Traditional proteomics studies can measure many protein biomarkers simultaneously from a single patient-derived sample, promising the possibility of syndromic diagnoses of multiple diseases sharing common symptoms. However, precious cellular-level information is lost in conventional bulk-scale studies that measure tissues comprising many types of cells. As single cells are the building blocks of organisms and are easier to biopsy than traditional bulk samples, performing proteomics on a single-cell level would benefit clinicians and patients. Single-cell proteomics, combined with mass spectrometry imaging, can be used to analyze cells in their microenvironment, preserving spatial information. We have previously used laser-capture microdissection to isolate single motor neurons from tissue and analyze them in our single-cell proteomics platform. However, our sampled population of cells was necessarily limited by the low throughput of the measurement platform, and by the sensitivity of our liquid chromatography-mass spectrometry system to debris introduced in the laser-capture microdissection isolation workflow. In the work described in this dissertation, we dramatically improved the throughput of single-cell proteomics, created a method for removing insoluble debris that clogged our liquid chromatography-mass spectrometry system, and developed a high-performance, low-cost method for nanoflow gradient formation. Together, these methodologies will increase the depth of information and the number of biological replicates that can measured in single-cell proteomics. We hope that these technologies will be applied to future liquid chromatography systems to enable large scale single-cell proteomics studies of tissues. This will reveal the cellular origins of disease on a multimolecular level, while keeping important spatial information. Thus, we expect the technologies and ideas developed here to play a key role in understanding the cellular proteomics in biomedical and clinical settings.

Single-Cell Proteomics (SCP) can provide unique insights into biological processes by resolving heterogeneity that is obscured by bulk measurements. Gains in the overall sensitivity and proteome coverage through improvements in sample processing and analysis increase the information content obtained from each cell, particularly for less abundant proteins. In addition to achieving in-depth proteome coverage from single cells, higher throughput measurements enable large-scale and statistically significant features within single cell populations. This dissertation focuses on method development to improve the sensitivity and throughput of SCP based on the nanoPOTS (nanodroplet Processing in One pot for Trace Samples) platform. The methods discussed here include miniaturization of bottom-up proteome sample preparation and liquid chromatography (LC) separations, implementation of an ultrasensitive latest-generation mass spectrometer, development of automated sample handling workflow, and combination of isotopic and isobaric labeling for higher order multiplexing. The miniaturization of sample preparation largely reduced protein loss during sample preparation and enabled in-depth single-cell proteomics. The sensitivity was further improved using a 20-μm-i.d. in-house-packed nanoLC column and the latest generation Orbitrap Eclipse Tribrid mass spectrometer. A >70% increase in proteome coverage was observed for single cells relative to previous efforts using a 30-μm-i.d. LC columns coupled to a previous-generation Orbitrap Fusion Lumos mass spectrometer. To make SCP and low-input proteome profiling accessible to more proteomics laboratories, a fully automated platform termed autoPOTS (automated Preparation in One pot for Trace Samples) was developed using only commercially available instrumentation for sample processing and analysis. AutoPOTS can be used to analyze 1–500 cells with a modest reduction in peptide coverage for 150 cells and a 24% reduction in coverage for single cells compared to the nanoliter preparation. To improve the throughput of SCP, a hyperplexing sample preparation and analysis method for Single-Cell Proteomics (hyperSCP) was developed using a combination of isotopic and isobaric labeling. This method can improve the throughput by at least 28 times with the same gradient compared to the label-free proteomics and can double or triple the throughput of standard tandem mass tag multiplexing.

Single Cell Proteomics (SCP) is an emerging discipline that contributes to a deeper understanding of individual cells' essential components. In biological systems, individual cells exhibit remarkable diversity, showcasing distinct proteomic profiles and functions. Mass Spectrometry (MS)-based techniques have become essential tools for exploring the proteomes of single cells with remarkable precision. While traditional bulk proteomics methods have been invaluable in revealing the overall protein composition of biological samples, they fall short in capturing the subtle nuances and heterogeneity among individual cells in a population. This limitation emphasizes the need for more targeted and detailed analyses to uncover the protein makeup of single cells. The MS-Based Single-Cell Proteomics technology serves as a valuable solution, providing comprehensive insight at the cellular level by analyzing proteins for identity, abundance, post-translational modifications, and interactions. This dissertation focuses on advancing single-cell proteomics through method development to enhance sensitivity and throughput. It presents a detailed protocol for a label-free single-cell proteomics workflow that integrates the cost-effective HP D100 Single Cell Dispenser and a one-hour, one-step sample preparation method. In contrast to the standard data-dependent acquisition method, the novel wide window acquisition (WWA) intentionally co-isolates and co-fragments adjacent precursors along with the selected precursor, using large isolation windows. Optimized WWA significantly increased the number of MS2-identified proteins by ≈40% compared to standard data-dependent acquisition. In a 40-minute LC gradient at ≈15 nL/min, an average of 3000 proteins per single HeLa cell was identified. Employing this platform, we compared protein expression in individual HeLa cells where the crucial autophagy gene, atg9a, was knocked out, and contrasted it with their isogenic wild-type parental line. To enhance throughput and robustness while preserving superior sensitivity at ultra-low flow rates, we developed an improved multi-column nanoLC system. This system features accelerated offline gradient generation, multiple storage sample loops with selective elution profiles, and allows for analysis as fast as every 20 minutes at 40 nL/min with close to 100% MS utilization time. Moreover, it enables continuous operation for up to 6 months without the need for column replacement. When applied to single-cell Multiple Myeloma treated with lenalidomide, this workflow identified an average of around 1300 unique protein groups.