Neuroproteomics (Neuromethods 57)

Neuromethods Series Editor Wolfgang Walz University of Saskatchewan Saskatoon, SK, Canada For further volumes: htt...

Author: Ka Wan Li

34 downloads 1023 Views 8MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Neuromethods

Series Editor Wolfgang Walz University of Saskatchewan Saskatoon, SK, Canada

For further volumes: http://www.springer.com/series/7657

wwwwwwwwwwwwwwwwww

Neuroproteomics Edited by

Ka Wan Li Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands

Editor Ka Wan Li Department of Molecular and Cellular Neurobiology Center for Neurogenomics and Cognitive Research VU University Amsterdam The Netherlands [email protected]

ISSN 0893-2336 e-ISSN 1940-6045 ISBN 978-1-61779-110-9 e-ISBN 978-1-61779-111-6 DOI 10.1007/978-1-61779-111-6 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011928074 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface to the Series Under the guidance of its founders Alan Boulton and Glen Baker, the Neuromethods series by Humana Press has been very successful since the first volume appeared in 1985. In about 17 years, 37 volumes have been published. In 2006, Springer Science+Business Media made a renewed commitment to this series. The new program will focus on methods that are either unique to the nervous system and excitable cells or which need special consideration to be applied to the neurosciences. The program will strike a balance between recent and exciting developments like those concerning new animal models of disease, imaging, in vivo methods, and more established techniques. These include immunocytochemistry and electrophysiological technologies. New trainees in neurosciences still need a sound footing in these older methods in order to apply a critical approach to their results. The careful application of methods is probably the most important step in the process of scientific inquiry. In the past, new methodologies led the way in developing new disciplines in the biological and medical sciences. For example, Physiology emerged out of Anatomy in the nineteenth century by harnessing new methods based on the newly discovered phenomenon of electricity. Nowadays, the relationships between disciplines and methods are more complex. Methods are now widely shared between disciplines and research areas. New developments in electronic publishing also make it possible for scientists to download chapters or protocols selectively within a very short time of encountering them. This new approach has been taken into account in the design of individual volumes and chapters in this series. Wolfgang Walz

v

wwwwwwwwwwwwwwwwww

Preface The field of Proteomics has various interesting emerging technologies that allow the quantitative analysis of hundreds to thousands of proteins in a biological system of interest. In the past few years, there was a steep increase in the application of proteomics to examine the molecular mechanisms underlying (mal-)functioning of the nervous system and brain disorders, which in many cases has yielded novel insights. Since neuroproteomics is a new research field involving the use of a number of high-end analytical instruments and technologies, both promises and pitfalls may not be well appreciated by the researchers. Equally, the way to design a proper proteomics experiment may not be obvious for an average neuroscientist. Presented in this single volume, Neuroproteomics has 21 chapters covering aspects of various dimensions of this new technology, together with some established methods that have supporting roles in the workflow of neuroproteomics. The contributors to book chapters in this volume are active researchers with ample experience in the techniques that they describe. Each chapter provides a step-by-step set of instructions on how to perform the experiments and advice in the NOTES that may help to optimize the experimental conditions. We have covered most of the recent proteomics methods and some of their applications. The chapters introduce readers to the various ways of designing successful neuroproteomics experiments, and the implication of the different experimental designs. The details of the technologies, which can be very analytical in nature, are not the focus of this book. Taken together, I am convinced that these chapters will be of great assistance to the readers wishing to design and execute their own proteomics experiments in an optimal way. Amsterdam, The Netherlands

Ka Wan Li

vii

wwwwwwwwwwwwwwwwww

Contents Preface to the Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v vii xi

Part I Introduction 1 Neuroproteomics: Deciphering Brain Function and Disorders . . . . . . . . . . . . . . . Ka Wan Li

3

Part II Fractionation of Brain Regions, Organelles, and Protein Complexes 2 Dissection of Rodent Brain Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sabine Spijker 3 Subcellular Fractionation of Brain Tissue Using Free-Flow Electrophoresis . . . . . Markus Islinger, Joachim Kirsch, Sabine Angermüller, Ramona Rotaru, Afsaneh Abdolzade-Bavil, and Gerhard Weber 4 Isolation of Synapse Subdomains by Subcellular Fractionation Using Sucrose Density Gradient Centrifugation . . . . . . . . . . . . . . . . . . . . . . . . . . Tatsuo Suzuki 5 Enrichment of Plasma Membranes from Small Brain Samples by Aqueous Polymer Two-Phase Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jens Schindler 6 Identification and Characterization of Protein Complexes from Total Brain and Synaptoneurosomes: Heterogeneity of Molecular Complexes in Distinct Subcellular Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Silvia De Rubeis and Claudia Bagni 7 Analysis of Protein Complexes by 2D Blue Native/SDS–PAGE and Antibody-Shift Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dong Yang, Xinyu Deng, Ying Jiang, and Fuchu He

13 27

47

63

69

81

Part III Analytical Tools for Neuroproteomics 8 Two-Dimensional Gel Electrophoresis-Based Proteomic Analysis of Brain Synapses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Karl-Heinz Smalla and Ursula Wyneken 9 Two-Dimensional BAC-SDS Polyacrylamide Gel Electrophoresis for the Fractionation and Identification of Synaptic Vesicle Proteins and the Presynaptic Active Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Joern Barth and Walter Volknandt

ix

x

Contents

10 iTRAQ-Based LC-LC MALDI TOF/TOF MS Quantitative Analysis of Membrane Proteins from Human Glioma . . . . . . . . . . . . . . . . . . . . . . Uroš Raj čević 11 OFFGEL-Isoelectric Focusing Fractionation for the Analysis of Complex Proteome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Emilie Ernoult and Catherine Guette 12 A 1D-PAGE/LC-ESI Linear Ion Trap Orbitrap MS Approach for the Analysis of Synapse Proteomes and Synaptic Protein Complexes . . . . . . . . Ning Chen, Roel C. vd Schors, and August B. Smit 13 SDS-PAGE Immunoblot Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patricia Klemmer 14 Phosphoproteomics by Highly Selective IMAC Protocol . . . . . . . . . . . . . . . . . . . Chia-Feng Tsai, Yi-Ting Wang, Pei-Yi Lin, and Yu-Ju Chen 15 Global Analysis of Ubiquitination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Meierhofer and Peter Kaiser 16 High-Throughput High-Content Functional Image Analysis of Neuronal Proteins Implicated in Parkinson’s Disease . . . . . . . . . . . . . . . . . . . . Eva Blaas and Ronald E. van Kesteren

127

145

159 169 181 197

211

Part IV Analysis of Specific Brain Tissues and Fluid 17 Neuropeptidomics of the Mammalian Brain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Fang Xie, Elena V. Romanova, and Jonathan V. Sweedler 18 In-Depth Analysis of the Cerebrospinal Fluid Proteome and Biomarker Discovery: Abundant Protein Depletion Sample Pretreatment Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Silvina A. Fratantoni and Connie R. Jimenez 19 Lipidomics of the Nervous System: Phospholipidomics by Liquid Chromatography Coupled to Mass Spectrometry or Tandem Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 Su Chen

Part V Bioinformatics 20 Bioinformatics Procedures for Analysis of Quantitative Proteomics Experiments Using iTRAQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 Pim van Nierop and Maarten Loos 21 Statistical Analysis of Spectral Count Data Generated by Label-Free Tandem Mass Spectrometry-Based Proteomics . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 Thang V. Pham and Connie R. Jimenez Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311

Contributors Afsaneh Abdolzade-Bavil • Beckton & Dickinson GmbH, Heidelberg, Germany Sabine Angermüller • Department of Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany Claudia Bagni • Center for Human Genetics, Katholieke Universiteit Leuven, Leuven, Belgium; Department of Molecular and Developmental Genetics, Leuven, Belgium; Department of Experimental Medicine and Biochemical Sciences,University of Rome “Tor Vergata”, Rome, Italy Joern Barth • AK Neurochemistry, Biocenter of JW Goethe-University, Frankfurt/Main, Germany Eva Blaas • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Ning Chen • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Su Chen • Chainon Neurotrophin Biotechnology Inc., Malta, NY, USA Yu-Ju Chen • Graduate Institute of Medicine and Center for Research Resources and Development, Kaohsiung Medical University, Kaohsiung, Taiwan Silvia De Rubeis • Center for Human Genetics, Katholieke Universiteit Leuven, Leuven, Belgium; Department of Molecular and Developmental Genetics, Leuven, Belgium Xinyu Deng • State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Beijing, China Emilie Ernoult • Laboratory of Oncopharmacology-Pharmacogenetics, Centre INSERM Régional de Recherche sur le Cancer U892, Centre Régional de Lutte Contre le Cancer Paul Papin, Angers, France Silvina A. Fratantoni • OncoProteomics Laboratory, Department of Medical Oncology, VU University Medical Center-Cancer Center Amsterdam, Amsterdam, The Netherlands Catherine Guette • Laboratory of Oncopharmacology-Pharmacogenetics, Centre INSERM Régional de Recherche sur le Cancer U892, Centre Régional de Lutte Contre le Cancer Paul Papin, Angers, France Fuchu He • State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Beijing, China Markus Islinger • Department of Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany Ying Jiang • State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Beijing, China Connie R. Jimenez • OncoProteomics Laboratory, Department of Medical Oncology, VU University Medical Center-Cancer Center Amsterdam, Amsterdam, The Netherlands xi

xii

Contributors

Peter Kaiser • Department of Biological Chemistry, College of Medicine, University of California, Irvine, CA, USA Ronald E. van Kesteren • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Joachim Kirsch • Department of Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany Patricia Klemmer • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Ka Wan Li • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Pei-Yi Lin • Graduate Institute of Medicine and Center for Research Resources and Development, Kaohsiung Medical University, Kaohsiung, Taiwan Maarten Loos • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands David Meierhofer • Max Planck Institute for Molecular Genetics, Berlin, Germany Pim van Nierop • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Thang V. Pham • OncoProteomics Laboratory, Department of Medical Oncology, VU University Medical Center-Cancer Center Amsterdam, Amsterdam, The Netherlands Uroš Rajčević • NorLux Neuro-Oncology Laboratory, CRP-Santé, Strassen, Luxembourg; Department of Genetic Toxicology and Cancer Biology, National Institute of Biology, Ljubljana, Slovenia Elena V. Romanova • Department of Chemistry and the Beckman Institute, University of Illinois, Urbana, IL, USA Ramona Rotaru • Department of Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany Jens Schindler • Institut für Biologie und Umweltwissenschaften, Arbeitsgruppe Neurogenetik, Carl-von-Ossietzky Universität Oldenburg, Oldenburg, Germany Roel C. vd Schors • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Karl-Heinz Smalla • Leibniz Institute for Neurobiology, Magdeburg, Germany August B. Smit • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Sabine Spijker • Department of Molecular and Cellular Neurobiology, Center for Neurogenomics and Cognitive Research, VU University, Amsterdam, The Netherlands Tatsuo Suzuki • Department of Neuroplasticity, Institute on Aging and Adaptation, Shinshu University Graduate School of Medicine, Matsumoto, Japan

Contributors

xiii

Jonathan V. Sweedler • Department of Chemistry and the Beckman Institute, University of Illinois, Urbana, IL, USA Chia-Feng Tsai • Graduate Institute of Medicine and Center for Research Resources and Development, Kaohsiung Medical University, Kaohsiung, Taiwan Walter Volknandt • AK Neurochemistry, Biocenter of JW Goethe-University, Frankfurt/Main, Germany Yi-Ting Wang • Graduate Institute of Medicine and Center for Research Resources and Development, Kaohsiung Medical University, Kaohsiung, Taiwan Gerhard Weber • FFE Service GmbH, Kirchheim, Germany Ursula Wyneken • Universidad de los Andes, Santiago, Chile Fang Xie • Department of Chemistry and the Beckman Institute, University of Illinois, Urbana, IL, USA Dong Yang • State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Beijing, China

Part I Introduction

wwwwwwwwwwwwwwwwww

Chapter 1 Neuroproteomics: Deciphering Brain Function and Disorders Ka Wan Li Abstract Neuroproteomics is a branch of proteomics that specifically studies qualitatively and/or quantitatively the tissue/organelle proteomes of the nervous system. This chapter introduces the various aspects of neuroproteomics, and outlines the range of methods that are commonly employed. Key words: Proteomics, Brain function, Brain disorder, Methods

1. Introduction The brain is the most complex and dynamically organized organ of the human body with a high degree of computation capability enabling the execution of a wide spectrum of physiological processes and behaviors based on integrative neurotransmission among neural systems and across brain regions. Abnormalities of neuronal activity and cellular dysfunction are known to underlie a large number of brain disorders, including neurodegenerative diseases, addiction, psychiatric disorders, and mental retardation. While brain diseases represent a major socio-medical burden, the underlying mechanisms remain largely unclear. Due to the complexity of the nervous system, the majority of the studies are focused on single to a few genes or proteins. However, protein function depends primarily on the coordination of extended series of molecular events, and less on the properties of single molecules. The classical approach does not yield sufficient insight into the mechanisms by which neural processes are accomplished or how they become dys-regulated.

Ka Wan Li (ed.), Neuroproteomics, Neuromethods, vol. 57, DOI 10.1007/978-1-61779-111-6_1, © Springer Science+Business Media, LLC 2011

3

4

Li

The necessity to study brain function at a systems biology level was recognized already decades ago. For example, the neuronal activity dependent alteration of global protein expression or protein phosphorylation patterns have been examined in various models, e.g., an invertebrate model of learning and memory and a rat model of addiction (1, 2). These classical approaches depended on the use of two-dimensional (2D) gel electrophoresis to display and quantify proteins present in the system of interest. In recent years, there were major advancements in analytical techniques (3), especially the improvement of scanning speed, mass accuracy, and sensitivity of mass spectrometry (MS). In conjunction with the use of capillary liquid chromatography (in the form of 1D or 2D separation of the analytes) coupled off- or on-line to the mass spectrometer, the quantitative analysis of (organellar) proteomes with good reproducibility and high protein coverage was realized. In parallel, the corresponding development of bioinformatics tools can now handle huge amount of data, and interpret the data with meaningful biological significance. In essence, neuroproteomics has come of age.

2. Category Modern neural science integrates the studies of molecular/cellular and network/system biology to explain brain function, behavior, and disease of the brain. While these studies require multidisciplinary approaches, neuroproteomics is emerging as the driving force that (potentially) shapes and even redefines the directions of the studies. Today, neuroproteomics is extensively used: (a) to register the expression changes of the neuronal proteomes underlying neuroplasticity or brain disorders (4, 5), (b) to describe the molecular mechanisms that are employed to organize neuronal organelles such as the synapse (6, 7), and (c) to discover and measure biomarkers of human diseases by screening of body fluids, in particular cerebrospinal fluid and/or serum (8). (a) Expression Proteomics. These studies aim at detecting the alteration of protein expression patterns of a particular sample type under different conditions. This requires a quantitative analysis of the sample proteome. This is the most common application of neuroproteomics and has been used in the studies of mechanisms underlying neuroplasticity, drug addiction, neurodegenerative diseases, mental retardation, and psychiatric disorders (see reviews (4–6, 9, 10)). (b) Functional Proteomics. This addresses how proteins are organized into complexes, and how multiprotein complexes form the organelles that drive physiological processes.

Neuroproteomics: Deciphering Brain Function and Disorders

5

Current functional neuroproteomics studies focus mainly on the structural organization of the synapse, the main building block of the neuronal communication channel. The proteomes of synaptic vesicle, presynaptic terminal, and postsynaptic density, have been examined by various proteomics techniques (see review (6)). Several synapse protein complexes, especially membrane receptors and ion channels and their associated proteins have been elucidated, yielding, in many cases, novel insights into synapse function and plasticity (see review (7)). Large-scale analysis of the synapse protein interactome and its dynamics under different experimental conditions is underway. Posttranslational modifications of proteins are well known factors that change protein functions and protein–protein interactions. The analysis of synapse phosphoproteome is now feasible (11). The techniques to detect other equally important modifications, such as ubiquitination and sumoylation, are under development. (c) Biomarker discovery. This involves the identification of the changes of (a) protein(s) in the biofluid that can be used as a marker of the disease or the progression of the disease. Cerebrospinal fluid contains proteins directly derived from the brain, and it is expected that the protein patterns there might reflect the status of the brain. Therefore, cerebrospinal fluid is the method of choice for screening of biomarkers for brain disorders (see Chap. 18) (8).

3. Methods The first consideration is sample preparation. There are reports that used the whole brain as input material. However, in brain region specific protein expression patterns exist (for instance, see Allen Brain Atlas for the brain regions-specific differential expression of genes; http://mouse.brain-map.org), which probably reflect the differences in protein expression in distinct regions. Thus, it is preferred to use a specific-brain region of interest (see Chap. 2), bear in mind that the amount of input may be lower thereby necessitating the pooling of sampled brain material from multiple animals. The earliest expression proteomics experiments used the extract of the brain regions as input. The high complexity of these samples (different cell types and different cell compartments being analyzed together) may well focus attention to the identified proteins according to high abundance and blur the data coming from individual cell compartments, which may hinder the interpretation of the data. To address this problem, recent

6

Li

studies generally incorporate additional step(s) to isolate relevant organelles (such as the synapse or mitochondria) or protein complexes for the proteomics studies (see Chaps. 3–7). Such approach often generates useful hypotheses, which can guide further functional studies (12). In all modern proteomics studies MS instruments are used as a detector to identify and quantify (trypsin-digested) peptides generated from the proteins of the samples. As the complexity of the samples is always too high to be handled by direct infusion into a mass spectrometer, peptides or proteins have to be partially separated before MS analysis. For most analysis, even a simple organelle such as the synapse, thousands of distinct proteins may be present. Therefore, 2D separation is routinely included in most proteomics studies to cope with the sample complexity. There are several commonly employed 2D separation methodologies at the protein level; for example 2D IEF-SDS gel electrophoresis (see Chap. 8) and 2D BAC-SDS gel electrophoresis (see Chap. 9) separating proteins into distinct spots, which are then digested and directly analyzed by MS/MS. The 2D BACSDS gel electrophoresis is suitable for the analysis of membrane proteins, but it has a lower resolution than the 2D IEF-SDS gel electrophoresis. The 1D-PAGE/LC–electrospray (ESI) MS/MS approach is another popular approach (see Chaps. 12 and 14). Here, proteins are separated on SDS gel electrophoresis, digested, and the peptides analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The LC run usually consists of the use of capillary reverse phase C18 column with a 1–2 h gradient of increasing acetonitrile concentration in low pH solvent, where peptides sequentially elute from the column according to their hydrophobicity. The column may be coupled online to an ESI mass spectrometer, or offline to a MADLI mass spectrometer. In a “shotgun” proteomics protocol, proteins are digested and peptides are separated by LC–LC and analyzed by MS. The first dimension, LC usually comprises of a strong cation exchange column. Its separation is orthogonal to the second dimensional LC with capillary reverse phase C18 column, thereby maximizing the separation power of the system (see Chap. 10). In recent years, separation of peptides by isoelectric focusing has been employed as the first dimension (see Chap. 11). This has the advantage of a higher resolution than the cation exchange chromatography. Expression proteomics requires quantitation across different samples. 2D IEF-SDS gel electrophoresis and 2D BAC-SDS gel electrophoresis can quantify proteins based on the intensity of the protein spots on the gels. For 1D-PAGE/LC-ESI MS/MS and shotgun proteomics, proteins can be quantified by label free

Neuroproteomics: Deciphering Brain Function and Disorders

7

quantitation of the tryptic peptides (see Chap. 21). In case of shotgun proteomics, relative quantitation may also be done with chemical tagging of each tryptic peptides across different samples with specific groups of isotopic labels, such as the iTRAQ reagents (see Chaps. 10, 11 and 20) (13). Here, up to eight samples can be compared in a single experiment. Stable isotope labeling with amino acids in cell culture (SILAC) is another popular method. SILAC relies on metabolic incorporation of given “light” or “heavy” form of the amino acid into proteins in two respective cell cultures, and use them for quantitative detection of changes of proteins and/or protein posttranslational modifications in response to various stimuli ((14); see Chap. 15). Often, the differences of expression patterns of proteins revealed by expression proteomics were further confirmed by quantitative immunoblotting (see Chap. 13). For years, much effort has been made to detect and quantify protein posttranslational modifications, for example the neuronal activity dependent changes of (sites of) phosphorylation that may underlie learning and memory. As phosphopeptides are minor constituents of all the peptides present in the sample, prior affinity isolation to enrich phosphopeptides is a critical step for the success of their subsequent detection. Affinity materials for phosphopeptides enrichment have been developed. The most used modes of isolation consist of IMAC and TiO2, respectively, through which hundreds to thousand of phosphopeptides can be subsequently detected by (tandem) LC-MS (11). Recently, the optimum condition for IMAC has been worked out, which further increases the specificity of the isolation (see Chap. 14) (15, 16). Other types of modifications including ubiquitination (see Chap. 15) and the closely related sumoylation (17) are equally important in controlling synapse function. Their detection by proteomics technologies are just starting to be addressed. There are several specific topics in neuroproteomics, for example neuropeptidomics and lipodomics. They are less studied, but equally important in the understanding of brain function. The studies generally use LC-MS/MS for the analysis (see Chaps. 17 and 19). Finally, a high through-put imaging technique is used to examine the neuronal proteins implicated in brain disorders (Chap. 16).

4. Perspectives Neuroproteomics promises to revolutionalize the ways we approach the understanding of brain function; i.e., via a systems biology approach and by interrogating the whole, or important

8

Li

parts of a proteome that underpins the system. This will provide a better view of how the individual organelles function and are dynamically regulated in order to drive a physiological process. Future proteomics studies will be used to identify important changes in protein expression and/or posttranslational modification underlying specific disease states. Neuroproteomics is not hypothesis-free. Biological questions need to be defined critically. This will guide the choice of sample preparation, the analytical tools that need to be used, and how to deal with interpretation of the results. Many of the proteomics tools are emerging technologies that may not be familiar to neuroscientists. To this end, the following chapters provide a comprehensive guide to neuroproteomics. We have covered most of the recent proteomics methods, and some of the applications. I believe that these chapters will be of assistance to the readers to design and execute their own proteomics experiments in an optimal way. References 1. Guitart, X., and Nestler, E. J. (1989) Identification of morphine- and cyclic AMPregulated phosphoproteins (MARPPs) in the locus coeruleus and other regions of rat brain: regulation by acute and chronic morphine, J Neurosci 9, 4371–4387. 2. Castellucci, V. F., Kennedy, T. E., Kandel, E. R., and Goelet, P. (1988) A quantitative analysis of 2-D gels identifies proteins in which labeling is increased following long-term sensitization in Aplysia, Neuron 1, 321–328. 3. Choudhary, C., and Mann, M. (2010) Decoding signalling networks by mass spectrometry-based proteomics, Nature reviews 11, 427–439. 4. Li, K. W., and Jimenez, C. R. (2008) Synapse proteomics: current status and quantitative applications, Expert review of proteomics 5, 353–360. 5. Lull, M. E., Freeman, W. M., VanGuilder, H. D., and Vrana, K. E. (2010) The use of neuroproteomics in drug abuse research, Drug and alcohol dependence 107, 11–22. 6. Li, K. W., and Smit, A. B. (2008) Subcellular proteomics in neuroscience, Front Biosci 13, 4416–4425. 7. Li, K. W., Klemmer, P., and Smit, A. B. (2010) Interaction proteomics of synapse protein complexes, Analytical and bioanalytical chemistry 397, 3195–3202. 8. Westman-Brinkmalm, A., Ruetschi, U., Portelius, E., Andreasson, U., Brinkmalm, G.,

Karlsson, G., Hansson, S., Zetterberg, H., and Blennow, K. (2009) Proteomics/peptidomics tools to find CSF biomarkers for neurodegenerative diseases, Front Biosci 14, 1793–1806. 9. Kovacech, B., Zilka, N., and Novak, M. (2009) New age of neuroproteomics in Alzheimer’s disease research, Cellular and molecular neurobiology 29, 799–805. 10. Bayes, A., and Grant, S. G. (2009) Neuroproteomics: understanding the molecular organization and complexity of the brain, Nat Rev Neurosci 10, 635–646. 11. Trinidad, J. C., Thalhammer, A., Specht, C. G., Lynn, A. J., Baker, P. R., Schoepfer, R., and Burlingame, A. L. (2008) Quantitative analysis of synaptic phosphorylation and protein expression, Mol Cell Proteomics 7, 684–696. 12. Van den Oever, M. C., Goriounova, N. A., Li, K. W., Van der Schors, R. C., Binnekade, R., Schoffelmeer, A. N., Mansvelder, H. D., Smit, A. B., Spijker, S., and De Vries, T. J. (2008) Prefrontal cortex AMPA receptor plasticity is crucial for cue-induced relapse to heroin-seeking, Nature neuroscience 11, 1053–1058. 13. Li, K. W., Miller, S., Klychnikov, O., Loos, M., Stahl-Zeng, J., Spijker, S., Mayford, M., and Smit, A. B. (2007) Quantitative proteomics and protein network analysis of hippocampal synapses of CaMKIIalpha mutant mice, Journal of proteome research 6, 3127–3133.

Neuroproteomics: Deciphering Brain Function and Disorders 14. Hilger, M., Bonaldi, T., Gnad, F., and Mann, M. (2009) Systems-wide analysis of a phosphatase knock-down by quantitative proteomics and phosphoproteomics, Mol Cell Proteomics 8, 1908–1920. 15. Ye, J., Zhang, X., Young, C., Zhao, X., Hao, Q., Cheng, L., and Jensen, O. N. (2010) Optimized IMAC-IMAC Protocol for Phosphopeptide Recovery from Complex Biological Samples, Journal of proteome research 9, 3561–3573.

9

16. Tsai, C. F., Wang, Y. T., Chen, Y. R., Lai, C. Y., Lin, P. Y., Pan, K. T., Chen, J. Y., Khoo, K. H., and Chen, Y. J. (2008) Immobilized metal affinity chromatography revisited: pH/acid control toward high selectivity in phosphoproteomics, Journal of proteome research 7, 4058–4069. 17. Martin, S., Nishimune, A., Mellor, J. R., and Henley, J. M. (2007) SUMOylation regulates kainate-receptor-mediated synaptic transmission, Nature 447, 321–325.

wwwwwwwwwwwwwwwwww

Part II Fractionation of Brain Regions, Organelles, and Protein Complexes

wwwwwwwwwwwwwwwwww

Chapter 2 Dissection of Rodent Brain Regions Sabine Spijker Abstract Dissection of brain tissue is an important step in sample preparation for (subcellular) proteomics studies. In this chapter, brain removal and separate dissection of multiple brain regions from a single brain are described in step-by-step protocol. This concerns dissection from fresh or frozen tissue of cerebellum, hippocampus, prefrontal cortex, and striatum. Key words: Dissection, Mouse, Brain, Hippocampus, Cerebellum, Cortex

1. Introduction The mammalian brain is structurally organized into distinct anatomical regions (Fig. 1) mostly exerting dedicated control over specific physiological and behavioral functions. Various brain structures of interest to neuroscientists are the olfactory bulb, which is involved in sensing odors, the prefrontal cortex, which is important for executive control together with the striatum, the latter of which is also involved in integration of movement and reward processing. The hippocampus controls formation and retrieval of associative and episodic memories, the cerebellum is involved in motor learning and coordination, and controls voluntary learned physical movements. The pons contains a white matter tract of cranial nerves that among others relays signals between the cerebrum and cerebellum, and the medulla controlling autonomic functions, and relaying information between the brain and spinal cord. At the cellular level, synaptic strength and neuroplasticity show clear brain region-specific features, as is evident from different plasticity mechanisms related to learning in cerebellum and hippocampus (1–3). Accordingly, the synapses across these brain regions show overlapping yet distinct proteomes (4, 5). Ka Wan Li (ed.), Neuroproteomics, Neuromethods, vol. 57, DOI 10.1007/978-1-61779-111-6_2, © Springer Science+Business Media, LLC 2011

13

14

Spijker

The majority of quantitative neuroproteomics studies today focus on the indicated brain regions (Fig. 1). In this chapter, I will discuss methods to dissect multiple brain regions from a single brain based on existing atlases (6, 7), as a way to allow region-specific (subcellular) proteomics analyses. 1.1. Fresh vs. Frozen Dissection

Fresh dissection of neuronal tissue has the advantage that particular brain regions can easily be dissected based on visual information, such as differences in color of adjacent tissues, and on the natural anatomical boundaries of certain regions present in the brain. Examples of these are the cerebellum that can be easily taken off from the medulla and pons (Figs. 1a and 2a), and are distinct in color, and the hippocampus that differs from the occipital cortex by color and because it lies loosely on the thalamus, basically only connected by the fornix (Fig. 1b). For other tissues, such as medial prefrontal cortex (mPFC) and striatum, dissection in both fresh and frozen tissue could be carried out, with frozen tissue likely yielding a slightly more accurate dissection, because it allows thinner sections to be made. In fresh tissue, dissection of these structures is more challenging as these regions are highly

Fig. 1. (a) Schematic representation of a sagittal section of the mouse brain. Indicated are specific brain regions. Dissection of the cerebellum, the hippocampus, the medial prefrontal cortex (mPFC; gray), and striatum (dorsal and ventral) will be discussed. (b) Schematic representation of the hippocampus and fornix in the brain.

Fig. 2. Dissection tools. (a) For mouse brain dissection, a large curved serrated (front) and a small curved serrated (back) forceps are used. (b) Examples (Fine Science tools (FST) 11003-12, FST 11152-10) are shown in comparison to a mouse brain. (c) Scalpel with holder for subsequent dissection of the mPFC and striatum.

Dissection of Rodent Brain Regions

15

interconnected, and change shape going from rostral to caudal. However, it should be noticed that with today’s biochemical methods for isolation of subcellular structures, brain slices should not be thinner than ~300–400 mm as they will float in the lysis buffer and hence it becomes hard to homogenate these properly.

2. Materials The products used are listed below. Comparable products from other suppliers should also be effective. Underlined is equipment visible on the photographical and schematic representation of the dissections (Figs. 3–14). 2.1. B rain Removal

1. Surgical scissors – Straight sharp/blunt 12 cm (Fine Science tools (FST) 14001-12). 2. Narrow pattern forceps – Curved 12 cm, 2 × 1.25 mm (FST 11003-12; Fig. 2a, b). 3. Iris scissors – Large loops, angled (FST 14107-09; Fig. 2b).

2.2. B rain Dissection

1. Metal iron-free plate (favorably 10 × 10 cm; 5–30 mm thick). 2. For dissection of a mouse brain: One small curved blunt forceps with serrated tips for dissection (Graefe Forceps – 0.8 mm tips curved serrated, 0.8 × 0.7 mm (FST 11052-10), or Graefe Forceps – 0.5 mm tips curved serrated, 0.5 × 0.4 mm (FST 11152-10; Fig. 2a, b)) and one larger forceps with curved

Fig. 3. Schematic representation of the mouse skull, showing the different bone plates. Indicated are the cuts to be made (dark gray hatched line) in the interparietal bone, as well as in the anterior part of the frontal bone.

16

Spijker

Fig. 4. Removal of the brain from the skull. Take care of the meninges (arrows) as these could rupture the brain while taking it out. Carefully cut the cranial nerves (arrowheads) upon taking out the brain.

blunt tips to hold the brain, e.g., the narrow pattern forceps (FST 11003-12). For a rat brain, larger forceps are needed, like the Graefe Forceps – 1.0 mm Tips Curved Serrated, 1 × 0.9 mm (FST 11652-10), Semken forceps curved Serrated – 13 cm, 1.3 × 1 mm (FST 11009-13), or FST 11003-12 for


17

Fig. 5. Dissection of the cerebellum, freed from the medulla, and the pons. The arrow indicates the colliculus inferior.

Fig. 6. Opening the cortex from the midline to free the hippocampus.

dissection and Narrow Pattern Forceps – Curved 14.5 cm, 2.25 × 1.45 mm (FST 11003-14) or Standard Pattern Forceps – Curved 12 cm, 2.5 × 1.35 mm (FST 11001-12) to hold the brain. 3. Razor blade (dissection of mPFC and striatum). 4. Scalpel handle (FST #10003-12 Scalpel Handle #3–12 cm) with scalpel (FST #10011-00) as illustrated in Fig. 2c (dissection of mPFC and striatum).

18

Spijker

Fig. 7. Removal of cortex from the right hippocampus. The arrow indicates a piece of cortex left behind on the hippocampus. Note the difference in color between the two structures.

Fig. 8. Removal of cortex from the left hippocampus.


19

Fig. 9. Dissection of the right hippocampus. Arrowhead indicates a piece of cortex. The double arrow indicates the direction in which the hippocampus is rolled to free it from the cortex.

Fig. 10. Removal of a piece of cortex left behind on the hippocampus (arrowhead ). Note the difference in color between the two structures.

Fig. 11. Dissection of the left hippocampus.

20

Spijker

Fig. 12. Placing back the cortex after hippocampal dissection in preparation of subsequent dissections. The brain could be snap-frozen or used for dissection of fresh tissue as presented in Figs. 13 and 14.

Fig. 13. Coronal slices are made in preparation of subsequent dissections of, e.g., the mPFC and striatum, as presented in Fig. 14. Arrowhead in 4 shows the anterior commissure. Anterior forceps of the corpus callosum (AFCC), indicated by an arrow, shine through the section from which the mPFC will be taken (see Fig. 14.1). Open arrow in 9 shows the genus corpus callosum (GCC).


21

Fig. 14. From coronal sections #1–3 (left), the mPFC could be dissected from section #1, and the striatum could be dissected from sections #2–3. b and b’ indicate the outline for dissection (orange line). c shows the dissected parts. AFCC, indicated by an arrow in 1b, shine through the section from which the mPFC will be taken. Open arrow in 2b shows the GCC. Open arrowhead in 2a indicates the boundary of the striatum and the cortex by the capsula externa.

3. Methods 3.1. B rain Removal

1. Use cervical dislocation to prevent pre- and postsynaptic effects of anesthesia (e.g., with protein abundance and phosphorylation (8–11)), and a surgical scissor to remove the head with a cut posterior from the ears. Using the scissors, make a midline incision in the skin. Flip the skin over the eyes to free

22

Spijker

the skull. Make a small incision (Iris scissors) on the top of the skull starting from the caudal part at the point of the (inter) parietal bone (Fig. 3), be careful not to cut through the brain. Make a firm cut through the most anterior part of the skull, between the eyes (frontal bone, Fig. 3). This enables to remove the brain more easily. 2. Tilt one side of the parietal bone with the curved narrow pattern forceps and break it off (Fig. 4.1–3). Do the same with the other side (Fig. 4.4). Most likely the frontal bone will remain. In that case, make a small incision that enables tilting and breaking off this bone plate. Be careful of the meninges that are surrounding the brain and that are between the brain and the skull; they could rupture the brain while breaking off the skull (arrow; Fig. 4.5–7). 3. When the brain is freed from meninges (Fig. 4.8), slide the curved narrow pattern forceps (closed) under anterior part of the brain (olfactory bulb) and tilt the brain gently upward (Fig. 4.9). Slide the forceps further down to break the optic nerves and other cranial nerves (arrowheads; Fig. 4.10, 11) and gently lift the brain out of the skull (Fig. 4.12). 4. Transfer brain to metal plate placed on ice to cool down the brain immediately. Wipe off excess blood. Note that these steps should be performed within 2–3 min. 3.2. Cerebellum Dissection

1. Place the brain with the dorsal side facing the metal plate (Fig. 5.1). 2. Lift with the curved narrow pattern forceps the medulla/ pons upward (Fig. 5.2). 3. Using the small curved forceps (Graefe Forceps – 0.5 mm tip), cut through the pons by closing the forceps around the tissue (Fig. 5.2, 3). 4. When the majority of the white tissue is removed (Fig. 5.4), turn the brain around with the ventral side facing the metal plate (Fig. 5.5). 5. Place the small forceps between the cortical lobes and the cerebellum (Fig. 5.5), and snap the cerebellum off from the colliculus inferior (arrow, Fig. 5.5). 6. Finally, remove possible remaining parts of the pons (Fig. 5.6). Note that all previous steps should be performed within 1 min.

3.3. Hippocampus Dissection

1. Place the brain with the ventral side facing the metal plate (Fig. 6.1). 2. Place the small curved forceps between the cerebral halves in a closed position. Gently hold the brain in position with the large curved forceps (Fig. 6.1, 2).


23

3. Gently open the forceps (Fig. 6.3), thereby slowing the opening of the cortical halves. 4. Repeat this process of placing the closed forceps in between the cortical halves, and opening the forceps (Fig. 6.4, 5). The initial white-colored part encountered is most likely the corpus callosum, under it is the hippocampus. 5. Once an opening is obtained for 60% along the midline, direct the forceps (closed position) 30–40° counterclockwise (Fig. 6.5) to open up the left cortex from the hippocampus by repeatedly opening the forceps. Thereafter repeat the same for the right cortex by pointing the forceps in a 30–40° clockwise direction (Fig. 6.6). 6. Repeat this movement on either side until the upper part of the hippocampus is visible (Fig. 6.7–9). 7. Using the large forceps, gently pick up the cortex (Fig. 7.1). 8. Turn the small forceps (point downward; Fig. 7.2) to free the hippocampus from the cortex without damaging the cortex. Remember to enter the tissue always with closed forceps. 9. Again repeat the process of opening and closing the small forceps while moving them to the caudal part of the hippocampus/cortex boundary (Fig. 7.3–5). 10. Once at the most caudal part of the hippocampus/cortex boundary, move the small forceps through the cortex (Fig. 7.5–8). Possible remainders of cortex (arrow, Fig. 7.8), visible from a more pink/yellow color than the hippocampus (gray, translucent), can be removed at that moment by snapping it off using the small forceps, or can be removed later. 11. Repeat steps 7–10 to remove the left cortex from the hippocampus (Fig. 8.1–8). Figure 8.8 presents the removal of a cortical piece from the hippocampus as explained in step 10. 12. Move the cortical halves anterior from the cortex to reveal the fornix (see Fig. 1b). Using the small forceps cut the hippocampus separate from the fornix (Fig. 8.9). 13. In addition, separate the two halves of the hippocampus (Figs. 8.10 and 9.1, 2). 14. Gently push with closed forceps the hippocampal halve to the side (Fig. 9.3), while keeping the brain in position with the larger forceps (Fig. 9.4). 15. Continue with step 14, until the hippocampus is lying sideways of the brain (Fig. 9.5, 6). 16. Using the small forceps roll the hippocampus out of the brain (direction of arrow, Fig. 9.7–10) to remove it from the cortex that was still adhered (arrowhead, Fig. 9.8).

24

Spijker

17. Inspect the hippocampus for pieces of cortex (yellow/pink), as visible in Fig. 10.1, 2 (arrowhead), and remove them (Fig. 10.3). 18. Repeat steps 14–17 for the left halve of the hippocampus (Fig. 11.1–6) to finally obtain both halves (Fig. 11.7). Note that together these steps should be performed within 2–3 min. 19. When the posterior part of the cortex is not needed as tissue sample, one could at step 6 instead of folding away the cortex, also dissect the cortex away from the medial to the lateral side using the small forceps. In this way, the hippocampus is easily accessible. This will also speed up the process of dissection. 3.4. Prefrontal Cortex and Striatum Dissection (Fresh)

1. After removal of the hippocampus, use the large forceps to fold back the cortex into the original position (Fig. 12.1–7). 2. Then, place the brain with the dorsal side facing the metal plate (Fig. 12.8, 9). We will now start to make coronal sections in which the prefrontal cortex and striatum are visible at different levels (Figs. 14 and 15). 3. Take a sharp razor blade and make sections, the first one being to cut off the olfactory bulb (Fig. 13.1–3). 4. The anterior commissure is well visible at this point (arrowhead, Fig. 13.4). 5. The first section (Fig. 13.5, 6) contains mainly motor cortex. 6. Be aware that the subsequent section contains the anterior forceps of the corpus callosum (AFCC) shining through (arrow, Figs. 13.6 and 14.1b), with a darker area in the middle that represents the mPFC. 7. Cut the section containing the mPFC (Fig. 13.7–9).

Fig. 15. Schematic representation of sections #1–3 from Fig. 14. Indicated are striatum (light gray) and mPFC (dark gray) for dissection. Ventricle is indicated in black.


25

8. In the remaining brain, the genus corpus callosum (GCC; open arrow, Figs. 13.9 and 14.2b) is now well visible, as well as the dorsal and ventral striatum that are separated from the adjacent cortex by the capsula externa (open arrowhead, Figs. 14.2a and 15). 9. After cutting this section (Fig. 13.10), the brain should now show the joining of the anterior commissure that is best visible in Fig. 14.3a. This marks the absence of the ventral striatum in this section. 10. The last section therefore contains only dorsal striatum (Fig. 13.11). 11. From each brain, the last three sections are used for dissection of the mPFC and striatum (Fig. 14, left, mPFC, section #1; striatum, sections #1–3). 12. For the mPFC, take section #1. The mPFC, containing the prelimbic and infralimbic cortex, is visible as a darker area between the AFCC (Fig. 14.1a, b, left section). Note that the infralimbic cortex ends when the GCC is present (Fig. 14.1a, b, right section). 13. Cut through the GCC to dissect the mPFC in a diamond-like shape (Fig. 14.1b, c). Be careful not to take along any material from the AFCC. 14. For the striatum, take sections #1–3. From section #1, the ventral striatum is visible as a darker structure surrounded by the somewhat lighter and less translucent cortex, as well as the AFCC (Fig. 14.1b). 15. From section #2, both the dorsal and ventral striatum have a darker appearance than the surrounding cortex (Fig. 14.2a). Around the midline, the septum, a structure similar in color as the cortex, separates the two striatal halves. 16. Dissect the striatum from the GCC and adjacent capsula externa (caudal and lateral), as well as from the ventricle and septum (medial), and cortex (ventral), as indicated by the natural borders (Fig. 14.2b, c). 17. In section #3, only the dorsal striatum is present, as the anterior commissure now connects both hemispheres. 18. Dissect the dorsal striatum from the corpus callosum and adjacent capsula externa (caudal and lateral), as well as from the ventricle and septum (medial), and cortex (ventral), as indicated by the natural borders (Fig. 14.3b, c). 3.5. Prefrontal Cortex and Striatum Dissection (Frozen)

1. After replacing the cortex halves into the original position (Fig. 12.1–7, see step 1 in Sect. 3.4), immediately snap-freeze the brain. This can be done in liquid nitrogen, or in isopentane cooled on dry ice. Be careful with the latter method, as isopentane is toxic and should be discarded properly.

26

Spijker

2. Store the tissue (−80°C) in aluminum foil to prevent freezedrying. 3. Prior to dissection, place the brain for at least 1 h at −20°C, preferably on a metal plate in a cryostat. 4. After the tissue reached a stable temperature, take a razor blade to make manual coronal sections. Because the slices are thinner than when sectioning fresh tissue, more precision can be obtained in this way. However, fresh tissue yields a higher degree of visual information on the natural boundaries of brain areas that is partially lost upon freezing. 5. Based on the same visual cues as described above, cut out the regions of interest with a scalpel. Collect the tissue in an eppendorf tube, and store at −80°C until further use. 6. Note that putting plastic tubing around the scalpel holder will reduce the transfer of body heat to the scalpel, and thereby prevents the tissue from being thawed. In addition, wearing double gloves better insulates from cold and prevents heat being carried over to the tissue. References 1. Zhuo M, Hawkins RD (1995): Long-term depression: a learning-related type of synaptic plasticity in the mammalian central nervous system. Rev Neurosci 6:259–277. 2. Teyler TJ, Discenna P (1984): Long-term potentiation as a candidate mnemonic device. Brain Res 319:15–28. 3. Jorntell H, Hansel C (2006): Synaptic memories upside down: bidirectional plasticity at cerebellar parallel fiber-Purkinje cell synapses. Neuron 52:227–238. 4. Olsen JV, Nielsen PA, Andersen JR, Mann M, Wisniewski JR (2007): Quantitative proteomic profiling of membrane proteins from the mouse brain cortex, hippocampus, and cerebellum using the HysTag reagent: mapping of neurotransmitter receptors and ion channels. Brain Res 1134:95–106. 5. Trinidad JC, Thalhammer A, Specht CG, Lynn AJ, Baker PR, Schoepfer R, et al (2008): Quantitative analysis of synaptic phosphorylation and protein expression. Mol Cell Proteomics 7:684–696. 6. Paxinos G, Franklin KBJ (2003): The mouse brain in stereotaxic coordinates: Academic press, San Diego.

7. Williams RW (1999): The Mouse Brain Library. http://www.mbl.org/atlas165/atlas165_start. html. 8. Futterer CD, Maurer MH, Schmitt A, Feldmann RE, Jr., Kuschinsky W, Waschke KF (2004): Alterations in rat brain proteins after desflurane anesthesia. Anesthesiology 100: 302–308. 9. Snyder GL, Galdi S, Hendrick JP, Hemmings HC, Jr. (2007): General anesthetics selectively modulate glutamatergic and dopaminergic signaling via site-specific phosphorylation in vivo. Neuropharmacology 53:619–630. 10. Westphalen RI, Hemmings HC, Jr. (2006): Volatile anesthetic effects on glutamate versus GABA release from isolated rat cortical nerve terminals: 4-aminopyridine-evoked release. J Pharmacol Exp Ther 316: 216–223. 11. Westphalen RI, Hemmings HC, Jr. (2006): Volatile anesthetic effects on glutamate versus GABA release from isolated rat cortical nerve terminals: basal release. J Pharmacol Exp Ther 316:208–215.

Chapter 3 Subcellular Fractionation of Brain Tissue Using Free-Flow Electrophoresis Markus Islinger, Joachim Kirsch, Sabine Angermüller, Ramona Rotaru, Afsaneh Abdolzade-Bavil, and Gerhard Weber Abstract Accurate annotation of protein identifications in organellar proteomics highly depends on the sample quality with special respect to contaminations from other subcellular compartments. In this respect, Freeflow electrophoresis (FFE) offers a valuable alternative to classical centrifugation techniques, since it relies on quite different physical parameters. During the last years, FFE has been successfully used for the separation of various organelles from different tissues, yet is largely unknown in the field of neurobiology. Here we present two separation schemes for the fractionation of a synaptic preparation from rat brain using different modes of FFE. Isotachophoresis (ITP), a focusing technique separating organelles according to their electrophoretic mobilities, was able to distribute the synaptosome sample into different subfractions: mitochondrial cross contaminations showed the highest electrophoretic mobility and migrated nearest to the anode of the FFE instrument; proximate to these, proteins of the presynaptic compartment accumulated, whereas nearest to the cathode of the instrument postsynaptic marker proteins were predominantly found. As a nonfocusing technique, zonal FFE does not possess a separation capacity comparable to ITP; however, due to a continuous separation mode, it is adapted to process higher sample amounts and can be used for large-scale separations. We applied zonal FFE to the same starting material as in ITP and were able to separate mitochondria from synaptic material of the preparation, thus offering a fast alternative to clean synaptosome preparations from residual mitochondrial contaminations. Key words: Synaptosomes, Mitochondria, Organelle separation, Electrophoresis, Proteomics

1. Introduction Organellar proteomics has been successfully used to extend our view on various subcellular compartments from different tissues, leading to the identification and localization of new proteins further to be characterized by functional analyses (1, 2). With respect to a complete elucidation of the mammalian brain, proteome mass spectrometric studies based on total tissue homogenates fail


27

28

Islinger et al.

to identify brain-specific proteins in greater numbers (3). By contrast, proteomic studies on subcellular compartments isolated from specific brain areas were able to comprise and expand the spectrum of neuronal or synaptic proteins (4, 5). Moreover, quantitative mass spectrometry was successfully applied to characterize functional proteome changes in synaptosomes during mouse development (6) or to identify the effect of potential neurotoxic agents (7). However, interorganellar contaminations, in synaptic preparations especially mitochondria, are still a major drawback for a correct allocation of low-abundant proteins and are of special concern in such a complex tissue as brain (5). Synaptosomes, generated from disrupted, resealing synapses during homogenization of brain tissue, have been used as the preferable object to investigate the physiology of neurotransmitter release and to identify and characterize the cellular protein structures involved in this process (8). Moreover synaptosomal preparations are typically the starting material for the preparation of the postsynaptic density (PSD) or synaptic vesicles used for the identification of synapse-specific protein constituents in recent proteomic studies (9) (see also Chap. 4). With regard to synaptosomal preparations isolated by standard centrifugation techniques, major contaminants are mitochondria, which prevent a definite identification of true synaptic proteins in the range of low-abundant proteins. On the other hand, the investigation of the status of synaptic mitochondria may be of interest with respect to the pathogenesis of Alzheimer’s disease (10, 11). Free-flow electrophoresis (FFE) has been repeatedly invented for the isolation of various organelles (12–18), and new developments during the last years have given new impetus on the susceptibility of this promising but still underestimated technique (19–26). However, despite the successful use of FFE for the isolation and subsequent proteomic characterization of subcellular compartments and protein complexes (19, 27–31), hitherto this technique is comparably disregarded in the field of neuroscience. In contrast to gel-based systems, the driving force for FFE separation is an electric field, which is applied perpendicular to a laminar flow of an aqueous, nonmatrix-assisted medium in the separation chamber (Fig. 1). Thus, the analytes are separated mainly according to their surface charge density, a physicochemical parameter unrelated to the organelle’s buoyant density. Thus, FFE offers the possibility to separate particles of similar density, which are beyond the resolution of mere gradient centrifugation. More important, centrifugation and FFE can be combined in a 2D-separation scheme expanding the capacity to isolate organelles of enhanced purity even from complex tissues such as brain. Basically, FFE can be carried out in three different modes, namely isoelectric focusing, isotachophoresis (ITP), and zonal electrophoresis. While isoelectric focusing remains restricted to the separation of proteins and protein

Subcellular Fractionation of Brain Tissue Using Free-Flow Electrophoresis

29

a +

L

+

C

L

+

C

L

C

S1

C

T

C E

C

T

– –

T

S2 E

–

mobility

T = Terminating ion, C/D/E = Analytes, L = Leading ion, S1/S2 = spacers

b

Zonal-FFE C2

C1

FF-Isotachophoresis C2

C1

C3

C3

+

–

+

–

anode

cathode

anode

cathode

S1

I1 I2

S2 S3 S4

I3

I4

I5 I6

separation medium stabilization media

S1

I7

I1 I2

S2 S3 S4

I3

I4

I5 I6 I7 I8

leader ion m. Sp.T. stabilization media

Fig. 1. (a) Schematic illustration of the principal of isotachophoresis (ITP). ITP is a discontinuous electrophoresis mode, where the analytes (C−, D−, E−) migrate in the electric field between leading (L−) and terminating ions (T−). During the electrophoresis, the analyte mixture is fractionated into its individual components according to differences in their electrophoretic mobility (a more detailed explanation is given in the text). (b) Main FFE-separation modes for organelle purification. Whereas, free-flow (FF) ITP is a focusing technique concentrating the sample during the run, in zonal FFE, consisting of a homogenous buffer system, the sample is increasingly diluted during its passage in the separation chamber. The abbreviations used for the various buffer inlets correspond to the terms used in the text section (C counterflow inlet; S sample inlet; I medium inlet).

