| This dissertation describes developments in mass spectrometry-driven proteomics with an emphasis on incorporating electron transfer dissociation (ETD) into the modern proteomics pipeline. ETD is different from the traditional peptide fragmentation method (i.e., collision-activated dissociation, CAD) in that it dissociates peptides with indifference to the presence of post translational modifications and is most effective for peptides of high charge state and low mass-to-charge ratio.;Chapter 1 is an introduction to the field of proteomics and mass spectrometry. The topics of sample introduction, instrumentation, gas-phase reactions, characterization of post translation modifications, and bioinformatics are described here. The next chapter discusses improvements to ETD to increase dissociation efficiency for dications by applying a supplemental activation step subsequent to the ETD reaction. Though effective, bioinformatic difficulties limited the utility of this approach. To circumvent these issues, a novel data-dependent decision tree-driven methodology was developed. As described in Chapter 3, this methodology takes advantage of the complementary nature of ETD and CAD to select on-the-fly the most appropriate dissociation method for every precursor.;The next two chapters utilize this decision tree-driven methodology to more comprehensively characterize the proteome. In Chapter 4, over 10,000 phosphorylation sites in the human embryonic stem cell phosphoproteome are identified. This work also resulted in the identification of several novel phosphorylation sites on transcription factors vital to the pluripotency of human embryonic stem cells. Chapter 5 takes advantage of the greater diversity of peptides that can be successfully identified when using both CAD and ETD in a single experiment. Here a multiple protease strategy results in a more than 2-fold improvement in the characterization of the Saccharomyces cerevisiae proteome. |
