Phyloproteomic classification of unsequenced organisms by top-down identification of bacterial proteins using capLC-MS/MS on an Orbitrap

Currently, most MS-based proteomic studies of bacteria and archea match experimental data to known amino acid sequences from the target organism. Top-down studies use a protein's molecular weight along with data gathered from MS/MS experiments to identify proteins by database matching. For Erwinia herbicola and Enterobacter cloacae, studied here, the necessary protein sequences are not available in protein sequence repositories. We apply top-down protein fragmentation, but match the experimental data with homologous proteins from related organisms with sequenced genomes, demonstrating considerable shared protein sequence between closely related bacteria. Using this homology-based approach, we are not only able to identify representative proteins, but are also able to place the two target bacteria in their correct phylogeny. Furthermore, we show that the unexpected mass delta between the experimental precursor and matched protein sequence can often be localized and characterized using accurate-mass precursor and fragment ion measurements. Finally, we demonstrate that proteins identified by top-down workflows provide strong experimental evidence for correct, missing, and misannotated bacterial protein sequences, not only in the analyzed organism, but also for homologous proteins in closely related species.