly matched the 3’ UTR of RUNX1, followed by CALN1 and DOT1L ( Fig. 6 B and Supplementary Tables 7 and 8 ). Fig. 6 miR-4466 directly regulates RUNX1 expression in HUVECs. (A) A Venn diagram shows target genes of miR-4466 predicted by the miRDB, miRWalk, miRPathDB, mirDIP databases. (B) Schematic diagram illustrating the predicted miR-4466 binding sites with the 3′ UTR of RUNX1. (C, D) Relative expression of target genes was detected by qRT-PCR in HUVECs (C) and HEK-293T cells (D) transfected with NC-mimics and miR-mimics. n = 3. E, F Relative expression of target genes was detected by qRT-PCR in HUVECs (E) and HEK-293T cells (F) transfected with NC-inhibitor and miR-inhibitor. n = 3. (G) Schematic representation of the wild-type (Wt) and mutant-type (Mut) binding site between the 3′ UTR of RUNX1 and miR-4466. (H) Relative luciferase activity of reporter constructs containing miRNA binding sites transfected with NC-mimics and miR-mimics in HUVECs. n = 3. (I) Western blot of RUNX1 and VEGFA protein expression in HUVECs after transfection with NC-mimics/inhibitor or miR-mimics/inhibitor for 48 h (Full-length blot are presented in Supplementary Figure). (J, K) Protein levels of RUNX1 (J) and VEGFA (K) were normalized to GAPDH protein expression level in (I). ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 6 miR-4466 directly regulates RUNX1 expression in HUVECs. (A) A Venn diagram shows target genes of miR-4466 predicted by the miRDB, miRWalk, miRPathDB, mirDIP databases. (B) Schematic diagram illustrating the predicted miR-4466 binding sites with the 3′ UTR of RUNX1. (C, D) Relative expression of target genes was detected by qRT-PCR in HUVECs (C) and HEK-293T cells (D) transfected with NC-mimics and miR-mimics. n = 3. E, F Relative expression of target genes was detected by qRT-PCR in HUVECs (E) and HEK-293T cells (F) transfected with NC-inhibitor and miR-inhibitor. n = 3. (G) Schematic representation of the wild-type (Wt) and mutant-type (Mut) binding site between the 3′ UTR of RUNX1 and miR-4466. (H) Relative luciferase activity of reporter constructs containing miRNA binding sites transfected with NC-mimics and miR-mimics in HUVECs. n = 3. (I) Western blot of RUNX1 and VEGFA protein expression in HUVECs after transfection with NC-mimics/inhibitor or miR-mimics/inhibitor for 48 h (Full-length blot are presented in Supplementary Figure). (J, K) Protein levels of RUNX1 (J) and VEGFA (K) were normalized to GAPDH protein expression level in (I). ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Subsequently, we conducted in vitro overexpression or inhibition of miR-4466 in HUVECs and HEK-293T cells ( Fig. S3A and B ). Notably, qRT-PCR assays revealed that overexpression of miR-4466 led to a decrease in RUNX1 expression, while inhibition of miR-4466 resulted in an increase in RUNX1 expression. However, neither overexpression nor knockdown of miR-4466 affected the expression of CALN1 and DOT1L ( Fig. 6 C–F). Additionally, bioinformatics analysis indicated two potential binding sites for miR-4466 in the 3’ UTR of RUNX2 in miRDB, while no binding was observed with RUNX3 in either TargetScan or miRDB ( Supplementary Tables 7 and 8 ). qRT-PCR analysis further showed that the expression levels of RUNX2 and RUNX3 were unaffected by miR-4466 overexpression or inhibition ( Fig. S3C and D ). To validate direct binding and repression effects, wild-type (Wt) and mutant (Mut) RUNX1-3′UTR sequences were synthesised and inserted into the GP-miRGLO vector ( Fig. 6 G). Dual luciferase reporter assays demonstrated that co-transfection of the Wt RUNX1-3′UTR luciferase reporter construct with miR-4466 mimics significantly reduced luciferase activity, while co-transfection of the Mut RUNX1-3′UTR luciferase reporter construct with miR-4466 mimics did not affect luciferase activity in HUVECs ( Fig. 6 H) and HEK-293T cells ( Fig. S3E ). These results suggest that miR-4466 directly regulates the expression of RUNX1 