Monocyte/macrophage-derived IL-15 activates STAT5 to trigger the EFNA1/NCOA2-positive feedback loop, facilitating retinal angiogenesis in high-glucose environments

Acquisition of public microarray data sets

This study used four microarray data sets from the Gene Expression Omnibus (GEO) database: GSE221521, GSE185011, GSE236333, and GSE179568. These data sets included blood and retinal tissue samples from healthy control groups, diabetic patients (DM group), and DR (DR group) patients. We downloaded the original count matrices (GSE236333 and GSE221521) and fragments per kilobase per million mapped fragments (FPKM) matrices (GSE185011 and GSE179568) and used R packages for data preprocessing, including background correction, normalization, and data filtering.

Initially, we obtained the original gene expression matrices from the GEO database for the following data sets: GSE221521, which includes blood samples from 50 healthy controls, 74 diabetic patients without retinopathy (DM group), and 69 DR (DR group) patients; GSE185011, containing peripheral blood mononuclear cell (PBMC) samples from 5 healthy controls and 5 patients with DR; GSE236333, studying the effects of the glucagon-like peptide 1 receptor agonist, liraglutide, on experimental DR in 6 healthy controls and 6 DR group retinal tissues; and GSE179568, including 7 retinal neovascularization (RNV) membranes (PDR), 10 macular puckers, and 7 macular holes samples. We used the R package limma function to analyze the differential gene sets between the DR group and the normal control group (Con) in the above data sets. We set a significant difference threshold of |log₂ fold change| >0.5 and p < 0.05 to screen for statistically significant differential genes.

Clinical sample collection

Human blood samples from 20 healthy individuals, 20 diabetic patients, and 20 patients with DR were collected from Hunan Children’s Hospital and Xiangya Hospital of Central South University. The clinical information of the participants is shown in Table 1. The blood samples were centrifuged at 1006 ×g for 10 min at 4 °C. This study was conducted according to the principles of the Declaration of Helsinki. Our research obtained approval from the Ethics Committee of Hunan Children’s Hospital after each patient signed an informed consent form.

Cell preparation and treatment

Human monocytic leukemia cells (THP-1) were purchased from the American Type Culture Collection (Manassas, VA). THP-1 cells have been identified by short tandem repeat (STR) analysis (Appendix 1) The process begins with the extraction of DNA from the cell sample, followed by a multiplex PCR amplification of multiple core STR loci using fluorescently labeled primers. The data are analyzed to determine the allele sizes and counts at each locus, producing a distinct STR profile. This profile is compared against reference databases (e.g., ATCC, DSMZ, JCRB, and RIKEN). STR analysis showed a match that confirms the cell line’s authenticity (Appendix 2). Macrophages were divided into two groups and stimulated and cultured for 48 h with glucose solutions at concentrations of 5 nM (Con) and 25 nM (HG). The level of IL-15 in culture medium was detected using the human IL-15 ELISA Kit (ab218266; Abcam, Cambridge, UK).

Human retinal microvascular endothelial cells (hRMECs) and human umbilical vein endothelial cells (HUVECs) were purchased from the American Type Culture Collection. hRMECs and HUVECs have been identified by STR analysis and immunofluorescence using von Willebrand factor as a biomarker of vascular endothelial cells (Appendix 2). Cells were maintained in a mixture of DMEM (glucose-free, #11966025; Thermo Fisher Scientific, Waltham, MA) and Ham’s F-12 Nutrient Mixture (10.0 mM d-glucose, #11765047; Thermo Fisher Scientific), containing normal-glucose concentrations (5 mM glucose, corresponding to normal human blood glucose levels), and were supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific) and 1% penicillin-streptomycin (P/S; Thermo Fisher Scientific) in a 37 °C, 5% CO₂ humidified incubator. hRMECs and HUVECs in the experimental group were cultured in medium containing high-glucose (HG) concentrations (25 mM glucose). In the coculture group, retinal endothelial cells were cocultured with THP-1 cells for 48 h.

