Developing New Methodologies for the Analysis of Proteins by Native Top-Down Mass Spectrometry

Abstract

Proteins are critical mediators of physiological function and human health. The complexity of the human proteome, which contains over 1 million individual proteoforms, presents a daunting challenge for modern measurement science. Such proteoforms are recruited into myriad multi-protein complexes, many of which remain refractory to standard functional and structural assays. Native proteomics seeks to deploy tools, such as native mass spectrometry (nMS) to identify and a characterize such complexes, but current nMS methods struggle to rapidly quantify closely related proteoforms and to sequence many important assemblies. To advance our understanding of biochemistry, new nMS methodologies are needed to improve the throughput and information content of nMS assays. MPs are challenging targets for nMS due to their hydrophobicity. Current nMS methods yield limited sequence coverages for MPs (< 20%). In Chapter 2, we explore the use of infrared (IR) photoactivation to improve the sequence coverage that can be obtained for MPs. We discovered that IR photoactivation can selectively liberate proteins from detergent micelles, dissociate detergent and lipid clusters, and ultimately enable greater sequence coverages to be obtained (40-60%). Native proteomics methodologies targeting MPs rely upon detergent exchange methodologies to prepare samples for nMS analysis but have yet to be evaluated quantitatively. In Chapter 3, we analyze the efficiency of standard detergent exchange methodologies to identify optimal approaches for native proteomics. Our results highlight the inability of current methods to completely exchange samples into a desired detergent. Furthermore, we note that exchange efficiency depends strongly upon the starting detergent, exchange method, and the MP contained within the sample. Furthermore, we were able to improve detergent exchange efficiencies by increasing the number of exchanges performed or by increasing the concentration of the target surfactant. Collision induced unfolding (CIU) is methodology based in ion mobility-mass spectrometry (IM-MS) capable of rapidly distinguishing between closely related protein isoforms and assessing their stabilities. In Chapter 4, we conduct an interlaboratory assessment of CIU reproducibility by partnering with three laboratories housing of the same IM-MS instrumentation. Using selected standard proteins prepared in our lab and shipped to our collaborators, we found CIU to be highly reproducible (~4-5% RMSD) across laboratories. Chapter 5 explores the first use of CIU for quantitative measurements to assess isomass pairs of biotherapeutics and sequence homologues in both standard and biological matrices that are indistinguishable by IM-MS alone. Our results cover three antibody pairs and include examples of mixed therapies that are provided to patients. Our CIU assays produce calibration curves with correlation coefficients ranging from 0.92- 0.99, limits of detection ranging from 300-5000 nM. Despite the wide adoption of CIU, it has generally remained a throughput limited technology. In Chapter 6, we explore the use of a RapidFireTM robotic system to automate and increase the throughput of CIU. We developed an automated, online desalting procedure, evaluate its efficiency, and develop methods to collect CIU in as little as 30 seconds for a variety of standard proteins, protein complexes, biotherapeutic antibodies, and RNA samples. The work described in this thesis greatly expands the potential applications of nMS by developing new approaches that allow for nMS to be better deployed for challenging targets, to be used for novel applications such as quantitation, whilst also creating methods that vastly improve the throughput of these technologies.