Abstract
In aging and diseased retina, Bruch's membrane (BrM) stiffens heterogeneously, impairing phagocytic clearance of photoreceptor outer segments by the overlying retinal pigment epithelial (RPE) monolayer and contributing to retinal degeneration. While substrate stiffness is known to influence RPE phagocytosis, conventional assays lack the spatial resolution to capture how mechanical cues coordinate behavior across the monolayer over time. Here, dynamic magnetorheological elastomer (MRE) substrates are combined with an automated machine-learning image-analysis pipeline to simultaneously quantify particle engagement kinetics, cell morphology, and spatial organization at single-cell resolution. This combination enables spatiotemporal observation of RPE mechanosensitivity inaccessible to population-level assays. Substrate stiffness shaped particle engagement kinetics, modulating engagement rate and maximum particle capacity. Soft substrates additionally increased cell area and eccentricity, suggesting stiffness-dependent cytoskeletal remodeling. Strikingly, cells with increased levels of particle engagement form spatially localized clusters whose organization was sensitive to substrate stiffness, revealing that mechanical cues coordinate engagement behavior at the tissue level rather than acting on cells independently. This work reveals how dynamic BrM stiffening drives coupled changes in particle engagement kinetics, cell morphology, and spatial coordination, highlighting the tissue-level complexity of RPE mechanosensitivity. Together, these findings reframe RPE particle engagement as a spatially coordinated, mechanosensitive process, providing new insight into how dynamic alterations in extracellular mechanical environments regulate cellular function. STATEMENT OF SIGNIFICANCE: The retinal pigment epithelium (RPE) clears cellular debris daily to maintain vision, a process that falters as the underlying tissue stiffens with age, a hallmark of macular degeneration. Existing tools capture only population averages on static surfaces, obscuring both the timing and spatial organization of the response. This study combines dynamically tunable substrates with automated machine-learning image analysis to quantify receptor-mediated particle engagement at single-cell resolution across time and space. Mechanical transitions, not stiffness magnitude alone, accelerate engagement rate while reducing total capacity. Cell shape predicts particle engagement activity in a stiffness-dependent manner, and highly active cells cluster spatially in patterns governed by the mechanical environment. These findings reframe RPE dysfunction as a spatially coordinated, mechanosensitive process with direct implications for early AMD.