Cell transfection

Small interfering RNAs (siRNAs) targeting STAT5, epithelial cell adhesion molecule 1 (EFNA1), and NCOA2 (siRNA-STAT5, siRNA-EFNA1, and siRNA-NCOA2) were synthesized by RiboBio (Guangzhou, China). The siRNA sequences are shown in Table 2. Before transfection, we prepared siRNAs at a working concentration (typically 50–100 nM) in transfection reagent (Lipofectamine 2000; Invitrogen, Carlsbad, CA). siRNA transfection reagents were added to the growth medium of hRMECs and HUVECs before incubation for 24 to 48 h. The full length of human EFNA1 DNA was synthesized by Sangon Biotech and cloned into the pcDNA3.1 vector (Takara, Shiga, Japan) to construct overexpression plasmids pcDNA-EFNA1. DNA Midiprep Kits (Thermo Fisher Scientific) were used for preparing the overexpression plasmids. Cell transfections were performed using Lipofectamine 3000 (Invitrogen). Cells were harvested 48 h posttransfection to conduct further experiments.

IL-15 knockout cells

To generate IL-15 knockout THP-1 cells, we employed the CRISPR/Cas9 technology. The human monocytic cell line THP-1 was used for this purpose. Initially, we designed and synthesized a single guide RNA (sgRNA) sequence targeting a conserved region within the coding region of the IL-15 gene. The sgRNA sequence is as follows: 5′-GGA GAA TCC ATT CCA ATA TA TGG-3′. The sgRNA sequence was cloned into a plasmid vector containing the Cas9 expression cassette, provided by RiboBio, and the plasmid was used according to the manufacturer’s instructions. For spCas9, the template for in vitro synthesis of guide RNA with T7 RNA polymerase can be prepared by annealing and primer extension of the following primers: IL15_guideRNA183fwT7crTarget: 5′-GAA ATT AAT ACG ACT CAC TAT AGG AGA ATC CAT TCC AAT ATA GTT TTA GAG CTA GAA ATA GCA AG-3′; IL15_guideRNAallT7common (constant primer used for all guide RNAs): 5′-AAA AGC ACC GAC TCG GTG CCA CTT TTT CAA GTT GAT AAC GGA CTA GCC TTA TTT TAA CTT GCT ATT TCT AGC TCT AAA AC-3′.

Following transfection of THP-1 cells with the CRISPR/Cas9 construct, knockout cells were selected using puromycin (2 µg/ml) for 7 days to enrich the population of successfully transfected cells. To generate a clonal cell line, the transfected pool was subsequently subjected to single-cell sorting by flow cytometry into 96-well plates. PCR and western blot assays were conducted to confirm the loss of IL-15 expression. The clone exhibiting complete IL-15 knockout was selected for subsequent experiments.

Flow cytometry assay

Human blood was collected into EDTA-treated tubes (Improve Medical, Guangzhou, China). PBMCs were isolated using Ficoll-Paque Plus density centrifugation (500g at room temperature without brake for 25 min). PBMCs were stained with phycoerythrin (PE)–conjugated anti-CD4 (1:100, #ab172730; Abcam), PE-conjugated anti-CD19 (1:400, #ab227688; Abcam), and PE-conjugated anti-CD163 (1:60, #ab182422; Abcam) antibodies at 4 °C for 1 h. The fluorescence of the stained cells was measured using a BD AccuriTM C6 cellometer (Becton, Dickinson, Franklin Lakes, NJ, USA). Data were analyzed with FlowJo software (Tree Star, Ashland, OR, USA).

RNA isolation and quantitative real-time PCR

Total RNA was extracted from the cells using TRIzol reagent (catalog number: 15,596,026; Invitrogen) and stored at −80 °C. After quantification of the RNA concentration with a NanoDrop 2000 spectrometer (Thermo Fisher Scientific), 1 μg total RNA was used to synthesize cDNA using a SureScript First-strand cDNA Synthesis Kit (QP057; GeneCopoeia, Guangzhou, China). A quantitative real-time PCR experiment was conducted with the QuantiTect Reverse Transcription kit (205,314; Qiagen, Hilden, Germany) in a 7500 Fast real-time PCR System (Thermo Fisher Scientific) under the following conditions: 94 °C for 10 min, followed by 40 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 40 s. Primer sequences are shown in Table 3. GAPDH was used as an internal control. Gene expressions were quantified with the 2^−ΔΔCT method [28].