30

Islinger et al.

complexes, organelles tend to aggregate while reaching a pH similar to their pI. The ITP and zonal FFE are suitable for the separation of membrane-surrounded particles. Both techniques rely on the net surface charge of the particles to be separated and their frictional resistance in the separation medium (background electrolyte), in short their electrophoretic mobility. In zonal FFE, the individual analytes migrate according to mutually different but constant velocities in a continuous electric field depending on the homogenously distributed background electrolyte. By contrast, in ITP the sample substances are separated into individual zones without any background electrolyte, and the electric field is only maintained by the analyzed substances. Since the electric current cannot be interrupted, the different substances assemble in stacks, which migrate with the same velocity. Due to these fundamental differences in the separation principal, both techniques offer individual assets and drawbacks for the separation of subcellular compartments which we will discuss in this chapter. More specific, we present two applications for the separation of mitochondria and synaptosomes from brain. The separation of a synaptosome sample by Free-flow (FF)-ITP resulted in an enrichment of mitochondria at the anodic side of the instrument; synaptosomes were further separated into fractions with higher percentages in presynaptic synaptosomes or postsynaptical membrane sheets. Alternatively, a mitochondria-enriched prefraction from an Optiprep-gradient was effectively purified from remaining synaptosomal contaminations using zonal FFE, underlining the suitability of FFE for the preparation of highly purified organelles as required for their subsequent proteomic characterization. Both examples show the capacity of FFE for the purification of subcellular compartments from neurons, but nevertheless should not stand as an end point. We encourage the readers to further modify and adapt the systems for their own applications.

2. Materials 2.1. Prefractionation of Synaptosomes and Mitochondria

1. Motor-driven Potter-Elvehjem tissue grinder with loose- fitting pestle (clearance 0.1–0.15 mm). 2. Homogenization buffer (HB): 320 mM sucrose, 5 mM HEPES, 1 mM ethylenediaminotetraacetic acid (EDTA), 2 mM phenylmethylsulphonylfluoride (PMSF), 1 mM dithiothreitol (DTT), 1 mM e-aminocaproic acid, pH adjusted to 7.4 with KOH. 3. Gradient Buffer (GB): 5 mM HEPES, 1 mM EDTA, 2 mM PMSF, 1 mM DTT, 1 mM e-aminocaproic acid, pH adjusted to 7.4 with KOH.


31

4. Refrigerated centrifuge with a fixed angle rotor, e.g., Beckman Avanti J-25 and JA-20 rotor. 5. Ultracentrifuge with a fixed-angle rotor and appropriate centrifugation tubes for synaptosomes or a vertical-type rotor for mitochondria: Beckman Optima LE-80K with a Beckman 45Ti rotor and 70 mL polycarbonate bottles 38 × 102 mm or a Beckman VTi 50 vertical-type rotor and Quick-seal polyallomer tubes 25 × 89 mm. 6. Animals: rats of 220–225 g body weight (~8 weeks old or mice at the same age) (Note 1). 2.2. Free-Flow Isotachophoresis

1. All chemicals used for the preparation of buffers, unless otherwise indicated, were purchased from Sigma-Aldrich, Munich, Germany and were of the highest purity available. 2. Anodic stabilization medium: 200 mM morpholinoethanol, 100 mM HCl, 0.1% hydroxypropylmethylcellulose (HPMC) (Notes 2 and 3). 3. Cathodic stabilization medium: 100 mM HEPES, 50 mM NaOH, 0.1% HPMC. 4. Leader ion medium: 20 mM morpholinoethanol, 10 mM HCl, 0.1% HPMC. 5. Spacer mix: Prepare a spacer stock solution of 4 mM acetic acid, 4 mM 2-hydroxyisobutyric acid (HIBA), 10 mM 2-(N-morpholino)ethanesulfonic acid (MES), 10 mM 4-pyridineethanesulfonic acid (PES), 10 mM glucuronic acid, 8 mM 3-(N-morpholino)-2-hydroxypropanesulfonic acid (MOPSO), 7.5 mM MOPS, 50 mM HEPES, 134 mM morpholinoethanol, 40 mM imidazole. Dilute to 15% (w/w) with counterflow medium. 6. Terminator ion medium: 20 mM HEPES, 10 mM NaOH, 0.1% HPMC. 7. Counterflow medium: 250 mM sucrose, 0.1% HPMC. 8. Electrode anode circuit: 200 mM morpholinoethanol, 100 mM HCl, 1% formaldehyde. 9. Electrode cathode circuit: 100 mM HEPES, 50 mM NaOH. 10. SPADNS solution: for the stripe test sulfanilic acid, azochromotrope (SPADNS) is diluted 1:100 in water (~50 mL of SPADNS solution is required per test). 11. pI-Marker: for the performance test, use a 1:10 dilution (v/v) of the BD™ Free-Flow electrophoresis pI-Marker in spacer medium. 12. Pump tubes (Tygon standard; Ismatec, Wertheim-Mondfeld, FRG): media, 6 tubes of 0.64 mm ID (E1–E5, E8) and 2 tubes of 0.38 mm ID (E6, E7); counterflow, 1 tube of

32

Islinger et al.

0.64 mm ID, 1 tube of 1.42 mm ID, 1 tube of 0.64 mm ID; sample, 1 tube of 0.51 mm. 13. Parameters given are suitable for a BD™ Free-Flow Electrophoresis System. However, with slight modifications, the procedure should be adaptable to the OCTOPUS and TECAN FFE instrument. 2.3. Zonal Free-Flow Electrophoresis

1. Anodic stabilization medium: 320 mM sucrose, 25 mM sodium acetate, 50 mM MES, 100 mM H2SO4, 250 mM morpholinoethanol, 0.1% HPMC. 2. Cathodic stabilization medium: 320 mM sucrose, 25 mM NaCl, 150 mM NaOH, 30 mM Tris, 200 mM 3-[[2-hydroxy1,1-bis(hydroxymethyl)ethyl]amino]-1-propanesulfonic acid (TAPS), 0.1% HPMC. 3. Separation medium: 320 mM sucrose, 2.5 mM NaCl, 7 mM TAPS 14 mM morpholinoethanol, 0.1% HPMC. 4. Counterflow: 320 mM sucrose, 0.1% HPMC. 5. Electrode anode circuit: 100 mM H2SO4, 250 mM morpholinoethanol, 25 mM sodium acetate. 6. Electrode cathode circuit: 150 mM NaOH, 200 mM TAPS, 25 mM NaCl. 7. Pump tubes (Tygon standard; Ismatec): Media, 7 tubes of 0.64 mm ID (E1–E7); counterflow, 1 tube of 0.86 ID, 1 tube of 1.42 ID, 1 tube of 0.86 ID; sample, 1 tube of 0.64 mm ID.

2.4. Immunoblotting

1. NuPAGE Novex Bis-Tris 4–12% polyacrylamide gels (Invitrogen, Karlsruhe, Germany), 20× NuPAGE MES SDS running buffer and 4× NuPAGE LDS sample buffer. 2. Blocking solution: 10% Fetal calf serum, 0.1% Tween-20 in phosphate-buffered saline (PBS), pH 7.4. 3. Primary antibodies: mouse anti-synaptophysin (SigmaAldrich), mouse anti-PSD 95 (Upstate, Lake Placid, USA), mouse anti-ATP-synthase (BD Biosciences, San Jose, USA), rabbit anti-ERp29 (Abcam, Cambridge, UK), rabbit antiNMDA 2A (Affinity Bioreagents, Golden, USA), mouse antiGlycine receptor (32). 4. Secondary antibodies: polyclonal goat anti-rabbit, polyclonal rabbit anti-mouse (Sigma-Aldrich). 5. Detection reagents: Amersham™ Hyperfilm ECL, Amersham™ ECL Western blotting detection reagents (GE Healthcare, Buckinghamshire, UK).

2.5. Electron Microscopy

1. Fixation solution: dilute 25% glutaraldehyde stock-solution (Sigma-Aldrich) to a final concentration of 2% in 0.15 M PIPES, pH 7.4.


33

2. Wash solution: 0.15 M PIPES, pH 7.4. 3. OsO4 (Polysciences, Eppelheim, Germany) (Note 4). 4. 75, 85, 95, and 100% (v/v) ethanol in H2O. 5. Epon 812 (Fluka, Munich, Germany). 6. BEEM-capsules (Beem, West Chester, USA). 7. Ultramicrotome: e.g., UltraCut, Reichert, Wien, Austria. 2.6. Measurement of Succinate Dehydrogenase Activity

1. TVBE buffer: 1 mM NaHCO3, 1 mM EDTA, 0.1% (w/v) ethanol, 0.01% (w/v) Triton X-100, pH adjusted to 7.6 with HCl. 2. Reaction buffer – prepare freshly, mixing the following stock solutions (volumes are given for one reaction): 100 mL 50 mM potassium phosphate buffer, pH 7.4, 50 mL 1% Triton WR-1339, 50 mL 2.5 mg/mL Nitrotetrazolium Blue chloride (NBT, prepare from a 50 mg/mL ethanol stock solution), 50 mL 100 mM sodium succinate. 3. Stop-solution: 2% SDS in H2O.

3. Methods As mentioned previously, FFE can be applied for the separation of membranous subcellular compartments in the modes of zone electrophoresis or ITP. In the zonal FFE mode, substances separate in a homogenous separation buffer, which is the source of the electrolytes conducting the electric current. The medium conductivity is practically constant during the separation process and deviations at the positions of the migrating analytes are negligible. Thus, during the run, sample components migrate with slightly different velocities according to their electrophoretic mobilities through the separation medium and are hence separated. However, during the run the sample is permanently diluted with separation buffer resulting in a fan-like separation profile, which distributes the individual analytes in a range of adjacent outlets (Fig. 1b). Thus, resolution in zonal FFE is depending, on the one hand, on the migrated distance of the sample substances in the electric field and, on the other hand, on the residence time in the separation chamber. Therefore, zonal FFE is usually performed in a mode of continuous buffer flow, collecting the separated analytes at the instrument outlet, while still injecting sample mixture at the bottom of the chamber. Consequently, separation conditions have been optimized to conduct separations in a high voltage electric field to increase the electrophoretic migration of the sample compounds and to permit a high flow through the separation chamber. This means that zonal FFE offers advantages in permitting high-throughput separation of large sample amounts

34

Islinger et al.

and to begin sample analysis while the separation is still continued. However, as a nonfocusing technique, zonal FFE is not able to separate the individual sample compounds into sharp-bounded fractions, a weakness, which can be overcome by adapting the separation conditions to optimize the purification of one or a few components of the sample as exemplified here for the separation synaptosomes from mitochondria. By contrast, ITP is carried out in a discontinuous electrolyte system formed by a leading ion with the highest electrophoretic mobility at the front zone, a terminating electrolyte with the lowest electrophoretic mobility at the rear zone, and the compounds to be separated in between (Fig. 1a). By application of a current, leading ions begin to migrate with the upmost velocity in the system. However, since no background electrolyte is present in the sample zone, the electric current, as a flow of charged particles, has to be maintained by the compounds in the sample. Consequently, the different sample compounds migrate initially in the electric field according to their electrophoretic mobility creating first overlapping, later individual zones. Hence, after reaching this steady state, the individual compounds migrate in sharp adjacent zones with the same velocity in the system. Therefore, to enhance the separation capacity of the system spacer, ions with different electrophoretic mobilities are added to the sample mixture to be separated to fit in between the otherwise adjacent analytes. To reach the stage of steady state migration in the FFE-system, ITP is not carried out in a mode of continuous flow through, but in an interval mode, where the sample is first rapidly injected into the separation chamber without application of a current. Thereafter, the electrophoresis is carried out at a reduced medium flow, reversing the flow direction every time the sample front reaches the upper and lower end of the separation chamber, respectively. Finally, when the different compounds of the sample have been focused into sharp individual zones separated by spacer ions at the anodic side of the separation chamber, the current is removed and the analytes are rapidly pumped out into the outlet tubes. Consequently, when the principal parameters of flow conditions have been established, a successful purification mainly depends on the selection of appropriate spacer ions with electrophoretic mobilities fitting between the compounds to be separated. Thus, the high-quality separation of two adjacent sample compounds depends on the existence of an appropriate spacer substance. As a focusing technique, the individual sample compounds are concentrated during the run, permitting the electrophoresis of comparably high sample concentrations. However, the interval FFE mode that is required to reach the state of analytes separated by sharp boundaries leads to a lower sample throughput if compared to zonal FFE. Thus, FF-ITP, while offering benefits in the separation quality if an adequate spacer


35

c omposition can be found, is not suitable for the separation of large sample amounts, e.g., as an initial purification step in a multistep purification process. 3.1. Preparation of Synaptosomeand MitochondriaEnriched Fractions from Rat Brain Homogenates

1. All steps should be performed at a temperature of 4°C. 2. Anesthetize rats (see Note 5), decapitate, insert scissors into the foramen magnum, and open the cranium along the squamosal sutures (see also Chap. 2). Rapidly excise and weigh the brain. For better homogenization, the tissue is cut into small pieces and suspended in ice cold HB at a ratio of 10 mL/g tissue. 3. Homogenize the tissue using ten strokes of 900 rpm in a motor-driven Potter-Elvehjem tissue grinder and a loose fitting pestle. 4. Centrifuge the homogenate at 1,000 × gav for 10 min at 4°C to sediment nuclei and cellular debris. Store the supernatant on ice and rehomogenize the pellet as before. Centrifuge again at 1,000 × gav and combine both supernatants. 5. Centrifuge at 12,000 × gav for 20 min, 4°C. Keep the pellet and discard the supernatant. 6. Resuspend the pellet with a glass rod adding drop-wise HB until a homogenous suspension is gained. Adjust to 12.5 mL with HB and top-load on a sucrose step gradient consisting of 20 mL 0.85 M sucrose, 5 mM HEPES, pH 7.4 and 7.5 mL 1.2 M sucrose, 5 mM HEPES, pH 7.4. Centrifuge for 2 h at 100,000 × gav, 4°C. 7. Aspirate the organellar fractions banding at the border to the 1.2 M sucrose cushion comprising the synaptosomal fraction (#A). Use this fraction in the ITP experiments described in Sect. 3.2. Mitochondria can be found enriched in the pellet of the gradient. 8. For a better recovery of mitochondria, it is recommended to separate the 12,000 × gav pellet on a sigmoidal Optiprepgradient. For this purpose, prepare four Optiprep-solutions in GB comprising densities of 1.07, 1.10, 1.13, and 1.15 g/ mL. Overlay Optiprep-solutions of 3 mL with a density of 1.15 g/mL, 10 mL with 1.13 g/mL, 13 mL with 1.10 g/ mL, and 4 mL with 1.07 g/mL into a Quick-seal centrifugation tube (Beckman), top-load 5 mL of the 12,000 × gav pellet suspended in HB. Overlay with GB and seal the tubes. 9. Centrifuge in a vertical-type rotor at an integrated force of 1,256 × 106 × g × min (gmax = 33,000) with slow acceleration/ deceleration at 4°C. 10. Puncture the tube with a syringe and aspirate the organelle fraction banding at a density of 1.120 g/mL (lowest band, #B).

36

Islinger et al.

3.2. Fractionation of Synaptosomes Using Free-Flow Isotachophoresis

1. Assemble the chamber as described in the operating manual of your FFE system. Use the 0.4 mm spacer designed for ITP as shown in Fig. 2. Start cooling and wait until the chamber has reached 10°C. 2. Fill the separation chamber at the inclined position with water and remove any trapped air bubbles, check for leakages. Bring the chamber to the horizontal position and perform a stripe test to insure laminar flow conditions. To do this feed inlets I2, I4, and I6 (Fig. 1b) with a 1:100 dilution of SPADNS in water and run media pump with 40 rpm. The SPADNSsolution should flow in parallel, even-colored stripes through the separation chamber. If this is not the case, clean and reassemble the system. 3. Change to separation media according to the following instructions (Note 6 and Fig. 1b): hang the inlet 1 into anodic stabilization medium, inlets 2–5 into leader ion medium, inlet 6 into spacer mix, inlet 7 into terminator ion medium, and inlet 8 into cathodic stabilization medium. Place counterflow tubings C1–3 into counterflow medium. Fill the electrolyte anode and cathode circuit with liquid circuit (+ve) and liquid circuit (−ve), respectively. Equilibrate the chamber for at least two transition times. 4. After medium change, carry out a performance test with a pI marker to ensure that your chamber has reached a separation capacity of high quality. To do this adjust the medium flow to a velocity of 180 mL/h. Apply pI-mix, supplemented with bromphenole blue and acridine orange in a 1:10 dilution in spacer medium via the right sample inlet (S4) at a flow rate of 3,000 mL/h for 2.30 min and change for another 30 s to spacer medium without pI-mix. Then, stop the sample pump

d = 2.8 cm I7 d = 2.8 cm

I1

I2

I3

I4

I5

I6

I8

Fig. 2. Spacer used for the ITP experiments. To permit the use of eight media inlets, the spacer shows asymmetric excavations at the cathodic side. Note, that the distance from sample inlet I6–I8 exhibits the same dimension as to standard inlet excavations.


37

and reduce the flow rate of the separation medium to 80 mL/h. Apply a current of 1,200 V with Imax set to 45 mA and Pmax to 60 W. Separate the sample for 7 min while changing the flow direction every minute. Thereafter, switch of the current and pump the sample toward the chamber outlet at a flow rate of 280 mL/h for another 3 min. When colored drops are visible at the collecting tube outlets (~1.30 min), position a microtiter plate on the collector’s drawer, tap fractionation plate, and slide in the drawer under the fractionation tubes. The pI marker should separate into seven discrete components focused each in 1–3 wells of the microtiter plate (Fig. 3a). 5. Pellet fraction #A (Sect. 3.1, step 7) at a velocity of 20,000 × gav and resuspend the pellet in 4 mL HB (brain of one rat). Dilute the sample 1:3 in spacer mix and repeat the separation procedure as described for the performance test (Sect. 3.2, step 4). To visualize your migration front during separation, SPADNS can be added. The amount of 1% SPADNS added in microliter is calculated as followed: volume of 1% SPADNS  5 × sample volume applied [mL] / velocity of the sample pump [mL/H]. To monitor the separation profile, measure the optical density of the microtiter plate at 420 nm (Notes 7 and 8). A typical separation of a synaptosomal sample is shown in Fig. 3b–d. 3.3. Purification of Brain Mitochondria by Zonal Free-Flow Electrophoresis

1. Set up the FFE system according to manufacturers operating manual and perform a stripe test as described under Sect. 3.2, step 2. 2. Change to the separation media according to the following instructions (Fig. 1b): hang inlets I2–I6 into separation medium, I1 into anodic stabilization medium, I7 into cathodic stabilization medium. Place counterflow tubings C1–3 into counterflow medium. Fill the electrolyte anode and cathode circuit with liquid circuit (+ve) and liquid circuit (−ve), respectively. Equilibrate the chamber for at least two transition times (~20 min). 3. After medium change, carry out a performance test. For this purpose, switch on high voltage (850 V, Imax: 45 mA, Pmax: 60 W) and set the media pump at 300 mL/h. Dilute the BD Free-Flow Electrophoresis system pI-Marker 1:2 in separation medium and equilibrate with 0.8% HPMC stock solution and solid sucrose to reach a final concentration of 0.1% and 320 mM, respectively. Apply the mixture via the right sample inlet (S4) at a flow rate of 800 mL/h. Begin to collect samples, when colored drops are visible at the collecting tube outlets (5 min) as described under Sect. 3.2, step 5.

38

Islinger et al.

a 1.00 0.80 0.60 0.40 0.20 0.00 0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

Fraction number O.D. 420 nm

b

O.D. 515 nm

O.D. 595 nm

0.15

0.10

0.05

0.00 0

5

10

15

20

25

30

35

40

45

50

55

Fraction number

c

d

60

65

70

75

80

85

90

95


39

The pI mix should distribute into five distinguishable peaks (Fig. 4a). 4. Spin down fraction #B (Sect. 3.1, step 10) from the Optiprep gradient at a velocity of 20,000 × gav, 4°C for 15 min. Suspend the pellet with a glass rod by drop wise addition of separation buffer until you reach a final protein concentration of 1.2– 1.3 mg/mL. 5. Switch on high voltage and media flow as described for the performance test. Apply the organelle suspension via inlet S4 at a flow rate of 1,800 mL/h (Note 9). Begin to collect the sample approximately 5 min after the organelle suspension reached the separation chamber. First use a standard 96-well plate to monitor the separation profile at 420 nm, and then use a 4 mL 96-well plate for further collection, since the probe is significantly diluted with separation buffer during the run. 3.4. Analysis of the FFE-Separation Using Immunoblotting

1. For subsequent analysis, it is recommended to concentrate the samples after separation in both FFE modes. The particulate nature of the probes allows to enhance protein concentrations pelleting the samples by centrifugation at 20,000 × gav, 4°C, 15 min. 2. For SDS-PAGE suspend the pellets in 50 mL HB and determine protein concentration. Load equal protein amounts to SDS-gels (e.g., 5–10 mg/lane). For separation we run NuPAGE 4–12% acrylamide gels in an Xcell SureLock Mini Cell with a MES-buffer system at a constant voltage of 150 V. 3. Gels were blotted onto PVDF-membrane with a constant current of 2.75 mA/cm2 using the discontinuous semidryblotting system introduced by Kyhse-Andersen (33). Blots were blocked in blocking solution for 1 h and incubated with

Fig. 3. Separation of a synaptosome fraction using free flow-ITP. (a) Separation profile of the bromphenol blue, acridine orange, pI marker mix described in the text. Note the sharp focusing of the dyes in the zone of ITP near the anodic side of the instrument. By contrast, dyes migrating with lower electrophoretic mobility than the terminating ions separated in the mode of zone electrophoresis and are not as properly focused. (b) Separation profile of the synaptosome fraction after FF-ITP. Synaptosomes prepurified on a sucrose gradient where separated according to the instruction in the text section. Afterward the fractions 21–41 were pelleted by centrifugation and analyzed by SDS-PAGE and immunoblotting. (c) Protein pattern of the organellar membranes produced by carbonate stripping of the individual FFE fractions depicted in (b). Note slight but significant changes in the polypeptide pattern from anode toward cathode. (d) Identification of marker proteins of the main subcellular compartments by immunoblotting. According to the staining for cytochrome C oxidase, mitochondria display the highest electrophoretic mobility and can be found concentrated nearest to the anode. As indicated by the differential detection of synaptophysin and PSD 95, synaptosomes are further separated into particles with higher presynaptic and postsynaptic proportions. However, as indicated by the differing concentrations of synaptic receptors NMDA 2A and glycine, synaptosomes from synapses with different physiological properties seem to possess also varying electrophoretic mobilities.

40

Islinger et al.

a

O.D. 420 nm

0.12 0.09 0.06 0.03 0 0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

70

75

80

85

90

95

Fraction number 420nm

b

515nm

O.D.

0.09

0.06 FFE-1 0.03

FFE-2

FFE-3

0 0

5

10

15

20

25

30

35

40

45

50

55

60

65

Fraction number

c

d Activity of Succinate dehydrogenase before and after FFE Sample

Enzyme activity [U/mg]

Gradient fraction

0.62

FFE-1 (44-51)

2.67

FFE-2 (52-57)

1.73

FFE-3 (61-68)

1.02

Fig. 4. Purification of mitochondria from the brain using zone electrophoresis. (a) Separation profile of the pI-mix using the same electrophoretic conditions as in the experiments with brain tissue. Due to the nonfocusing separation conditions in zone FFE, the individual pI-markers disperse into relatively broad zones. (b) Separation profile of brain mitochondria prepurified by density gradient centrifugation. For further analysis, the FFE-fractions were pooled into the samples FFE-1 to FFE-3 (44–51, 52–57, 61–68) and pelleted by centrifugation. (c) Immunoblots demonstrating the further purification of mitochondria (identified by ATP-synthase a) from synaptosomal (synaptophysin) and microsomal (ERp29) contaminants in fraction FFE-1 and FFE-2. GF Gradient fraction; P 12,000 × g pellet. (d) Activity measurements for succinate dehydrogenase substantiate the purification indicated by immunoblotting. FFE-1 exhibits a ~4-fold higher activity than the initial mitochondria fraction from the Optiprep gradient.


41

the primary antibodies for 1 h at RT or overnight at 4°C. After washing in PBS/0.1% Tween 20/1% FCS correspondent secondary antibodies were applied again for 1 h at RT. Protein bands were visualized on Amersham™ Hyperfilm ECL using Amersham™ ECL Western blotting detection reagents (GE Healthcare, Buckinghamshire, UK). 3.5. Electron Microscopy

1. For electron microscopy, pool and pellet samples belonging to an individual peak as described above. Thereafter, overlay the pellet with 2% glutaraldehyde in 10 mM HEPES, 320 mM sucrose, pH 7.4, and incubate for 30 min at RT. 2. Remove fixative and mount pellets on 1.5% agarose. Carefully dislocate the pellet from the reaction tube and slice into pieces of approximately 1 mm3. 3. Wash the agarose cubes 3 times in 0.15 M Pipes and postfix for 5 min in 2% OsO4 (Note 6). Add an equal volume of a 300 mg/mL K4[Fe[CN]6]-solution and incubate in closed reaction tubes for another 60 min. 4. Wash with deionized water and dehydrate in an ascending series of 75, 85, 95, and 100% (v/v) ethanol. 5. Remove ethanol by 3× 15 min incubation in propylene oxide, and then incubate in a 1:2 solution of Epon 812 in propylene oxide for 30 min. Finally switch to pure Epon 812 and incubate overnight in the coldroom. 6. Fill into BEEM-capsules and let polymerize for 2 days at 60°C. 7. Cut ultrathin sections of 100–200 nm on an ultra microtome. Electron microscopical analyses of fractions separated by both protocols are shown in Fig. 5.

3.6. Measurement of Succinate Dehydrogenase Activity

1. Pool and pellet samples as described above and suspend in 50–100 mL HB. 2. Dilute the samples in TVBE-buffer to a concentration resulting in a reaction in the linear range of the color change (begin with 1:100). 3. For a 1.5 mL reaction volume, mix 250 mL reaction buffer and 250 mL sample-solution and incubate 30 min at 37°C. For the blank, replace sample by TVBE-buffer. 4. Terminate the reaction by adding 1 mL SDS-stop-solution. 5. Measure the extinction at a wavelength of 630 nm and calculate the enzyme activity according to the following formula: (ODsample - ODblank ) ´ 111.1 ´ (reaction volume / sample volume

´ reaction time) ´ dilution.

42

Islinger et al.


43

4. Notes 1. The total brain of one rat was used for each experiment. However, the separation scheme can be downscaled to purify organelles from more restricted brain areas. 2. For the preparation of all FFE-solutions, use fully deionized (resistivity of 18.2 megohm/cm) and degassed water. 3. It is recommended to use a 0.8% (w/w) HPMC stock solution to prepare the different media. For this purpose, fill 992 mL of H2O in a 2 L beaker with a wide orifice. Stir vigorously using a large magnetic stir bar. 8 g HPMC have to be added step-wise in small quantities during 5–10 min. Agitation of the solution should be continued for more than 10 h, preferably overnight. As HPMC-solutions are prone to bacterial growth, stock solutions should be consumed within 2–3 weeks. 4. Osmium tetroxide (OSO4) is highly poisonous even at low exposure levels. Inhalation at concentrations well below those at which a smell can be perceived can lead to pulmonary edema, and subsequent death. Moreover, osmium tetroxide can stain the cornea leading to irreversible blindness. Use only under a safety hood and wear appropriate protective clothing. 5. Anaesthetization of the animal, e.g., by ip injection of Nembutal or chloralhydrate must be carried out by a properly trained operator with the appropriate license. 6. Media temperature should not exceed the working in the separation chamber temperature by more than 15°C. To avoid degassing, the media temperatures should not be kept below the chamber temperature. It is recommended to use media with a temperature similar to room temperature. 7. The composition of the spacer mix may be varied to improve the purification of a specific organelle of interest. Concentrations of the individual spacers can be changed, e.g., to broaden the gap between to adjacent peaks or new spacer

Fig. 5. Electron microscopical analysis of the FFE-experiments. (a–f) Ultrastructural analysis of a synaptosomal fraction separated by FF-ITP: (a, d) Fraction 27 is characterized by a mixture of moderately sized mitochondria and presynaptic synaptosomes. (b, e) Fraction 29 of the FF-ITP separation comprises a mixture of sporadic giant mitochondria and a high number of resealed, presynaptic terminals. (c, f) Fraction 35 is virtually free of mitochondria. In addition to reduced numbers of presynaptic terminals, sheets of membranous structures can be frequently observed. According to the increased abundance of PSD 95 on immunoblots, these may represent postsynaptic membranes. (g–l) Ultrastructural analysis of brain mitochondria purified by zone electrophoresis: (g, j) Mitochondrial fraction after density gradient centrifugation. The sample is still contaminated with a high degree of vesicular and membranous material. Mitochondria can be found in a condensed state. (h, k) Fraction FFE-1 shows a high concentration of well-preserved mitochondria. (i, l) In fraction FFE-3 shows the bulk of vesicular material found also in the prefraction. Mitochondria are of much lower abundance than in FFE-1. The magnification bars to the right are representative to all the images in one line.

44

Islinger et al.

substances with different electrophoretic mobilities can be used to replace those applied in this separation. However, the summated molarity of all spacers in the mixture should not be changed. 8. Interval FF-ITP can be run with high reproducibility. Thus, the fractions of several individual runs can be pooled for a subsequent detailed biochemical, electron microscopical, or proteomic analysis. 9. Higher protein concentrations can lead to sample agglutination during the run. If protein aggregation still occurs, stop media pump and reverse the flow direction of the sample pump to aspirate the agglutinated material. Proceed with separation with a reduced flow rate of the sample pump.

Acknowledgments We thank Medea Krapp, Heribert Mohr, Ute Sukopp, and Inge Frommer for their excellent technical assistance. We are grateful to Prof. Alfred Völkl for carefully reading through this manuscript. References 1. Brunet S, Thibault P, Gagnon E, et al. Organelle proteomics: looking at less to see more. Trends Cell Biol 2003;13:629–38. 2. Au CE, Bell AW, Gilchrist A, et al. Organellar proteomics to create the cell map. Curr Opin Cell Biol 2007;19:376–85. 3. Tribl F, Meyer HE, Marcus, K. Analysis of organelles within the nervous system: impact on brain and organelle functions. Expert Rev Proteomics 2008;5:333–51. 4. Grant SG. The synapse proteome and phosphoproteome: a new paradigm for synapse biology. Biochem Soc Trans 2006;34:59–63. 5. Li, KW, Jimenez CR. Synapse proteomics: current status and quantitative applications. Expert Rev Proteomics 2008;5:353–60. 6. Venable JD, Wohlschlegel J, McClatchy DB, et al. Relative quantification of stable isotope labeled peptides using a linear ion trap-Orbitrap hybrid mass spectrometer. Anal Chem 2007; 79:3056–64. 7. Mello CF, Sultana R, Piroddi M, et al. Acrolein induces selective protein carbonylation in synaptosomes. Neuroscience 2007;147:674–9. 8. Ghijsen WE, Leenders AG, Lopes da Silva FH. Regulation of vesicle traffic and neurotransmitter release in isolated nerve terminals. Neurochem Res 2003;28:1443–52.

9. Bai F, Witzmann FA. Synaptosome proteomics. Subcell Biochem 2007;43:77–98. 10. Boyd-Kimball D, Castegna A, Sultana R, et al. Proteomic identification of proteins oxidized by Abeta(1–42) in synaptosomes: implications for Alzheimer’s disease. Brain Res 2005;1044:206–15. 11. Gillardon F, Rist W, Kussmaul L, et al. Proteomic and functional alterations in brain mitochondria from Tg2576 mice occur before amyloid plaque deposition. Proteomics 2007;7:605–16. 12. Barkla BJ, Vera-Estrella R, Pantoja O. Enhanced separation of membranes during free flow zonal electrophoresis in plants. Anal Chem 2007;79:5181–7. 13. Eubel H, Lee CP, Kuo J, et al. Free-flow electrophoresis for purification of plant mitochondria by surface charge. Plant J 2007;52:583–94. 14. Marsh M. Endosome and lysosome purification by free-flow electrophoresis. Methods Cell Biol 1989;31:319–34. 15. Marsh M, Kern H, Harms E, et al. Co-fractionation of BHK-21 cell endosomes and lysosomes by free-flow electrophoresis. Prog Clin Biol Res 1988;270:21–33. 16. Völkl A, Mohr H, Fahimi HD. Peroxisome subpopulations of the rat liver. Isolation by

Subcellular Fractionation of Brain Tissue Using Free-Flow Electrophoresis immune free flow electrophoresis. J Histochem Cytochem 1999;47:1111–8. 17. Zischka H, Larochette N, Hoffmann F, et al. Electrophoretic analysis of the mitochondrial outer membrane rupture induced by per meability transition. Anal Chem 2008;80: 5051–8. 18. Zischka H, Weber G, Weber PJ, et al. mproved proteome analysis of Saccharomyces cerevisiae mitochondria by free-flow electrophoresis. Proteomics 2003;3:906–16. 19. Islinger M, Li KW, Loos M, et al. Peroxisomes from the heavy mitochondrial fraction: isolation by zonal free flow electrophoresis and quantitative mass spectrometrical characterization. J Proteome Res 2010;9:113–24. 20. Malmstrom J, Lee H, Nesvizhskii AI, et al. Optimized peptide separation and identification for mass spectrometry based proteomics via free-flow electrophoresis. J Proteome Res 2006;5:2241–9. 21. Moritz, RL, Simpson RJ. Liquid-based free-flow electrophoresis-reversed-phase HPLC: a proteomic tool. Nat Methods 2005;2:863–73. 22. Völkl A, Mohr H, Weber G, et al. Isolation of rat hepatic peroxisomes by means of immune free flow electrophoresis. Electrophoresis 1997;18:774–80. 23. Weber G, Bauer J. Counterbalancing hydrodynamic sample distortion effects increases resolution of free-flow zone electrophoresis. Electrophoresis 1998;19:1104–9. 24. Weber G, Bocek P. Stability of continuous flow electrophoresis. Electrophoresis 1998;19: 3094–5. 25. Weber G, Bocek P. Interval isotachophoresis for purification and isolation of ionogenic species. Electrophoresis 1998;19:3090–3.

45

26. Weber G, Islinger M, Weber P, et al. Efficient separation and analysis of peroxisomal membrane proteins using free-flow isoelectric focusing. Electrophoresis 2004;25:1735–47. 27. Cutillas PR, Biber J, Marks J, et al. Proteomic analysis of plasma membrane vesicles isolated from the rat renal cortex. Proteomics 2005;5:101–12. 28. Drews O, Wildgruber R, Zong C, et al. Mammalian proteasome subpopulations with distinct molecular compositions and proteolytic activities. Mol Cell Proteomics 2007;6: 2021–31. 29. Eubel H, Meyer EH, Taylor NL, et al. Novel proteins, putative membrane transporters, and an integrated metabolic network are revealed by quantitative proteomic analysis of Arabidopsis cell culture peroxisomes. Plant Physiol 2008;148:1809–29. 30. Huang S, Taylor NL, Narsai R, et al. Experimental analysis of the rice mitochondrial proteome, its biogenesis, and heterogeneity. Plant Physiol 2009;149:719–34. 31. Zischka H, Braun RJ, Marantidis EP, et al. Differential analysis of Saccharomyces cerevisiae mitochondria by free flow electrophoresis. Mol Cell Proteomics 2006;5:2185–200. 32. Pfeiffer F, Simler R, Grenningloh G, et al. Monoclonal antibodies and peptide mapping reveal structural similarities between the subunits of the glycine receptor of rat spinal cord. Proc Natl Acad Sci USA 1984;81: 7224–7. 33. Kyhse-Andersen J. Electroblotting of multiple gels: a simple apparatus without buffer tank for rapid transfer of proteins from polyacrylamide to nitrocellulose. J Biochem Biophys Methods 1984;10:203–9.

wwwwwwwwwwwwwwwwww

Chapter 4 Isolation of Synapse Subdomains by Subcellular Fractionation Using Sucrose Density Gradient Centrifugation Tatsuo Suzuki Abstract A protocol presents a purification of postsynaptic density (PSD), from rat brain by subcellular fractionation using solubilization of membrane with Triton X-100 and sucrose density centrifugation. The protocol also includes purification of other synapse subdomains such as synaptosome, synaptic plasma membrane, P1 (nuclei and cell debris), P2 (crude mitochondria fraction), S3 (soluble fraction), and P3 (microsomal fraction). The method presented in this text is the one established by Siekevitz group. The PSDs obtained by this method are mainly excitatory type I PSDs. The method has been widely used and is useful for biochemical analyses such as identification of proteins associated with these subdomains by proteomics methods and western blotting, and morphological analyses at the electron microscopic level. Key words: Synaptosome, Synaptic plasma membrane, Postsynaptic density, Subcellular fractionation, Detergent-insoluble cytoskeleton, Detergent-insoluble membrane

1. Introduction Isolation of subcellular compartment is a useful approach to analyze the subcellular complexes at the molecular level. We describe methods to isolate synapse subdomains including P1 (nuclei and cell debris), P2 (crude mitochondria fraction), synaptosome, synaptic plasma membrane (SPM), and postsynaptic density (PSD) by subcellular fractionation using density gradient centrifugation. History of development of method for isolation of synaptic complex and PSD is concisely summarized previously (1). The method for PSD purification established by Siekevitz’s laboratory (2–4) has been widely used. The methods introduced in this text are basically the same as those used in his laboratory. Both short and long procedures are stated. The method is


47

48

Suzuki

a pplicable to brains from at least dog, rat, mouse, and human (2, 5, 6), different brain regions (4, 7–10) and brains in various developmental stages (11). Protein yields of synaptosome, SPM, and PSD (short and long procedures, respectively) are approximately 15, 6, and 0.26 and 0.1 mg per 1 g original forebrain of adult rat, respectively, but may fluctuate by unknown factor(s). Protein yields of these fractions change depending on the age of the animals used (11). The short procedure, in which PSD is purified from Triton X-100 (TX-100)-treated synaptosomes, has been widely used and is now a standard method. Protein profiles of the PSD isolated by short and long procedures in one-dimensional gel are similar but not identical (Fig. 1) (2, 12). Contents of neurofilament proteins, at least partly contaminants (12), are higher in the PSD prepared by short procedure (12). Protein yield is also different (12). Major constituent proteins are the same between the two preparations, while the mass spectrometric (MS) analysis revealed that only 70% proteins in these two PSD fractions are common (Suzuki, unpublished data). Purified PSD fraction also contains mRNAs encoding various kinds of proteins (13).

Fig. 1. Protein profiles of postsynaptic density (PSD) fractions purified by short and long procedures. PSD fractions were purified from rat forebrains (Wistar male, 6 weeks old) and separated by 7–17% gradient polyacrylamide gel. s-PSD and m-PSD refer to PSD fractions prepared by short and long procedures via TX-100 treatment of synaptosome and synaptic plasma membrane (SPM), respectively. Molecular weights are shown in kDa on the left.

Isolation of Synapse Subdomains by Subcellular Fractionation

49

In the early methods to prepare synaptic junctional complex and PSD, p-iodonitrotetrazolium violet (INT) was used to separate mitochondria by producing heavy formazan in mitochondria (14–16). However, it was found that INT causes undesirable oxidation of proteins and artificially cross-links synaptic proteins (16–19). Structures of the isolated PSD are tightened by disulfide bonds formed during the PSD isolation using INT. It is suggested that the artificial disulfide bonding of PSD proteins during isolation may occur even in the absence of INT (1, 20, 21). Artificial cross-linking of postsynaptic proteins during isolation gives resistance of the isolated PSD to various treatments including detergent solubilization (1, 20, 21). Blocking of disulfide formation is required for preparing PSD for analyses of its structural and physiological properties. It is desirable to prepare synaptic subcompartments from freshly dissected brains. PSD fraction can also be prepared from frozen brains (3), which is convenient, in particular, when purifying it from human specimens. However, special attention should be paid when collecting brain tissue, because some proteins, in particular Ca2+/calmodulin-dependent protein kinase II (CaMKII), accumulate to PSD in a short duration after decapitation (22). Accumulation of CaMKII is accelerated at room temperature or 37°C. Tubulin also accumulates to PSD fraction in a relatively long time period at 4°C after decapitation (23). Attention should also be paid to “cold-induced exodus of postsynaptic proteins” (24). Exposure of neuron to coldness causes rapid disassembly of unstable microtubules that are present in the spine and associated with PSD. Various proteins also exit from PSD, and spine morphology, at least some, may change by this microtubule disassembly. The method stated in this text is useful to prepare the fraction enriched in the PSDs of asymmetric type I excitatory synapses, but not of the inhibitory neurons, such as those in the cerebellum. Protein yield of cerebellar PSD fluctuates and sometimes very low by unknown reason. Preparation of type II inhibitory PSD has been reported recently (25). Method using sonication but not detergent has also been reported (26), but up to now the method has been reported only once to the best of author’s knowledge. “One-Triton” PSD and “Two-Triton” PSD are prepared as a pellet after centrifugation of TX-100-treated synaptosomes (27–29). “One-Triton” PSD contains detergent-resistant membrane with light buoyant density (DRM, membrane raft fraction), which is also TX-100 insoluble at 4°C and floats on the 1.0 M sucrose layer (25, 30). Recently, it is demonstrated that “One-Triton PSD” also contains type II GABAergic inhibitory PSD (25). Nonionic detergent TX-100 is usually used to purify PSD fraction. High quality TX-100 should be used. Other detergent,

50

Suzuki

such as deoxycholate (31, 32), octyl glucoside (1, 33), and N-lauroyl sarcosinate (NLS) (16, 27, 32), has also been used. NLS, a strong ionic detergent, nearly completely solubilizes PSD components when oxidation is prevented with 1 mM N-ethylmaleimide (NEM) during isolation of PSD (1). Deoxycholate-insoluble PSD shows clearly a lattice-like core PSD structure (32, 34), which is broken after NLS treatment (32). Octyl glucoside is effective to solubilize rapidly the whole membrane, both raft and nonraft domains (35, 36), and generally does not affect protein–protein interactions. Presynaptic structure is unstable in alkaline solution, while postsynaptic structures are resistant (37). Therefore, the synaptic junctional structures composing of both pre and postsynaptic cytoskeletal structures can be prepared when synaptosome is solubilized with TX-100 at slightly acidic conditions (37).

2. Materials Use distilled, double-distilled, distilled-and-deionized, or equivalent grade water. Using ultrapure water sometimes results in lowprotein yield of PSD (see Note 1). All chemicals should be of reagent grade. All stocks and working solutions are kept at −20 to −30°C between uses to prevent bacterial and fungal growth. Make sure to mix up the solutions homogeneously after defreezing them, especially those containing dense sucrose solutions. All solutions should be kept at 4°C or on ice during the subfractionation. TX-100 is susceptible to autoxidation upon exposure to air. Store the unused solution sealed and also avoid storage in direct light (see Note 2). PSD material is extremely sticky to glass and cellulose nitrate and tend to aggregate very easily (2). Therefore, the use of plastic (polyallomer) tubes and pipettes, in particular, after TX-100 treatment, is necessary to avoid undesirable absorption of PSD proteins to glasses. Add protease inhibitors, phosphatase inhibitors, oxidization inhibitors, or RNase inhibitors as required. Addition of protease inhibitors results in increased yield of PSD proteins. It is desirable to purify PSD in the presence of iodoacetamide or NEM, which prevents harmful oxidation during the purification (1, 20, 21). PSDs prepared in the presence of iodoacetamide are different from those prepared in the absence of iodoacetamide in detergent solubility and aggregation state of PSD. Addition of dithiothreitol interferes with endogenous disulfide bondages necessary for the formation of normal PSD configuration (20, 21).


2.1. Preparation of P1, P2, Synaptosome, and PSD Fraction (Short Procedure)

51

1. 1 M MgCl2 stock. Dissolve 20.33 g of MgCl2·6H2O (MW, 203.30) in 100 mL H2O. 2. 1 M CaCl2 stock. Dissolve 14.70 g of CaCl2·2H2O (MW, 147.02) in 100 mL H2O. 3. 100 mM NaHCO3 stock. Dissolve 1.68 g of NaHCO3 in H2O and make up to 200 mL with H2O. 4. 1 M Tris–HCl (pH 8.1) stock. Dissolve 24.2 g of Tris(hydroxymethyl)aminomethane in H2O (~150 mL) and adjust pH to 8.1 by HCl, and make up to 200 mL with H2O. 5. 0.5 M HEPES/KOH (pH 7.4) stock. Dissolve 11.8 g of HEPES, adjust pH to 7.4 with KOH solution, and make up to 100 mL with H2O. 6. Solution A (0.32 M sucrose, 1 mM MgCl2, 0.5 mM CaCl2, 1 mM NaHCO3). Dissolve 109.6 g of sucrose in H2O. Add 10 mL of 100 mM NaHCO3, 1 mL of 1 M MgCl2, and 0.5 mL of 1 M CaCl2. Make up to 1,000 mL with H2O. 7. Solution B (0.32 M sucrose, 1 mM NaHCO3). Dissolve 109.6 g of sucrose in H2O. Add 10 mL of 100 mM NaHCO3. Make up to 1,000 mL with H2O. 8. 1% TX-100, 0.32 M sucrose, 12 mM Tris–HCl (pH 8.1). Dissolve 109.6 g of sucrose in H2O. Add 10 g of TX-100 (Sigma) and 12 mL of 1 M Tris–HCl (pH 8.1). Make up to 1,000 mL with H2O. 9. 1% TX-100, 150 mM KCl. Dissolve 2 g of TX-100 and 2.26 g of KCl in H2O. Make up to 200 mL with H2O. 10. 10 mM HEPES/KOH (pH 7.4)-40% glycerol. Dilute 4 mL of 0.5 M HEPES/KOH (pH 7.4) in H2O (~80 mL). Add 80 g of glycerol and make up to 200 mL with H2O. 11. Sucrose solution (1.0, 1.4, 1.5, and 2.1 M). Dissolve sucrose (68.5, 95.8, 102.7, and 143.8 g for 1.0, 1.4, 1.5, 2.1 M sucrose solutions, respectively) in H2O. Add 2 mL of 100 mM NaHCO3 to each solution and make up to 200 mL with H2O. 12. 1 mM NaHCO3. Dilute 100 mM NaHCO3 into H2O. 400 mL/20 g of starting brain is required. 13. Plastic disposable pipettes, e.g., Liquipette, polyethylene transfer pipettes of 4 mL capacity, thin stem, 7 mL capacity, with scale, and 6 mL capacity 9″ long (Elkay, Shrewsbury, MA), or other plastic Pasteur pipette such as 3 mL (with scale) etc.

52

Suzuki

2.2. Preparation of SPM and PSD Fraction (Long Procedure)

The long procedure requires solutions used in Sect. 2.1 and additional solutions listed below. 1. Sucrose solutions (0.85, 1.0, and 1.2 M). Dissolve sucrose (58.2, 68.5, and 82.2 g for 0.85, 1.0, and 1.2 M sucrose solutions, respectively) in H2O, add 2 mL of 100 mM NaHCO3 and make up to 200 mL with H2O. 2. 0.5 mM HEPES/KOH (pH 7.4). Dilute 0.5 M stock in H2O. About 250–500 mL/20–25 g brain is required for SPM preparation. 3. 1 mM NaHCO3. Dilute 100 mM NaHCO3 into H2O. About 150 mL/20–25 g of starting brain is required.

2.3. Preparation of S3 and P3 Fraction

No additional reagent or solution is necessary.

3. Methods 3.1. Preparation of P1, P2, Synaptosome, and PSD Fraction (Short Procedure) from Rat Forebrain

The method is based on those developed by Siekevitz’s group (2–4). Protocol for PSD purification (short procedure) using 20–25 g forebrain as starting material is described below. The maximum amount of forebrains is about 25 g due to the limitation of capacity of ultracentrifuge. All the processes are carried out at 4°C. The procedure is outlined in Fig. 2. 1. Collect rat forebrains by decapitation and quick dissection (see Note 3). Place forebrains immediately after dissection in a beaker placed on ice. Weigh the pooled brains (weight of the container is better measured before pooling tissues). Proceed for step 2 or freeze and keep the forebrains at −80°C until use. 2. Chop forebrains into small pieces (about less than 2 × 2 × 2 mm) with scissors. When using frozen brains, dip frozen brains into small amount of cooled solution A (approximately a few milliliter) in a beaker, chop or scrape them by scissors. Add solution A to make 80 mL suspension. Keep the suspension on ice for at least 20 min when using frozen brains (see Note 4). 3. Homogenize the suspension at 1,000 rpm with six or seven up-and-down motions with a motor-operated Teflon/glass homogenizer using a loose-fitting pestle (see Note 5) while cooling the container in ice water. Recover suspension into a new beaker and dilute to 200 mL with solution A. (Start preparing sucrose layers necessary at step 9 during centrifugations at steps 3 (or 4) to 8. To make sucrose layers, auto-pipette using 25 mL transfer pipette is convenient.)


53

Fig. 2. Purification of synaptosome and PSD by subcellular fractionation using sucrose density gradient centrifugation (PSD purification by short procedure). Examples of centrifugation conditions (rotors and speed) are indicated. Rotors marked with ‡1, ‡2, and ‡3 are those for centrifuges of Avanti J-25 (Beckman), L5-50 (Beckman), and himac CP60E (Hitachi), respectively. Steps using ultracentrifuge are numbered with roman characters and surrounded with parenthesis. Volumes and number of centrifuge tubes used are those for purifying PSD from 20 to 25 g forebrains of rats. *, **, and *** Positions where synaptosome and PSD before and after TX-100/KCl treatment, respectively, are collected. Cf Centrifugation; UCf ultracentrifugation; av average; sol solution; vol volume.

54

Suzuki

4. Centrifuge at 1,400 × gav or 1,475 × gmax for 10 min (JA18; 3,160 rpm, 4 tubes). Save supernatant in a beaker placed on ice or at 4°C. 5. Dilute the pellet with solution A and make 80 mL suspension. Homogenize with three strokes as in step 3. Dilute with solution A to 200 mL. 6. Centrifuge at 710 × gav or 755 × gmax for 10 min (JA18; 2,260 rpm, 4 tubes). Collect supernatant. Pellet is P1. 7. Combine supernatants obtained in steps 4 and 6 and centrifuge at 710 × gav or 755 × gmax for 10 min (JA18; 2,260 rpm, 6 tubes) (see Note 6). 8. Collect supernatant (S1) and centrifuge at 13,800 × gav or 17,300 × gmax for 10 min (JA18; 10,820 rpm, 6 tubes). Supernatant and pellet obtained in this step are S2 and P2, respectively. 9. Resuspend the P2 and gently hand homogenize with a Dounce homogenizer or Teflon-glass homogenizer in solution B (48 mL). Layer the suspension on gradients composed of 1.0 and 1.4 M sucrose, and centrifuge at 82,500 × gav for 70 min (SW 28, 25,000 rpm, 4–6 tubes) (see Note 7). 10. Collect the bands in the interface between 1.0 and 1.4 M sucrose layer (synaptosome fraction) (see Note 8) into a small beaker with a plastic pipette of 4 mL capacity with thin stem (see Note 9). Measure the volume of the synaptosome suspension, if necessary. Protein concentration of synaptosome just after recovered from the interface band is approximately 5 mg protein/mL. Save aliquot of synaptosome suspension after dilution to make about 2.5 mg/mL (just an example), if necessary. 11. Pour 1 mM NaHCO3 into a large beaker to make the final volume after mixing of the synaptosome suspension 400 mL/20 g starting forebrains (see Note 10). Place a stirrer bar into the beaker. Add synaptosome suspension dropwise into 1 mM NaHCO3 in a beaker while stirring. Continue to stir for about 20–40 min at 4°C (see Note 11). 12. Add slowly 400 mL/20 g of starting forebrains of 1% TX-100/0.32 M sucrose/12 mM Tris–HCl (pH 8.1) (final 0.5% TX-100, 0.16 M sucrose, 6 mM Tris–HCl) with constant stirring. Take 1 min to add the TX-100 solution. Continue to stir at 4°C. Total time of treatment with TX-100 (from starting addition of TX-100 to starting next centrifugation) should be 15 min. Therefore, transfer the solution to the transparent centrifuge tubes (see Note 12) at about 12 min after starting addition of TX-100 (see Note 13).


