QUANTITATIVE PHOSPHOSPROTEOMICS ANALYSIS OF BASAL T CELL SIGNALING PATHWAYS AND OPTIMIZED METHODOLOGY TO IMPROVE PROTEOMICS SEQUENCING DEPTH By Zhuo Chen B.S., Peking University, China, 2008 M.S., Peking University, China, 2010 A Dissertation Submitted in Partial Fulfillment of the Requirements for The Degree of Doctor of Philosophy in the Department of Chemistry at Brown University Providence, Rhode Island May 2016 © Copyright 2016 by Zhuo Chen All Rights Reserved This dissertation by Zhuo Chen is accepted in its present form by the Department of Chemistry as satisfying the dissertation requirements for the Degree of Doctor of Philosophy. Date ___________ ____________________________ Dr. Arthur R. Salomon, Director Recommended to the Graduate Council Date ___________ ____________________________ Dr. David E. Cane, Reader Date ___________ ____________________________ Dr. Wolfgang Peti, Reader Approved by the Graduate Council Date ___________ ____________________________ Dean of the Graduate School iii CURRICULUM VITAE Zhuo Chen was born in Tianjin, China in 1988. After completing his high school degree at No. 5 High School in Beijing, he attended Peking University as an undergraduate student in 2004. He spent four years in the School of Pharmaceutical Sciences and received his Bachelor of Science degree in 2008. Then, he joined the State Key Laboratory of Natural and Biomimetic Drugs in Peking University and completed a Master’s degree with the thesis: Synthesis and Biological Research of c-di-GMP and Its Analogues. In 2010, he moved to Providence, RI and started his Chemistry Ph.D. program at Brown University. He first entered Dr. Bazemore-Walker’s group (2010-2013) and then joined Dr. Salomon’s group (2013-2015), to conduct research in the field of proteomics. His research focused on using quantitative phosphoproteomics approach to investigate T cell signaling pathway. He was also interested in optimizing the methodology of LC-MS/MS based proteomics to improve both the depth of peptide sequencing and the reproducibility of protein quantification. PUBLICATIONS: 1. Chen, Z.; Clifton, J.; Salomon, A. “Standard Pressure LC-MS/MS Using 1.9 µm C18 Particles at Room Temperature Improves Quantitation Reproducibility with Increased Proteomic Sequencing Depth” manuscript in preparation. 2. Chen, Z.*; Courtney, A.*; Salomon, A.; Weiss, A. “The Regulatory Role of Csk and CD45 in the Basal T cell Signaling Pathway” manuscript in preparation. (* These authors contributed equally to this research) 3. Chen, Z.; Salomon, A. “An Optimized Protocol for the Use of DMSO Cosolvent with Q Exactive Mass Spectrometer to Improve Peptide Sequencing Depth Significantly.” manuscript in preparation. iv 4. Patent: Yang, Z.*; Xing, L.; Chen, Z.; Wang, M.; Zhang, L. “A New One-pot Phosphoramidite Method for the Synthesis of Cyclic Diguanylate”, China, 2010, No. 201010135002.3. CONFERENCE PRESENTATIONS: 1. Zhuo Chen, Arthur Salomon*. An Optimized Protocol for the Use of DMSO Cosolvent with Q Exactive Mass Spectrometer to Improve Peptide Sequencing Depth Significantly. 63rd ASMS Conference, 2015, St. Louis, MO poster presentation 2. Zhuo Chen, Na Qi, Lei Xing, Meng Wang, Zhen-Jun Yang*, Li-He Zhang. New Strategies for the Syntheses of cIDPRE and c-di-GMP Analogues. 10th International Tetrahedron Symposium, 2010, Beijing, China poster presentation 3. Zhuo Chen, Lei Xing, Meng Wang, Zhenjun Yang*, Lihe Zhang. Synthesis and Biological Research of c-di-GMP and Its Analogs. 6th National Conference on Chemical Biology, 2009, Xiamen, China poster presentation v Abstract of “Quantitative Phosphosproteomics Analysis of Basal T cell Signaling Pathways and Optimized Methodology to Improve Proteomics Sequencing Depth” by Zhuo Chen, Ph.D., Brown University, May 2016 The basal T cell signaling is important in the T cells maturation and differentiation. The kinase-phosphatase pair Csk and CD45 regulate the basal signaling by modulating the activity of Src family kinases Lck and Fyn. In this study, we deployed a mass spectrometry-based quantitative phosphoproteomic approach to identify the signaling through Csk and CD45 during T cell basal signaling by comparing the wide-scale phosphorylation patterns in Csk and CD45 singly or doubly deficient cell lines with wild type Jurkat T cells. The results reveal that signaling cascades and cytoskeletal dynamics in the basal state of T cells share components with conventional TCR signaling. We also propose a new model of negative regulation of Fyn SH2 domain phosphorylation (on Y185, Y213, Y214) by CD45. Furthermore, based on the many hyperphosphorylated tyrosine sites detected in Csk and CD45 doubly deficient cells, we propose a synergistic regulation model of Lck kinase loop and Fyn SH2 domain, which regulate integrin- mediated bidirectional signaling and cytoskeletal dynamics in T cells. The wide-scale adoption of sub-2 µm particles in HPLC columns has been hampered by the necessity for ultra-high pressure liquid chromatography or a column heating apparatus. We introduce a new strategy to fabricate a 50 cm-long, 1.9 µm particle C18 column, which was packed under 100 Bar and routinely operated below 300 Bar. Compared with 3 µm particles, the column with the 1.9 µm particles could detect 330% more peptides with statistically significant changes from differentially stimulated T cells. vi We thus provide an inexpensive improvement for single-run LC-MS/MS analysis to optimize sequencing depth, dynamic range, sensitivity, and reproducibility. This study also highlights the importance of the statistical analysis of quantitative proteomic data instead of a sole focus on peptides identification yields. Another effort on increasing proteome coverage is to optimize the protocol for the use of DMSO cosolvent with Q Exactive mass spectrometer. DMSO cosolvent was reported to improve the number of peptide identification with Orbitrap, but was found to contaminate Q Exactive. Our proposed protocol minimized the contamination while still increasing peptide identification by at least 10%. vii ACKNOWLEDGEMENT First of all, I would like to thank my graduate advisor, Professor Arthur R. Salomon, for his excellent supervision and support. He is a kind and inspiring mentor who provides insightful guidance while granting me sufficient freedom to work the project out. It is my pleasure and honor for being a member of his lab. I also thank my previous advisor, Professor Bazemore-Walker, who taught me the fundamental knowledge about proteomics and preliminarily trained me to be an eligible researcher. Next, I would like to thank my committee member, Professor David E. Cane and Professor Wolfgang Peti for their valuable advice and support. They continuously helped me for my research project defense, original research proposal and dissertation defense. Thanks also go to Dr. James G. Clifton, Dr. Qinqin Ji and Dr. Michael B. Ellisor. These three intelligent researchers taught me a lot in instrument operation, data analysis, cell culture, and biochemical experiments. Their selfless mentorship and assistance helped me overcome numerous difficulties during my Ph.D. study. Many thanks to all the colleagues I have worked with in Salomon lab (Tao Gu, Qinqin Ji, Ynes Helou, Judson Belmont) and Bazemore-Walker lab (Chao Gong, Yuan Cao, Hongbo Gu, Michael Ellisor, Chloe Poston, Yiying Zhu, Shumin Yao, Anthony Bui). I really enjoyed my time being with you guys. It will be my very precious memory. Last but not the least, I must thank my beloved parents and girlfriend, for their unconditional love and support. Making them happy is my first task of the whole life. This thesis is a gift for them. viii TABLE OF CONTENTS CHAPTER 1: GENERAL INTRODUCTION ................................................................1 1.1 MASS SPECTROMETRY-BASED PROTEOMICS ................................................2 1.2 QUANTITAITVE PHOSPHOPROTEOMICS ..........................................................7 1.3 T CELL SIGNALING ..............................................................................................11 1.4 DEVELOPMENT OF LIQUID CHROMATOGRAPHY TECHNOLOGY IN PROTEOMICS RESEARCH ...................................................................................15 1.5 SUMMARY .............................................................................................................20 1.6 REFERENCES.........................................................................................................21 CHAPTER 2: QUANTITATIVE PHOSPHOPROTEOMICS ANALYSIS REVEALS THE ROLES OF CSK AND CD45 IN BASAL T CELL SIGNALING .29 2.1 INTRODUCTION ....................................................................................................30 2.2 MATERIALS AND METHODS..............................................................................34 2.2.1 Cell Culture and Lysis .......................................................................................34 2.2.2 Western Blot Analysis ........................................................................................34 2.2.3 Protein Reduction, Alkylation, Digestion, and Peptide Immunoprecipitation ..35 2.2.4 Automated Nano-LC/MS....................................................................................37 2.2.5 Data Analysis.....................................................................................................38 2.2.6 Quantitation of Relative Phosphopeptide Abundance .......................................39 2.3 RESULTS AND DISCUSSION ...............................................................................41 2.3.1 Label-free quantification and statistical analysis .............................................43 2.3.2 LCK-mediated phosphorylation events in the basal T cell signaling pathway and actin cytoskeleton regulators ...............................................................................46 2.3.3 A possible co-regulatory role of Lck and CD45 in the integrin-mediated T cell signaling pathway .......................................................................................................54 2.4 CONCLUSION ........................................................................................................62 2.5 REFERENCES ........................................................................................................63 ix CHAPTER 3: STANDARD PRESSURE LC-MS/MS USING 1.9 µm C18 PARTICLES AT ROOM TEMPERATURE IMPROVES QUANTIFICATION REPRODUCIBILITY WITH INCREASED PROTEOMICS SEQUENCING DEPTH ..............................................................................................................................69 3.1 INTRODUCTION ....................................................................................................70 3.2 MATERIALS AND METHODS..............................................................................72 3.2.1 Cell Culture, T cell stimulation and Lysis .........................................................72 3.2.2 Protein Reduction, Alkylation, Digestion and Desalting ..................................73 3.2.3 TiO2 Phosphopeptide Enrichment .....................................................................74 3.2.4 Fabrication of Packed Capillary Columns........................................................74 3.2.5 LC-MS/MS Analysis...........................................................................................75 3.2.6 Data Analysis.....................................................................................................76 3.2.7 Quantitation of Relative Phosphopeptide Abundance .......................................78 3.3 RESULTS .................................................................................................................79 3.3.1 Fabrication of Fritless Column Packed with 1.9 µm C18 Particles ..................79 3.3.2 Operation of an In-house Fabricated 50 cm-long Column Packed with 1.9 µm C18 Particles ...............................................................................................................81 3.3.3 Evaluation of Quantitative Data........................................................................83 3.4 DISCUSSION ..........................................................................................................90 3.5 CONCLUSION ........................................................................................................92 3.6 REFERENCES .........................................................................................................93 CHAPTER 4: AN OPTIMIZED PROTOCOL FOR THE USE OF DMSO COSOLVENT WITH Q EXACTIVE MASS SPECTROMETER TO IMPROVE PEPTIDE SEQUENCING DEPTH SIGNIFICANTLY ..............................................97 4.1 INTRODUCTION ....................................................................................................98 4.2 MATERIALS AND METHODS............................................................................100 4.2.1 Cell Culture and Lysis .....................................................................................100 4.2.2 Protein Reduction, Alkylation, Digestion, and Desalting ...............................101 4.2.3 TiO2 Phosphopeptide Enrichment ...................................................................102 4.2.4 Automated Nano-LC/MS..................................................................................102 x 4.2.5 Data Analysis...................................................................................................103 4.3 RESULTS AND DISCUSSION .............................................................................104 4.3.1 Effect of 5% DMSO Cosolvent on Peptide Identification with Q Exactive and DMSO-induced Q Exactive Mass Spectrometer Contamination .............................104 4.3.2 Minimization of DMSO Contamination with an Optimized Protocol .............106 4.4 CONCLUSION ......................................................................................................108 4.5 REFERENCES ......................................................................................................109 CHAPTER 5: GENERAL INTRODUCTION ............................................................111 5.1 SUMMARY OF RESULTS ...................................................................................112 5.2 FUTURE WORK ...................................................................................................115 xi LIST OF FIGURES Figure 1.1: Peptide backbone fragmentation .......................................................................4 Figure 1.2: A sample MS/MS spectrum of peptide in bottom-up proteomics ....................6 Figure 1.3: Strategies for phosphopeptides enrichment .......................................................9 Figure 1.4: A Proximal signaling complex associates upstream protein tyrosine kinases with downstream pathways .......................................................................................13 Figure 1.5: Overview of TCR signaling pathways ...........................................................14 Figure 1.6: Schematic diagram of slurry packing method ................................................16 Figure 1.7: Performance of different analytical strategies based on two parameters ........20 Figure 2.1: A proposed mechanism by which Csk and CD45 regulate SFKs activity .....32 Figure 2.2: Phosphorylation level in Csk- or/and CD45- deficient cell lines ....................42 Figure 2.3: Csk and CD45 expression in wild type and protein-knock-down Jurkat cells 42 Figure 2.4: Pairwise replicate analysis of selected ion chromatography (SIC) peak areas of different cell clones ...............................................................................................44 Figure 2.5: Volcano plots for q-values versus peak area changes between Csk- or/and CD45- deficient cells and wild type Jurkat cells ........................................................46 Figure 2.6: Quantitative phosphoproteomic analysis of known TCR signaling proteins ..48 Figure 2.7: Quantitative phosphoproteomic analysis of actin cytoskeleton proteins ........51 Figure 2.8: Tyrosine sites that are highly phosphorylated in J.Csk/CD45 cells ................55 Figure 3.1: HPLC separation profile of 50 cm-long/1.9 µm C18 column ..........................82 Figure 3.2: Comparison of total ion chromatogram (TIC) and five representative peptide selected ion chromatograms (SIC) between two 50 cm-long columns......................85 Figure 3.3: Pairwise replicate comparison of selected ion chromatography (SIC) peak xii areas from different analytical columns .....................................................................86 Figure 3.4: Histogram of the q-value obtained from three different analytical columns ..87 Figure 3.5: Normalized peak area distribution of three different analytical columns .......88 Figure 3.6: Fold change distribution of three different analytical columns .......................89 Figure 3.7: Schematic diagram of “keystone effect” .........................................................91 Figure 4.1: Proposed mechanism by which DMSO increases electrospray response .......98 Figure 4.2: Effect of 5% DMSO cosolvent on peptide identification with Q Exactive...105 Figure 4.3: Schematic diagram of electrospray process and the further desolvation occurring in the heated capillary inlet of Q Exactive ..............................................107 Figure 4.4: Peptide identification results under different DMSO concentrations and capillary temperatures of Q Exactive .......................................................................108 xiii LIST OF TABLES Table 2.1: List of tyrosine phosphorylation sites related to Lck kinase activity ..............53 Table 2.2: List of hyperphosphorylated tyrosine sites in J.Csk/CD45 cells .....................61 Table 3.1: Size of electrospray tip determined by laser puller parameters ........................80 Table 3.2: Comparison of 15 cm-long columns pack with 1.9 µm and 3 µm particles .....81 Table 3.3: Comparison of Jurkat peptide identification using in-house fabricated 50 cm- long, 1.9 µm C18 column under different LC gradients ..............................................83 Table 3.4: Overview of phosphopeptide PSM yield with different analytical conditions .84 Table 3.5: Quantitative comparison of TiO2 enriched unstimulated and CD3/4 stimulated Jurkat cells with different analytical columns ............................................................87 Table 3.6: Peak area distribution of different analytical columns .....................................88 xiv LIST OF ABBREVIATIONS ACTC cardiac muscle alpha actin Arp2/3 actin-related proteins 2/3 BSA bovine serum albumin CBL casitas B-lineage lymphoma CD3/31/45 cluster of differential 3/31/45 Csk C-terminal Src kinase CYFIP-1 cytoplasmic FMR1-interacting protein 1 DMSO dimethyl sulfoxide DTT dithiothreitol FDR false discovery rate GAPDH glyceraldehyde-3-phosphate dehydrogenase GSK3B glycogen synthase kinase 3 beta ITAM immunoreceptor tyrosine activation motif Itk IL2-inducible T-cell kinase JAM-1 junctional adhesion molecule 1 KEGG Kyoto Encyclopedia of Genes and Genomes LAT linker for activation of T cells Lck lymphocyte-specific protein tyrosine kinase MHC major histocompatibility complex MYPT1 myosin phosphatase target subunit 1 NF-AT nuclear factor for activated T cells NF-κB nuclear factor kappa B xv N-WASp neuronal Wiskott–Aldrich Syndrome protein PECAM-1 platelet/endothelial cell adhesion molecule 1 PI(4,5)P2 phosphatidylinositol 4,5-bisphosphate PI3K Phosphoinositide 3-kinase PIP5K3 1-phosphatidylinositol-4-phosphate 5-kinase PKC-α protein kinase C alpha PKC-β1 protein kinase C beta 1 PLCγ1 phospholipase C gamma 1 PPP1CB protein phosphatase 1, catalytic subunit, beta isoform PSM peptide spectrum match PTM post-translational modification ROCK2 Rho-associated coiled-coil forming kinase 2 SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis SFK Src family kinase SH2 Src-homology 2 SHP-1 SH2 domain-containing protein tyrosine phosphatase 1 SIC selected ion chromatogram SILAC stable isotope labeling by amino acids in cell culture SLP-76 SH2 domain-containing leukocyte phosphoprotein of 76 kDa TCR T cell receptor UHPLC ultra-high pressure liquid chromatography ZAP-70 zeta-chain-associated protein kinase 70 xvi Chapter 1 GENERAL INTRODUCTION 1 1.1 MASS SPECTROMETRY-BASED PROTEOMICS The term “proteomics” is the comprehensive and high-throughput study of proteins expressed in a given cell, tissue or organism under a specific set of environmental conditions (1, 2). Since proteins serve as the fundamental component of the cellular machinery, the elucidation of the structure, function, post-translational modifications (PTMs) and interactions of proteins are of high importance in biological research. Although huge technical advances have been realized in genomics for complete DNA sequencing, the development of proteomic technology remains an emerging field. Also, instead of the original belief that mRNA expression levels are directly related to protein abundance levels, this correlation was found to be indirect at best as mRNA transcripts are not necessarily translated into proteins (3). Thus, the ultimate mission of proteomics is to identify and quantify the proteins and their PTMs, to provide a better understanding of the molecular mechanism and signal networks involved in various cellular processes at the protein level. The start of proteomics research could be dated back to the 1970s when traditional biochemical approaches involving two-dimensional gel electrophoresis followed by Edman degradation sequencing (4) were employed for protein identification (5, 6). This labor-intensive and time-consuming methodology required large sample amounts and the results were prone to be compromised by the inefficient yield of the sequencing chemical reaction. 