Western blot analysis

Total proteins were extracted with RIPA lysis buffer (Sigma-Aldrich, St. Louis, MO), and the protein concentration in each extract was determined using a BCA quantification kit (Beyotime, Shanghai, China). Next, an equal amount of protein from each extract was separated by 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and the protein bands were transferred onto polyvinylidene fluoride membranes, which were subsequently blocked with 5% fat-free milk at room temperature for 2 h. Subsequently, the protein bands were probed with primary antibodies (Abcam) against IL-15 (1:200 dilution), p-STAT1 (1:200 dilution), STAT1 (1:200 dilution), p-STAT5 (1:200 dilution), STAT5 (1:200 dilution), EFNA1 (1:200 dilution), VEGF (1:500 dilution), p-AKT (1:500 dilution), p-AKT (1:500 dilution), p-Vav (1:200 dilution), NCOA2 (1:200 dilution), and β-actin (1:500 dilution) at 4 °C overnight, after which, the membranes were incubated with horseradish peroxidase–conjugated goat antirabbit secondary antibodies for 2 h at room temperature. The immunostained protein bands were detected with an enhanced chemiluminescence kit (Millipore, Burlington, MA). X-ray films were used for signal detection and quantification. To ensure the comparability of the western blot experimental results, all images were uniformly adjusted for brightness and contrast using Photoshop (Adobe, San Jose, CA) software.

Cell Counting Kit-8 assay

Cell viability was determined using the Cell Counting Kit-8 assay (CCK8; Dojindo, Kumamoto, Japan). Each assay was conducted in triplicate at 37 °C and using 10 µl of the CCK8 solution. The absorbance of each assay well at 450 nm was detected with a microplate reader. Proliferation rate (%) = [OD(Treatment) – OD(Control)] / OD(Control) * 100%.

5-Ethynyl-2′-deoxyuridine staining assay

Cells from different groups were cultured in 96-well plates for 24 h at 37 °C. The following day, the cells were incubated for 2 h with 50 µl 5-ethynyl-2′-deoxyuridine (EdU; Beyotime) solution at 37 °C and then fixed for 15 min with 4% paraformaldehyde at room temperature. Next, the cellular nuclei were stained with DAPI solution (Sigma-Aldrich), and images of EdU-positive cells were acquired under a microscope.

Wound-healing assay

A wound-healing assay was performed to assess cell migration capacity as follows: cells from different groups (5 × 10⁵ cells/well) were seeded in a 6-well plate and starved for 24 h. Subsequently, cells were incubated with 10% FBS containing medium, and a plastic tip was used to create wounds across the cell monolayer. After 24 h, wound closure was photographed under an inverted microscope (magnification, ×100). The inhibitory rate was estimated according to the wound closure in each group.

Tube formation assay

Frozen vials of Matrigel membrane matrix (Thermo Fisher Scientific) were thawed on ice at +4 °C overnight. Matrigel was added (50 µl/well) into 96-well plates (TPP Techno Plastic Products AG, Trasadingen, Switzerland) and allowed to polymerize at 37 °C for 30 min. Meanwhile, cells were detached from wells/flasks using trypsin-EDTA (Gibco, Life Technologies, Carlsbad, CA). Using basal Vasculife medium (2% fetal bovine serum/FBS) with or without IL-6/sIL-6R (100 ng/ml each), suspension of cells was prepared, and 10⁴ cells per well were added to the polymerized Matrigel in duplicate. The plates were placed in an IncuCyte S3 Live Cell Analyses System (Sartorius AG, Göttingen, Germany), and the tube formation progress was monitored over 24 h by taking pictures every hour (three images per well). The images (10× magnification) were subsequently analyzed using the Angiogenesis-Analyzer macro written for ImageJ (National Institutes of Health, Bethesda, MD).

Co-immunoprecipitation

The anti-STAT5, anti-NCOA2, and normal IgG were added to 1-mg cell lysates precleared with Protein G Agarose beads (Sigma-Aldrich) and then incubated at 4 °C overnight. The co-immunoprecipitation complexes were harvested and analyzed by western blot analysis.

Chromatin immunoprecipitation assay

The relationship between STAT5 protein and the EFNA1 promoter was determined by performing a chromatin immunoprecipitation assay with a chromatin immunoprecipitation (ChIP) assay kit (Upstate Biotechnology, Lake Placid, NY). In brief, the cells transfected with STAT5 or vector at 80% to 90% confluency were incubated for 15 h with 0.1 μg/ml doxycycline and cross-linked in 1% formaldehyde. After sonication, the precleaned supernatants were incubated with streptavidin-coated magnetic beads conjugated with human anti-STAT5 or anti-IgG antibody (Sigma-Aldrich) for 20 min. The beads were washed, digested with proteinase K, and used for extracting eluted DNA. Next, the enrichment of EFNA1 was analyzed via PCR for the ChIP assay.

