Beauty hidden in lipids. Proteomics of membrane proteins based on their transmembrane domains.
Membrane proteins are coded by approximately one third of human genes but have been notoriously under-represented in proteomic analyses due to their low expression levels and especially because of their hydrophobicity. Many technical attempts have been made to increase the number of transmembrane proteins identified in proteomic analyses. Multi-step isolation of membrane-enriched fractions improves the number of identified TM proteins but major contamination by membrane-associated proteins or abundant cellular proteins remains a critical obstacle. The amphipathic nature of transmembrane proteins prevents effective solubilzation of intact proteins in both polar and non-polar solvents. Peptides from soluble domains of TM proteins can be enriched but are also inevitably contaminated by associated proteins. On the other side the hydrophopic transmembrane domains are protected by the phospholipid bilayer and can be isolated after effective enzymatic degradation of all contaminating non-membrane proteins. Based on the concept by Adele Blackler and Christine Wu [1] we are testing and developing an improved method for proteomic analysis of transmembrane proteins based on single-step enrichment of membrane fraction, opening of membrane vesicles at low pH and enzymatic digestion of all non-TM proteins. Membrane sheets are then solubilized, TM domains are redigested by CNBr, purified [2] and identified by MS/MS. Proteins with at least one predicted transmembrane domain represent almost 80 percent of all identifications in our samples from HL60 leukemia cells processed as described. Most of the proteins were identified by their hydrophobic peptides originating from the predicted transmembrane domains.
Supported by the Ministry of Education, Youth and Sports (VZ 0021620806 and LC06044)
[1] Blackler A, et al. J. Proteome Res. 2008, 7, 3028-34
[2] Rey M, et al.. Anal Chem. 2010, 82, 5107-16