2 With the development of mass spectrometry (MS) instrumentation, the availability of complete organismal protein sequence databases, and bioinformatics technology, MS-based proteomics has become a more efficient, accurate and sensitive tool for wide-scale proteomic observation. Mass spectrometers consist of three basic parts: the ionization source, the mass analyzer and the ion detector. Biomolecules, such as proteins and peptides, are first ionized via electrospray ionization (ESI) or matrix assisted laser desorption ionization (MALDI), which can preserve structural information because of the soft manner of ionization. Then, the mass-to-charge ratio (m/z) of ions are measured by the mass analyzer, which typically determines the sequencing rate and accuracy of mass spectrometers. Currently, common mass analyzers include: ion trap (IT), quadrupole (Q), time-of-flight (TOF), ion cyclotron resonance (ICR) and Orbitrap (7-10). Each of these detectors differs in resolution, mass accuracy, sensitivity, scan rate, and dynamic range (11). Proteomics researchers select the ideal mass spectrometer based on their analysis requirements and budgets. After the m/z of precursor peptide or protein ions are determined, the selected precursor ions will undergo fragmentation at amide bonds between each amino acid in the protein or peptide, which can be achieved via various methods and thereby produce different ion types. The most widely employed fragmentation modes are summarized in Figure 1.1. The MS spectra of precursor ions and the MS/MS spectra of fragment ions are processed by searching against protein databases to match the experimental spectra with theoretical fragmentation spectra based 3 Figure 1.1 Peptide backbone fragmentation (12). Tandem mass spectrometric (MS/MS) techniques and nomenclature of polypeptide fragment ions (a, b, c ions for charged N-terminal fragments and x, y, z ions for charged C-terminal fragments) are summarized. on the known sequences of the proteome studied to obtain the sequences of peptides or proteins. Currently, two major strategies are developed in proteomics research: the top-down and the bottom-up approaches. In the top-down approach, the molecular weight of an intact protein is measured while the protein sequencing and any PTMs are localized by dissociation of protein ion in the tandem mass spectrometer with various approaches (13), such as electron capture dissociation (ECD) (14) and electron transfer dissociation (ETD) (15). Ideally, top-down can provide 100% sequence coverage, which would facilitate PTM localization (16). However, this strategy is highly limited by the complexity of the MS/MS spectra acquired from the multiply charged protein fragments, the lack of 4 comprehensive fragmentation, and the requirement for expensive, high-resolution mass spectrometers. Moreover, the difficulty of large molecule separation in HPLC and the poor solubility of protein in MS-compatible buffers hinder the LC-MS/MS analysis of large proteins (MW > 50 kDa) and hydrophobic proteins (e.g. membrane proteins). On the other hand, the bottom-up approach focuses on analyzing proteolytic peptides that are produced from site-specific proteases such as trypsin (17), Lys-C (18), Lys-N (19) and Arg-C (20). The peptide sequencing information can provide a “mass fingerprint” to identify the sequence of the protein and by inference, the identity of the original protein. Compared to top-down, the bottom-up approach offers several advantages for wide-scale, high-throughput protein identification. Firstly, because of the better solubility in the common solvents, peptides are more compatible with LC-MS/MS than proteins. Thus, the complex peptide mixtures can be separated effectively using one or more dimensions of liquid chromatography, to maximize the efficiency of mass spectrometry analysis. Secondly, MS/MS spectra from short peptides are much easier to interpret (Figure 1.2) compared to the complicated MS/MS spectra from intact proteins. 5 Figure 1.2 A sample MS/MS spectrum of peptide in bottom-up proteomics (21). Underlined nominal masses above and below the sequence denote the b and y ions respectively that were annotated from the spectrum. In the common workflow of bottom-up proteomics, a mixture of peptides is usually separated by a reversed-phase chromatography according to the hydrophobicity of the peptides contained in the mixture. Then, the eluted peptides are ionized by electrospray and transmitted into the mass spectrometer for m/z measurement. Further MS/MS fragmentations are performed on the selected precursor ions with different methods such as collision-induced dissociation (CID). Sequence information of peptides can be deduced from the MS/MS spectra. With the rapid development of computational methodologies, peptide sequencing from the mass spectrometry data are typically achieved by the database searching method. The most popular searching programs include Mascot (22), Sequest (23), X!Tandem (24) and ProteinPilot (25). Although each of them employs different algorithms to process data, they share the same idea of the 6 characterization of peptide sequences by matching experimental MS/MS spectra with theoretical spectra generated in silico from proteomic sequence databases such as UniProt (26) and NCBInr (22). With the goal of reducing false positive identifications, quality scores are commonly calculated to indicate the confidence of the spectral match. Researchers can filter the peptide assignments by quality scores to remove the peptide spectrum matches (PSMs) with low significance. Certainly, quality scores serve as a probability-related factor and is not a completely reliable indicator of true identifications. Thus, various methods have been developed to fix the statistical error by filtering the peptide identification result with minimum threshold values corresponding to an acceptable false discovery rate (FDR) (27-29). As the mass accuracy improves on modern mass spectrometers, the confidence of sequence assignment dramatically improves, making statistical methods more robust. 1.2 QUANTITATIVE PHOSPHOPROTEOMICS Protein phosphorylation plays a significant role in numerous cellular processes, including metabolism, cell division, signal transduction and enzymatic activity (30). The identification of phosphorylation sites is very important for understanding their precise molecular mechanisms and cellular functions. Nowadays, mass spectrometry-based proteomics has already become an effective approach to identify phosphorylation sites and dynamic changes of the phosphorylation status (31). The major difficulty in 7 phosphoproteomics research is the extremely low abundance of phosphopeptides, which are thus barely detectable without enrichment prior to MS analysis. The wide-scale identification of tyrosine phosphorylation is especially limited by the lower abundance and diversity of this particular modification with significant competition from serine and threonine phosphorylation sites which are also enriched by many of the methods. In normal growing cells, the majority of phosphorylation occurs on serine (90%) and threonine (10%) residues, whereas only a small fraction occurs on tyrosine (0.05%) residues (32). Several methods have already been developed to enrich for phosphopeptides from complex mixtures such as whole cell lysates (Figure 1.3). One strategy is immunoprecipitation, which uses pan-specific antibodies to purify a certain type of phosphopeptide. This method is particularly effective for analyzing low abundance tyrosine phosphorylation, since highly selective phosphotyrosine-specific antibodies are available. This approach has been successfully applied to study T cell signaling in Jurkat cells (33-35). The primary limitation of this approach is the low yield of the capture step, necessitating the use of somewhat large biological samples (typically 100 million cells). Another common strategy is affinity chromatography. To date, immobilized metal ion affinity chromatography (IMAC) and titanium dioxide (TiO2) chromatography are the most widely used approaches for general phosphopeptide enrichment. IMAC utilizes a metal ion (Fe3+, Al3+, Ga3+ or Co3+) which is immobilized to nitrilotriacetic acid (NTA) 8 Figure 1.3 Strategies for phosphopeptides enrichment (36). or iminodiacetic acid (IDA) coated beads (37). The metal ions can coordinate negatively charged phosphopeptides in the mobile phase. TiO2 enrichment is based on the similar principle as IMAC but the metal need not be chelated (38). In this method, adding organic acids can reduce non-specific binding of acidic peptides and improve the selectivity of the enrichment. It is generally observed that TiO2 chromatography favors enrichment of mono-phosphorylated peptides while IMAC has a preference for multi-phosphorylated peptides (36). The third strategy is ion exchange chromatography, among which strong cation exchange chromatography (SCX) is a popular choice. Under the low pH environment of SCX, phosphopeptides elute before non-phosphorylated 9 peptides with multiple positive charges. In conclusion, each technique has respective advantages, but none is sufficient to identify the entire phosphoproteome. An in-depth phosphoproteomic analysis can require a combination of two or three different methods as outlined in Figure 1.3 (39, 40). Another important field of phosphoproteomics research is quantitative analysis. The ability to monitor changes in phosphorylation allows identification of up- or down-regulated phosphorylation sites in response to stimuli. This type of data can be useful in the understanding of cellular signaling pathways. In mass spectrometry, the ionization efficiency of different proteins or peptides can vary dramatically because of factors such as the chemical properties of individual amino acids in the sequence and charge states. The signal intensities in the MS spectra do not directly reflect the abundance. Thus, the quantitation in mass spectrometry is relative and instead of absolute quantification, the quantitative comparisons between two chemically identical proteins or peptides in different cellular states provides biologically important information. There are generally two methods for relative quantification: In the isotope labeling method, an isotopic label can be introduced to the phosphoprotein (or phosphopeptide) either in cell culture (e.g. stable isotope labeling by amino acids in cell culture, SILAC) (41), or by chemical modification (e.g. isotope-code affinity tag, ICAT; isobaric tags for relative and absolute quantification, iTRAQ). An alternative quantification strategy is label-free quantitative approach, which relies on the present availability of high-resolution mass 10 spectrometry and analytical software (42). In this method, a synthetic peptide standard is spiked into samples and co-purified with the peptide mixtures. Quantification is done by comparing the peak areas of a particular peptide that have been normalized to an exogenous standard spiked at the same levels into each sample. Although quantitation by the stable isotope labeling technique offers slightly lower variation then label-free quantitation, it suffers from some disadvantages. These include the necessity of much more starting material, the cost of expensive isotope labels, the complexity of experimental procedures and the potential biological influence on the cells under investigation. Furthermore, the total number of sequenced peptides is reduced since all the types of labeled peptides trigger redundant MSMS spectra due to the mass shift (43). In contrast, label-free technique is faster, cheaper, with higher dynamic range of quantification and able to provide more analytical depth (42). 1.3 T CELL SIGNALING T cells are lymphocytes that play a critical role in eliminating pathogens and mediating the adaptive immune response. To understand the activation and regulation of T cells is thus of high biological importance. Activation of T cells occurs upon the ligation of T cell receptor (TCR), which is expressed on the cell surface as a key characteristic of T cells, by major histocompatibility complex (MHC) molecules in complex with processed peptides on antigen presenting cells (APCs), resulting in the 11 secretion of cytokines that regulate the immune response. Instead of a linear pathway, T cell signal transduction is a complicated branching network (44). The TCR is a complex oligomer composed of a αβ heterodimer along with CD3 δ/ε/ζ/γ subunit chains (45). The αβ heterodimer is the ligand-binding unit, which recognizes the MHC-peptide, but it cannot perform signal transduction independently. Instead, the signal transduction is mediated by the cytoplasmic domains of the associated CD3 subunits and its ζ chains (46). These chains contain a common phosphorylatable motif designated as immunoreceptor tyrosine-based activation motifs (ITAMs). ITAMs are important for the coupling of TCR ligation to intracellular tyrosine kinases and are essential for all subsequent TCR signaling responses (47). Because TCR has no enzymatic activity, Src family kinases (SFKs) initiate the phosphorylation cascade. Lck (56 kDa) and Fyn (59 kDa) are two important SFKs (Lck is of particular importance) and have been intensively studied (48). They phosphorylate ITAMs and thus provide docking sites for the recruitment of Syk family tyrosine kinase ζ-chain-associated protein kinase 70 (ZAP-70) (49). Subsequent phosphorylation and activation of ZAP-70 by Lck initiates a cascade of phosphorylation of downstream adaptor proteins. This phosphorylation leads to the formation of a signalosome complex (Figure 1.4) nucleated by linker for activation of T cells (LAT) (50) and SH2 domain-containing leukocyte phosphoprotein of 76 kDa (SLP-76) (34), both of which contain the most important targets of ZAP-70. These two adapters lack enzymatic activity but are fundamental for the activation of multiple 12 Figure 1.4 A Proximal signaling complex associates upstream protein tyrosine kinases with downstream pathways (51). downstream pathways via scaffold effector molecules. The absence of either LAT or SLP-76 causes a nearly complete loss of TCR signal transduction (52, 53). Once LAT and SLP-76 are phosphorylated, the signaling protein PLCγ1 is bound to the signalosome complex and phosphorylated by Itk (54). Activated PLCγ1 then hydrolyzes the phosphatidylinositol 4,5-biphosphate (PIP2) to yield two intracellular second messengers, inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 then diffuses into the cytoplasm and binds to receptors on the endoplasmic reticulum, mediating activation of protein kinase C through intracellular Ca2+ mobilization that is essential for downstream signal transductions including activation of the transcription factor NF-AT (55). DAG is a key messenger to regulate two major signaling pathways involving Ras and PKCθ. Ras is required for the activation of the serine-threonine kinase Raf-1, which initiates the phosphorylation of mitogen-associated protein kinase (MAPK) and activates Erk1 13 (MAPK extracellular signal-regulated kinase 1) and Erk2. Erk’s kinase activity results in the activation of transcription factor Elk1, which initiates the activation of activator protein-1 (AP-1) (Jun/Fos) transcription complex and the transcriptional activation of signal transducer and activator of transcription 3 (STAT3) (56). DAG also regulates the Protein kinase C family member PKCθ. Upon T-cell activation, PKCθ is activated by being recruited to the plasma membrane through a lipid-binding domain specific on DAG (57) and subsequently regulates the important activation of NF-κB, which allows cytoplasmic NF-κB to translocate into the nucleus (58). In summary, both DAG pathways and IP3-involved Ca2+ signaling pathways play critical roles in T cell cytokine production and cell proliferation. An overview of the entire TCR signaling pathway is shown in Figure 1.5. Figure 1.5 Overview of TCR signaling pathways (59) 14 1.4 DEVELOPMENT OF LIQUID CHROMATOGRAPHY TECHNOLOGY IN PROTEOMICS RESEARCH With the rapid development of liquid chromatography methodology, especially the wide-scale use of high-pressure liquid chromatography (HPLC), a complex mixture of biomolecules can be separated efficiently within a short time frame. Furthermore, computers can easily achieve the automatic operation of LC-MS/MS, which has become the most commonly used method for proteomics research. Currently, nanoHPLC is the mainstream of separation techniques in proteomics. “Nano” refers to flow rates in the nL/min level (as opposed to µL/min level for traditional HPLC columns). Compared to traditional HPLC, advantages of nanoHPLC include higher sensitivity, minimal sample loss, and better separation efficiency. When handling tiny amounts of sample, nanospray has significantly better compatibility with concentration-sensitive detectors like mass analyzers such as the orbitrap. A capillary HPLC column with a small inner diameter (ID) (µm level) is typically required to achieve nanoHPLC. To fabricate a capillary column for LC-MS/MS, fused-silica tubing (60) or polyether ether ketone (PEEK) (61) are usually employed as the column body material. A tapered end of the column is necessary to form a nanospray tip, which could be either in situ constructed with a laser-based micropipette puller or prefabricated and connected to a capillary column (60). The HPLC column also frequently contains a porous frit, as a stopper to confine the stationary phase while allowing the mobile phase 15 to freely pass. Various methodologies have been developed for frit preparation (62). These frits are typically constructed with a permanently fixed frit, such as sol-gel technology, organic monoliths and sintering, whereas commercially available frits (e.g. stainless-steel frit) can also be directly capped to the column. When the column body is integrated with nanospray tip and frit, researchers can finish fabrication of an HPLC column by packing the stationary phase (usually silica particles) into the body. The most widespread packing method is slurry packing (Figure 1.6): the packing particles are suspended in solvent to form the slurry that is placed into a packing reservoir. The column is held upright by the reservoir with its open bottom end inserted into the slurry. When the sealed reservoir is pressurized with inert gas, the slurry is packed into the column. As the solvent passes the frit, the particles are kept inside column by the frit. Figure 1.6 Schematic diagram of slurry packing method: (A) Instrumentation of capillary HPLC column packing; (B) Cross-section view of packing reservoir. 16 Over the past decade, packing materials for use with nanoHPLC columns have evolved from irregularly shaped particles to spherical particles while the sizes (in diameter) have decreased from 30-100 µm to 3-5 µm and even less than 2 µm. The particle size of the HPLC column primarily determines the column efficiency (reflected by plate number N). Reducing the particle size decreases the diffusion distances of solutes while minimizing the band spreading, and thus improves the column efficiency. Nowadays, to achieve the maximal separation capacity, people are increasingly interested in HPLC columns packed with sub-2 µm particles (63). Besides the commonly used fully porous silica particles (e.g. C18 particles), other types of particles for HPLC stationary phase have also been developed to improve the column efficiency. Superficially porous (also known as fused-core, core-shell) particles are composed of a solid inner core with a thin porous shell (64). The much shorter diffusion path of these particles minimizes peak broadening while their large size generates much lower backpressure for easier operation. In 2012, Phenomenex developed a new AerisTM 3.6µm 100Å core-shell particle, which demonstrates better performance than sub-2µm fully porous columns in peptide separation (65). Because its backpressure is very low, one can attempt the longer column length for more efficient separation. Another stationary phase is called monolithic columns, which consist of a continuous piece of macroporous material (66, 67). They are typically prepared by in situ polymerization. Because of the elevated permeability, monolithic columns generate much 17 lower backpressure than fully porous particles and can be used in very long columns (meter lengths). However, compared to the fully porous silica particles, chemical modification (such as C18, C8, HILIC) techniques of both the superficially porous particles and the monolithic columns have not been well developed and thus the scope of application is limited (68). Also, these particles have lower compatibility with high pressure and harsh pH (< 2 or > 9). Among various commercially available packing particles, sub-2µm fully porous silica particles currently attract substantial research interest for their excellent column efficiency. Certainly, the application of smaller particles increases the column backpressure and thus requires higher operation pressure. To operate the columns packed with sub-2 µm particles, an ultra-high pressure liquid chromatography (UHPLC) system, which delivers pressures higher than the conventional 400 bar, needs to be employed. Because of the ultra-fast separation and excellent peak capacity, UHPLC has already become an emerging technique in both pharmaceutical industry and research laboratories. It is also widely applied for proteomics study by coupling to mass spectrometry and the obtained proteome coverage improves considerably compared to using conventional HPLC (60, 69, 70). Nevertheless, the cost of instrumentation and consumables of UHPLC are significantly higher than conventional HPLC. Many laboratories, which are only interested in routine protein identification but not concentrated on separation technology, are still not equipped with costly UHPLC systems. Besides the 18 instrumentation limit, the application of UHPLC under elevated backpressure can also result in significant alteration of solvent properties to cause frictional heating (71) and solvent compressibility effects (72), which may compromise the reproducibility of peptide retention time or partial loss of column efficiency. Another approach to handle the backpressure issue of sub-2 µm particles is to elevate the column temperature (70, 73). As temperature is elevated, the backpressure of the column decreases because of the lower mobile phase viscosity and the separation time can be shortened (73). By heating the column up to 60 °C, Mann’s group successfully operated 50 cm-long columns (packed with 1.8 µm particles) under conventional HPLC pressure to achieve a unidimensional LC-MS/MS analysis with high proteome coverage (74). However, the potential degradation of peptides and conformation change of proteins under the high temperature could be a critical concern in the proteomics study. Ruta et al. studied the thermostability of peptides and proteins in HPLC at elevated temperature. They found that at 90 °C, the stability of high MW peptides and proteins could be compromised, especially