Biologic and technical reproducibility in proteomic analyses of human vitreous.

Abstract

Purpose : Prior studies have shown that proteomic analysis of vitreous can identify proteins and pathways that play critical roles in retinal disease. In order to confidently identify novel pathways in disease, both technical and biological variation must be known. Therefore, the goal of the present study was to determine biological and technical variability in vitreous proteomic studies.

Methods : Vitreous was derived from patients undergoing repair of epiretinal membrane. Samples selected were derived from patients negative for a history of other intraocular diseases, cancer, and diabetes. Each sample was run on SDS-PAGE to assess its integrity. Abundant proteins were depleted from samples using a Pierce Top 12 protein depletion spin column (ThermoFisher). Protein concentration was determined by DC Protein Assay (Bio-Rad). Samples were processed individually to define variation between patients. Additional samples were combined, aliquoted, and treated as separate pools in order to capture all aspects of technical variation starting at the protein depletion step. All samples were then subjected to tandem mass tag mass spectrometry. This study was approved by the University of Michigan and Penn State College of Medicine institutional review boards and adhered to the tenets of the Declaration of Helsinki.

Results : Following SDS-PAGE and protein staining, each sample had a prominent 55 kDa band corresponding to albumin. This band was absent following abundant protein depletion. No samples had any signs of degradation. The average protein concentration across individual samples was 3.96 µg/µL before and 0.76 µg/µL after depletion, yielding an average recovery rate of 12.83%. The protein concentration across the pooled and then individually aliquoted vitreous was 5.05 µg/µL before depletion and 0.71 µg/µL after depletion, yielding an average recovery rate of 11.85%. Across all samples, the total number of unique proteins and peptides were 1,152 and 12,789, respectively. Analysis of variation in protein abundance revealed that the variability between patients was significantly greater than the observed technical variability.

Conclusions : These data provide a foundation for interpretation of future vitreous proteomic studies. Future studies will assess the magnitude of change across experimental groups as well as test and validate algorithms to generate normalized data matrices across multiple batches of samples and experiments.

This is a 2020 ARVO Annual Meeting abstract.