55

13. Centrifuge at 32,800 × gav or 48,200 × gmax for 20 min (JA14; 14,000 rpm × 45 min, four 250 mL tubes). (Prepare sucrose layers required at step 14 by using a plastic Pasteur pipette with scale.) Discard upper large portion of supernatant by slow decantation. Discard supernatant using a plastic pipette of 6 mL with 9″-long so that about 2 mL supernatant remains in the tube. Be very careful not to disturb the pellet. Recover pellet with plastic pipette by peeling and aspirating the pellet as a mass. Collect the pellet as small a volume as possible. Resuspend the pellet in solution B. Gently hand homogenize the pellet with a Dounce homogenizer or loose Teflon-glass homogenizer. 14. Layer the solution on gradients composed of 1.0, 1.5, and 2.1 M sucrose, and centrifuge at 201,800 × gav for 120 min (SW40, 40,000 rpm or RPS40T; 25,000 rpm 315 min, 4 tubes) (see Notes 14 and 15). (In the latter case, next step begins next morning). 15. Recover PSD fraction (**) with a plastic pipette (4 mL with thin stem) into 15 mL plastic tube. Dilute with cold H2O to 4 mL/1 tube and mix homogeneously. Add equal volume [4 mL/tube] of 1% TX-100/150 mM KCl (final 0.5% TX-100, 75 mM KCl) and mix homogeneously. Stand for 60 min (see Note 16). 16. Layer the solution on gradients composed of 1.5 and 2.1 M sucrose, and centrifuge at 30,000 rpm (SW40) for 20 min (RPS40T; 30,000 rpm, 2 tube) (see Note 14). 17. Retrieve PSD fraction (***) with a plastic pipette (4 mL with thin stem) into 1.5 mL Eppendorf microfuge tubes. Dilute with more than an equal volume of cold H2O (see Note 17). Centrifuge at >10,000 × g for 20 min. (Swing rotor is favorable.) 18. Discard supernatant and weigh the PSD material. Add equal amount of 10 mM HEPES/KOH (pH 7.4)/40% glycerol and mix homogeneously (see Note 18). Divide into small aliquots and keep them in plastic tubes at −80°C until use. 3.2. Preparation of SPM and PSD Fraction (Long Procedure)

Protocol (long procedure) for PSD purification using 20–25 g forebrain as starting material is described below. All the processes are carried out at 4°C. The procedure is outlined in Fig. 3. Steps 6–11 are the same as steps 12–18 of short procedure except for volumes of the samples and the number of centrifuge tubes used. 1. Prepare synaptosome fraction following the protocol described in Sect. 3.1. 2. Pour 0.5 mM HEPES/KOH (pH 7.4) into a large beaker to make the final volume after mixing the synaptosome suspension 400 mL. Place a stirrer bar into the beaker. Add synaptosome suspension (~50 mL) dropwise into the HEPES/KOH

56

Suzuki

Fig. 3. Purification of SPM and PSD by subcellular fractionation using sucrose density gradient centrifugation (PSD purification by long procedure). Protocol to prepare synaptosome and steps after sucrose gradient ultracentrifugation (II) are the same as those shown in Fig. 2. Comments and abbreviations are the same as in Fig. 2.

buffer in a beaker while stirring. Continue to stir for about 45 min at 4°C. 3. Centrifuge at 32,800 × gav or 48,200 × gmax for 20 min (JA14; 14,000 rpm × 45 min). Collect pellet and resuspend in solution B. 4. Layer the suspension on gradients composed of 0.85, 1.0, and 1.2 M sucrose, and centrifuge at 82,500 × gav for 70 min (SW 28, 25,000 rpm). Use 4 tubes. 5. Collect materials in the 1.0–1.2 M sucrose interface (#). Volume of this suspension is usually 20–30 mL (*approximately 3–4 mg protein/mL). Save aliquot if necessary. Dilute SPM suspension with 1 mM NaHCO3 (final volume, 50–120 mL [average 80 mL/20 g forebrain]) (see Note 19). 6. Treat the SPM suspension by adding equal volume of TX-100 as stated in Sect. 3.1, step 12. 7. Centrifuge at 32,800 × gav or 48,200 × gmax for 20 min (JA14; 14,000 rpm × 45 min, 1 or 2 tubes). Collect pellet and resuspend in solution B as stated in Sect. 3.1, step 13.


57

8. Layer the solution on the top of the sucrose gradient and centrifuge at 201,800 × gav × 120 min (SW40, 40,000 rpm or RPS40T; 25,000 rpm 315 min, 1 tube with sample). 9. Recover PSD fraction (**) as stated in Sect. 3.1, step 15. 10. Centrifuge at 30,000 rpm (SW40) 20 min (RPS40T; 30,000 rpm, 1 tube) as stated in Sect. 3.1, step 16. 11. Retrieve PSD fraction (***), process, and save as stated in Sect. 3.1, steps 17 and 18. 3.3. Preparation of S3 and P3 Fraction

Centrifuge S2 material at 100,000 × g for 1 h. Supernatant and pellet obtained are S3 and P3 fractions, respectively.

4. Notes 1. Subtle changes in ionic strength and metal concentration may affect sedimentation of subcellular organelles and protein complexes. It is not necessary to use ultra pure water, such as nanopure or miliQ water, for this subfractionation, and the use of ultra pure water may sometimes result in low yield of synaptosome and PSDs. Some unidentified factor(s) affect on the sedimentation and/or are necessary for stabilization of PSD protein complex. 2. Commercial TX-100 has been found to contain impurity with oxidizing activity (38). 3. If brains are homogenized or rapidly frozen in liquid nitrogen within 30 s to 1 min after decapitation, content of CaMKII, both a and b, is very low in the PSD fraction (22). Neurofilament content is increased in such PSD fraction. 4. Defrozen and chopped brains should be kept in cooled solution A for at least 20 min to depolymerize actin cytoskeleton. Inadequate depolymerization causes unfavorable sedimen tation. 5. Literatures (1, 2) recommend loose homogenizer (e.g., Teflon-glass homogenizer with a clearance of 0.25 mm or Dounce homogenizer with a loose-fitting pestle) to preserve morphological integrity of PSD. However, 0.25 mm clearance homogenizer does not appear to be a must. 6. It is very difficult to separate clearly the supernatant and pellet from total brain homogenate by centrifugation at 755 × gmax. Therefore, the first centrifugation was carried out at 1,475 × gmax. Supernatant obtained in the first centrifugation and the second centrifugation at 755 × gmax is combined, centrifuged again at 755 × gmax, and thus, S1 fraction was obtained. Removing

58

Suzuki

755 × gmax pellet is important to minimize contamination of nuclear materials to synaptic fractions (39). Methods omitting this step (e.g., one step purification of synaptosome) cannot avoid large amounts of contamination of nuclear proteins. 7. The first sucrose gradient was originally composed of 0.85, 1.0, and 1.2 M sucrose (2, 4), but can be replaced by those composed of 1.0 and 1.4 M sucrose with equivalent result (3). 8. Use fresh unfrozen brain as starting tissues for functional analysis of synaptosome. It is required to incubate synaptosome suspension in normotonic buffer to bring the terminals to a physiological steady state (40). Synaptosomes recovered from the sucrose gradient and are not incubated in normotonic buffer are shrunken due to high osmotic pressure. 9. Using disposable plastic pipettes to collect synaptosome, SPM, and PSD enriched bands after sucrose gradient centrifugation is convenient. See also 2.1.13. Be careful not to warm the plastic pipette (it means protein sample) by holding it with warm hand or fingers with wide contact areas for long time. 10. Fixed volume (400 mL for 20 g starting tissue) of the synaptosome suspension just before the TX-100 treatment is based on the protein concentration (2) estimated by the A260 and A280 using nomogram (distributed by California Corporation for Biochemical Research, LA) based on the equation by Warburg and Christian (41). Dilute synaptosome solution by 40-times for measurement of A260 and A280 (This also applies to SPM solution). The protein concentration estimated by Warburg–Christian method is about fourfold of the value obtained by Lowry method using BSA as standard. (The values were 4.3 ± 1.2 [n = 7] and 3.3 ± 1.2 [n = 12] folds for synaptosome and SPM fractions, respectively.) Therefore, protein concentration of the synaptosome suspension in 400 mL/20 g original forebrain is approximately 1 mg protein/mL (not 4 mg protein/mL as written in the original paper). Volume should be changed when starting from other parts of the brain, such as cerebellum. 11. This process is required before treatment with TX-100 and important to obtain good yield of PSD proteins, although the reason is unknown. Omitting this process may bring low yields of PSD. 12. Use transparent centrifuge tube to see the pellet clearly with the naked eye. The pellet obtained is very soft and easy to disturb. It is required to collect the pellet in a small volume to load on the top layer of the next sucrose density gradient. 13. The duration of TX-100 treatment affects the recovery of PSD.


59

14. PSD is extremely sticky to glass and cellulose nitrate tubes (2). Use polyallomer centrifuge tubes (2) to prevent adherence of PSD to the tubes. 15. Keep temperature to be around 4°C during ultracentrifugation. Raise of temperature loses some enzyme activity. 16. Inadequate treatment at this step leaves membrane materials to the final PSD preparation. 17. Repeat wash once or twice if complete removal of TX-100 is required. 18. Glycerol (20–50%) should be added to prevent artificial aggregation of the PSD proteins during storage at −80°C. Again, PSD material is extremely sticky to glass and cellulose nitrate, and tend to aggregate very easily, in particular, after freezing and defreezing. 19. The volume was determined by protein concentration read off with the nomogram, which was described in the original method for the PSD purification (2). The volume (average ± SD) was 80.3 ± 25.7 mL/20 g forebrain (n = 13) (see also Note 10).

Acknowledgments The author learned the method of PSD purification in the Philip Siekevitz laboratory, Rockefeller University, New York. The author heartily thanks Dr. Philip Siekevitz and Marie LeDoux for their instruction. References 1. Somerville, R. A., Merz, P. A., and Carp, R. I. (1984) The effects of detergents on the composition of postsynaptic densities, J Neurochem 43, 184–191. 2. Cohen, R. S., Blomberg, F., Berzins, K., and Siekevitz, P. (1977) The structure of postsynaptic densities isolated from dog cerebral cortex. I. Overall morphology and protein composition, J Cell Biol 74, 181–203. 3. Wu, K., Carlin, R., and Siekevitz, P. (1986) Binding of L-[3H]glutamate to fresh or frozen synaptic membrane and postsynaptic density fractions isolated from cerebral cortex and cerebellum of fresh or frozen canine brain, J Neurochem 46, 831–841. 4. Carlin, R. K., Grab, D. J., Cohen, R. S., and Siekevitz, P. (1980) Isolation and characterization of postsynaptic densities from various

brain regions: enrichment of different types of postsynaptic densities, J Cell Biol 86, 831–845. 5. Kim, T. W., Wu, K., and Black, I. B. (1995) Deficiency of brain synaptic dystrophin in human Duchenne muscular dystrophy, Ann Neurol 38, 446–449. 6. Hahn, C. G., Banerjee, A., Macdonald, M. L., Cho, D. S., Kamins, J., Nie, Z., BorgmannWinter, K. E., Grosser, T., Pizarro, A., Ciccimaro, E., Arnold, S. E., Wang, H. Y., and Blair, I. A. (2009) The post-synaptic density of human postmortem brain tissues: an experimental study paradigm for neuropsychiatric illnesses, PLoS ONE 4, e5251. 7. Suzuki, T., Okumura-Noji, K., Tanaka, R., Ogura, A., Nakamura, K., Kudo, Y., and Tada, T. (1993) Characterization of protein kinase

60

Suzuki

C activities in postsynaptic density fractions prepared from cerebral cortex, hippocampus, and cerebellum, Brain Res 619, 69–75. 8. Kim, T. W., Wu, K., Xu, J. L., and Black, I. B. (1992) Detection of dystrophin in the postsynaptic density of rat brain and deficiency in a mouse model of Duchenne muscular dystrophy, Proc Natl Acad Sci USA 89, 11642–11644. 9. Wu, K., and Black, I. B. (1987) Regulation of molecular components of the synapse in the developing and adult rat superior cervical ganglion, Proc Natl Acad Sci USA 84, 8687–8691. 10. Wu, K., and Siekevitz, P. (1988) Neurochemical characteristics of a postsynaptic density fraction isolated from adult canine hippocampus, Brain Res 457, 98–112. 11. Suzuki, T., Mitake, S., Okumura-Noji, K., Shimizu, H., Tada, T., and Fujii, T. (1997) Excitable membranes and synaptic transmission: postsynaptic mechanisms. Localization of alpha-internexin in the postsynaptic density of the rat brain, Brain Res 765, 74–80. 12. Matus, A., Pehling, G., Ackermann, M., and Maeder, J. (1980) Brain postsynaptic densities: the relationship to glial and neuronal filaments, J Cell Biol 87, 346–359. 13. Suzuki, T., Tian, Q. B., Kuromitsu, J., Kawai, T., and Endo, S. (2007) Characterization of mRNA species that are associated with postsynaptic density fraction by gene chip microarray analysis, Neurosci Res 57, 61–85. 14. Cotman, C. W., and Taylor, D. (1972) Isolation and structural studies on synaptic complexes from rat brain, J Cell Biol 55, 696–711. 15. Nieto-Sampedro, M., Bussineau, C. M., and Cotman, C. W. (1981) Optimal concentration of iodonitrotetrazolium for the isolation of junctional fractions from rat brain, Neurochem Res 6, 307–320. 16. Cotman, C. W., Banker, G., Churchill, L., and Taylor, D. (1974) Isolation of postsynaptic densities from rat brain, J Cell Biol 63, 441–455. 17. Kelly, P. T., and Montgomery, P. R. (1982) Subcellular localization of the 52,000 molecular weight major postsynaptic density protein, Brain Res 233, 265–286. 18. Kelly, P. T., and Cotman, C. W. (1976) Intermolecular disulfide bonds at central nervous system synaptic junctions, Biochem Biophys Res Commun 73, 858–864. 19. Kelly, P. T., and Cotman, C. W. (1981) Developmental changes in morphology and molecular composition of isolated synaptic junctional structures, Brain Res 206, 251–257.

20. Lai, S. L., Chiang, S. F., Chen, I. T., Chow, W. Y., and Chang, Y. C. (1999) Interprotein disulfide bonds formed during isolation process tighten the structure of the postsynaptic density, J Neurochem 73, 2130–2138. 21. Sui, C. W., Chow, W. Y., and Chang, Y. C. (2000) Effects of disulfide bonds formed during isolation process on the structure of the postsynaptic density, Brain Res 873, 268–273. 22. Suzuki, T., Okumura-Noji, K., Tanaka, R., and Tada, T. (1994) Rapid translocation of cytosolic Ca2+/calmodulin-dependent protein kinase II into postsynaptic density after decapitation, J Neurochem 63, 1529–1537. 23. Carlin, R. K., Grab, D. J., and Siekevitz, P. (1982) Postmortem accumulation of tubulin in postsynaptic density preparations, J Neurochem 38, 94–100. 24. Cheng, H. H., Huang, Z. H., Lin, W. H., Chow, W. Y., and Chang, Y. C. (2009) Coldinduced exodus of postsynaptic proteins from dendritic spines, J Neurosci Res 87, 460–469. 25. Li, X., Serwanski, D. R., Miralles, C. P., Bahr, B. A., and De Blas, A. L. (2007) Two pools of Triton X-100-insoluble GABA(A) receptors are present in the brain, one associated to lipid rafts and another one to the post-synaptic GABAergic complex, J Neurochem 102, 1329–1345. 26. Ratner, N., and Mahler, H. (1983) Isolation of postsynaptic densities retaining their membrane attachment, Neuroscience 9, 631–644. 27. Cho, K. O., Hunt, C. A., and Kennedy, M. B. (1992) The rat brain postsynaptic density fraction contains a homolog of the Drosophila discs-large tumor suppressor protein, Neuron 9, 929–942. 28. Walikonis, R. S., Jensen, O. N., Mann, M., Provance, D. W., Jr., Mercer, J. A., and Kennedy, M. B. (2000) Identification of proteins in the postsynaptic density fraction by mass spectrometry, J Neurosci 20, 4069–4080. 29. Murphy, J. A., Jensen, O. N., and Walikonis, R. S. (2006) BRAG1, a Sec7 domain-containing protein, is a component of the postsynaptic density of excitatory synapses, Brain Res 1120, 35–45. 30. Suzuki, T. (2002) Lipid rafts at postsynaptic sites: distribution, function and linkage to postsynaptic density, Neurosci Res 44, 1–9. 31. Blomberg, F., Cohen, R. S., and Siekevitz, P. (1977) The structure of postsynaptic densities isolated from dog cerebral cortex. II. Characterization and arrangement of some of the major proteins within the structure, J Cell Biol 74, 204–225.

Isolation of Synapse Subdomains by Subcellular Fractionation 32. Matus, A. I., and Taff-Jones, D. H. (1978) Morphology and molecular composition of isolated postsynaptic junctional structures, Proc R Soc Lond B Biol Sci 203, 135–151. 33. Gurd, J. W., Gordon-Weeks, P., and Evans, W. H. (1982) Biochemical and morphological comparison of postsynaptic densities prepared from rat, hamster, and monkey brains by phase partitioning, J Neurochem 39, 1117–1124. 34. Matus, A. (1981) The postsynaptic density, Trends Neurosci 4, 51–53. 35. Garner, A. E., Smith, D. A., and Hooper, N. M. (2008) Visualization of detergent solubilization of membranes: implications for the isolation of rafts, Biophys J 94, 1326–1340. 36. Shogomori, H., and Brown, D. A. (2003) Use of detergents to study membrane rafts: the good, the bad, and the ugly, Biol Chem 384, 1259–1263. 37. Phillips, G. R., Huang, J. K., Wang, Y., Tanaka, H., Shapiro, L., Zhang, W., Shan, W. S., Arndt, K., Frank, M., Gordon, R. E., Gawinowicz, M.

61

A., Zhao, Y., and Colman, D. R. (2001) The presynaptic particle web: ultrastructure, composition, dissolution, and reconstitution, Neuron 32, 63–77. 38. Chang, H. W., and Bock, E. (1980) Pitfalls in the use of commercial nonionic detergents for the solubilization of integral membrane proteins: sulfhydryl oxidizing contaminants and their elimination, Anal Biochem 104, 112–117. 39. Adam, R. M., Yang, W., Di Vizio, D., Mukhopadhyay, N. K., and Steen, H. (2008) Rapid preparation of nuclei-depleted detergent-resistant membrane fractions suitable for proteomics analysis, BMC Cell Biol 9, 30. 40. Fried, R. C., and Blaustein, M. P. (1978) Retrieval and recycling of synaptic vesicle membrane in pinched-off nerve terminals (synaptosomes), J Cell Biol 78, 685–700. 41. Warburg, O., and Christian, W. (1941) Isolierung and Kristallisation des Garungsferment, Biochem Z 310, 384–421.

wwwwwwwwwwwwwwwwww

Chapter 5 Enrichment of Plasma Membranes from Small Brain Samples by Aqueous Polymer Two-Phase Systems Jens Schindler Abstract Plasma membrane (PM) proteins take center-stage in most of the fundamental processes of the nervous system. They impart specificity to the formation of neuronal circuits, and determine the mode of neurotransmission. To accomplish these tasks, they demonstrate spatially and temporally restricted expression and regulation. Consequently, molecular analysis has to focus on functionally or anatomically distinct areas of the nervous system in order to get physiological relevant insights into PM proteomes. This constriction results most often in minute amounts of sample. The low abundance of PM proteins when compared to other subcellular proteomes such as those from mitochondria and cell nuclei further aggravates their analysis. Finally, posttranslational modifications of PM residing proteins differ from those observed in the secretory, endocytic, and recycling pathways. Taken together, all these properties of PM proteins require a protocol, which fosters their selective enrichment from small samples at high yield and high purity. Here, I will present a protocol based on partitioning of subcellular membranes in aqueous polymer two-phase systems, which fulfills these criteria. In contrast to previous protocols, I introduced an adaptation to preserve phosphorylation sites. The protocol was previously shown to provide profound insights into physiological relevant PM proteomes. By doing so, the protocol furthermore represents an excellent starting point towards identification of novel drug targets or key regulatory mechanisms such as phosphorylation. Key words: Nervous system, Plasma membrane, Aqueous polymer two-phase system, Counter current distribution, Phosphoproteome

1. Introduction Proper function of the nervous system heavily relies on PM proteins. This important protein class encompasses all neurotransmitter receptors as well as many transporters and ion channels. Furthermore, PM proteins participate in axonal guidance, cellcell adhesion, and in the synaptic release and reuptake machinery. They are therefore key players in developmental processes and in neurotransmission. Their malfunction is often associated with Ka Wan Li (ed.), Neuroproteomics, Neuromethods, vol. 57, DOI 10.1007/978-1-61779-111-6_5, © Springer Science+Business Media, LLC 2011

63

64

Schindler

diseases such as epilepsy, sensory deficits, and mood disorders. In agreement with their fundamental role, PM proteins account for approximately 70% of all known drug targets (1). This also puts them into prime focus of pharmaceutical research. Plasma membrane proteins are differently distributed and regulated throughout the nervous system. This differential molecular build-up of neuronal population forms one of the bases which enable the nervous system to perform its various distinct tasks such as motor control, processing of sensory information, or memory formation. So far, however, in-depth information of the PM protein composition of discrete neuronal populations is scarce. Likewise, the posttranslational modifications involved in the tight regulation of these proteins are still largely unknown. In the area of highly sophisticated mass spectrometry tools for proteomic studies, this lag in knowledge is mainly caused by the lack of appropriate protocols to enrich this category of proteins. PM proteins are low abundant with respect to the entire cellular proteome. They represent only a faint amount of 0.4–2.5% of the protein content of a cell (2). This renders the identification of PM proteins and their posttranslational modifications extremely difficult in the presence of contaminating compartments. To scope with this challenge, protocols are required, which foster the separation of PM from intracellular membranes with high yield and high purity. In the 1950, P.A. Albertsson published the separation of biological samples using aqueous polymer two-phase systems (3–5). This separation technique exploits differences in surface properties of biological membranes, which are mainly caused by distinct lipid compositions. These differences translate into different partitioning of membranes in one of the two phases under appropriate conditions (6, 7). In aqueous polymer two-phase systems which consist of polyethylene glycol (PEG) and dextran, PM tend to partition into the PEG-enriched top phase. All other membranes display a considerably lower affinity for the PEG-phase. This difference is exploited to selectively enrich PM at high yield with low contaminations by intracellular membranes (for review see (8, 9)). Due to their high content of water, aqueous polymer two-phase systems also preserve protein structure and function. Here, I report a protocol for the enrichment of plasma membranes from small brain samples which is based on a protocol published by Schindler et al. (10). A distinct feature of this novel protocol is its preservation of phosphorylated sites, due to the adaptation of the two-phase system to phosphatase inhibitors. In contrast to previous protocols, it therefore enables in-depth insight into phosphorylation, a key regulatory modification of great interest to many neuroscientists.

65

Enrichment of Plasma Membranes from Small Brain

2. Materials All solutions should be prepared with double distilled water. 1. Glass-teflon homogenizer. 2. Dextran: 20% (w/w) Dextran T500 (see Note 1). 3. PEG: 40% (w/w) PEG 3350. 4. Tris/SO42−: 250 mM Tris, pH 7.8, adjusted with H2SO4. 5. Protease/phosphatase Inhibitor: one tablet of each, complete mini and PhosStop (Roche Applied Sciences), in 1 mL of water (see Note 2). 6. IAA: 10% (w/v) iodoacetamide.

3. Methods 1. Prepare two two-phase systems labeled “A” and “B–J” as indicated in Table 1 without IAA and protease/phosphatase inhibitor. Vortex vigorously and store at 4°C over night. 2. Complete the two-phase systems with IAA and protease/ phosphatase inhibitor as indicated in Table 1. Vortex vigorously and centrifuge at 750×g for 5 min to accelerate phase separation. 3. Separate top phase and bottom phase of the two-phase system labeled with “B–J.” 4. Label nine tubes with “B,” “C,” …, “J.” Add 0.525 g of the bottom phase from step 3 to each tube.

Table 1 Composition of two-phase systems A (g)

B–J (g)

0.465

4.185

0.233

2.093

Tris/SO4

0.300

2.700

IAA

0.030

0.068

Protease/phosphatase inhibitor

0.300

2.700

Water

0.023

1.755

Dextran PEG 2−

66

Schindler

5. Add 0.150 g of brain tissue to two-phase system “A.” Homogenize the tissue in a glass Teflon homogenizer by 50 strokes with rotation at 500 rpm. 6. Aspirate the homogenate with a syringe through a 20 gauge needle for 15 times (see Note 3). 7. Centrifuge the sample at 750×g for 5 min to accelerate phase separation. 8. Remove the top phase from tube “A” and transfer it to tube “B” (Fig. 1, see Note 4). Add fresh top phase (step 3) to tube “A.” In the entire procedure, each newly added volume should equal that of the previously removed one. Mix both tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 9. Transfer the top phase from tube “B” to tube “C,” and the top phase from tube “A” to tube “B.” Add fresh top phase (step 3) to tube “A.” Mix the tubes by vigorous vortexing and centrifuge at 750×g for 5 min.

Fig. 1. Scheme of the plasma membrane enrichment procedure. The “step numbers” refer to the steps in the method section. Italic numbers indicate the order in which the top phases have to be transferred. (Step 8) Top phase “A” is transferred onto bottom phase “B.” Bottom phase “A” is reextracted with fresh top phase. (Step 9) Top phase “B” is transferred onto bottom phase “C” and top phase “A” is transferred onto bottom phase “B.” Bottom phase “A” is reextracted with fresh top phase. In the end (step 17), PM are enriched in top phases “I” and “J”.

Enrichment of Plasma Membranes from Small Brain

67

10. Transfer the top phase from tube “C” to tube “D,” the top phase from tube “B” to tube “C,” and the top phase from tube “A” to tube “B.” Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 11. Transfer the top phase from tube “D” to tube “E,” the top phase from tube “C” to tube “D,” and so forth. Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 12. Transfer the top phase from tube “E” to tube “F,” the top phase from tube “D” to tube “E,” and so forth. Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 13. Transfer the top phase from tube “F” to tube “G,” the top phase from tube “E” to tube “F,” and so forth. Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 14. Transfer the top phase from tube “G” to tube “H,” the top phase from tube “F” to tube “G,” and so forth. Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 15. Transfer the top phase from tube “H” to tube “I,” the top phase from tube “G” to tube “H,” and so forth. Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 16. Transfer the top phase from tube “I” to tube “J,” the top phase from tube “H” to tube “I” and so forth. Add fresh top phase (step 3) to tube “A.” Mix all tubes by vigorous vortexing and centrifuge at 750×g for 5 min. 17. Top phases of two phase systems “I” and “J” are highly enriched for PM. Transfer the top phases from “I” and “J” to a new tube. Dilute both top phases fivefold with 50 mM Tris/ SO42− and spin at 100,000×g for 1 h to recover the PM in the pellet. All other phases can be discarded.

4. Notes 1. Dextran can contain up to 10% of water. To assure the correct final concentration of dextran in the two-phase systems, dextran should be freeze-dried prior to usage. Freeze-dried dextran can be stored at −20°C in a tightly closed plastic tube wrapped with parafilm. Prior to opening, let it come to room temperature.

68

Schindler

2. The protocol is optimized for the use of complete mini (Roche Applied Sciences) and PhosStop (Roche Applied Sciences) as protease as well as phosphatase inhibitors. Inhibitors from other providers may need further adaptation of the protocol. 3. The better the homogenization, the higher the yield. 4. The interphase is always considered to be part of the bottom phase. References 1. Hopkins, A. L., and Groom, C. R. (2002) The druggable genome. Nat Rev Drug Discov. 1, 727–730 2. Evans, H. W., in: Rickwood, D. (Eds.), Preparative centrifugation, IRL Press, Oxford 1991, pp. 233–270 3. Albertsson, P. A. (1956) Chromatography and partition of cells and cell fragments. Nature 177, 771–774 4. Albertsson, P. A. (1958) Partition of proteins in liquid polymer-polymer two-phase systems. Nature 182, 709–711 5. Albertsson, P. A. (1958) Particle fractionation in liquid two-phase systems; the composition of some phase systems and the behaviour of some model particles in them; application to the isolation of cell walls from microorganisms. Biochim. Biophys. Acta 27, 378–395

6. Yamaji-Hasegawa, A., and Tsujimoto, M. (2006) Asymmetric distribution of phospholipids in biomembranes. Biol Pharm. Bull. 29, 1547–1553 7. Zachowski, A. (1993) Phospholipids in animal eukaryotic membranes: transverse asymmetry and movement. Biochem. J. 294, 1–14 8. Persson, A., and Jergil, B. (1995) The purification of membranes by affinity partitioning. FASEB Journal 9, 1304–1310 9. Schindler, J., and Nothwang, H. G. (2006) Aqueous polymer two-phase systems: Effective tools for plasma membrane proteomics. Proteomics 6, 5409–5417 10. Schindler, J., Lewandrowski, U., Sickmann, A., and Friauf, E. (2008) Aqueous polymer two-phase systems for the proteomic analysis of plasma membranes from minute brain samples. J Proteome Res 7, 432–442

Chapter 6 Identification and Characterization of Protein Complexes from Total Brain and Synaptoneurosomes: Heterogeneity of Molecular Complexes in Distinct Subcellular Domains Silvia De Rubeis and Claudia Bagni Abstract Neurons are highly polarized cells characterized by subcellular microdomains: the synapses. These compartments are specialized structures and are, for certain cellular pathways, independent from the cell body. To achieve such a functional specificity, including local mRNA translation, different molecular complexes are transported along the dendrites and locally regulated. Characterization of such a molecular diversity may help to elucidate neuronal functions as well as detect differences in neuronal dysfunctions. Here, we describe a method to specifically dissect a molecular complex according to the neuronal subcellular compartment. Specifically, the complexes are isolated by immunoprecipitation of the protein of interest from brain lysates or from purified synapses (synaptoneurosomes) and identified by mass spectrometry analysis. Key words: Immunoprecipitation, Synaptoneurosomes, Synaptic proteomic, Neurites, mRNP transport

1. Introduction 1.1. Subcellular Neuronal Compartments

Neurons are one of the most fascinating examples of polarized cells. The highly specialized morphology of each neuron is essential to create and maintain the neural circuits which underlie the proper functioning of the brain. Early during development, neurons acquire a polarity. Two types of neurites arise from the cell body, each with specific functions: a single axon, which routes the neuron’s output, and one or multiple dendrites, which integrate inputs from other neurons (1, 2). In order to achieve their functional differentiation, the two different neurites acquire an array of morphological and biochemical specificities, including changes in the cytoskeleton,


69

70

De Rubeis and Bagni

membrane, and secretory pathway components (3, 4). Besides the cell body, which ensures the basic metabolic functions, axons and dendrites possess subcellular compartments working as specialized domains: the synapses. In the cortical area of the brain, most of the excitatory synapses are formed by a bouton and a dendritic spine (5). The bouton is a small axonal varicosity representing the presynaptic terminal, while the spine arises from the dendritic shaft and represents the postsynaptic compartment (5). Presynaptic compartments are characterized by the presence of an active zone with the synaptic vesicles containing the neurotransmitters (6). The postsynaptic sites are defined by a local thickening beneath the membrane, the so-called postsynaptic density (PSD), which links the neurotransmitters receptors to signaling proteins and cytoskeleton (7). The synapses are highly dynamic structures which undergo morphological changes in response to activity, underlying forms of experience-driven plasticity (5). Such remodeling involves both the architecture and the biochemistry of the neuron, and it is thought to be crucial for sustaining and consolidating complex phenomena such as learning and memory. Therefore, some regulatory pathways such as protein synthesis are also present at synapses and are finely tuned with synaptic activity (8). Considering both the structural and molecular differentiation of the synapses, many studies have been aiming at the identification of the molecular complexes giving specificity to the synapses. Here, we describe a method to characterize molecular complexes present in the cytoplasm of a neuronal cell (from whole brain extracts) and in isolated synapses (from synaptoneurosomes extracts), by combining immunoprecipitation assays and mass spectrometry analysis. 1.2. Principle of Synaptoneurosomes Isolation

Synaptoneurosomes are subcellular particles that can be isolated from the brain tissue after homogenization in a buffer isoosmotic with the plasma. They appear as spherical or elongated structures coated by a membrane, which seals off the particle when the axonal termini are fractured during the homogenization. The particle mainly represents the presynapse (synaptosome), joined to a part or to the entire postsynapse (neurosome) (9). According to the homogenization procedure, it is possible to preserve mainly the presynaptic compartment or both pre- and postsynaptic compartments (9). A synaptoneurosome has a mean diameter of 1.6 mm (9, 10). Both pre- and postsynaptic compartments retain their morphological, functional, and biochemical identities. Electron microscopy studies on isolated synaptoneurosomes revealed that the most prominent feature of the presynaptic side is the active zone, composed by synaptic vesicles (11, 12). One or two mitochondria can also be observed, supplying the energy for the local metabolism (13). In the postsynaptic compartment, the above-mentioned

Identification and Characterization of Protein Complexes

71

PSD is well preserved. Moreover, free or membrane-associated polyribosomes beneath the PSD, as well as specific mRNAs targeted to the synapses, have been detected (14). Since the architecture of the synapse is maintained, the synaptoneurosomes are viable and metabolically active for few hours after isolation. In fact, in isotonic media, the particles have a high membrane potential and low intracellular calcium concentration. The local mitochondria can sustain the metabolism through ATP production (13, 15). Moreover, synaptoneurosomes can undergo calcium-dependent release of the neurotransmitters and endocytotic recycling as in living synapses (12, 16–18). Furthermore, the postsynapse possess functional neurotransmitter receptors and possibly intact signal transduction machinery. Synaptic stimulation with KCl or glutamate can activate local downstream events, such as protein synthesis (19–22). Finally, synaptoneurosomes are competent not only for basic neurotransmission but they can also sustain for some forms of glutamatergic synaptic plasticity (23). Therefore, synaptoneurosomes can be effectively used as a model for isolated synapses. Several methods have been employed to purify synaptoneurosomes from whole rodent brain (10, 20, 24). The traditional procedure exploits the separation of particles deriving from brain homogenate, through an isoosmotic density gradient (see Chap. 4) (24). The homogenate is loaded onto a sucrose gradient that is then centrifuged at high speed. During centrifugation, the particles sediment in a specific fraction along the gradient, according to the size and weight. The bottom fraction (P1) contains nuclei and cell debris, whereas the middle fraction (P2) includes myelin fragments, synaptoneurosomes and free mitochondria. On top of the gradient, the microsomes, ribosomes, and smaller entities form two distinct fractions (P3 and P4). The middle fraction (P2) is further separated to yield purified synaptoneurosomes. Here, we used a modified protocol based on an isoosmotic Percoll®/sucrose discontinuous gradient (25, 26). This method has several advantages, including reduced preparation time (about 1 h), maintenance of the isoosmolarity, and use of nontoxic material to obtain metabolically active synaptoneurosomes (25). The crude brain homogenate is separated upon centrifugation in a pellet, containing nuclear and cell debris, and a supernatant. The supernatant is further centrifuged, yielding a crude pellet that is resuspended in a Percoll®/sucrose solution in order to create a floating gradient upon centrifugation. The purified synaptoneurosomes can be recovered from the top layer of the gradient (26). 1.3. Principle of Protein Complex/es Immunoprecipitation

Immunoprecipitation (IP) is a method to selectively isolate a protein and its co-interacting partners from a tissue homogenate or from cell lysates (see Fig. 1). The assay exploits the specific binding

72

De Rubeis and Bagni

Fig. 1. Scheme to illustrate the immunoprecipitation of different complexes from distinct neuronal compartments. After brain homogenization (i.e., cortex as depicted in red in the left panel) or synaptoneurosomes isolation (right panel), the protein of interest (green) and its complex/es are immunoprecipitated with specific antibodies and then captured by the beads (in light blue) conjugated to either protein A or G (in pink). The molecular complexes are then eluted from the beads and analyzed by mass spectrometry.

occurring between the variable regions of an antibody and the antigen. The immune complex, composed by the antibody and the target protein, can be bound through the constant domain (Fc fragment) of the immunoglobulins to Protein A or G purified from Staphylococcus aureus and Streptococci bacteria, respectively (Fig. 1). Alternatively, a secondary antibody raised against the Fc region of the primary antibody can be used (not shown).


73

Proteins A/G or the secondary antibody are conjugated to a solid support, such as sepharose, agarose, or magnetic beads, which tightly capture the molecular complex/es. The beads can be recovered by centrifugation or magnet. The molecular complex/es captured by the beads can be subsequently eluted and analyzed by Western blot or by mass spectrometry (Fig. 1). Many parameters contribute to define the specificity/ efficiency of the immunoprecipitation; the most important one is the design of the antibody: 1. The antibody needs to bind a conformational epitope, i.e., exposed in the native tertiary structure of the protein. Therefore, antibodies raised against linear epitopes, such as small peptides, might not be able to recognize the entire protein in its folded structure. In this case, the linear epitope is masked in the native protein and becomes exposed only upon denaturation. 2. The epitope should be accessible, meaning that regions with a high probability of protein–protein interactions should not be chosen, since they will probably not be available to interact with the antibody. 3. The epitope should be specific for the protein of interest. This issue is very important when the protein is part of a family and shares high amino acid homology with the other family members (27).

2. Materials The materials to be used are listed below. Similar products from other suppliers can also be used. 2.1. Brain Extracts

1. Protease inhibitor cocktail (Sigma). It contains a mixture of protease inhibitors, namely 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF), pepstatinA, E-64, bestatin, leupeptin, and aprotinin. Such inhibitors cover a broad range of specificity (serine, cysteine, aspartic proteases and aminopeptidases). Aliquots can be stored at −20°C. 2. Na3VO4 and b-glycerophosphate or other phosphatases inhibitors. Aliquots can be stored at −20°C. 3. RNase OUT™ Ribonuclease (RNase) Inhibitor (Invitrogen). Store at −20°C. 4. 10 cm3 glass-Teflon® douncer.

2.2. Synaptoneu rosomes Isolation

1. Percoll™ (Amersham Biosciences). Store at 4°C for 3 months maximum. 2. 5 mL syringes.

74

De Rubeis and Bagni

2.3. Immunoprecipi tation

1. Antibodies against the target protein. 2. Protein A or protein G-sepharose™ beads (Amersham). The beads can be purchased as lyophilized powder or resuspended in 20% ethanol. The lyophilized powder should be extensively washed according to manufacturer’s indications and afterward resuspended in 20% ethanol. It can be kept at 4°C. 3. Trifluoroacetic acid (25%, Applied Biosystems). Store at 4°C.

3. Methods 3.1. Brain Extracts

1. After appropriate anesthesia, remove the mouse brain from the skull, and dissect it in PBS 1× on ice. 2. Homogenize the brain tissue in 3 mL/400 mg tissue of icecold extraction buffer (100 mM NaCl, 10 mM MgCl2, 10 mM Tris–HCl pH = 7.4, 1% Triton X-100, 1 mM DTT, 1× protease inhibitor cocktail, 0.5 mM Na3VO4, 5 mM b-glycerophosphate, and 40 U/mL RNase inhibitor). Homogenization is performed with a glass-Teflon® douncer (about 12 strokes, see Note 1). 3. Incubate the lysate 5 min on ice. 4. Centrifuge 8 min at 12,000×g at 4°C. 5. Carefully recover the supernatant and collect it. The extracts can be used immediately or stored for long time at −80°C upon quick freezing in liquid nitrogen.

3.2. Synaptoneurosomes Isolation

The protocol reported here is slightly modified from a previously described procedure (25, 26). 1. Isolate the brain and dissect the area of interest on ice (see Chap. 2). 2. Homogenize the tissue or six cortices in 10 mL of ice-cold homogenizing buffer (0.32 M sucrose, 1 mM EDTA, 1 mg/ mL BSA, 5 mM HEPES pH = 7.4) in a glass-Teflon® douncer (about ten strokes at 200–250 rpm on ice, see Note 2). 3. Centrifuge 10 min at 3,000×g at 4°C and recover the supernatant (see Note 3). 4. Centrifuge 12 min at 14,000×g at 4°C and discard the supernatant. 5. Resuspend pelleted synaptoneurosomes in 550 mL of KrebsRinger buffer (140 mM NaCl, 5 mM KCl, 5 mM glucose, 1 mM EDTA, 10 mM HEPES pH = 7.4) (see Note 4).


75

6. Transfer the synaptoneurosomes in Krebs-Ringer buffer in 1.5 mL tubes and add 450 mL of Percoll™ (final 45% v/v). Mix by gently inverting the tube. 7. Centrifuge 2 min at 14,000 rpm at 4°C to separate the flotation gradient. 8. The synaptoneurosomes are enriched on the top surface of the gradient, and they can be carefully recovered removing the underlying solution with a 5 mL syringe (see Note 5). 9. Gently resuspend the pellet in 1 mL of Krebs-Ringer buffer (see Note 4). 10. Spin 30 s at 14,000 rpm at 4°C and discard the supernatant. 11. Gently resuspend the pellet containing the synaptoneurosomes in 600 mL of Krebs-Ringer buffer (see Note 4). As before, the synaptoneurosomes can be immediately used or freeze in liquid nitrogen and subsequently stored at −80°C. 3.3. Immunoprecipi tation

The protocol reported below is slightly modified from (28). 1. Use 500 mg of brain extracts or 300 mg of purified synaptoneurosomes and add 5 mM HEPES pH = 7.4 to a final volume of 250 mL (see Note 6). 2. Add 250 mL of solubilization buffer (300 mM NaCl, 45 mM HEPES pH = 7.4, 2% Triton X-100) and keep the samples 1 h at 4°C on a wheel. 3. Centrifuge 20 min at 14,000 rpm at 4°C and recover the supernatant. Repeat this step one more time. 4. Recover the supernatant and incubate with the specific antibody or with purified immunoglobulins overnight at 4°C on a wheel (see Notes 6 and 7). 5. For each sample, wash 20 mL of protein A/G sepharose beads with 1 mL of washing buffer (25 mM HEPES pH = 7.4, 150 mM NaCl, 0.1% Triton X-100), centrifuge 1 min at 2,500 rpm at 4°C and remove the supernatant. Repeat three more times (see Note 8). 6. Add the samples + antibody complexes to the beads and incubate 1 h on a wheel at 4°C (Note 9). 7. Wash the beads four times with 1 mL of washing buffer and one time with 1 mL of 5 mM HEPES pH = 7.4. 8. Add 110 mL of elution buffer (0.2% trifluoroacetic acid) to the beads, shake 5 min, and recover the elute upon centrifugation for 2 min at 2,500 rpm at 4°C (see also Note 10). 9. Repeat step 8, mix the two elutions and dry the sample. The immunoprecipitated complexes are now ready for further analysis, for example by mass spectrometry (see Chap. 12) (see Note 11).

76

De Rubeis and Bagni

4. Notes 1. The volume of the homogenization buffer can be adjusted according to the size/weight of the tissue. 3 mL homogenization buffer is normally used for an entire brain from a 3 weeks-old mouse (around 400 mg). If specific brain regions will be used as starting material, use 0.75 mL of homogenizing buffer per 100 mg of tissue. 2. The volumes indicated apply to the entire cortex or the cerebellum from one mouse brain. If hippocampus or striatum is used, we suggest to homogenate the two hippocampi or two striata from a single mouse brain in 3 mL of buffer. In the latter case, the volumes for all the subsequent steps should be adjusted. 3. After centrifugation, the supernatant can be recovered by simply pouring it into a new tube, carefully avoiding the white phase. 4. Resuspend the pellet very gently to avoid synaptoneurosomes bursting. First, pour the appropriate volume of buffer along the wall of the tube without touching the pellet, then resuspend the solution by gently pipetting up and down, to avoid producing bubbles. 5. The separation of the Percoll gradient produces a floating, white material, which is enriched in synaptoneurosomes. To recover it, slide the needle of the syringe along the wall of the tube and slowly aspirate the solution. 6. The amount of antibody to use depends on the specificity and efficiency of the antibody to immunoprecipitating the protein of interest, as discussed in the introduction. Anti-serum (polyclonal antibodies) or hybridoma supernatant (monoclonal antibodies) can be used for immunoprecipitations. However, affinity-purified antibodies present many advantages, such as improved efficiency, low background (increased specificity), and reduced amount of initial material. It is important, before proceeding with the method described here, to set up the most efficient conditions for immunoprecipitation (see Note 9). 7. In order to exclude the possibility of pulling-down nonspecific protein complexes, perform in parallel a control immunoprecipitation using extracts from a transgenic mouse knocked out for the gene of interest or from cells knocked down for that gene. Other suitable controls are (1) IP with pre-immune serum from the same animal used to produce the immune serum and (2) IP with commercial purified immunoglobulins. The same amount of specific antibody and nonspecific serum/immunoglobulins should be used. While the concentration of purified immunoglobulins can be determined by a quantitative analysis such as a Bradford assay, the


77

antisera would require a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and subsequent Coomassie staining analysis. 8. Beads need to be washed prior to use in order to remove all the traces of the preservative (20% ethanol). In case of background, we suggest to saturate the nonspecific binding sites on the sepharose beads by washing three times with PBS 1× and incubating afterward the beads in PBS 1× BSA 0.1% for 1 h at 4°C on a wheel. 9. As mentioned in the introduction, either protein A or protein G-conjugated sepharose beads can be used. The relative affinity of such proteins for various antibody species and isotypes is reported in Table 1. Moreover, although secondary antibodies

Table 1 Relative affinity of immobilized Protein A and Protein G for different antibody species and subclasses of polyclonal and monoclonal antibodies. Modified from (29) Monoclonal

Protein G

Protein A

Human IgG1 IgG2 IgG3 IgG4

High High High High

High High – High

Mouse IgG1 IgG2 IgG21 IgG3

High High High High

Low High High Medium

Rat IgG1 IgG2 IgG21 IgG2 Human Mouse Rat Hamster Guinea pig Rabbit Horse Cow Pig Sheep Goat Chicken

Low High Medium Medium High Medium Medium Medium Medium High High High High Medium Medium Low

– – – Low High Medium Low Low High High Medium Medium High Low – –

78

De Rubeis and Bagni

can also be directly conjugated to the beads, we suggest the use of Protein A or G because of their high affinity to capture the antibody/antigen complex. 10. As mentioned in Note 6, the experimental conditions can be optimized, in order to solve two common problems, namely low yield and low specificity. In order to increase the yield of the immunoprecipitate, both amount of antibody and extracts can be increased, especially if the protein of interest is rare. If the recovery of the antigen/antibody complex is low, check the compatibility between the antibody and the protein A/G (see Note 8 and Table 1). In order to reduce the background due to nonspecific interactions, several steps can be optimized. First, a preclearing of the extracts can be performed before the incubation with the antibody. Briefly, the extracts are incubated with the beads for 1 h at 4°C, and then the recovered extracts are incubated with protein A/G conjugated with the specific antibody. Second, the time of incubation of the extracts with the antibodies can be reduced to 1 h. Third, the sepharose beads can be presaturated with PBS BSA 0.1%, as suggested in Note 8. 11. Proteins can also be dislodged from the beads with SDS-PAGE sample buffer, separated by 1D PAGE, and analyzed by immunoblotting (Chap. 13), or mass spectrometry (Chap. 12).

Acknowledgments This study was supported by Telethon, Compagnia di San Paolo, COFIN and FWO. References 1. Arimura, N., and Kaibuchi, K. (2007) Neuronal polarity: from extracellular signals to intracellular mechanisms. Nat Rev Neurosci 8, 194–205. 2. da Silva, J. S., and Dotti, C. G. (2002) Breaking the neuronal sphere: regulation of the actin cytoskeleton in neuritogenesis. Nat Rev Neurosci 3, 694–704. 3. Ye, B., Zhang, Y., Song, W., Younger, S. H., Jan, L. Y., and Jan, Y. N. (2007) Growing dendrites and axons differ in their reliance on the secretory pathway. Cell 130, 717–29. 4. Conde, C., and Caceres, A. (2009) Microtubule assembly, organization and dynamics in axons and dendrites. Nat Rev Neurosci 10, 319–32. 5. Holtmaat, A., and Svoboda, K. (2009) Experience-dependent structural synaptic

lasticity in the mammalian brain. Nat Rev p Neurosci 10, 647–58. 6. Ziv, N. E., and Garner, C. C. (2004) Cellular and molecular mechanisms of presynaptic assembly. Nat Rev Neurosci 5, 385–99. 7. Kennedy, M. B., Beale, H. C., Carlisle, H. J., and Washburn, L. R. (2005) Integration of biochemical signalling in spines. Nat Rev Neurosci 6, 423–34. 8. Steward, O., and Schuman, E. M. (2003) Compartmentalized synthesis and degradation of proteins in neurons. Neuron 40, 347–59. 9. De Rubeis, S., and Bagni, C. (2008) Synaptosome. Encyclopedia of Neuroscience Eds Binder MD, Hirokawa N, Windhorst U, Hirsch, MC.

Identification and Characterization of Protein Complexes 10. Hollingsworth, E. B., McNeal, E. T., Burton, J. L., Williams, R. J., Daly, J. W., and Creveling, C. R. (1985) Biochemical characterization of a filtered synaptoneurosome preparation from guinea pig cerebral cortex: cyclic adenosine 3’:5’monophosphate-generating systems, receptors, and enzymes. J Neurosci 5, 2240–53. 11. Verhage, M., McMahon, H. T., Ghijsen, W. E., Boomsma, F., Scholten, G., Wiegant, V. M., and Nicholls, D. G. (1991) Differential release of amino acids, neuropeptides, and catecholamines from isolated nerve terminals. Neuron 6, 517–24. 12. Takei, K., Mundigl, O., Daniell, L., and De Camilli, P. (1996) The synaptic vesicle cycle: a single vesicle budding step involving clathrin and dynamin. J Cell Biol 133, 1237–50. 13. Nicholls, D. G. (2003) Bioenergetics and transmitter release in the isolated nerve terminal. Neurochem Res 28, 1433–41. 14. Rao, A., and Steward, O. (1991) Evidence that protein constituents of postsynaptic membrane specializations are locally synthesized: analysis of proteins synthesized within synaptosomes. J Neurosci 11, 2881–95. 15. Choi, S. W., Gerencser, A. A., and Nicholls, D. G. (2009) Bioenergetic analysis of isolated cerebrocortical nerve terminals on a microgram scale: spare respiratory capacity and stochastic mitochondrial failure. J Neurochem 109, 1179–91. 16. Nicholls, D. G., and Sihra, T. S. (1986) Synaptosomes possess an exocytotic pool of glutamate. Nature 321, 772–3. 17. Anggono, V., Smillie, K. J., Graham, M. E., Valova, V. A., Cousin, M. A., and Robinson, P. J. (2006) Syndapin I is the phosphorylation-regulated dynamin I partner in synaptic vesicle endocytosis. Nat Neurosci 9, 752–60. 18. Serulle, Y., Sugimori, M., and Llinas, R. R. (2007) Imaging synaptosomal calcium concentration microdomains and vesicle fusion by using total internal reflection fluorescent microscopy. Proc Natl Acad Sci USA 104, 1697–702. 19. Scheetz, A. J., Nairn, A. C., and ConstantinePaton, M. (2000) NMDA receptor-mediated control of protein synthesis at developing synapses. Nat Neurosci 3, 211–6.

