Abstract
Age-related macular degeneration (AMD), particularly its non-neovascular (dry) form, is a progressive retinal disorder that causes central vision loss and substantial impairment in daily life. Inflammation and immune dysregulation are recognized as core drivers of AMD, yet the contribution of PANoptosis, a form of programmed cell death that integrates pyroptosis, apoptosis, and necroptosis, remains unclear. In this study, we integrated human single-cell transcriptomic and bulk microarray datasets from the retina and retinal pigment epithelium-choroid to characterize PANoptosis-related transcriptional changes in dry AMD. Dimensionality reduction, cell-type annotation, and PANoptosis gene-set scoring revealed a distinct PANoptosis signature enriched in AMD, with particularly strong activation in myeloid populations. By combining differential expression analysis with machine learning-based feature selection, we identified four PANoptosis-related genes (PON2, BNIP3, EPHB6, and TPD52) that robustly distinguished AMD from control samples and were associated with an altered immune microenvironment. Genetic instrument analysis further suggested a positive association between TPD52 expression and AMD risk. At the cellular level, our data highlighted macrophages, especially pro-inflammatory M1-like macrophages, as key coordinators of PANoptosis-related pathways in dry AMD. To validate these findings in vivo, we used a sodium iodate-induced mouse model of dry AMD and observed significant dysregulation of PON2, BNIP3, EPHB6, and TPD52 in the retina by RT-qPCR, consistent with the human transcriptomic results and supporting their involvement in retinal degeneration and inflammation. Together, these findings implicate PANoptosis as an important and previously underappreciated component of dry AMD pathophysiology, define a four-gene PANoptosis-related signature with diagnostic potential, and suggest new molecular targets for therapeutic intervention.