by targeting its 3′-UTR. Previous studies have indicated that RUNX1 functions as a transcriptional repressor of VEGFA and restricts angiogenesis. Consequently, we hypothesised that miR-4466 may promote angiogenesis by modulating VEGFA expression through targeting RUNX1. Western blot analysis revealed a significant downregulation of RUNX1 following miR-4466 overexpression ( Fig. 6 I and J). Interestingly, we observed a notable increase in intracellular mRNA ( Fig. 5 G) and protein ( Fig. 6 I–K) levels of VEGFA. Concurrently, there was an increase in extracellular VEGFA protein levels in HEK-293T cells ( Fig. S3F ). In contrast, inhibition of miR-4466 produced the opposite effect ( Fig. 6 I–K). Overall, these findings confirm that miR-4466 promotes VEGFA expression by directly targeting RUNX1. To investigate whether miR-4466-mediated angiogenesis directly depends on RUNX1, a series of rescue experiments were conducted using vectors that overexpressed RUNX1 in conjunction with miR-4466 mimics. Phenotypic analysis demonstrated that upregulation of miR-4466 enhanced the proliferation, invasion/migration, and tube formation capabilities of HUVECs in vitro. Remarkably, the pro-angiogenic effects of miR-4466 overexpression were counteracted by RUNX1 overexpression ( Fig. 7 A–E). Furthermore, miR-4466 overexpression led to a decrease in RUNX1 expression, which corresponded with increased VEGFA levels. Importantly, co-expression of miR-4466 and RUNX1 significantly reduced the VEGFA expression induced by miR-4466 ( Fig. 7 F–H). This observation was consistent with the levels of secreted VEGFA in the cell supernatant, as demonstrated by the ELISA results ( Fig. 7 I). To sum up, these findings confirm that miR-4466 promotes angiogenesis in HUVECs by directly targeting RUNX1. Fig. 7 RUNX1 functioned as transcriptional inhibitors of VEGFA. (A) Representative images on the proliferation, invasion, migration, and tube formation abilities of in HUVECs transfected with none, NC-mimics, miR-mimics alone or in combination with either empty pcDNA 3.1 vector or pcDNA 3.1-RUNX1, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (B) Quantitative analysis of EdU positive cells. (C) Statistical analysis of the number of invasive cells in per field at different groups. (D) Quantitative analysis of the wound healing assay results. (E) Quantification of tubule formation, defined as total vessels length. (F) Western blot of RUNX1 and VEGFA protein expression in HUVECs, with grouping consistent with the phenotypic experiments mentioned above (Full-length blot are presented in Supplementary Figure). (G, H) Protein levels of RUNX1 (G) and VEGFA (H) were normalized to GAPDH protein expression level in (F). n = 3. I ELISA of VEGFA protein expression in the indicated cell supernatants. n = 3. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05. Fig. 7 RUNX1 functioned as transcriptional inhibitors of VEGFA. (A) Representative images on the proliferation, invasion, migration, and tube formation abilities of in HUVECs transfected with none, NC-mimics, miR-mimics alone or in combination with either empty pcDNA 3.1 vector or pcDNA 3.1-RUNX1, arranged sequentially from top to bottom. Scale bars are included in each set of images. n = 3. (B) Quantitative analysis of EdU positive cells. (C) Statistical analysis of the number of invasive cells in per field at different groups. (D) Quantitative analysis of the wound healing assay results. (E) Quantification of tubule formation, defined as total vessels length. (F) Western blot of RUNX1 and VEGFA protein expression in HUVECs, with grouping consistent with the phenotypic experiments mentioned above (Full-length blot are presented in Supplementary Figure). (G, H) Protein levels of RUNX1 (G) and VEGFA (H) were normalized to GAPDH protein expression level in (F). n = 3. I ELISA of VEGFA protein expression in the indicated cell supernatants. n = 3. ∗ P <  0.05, ∗∗ P <  0.01, ∗∗∗ P   0.05.

