Personalized proteomic profiles enabled by advances in mass spectrometry-based proteomics

Abstract

The major aim of the work presented in this thesis was to generate personalized proteomics profiles by improving the chromatographic aspects of the proteomic experiment. In the first chapter an overview of proteomics is given and several practical aspects of a proteomic workflow are highlighted. An important aspect is the high complexity and the intrinsic dynamic range of the samples analyzed in proteomics. In order to reduce this complexity and perform a comprehensive study a separation strategy (involving one or more separation steps) is required. A detailed description of reversed phase chromatography is given in this chapter as is the most widely used chromatographic methodology in proteomics. Furthermore, an overview of mass spectrometers is given which includes details of both ionization sources and mass analyzers. Protein quantification is also described as in chapter 3 a quantitative proteomics workflow was applied. A separate section for intestinal stem cell biology describes the characteristics of adult stem cells, in general, and of intestinal stem cells, in particular. The impact of the identification of intestinal stem cells is also addressed and how this led to the development of organoids. Moreover, applications of organoids to investigate disease, colorectal cancer in this case, is further explored. In the second chapter we aimed to build an in-house ultra-high pressure liquid chromatography system with a high separation efficiency and resolution. An easy to implement design was constructed. Particles with small diameter (1.8 μm) were used in the stationary phase with the aim of obtaining an increased efficiency compared to conventionally used 3 μm particles. Thanks to advancements in HPLC system design, particularly with respect to generating and maintaining pressures, we were able to implement columns containing smaller particles (1.8 μm) in dimensions of 50 μm x 40 cm. Several flow rates were studied and gradients ranging from 23 to 458 minutes were optimized. The level of performance of this system was tested by coupling it to an Orbitrap Velos achieving over 4500 proteins with the longest analysis time. Furthermore, a mass spectrometer with faster sequencing speed was used (TripleTOF) which enables us to identify over 1400 proteins in a 23-minute gradient. In the third chapter we performed personalized proteome profiles of healthy and tumor human colon organoids. Peptides were extracted from the organoid samples and dimethyl labeled in order to obtain quantitative information. The optimized reversed phase chromatography described in chapter 2 was used here after SCX prefractionation, to perform a comprehensive proteomics study. Furthermore, two state-of-the art mass spectrometers were used which allowed us to perform optimal peptide sequencing for each SCX fraction. Proteins extracted from organoids grown from seven patients were analyzed and the data was combined with transcriptomics data. Although generic features of colorectal cancer could be obtained across all the patients, our data clearly revealed that each patient possesses a distinct organoid signature at the proteomic level. Key to a comprehensive proteomics study is knowing how PTMs are interconnected, it is important to get an insight into the specific proteoforms present. This is a problem in bottom-up shotgun proteomics approaches since this interconnectedness is removed by the proteolysis step and the generation of peptides. Sadly, it is currently suboptimal to perform full protein sequencing (top down) in a high-throughput manner. A compromise is found in middle-down proteomics, which tries to sequence larger peptides (ideally above 3 kDa) in a shotgun-type experiment. In the fourth chapter we set out to optimize a workflow for the emerging technology of middle-down proteomics. Starting from the sample preparation to the state-of-the- art chromatographic separations and alternative MS fragmentation techniques for the identification of larger peptides are covered. The improvements made to the workflow clearly boost the detection and sequence coverage of middle-range peptides making a step forward towards the identification of specific proteoforms.