Statistical analysis

All statistical data were analyzed using GraphPad Prism 8 software (GraphPad Prism, La Jolla, CA), and quantitative data are presented as the mean ± standard deviation. Comparisons between two groups were performed using the Student t test, and comparisons between multiple groups were performed by one-way analysis of variance, followed by Dunnett’s post hoc test. A p value <0.05 was considered statistically significant.

PBMCs may influence gene expression in the retinal tissues of patients with DR

We employed the DR-related data sets GSE236333 and GSE179568 for an intersection analysis. This analysis led to the identification of 72 genes that were upregulated in the retinal tissues across both data sets. In a similar manner, we performed an intersection analysis using the data sets GSE185011 and GSE221521 to screen for upregulated genes in the blood and PBMCs of patients with DR (Figure 1A, B). A total of 27 upregulated genes were found within the intersection area. To investigate whether the upregulated genes in the blood and PBMCs of patients with DR affect gene expression in retinal tissues, we constructed a protein-protein interaction network of these differentially expressed genes. This was achieved by using the STRING database in conjunction with the CytoNCA plugin in Cytoscape. The results demonstrated that three genes (ATM, IL15, and CD36) in peripheral blood samples exhibited the strongest associations with the most upregulated genes in the retinal tissues. These included the five most significantly upregulated genes: STAT1, CD274, STAT5A, TMP1, and FN1 (Figure 1C). Notably, STAT1 and STAT5A are well recognized as transcription factors. Through predictions from the CHEA and ENCODE databases, we discovered that SBNO2 and EFNA1 might serve as their downstream target genes (Figure 1D). Moreover, both of these genes were upregulated in the retinal tissues of patients with DR. Gene functional enrichment analysis (GO-BP) revealed that the 72 upregulated genes in the retinal tissues of patients with DR were associated with various biological processes, including cytokine production, angiogenesis, wound healing, and immune responses (Figure 1E). We also presented the expression levels of STAT5A, EFNA1, and STAT1 in data sets GSE236333 and GSE179568 (Figure 1F, H), as well as the expression of IL-15 in data sets GSE185011 and GSE221521 (Figure 1G, I). Collectively, we hypothesized that upregulated genes, such as IL-15, in the blood and PBMCs of patients with DR influence the expression of genes such as STAT5A and STAT1 in retinal tissues. Subsequently, STAT5A and STAT1 induce the expression of EFNA1, which may be involved in certain pathological processes of DR.

Elevated expression of EFNA1 in retinal endothelial cells

This study analyzed a single-cell RNA sequencing data set, GSE245561, of patients with DR. Within the retinal fibrovascular membrane tissues, we successfully identified five distinct cell types: endothelial cells, T cells, stromal cells, macrophages, and epithelial cells (Figure 2A, B). Notably, we observed that the EFNA1 gene exhibited a high level of expression in endothelial cells (Figure 2C, D). To further elucidate the role of EFNA1 in endothelial cells, we partitioned the endothelial cells into two primary groups: a negative group consisting of 653 cells without EFNA1 expression and a positive group comprising 1,075 cells with EFNA1 expression (Figure 2E, F). Subsequently, based on the median expression level of EFNA1, we subdivided the positive group into low and high EFNA1 expression subgroups (Figure 2E, F). By applying the FindMarkers algorithm for differential gene analysis between these groups, we established a threshold for screening differential genes. The analysis revealed that between the high and low EFNA1 expression groups, there were 90 downregulated genes, 1,572 genes with stable expression, and 76 upregulated genes (Figure 2G). We then presented the top five genes (CXCR4, NCOA2, INHBB, EFNB2, and SAT1) in the high EFNA1 expression group and the top five genes (CDH11, DLC1, ACTN1, NR2F2, and PRCP) in the low EFNA1 expression group (Figure 2H, I). The expression of these genes was closely correlated with the expression of EFNA1 (Figure 2I), suggesting that these genes are under the regulatory control of EFNA1.