This study was approved by the Ethical Committee of Jiangsu Province Hospital with Integration of Chinese and Western Medicine (2023-LWKYZ-055). All patients provided written informed consent prior to tissue collection. The research was carried out in according to the Declaration of Helsinki. Seven patients with AM and seven control patients with subserosal fibroid, who underwent hysterectomy, were diagnosed using transvaginal ultrasound and histopathological examination. The demographic and clinical data of patients and controls are summarized in Supplementary Table 1 . Adenomyotic lesions and normal endometrial tissues were obtained and immediately transported to the laboratory in ice-cold phosphate buffer solution (PBS) within 30 min for the primary culture of eMSCs. Eutopic and ectopic eMSCs (Eu-eMSCs and Ec-eMSCs) were obtained from matched eutopic endometrium and ectopic lesions of AM patients, while normal eMSCs (N-eMSCs) were derived from the endometrium of control patients. Primary eMSCs were isolated from tissues following established protocols with minor modifications [ 15 ]. Briefly, the tissues were minced into fragments and dissociated in 1 mg/mL collagenase type II (Sigma-Aldrich, USA) and deoxyribonuclease I (DNase I) (Sigma-Aldrich, USA) at 37 °C for 30 min. After filtration with 70 μm sterile cell sieves, the cell suspension was centrifuged at 300× g for 3 min at room temperature, and the supernatant was discarded. The cell pellet was resuspended in DMEM/F12 (1:1) (Gibco, USA) containing 10 % fetal bovine serum (FBS, ExCell Bio, China) and 1 % penicillin-streptomycin (Pen-Strep, Gibco, USA). The cells were then seeded into a T-25 flask and incubated in a humidified incubator containing 5 % CO2 at 37 °C. Once confluent, phycoerythrin (PE)-conjugated anti-human SUSD2+ eMSCs were sorted using anti-PE MicroBeads (Miltenyi Biotec, Germany) and subsequently cultured in DMEM/F12 (1:1) supplemented with 10 % mesenchymal stem cell-qualified FBS (Viva Cell, China) and 1 % Pen-Strep, and incubated in a humidified incubator containing 5 % CO2 at 37 °C. Flow cytometry was then performed on a FACSVerse (BD Biosciences, USA) to analyse the expression of positive cell surface markers (CD73, CD90, and CD105) and the negative cell surface marker human leukocyte antigen-DR (HLA-DR). All antibodies used for flow cytometry analyses were purchased from BioLegend (San Diego, CA, USA). Human umbilical vein endothelial cells (HUVECs) and human embryonic kidney 293T (HEK-293T) cells were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). HUVECs were cultured in complete endothelial cell medium (ECM) (ScienCell, CA, USA), consisting of basal medium supplemented with 5 % FBS and 1 % endothelial cell growth supplement. HEK-293T cells were cultured in high-glucose DMEM supplemented with 10 % FBS. All cells were maintained in a humidified incubator containing 5 % CO2 at 37 °C. When eMSCs reached approximately 70 % confluence, the medium was replaced with DMEM/F12 containing 10 % exosome-depleted FBS (Exo-FBS) (System Biosciences, USA) for 48 h. The cell culture supernatants were then collected and sequentially centrifuged at 300 g, 2000 g, and 10,000 g to remove cells and debris. The exosomes were then pelleted at 100,000 g for 70 min, washed with PBS, and re-centrifuged. The final pellet was resuspended in PBS and stored at −80 °C for further analysis. Transmission electron microscopy (TEM, Hitachi H7800, Japan) was utilised to observe the morphology of eMSC-derived exosomes. The size and concentration of the exosomes were determined using nanoparticle tracking analysis (NTA, Nanosight 300, UK). Western blot analysis was performed on the lysates of eMSCs and eMSC-derived exosomes to detect the expression of positive exosomal markers (CD9, CD63, and TSG101) and the negative marker calnexin (Proteintech, China). For the cell internalisation assay, eMSC-derived exosomes were labelled with PKH26 lipid dyes (red) (Yeasen Biotechnology, China) according to the manufacturer's protocol. HUVECs or paraformaldehyde (PFA) (4 % PFA in PBS)-treated HUVECs were then co-cultured with labelled eMSC-derived exosomes or an equal volume of PBS in basal ECM containing 5 % Exo-FBS at 37 °C for 48 h. Following this, HUVECs were fixed, and the nuclei were counterstained with DAPI (Biosharp, China). Imaging of the cells in the plate was subsequently performed using an inverted fluorescence microscope (Olympus IX71, Japan). To directly observe the transfer of miRNA by exosomes, eMSCs were transfected with Cy3-labelled miR-4466. The exosomes derived from eMSCs were labelled with PKH67 lipid dyes (green) and incubated with