79

20. Bagni, C., Mannucci, L., Dotti, C. G., and Amaldi, F. (2000) Chemical stimulation of synaptosomes modulates alpha-Ca2+/ calmodulin-dependent protein kinase II mRNA association to polysomes. J Neurosci 20, RC76. 21. Takei, N., Inamura, N., Kawamura, M., Namba, H., Hara, K., Yonezawa, K., and Nawa, H. (2004) Brain-derived neurotrophic factor induces mammalian target of rapamycindependent local activation of translation machinery and protein synthesis in neuronal dendrites. J Neurosci 24, 9760–9. 22. Napoli, I., Mercaldo, V., Boyl, P. P., Eleuteri, B., Zalfa, F., De Rubeis, S., Di Marino, D., Mohr, E., Massimi, M., Falconi, M., Witke, W., Costa-Mattioli, M., Sonenberg, N., Achsel, T., and Bagni, C. (2008) The fragile X syndrome protein represses activity-dependent translation through CYFIP1, a new 4E-BP. Cell 134, 1042–54. 23. Corera, A. T., Doucet, G., and Fon, E. A. (2009) Long-term potentiation in isolated dendritic spines. PLoS One 4, e6021. 24. Gray, E. G., and Whittaker, V. P. (1962) The isolation of nerve endings from brain: an electron-microscopic study of cell fragments derived by homogenization and centrifugation. J Anat 96, 79–88. 25. Nagy, A., and Delgado-Escueta, A. V. (1984) Rapid preparation of synaptosomes from mammalian brain using nontoxic isoosmotic gradient material (Percoll). J Neurochem 43, 1114–23. 26. Lopes, L. V., Cunha, R. A., and Ribeiro, J. A. (1999) Cross talk between A(1) and A(2A) adenosine receptors in the hippocampus and cortex of young adult and old rats. J Neurophysiol 82, 3196–203. 27. Ferrari, F., Mercaldo, V., Piccoli, G., Sala, C., Cannata, S., Achsel, T., and Bagni, C. (2007) The fragile X mental retardation protein-RNP granules show an mGluR-dependent localization in the post-synaptic spines. Mol Cell Neurosci 34, 343–54. 28. Klemmer, P., Smit, A. B., and Li, K. W. (2009) Proteomics analysis of immuno-precipitated synaptic protein complexes. J Proteomics 72, 82–90. 29. Harlow, E., and Lane, D. E. (1988) Antibodies, A Laboratory Manual. Cold Spring Harb Lab, N.Y., 617–618.

wwwwwwwwwwwwwwwwww

Chapter 7 Analysis of Protein Complexes by 2D Blue Native/SDS–PAGE and Antibody-Shift Assay Dong Yang, Xinyu Deng, Ying Jiang, and Fuchu He Abstract Blue native polyacrylamide gel electrophoresis (BN-PAGE) is a powerful strategy for the separation of native multiprotein complex (MPC). Combined with other techniques, such as SDS–PAGE, immunoblotting, mass spectrometry, and antibody-shift assay, BN-PAGE has been widely used in MPC research. Here, we describe the materials and methods for the basic techniques mentioned above and provide the extensive practical advice and potential pitfalls of BN-PAGE. Key words: BN-PAGE, Multiprotein complex, Antibody-shift assay

1. Introduction Blue native polyacrylamide gel electrophoresis (BN-PAGE) is originally introduced by H. Schägger and G. von Jagow for the research of multiprotein complexes (MPC) in solubilized mitochondria or extracts of heart muscle tissue, lymphoblast, yeast, and bacteria (1). Its name “blue native” comes from the Coomassie Brilliant Blue (CBB) G-250 dye used in this system, which makes the separated slices colored in beautiful blue without further dyeing. BN-PAGE can be used for one-step isolation of protein complexes from biological membranes and total cell and tissue homogenates (2), with the resolution of which is much higher than that of other methods such as gel filtration or sucrose-gradient ultracentrifugation (1, 3). Since 1991, BN-PAGE has been used for the analysis of a variety of samples including plant and animal mitochondria (4, 5), whole cellular lysates (6, 7), plasma membrane (8–10), cytosol samples (11, 12), brain subcellular fractions (13, 14), and blood plasma (15). These studies showed


81

82

Yang et al.

that the resolving power of BN-PAGE can further be increased by pretreatment or fractionation of samples. BN-PAGE is a charge shift method, in which the electrophoretic mobility of an MPC is determined by the negative charge of the bound Coomassie dye and the size and shape of the complex (1, 3, 16). Anionic CBB is added to the supernatant and/or to the cathode buffer for BN-PAGE in order to bind to proteins in the sample and to modify their net charge and their solubility (17). Besides, the mild detergents and buffers used in the system can provide a native condition for protein complexes separation. Dodecyl-b-d-maltoside (DDM) and Triton X-100 are the most popular detergents used in BN-PAGE (3, 18, 19), and digitonin is regarded as the mildest detergent that is recommended recently (20, 21). 2-(bis(2-hydroxy-ethyl)amino)-2-(hydroxymethyl)-1,3propanediol (Bis–Tris) (1, 3) buffer helps to stabilize pH 7.0–7.5 in the native gel, and the zwitterionic compound 6-aminohexanoic acid in the buffer can improve membrane solubilization by neutral detergents (1). BN-PAGE can be coupled to other gel-based separation methods for multidimensional separation. The most commonly used technique consists of BN-PAGE and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) because it resolves subunits of the protein complexes efficiently (22–24). It is reported that the N-tris(hydroxymethyl)methyl-glycine (Tricine)-SDS–PAGE is preferred to other SDS electrophoresis systems. In this system the BN-PAGE protein bands are concentrated to sharper protein spots (4). Two orthogonal native dimensions with different detergent contents have been applied to resolve protein complex spots (2, 20). Three dimensional separation has also been reported. The protein complex slices/ spots from the BN-PAGE were electroeluted, detached from Coomasssie blue, and subsequently separated by 2D isoelectric focusing (IEF)/SDS–PAGE (25). Finally, prelabeling of different protein fractions by differential gel electrophoresis (DIGE) has been recently introduced into BN-PAGE studies and showed its advantages for comparative research (26, 27). For different research aims, different methods can be used after the BN-PAGE studies. For hypothesis-based experiment, simple immunoblotting experiment can be used to detect the known proteins on the one dimensional or multiple dimensional BN gels. For global analysis, proteomics is the method of choice to reveal the identity of the proteins. Protein complexes or subunits can also be recovered from the gels by electroelution or by diffusion (9). These proteins can be analyzed by 2D crystallization, atomic force microscopy researches, or as antigens to generate antibodies (28, 29). There are two variant forms of BN-PAGE. Colorless-native PAGE (CN-PAGE or CNE) is a method with the experimental setup very similar to that of BN-PAGE, except that it does not

Analysis of Protein Complexes by 2D Blue Native/SDS–PAGE and Antibody-Shift Assay

83

contain CBB. CNE is milder than BN-PAGE, and is particularly useful for in-gel catalytic activity assays and the analyses of fluorescent-labeled proteins (8, 30). As protein complexes migrate according to their intrinsic charges, only acidic proteins that move toward the anode would be resolved within the gel; i.e., basic protein complexes that are moved toward cathode and away from the gel could not be analyzed. Antibody Super Shift Assay (also called Antibody Super Migration) is often used to confirm the protein complex detected in BN gel. Antibody against the main component of the protein complex is added into the sample before BN analysis. The binding of the antibody to the protein complex causes an increase in mass. The complex should therefore be detected on the BN gel with a higher mass corresponding to the mass of the antibody than the original protein complex (13, 18). In the years since its development, BN-PAGE has become a very important tool in MPCs research (12, 31). Here more extensive practical advice, including potential pitfalls, of BN-PAGE is provided. This chapter explains the basic techniques of sample preparation, the first-dimension blue native gels, second-dimension SDS–PAGE, immunoblotting, and antibody super migration assay.

2. Materials The products used are listed below. Comparable products from other suppliers should also be effective. 2.1. Equipments

1. Hoefer SE 245 Dual Gel Caster (GE Healthcare, San Francisco, CA, USA). 2. SE 250 Mini-Vertical Unit for two slab gels (GE Healthcare). 3. Glass and alumina plates, 10 × 8 cm, for SE 250 (GE Healthcare). 4. Spacers and teflon combs for SE 250, 1.5 mm (GE Healthcare). 5. SG 15 Gradient maker, 15 mL total volume (GE Healthcare). 6. Mini-PROTEAN II system including Mini-PROTEAN II cell, 1.5 mm spacer glass plates, and 1.5 m IPG well combs (BIO-RAD, Hercules, CA, USA). 7. Microcon® Centrifugal Filter Units Corporation, Billerica, MA, USA).

2.2. Special Reagents

YM-5

(Millipore

All the reagents below must be of at least electrophoresis grade. 1. 2-[bis(2-hydroxy-ethyl)amino]-2-(hydroxymethyl)-1, 3-propanediol (Bis–Tris).

84

Yang et al.

2. 6-aminohexanoic acid. 3. Serva Blue G. 4. Dodecyl-b-d-maltoside (DDM). 5. Digitonin. 6. Protein and Peptide Molecular Weight Marker, HMW-Native (GE Healthcare). 2.3. Solutions 2.3.1. Stock Solutions

49.5% T, 3% C Acrylamide. 24 g acrylamide, 0.75 g bisacrylamide, 50 mL H2O, store at room temperature. 30% T, 3% C Acrylamide. 29 g acrylamide, 1 g bisacrylamide, 100 mL H2O, store at room temperature. 3× Gel buffer. 150 mM Bis–Tris–HCl, 1.5 M 6-amino-caproic acid. Adjust pH to 7.0 with HCl at 4°C, store at 4°C. 75% (w/v) Glycerol. Store at 4°C. 10× Cathode buffer. 500 mM Tricine, 150 mM Bis–Tris, at 4°C. No need to adjust pH. 5× Anode buffer 250 mM Bis–Tris–HCl, pH 7.0. Adjust pH to 7.0 with HCl at 4°C and store at 4°C. 2× Bis–Tris ACA. 200 mM Bis–Tris–HCl, 1 M 6-amino-caproic acid, pH 7.0. Adjust pH to 7.0 with HCl at 4°C, store at 4°C. 50BTH40G. 50 mM Bis–Tris–HCl, 40% (w/v) Glycerol, pH 7.0. Adjust pH to 7.0 with HCl at 4°C and store at 4°C. 10× BN-sample buffer. Serva Blue G 50 mg, 2× Bis–Tris ACA 500 mL, 75% glycerol 400 mL, H2O 100 mL. Store at −20°C.

2.3.2. Working Solutions

1× Cathode without dye, 1× anode buffer. Store at 4°C. 1× Cathode with dye. 0.01% Serva Blue G in 1× cathode buffer. Store at 4°C.


85

Solubilization buffer. 50BTH40G 100 mL, 10% detergent 40 mL and H2O 60 mL. Detergent: Triton X-100, DDM or digitonin. Store at 4°C. Immunoblotting transfer buffer. 25 mM Tris, 192 mM glycine, 20% methanol.

3. Methods 3.1. Sample Preparation

Add the solubilization buffer into samples (see Note 1). Put the samples on ice and gently vortex them each 5 min for 30 min (DDM, TX-100) or 30–60 min (Digitonin). Spin at 40,000– 100,000×g for 30 min and transfer the supernatant to new tubes. Adjust the buffer volume to get about 5 mg/mL protein per sample. If the sample is whole cellular lysate, simply add 2% DDM instead of solubilization buffer. Use water to adjust the protein concentration of the plasma sample. Cytosol or cytoplasm sample needs to be concentrated to about 5 mg/mL using Microcon YM-3 tube.

3.2. BN Gel Preparation

The gel dimension is 8.0 × 10.0 cm. Use SG 15 Gradient maker to make 5–13.5% gradient gels (Table 1) (see Note 2), or try different concentrations of gels for different samples (see Note 3). Gel and gradient preparation is carried out at 4°C to slow polymerization. The gels must be prepared more than 6 h and less than 48 h before use and store at 4°C. 1.5 mm gels are recommended because more samples can be loaded and more subunits can be found in the second dimensional gels.

Table 1 Casting of gradient BN gel with 1.5 mm spacer Stacking gel (4%, for two gels)

Gradient gel (5%, for one gel)

Gradient gel (13.5%, for one gel)

49.5% Acrylamide

0.485 mL

0.404 mL

1.09 mL

3× Gel buffer

2 mL

1.333 mL

1.33 mL

75% Glycerol

–

0.267 mL

1.067 mL

H2O

3.458 mL

1.97 mL

0.487 mL

TEMED

12 mL

3.8 mL

3.8 mL

10% AP

22.5 mL

10 mL

10 mL

Total volume

6 mL

4 mL

4 mL

86

Yang et al.

3.3. First Dimensional BN-PAGE

Add 1/10 10× BN-sample buffer into samples and keep on ice for about 5 min. If the samples are complicated, about 80 mg proteins per sample should be loaded into the gel. At the beginning, run the electrophoresis at very low voltage (5–10 V) until all the proteins enter the separation gel and then change the voltage to 50 V until the dye front runs off the bottom of the gel. The gels can be further stained by CBB or be directly used for second dimensional separation or immunoblotting (see Note 6).

3.4. Second Dimensional SDS–PAGE

Use another gradient maker to make 6–12% gels (Table 2), or try different concentrations of gels for different samples (see Note 4). Do not forget to use the IPG combs to make the gels. The whole BN gel lane or interested slices can be cut out and put into common 1× SDS–PAGE loading buffer, incubate at room temperature for 30 min to 1 h. Then put the gel segment on the SDS–PAGE gel orthogonally to the dimension of BN-PAGE. Using hands with new gloves to push the BN gel into the PAGE gel cell is recommended because human hands are the most available and soft tools we can use. Run the electrophoresis at 15 V until all the proteins enter the separation gel and then change the voltage to 50 V until the dye front reaches the b ottom of the gel.

3.5. Immunoblotting After BN-PAGE

Additional steps must be taken for the immunoblotting after BN-PAGE because the CBB in the gels may disturb the interaction between antibodies and proteins (see Note 5). One step is to change the cathode buffer containing dye with cathode buffer

Table 2 Casting of gradient SDS–PAGE gel with 1.5 mm spacer Stacking gel (5%)

Gradient gel (6%)

Gradient gel (12%)

H2O

1.4 mL

2.65 mL

1.65 mL

30% Acrylamide

0.33 mL

1 mL

2 mL

1.5 mol/L Tris (pH8.8)

–

1.25 mL

1.25 mL

1.0 mol/L Tris (pH6.8)

0.25 mL

–

–

10% SDS

20 mL

50 mL

50 mL

10% AP

20 mL

50 mL

50 mL

TEMED

12 mL

4 mL

2 mL

Total volume

2 mL

5 mL

5 mL

Use mini alumina and glass plates Sample gel height should be ~0.7 cm Make gel 1 day before use and store at 4°C


87

without dye when the dye front runs to one-third long of the gel and then continues to finish the electrophoresis. Assemble blotting sandwich without PVDF membrane first and only use more filter papers on the anode side. After 5 min at 50 V transferring, change the filter papers once and run another 5 min, change the filter papers again and put PVDF membrane into the sandwich, continue the transfer for 2–3 h. The following steps are routine steps (see Chap. 13). 3.6. Antibody Super Migration Assay

When we know a major component in a BN-PAGE slice, we can use its antibody to perform antibody super migration assay. Usually 10–20 mg antibodies should be added into per sample. Incubate the samples on ice for about 20 min, mix with sample buffer, and load them into BN gel. In ideal state, comparing to the results without antibody added, the original slice will be disappeared and new slice(s) will be presented above the original place. If the super migration slice(s) is (are) further identified to contain the target protein or exactly have same subunits with the original slice, we can confirm that the original slice is a real protein complex. The appearance of multiple super migration slices means there are more than one copy of the target protein in this complex or the BN-PAGE slice is formed by more than one protein complex containing the target protein. If there is still weak slice in the original place, it means that the antibody added is not superfluous or there is (are) comigration protein complex(es).

4. Notes 1. Sample preparation depends on the tissue or cell type. Membrane fractions of interest are isolated according to standard procedures. Fractions may include total membranes of cells or membranes of subcellular fractions: e.g., organelles. Organelles often can be efficiently prepared by combining differential centrifugations and density centrifugations. Schägger’s group also has provided an excellent protocol for preparation from several human tissues (32). The tissue is homogenized using a tight-fitting glass–Teflon homogenizer in a MOPS sucrose buffer containing protease inhibitors (440 mM sucrose, 20 mM Mops, 1 mM EDTA, 0.2 mM phenylmethylsulfonyl fluoride) and an enriched mitochondrial fraction is collected by centrifuging at 20,000×g for 20 min. 2. Common gel types are acrylamide gradient gels, although, very rarely, nongradient uniform acrylamide gels have been used for BN-PAGE. Uniform acrylamide gels can be optimal for separation of two protein complexes with similar masses,

88

Yang et al.

but several attempts may be required to find the special acrylamide concentration for the specific narrow mass range. 3. Gradient gel range of BN-PAGE. Standard 5–13.5% BN-PAGE gives good resolution of complexes between 100 kDa and 1 MDa. Other gradient gels may be preferred to focus on a specific complex or subcomplex. For example, a 10–20% gradient gel was employed to resolve protein complexes between 20 and 200 kDa. Agarose gels are more appropriate for the analysis of complexes larger than 1 MDa. Pyruvate dehydrogenase complex (PDC) is approximately 7 MDa; therefore, we substituted agarose for acrylamide to resolve PDC. The other crucial modification was a lowered salt concentration (0.05 M aminocaproic acid) to prevent dissociation of the complex. A high salt concentration aids extraction of membrane-embedded complexes but is unnecessary in the case of PDC, which is located in the mitochondrial matrix. The exact protocol of blue native agarose gel electrophoresis (BN-AGE) is described by Henderson and colleagues (33). Briefly, a very thin (3–4 mm) gel of 2.5–4% high-gel-strength agarose (SeaKem Gold, Flowgen) was poured into a standard horizontal minigel apparatus (Flowgen). Running buffer was blue cathode buffer B. Electrophoresis was performed at relatively high current (5–10 mA) to minimize smearing of complexes. Second-dimension SDS– PAGE was performed as described above, except using thicker spacers (1.5 mm) due to the thickness of the agarose. Since the agarose gel must be thin, this does not allow for loading of a large volume of sample (typically 8–10 mL). Whatever the gradient or gel matrix, we routinely use a minigel system (e.g., Bio-Rad, Hercules, CA, Mini-PROTEAN II) for PAGE and any horizontal gel rig (e.g., Flowgen) for agarose to shorten electrophoresis run times. Additional benefits are that a minigel system saves on chemicals and sample. The latter is particularly important in the case of precious human tissue from diseased individuals. 4. The second dimension is a SDS–polyacrylamide gel. Depending on the molecular weight range of interest, a glycine or tricine SDS–polyacrylamide gel can be used. A 10% tricine SDS-gel without a spacer gel is a good starting point as it can resolve polypeptides across the range 20–200 kDa. The addition of a spacer gel and an increase in acrylamide concentration extend the useful separation range to 1–100 kDa. 5. Immunodetection of protein complexes from a blot of one-dimensional BN-PAGE does not always give good results. One reason for this is that epitopes recognized by the antibody may be hidden in the complex. Another potential problem is nonspecific signal resulting from the blue dye, and the


89

high local concentration of dye bound to the target protein may interfere with antibody binding. Incubating the gel with dissociating reagent before blotting can improve immunodetection, as epitopes are liberated that were previously hidden. Stripping the blot by incubating it 30 min at 50°C in stripping buffer (100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris– HCl, pH 6.7) to remove Serva Blue G decreases nonspecific binding of antibody. However, in our hands a blot of the second-dimension SDS-gel almost always enhances the signal obtained. 6. There are also other choices for the separation of components obtained from the first dimension of BN-PAGE, such as a second dimension of IEF followed by a third dimension of tricineSDS–PAGE for the separation of subunits of complexes (9). The procedure for 3D-BN-PAGE (9, 25): (1) Electroelute proteins from BN gels as described above. (2) Precipitate proteins by adding a 10% (vol/vol) trichloroacetic acid (TCA) solution (final TCA concentration: 3.5%); incubate for 10 min at 0.1°C and centrifuge at 13,000×g for 10 min at 0.1°C. (3) Resuspend proteins in acetone supplemented with 20 mM DTT, 1 mM PMSF; incubate for 1 h at −20.1°C; and centrifuge at 13,000×g for 15 min at 4.1°C. (4) Repeat step 3. (5) Dry the pellet using a vacuum centrifuge. (6) Resuspend the pellet in “lysis solution” for IEF. (7) Carry out IEF according to standard procedures. All basic strategies for IEF are suitable for the separation of proteins electroeluted from blue native gels, like tube-gel separations using mobile ampholytes 50 or separations on gel strips using immobilized pH gradients 51. (8) Horizontally transfer the IEF gel strip onto an SDS polyacrylamide gel and carry out SDS–PAGE according to standard procedures. Three-dimensional BN/IEF/SDS–PAGE allows separation of isoforms of subunits of protein complexes, which often have very similar molecular masses but differ with respect to their isoelectric points. Especially in mammals and higher plants, subunits of protein complexes are sometimes encoded by small protein families. For instance, the preprotein translocase of the outer mitochondrial membrane from Arabidopsis includes four isoforms for the preprotein receptor TOM20, which can be separated nicely by 3D BN/IEF/SDS–PAGE.

Acknowledgments This work was partially supported by National Natural Science Foundation of China for Creative Research Groups (30621063) and International Scientific Collaboration Program (2009DFB33070)

90

Yang et al.

References 1. Schägger H, von Jagow G. (1991) Blue native electrophoresis for isolation of membrane protein complexes in enzymatically active form. Anal Biochem 199, 223–231. 2. Wittig I, Braun HP, Schägger H. (2006) Blue native PAGE. Nat Protoc.1, 418–428. 3. Schägger, H., Cramer, W. A., von Jagow, G. (1994) Analysis of molecular masses and oligomeric states of protein complexes by blue native electrophoresis and isolation of membrane protein complexes by twodimensional native electrophoresis. Anal. Biochem. 217, 220–230. 4. Aufurth S, Schagger H, Muller V. (2000) Identification of subunits a, b, and c1 from Acetobacterium woodii Na+−F1F0-ATPase. Subunits c1, c2, AND c3 constitute a mixed c-oligomer. J Biol Chem 275, 33297–33301. 5. Kruft V, Eubel H, Jansch L, Werhahn W, Braun HP. (2001) Proteomic approach to identify novel mitochondrial proteins in Arabidopsis. Plant Physiol 127, 1694–1710. 6. Camacho-Carvajal MM, Wollscheid B, Aebersold R, Steimle V, Schamel WW. (2004) Two-dimensional Blue native/SDS gel electrophoresis of multi-protein complexes from whole cellular lysates: a proteomics approach. Mol Cell Proteomics 3, 176–182. 7. Liu K, Qian L, Wang J, Li W, Deng X, Chen X, et al. (2009) Two-dimensional blue native/ SDS–PAGE analysis reveals heat shock protein chaperone machinery involved in hepatitis B virus production in HepG2.2.15 cells. Mol Cell Proteomics 8, 495–505. 8. Hellwig S, Schamel WW, Pflugfelder U, Gerlich B, Weltzien HU. (2005) Differences in pairing and cluster formation of T cell receptor alpha- and beta-chains in T cell clones and fusion hybridomas. Immunobiology 210, 685–694. 9. Nicke A, Baumert HG, Rettinger J, Eichele A, Lambrecht G, Mutschler E, et al. (1998) P2X1 and P2X3 receptors form stable trimers: a novel structural motif of ligand-gated ion channels. EMBO J 17, 3016–3028. 10. Nicke A, Kerschensteiner D, Soto F. (2005) Biochemical and functional evidence for heteromeric assembly of P2X1 and P2X4 subunits. J Neurochem 92, 925–33. 11. Hedman E, Widen C, Asadi A, Dinnetz I, Schroder WP, Gustafsson JA, et al. (2006) Proteomic identification of glucocorticoid receptor interacting proteins. Proteomics 6, 3114–3126. 12. Deng X, Li W, Chen N, Sun Y, Wei H, Jiang Y, et al. (2009) Exploring the priming mechanism

of liver regeneration: proteins and protein complexes. Proteomics 9, 2202–2216. 13. Evin G, Canterford LD, Hoke DE, Sharples RA, Culvenor JG, Masters CL. (2005) Transition-state analogue gamma-secretase inhibitors stabilize a 900 kDa presenilin/nicastrin complex. Biochemistry 44, 4332–4341. 14. Culvenor JG, Ilaya NT, Ryan MT, Canterford L, Hoke DE, Williamson NA, et al. (2004) Characterization of presenilin complexes from mouse and human brain using Blue Native gel electrophoresis reveals high expression in embryonic brain and minimal change in complex mobility with pathogenic presenilin mutations. Eur J Biochem 271, 375–385. 15. Deng XY, Li WR, Sun YW, Wei HD, Jiang Y, He FC. (2009) Exploring rat plasmatic proteomes: what triggered the liver regeneration? Protein Pept Lett 16, 698–705. 16. Schamel, W. W. A., Reth, M. (2000) Monomeric and oligomeric complexes of the B cell antigen receptor. Immunity 13, 5–14. 17. Wittig I, Schagger H. (2008) Features and applications of blue-native and clear-native electrophoresis. Proteomics 8, 3974–3990. 18. Zhang P, Battchikova N, Paakkarinen V, Katoh H, Iwai M, Ikeuchi M, et al. (2005) Isolation, subunit composition and interaction of the NDH-1 complexes from Thermosynechococcus elongatus BP-1. Biochem J 390(Pt 2), 513–520. 19. Lasserre JP, Beyne E, Pyndiah S, Lapaillerie D, Claverol S, Bonneu M. (2006) A complexomic study of Escherichia coli using two-dimensional blue native/SDS polyacrylamide gel electrophoresis. Electrophoresis 27, 3306–3321. 20. Schägger H, Pfeiffer K. (2000) Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. EMBO J 19, 1777–1783. 21. Wittig I, Schagger H. (2005) Advantages and limitations of clear-native PAGE. Proteomics 5, 4338–4346. 22. Devreese B, Vanrobaeys F, Smet J, Van Beeumen J, Van Coster R. (2002) Mass spectrometric identification of mitochondrial oxidative phosphorylation subunits separated by two-dimensional blue-native polyacrylamide gel electrophoresis. Electrophoresis 23, 2525–2533. 23. Millar AH, Eubel H, Jansch L, Kruft V, Heazlewood JL, Braun HP. (2004) Mitochondrial cytochrome c oxidase and succinate dehydrogenase complexes contain plant specific subunits. Plant Mol Biol 56, 77–90. 24. Reifschneider NH, Goto S, Nakamoto H, Takahashi R, Sugawa M, Dencher NA, et al. (2006) Defining the mitochondrial proteomes from five rat organs in a physiologically significant

Analysis of Protein Complexes by 2D Blue Native/SDS–PAGE and Antibody-Shift Assay context using 2D blue-native/SDS–PAGE. J Proteome Res 5, 1117–1132. 25. Werhahn W, Braun HP. (2002) Biochemical dissection of the mitochondrial proteome from Arabidopsis thaliana by three-dimensional gel electrophoresis. Electrophoresis 23, 640–646. 26. Reisinger V, Eichacker LA. (2007) How to analyze protein complexes by 2D blue native SDS–PAGE. Proteomics 7 Suppl 1, 6–16. 27. Heinemeyer J, Scheibe B, Schmitz UK, Braun HP. (2009) Blue native DIGE as a tool for comparative analyses of protein complexes. J Proteomics 72, 539–544. 28. Poetsch A, Neff D, Seelert H, Schagger H, Dencher NA. (2000) Dye removal, catalytic activity and 2D crystallization of chloroplast H(+)-ATP synthase purified by blue native electrophoresis. Biochim Biophys Acta 1466, 339–349. 29. Seelert H, Dencher NA, Muller DJ. (2003) Fourteen protomers compose the oligomer III

91

of the proton-rotor in spinach chloroplast ATP synthase. J Mol Biol 333, 337–44. 30. Wittig I, Karas M, Schagger H. (2007) High resolution clear native electrophoresis for ingel functional assays and fluorescence studies of membrane protein complexes. Mol Cell Proteomics 6, 1215–1225. 31. Meyer B, Wittig I, Trifilieff E, Karas M, Schägger H. (2007) Identification of two proteins associated with mammalian ATP synthase. Mol Cell Proteomics 6, 1690–1699. 32. Wittig I, Carrozzo R, Santorelli FM, Schagger H. (2007) Functional assays in high-resolution clear native gels to quantify mitochondrial complexes in human biopsies and cell lines. Electrophoresis, 28, 3811–3820. 33. Henderson NS, Nijtmans LG, Lindsay JG, Lamantea E, Zeviani M, Holt IJ. (2000) Separation of intact pyruvate dehydrogenase complex using blue native agarose gel electrophoresis.Electrophoresis 21, 2925–2931.

wwwwwwwwwwwwwwwwww

Part III Analytical Tools for Neuroproteomics

wwwwwwwwwwwwwwwwww

Chapter 8 Two-Dimensional Gel Electrophoresis-Based Proteomic Analysis of Brain Synapses Karl-Heinz Smalla and Ursula Wyneken Abstract Neuroproteomic technologies are crucial for a deeper understanding of molecular mechanisms underlying neuronal plasticity and pathologies of the central nervous system. For a comprehensive neuroproteomic analysis, high-resolution high-throughput techniques are indispensable. The most commonly used system for 2D gel electrophoresis is based on the combination of isoelectric focussing and sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Nevertheless, the analysis of complex samples derived from biochemically prepared synaptic fractions raises some crucial challenges. Therefore, we describe the general 2D gel electrophoresis procedure, including sample preparation, gel casting, electrophoresis condition, and the detection of proteins on the gel with mass spectrometry compatible silver staining, sypro-ruby or colloidal Coomassie Brilliant Blue. Key words: 2D gel electrophoresis, Sample preparation, Isoelectric focussing, SDS-PAGE, Protein staining, Image analysis

1. Introduction In the last 2 decades, many proteomic analyses of central nervous system (CNS) synapses have been carried out by using 2D gel electrophoretic separation technologies. While several separation technologies based on different physico-chemical separation principles have been established, the majority of 2D gel-based studies relies on the combination of isoelectric focusing (IEF) with sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) as reported by Klose (1) and O’Farrell in 1975 (2). IEF separates proteins according to their isoelectric point in the first dimension, whereas SDS-PAGE separates proteins according to their masses in the second dimension. A typical 2D gel can resolve complex protein extracts into hundreds or even up to Ka Wan Li (ed.), Neuroproteomics, Neuromethods, vol. 57, DOI 10.1007/978-1-61779-111-6_8, © Springer Science+Business Media, LLC 2011

95

96

Smalla and Wyneken

Fig. 1. Workflow for 2D gel-based separation using isoelectric focusing for first dimension and SDS-PAGE for second dimension separation. The description of methods in the text is following this outline.

thousands of single protein spots. This type of 2D gel system (IEF/SDS-PAGE) is superior to other gel-based systems in resolution power and reproducibility. Therefore, it is preferably used to quantify differences in functional proteomics studies. The workflow for IEF/SDS-PAGE 2D gel electrophoresis is depicted in Fig. 1. Some general parameters related to different stages of the workflow should be considered first before going into details: 1. The sensitivity of the chosen staining methods and the size of immobilised pH gradient (IPG) strips determine which volume and protein amount should be applied for separation. Typical amounts for a complex protein mixture to be loaded on 18 cm IPG strips are 600–1,000 mg protein for Coomassie Brilliant Blue G-250 staining and 300–500 mg protein for SYPRO™ Ruby staining or Mass Spectrometry-compatible silver staining (staining protocols see below) while the required sample volumes per IPG strip are clearly defined in the manufacturer’s protocols. Thus, typical protein concentrations of samples are 1.5–2.5 mg/mL for Coomassie staining and 0.7–1.5 mg/mL for silver staining. 2. Choose the separation range for IEF. Commercially available IPG strips might have either a broad or narrow pH separation range. If there is no hypothesis-driven restriction, use a broad separation range like for instance GE Healthcare IPG strips 3-11NL, at least for pilot experiments. 3. For the SDS-PAGE, homogeneous or gradient gels can be used. For best reproducibility, preferably use homogenous gels.

Two-Dimensional Gel Electrophoresis-Based Proteomic Analysis of Brain Synapses

97

The acrylamide monomer concentration will define the optimum separation range. For a SDS-PAGE separation range of 10–100 kDa, a monomer concentration of 11% is a good choice, nevertheless resolution within this system is optimised between 10–80 kDa and changes in monomer concentrations might change optimum separation ranges accordingly. Decreasing acrylamide monomer concentrations might improve resolution in higher molecular weight ranges. Several methods for the purification or enrichment of synaptic or subsynaptic structures have been described (8, 9, 10). An example for the resolution power of IEF/SDS-PAGE is shown in Fig. 2. There are some limitations to IEF/SDS-PAGE, i.e. the poor presentation of transmembrane proteins, high molecular weight proteins (Mr >100 kDa) and proteins with pI beyond the pH range of the IEF gel strip. Hence, if a study focuses on large scale identification of protein components in synaptic or subsynaptic proteomes, alternative combinations of first dimension separations followed SDS-PAGE are the better choice (3, 4). For this blue native gel electrophoresis or 16-benzyldimethyl-n-hexadecylammonium chloride (5, 6, 7) electrophoresis in combination with

Fig. 2. Separation of synaptosomes derived from rat brain cortex. Six hundred micrograms of protein were separated first in an IPG strip 3-11NL and in the second dimension separated with SDS-PAGE 11% gel. Gels were silver stained (11). Synaptosomes were prepared as described in (8, 9).

98

Smalla and Wyneken

second dimension SDS-PAGE has been applied very successfully to study samples derived from brain tissue and subsynaptic structures (see Chaps. 7 and 9).

2. Materials Take care when handling the chemicals mentioned below and read the Material and Safety Data Sheets thoroughly. Take all measures to protect yourself since several hazardous reagents are used which cannot be substituted by less harmful substances. Wearing protective gloves throughout all experimental procedures is favourable. Products used in the following protocols are listed below, comparable products from other suppliers should be also effective. 2.1. Sample Preparation

All Chemicals PlusOne grade (GE Healthcare) if available.

2.1.1. Chemicals/Solutions

–– Acetone p.A. –– Parafilm –– Ethanol (96% or absolute) p.A.; final concentration: 80/20% (v/v) ethanol/water. –– 1 mM Tris/HCl, pH 7.4. –– Centrifuge tubes (5–15 mL) resistant to acetone, ethanol and mechanically stable to 10,000×g forces; glass tubes (e.g. “Corex” tubes from Thermo Scientific/Sorvall) work very well. Preferably use swing-out rotors. –– Sample buffer: use chemicals of “PlusOne grade” (GE Healthcare). Component

Final concentration

Amount/10 mL

Ultrapure Ureaa

9 M

5.4 g

CHAPS

4.0% (w/v)

0.4 g

Triton X-100

0.5%

50 mg

Tris

20 mM

0.024 g

DTT

64 mM

100 mg

IPG buffer (same pH range as the IPG strip)

0.5% (v/v)

50 mL

Bromophenol blue

Approximately 0.002%

(a few grains)

Deionised water See Note 1

a

to 10 mL


99

–– Store in 1.0 mL aliquots at –80°C. If protein quantification by a Bradford Assay is planned, add the IPG buffer after quantification. –– Bradford assay from Bio-Rad (Cat.-No. 500-0002) with bovine serum albumin as standard. 2.1.2. Equipment

–– For Bradford assays, a spectral photometer set to 595 nm is required. –– For drying of samples, a lyophiliser (e.g. FreeZone2.5 from Labconco in our lab) is very useful.

2.2. Isoelectric Focusing

–– IPG strips (choose length of strips and pH range). –– IPG cover fluid.

2.2.1. Chemicals/ Solutions/Consumables 2.2.2. Equipment

–– Curved tweezers with extra fine tips for handling of IPG strips. –– IPG strip holder. –– Tray(s) for IPG strips. –– Equipment for IEF: IPGphor (GE Healthcare Life Sciences), see also Note 2.

2.3. Sodium Dodecyl SulphatePolyacrylamide Gel Electrophoresis 2.3.1. Chemicals/Solutions

All chemicals should be of PlusOne grade (GE Healthcare) if available, see Note 3. –– Equilibration buffer stock solution: 50 mM Tris/HCl pH 8.8, 6 M urea, 30% (v/v) glycerol, 2% (w/v) SDS, Bromophenol blue 0.002%. For 200 mL use: 72.07 g urea, 60 mL 100% glycerol, 40 mL 10% SDS, 10 mL 1.5 M Tris/HCl pH 8.8 (Bio-Rad Cat.-No. 161-0798), add deionised water to 200 mL. It can be stored in 50 mL aliquots at −20°C. –– Equilibration buffer 1 contains DTT (100 mg/10 mL). –– Equilibration buffer 2 contains Iodoacetamide (250 mg/ 10 mL). –– Gel buffer: stock solution: 1.5 M Tris/HCl pH 8.8, store at 4°C. Either take ready-to-use Bio-Rad (Cat.-No. 161-0798) or take home-made buffer (see Note 4). –– 10% SDS (10 g SDS dissolved in deionised water and filled up to 100 mL), store at room temperature. –– Acrylamide stock solution (30.8% T, 2.6% C): use 30% Duracryl from Genomic Solutions (Cat.-No. 80-0085), store at 4°C (see Note 5). –– TEMED (N,N,N ¢,N ¢-Tetramethylethan-1,2-diamin), store at 4°C.

100

Smalla and Wyneken

–– 10% Ammonium persulphate (APS) (1.1 g APS fill up to 11 mL with deionised water), prepare freshly before use. –– Displacing solution for casting gels, prepare freshly before use. –– 375 mM Tris/HCl pH 8.8, 50% glycerol, bromophenol blue, for 200 mL. Tris/HCl (1.5 M, pH 8.8)

50 mL

Glycerol (100%)

100 mL

Bromophenol blue

2 mg

Deionised water

50 mL

–– SDS-PAGE running buffer: Use 10× TGS-buffer stock solution (Bio-Rad Cat.-No. 161-0772) diluted 1:10 (v/v) with deionised water; keep stock solution at room temperature, running buffer is kept at 4°C; alternatively home-made buffers can be used (see Note 6). –– Prepare 100 mL of 1% agarose solution in SDS-PAGE running buffer, for use within a few days store at room temperature but for longer times keep at 4°C. –– Recipe for casting 11 vertical DALT slab gels, total volume 1,100 mL. Component

11% gel (in mL)

12% gel (in mL)

13% gel (in mL)

Duracryl 30%

403

440

477

Tris/HCl (1.5 M, pH 8.8)

275

275

275

Water

400

363

326

11

11

11

10% SDS TEMED 10% APS 2.3.2. Equipment

0.4 11

0.5 11

0.5 11

–– Tray for equilibration of IPG strips. –– Hoefer Ettan ISODalt electrophoresis tank. –– Power supply Hoefer EPS2A200 (max 200 V, 2A). –– GE Healthcare Thermostat MultiTemp III. –– Glass plate cassettes (gel size 200 × 250 × 1.5 mm). –– ISODalt multigel caster.

2.4. Protein Staining 2.4.1. Chemicals/Solutions for Mass SpectrometryCompatible Silver Staining

–– Fixing solution: 50% ethanol, 5% acetic acid, store at room temperature. –– Washing 1: 50% ethanol, store at room temperature. –– Washing 2: deionised water.


101

–– Sensitiser: 0.02% (w/v) sodiumthiosulphate in deionised water (0.6 g/3 L), always prepare fresh. –– Silver impregnation solution: 0.15% (w/v) silver nitrate in deionised water (4.5 g/3 L), always prepare fresh. –– Developer solution: 0.04%(v/v) formaldehyde, 2% (w/v) sodium carbonate in deionised water (60 g sodium carbonate, 1.2 mL formaldehyde for 3 L developer solution), make always fresh. –– Stopper: 5% (v/v) acetic acid in deionised water, keep at room temperature. –– Storage solution: 0.2% (v/v) acetic acid, 10% (v/v) methanol in deionised water; can be kept at room temperature for weeks (11). 2.4.2. Chemicals/Solutions for SYPRO™ Ruby staining

–– Fixing solution for five large format gels: 5 L of 50% methanol/7% acetic acid in deionised water. –– SYPRO™ Ruby gel stain (Invitrogen/Molecular Probes, Cat.-No. S 12000, 1 L). –– Wash solution for five large format gels: 5 L of 10% methanol/7% acetic acid in deionised water.

2.4.3. Chemicals/Solutions for Colloidal Coomassie Brilliant Blue G-250 Staining

–– Fixing solution: 50% (v/v) ethanol, 3% (v/v) phosphoric acid in deionised water (1.5 L ethanol; 106 mL phosphoric acid (85%) for 3 L fixing solution), stable at room temperatures for a few weeks. –– Staining solution: 34% (v/v) methanol, 3% (v/v) phosphoric acid, 15% (w/v) ammonium sulphate, 0.1% (w/v) Coomassie Brilliant Blue G-250 in deionised water (340 mL methanol, 35.3 mL phosphoric acid (85%), 150 g ammonium sulphate, 1 g Coomassie Brilliant Blue, add water to 1 L).

2.4.4. Equipment for Staining Procedures

–– Plastic trays for incubations 25 × 28 × 10 (height) cm, do not treat more than five gels in one tray. –– Orbital mixer or rocking platform mixer.

2.5. Data Evaluation/ Image Analysis Equipment

–– Scanner for calibrated digitisation of protein staining patterns (gel images), in our lab a Bio-Rad GS-800 scanner is used. –– For visualisation of SYPRO™ Ruby staining a Fluorescence Scanner is required, we use the Bio-Rad Molecular Imager FX. –– Computer including software for matching and quantifications of protein patterns; in our lab we use PDQuest from Bio-Rad (see Note 7).

102

Smalla and Wyneken

3. Methods 3.1. Sample Preparation for IEF

In general, samples for the separation in 2D gel electrophoresis contain several contaminating compounds in varying quantities depending on the protocols used to isolate synaptic protein fractions. Such contaminants could be lipids, polysaccharides, ions from buffer components or ionic detergents. After removal of at least the majority of contaminants, samples are solubilised within a suitable sample buffer with subsequent removal of insoluble material. High standards in reproducible sample preparations are an absolute requirement for high-quality separations, since most of these contaminants interfere extremely with IEF. Therefore, we established in our lab a simple but very effective method for removal of contaminants interfering with IEF and thus further steps of 2D gel separations of synaptic protein fractions (see Note 8). The procedure consists basically of four steps: (1) protein precipitation and extraction of lipids, (2) washing the protein pellet with further extraction of lipids and removal of ionic compounds, (3) sample lyophilisation and solubilisation with IEF sample buffer and (4) removal of insoluble material by high-speed centrifugation (see Note 9). This can be achieved as follows: 1. Add a fourfold volume of pre-chilled (−20°C) acetone to the protein sample, mix well and leave overnight at −20°C, use glass tubes, seal the tubes with Parafilm. Usually, almost immediately a precipitate starts to form, but it is better to wait for some hours or overnight at least to finalise this first precipitation step. Precipitation should be standardised, we prefer to do the precipitation overnight. Further precipitation or washing steps do not require longer incubation times. 2. Spin samples at 10,000×g for 20 min, discard the supernatant and resuspend the pellet again (use the same volume as at step 1) in pre-chilled (−20°C) acetone carefully. 3. Repeat step 2 twice. 4. Resuspend the remaining pellet in pre-chilled (−20°C) 80% ethanol. Spin samples at 10,000×g for 20 min, discard the supernatant and resuspend the pellet again (use the same volume as at step 1) in pre-chilled (−20°C) 80% ethanol carefully. Repeat this four times. 5. The remaining pellet is freeze-dried or left overnight in a clean-bench with ventilation to dry samples. 6. Add the required amount of sample buffer (without IPG buffer) to the samples to yield approximate final protein concentrations of 1–2 mg/mL for silver staining or SYPRO™ Ruby staining, and 2–4 mg/mL for Coomassie Brilliant Blue staining.


103

7. Spin samples in Eppendorf tubes for 30 min or longer to remove unsolubilised material at maximum speed (20,000×g with an Eppendorf 5415R centrifuge). 8. Determine protein concentration with Bradford assay from Bio-Rad (Cat.-No. 500-0002) with bovine serum albumin as standard. Light absorption at 595 nm is measured with a spectral photometer. 9. Bring samples to the same protein concentration by adding sample buffer (without IPG buffer); finally add IPG buffer to a final concentration of 0.5%. 3.2. First Dimension: Isoelectric Focussing 3.2.1. Sample Loading

1. Select the stripholder(s) or multichannel tray corresponding to the IPG strip length chosen for the experiment. 2. Pipette the appropriate volume of sample-containing rehydration solution into each holder (or lane of the tray). For an 18 cm IPG strip 400 mL and for a 7 cm strip 125 mL or for other strip lengths according to the manufacturer’s recommendation. Deliver the solution slowly at a central point in the strip holder channel away from the sample application wells. Remove any larger bubbles. 3. Take the strips out of the freezer (−20°C) and give a few minutes until they have room temperature (see Note 11). Mark the IPG strips with a number on the square end of the strip for sample identification. 4. Remove the protective cover from the IPG strip. Position the IPG strip with the gel side down and the pointed (anodic) end of the strip directed towards the pointed end of the strip holder. Introduce the pointed end first, lower the IPG strip onto the solution. To help coating the entire strip gently lift and lower the strip and slide it back and forth along the surface on the solution, tilting the strip holder slightly as needed to ensure complete and even wetting. Finally, lower the cathodic (square) end of the IPG strip into the channel, making sure that the gel contacts the strip holder electrodes at each end. Be careful not to trap air bubbles beneath the strip. 5. Apply 600 mL IPG cover fluid (for an 18 cm strip; for other strip length accordingly) to minimise evaporation and urea crystallisation. Pipette the fluid dropwise into the strip holder until the entire IPG strip is covered. Place the cover on the strip holder. 6. Allow the IPG strips to rehydrate. Preferably this is done on the IPGphor unit platform for not less than 10 h, ensure that the holder is on a level surface. Best is to program the rehydration period as the first step of an IPGphor protocol running the IPG strips at low voltage as given in the protocol below.

104

Smalla and Wyneken

3.2.2. Running First Dimension Separation IEF

Step Rehydration

Most important requirements for the first dimension IEF separation are to keep temperature at 20°C and to limit maximum currents per strip to 50 mA. The separation protocol could be organised either in step-wise or in gradual changes of applied voltage – as a standard procedure we use voltage steps (step-n-hold) but in some cases a gradual increase to the final voltage might improve separation especially for difficult samples still containing low amounts of contaminants. Our standard protocol for 18 cm IPG strips 3-11NL is as indicated below. It can also be applied to other 18 cm IPG strips with gradients pH 3–10, pH 3–10NL or pH 4–7 (see Note 10). Select 50 mA per IPG strip and for both rehydration and IEF 20°C. The rehydration step may also performed separately outside the IPGphor. Voltage (V)

Step duration

Volt-hours

Gradient type

0

10:00 h

S1

30

12:00 h

360

Step-n-hold

S2

500

1:00 h

500

Step-n-hold

S3

1,000

1:00 h

1,000

Step-n-hold

S4

8,000

74,000 Vhrs

74,000

Step-n-hold

1. Clean the IPGphor platform properly with soft tissue to remove residual oil and dust particles. Position the stripholders properly on the IPGphor platform. 2. Start the program, it is running approximately 23 h or more. 3. After IEF proceed to the second dimension separation immediately or store the IPG strips at −40 to −80°C in tubes or between two plastic sheets. 3.3. Second Dimension: SDSPolyacrylamide Gel Electrophoresis 3.3.1. Equilibration of IPG Gel Strips

After IEF, IPG gel strips have to be equilibrated before running second dimension separations. For this, two subsequent equilibration steps are required in buffers containing SDS: The first equilibration buffer contains 1% DTT but the second contains 260 mM iodoacetamide to remove excess DTT (see Notes 12 and 13). Procedure for ten strips: 1. Dissolve 250 mg DTT in 25 mL of equilibration buffer to get equilibration buffer 1. 2. Pipette 2.4 mL equilibration buffer 1 (containing DTT) in each well of the tray for IPG strips. 3. Take out the focused IPG strips one by one and remove the coverfluid, avoid damage to the gel and carefully wipe off most of the cover fluid on the plastic side of the strips with soft tissue.


105

4. Place the strips individually in a well of the equilibration tray with the gel side faced up. 5. Place the equilibration tray on a rocking shaker. Equilibrate for a maximum of 15 min at room temperature. 6. Dissolve 625 mg of iodoacetamide in 25 mL equilibrium buffer to get equilibration buffer 2. 7. Take the IPG strips out of the tray and remove residual equilibration buffer 1 with soft tissue. 8. Add 2.4 mL equilibration buffer 2 into each well of the equilibration tray. 9. Place the strips individually in a well of the equilibration tray with the gel side faced up. 10. Place the equilibration tray on a rocking shaker. Equilibrate for a maximum of 15 min at room temperature. 3.3.2. Casting Homogeneous SDSPolyacrylamide-Gels

1. Make sure that the casting system and the glass cassettes are clean, dry and free from any polymerised acrylamide. 2. Prepare a sufficient volume of gel overlay solution. For each cassette 1.0 mL is needed. 3. Prepare 200 mL of displacing solution. 4. Prepare the acrylamide solution without both, TEMED and 10% APS. 5. Load the gel caster with glass cassettes, separator sheets and filler blocks, if required (for eleven 1.5 mm gels use four big and one small filler block). Place a gel label in each cassette (see Note 14). Do not forget to insert the rubber and the sponge into the casting chamber to avoid air bubbles. 6. Connect the feed tube to a funnel held in a stand at a level of about 30 cm above the top of the gel caster. Insert the other end of the feed tube in the grommet in the bottom of the balance chamber. 7. Load the balance chamber with 200 mL displacing solution. 8. Add the appropriate volume of TEMED and 10% APS only when you are ready to pour the gel, not before. 9. Pour the gel solution into the funnel slowly, taking care to avoid introducing any air bubbles into the feed tube line. 10. When pouring is complete, immediately remove the feed tube from the balance chamber grommet giving way for the displacing solution to shift all the gel solution upwards into the glass cassettes. If the V-well in the caster is not completely filled and the level of gel solution in the cassettes is more than 1 cm below the cassette top, you may add up to 50 mL more displacing solution to the balance chamber.