with the long analysis time. High temperatures have also been linked to increased carbamylation of peptides with residual urea used in the digestion of proteins to peptides. They proposed that 60 °C could be an acceptable temperature for peptide analysis (70). In Figure 1.7, the throughput and resolving power of different emerging liquid chromatography technologies are compared with the conventional HPLC. Although each 19 of them provides a highly improved performance, they may still suffer from various shortcomings and be only appropriate with specific research topics. Exploring the methodology to keep the optimal balance between efficiency and instrumentation accessibility attracted our research interest. Figure 1.7 Performance of different analytical strategies (75) based on two parameters: (A) required gradient time to achieve a peak capacity of 100; (B) maximal peak capacity under a 3 hr gradient time. 1.5 SUMMARY The primary goal of this thesis was to elucidate the phosphorylation events of signaling transduction in the basal state of T cells. To achieve this goal, we deployed a mass spectrometry-based quantitative phosphoproteomics approach and elucidated important regulatory roles of the kinase-phosphatase pair Csk and CD45 in the basal T cell signaling pathway (Chapter 2). Our second interest was to improve the efficiency of 20 LC-MS/MS. We managed to pack a 50 cm-long HPLC column with sub-2 µm C18 particles and operate the column under standard pressure with ambient temperature (Chapter 3). The new column permitted the improved performance on both peptides sequencing and quantitative reproducibility. We also optimized the protocol for the use of DMSO cosolvent with a Q Exactive mass spectrometer (Chapter 4), in order to increase peptide identification yield while minimizing the contamination of DMSO. 1.6 REFERENCES 1. Wilkins, M. R., Pasquali, C., Appel, R. D., Ou, K., Golaz, O., Sanchez, J. C., Yan, J. X., Gooley, A. A., Hughes, G., Humphery-Smith, I., Williams, K. L., and Hochstrasser, D. F. (1996) From proteins to proteomes: large scale protein identification by two-dimensional electrophoresis and amino acid analysis. Bio/technology 14, 61-65 2. Anderson, N. L., and Anderson, N. G. (1998) Proteome and proteomics: new technologies, new concepts, and new words. Electrophoresis 19, 1853-1861 3. Gygi, S. P., Rochon, Y., Franza, B. R., and Aebersold, R. (1999) Correlation between protein and mRNA abundance in yeast. Molecular and cellular biology 19, 1720-1730 4. Edman, P. (1949) A method for the determination of amino acid sequence in peptides. Archives of biochemistry 22, 475 5. O'Farrell, P. H. (1975) High resolution two-dimensional electrophoresis of proteins. The Journal of biological chemistry 250, 4007-4021 6. Ames, G. F., and Nikaido, K. (1976) Two-dimensional gel electrophoresis of membrane proteins. Biochemistry 15, 616-623 7. Marshall, A. G., Hendrickson, C. L., and Jackson, G. S. (1998) Fourier transform ion 21 cyclotron resonance mass spectrometry: a primer. Mass spectrometry reviews 17, 1-35 8. Mirsaleh-Kohan, N., Robertson, W. D., and Compton, R. N. (2008) Electron ionization time-of-flight mass spectrometry: historical review and current applications. Mass spectrometry reviews 27, 237-285 9. Michalski, A., Damoc, E., Hauschild, J. P., Lange, O., Wieghaus, A., Makarov, A., Nagaraj, N., Cox, J., Mann, M., and Horning, S. (2011) Mass spectrometry-based proteomics using Q Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer. Molecular & cellular proteomics : MCP 10, M111 011015 10. Zubarev, R. A., and Makarov, A. (2013) Orbitrap mass spectrometry. Analytical chemistry 85, 5288-5296 11. Yates, J. R., Ruse, C. I., and Nakorchevsky, A. (2009) Proteomics by mass spectrometry: approaches, advances, and applications. Annual review of biomedical engineering 11, 49-79 12. Zhurov, K. O., Fornelli, L., Wodrich, M. D., Laskay, U. A., and Tsybin, Y. O. (2013) Principles of electron capture and transfer dissociation mass spectrometry applied to peptide and protein structure analysis. Chemical Society reviews 42, 5014-5030 13. Kelleher, N. L. (2004) Top-down proteomics. Analytical chemistry 76, 197A-203A 14. Sze, S. K., Ge, Y., Oh, H., and McLafferty, F. W. (2002) Top-down mass spectrometry of a 29-kDa protein for characterization of any posttranslational modification to within one residue. Proceedings of the National Academy of Sciences of the United States of America 99, 1774-1779 15. Syka, J. E., Coon, J. J., Schroeder, M. J., Shabanowitz, J., and Hunt, D. F. (2004) Peptide and protein sequence analysis by electron transfer dissociation mass spectrometry. Proceedings of the National Academy of Sciences of the United States of America 101, 9528-9533 16. Siuti, N., and Kelleher, N. L. (2007) Decoding protein modifications using top-down mass spectrometry. Nature methods 4, 817-821 17. Keil-Dlouha, V. V., Zylber, N., Imhoff, J., Tong, N., and Keil, B. (1971) Proteolytic activity of pseudotrypsin. FEBS letters 16, 291-295 22 18. Jekel, P. A., Weijer, W. J., and Beintema, J. J. (1983) Use of endoproteinase Lys-C from Lysobacter enzymogenes in protein sequence analysis. Analytical biochemistry 134, 347-354 19. Taouatas, N., Heck, A. J., and Mohammed, S. (2010) Evaluation of metalloendopeptidase Lys-N protease performance under different sample handling conditions. Journal of proteome research 9, 4282-4288 20. Rahmatullah, M., Gopalakrishnan, S., Radke, G. A., and Roche, T. E. (1989) Domain structures of the dihydrolipoyl transacetylase and the protein X components of mammalian pyruvate dehydrogenase complex. Selective cleavage by protease Arg C. The Journal of biological chemistry 264, 1245-1251 21. Zee, B. M., Levin, R. S., Dimaggio, P. A., and Garcia, B. A. (2010) Global turnover of histone post-translational modifications and variants in human cells. Epigenetics & chromatin 3, 22 22. Perkins, D. N., Pappin, D. J., Creasy, D. M., and Cottrell, J. S. (1999) Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20, 3551-3567 23. Eng, J. K., McCormack, A. L., and Yates, J. R. (1994) An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. Journal of the American Society for Mass Spectrometry 5, 976-989 24. Craig, R., and Beavis, R. C. (2004) TANDEM: matching proteins with tandem mass spectra. Bioinformatics 20, 1466-1467 25. Jagtap, P., Bandhakavi, S., Higgins, L., McGowan, T., Sa, R., Stone, M. D., Chilton, J., Arriaga, E. A., Seymour, S. L., and Griffin, T. J. (2012) Workflow for analysis of high mass accuracy salivary data set using MaxQuant and ProteinPilot search algorithm. Proteomics 12, 1726-1730 26. O'Donovan, C., Martin, M. J., Gattiker, A., Gasteiger, E., Bairoch, A., and Apweiler, R. (2002) High-quality protein knowledge resource: SWISS-PROT and TrEMBL. Briefings in bioinformatics 3, 275-284 27. Keller, A., Nesvizhskii, A. I., Kolker, E., and Aebersold, R. (2002) Empirical 23 statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Analytical chemistry 74, 5383-5392 28. Storey, J. D., and Tibshirani, R. (2003) Statistical significance for genomewide studies. Proceedings of the National Academy of Sciences of the United States of America 100, 9440-9445 29. Elias, J. E., and Gygi, S. P. (2007) Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature methods 4, 207-214 30. Hunter, T. (2007) The age of crosstalk: Phosphorylation, ubiquitination, and beyond. Mol Cell 28, 730-738 31. Beausoleil, S. A., Jedrychowski, M., Schwartz, D., Elias, J. E., Villen, J., Li, J. X., Cohn, M. A., Cantley, L. C., and Gygi, S. P. (2004) Large-scale characterization of HeLa cell nuclear phosphoproteins. Proceedings of the National Academy of Sciences of the United States of America 101, 12130-12135 32. Hunter, T., and Sefton, B. M. (1980) Transforming Gene-Product of Rous-Sarcoma Virus Phosphorylates Tyrosine. P Natl Acad Sci-Biol 77, 1311-1315 33. Helou, Y. A., Nguyen, V., Beik, S. P., and Salomon, A. R. (2013) ERK Positive Feedback Regulates a Widespread Network of Tyrosine Phosphorylation Sites across Canonical T Cell Signaling and Actin Cytoskeletal Proteins in Jurkat T Cells. Plos One 8 34. Ji, Q., Ding, Y., and Salomon, A. R. (2015) SRC homology 2 domain-containing leukocyte phosphoprotein of 76 kDa (SLP-76) N-terminal tyrosine residues regulate a dynamic signaling equilibrium involving feedback of proximal T-cell receptor (TCR) signaling. Molecular & cellular proteomics : MCP 14, 30-40 35. Nguyen, V., Cao, L., Lin, J. T., Hung, N., Ritz, A., Yu, K., Jianu, R., Ulin, S. P., Raphael, B. J., Laidlaw, D. H., Brossay, L., and Salomon, A. R. (2009) A new approach for quantitative phosphoproteomic dissection of signaling pathways applied to T cell receptor activation. Molecular & cellular proteomics : MCP 8, 2418-2431 36. Thingholm, T. E., Jensen, O. N., and Larsen, M. R. (2009) Analytical strategies for phosphoproteomics. Proteomics 9, 1451-1468 24 37. Michel, H. P., and Bennett, J. (1987) Identification of the Phosphorylation Site of an 8.3 Kda Protein from Photosystem-Ii of Spinach. FEBS letters 212, 103-108 38. Larsen, M. R., Thingholm, T. E., Jensen, O. N., Roepstorff, P., and Jorgensen, T. J. D. (2005) Highly selective enrichment of phosphorylated peptides from peptide mixtures using titanium dioxide microcolumns. Molecular & Cellular Proteomics 4, 873-886 39. Alpert, A. J. (2008) Electrostatic repulsion hydrophilic interaction chromatography for isocratic separation of charged solutes and selective isolation of phosphopeptides. Analytical chemistry 80, 62-76 40. Gruhler, A., Olsen, J. V., Mohammed, S., Mortensen, P., Faergeman, N. J., Mann, M., and Jensen, O. N. (2005) Quantitative phosphoproteomics applied to the yeast pheromone signaling pathway. Molecular & Cellular Proteomics 4, 310-327 41. Ibarrola, N., Kalume, D. E., Gronborg, M., Iwahori, A., and Pandey, A. (2003) A proteomic approach for quantitation of phosphorylation using stable isotope labeling in cell culture. Analytical chemistry 75, 6043-6049 42. Filiou, M. D., Martins-de-Souza, D., Guest, P. C., Bahn, S., and Turck, C. W. (2012) To label or not to label: applications of quantitative proteomics in neuroscience research. Proteomics 12, 736-747 43. Neilson, K. A., Ali, N. A., Muralidharan, S., Mirzaei, M., Mariani, M., Assadourian, G., Lee, A., van Sluyter, S. C., and Haynes, P. A. (2011) Less label, more free: approaches in label-free quantitative mass spectrometry. Proteomics 11, 535-553 44. Brownlie, R. J., and Zamoyska, R. (2013) T cell receptor signalling networks: branched, diversified and bounded. Nat Rev Immunol 13, 257-269 45. Weiss, A. (1993) T-Cell Antigen Receptor Signal Transduction - a Tale of Tails and Cytoplasmic Protein-Tyrosine Kinases. Cell 73, 209-212 46. Weiss, A., and Littman, D. R. (1994) Signal-Transduction by Lymphocyte Antigen Receptors. Cell 76, 263-274 47. Reth, M. (1989) Antigen Receptor Tail Clue. Nature 338, 383-384 48. Palacios, E. H., and Weiss, A. (2004) Function of the Src-family kinases, Lck and 25 Fyn, in T-cell development and activation. Oncogene 23, 7990-8000 49. Au-Yeung, B. B., Deindl, S., Hsu, L. Y., Palacios, E. H., Levin, S. E., Kuriyan, J., and Weiss, A. (2009) The structure, regulation, and function of ZAP-70. Immunol Rev 228, 41-57 50. Zhang, W. G., Sloan-Lancaster, J., Kitchen, J., Trible, R. P., and Samelson, L. E. (1998) LAT: The ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell 92, 83-92 51. Kane, L. P., Lin, J., and Weiss, A. (2000) Signal transduction by the TCR for antigen. Curr Opin Immunol 12, 242-249 52. Koretzky, G. A., Abtahian, F., and Silverman, M. A. (2006) SLP76 and SLP65: complex regulation of signalling in lymphocytes and beyond. Nat Rev Immunol 6, 67-78 53. Zhang, W. G., Sommers, C. L., Burshtyn, D. N., Stebbins, C. C., DeJarnette, J. B., Trible, R. P., Grinberg, A., Tsay, H. C., Jacobs, H. M., Kessler, C. M., Long, E. O., Love, P. E., and Samelson, L. E. (1999) Essential role of LAT in T cell development. Immunity 10, 323-332 54. Serrano, C. J., Graham, L., DeBell, K., Rawat, R., Veri, M. C., Bonvini, E., Rellahan, B. L., and Reischl, I. G. (2005) A new tyrosine phosphorylation site in PLC gamma 1: The role of tyrosine 775 in immune receptor signaling. J Immunol 174, 6233-6237 55. Hogan, P. G., Chen, L., Nardone, J., and Rao, A. (2003) Transcriptional regulation by calcium, calcineurin, and NFAT. Gene Dev 17, 2205-2232 56. Genot, E., and Cantrell, D. A. (2000) Ras regulation and function in lymphocytes. Curr Opin Immunol 12, 289-294 57. Yasuda, K., Nagafuku, M., Shima, T., Okada, M., Yagi, T., Yamada, T., Minaki, Y., Kato, A., Tani-ichi, S., Hamaoka, T., and Kosugi, A. (2002) Fyn is essential for tyrosine phosphorylation of Csk-binding protein/phosphoprotein associated with glycolipid-enriched microdomains in lipid rafts in resting T cells. J Immunol 169, 2813-2817 58. Stefanova, I., Hemmer, B., Vergelli, M., Martin, R., Biddison, W. E., and Germain, R. N. (2003) TCR ligand discrimination is enforced by competing ERK positive and 26 SHP-1 negative feedback pathways. Nat Immunol 4, 248-254 59. Schwartzberg, P. L., Finkelstein, L. D., and Readinger, J. A. (2005) TEC-family kinases: Regulators of T-helper-cell differentiation. Nat Rev Immunol 5, 284-295 60. Shen, Y., Zhao, R., Berger, S. J., Anderson, G. A., Rodriguez, N., and Smith, R. D. (2002) High-efficiency nanoscale liquid chromatography coupled on-line with mass spectrometry using nanoelectrospray ionization for proteomics. Analytical chemistry 74, 4235-4249 61. Tyrrell, E., Hilder, E. F., Shalliker, R. A., Dicinoski, G. W., Shellie, R. A., Breadmore, M. C., Pohl, C. A., and Haddad, P. R. (2008) Packing procedures for high efficiency, short ion-exchange columns for rapid separation of inorganic anions. Journal of chromatography. A 1208, 95-100 62. Cheong, W. J. (2014) Fritting techniques in chromatography. Journal of separation science 37, 603-617 63. Eugster, P. J., Guillarme, D., Rudaz, S., Veuthey, J. L., Carrupt, P. A., and Wolfender, J. L. (2011) Ultra high pressure liquid chromatography for crude plant extract profiling. Journal of AOAC International 94, 51-70 64. Guiochon, G., and Gritti, F. (2011) Shell particles, trials, tribulations and triumphs. Journal of chromatography. A 1218, 1915-1938 65. Fekete, S., Berky, R., Fekete, J., Veuthey, J. L., and Guillarme, D. (2012) Evaluation of a new wide pore core-shell material (Aeris WIDEPORE) and comparison with other existing stationary phases for the analysis of intact proteins. Journal of chromatography. A 1236, 177-188 66. Minakuchi, H., Nakanishi, K., Soga, N., Ishizuka, N., and Tanaka, N. (1996) Octadecylsilylated porous silica rods as separation media for reversed-phase liquid chromatography. Analytical chemistry 68, 3498-3501 67. Luo, Q., Shen, Y., Hixson, K. K., Zhao, R., Yang, F., Moore, R. J., Mottaz, H. M., and Smith, R. D. (2005) Preparation of 20-microm-i.d. silica-based monolithic columns and their performance for proteomics analyses. Analytical chemistry 77, 5028-5035 68. Guillarme, D., Ruta, J., Rudaz, S., and Veuthey, J. L. (2010) New trends in fast and 27 high-resolution liquid chromatography: a critical comparison of existing approaches. Analytical and bioanalytical chemistry 397, 1069-1082 69. Cristobal, A., Hennrich, M. L., Giansanti, P., Goerdayal, S. S., Heck, A. J., and Mohammed, S. (2012) In-house construction of a UHPLC system enabling the identification of over 4000 protein groups in a single analysis. The Analyst 137, 3541-3548 70. Ruta, J., Guillarme, D., Rudaz, S., and Veuthey, J. L. (2010) Comparison of columns packed with porous sub-2 microm particles and superficially porous sub-3 microm particles for peptide analysis at ambient and high temperature. Journal of separation science 33, 2465-2477 71. Jerkovich, A. D., Mellors, J. S., and Jorgenson, J. W. (2003) The use of micron-sized particles in ultrahigh-pressure liquid chromatography. Lc Gc Eur 16, 20-23 72. Foley, J. P., Crow, J. A., Thomas, B. A., and Zamora, M. (1989) Unavoidable Flow-Rate Errors in High-Performance Liquid-Chromatography. Journal of chromatography 478, 287-309 73. Xiang, Y., Liu, Y., and Lee, M. L. (2006) Ultrahigh pressure liquid chromatography using elevated temperature. Journal of chromatography. A 1104, 198-202 74. Thakur, S. S., Geiger, T., Chatterjee, B., Bandilla, P., Frohlich, F., Cox, J., and Mann, M. (2011) Deep and highly sensitive proteome coverage by LC-MS/MS without prefractionation. Molecular & cellular proteomics : MCP 10, M110 003699 75. Guillarme, D., and Veuthey, J. L. (2011) The Theory and Practice of UHPLC and UHPLC-MS. In: Byrdwell, W. C., and Holcapek, M., eds. Extreme Chromatography: Faster, Hotter, Smaller, pp. 1-46 28 Chapter 2 QUANTITATIVE PHOSPHOPROTEOMICS ANALYSIS REVEALS THE ROLES OF CSK AND CD45 IN BASAL T CELL SIGNALING 29 2.1 INTRODUCTION In T cell signaling pathways, Src family kinases (SFKs) are among the first signaling molecules to initiate the downstream signaling cascades. Lck (56 kDa) and Fyn (59 kDa) are two important SFKs that have been extensively studied (1, 2). The SFK family members consist of N-terminal binding sites for saturated fatty acid, a unique region, a Src-homology 3 (SH3) domain, an SH2 domain, a tyrosine kinase domain, and a C-terminal negative regulatory domain (1). The unique region may contribute to the distinct localization of individual family members: for example, Lck is typically located at the plasma membrane while Fyn is largely found at intracellular structures, such as the mitotic spindle (3). Instead of kinase activity, the SH2 and SH3 domains function to facilitate association of other macromolecular complexes with Lck and Fyn (4). Upon T cell activation and co-receptor clustering, Lck molecules transphosphorylate their activation loop tyrosines (Y394) in their kinase domain and proximally localize to phosphorylate-specific tyrosine residues within ITAMs of CD3 and TCR ζ chains, recruiting ZAP-70 to ITAMs and activating the downstream signaling molecules. In contrast, Lck is down regulated via its C-terminal negative regulatory tyrosine Y505. When Y505 is mutated, a hyperactive T cell phenotype is observed, which leads to tumor formation in vivo (5). The phosphorylation of Lck Y505 is reciprocally regulated by the kinase-phosphatase pair Csk (C-terminal Src kinase) and CD45. The only confirmed 30 substrate of the ubiquitously expressed kinase Csk is the inhibitory tyrosine of the SFKs (6). Csk phosphorylates Y505 of Lck and thus inactivates Lck activity. Upon deletion or inhibition of Csk, thymocytes were able to autonomously pass through beta selection and positive selection checkpoints without TCR expression (7, 8). This abnormal thymic development produces TCR-like signals in the absence of TCR engagement. CD45, a receptor-like tyrosine phosphatase expressed on all nucleated hematopoietic cells, specifically dephosphorylates Lck Y505 and counteracts the function of Csk kinase to positively regulate the activity of Lck (9). CD45-null T cell lines exhibit constitutive hyperphosphorylation of Y505 and drastic suppression of TCR signaling (10, 11). Also, CD45-deficient mice suffer from a partial block at the pre-TCR beta-selection checkpoint and a nearly complete block at positive selection. The mutation of Lck Y505 to phenylalanine rescues both TCR signaling and T cell development in theses animals, indicating the critical role of this tyrosine as an important substrate of CD45 (12). The mechanism by which Csk and CD45 regulate Lck activity is well established as shown in Figure 2.1. Once Lck Y505 is phosphorylated, it interacts with its own SH2 domain to form a closed, inactive conformation. The SH3 domain can further stabilize this closed conformation via intramolecular interactions (13). Dephosphorylation of Y505 by CD45 can open this conformation to generate a ‘primed’ (opened, but not activated) Lck, which can be further activated to perform the kinase activity by phosphorylation of Y394. To terminate TCR signal transduction by Lck, CD45 can also play a negative regulatory role 31 through dephosphorylation of Y394. This negative regulation is typically found in T cells upon TCR stimulation instead of resting (unstimulated) T cells (14). In summary, both Csk and CD45 are indispensable for modulating the activity of SFKs in T cells and setting a threshold for TCR signals, to adjust the receptor sensitivity to a proper level (15). Figure 2.1 A proposed mechanism by which Csk and CD45 regulate SFKs activity (16). Although the TCR-triggered signaling pathway plays an important role in mediating adaptive immune responses, another type of signal transduction in the basal state of T cells is also critical during thymic development (14). Unlike the TCR inducible pathway, the basal signaling pathway is continuously active without the requirement for TCR ligation with MHCs and thus plays an important role in pre-TCR beta-selection, a key 32 step in T cell survival and differentiation (17). Moreover, basal signaling may sensitize TCR receptors by speeding up the responses of T cells to stimuli to enhance ligand recognition (18, 19). Many studies have identified the fundamental role of Lck in basal signal transduction in the absence of TCR ligation (8, 14, 20). Csk and CD45 regulate the activity of Lck via its inhibitory site Y505. Csk only negatively regulates basal but not TCR inducible signaling, because Csk is rapidly detached from the proximity of its substrate upon TCR ligation by an unresolved mechanism (21, 22). CD45 may either positively (via Lck Y505) or negatively (via Lck Y394) regulate Lck activity. Nevertheless, in the basal signaling pathway, CD45 is found to mostly play a positive regulatory role to counteract Csk inhibitory phosphorylation of Lck Y505 (16). Both of these two regulators establish a dynamic equilibrium to maintain the basal steady-state of T cells. Studies of the TCR inducible signaling pathway have been well documented, but the ligand-independent basal signaling pathway is less understood. Considering its biological importance in thymocyte maturation and maintenance of the T cell steady-state, we deployed a mass spectrometry-based quantitative phosphoproteomic approach to identify the signaling through Csk and CD45 during T cell basal signaling. We were particularly interested in elucidating the regulatory role of the important kinase-phosphatase pair Csk and CD45 by comparing the wide-scale phosphorylation patterns in Csk and CD45 singly or doubly deficient cell lines with wild type parental Jurkat T cells. 