Targeted Workflow Investigating Variations in the Tear Proteome by Liquid Chromatography Tandem Mass Spectrometry

Abstract

Proteins in tears have an important role in eye health and have been shown as a promising source of disease biomarkers. The goal of this study was to develop a robust, sensitive, and targeted method for profiling tear proteins to examine the variability within a group of healthy volunteers over three days. Inter-individual and inter-day variabilities were examined to contribute to understanding the normal variations in the tear proteome, as well as to establish which proteins may be better candidates as eventual biomarkers of specific diseases. Tear samples collected on Schirmer strips were subjected to bottom-up proteomics, and resulting peptides were analyzed using an optimized targeted method measuring 226 proteins by liquid chromatography-scheduled multiple reaction monitoring. This method was developed using an in-house database of identified proteins from tears compiled from high-resolution data-dependent liquid chromatography tandem mass spectrometry data. The measurement of unique peptide signals can help better understand the dynamics of each of these proteins in tears. Some interesting trends were seen in specific pathways or protein classes, including higher variabilities for those involved in glycolysis, glutathione metabolism, and cytoskeleton proteins and lower variation for those involving the degradation of the extracellular matrix. The overall aim of this study was to contribute to the field of tear proteomics with the development of a novel and targeted method that is highly amenable to the clinical laboratory using high flow LC and commonly used triple quadrupole mass spectrometry while ensuring that protein quantitation was reported based on unique peptides for each protein and robust peak areas with data normalization. These results report on variabilities on over 200 proteins that are robustly detected in tear samples from healthy volunteers with a simple sample preparation procedure.

Discovery and Verification of Sjögren’s Syndrome Protein Biomarkers in Tears by Targeted LC-MRM

“…Untargeted SWATH LC-HRMS/MS analysis quantified 306 proteins, of which 48 were found to be significantly changing in SS-DED patients. A previously developed LC-sMRM (scheduled MRM) method targeting 226 proteins was also applied and 114 of these were significantly changing in the SS-DED group. Considering the overlap between untargeted and targeted analyses, a final list of 124 putative biomarkers were selected, based on criteria of having p -values <0.05 and fold change higher than |1.5|.…”

“…Tear protein composition between individuals has been shown to vary naturally without the presence of disease and some proteins have been shown to be more variable than others . Taking note of these normal variations can be useful to help prioritize which proteins could serve as robust biomarkers.…”

“…Pooled samples (one pool per sample group) were separated into 5 fractions with OASIS MCX cartridges (Waters Limited, 30 mg/1 mL) as described previously and analyzed in information dependent acquisition (IDA) mode to generate the ion library database for subsequent quantitative SWATH data processing. TOF-MS data was acquired from m / z 120 to 1250 (250 ms accumulation time) followed by MS/MS ( m / z 80–1500, 50 ms each) on the 15 most intense ions with collision–offset voltage (CE) of 30 ± 10 V, and dynamic background subtraction.…”

“…Different scheduled multiple reaction monitoring (sMRM) methods were used on the three different cohorts of patients. A previously developed method containing unique peptides from 226 proteins with 2 transition per peptide was applied on the first cohort of patients. Subsequent sMRM methods covering 128 and 73 proteins were developed with 2 unique peptides per protein, whenever possible, and applied for the analysis of the second and third cohorts, respectively (Table S2).…”

“…Sciex Analyst TF 1.7.1 software was used for LC-MRM data acquisition. Target peptides to optimize in subsequent LC-MRM analyses were chosen based on IDA MS/MS data, previously optimized MRM transitions and the public repository SRM Atlas (Institute for Systems Biology). neXtProt (Swiss Institute of Bioinformatics) was used to verify selected peptides were unique to specific target proteins.…”