Inhibition of Retinal Neovascularization by BEZ235: Targeting the Akt/4EBP1/Cyclin D1 Pathway in Endothelial Cells

Abstract

Purpose: To investigate the therapeutic efficacy of BEZ235, a dual PI3K/mTOR inhibitor, in suppressing pathological neovascularization in an oxygen-induced retinopathy (OIR) mouse model and explore the role of cyclin D1 in endothelial cell cycle regulation.

Methods: Single-cell RNA sequencing was performed to analyze gene expression and cell-cycle alterations in retinal endothelial cells under normoxic and OIR conditions. The effects of BEZ235 on human umbilical vein endothelial cells (HUVECs) and human retinal microvascular endothelial cells (HRMECs) were evaluated by assessing cell viability, cell-cycle progression, proliferation, migration, and tube formation. In the OIR mouse model, retinal neovascularization was evaluated by retinal flatmount immunofluorescence staining, hematoxylin and eosin (H&E) staining, quantitative reverse-transcription polymerase chain reaction (RT-qPCR), and western blot analyses. The in vivo toxicity of BEZ235 was evaluated by electroretinography (ERG) and histological examination of the heart, liver, spleen, lungs, and kidneys.

Results: In vitro, BEZ235 significantly inhibited cell cycle progression by downregulating cyclin D1 at both mRNA and protein levels, inducing G0/G1 phase arrest. This led to significant reductions in cell viability, proliferation, migration, and tube formation. In the OIR model, BEZ235 substantially decreased neovascularization and improved vascular organization. BEZ235 mediates its effects by inhibiting the PI3K/Akt/mTOR pathway, reducing Akt and 4E-binding protein 1 (4EBP1) phosphorylation levels, thus downregulating cyclin D1 expression. ERG and histological examination suggested that BEZ235 did not induce evident retinal or systemic toxicity at the dosage used to inhibit retinal neovascularization.

Conclusions: BEZ235 effectively inhibits retinal neovascularization by downregulating cyclin D1 via 4EBP1 phosphorylation inhibition, highlighting its potential as a promising therapeutic agent for retinal neovascularization diseases.

Figure 1.

Single-cell RNA-seq analysis of retinal cell populations in normoxia and OIR conditions from GSE150703. (A) t-SNE plot representing 31,271 retinal cells from the normoxia P14, normoxia P17, OIR P14, and OIR P17 groups, classified into 22 clusters. (B) t-SNE plots showing retinal cell distribution across samples (left) and by cell type (right). (C) Violin plots depicting the expression levels of key marker genes used to identify retinal cell types. (D) Stacked bar plots representing the proportion of cell-cycle phases (G1, G2/M, and S) by cell type and sample in normoxia and OIR conditions at P14 and P17. (E) Bubble plot illustrating the expression of cyclins and CDKs across different retinal cell types and samples. Notably, CCND1 is significantly upregulated in endothelial cells at OIR P17. The color of each dot indicates gene expression levels, ranging from red (high) to blue (low), and dot size represents the percentage of cells expressing the gene.

Figure 2.

PI3K/mTOR pathway modulation of CCND1-mediated endothelial cell-cycle progression. (A) Immunofluorescent staining of retinal flatmounts shows increased cyclin D1 expression in the OIR P17 group compared to the control P17 group, particularly colocalizing with CD31. Fluorescence intensity quantification is presented on the right (mean ± SEM; n = 4 per group). **P < 0.01. Scale bar: 25 µm. (B) Bubble plot illustrating phase-specific expression of cell-cycle–related genes in endothelial cell samples, with marked upregulation of CCND1 in the G1 phase of the OIR P17 group. (C) KEGG pathway analysis using GSEA in G1 phase endothelial cells (OIR P17 vs. normoxia P17), showing significant enrichment of pathways. (D) GSEA enrichment plot highlighting significant enrichment of pathways containing CCND1, particularly the PI3K/Akt signaling pathway. (E) t-SNE plot comparing CNVECs and CECs. (F) t-SNE plot showing CCND1 expression in CNVECs and CECs, with CCND1 upregulated in CNVECs. (G) Cell-cycle analysis of CNVECs revealed a coordinated peak in the expression of CCND1, PI3K, and mTOR pathways during the G1 to S phase transition.

Figure 3.

Effects of BEZ235 on viability, proliferation, and cell-cycle progression of HRMECs. (A) The CCK-8 assay was used to assess HRMEC viability after treatment with BEZ235 at concentrations of 2 nM, 4 nM, 8 nM, 16 nM, 32 nM, 64 nM, 128 nM, and 256 nM for 24 hours. The calculated IC₅₀ was 9.039 nM (mean ± SEM; n = 3 per group). (B) Trypan blue dead cell count revealed that the HRMEC death rate in the BEZ235-treated groups (1 nM and 10 nM) was not significantly different from that of the vehicle group (mean ± SEM; n = 3 per group); ns, not significant. (C) Flow cytometry showed that, after 24 hours of treatment with BEZ235 (1 nM and 10 nM), cell-cycle distribution was analyzed. BEZ235 at 10 nM significantly increased the percentage of cells in the G0/G1 phases and reduced the proportion in the S phase, suggesting G0/G1 cell-cycle arrest (mean ± SEM; n = 3 per group). **P < 0.01. (D) Ki67 immunofluorescence staining showed that BEZ235 at 10 nM significantly reduced the proportion of Ki67-positive cells compared to the vehicle group, indicating suppressed proliferation (mean ± SEM; n = 6 per group). ***P < 0.001. Scale bar: 100 µm.