33 2.2 MATERIALS AND METHODS 2.2.1 Cell Culture and Lysis CRISPR/Cas9 targeted editing was used to disrupt CSK and CD45 in "wild-type" parental Jurkat T cells by Dr. Adam Courtney in the lab of Prof. Arthur Weiss (University of California San Francisco). Three separate clones were analyzed of the J.Csk (Csk-deficient; 3C10, 4G2, 4A8), J.CD45 (CD45-deficient; 1G11, 2F5, 3A2), and J.Csk/CD45 (Csk/CD45-deficient; 1C5, 2B10, 2F10) cells. The cells were maintained in RPMI 1640 medium (HyClone, Logan, UT) supplemented with 10% heat-inactivated undialyzed FBS (HyClone, Logan, UT), 2 mM L-glutamine, 100 U/ml penicillin G, and 100 µg/ml streptomycin (HyClone, Logan, UT) in a humidified incubator with 5% CO2 at 37°C. Cells were grown for eight doublings before harvest, and then placed in lysis buffer (8 M urea, 1 mM sodium orthovanadate, 20 mM HEPES, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, pH 8.0) for 20 min at 4°C. The lysate was sonicated at a 30 watt output with 2 bursts of 30 seconds each and cleared at 14,000 g for 15 min at 4°C. In total, five biological replicates for each cell clone were performed and treated separately. 2.2.2 Western Blot Analysis Total cellular protein from 8 M urea cell lysates was diluted 1:1 with a 2× sample loading buffer (4% SDS, 125 mM Tris-HCl, pH 6.8, 20% v/v glycerol, 5% 2- 34 mercaptoethanol, 0.01% bromophenol blue) for each sample. Equal amounts of protein, as measured by the DC protein Assay (Bio-Rad, Hercules, CA), were separated by 4-20% gradient SDS-PAGE on Precise Tris-HEPES gel (Thermo Fisher Scientific, Waltham, MA), and electroblotted onto an Immobilon membrane (EMD Millipore, Billerica, MA). The membrane was blocked for 45 minutes in blocking buffer (3% BSA in TBS/0.1% Tween-20) at room temperature and then incubated with the primary antibody overnight at 4°C. Primary antibodies used in this study were 1:1000 Mouse Anti-Csk (BD Biosciences, Franklin Lakes, NJ), 1:2000 Mouse monoclone anti-CD45 (9.4) ascites (kind gift of Prof. Arthur Weiss), 1:10000 anti-GAPDH produced in rabbit (Sigma-Aldrich, St. Louis, MO). The membrane was washed 3 times for 5 minutes at room temperature in TBS/0.1% Tween-20. The membrane was then incubated with anti-mouse IgG and anti-rabbit IgG (Li-Cor, Lincoln, NE) for 1 hr in blocking buffer (3% BSA in TBS/0.1% Tween-20) at room temperature in the dark and washed 4 times for 5 minutes with TBS/0.1% Tween-20. Bands were visualized using an Odyssey Imaging System (Li-Cor, Lincoln, NE). 2.2.3 Protein Reduction, Alkylation, Digestion, and Peptide Immunoprecipitation Protein concentration was measured using the DC Protein Assay (Bio-Rad, Hercules, CA). Next, clear cell lysate was reduced with 45 mM DTT for 20 min at 60°C and alkylated with 100 mM iodoacetamide for 15 min at room temperature in the dark. Cell 35 lysate was then diluted 4-fold with 20 mM HEPES buffer, pH 8.0 and digested with sequencing grade modified trypsin (Promega, Madison, WI) in a 1:100 (w/w) trypsin : protein ratio overnight at room temperature. Typtic peptides were acidified to pH 2.0 with 20% trifluoroacetic acid (TFA), cleared at 1800 g for 5 min at room temperature, and desalted using C18 Sep-Pak plus cartridges (Waters, Milford, MA) as previously described (23), with the exception that TFA was used instead of acetic acid. Eluents containing peptides were lyophilized for 48 hours to dryness. Peptide immunoprecipitation for phosphotyrosine containing peptides was performed using p-Tyr-100 phosphotyrosine antibody beads (Cell Signaling Technology). Dry peptides from each replicate of different cell clones were respectively reconstituted in ice-cold immunoaffinity purification (IAP) buffer (5 mM MOPS pH 7.2, 10 mM sodium phosphate, 50 mM NaCl) and further dissolved through gentle shaking for 30 minutes at room temperature and brief sonication in a sonicator water bath. Prior to peptide immunoprecipitation, a 10 pmol fraction of synthetic phosphopeptide LIEDAEpYTAK was added to each replicate as an exogenous quantitation standard. Peptide solutions were then cleared at 1800 g for 5 min at room temperature, combined with p-Tyr-100 phosphotyrosine antibody beads, and incubated for 2 hr at 4 °C. Beads were then washed three times with IAP buffer and twice with cold ddH2O, and eluted with 0.15% TFA. Eluted peptides were then desalted using C18 Zip Tip pipette tips (Millipore Corporation Billerica, MA) according to the manufacturer’s instructions. 36 2.2.4 Automated Nano-LC/MS Tryptic peptides were analyzed by a fully automated phosphoproteomic technology platform (24, 25). The nanoLC-MS/MS experiments were performed with an Agilent 1200 Series Quaternary HPLC system (Agilent Technologies, Santa Clara, CA) connected to a Q Exactive Plus mass spectromter (Thermo Fisher Scientific, Waltham, MA). Phosphopeptides were eluted into the mass spetrometer through a PicoFrit analytical column (360 µm outer diameter 75 µm inner diameter-fused silica packed on a pressure bomb with 15 cm of 3 µm Monitor C18 particles; New Objective, Woburn, MA) with a reversed phase gradient (0-70% 0.1 M acetic acid in acetonitrile in 60 min, with a 90 min total method duration). The electrospray ion source was operated at 2.0 kv in a split flow configuration. The Q Exactive Plus was operated in the data dependent mode using a top-9 data dependent method. Survey full scan MS spectra (m/z 400-1800) were acquired at a resolution of 70,000 with an AGC target value of 3×106 ions or a maximum ion injection time of 200 ms. Peptide fragmentation was performed via higher-energy collision dissociation (HCD) with the energy set at 28 NCE. The MS/MS spectra were acquired at a resolution of 17,500, with a targeted value of 2×104 ions or a maximum integration time of 200 ms. The underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 1.0%. The ion selection abundance threshold was set at 1.0×103 with charge state exclusion of unassigned and z =1, or 6-8 ions and dynamic exclusion time of 30 sec. 37 2.2.5 Data Analysis Peptide spectrum matching of MS/MS spectra was performed against a human-specific database (UniProt; downloaded 2/1/2013) using MASCOT v. 2.4 (Matrix Science, Ltd, London W1U 7GB UK). A concatenated database containing 144,156 “target” and “decoy reversed” sequences was employed to estimate the false discovery rate (FDR) (26). Msconvert from ProteoWizard (v. 3.0.5047), using default parameters and with the MS2Deisotope filter on, was employed to create peak lists for Mascot. Mascot database searches were performed with the following parameters: trypsin enzyme cleavage specificity, 2 possible missed cleavages, 10 ppm mass tolerance for precursor ions, 20 mmu mass tolerance for fragment ions. Search parameters specified a dynamic modification of phosphorylation (+79.9663 Da) on serine, threonine, and tyrosine residues, a dynamic modification of methionine oxidation (+15.9949 Da), and static modification of carbamidomethylation (+57.0215 Da) on cysteine. Mascot results were filtered by Mowse Score (>10). Peptide assignments from the database search were filtered down to 1% false discovery rate (FDR) by a logistic spectral score, as previously described (26, 27). To validate the position of the phosphorylation sites, the Ascore algorithm (28) was applied to all data, and the reported phosphorylation site position reflected the top Ascore prediction. 38 2.2.6 Quantitation of Relative Phosphopeptide Abundance Relative quantification of phosphopeptide abundance was performed via calculation of selected ion chromatograms (SIC) peak areas. Retention time alignment of individual replicate analyses was performed as previously described (29). Peak areas were calculated by inspection of SICs using in-house software programmed in R 3.0 based on the Scripps Center for Metabolomics’ XCMS package (version 1.40.0). This approach performed multiple passes through XCMS’s central wavelet transformation algorithm (implemented in the centWave function) over increasingly narrower ranges of peak widths, and used the following parameters: mass window of 10 ppm, minimum peak widths ranging from 2 to 20 sec, maximum peak width of 80 sec, signal to noise threshold of 10 and detection of peak limits via descent on the non-transformed data enabled. For cases when centWave did not identify an MS peak, we used the getPeaks function available in XCMS to integrate in a pre-defined region surrounding the maximum intensity signal of the SIC. SIC peak areas were determined for every phosphopeptide that was identified by MS/MS. In the case of a missing MS/MS for a particular peptide, in a particular replicate, the SIC peak area was calculated according to the peptide's isolated mass and the retention time calculated from retention time alignment. A minimum SIC peak area equivalent to the typical spectral noise level of 1000 was required of all data reported for label-free quantitation. Individual SIC peak areas were normalized to the exogenously spiked phosphopeptide LIEDAEpYTAK peak 39 area. The LIEDAEpYTAK phosphopeptide was added in the same amount to each replicate of each cell clone and accompanying cellular phosphopeptides through the phosphotyrosine peptide enrichment and reversed-phase elution into the mass spectrometer. A ratio heatmap was generated as previously described (23). Each row represented an individual phosphopeptide and each column represented the ratio of phosphopeptide abundance differences between a specific cell clone (three each for J.Csk, J.CD45, J.Csk/CD45) and wild type Jurkat clone E6-1. Each ratio was generated from the average of the LIEDAEpYTAK-normalized SICs in the five biological replicate experiments. A black color represented a ratio of 1 between the two cell types. A red color represented less abundance, and green represented higher abundance of the given phosphopeptide in a specific Jurkat-derived cell line compared to wild type Jurkat. The magnitude of change of the heatmap color was calculated as described (23). Blanks in the heatmap indicated that a clearly defined SIC peak was not observed for that phosphopeptide in any of the replicate analyses for that cell clone. Based on the determined unpaired p-values, q-values were also calculated for each ratio heatmap square to assess the statistical significance of phosphopeptide abundance changes between Jurkat-derived cell lines (J.Csk, J.CD45 or J.Csk/CD45) and wild type Jurkat. A white dot on a ratio heatmap square indicated that a significant difference (q-value < 0.01) was observed. 40 2.3 RESULTS AND DISCUSSION To elucidate the role of Csk and CD45 in the T cell basal signaling pathway, a quantitative phosphoproteomic analysis was applied to compare the wild type Jurkat E6-1 cell line with three clones of each of three Jurkat-derived cell lines (J.Csk (Csk-deficient), J.CD45 (CD45-deficient), J.Csk/CD45 (Csk/CD45-deficient)). Western blot quantitation of phosphorylation level in the whole cell lysate and the key regulatory sites Lck Y505 and Y394 in these cell lines was performed by Dr. Adam Courtney in the Weiss lab (Figure 2.2). In our project, each of these Jurkat-derived cell lines comes with three cell clones: J.Csk (3C10, 4G2, 4A8), J.CD45 (1G11, 2F5, 3A2), J.Csk/CD45 (1C5, 2B10, 2F10). The expression level of Csk and CD45 in these cell clones was verified by Western blot (Figure 2.3), to confirm the protein knock down was complete. For each cell clone, five biological replicate experiments were performed. The resulting phosphoproteomic data represents a wide-scale view of the tyrosine phosphorylation events in the T cell basal signaling pathway with or without Csk and CD45. 41 1 2 3 4 5 6 7 8 1. Jurkat wild type 2. J.Csk 3C10 3. J.CD45 clone 1 Blot 1A anti pY (4G10) 4. J.CD45 clone 2 5. J.Csk/CD45 clone 1 6. J.Csk/CD45 clone 2 7. J.Csk/CD45 clone 3 8. J.Csk/CD45 clone 4 Blot 1B anti Lck pY505 Blot 2A anti Lck pY394 Blot 2B anti GAPDH Figure 2.2 Phosphorylation level in Csk- or/and CD45- deficient cell lines. The Western blot results were provided by Arthur Weiss Lab from University of California San Francisco. Cell lines and antibodies for each blot are marked in figure. GAPDH was used as a loading control. Figure 2.3 Csk and CD45 expression in wild type and protein-knock-down Jurkat cells. Protein lysates from J.Csk, J.CD45, J.Csk/CD45, and wild type Jurkat were separated by SDS-PAGE and immunoblotted with anti-Csk antibody or anti-CD45 antibody. GAPDH was used as a loading control. 42 2.3.1 Label-free quantification and statistical analysis After collecting the phosphopeptide proteomic data, high quality sequence assignments were determined using stringent criteria as described in Materials and Methods. In all, 2282 unique tyrosine phosphorylation sites residing on 1290 proteins were identified at 1% FDR, of which 1935 unique tyrosine phosphorylation sites on 1080 proteins showed statistically significant changes (q-value < 0.01) between wild type Jurkat cells and at least one clone of Csk- or/and CD45- deficient cell line. Relative quantitation of phosphopeptide abundance via normalized SIC peak areas was performed for each phosphopeptide in each cell clone. Pairwise replicate analysis indicated the high degree of quantitative reproducibility between individual LC-MS runs for label-free quantitation (average R value of 0.833 ± 0.057; Figure 2.4). 43 Figure 2.4 Pairwise replicate analysis of selected ion chromatography (SIC) peak areas of different cell clones. Each dot in the scatterplot represents the log10 (normalized IC peak area) of a single phosphopeptide in two different replicates (ten possible pairs in total: Rep1:Rep2, Rep1:Rep3, Rep1:Rep4, Rep1:Rep5, Rep2:Rep3, Rep2:Rep4, Rep2:Rep5, Rep3:Rep4, Rep3:Rep5, Rep4:Rep5.). Dot density is indicated by color (from low to high: gray, blue, green, yellow, orange and red). Cell lines name and correlation coefficient are marked in the figure respectively. The calculation of correlation coefficient was based on the combination of all the 10 pairwise replicate to replicate comparisons. 44 Ratios were calculated by comparing the normalized SIC peak area for each phosphopeptide from each clone of Csk- or/and CD45- deficient cell lines to wild type Jurkat cells. A total of five replicate experiments were performed and SIC peak areas were calculated from the average values. q-values for multiple hypothesis tests were calculated based on p-values from two-tailed unpaired Student’s t tests and significantly changed phosphopeptides were conservatively attributed to those with q-value < 0.01. Volcano plots depicting the relationship between the magnitude of the ratio (fold change) and q-value for each clone are shown in Figure 2.5. A substantial number of significantly changed phosphopeptides (q-value < 0.01) were identified. The results also indicated that most tyrosine phosphopeptides showed elevated phosphorylation in Csk-deficient and Csk/CD45-deficient cell lines and reduced phosphorylation in CD45-deficient cell lines, which were consistent with the previously defined regulatory role of Csk and CD45 in the basal signaling pathway and our Western blot results (Figure 2.2). 45 Figure 2.5 Volcano plots for q-values versus peak area changes between Csk- or/and CD45- deficient cells and wild type Jurkat cells. Each volcano plot represents the comparison of a specific cell clone (as marked in title respectively) with wild type Jurkat cell line. Cyan points are considered to be statistically significant with q-values < 0.01. 2.3.2 Lck-mediated phosphorylation events in the basal T cell signaling pathway and actin cytoskeletal regulators Among all the confidently identified phsphopeptides in this study, 158 tyrosine phosphorylation sites on 49 proteins were found within the subset of proteins annotated 46 in KEGG as T cell receptor signaling pathway proteins, of which 134 unique tyrosine phosphorylation sites on 38 proteins showed statistically significant changes (q-value < 0.01) between wild type Jurkat cells and at least one clone of Csk- or/and CD45- deficient cell line (Figure 2.6). Western blot quantitation (Figure 2.2) showed that compared to wild type Jurkat cells, the phosphorylation of inhibitory tyrosine Lck Y505 was reduced in both J.Csk and J.Csk/CD45 cells because of the loss of Csk kinase activity, whereas Y505 is hyperphosphorylated in J.CD45 cells. As a result, the phosphorylation of Lck Y394 was elevated in J.Csk, J.Csk/CD45 cells and down-regulated in J.CD45 cells. Our phosphoproteomics data were exactly consistent with this observation on Lck Y394. However, for Lck Y505, although we also detected a nearly 30-fold increase in phosphorylation among all the J.CD45 cell clones, the reproducible quantitation of this phosphopeptide was hampered by its low abundance in the J.Csk nor J.Csk/CD45 clones (30). Being located at the C terminus of the Lck protein, the tryptic peptide containing this site does not contain any arginine or lysine amino acids, reducing the peptide's ionization efficiency. Thus, the peak areas of this peptide were extremely low (E3 level compared to the most common E5-E7 level) in all the cell lines except for the J.CD45 clones. 47 48 Figure 2.6 Quantitative phosphoproteomic analysis of known TCR signaling proteins. Listed is a portion of the data collected, representing proteins annotated in KEGG as T cell receptor signaling proteins. Ratio heatmap is generated from five biological replicate experiments. Fold changes of phosphopeptide abundance between a specific cell clone (three each for J.Csk, J.CD45, J.Csk/CD45) and wild type Jurkat clone E6-1 are represented according to the ratio heatmap color key. Black signifies no change. Red represents reduced phosphorylation while green represents elevated phosphorylation in specific Jurkat-derived cell lines compared to wild type Jurkat. A white dot on a ratio heatmap square indicated that a significant difference (q-value < 0.01) was observed. Because the Lck kinase activity largely depends on the phosphorylation of Y394, we identified a series of statistically significant changes (elevated in J.Csk, J.Csk/CD45 cells and down-regulated in J.CD45 cells with q-value < 0.01) in phosphorylation of downstream adaptor and signaling proteins that depend on Lck kinase activity. These tyrosine sites and proteins included important TCR signaling components, such as ZAP-70 (Y492, Y493), SLP-76 (Y483, Y532), Vav1 (Y541, Y791), PI3K (regulatory subunit α Y556, regulatory subunit γ Y407, catalytic subunit β Y505), Itk (Y146, Y512), PLC-γ1 (Y783, Y833), NF-κB1 (Y241). (30-33) These observations supported the previous suggestion that the basal signaling pathway shares components with the conventional TCR signaling. The absence of Csk generates ligand-independent TCR-like signals via the activation of Lck (20). We also detected significant elevation in J.Csk, J.Csk/CD45 cells and down-regulation in J.CD45 cells in phosphopeptide abundance of SHP-1 (Y564) and CBL (Y552, Y674). SHP-1 and CBL are proteins reported to function in negative feedback loops in TCR signaling (34, 35). In the basal signaling pathway, 49 these two proteins may play a role as negative regulators, to counteract the positive regulation and establish the steady-state level of signaling in unstimulated T cells. Our results also suggested the interplay between basal T cell signaling and proteins involved in the actin cytoskeletal pathway. Actin cytoskeletal dynamics is fundamental in many cellular processes such as cell motility, cell signaling, and cytokinesis. The polymerization and rearrangement of the actin filaments is necessary for the formation of immunological synapse (36) and regulated by multiple signaling proteins including Lck and Itk (37, 38). In this study, a total of 95 tyrosine phosphorylation sites on 40 proteins were found within the subset of proteins annotated in KEGG as actin cytoskeletal pathway proteins, of which 72 unique tyrosine phosphopeptides on 27 proteins showed statistically significant changes (q-value < 0.01) between wild type Jurkat cells and at least one clone of Csk- or/and CD45- deficient cell lines (Figure 2.7). A portion of these tyrosine phosphorylation sites showed the same directionality of change as Lck Y394 (significantly up-regulated in J.Csk, J.Csk/CD45 cells and significantly down-regulated in J.CD45 cells), including ACTC (Y55, Y93), ARHGEF6 (Y644, Y666), CYFIP-1 (Y108), MYPT1 (Y762, Y758), Paxillin (Y88, Y118), PIP5K3 (Y1772), Profilin-1 (Y129), Profilin-2 (Y99), PPP1CB (Y304, Y306), and ROCK2 (Y722). Some of these sites have been previously reported to be stimulated upon TCR ligation (39, 40). Therefore, in the basal tone of unstimulated T cells, actin cytoskeletal dynamics may also 50 Figure 2.7 Quantitative phosphoproteomic analysis of actin cytoskeleton proteins. Listed is a portion of the data collected, representing proteins annotated in KEGG as actin cytoskeletal pathway proteins. Ratio heatmap is generated from five biological replicate experiments. Fold changes of phosphopeptide abundance between a specific cell clone (three each for J.Csk, J.CD45, J.Csk/CD45) and wild type Jurkat clone E6-1 are represented according to the ratio heatmap color key. Black signifies no change. Red represents reduced phosphorylation while green represents elevated phosphorylation in specific Jurkat-derived cell lines compared to wild type Jurkat. A white dot on a ratio heatmap square indicated that a significant difference (q-value < 0.01) was observed. 51 be regulated by mechanisms very similar to those involved in TCR inducible signaling pathway and Lck is clearly an important mediator in this process. To identify more tyrosine phosphorylation sites probably participating in the Lck-mediated signaling events in unstimulated T cells, we explored all the data in addition to those annotated in KEGG as T Cell Receptor Signaling or actin cytoskeletal pathway proteins. To be considered as regulated by Lck kinase activity, selected candidates were required to meet the criteria as below: compared to wild type Jurkat, the phosphorylation level of tyrosine sites in each cell clone of J.Csk and J.Csk/CD45 showed > 2-fold change with q-value < 0.05; whereas in each clone of J.CD45, the same sites showed < 0.5-fold change with q-value < 0.05 or could not been detected. In total, 117 tyrosine phosphorylation sites fulfilled all of these cutoffs (Table 2.1) and could be interesting for future biological study of Lck-mediated T cell basal signaling pathway. 