HUVECs for 24 h. Images were captured using an inverted fluorescence microscope. To simulate exosome-mediated intercellular communication between donor and recipient cells, an in vitro indirect co-culturing system was established, with eMSCs and HUVECs inoculated, respectively, in the upper and lower chambers of a 6-well Transwell co-culture system (0.4 μm pore with a polyester membrane, LABSELECT, China) with the corresponding complete culture medium at 37 °C. When the cells reached approximately 60–80 % confluence, they were co-cultured at 37 °C for 48 h in DMEM/F12 containing 2 % Exo-FBS. To investigate whether eMSCs promote angiogenesis through exosomes, eMSCs were pre-treated with or without the exosomal inhibitor GW4869 (Yeasen Biotechnology, China) at 37 °C for 0.5 h before co-culturing, as described above. To further study the effect of eMSC-derived exosomes on HUVECs, the eMSC-derived exosomes or an equal volume of PBS were co-cultured with HUVECs at 37 °C for 48 h in DMEM/F12 containing 2 % Exo-FBS. After co-culturing, the recipient HUVECs were harvested and used for further assays. Cell proliferation were assessed by using an EdU incorporation assay kit (RiboBio incorporation, Guangdong, China) according to the manufacturer's instructions. Briefly, 2 × 10 4  cells per well were seed into 48-well plates for 48 h, and then 50 μM of EdU was added to each well and cultured for additional 2 h. The cells were fixed with 4 % formaldehyde (PFA) for 30 min and treated with 0.5 % Triton X-100 for 30 min. After washing trice with PBS, 200 μL of Apollo reaction cocktail (1 × ) was added and incubated for 30 min in the dark. After incubating with 100 μL of Hoechst 33342 for 30 min, the cells were visualized and captured under an inverted fluorescence microscope (Olympus, IX-71, Tokyo, Japan). The percentage of EdU-positive cells was subsequently analysed using Image J software (NIH, Bethesda, MD). All experiments were done in triplicate and three independent repeating experiments were performed. The invasion and migration capacities of HUVECs were evaluated using the Matrigel invasion assay and wound healing assay, respectively. For the Matrigel invasion assay, HUVECs were seeded onto the upper chamber (8 μm pore polycarbonate membrane, LABSELECT, China) pre-coated with Matrigel matrix (BD Biosciences, USA) in basal ECM. The lower chamber was filled with complete ECM. After 24 h of incubation, the invasive cells were fixed with methanol, stained with crystal violet, and quantified using Image J. For the wound healing assay, HUVECs were resuspended in complete ECM medium and seeded into a 2-well culture insert (ibidi, Germany). Once the cells reached full confluence, the inserts were removed to observe and record the distance of cell migration. The relative distance of the cells to the scratched area was measured using Image J software, and the relative cell migration rate (%) was calculated as follows: relative distance of cell migration/original distance of the scratched area × 100 %. The tube formation capacity of HUVECs was assessed using the μ-slide angiogenesis assay. HUVECs were seeded into the inner wells of a μ-Slide 15-well 3D (ibidi, Germany) pre-coated with Matrigel and cultured for 4 h. After staining with calcein-acetoxymethyl (calcein-AM) (Yeasen Biotech, China) for 15 min, tube formation was observed and recorded using a fluorescence inverted microscope. The total length of vessels was measured using Angio Tool software [ 16 ]. Total RNA was isolated using TRIzol reagent (Invitrogen, USA), and complementary DNA (cDNA) was synthesised using either the HiScript® III RT SuperMix kit or the miRNA 1st Strand cDNA Synthesis Kit. qRT-PCR was performed using Taq Pro Universal SYBR qPCR Master Mix. All reagents for the qRT-PCR were purchased from Vazyme (Vazyme Biotech Co., Ltd, China). The relative expression levels of mRNA or miRNA were determined using the 2 −ΔΔCt method. GAPDH was used as an internal reference for mRNA, and U6 was used as an internal reference for miRNA. All primer sequences are listed in Supplementary Tables 2 and 3 Cellular or exosomal protein samples were extracted using Cell Lysis Buffer (Beyotime, China) supplemented with a phosphatase and protease inhibitor cocktail (Thermo Fisher Scientific, USA). Protein concentrations were quantified using the bicinchoninic acid (BCA) assay (Thermo Fisher Scientific, USA), and 5–20 μg of protein was loaded onto 10 % or 12 % sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were then transferred onto polyvinylidene difluoride (PVDF) membranes (Roche, Switzerland) and blocked with 5 % skimmed milk in Tris-buffered saline (TBS) or