Figure 4.

BEZ235 inhibits HRMEC migration, tube formation, and cyclin D1 expression. (A) Scratch assay showed that treatment with 10-nM BEZ235 significantly reduced the wound healing capacity of HRMECs compared to the vehicle group (mean ± SEM; n = 6 per group). ***P < 0.001. Scale bar: 500 µm. (B) The Transwell migration assay revealed that HRMECs treated with 10-nM BEZ235 exhibited significantly fewer migrated cells compared to the vehicle group (mean ± SEM; n = 4 per group). ***P < 0.001. Scale bar: 100 µm. (C) The tube formation assay showed that 10-nM BEZ235 significantly reduced the number of capillary-like tubes formed by HRMECs (mean ± SEM; n = 4 per group). ***P < 0.001. Scale bar: 500 µm. (D) Cyclin D1 immunofluorescence staining indicated that BEZ235 at 10 nM significantly decreased the proportion of cyclin D1-positive cells (mean ± SEM; n = 6 per group). ***P < 0.001. Scale bar: 100 µm.

Figure 5.

BEZ235 reduces Akt and downstream target 4EBP1 phosphorylation and downregulates cyclin D1. (A) RT-qPCR analysis showed that BEZ235 at 10 nM and 100 nM significantly reduced the mRNA expression levels of RPS6KB1 (S6K1) and CCND1 while not affecting EIF4EBP1 (4EBP1), as compared to the DMSO group (mean ± SEM; n = 4 per group). ***P < 0.001. (B) Western blot analysis indicated the impact of BEZ235 on the Akt/mTOR pathway, focusing on the key downstream targets of S6K1 and 4EBP1. BEZ235 treatment reduced Akt and 4EBP1 phosphorylation levels while also downregulating cyclin D1 expression. S6K1 phosphorylation remained unaffected despite a decrease in total S6K1 protein expression. (C) Quantification of western blot results showed the relative protein levels of p-Akt/Akt, p-S6K1/S6K1, p-4EBP1/4EBP1, cyclin D1/β-actin (mean ± SEM; n = 3 per group). *P < 0.05, ***P < 0.001.

Figure 6.

KEGG and Reactome pathway analyses of differentially expressed genes in HRMECs treated with BEZ235. HRMECs were treated with 10-nM BEZ235 for 24 hours, followed by RNA extraction and transcriptome sequencing. Differentially expressed genes were identified and subjected to KEGG pathway analysis and Reactome pathway analysis. KEGG pathway analyses of upregulated (UP) and downregulated (DOWN) genes are shown in A and B, respectively. Reactome pathway analyses of UP and DOWN genes are shown in C and D, respectively. The x-axis represents gene ratio, and the y-axis shows the significantly altered pathways. The results illustrate key pathway changes induced by BEZ235 in HRMECs, providing insights into its regulatory mechanisms.

Figure 7.

Impact of BEZ235 on retinal neovascularization in an OIR mouse model. (A) Isolectin B4 staining demonstrated vascular organization in retinal flatmounts from different groups. Yellow indicates avascular zones, and red indicates neovascular areas. BEZ235 treatment significantly mitigated both neovascular and avascular areas compared to vehicle (corn oil/DMSO) group, as quantified on the right (mean ± SEM; n = 4 per group). *P < 0.05, **P < 0.01. Scale bars: 1000 µm and 100 µm. (B) H&E staining shows retinal layers with marked neovascular growth in OIR mice, identified by red arrows as pathological neovascular cells. BEZ235 notably reduced preretinal neovascular cells compared to the vehicle group (mean ± SEM; n = 4 per group). **P < 0.01. Scale bar: 100 µm.

Figure 8.

Impact of BEZ235 on cyclin D1 and p-4EBP1 expression in OIR retinas. (A) RT-qPCR analysis shows increased CCND1 mRNA levels in OIR conditions, significantly reduced by BEZ235 treatment (mean ± SEM; n = 4 per group). *P < 0.0. (B) Western blot analysis revealed a reduction cyclin D1 protein level with BEZ235 treatment in OIR conditions, confirming its role in protein expression downregulation (mean ± SEM; n = 3 per group). *P < 0.05, ***P < 0.001. (C, D) With retinal section immunofluorescence, staining for CD31, p-4EBP1, and cyclin D1 in retinal sections demonstrated elevated expression in OIR conditions, significantly reduced in CD31-positive endothelial cells following BEZ235 treatment. Quantitative analysis is provided on the right (mean ± SEM; n = 4 per group). *P < 0.05, **P < 0.01. Scale bar: 100 µm.

Figure 9.

In vivo toxicity evaluation of BEZ235 using visual electrophysiology and H&E staining of visceral sections. Mice were administered BEZ235 or vehicle from P12 to P17. ERG was performed at P21, and histological examinations were conducted at P17. (A) ERG results showed that BEZ235 had no significant effect on retinal function in living mice (mean ± SEM; n = 6 per group). (B) H&E-stained sections of the heart, liver, spleen, lung, and kidney showed no evidence of systemic toxicity from BEZ235 treatment. Scale bar: 100 µm.