Plasma biomarkers of the amyloid pathway are associated with geographic atrophy secondary to age-related macular degeneration.

• Main Authors: Kameran Lashkari, Gianna C Teague, Ursula Beattie, Joanna Betts, Sanjay Kumar, Megan M McLaughlin, Francisco J López
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2020-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0236283

Geographic atrophy (GA) is an advanced form of dry age-related macular degeneration (AMD), in which local inflammation and hyperactivity of the complement pathway have been implicated in its pathophysiology. This study explores whether any surrogate biomarkers are specifically associated with GA. Plasma from subjects with GA, intermediate dry AMD and non-AMD control were evaluated in 2 cohorts. Cohort 1 was assayed in a 320-analyte Luminex library. Statistical analysis was performed using non-parametric and parametric methods (Kruskal-Wallis, principal component analysis, partial least squares and multivariate analysis of variance (MANOVA) and univariate ANCOVAs). Bioinformatic analysis was conducted and identified connections to the amyloid pathway. Statistically significant biomarkers identified in Cohort 1 were then re-evaluated in Cohort 2 using individual ELISA and multiplexing. Of 320 analytes in Cohort 1, 273 were rendered measurable, of which 56 were identified as changing. Among these markers, 40 were identified in univariate ANCOVAs. Serum amyloid precursor protein (sAPP) was analyzed by a separate ELISA and included in further analyses. The 40 biomarkers, sAPP and amyloid-β (Aβ) (1-42) (included for comparison) were evaluated in Cohort 2. This resulted in 11 statistically significant biomarkers, including sAPP and Aβ(1-40), but not Aβ(1-42). Other biomarkers identified included serum proteases- tissue plasminogen activator, tumor-associated trypsinogen inhibitor, matrix metalloproteinases 7 and 9, and non-proteases- insulin-like growth factor binding protein 6, AXL receptor tyrosine kinase, omentin, pentraxin-3 and osteopontin. Findings suggest that there is a preferential processing of APP to Aβ(1-40) over Aβ(1-42), and a potential role for the carboxylase activity of the γ-secretase protein, which preferentially splices sAPPβ to Aβ(1-40). Other markers are associated with the breakdown and remodeling of the extracellular matrix, and loss of homeostasis, possibly within the photoreceptor-retinal pigment epithelium-choriocapillaris complex. These data suggest novel disease pathways associated with GA pathogenesis and could provide potential novel targets for treatment of GA.