106

Smalla and Wyneken

11. After removal of the feed tube from the caster, immediately deliver carefully 1.0 mL of overlay buffer to the surface of each gel (see Note 15). The overlay should spread evenly across the cassette with a minimum of mixing resulting in a smooth, flat gel top surface. 12. Wrap the open part of the caster with (plastic) foil and allow the gels to polymerise for at least 1 h (see Note 16). 3.3.3. Loading IPG strips onto DALT Slab Gels

1. Dip the IPG gel strips in SDS electrophoresis buffer to lubricate it and wash residual coverfluid away. 2. Place the IPG strip between the plates touching the surface of the second dimension gel, with the plastic side backing against one of the glass plates (see Note 17). Insert the IPG strips always in the same orientation, e.g. putting always the acidic (edged) end to the same side of the gel as the gel label in order to have a convenient reference. Use a thin plastic backing from the package of the IPG strips to push against the plastic backing of the strips (but not the gel itself) and move the strip down into contact with the surface of the second dimension gel. Avoid trapping air bubbles between the plastic backing and the glass or cutting into the slab gel with the strip. The gel face must not touch the opposite glass plate. 3. Load a protein molecular weight marker (either unstained Bio-Rad Cat.-No. 161-0363 or prestained Bio-Rad Cat.-No. 161-0373 or 161-0374) on a precut filter piece (approximately 5 × 5 mm, approximately 0.5–1 mm thick), for silver stained or SYPRO™ Ruby stained gels 2 mL and for Coomassie Brilliant Blue stained gels or western blots 7 mL. Position the marker right from the basic end of IPG gel strip. 4. Melt the agarose at 100°C for approximately 10 min. Allow the agarose to cool to 40–50°C and then deliver the agarose sealing solution onto the IPG strip to seal it into place. Avoid air bubbles when sealing with agarose. Wait for approximately 5 min to allow the agarose to solidify completely before proceeding.

3.3.4. Running Second Dimension SDS-PAGE

1. Prepare SDS-PAGE running buffer and precool the buffer (for ISODALT 22 L buffer required) in the electrophoresis chamber for a few hours at 16°C. 2. Carefully load the cassettes after agarose sealing has fully solidified. The cassettes are correctly loaded in running orientation in the DALT tank slot with the IPG strips vertical along the left side (cathode side, negative pole, red cable connector) and the rubber cassette hinge to the bottom. Dip the hinge side of the cassette into the tank buffer first to lubricate it before inserting into the flap seals. Use both hands to slide the cassettes firmly to the bottom (see Note 18).


107

3. Adjust the buffer level to the top edge of the cassettes after all cassettes are loaded in position. Check that the tank buffer level is even with the uppermost spacer of the cassette neither above the top of the cassette nor below the level of the top edge of the gel. 4. Close the lid and attach the electrical leads to make proper electrical contact with the power supply. Migration proceeds towards the red side (positive pole, right chamber). 5. Set the power supply to 80 V constant voltage for overnight runs (it takes about 18–20 h for 11% gels) 6. SDS-PAGE separation is finished, when the bromophenol band reaches the right end of the gel. Switch off the power supply and take the first cassette out of the tank. 7. Open the cassette with a plastic tool which fits between the glass plates but is also stable enough to open the cassette by twisting the tool slowly (see Note 19). Transfer the gel to a tray containing fixing solution for protein staining, shake the gel(s) gently. Continue with the other gels the same way but put not more then five gels into one tray. Use excess fixing solution, at least a minimum of 2.5 L for 5 gels. 8. Put the cassettes into a wash tank, take care that the glass plates do not get dry before they have been cleaned thoroughly. 3.4. Protein Staining

Various protocols for protein stainings that differ in sensitivity and compatibility with subsequent analysis methods are available. For identification of proteins from gel spots, mass spectrometric analysis is indispensable. Therefore, all procedures have to be compatible with mass spectrometry. For quantitative comparison of gels from different groups, we use the more sensitive silver staining (detection limit is approximately 1 ng protein) but for selecting samples based on the results of image analysis we take Coomassie Brilliant Blue stained gels with higher protein loads (detection limit is approximately 40 ng protein). Staining with fluorescent dyes like SYPRO™ Ruby is an interesting but also very expensive alternative to the silver staining or Coomassie staining. The sensitivity is as high as for silver stains (detection limit is approximately 1–2 ng protein) and the dynamic (linear) range is higher than with conventional stainings. In addition, SYPRO™ Ruby stained gels can be stained afterwards with Coomassie stains when manual picking of spots is required (Note 20).

3.4.1. MS: Compatible Silver Staining

Several MS compatible protocols for silver staining are available, the most frequently used method was described by Shevchenko et al. (11) and a slightly modified variant of it is described here. If only detection of protein but not quantitation of protein

108

Smalla and Wyneken

amounts is required, the much more sensitive staining method introduced by Heukeshoven and Dernick (13) is a good alternative. The recommended incubation times and volumes are optimised for “large scale” silver staining of ten 2D gels (20 × 25 cm × 1.5 mm) in two separate boxes with five gels per box. All incubation volumes are 1.5 L per box with five gels. All incubation steps are performed preferably on a circular shaker. To retain the gels in the box when decanting solutions, we use a home-made “retainer” with holes. 1. Fixation of proteins in the gels is achieved by incubation in fixing solution for a minimum of 60 min. 2. Wash in 50% ethanol for 15 min. 3. Wash three times in distilled water for 15 min. Prepare already at this point 2 × 3 L of developer solution (but without the formaldehyde) for ten gels, since dissolving of sodium carbonate requires a longer time. 4. Sensitise gels for 3 min in freshly prepared sensitiser solution. Sensitisation depends on the ambient temperature in the lab, at temperatures above 20°C sensitisation time must be shorter. For reproducible results stick to time! 5. Wash gels with distilled water for 2 min and another time with distilled water for 1 min. Stick to time! 6. Impregnation is done in freshly prepared silver impregnation solution for 30–45 min. Gels might become slightly yellowish (see Note 21). 7. Wash gels with distilled water for 2 min and another time with distilled water for 1 min. Stick to time! 8. Develop (do not forget to add formaldehyde to the prepared developer solution: 1.2 mL 37% formaldehyde to 3 L solution) gels in 1.5 L developer solution until the liquid turns yellow. Then quickly refresh developer and develop until the desired intensity is achieved. Normally this takes not more than 2 min (see Note 22). 9. Stop development transferring the gels to clean boxes with stopper solution, incubate for 20–60 min. 10. Store the gels in storage solution. Gels can be kept for a long time (at least for up to 1 year, best at 4°C) (see Note 23). 3.4.2. SYPRO™ Ruby Protein Staining

All steps are performed on a rocking mixer at room temperature, choose shaking frequencies that the gels are moving slightly but not damaged by their movements. For staining and washing, protect the container from light, until gels can be scanned. 1. Fix gels in large volume of fixing solution (approximately 10–20-fold gel volume) for 30 min, then change fixing solution and fix for another 30 min or longer.


109

2. Stain the gels with SYPRO™ Ruby gel stain for 12 h in a dark container (e.g. you can wrap the container with tinfoil). Up to five gels can be stained in one container. Staining can also be prolonged but normally there is no increase in sensitivity after 12 h of staining but an increase in speckles and background. Use 8–10 times of the gel volume of SYPRO™ Ruby stain. For qualitative stainings, SYPRO™ Ruby gel stain solution can be re-used although there might be slight loss in sensitivity. 3. Wash gels in a dark container with 10–20 times of the gel volume in washing solution for at least 30 min. Background can be reduced by additional exchange of washing solution and further washing for 30 min. For more detailed information, see also (15). 3.4.3. Coomassie Brilliant Blue G-250 Staining

Staining methods with Coomassie Brilliant Blue dyes are compatible with MS analysis, the protocol described below is often used as a standard procedure because of very simple handling and good sensitivity (see also Note 24). As with silver staining, incubation volumes are minimum 2.5 L per box with five gels. All incubation steps are performed preferably on a circular shaker. A “retainer” for gels as described in the silver staining part can be useful for exchange of solutions. 1. Fixation is done overnight or longer. Gels can also be kept for a long time in fixing solution prior to staining. 2. Wash the gel(s) three times with deionised water. 3. Stain gel(s) in Coomassie staining solution for at least several hours to days. No destaining is necessary. 4. For storage keep the gels in 5% acetic acid (see Note 25).

3.5. Data Evaluation and Image Analysis

For quantitative comparison of 2D gels, a minimum of four samples per group is an absolute requirement for statistical analysis of a data set, but in most cases more independent samples per group are needed to overcome general problems of biological variance within a group. A manual comparison of the gel patterns either within a single group of samples or the comparisons of samples belonging to different groups cannot be done without computational image analysis support because of the complexity of various inherent problems. To give a few examples, besides biological variance it is almost impossible to have zero variation in the samples because of deviations caused by (even very small) differences in sample preparation, gel properties and local disturbances in electrophoretic running conditions for instance. Furthermore, comparison of spots requires correct matching (and correct warping of gels), determination of local optical densities and for comparison of gels a normalisation of the absolute to relative optical densities. It is not the intention of this paper to explain

110

Smalla and Wyneken

details and all difficulties in 2D gel data interpretation but to express the need for computer-assisted evaluation of such data. Careful 2D gel analysis requires as a first step the generation of a digital image from the gels. For this, the scanner should have the option for (self-) calibration, a sufficient resolution and should be resistant to the chemicals in use. It is of advantage, if the scanner has options for the use of different optical filters in order to optimise the contrast of digital images in a suitable way. Several scanners meeting these prerequisites are commercially available. Even more important for 2D gel image analysis according to Good Laboratory Practice guidelines is the use of an image analysis software for processing and comparing these images which in addition always keeps the original optical density data unprocessed. Examples of software meeting these requirements are PDQuest (Bio-Rad) or ImageMaster™ 2D Platinum (GE Healthcare).

4. Notes 1. The purity of urea is extremely important since iso-cyanate, a urea break-down product, covalently modifies lysine residues. This reaction (carbamylation) induces changes in the isoelectric point. It depends mainly on temperature; therefore, avoid elevated temperatures during sample preparation when urea breakdown products are present. 2. IEF-equipments from other companies can be also used, e.g. the Protean IEF Cell from Bio-Rad. 3. We used reagents from the companies indicated. Ultrapure reagents are available from various companies and may also work very well. However, purity of reagents is very critical for the quality of 2D gel separations and therefore purity should be best specified as suitable for proteomics. 4. Dissolve 182 g Tris in approximately 700 mL deionised water, adjust pH with 6 N HCl (approximately 50 mL) and add deionised water up to 1,000 mL. 5. Ready-to-use stock solutions from other companies are suitable as well, e.g. from Bio-Rad (Cat.-No. 161-0158). 6. Final concentrations for the SDS electrophoresis running buffer are 25 mM Tris, 192 mM Glycine, 0.1% SDS. Thus, for a tenfold concentrated stock solution, use 30.3 g Tris, 144.0 g Glycine and 10.0 g SDS and add deionised water to a total volume of 1 L. It is important NOT to adjust the pH of this solution! 7. There are several computer programs for image analysis available as commercial products or as freeware (see also Sect. 3.5).


111

8. Besides in-house developed sample preparation clean-up protocols are also commercially available. Sample clean-up kits can be used, e.g. from GE Healthcare, Bio-Rad or Thermo Scientific Pierce. 9. If possible it is always better to start with higher total protein amount than later required for the 2D gel separations. This will sometimes lead to higher protein concentrations in the solubilised sample but this way it is easier to adjust or standardise equal sample loading. Samples in sample buffer can be stored finally up to 1 year at −80°C but after thawing they must be (as freshly prepared samples) centrifuged for 30 min at 20,000×g or higher to remove insoluble material. 10. For strips with other length, the protocols can be adapted according to the guidelines from GE Healthcare (12). In most cases, separation is improved when taking slightly longer focussing times than those recommended therein. 11. IPG strips are normally stored at −20°C. Before using the strips, give them a few minutes time to reach room temperature, otherwise the sample buffer will immediately crystallise when the strips are too cold. 12. Treatment with iodoacetamide is very important since excess DTT can be responsible for point streaking. 13. Before starting equilibration of samples, make sure that the second dimension gels are ready to use. Also prepare second dimension running buffer and precool it to 16°C at this point. 14. When printing labels, use a laser printer not an inkjet printer. For later orientation, it is useful to put the labels always in the sameposition; we put it always on the left side, where we also have always the acidic end of the strip. 15. It is important to use equal volumes of overlay solution. 16. If you are planning to run the second dimension immediately after casting, prepare now and not later the second dimension electrophoresis buffer and precool it to 16°C. 17. This alignment of the IPG strips is very important to avoid vertical double spotting. 18. Special care is required since the glass plates slip easily once your hands are immersed in tank buffer. 19. Never use metal tools for this purpose, it will produce cracks in the glass plates (14). We are using for that purpose the “wonder wedge” from Hoefer (Product Code: SE1514). 20. Another powerful approach for quantitative proteomics can be 2D Fluorescence difference gel electrophoresis (DIGE) in which the two samples to be compared are first labelled each with a different fluorescent dye, then mixed and run

112

Smalla and Wyneken

on the same 2D gel. Detailed information can be found at manufacturers websites (15). 21. Collect the used impregnation solution in a separate waste container. 22. If five gels are developed within one box, gels have to be shaken more vigorously while watching the first incubation, best do that manually. If other than duracryl gels are used, reduce the number of gels per box because they are mechanically less stable (at least only two per box). 23. Digital gel images for quantitation are best acquired by scanning later the day after staining, afterwards there is a slight loss in staining contrast. 24. However, some commercial staining solutions as, e.g. Imperial Protein Stain (Thermo Scientific Pierce) might have higher sensitivity and can be used as well. 25. Destaining to reduce background is normally not required but can be achieved by shaking the gels in deionised water for several hours. Gels can be stored for several months in 5% acetic acid at 4°C.

Acknowledgements This work was supported by bilateral programs of the Deutsche Forschungsgemeinschaft (DFG 444 CHL-113/32/0-1) and the Bundesministerium für Bildung und Forschung (BMBF CHL 06/027) with Conicyt (Chile), Proyecto Anillo 09-06 (PBCT, Conicyt, Chile) and the European Union (ZVOH-TP3/2, ZVOH 6/1). References 1. Klose J (1975) Protein mapping by combined isoelectric focusing and electrophoresis of mouse tissues. A novel approach to testing for induced point mutations in mammals. Humangenetik 26:231–243. 2. O’Farrell PH (1975) High resolution twodimensional electrophoresis of proteins. J Biol Chem 250:4007–4021. 3. Schägger H, von Jagow G (1991) Blue native electrophoresis for isolation of membrane protein complexes in enzymatically active form. Anal Biochem 199:223–231. 4. Wittig I, Braun HP, Schägger H (2006) Blue native PAGE. Nat Protoc 1:418–428. 5. Hartinger J, Stenius K, Hogemann D, Jahn R (1996) 16-BAC/SDS-PAGE: a two-dimensional

gel electrophoresis system suitable for the separation of integral membrane proteins. Anal Biochem 240:126–133. 6. Burre J, Volknandt W (2007) The synaptic vesicle proteome. J Neurochem 101:1448–1462. 7. Wenge B, Bonisch H, Grabitzki J, Lochnit G, Schmitz B, Ahrend MH (2008) Separation of membrane proteins by two-dimensional electrophoresis using cationic rehydrated strips. Electrophoresis 29:1511–1517. 8. Suzuki T (2011) Isolation of synapse subdomains by subcellular fractionation using sucrose density gradient centrifugation. Chapter 4 of this book. 9. Carlin RK, Grab DJ, Cohen RS, Siekevitz P (1980) Isolation and characterization of

Two-Dimensional Gel Electrophoresis-Based Proteomic Analysis of Brain Synapses postsynaptic densities from various brain regions: enrichment of different types of postsynaptic densities. J Cell Biol 86:831–845. 10. Smalla KH, Matthies H, Langnase K, Shabir S, Bockers TM, Wyneken U, Staak S, Krug M, Beesley PW, Gundelfinger ED (2000) The synaptic glycoprotein neuroplastin is involved in long-term potentiation at hippocampal CA1 synapses. Proc Natl Acad Sci USA 97:4327–4332. 11. Shevchenko A, Wilm M, Vorm O, Mann M (1996) Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem 68:850–858. 12. Görg A (2004) 2-D Electrophoresis: Principles and Methods. http://www5gelifesciencescom/aptrix/upp00919nsf/Content/4EE0

113

93B6C6B7BB18C1257628001D252E/$file /80642960pdf. 13. Heukeshoven J, Dernick R (1985) Simplified method for silver staining of proteins in polyacrylamide gels and the mechanism of silver staining. Electrophoresis 6:103–112. 14. Swatton JE, Prabakaran S, Karp NA, Lilley KS, Bahn S (2004) Protein profiling of human postmortem brain using 2-dimensional fluorescence difference gel electrophoresis (2-D DIGE). Mol Psychiatry 9:128–143. 15. GE Healthcare Web page http://www1gelifesciencescom/aptrix/upp00919nsf/Content/ Proteomics%20DIGE%7EProteomics%20 DIGE%20Introduction?OpenDocument&hom etitle=Proteomics or DyeAGNOSTICS Web page: http://www.dyeagnostics.com/wp/?page_id=6.

wwwwwwwwwwwwwwwwww

Chapter 9 Two-Dimensional BAC-SDS Polyacrylamide Gel Electrophoresis for the Fractionation and Identification of Synaptic Vesicle Proteins and the Presynaptic Active Zone Joern Barth and Walter Volknandt Abstract Recent progress in mass spectrometry allowed the identification of numerous proteins in complex soluble mixtures; however, the analysis of membranous subproteomes is still a challenging task. Integral membrane proteins, in particular those with more than one transmembrane domain, have been difficult to separate by 2D electrophoretic methods due to their tendency to aggregate. The conventionally used 2D technique based on isoelectric focussing in the first-dimension and SDS-PAGE in the second-dimension is a powerful method to separate soluble protein samples, albeit it has major drawbacks in the separation of hydrophobic proteins. An alternative electrophoretic technique especially suitable for the analytical and preparative separation of membrane proteins is 2D 16-BAC/SDS electrophoresis. The first-dimension involves discontinuous gel electrophoresis towards the cathode in an acidic buffer system using the cationic detergent benzyldimethyl-n-hexadecylammonium chloride (16-BAC). The separation in the second-dimension towards the anode is based on the anionic detergent SDS. Using this system in combination with subsequent MALDI-TOF mass spectrometry enabled us to identify the proteinaceous inventory of synaptic vesicles from rat brain. Moreover, applying these techniques to plasma membranedocked synaptic vesicles allowed us to identify numerous proteins of the presynaptic active zone. Here, we describe the application of 2D 16-BAC-SDS polyacrylamide gel electrophoresis for the fractionation and identification of synaptic vesicle proteins and the presynaptic active zone. Key words: 16-BAC, Protein separation, 2D gel electrophoresis, Synaptic vesicles proteins, Active zone, Cationic detergents, SDS-PAGE

1. Introduction 1.1. Synaptic Vesicles and the Active Zone

The nerve terminal proteome governs neurotransmitter release as well as the structural and functional dynamics of the presynaptic compartment. Neurotransmission is based primarily on the


115

116

Barth and Volknandt

regulated release of neurotransmitters from synaptic vesicles. Synaptic vesicles carry out essential presynaptic tasks such as the uptake, storage, and Ca2+-regulated release of neurotransmitters. Upon arrival of an action potential, synaptic vesicles docked and primed at the active zone of the presynaptic plasma membrane fuse with the membrane and release neurotransmitters into the synaptic cleft. Neurotransmitters then diffuse to the postsynaptic membrane to activate neurotransmitter receptors. Synaptic vesicles are retrieved by endocytosis to restore the primed vesicle pool (1, 2). These essential tasks are controlled by a unique inventory of integral membrane proteins and also of membrane-associated proteins that attach and detach during the vesicle cycle in a timedependent manner (reviewed in (3, 4)). The exploration of the proteinaceous inventory of synaptic vesicles and the active zone is essential for an in depth understanding of the molecular mechanisms of neurotransmitter release. 1.2. Subcellular Fractionation and Immunoisolation of Synaptic Proteins

In order to identify specific presynaptic subproteomes, we used subcellular fractionation and a monoclonal antibody against the synaptic vesicle protein SV2 for immunoaffinity purification of two major synaptosome-derived synaptic vesicle-containing fractions: one sedimenting at lower and one sedimenting at higher sucrose density. The less dense fraction contains free synaptic vesicles, the denser fraction synaptic vesicles as well as components of the presynaptic membrane compartment. These immunoisolated fractions were analyzed using the cationic BAC polyacrylamide gel system in the first and SDS-PAGE in the second-dimension. Proteins spots were subjected to analysis by matrix assisted laser desorption ionization time of flight mass spectrometry (MALDITOF MS). We identified a large variety of proteins in the free vesicle fraction and in the plasma membrane-containing denser fraction. Synaptic vesicles contain a considerably larger number of protein constituents than previously anticipated. The plasma membrane-containing fraction contains synaptic vesicle proteins, components of the presynaptic fusion, and retrieval machinery and numerous other proteins potentially involved in regulating the functional and structural dynamics of the nerve terminal.

1.3. Protein Separation and Identification

In proteomic studies using conventional 2D electrophoresis (5), integral membrane proteins are notoriously underrepresented (6, 7). On a global proteomic level emphasizing cytosolic proteins, numerous modifications comprise refinements of the conventional 2D polyacrylamide gel electrophoresis (PAGE) (8, 9). Several efforts have been reported to overcome the limitations of the original procedure for example by the introduction of immobilized pH gradients. Despite these attempts IEF still has the disadvantage that it cannot accommodate large amounts of protein or resolve proteins which are poorly soluble in nonionic detergents. Furthermore, proteins which are posttranslationally

Two-Dimensional BAC-SDS Polyacrylamide Gel Electrophoresis for the Fractionation

117

modified (charged groups or phosphorylation) may migrate as a family of spots. Even if solubilization is successful, proteins often aggregate and precipitate around pH values close to their isoelectric point. In addition, heterogeneity in charge as often found in glycosylated membrane proteins leads to streaking and smears in the first-dimension (see Fig. 1). The development of techniques especially designed for the separation of hydrophobic proteins is thus an essential task (reviewed in (10, 11)).

Fig. 1. Silver staining of synaptic vesicle proteins separated by IEF/SDS-PAGE. 60 mg of proteins were loaded.

118

Barth and Volknandt

1.4. Two-Dimensional 16-BAC/SDS Gel Electrophoresis

In 1983 Macfarlane (12) developed a novel electrophoretic system to separate methylated proteins at acidic pH by evaluating a variety of cationic detergents including benzyldimethyl-n-hexadecyl ammonium chloride (16-BAC). Later on Macfarlane (13) improved this technique for the separation of a complex protein mixture in high capacity and at high resolution. Hartinger et al. (14) optimized this method for the separation of synaptic vesicle proteins. Application of this method leads to the clear focusing of integral membrane proteins in individual spots (see Fig. 2). 16-BAC gels allow for high loading capacity of up to 100 mg protein, as well as high resolution power for complex protein mixtures. In particular, integral membrane proteins with more than one transmembrane domain can be resolved almost quantitatively by the cationic 16-BAC in the first- and the anionic SDS in the second-dimension (15–17). Furthermore, DNA/RNA does not need to be removed from the sample. Thus, whole cells or nuclei can be extracted directly with sample buffer (13). Another advantage is the separation of proteins with charge heterogeneity as well as proteins with several posttranslational modifications. In addition, 16-BAC/SDS-PAGE is easy to handle and highly reproducible (see Fig. 3). 16-BAC/SDS-PAGE is an

Fig. 2. Immunodetection of the integral membrane protein synaptobrevin-2 after IEF/ SDS-PAGE (left) or 16-BAC/SDS-PAGE (right) separation. In both cases 10 mg of proteins were loaded.

Fig. 3. Comparison of two independent 16-BAC/SDS-PAGE experiments with immuno isolated synaptic vesicles docked to the plasma membrane stained with Coomassie Blue. Apparent molecular mass markers are shown on the left of each gel.


119

Fig. 4. Illustration of synaptic vesicle proteins separated by 16-BAC/SDS-PAGE and identified by mass spectrometry and/ or Western blots. Not all identified protein spots are highlighted.

ideal tool for 2D immunodetection of hydrophobic proteins (18, 19) and for the analysis by 2D difference gel electrophoresis (20). MALDI-TOF MS analysis of individual protein spots identified 72 synaptic vesicle proteins (see Fig. 4) and 81 proteins in the active zone. Only, less than 4% of protein spots escaped identification.

2. Materials The products used are listed below. Comparable products from other suppliers should also be effective. 2.1. Sample Preparation

Synaptic vesicles or active zone membranes from rat brain were immunoisolated using magnetic beads (Dynabeads M-280, Dynal) as described in detail (18, 20). 1. Use methanol and chloroform for protein precipitation. 2. 16-BAC solubilizing buffer consists of 2.95 M urea (Roth), 5% 16-BAC (w/v, Sigma), 5% glycerol (Roth), 37.5 mM

120

Barth and Volknandt

dithio-dl-threitol (AppliChem), and 0.025% Pyronin Y (w/v) (Sigma-Aldrich). 3. For samples which will not been precipitated prepare twofold concentrated 16-BAC solubilizing buffer. 2.2. 16-BAC Gel Electrophoresis (First-Dimension)

1. Mini-slab gels (Biometra), power supply, combs, glass plates, and samples. 2. The composition of separating gel (8%, approximately 40 mL) is the following: 7.5 g urea, 8 mL 40% acrylamide solution (29:1; Roth), 2.4 mL 1% N,N ¢-methylenebisacrylamide (Roth), 10 mL 300 mM potassium phosphate buffer pH 2.1, 11 mL H2O (double distilled; see Note 1), 2 mL 80 mM ascorbic acid (Merck), 64 mL 5 mM ferrous sulfate (Sigma), 0.4 mL 250 mM 16-BAC, and 1.6 mL H2O2 (1:1,200, diluted from a 30% stock solution, AppliChem). 3. Composition of stacking gel: 1 g urea, 1 mL 40% acrylamide (29:1), 2.4 mL 1% N,N ¢-methylenebisacrylamide, 2.5 mL 500 mM potassium phosphate buffer pH 4.1, 3 mL H2O (double distilled), 0.5 mL 80 mM ascorbic acid, 70 mL 250 mM 16-BAC (w/v), and 0.5 mL H2O2 (1:750, diluted from a 30% stock solution). 4. The electrode buffer contains: 2.5 mM 16-BAC, 150 mM glycine (Roth), 50 mM phosphoric acid (Roth).

2.3. Staining, Fixation, and Reequilibration of Separated Proteins

1. Fixative contains: 35% isopropanol (v/v) and 10% acetic acid (v/v, Roth). 2. Staining solution: 0.15% Coomassie Brilliant Blue R250 (w/v; Bio-Rad) in fixative. 3. Reequelibration in 100 mM Tris/HCl, pH 6.8 (Roth) and in SDS sample buffer consisting of 1% glycerol (w/v), 2% SDS (w/v, AppliChem), 0.08 M Tris/HCl, pH 6.8, 0.0012% bromophenol blue (w/v; Serva), 100 mM 1,4-dithio-dl-threitol (AppliChem).

2.4. SDS-PAGE (Second-Dimension)

The same experimental setup and materials like minigels, combs etc. as for the first-dimension can be employed in the seconddimension. 1. Separating gel (e.g., 15%, approximately 15 mL): 7.5 mL 30% acrylamide (29:1), 1.3 mL 1% N,N¢-methylenebisacrylamide, 5.6 mL 1 M Tris/HCl, pH 8.7, 0.5 mL H2O (double distilled), 75 mL 20% SDS (w/v), 15 mL TEMED (Roth), and 75 mL 10% ammonium persulfate (w/v; Roth). 2. Stacking gel (5%, approximately 5 mL): 835 mL 30% acrylamide, 650 mL 1% N,N ¢-methylenebisacrylamide, 2.5 mL 0.25 M Tris/HCl, pH 6.8, 975 mL H2O (double distilled), 25 mL


121

20% SDS (w/v), 5 mL TEMED, and 25 mL 10% ammonium persulfate (w/v). 3. Electrophorese buffer: 100 mM Tris, 100 mM glycine, 70 mM SDS, mixture of all components results in pH 8.0. For mass spectrometry individual protein spots have to be excised using sterile scalpel blades.

3. Methods 3.1. Sample Preparation

1. Prepare one- or twofold 16-BAC solubilizing buffer directly (see Note 2) before use. The buffer needs to be stored at 60°C to avoid solidification. 2. Samples can be solubilized directly (see Note 3) by adding an equal volume of the twofold concentrated 16-BAC solubilizing buffer. Alternatively, samples (100 mL) can be precipitated using a fourfold volume (400 mL) of methanol (see Note 4). After vortexing, samples are centrifuged for 1 min at 9,000×g. Then add chloroform (200 mL) to yield a ratio of 2:1 methanol:chloroform, vortex, and recentrifuge. To induce phase separation add H2O (300 mL) at a ratio of 2:1:1.5 methanol:chloroform:H2O. Carefully discard the upper phase containing the methanol, and add 300 mL of methanol, vortex, and recentrifuge for 10 min at 9,000×g. Remove the supernatant and air-dry the protein pellet prior to solubilization in 16-BAC buffer. After mixing, the material should be incubated at 60°C for 5 min (see Note 5) and then loaded on top of the first-dimension gel.

3.2. 16-BAC Discontinuous Gel (First-Dimension)

16-Bac gels were prepared in slab minigels (see Note 6) similarly to conventional SDS-PAGE (see Note 7). 1. Prepare separation gels by mixing the compounds described in materials. It is important to note, that the ascorbic acid and ferrous sulfate solutions have to be freshly prepared (see Note 8). Initiate polymerization by adding H2O2. To obtain a smooth gel surface and to avoid dehydration carefully overlay the separating gel with 75 mM potassium phosphate buffer, pH 2.1 (see Note 9). After polymerization (approximately 30 min) decant potassium phosphate buffer. 2. For the stacking gel solution prepare ascorbic acid and ferrous sulfate solutions freshly. The stacking gel solution is poured onto the top of the separation gel and the appropriate comb is inserted. Polymerization is complete after 30 min (see Note 10).

122

Barth and Volknandt

3. After washing the wells with the acidic electrode buffer load the samples. 4. For the first-dimension electrophoresis towards the cathode set the initial current to 10 mA until the dye front enters the separating gel, then switch to 20 mA until the dye front has left the gel completely or the Schlieren line is about to leave the gel (approximately 2.5 h). 3.3. Fixation, Staining, and Reequilibration

1. Fix proteins for 30 min in 35% isopropanol and 10% acetic acid (see Note 11). 2. Subsequently, stain the gel with Coomassie R250 (see Note 12) and cut the gel strips of interest carefully without touching the gel (see Note 1). 3. For reequilibration incubate the gel strips three times for 10 min in 100 mM Tris/HCl, pH 6.8. Prior to SDS-PAGE incubate the gel strips for 5 min in SDS sample buffer.

3.4. SDS-PAGE (Second-Dimension)

1. For the discontinuous SDS-PAGE, either uniform 10 or 15%, see materials, or linear 4–15% acrylamide concentrations, overlay the separating gel with stacking gel solution. 2. Prepare the stacking gel with a large well to accommodate the gel strip from the first-dimension and an additional small well for the molecular mass marker. 3. Place the equilibrated gel strip vertically into the large well such that it touches the separating gel without any gaps (see Fig. 1). 4. Subsequently, overlay the gel with fivefold concentrated sample buffer and incubate for 5 min. 5. Perform the electrophoresis towards the anode with an initial current of 5 mA until the second dye front (bromophenol blue) reaches the separating gel. The first dye front is Coomassie Blue eluted from the equilibrated gel strip (see Note 13). Separation is carried out at 20 mA until the second dye front is about to leave the gel. 6. Visualization of protein spots is achieved using Coomassie or a mass spectrometry compatible silver stain (21).

4. Notes 1. For the identification of proteins by mass spectrometry it is of uttermost importance to avoid contamination of samples or material by hair or skin particles. Protein identification is often biased by contamination of samples with keratins. It is


123

recommended to work under sterile conditions. In addition, wear a plastic lab coat, long gloves reaching up to the elbows, cover the hair with a plastic hat, and wear a mouth mask. 2. Prolonged storage of the buffer (more than 1 day), boiling the sample, or storage of the protein sample in sample buffer (including frozen storage) leads to significant loss of resolution and protein degradation and should be avoided. 3. High concentration of salts (e.g., 2 M NaCl) or nonionic detergents (e.g., 1% Triton-X) in the sample do not interfere with the resolution of proteins. 4. For protein precipitation avoid trichloracetic acid and perchloric acid because both precipitate with 16-BAC, and should not be used unless they can be removed quantitatively. Methanol/chloroform precipitation according to Wessel and Flügge (22) can be used alternatively. 5. Samples in 16-BAC solubilizing buffer should not be incubated for longer time periods than 5 min and avoid higher temperatures than 60°C because membrane proteins then tend to aggregate fast. 6. Do not use tube gels for first-dimension electrophoresis since they are difficult to remove from the tubes. 7. Molecular weight standards can be reduced by heating them briefly at neutral pH with 2-mercaptoethanol in 16-BAC and urea. 8. Use hydrogen peroxide, ferrous sulfate, and ascorbic acid in 16-BAC gels instead of ammonium peroxide as a polyacrylamide polymerizing reagent, which precipitates with cationic detergents (23). 9. The separating 16-BAC gel should have a rather acidic pH between 1.5 and 2.1, because at higher (e.g., pH 4.0) proteins migrate considerably more slowly, so that even at low concentrations the low molecular weight proteins do not migrate at or near the line of discontinuity (12). 10. Better results have been reported for overnight polymerization of 16-BAC gels by Hartinger and coworkers (14). 11. To assess the separation quality and to avoid loss of protein fixation is performed for 30 min in 35% isopropanol and 10% acetic acid prior to SDS-PAGE. 12. Cationic detergents have the propensity to precipitate Coomassie blue, nevertheless Coomassie blue stains 16-BAC gels in about half the time needed for SDS gels and it stains protein bands more intensely. 13. Transfer efficacy of proteins from the first- to the seconddimension can be evaluated by silver staining of the gel strip.

124

Barth and Volknandt

Acknowledgments We are indebted to Drs Jaqueline Burré and Marco Morciano for sharing data with us. References 1. Littleton, J. T. (2006) Mixing and matching during synaptic vesicle endocytosis. Neuron 51, 149–151. 2. Ryan, T. A. (2006) A pre-synaptic to-do list for coupling exocytosis to endocytosis. Curr. Opin. Cell Biol. 18, 416–421. 3. Li, L., Chin, L.-S. (2003) The molecular machinery of synaptic vesicle exocytosis. Cell. Mol. Life Sci. 60, 942-960. 4. Südhof, T. C. (2004) The synaptic vesicle cycle. Annu. Rev. Neurosci. 27, 509–547. 5. O’Farrell, P. H. (1975) High resolution twodimensional electrophoresis of proteins. J. Biol. Chem. 250, 4007–4021. 6. Rabilloud, T., Blisnick, T., Heller, M., Luche, S., Aebersold, R., Lunardi, J., Braun-Breton C. (1999) Analysis of membrane proteins by twodimensional electrophoresis: comparison of the proteins extracted from normal or Plasmodium falciparum-infected erythrocyte ghosts. Electrophoresis 20, 3603–3610. 7. Santoni, V., Molloy, M., Rabilloud T. (2000) Membrane proteins and proteomics: un amour impossible? Electrophoresis 21, 1054–1070. 8. Kashino, Y. (2003) Separation methods in the analysis of protein membrane complexes. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 797, 191–216. 9. Luche, S., Santoni, V., Rabilloud T. (2003) Evaluation of nonionic and zwitterionic detergents as membrane protein solubilizers in twodimensional electrophoresis. Proteomics 3, 249–253. 10. Burré, J., Volknandt, W. (2007) The synaptic vesicle proteome. J. Neurochem. 101, 1448–1462. 11. Braun, R. J., Kinkl, N., Beer, M., Ueffing, M. (2007) Two-dimensional electrophoresis of membrane proteins. Anal. Bioanal. Chem. 389, 1033–1045. 12. Macfarlane, D. E. (1983) Use of benzyldimethyl-n-hexadecylammonium chloride (“16BAC”), a cationic detergent, in an acidic polyacrylamide gel electrophoresis system to detect base labile protein methylation in intact cells. Anal. Biochem. 132, 231–235.

13. Macfarlane, D. E. (1989) Two dimensional benzyldimethyl-n-hexadecylammonium chloride → sodium dodecyl sulfate preparative polyacrylamide gel electrophoresis: a high capacity high resolution technique for the purification of proteins from complex mixtures. Anal. Biochem. 176, 457–463. 14. Hartinger, J., Stenius, K., Högemann, D., Jahn, R. (1996) 16-BAC/SDS-PAGE: a twodimensional gel electrophoresis system suitable for the separation of integral membrane proteins. Anal. Biochem. 240, 126–133. 15. Peters, C., Bayer, M. J., Bühler, S., Andersen, J. S., Mann, M., Mayer, A. (2001) Transcomplex formation by proteolipid channels in the terminal phase of membrane fusion. Nature 409, 581–588. 16. Yamaguchi, Y., Miyagi, Y., Baba, H. (2008) Twodimensional electrophoresis with cationic detergents, a powerful tool for the proteomic analysis of myelin proteins. Part 1: technical aspects of electrophoresis. J. Neurosci. Res. 86, 755–765. 17. Zahedi, R. P., Meisinger, C., Sickmann, A. (2005) Two-dimensional benzyldimethyl-nhexadecylammonium chloride/SDS-PAGE for membrane proteomics. Proteomics 5, 3581–3588. 18. Morciano, M., Burré, J., Corvey, C., Karas, M., Zimmermann, H., Volknandt, W. (2005) Immunoisolation of two synaptic vesicle pools from synaptosomes: a proteomics analysis. J. Neurochem. 95, 1732–1745. 19. Morciano, M., Beckhaus, T., Karas, M., Zimmermann, H., Volknandt, W. (2009) The proteome of the presynaptic active zone: from docked synaptic vesicles to adhesion molecules and maxi-channels. J. Neurochem. 108, 662–675. 20. Burré, J., Beckhaus, T., Schägger, H., Corvey, C., Hofmann, S., Karas, M., Zimmermann, H., Volknandt, W. (2006) Analysis of the synaptic vesicle proteome using three gel-based protein separation techniques. Proteomics 6, 6250–6262. 21. Rais, I., Karas, M., Schägger, H. (2004). Two-dimensional electrophoresis for the isolation of integral membrane proteins and mass

Two-Dimensional BAC-SDS Polyacrylamide Gel Electrophoresis for the Fractionation spectrometric identification. Proteomics 4, 2567–2571. 22. Wessel, D., Flügge, U. I. (1984) A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal. Biochem. 138, 141–143.

125

23. Eley, M. H., Burns, P. C., Kannapell, C. C., Champbell, P. S. (1979) Cetyltrimethylammo nium bromide polyacrylamide gel electrophoresis: estimation of protein subunit molecular weights using cationic detergents. Anal. Biochem. 181, 411–419.

wwwwwwwwwwwwwwwwww

Chapter 10 iTRAQ-Based LC-LC MALDI TOF/TOF MS Quantitative Analysis of Membrane Proteins from Human Glioma Uroš Rajčević Abstract Various proteomic approaches are being applied in brain tumor proteomics with regard to targeted proteins of interest, to discover phenotype specific markers which could facilitate diagnosis as well as potential antitumor drug targets. iTRAQ technology is a multiplexing protein quantitation strategy that provides relative and absolute measurements of protein abundance in complex mixtures based on the differential labeling of the proteins. Combined with membrane enrichment methodology, separation of the labeled peptides by two-dimensional liquid chromatography linked to tandem mass spectrometry and adequate data-mining, it can provide an excellent tool in search for novel isoform- and species-specific biomarkers and drug targets in various biotechnological and biomedical applications. Glioblastoma (GBM) is the most frequent primary brain tumor diagnosed in adults and remains one of the most lethal forms of human cancer. No biomarkers can distinguish different cell populations within GBMs or predict the potential of low grade gliomas to develop into malignant angiogenic gliomas. In our study, we have used iTRAQ-based technology to search for novel biomarkes of GBM. Key words: iTRAQ, Glioblastoma, Biomarkers, Glioma invasion, Angiogenesis, Subproteome, Quantitative mass spectrometry, 2D-LC

1. Introduction 1.1. Proteomics and Mass Spectrometry

Rapid developments in genomics and proteomics in the past two decades have introduced novel technologies enabling a thorough insight into the functional effectors of cellular processes. In biomedicine, these technologies lead to the identification of biological markers which may provide the starting point for the development and identification of diagnostic tests and therapeutic targets. To identify and validate reliable tumor markers within the proteome, it is necessary, prior to tandem mass spectrometry, to reduce sample complexity. This can be done by robust fractionation and separation techniques. Proteomics represents the large-scale


127

128

Rajčević

analysis of protein expression, posttranslational modifications and protein–protein interactions, thereby providing a link between gene and cellular function. Mass spectrometry (MS), in which biomolecules are ionized and their mass is measured following their specific trajectories in a vacuum system, has become the method of choice for the detection, identification, and quantitation of proteins in complex biological samples (reviewed in (1)). 1.2. Quantitative MS and LC-MS Using Sample Labeling Approaches

In quantitative proteomics, quantifying changes in protein expression between biological samples is a key requirement to determine the differences in cell state at the molecular level. Numerous methods for relative quantification have been developed using either stable isotope labels or chemical labels (reviewed in (2)). In these methods proteins or peptides are differentially labeled by isotopically “light” and “heavy” labels. The samples are pooled, separated by (usually) several dimensions of liquid chromatography (LC), and peptides common to all the samples appear in mass spectra as peak pairs (triplets, etc.), differing in mass determined by isotope label used. Molecular ion intensity as determined by peak heights or peak areas is compared to an estimated abundance of the peptides between the compared samples. Here, we compare four labeling methods used in the quantitative proteomics with regards to the label type, labeling chemistry, and the possible applications.

1.2.1. Stable Isotope Labeling of Amino Acids in Cell Culture

Stable isotope labeling of amino acids in cell culture (SILAC) is a metabolic labeling approach allowing the combination of up to three isotopic labels (e.g., deuterium, 13C, 15N) and therefore combining up to three samples in an assay. Amino acids containing stable isotopes are supplemented to cell culture media (3), or fed to laboratory animals (4) and metabolically incorporated in the native proteins. Isotope labeled peptides are identified and quantified in the MS spectra as precursor ion pairs (or triplets) differing in mass by known amount and their relative abundance is measured by comparing peak intensities or areas. This approach has been used to identify protein–protein interactions involved in EGFR signaling (5) and in insulin-dependent GLUT4 interactions (6). Thus, SILAC makes an excellent approach to be used with in vitro and in vivo laboratory models requiring up to three protein abundance comparisons in an assay.

1.2.2. Heavy Water Labeling

Tryptic degradation of proteins in the presence of light (H2O16) or heavy (H2O18) water results in enzymatic labeling of peptides with light (16O) or heavy (18O) oxygen by natural exchange of two oxygen atoms from the carboxyl terminus of the degraded peptides with two oxygen atoms from the surrounding water molecules (7). The heavy oxygen isotope labeled peptide differs from light oxygen isotope labeled peptide by 4 Da and can be detected by a corresponding mass shift in the mass spectrum.

iTRAQ-Based Analysis of Membrane Proteins from Human Glioma

129

1.2.3. Isotope-Coded Affinity Tags

The isotope-coded affinity tags (ICAT) labels (8) are biotin-tagged chemicals that are added onto the cysteine residues in proteins. Protein samples are labeled with isotopically different labels, combined, and digested. The biotin tag allows for enrichment of ICAT labeled peptides by affinity chromatography and thus reducing sample complexity, while keeping high protein coverage, a clear advantage over other techniques described in this section. Enriched sample is separated by (multidimensional) separation techniques and analyzed by either electro spray ionization tandem mass spectrometry (ESI MS/MS) or matrix assisted laser desorption/ionization tandem mass spectrometry (MALDI MS/MS). Co-immunoprecipitation of proteins followed by analysis using ICAT LC-MS/MS technology has identified differences in the composition of subunits of 20 S proteosomes (9), proteins interacting with transcription factors and large polymerase II preinitiation complex (10, 11) and the mSin3 corepressor complex (12). Organelle purification in conjunction with ICAT labeling and nanoLC-MS/MS provided a first comprehensive description of the protein constituents in the postsynaptic density of rat brain (13).

1.2.4. Isobaric Tags for Relative and Absolute Quantification of Peptides

Isobaric tags for relative and absolute quantification (iTRAQ) technology is a multiplexing protein quantitation strategy that provides relative and absolute measurements of proteins in complex mixtures (14). In this method, isobaric tags, different for each sample, are covalently linked to amine residues on peptides (N-terminal amine and lysine side chain amine). Identical peptides labeled with the different iTRAQ reagents exhibit the same parent ion in MS. Differentially labeled intact peptide masses are indistinguishable in MS mode, but upon fragmentation of the attached tag in MS/MS mode produce diagnostic fragment peaks as low molecular mass reporter ions that provide relative quantitative information on proteins and enable simultaneous identification and quantitation. iTRAQ reagent strategy currently contains up to eight isotopically distinct tags (a big advantage over three in SILAC and two in ICAT or heavy water) enabling time course and disease progression experiments in a multiplexed fashion. Straightforward labeling protocol, efficient and ubiquitous labeling of the peptides has rendered iTRAQ increasingly popular. Using an internal standard in the samples allows also an absolute quantitation. iTRAQ technology has been applied to differential protein profiling in lung cancer where 51 differentially expressed proteins were found during epithelial-mesenchymal transition (15). In a search for biomarkers in endometrial cancer, nine potential candidates have been discovered, using a combination of differentially labeled tags, iTRAQ, and cleavable ICAT (cICAT) (16). iTRAQ fourplex has also been used in a quantitative proteomics assay to reveal protein networks of hippocampal

130

Rajčević

synapses of CaMKII alpha mutant mice to find out a threefold decrease of calcium/calmodulin-dependent kinase II R in the synaptic membrane fraction of the 3¢UTR mutant mice. In the same study, using the quantified proteins a synaptic protein interaction network has been constructed and the synaptic proteins were organized into ten (interconnecting) functional groups belonging to the pre- and postsynaptic compartment (17). In glioma, iTRAQ technology has been used to screen the wild type (wt) EGFR and EGFRvIII mutant expressing glioblastomas to reveal the correlation between the EGFRvIII mutant positivity and decreased expression of collapsin response mediator protein 1. The authors showed that the loss of collapsin response mediator protein 1 contributes to the increased invasive phenotype of human GBMs expressing mutant EGFRvIII (18). In our own setup, the method presented here was used on a particular rat xenograft model of glioma developed by Sakariassen and collaborators (19). Since more than a third of all known biomarkers as well as more than two thirds of known and potential antitumor protein targets are membrane related proteins (20–23) membrane enriched fractions of brain tumor tissue are most valuable source of biomarker/drug target information and are thus used in this assay (Fig. 1). iTRAQ fourplex experiments were performed comparing light membrane fractions of different experimental phenotypes of glioma deriving from two patients, and plasma membranes of experimental phenotypes of glioma of two patients to reveal novel biomarkers and targets for antiangiogenic therapy (24). 1.3. Proteomics of Gliomas

Gliomas are histologicaly heterogeneous and invasive brain tumors showing glial-like morphologies/characteristics. The World Health Organization (WHO) classifies gliomas on the basis of histological features into four prognostic grades. Grade III and IV tumors are considered malignant gliomas. Glioblastoma (Grade IV glioma) is the most frequent primary brain tumor diagnosed in adults, with 12–15 months median survival of GBM patients. Therefore, this brain tumor remains one of the most lethal forms of human cancer. No underlying cause has been identified for malignant gliomas (Reviewed in (25, 26)). Invasion/infiltration and angiogenesis are two defining hallmarks of GBM largely responsible for the aggressive nature of the tumor. Highly infiltrative glioma cells escape surgical resection and lead to tumor recurrence. Limited oxygen supply in the tumor microenvironment is among the factors influencing recruitment of new blood vessels from preexisting vessels, a process termed angiogenesis. Absence of angiogenesis is considered a rate-limiting factor in solid tumors. High grade gliomas show extensive infiltration of the adjacent normal brain but they are also among the most vascularized neoplasms (27–29). Currently, no biomarkers can distinguish different cell populations within GBMs (e.g., tumor cells showing infiltrative growth from those triggering angiogenesis) or predict the propensity of


131

Fig. 1. Workflow of the procedure. Following the blending of the tissue samples in homogenization buffer, the homogenate is ultracentrifuged in sucrose step gradient resulting in three fractions of interest – the light membranes, the plasma membranes, and mitochondria. These are labeled with different iTRAQ labels and separated in two dimensions of liquid chromatography followed by MALDI TOF/TOF MS. The protein identification and quantification is performed using GPS software. Isoform-specific (and species-specific if necessary) proteins are separated by sequence differences. Finally, the data from different membrane fractions are merged and data mined. (*) iTRAQ labels are very useful in time-course experiments, where several time points of tumor development are differentially labeled and compared at a quantitative proteomic level.

low grade (nonangiogenic) gliomas to develop into malignant angiogenic gliomas. With all the known problems of two-dimensional gel-based glioma proteomics (Reviewed in (30)), liquid chromatography based separation of intact proteins (31–33) or peptides (34) followed by mass spectrometry has emerged not only as an alternative method yielding higher number of identified proteins

132

Rajčević

but also as a way to increase the slice of the proteome that can be routinely analyzed. Efforts have been made to minimize the normal (nontumor) tissue contamination of the tumors by microdissection of the samples prior to proteomics. This offers the opportunity to reduce the noise, to compare different regions of the same tumor (35, 36), or even to focus into particular features of the tumor (e.g., blood vessels (37)). Finally, the proteomics also enables us to study at a large scale, the posttranslational modifications (PTM) of proteins. These modifications, a consequence of the signals from the signaling pathways, have a crucial role in the activity of specific proteins. The information on the PTMs is thus essential to understand the underlying role of tumor biology of glioma and has a strong potential in the clinic in the future. One of the most studied examples of PTM analyses by proteomics includes EGFRvIII (38).

2. Materials The products used are listed below. Comparable products from other suppliers should also be effective. 2.1. Tissue Processing and Fractionation

1. Polytron homogenizer (Kinematica, Lucerne, Switzerland). 2. Homogenizing buffer 0.32 M sucrose, 5 mM HEPES, pH 7.4, protease inhibitor mix (Amersham Biosciences, Piscataway, NJ, USA) (store at 4°C up to 2 weeks). 3. Buffer containing 0.32 M Sucrose, 5 mM NaH2PO4, pH 8.1, protease inhibitor mix (Amersham Biosciences, Piscataway, NJ, USA) (store at 4°C up to 2 weeks). 4. 1.4 M sucrose in water (store at 4°C up to 2 weeks). 5. 1.1 M sucrose in water (store at 4°C up to 2 weeks). 6. OPTIMA™ L-100 K ultracentrifuge, SW 40 Ti rotor, SW 32.1 Ti rotor, 17 mL ultracentrifuge tubes (all Beckman Coulter, Fullerton, CA, USA). 7. 5 mM HEPES, pH 8.1 (store at 4°C). 8. Bradford reagent (Biorad, Hercules, CA, USA).

2.2. Sample Labeling

1. iTRAQ® Reagents Multiplex Kit (Applied Biosystems/MDS Sciex, Foster City, CA, USA). 2. Rapigest detergent solution (Waters, Milford, MA, USA). 3. Trypsin (Modified sequencing grade; Promega, Madison, WI, USA). 4. 1% (v/v) trifluoroacetic acid (TFA).


