Abstract
Purpose: :
To employ label free quantitative proteomics in the analysis of protein extracted from the Schirmer strips from normal patients and patients with post-menopausal dry eye to determine quantitative differences in proteins for the detection of protein biomarkers. This data will be compared against our previous findings using iTRAQ and correlated with matching lipid results from the same patients.
Methods: :
Protein samples were collected from two groups: postmenopausal women with dry eye (n = 25), postmenopausal women without dry eye (n = 25). Proteins were extracted from Schirmer strips, each sample was individually digested with trypsin and then analyzed by LC-MS/MS. Each data set is analyzed independently for protein identification as well as the measurement of peak intensity, spectral or peptide counts. Peptide peak intensity, spectral or peptide count for the same protein can be compared between different analyses to obtain the relative quantitation for protein differential expression.
Results: :
On average ~400 unique proteins are identified. Notable differences in protein expression levels are observed between the different patient groups and disease states with statistical relevance using label free quantitation and iTRAQ. Proteins are detected as up- or down regulated and potential biomarkers that are unique within the patient categories are demonstrated. For example, 22 proteins were identified as showing significant changes in protein expression in the severe dry eye group compared to normal patients using iTRAQ. Of the 22 proteins, 4 were up-regulated and 18 were down-regulated. Proteins of interest include lysozyme, lipocalin and mammoglobin B. These proteins of interest have numerous functions and protective roles including front line defense, tear film stability, lipid scavengers and products of inflammation.
Conclusions: :
The major goal of dry eye disease biomarker discovery, and in this case protein profiling, is to identify disease-specific proteins from human patients. Here we used a label free approach in combination with bioinformatic calculations to interrogate data from two sample groups, and discovered potential protein candidates demonstrating up- or down-regulation.