Abstract
Age-related macular degeneration (AMD) is a leading cause of irreversible vision loss worldwide and is driven in part by dysfunction of the retinal pigment epithelium (RPE) and its supporting extracellular matrix, Bruch’s membrane (BM). BM is a thin, specialized matrix that regulates structural support and molecular transport across the outer blood-retinal barrier, yet it undergoes age-related thickening and biochemical alteration that contribute to RPE dysfunction and retinal degeneration. Despite its importance, current in vitro AMD models do not adequately recapitulate the structural and biochemical complexity of native BM, limiting their physiological relevance and translational utility.
In this study, we developed a tissue-engineered BM-mimetic platform using electrospun polyethylene glycol (PEG) hydrogel scaffolds functionalized with adhesive ligands and evaluated their ability to support RPE growth and function. Scaffolds were fabricated with three ligand conditions, RGD, Fn4G, and Fn9*, chosen to provide distinct cell-adhesive cues, and were electrospun for 15, 30, or 45 minutes to generate variation in scaffold architecture. Human ARPE-19 cells were seeded onto the scaffolds, and scaffold performance was assessed through structural characterization, cell attachment, proliferation, morphology, and ZO-1 localization as a marker of tight junction organization.
Fiber characterization demonstrated that scaffold architecture was influenced by both ligand type and electrospinning duration. Alignment varied significantly across conditions, with PEG-RGD and PEG-Fn9* scaffolds showing peak fiber alignment at 30 minutes, while PEG-only and PEG-Fn4G scaffolds showed greater alignment at 15 minutes followed by increased randomization over time. Fiber thickness also differed significantly among scaffold groups, particularly at earlier time points, although these changes did not directly correlate with alignment, suggesting that thickness and alignment contribute independently to scaffold structure.
Functional analyses showed that PEG-based scaffolds generally promoted improved RPE behavior relative to uncoated glass control surfaces. ZO-1 immunofluorescence revealed stronger junctional localization on multiple scaffold conditions, especially PEG-RGD at 15 minutes and PEG-Fn9* at 30 minutes. Compared with control, PEG-RGD at 15 minutes demonstrated a 59.5% increase in mean peak intensity and a 2.6-fold increase in peak-to-center ratio, while PEG-Fn9* at 30 minutes produced the highest peak-to-center ratio observed. Across all scaffold groups, PEG-based materials supported greater border-localized ZO-1 staining than control, indicating improved tight junction organization and a more epithelial phenotype.
Crystal violet analysis suggested that both ligand identity and scaffold architecture influenced early cell attachment, although variability within groups limited statistical significance. Proliferation studies showed that scaffold biomaterial composition significantly affected ARPE-19 growth, with PEG-RGD and PEG-Fn9* scaffolds generally supporting higher proliferation than control, while PEG-Fn4G maintained more moderate but stable growth. Morphological analysis further showed that all scaffold conditions yielded significantly lower aspect ratios than control, indicating a rounder, more cobblestone-like morphology consistent with native RPE phenotype.
Overall, these findings demonstrate that combining tunable electrospun PEG scaffold architecture with defined adhesive ligand presentation enhances RPE attachment, proliferation, morphology, and tight junction organization. This engineered platform offers a reproducible and physiologically relevant model of BM that may improve in vitro retinal disease studies and support future therapeutic development for AMD.