2.3. Two Dimensional Liquid Chromatography and MALDI-TOF/TOF Mass Spectrometry

133

1. Strong cation exchange (SCX) column (Poly SULFOETHYL Aspartamide), 100 × 2.1 mm, 3 mm, 300Å (Poly LC, Columbia, MD, USA). 2. Fraction collector (Dionex, Sunnyvale, CA, USA). 3. 0.1% (v/v) TFA. 4. Trap column (1 mm × 300 mm ID, C18; Dionex, Sunnyvale, CA, USA). 5. Capillary reverse phase (RP) column (reprosil, nano pur C18 Aqua 3 mm material, 100 m ID; 15 cm column). 6. MALDI matrix (6 mg/mL a-cyano hydroxy-cinnamic acid (Sigma, St. Louis, MO, USA) in 33.3 % (v/v) acetonitrile, 1 mM NH4 citrate). 7. Robot Microfraction Collector (Dionex, Sunnyvale, CA, USA). 8. 4800 MALDI TOF/TOF Analyzer, operated under 4000 Series Explorer v.3.5.1 software; MALDI target plate, 4700 Cal Mix containing six calibrant peptides (all Applied Biosystems/MDS SCIEX, Foster City, CA, USA).

3. Methods 3.1. Tissue Processing and Fractionation

1. Membrane fractions were prepared as described (39) (see also Chap. 4) with minor modifications. 2. Homogenize 80–150 mg of tissue samples at a time with a Polytron homogenizer at medium speed, 1.5 min each in a 10 mL/g tissue homogenizing (H) buffer and centrifuge the homogenate for 10 min at 1,000×g at 4˚C to remove the cell debris (see Notes 1 and 2). 3. Collect the supernatant 1 (S1). Add 0.7 mL H buffer to the pellet (P1), homogenize briefly with Polytron and spin 10 min 1,000×g at 4°C. Remove supernatant (S¢) and pool it with S1. 4. Centrifuge S1 + S¢ for 30 min at 4°C at 16,000×g. 5. Homogenize the resulting pellet (P2) in 0.7 mL of the same buffer and centrifuge P2 for 30 min at 4°C, 20,000×g on a tabletop centrifuge. 6. Discard S2¢. Resuspend the resulting pellet P2¢ – the crude membrane fraction – in a buffer containing 0.32 M Sucrose, 5 mM NaH2PO4, pH 8.1, protease inhibitor mix. 7. Build a sucrose step gradient in a 17 mL ultracentrifuge tube with 1.4 M sucrose in water at the bottom (5–7 mL), followed by a layer of 1.1 M sucrose (7 mL) (see Note 3).

134

Rajčević

8. Apply the crude membrane fraction (P2¢) to the top of a sucrose step gradient and centrifuge the tubes at 85,000×g (Beckmann SW 40 Ti rotor) for 2 h at 4°C. 9. Carefully collect the two interphase fractions between 0.32 and 1.1 M sucrose (light membranes) and between 1.1 and 1.4 M sucrose (plasma membranes), respectively (see Note 3). Mitochondrial membrane enriched fraction is pelleted at the bottom. 10. Add the appropriate volume of HEPES to dilute the light membrane and plasma membrane fractions to cca. 0.32 M sucrose and centrifuge at 86,000×g (Beckman SW 32.1 Ti rotor) at 4°C for 2 h to pellet the membranes. 11. Resuspend the two pelleted membrane fractions in 100– 110 mL 5 mM HEPES, pH 8.1. 12. Estimate the protein concentrations by Bradford reagent and absorbance measurement at 595 nm. 13. Membrane suspensions can be stored at this point at −20°C for short period or at −80°C for long term. 3.2. Sample Labeling

1. Spin 100 mg of membrane suspensions and vacuum-dry the pellet. 2. Resuspend the pellet in 25 mL Rapigest detergent solution dissolved in 125 mL Dissolution buffer from the iTRAQ kit per vial of detergent. 3. Add 2 mL of reducing agent from the iTRAQ kit to the sample and incubate on vortex for 2 h at room temperature. 4. Spin down, add 1 mL Cysteine blocking agent from the kit and incubate for 10 min at room temperature (net volume ± 30 mL) (see Note 4). 5. Add 2 mg of trypsin (volume of 5 mL; 20 mg of trypsin is resolved in 50 mL of 50 mM acetic acid to a conc. of 0,4mg/mL) and incubate 12 h at 37°C to digest the proteins. 6. Spin down (see Note 5). Label the four digested samples by four different iTRAQ reagents dissolved in 80 mL of ethanol according to the manufacturer’s instructions and vortex at room temperature for 4 h. 7. After the labeling, pool the four differentially labeled samples together and add 400 mL of 1% (v/v) TFA to drop the pH of the samples to around 2. The low pH causes the cleaveage of the Rapigest detergent. 8. Vortex for 1 h at room temperature, centrifuge the samples on a tabletop centrifuge, at max speed for 20 min. 9. Transfer the supernatant to a new tube and vacuum dry it. Discard the pellet (see Note 6).


3.3. Two Dimensional Liquid Chromatography Separation and MALDI TOF/TOF MS

135

The sample set is analyzed on a 4800 Proteomics Analyzer – a MALDI TOF/TOF instrument – thus the calibration and tuning presented here are meant for that instrument. The procedure can as well be performed using other MALDI TOF/TOF and ESI MS/MS instruments. 1. Resuspend the iTRAQ labeled peptide pools in 250 mL of solvent A (10 mM KH2PO4, 20% (v/v) acetonitrile, pH 2.9). 2. Inject 200 mL onto a SCX column (Poly SULFOETHYL Aspartamide), 100 × 2.1 mm, 3 mm, 300Å (Poly LC, Columbia, MD, USA). The flow rate is 200 mL/min. The gradient is increased linearly from 0 to 60% solvent B (solvent A + 500 mM KCl) in 20 min, and then to 100% B in 5 min. 3. Collect the fractions every minute in a fraction collector and vacuum dry. 4. For the second dimension of LC separation, redissolve each SCX fractions in 20 mL 0.1% (v/v) TFA. 5. Inject 10 mL of a SCX fraction on a trap column at a flow rate of 30 mL/min with solvent A (5% (v/v) acetonitrile, 0.05% (v/v) TFA). The peptides are then directed to a capillary reverse phase column at a flow rate of 400 nL/min for separation. The concentration of solvent B (80% (v/v) acetonitrile, 0.04% v/v TFA) is increased linearly from 0 to 5% in 6 min, to 45% in 45 min and finally to 95% in 1 min (Fig. 2). 6. Premix the reverse phase fraction eluate with the MALDI matrix (see Note 7), delivered at a flow rate of 1.5 mL/min and spot them off-line onto a MALDI target plate every 15 s using a Probot Microfraction Collector, for a total of 192 spots per SCX fraction (Fig. 3). 7. For calibration and tuning purposes, spot 4700 Cal Mix (Applied Biosystems/MDS SCIEX) containing six calibrant peptides in the m/z range of 904–3,658 on 13 dedicated calibration spot locations distributed over each MALDI target tplate. 8. Create automatic jobs for two 384-spot MALDI target plates at a time. 9. In MS analyses acquire 1,250 shots (25 subspectra at randomized locations over the whole spot, 50 shots/spectrum) and average over an m/z window of 800–3,000. 10. For each spot, automatically select up to 30 of the strongest precursor peaks with a minimal signal-to-noise ratio of 35 for MS/MS analysis, using a job-wide interpretation method which excludes multiple analyses of the same precursor with a 200-ppm tolerance. 11. In MS/MS analyses, acquire from the weakest to the strongest selected precursor for each spot. For each precursor, acquire

136

Rajčević

Fig. 2. Strong cation exchange-liquid chromatography. iTRAQ labeled peptides are separated based on their charge at low pH in a linear gradient of increasing salt concentration. Free iTRAQ reagents are less retained on the column than the iTRAQ labeled peptides, and eluted completely by 12 min. SCX fractions from 13 to around 26 min contain most peptides. Notice that SCX has a lower resolving power but a higher binding capacity than reverse phase-liquid chromatography, and therefore is chosen as the first dimensional fractionation step. The SCX fractions are vacuum-dried, resolved in 20 mL 0.1% TFA and further subjected to the second dimension separation with the nano-reverse phase-liquid chromatography.

and average 2,500 shots (50 subspectra at randomized locations over the whole spot, 50 shots/spectrum). 12. Perform all MS/MS fragmentation (CID-collision induced dissociation) with air at medium pressure (~1 × 10−6 torr) and using 2 keV collision energy. 3.4. Analysis of the Spectra and Extraction of Isoform- and Species-Specific Proteins

(see Chap. 20 for the detail description of the background of the analysis) 1. Search the MS/MS spectra against databases with trypsin specificity, allowing one missed cleavage and fixed iTRAQ modifications at lysine residue and the N-termini of the peptides, using GPS Explorer v.3.6. (Applied Biosystems/MDS SCIEX) which incorporates Mascot (v.2.1.) search algorithm (Matrix


137

Fig. 3. Representative example of the capillary reverse phase – liquid chromatography of a SCX fraction. The gradient of increasing concentration of acetronitrile is chosen to maximize the separation of peptides across the separation time. Here, the separation time is around 1 h, with 192 fractions collected on the MALDI metal plate. A 2 h run time with ³384 fractions often give a better result, but with a considerable longer MALDI MS/MS analysis time.

Science Inc., Boston, MA, USA). Set the mass tolerance to 100 ppm for precursor ions and 0.5 Da for fragment ions. 2. Next, retrieve the precursor protein sequences of all peptides from the respective databases, and cluster together as a single protein cluster (blastclust; ftp://ftp.ncbi.nih.gov/blast/ documents/blastclust.html) NCBI sequence that shares more than 90% similarity over 85% of the sequence length with a Swissprot sequence. 3. Compare the sequences of the peptides with the sequence of the protein clusters. Peptides that match the sequence of more than one cluster are considered isoform nonspecific. These nonspecific peptides are not considered for protein identification and quantification (see Notes 8–10). 4. Obtain and correct Peak areas for each iTRAQ signature peak (m/z 114.1, 115.1, 116.1, and 117.1) according to the manufacturer’s instruction to account for the differences in isotopic overlap, and log2 transform them. 5. Normalize the possible systematic differences in the starting amounts or labeling efficiencies between samples by subtracting the mean peak area of a sample from each individual iTRAQ signature peak (Fig. 4).

138

Rajčević

Fig. 4. Representative examples of iTRAQ MALDI TOF/TOF mass spectrum. Each reverse phase fraction spotted to a MALDI plate is resolved by MALDI TOF/TOF MS. MS/MS spectrum of a peptide (1,192.56 Da, highlighted in a box (a) or 1,465.31 Da (b)) of four different samples (labeled with four different iTRAQ labels) is shown. A peptide differentially labeled with different iTRAQ reagents in the four samples is indistinguishable in MS mode. Upon MS/MS, the peptide is dissociated into sequence-informative fragment ions for protein identification, and a series of iTRAQ reagent signature ions with masses 114–117 (see enlarged section). The peak area of each iTRAQ reagent signature ions are used for relative quantitation (see Chap. 20).


139

6. To compare the abundance of peptides across multiple iTRAQ experiments, standardize peptide quantity values within each experiment to scores around 0 by subtracting the mean peak area of all four samples. Use these normalized and standardized peptide quantity values to calculate mean protein quantity values.

4. Notes 1. Keep the tissue samples deep frozen until the very beginning of the homogenization. Still frozen, submerge them in the H buffer and start. 2. Make sure the homogenate is spun at 1,000×g and not more, since higher speed will result in loss of the membranes. 3. Following the ultracentrifugation step, three fractions are clearly visible in the tube: the surface layer (on top of 0.32 M sucrose) composed mostly of myelin, the light membrane enriched fraction (visible as whitish cloud between 0.32 and 1.1 M sucrose) composed of lighter membranous vesicles, parts of endoplasmic reticuli, etc., and plasma membranes enriched fraction (a cloud between 1.1 and 1.4 M sucrose). Mitochondria are largely pelleted down to the bottom of the tube and can be preserved for further investigation. 4. Alternatively, reduction and alkylation of cysteine containing peptides can be performed with 2.16–2.7 mL of 50 mM (4–5 mM final concentration) Tris(2-carboxyethyl) phosphine-TCEP (incubate for 1 h at 55°C) and 2.7–10.8 mL 50 mM (5–20 mM final concentration) S-methyl methanethiosulfonate-MMTS (15 min, RT), respectively. 5. After tryptic digestion of the proteins the solution should be spun down and there must be no pellet left. If there is an insoluble pellet, repeat the “Rapigest” step. 6. After the iTRAQ labeling the Rapigest is cleaved by TFA. The pellet indicates the cleaved Rapigest and should be discarded after carefully removing the supernatant with the sample. 7. MALDI matrix recrystallization has a crucial role in MS spectrum quality. Dissolve the CHCA in 100% ethanol (best quality) which is indirectly heated with boiling water. To make it saturated keep adding matrix until no more will dissolve. Decant the dissolved matrix to another tube or vial and put it in the freezer (normal freezer, −20°C or so) for a couple of days. After a few days you should see a very light, pale yellow, colored precipitate of CHCA matrix. Remove the

140

Rajčević

supernatant, scrape out the matrix, put it onto Whatman filter paper, break it up into little bits, and transfer it to a Buchner funnel under a partial vacuum. Pour over a few volumes of ice cold ethanol to wash off any of the previous ethanol still on the dried matrix. Transfer the matrix onto a fresh filter paper and air dry. Transfer the dried, pure matrix to a dark brown tube for storage in the fridge. 8. The approach described here allows not only the identification and quantification of isoform-specific but also species-specific peptides. This is of crucial importance in the proteomics of two species xenograft animal models to be able to distinguish the host from tumor peptide. If such models are used, annotate the spectra twice; once according to the highest Mascot confidence interval (CI) in the species-specific Swissprot database (e.g., mouse, rat, monkey, etc.) from “host animal” and once according to the human Swissprot database. 9. If a spectrum is not annotated using the well curated speciesspecific Swissprot databases, perform Mascot searches in the larger but more redundant species-specific NCBI databases. Thus, the mass spectra will be annotated twice, once for each species, using species-specific SwissProt and NCBI databases. 10. To distinguish between the isoforms of the peptides or the species, which is critical in host-tumor proteomic studies, the clustering of NCBI sequence that shares more than 90% similarity over 85% of the sequence length with a Swissprot sequ ence together as a single protein cluster, is an essential step. Clustering of Swissprot sequences must not be permitted to prevent the clustering of isoform and species nonspecific sequences. Peptides matching the sequence of more than one cluster, when comparing the sequences of the peptides with the sequence of the protein clusters, are not considered species or isoform-specific and are not considered for protein identification and quantification. They however, do present an interesting source of information and should not be discarded.

Acknowledgements The work presented in this chapter was supported by EU FP6 Integrated project grant “Angiotargeting” (contract number 504743), by CRP-Santé, the Research Ministry (MCESR) in Luxembourg and by grant AFR-PDR-08-007 from the FNR, Luxembourg. The author is grateful to Dr. S.P. Niclou and Prof. R. Bjerkvig NorLux Neuro-Oncology laboratory, CRP-Santé, Luxembourg and Department of Biomedicine, University in


141

Bergen, Norway for their invaluable support with this project and critical review of the manuscript. Drs. C. R. Jimenez and J. Knol, Oncoproteomics Laboratory, VU University Medical Center in Amsterdam, The Netherlands are kindly acknowledged for their help. References 1. Rajcevic, U., Niclou, S. P. and Jimenez, C. R. (2009) Proteomics strategies for target identification and biomarker discovery in cancer Front Biosci 14, 3292–303. 2. Bantscheff, M., Schirle, M., Sweetman, G., Rick, J. and Kuster, B. (2007) Quantitative mass spectrometry in proteomics: a critical review Analytical and bioanalytical chemistry 389, 1017–31. 3. 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, 376–86. 4. Kruger, M., Moser, M., Ussar, S., Thievessen, I., Luber, C. A., Forner, F., Schmidt, S., Zanivan, S., Fassler, R. and Mann, M. (2008) SILAC mouse for quantitative proteomics uncovers kindlin-3 as an essential factor for red blood cell function Cell 134, 353–64. 5. Blagoev, B., Kratchmarova, I., Ong, S. E., Nielsen, M., Foster, L. J. and Mann, M. (2003) A proteomics strategy to elucidate functional protein-protein interactions applied to EGF signaling Nat Biotechnol 21, 315–8. 6. Foster, L. J., Rudich, A., Talior, I., Patel, N., Huang, X., Furtado, L. M., Bilan, P. J., Mann, M. and Klip, A. (2006) Insulin-dependent interactions of proteins with GLUT4 revealed through stable isotope labeling by amino acids in cell culture (SILAC) J Proteome Res 5, 64–75. 7. Yao, X., Freas, A., Ramirez, J., Demirev, P. A. and Fenselau, C. (2001) Proteolytic 18O labeling for comparative proteomics: model studies with two serotypes of adenovirus Anal Chem 73, 2836–42. 8. Gygi, S. P., Rist, B., Gerber, S. A., Turecek, F., Gelb, M. H. and Aebersold, R. (1999) Quantitative analysis of complex protein mixtures using isotope-coded affinity tags Nat Biotechnol 17, 994–9. 9. Froment, C., Uttenweiler-Joseph, S., Bousquet-Dubouch, M. P., Matondo, M., Borges, J. P., Esmenjaud, C., Lacroix, C., Monsarrat, B. and Burlet-Schiltz, O. (2005) A quantitative proteomic approach using two-dimensional gel electrophoresis and

isotope-coded affinity tag labeling for studying human 20S proteasome heterogeneity Proteomics 5, 2351–63. 10. Brand, M., Ranish, J. A., Kummer, N. T., Hamilton, J., Igarashi, K., Francastel, C., Chi, T. H., Crabtree, G. R., Aebersold, R. and Groudine, M. (2004) Dynamic changes in transcription factor complexes during erythroid differentiation revealed by quantitative proteomics Nat Struct Mol Biol 11, 73–80. 11. Ranish, J. A., Yi, E. C., Leslie, D. M., Purvine, S. O., Goodlett, D. R., Eng, J. and Aebersold, R. (2003) The study of macromolecular complexes by quantitative proteomics Nat Genet 33, 349–55. 12. Shiio, Y., Rose, D. W., Aur, R., Donohoe, S., Aebersold, R. and Eisenman, R. N. (2006) Identification and characterization of SAP25, a novel component of the mSin3 corepressor complex Mol Cell Biol 26, 1386–97. 13. Li, K. W., Hornshaw, M. P., Van Der Schors, R. C., Watson, R., Tate, S., Casetta, B., Jimenez, C. R., Gouwenberg, Y., Gundelfinger, E. D., Smalla, K. H. and Smit, A. B. (2004) Proteomics analysis of rat brain postsynaptic density. Implications of the diverse protein functional groups for the integration of synaptic physiology The Journal of biological chemistry 279, 987–1002. 14. 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., BartletJones, M., He, F., Jacobson, A. and Pappin, D. J. (2004) Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents Mol Cell Proteomics 3, 1154–69. 15. Keshamouni, V. G., Michailidis, G., Grasso, C. S., Anthwal, S., Strahler, J. R., Walker, A., Arenberg, D. A., Reddy, R. C., Akulapalli, S., Thannickal, V. J., Standiford, T. J., Andrews, P. C. and Omenn, G. S. (2006) Differential protein expression profiling by iTRAQ-2DLCMS/MS of lung cancer cells undergoing epithelial-mesenchymal transition reveals a migratory/invasive phenotype J Proteome Res 5, 1143–54.

142

Rajčević

16. de Souza, G. A., Godoy, L. M. and Mann, M. (2006) Identification of 491 proteins in the tear fluid proteome reveals a large number of proteases and protease inhibitors Genome biology 7, R72. 17. Li, K. W., Miller, S., Klychnikov, O., Loos, M., Stahl-Zeng, J., Spijker, S., Mayford, M. and Smit, A. B. (2007) Quantitative proteomics and protein network analysis of hippocampal synapses of CaMKIIalpha mutant mice J Proteome Res 6, 3127–33. 18. Mukherjee, J., DeSouza, L. V., Micallef, J., Karim, Z., Croul, S., Siu, K. W. and Guha, A. (2009) Loss of collapsin response mediator Protein1, as detected by iTRAQ analysis, promotes invasion of human gliomas expressing mutant EGFRvIII Cancer Res 69, 8545–54. 19. Sakariassen, P. O., Prestegarden, L., Wang, J., Skaftnesmo, K. O., Mahesparan, R., Molthoff, C., Sminia, P., Sundlisaeter, E., Misra, A., Tysnes, B. B., Chekenya, M., Peters, H., Lende, G., Kalland, K. H., Oyan, A. M., Petersen, K., Jonassen, I., van der Kogel, A., Feuerstein, B. G., Terzis, A. J., Bjerkvig, R. and Enger, P. O. (2006) Angiogenesis-independent tumor growth mediated by stem-like cancer cells Proc Natl Acad Sci USA 103, 16466–71. 20. Hopkins, A. L. and Groom, C. R. (2003) Target analysis: a priori assessment of druggability Ernst Schering Res Found Workshop 11–7. 21. Josic, D. and Clifton, J. G. (2007) Mammalian plasma membrane proteomics Proteomics 7, 3010–29. 22. Josic, D., Clifton, J. G., Kovac, S. and Hixson, D. C. (2008) Membrane proteins as diagnostic biomarkers and targets for new therapies Curr Opin Mol Ther 10, 116–23. 23. Rabilloud, T. (2003) Membrane proteins ride shotgun Nat Biotechnol 21, 508–10. 24. Rajcevic, U., Petersen, K., Knol, J. C., Loos, M., Bougnaud, S., Klychnikov, O., Li, K. W., Pham, T. V., Wang, J., Miletic, H., Peng, Z., Bjerkvig, R., Jimenez, C. R. and Niclou, S. P. (2009) iTRAQ based proteomic profiling reveals increased metabolic activity and cellular crosstalk in angiogenic compared to invasive Glioblastoma phenotype Mol Cell Proteomics. 25. Terzis, A. J., Niclou, S. P., Rajcevic, U., Danzeisen, C. and Bjerkvig, R. (2006) Cell therapies for glioblastoma Expert Opin Biol Ther 6, 739–49. 26. Wen, P. Y. and Kesari, S. (2008) Malignant gliomas in adults N Engl J Med 359, 492–507. 27. Carmeliet, P. and Jain, R. K. (2000) Angiogenesis in cancer and other diseases Nature 407, 249–57.

28. Jain, R. K., di Tomaso, E., Duda, D. G., Loeffler, J. S., Sorensen, A. G. and Batchelor, T. T. (2007) Angiogenesis in brain tumours Nat Rev Neurosci 8, 610–22. 29. Reiss, Y., Machein, M. R. and Plate, K. H. (2005) The role of angiopoietins during angiogenesis in gliomas Brain Pathol 15, 311–7. 30. Bogler, O. and Sawaya, R. (2008) Biomarkers and cancer stem cells in primary brain tumors. Foreword Curr Probl Cancer 32, 95–6. 31. Billecke, C., Malik, I., Movsisyan, A., Sulghani, S., Sharif, A., Mikkelsen, T., Farrell, N. P. and Bogler, O. (2006) Analysis of glioma cell platinum response by metacomparison of twodimensional chromatographic proteome profiles Mol Cell Proteomics 5, 35–42. 32. Hamler, R. L., Zhu, K., Buchanan, N. S., Kreunin, P., Kachman, M. T., Miller, F. R. and Lubman, D. M. (2004) A two-dimensional liquid-phase separation method coupled with mass spectrometry for proteomic studies of breast cancer and biomarker identification Proteomics 4, 562–77. 33. Lubman, D. M., Kachman, M. T., Wang, H., Gong, S., Yan, F., Hamler, R. L., O’Neil, K. A., Zhu, K., Buchanan, N. S. and Barder, T. J. (2002) Two-dimensional liquid separationsmass mapping of proteins from human cancer cell lysates J Chromatogr B Analyt Technol Biomed Life Sci 782, 183–96. 34. Wang, Y. C., Choi, M. H. and Han, J. (2004) Two-dimensional protein separation with advanced sample and buffer isolation using microfluidic valves Anal Chem 76, 4426–31. 35. Furuta, M., Weil, R. J., Vortmeyer, A. O., Huang, S., Lei, J., Huang, T. N., Lee, Y. S., Bhowmick, D. A., Lubensky, I. A., Oldfield, E. H. and Zhuang, Z. (2004) Protein patterns and proteins that identify subtypes of glioblastoma multiforme Oncogene 23, 6806–14. 36. Guo, T., Wang, W., Rudnick, P. A., Song, T., Li, J., Zhuang, Z., Weil, R. J., DeVoe, D. L., Lee, C. S. and Balgley, B. M. (2007) Proteome analysis of microdissected formalin-fixed and paraffin-embedded tissue specimens J Histochem Cytochem 55, 763–72. 37. Mustafa, D. A., Burgers, P. C., Dekker, L. J., Charif, H., Titulaer, M. K., Smitt, P. A., Luider, T. M. and Kros, J. M. (2007) Identification of glioma neovascularization-related proteins by using MALDI-FTMS and nano-LC fractionation to microdissected tumor vessels Mol Cell Proteomics 6, 1147–57. 38. Huang, P. H., Mukasa, A., Bonavia, R., Flynn, R. A., Brewer, Z. E., Cavenee, W. K., Furnari, F. B. and White, F. M. (2007) Quantitative analysis of EGFRvIII cellular signaling networks reveals

iTRAQ-Based Analysis of Membrane Proteins from Human Glioma a combinatorial therapeutic strategy for glioblastoma Proc Natl Acad Sci USA 104, 12867–72. 39. Li, K.W, Hornshaw, M. P., van Minnen, J., Smalla, K. H., Gundelfinger, E. D. and Smit, A. B. (2005) Organelle proteomics of rat

143

synaptic proteins: correlation-profiling by isotope-coded affinity tagging in conjunction with liquid chromatography-tandem mass spectrometry to reveal post-synaptic density specific proteins J Proteome Res 4, 725–33.

wwwwwwwwwwwwwwwwww

Chapter 11 OFFGEL-Isoelectric Focusing Fractionation for the Analysis of Complex Proteome Emilie Ernoult and Catherine Guette Abstract The proteomic analysis at the peptide level is increasingly becoming a method of choice for complex samples. The success, however, depends on the development of attractive peptide fractionation methodologies to decrease the sample complexity prior to mass analysis. Recently, the OFFGEL technology has emerged as being of great interest in shotgun proteomics. Using the Agilent 3100 OFFGEL fractionator, OFFGEL allows the in-solution separation of peptides from various biological sources by isoelectric focusing in up to highly resolved 24 fractions. In this chapter, we describe a detailed experimental protocol of a quantitative proteomic workflow including OFFGEL as a crucial step to improve the proteome coverage. We provide the technical details concerning sample preparation, trypsin digestion, iTRAQ labelling, OFFGEL-IEF, RP nano-LC separation, MALDI-TOF/TOF MS/MS, and pI analysis. We rely on our recently published data to illustrate the use of OFFGEL as a powerful method for peptide fractionation and as a filtering tool for pI-based validation of peptide identification. Key words: OFFGEL, Peptide fractionation, Preparative IEF, Shotgun proteomics, Proteome coverage, iTRAQ labelling, Biomarker discovery, Methodology

1. Introduction Face to the high complexity and large dynamic range of proteomics samples, efficient and reproducible separation has become an essential step in the strategies to analyze such samples. In a typical proteomics experiment, a sample of interest is separated either at the protein level or at the peptide level after enzymatic digestion of the proteins, followed by protein identification by mass spectrometry (MS). Isoelectric focusing (IEF) is a high-resolution electrophoretic technique used to separate and concentrate amphoteric biomolecules


145

146

Ernoult and Guette

at their isoelectric point (pI) in a pH gradient and under the application of an electric field. In comparison with other electrophoretic separation techniques, IEF offers the highest resolution, due to the inherent nature of the focusing process: it is a dynamic process resulting from the constant equilibrium between diffusion and migration. IEF, thus, is useful for preparative or semipreparative purposes. In the context of the evolution of gel-based separations towards gel-free strategies, many techniques have recently been introduced to allow the fractionation of proteins and peptides in solution, such as continuous free-flow electrophoresis (see also Chap. 3) (1), Rotofor (2), multicompartment electrolysers (MCE) (3, 4) and off-gel electrophoresis (OFFGEL, OGE), which was first described by Ros et al. (5) as a free-flow technique to purify proteins according to pI and to isolate the protein fraction of interest, in a one-chamber device. The latter technique was later adapted to a more versatile multicompartment format, to recover fractions of well-defined pI values at the end of the separation, and submit them to further analysis or detection (6). The particular advantages of multicompartment OFFGEL-IEF are: (1) the low volumes used (100–300 mL per compartment), positioning it as a semi-preparative device, useful not only for prefractionation purposes but also for analytical uses, and (2) the direct recovery of liquid fractions, making it fit elegantly into the usual LC-MS workflow (7). This technique was demonstrated to be of great interest in shotgun proteomics (8). Not only is IEF a high-resolution and high-capacity separation method for peptides but it also provides additional physicochemical information such as their isoelectric point (9, 10). The pI value provided is used as an independent validating and filtering tool during database search for MS/MS peptide sequence identification (11). Recently, investigations have shown the compatibility between OFFGEL technology and iTRAQ quantitation methods (12–14). As we have recently demonstrated that iTRAQ reagent by itself favours the MALDI ionization of peptides and that peptide OFFGEL fractionation after iTRAQ labelling improves proteome coverage (14, 15), we propose here a detailed protocol for the analysis of complex proteomes.

2. Materials The reagents and the equipment used in each step of the protocol are listed below. All solutions must be prepared using high-quality ultra-pure water, Milli-Q water (Millipore), or equivalent (minimum 18.2 MW/cm).

OFFGEL-Isoelectric Focusing Fractionation for the Analysis

147

2.1. Protein Sample(s)

Protein sample(s) are obtained from various sources, e.g., from nervous tissue or neuronal cell culture, extracted on ice using a denaturant lysis buffer containing a protease inhibitor cocktail, after extensive washing with phosphate-buffered saline (PBS). A typical lysis buffer consists of 7 M urea, 2 M thiourea, 4% (w/v) CHAPS, aliquoted and conserved at −20°C, but other components can also be used (see Notes 1 and 2). Sonication can help in lysis achievement and DNA-bound proteins recovery (see Note 3). Protein extract(s) must be kept at −80°C, or at −20°C for very short term and can be aliquoted for a better storage.

2.2. Trypsin Digestion

1. 100% acetone stored at −20°C at least 30 min before use in a polypropylene plastic tube. 2. Digestion buffer, used when iTRAQ labelling is not later necessary: 50 mM ammonium bicarbonate, pH 7.8, aliquoted and conserved at −20°C; or iTRAQ labelling compatible digestion buffer: triethylammonium bicarbonate buffer (TEAB) 1.0 M, pH 8.5 (cat. no. T7408, Sigma-Aldrich) diluted to 0.5 M and conserved at +4°C. 3. Protein quantification kit: FluoroProfile Protein Quantification Kit (cat. no. FP0010, Sigma-Aldrich) or 2D Quant Kit (cat. no. 80-6483-56, GE Healthcare) and Protein Standard (cat. no. P5619, Sigma-Aldrich). Comparable quantification kits from other suppliers should also be effective, but their compatibility with digestion buffer must be tested. 4. Reducing reagent: Tris-(2-carboxyethyl)phosphine (TCEP, provided in the iTRAQ Reagents Multiplex Kit or cat. no. C4706, Sigma-Aldrich), prepared as a 50 mM stock solution conserved at −20°C. 5. Cysteine-blocking reagent: methyl methanethiosulfonate (MMTS, provided in the iTRAQ Reagents Multiplex Kit or cat. no. 64306, Sigma-Aldrich), prepared as a 200 mM stock solution in isopropanol conserved at −20°C. 6. Trypsin (Trypsin, 25 mg with CaCl2, TPCK treated to inactivate extraneous chymotryptic activity, cat. no. 4352157, Applied Biosystems) conserved at −20°C. Trypsin for proteomic analysis from other suppliers should also be effective but must be tested. 7. Heating block or water bath, 60 and 37°C.

2.3. iTRAQ Labelling, Optional

1. iTRAQ Reagents Multiplex Kit (cat. no. 4352135, Applied Biosystems) or iTRAQ Reagents 8Plex Multiplex Kit (cat. no. 4390812, Applied Biosystems) conserved at −20°C. 2. Precision pH paper, basic (cat. no. P4536, Sigma-Aldrich). 3. ZipTip mC18 (cat. no. ZTC18M096, Millipore).

148

Ernoult and Guette

4. Solvents: ethanol (4Plex labelling) or isopropanol (8Plex labelling), acetonitrile. 5. Trifluoroacetic acid (TFA), 0.1% prepared in a glass bottle and conserved at +4°C. 6. a-Cyano-4-hydroxycinnamic acid (aCHCA) matrix (cat. no. M101, LaserBio Labs). 2.4. RP Desalting

1. Oasis HLB cartridges (Waters). 2. Solvents: methanol or acetonitrile. 3. Syringe, 5 mL. 4. Centrifugal vacuum concentrator.

2.5. OFFGEL-IEF

1. 3100 OFFGEL Fractionator (Agilent Technologies). 2. 3100 OFFGEL Fractionator Kit (Agilent Technologies), containing frames and cover seals. 3. IPG strips, pH 3–10, 13 or 24 cm (provided in the 3100 OFFGEL Fractionator Kit or cat. no. 17-6001-14 or 17-6002-44, GE Healthcare, see Note 4), stored at −20°C. 4. Glycerol 87% (cat. no. 17-1325-01, GE Healthcare). 5. IPG Buffer pH 3–10 (provided in the 3100 OFFGEL Fractionator Kit or cat. no. 17-6000-87, GE Healthcare). 6. Peptide OFFGEL Stock Solution (1.25×): 3% (v/v) glycerol, 1.2% (v/v) IPG Buffer pH 3–10, extemporaneously prepared (2 or 4 mL, respectively, for one 13 or 24-cm strip). 7. Peptide IPG strip rehydration solution: Peptide OFFGEL stock solution (1.25×) diluted to 1× (0.5 or 1.0 mL, respectively, for one 13 or 24-cm strip). 8. Paper electrode (cat. no. 80-6499-14, GE Healthcare), cut and thinned to make it enter in tray lane. 9. Mineral oil (cat. no. 17-1335-01, GE Healthcare). 10. OFFGEL rinsing solution: 2.5 or 5.0 mL (respectively for one 13 or 24-cm strip) ultra-pure water–methanol–formic acid (49:50:1). 11. Centrifugal vacuum concentrator. 12. Cleaning: neutral detergent (cat. no. 80-6452-78, GE Healthcare), toothbrush, hot 1% (w/v) SDS (60°C).

2.6. Capillary RPLC Separation

1. Capillary reversed-phase HPLC system: Ultimate 3000 nanoLC (Dionex) equipped with a C18 column (PepMap100, 3 mm, 100Å, 75 mm id × 15 cm, Dionex) or other systems using a capillary C18 column. 2. May be connected to a Probot microfraction collector (Dionex) for MALDI-TOF/TOF analysis.


149

3. Buffer A: 2% acetonitrile in ultra-pure water with 0.05% TFA. Buffer B: 80% acetonitrile in ultra-pure water with 0.04% TFA. 4. MALDI sample plates (1,664 spots per plate, Applied Biosystems). 5. Matrix: 2 mg/mL aCHCA matrix (cat. no. M101, LaserBio Labs) in 70% acetonitrile/30% ultra-pure water with 0.1% TFA. It contains 0.5 mL/mL matrix of a MS internal calibrant: Glu1-fibrinopeptide (cat. no. F3261, Sigma-Aldrich), stock solution at 50 mg/mL in acetonitrile–ultra-pure water with 0.1% TFA (50/50). 2.7. MS/MS

1. Mass spectrometer able to measure at low m/z: Applied Biosystems 4800 Analyzer (MALDI-TOF/TOF). Apparatus from other suppliers should also be effective, but MALDITOF/TOF provides more detailed peptide sequence analysis. 2. Analysis software such as ProteinPilot Software (Applied Biosystems) fully compatible with iTRAQ quantification.

2.8. pI Analysis

The pI value of each identified peptide can be calculated with the assistance of several free software programs: ●●

●●

●●

Compute pI/MW (http://expasy.org/tools/pi_tool.html). Peptide pI and MW calculation (http://www.fmi.ch/ members/reto.portmann/pepeval.html). Peptide property calculator (http://www.innovagen.se/ custom-peptide-synthesis/peptide-property-calculator/ peptide-property-calculator.asp).

3. Methods An overview of the method is presented in Fig. 1. During the procedure, avoid too many repeated cycles of freezing–thawing of protein samples. Run samples to compare at the same time, in same experimental conditions. 3.1. Trypsin Digestion (2.5 Days)

1. As protein sample(s) may contain potential interfering substances with trypsin digestion and/or iTRAQ labelling (see Notes 1 and 2), a clean-up is realized by precipitating the protein sample(s) in six volumes cold acetone at −20°C overnight (see Notes 5 and 6) in polypropylene plastic 1.5-mL Eppendorf tubes (see Note 7).

150

Ernoult and Guette Protein extract

Acetone clean up Reducing Cystein-blocking

Trypsin digestion

iTRAQ labelling

Dessalting

OFFGEL-IEF fractionation

Nano RPLC MS/MS

Fig. 1. Overview of the workflow based on the use of iTRAQ labelling and peptide OFFGEL fractionation to improve proteome coverage of complex samples.

2. The next day, pellet the sample(s) by centrifugation at 16,000×g for 15 min at 4°C and decant the acetone. Air-dry the pellet for 2 min (see Note 8). 3. Re-suspend the pellet in the adequate digestion buffer (see Note 9) in a volume that ideally permits to have a protein concentration of more than 5 mg/mL. Pellet insoluble materials by centrifugation (16,000×g for 15 min at 4°C). Samples can be stored at −80°C until use. 4. Determine the protein concentration using the sample buffercompatible FluoroProfile Protein Quantification Kit – by diluting samples at least 1/200 in ultra-pure water – or the 2D-Quant kit (see Note 10), using BSA as the standard. 5. For up to 100 mg protein reaction, adjust, if possible, the volume of digestion buffer to 20 mL (see Note 11). 6. Add reducing reagent in a volume equivalent to 1/10 of the starting volume (final concentration of about 5 mM). Vortex, spin and incubate the tubes at 60°C for 1 h. 7. Spin and add cysteine-blocking reagent in a volume equivalent to 1/20 of the starting volume (final concentration of about 10 mM). Vortex, spin and incubate the tubes at room temperature for 10 min.


151

8. Reconstitute 25 mg trypsin with 25 mL ultra-pure water and add 10 mL (10 mg) trypsin to each 100 mg protein reaction. Vortex, spin and incubate at 37°C for 40 h. At the end of the digestion, spin to bring the protein digest to the bottom of the tube(s). Samples can be stored at −80°C until use. 3.2. iTRAQ Labelling, Optional (3.5 h)

1. Check the pH of all protein digest samples by adding a very small amount of samples to a pH strip. pH must be between 8.0 and 9.0 (ideally at 8.5). 2. Spin each vial of iTRAQ reagent previously let to reach room temperature. 3. Add 70 mL ethanol to each vial of iTRAQ Reagent – 4Plex or 50 mL isopropanol to each vial of iTRAQ Reagent – 8Plex (see Note 12). Vortex and spin. 4. Transfer one iTRAQ Reagent (about 90 mL) to one 100 mg protein digest (33 mL). 5. Vortex, spin and incubate at room temperature for 3 h. 6. Verify the labelling quality and the protein quantification precision by pooling 1 mL of each sample of an experiment and analysing the pool by MS/MS after desalting using ZipTip devices according to standard procedures. 7. Stop labelling reaction by adding 120 mL ultra-pure water to each tube. Samples can be stored at −80°C until use. 8. Combine samples labelled with different iTRAQ reagents. Vortex and spin. This labelled pool can be stored at −80°C until use.

3.3. RP Desalting (3–4 h)

This step is absolutely essential before OFFGEL-IEF (see Note 13). 1. Connect the Oasis HLB cartridge to a syringe. 2. Condition the RP by drawing through methanol or acetonitrile and equilibrate with ultra-pure water. 3. Load the sample and wash with ultra-pure water. 4. Elution: draw through 1 mL methanol or acetonitrile. Collect the eluate in a polypropylene plastic 2 mL tube. 5. Evaporate in a centrifugal vacuum concentrator until no liquid is observable (see Note 14). Samples can be stored at −80°C until use.

3.4. OFFGEL-IEF (16–24 h)

This section follows the supplier manual “Quick Start Guide 4.0,” updated on 1/22/2009 (Agilent Technologies) with some modifications. 1. Prepare the peptide OFFGEL stock solution (1.25×) and peptide IPG strip rehydration solution.

152

Ernoult and Guette

2. Prepare the sample; re-suspend the dried sample (400 mg, see Note 15) in 0.36 or 0.72 mL ultra-pure water (respectively for one 13- or 24-cm strip). Vortex vigorously. Add 1.44 or 2.88 mL (respectively for one 13- or 24-cm strip) of the peptide OFFGEL stock solution (1.25×). For the 24-fractions device, sample transfer in a larger tube is required. Vortex. 3. Place tray in a good orientation (you must be able to read figures), remove the plastic film from the 13- or 24-cm in length IPG strip gel and place the IPG strip in the tray (see Note 16) with the gel side up and the positive extremity on the left until the strip touches the edge of the tray. Delicately click the frame (with 12 or 24 wells according to the strip length, 13 or 24 cm, respectively) on the gel starting from the left against the mechanical stop. 4. Rehydrate the gel by pipetting 20 mL peptide IPG strip rehydration solution into each of the wells. Tap the tray on the desk to ensure that the rehydration solution reaches the gel. Rehydrate both gel extremities (see Note 17). Wait for 15 min and then absorb with a pipette tip the excess of solution at the strip extremities. 5. Wet two paper electrode pieces with ultra-pure water and press on absorbent paper to eliminate the liquid excess. Put one paper on each strip extremity (see Note 13), letting a little interspace between the paper and the frame. 6. Load 150 mL sample in each of the 12 or 24 wells. Gently re-cover with appropriate frame cover seal. 7. Place the tray on the OFFGEL platform. Pipette mineral oil onto each end of the strip, both sides of the covered frame. Wait 1 min for the oil to slip along the frame. Repipette mineral oil. Paper electrode pads must be re-covered. Pipette 2 mL mineral oil in all unused tray lanes. 8. Assembly first the anode electrode and slowly wedge the tray in the anode connector. Then, position the cathode electrode close to the frame in a good orientation (you must be able to read figures). Electrodes must be in contact with paper electrode pads. Re-position the electrode pads if necessary. 9. Close the lid, set run temperature to 20°C and start the appropriate pre-entered program (“OG12PE” or “OG24PE”). 10. At the end of the OFFGEL fractionation (flashing blue light), stop the apparatus, remove electrodes and carefully remove the frame cover seal without disassembling the frame. Pipette each well fraction into an annotated polypropylene plastic 1.5-mL Eppendorf tube. 11. Rinse each well by letting incubate 200 mL OFFGEL rinsing solution for 30 min. Pipette each rinsing fraction and combine with the corresponding peptide fraction.


153

12. Evaporate in a centrifugal vacuum concentrator until a glycerol pellet is observable. The fractions can be stored at −80°C until use. 13. To avoid cross-contamination between samples, take care to clean the tray, frames, cover seals and electrodes with a neutral detergent using a toothbrush. Then, rinse with water, followed by rinsing with 1% SDS at 60°C and, to finish, extensively with ultra-pure water to eliminate salts and SDS. Blot oil drops on the OFFGEL platform. 3.5. Capillary RPLC Separation (6.5 h for Three OFFGEL Fractions)

1. Reconstitute up to three evaporated OFFGEL fractions in 20 mL ultra-pure water with 0.1% TFA. Vortex vigorously and transfer to appropriate chromatography vials. 2. Inject successively each sample onto the nanoLC system at a relatively slow flow rate (300 nL/min). 3. Peptides are desalted for 3 min using only Buffer A on the precolumn. 4. The separation uses the following gradient: 0–20% Buffer B in 10 min, 20–50% Buffer B in 65 min and 50–100% Buffer B in 20 min. Chromatograms are recorded at 214 nm. 5. Peptide fractions are collected for 90 min using a Probot microfraction collector. The matrix is continuously added to the column effluent via a micro “T” mixing piece at 1.2 mL/ min flow rate. After 11-min run, a start signal is sent to the Probot to initiate fractionation. The fractions are collected for 10 s and spotted on a MALDI sample plate. Up to three OFFGEL fractions can be separated on one MALDI plate. Plates can be conserved, protected from moisture, for up to 2 weeks at room temperature before MS/MS analyses.

3.6. MS/MS (30–48 h per MALDI Plate)

All LC-MALDI sample positions are screened in MS-positive reflector mode using 1,500 laser shots. The fragmentation of automatically selected precursors is performed at collision energy of 1 kV using air as collision gas (pressure of ~2 × 10−6 Torr). MS spectra are acquired between m/z 800 and 4,000. The parent ion of Glu1-fibrinopeptide at m/z 1,570.677 diluted in the matrix (3 fmole per spot) is used for internal calibration. Up to 12 of the most intense ion signals per spot position having a S/N >12 are selected as precursors for MS/MS acquisition. Data processing for peptide and protein identification and iTRAQ quantification is performed using the ProteinPilot software using the Paragon algorithm (16). Each MS/MS spectrum is searched against the Uniprot/Swissprot database for specific specie with the fixed modification of MMTS-labelled cysteine parameter enabled. Other parameters such as tryptic cleavage specificity, precursor ion mass accuracy and fragment ion mass accuracy are MALDI 4800 builtin functions of ProteinPilot software. The ProteinPilot software

154

Ernoult and Guette

calculates a confidence percentage (the unused score) that reflects the probability that the hit is a false positive, meaning that at 95% confidence level, there is a false-positive identification chance of about 5%. While this software automatically accepts all peptides having an identification confidence level >1%, only proteins having at least one peptide above 95% confidence are initially recorded. The low-confidence peptides cannot give a positive protein identification by themselves but may support the presence of a protein identified using other peptides with higher confidence. Searches against a concatenated database containing both forward and reversed sequences provide estimation of the false discovery rate. 1. The identified peptides with a confidence score superior to 95% are selected.

3.7. pI Analysis

2. Their pI are calculated taking into account possible modifications (N and Q deamidation, formyl, acetyl, pyroglu N-term). 3. The average pI value is calculated with all the peptides of each OFFGEL fraction (Fig. 2). A standard deviation is also evaluated for each fraction. This error may be due to a peptide misidentification (false positive), but it also may be due to the

11 free-labelled peptides 10

iTRAQ peptides

Isoelectric point

9 8 7 6 5 4 3 1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Fraction number

Fig. 2. Analysis of SH-SY5Y peptides pI after OFFGEL fractionation and MALDI-MS/MS identification. The average experimental pH of all peptides in a single fraction after filtering for false positive is presented as bars; white bars: free-labelled peptides; dark bars: iTRAQ peptides. Error bars indicate the SD of each fraction’s experimental pI. The broken line is based on the theoretical pI values for an IPG strip of 24 cm ranging from pH 3–10; Agilent Technologies provided the theoretical pI values.


155

Bjellqvist algorithm (9). This algorithm is known to overestimate pI (17) values in the acidic range and to take not into account the iTRAQ groups in N-term position and/or on the lateral lysine chain. For these reasons, only the peptides showing an experimental pI deviating by more than one pI unit from the average value are excluded. For example, in fraction 3 of a neuroblastoma cell extract (14), we identified 1,310 unique peptides. The pI average of all the peptides is 4.28 with a standard deviation of 0.45 (see representative examples in Table 1). We kept peptides that had a pI value between 3.28 and 5.28. In these conditions, only 12 peptides were not validated and we obtained a new pI average value of 4.25 with a standard deviation of 0.13.

4. Notes 1. Some lysis buffer components may interfere with trypsin digestion and/or iTRAQ labelling. Avoid thiols (dithiotreitol, mercaptoethanol, etc.) as they interfere with cysteine blocking. High amounts of denaturants and detergents inhibit trypsin activity. Primary amines (e.g., ammonium bicarbonate) react with iTRAQ reagents and affect peptide labelling efficacy. Sample clean-up by acetone precipitation is highly recommended before trypsin digestion. 2. Avoid charged detergents, especially sodium dodecyl sulphate (SDS), as they influence peptide pI-fractionation during OFFGEL isoelectrofocalization. In case when SDS is necessary for protein extraction, clean-up the sample mixture using cation-exchange chromatography. 3. When urea is used in lysis buffer, keep protein samples in ice during sonication and run them for repeated short pulses. Temperature over 37°C causes urea to hydrolyze into isocyanate, which modifies proteins by carbamylation. 4. Neither Mann’s team (7) nor our team observed quality differences between Agilent and GE Healthcare IPG strips. 5. When iTRAQ labelling is planned, clean protein samples with acetone, not using acidic-based precipitation procedure (trichloroacetic acid or some commercially available kits). 6. Acetone precipitation encounters some protein loss. Prepare a larger quantity of starting biological material. 7. We suggest to aliquot in 200-mL protein lysates and to add acetone. It will be then easier to re-solubilize samples in 1.5-mL tubes.

i@N-term; Me(E)@11 i@N-term; Me(E)@13 i@N-term; Me(E)@7 i@N-term; Me(E)@7; Pro->p(P)@8

IPNPDFFEDLEPFR

AVFVDLEPTVIDEVR

AVFVDLEPTVIDEIR

AVFVDLEPTVVDEVR

3.92

3.92

3.92

3.92

3.92

……………………………………………………………….. ………………………………………………………………..

………………………………………………………………..

………………………………………………………………..

SMEAEMIQLQEELAAAER

3.98

3.93

3.93 i@N-term; O(P)@12; O(P)@15; O(P)@18; O(P)@21

i@N-term; Me(D)@10

TLFELAAESDVSTAIDLFR

3.92

3.9

i@N-term; MC@3; MC@4; i(K)@12

i@N-term; i(Y)@8

DQGTYEDYVEGLR

GPQGHQGPAGPPGPPGPPGPPGVSGGGYDFGYDGDFY

Gln->p@N-term

QHFISFDTDR

3.84

LICCDILDVLDK

i@N-term; Ser->Ox(S)@4; D(N)@5; D(N)@6

DLISDDEQLPMLGR

3.84

3.93

i@N-term; Me(R)@16

DQEGQDVLLFIDNIFR

3.8

i@N-term; i(Y)@2

Gln->p@N-term; Me(E)@5

QEYDEAGPSIVHR

3.8

3.3

DYDSLAQPGFFDR

Gln->p@N-term

QEYDESGPSIVHR

3.93

Formyl@N-term; D(Q)@4

LIGEIVSSITASLR

pI

TTQVPQFILDDFIQNDR

Modifications

Identified peptides

Table 1. Representative examples of identified peptides in fraction 3 of one OFFGEL experiment with a neuroblastoma cell extract. In total 1310 unique peptides are identified. The peptides are sorted based on their pI; the first 17 peptides with pI < 4 are shown as example. (i: iTRAQ-4plex; Me: methyl; MC: methio(C); O: oxidation; D: deamidated; P: pyru-Glu; Ox: Oxoalanine).