52 Table 2.1 List of tyrosine phosphorylation sites related to Lck kinase activity* * Fold change in this table represents the average value of three cell clones from five biological replicate experiments. 53 2.3.3 A possible co-regulatory role of Lck and CD45 in the integrin-mediated T cell signaling pathway The Csk- or/and CD45- deficient cell lines employed in this project provided comprehensive information about the pathway targets of Csk or CD45. For example, we hardly identified any sites showing significantly reduced phosphorylation level in all clones of both J.Csk and J.Csk/CD45 cells except for Lck Y505, which was suggested to be the only protein target of Csk kinase by previous studies (6, 41). Thus, our results support the conclusion that Csk specifically phosphorylates the inhibitory site Lck Y505 and negatively regulates all the downstream signaling events via Lck kinase activity. Similarly, if some tyrosine sites were the pathway substrates of CD45 phosphatase in a Csk/Lck-independent manner, we would expect to detect significantly increased phosphorylation in all clones of both J.CD45 and J.Csk/CD45 cells with a similar magnitude. In our data, Fyn Y185, Y213, Y214 showed this regulatory pattern. These three sites are located within the SH2 domain of Fyn and may interact with and activate downstream signaling molecules. The highly homologous and similarly located, Y190 in Lck SH2 domain was not found up-regulated upon the deficiency of CD45. Thus, the regulation by CD45 of the SH2 domain of Src family kinases may be specifically targeted to Fyn instead of Lck. Interestingly, among proteins annotated in KEGG as T cell receptor signaling or actin cytoskeletal pathway, we identified a series of tyrosine sites (Figure 2.8), on which 54 > 20 <-20 -11 11 Jurkat Derived Cells Jurkat Derived Cells -4 4 1 Wild Type Jurkat Wild Type Jurkat Ratio Heatmap Fold Change Key J.Csk/ Protein Phospho- J.Csk J.CD45 CD45 Average Fold Change of 3 Clones Name sites 2B10 2F10 1G11 3C10 J.Csk J.CD45 J.Csk/CD45 3A2 1C5 4G2 4A8 2F5 CD31 Y713 1.8 2.5 31.4 CD31 Y690 29.4 8.8 374.3 Cofilin Y140 2.9 1.2 16.3 IGF-1 Receptor Y1165 8.1 2.8 30.4 Integrin β1 Y783 4.9 0.8 8.6 Integrin β1 Y795 2.1 1.3 14.2 JAM1 Y280 3.8 1.5 93.8 PAK1 Y131 22.4 7.7 111.5 PAK2 Y130 2.9 3.2 13.9 PIP5K1A Y42 59.3 4.3 456.0 PIP5K1A Y483 3.0 2.0 54.3 PKB-β Y475 4.9 5.5 86.9 PKC-α Y365 1.9 3.5 17.7 PKC-β1 Y368 8.9 11.2 70.6 N-WASp Y256 2.4 3.6 11.1 SMAD2 Y151 1.5 5.7 31.1 SAP97 Y399 5.8 4.7 38.3 Figure 2.8 Tyrosine sites that are highly phosphorylated in J.Csk/CD45 cells. Proteins in this list are annotated in KEGG as T Cell Receptor Signaling or actin cytoskeletal pathway proteins. In the ratio heatmap, black signifies no change. Red represents reduced phosphorylation while green represents elevated phosphorylation in signaling protein deficient Jurkat-derived cell lines compared to wild type Jurkat. A white dot on a square indicated that a significant difference (q-value < 0.01) was observed. Fold change in the table represents the average value of three cell clones from five biological replicate experiments. the phosphorylation levels were moderately elevated (or even not significantly changed in some clones) in either J.CSK or J.CD45 cells, but extremely elevated in J.Csk/CD45 cells (many of them showed greater than 50-fold change). This unique pattern in the cells lacking both proteins differs significantly from the pattern on the activation site of Lck Y394 and suggests the regulation of independent pathways. We found that integrin β1 (Y783, Y795) and PECAM-1 (Y690, Y713) may play a central role amongst the phosphopeptides showing this pattern since they are reported to interact with most of the 55 other similarly regulated proteins. The precise pathways regulated by Csk and CD45 in this group of proteins remains to be elucidated. Integrin β1 (also known as CD29) is a subunit of the broader integrin family of proteins, which are heterodimeric cell surface adhesion receptors (42). Upon binding the cell or extracellular matrix (ECM) ligands, activated integrins control cytoskeletal remodeling and initiate signaling cascades that regulate cell motility and growth (43). Integrin-mediated cell adhesion is important for the migration of T lymphocytes into sites of inflammation or secondary lymphoid organs (44). Y783 and Y795 of integrin β1 are located in the cytoplasmic tail and known to be essential for the coupling of integrin β1 to the actin cytoskeleton and transducing integrin signals (43). Thomas et al. proposed a mechanism by which CD45 negatively regulates integrin-mediated adhesion via dephosphorylation of the high affinity binding sites in the SH2 and SH3 domains of Src-family kinases (SFKs), because the phosphorylation of these sites are important for their recruitment to the integrin binding sites (45). This regulation is unrelated to the C-terminal inhibitory sites or activation loop sites of SFKs. Our data supports this mechanism and further specifies the possible sites involved in this process. It may be Fyn SH2 domain (Y185, Y213, Y214), not Lck, that activate integrin by binding and phosphorylating integrin β1 (Y783, Y795). Independent of this pathway, our results are consistent with a model that elevated Lck kinase activity from the deficiency of Csk could also stimulate integrin β1 Y783 and Y795. Woods et al. have demonstrated that 56 upon TCR stimulation, activation of Lck led to increasing integrin β1-mediated cell adhesion via PI3K-dependent relocalization and activation of Itk (44). In summary, the observed hyperphosphorylaton of integrin β1 in J.Csk/CD45 cells may result from the elevated activity of either Lck (from Csk-deficiency) or Fyn (from CD45-deficiency). PECAM-1 Y690 and Y713 are another possible target of Lck and Fyn synergistic regulation. PECAM-1 (also known as CD31) is a transmembrane endothelial adhesion protein that negatively regulates cell migration by tyrosine phosphorylation (46). These two sites have been reported to specifically bind, in a tyrosine phosphorylation dependent manner, to the SH2 domain of Fyn rather than PLC-γ or PI3K. The molecular mechanism of this interaction is that both Tyr690-Thr-Glu-Val and Tyr713-Ser-Glu-Val motifs of PECAM-1 contain an acidic residue at the pTyr+2 and a hydrophobic residue at the pTyr+3 position, meeting the sequence requirements for binding to the Fyn SH2 domain (47, 48). PECAM-1 Y690 and Y713 were also found phosphorylated upon overexpression of Lck, although the mechanism was unclear (49). Furthermore, numerous studies have demonstrated the mutual modulation of PECAM-1 and integrin β1 activities via a set of signaling molecules (50, 51). These two proteins are highly important mediators transducing bidirectional signals: outside-in signaling (by binding ECM ligands to stimulate downstream intracellular signaling) and inside-out signaling (by binding cytoplasmic proteins to change the integrin affinity for ECM ligands). 57 Most other proteins containing hyperphosphorylated tyrosine sites in J.Csk/CD45 cells were found related to integrin-mediated signaling transduction. For example, PKC-α is a key protein mediating integrin-dependent cell migration. The observed hyperphosphorylated PKC-α Y365 is located within the kinase domain and spatially adjacent to 12-amino-acid motif (aa 313 to 325) within the V3 hinge domain, a key region for association with integrin β1 to regulate cell polarization and directional motility (52). Thus, PKC-α Y365 could be the site to stimulate downstream signals upon activation of integrin β1. PKC-β1 was also hyperphosphorylated at kinase domain Y368 in the J.Csk/CD45 cells. Although the study of this site is very limited, it may play a similar role as PKC-α Y365 considering their high homology as protein kinase C family members and the similar position in the protein structure. Another identified integrin-associated site is JAM-1 Y280. Its phosphorylation could recruit Csk to the integrin-Src complex and inhibit the Src kinase activity (53). Thus, our results suggested a possible mechanism that activated Lck strengthens while CD45 removes this negative feedback regulation of Lck kinase activity respectively via modulating the phosphorylation of integrin Y783 and Y795. Moreover, the hyperphosphorylated sites PIP5K1A (PIP 5-kinase type-1 α, Y42, Y483), cofilin (Y140), and N-WASp (Y256) are functionally related to integrin-mediated actin assembly. Upon, integrin activation, PIP 5-kinase is phosphorylated to bind and activate talin, which in turn elevates the synthesis of phosphatidylinositol 58 4,5-bisphosphate (PI(4,5)P2) (54). PI(4,5)P2 can then inhibit the activity of cofilin, an important actin-binding protein regulating actin filament dynamics and reorganization, by phosphorylating Ser-3 near the N-terminus to slow down actin filament depolymerization in cells (55). Our data indicated another possible phosphorylation site Y140 that may also be involved in cofilin-regulated actin dynamics. On the other hand, PI(4,5)P2 stimulates Arp2/3 complex-induced actin polymerization by phosphorylating N-WASp Y256 (56), which was also identified in this project. Other hyperphosphorylated sites in J.Csk/CD45 cells are important components of various signaling pathways. The elevated phosphorylation on these sites may result from either the direct co-regulation of Lck and CD45/Fyn or the convergence of integrin-mediated signaling with other pathways. PAK1 (p21 activated kinase 1, Y131) and PAK2 (Y130) were reported as key intermediate in p21 GTPase-dependent signaling pathways to mediate the interactions between small G proteins and intracellular kinases. Y131 and Y130 are the major phosphoacceptor sites of PAK1 and PAK2 respectively (57). On the other aspect, PKB-β and GSK3B are both involved in the PI3K/Akt pathway, which is activated by IGF-1 receptor signaling (58). PKB-β inhibits the activity of GSK3B by reducing the phosphorylation of Y216 (59), which could be detected in this project as a 0.3-fold significant decrease in J.Csk/CD45 cells. Thus, the identified PKB-β Y475 may mediate this negative regulation. 59 Finally, we explored all the data in addition to those annotated in the KEGG as T Cell Receptor Signaling or actin cytoskeletal pathway proteins, to select hyperphosphorylated sites in J.Csk/CD45 cells meeting the criteria as below: compared to wild type Jurkat, the phosphorylation level of tyrosine sites in each cell clone of J.Csk and J.CD45 showed > 0.5-fold change and < 20-fold change; whereas in each clone of J.Csk/CD45, the same sites showed > 5-fold change with q-value < 0.05. Moreover, the average ratio of three J.Csk/CD45 clones was required to show > 10-fold change. In total, 103 tyrosine phosphorylation sites fulfilled all of these cutoffs (Table 2.2) and could help the systematic understanding of integrin-mediated signaling pathways that is synergistically regulated by Lck and CD45/Fyn. For example, talin was reported to link integrins to the actin cytoskeleton. Its function is fulfilled by association with PIP 5-kinase to elevate the (PI(4,5)P2) level (54). Talin Y26 listed in Table 2.2 may participate in this interaction. 60 Table 2.2 List of hyperphosphorylated tyrosine sites in J.Csk/CD45 cells* * Fold change in this table represents the average value of three cell clones from five biological replicate experiments. 61 2.4 CONCLUSION With quantitative proteomic analysis of phosphorylation events in Csk and CD45 deficient cell lines, we confidently identified a series of tyrosine phosphopeptides involved in the Lck-mediated T cell basal signaling pathway. From the proteomics view, we validated the high overlap of signal components between basal and TCR-induced signaling pathway. Unlike the regulatory role of Csk simply on Lck kinase activity, CD45 showed a more complicated pattern of substrate specificity based on our results. We proposed a new model of negative regulation of Fyn SH2 domain phosphorylation by CD45. We identified Fyn Y185, Y213, and Y214 as potential sites of CD45 regulation. Furthermore, we detected many hyperphosphorylated tyrosine sites in J.Csk/CD45 cells, which could be explained by the synergistic regulation of activated Lck kinase loop (from Csk-deficiency) and Fyn SH2 domain (from CD45-deficiency). Proteins with this pattern of regulation were found to have over-representation in proteins with functions in integrin-mediated signaling and regulation of cytoskeletal dynamics. Since the hyperphosphorylation observed in J.Csk/CD45 cells was highly unique resulting from the knockdown of two important regulatory proteins, more than 100 identified tyrosine sites showing this regulation pattern may share common pathway regulation. By further analyzing these sites individually through directed mutagenesis, we will gain a more systematic view of basal signaling cascades occurring in T cells. 62 2.5 REFERENCES 1. Palacios, E. H., and Weiss, A. (2004) Function of the Src-family kinases, Lck and Fyn, in T-cell development and activation. Oncogene 23, 7990-8000 2. Thomas, S. M., and Brugge, J. S. (1997) Cellular functions regulated by Src family kinases. Annu Rev Cell Dev Bi 13, 513-609 3. Ley, S. C., Marsh, M., Bebbington, C. R., Proudfoot, K., and Jordan, P. (1994) Distinct Intracellular-Localization of Lck and Fyn Protein-Tyrosine Kinases in Human T-Lymphocytes. J Cell Biol 125, 639-649 4. Caron, L., Abraham, N., Pawson, T., and Veillette, A. (1992) Structural Requirements for Enhancement of T-Cell Responsiveness by the Lymphocyte-Specific Tyrosine Protein-Kinase P56lck. Mol Cell Biol 12, 2720-2729 5. Abraham, K. M., Levin, S. D., Marth, J. D., Forbush, K. A., and Perlmutter, R. M. (1991) Thymic Tumorigenesis Induced by Overexpression of P56lck. P Natl Acad Sci USA 88, 3977-3981 6. Veillette, A., Latour, S., and Davidson, D. (2002) Negative regulation of immunoreceptor signaling. Annu Rev Immunol 20, 669-707 7. Schmedt, C., Saijo, K., Niidome, T., Kuhn, R., Aizawa, S., and Tarakhovsky, A. (1998) Csk controls antigen receptor-mediated development and selection of T-lineage cells. Nature 394, 901-904 8. Schmedt, C., and Tarakhovsky, A. (2001) Autonomous maturation of alpha/beta T lineage cells in the absence of COOH-terminal Src kinase (Csk). J Exp Med 193, 815-825 9. Hermiston, M. L., Xu, Z., and Weiss, A. (2003) CD45: A critical regulator of signaling thresholds in immune cells. Annu Rev Immunol 21, 107-137 10. Byth, K. F., Conroy, L. A., Howlett, S., Smith, A. J. H., May, J., Alexander, D. R., and Holmes, N. (1996) CD45-null transgenic mice reveal a positive regulatory role for CD45 in early thymocyte development, in the selection of CD4(+)CD8(+) thymocytes, and in B cell maturation. J Exp Med 183, 1707-1718 63 11. Stone, J. D., Conroy, L. A., Byth, K. F., Hederer, R. A., Howlett, S., Takemoto, Y., Holmes, N., and Alexander, D. R. (1997) Aberrant TCR-Mediated signaling in CD45-null thymocytes involves dysfunctional regulation of Lck, Fyn, TCR-zeta and ZAP-70. J Immunol 158, 5773-5782 12. Pingel, S., Baker, M., Turner, M., Holmes, N., and Alexander, D. R. (1999) The CD45 tyrosine phosphatase regulates CD3-induced signal transduction and T cell development in recombinase-deficient mice: restoration of pre-TCR function by active p56(lck). Eur J Immunol 29, 2376-2384 13. Zamoyska, R., Basson, A., Filby, A., Legname, G., Lovatt, M., and Seddon, B. (2003) The influence of the src-family kinases, Lck and Fyn, on T cell differentiation, survival and activation. Immunol Rev 191, 107-118 14. Zikherman, J., Jenne, C., Watson, S., Doan, K., Raschke, W., Goodnow, C. C., and Weiss, A. (2010) CD45-Csk Phosphatase-Kinase Titration Uncouples Basal and Inducible T Cell Receptor Signaling during Thymic Development. Immunity 32, 342-354 15. Hermiston, M. L., Zikherman, J., and Zhu, J. W. (2009) CD45, CD148, and Lyp/Pep: critical phosphatases regulating Src family kinase signaling networks in immune cells (vol 228, pg 288, 2009). Immunol Rev 229, 387-387 16. Tan, Y. X., Zikherman, J., and Weiss, A. (2013) Novel tools to dissect the dynamic regulation of TCR signaling by the kinase Csk and the phosphatase CD45. Cold Spring Harb Symp Quant Biol 78, 131-139 17. Becker, A. M., DeFord-Watts, L. M., Wuelfing, C., and van Oers, N. S. (2007) The constitutive tyrosine phosphorylation of CD3zeta results from TCR-MHC interactions that are independent of thymic selection. J Immunol 178, 4120-4128 18. Stefanova, I., Dorfman, J. R., and Germain, R. N. (2002) Self-recognition promotes the foreign antigen sensitivity of naive T lymphocytes. Nature 420, 429-434 19. Artomov, M., Kardar, M., and Chakraborty, A. K. (2010) Only signaling modules that discriminate sharply between stimulatory and nonstimulatory inputs require basal signaling for fast cellular responses. J Chem Phys 133, 105101 20. Roose, J. P., Diehn, M., Tomlinson, M. G., Lin, J., Alizadeh, A. A., Botstein, D., Brown, P. O., and Weiss, A. (2003) T cell receptor-independent basal signaling via Erk 64 and Abl kinases suppresses RAG gene expression. PLoS Biol 1, E53 21. Dobenecker, M. W., Schmedt, C., Okada, M., and Tarakhovsky, A. (2005) The ubiquitously expressed Csk adaptor protein Cbp is dispensable for embryogenesis and T-cell development and function. Mol Cell Biol 25, 10533-10542 22. Borger, J. G., Filby, A., and Zamoyska, R. (2013) Differential polarization of C-terminal Src kinase between naive and antigen-experienced CD8+ T cells. J Immunol 190, 3089-3099 23. Nguyen, V., Cao, L., Lin, J. T., Hung, N., Ritz, A., Yu, K., Jianu, R., Ulin, S. P., Raphael, B. J., Laidlaw, D. H., Brossay, L., and Salomon, A. R. (2009) A new approach for quantitative phosphoproteomic dissection of signaling pathways applied to T cell receptor activation. Molecular & cellular proteomics : MCP 8, 2418-2431 24. Yu, K., and Salomon, A. R. (2009) PeptideDepot: flexible relational database for visual analysis of quantitative proteomic data and integration of existing protein information. Proteomics 9, 5350-5358 25. Yu, K., and Salomon, A. R. (2010) HTAPP: high-throughput autonomous proteomic pipeline. Proteomics 10, 2113-2122 26. Elias, J. E., and Gygi, S. P. (2007) Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature methods 4, 207-214 27. Yu, K., Sabelli, A., DeKeukelaere, L., Park, R., Sindi, S., Gatsonis, C. A., and Salomon, A. (2009) Integrated platform for manual and high-throughput statistical validation of tandem mass spectra. Proteomics 9, 3115-3125 28. Beausoleil, S. A., Villen, J., Gerber, S. A., Rush, J., and Gygi, S. P. (2006) A probability-based approach for high-throughput protein phosphorylation analysis and site localization. Nature biotechnology 24, 1285-1292 29. Demirkan, G., Yu, K., Boylan, J. M., Salomon, A. R., and Gruppuso, P. A. (2011) Phosphoproteomic profiling of in vivo signaling in liver by the mammalian target of rapamycin complex 1 (mTORC1). PloS one 6, e21729 30. Sjolin-Goodfellow, H., Frushicheva, M. P., Ji, Q., Cheng, D. A., Kadlecek, T. A., 65 Cantor, A. J., Kuriyan, J., Chakraborty, A. K., Salomon, A. R., and Weiss, A. (2015) The catalytic activity of the kinase ZAP-70 mediates basal signaling and negative feedback of the T cell receptor pathway. Sci Signal 8, ra49 31. Ji, Q., Ding, Y., and Salomon, A. R. (2015) SRC homology 2 domain-containing leukocyte phosphoprotein of 76 kDa (SLP-76) N-terminal tyrosine residues regulate a dynamic signaling equilibrium involving feedback of proximal T-cell receptor (TCR) signaling. Molecular & cellular proteomics : MCP 14, 30-40 32. Leevers, S. J., Vanhaesebroeck, B., and Waterfield, M. D. (1999) Signalling through phosphoinositide 3-kinases: the lipids take centre stage. Curr Opin Cell Biol 11, 219-225 33. Helou, Y. A., Petrashen, A. P., and Salomon, A. R. (2015) Vav1 Regulates T-Cell Activation through a Feedback Mechanism and Crosstalk between the T-Cell Receptor and CD28. J Proteome Res 14, 2963-2975 34. Stefanova, I., Hemmer, B., Vergelli, M., Martin, R., Biddison, W. E., and Germain, R. N. (2003) TCR ligand discrimination is enforced by competing ERK positive and SHP-1 negative feedback pathways. Nat Immunol 4, 248-254 35. Murphy, M. A., Schnall, R. G., Venter, D. J., Barnett, L., Bertoncello, I., Thien, C. B., Langdon, W. Y., and Bowtell, D. D. (1998) Tissue hyperplasia and enhanced T-cell signalling via ZAP-70 in c-Cbl-deficient mice. Mol Cell Biol 18, 4872-4882 36. Dustin, M. L., and Cooper, J. A. (2000) The immunological synapse and the actin cytoskeleton: molecular hardware for T cell signaling. Nat Immunol 1, 23-29 37. Morgan, M. M., Labno, C. M., Van Seventer, G. A., Denny, M. F., Straus, D. B., and Burkhardt, J. K. (2001) Superantigen-induced T cell:B cell conjugation is mediated by LFA-1 and requires signaling through Lck, but not ZAP-70. J Immunol 167, 5708-5718 38. Dombroski, D., Houghtling, R. A., Labno, C. M., Precht, P., Takesono, A., Caplen, N. J., Billadeau, D. D., Wange, R. L., Burkhardt, J. K., and Schwartzberg, P. L. (2005) Kinase-independent functions for Itk in TCR-induced regulation of Vav and the actin cytoskeleton. J Immunol 174, 1385-1392 39. Ji, Q., and Salomon, A. R. (2015) Wide-scale quantitative phosphoproteomic analysis reveals that cold treatment of T cells closely mimics soluble antibody stimulation. J Proteome Res 14, 2082-2089 66 40. Helou, Y. A., Nguyen, V., Beik, S. P., and Salomon, A. R. (2013) ERK positive feedback regulates a widespread network of tyrosine phosphorylation sites across canonical T cell signaling and actin cytoskeletal proteins in Jurkat T cells. PloS one 8, e69641 41. Cooper, J. A., and Howell, B. (1993) The When and How of Src Regulation. Cell 73, 1051-1054 42. Kinashi, T. (2005) Intracellular signalling controlling integrin activation in lymphocytes. Nat Rev Immunol 5, 546-559 43. Simpson, M. A., Bradley, W. D., Harburger, D., Parsons, M., Calderwood, D. A., and Koleske, A. J. (2015) Direct interactions with the integrin beta1 cytoplasmic tail activate the Abl2/Arg kinase. J Biol Chem 290, 8360-8372 44. Woods, M. L., and Shimizu, Y. (2001) Signaling networks regulating beta1 integrin-mediated adhesion of T lymphocytes to extracellular matrix. J Leukoc Biol 69, 874-880 45. Thomas, M. L., and Brown, E. J. (1999) Positive and negative regulation of Src-family membrane kinases by CD45. Immunol Today 20, 406-411 46. Gratzinger, D., Barreuther, M., and Madri, J. A. (2003) Platelet-endothelial cell adhesion molecule-1 modulates endothelial migration through its immunoreceptor tyrosine-based inhibitory motif. Biochem Biophys Res Commun 301, 243-249 47. Masuda, M., Osawa, M., Shigematsu, H., Harada, N., and Fujiwara, K. (1997) Platelet endothelial cell