Logo Logo
Help
Contact
Switch language to German
True single-cell proteomics using advanced ion mobility mass spectrometry
True single-cell proteomics using advanced ion mobility mass spectrometry
In this thesis, I present the development of a novel mass spectrometry (MS) platform and scan modes in conjunction with a versatile and robust liquid chromatography (LC) platform, which addresses current sensitivity and robustness limitations in MS-based proteomics. I demonstrate how this technology benefits the high-speed and ultra-high sensitivity proteomics studies on a large scale. This culminated in the first of its kind label-free MS-based single-cell proteomics platform and its application to spatial tissue proteomics. I also investigate the vastly underexplored ‘dark matter’ of the proteome, validating novel microproteins that contribute to human cellular function. First, we developed a novel trapped ion mobility spectrometry (TIMS) platform for proteomics applications, which multiplies sequencing speed and sensitivity by ‘parallel accumulation – serial fragmentation’ (PASEF) and applied it to first high-sensitivity and large-scale projects in the biomedical arena. Next, to explore the collisional cross section (CCS) dimension in TIMS, we measured over 1 million peptide CCS values, which enabled us to train a deep learning model for CCS prediction solely based on the linear amino acid sequence. We also translated the principles of TIMS and PASEF to the field of lipidomics, highlighting parallel benefits in terms of throughput and sensitivity. The core of my PhD is the development of a robust ultra-high sensitivity LC-MS platform for the high-throughput analysis of single-cell proteomes. Improvements in ion transfer efficiency, robust, very low flow LC and a PASEF data independent acquisition scan mode together increased measurement sensitivity by up to 100-fold. We quantified single-cell proteomes to a depth of up to 1,400 proteins per cell. A fundamental result from the comparisons to single-cell RNA sequencing data revealed that single cells have a stable core proteome, whereas the transcriptome is dominated by Poisson noise, emphasizing the need for both complementary technologies. Building on our achievements with the single-cell proteomics technology, we elucidated the image-guided spatial and cell-type resolved proteome in whole organs and tissues from minute sample amounts. We combined clearing of rodent and human organs, unbiased 3D-imaging, target tissue identification, isolation and MS-based unbiased proteomics to describe early-stage β-amyloid plaque proteome profiles in a disease model of familial Alzheimer’s. Automated artificial intelligence driven isolation and pooling of single cells of the same phenotype allowed us to analyze the cell-type resolved proteome of cancer tissues, revealing a remarkable spatial difference in the proteome. Last, we systematically elucidated pervasive translation of noncanonical human open reading frames combining state-of-the art ribosome profiling, CRISPR screens, imaging and MS-based proteomics. We performed unbiased analysis of small novel proteins and prove their physical existence by LC-MS as HLA peptides, essential interaction partners of protein complexes and cellular function.
Single-cell Proteomics, Proteomics, Lipidomics, Single-cell sequencing, Interactomics, Mass spectrometry, Liquid chromatography
Brunner, Andreas-David
2021
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Brunner, Andreas-David (2021): True single-cell proteomics using advanced ion mobility mass spectrometry. Dissertation, LMU München: Faculty of Chemistry and Pharmacy
[thumbnail of Brunner_Andreas-David.pdf]
Preview
PDF
Brunner_Andreas-David.pdf

60MB

Abstract

In this thesis, I present the development of a novel mass spectrometry (MS) platform and scan modes in conjunction with a versatile and robust liquid chromatography (LC) platform, which addresses current sensitivity and robustness limitations in MS-based proteomics. I demonstrate how this technology benefits the high-speed and ultra-high sensitivity proteomics studies on a large scale. This culminated in the first of its kind label-free MS-based single-cell proteomics platform and its application to spatial tissue proteomics. I also investigate the vastly underexplored ‘dark matter’ of the proteome, validating novel microproteins that contribute to human cellular function. First, we developed a novel trapped ion mobility spectrometry (TIMS) platform for proteomics applications, which multiplies sequencing speed and sensitivity by ‘parallel accumulation – serial fragmentation’ (PASEF) and applied it to first high-sensitivity and large-scale projects in the biomedical arena. Next, to explore the collisional cross section (CCS) dimension in TIMS, we measured over 1 million peptide CCS values, which enabled us to train a deep learning model for CCS prediction solely based on the linear amino acid sequence. We also translated the principles of TIMS and PASEF to the field of lipidomics, highlighting parallel benefits in terms of throughput and sensitivity. The core of my PhD is the development of a robust ultra-high sensitivity LC-MS platform for the high-throughput analysis of single-cell proteomes. Improvements in ion transfer efficiency, robust, very low flow LC and a PASEF data independent acquisition scan mode together increased measurement sensitivity by up to 100-fold. We quantified single-cell proteomes to a depth of up to 1,400 proteins per cell. A fundamental result from the comparisons to single-cell RNA sequencing data revealed that single cells have a stable core proteome, whereas the transcriptome is dominated by Poisson noise, emphasizing the need for both complementary technologies. Building on our achievements with the single-cell proteomics technology, we elucidated the image-guided spatial and cell-type resolved proteome in whole organs and tissues from minute sample amounts. We combined clearing of rodent and human organs, unbiased 3D-imaging, target tissue identification, isolation and MS-based unbiased proteomics to describe early-stage β-amyloid plaque proteome profiles in a disease model of familial Alzheimer’s. Automated artificial intelligence driven isolation and pooling of single cells of the same phenotype allowed us to analyze the cell-type resolved proteome of cancer tissues, revealing a remarkable spatial difference in the proteome. Last, we systematically elucidated pervasive translation of noncanonical human open reading frames combining state-of-the art ribosome profiling, CRISPR screens, imaging and MS-based proteomics. We performed unbiased analysis of small novel proteins and prove their physical existence by LC-MS as HLA peptides, essential interaction partners of protein complexes and cellular function.