Quantitative Proteomics via High Resolution MS Quantification: Capabilities and Limitations

Recent improvements in the mass accuracy and resolution of mass spectrometers have led to renewed interest in label-free quantification using data from the primary mass spectrum (MS1) acquired from data-dependent proteomics experiments. The capacity for higher specificity quantification of peptides from samples enriched for proteins of biological interest offers distinct advantages for hypothesis generating experiments relative to immunoassay detection methods or prespecified peptide ions measured by multiple reaction monitoring (MRM) approaches. Here we describe an evaluation of different methods to post-process peptide level quantification information to support protein level inference. We characterize the methods by examining their ability to recover a known dilution of a standard protein in background matrices of varying complexity. Additionally, the MS1 quantification results are compared to a standard, targeted, MRM approach on the same samples under equivalent instrument conditions. We show the existence of multiple peptides with MS1 quantification sensitivity similar to the best MRM peptides for each of the background matrices studied. Based on these results we provide recommendations on preferred approaches to leveraging quantitative measurements of multiple peptides to improve protein level inference. 1. Introduction The ability to identify and quantify hundreds or even thousands of peptides in a single data dependent LC/MS analysis of a proteolytic digest from a complex mixture such as serum or plasma established mass spectrometry-based proteomics as a tool of choice for biomarker discovery research. These are the ultimate hypothesis-neutral experiments that have the promise of finding solutions for major gaps in drug discovery and other biological inquiries. In the past several years, remarkable advances have been made to better support these efforts on multiple fronts, including sample preparation, instrumentation, and data processing. Proteins and peptides that are part of specific biological pathways tend to exist in low nanomolar or lower concentrations, while the abundant proteins in human serum reach concentrations in the hundreds of micromolar range. For example, a typical concentration of apolipoprotein A1 in human serum is around 30？μM while that of ghrelin is around 100？pM. This five-plus log difference in dynamic range is well beyond the current capabilities of analytical mass spectrometry methods. In order to measure peptides of such low abundance, enrichment of the target analyte using antibodies or other methods before analysis