PBS with 0.1 % Tween-20 at room temperature for 2 h. Subsequently, the membranes were incubated overnight with primary antibodies at 4 °C, followed by incubation with a corresponding horseradish peroxidase (HRP)-conjugated secondary antibody at room temperature for 2 h. Finally, protein bands were visualized using the LumiGLO® chemiluminescent detection kit (Cell Signaling Technology, Boston, MA, USA). All antibodies used for western blotting are listed in Supplementary Table 4 . Grayscale quantitative analysis was conducted using Image J software. Extracellular VEGFA levels in the supernatants of HUVECs and HEK-293T cells were quantified using the Human VEGFA ELISA Kit (Elabscience, TX, USA) according to the manufacturer's instructions. Supernatants were centrifuged to remove debris and applied to ELISA plates pre-coated with anti-VEGFA antibodies. Following incubation with detection reagents and substrate, absorbance was read at 450 nm. VEGFA concentrations were determined from a standard curve generated with recombinant VEGFA. All samples were measured in triplicate, with data expressed as mean ± standard deviation (SD). For the sequencing of exosomal miRNA, cell culture supernatants from N-eMSCs (n = 3) and Ec-eMSCs (n = 3) were collected as previously described. The analysis was conducted by Wayen Biotechnologies (Shanghai, China) following standard procedures. Total miRNA of exosomes was extracted by miRNeasy Micro Kit (Qiagen, USA). The quality of the extracted RNA was assessed with the Agilent 4200 TapeStation (Agilent Technologies, USA). MiRNA library construction was performed using the QIAseq miRNA Library Kit (Qiagen, Germany), followed by miRNA-seq on the Illumina Next Generation Sequencing (NGS) system, specifically the NovaSeq™ 6000 (USA). Differentially expressed miRNAs (DE-miRNAs) were identified using the edgeR package of R software with cutoff criteria set at |log2 fold change (log2FC)| > 1 and P value < 0.05. The results were visualized through rainbow plots, volcano plots, and heatmaps. The raw miR-seq data have been uploaded to the GEO database (GSE261030). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed using clusterProfiler package (version 4.4.2). The putative target genes of candidate miR-4466 were identified by overlapping four databases (miRDB, miRWalk, miRPathDB, and mirDIP). The intersecting genes were presented using EVenn ( http://www.ehbio.com/test/venn ). The interaction between miRNA and its target gene was predicted using TargetScan software (version 8.0) by searching for conserved or non-conserved 6- to 8-mer sites that match the seed region of miR-4466. The miRNA sequences and overexpressed RUNX1 plasmid (pcDNA3.1-RUNX1) were synthesised by GenePharma (Shanghai, China). HUVECs were seeded in 24-well or 6-well plates and incubated overnight in basal ECM (serum-free and antibiotic-free). In the cell transfection experiments, miR-4466 mimics and inhibitors (GenePharma, China) were used to specifically upregulate or suppress miR-4466 expression, respectively. The corresponding negative controls (NC-mimics and NC-inhibitors) are scrambled sequences with no known targeting effects in human cells. Transfection was performed with miR-4466/negative control (NC)-mimics (20 nM), miR-4466/NC-inhibitor (20 nM), 50 ng plasmid (pcDNA3.1-RUNX1 or an empty pcDNA3.1 plasmid), either alone or in combination, using Lipofectamine® 3000 Transfection Reagent (Invitrogen). After transfection, HUVECs were harvested at the indicated time points for further analyses. The sequences for miRNA mimics and inhibitors used are listed in Supplementary Table 5 . Wild-type (Wt) and mutant (Mut) 3′UTR sequences of RUNX1 were designed, synthesised, and inserted into the GP-miRGLO dual-luciferase vector (Genepharma, China). The recombinant reporter vectors (50 ng) were separately co-transfected with miR/NC-mimics (20 nM) into HEK-293T cells or HUVECs in a 24-well plate. At 48 h post-transfection, luciferase activity was assessed using the Dual-Luciferase Reporter Assay System (GenePharma). The firefly luciferase (FLU) activity was normalized to the Renilla luciferase activity (RLU) and presented as relative luciferase activity. All statistical data were analysed using GraphPad Prism (version 10.1.0; GraphPad Software, San Diego, CA, USA). Experimental data were presented as mean ± standard deviation (SD) with three biological replicates. Continuous variable use Student's t -test or Mann-Whitney U, Categorical variable were compared between groups with use of the χ 2 test or Fisher's exact test, as appropriate. Statistical significance was set at P  < 0.05 for all statistical analyses.

