Protein Analysis by Liquid Chromatography-Tandem Mass Spectrometry: Detecting and Profiling Protein S-Palmitoylation and Host Cell Protein Characterization

Proteins perform many biological functions and characterizing their role in the cellular environment is critical in understanding a protein’s role in disease. ‘Bottom-up’ liquid chromatography-tandem mass spectrometry (LC-MS/MS) offers the most effective strategy for large-scale global analysis of highly complex protein mixtures, i.e., proteomics. However, LC-MS/MS workflow strategies must be carefully adapted to maintain experimental reproducibility and overcome inherent limitations in detecting low abundance proteins in concentrated mixtures and labile posttranslational modifications (PTMs). This dissertation presents LC-MS/MS based analytical strategies for the direct detection and site-specific profiling of S-palmitoylated peptides, and for characterizing low abundance host cell proteins (HCPs) in purified biopharmaceutical drug substances (DS). Chapter 1 presents an introduction to quantitative LC-MS/MS-based protein analysis, including instrumentation and techniques relevant to the work presented in Chapters 2 and 3. Chapter 2 discusses S-palmitoylation, the reversible and dynamic modification of a cysteine sulfhydryl with a 16-carbon fatty acid. The dynamic cycling between a protein’s palmitoylated and depalmitoylated states regulates several intracellular events such as protein activity, localization, and protein-protein interactions. N-Ras, a well-known driver of many cancers, is S-palmitoylated near its C-terminus, and despite the therapeutic potential of targeting this modification, direct detection of this modified peptide by LC-MS/MS has not previously been successful. In fact, robust methods for direct LC-MS/MS-based analysis of S-palmitoylated proteins have been lacking. We addressed this issue by developing an LC-MS/MS “bottom-up” workflow that mitigates sample processing issues associated with S-palmitoylation, including peptide solubility, stability, LC column retention, and MS/MS-based sequence analysis. We successfully applied this workflow to directly detect and annotate endogenous acylation sites on recombinant N-Ras. Chapter 3 discusses the adaptation of an LC-MS/MS workflow for HCP characterization in biopharmaceutical drug substances. HCPs are native proteins derived from a host organism used to express biotherapeutic proteins and may be co-purified with a drug substance as process related impurities. The presence of these impurities in drug formulations may present problems in product performance and affect the health of patients. Therefore, effective monitoring of HCP levels in drug substances is essential. Common procedures for quantifying HCP content are limited to immunocapture detection methods, but serious efforts are underway in developing LC-MS/MS strategies for characterizing HCPs. Problematically, HCPs can be 106-fold less concentrated than the biotherapeutic protein in downstream products presenting a dynamic range challenge for electrospray ionization-MS. We developed a generalized, semi-automated, and plate-based hydrophilic interaction liquid chromatography (HILIC) fractionation workflow to deplete biotherapeutic protein from DS with limited sample-specific optimization. Our method is practical, quantitative, and demonstrates industry leading HCP detection sensitivity by LC-MS/MS. Chapter 4 provides a concluding summary of the novel contributions of our work and the broader impact on the field of LC-MS/MS based protein analysis. The developed method for direct detection of palmitoylated peptides promises to expand analytical profiling capabilities of an under-represented PTM. Furthermore, the demonstrated robustness and dynamic range improvements for HCP analysis show promise for assisting downstream process development of biotherapeutic drugs as part of an MS-based analytical platform.PhDChemical BiologyUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/163145/1/jcrellin_1.pd