156 Ernoult and Guette


157

8. After acetone decantation, do not over-dry the protein pellet, it would not dissolve in the buffer. 9. Re-solubilizing of proteins after precipitation is critical. Active vortexing and pipetting will help to dissolve pellets. Do not concentrate proteins too much in buffer. 10. Particular care should be taken for protein content estimation of samples, particularly when differential quantification between different extracts using iTRAQ labelling is necessary. According to our experience, a maximum of 30% variation in iTRAQ ratios (calculated by ProteinPilot Software, Applied Techno logies) is tolerated. Quantitation must be accomplished just prior trypsin digestion, after acetone precipitation. 11. A protein concentration of 5 mg/mL is ideal for iTRAQ labelling (it is recommended to label up to 100 mg protein per iTRAQ reagent tube). In case when concentration is lower and sample volume, therefore, higher, proceed to trypsin digestion (adjust reducing and cysteine-blocking reagents volumes) and evaporate digestion buffer (in a centrifugal vacuum concentrator) after trypsin digestion, before iTRAQ labelling. Then, re-suspend the peptide pellet in 20–30 mL TEAB 0.5 M, pH 8.5. 12. A minimal organic solvent concentration of 65% is essential for iTRAQ labelling to succeed. The digest volume must not be too large (see Note 11). 13. Salts greatly interfere with OFFGEL isoelectrofocalization. Salt concentration in samples should not exceed 10 mM. Desalting just prior OFFGEL fractionation prevents current variations and guarantees a completed run in 24 h (pH 3–10 fractionation). If run is extended, paper electrode pads, which absorb water, ions and non-focusing peptides during the run, can be changed. By modifying the run program, OFFGEL can be used to first desalt a protein (peptide) solution containing up to 100 mM salts (18). 14. When iTRAQ labelling is carried out, a brown pellet may appear at the end of the evaporation. 15. According to Agilent Technologies, up to 5 mg proteins can be loaded, but the quality of the fractionation should be tested (see Note 18). 16. If a few IPG strips are run in parallel, position them in the middle of the tray. 17. The good rehydration of IPG strip extremities ensures a perfect sealing of the system and avoids sample leakage. 18. The quality of the fractionation could be evaluated by looking at the number of fractions in which each distinct peptide could be found. 95% of the identified peptides must be present in one or two adjacent fractions.

158

Ernoult and Guette

References 1. Hannig, K. (1982) New Aspects in Preparative and Analytical Continuous Free-Flow Cell Electrophoresis Electrophoresis 3, 235–43. 2. Bier, M. (1998) Recycling isoelectric focu-sing and isotachophoresis Electrophoresis 19, 1057–63. 3. Righetti, P. G., Wenisch, E., Faupel, M. (1989) Preparative Protein-Purification in a MultiCompartment Electrolyzer with Immobiline Membranes Journal of Chromato-graphy 475, 293–309. 4. Righetti, P. G., Wenisch, E., Jungbauer, A., Katinger, H., Faupel, M. (1990) Preparative Purification of Human Monoclonal-Antibody Isoforms in a Multicompartment Electrolyzer with Immobiline Membranes Journal of Chromatography 500, 681–96. 5. Ros, A., Faupel, M., Mees, H., van Oostrum, J., Ferrigno, R., Reymond, F., Michel, P., Rossier, J. S., Girault, H. H. (2002) Protein purification by Off-Gel electrophoresis Proteomics 2, 151–6. 6. Michel, P. E., Reymond, F., Arnaud, I. L., Josserand, J., Girault, H. H., Rossier, J. S. (2003) Protein fractionation in a multicompartment device using Off-Gel (TM) isoelectric focusing Electrophoresis 24, 3–11. 7. Hubner, N.C., Ren, S., Mann, M. (2008) Peptide separation with immobilized pI strips is an attractive alternative to in-gel protein digestion for proteome analysis Proteomics 8, 4862–72. 8. Essader, A.S., Cargile, B.J., Bundy, J.L. Jr. (2005) A comparison of immobilized pH gradient isoelectric focusing and strong-cationexchange chromatography as a first dimension in shotgun proteomics Proteomics 5, 24–34. 9. Bjellgvist, B., Hughes, G.J., Pasquali, C., Paquet, N., Ravier, F., Frutiger, S., Hughes, G.J., Hochstrasser, D. (1993) The focusing positions of polypeptides in immobilized pH gradients can be predicted from their amino acid sequences Electrophoresis 14, 1023–31.

10. Krügsveld, J., Gauci, S., Dormever, W., Heck, A.J. (2006) In-gel isoelectric focusing of peptides as a tool for improved protein identification J. Proteome Res. 5, 1721–30. 11. Cargile, B.J., Bundy, J.L., Freeman, T.W., Stephenson, J.L. Jr. (2004) Immobilized pH gradients as a first dimension in shotgun proteomics and analysis of accuracy of pI predictability of peptides J. Proteome Res. 3, 112-9. 12. Lengqvist, J., Uhlen, K., Lehtio, J. (2007) iTRAQ compatibility of peptide immobilized pH gradient isoelectric focusing Proteomics 7, 1746–52. 13. Chenau, J., Michelland, S., Sidibe, J., Seve, M. (2008) Peptides OFFGEL electrophoresis: a suitable pre-analytical step for complex eukaryotic samples fractionation compatible with quantitative iTRAQ labeling Proteome Sci. 26, 6–9. 14. Ernoult, E., Gamelin, E., Guette, C. (2008) Improved proteome coverage by using iTRAQ labelling and peptide OFFGEL fractionation Proteome Sci. 6, 27–40. 15. Ernoult, E., Bourreau, A., Gamelin, E., Guette, C. (2010) A proteomic approach for plasma biomarker discovery with iTRAQ labelling and OFFGEL fractionation J. Biomed. Biotechnol. 2010:927917. 16. Shilov, I.V., Seymour, S.L., Patel, A.A., Loboda, A. et al. (2007) The Paragon Algorithm, a next generation search engine that uses sequence temperature values and feature probabilities to identify peptides from tandem mass spectra Mol. Cell. Proteomics 6, 1638–55. 17. Cargile, B.J., Talley, D.L., Stephenson, J.L. (2004) Immobilized pH gradients as a first dimension in shotgun proteomics and analysis of the accuracy of pI predictability of peptides Electrophoresis 25, 936–45. 18. Arnaud, I.L., Josserand, J., Jensen, H., Lion, N., Roussel, C., Girault, H.H. (2005) Salt removal during Off-Gel electrophoresis of protein samples Electrophoresis 26, 1650–8.

Chapter 12 A 1D-PAGE/LC-ESI Linear Ion Trap Orbitrap MS Approach for the Analysis of Synapse Proteomes and Synaptic Protein Complexes Ning Chen, Roel C. vd Schors, and August B. Smit Abstract The 1D-PAGE/LC-ESI MS/MS approach is widely used for the qualitative and quantitative analysis of proteomes ranging from low to high complexity. As the first dimension of separation is based on SDSPAGE, this method is compatible with the analysis of all classes of proteins including hydrophobic membrane proteins and proteins with extreme pI. 1D PAGE has the possibility to reduce protein complexity by cutting the sample gel lane into a large number of slices. The second dimension of separation is reverse phase C18 chromatography, which is coupled online to an ESI Linear Ion Trap Orbitrap MS. In this configuration the MS spectra are taken at high resolution in the Orbitrap, which improves the confidence of peptide identification. In parallel, the MS/MS spectra are generated by the Linear Ion Trap with high sensitivity, which allows the interrogation of larger numbers of peptides and thus yields improved protein identification in a given period of time. Key words: SDS-PAGE, Proteomics, LC-MS/MS, Orbitrap, Synapse

1. Introduction Neuroproteomics generally focuses on (quantitative) studies of protein constituents in the nervous system (1), such as in specific organelles, for instance synapse subdomains, or in protein complexes (2–4). Only in some occasions total extracts from certain brain areas are analyzed (5). Since the brain has a high protein complexity, probably the highest from all organs in the body, the number of distinct proteins that one would like to investigate in each experiment is usually large. A confounding issue therefore is to find ways to reduce the complexity of the samples using separation


159

160

Chen, vd Schors, and Smit

steps to a level that is retractable by mass spectrometry, but without sacrificing total protein coverage during the separation process. To address the issue of sample complexity, a number of twodimensional separation strategies have been developed (see Chaps. 8–11). One of the popular 2D separation workflows involves fractionation of proteins by SDS-PAGE, followed by capillary reverse phase liquid chromatography MS/MS analysis of the tryptic peptides from the in-gel digest of each gel sample (6). There are several distinct advantages of this approach. (a) It effectively handles all classes of proteins, including the hydrophobic membrane proteins such as receptors and ion channels, and proteins with extreme isoelectric focussing point. (b) It separates proteins into size classes, which provides useful additional information for protein identification. (c) It has the flexibility to reduce protein complexity by simply cutting the sample gel lane into a selectable number of slices. (d) SDS-PAGE is a standard laboratory technique, which is readily accessible in a typical biochemical laboratory environment. The samples thus prepared by life scientists can be sent to a core facility for further proteomics analysis. The trypsin digestion of proteins from a single gel slice may generate thousands of peptides. These peptides are partially separated by a capillary reverse phase chromatography column. Peptides eluted from the column are then subjected to mass spectrometric analysis. They can be processed offline using a MALDI TOF/TOF MS (Chap. 10–11), or online using a mass spectrometer via electrospray. Here, we describe the use of an online Linear Ion Trap Orbitrap MS approach. In this configuration the MS spectra are taken at high resolution in the Orbitrap, which improves the confidence of peptide identification. In parallel, the MS/MS spectra are generated by the Linear Ion Trap with high sensitivity, which allows the sampling of a larger number of peptides in a given time window (7).

2. Materials 2.1. SDS: PAGE and Trypsin Digestion

100 mM ammonium bicarbonate, pH 7.8~8.0 1. Ammonium bicarbonate (NH4HCO3; 99%; Fluka, Steinheim, Germany); 0.78 g 2. Deionized water; fill up to 100 mL 50 mM ammonium bicarbonate/50% (v/v) acetonitrile 1. Acetonitrile (ultragradient HPLC grade; JT Baker, Deventer, Holland); 50 mL 2. 100 mM NH4HCO3; 50 mL

A 1D-PAGE/LC-ESI Linear Ion Trap Orbitrap MS Approach

161

0.1% Trifluoroacetic acid/50% acetonitrile 1. Trifluoroacetic acid (TFA; Protein sequence grade; Applied Biosystems, Warrington, UK); 0.1 mL 2. Acetonitrile; 50 mL 3. Deionized water; fill up to 100 mL Trypsin solution 1. 10 mg/mL trypsin (sequence grade; Promega, Madison, USA) in 50 mM ammonium bicarbonate Vertical electrophoresis system (e.g., Mini-PROTEAN Tetra Electrophoresis System from Bio-rad). 2. Premixed Gel-casting buffer, premixed sample loading buffer and premixed running buffer (e.g., from Bio-rad) 2.2. LC-MS Analysis

Capillary HPLC system with a reverse phase C18 column. 1. A capillary HPLC system with autosampler/autoinjector (we use an Eksigent nano LC-ultra 1D plus HPLC system equipped with a 5-mm Pepmap 100 C18 (Dionex) trapping column (300 mm ID, 5 mm particle size) and a capillary C18 column. We use a 200-mm home-made Alltima C18 analytical column (100 mm ID, 3 mm particle size). 2. HPLC solvents. Solvent A: 95% deionized water/5% acetonitrile, 0.1% acetic acid; solvent B: 95% acetonitrile/5% deionized water, 0.1% acetic acid. 3. Reconstitute solution: 0.1% acetic acid. 4. 200 mL conical vial that fits the autosampler and the corresponding cap. Mass spectrometer 1. Electrospray LTQ-Orbitrap from Thermo Fisher Scientific 2. Stainless Steel Nano-bore emitter (Proxeon, 30 mm ID)

3. Methods 3.1. Fractionation of Proteins by SDS-PAGE

1. Mix the sample with 5× loading buffer (1:1), and add deionized water to a final volume of about 25 mL, and heat at 90°C for 5 min. Then add 5 mL aliquot of 30% acrylamide stock solution to the sample (8), and incubate for 30 min at room temperature. The cysteinyl residues may be reduced by dithiothreitol in loading buffer and alkylated by unpolymerised acrylamide (see notes 1 and 2).

162


2. Prepare a 10 or 12% SDS polyacrylamide gel, run samples, and stain the gel with colloidal Coomassie blue, as described in Chap. 8. 3. In case of low sample complexity, such as a sample resulting from an immuno-precipitation experiment (see Chap. 6), cut the gel into five slices for each gel lane; two above the antibody heavy chain, one covering the antibody heavy chain, one between the antibody heavy and light chain, and the bottom slice containing the antibody light chain (see Fig. 1). 4. The cutting places on the gel are defined according to the position of the marker proteins; the first cut should be at 120 kD, the second cut at 60 kD, the third cut at 45 kD, and the last cut at 30 kD. 5. For samples containing ~1,000 proteins with intermediate complexity, for example a synapse enriched fraction, we generally cut 12 gel slices of equal sizes per sample to reduce sample complexity (Fig. 2). For more even higher complexity samples it is possible to run them in a large format gel to increase sample loading, and then cut each gel lane into 12–20 pieces.

Fig. 1. Analysis of protein complexes fractionated on a 10% SDS gel. The molecular weight marker was loaded on the left lane (lane 1) followed by the four different IP samples (lanes 2–4) and a control IP sample without antibody (lane 5). The Coomassie blue stained gel revealed a low level of protein complexity. Therefore, the gel from each lane was cut into five pieces as indicated by the horizontal dotted lines. The cut was guided by the position of the marker proteins (see text).


163

Fig. 2. Analysis of proteins from synapse subdomains. A synaptosome-enriched sample was fractionated by free flow electrophoresis (FFE) (Chap. 3); individual fractions were run on a 10% gel. The Coomassie blue stained gel revealed an intermediate complexity protein pattern across all fractions. Therefore, the gel from each lane was cut into twelve pieces as indicated by the horizontal dotted lines to reduce protein complexity in the subsequent MS analysis.

3.2. In-Gel Digestion

1. Excise protein bands and cut each gel piece into small fragments using a scalpel, and transfer them to a 1.5-mL Eppendorf tube. 2. Add 500 mL of 50 mM NH4HCO3/50% acetonitrile to the gel fragments, vortex for 20 min. 3. Remove the solution and discard. 4. Add 500 mL 100% acetonitrile and vortex for 20 min. The gel fragments should turn white and shrink. 5. Remove acetonitrile and discard. 6. Add 500 mL 50 mM NH4HCO3, incubate for 5 min at room temperature. 7. Remove NH4HCO3 and discard. 8. Add 500 mL of 50 mM NH4HCO3/50% acetonitrile, incubate overnight to destain completely the gel fragments. 9. Remove the solution and discard. 10. Add 500 mL 100% acetonitrile and vortex for 20 min.

164


11. Remove the solution. 12. Dry gel fragments for 30 min in a SpeedVac. 13. Reswell the dried gel fragments at 4°C for 45 min in buffer containing trypsin. The gel fragments should just be covered (50–120 mL trypsin solution/slice, note 3). 14. Add 100 mL 50 mm NH4HCO3. 15. Digest overnight at 37°C or 2 h at 55°C. 16. Remove solution containing the trypic peptides diffused out of the gel piece and transfer it into a 1.5-mL eppendorf tube. 17. Add 200 mL 0.1% TFA/50% acetonitrile to the gel piece and incubate for 20 min. This step extracts the remaining peptides from the gel piece. Remove the solution, and put into the same eppendorf tube that contains the tryptic peptides diffused out of the gel piece first. Repeat this step once and pool the solution. 18. Dry the peptide solution in SpeedVac (note 4). The samples can be stored at −20°C for months before further analysis. 3.3. LC-MS/MS Analysis of the Tryptic Peptides

1. Redissolve the sample in 15 mL 0.1% acetic acid, vortex for 5 min. 2. Transfer the solution to a sample vial, cap the vial and put it into the loading tray at 4°C. 3. Load 10 mL of samples into the injection loop. 4. Separate the peptides on the capillary reverse phase C18 column with increasing acetonitrile concentration (note 5). A longer liquid chromatography run time may be necessary for complexity samples (note 1). 5. Position the emitter in the ion source to initiate and maintain a stable electrospray. The spray tip, which has a small diameter, is susceptible to blockade, resulting in an unstable spray with significant negative impact on the detection of peptides in the mass spectrometer (note 6). 6. Operate LTQ-Orbitrap in data dependent mode to automatically switch between MS and MS/MS. MS spectra in the range of m/z 330–2,000 can be acquired in the Orbitrap at a FWHM resolution of 30,000 after accumulation to 500,000 in the linear ion trap with one microscan. The three most abundant precursor ions are selected for fragmentation by CID with an isolation width of 2 Da. CID is performed in the linear ion trap after accumulation to 50,000 with one microscan. 7. Process the MS data for protein identification and quantitation (see Chap. 21 for spectral count quantitation).


165

4. Notes 1. To facilitate identification of cysteine containing peptides, postalkylation with iodoacetamide before the in-gel digestion is often included in the workflow. However, the unpolymerised acrylamide in the gel may also alkylate cysteine, yielding mixtures of Cys-S-beta-propionamide from acrylamide derivatization, and S-carboxyamidomethylcysteine from iodoacetamide derivatization. Alternatively, extra acrylamide (5 mL aliquot of 30% acrylamide stock solution per sample (8)) can be added to the samples after solubilisation and reduction; and the cysteinyl residues are allowed to alkylate with acrylamide during gel electrophoresis. 2. We compared the effectiveness of in-gel digestion experiments on the same synaptosome sample with different alkylation methods, including the derivatization with iodoacetamide or extra acrylamide before SDS-PAGE, and with iodoacetamide on gel after SDS-PAGE. We also tested the effect of short and long liquid chromatography time on the number of proteins identified (Fig. 3; see note 5 for detail). The number of identified peptides is quite similar across all alkylation treatments, except alkylation with iodoacetamide on destained gel fragments which gives a slightly lower number of identified peptides. The cysteine containing peptides were prone to be identified in samples alkylated by iodoacetamide or extra acrylamide. Considering the favorable number of cysteine containing peptides detected and the ease of the treatment, we use the in-situ alkylation with added acrylamide as our method of choice. 3. The solution volume is usually depending on the size of the gels. For the destaining, dehydration, washing, and peptide extraction steps, an extra amount of solution can be used to get to a high efficiency. For the trypsin incubation step, an extra amount will result in increased self-digestion. Therefore, the trypsin solution should just cover the reswelled gel fragments. 4. The peptides should not be dried too long to avoid sample loss. 5. A short linear gradient of 5% solvent A to 45% solvent B in 50 min is sufficient for relatively low complexity samples containing only tens of proteins, such as the samples from immuno-precipitation experiments. A longer LC run is required for more complex samples. We use the following gradient with linear increase in solvent B for the analysis of synapse enriched fraction: 0 min, 5% solvent B; 95 min, 40% solvent B; 100 min, 90% solvent B (see Fig. 3).

166


Fig. 3. Effects of different alkylation methods and liquid chromatography run times on the peptide identification. Synaptosome proteins were separated on SDS-PAGE, and the gel slice between 45 and 60 kDa was cut out for digestion. Each sample was split into two equal parts and analysed separately by LC-MS/MS with a 60-min LC run and a 120-min LC run, respectively. (a) Number of identified peptides with different alkylation treatments. (b) Number of cysteine containing peptides identified with different alkylation treatments. Acry before SDS-PAGE, extra acrylamide (5 mL aliquot of 30% acrylamide stock solution) was added to the sample after solubilization and reduction in normal loading buffer containing 0.1 M dithiothreitol. IoAA before SDS-PAGE, sample was solubilized and reduced in loading buffer containing 0.015 M dithiothreitol at 70°C. The alkylation was performed by adding 5 mL aliquot of 300 mM iodoacetamide to 25 mL sample and incubation at room temperature in the dark for 30 min (9). Then the samples were separated by SDS-PAGE. IoAA on whole gel, the whole gel was incubated in 55 mM idoacetamide solution in dark for 30 min after electropheresis. IoAA after destain, the destained gel fragments were reswelled in 55 mM idoacetamide solution and incubated in dark for 30 min. No extra Acry or IoAA, no extra acrylamide or iodoacetamide was added to samples or gels.


167

6. In the present configuration, the diameters of the reverse phase C18 column and the spray tip are larger than most reported capillary LC-ESI MS/MS setups. Our flow rate is therefore also higher. This setup gains higher stability and is less susceptible to blockade. The trade-off is that the sensitivity of the analysis may be slightly lower. References 1. Bayes, A., and Grant, S. G. (2009) Neuroproteomics: understanding the molecular organization and complexity of the brain, Nat Rev Neurosci 10, 635–646. 2. Li, K. W., and Jimenez, C. R. (2008) Synapse proteomics: current status and quantitative applications, Expert review of proteomics 5, 353–360. 3. Li, K. W., Klemmer, P., and Smit, A. B. (2010) Interaction proteomics of synapse protein complexes, Analytical and bioanalytical chemistry, 397, 3195–3202. 4. Klemmer, P., Smit, A. B., and Li, K. W. (2009) Proteomics analysis of immuno-precipitated synaptic protein complexes, Journal of proteomics 72, 82–90. 5. Li, K. W., Jimenez, C. R., van der Schors, R. C., Hornshaw, M. P., Schoffelmeer, A. N., and Smit, A. B. (2006) Intermittent administration of morphine alters protein expression in rat nucleus accumbens, Proteomics 6, 2003–2008.

6. Choudhary, C., and Mann, M. Decoding signalling networks by mass spectrometry-based proteomics, Nature reviews 11, 427–439. 7. Yates, J. R., Cociorva, D., Liao, L., and Zabrouskov, V. (2006) Performance of a linear ion trap-Orbitrap hybrid for peptide analysis, Analytical chemistry 78, 493–500. 8. Mineki, R., Taka, H., Fujimura, T., Kikkawa, M., Shindo, N., and Murayama, K. (2002) In situ alkylation with acrylamide for identification of cysteinyl residues in proteins during one- and two-dimensional sodium dodecyl sulphatepolyacrylamide gel electrophoresis, Proteomics 2, 1672–1681. 9. Atrih, A., Turnock, D., Sellar, G., Thompson, A., Feuerstein, G., Ferguson, M. A., and Huang, J. T. (2010) Stoichiometric quantification of Akt phosphorylation using LC-MS/MS, Journal of proteome research 9, 743–751.

wwwwwwwwwwwwwwwwww

Chapter 13 SDS-PAGE Immunoblot Analysis Patricia Klemmer Abstract Immunoblotting is a core method in molecular neuroscience to detect and/or quantify single proteins in a complex mixture. The whole procedure covers the solubilization of proteins, their separation by mass on an SDS-polyacrylamide gel, the electrophoretic transfer to a membrane, and the immunodetection of the target proteins by using specific antibodies. Here we provide protocols for SDS-PAGE followed by electroblotting combined with immunostaining. Useful hints are included for method optimization. Key words: Western blotting, SDS-polyacrylamide gel

1. Introduction Immunoblotting, also known as Western blotting, was first described by Towbin and colleagues in 1979 (1). Using this method, single to a few proteins can be detected and quantified in a complex mixture. Over the last 3 decades, this rather simple and inexpensive method has become one of the most widely used techniques in molecular neuroscience. Examples of its applications can be found in Chaps. 3 and 9. The workflow of immunoblotting comprises three steps (a) separation of proteins by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), (b) electrophoretic transfer of the proteins from the gel onto a membrane, and (c) immunodetection of the protein of interest on the membrane (Fig. 1). (a) The separation of proteins by SDS-PAGE. In order to separate proteins by their size, all higher order protein structures need to be removed during sample preparation. Therefore, a reducing agent such as DTT and a strong


169

170

Klemmer

Fig. 1. Overview of the SDS-PAGE immunoblot analysis. The proteins are separated by SDS-PAGE (a) and transferred to a blotting membrane (b). Using specific antibody (c) the target protein is detected as an immunoreactive band in sample 2 (d). For details see text. MW molecular weight marker proteins. 1 and 2 are samples containing different protein constituents. These proteins are not visible (a–c) but are drawn as bands to illustrate their presence in the samples.

anionic detergent, mostly SDS, is included in the sample buffer. In combination with heating, near to complete denaturation occurs to every protein in the sample. All proteins are coated globally with SDS resulting in a negative charge per mass unit. After loading the samples on an SDS-polyacrylamide gel and applying an electric field, the negatively charged proteins migrate through the pores in the gel toward the positive electrode (anode). Each protein move through the gel at different speed that depends on their size; smaller proteins encounter less resistance while moving through the pores and therefore move a longer distance in the gel over a given time than the bigger proteins, which encounter more resistance. In general, the percentage of acrylamide used and the extent of the crosslinking between polymers in the gel will determine the pore size and as such the running behavior of the protein. Mostly, uniform

SDS-PAGE Immunoblot Analysis

171

PAGE is used, but also gradient gels can be generated in order to optimize separation of proteins in certain size ranges. (b) Electroblotting of proteins. In principle, proteins in a gel can be immunodetected directly. However, the gel is not a good matrix for antibody incubation and detection, and the resulting signal intensity is poor. So following gel electrophoresis, the proteins are transferred under an electric field to, and immobilized on, a thin solid supporter, usually nitrocellulose or polyvinylidene fluoride (PVDF) membrane. Proteins are often electroblotted in a tank of transfer buffer. Alternatively, semidry procedures can be used. The latter method has been suggested to be more efficient to transfer whole protein complexes separated by 2D blue-native -PAGE (see also Chap. 7). (c) Immunodetection of the protein of interest on the membrane. After electroblotting, the proteins of interest are detected using a specific antibody. As the free binding sites on the membrane can also absorb antibody, these sites have to be blocked before antibody incubation. The commonly used reagents for blocking are nonfat dried milk, BSA, and/or a low concentration of a mild detergent such as 0.1% Tween20. Usually a two-step immunostaining protocol is followed. This requires some more time than the direct immunostaining with a target specific and labeled antibody. On the other hand, the two-step protocol is considerably more sensitive, and therefore it is adopted in nearly all immunoblotting experiments. In the first step, the specific antibody binds to its target protein. In the second step, the labeled secondary antibody binds to the primary one. The label conjugated to the secondary antibody could be an enzyme (Horseradish Peroxidase or Alkaline Phosphatase), which converts the subsequently added substrate reagents to a visible signal, or might be a fluorescent dye that can be monitored directly. The immunoreactive band that appears on the blot should correspond to the size of the protein. A mismatch with the expected molecular weight may give additional information. Smaller size can point toward possible cleavage of the target protein, while higher molecular weight argues for multimer formation, posttranslational modifications, or an enrichment of charged amino acid residues. Could immunoblotting be applied to a proteome-wide scale? Today, immunoblotting plays mainly a supportive role in proteomics, i.e., it is used as an independent means to confirm the expression differences of individual proteins revealed by the quantitative proteomics experiment. Nevertheless, there is scope for large-scale immunoblotting analysis. Since PVDF membrane is mechanical and chemically

172

Klemmer

robust, fixed proteins can be stained using reversible chemical staining to control for loading differences between samples and this information can be used for normalization purposes (reviewed in (2)). Additionally, it is possible to remove antibodies and reuse the membrane to detect other target proteins. By choosing appropriate SDS-PAGE conditions, multiple proteins can be detected on a single blot, as long as they do not overlap by their molecular weight. Furthermore, using primary antibodies raised in different animals (e.g., mouse and rabbit), the applied secondary antibodies can be labeled with diverse enzymes or fluorescent dyes and so visualized simultaneously. By applying this it can be foreseen that it is possible to perform “proteome-wide” quantitative immunoblotting analysis. A prerequisite is that the large number of antibodies will become available at affordable price. For instance with the establishment of the Neuromab consortium (http://neuromab.ucdavis.edu/) that serves neuroscience community with low-cost monoclonal antibodies against mainly synaptic proteins, the synaptic proteomics scale immunoblotting is within reach.

2. Materials and Solutions/ Buffers

2.1. Sample Preparation

The following buffers and solutions provide a general starting point. The reader is encouraged to change some of the buffer/ solution conditions that may give better result for their particular sample. Hints are given in the “Notes” part. 1. Protein samples should be stored at −80°C. While thawing, they have to be placed on ice to prevent protein degradation. Samples already mixed with SDS sample buffer can be stored at −20°C for 1 month. During thawing, they should not be placed on ice, otherwise SDS will precipitate and a proper coverage of proteins by SDS and thereby a proper electrophoretic separation cannot be guaranteed. 2. 1.0 M Tris/HCl, pH 6.8: Dissolve 6.05 g Tris in 40 mL distilled water and adjust the pH to 6.8 with 6 N HCl. Fill up the buffer to a total volume of 50 mL and store the buffer at 4°C. 3. 5 × SDS sample buffer: Dissolve the following substances in 3.75 mL 1 M Tris/HCl, pH 6.8 plus 3.25 mL distilled water: 1.5 g SDS, 0.015 g brome phenol blue, and 1.16 g DTT. Fill up to 7.5 mL and add 7.5 mL Glycerol. SDS sample buffer is stored in aliquots at −20°C. After thawing SDS sample buffer can be used up to 3 days, but keep it at room temperature, otherwise SDS will precipitate (Note 1).


2.2. SDS-PAGE

173

1. 30% Acrylamide/Bis Solution, 37.5:1 Mixture (30% T, 2.67% C) [Bio-Rad]. This solution needs to be stored at 4°C. 2. 1.5 M Tris/HCl, pH 8.8: Dissolve 45.4 g Tris in 200 mL distilled water and adjust the pH to 8.8 with 6 N HCl. Fill up the buffer to a total volume of 250 mL and store the buffer at 4°C. 3. 1.5 M Tris/HCl, pH 6.8: Dissolve 45.4 g Tris in 200 mL distilled water and adjust the pH to 6.8 with 6 N HCl. Fill up the buffer to a total volume of 250 mL and store the buffer at 4°C. 4. 10% SDS: Dissolve 10 g SDS in 90 mL water with gentle stirring and bring to 100 mL with distilled water. Store the solution at room temperature. 5. 10% APS: Dissolve 10 g APS [Bio-Rad] in 100 mL distilled water. Aliquot the solution and store at −20°C. APS is an initiator for polymerization of acrylamide and it is known to be quite instable. So, take only as much APS as needed and use it within 1 day. 6. TEMED [Bio-Rad]: TEMED, a free radical stabilizer, is included to promote polymerization while casting an SDSpolyacrylamide gel. 7. 1 M Tris/HCl, pH 6.8: Dilute 1.5 M Tris/HCl, pH 6.8 to a final concentration of 1 M. 8. Electrophoresis running buffer: Dilute a 10 × Tris/Glycine/ SDS-Buffer [Bio-Rad] to a 1 × Buffer (Note 2). 9. Protein Molecular Weight Marker: PageRuler™ Plus Prestained Protein Ladder (e.g., Fermentas). 10. Overlay buffer: Saturate n-butanol with distilled water, store at room temperature. 11. Homogenous 10% SDS-polyacrylamide gel. Separation gel: 1.9 mL distilled water 1.7 mL 30% Acrylamide/Bis solution 1.3 mL 1.5 M Tris/HCl, pH 8.8 50 mL 10% SDS 50 mL 10% APS 2 mL TEMED Stacking gel: 1.4 mL distilled water 330 mL 30% Acrylamide/Bis solution 250 mL 1.5 M Tris/HCl, pH 6.8 20 mL 10% SDS 20 mL 10% APS 2 mL TEMED

174

Klemmer

2.3. Electroblotting

1. 10 × CAPSO/Tris: Dissolve 22.1 g CAPSO and 45.4 g Tris in 800 mL distilled water. Fill up to 1 L. 2. 100% Methanol. 3. Transfer Buffer: Mix 100 mL 10 × CAPSO/Tris and 100 mL Methanol with 800 mL distilled water. Store at 4°C (Note 3). 4. PVDF membrane (Bio-Rad) (6.5 × 8 cm for Mini-SDSpolyacrylamide gel) (Note 4). 5. Blotting Filter paper (8 × 10 cm for Mini-SDS-polyacrylamide gel).

2.4. Immunostaining

1. 10 × TBS (Tris-buffered saline): Dissolve 80 g NaCl and 34 g Tris in 800 mL distilled water. Adjust pH to 7.4 and fill up to 1 L (Note 5). 2. 20% Tween: Dissolve 10 mL Tween in 40 mL distilled water. Store at 4°C (Note 6). 3. TBS-T (1 × TBS, 0.05% Tween): Mix 100 mL 10 × TBS with 2.5 mL 20% Tween and fill up to 1 L with distilled water. 4. Blocking Solution: Mix 5 g milk powder with 100 mL TBS-T (Note 7). 5. Primary antibody is a specific antibody, which recognizes the protein of interest. Dilute the primary antibody 1:1,000 in 2 mL TBS-T. The dilution depends on the binding efficiency of the antibody, the expression level of the protein, and the sensitivity of detection method. Secondary antibody: Prepare a 1:10,000 dilution of an Alkaline-Phosphates conjugated antibody, which recognizes the correct species of the primary antibody, with 2 mL TBS-T. 6. ECF Substrate: ECF™ Western blotting reagent pack (GE Healthcare Life Sciences).

2.5. Stripping

1. Stripping Buffer (0.1 M Glycine/HCl, pH 2.5–3.0): Dissolve 3.75 g Glycine in 200 mL distilled water and adjust pH to 2.5 using 6 N HCl. Fill up to 250 mL.

3. Methods In the following section, we will describe the key elements to perform an SDS-PAGE Immunoblotting analysis. In our lab, we are using the SDS-PAGE and blotting equipments from Bio-Rad, the Mini-protean system. 3.1. Sample Preparation

1. While thawing, place the protein sample (not including SDS sample buffer) on ice to prevent protein degradation. 2. Vortex the protein sample to get a homogeneous mixture.


175

3. Mix sample with 5 × SDS sample buffer, and eventual fill up with water. Note that the maximum sample volume which can be loaded is 30 mL using a 1 mm thick Mini-SDSpolyacrylamide gel with ten wells. 4. Vortex the sample and shortly spin. 5. To improve the denaturing process heat sample for 5 min at 90°C. 6. Centrifuge sample to spin down evaporated solution in tube lid. 7. Store the sample at room temperature until loading the gel. Samples mixed with SDS sample buffer should not be placed on ice, otherwise SDS will precipitate and a proper coverage of proteins by SDS cannot be guaranteed. 3.2. SDS-PAGE 3.2.1. Gel Preparation

1. The glass plates should be cleaned with soap to remove fats/ oils and other residues. These substances can disturb electrophoretic separation and will be transferred at the blotting membrane resulting in an increased background staining. 2. Fix a 1 mm spacer plate together with one short glass plate for Mini-Proteane gels in the casting frame and place at the casting tray. 3. Cast a 10% SDS-polyacrylamide gel by mixing all required buffers and solutions for the separation gel (Note 2). 4. Mix the solution gently and pour 4.5 mL of it between the glass plates. 5. Cover the solution with 100 mL Overlay Buffer. 6. When polymerization is completed (~30 min), remove Overlay Buffer by washing with excess water. 7. Pour out the washing water, and fill the space between the glass plates completely with the 4% Stacking gel. Insert a 1-mm comb with ten wells immediately. 8. Wait till Polymerization is finished (~10 min) and continue with electrophoretic separation. 9. To prevent drying of the gel, it should be placed as soon as possible into the running tank with Running Buffer or it should be covered by wetted tissue. Casted gels can be stored in the fridge up to 1 week, if covered with wet tissue.

3.2.2. Electrophoretic Separation

1. Disassemble the casting tray and wash the gel cassette with distilled water to remove all gel particles and solution residues at the outside. 2. Insert gels in a SDS-PAGE running tank and fill both compartments with 1× Running buffer. 3. Remove the comb and splash the wells with running buffer to remove unpolymerized acrylamide and salts by using a syringe.

176

Klemmer

4. Load one well with 3 mL Molecular Weight marker and the others with the samples. Empty wells of the gel should be filled with 1 × SDS sample buffer to minimize diffusion of proteins toward empty lanes. 5. Start the gel electrophoresis with 80 V until the running front marked by brome phenol blue passes the border between stacking and separation gel (~15 min). 6. Then the voltage can be increased to 120 V (~45 min) (Note 8). 7. Stop Run when the blue front almost runs off. 8. Continue with the Coomassie colloidal staining procedure or protein transfer to a PVDF membrane. 3.3. Electroblotting

1. Place a stirrer and a cooling unit in the transfer cell and fill it until half with transfer buffer. 2. While working with the SDS-polyacrylamide gel and blotting membrane always wear gloves, because oil/fat on the hands can interfere with the chemiluminescence signal and can increase background staining on the blot membrane. 3. Remove the SDS-polyacrylamide gel from the gel cassette and put it in a clean plastic container. 4. Equilibrate the gel in transfer buffer, while preparing the membrane. 5. Wet a PVDF membrane in Methanol for 1 min. Afterward soak the membrane in water until membrane stays under the water surface. Then transfer the membrane into a container filled with transfer buffer. 6. Prepare the “Blot-Sandwich” by placing the black side of a cassette in a bowl filled with transfer buffer. Then follow with a prewetted filter pad and two filter papers. Place on top the equilibrated gel and then the PVDF membrane. All air bubbles between the gel surface and blotting membrane need to be carefully removed by pushing gently against the surface or using a roller. Cover with two filter papers and remove air bubbles again using a roller. Finish with a prewetted filter pad and close the cassette. 7. Place “Blot-Sandwich” in the transfer tank. Fill up the tank with transfer buffer. 8. Transfer proteins from Gel to the PVDF membrane at 150 mA (constant current at 4°C) for 2.5 h. The protein transfer should be performed with cold Transfer Buffer and at 4°C to reduce the diffusion effect during protein transfer from SDS-polyacrylamide gel to PVDF membrane (Notes 9 and 10).


3.4. Immunostaining

177

1. Place membrane in a 50 mL Falcon-Tube after transfer is finished. Make sure that proteins face the inside of the tube. 2. Block membrane with 5 mL Blocking Solution for 1 h at RT (Note 6 and 7). 3. Wash membrane 3 × 5 min with TBS-T (Notes 4 and 5). 4. Dilute the primary antibody to appropriate concentration in TBS-T. 5. Incubate membrane with primary antibody for 1 h at room temperature. However, the incubation time with primary antibodies depends on the expression level of the target protein, the affinity of the antibody and the sensitivity of the detection method, and might need optimization. 6. Wash membrane 3 × 15 min with TBS-T. 7. Incubate membrane with secondary antibody for 1 h at room temperature. 8. Wash membrane 3 × 15 min with TBS-T to remove unbound secondary antibody. 9. Incubate membrane with ECF for 5 min and wash shortly in TBS-T to reduce background staining caused by unbound ECF. 10. Scan the chemiluminescence signal with a FLA 5,000 instrument (Fujifilm; laser 453 nm, filter LPG) or equivalent and analyze with Quantity One software from Bio-Rad.

3.5. Stripping

1. Wash membrane in two changes with 25 mL Stripping Buffer for 30 min. 2. For reprobing the membrane, follow the instruction of immunostaining.

4. Notes 1. In some cases, the immunoreactive band on the immunoblot may be diffused. This might due to the presence of a series of posttranslational modifications on the protein, such as different degrees of glycosylation. Here, the result can be improved simply by the removal of the modifications during sample preparation. Sometimes, changing the constituents of the sample buffer may also be helpful, for example, DTT may be replaced by b-Mercaptoethanol. 2. The achieved separation range during SDS-PAGE depends on the pore size of the gel defined by acrylamide concentration and the buffer composition. It is necessary to choose the right conditions for every protein depending on its molecular weight (Table 1).

178

Klemmer

Table 1 Composition of 5 mL separation gel and the corresponding optimal separation ranges for proteins Compounds

6% Gel

8% Gel

10% Gel

12% Gel

15% Gel

Distilled water

2.6

2.3

1.9

1.6

1.1

30% acrylbisacrylamide mix

1.0

1.3

1.7

2.0

2.5

1.5 M Tris/ HCl (pH 8.8)

1.3

1.3

1.3

1.3

1.3

10% SDS

0.05

0.05

0.05

0.05

0.05

10% APS

0.05

0.05

0.05

0.05

0.05

TEMED

0.004

0.003

0.002

0.002

0.002

Optimal separation range of proteins

60–250 kDa

40–100 kDa

20–70 kDa

20–60 kDa

10–40 kDa

3. The efficiency of electrotransblotting depends on the chosen transfer buffer and applied parameters to acquire an electric field. Tris-Glycine buffer containing some SDS or carbonate buffer at pH 9.9 is a good alternative to CAPSO/Tris (3). 4. In some cases, immunostaining of PVDF membranes can result in high background staining. This may be alleviated by choosing appropriate blocking solutions and antibody diluents (Notes 5–7). Nitrocellulose, known for excellent protein binding and retention capabilities, is a good alternative to reducing the high background problem. Nevertheless, PVDF membrane shows a better linearity range of a protein dilution series, and a higher mechanical robustness, which allows reversible protein staining and stripping of the membrane. 5. For immunostaining, both PBS and TBS are commonly used. However, PBS can interfere with AP signals, in which case TBS should be used. 6. Nonionic detergents like Tween-20 and NP40 are used as additives to decrease nonspecific binding . However, it has been reported that Tween-20 (≥0.3%) can remove proteins from the matrix. So it is necessary to find a good balance between background staining and signal intensity for your antibody-protein combination. 7. Several blocking solutions have been reported with different effects on background staining and immunosignal intensities (4).


179

Membranes can be blocked with nonfat dried milk, BSA, fetal calve serum, and gelatine. 8. The quality of the electrophoresis can be examined by the running pattern of the prestained molecular weight marker. 9. The size of the target protein does not only determine the acrylamide percentage of the SDS-PAGE used, it also should be considered for the electrophoretic transfer. Small proteins are easy to elute from the gel, but they may migrate through the blotting membrane, so membranes with small pore size should be used. This can be checked by placing a second membrane behind the first to capture proteins that are transferred through the first membrane. High molecular weight proteins on the other hand are transferred slowly and may be retained in the gel even after prolonged transblotting. The complete protein transfer can be confirmed by staining the SDS-polyacrylamide gel with Coomassie. Several possibilities have been discussed to ensure a complete transfer across a wide range of protein sizes (3). Changing the transfer buffer (Note 3) or transfer parameters could be useful. We usually transfer overnight with a constant voltage at 40 V, when proteins bigger than 90 kDa will be transferred. 10. The prestained molecular weight marker evaluates a successful protein transfer. However, this is a rather crude measure, because it gives only information for a single lane. Reversible chemical protein stains (reviewed in (2)) are quiet useful, because they confirm successful blotting over the entire membrane and do not interfere with further immunodetections. In addition, they can be used as a loading control for normalization purpose instead of “housekeeping” proteins. In our lab, we use Ponceau S staining or the reversible protein detection kit from Sigma. References 1. Towin, H., et al. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. 76, 4350–4354 2. Westermeier, R. and Marouga, R. (2005) Protein Detection Methods in Proteomics Research. Bioscience Report 25, 19–32 3. Bolt, M. and Mahoney, P. (1997) Highefficiency blotting of proteins of diverse sizes

f ollowing sodium dodecyl sulphatepolyacrylamide gel electrophoresis. Anal. Biochem. 247, 185–192 4. Spinola, S. M. and Cannon, J. G. (1985) Different Blocking Agents Cause Variation in the Immunologic Detection of Proteins Transferred to Nitrocellulose Membranes, Journal of Immunological Methods 81, 161–165

wwwwwwwwwwwwwwwwww

Chapter 14 Phosphoproteomics by Highly Selective IMAC Protocol Chia-Feng Tsai, Yi-Ting Wang, Pei-Yi Lin, and Yu-Ju Chen Abstract Protein phosphorylation plays an important role in biological process such as cell differentiation, cell cycle control, metabolism, and apoptosis. Toward global analysis of the phosphoproteome, enrichment is an essential step to overcome analytical challenges associated with the nature of phosphoprotein, including their dynamic modification patterns, substoichiometric concentrations, heterogeneous forms of phosphoproteins, and low mass spectrometric response. Here, based on detailed evaluation of the capture and release mechanism in immobilized metal affinity chromatography (IMAC), we provide a pH/acid-controlled IMAC protocol for phosphopeptide purification with high specificity and lower sample loss. Based on a model study on non-small-cell lung cancer cell, better than 90% phosphopeptide enrichment specificity can be achieved without the use of commonly adapted methyl esterification procedure. In addition, the protocol is compatible to fractionation using SDS-PAGE. We have successfully employed the pH/acid-controlled IMAC enrichment strategy to characterize over 2,360 nondegenerate phosphopeptides and 2,747 phosphorylation sites in H1299 lung cancer cell line. We expect that the simple and reproducible IMAC protocol can be applied, fully automated or manual, for large-scale identification of the vastly under-explored phosphoproteome associated with neurodegenerative diseases. Key words: IMAC, SDS-PAGE, Phosphoproteomics, Mass spectrometry

1. Introduction 1.1. Significance of Protein Phosphorylation

Protein phosphorylation and dephosphorylation, reversible modification catalyzed by kinases and phosphatases, are key steps in cellular signaling to initiate functions such as signal transduction, cell differentiation, cellular development, cell cycle control, and metabolism. An increasing number of human diseases have been discovered to involve mutations, overexpression, or malfunction of protein kinases and phosphatases as well as of their regulators and effectors (1). In the nervous system, for example, protein phosphatases are comprised in dynamic protein complexes in specialized subcellular locations, initiating timely dephosphorylation


181

182

Tsai et al.

of various neuronal phosphoproteins. This regulates the reaction of individual synapses to neural activity and controls synaptic plasticity and affects enzymes related to learning and memory. Thus, dysfunction of these protein phosphatases caused cognitive shortage associated with aging and dementias or neurodegenerative diseases (2). To elucidate the molecular basis of these processes, it is crucial to identify the specific phosphorylation sites and quantify their temporal and dynamic changes. 1.2. Principle of Immobilized Metal Ion Affinity Chromatography

Recently, mass spectrometry (3) has emerged as a reliable and sensitive method to identify protein phosphorylation sites. Despite the advances in MS, characterization of site-specific phos phorylation has been challenged by the technical difficulty associated with their dynamic modification patterns, low-abundant presence in cell, and heterogeneous forms of phosphoprotein (4). Methodologies that specifically enrich the transient phospho- subproteome in a robust, comprehensive manner are important to study the phosphorylation-dependent cellular signaling. On the basis of the interaction between metal ion, such as Zn(II), Fe(III), Ga(III), and phosphate group (Fig. 1a), immobilized metal affinity chromatography (IMAC) was originally discovered by Andersson and Porath as an easy, economic, and generic enrichment protocol for all types of phosphorylation (5–7) and rapidly evolved for large-scale enrichment protocol for phosphoproteomics study (8, 9). To chelate the metal ions, nitrilotriacetic acid (NTA) and iminodiacetic acid (IDA) are the two most commonly used compounds to conjugate to a solid support such as chromatographic resins or magnetic beads. After the NTA/IDA moiety chelates metal ions, unoccupied coordination sites on metal ion will interact with negatively charged phosphopeptides. The metal ion has an empty orbital as the electron acceptor, which can interact with spare electron pairs in the negatively charged peptides through electrostatic interactions and coordination bonding. As shown in Fig. 1b, nonspecific capture of peptides containing acidic amino acids, such as glutamic acid, aspartic acid, histidine, and cysteine, has been criticized as a serious drawback in IMAC. To reduce nonspecific binding, optimized protocols to change buffer pH (6, 10) or washing conditions (11, 12) have been reported with various degrees of specificity for enrichment and identification of phosphopeptides. Despite the reported successes of standard proteins (6, 13), the performance of IMAC at the proteome-wide level has low specificity (60–70%) and thus yields low number of identified phosphoproteins. Although methyl esterification before IMAC enrichment can eliminate nonspecific binding (11, 14–16), however, the additional chemical reaction may cause serious sample loss (3, 17).

Phosphoproteomics by Highly Selective IMAC Protocol

183

Fig. 1. Factors that affect immobilized metal affinity chromatography (IMAC) performance. (a) Specific binding between phosphopeptide and metal ions is dependent on the pH in solution. (b) The performance may be complicated by competitive binding from acidic peptides and substances containing carboxylic acid group in solution.

The IMAC method is easily coupled with separation techniques to reduce sample complexity and increase coverage for phosphopeptide enrichment. Tandem purification by either cation exchange chromatography (SCX) (18) or anion exchange chromatography (SAX) (19) in combination with IMAC has been interfaced to increase the enrichment specificity up to 75%. More recently, hydrophilic interaction chromatography (HILIC) combined with IMAC was reported as an effective strategy for largescale identification of phosphoproteins (20). On the basis of separation of the strong hydrophilic phosphate group in the first HILIC fraction, subsequent enrichment by IMAC was reported to achieve nearly 99% purification specificity without additional derivatization or chemical modification. From the chemical standpoint, enrichment vs. nonspecific binding is determined by the binding affinity between metal ion and phosphopeptides and the competition by nonphosphopeptides or components in the buffer solution. A rational IMAC protocol can be designed based on the chemical nature of the system components and their contribution to specific or competitive binding. In our previous study (21), we reported a pH/acid-controlled IMAC protocol for comprehensive profiling of the phosphoproteome with low sample loss and high- enrichment selectivity. The results indicated that one-step IMAC method with low sample loss and high specificity can be rationally designed by controlling salt, pH, and the structure and concentration of organic acid used in the protocol.

184

Tsai et al.

By incorporation with either high-resolution liquid c hromatography (LC) or SDS-PAGE fractionation, the optimized single-step pH/acid-controlled IMAC protocol can significantly increase the number of phosphoprotein identifications for more in-depth phosphoproteomic profiling. On the proof-of-principle demonstration for the analysis of non-small-cell lung cancer cells, our results contradicted the prevailing view that IMAC has low selectivity for phosphopeptide enrichment. A total of 926 phosphorylation sites (386 phosphoproteins/637 phosphopeptides) with 96% selectivity were identified in a single LC-MS/MS analysis (21).

2. Material 2.1. Reagent

The chemicals used are listed below. It is important to obtain the highest purity of all chemicals. Comparable products from other suppliers should also be effective. 1. Standard protein (a) a-Casein (Sigma Aldrich) (b) b-Casein (Sigma Aldrich) 2. Cell culture and lysate (a) Human non-small cell lung carcinoma cell line (H1299) (b) RPMI 1640 medium (HyClone Logan) (c) Fetal bovine serum (GibooBRL) (d) Penicillin G (GibooBRL) (e) Hydrogenperoxide (Merck) (f) Sodium orthovanadate (Sigma Aldrich) 3. Enzymatic digestion (a) Triethylammonium bicarbonate, TEABC (Sigma Aldrich) (b) Trifluoroacetic acid, TFA (Sigma Aldrich) (c) Acetonitrile, HPLC grade (Merck) (d) Modified trypsin (Promega) (e) Formic acid, FA (Sigma Aldrich) (f) Ammonium persulfate, APS (Amersham Pharmacia) (g) N,N,N ¢,N ¢-tetramethylenediamine, TEMED (Amersham Pharmacia) (h) Acrylamide/bisacrylamide, 40%, v/v, 29:1 (Bio-Rad) 4. Material of IMAC protocol (a) 500 mm i.d. PEEK™ tubing (Upchurch Scientific/ Rheodyne)


185

(b) 0.5 mm SST frit disk (Upchurch Scientific/Rheodyne) (c) Precolumn Filter (Upchurch Scientific/Rheodyne) (d) Acetic acid, AA (J. T. Baker) (e) Ni-NTA spin column (Qiagen) (f) Ethylenediaminetetraacetic acid, EDTA (Merck) (g) Iron chloride, FeCl3 (Sigma Aldrich) (h) Ammonium phosphate, NH4H2PO4 (Sigma Aldrich)

3. Procedure 3.1. Preparation of Cell Lysate

1. The human non-small lung carcinoma cell line (H1299) was cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum and 1% penicillin G at 37°C in a 5% CO2 atmosphere. 2. Cells were stimulated with or without 500 mM pervanadate (pH = 10 with 0.14% H2O2) for 50 min before harvesting cells. 3. Cells were washed with ice-cold PBS for 3 times and lysed in modified RIPA buffer (10 mM Tris–HCl, pH 7.4, 158 mM NaCl, 1 mM EGTA, 1% Triton X-100, 1% deoxycholate, 0.1% SDS, 1 mM DTT, 100 mM sodium vanadate, 100 mM sodium fluoride, and 100 mM protease inhibitor) (see Note 1). 4. Measure protein concentration using Bradford or BCA protein assay. For H1299 cell lysate, the concentration should be approximately 5–10 mg/mL. At this point, samples can be stored at −80°C for several weeks.