Abnormal angiogenesis is a key factor in the initiation of metastasis and ectopic growth of endometrial cells, contributing to the symptoms of abnormal uterine bleeding (AUB) and infertility in patients with adenomyosis (AM). However, the precise molecular mechanisms underlying abnormal angiogenesis in AM remain unclear. While previous studies have implicated exosomes in adenomyosis progression, the role of ectopic eMSC-derived exosomes, particularly their miRNA cargo, in aberrant angiogenesis remains unexplored. In this study, we demonstrated that Ec-eMSCs have a significant impact on the angiogenesis process via paracrine signalling, with their pro-angiogenic effect primarily mediated by exosomes. Mechanistically, miR-4466 promotes angiogenesis by directly targeting RUNX1 to activate VEGFA signalling. These findings may provide a novel therapeutic target for the treatment of AM ( Fig. 8 ). Fig. 8 Schematic diagram of the mechanism of Ec-eMSCs-derived exosomal miR-4466 promotes angiogenesis by targeting RUNX1 in AM. The graphical abstract was created with BioRender.com ( https://biorender.com/ ). Fig. 8 Schematic diagram of the mechanism of Ec-eMSCs-derived exosomal miR-4466 promotes angiogenesis by targeting RUNX1 in AM. The graphical abstract was created with BioRender.com ( https://biorender.com/ ). Previous research has highlighted the importance of abnormal differentiation of eMSCs and alterations in the niche environment in the development and progression of AM [ 17 ], although the specific mechanisms remain unclear. Traditionally, MSCs interact with neighbouring cells primarily through paracrine signalling, involving the release of factors such as VEGFA, hepatocyte growth factor (HGF), insulin-like growth factor (IGF), miRNAs, and exosomes [ [18] , [19] , [20] ]. A recent study showed that perivascular eMSCs from normal endometrium can effectively induce neovascularisation by releasing the pro-angiogenic factor VEGFA [ 21 ]. Furthermore, Domnina et al. demonstrated that eMSCs derived from menstrual blood enhance the paracrine secretion of VEGFA, thereby facilitating wound closure in three-dimensional configurations [ 22 ]. While previous research has established that eMSCs can exhibit pro-angiogenic properties through paracrine signalling [ 23 ], studies specifically examining the paracrine effects of Ec-eMSCs are limited. One study showed that co-culture with vascular smooth muscle cells (SMSCs) modifies the miRNA content in exosomes derived from endothelial cells (ECs), thereby facilitating vascular maturation [ 24 ]. Another study highlighted that MSCs enhance angiogenesis through the secretion of exosomes that deliver pro-angiogenic miRNAs and modulate target genes in recipient cells [ 25 ]. Consequently, we hypothesised that Ec-exo deliver miRNAs to HUVECs and mediate angiogenesis. In this study, we established a co-culture system to analyse the paracrine pro-angiogenic effects of eMSCs on HUVECs. Our results indicated that Ec-eMSCs have a potent effect on angiogenesis via paracrine signalling. Blocking experiments with GW4869 further demonstrated that the paracrine pro-angiogenic effect of Ec-eMSCs is primarily mediated by exosomes. Notably, Ec-exo exert a pronounced pro-angiogenic effect on HUVECs in vitro. These results provide evidence that Ec-eMSCs exert their paracrine pro-angiogenic effect through exosomes, contributing to abnormal angiogenesis in AM. The findings above prompt an investigation into the potential mechanisms underlying the pro-angiogenic effect of Ec-exo. Exosomes are known to harbour a variety of bioactive substances, highlighting their role in various diseases. Extensive evidence suggests that exosomes can compromise vascular integrity and increase vascular permeability, ultimately promoting tumour progression and metastasis [ 26 ]. Additionally, exosomes have been implicated in cardiovascular diseases through the transfer of metabolites, particularly miRNAs, which facilitate intercellular communication [ 27 ]. However, there are only a few