Reversible labeling of cysteine‐containing peptides allows their specific chromatographic isolation for non‐gel proteome studies

K Gevaert, B Ghesquière, A Staes, L Martens… - …, 2004 - Wiley Online Library
Proteomics, 2004Wiley Online Library
We report upon a novel procedure to specifically isolate cysteine‐containing peptides from a
complex peptide mixture. Cysteines are converted to hydrophobic residues by mixed
disulfide formation with Ellman's reagent. Proteins are subsequently digested with trypsin
and the generated peptide mixture is a first time fractionated by reverse‐phase high‐
performance liquid chromatography. Cysteinyl‐peptides are isolated out of each primary
fraction by a reduction step followed by a secondary peptide separation on the same …
Abstract
We report upon a novel procedure to specifically isolate cysteine‐containing peptides from a complex peptide mixture. Cysteines are converted to hydrophobic residues by mixed disulfide formation with Ellman's reagent. Proteins are subsequently digested with trypsin and the generated peptide mixture is a first time fractionated by reverse‐phase high‐performance liquid chromatography. Cysteinyl‐peptides are isolated out of each primary fraction by a reduction step followed by a secondary peptide separation on the same column, performed under identical conditions as for the primary separation. The reducing agent removes the covalently attached group from the cysteine side chain, making cysteine‐peptides more hydrophilic and, thereby, such peptides can be specifically collected during the secondary separation and are finally used to identify their precursor proteins using automated liquid chromatography tandem mass spectrometry. We show that this procedure efficiently isolates cysteine‐peptides, making the sample mixture less complex for further analysis. This method was applied for the analysis of the proteomes of human platelets and enriched human plasma. In both proteomes, a significant number of low abundance proteins were identified next to extremely abundant ones. A dynamic range for protein identification spanning 4–5 orders of magnitude is demonstrated.
Wiley Online Library