adhesion molecule-1 is a major SH-PTP2 binding protein in vascular endothelial cells. FEBS Lett 408, 331-336 48. Songyang, Z., Shoelson, S. E., Chaudhuri, M., Gish, G., Pawson, T., Haser, W. G., King, F., Roberts, T., Ratnofsky, S., Lechleider, R. J., and et al. (1993) SH2 domains recognize specific phosphopeptide sequences. Cell 72, 767-778 49. Newman, P. J., and Newman, D. K. (2003) Signal transduction pathways mediated by PECAM-1: new roles for an old molecule in platelet and vascular cell biology. Arterioscler Thromb Vasc Biol 23, 953-964 67 50. Newman, P. J. (1997) The biology of PECAM-1. J Clin Invest 99, 3-8 51. Jackson, D. E., Ward, C. M., Wang, R., and Newman, P. J. (1997) The protein-tyrosine phosphatase SHP-2 binds platelet/endothelial cell adhesion molecule-1 (PECAM-1) and forms a distinct signaling complex during platelet aggregation. Evidence for a mechanistic link between PECAM-1- and integrin-mediated cellular signaling. J Biol Chem 272, 6986-6993 52. Parsons, M., Keppler, M. D., Kline, A., Messent, A., Humphries, M. J., Gilchrist, R., Hart, I. R., Quittau-Prevostel, C., Hughes, W. E., Parker, P. J., and Ng, T. (2002) Site-directed perturbation of protein kinase C- integrin interaction blocks carcinoma cell chemotaxis. Mol Cell Biol 22, 5897-5911 53. Naik, M. U., Caplan, J. L., and Naik, U. P. (2014) Junctional adhesion molecule-A suppresses platelet integrin alphaIIbbeta3 signaling by recruiting Csk to the integrin-c-Src complex. Blood 123, 1393-1402 54. DeMali, K. A., Wennerberg, K., and Burridge, K. (2003) Integrin signaling to the actin cytoskeleton. Curr Opin Cell Biol 15, 572-582 55. Mizuno, K. (2013) Signaling mechanisms and functional roles of cofilin phosphorylation and dephosphorylation. Cell Signal 25, 457-469 56. Wu, X., Suetsugu, S., Cooper, L. A., Takenawa, T., and Guan, J. L. (2004) Focal adhesion kinase regulation of N-WASP subcellular localization and function. J Biol Chem 279, 9565-9576 57. Mayhew, M. W., Jeffery, E. D., Sherman, N. E., Nelson, K., Polefrone, J. M., Pratt, S. J., Shabanowitz, J., Parsons, J. T., Fox, J. W., Hunt, D. F., and Horwitz, A. F. (2007) Identification of phosphorylation sites in betaPIX and PAK1. J Cell Sci 120, 3911-3918 58. Kang, J. H., Song, K. H., Woo, J. K., Park, M. H., Rhee, M. H., Choi, C., and Oh, S. H. (2011) Ginsenoside Rp1 from Panax ginseng exhibits anti-cancer activity by down-regulation of the IGF-1R/Akt pathway in breast cancer cells. Plant Foods Hum Nutr 66, 298-305 59. Wang, L., Zuo, B., Xu, D., Ren, Z., Zhang, H., Li, X., Lei, M., and Xiong, Y. (2012) Alternative splicing of the porcine glycogen synthase kinase 3beta (GSK-3beta) gene with differential expression patterns and regulatory functions. PloS one 7, e40250 68 69 Chapter 3 STANDARD PRESSURE LC-MS/MS USING 1.9 µm C18 PARTICLES AT ROOM TEMPERATURE IMPROVES QUANTIFICATION REPRODUCIBILITY WITH INCREASED PROTEOMICS SEQUENCING DEPTH 69 3.1 INTRODUCTION In recent decades, mass spectrometry (MS)-based proteomics has played an indispensable role in wide-scale analysis of complex protein mixtures. The technology has been employed in the wide-scale characterization of post-translational modifications and protein interactions (1, 2). A major challenge in proteomic identification and quantitation is the large dynamic range of protein abundance contained in cellular lysates: four orders of magnitude in yeast and at least seven orders of magnitude in human cells (3, 4). Improvements in peptide separation technology are crucial for complex proteome characterization. To analyze the proteome at increased depth, two-dimensional separation based on gels or liquid chromatography have been used (5-8). With these methods, LC-MS/MS is preceded by an additional mode of fractionation, such as strong cation exchange (SCX), SDS-PAGE, isoelectric focusing (IEF), and high pH RPLC. Although these methods can reduce the complexity of the sample during LC-MS/MS and thereby increase sequencing depth, they do require more starting material (up to mg level) and each additional separation step decreases the quantitative reproducibility of replicate analyses. In recent years, improvements in instrument sampling rates such as with the Orbitrap mass spectrometer and improved chromatographic peptide separations have led some researchers to explore the utility of complex proteome characterization without 70 peptide prefractionation (3, 9, 10). Compared to traditional 3 µm or 5 µm reversed-phase C18 particles, sub-2 µm C18 particles decrease the diffusion distances of solutes while minimizing the band spreading and improving column efficiency of reversed-phase separations. Another strategy is to operate a long column (50 cm or even longer), which provides more theoretical plates, with a long LC gradient (several hours or more) (11). A major technical difficulty in using sub-2 µm particles and long columns is the high column backpressure, which can be overcome by ultra-high pressure liquid chromatography (UHPLC) (12-14) or a column heater (11). In this report, an improved strategy for analytical column fabrication that reduces backpressure using sub-2 µm particles was explored. Although many recent reports have impressively expanded the depth of proteomic sequencing, far fewer reports have examined the relationship between sequencing depth and reproducibility of proteomic quantitation through replicate analyses. Although prefractionation of proteomic samples prior to LC-MS can increase the yield of peptide identifications, this strategy complicates the quantitative data analysis and increases the overall cost and instrument time required. For analyses without prefractionation, minimization of the necessity for specialized equipment can provide higher accessibility of sensitive proteomic capabilities to more labs. A true wide-scale understanding of a biological system requires statistical analysis of biological replicates to correctly understand biological variation. Mass spectrometry-based protein quantification methods 71 can be either label-free or incorporate a stable isotope label such as stable isotope labels with amino acids in cell culture (SILAC) (15), or isobaric tags for relative and absolute quantitation (iTRAQ) (16). Optimization of quantitative reproducibility with each technique while maintaining the highest possible sequencing depth will maximize the number of biologically relevant findings. In this project, we investigate the sensitivity and quantitative reproducibility of a newly designed 50 cm-long fritless column packed with 1.9 µm C18 particles. The novel application of this column configuration with 1.9 µm particles permits high-resolution peptide separations using a conventional HPLC system at standard pressures (< 300 Bar) under room temperature while providing highly reproducible quantitation. 3.2 MATERIALS AND METHODS 3.2.1 Cell Culture, T cell stimulation and Lysis Jurkat clone E6-1 was obtained from American Tissue Culture Collection (Manassas, VA). The cells were maintained in RPMI 1640 medium (Gibco, Grand Island, NY) supplemented with 10% heat-inactivated undialyzed FBS (HyClone, Logan, UT), 2 mM L-glutamine, 100 U/ml penicillin G, and 100 µg/ml streptomycin (Gibco, Grand Island, NY) in a humidified incubator with 5% CO2 at 37 °C. Cells were grown for eight doublings before harvest, and then placed in lysis buffer (8 M urea, 1 mM sodium orthovanadate, 20 mM HEPES, 2.5 mM sodium pyrophosphate, 1 mM 72 β-glycerophosphate, pH 8.0) for 20 min at 4 °C. The lysate was sonicated at a 30 watt output with 2 bursts of 30 sec each and cleared at 14,000 g for 15 min at 4 °C. (For the CD3/4 stimulated Jurkat cells, after harvest, cells were treated with anti-CD3 and anti-CD4 antibody in PBS (clone OKT3 and OKT4; eBioscience, San Diego, CA) for 3 min at 37 °C. Then, the cell lysis was performed as described above.) 3.2.2 Protein Reduction, Alkylation, Digestion and Desalting Protein concentration was measured using the DC Protein Assay (Bio-Rad, Hercules, CA). Next, clear cell lysate was reduced with 45 mM DTT for 20 min at 60 °C and alkylated with 100 mM iodoacetamide for 15 min at room temperature in the dark. Cell lysate was then diluted 4-fold with 20 mM HEPES buffer, pH 8.0 and digested with sequencing grade modified trypsin (Promega, Madison, WI) in a 1:100 (w/w) trypsin : protein ratio overnight at room temperature. Typtic peptides were acidified to pH 2.0 with 20% trifluoroacetic acid (TFA), cleared at 1800 g for 5 min at room temperature, and desalted using C18 Sep-Pak plus cartridges (Waters, Milford, MA) as previously described (17), with the exception that TFA was used instead of acetic acid. Eluents containing peptides were lyophilized for 48 hr to dryness. 73 3.2.3 TiO2 Phosphopeptide Enrichment Phosphopeptides were enriched with Titansphere Phos-TiO tips (GL Sciences, Tokyo Japan) following the manufacturer’s protocol with some modifications. The condition buffers (containing TFA, CH3CN and lactic acid) and elution buffers (1% NH4OH in water and 40% CH3CN) were prepared first. Then, the condition buffer was added to the Phos-TiO tips (centrifuge at 3000g, 22 ˚C). Once the conditioning was finished, desalted tryptic peptides from Jurkat total lysates (CD3/4 stimulated or unstimulated) were mixed with a synthetic phosphoserine standard (FQpSEEQQQTEDELQDK, AnaSpec, San Jose, CA) at a ratio of 5 fmol standard : 1 µg sample. The mixture was loaded onto tips using centrifugation at 1000 g at 22 ˚C. After loading, the column was washed with condition buffers followed by elution buffers. Acetic acid was used to acidify TiO2 enriched samples, which were dried almost to completeness. 3.2.4 Fabrication of Packed Capillary Columns The fused-silica capillary tubing (360 µm O.D. x 75 µm I.D., Polymicro Technologies, Phoenix, AZ) with integrated nanospray tip was generated with a laser-based micropipette puller (Model P-2000, Sutter Instruments, Novato, CA). The inner diameter at the tip end was approximately 8 µm, when setting the parameters of laser puller as: heat 350, filament 0, velocity 18, delay 126 and pull strength 10. An 74 analytical column with integrated nanospray tip similar in configuration to the laser-pulled tip was also obtained from New Objective (FS360-75-8-N-5-C70, Woburn, MA). A slurry of 30-40 mg ReproSil-Pur C18-AQ 1.9µm particles (Dr. Maisch GmbH, Ammerbuch, Germany) in 1 ml methanol was sonicated for 10 min after incubation overnight (18). Then, the vial containing the slurry suspended with a micro-stirbar was placed in a stainless steel pressure bomb (GNF Commercial Systems). The column of 50 cm length was packed within 4-6 hr at ~1500 psi of helium. The 3 µm C18 columns were packed using the same protocol: 3 µm Monitor C18 particles were suspended in CH3CN containing 5% isopropanol and packed into 75 µm I.D fused-silica capillary tubing with integrated nanospray tip (PF360-75-10-N-5, New Objective, Woburn, MA). 3.2.5 LC-MS/MS Analysis Tryptic peptides were analyzed by a fully automated phosphoproteomic technology platform (19, 20). The nanoLC-MS/MS experiments were performed with an Agilent 1200 Series Quaternary HPLC system (Agilent Technologies, Santa Clara, CA) connected to a Q Exactive mass spectromter (Thermo Fisher Scientific, Waltham, MA). The lyophilized Jurkat tryptic peptides were reconstituted in buffer A (0.1 M acetic acid) at a concentration of 1 µg/µl and 2 µl was injected for each analysis. For the phosphopeptide replicate quantitation experiment, 100 µg tryptic peptides were loaded onto Phos-TiO tips. Then, the dried eluted peptides were reconstituted in 10 µl buffer A 75 and analyzed using LC-MS/MS. The peptides were separated through a linear reversed-phase gradient (2 hr, 3 hr, 4.5 hr, 6 hr, 8 hr) from 0% to 40% buffer B (0.1 M acetic acid in CH3CN) at a flow rate 60 nl/min for 1.9 µm C18 column (100 nl/min for 3µm C18 column) with an additional 1hr wash and re-equilibration time. The electrospray ion source was operated at 2.0 kv in a split flow configuration. The Q Exactive was operated in the data dependent mode using a top-9 data dependent method. Survey full scan MS spectra (m/z 400-1800) were acquired at a resolution of 70,000 with an AGC target value of 3×106 ions or a maximum ion injection time of 200 ms. Peptide fragmentation was performed via higher-energy collision dissociation (HCD) with the energy set at 28 NCE. The MS/MS spectra were acquired at a resolution of 17,500, with a targeted value of 2×104 ions or a maximum integration time of 250 ms. The underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 1.0%. The ion selection abundance threshold was set at 8.0×102 with charge state exclusion of unassigned and z =1, or 6-8 ions and dynamic exclusion time of 20 sec. 3.2.6 Data Analysis Peptide spectrum matching of MS/MS spectra from whole cell lysate Jurkat tryptic digest samples was performed against a human-specific database (UniProt; downloaded 2/1/2013) using MASCOT v. 2.4 (Matrix Science, Ltd, London W1U 7GB UK). A 76 concatenated database containing 144,156 “target” and “decoy reversed” sequences was employed to estimate the false discovery rate (FDR) (21). Msconvert from ProteoWizard (v. 3.0.5047), using default parameters and with the MS2Deisotope filter on, was employed to create peak lists for Mascot. Mascot database searches were performed with the following parameters: trypsin enzyme cleavage specificity, 2 possible missed cleavages, 10 ppm mass tolerance for precursor ions, 20 mmu mass tolerance for fragment ions. Search parameters permitted variable modification of methionine oxidation (+15.9949 Da), and static modification of carbamidomethylation (+57.0215 Da) on cysteine. The resulting peptide spectrum matches (PSMs) were reduced to sets of unique PSMs by eliminating lower scoring duplicates. Next, score thresholding was performed to reduce the sets’ FDRs to ≤ 1%. Protein identifications were derived from the PSMs using a script that employed strict parsimony (i.e., the minimum number of proteins that explained all of the PSMs), with the restriction that the protein score be greater than 20. Peak lists from the TiO2-enriched phoshopeptide experiments were generated using identical parameters to those created for the total lysate analysis. Mascot search parameters were identical to those for the total lysate analysis but with the addition of variable modification of phosphorylation (+79.9663 Da) on serine, threonine, and tyrosine residues. Mascot results were filtered by Mowse Score (>20). Peptide assignments from the database search were filtered down to 1% false discovery rate 77 (FDR) by a logistic spectral score, as previously described (21, 22). To validate the position of the phosphorylation sites, the Ascore algorithm (23) was applied to all data, and the reported phosphorylation site position reflected the top Ascore prediction. 3.2.7 Quantitation of Relative Phosphopeptide Abundance Relative quantification of phosphopeptide abundance was performed via calculation of selected ion chromatogram (SIC) peak areas. Retention time alignment of individual replicate analyses was performed as previously described (24). Peak areas were calculated by inspection of SICs using in-house software programmed in R 3.0 based on the Scripps Center for Metabolomics’ XCMS package (version 1.40.0). This approach performed multiple passes through XCMS’s central wavelet transformation algorithm (implemented in the centWave function) over increasingly narrower ranges of peak widths, and used the following parameters: mass window of 10 ppm, minimum peak widths ranging from 2 to 20 seconds, maximum peak width of 80 seconds, signal to noise threshold of 10 and detection of peak limits via descent on the non-transformed data enabled. For cases when centWave did not identify an MS peak, we used the getPeaks function available in XCMS to integrate in a pre-defined region surrounding the maximum intensity signal of the SIC. SIC peak areas were determined for every phosphopeptide that was identified by MS/MS. In the case of a missing MS/MS for a particular peptide, in a particular replicate, the SIC peak area was calculated according to 78 the peptide's isolated mass and the retention time calculated from retention time alignment. A minimum SIC peak area equivalent to the typical spectral noise level of 1000 was required of all data reported for label-free quantitation. Individual SIC peak areas were normalized to the peak area of the exogenously spiked phosphopeptide FQpSEEQQQTEDELQDK added prior to phosphopeptide enrichment and reversed-phase elution into the mass spectrometer. Quantitative analysis was applied to five replicate experiments. To select phosphopeptides that show a statistically significant change in abundance between CD3/4 stimulated and unstimulated cells, q-values for multiple hypothesis tests were calculated based on p-values from two-tailed unpaired Student’s t tests using the R package QVALUE as previously described (25, 26). 3.3 RESULTS 3.3.1 Fabrication of Fritless Column Packed with 1.9 µm C18 Particles To avoid the high backpressure caused by the existence of a frit, we managed to pack the 1.9µm C18 particles into a fritless integrated electrospray tip, which could be economically prepared from fused-silica capillary tubing with a laser puller. The shape and diameter of tip was optimized reproducibly by manipulating the laser puller parameters (Table 3.1). We found that the tip with 8 µm in diameter could successfully retain 1.9 µm C18 particles in column when applying methanol as slurry liquid for column packing. With this fritless strategy, a 15 cm-long 1.9 µm C18 analytical column could be 79 fabricated within 30 min under 1500 psi. The column packing bed was stable without observable loss after several weeks use. Table 3.1 Size of electrospray tip determined by laser puller parameters Heat Filament Velocity Delay Pull Tip Size a 420 0 20 130 10 4 µm 385 0 19 128 10 5~6 µm 350 0 18 126 10 8 µm a Tip size was estimated by comparing the tip with 5µm and 10µm C18 particles under microscope. LC-MS/MS tests with the Orbitrap Velos mass spectrometer indicated considerably better performance of this fritless 1.9 µm C18 analytical column than the traditional 3 µm C18 analytical column (Table 3.2) under the same analytical conditions (such as flow rate and LC gradient). The 1.9 µm column provided significantly narrower peaks and higher peptides identification yield. Its superior performance was consistent during replicate experiments (five replicates of BSA test and three replicates of Jurkat test). We also validated the reproducibility of our fabrication methodology by packing a batch of 1.9 µm columns, which showed nearly identical performance on LC separation and peptides identification. 80 Table 3.2 Comparison of 15 cm-long columns pack with 1.9 µm and 3 µm particles 15 cm-long 1.9 µm 15 cm-long 3 µm C18 Column C18 Column S*LHTLFGDELCK 1.91 / 6.34E7 3.11 / 5.64E7 FWHM and Peak area of Three Characteristic Peptides HLVDEPQNLIK 1.89 / 6.22E7 3.04 / 5.74E7 from 0.2 pmol BSA Test a LVNELTEFAK 1.80 / 1.02E8 3.00 / 9.09E7 Total PSMs (Unique PSMs) 26 min 2%-70% B 1981 (1682) 1376 (1127) identified from 2 µg Jurkat- 60 min 2%-70% B 3213 (2752) 2030 (1581) derived tryptic peptides under different LC gradients b 180 min 2%-70% B 4388 (3235) 3784 (2488) a b The results are the average value of five replicates. The results are the average value of three replicates. 3.3.2 Operation of an In-house Fabricated 50 cm-long Column Packed with 1.9 µm C18 Particles With the fritless packing strategy described above, a 50 cm-long 1.9 µm C18 column with integrated electrospray tip was successfully packed within several hours under a pressure of only 1500 psi (~100 Bar). Reversed phase separations were performed with a this column using an Agilent 1200 quaternary pump operated under 300 Bar (Figure 3.1) without the requirement of an UHPLC or a column heater. Although there was not a frit at the head of the column, there was no observable loss of packing bed after several weeks use. To verify the stability of this fritless column, the HPLC pressure was cycled 81 from 0 to ~260 bar a total of 100 times. The packing bed length and flow rate at the electrospray tip were unaltered through 100 cycles. Figure 3.1 HPLC separation profile of 50 cm-long/1.9 µm C18 column. Red: the pressure curve during a LC-MS/MS analysis; blue: % B at HPLC pump. LC gradient including 3 hr linear gradient (0%-40% B) with additional wash (up to 98% B) and equilibration (0% B); gray: total ion chromatogram (TIC) of the LC-MS/MS analysis of TiO2-enriched Jurkat tryptic peptides. The optimal reversed-phase gradient length of this newly constructed column was investigated (2 hr, 3 hr, 4.5 hr, 6 hr, 8 hr) with a whole cell lysate from the human T cell line Jurkat. The results indicated that a 3 hr linear gradient (0% to 40% buffer B) provided the best balance between sequencing depth and acquisition time (Table 3.3). Using a 3 hr gradient and triplicate analyses, 26,894 unique peptide-spectrum matches (PSMs) and 4165 protein groups were identified. 82 Table 3.3 Comparison of Jurkat peptide identification using in-house fabricated 50 cm- long, 1.9 µm C18 column under different LC gradients LC gradient Average yield from a single runa Aggregate yield from triplicate analyses time (hr) # Unique PSMs # Protein groups # Unique PSMs # Protein groups 2 15324 ± 128 2987 ± 25 22059 3630 3 18330 ± 168 3426 ± 38 26894 4165 4.5 20706 ± 408 3767 ± 52 29327 4620 6 21876 ± 369 3856 ± 56 30403 4679 8 22427 ± 559 3843 ± 71 28716 4638 a The value after “±” indicates the standard deviation of the triplicate experiments. 