reports on the role and mechanisms of eMSC-derived exosomes in AM. Jiang et al. found that exosomes derived from eutopic endometrial cells can promote the proliferation and migration of adenomyotic myometrial cells via the IL-6/JAK2/STAT3 pathway [ 12 ]. Nevertheless, the specific mechanism by which exosomes exert their effects remains unclear. Recently, a study showed that miR-25-3p found in extracellular vesicles (EVs) from eutopic endometrium in AM could induce macrophage polarization towards the M2 phenotype, subsequently promoting epithelial-mesenchymal transition (EMT) in endometrial cells [ 28 ]. Notably, unlike miR-25-3p from eutopic endometrium-derived EVs that promote EMT via macrophage polarization, miR-4466 from Ec-exo directly targets RUNX1 to enhance VEGFA-driven angiogenesis, highlighting a distinct mechanism. These findings suggest that Ec-eMSCs might exert their pro-angiogenic effects through exosomal miRNAs. As anticipated, GO and KEGG pathway enrichment analyses indicated that the potential target genes of DE-miRNAs in Ec-eMSCs are involved in the positive regulation of angiogenesis and the VEGF signalling pathway. Notably, among these DE-miRNAs, miR-4466 was the most significantly upregulated in Ec-exo. However, current evidence remains limited with only a single epidemiological study to date having identified miR-4466 along with four specific miRNA species (miR-4516, miR-6090, miR-4763-3p, and miR-4281) as independent genetic risk factors for attention-deficit/hyperactivity disorder (ADHD) [ 29 ]. Strikingly, our findings systematically validated that miR-4466 encapsulated within exosomes derived from Ec-eMSCs can be efficiently internalised by HUVECs via exosome-mediated intercellular communication. Furthermore, functional assays demonstrated that miR-4466 exerts a pronounced pro-angiogenic effect on HUVECs in vitro, as evidenced by enhanced endothelial migration/invasion, tube formation capacity. As is well known, VEGFA is one of the most important regulators of tumour-associated angiogenesis in cancer [ 30 ]. We observed that overexpression of miR-4466 significantly increased both intracellular and extracellular VEGFA expression, while inhibition of miR-4466 had the opposite effect. These results indicate that the expression of pro-angiogenic VEGFA is dependent on miR-4466. To our knowledge, this is the first study to demonstrate the pro-angiogenesis role of exosomal miR-4466 from Ec-eMSCs in AM. To further elucidate the mechanism by which miR-4466 influences angiogenesis, we first predicted potential target genes of miR-4466 through bioinformatics analysis. These predictions were then validated in HUVECs and HEK-293T cells overexpressing or inhibiting miR-4466. Notably, we observed that only RUNX1 expression was significantly negatively correlated with miR-4466 expression. The direct interaction between miR-4466 and RUNX1 was subsequently confirmed through a dual-luciferase reporter assay. Collectively, these results indicate that RUNX1 is a direct target of miR-4466. RUNX1, a well-characterised member of the RUNX family, is involved in various cellular processes, including growth, invasion, migration, apoptosis, angiogenesis, and tumourigenesis [ 31 ]. Specifically, RUNX1 has been identified as a transcriptional inhibitor of VEGFA, which plays a role in wound healing and regulates critical processes in cancer progression, including cell proliferation, metastasis, and angiogenesis [ [32] , [33] , [34] ]. However, it remains unclear whether miR-4466 regulates VEGFA expression through RUNX1 in HUVECs. Our rescue experiments validated that the pro-angiogenic effect of miR-4466 is dependent on RUNX1. Importantly, we confirmed that miR-4466 induces VEGFA expression in a RUNX1-dependent manner. Similar to our findings, previous studies have reported that chondrocyte-derived exosomal miR-214-3p can activate VEGFA by interacting with RUNX1, thereby promoting angiogenesis in HUVECs [ 35 ]. These findings suggest that targeting the miR-4466/RUNX1/VEGFA axis may offer a