**Background:** Choroidal ischemia (CI) is a pathological condition characterized by thinning and atrophy of the choroidal vasculature, which supplies 85% of ocular blood flow and is critical for outer retinal health. CI is implicated in dry age-related macular degeneration (AMD), myopic maculopathy, and other degenerative eye diseases. Current treatments are limited; for dry AMD, only Pegcetacoplan (Syfovre) was approved in 2023. Human pluripotent stem cell-derived endothelial cells (hPSC-ECs) have shown revascularization potential in other ischemic models. This study hypothesized that hPSC-ECs could repair choroidal vasculature and rescue atrophy in a rat CI model.
**Methods:** hPSC-ECs were differentiated from H1, H9 hESCs and hiPSCs using a two-stage, chemically defined system (AATS medium) with BMP4, CHIR99021, VEGF-A, and bFGF. Cells were characterized by immunostaining for CD31, CA4, RGCC, and PV-1; qPCR for choroidal EC markers; Dil-Ac-LDL uptake; tube formation; and SEM for fenestrations. A rat CI model was established by suprachoroidal injection of 10 µg NaIO3. CI was confirmed by H&E staining, immunofluorescence of choroidal flatmounts, and fluorescein/indocyanine green angiography (FA/ICGA). At 7 days post-CI, 5 × 10^4 hPSC-ECs, culture medium, or no injection were administered into the suprachoroidal space. Outcomes were assessed at 7, 14, 30, 60, and 90 days post-transplantation via FA/ICGA (vessel junction density using AngioTool), H&E (normalized choroid thickness, NCT), and immunofluorescence (human-specific hCD31 and pan-CD34). Electroretinography (ERG) was performed at 30 days. Single-cell RNA-seq was conducted on hPSC-ECs cocultured with ex vivo ischemic rat choroid for 48 hours.
**Key Results:** hPSC-ECs expressed choroidal EC markers (CD31, CA4, RGCC, PV-1) and formed fenestrations (100–200 nm). The CI model showed ~50% reduction in NCT at days 7, 30, and 90 (p < 0.00005) and significantly decreased vessel junction density (p < 0.0005). After transplantation, hCD31+ human ECs were detected in choroidal flatmounts up to 90 days, forming chimeric vessels with rat ECs. FA/ICGA revealed new vessel formation in the ischemic area from day 7 onward, with significantly higher junction density in EC-transplanted eyes versus medium-injected eyes at days 14–90 (p < 0.05 to p < 0.00005). NCT in EC-transplanted eyes returned to normal levels by day 30, while medium-injected eyes remained thin (p < 0.0005). ERG showed improved dark-adapted 0.01 a-wave and b-wave amplitudes in EC-transplanted rats compared to medium-injected rats (p < 0.05). No inflammation, hemorrhage, or choroidal neovascularization was observed. In contrast, UC-MSC transplantation caused severe intraocular inflammation. scRNA-seq of cocultured hPSC-ECs revealed upregulation of angiogenesis genes (VEGFA, HSPG2, EFNB2), immune-modulatory genes (ANXA1, ICAM1, IL1A), and neuroprotective genes (SEMA6A, SEMA5A, SEMA3F, GRN).
**Clinical Implications:** This study provides proof-of-concept that hPSC-ECs can regenerate choroidal vasculature and improve retinal function in a CI model that mimics dry AMD. The cells demonstrated low immunogenicity, integration into all choroidal layers, and no adverse effects up to 90 days. The paracrine secretion of neuroprotective and anti-angiogenic factors (e.g., SEMA3F, GRN) suggests additional benefits beyond vessel repair. These findings support further development of hPSC-EC transplantation as a potential therapy for choroidal ischemia in AMD and myopic maculopathy. Limitations include the small proportion of retained cells, unknown host tissue changes, and need for long-term safety monitoring for CNV risk.