HG induced the expression of IL-15 in macrophages

In this study, we collected blood samples from healthy individuals, diabetic patients, and patients with DR and isolated PBMCs from them. IL-15 expression levels were significantly higher in PBMCs from patients with DR than in normal diabetic patients and healthy individuals (Figure 3A, B). PBMCs​ refer to cells with a single nucleus and mainly include lymphocytes and monocytes/macrophages. Lymphocytes are divided into T and B lymphocytes. To determine which cell type had the most pronounced IL-15 upregulation in patients with DR, we flow-sorted PBMCs (Figure 3C) and measured IL-15 mRNA expression levels in the three cell types (Figure 3D). Lymphocyte marker CD4, pan-B lymphocyte marker CD19, and monocytes/macrophages marker CD163 all had enrichment levels above 90%, with monocytes/macrophages having the highest levels of IL-15. A monocyte/macrophage cell line (THP-1) was used in the study in vitro. THP-1 cells were treated with different concentrations of glucose solution, and the expression level of IL-15 was detected. There was a significant difference in the expression level of IL-15 between the cells treated with 5-nM and 25-nM concentrations (Figure 3E). THP-1 cells were stimulated with 5-nM and 25-nM concentrations of glucose solution. By western blot and the enzyme-linked immunosorbent assay, we found that HG stimulation significantly increased the production and secretion of IL-15 of THP-1 cells, consistent with clinical studies showing that IL-15 expression was elevated in monocytes/macrophages from patients with DR (Figure 3F, G).

IL-15 promotes the proliferation, migration, and angiogenesis of retinal vascular endothelial cells and interacts with the HG environment

To evaluate the effect of IL-15 on the function of retinal vascular endothelial cells (hRMECs and HUVECs) in an HG environment, we assessed the impact of different concentrations of IL-15 (0, 10, 50, 200, 1,000 nM) at different time points (0, 12, 24, 48 h) on endothelial cell viability using the CCK8 assay. The results showed that when treated with 200 nM IL-15 for 48 h, cell viability reached the highest level, so this condition was used for subsequent experiments (Figure 4A). Subsequently, we treated hRMECs and HUVECs with 25 nM glucose solution and 200 nM IL-15. The results showed an increase in cell proliferation after 48 h of treatment with IL-15 alone or HG and a further increase after combined treatment of the two (Figure 4B, C). The wound-healing assay revealed an increase in the migration capacity of hRMECs and HUVECs after treatment with IL-15 alone or HG and a further increase after combined treatment with IL-15 and HG (Figure 4D). To determine the effect of combined IL-15 and HG treatment on endothelial cell angiogenesis, we examined the ability of cells in each group to form tubes. The results showed that the tube formation ability of hRMECs and HUVECs was improved after treatment with IL-15 or HG alone, and it further improved after combined treatment of the two (Figure 4E). Moreover, treatment with IL-15 alone or in combination with HG significantly increased the mRNA expression levels of STAT1, STAT5, and EFNA1 in hRMECs and HUVECs (Figure 4F). With β-actin as the internal control, p-STAT1, STAT1, p-STAT5, STAT5, and EFNA1 protein levels were increased by IL-15 in hRMECs and HUVECs. HG increased STAT1 and EFNA1 but had a minor effect on the p-STAT1, p-STAT5, and STAT5 protein levels. Interestingly, cotreatment with IL-15 and HG exerted a stronger promoting effect on the protein levels of p-STAT1, STAT1, p-STAT5, STAT5, and EFNA1 than treatment with IL-15 alone. The results suggest that the effect of IL-15 on endothelial cells was more pronounced under HG conditions (Figure 4F, G). In conclusion, IL-15 promotes the proliferation, migration, and angiogenic capacity of retinal vascular endothelial cells, and an HG environment exacerbates these effects of IL-15.

MDM-secreted IL-15 promotes the proliferation, migration, and angiogenesis of vascular endothelial cells under HG conditions

