Proteomics approach identifies aqueous humor biomarkers in retinal diseases

Geographic atrophy (GA) secondary to age-related macular degeneration (AMD) is one of the main causes of blindness in the elderly population, but the molecular pathophysiology is difficult to study due to limited access to retinal tissue. We investigated aqueous humor (AH) as an accessible surrogate for studying retinal pathophysiology.

We applied affinity-based Olink proteomics on AH samples obtained from 30 non-AMD control, 30 intermediate AMD (iAMD) and 28 GA subjects to identify AH biomarkers associated with GA. Quantile normalization was applied to the Olink data, followed by differential abundance analysis using the limma R package. To contextualize our findings, we cross-referenced the identified proteins to gene expression datasets and AH proteomics data from diabetic retinopathy (DR) subjects.

Our differential abundance analysis reveals 82 significantly altered proteins in GA compared to non-AMD control. Cross-referencing with gene expression datasets indicates a majority of them are robustly expressed in the retina, particularly in retinal pigment epithelium cells. Comparison with AH proteomics data from DR subjects reveals both unique and shared biomarkers between GA and DR. Integrating these findings, we identify SMOC2 and IL-6 as top candidate GA biomarkers, warranting further investigation.

Our integrative analysis demonstrates a robust framework for AH biomarker discovery and identifies SMOC2 and IL-6 as promising biomarkers for GA. Our findings underscore the potential of AH proteomic profiling to advance our understanding of the underlying pathophysiology of retinal diseases.