3.3.3 Evaluation of Quantitative Data Although the number of peptide identifications (PSM yield) is an important metric of any proteomic experiment, it is influenced by many variables, such as the type of cells analyzed and their mode of stimulation, the protocol used to prepare the sample, the HPLC gradients employed to separate the peptides and the capabilities of the mass spectrometer used to collect the data, and the methodology used for post-acquisition data analysis. Measures of quantitative reproducibility are essential to maximize extraction of biologically relevant insights from proteomic experiments. To determine the quantitative reproducibility of our newly constructed analytical column, Jurkat T cells stimulated with CD3/4 antibody crosslink were compared to control unstimulated cells after phosphopeptide enrichment by the TiO2 method from whole cell lysates. Using a 3 hr linear gradient (0% to 40% buffer B), a 50 cm column containing 1.9 µm particles was 83 compared to 15 and 50 cm columns constructed with 3 µm particles. The 15 cm, 3 µm C18 column was included since this configuration represents a commonly used column type employed in proteomic labs. For each specific column configuration, label-free quantitative analysis with retention time alignment (20, 24) was applied to five technical replicate phosphoproteomic experiments to quantitate phosphopeptides from control and CD3/4 stimulated Jurkat cells. All selected ion chromatogram peak areas from a given LCMS run were normalized to the signal intensity of an exogenously spiked phosphopeptide standard that was added to every sample at equal levels and copurified with the phosphopeptides. The results revealed that the number of unique phosphopeptides identified at a 1% FDR confidence level increased by 44% for unstimulated cells and 56% for CD3/4 stimulated cells using the column packed with 1.9 µm particles compared to 3 µm particles (Table 3.4). Table 3.4 Overview of phosphopeptide PSM yield with different analytical conditions Column Type Number of unique phosphopeptides a Gradient time (hr) Length (cm) Particle size (µm) Unstimulated CD3/4 stimulated 50 1.9 3 6293 ± 84 7502 ± 99 50 3 3 4344 ± 191 4782 ± 200 15 3 3 3939 ± 67 4249 ± 50 a Number of unique phosphopeptides is the average result of five replicates, value after “±” indicates the standard deviation. Compared to the 3 µm particles, the 1.9 µm particles decreased the SIC peak 84 spreading (Figure 3.2). Pairwise replicate analysis revealed that the 50 cm-long 1.9µm particle column provided a better degree of correlation (R > 0.91) than the 3 µm column, especially in the region of low peak areas (< 1E7) (Figure 3.3). Figure 3.2 Comparison of total ion chromatogram (TIC) and five representative peptide selected ion chromatograms (SIC) between two 50 cm-long columns packed with (A) 1.9 µm C18 particles and (B) 3 µm C18 particles. The data was obtained from a 3 hr LC-MS/MS analysis of TiO2-enriched Jurkat tryptic peptides as described in Methods section. SICs were randomly selected at a variety of abundance levels. Peptide sequences of five SICs are marked in the figure respectively. The retention time window (x axis range) of all the SICs are set at 2 min. 85 Figure 3.3 Pairwise replicate comparison of selected ion chromatography (SIC) peak areas from different analytical columns: (A) 50 cm-long/1.9 µm C18 columns (B) 50 cm-long/3 µm C18 columns (C) 15 cm-long/3 µm C18 columns. Each dot in the scatterplot represents the log10 (normalized XIC peak area) of a single phosphopeptide in two different replicates (ten possible pairs in total: Rep1:Rep2, Rep1:Rep3, Rep1:Rep4, Rep1:Rep5, Rep2:Rep3, Rep2:Rep4, Rep2:Rep5, Rep3:Rep4, Rep3:Rep5, Rep4:Rep5.). Dot density is indicated by color (from low to high: gray, blue, green, yellow, orange and red). The sample types (CD3/4 stimulated or unstimulated (control)) and correlation coefficient are marked in the figure respectively. The calculation of correlation coefficient was based on the combination of all the 10 pairwise replicate to replicate comparisons. The magnitude of the q-value was greatly reduced when comparing 1.9 µm column to the 3 µm columns, suggesting an increase in the reproducibility of replicate analyses (Figure 3.4). 86 A B C Figure 3.4 Histogram of the q-value obtained from three different analytical columns. Most importantly, at a strict cutoff of q-value < 0.01 for the replicate analyses, the 50 cm-long, 1.9 µm bead column was able to detect noticeably more significantly changed phosphopeptides (330% increased detection) and unique phosphorylation sites (270% increased detection) than columns using 3 µm particles (Table 3.5). Table 3.5 Quantitative comparison of TiO2 enriched unstimulated and CD3/4 stimulated Table 1. Quantitative comparison of TiO2 enriched unstimulated and CD3/4 stimulated Jurkat cells with Jurkat cells with different analytical columns different analytical columns Total # unique # Unique phosphopeptide-spectrum # Unique phosphosites c on Column length / matches with q < 0.01 and: peptides with q < 0.01 phosphopeptide- Particle Size spectrum matches a All > 2-fold b > 10-fold b pSer pThr pTyr 50 cm/ 1.9 µm 23131 16106 7900 660 9411 2014 416 50 cm / 3 µm 15173 4860 2117 221 3388 725 142 15 cm / 3 µm 14494 3749 1391 191 2743 554 105 a Includes all identified unique phospeptide-spectrum matches in CD3/4 stimulated and unstimulated Jurkat from five replicates. b Indicates number of significantly changed phospeptide-spectrum matches (q-value < 0.01) with a 2 or 10 fold increase or decrease when comparing average peak area between CD3/4 stimulated and unstimulated Jurkat. c One specific phosphosite in a protein is counted as one, even though it repeatedly appears in different PSMs with different charge states or miss cleavage. The 1.9 µm C18 column also greatly improved the identification of low abundance peptides with significant change (Figure 3.5). When examining the magnitude of the 87 calculated SIC peak areas, more than 30% of the identified significantly changed phosphopeptides by 50 cm-long 1.9 µm C18 column were below 1E7, while this percentage was less than 14% in 3 µm C18 column configurations (Table 3.6). Figure 3.5 Normalized peak area distribution of three different analytical columns. Peak area is the average value of the five replicate experiments. This distribution includes only the identified phosphopeptides with significant change (q < 0.01) when comparing the CD3/4 stimulated and unstimulated Jurkat cells. Table 3.6 Peak area distribution of different analytical columns Range of normalized peak area a Column type < 1E7 1E7 - 5E7 5E7 – 1E8 > 1E8 50 cm-long / 1.9 µm C18 column 30.2% 41.1% 12.3% 16.4% 50 cm-long / 3 µm C18 column 13.9% 36.8% 17.7% 31.6% 15 cm-long / 3 µm C18 column 13.2% 37.5% 19.7% 29.6% a Peak area is the average value of five replicates. This distribution includes all the identified phosphopeptides with significant change (q < 0.01) in CD3/4 stimulated and unstimulated Jurkat cells. 88 The dynamic range of the fold change calculated between the stimulated and unstimulated cells was compared for the 1.9 µm and 3 µm column configurations (Table 3.5 and Figure 3.6). The 1.9 µm C18 column reproducibly detected 373% more phosphopeptides with greater than 2 fold significant change and 299% more phosphopeptides with greater than 10 fold significant change compared to the 3 µm C18 column. The improved ability of the 1.9 µm C18 column to detect large fold changes in phosphopeptide abundance may be attributed to this column's improved ability to reproducibly quantitate low abundance peptides. Figure 3.6 Fold change distribution of three different analytical columns. Fold change is defined as the ratio of normalized peptide peak areas between CD3/4 stimulated and unstimulated Jurkat cells. This distribution includes all the identified phosphopeptides with significant change (q < 0.01). 89 3.4 DISCUSSION In this study, a fritless column fabrication strategy was employed to prepare a 50 cm-long 1.9 µm C18 reversed-phase column, which could be packed under 100 Bar and operated under 300 Bar at room temperature. Fritless columns were first employed in the capillary electrochromatography (CEC) field (27, 28). Without the existence of a frit, the particles can be well retained in the column because of the “keystone effect” (Figure 3.7) (29). In the context of 1.9 µm particles, a keystone effect was achieved in this study with pre-aggregation of the particles in the methanol packing slurry after sonication. The tapered 8 µm electrospray tip in the fritless column is totally compatible with sub-2 µm LC-MS/MS producing highly sensitive detection of peptides and reproducible quantitation. Unlike common C18 column preparation strategies employing low polarity solvents (e.g. chloroform), the incubation of hydrophobic C18 particles after sonication in a hydrophilic packing solution of methanol allowed for the pre-aggregation of particles in the slurry. These aggregated 1.9 µm particles could then more effectively be trapped in the 8 µm fritless tip while providing higher column flow rates at lower backpressures. Selection of an optimal slurry concentration was also important in the construction of the fritless column. Bruns et al. (30) found that increasing the slurry concentration could suppress particle mobility, attenuating the transcolumn bed heterogeneities and thus improving separation efficiency. Nevertheless, packing at a bead density higher than optimal leads to large packing voids, significantly increased the eddy dispersion. In this 90 study, we determined that a bead concentration of 30-40 mg/ml was sufficient to prevent small particles from freely passing through the tip while providing optimal column efficiency. Figure 3.7 Schematic diagram of “keystone effect”. The black arrows represent the direction of packing slurry. During the packing, as the particles become denser at the taper, these first particles (as “keystone”) block the following particles and allow the packing bed to grow longer. The investigation of optimal LC gradient for the 50 cm-long 1.9 µm C18 column revealed that the peptide and protein identifications did not improve linearly as gradient length increased. This finding was consistent with our finding that a longer than optimal gradient was shown to broaden the SIC peaks, leading to increasing peak overlap and a reduction in peak intensities. Our phosphoproteomic analysis determined the influence of particle size and column length on the reproducibility of peptide quantification. In addition to expanded proteome coverage, the column packed with 1.9 µm particles provided a 330% increase in the detection of significant quantitative changes in the proteome. This column also 91 demonstrated improved discrimination of significant quantitative changes in the proteome for low abundance peptides. These advantages highlighted the real world benefits of the improved column efficiency obtained with the smaller particles. By comparing the 50 cm column and 15 cm column with the same particle and LC gradient, we found that the longer column also improved the peptide identification and quantification in many respects, although to a much lesser extent than particle size. The enlarged peptide elution window and peak capacity of the longer column could explain these improvements. Combining the merits of small particles and long column, our 50 cm-long fritless column represents a highly efficient analytical tool in quantitative proteomics research. When evaluating the performance of new proteomic methods applied to biological systems, proteomic researchers should prioritize improvements in the number of significant changes detected from a biological stimulation over a sole focus on improvements in the PSM yield. 3.5 CONCLUSION In conclusion, the column fabrication methodology reported here represents an inexpensive and highly accessible approach to prepare a long capillary HPLC column packed with sub-2 µm particles. This self-packed column can provide excellent performance for both peptide identification and quantification with high dynamic range and sensitivity. Moreover, the routine operation of this column does not require less 92 frequently available UHPLC equipment, high-pressure column packing systems, or a column heating apparatus to pack or operate. Our hope is that this advance will provide broader access of highly sensitive and reproducible proteomics capabilities to a wider range of proteomic labs. 3.6 REFERENCES 1. Aebersold, R., and Mann, M. (2003) Mass spectrometry-based proteomics. Nature 422, 198-207 2. Choudhary, C., and Mann, M. (2010) Decoding signalling networks by mass spectrometry-based proteomics. Nat. Rev. Mol. Cell Biol. 11, 427-439 3. Pirmoradian, M., Budamgunta, H., Chingin, K., Zhang, B., Astorga-Wells, J., and Zubarev, R. A. (2013) Rapid and deep human proteome analysis by single-dimension shotgun proteomics. Mol. Cell. Proteomics 12, 3330-3338 4. Zubarev, R. A. (2013) The challenge of the proteome dynamic range and its implications for in-depth proteomics. Proteomics 13, 723-726 5. Gilar, M., Olivova, P., Daly, A. E., and Gebler, J. C. (2005) Orthogonality of separation in two-dimensional liquid chromatography. Anal. Chem. 77, 6426-6434 6. Gilar, M., Olivova, P., Daly, A. E., and Gebler, J. C. (2005) Two-dimensional separation of peptides using RP-RP-HPLC system with different pH in first and second separation dimensions. J Sep Sci 28, 1694-1703 7. Zhou, F., Sikorski, T. W., Ficarro, S. B., Webber, J. T., and Marto, J. A. (2011) Online nanoflow reversed phase-strong anion exchange-reversed phase liquid chromatography-tandem mass spectrometry platform for efficient and in-depth proteome sequence analysis of complex organisms. Anal. Chem. 83, 6996-7005 93 8. Gautier, V., Mouton-Barbosa, E., Bouyssie, D., Delcourt, N., Beau, M., Girard, J. P., Cayrol, C., Burlet-Schiltz, O., Monsarrat, B., and Gonzalez de Peredo, A. (2012) Label-free quantification and shotgun analysis of complex proteomes by one-dimensional SDS-PAGE/NanoLC-MS: evaluation for the large scale analysis of inflammatory human endothelial cells. Mol. Cell. Proteomics 11, 527-539 9. Nagaraj, N., Kulak, N. A., Cox, J., Neuhauser, N., Mayr, K., Hoerning, O., Vorm, O., and Mann, M. (2012) System-wide perturbation analysis with nearly complete coverage of the yeast proteome by single-shot ultra HPLC runs on a bench top Orbitrap. Mol. Cell. Proteomics 11, M111 013722 10. Cristobal, A., Hennrich, M. L., Giansanti, P., Goerdayal, S. S., Heck, A. J., and Mohammed, S. (2012) In-house construction of a UHPLC system enabling the identification of over 4000 protein groups in a single analysis. Analyst 137, 3541-3548 11. Thakur, S. S., Geiger, T., Chatterjee, B., Bandilla, P., Frohlich, F., Cox, J., and Mann, M. (2011) Deep and highly sensitive proteome coverage by LC-MS/MS without prefractionation. Mol. Cell. Proteomics 10, M110 003699 12. MacNair, J. E., Lewis, K. C., and Jorgenson, J. W. (1997) Ultrahigh-pressure reversed-phase liquid chromatography in packed capillary columns. Anal. Chem. 69, 983-989 13. MacNair, J. E., Patel, K. D., and Jorgenson, J. W. (1999) Ultrahigh-pressure reversed-phase capillary liquid chromatography: isocratic and gradient elution using columns packed with 1.0-micron particles. Anal. Chem. 71, 700-708 14. Yin, X., Liu, X., Zhang, Y., Yan, G., Wang, F., Lu, H., Shen, H., and Yang, P. (2014) Rapid and sensitive profiling and quantification of the human cell line proteome by LC-MS/MS without prefractionation. Proteomics 14, 2008-2016 15. Ong, S. E., Blagoev, B., Kratchmarova, I., Kristensen, D. B., Steen, H., Pandey, A., and Mann, M. (2002) Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell. Proteomics 1, 376-386 16. Ross, P. L., Huang, Y. N., Marchese, J. N., Williamson, B., Parker, K., Hattan, S., Khainovski, N., Pillai, S., Dey, S., Daniels, S., Purkayastha, S., Juhasz, P., Martin, S., Bartlet-Jones, M., He, F., Jacobson, A., and Pappin, D. J. (2004) Multiplexed protein 94 quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol. Cell. Proteomics 3, 1154-1169 17. Nguyen, V., Cao, L., Lin, J. T., Hung, N., Ritz, A., Yu, K., Jianu, R., Ulin, S. P., Raphael, B. J., Laidlaw, D. H., Brossay, L., and Salomon, A. R. (2009) A new approach for quantitative phosphoproteomic dissection of signaling pathways applied to T cell receptor activation. Molecular & cellular proteomics : MCP 8, 2418-2431 18. Protocol for fritless column preparation kindly provided by personal correspondence from Tobias Walther and Romain Christiano. Protocol for fritless column preparation kindly provided by personal correspondence from Tobias Walther and Romain Christiano 19. Yu, K., and Salomon, A. R. (2009) PeptideDepot: flexible relational database for visual analysis of quantitative proteomic data and integration of existing protein information. Proteomics 9, 5350-5358 20. Yu, K., and Salomon, A. R. (2010) HTAPP: high-throughput autonomous proteomic pipeline. Proteomics 10, 2113-2122 21. Elias, J. E., and Gygi, S. P. (2007) Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature methods 4, 207-214 22. Yu, K., Sabelli, A., DeKeukelaere, L., Park, R., Sindi, S., Gatsonis, C. A., and Salomon, A. (2009) Integrated platform for manual and high-throughput statistical validation of tandem mass spectra. Proteomics 9, 3115-3125 23. Beausoleil, S. A., Villen, J., Gerber, S. A., Rush, J., and Gygi, S. P. (2006) A probability-based approach for high-throughput protein phosphorylation analysis and site localization. Nature biotechnology 24, 1285-1292 24. Demirkan, G., Yu, K., Boylan, J. M., Salomon, A. R., and Gruppuso, P. A. (2011) Phosphoproteomic profiling of in vivo signaling in liver by the mammalian target of rapamycin complex 1 (mTORC1). PloS one 6, e21729 25. Storey, J. D. (2003) The positive false discovery rate: A Bayesian interpretation and the q-value. Annals of Statistics 31, 2013-2035 26. Storey, J. D., and Tibshirani, R. (2003) Statistical significance for genomewide studies. Proc. Natl. Acad. Sci. U. S. A. 100, 9440-9445 95 27. Lord, G. A., Gordon, D. B., Myers, P., and King, B. W. (1997) Tapers and restrictors for capillary electrochromatography and capillary electrochromatography mass spectrometry. J. Chromatogr. A 768, 9-16 28. Mayer, M., Rapp, E., Marck, C., and Bruin, G. J. M. (1999) Fritless capillary electrochromatography. Electrophoresis 20, 43-49 29. Gaspar, A., Piyasena, M. E., and Gomez, F. A. (2007) Fabrication of fritless chromatographic microchips packed with conventional reversed-phase silica particles. Anal. Chem. 79, 7906-7909 30. Bruns, S., Franklin, E. G., Grinias, J. P., Godinho, J. M., Jorgenson, J. W., and Tallarek, U. (2013) Slurry concentration effects on the bed morphology and separation efficiency of capillaries packed with sub-2 mum particles. J. Chromatogr. A 1318, 189-197 96 Chapter 4 AN OPTIMIZED PROTOCOL FOR THE USE OF DMSO CO- SOLVENT WITH Q EXACTIVE MASS SPECTROMETER TO IMPROVE PEPTIDE SEQUENCING DEPTH SIGNIFICANTLY 97 4.1 INTRODUCTION HPLC-ESI-MS/MS has become a powerful platform in mass spectrometry-based proteomics for protein identification from complex biological samples. Among the various efforts to optimize its performance, Kuster’s group in 2013 (1) reported an attractive improvement by adding 5% DMSO to the LC mobile phase. They found that DMSO could enhance the electrospray response and increase the sensitivity of peptide detection. The concomitant reduction of injection time indicated that the higher signal intensity was due to better electrospray efficiency. Based on classical electrospray theory, they proposed that DMSO decreases the surface tension and elevates the Rayleigh limit. As a result, the charged droplets become destabilized faster and droplet fission accelerates (Figure 4.1). Certainly, the number of identified peptides, especially low-abundance peptides, benefits much from the improved ionization efficiency. Surface Density r Boiling tension g [g/ml] point [ C] [mN/m] 28 0.78 80 73 1.00 100 44 1.10 180 43 0.97 n.a. 47 0.98 n.a. Eq. (1) Initial droplets Rayleigh limit Ionization by the ion evaporation (IEM) ~ 5% smaller ~10-20% larger and/or charge residue model (CRM) DMSO ACN + water Eq. (2) Figure 4.1 Proposed mechanism by which Supplementary DMSO Fig. 14 | Working modelincreases electrospray explaining the increased response electrospray response (1).under of peptides DMSO conditions. (a) Schematic representation of the electrospray process. (b) Surface tension, density and boiling point data. (c) Nanoelectrospray equation (1) and Rayleigh limit (2). In the presence of 5% DMSO, initial droplets are ~5% smaller and at the Rayleigh limit 10-20% larger than under no DMSO conditions. As the solvent evaporates (ACN first as azeotrope with water followed by water alone), the DMSO content of droplets 98 rapidly increases and drastically increases the Rayleigh limit. Thus, the presence of even low initial DMSO concentrations will lead to instable droplets more quickly which in turn leads to faster sequestering of peptides into charged droplets (ideally one peptide molecule per droplet) and increased ionization efficiency. This DMSO experiment was performed on an Orbitrap Elite mass spectrometer which is a hybrid linear ion trap/orbitrap instrument, while the Q Exactive mass spectrometer widely employed by current proteomics researchers dispenses with the ion trap while maintaining the orbitrap. In the Q Exactive instrument (2, 3), ions from an atmospheric pressure ion source (API) are transferred efficiently via the stacked-ring ion guides (S-lens). Then, the ions pass through a bent flatpole, which has 2-mm gaps between its rods to eject solvent droplets and other neutral species. After that, a quadrupole mass filter with fast switching times can nearly instantaneously select ions within the desired m/z range. The selected ions are further fragmented and analyzed at a high-energy collision-induced dissociation (HCD) cell and an Orbitrap analyzer respectively. The major difference between the Q Exactive and the Orbitrap Elite is that the Q Exactive utilizes the quadrupole mass filter, instead of the linear ion trap (4), to perform precursor ion selection. Although the data dependent cycle speed is significantly improved in this configuration, the quadrupole is a vulnerable part for contamination. Studies have found that the performance and accuracy of the Q Exactive can be compromised after a period of use (3). These researchers proposed that due to the abundant peptide ion beam entering the instrument, peptides excluded from selection could gradually coat and impair the performance of the rods of the quadrupole. The Q Exactive needs to be cleaned and baked out more frequently than the hybrid Orbitrap mass spectrometers to maintain optimal performance. 