Identifying Peptides with Higher Order Charge States: An Optimized Approach for Characterizing Membrane Proteins

摘要

Analysis of protein digests using ion trap instrumentation has been a mainstay for shotgun proteomics. Probability based searching algorithms of collision-induced dissociation (CID) data acquired at unit mass resolution are effective for the identification of peptides possessing 1+, 2+ and 3+ charge-states. However, information from larger peptides possessing higher order charge states (i.e., >=4+) is typically lost due to the complexity of the CID spectra (e.g., poor identification of fragment ion charge state), incomplete fragmentation of higher molecular weight precursor ions, and an inability of searching algorithms to effectively model and score the data. Data dependent selection of peptide ions possessing higher order charge-states followed by subsequent analysis of CID fragments at high resolution and mass accuracy represents a complementary strategy for protein identification. Integration of this approach into a shotgun proteomics workflow for characterizing membrane proteins is presented. Membranes of Francisella tularensis subspecies novicida (Fn) were isolated from lysed spheroblasts, and subjected to a sucrose step gradient by centrifugation. The inner membranes (20-53percent sucrose gradient interface) were harvested and frozen at -80 deg C, prior to their digestion by either trypsin or Glu-C. The resulting peptide mixtures were analyzed by nano-LC-MS/MS on a 30 cm long, 75 (mu)m i.d. column packed with C18 employing a gravity-pulled ESI tip. Data was acquired on an LTQ-Orbitrap (Thermo) hybrid mass spectrometer in a variety of data dependent acquisition modes, with CID fragments analyzed in either the LTQ or Orbitrap mass analyzers. CID mass spectra were analyzed by database searching algorithms SEQUEST (Thermo) and Phenyx (Geneva Bioinformatics).