To further investigate the effect of macrophage-secreted IL-15 on the function of hRMECs and HUVECs under HG conditions, we stimulated vascular endothelial cells and THP-1, a type of monocyte/macrophage (MDM), with HG solution separately, knocked out the IL-15 gene in HG-treated THP-1, and then cocultured the two cells. The knockout of the IL-15 gene in THP-1 was validated by PCR and western blot tests (Appendix 2). The experiment was divided into five groups: control group (Con), endothelial cells cocultured with THP-1 (MDM) group, endothelial cells cocultured with HG-stimulated THP-1 (HG MDM) group, endothelial cells cocultured with HG-stimulated THP-1 with IL-15 knockout (HG MDM–IL-15 KO) group, and endothelial cells treated with HG and cocultured with HG-stimulated THP-1 (HG + HG MDM) group. CCK8 and EdU staining results showed that after 48 h of coculture, endothelial cell proliferation did not change significantly in the MDM group, whereas endothelial cell proliferation in the HG MDM group was significantly increased. After IL-15 knockout, this increase was significantly reversed, suggesting that HG-stimulated macrophages promote endothelial cell proliferation by secreting IL-15. Compared to the HG MDM group, the proliferation ability in the HG + HG MDM group was also significantly increased, indicating that the HG environment of endothelial cells could promote cell proliferation (Figure 5A, B). In addition, the migration rate of endothelial cells was increased in the HG MDM group compared with the Con group, and this increase was significantly inhibited by IL-15 knockout. Endothelial cell migration was more pronounced in the HG + HG MDM group than in the HG MDM group (Figure 5C). The enhanced angiogenic ability of HG MDMs was reversed after IL-15 knockout. The angiogenic capacity of endothelial cells in HG + HG MDMs was also higher than that in the HG MDM group (Figure 5D). The levels of p-STAT1, STAT1, p-STAT5, STAT5, and EFNA1 were significantly increased in endothelial cells after coculture with HG-stimulated THP-1 cells, especially for STAT5, which showed the most significant increase. Knockout of IL-15 in HG-stimulated THP-1 markedly reversed the high expression of p-STAT1, STAT1, p-STAT5, STAT5, and EFNA1 in endothelial cells (Figure 5E, F). Overall, our experimental results indicate that macrophage-secreted IL-15 plays a key regulatory role in endothelial cell function under HG conditions. The absence of IL-15 can significantly reverse the enhancement of endothelial cell proliferation, migration, and angiogenesis induced by HG-stimulated macrophages.

Interference with STAT5 inhibits IL-15–induced EFNA1 expression, inhibiting the proliferation, migration, and angiogenesis of retinal vascular endothelial cells

To understand the effect of the STAT5/EFNA1 signaling axis on endothelial cell function, we treated hRMECs and HUVECs with IL-15 and transfected siRNA-STAT5 to knock down STAT5 expression. This study designed three siRNAs to knock down STAT5. The PCR test showed that siRNA1 had the highest efficiency to downregulate STAT5. siRNA1 was used in the following study for STAT5 knockdown (Appendix 2). After 48 h of culture under normal glucose conditions, we found that IL-15 treatment significantly increased the protein expression levels of p-STAT5, STAT5, and EFNA1 in hRMECs and HUVECs. After knocking down STAT5 expression, the increase in p-STAT5 and EFNA1 expression induced by IL-15 was significantly reversed (Figure 6A). As indicated by CCK8 testing, IL-15 treatment significantly improved the proliferation ability of hRMECs and HUVECs. After STAT5 expression was knocked down, the increase in endothelial cell proliferation induced by IL-15 was reversed (Figure 6B). Knockdown of STAT5 alone had a moderate effect on the viability of endothelial cells. EdU staining experiments showed consistent results from CCK8 testing (Figure 6C). In terms of cell migration ability, IL-15 treatment significantly promoted the migration of hRMECs and HUVECs, and the enhancement of migration ability was significantly reversed after STAT5 knockdown (Figure 6D, E). Finally, the angiogenesis experiment showed that IL-15 treatment enhanced the angiogenic ability of hRMECs and HUVECs, but this enhancement was significantly reversed after STAT5 knockdown, indicating that IL-15 regulates angiogenesis through the STAT5 pathway (Figure 6F). In summary, IL-15 promotes the proliferation, migration, and angiogenesis of retinal vascular endothelial cells by upregulating STAT5 protein.

