Purpose
To determine candidate serum protein biomarkers of age-related macular degeneration (AMD).
Methods
We analyzed pooled serum samples collected from subjects who had participated in the Age-Related Eye Disease Study. We pooled the samples into the following analytic groups, based upon the participants’ clinical AMD grade at the time of sample collection: No AMD, n=27; intermediate AMD (large drusen {LDr}), n=34; late AMD-neovascular (NV), n=34; and late AMD-geographic atrophy (GA), n=29. Samples were immunodepleted of the14 most abundant serum proteins using a Sigma IgY-14 column. One mg of protein/pooled sample was digested with trypsin in Rapigest, cation-exchange fractionated into 40 fractions, and analyzed using liquid chromatography/tandem mass spectrometry with a LTQ Velos ion trap mass spectrometer. Peptide and protein identification used SEQUEST and a UniProt database of human proteins amended with reversed decoy sequences to control false discovery. The total MS2 spectrum counts of peptides were used to measure differences in protein abundance between samples using a Chi-squared test and Z-score transformation of the data (p<0.01).
Results
Of the 717 identified proteins, 383 (53%) exceeded the minimum spectral count cutoff of 2.5. NV, GA, and LDr samples were more similar to each other than to No AMD controls. The number of differentially abundant proteins (AMD vs. control) was 47, 57 and 43 for NV, GA, and LDr respectively. Of these, 28 proteins passed differential criteria in all three disease groups vs. No AMD controls. We also observed differentially abundant proteins between AMD groups: 12 in NV vs. LDr, 14 in LDr vs. GA, and 12 in NV vs. GA. Several biological pathways were represented by differentially abundant proteins such as complement activation (including CR1, C4BPB, and SERPING1), lipid transport/regulation, retinol transport, oxidative enzymatic components and inflammatory mediators.
Conclusions
These data validate the concept that protein biomarkers reflect disease processes in AMD and suggest the feasibility of using signature biomarker patterns to identify disease-related parameters. Differentially abundant proteins between the NV, GA, and LDr, phenotypes may lead to a better understanding of the progression of AMD.
Keywords: 412 age-related macular degeneration •
663 proteomics •
609 neovascularization