novel anti-angiogenic strategy for adenomyosis, potentially reducing abnormal uterine bleeding and improving fertility outcomes. Nevertheless, this study has several limitations that need addressing. First, despite these findings, this study was limited to in vitro experiments, and in vivo experiments have not yet been conducted. Second, although we focused on miR-4466, other upregulated miRNAs in Ec-exo may also contribute to angiogenesis or other pathological processes, warranting further investigation. Third, we did not explore the potential effects of miR-4466 on other phenotypic characteristics of HUVECs, such as apoptosis or adhesion. Fourth, the use of HUVECs, while standard for angiogenesis assays, may not fully recapitulate the endometrial endothelial microenvironment. Future studies using patient-derived endothelial cells or in vivo models would strengthen our conclusions. Finally, this study assessed the functional role of primary adenomyotic eMSC-derived exosomal miR-4466 primarily in HUVECs. Future studies should use in vitro AM organoid or in vivo AM animal models to further validate it. In summary, our study not only unveils a novel exosomal miRNA-mediated mechanism in adenomyosis-associated angiogenesis but also provides a foundation for miRNA-based therapeutics aimed at mitigating vascular abnormalities in this debilitating condition.

Adenomyosis (AM) is a common gynecological disorder characterized by the infiltration of the endometrium into the myometrium, leading to abnormal uterine bleeding (AUB), dysmenorrhoea, and infertility, severely impacting the quality of life for women of reproductive age [ 1 , 2 ]. Existing evidence indicates a substantial increase in microvascular density (MVD) in both eutopic and ectopic endometrium, which may lead to the excessive formation of superficial, disorganised, and hyperpermeable vessels [ 3 , 4 ]. Abnormal angiogenesis accelerates disease onset and progression, exacerbating symptoms in AM patients [ [4] , [5] , [6] ]. However, the precise mechanisms underlying abnormal angiogenesis in adenomyotic lesions remain poorly understood. Perivascular human endometrial mesenchymal stem cells (eMSCs), located around blood vessels, play a critical role in the pathogenesis of AM [ 5 , 7 , 8 ]. Although eMSCs derived from ectopic lesions have demonstrated enhanced migratory and invasive capacities, potentially contributing to the progression of AM [ 9 ], their role in abnormal angiogenesis remains unclear. Recent research has shown that eMSCs possess pro-angiogenic properties through the paracrine and autocrine effects of VEGFA [ 10 ]. It is well established that exosomes exert paracrine effects by facilitating intercellular communication through their contents, such as proteins, mRNAs, and microRNAs (miRNAs), among others [ 11 ]. Exosomes derived from primary endometrial cells have been found to enhance the proliferation and migration of adenomyotic myometrial cells [ 12 ]. Furthermore, extracellular vesicles (EVs) derived from primary endometrial cells of the eutopic endometrium in patients with AM can induce the epithelial-mesenchymal transition (EMT) process in endometrial epithelial cells (EECs) [ 13 ]. Similarly, Wang et al. demonstrated that adenomyotic lesion-derived EVs induce EMT and enhance the invasion of EECs [ 14 ]. However, there are currently no reports investigating the role and mechanism of ectopic eMSC-derived exosomes (Ec-exo) in the abnormal angiogenesis of AM. In this study, we compared the differential expression of miRNAs in normal eMSC-derived exosomes (N-exo) and Ec-exo using miRNA sequencing. Exosomal miR-4466 was selected as a candidate miRNA to explore the angiogenic role and potential mechanisms of Ec-exo. In conclusion, our study elucidated that ectopic eMSC-derived exosomal miR-4466 promotes angiogenesis by targeting the RUNX1/VEGFA axis in AM. These findings may offer new insights into therapeutic targets and strategies for the anti-angiogenic treatment of AM.

The authors declare no competing interests.