99 One important component of the Q Exactive to control the amount of analytes and solvent entering the mass spectrometer is a heated capillary inlet, as an electrospray ionization mass spectrometry interface. After the analytes in the LC solvent are ionized by electrospray to form the aerosol plume (electrosprayed droplets), they enter the mass spectrometer from the heated capillary inlet, where the desolvation occurs along with the ion transmission to release the gas-phase analyte ions. High capillary temperature elevates the ionization efficiency by improving the desolvation, but compromises the ion transmission, which could result from the increased diffusion to the capillary wall (5). Therefore, the MS intensity reflects the interplay between ion transmission and desolvation efficiency. In this study, we investigated the effect of DMSO cosolvent on Q Exactive peptide identification yields and identified a rapid mass spectrometer contamination issue caused by high concentrations of DMSO. This protocol optimization study revealed that elevation of capillary temperature and reduction of DMSO concentration could minimize contamination while increasing confident peptide identification. 4.2 MATERIALS AND METHODS 4.2.1 Cell Culture and Lysis Jurkat clone E6-1 was obtained from American Tissue Culture Collection (Manassas, VA). The cells were maintained in RPMI 1640 medium (Gibco, Grand Island, NY) 100 supplemented with 10% heat-inactivated undialyzed FBS (HyClone, Logan, UT), 2 mM L-glutamine, 100 U/ml penicillin G, and 100 µg/ml streptomycin (Gibco, Grand Island, NY) in a humidified incubator with 5% CO2 at 37 °C. Cells were grown for eight doublings before harvest, and then placed in lysis buffer (8 M urea, 1 mM sodium orthovanadate, 20 mM HEPES, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, pH 8.0) for 20 min at 4 °C. The lysate was sonicated at a 30 watt output with 2 bursts of 30 seconds each and cleared at 14,000 g for 15 min at 4 °C. SILAC labeled MCF-7 whole cell lysate (labeled with 13C6 L-lysine and 13C6 L-arginine, heavy and light 1:1 mixing) was generously provided by Northwestern University. 4.2.2 Protein Reduction, Alkylation, Digestion, and Desalting Protein concentration was measured using the DC Protein Assay (Bio-Rad, Hercules, CA). Next, the clear cell lysate was reduced with 45 mM DTT for 20 min at 60°C and alkylated with 100 mM iodoacetamide for 15 min at room temperature in the dark. Cell lysate was then diluted 4-fold with 20 mM HEPES buffer, pH 8.0 and digested with sequencing grade modified trypsin (Promega, Madison, WI) in a 1:100 (w/w) trypsin : protein ratio overnight at room temperature. Typtic peptides were acidified to pH 2.0 with 20% trifluoroacetic acid (TFA), cleared at 1800 g for 5 min at room temperature, and desalted using C18 Sep-Pak plus cartridges (Waters, Milford, MA) as previously described (6), with the exception that TFA was used instead of acetic acid. Eluents 101 containing peptides were lyophilized for 48 hours to dryness. 4.2.3 TiO2 Phosphopeptide Enrichment Phosphopeptides were enriched with Titansphere Phos-TiO tips (GL Sciences, Tokyo Japan) following the manufacturer’s protocol with some modifications. The condition buffers (containing TFA, CH3CN and lactic acid) and elution buffers (1% NH4OH in water and 40% CH3CN) were prepared first. Then, the condition buffer was added to the Phos-TiO tips (centrifuge at 3000 g, 22 ˚C). Once the conditioning was finished, desalted tryptic peptides from Jurkat total lysates were mixed with a synthetic phosphoserine standard (FQpSEEQQQTEDELQDK, AnaSpec, San Jose, CA) at a ratio of 5 fmol standard : 1 µg sample. The mixture was loaded onto tips using centrifugation at 1000 g at 22 ˚C. After loading, the column was washed with condition buffers followed by elution buffers. Acetic acid was used to acidify TiO2 enriched samples, which were dried almost to completeness. 4.2.4 Automated Nano-LC/MS Tryptic peptides were analyzed by a fully automated phosphoproteomic technology platform (7, 8). The nanoLC-MS/MS experiments were performed with an Agilent 1200 Series Quaternary HPLC system (Agilent Technologies, Santa Clara, CA) connected to a Q Exactive mass spectrometer (Thermo Fisher Scientific, Waltham, MA). 102 Phosphopeptides were eluted into the mass spectrometer through a PicoFrit analytical column (360 µm outer diameter 75 µm inner diameter-fused silica packed on a pressure bomb with 15 cm of 3 µm Monitor C18 particles; New Objective, Woburn, MA) with a reversed-phase gradient (0-70% 0.1M acetic acid in acetonitrile in 60 minutes, with a 90 min total method duration). For the LC-MS/MS tests with DMSO cosolvent, 0.5%， 1%， or 5% HPLC-grade DMSO was added to both solvent A (0.1 M acetic acid) and solvent B (0.1 M acetic acid in CH3CN). The electrospray ion source was operated at 2.0 kv in a split flow configuration. The Q Exactive was operated in the data dependent mode using a top-9 data dependent method. Survey full scan MS spectra (m/z 400-1800) were acquired at a resolution of 70,000 with an AGC target value of 3×106 ions or a maximum ion injection time of 200 ms. Peptide fragmentation was performed via higher-energy collision dissociation (HCD) with the energy set at 28 NCE. The MS/MS spectra were acquired at a resolution of 17,500, with a targeted value of 2×104 ions or a maximum integration time of 200 ms. The underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 1.0%. The ion selection abundance threshold was set at 1.0×103 with charge state exclusion of unassigned and z =1, or 6-8 ions and dynamic exclusion time of 30 sec. 4.2.5 Data Analysis Peptide spectrum matching of MS/MS spectra was performed against a 103 human-specific database (UniProt; downloaded 2/1/2013) using MASCOT v. 2.4 (Matrix Science, Ltd, London W1U 7GB UK). A concatenated database containing 144,156 “target” and “decoy reversed” sequences was employed to estimate the false discovery rate (FDR) (9). Msconvert from ProteoWizard (v. 3.0.5047), using default parameters and with the MS2Deisotope filter on, was employed to create peak lists for Mascot. Mascot database searches were performed with the following parameters: trypsin enzyme cleavage specificity, 2 possible missed cleavages, 10 ppm mass tolerance for precursor ions, 20 mmu mass tolerance for fragment ions. Search parameters specified a dynamic modification of phosphorylation (+79.9663 Da) on serine, threonine, and tyrosine residues, a dynamic modification of methionine oxidation (+15.9949 Da), and static modification of carbamidomethylation (+57.0215 Da) on cysteine. Mascot results were filtered by Mowse Score (>20). Peptide assignments from the database search were filtered down to 1% false discovery rate (FDR) by a logistic spectral score, as previously described (9, 10). 4.3 RESULTS AND DISCUSSION 4.3.1 Effect of 5% DMSO Cosolvent on Peptide Identification with Q Exactive and DMSO-induced Q Exactive Mass Spectrometer Contamination When adding 5% DMSO to the HPLC mobile phase, we found that peptide spectrum matches with the Q Exactive were increased by 20%-25% in all the five 104 biological replicates of SILAC labeled MCF-7 whole cell lysate sample. The percentage of unique PSMs was also increased, which indicated improved identification of low-abundance peptides (Figure 4.2). These improved yields were consistent in magnitude with the study of Kuster’s group using the Orbitrap Elite mass spectrometer and validated that proteome coverage of Q Exactive could also benefit from the improvement of electrospray efficiency by DMSO cosolvent. Figure 4.2 Effect of 5% DMSO cosolvent on peptide identification with Q Exactive. The results are the average value of five biological replicates analysis of SILAC labeled MCF-7 whole cell lysate sample. Standard deviations are marked as error bars in the chart. However, after applying 5% DMSO on LC-MS/MS for two days, we detected that the m/z deviation of peptides gradually shifted from less than 2 ppm to beyond 100 ppm. In 0.2 pmol BSA tryptic digest test, three characteristic peptides S*LHTLFGDELCK, HLVDEPQNLIK, LVNELTEFAK deviated in m/z by 121.7 ppm, 120.6 ppm, and 121.1 105 ppm respectively. This huge mass inaccuracy greatly impacted the performance of the Q Exactive and lowered the peptide identification yield from more than 3000 to less than 100. This issue was an accumulative and irreversible hardware defect, which could only be rehabilitated by completely cleaning the quadrupole and S-lens with methanol and baking out the Q Exactive for more than 16 hours. Simply performing a calibration or ceasing DMSO addition could not solve it. After 24-hour LC-MS/MS with a freshly cleaned and calibrated Q Exactive, addition of 0.5%, 2%, 3%, 5% DMSO caused the m/z deviation by 0.98 ppm, 2.22 ppm, 55.18 ppm, and 110.89 ppm respectively (calculated from the average m/z deviation of three characteristic BSA peptides S*LHTLFGDELCK, HLVDEPQNLIK, LVNELTEFAK). Thus, we concluded that the mass inaccuracy issue of Q Exactive results from the contamination of DMSO entering the mass spectrometer. In the regular progress of Q Exactive contamination, peptides excluded from mass selection could gradually coat the rods of the quadrupole. DMSO may aggravate this coating by either electrostatic interaction or directly depositing on the quadrupole because of its high viscosity and elevated boiling point compared to water. The contamination of the quadrupole could impact its function for m/z selection that could lead to the mass inaccuracy. 4.3.2 Minimization of DMSO Contamination with an Optimized Protocol To minimize this contamination issue, decreasing the amount of DMSO is the most 106 straightforward solution. The heated capillary inlet is an important component of the Q Exactive to control the amount of analytes and solvent entering the mass spectrometer (Figure 4.3). Since higher capillary temperature improves the desolvation of electrospray Figure 4.3 Schematic diagram of electrospray process and the further desolvation occurring in the heated capillary inlet of Q Exactive. droplets, we assumed that it thereby reduces the DMSO entering Q Exactive. Thus, we improved the capillary temperature from original 320 °C to 380 °C. Meanwhile, we reduced the concentration of DMSO cosolvent to a lower level than the original 5%. After one week of LC-MS/MS analyses using 0.5% or 1% DMSO, we did not observe any mass deviation beyond 2 ppm mass tolerance. If using 2% DMSO at 380 °C for several days, mass inaccuracy was found occasionally (or even rarely) with the deviation less than 3 ppm. However, 5% DMSO at 380 °C could still cause significant mass inaccuracy (~ 100 ppm) after several days. Then, we evaluated the influence of capillary temperature and DMSO concentration 107 on peptide identification yield. We found that 0.5% and 1% DMSO increased the peptide identification of SILAC labeled MCF-7 sample by 8% and 13% respectively, which could be a valuable improvement in proteomics study, although less significant than using 5% DMSO. Elevated capillary temperature, which in theory may lower ion transmission, did not cause observable change of proteome coverage. Then, we applied another Q Exactive instrument (same model) and sample type to validate this observation. The results indicated that 1% DMSO increased peptide identification of label-free TiO2-enriched Jurkat sample by ~15% (phosphopeptide identification by ~10%). The differences between 320 °C and 380 °C of capillary temperature were also negligible. A B Figure 4.4 Peptide identification results under different DMSO concentrations and capillary temperatures of Q Exactive. The loading samples are: (A) SILAC labeled MCF-7 whole cell lysate; (B) label-free TiO2-enriched Jurkat whole cell lysate. 4.4 CONCLUSION Compared to the hybrid Orbitrap instrument, the Q Exactive is more vulnerable to contamination because of its hardware configuration. The addition of 5% DMSO to the 108 LC mobile phase causes a previously unappreciated severe contamination of the Q Exactive mass spectrometer. Certainly, sufficient studies have indicated that DMSO can significantly enhance the electrospray response, probably by decreasing the surface tension and elevating the Rayleigh limit. Thus, if researchers are still interested in the use of DMSO cosolvent, some optimizations need to be performed. Our research found that elevating capillary temperature while reducing the DMSO concentration could minimize the Q Exactive contamination to the lowest level (not observable within several weeks). This optimization offers proteomics researchers a new option, which improves confident peptide identification using the Q Exactive by at least 10% without the unacceptable mass deviation issue. 4.5 REFERENCES 1. Hahne, H., Pachl, F., Ruprecht, B., Maier, S. K., Klaeger, S., Helm, D., Medard, G., Wilm, M., Lemeer, S., and Kuster, B. (2013) DMSO enhances electrospray response, boosting sensitivity of proteomic experiments. Nature methods 10, 989-991 2. Michalski, A., Damoc, E., Hauschild, J. P., Lange, O., Wieghaus, A., Makarov, A., Nagaraj, N., Cox, J., Mann, M., and Horning, S. (2011) Mass spectrometry-based proteomics using Q Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer. Molecular & cellular proteomics : MCP 10, M111 011015 3. Scheltema, R. A., Hauschild, J. P., Lange, O., Hornburg, D., Denisov, E., Damoc, E., Kuehn, A., Makarov, A., and Mann, M. (2014) The Q Exactive HF, a Benchtop mass spectrometer with a pre-filter, high-performance quadrupole and an ultra-high-field Orbitrap analyzer. Molecular & cellular proteomics : MCP 13, 3698-3708 109 4. Olsen, J. V., Schwartz, J. C., Griep-Raming, J., Nielsen, M. L., Damoc, E., Denisov, E., Lange, O., Remes, P., Taylor, D., Splendore, M., Wouters, E. R., Senko, M., Makarov, A., Mann, M., and Horning, S. (2009) A dual pressure linear ion trap Orbitrap instrument with very high sequencing speed. Molecular & cellular proteomics : MCP 8, 2759-2769 5. Page, J. S., Marginean, I., Baker, E. S., Kelly, R. T., Tang, K., and Smith, R. D. (2009) Biases in ion transmission through an electrospray ionization-mass spectrometry capillary inlet. Journal of the American Society for Mass Spectrometry 20, 2265-2272 6. Nguyen, V., Cao, L., Lin, J. T., Hung, N., Ritz, A., Yu, K., Jianu, R., Ulin, S. P., Raphael, B. J., Laidlaw, D. H., Brossay, L., and Salomon, A. R. (2009) A new approach for quantitative phosphoproteomic dissection of signaling pathways applied to T cell receptor activation. Molecular & cellular proteomics : MCP 8, 2418-2431 7. Yu, K., and Salomon, A. R. (2009) PeptideDepot: flexible relational database for visual analysis of quantitative proteomic data and integration of existing protein information. Proteomics 9, 5350-5358 8. Yu, K., and Salomon, A. R. (2010) HTAPP: high-throughput autonomous proteomic pipeline. Proteomics 10, 2113-2122 9. Elias, J. E., and Gygi, S. P. (2007) Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature methods 4, 207-214 10. Yu, K., Sabelli, A., DeKeukelaere, L., Park, R., Sindi, S., Gatsonis, C. A., and Salomon, A. (2009) Integrated platform for manual and high-throughput statistical validation of tandem mass spectra. Proteomics 9, 3115-3125 110 Chapter 5 CONCLUSION 111 5.1 SUMMARY OF RESULTS TCR ligation-independent signal transduction in the basal state of T cells is critical for pre-TCR beta-selection, a key step of T cells survival and differentiation. Previous studies have identified that the basal signaling pathway shares some components with conventional TCR signaling and that the kinase-phosphatase pair Csk and CD45 are important regulators of the basal signaling by modulating the SFKs kinase activity through inhibitory tyrosine Lck Y505. However, compared to the well-documented TCR- inducible signaling pathway, global understanding of the ligand-independent basal signaling pathway is still insufficient. Moreover, whether other tyrosine phosphorylation sites in addition to Lck Y505 are involved in the regulation of Csk and CD45 remains to be elucidated. Therefore, we deployed a mass spectrometry-based quantitative phosphoproteomic approach to identify the signaling through Csk and CD45 during T cell basal signaling by comparing the wide-scale phosphorylation patterns in Csk and CD45 singly or doubly deficient cell lines with wild type parental Jurkat T cells. In total 1935 unique tyrosine phosphorylation sites on 1080 proteins were confidently identified showing statistically significant changes (q-value < 0.01) between wild type Jurkat cells and Csk or/and CD45 deficient cell line. Some of these sites displayed a direct relationship with Lck kinase activity. By analyzing the phosphorylation patterns in different cell lines, we verified that the signaling cascades and cytoskeletal dynamics in the basal state of T cells share components with conventional TCR signaling. A series of 112 proteins and tyrosine sites, which are modulated by Lck kinase activity and participate the signaling transduction in the basal state of T cells, were identified. Our results not only validated the simple regulatory role of Csk via Lck kinase activity, but also proposed a more complicated pattern of CD45 substrate specificity. We proposed a new model of negative regulation of Fyn SH2 domain phosphorylation by CD45 and identified Fyn Y185, Y213, and Y214 as potential sites of CD45 regulation. Furthermore, we detected many hyperphosphorylated tyrosine sites in Csk and CD45 double-mutant cells, which could be explained by the synergistic regulation of activated Lck kinase loop (from Csk-deficiency) and Fyn SH2 domain (from CD45-deficiency). Proteins with this pattern of regulation were found to have over-representation in proteins with functions in integrin-mediated signaling and regulation of cytoskeletal dynamics. The pivotal role of integrin β1 (Y783, Y795) and PECAM-1 (Y690, Y713) was thus highlighted. Since the hyperphosphorylation observed in J.Csk/CD45 cells was highly unique resulting from the knockdown of two important regulatory proteins, more than 100 identified tyrosine sites showing this regulation pattern may share common pathway regulation. To summarize, our results provided a more systematic view of basal signaling cascades occurring in T cells and a better understanding of the regulatory mechanisms of Csk and CD45. Because of the high overlap of signal components between basal and TCR-induced signaling pathway, integrin-mediated signaling and regulation of cytoskeletal dynamics that were found in the basal state of T cells may also function upon TCR activation. Further 113 biological studies based on the results in this project may shed light on the migration of activated T lymphocytes. Considering the strength of mass spectrometry on quantitatively revealing the global signaling transduction and protein interactions, we are also interested in developing the methodology to increase the proteome coverage while keeping the high quantification reproducibility. In Chapter 3, we focused on the application of sub-2 µm particles in LC-MS/MS. Compared to traditional 3 µm or 5 µm reversed-phase C18 particles, sub-2 µm C18 particles improve HPLC column efficiency by decreasing the diffusion distances of solutes and minimizing the band spreading. A major technical difficulty in using sub-2 µm particles is the high column backpressure, which is typically overcome by ultra-high pressure liquid chromatography (UHPLC) or a column heater. With the fritless packing strategy, we successfully packed a 50 cm-long capillary HPLC column with 1.9 µm particles under 100 Bar. The routine operation of this column does not require less frequently available UHPLC equipment, high-pressure column packing systems, or a column heating apparatus. Compared with traditional C18 columns packed with 3 µm particles, the column with the 1.9 µm particles could detect 330% more peptides with statistically significant changes from differentially stimulated T cells. Not only focusing on peptide identifications yield, we also evaluated the performance of this new column based on quantitative reproducibility, which is essential to maximize extraction of biologically relevant insights from proteomic experiments. Our study proposed an 114 alternative mode of evaluation of proteomic methodology that focuses on the yield of statistically significant changes from a biologically relevant stimulation. In Chapter 4, we tested the feasibility of using DMSO cosolvent with Q Exactive mass spectrometer to improve proteomics sequencing depth. 5% DMSO cosolvent was found to enhance the electrospray response and increase the sensitivity of peptide detection on an Orbitrap Elite mass spectrometer. Our study identified the contamination issue of 5% DMSO cosolvent with Q Exactive mass spectrometer. By optimizing the capillary temperature and DMSO concentration, we proposed a protocol that minimized the contamination of DMSO while increasing peptide identification by at least 10%. In both Chapter 3 and Chapter 4, our hope is to make technical improvements, which provide broader access of highly sensitive and reproducible proteomics capabilities to a wider range of proteomic labs not equipped with the most advanced instruments. 5.2 FUTURE WORK In Chapter 2, with the mass spectrometry-based quantitative phosphoproteomics technologies, we identified numerous tyrosine phosphorylation sites that may participate in the Lck-mediated T cell basal signaling and the integrin-mediated cell adhesion. The following biochemical experiments are required to confirm our proteomics results and the proposed role of these tyrosine sites. Immunoblotting, coimmunoprecipitation, and site-specific mutagenesis will respectively help validate the conclusion in this project 115 about the protein abundance, protein interaction and signal transduction mechanism. From our data, two major questions remain unanswered: 1) whether all the phosphorylation changes in J.Csk cells result from the improved Lck kinase activity? (i.e. Whether Lck Y505 is the only substrate of Csk?) 2) Is the negative regulatory role of CD45 on the hyperphosphorylated tyrosine sites in this project fulfilled via the phosphorylation of Fyn SH2 domain? To clarify these questions, an Lck Y505 mutant cell line and a Fyn Y185, Y213, Y214 triple mutant cell line can be developed and compared with our current results. In Chapter 3, our in-house fabricated 50 cm-long fritless column packed with 1.9 µm C18 particles have shown substantial advantages in providing high peptide identification yield and quantitative reproducibility. However, under standard HPLC pressure (< 300 Bar), the flow rate of the column was compromised at 60 nL/min. Thus, we are interested to test the performance of this column on a UHPLC instrument, which can increase the column flow rate to the optimal level (at least several hundreds nL/min). Under the optimal flow rate, we can further evaluate the advantages and the highest potential of our in-house fabricated fritless column by comparing it with a similar configured fritted column. 116