HG and IL-15 promote the proliferation, migration, and angiogenesis of retinal vascular endothelial cells by the upregulation of EFNA1

As a downstream target of STAT5, EFNA1 has been shown to promote angiogenesis through interacting with molecules such as Vav and AKT. To further validate the role and mechanism of EFNA1 in DR, we treated hRMECs and HUVECs with HG stimulation and IL-15 and transfected them with siRNA-EFNA1. This study also designed three siRNAs to knock down EFNA1. The PCR test showed that siRNA1 had the highest efficiency to downregulate EFNA1. siRNA1 was used in the following study for EFNA1 knockdown (Appendix 2). After HG stimulation and IL-15 treatment, we found that the phosphorylation levels of AKT and Vav in hRMECs and HUVECs cells were significantly increased, and this increase in phosphorylation levels induced by HG and IL-15 was significantly reversed after EFNA1 expression was knocked down (Figure 7A). CCK8 and EdU staining experiments showed that HG and IL-15 treatment significantly improved the cell proliferation ability of hRMECs and HUVECs. However, when EFNA1 expression was knocked down, the increase in cell proliferation ability induced by HG and IL-15 was significantly inhibited (Figure 7B, C). The cell wound-healing assay indicated that HG and IL-15 treatment significantly promoted the migration ability of hRMECs and HUVECs, and this enhancement in migration ability was significantly reversed after EFNA1 knockdown (Figure 7D, E). Finally, the angiogenesis experiment showed that HG and IL-15 treatment significantly improved the tubule formation ability of hRMECs and HUVECs, and this enhancement in angiogenesis induced by HG and IL-15 was reversed after EFNA1 knockdown (Figure 7F). In summary, our experimental results indicate that IL-15 and HG promote the proliferation, migration, and angiogenesis of retinal vascular endothelial cells by upregulating EFNA1 expression. These findings further reveal the importance of the IL-15/EFNA1 pathway in DR and provide new potential targets for the treatment of vascular complications in DR.

NCOA2 as a transcriptional coactivator induces positive feedback regulation of STAT5/EFNA1

In Figure 2H, I, we screened the top 10 genes. Their expression was most closely correlated with EFNA1 expression. For example, a higher expression of NCOA2 was observed in the high-EFNA1-expression group, while a lower expression of NCOA2 was observed in the low-EFNA1-expression group. This suggested that NCOA2 expression was influenced by EFNA1. NCOA2 is a transcriptional coactivator that can enhance the binding efficiency of transcription factors to DNA, thereby synergistically regulating gene expression. We hypothesized that NCOA2 can conversely influence EFNA1 expression by acting as a transcriptional coactivator of some transcription factors, such as STAT5. To understand the mutual relationship between EFNA1 and NCOA2, we overexpressed EFNA1 and knocked down NCOA2 in hRMECs and HUVECs. Three siRNAs have been designed to knock down NCOA2. The PCR test showed that siRNA2 had the highest efficiency to downregulate NCOA2. siRNA2 was used in the following study for NCOA2 knockdown (Appendix 2). The PCR and western blot results showed that EFNA1 overexpression increased not only EFNA1 but also NCOA2 in hRMECs and HUVECs. After transfecting siRNA-NCOA2, the expression of both EFNA1 and NCOA2 in the cells was decreased. NCOA2 knockdown also reversed the increase in EFNA1 and NCOA2 expression in cells with EFNA1 overexpression. These results suggest a mutually promoting effect of EFNA1 and NCOA2 (Figure 8A, B). This study also knocked down NCOA2 in IL-15–treated endothelial cells. IL-15 increased the expression of EFNA1 in the cells. Knocking down NCOA2 partially reversed an IL-15–induced increase in EFNA1 expression (Figure 8C). A co-immunoprecipitation experiment showed that IL-15 treatment significantly increased the interaction between STAT5 and NCOA2, and this interaction was significantly weakened after NCOA2 knockdown. However, NCOA2 knockdown did not influence STAT5 protein level (Figure 8D). ChIP experimental results showed that IL-15 treatment significantly increased the binding of STAT5 protein to the DNA in the promoter region of the EFNA1 gene, and this binding was weakened after NCOA2 knockdown, indicating that NCOA2 can promote the binding of STAT5 to the EFNA1 gene (Figure 8E). Functionally, CCK8 and EdU staining experiments showed that IL-15 treatment could significantly increase the proliferation ability of hRMECs and HUVECs, and this increase in proliferation ability was significantly reversed after NCOA2 knockdown (Figure 8F, G). The cell wound-healing assay (Figure 8H, I) and angiogenesis experiment (Figure 8J) also showed similar results, indicating that the promoting effect of IL-15 was weakened after NCOA2 knockdown. In summary, our findings reveal that NCOA2 acts as a transcriptional activator to promote IL-15–induced STAT5 binding to EFNA1 and interacts with EFNA1 in a positive feedback manner, thereby regulating the proliferation, migration, and angiogenic ability of retinal endothelial cells.