3.2. SDS-PAGE Separation and In-Gel Digestion (See Note 2)

1. For large-scale identification of phosphoprotein in H1299 cell, 4 mg of cell lysate was separated by 10% SDS-PAGE (8.6 cm × 6.8 cm × 1.5 mm). The protein loading capacity in SDS-PAGE is dependent on the gel size. The gel capacity in this protocol is approximately 4 mg. If your sample is less than 2 mg, we suggest skipping the Sect. 3.2 and directly adapting the Sect. 3.3. 2. To visualize SDS-PAGE image, a gel slice of roughly 2-cm wide was cut and stained with coomassie blue (shown in Fig. 6a). 3. The remaining gel was then cut into ten gel slices based on molecular weight. Each band was cut into pieces of approximately 0.9 mm3, washed with MilliQ water, and destained twice with 25 mM TEABC (pH 8) in 50% (v/v) ACN for 15 min. Gel slices were dehydrated with 100% ACN and dried by vacuum centrifugation at room temperature.

186

Tsai et al.

4. Dried gel pieces were rehydrated in approximately 400–500 mL 25 mM TEABC (pH = 8) containing 10 ng/mL trypsin until the gel pieces were fully immersed. The digestion was carried out at 37°C overnight. 5. Tryptic peptides were extracted 3 times with 5% (v/v) FA in 50% (v/v) ACN for 30 min and dried completely by vacuum centrifugation at room temperature. After this step is completed, please proceed to Sect. 3.4. At this point, samples can be stored at −30°C for several weeks. 3.3. Gel-Assisted Digestion (22)

1. The proteins were incorporated into a gel directly in the Eppendorf vial with acrylamide/bisacrylamide solution (40%, v/v, 29:1), 10% (w/v) ammonium persulfate, 100% TEMED at a 14:5:0.7:0.3 ratio (v/v). The optimal protein concentration should be around 6 mg/mL. If the protein concentration is too high, it should be diluted to 6 mg/mL. 2. The gel was cut into small pieces (ca. 1 × 1 mm). Note that smaller pieces may clog pipette tips. 3. The small gel pieces were washed 3–5 times with 1 mL of TEABC containing 50% (v/v) ACN. The gel has to be washed until no bubbles were generated from detergents. 4. The gel samples were further dehydrated with 100% ACN and then completely dried by vacuum centrifugation at room temperature. 5. Proteolytic digestion was then performed with trypsin (protein/trypsin = 50:1, g/g) in 25 mM TEABC with incubation overnight at 37°C. 6. Tryptic peptides were extracted 3 times with 5% (v/v) FA in 50% (v/v) ACN for 30 min. 7. The extraction peptides were dried completely by vacuum centrifugation at room temperature. At this point, samples can be stored at −30°C for several weeks.

3.4. Preparation of Automatic IMAC Purification System

1. Suspend NTA-Ni2+ bead from one spin column in 1 mL 0.5% acetic acid, pH 3.0. 2. One end of the IMAC column was capped with a 0.5 mm frit disk enclosed in stainless steel column-end fitting. 3. Load the Ni-NTA resin by packing 250 mL Ni-NTA resin into a 5- or 10-cm micro column (500 mm i.d. PEEK™ tubing) by using syringe manually under very gentle pressure (1 drop/2 s). 4. The packed column was connected to the autosampler and HP1100 solvent delivery system with a flow rate of 13 mL/ min (the expected pressure is below 300 psi and the upper limited pressure must be less than 600 psi). The detail configuration is shown is Fig. 2.


187

Fig. 2. The detailed configuration of automatic IMAC system.

3.5. Phosphopeptides Purification Procedure (21)

1. Ni2+ ions were removed with 100 mL of 50 mM EDTA in 1 M NaCl. 2. Equilibrate with loading buffer (6% acetic acid, pH 3.0) for 20 min. 3. Activate the column with 100 mL of 0.2 M FeCl3. 4. Equilibrate with loading buffer for 25 min. In this step, the use of organic acid, acetic acid, can occupy the surface of metal ion and minimize the acidic nonspecific peptide binding. 5. The tryptic peptides were reconstituted in 100 mL loading buffer and pH was adjusted to 3.0. If the pH of sample is above 3.0, the pH can be adjusted by 100% acetic acid. If the pH of sample is below 3.0, the pH can be adjusted by 1 N NaOH. 6. Before sample loading, the sample must be centrifuged at 16,000 × g for 5 min at 4°C to remove gel debris and transfer the supernatant into new tubes for loading into IMAC column. The tryptic peptides were loaded into IMAC column. For the remaining protein analysis, the flow-through can be collected in a 0.5 mL vial for further analysis. 7. Equilibrate with loading buffer for 15 min. In this step, the loading buffer serves as washing buffer to compete with acidic amino acids-containing nonphosphopeptide. 8. The nonspecifically bound peptides were removed with 100 mL washing solution consisting of 75% (v/v) loading buffer and 25% (v/v) ACN. 9. Equilibrate with loading buffer for 20 min.

188

Tsai et al.

10. The bound phosphopeptides were eluted from the IMAC column with 100 mL 200 mM NH4H2PO4 and collected into 0.5 mL vial for 20 min for further analysis. 11. Eluted peptide samples were dried by vacuum centrifugation at room temperature (see Note 3). 12. If you want to regenerate the column, go back to step 1 and repeat the whole process. 3.6. Peptide Desalting and Concentration

1. The dried phosphopeptides were acidified by 20 mL 0.1% TFA. 2. Desalting with Zip-Tip C18 pipette tips (Millipore) according to the manufacturer’s instructions. Wash and condition the material with 60 mL of 100% ACN and then with 100 mL of 50% ACN 0.1% TFA. 3. Equilibrate with 200 mL of 0.1% TFA. 4. Load sample in 0.1% TFA. 5. Wash/desalt with 300 mL of 0.1% TFA. 6. For matrix-assisted laser desorption/ionization (MALDI) mass spectrometry analysis, the eluted phosphopeptides with 50% ACN (v/v), 0.1% v/v TFA (v/v) can be directly mixed with matrix solution of DHB (10 mg/mL, saturated in 50% ACN (v/v), 0.1 % v/v TFA (v/v)) and directly spotted onto the sample plate. For LC-MS/MS analysis, the peptide can be eluted with 50% ACN (v/v), 0.1% v/v TFA (v/v) and then dried under vacuum for further analysis (see Note 4).

3.7. MALDI-TOF MS Analysis

For MALDI MS analysis on a 4800 MALDI TOF/TOF Analyzer (Applied Biosystems, Foster City, CA), 0.5 mL of ZipTip eluate was mixed with 0.5 mL of matrix solution (20 mg/mL DHB in 50% ACN and 1% H3PO4) and spotted onto stainless steel plates and air-dried. Acquisition was carried out in the positive reflector mode with an acceleration voltage of 20 kV, 16% grid voltage, and a low-mass gate of 1,000 Da. A typical spectrum was generated by 1,000 laser shots. Raw spectra were processed for baseline subtraction and noise removal using Data-Explorer software (Applied Biosystems). Data analysis is detailed in the Anticipated Results.

3.8. LC-MS/MS Analysis

Before LC-MS/MS analysis, the reconstituted phosphopeptides must be centrifuged at 16,000 × g for 5 min at 4°C and transfer the supernatant into new tubes for LC-MS/MS analysis. Purified phosphopeptides were reconstituted in 4 mL of buffer A (0.1% FA in H2O) and analyzed by LC-Q-TOF MS (Q-TOF Premier, Waters Corp). Samples were injected onto a 2 cm × 180 mm capillary trap column and separated by a 75 mm × 25 cm nanoACQUITY™ 1.7 mm BEH C18 column. The bound peptides were eluted with a linear gradient of 0–80% buffer B (buffer A, 0.1% FA in H2O; buffer B, 0.1% FA in ACN). Use the liquid chromatography gradient described in Table 1 for complex samples

88

85

80

75

70

65

60

55

20

20

99

1

12

20

25

33

40

47

57

60

63

65

1

80

80

45

40

35

30

25

20

15

12

105

103

95

85

75

65

55

45

30

20

1

Time (min)

99

20

20

55

60

65

70

75

80

85

88

Mobile phase A (%)

Mobile phase B (%)

Mobile phase A (%)

Time (min)

120 min Gradient

80 min Gradient

1

80

80

45

40

35

30

25

20

15

12

Mobile phase B (%)

170

167

160

155

145

126

100

75

45

25

1

Time (min)

99

20

20

55

60

65

70

75

80

85

88

Mobile phase A (%)

180 min Gradient

1

80

80

45

40

35

30

25

20

15

12

Mobile phase B (%)

170

167

160

155

145

125

100

75

45

25

1

Time (min)

99

20

20

55

60

65

70

75

80

85

88

Mobile phase A (%)

210 min Gradient

Table 1 Gradient used for separation of tryptic phosphopeptides by reverse phase LC-MS/MS

1

80

80

45

40

35

30

25

20

15

12

Mobile phase B (%)

255

250

240

225

200

165

130

100

75

50

1

Time (min)

99

20

20

55

60

65

70

75

80

85

88

1

80

80

45

40

35

30

25

20

15

12

Mobile Mobile phase A phase B (%) (%)

270 min Gradient

190

Tsai et al.

a nalysis. MS was operated in ESI positive V mode with a resolving power of 9,000. NanoLockSpray source was used for accurate mass measurement and the lock mass channel was sampled every 30 s. The mass spectrometer was calibrated with a synthetic human [Glu1]-Fibrinopeptide B solution (500 fmol/mL, from Sigma Aldrich) delivered through the NanoLockSpray source. Data acquisition was operated in the data directed analysis. The method included a full MS scan (m/z 400–1,600, 0.6 s) and three MS/MS (m/z 100–1,990, 1.2 s each scan) sequentially on the three most intense ions present in the full scan mass spectrum. The resulting MS/MS data was exported to *pkl (Proteinlynx GlobalServer 2.2.5) data file format. We performed the peptide identification and phosphorylation site assignment using an inhouse version of Mascot v. 2.2 (Matrix science). To evaluate the protein identification false discovery rate, we repeated the search using identical search parameters and validation criteria against a randomized decoy database (23) created by Mascot (from the 68,161 sequence). As the vast majority of phosphopeptides identified showed significant neutral loss of phosphoric acid, the peak with loss of 98 Da, corresponding to the loss of H3PO4 from peptide fragment, was annotated as an additional characteristic signal for phosphopeptides (as shown in Fig. 3). The phosphorylation site of phosphoserine, phosphothreonine, and phosphotyrosine can be determined by the characteristic mass difference of 69, 83, and 243 Da, respectively.

FQ KD

pSE Q

E L

E

Q

Q

Q

T

D

E

T

Q

EDEL Q

Q

QDK E

pS

E

Q

bMax yMax

F

ymax – H3PO4

ym max

y112

bm max

y14

y13 b13--98 8

20

y144-98

y11

y10

y8 y9 b9-98

y5

b4-98 8

b8-98 8

y2

40

b2 y 3

y4

60

b7-9 98 y7

b5-9 98

80

y6 b6-98

100

Intensity (%)

3.9. Database Search and Phosphorylation Site Determination

0 500

1000

1500

2000

m/z

Fig. 3. The annotated MS/MS spectrum of standard phosphopeptide, FQpSEEQQQTED ELQDK enriched from b-casein.


191

4. Anticipated Results 4.1. Purification Specificity and Recovery of IMAC

The pH/acid-controlled IMAC protocol is a simple and efficient method for enrichment of phosphopeptides from phosphoproteins as well as from highly complex samples. To illustrate the anticipated results, we prepared a peptide mixture containing tryptic peptide digest of two standard phosphoproteins (a- and b-casein). Figure 4 shows the results obtained before and after enrichment of phosphopeptide enrichment using IMAC protocol and MALDI mass spectrometry. Before affinity extraction, the presence of a number of nonphosphopeptides in the mixture seriously suppressed the detection of phosphopeptides (Fig. 4a). After enrichment, nine phosphopeptides were clearly observed in MALDI mass spectrum (Fig. 4b). It is noted that LC-MS/MS can also be used for the same purpose. The assignment of these peptides can be found in Table 2.

Fig. 4. MALDI-TOF mass spectra of 20 pmol of a-casein and b-casein (a) before and (b) after phosphopeptide enrichment by IMAC. The symbols, a and b, represent purified phosphopeptides. The detailed annotation is listed in Table 2.

192

Tsai et al.

Table 2 Summary of the phosphopeptides from the digested a- and b-casein standard proteins detected by MALDI-TOF MS Number of phosphate group [M + H]+ (Da) Symbol

Phosphopeptide sequence TVDMEpSTEVFTK

(a-S2, 153–164)

1

1,466.61

a1

VPQLEIVPNpSAEER

(a-S1, 121–134)

1

1,660.79

a2

YLGEYLIVPNpSAEER

(a-S1, 104–119)

1

1,832.83

a3

DIGpSEpSTEDQAMEDIK

(a-S1, 58–73)

2

1,927.69

a4

YKVPQLEIVPNpSAEER

(a-S1, 119–134)

1

1,951.95

a5

FQpSEEQQQTEDELQDK

(b, 33–48)

1

2,061.83

b1

NTMEHVpSpSpSEEpSIISQETYK

(a-S2, 17–36)

4

2,619.04

a6

QMEAEpSIpSpSpSEEIVPNpSVEAQK

(a-S1, 74–94)

5

2,720.91

a7

NANEEEYpSIGpSpSpSEEpSAEVATEEVK (a-S2, 61–85)

4

3,008.01

a8

4.2. Optimization of LC-MS/MS for Complex Sample

The performance of the optimized IMAC was demonstrated on a proteomic scale. The H1299 lung cancer cell lysate was digested with trypsin, subjected to IMAC purification, and analyzed by LC-MS/MS. For the complex proteomic sample, optimization of LC gradient can efficiently increase the identification coverage of the phosphoproteome. As shown in Fig. 5, single LC-MS/MS analysis of 550 mg lysate using a 80-, 120-, 180-, 210-, or 270min gradient identified 244 phosphoproteins/361 phos phopeptides, 258 phosphoproteins/403 phosphopeptides, 286 phosphoproteins/442 phosphopeptides, 386 phosphoproteins/637 phosphopeptides, and 376 phosphoproteins/596 phosphopeptide, respectively. The extension of LC gradient effectively increased the number of phosphoproteins. The false discovery rates as determined by decoy database search ranged from 0.42 to 0.94%. Most importantly, nearly 100% enrichment specificity (96–97%) from the cell lysate was observed under all conditions.

4.3. SDS-PAGE Fractionation

To exemplify the performance of the SDS-PAGE/IMAC approach for large-scale phosphorylation studies, we applied this method to H1299 lung cancer cell line. The gel shown in Fig. 6a was cut


193

Fig. 5. Increase number of identified phosphoproteins from 550 mg of H1299 cell lysate by extended gradient in LC-MS/MS. The number of identified phosphoproteins is shown in bar graph (left y-axis) with phosphopeptides indicated within each bar. The enrichment specificity is shown in line graph (right y-axis). The analysis was performed using the pH/acid-controlled IMAC procedure and single LC-MS/MS analysis. The LC gradients were performed at 80, 120, 180, 210, and 270 min (figure adapted from Ref. (21)).

into ten slices and subjected to trypsin digestion, IMAC purification, and 80 min LC-MS/MS analysis. Using the MASCOT search engine (p F5 »40 kDa) than in the lower molecular-weight gel bands (Fig. 6b). Taking into account the nondegenerate phosphopeptides (indicated in each bar Fig. 6b) and the corresponding number of phosphoproteins in each gel band, the average number of phosphopeptides per protein ranged from 1.5 to 2.2 phosphopeptides. For all gel slices, monophosphorylation sites represented the major group (>76%). However, doubly phosphopeptides and polyphosphopeptides ranged from 4.8 to 24.3% (Fig. 6c).

194

Tsai et al.

Fig. 6. Summary of identified phosphoproteins in H1299 non-small-cell lung cancer cells by SDS-PAGE fractionation and LC-MS/MS. (a) The percentage of identified phosphoproteins in each gel slice (F1-F10) represented by different colors is plotted as a function of molecular mass (kDa). (b) The number of phosphoproteins (left y-axis), phosphopeptides (indicated within the bar), and the average number of phosphorylation peptides per identified phosphoprotein (indicated on the top of bar) is compared in each gel slice. (c) Distribution of mono-, di-, and polyphosphorylated phosphopeptides is shown in each gel slice (figure adapted from Ref. (21)).

5. Notes 1. During the lysis, cells and lysates should be kept at 4°C at all times. 2. In experimental procedure Sect. 3.2, this is an optional step to either fractionate the complex sample or purify target protein. If the sample did not need to use SDS-PAGE fractionation strategy, you can directly adapt the procedure of Sect. 3.3. 3. Due to the labile nature of the phosphate group, it is recommended that the eluted phosphopeptides are stored at −30°C no more than 1 week. 4. It is recommended to continue to next step for mass spectrometric analysis and does not stop at this step. If the purified peptides need to be stored at this point, store them at −30°C for only 2–3 days.


195

5. For effective phosphopeptides purification and avoiding the contamination, we suggest a new packed column for standard phosphopeptides purification for 5 times (you can store the column at 4°C). If the packed column was used to purify complex sample such as cell line or tissue, we suggest one column for phosphopeptides purification only one time. 6. It is possible to perform the IMAC purification directly in the Ni-NTA spin column. The purification specificity is more than 80%. But the automatic phosphopeptides purification system can provide more sensitivity and stability than spin column. 7. This protocol can be used for SILAC-based quantitation or label-free quantitation for quantitative phosphoproteomics analysis. Specifically for label-free quantitation, this automatic phosphopeptides purification can minimize the system variation between different batches. The quantitation accuracy is around 10–12% for standard phosphoprotein (data not shown).

Acknowledgments This work was supported by Academia Sinica and the National Science Council in Taiwan. We thank Dr. Jeou-Yuan Chen for providing human non-small cell lung carcinoma cell line (H1299). References 1. Hunter, T. (1995) Protein kinases and phosphatases: The Yin and Yang of protein phosphorylation and signaling, Cell 80, 225–236. 2. Mansuy, I. M., and Shenolikar, S. (2006) Protein serine/threonine phosphatases in neuronal plasticity and disorders of learning and memory, Trends in Neurosciences 29, 679–686. 3. Ian, I. S., Ty, T., and Daniel, F. (2001) 18O Labeling: a tool for proteomics, Rapid Communications in Mass Spectrometry 15, 2456–2465. 4. Mann, M., Ong, S.-E., Grønborg, M., Steen, H., Jensen, O. N., and Pandey, A. (2002) Analysis of protein phosphorylation using mass spectrometry: deciphering the phosphoproteome, Trends in Biotechnology 20, 261–268. 5. Andersson, L., and Porath, J. (1986) Isolation of phosphoproteins by immobilized metal

(Fe3+) affinity chromatography, Analytical Biochemistry 154, 250–254. 6. Posewitz, M. C., and Tempst, P. (1999) Immobilized Gallium(III) Affinity Chromatography of Phosphopeptides, Analytical Chemistry 71, 2883–2892. 7. David, C. A. N., Townsend, R. R., Christine, R. R., Verkman, A. S., Elmer, M. P., and Darren, B. G. (1997) Evidence for phosphorylation of serine 753 in CFTR using a novel metal-ion affinity resin and matrix-assisted laser desorption mass spectrometry, Protein Science 6, 2436–2445. 8. 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, Mol Cell Proteomics 4, 310–327.

196

Tsai et al.

9. Ficarro, S. B., McCleland, M. L., Stukenberg, P. T., Burke, D. J., Ross, M. M., Shabanowitz, J., Hunt, D. F., and White, F. M. (2002) Phosphoproteome analysis by mass spectrometry and its application to Saccharomyces cerevisiae, Nat Biotech 20, 301–305. 10. Corthals, G. L., Aebersold, R., Goodlett, D. R., and Burlingame, A. L. (2005) Identification of Phosphorylation Sites Using Microimmobi lized Metal Affinity Chromatography, in Abelson, J. N., Simon, M. I., Colowick, S. P., Kaplan, N. O. (eds.) Methods in Enzymology, pp 66–81, Academic Press. 11. Ndassa, Y. M., Orsi, C., Marto, J. A., Chen, S., and Ross, M. M. (2006) Improved Immobilized Metal Affinity Chromatography for LargeScale Phosphoproteomics Applications, Journal of Proteome Research 5, 2789–2799. 12. Kokubu, M., Ishihama, Y., Sato, T., Nagasu, T., and Oda, Y. (2005) Specificity of Immobilized Metal Affinity-Based IMAC/C18 Tip Enrichment of Phosphopeptides for Protein Phosphorylation Analysis, Analytical Chemistry 77, 5144–5154. 13. Seeley, E. H., Riggs, L. D., and Regnier, F. E. (2005) Reduction of non-specific binding in Ga(III) immobilized metal affinity chromatography for phosphopeptides by using endoproteinase glu-C as the digestive enzyme, Journal of Chromatography B 817, 81–88. 14. Salomon, A. R., Ficarro, S. B., Brill, L. M., Brinker, A., Phung, Q. T., Ericson, C., Sauer, K., Brock, A., Horn, D. M., Schultz, P. G., and Peters, E. C. (2003) Profiling of tyrosine phosphorylation pathways in human cells using mass spectrometry, Proceedings of the National Academy of Sciences of the United States of America 100, 443–448. 15. Lee, J., Xu, Y., Chen, Y., Sprung, R., Kim, S. C., Xie, S., and Zhao, Y. (2007) Mitochondrial Phosphoproteome Revealed by an Improved IMAC Method and MS/MS/MS, Mol Cell Proteomics 6, 669–676. 16. Kim, J.-E., Tannenbaum, S. R., and White, F. M. (2005) Global Phosphoproteome of HT-29

Human Colon Adenocarcinoma Cells, Journal of Proteome Research 4, 1339–1346. 17. Speicher, K. D.; Kolbas, O.; Harper, S.; Speicher, D. W. (2000) Systematic analysis of peptide recoveries from in-gel digestions for protein identifications in proteome studies., J. Biomol. Tech. 11, 74–86. 18. Villén, J., Beausoleil, S. A., Gerber, S. A., and Gygi, S. P. (2007) Large-scale phosphorylation analysis of mouse liver, Proceedings of the National Academy of Sciences 104, 1488–1493. 19. Nuhse, T. S., Stensballe, A., Jensen, O. N., and Peck, S. C. (2003) Large-scale Analysis of in Vivo Phosphorylated Membrane Proteins by Immobilized Metal Ion Affinity Chromatography and Mass Spectrometry, Mol Cell Proteomics 2, 1234–1243. 20. McNulty, D. E., and Annan, R. S. (2008) Hydrophilic Interaction Chromatography Reduces the Complexity of the Phosphoproteome and Improves Global Phosphopeptide Isolation and Detection, Mol Cell Proteomics 7, 971–980. 21. Tsai, C.-F., Wang, Y.-T., Chen, Y.-R., Lai, C.-Y., Lin, P.-Y., Pan, K.-T., Chen, J.-Y., Khoo, K.-H., and Chen, Y.-J. (2008) Immobilized Metal Affinity Chromatography Revisited: pH/Acid Control toward High Selectivity in Phosphoproteomics, Journal of Proteome Research 7, 4058–4069. 22. Han, C.-L., Chien, C.-W., Chen, W.-C., Chen, Y.-R., Wu, C.-P., Li, H., and Chen, Y.-J. (2008) A Multiplexed Quantitative Strategy for Membrane Proteomics: Opportunities for Mining Therapeutic Targets for Autosomal Dominant Polycystic Kidney Disease, Mol Cell Proteomics 7, 1983–1997. 23. Elias, J. E., and Gygi, S. P. (2007) Targetdecoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry, Nat Meth 4, 207–214.

Chapter 15 Global Analysis of Ubiquitination David Meierhofer and Peter Kaiser Abstract The covalent attachment of the small protein ubiquitin to other proteins is known to control a host of biological pathways and is emerging as an important regulatory factor in various processes specific to the nervous system. Ubiquitination is also tightly linked to most neurodegenerative disorders. A quantitative, proteome-wide view of the dynamic changes in ubiquitin modification associated with neuronal activity states and various stages of neurodegenerative disorders is therefore desired. Advances in quantitative mass spectrometry and the development of new biological tools make these approaches feasible for many laboratories. We describe here a combination of SILAC-based (stable isotope labeling by amino acids in cell culture) quantitative mass spectrometry and tandem-affinity purification to detect systemwide changes in ubiquitination and ubiquitin chain topologies that will be useful to probe the role of ubiquitin in the nervous system. Key words: Ubiquitination, SILAC, HB-tag, Tandem-affinity purification, Ubiquitin-chain topology, Mass spectrometry

1. Introduction We are only beginning to understand the role of the ubiquitin/ proteasome system in processes such as nervous system development, neuronal plasticity, or neurodegenerative disorders (1–3). Nevertheless, the importance of protein modification with ubiquitin is evident, and proteome-wide approaches describing dynamic changes in ubiquitination profiles promise to advance our understanding of the role of the ubiquitin/proteasome system in neuronal processes. Ubiquitin, a 76 amino acid protein, is highly conserved among Eukaryotes and is best known for its function in labeling other proteins for degradation by the 26S proteasome. Ubiquitination, the covalent attachment of ubiquitin to other


197

198

Meierhofer and Kaiser

proteins, has also nonproteolytic functions such as direct modulation of protein activity or protein localization, which is often caused by changes in protein interaction partners (4). Ubiquitination involves a cascade of reactions that are catalyzed by the E1 (ubiquitin-activating enzyme), E2 (ubiquitin-conjugating enzymes), and E3 (ubiquitin ligases) enzymes, and results in the formation of an isopeptide bond between the carboxyl-terminus of ubiquitin and typically the e-amino-group of a lysine residue in substrate proteins (Fig. 1). Generally, we distinguish between mono, multi, and polyubiquitinated substrates (Fig. 1). While the former two describe the linkage of single ubiquitin molecules to one or more lysine residues in substrates, polyubiquitination involves the formation of ubiquitin chains (5). They are formed through isopeptide linkages between one out of seven different lysine residues in a substrate-anchored ubiquitin and the carboxylterminus of a new ubiquitin moiety. Depending on the specific lysine residue used for ubiquitin chain linkages, different chain topologies with distinct signaling functions are formed (6). Different ubiquitin chain topologies as well as precise ubiquitination sites in substrates can be detected using the mass spectrometric analysis we describe below (Table 1). The basis for this is a 114-Da mass shift due to the C-terminal diglycine motif from ubiquitin that remains linked to ubiquitin-acceptor residues (usually lysines) after trypsin digestion (7, 8). Like other posttranslational modifications, ubiquitination is a reversible modification due to the function of ubiquitin carboxylterminal hydrolases (UCHs) and ubiquitin-specific processing proteases (UBPs). The balance between ubiquitination, deubiquitination, and degradation forms the basis for the highly dynamic character of the ubiquitin proteome (9).

a

b

c

Ub ATP

Ub

Ub

E1 E1

Ub

E1 E1

E2 E2

E2 E2

E3

E2 Substrates

Ub

Mono-Ub

Multi- Ub

Substrate

Substrate

Ub

Ub

Poly-Ub Substrate

Ub Ub

Ub

Ub

Ub

Fig. 1. The three steps of the ubiquitination cascade: (a) Ubiquitin is activated by an E1 ubiquitin-activating enzyme (two E1 enzymes are known). (b) Transfer of ubiquitin from E1 to an ubiquitin-conjugating enzyme E2 (over 30 different E2 enzymes are known). (c) An Isopeptide bond between a substrate lysine and the C-terminal glycine of ubiquitin is created (hundreds of E3 ubiquitin-ligases are known). Substrates can either be mono, multi, or polyubiquitinated. Ub ubiquitin.

Global Analysis of Ubiquitination

199

Table 1 Tryptic signature fragments of ubiquitin chain linkage types (human) Linkage type

Sequence

Precursor MH+(Da)

K6

MQIFVK(GlyGly)TLTGK

1379.772

K11

TLTGK(GlyGly)TITLEVEPSDTIENVK

2402.266

K27

TITLEVEPSDTIENVK(GlyGly)AK

2101.102

K29

AK(GlyGly)IQDK

816.457

K33

IQDK(GlyGly)EGIPPDQQR

1637.824

K48

LIFAGK(GlyGly)QLEDGR

1460.786

K63

TLSDYNIQK(GlyGly)ESTLHLVLR

2244.198

Precursor masses contain the additional 114.04293 Da for the (GlyGly) remnant originating from ubiquitination of the lysine. Masses can differ due to additional modifications, miss cleavages or SILAC labeling, etc.

The complexity of nervous system function and neurodegeneration requires quantitative, system-level approaches for description of the role of the ubiquitin/proteasome in these processes. Such approaches have been developed for yeast, mammalian cells, and mice (8, 10–14). Typically tagged ubiquitin is expressed in cells, and ubiquitinated proteins are purified based on the affinity tag fused to ubiquitin (15). One important aspect in the analyses of ubiquitinated proteins is that purification is performed under completely denaturing conditions to disrupt noncovalent protein interactions. The denaturing purification conditions avoid copurification of proteins associated with ubiquitinated proteins and thus not only reduce sample complexity but also allow identification of ubiquitination substrates by protein ID without the inefficient detection of ubiquitin attachment sites. We describe here one of the strategies that has been applied to the quantitative detection of changes in proteome-wide ubiquitination profiles. The protocol presents a simple and efficient method to purify and analyze ubiquitinated proteins under fully denaturing conditions using the histidine-biotin (HB) tandemaffinity tag (11, 14, 16). Ubiquitin is fused to the HB-tag consisting of an RGS-hexahistidine motive followed by a bacterially derived biotinylation signal (Fig. 2). The biotinylation signal induces the attachment of biotin to a specific lysine residue in the tag in vivo (17). Cells use HB-ubiquitin like the endogenous ubiquitin in protein ubiquitination, and ubiquitinated proteins can be sequentially purified by Ni2+-chelate chromatography and binding to streptavidin sepharose (Fig. 2). Importantly, both purification steps tolerate highly denaturing conditions such as 8 M urea or 6 M guanidinium (14, 16, 18). The high-affinity

200


a

HB-ubiquitin affinity purification pCMV

RGS-6xhis

biotin

UBIQUITIN

IRES

Purr

b Ni2+

1. 2.

Purification on Ni2+ resin

6his

bio UBIQUITIN

ubiquitinated target protein

Elution from Ni2+ resin

3.

Purification on streptavidin

4.

On-bead trypsin digestion

5.

Mass spectrometry

streptavidin

6his bio UBIQUITIN

ubiquitinated target protein

Fig. 2. Tandem-affinity purification of HB-tagged ubiquitin. (a) The HB-ubiquitin-tag consists of an RGS-hexahistidine motif (RGS6xHis) followed by a bacterially derived in vivo biotinylation signaling peptide (biotin). The biotinylation signal peptide induces attachment of biotin in vivo and allows purification of HB- ubiquitinated proteins on streptavidin resins. HB-ubiquitin is expressed in the retroviral vector (pQCXIP). Expression is driven by the CMV promoter, and selection of clones with high expression levels is facilitated by an internal ribosomal entry side (IRES) controlling translation of the puromycin selection marker. (b) Flow diagram of the two-step purification process. Ubiquitinated proteins are sequentially purified on Ni2+ sepharose and streptavidin sepharose. Both purification steps are performed under fully denaturing conditions to minimize background binding. Because of the irreversible nature of the streptavidin-biotin interaction, “onbead” tryptic digestion is used to elute peptides from the streptavidin sepharose for mass spectrometric identification.

interaction between the biotin attached to the HB-tag and streptavidin allows exceptionally stringent purification conditions, but prevents efficient elution (19). “On-bead” tryptic digest is thus used to release peptides from the streptavidin sepharose beads for analysis by mass spectrometry (14, 16). Tandem-affinity purification of ubiquitinated proteins is an effective approach to identify ubiquitin profiles. However, detection of system-wide changes in ubiquitination requires quantitative mass spectrometric strategies. We have used stable isotope labeling with amino acids in cell culture (SILAC) (20–23) (Fig. 3) for quantitative detection of changes in global ubiquitin profiles and changes in ubiquitin-chain topologies in response to various conditions (11). The SILAC approach is ideal for this purpose as it lets the investigator combine the different samples before purification to ensure absolutely identical conditions during the entire process.


Light 12C,14N

201

Heavy

lys/arg

13C,15N

lysis

lys/arg

lysis mix 1:1

Purification and trypsin digest Identification by mass spectrometry 12C14N

Intensity

300

L/H ratio 5:1 200 13C15N

100 0 500

502

504

506

508

m/z

Fig. 3. SILAC strategy for quantitative comparison of proteome-wide ubiquitination profiles. The “light” (12C, 14N arginine and lysine growth medium) and “heavy” (13C, 15N arginine and lysine growth medium) samples are grown under two different conditions according to the experimental plan. Equal amounts of total cell lysates from both samples are mixed and purified under denaturing conditions. Purified proteins are digested with trypsin and prepared for analysis by mass spectrometry. Peptides with identical sequences from the labeled and unlabeled sample are detected as pairs, because 13C15N-arg/lys incorporation results in a defined mass shift. The peak heights of the SILAC pairs are used to calculate the light/heavy (L/H) ratios, which are equivalent to relative abundance changes of the peptide/ protein (SILAC stable isotope labeling by amino acids in cell culture). See also Note 1.

Protein ubiquitination has gained significant attention as a posttranslational modification that controls protein abundance, activity, and localization. Recent advances in affinity purification strategies and quantitative mass spectrometry have allowed proteome-wide descriptions of the dynamics of the ubiquitin/proteasome system. We hope that the protocol we describe here will contribute to our understanding of the role of ubiquitination in normal and pathogenic processes of the nervous system.

202


2. Materials 2.1. Mammalian Cell Culture and Lysis

Quantitative analysis of ubiquitination profile changes is described using the SILAC procedure. For nonquantitative applications, standard growth media can be used. The described purification protocol is identical for both the SILAC strategy and the nonquantitative approach. 1a. Nonquantitative analysis: Dulbeco’s Modified Eagle’s Medium (DMEM; Mediatech, Herndon, VA, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, GIBCO, Bethesda, MA, USA), 1% penicillin/streptomycin (GIBCO, Bethesda, MA, USA). 1b. SILAC-based quantitative experiments: SILAC DMEM medium lacking lysine and arginine (Thermo Scientific, Rockford, IL) supplemented with 10% (v/v) dialyzed fetal bovine serum (FBS, GIBCO, Bethesda, MA, USA), 1% penicillin/streptomycin (GIBCO, Bethesda, MA, USA). The heavy medium is supplemented with 0.028 mg/mL 13C615N4 arginine, and 0.073 mg/mL 13C615N2 lysine (isotopic purity >98 atom %) (Cambridge Isotope Labeling, Andover, MA). The light medium contains the same amount of 12C14N arginine and 12C14N lysine (Sigma Co., St Louis, MO, USA). For triple-labeled SILAC experiments see Note 1. 2. Antibiotics according to selection marker. 3. 10× PBS buffer: 0.58 M Na2HPO4, 0.17 M NaH2PO4, 0.68 M NaCl, pH 7.4. 4. 27 G needles to shear DNA (Becton Dickinson, NJ, USA). 5. Sonicator (VWR, West Chester, Pennsylvania, USA). 6. 10 mM MG132 (American Peptide, Sunnyvale, CA) dissolved in DMSO, keep frozen at −20°C.

2.2. Purification of HB-Ubiquitinated Proteins

Check the pH of all buffers before use and readjust accordingly. 1. Buffer A-8: 8 M Urea, 300 mM NaCl, 50 mM sodium phosphate buffer pH 8.0 (0.68 mL of 0.5 M NaH2PO4 and 9.32 mL of 0.5 M Na2HPO4 in 100 mL), 0.5% (v/v) Nonidet P-40, pH 8.0. 2. Buffer A-6.3: 8 M Urea, 300 mM NaCl, 50 mM sodium phosphate buffer pH 6.3 (8.22 mL of 0.5 M NaH2PO4 and 1.78 mL of 0.5 M Na2HPO4 in 100 mL), 0.5% (v/v) Nonidet P-40, pH 6.3. 3. Buffer A-6.3-imidazole: same as buffer A-6.3, but also containing 10 mM Imidazol, pH 6.3. 4. Buffer B: 8 M Urea, 200 mM NaCl, 50 mM sodium phosphate buffer pH (9.21 mL of 0.5 M NaH2PO4 and 0.79 mL


203

of 0.5 M Na2HPO4 in 100 mL), 2% (w/v) SDS, 10 mM EDTA, 100 mM Tris, pH 4.3. 5. Buffer C: 8 M Urea, 0.2 M NaCl, 0.2% (w/v) SDS, 100 mM Tris, pH 8.0. 6. Buffer D: 8 M Urea, 0.2 M NaCl, 100 mM Tris, pH 8.0. 7. Ni2+ Sepharose™ 6 Fast Flow beads (GE Healthcare). 8. Immobilized Streptavidin beads (Pierce, Rockford, IL, USA). 9. Poly-Prep® Chromatography Columns (Bio-Rad, Hercules, CA, USA). 10. 25 mM NH4HCO3 buffer, pH 8.0. 11. HPLC-grade H2O. 12. PMSF, 0.5 M solution in isopropanol, stored at 4°C. 2.3. Western Blot Analysis

1. 4× SDS sample buffer: 250 mM Tris-HCl, pH 6.8, 8% (w/v) SDS, 300 mM DTT, 30% (v/v) glycerol, 0.02% (w/v) bromophenol blue. 2. Antibodies for detection of the HB-tag: RGS-His Antibody (Catalog number 34610; Qiagen, Valencia, CA, USA), 1:2,000 in blocking buffer, or Horseradish PeroxidaseConjugated Streptavidin (Catalog number PI21126; Fisher, Pittsburgh, PA, USA), 1:10,000 in TBS-T buffer. 3. 10× TBS-T (Tris-buffered saline with Tween 20), 10× stock: 1.37 M NaCl, 27 mM KCl, 250 mM Tris-HCl, pH 7.4, 1% (v/v) Tween 20. 4. Blocking buffer: 5% (w/v) nonfat dry milk in 1× TBS-T.

2.4. Sample Preparation for MudPIT Analysis

1. 25 mM NH4HCO3. 2. 0.4 mg/mL trypsin (Promega) in 1 mM trifluoroacetic acid (TFA) (see Note 2). 3. Trypsin, (Promega, Madison, WI, USA). 4. “Slick Tubes” (Catalog number 16-8110-03P, PGC Scientific, NC, USA). 5. Strong cation exchange PolySULFOETHYL column (The Nest Group, Inc., Southborough, MA, USA). 6. Buffer C18-A: 2% (v/v) acetonitrile (ACN), 98% (v/v) H2O, 0.1% (v/v) formic acid (FA). 7. Buffer C18-B: 98% (v/v) ACN, 2% (v/v) H2O, 0.1% (v/v) FA. 8. Buffer SCX-A: 5 mM KH2PO4, 30% (v/v) ACN, 0.1% (v/v) FA, adjusted to pH 2.7 with FA. 9. Buffer SCX-B: 5 mM KH2PO4, 350 mM KCl, 30% (v/v) ACN, 0.1% (v/v) FA, adjusted to pH 2.7 with FA. 10. 0.1% (v/v) TFA. 11. Vivapure C-18 Microcolumn (Sartorius, Göttingen, Germany).

204


12. Pepmap C18 capillary column (Dionex, Bannockburn, IL, USA) (length: 15 cm, ID 75 mm), or capillary columns from another vender, or self-packed. 13. 500 fmol/mL BSA tryptic digest (Michrom Bioresources, Auburn, CA, USA).

3. Methods 3.1. Plasmid

3.2. Growth and Lysis of Mammalian Cells Expressing HB-Ubiquitinated Proteins

The retroviral vector pQCXIP (BD Biosciences) expressing human HB-ubiquitin (11) (Fig. 2) can be used to generate cell lines stably expressing HB-tagged ubiquitin. Viral particles are generated in 293 GP2 packaging cells and used to transduce cells according to standard protocols in order to establish a stable cell line expressing HB-ubiquitin. Expression of HB-ubiquitin should be tested by immunoblotting using the RGS-His antibody, and stable cell lines should be periodically maintained with the appropriate antibiotic selection. 1. Grow cells expressing HB-ubiquitin in DMEM (or SILAC DMEM) to about 90% confluency. Use five to ten 150 mm plates. The amount of plates required for the experiment depends on the protein yield of the specific cell line. We recommend growing enough cells to obtain between 10 and 20 mg of total protein. 2. If desired, the proteasome inhibitor MG132 can be used to accumulate ubiquitinated proteins. Cells are incubation at 37°C with 10 mM MG132 dissolved in DMSO for 1.5 h before harvesting. 3. Wash cells on plates twice with 5 mL of ice-cold 1× PBS, pH 7.4. 4. Lyse the cells “on plate” by adding buffer A with 1% PMSF. Pipette several times up and down to make sure that all cells detach and lyse. The lysate gets very viscous due to the presence of chromosomal DNA. Keep the volume as low as possible (e.g., transfer the harvested lysate to the next dish and harvest again). For a total of ten 150 mm plates of cells harvested, use a maximum of 20 mL lysis buffer. 5. Shear the DNA by passing the lysates several times through a 27-G needle, or sonicate on ice for 30 s intervals, until the viscosity is similar to that of water. 6. Centrifuge the lysate at 25,000 × g for 30 min at 4°C. Transfer the clarified supernatant to a fresh tube. Measure the protein concentration. For SILAC experiments, mix equal amounts of proteins from the heavy and the light sample. Save 50 mL aliquots of the lysate for analysis.


3.3. Tandem-Affinity Purification of HB-Ubiquitinated Proteins

205

To distinguish specific ubiquitinated proteins from background proteins, we recommend parallel processing of a sample from cells that do not express HB-ubiquitin. Proteins that are detected in both, the tagged and untagged cell line, are most likely nonspecific background proteins. 1. Use 70 mL of Ni2+ sepharose beads for each 1 mg of protein lysate. Wash the beads 3× with at least 5 bead volumes of buffer A-8 (without PMSF). Pellet the beads by centrifugation at 100 × g for 1 min in a microcentrifuge, and remove the supernatant. 2. Add Ni2+ sepharose beads to the lysate, followed by imidazole to a final concentration of 10 mM to reduce nonspecific binding. 3. Incubate on a rocking platform at room temperature for 4 h, or overnight. 4. Pellet the beads by centrifugation at 100 × g for 1 min, and remove the supernatant. Save 50 mL for analysis (Ni-unbound fraction). 5. Wash the Ni2+ sepharose beads sequentially with 20 bead volumes of buffer A-8, buffer A-6.3, and buffer A-6.3-imidazole, respectively. 6. Elute HB ubiquitinated proteins 2× with 5 bead volumes of buffer B, make sure that the pH is correctly adjusted. Incubate for at least 10 min at room temperature for each elution step, and pool the eluates. Save 50 mL for analysis (Ni2+ eluate). 7. Adjust the pH of the eluate to pH 8.0 (add ~25 mL of 1 M NaOH to each 1 mL eluate). 8. Prepare streptavidin sepharose by washing with 2 × 3 mL of buffer C. Use 15 mL of beads for each 1 mg protein in the whole cell lysate used in the first step of purification. 9. Incubate the Ni2+ sepharose eluate with streptavidin beads in a Poly-Prep chromatography column on a rocking platform overnight at room temperature. 10. Drain the column by gravity. Save 50 mL of the flow-through for analysis (streptavidin unbound fraction). 11. Wash the streptavidin beads (in the Poly-Prep chromatography column) sequentially with 2 × 25 bead volumes of bufferC and buffer-D, respectively. 12. Add 3 mL of 25 mM NH4HCO3, pH 8.0 and allow the column to drain by gravity. 13. Add 1 mL of 25 mM NH4HCO3, pH 8.0 with the Poly-Prep column closed at the bottom, resuspend the beads by pipetting up and down, and transfer the beads to a “slick tube.”

206


14. Collect the beads at the bottom of the tubes by centrifugation at 100 × g for 1 min in a microfuge. Carefully discard the supernatant and wash the beads again with 1.5 mL of 25 mM NH4HCO3, pH 8.0. 15. Collect the beads at the bottom of the tube as above and add 25 mM NH4HCO3 buffer (approximately 50% of the bead volume) in preparation for the trypsin digest (Sect. 15.3.4). 3.4. “On-Bead” Digestion for Mass Spectrometric Analysis

1. Dissolve 20 mg of trypsin in 50 mL of 1 mM TFA in the original glass tube (see Note 2). 2. Incubate the sample (from step 15, Sect. 15.3.3) with trypsin at 37°C for 12–16 h on a rocking platform. Use 1 mg of trypsin for every 2 mg of whole cell lysate used in the first purification step. 3. Carefully collect the supernatant and add FA to a final concentration of 1% (v/v). 4. Reextract tryptic peptides from the beads 2–3× by adding approximately 50% of the bead volume of 25% (v/v) ACN, 0.1% (v/v) FA to the beads. 5. Pool the extracted peptides and concentrate to about 5 mL using a SpeedVac. 6. Add 100 mL of H2O and concentrate (SpeedVac) to about 5 mL. Repeat this step once more. 7. For a 1D analysis, add 5% (v/v) ACN and 2% (v/v) FA to a final volume of 10 mL, for a 2D analysis to a final volume of 100 mL. 8. Samples can be stored frozen at this step, or used immediately for 1D analysis (Sect. 15.3.6), or further separated on an SCX column for complex peptide mixtures (2D-MS analysis, Sect. 15.3.7).

3.5. Western Blot Analysis of Purification

To analyze the efficiency of the purification, the small samples collected from the different purification steps should be analyzed by immunoblotting. 1. To evaluate the efficiency of binding to Ni2+ sepharose, analyze 20 mL of the collected whole cell lysates and 20 mL of the Ni-unbound fraction (step 4, Sect. 15.3.3) by immunoblotting with the anti-RGS-His antibody. 2. To analyze the efficiency of binding to streptavidin, use 10 mL of the Ni2+ eluate (step 6, Sect. 15.3.3) and 10 mL of the streptavidin unbound fraction (step 10, Sect. 15.3.3) for immunoblot analysis with the anti-RGS-His antibody, or horseradish peroxidase-conjugated streptavidin (see Note 3).

3.6. Mass Spectrometry

1. Condition the Pepmap C18 column sequentially in buffer C18-B for 30 min, and buffer C18-A for 30 min. Adjust the column flow rate to 250 nL/min.


207

2. Inject 1 mL of 500 fmol/mL BSA digest and separate the sample at the following gradient using buffers C18-A and C18-B: 0% buffer C18-B for 5 min, 0–35% buffer C18-B for 80 min, 80% buffer C18-B for 5 min, and 0% buffer C18-B for 30 min. Inject 1 mL of 10–50 fmol/mL BSA digest using the same gradient. The column is now ready to be used. 3. The protein digest from step 8, Sect. 15.3.4 can be directly injected onto the column for LC-MS/MS analysis. 3.7. Identification of Proteins by MudPIT Analysis

Peptide separation by ion-exchange chromatography for complex sample mixtures 1. Separate samples (generated at step 8, Sect. 15.3.4) on a strong cation exchange column with the following gradient: 0–5% buffer SCX-B for 2 min, 5–35% buffer SCX-B for 30 min, and 35–100% buffer SCX-B for 10 min. Collect the flow-through and each peak fractions. Collect about 10–20 fractions. 2. Concentrate each fraction to about 5–10 mL in a SpeedVac. 3. Add 180 mL of 0.1% (v/v) TFA to each fraction. 4. Desalt the samples using Vivapure C-18 Microcolumns according to the manufacturer’s instructions. 5. After desalting, concentrate samples to 1–2 mL (using a SpeedVac) and add 10 mL of 5% ACN (v/v), 2% (v/v) FA to each sample for LC-MS/MS analysis.

3.8. LC-MS/MS Analysis

1. Each desalted fraction from step 5, Sect. 15.3.7 can be analyzed by LC-MS/MS (as described in Sect. 15.3.6). 2. Automatically submit the acquired LC-MS/MS data to commercially available search engines, such as MASCOT, Protein Prospector, and/or SEQUEST, for database searching, protein identification, and characterization of posttranslational modifications.

4. Notes 1. A triple SILAC procedure can also be used (24). This allows comparison of three different samples. In case of the triple SILAC strategy light growth media (12C14N arginine and 12C14N lysine), medium growth media (4,4,5,5-D4-lysine and 13C6arginine), and heavy growth media (13C615N2-lysine and 13 C615N4-arginine) are used. 2. The amount of trypsin suggested is a rough estimate. To determine the amount of trypsin required more accurately, measure the protein concentration in the Ni2+ eluate (step 6,

208


Sect. 15.3.3) and in the streptavidin unbound fraction (step 10, Sect. 15.3.3). The difference in protein concentration in these two fractions is a good approximation of the amount of protein bound to the streptavidin beads. For every 1 mg of protein bound to the streptavidin beads use between 0.01 and 0.02 mg of trypsin (the ratio of protein/trypsin is between 1/100 and 1/50). Diluted trypsin solution should be prepared immediately before the digestion and can be stored at −20°C. 3. The RGS-His antibody detects HB ubiquitinated proteins and can be used to measure the efficiency of the purification steps. Horseradish peroxidase-conjugated streptavidin can be used to analyze the purification after elution from the Ni2+ sepharose. However, it is not very useful for analysis of the first purification step because all eukaryotic cells express between four and six endogenous biotinylated proteins. The endogenous biotinylated proteins are relatively abundant and can complicate interpretation of the Western blot results. However, endogenous biotinylated proteins are lost during the first purification step and horseradish peroxidase- conjugated streptavidin is useful for the detection of HB ubiquitinated proteins after the first purification step, providing important information about the efficiency of biotinylating HB-ubiquitinated target proteins.

Acknowledgments This work was supported by NIH (GM66164 and CA113823). David Meierhofer was an Erwin Schrödinger fellow supported by the FWF Austria. References 1. Patrick, G. N. (2006) Synapse formation and plasticity: recent insights from the perspective of the ubiquitin proteasome system. Curr Opin Neurobiol 16, 90–4. 2. Paul, S. (2008) Dysfunction of the ubiquitinproteasome system in multiple disease conditions: therapeutic approaches. Bioessays 30, 1172–84. 3. Segref, A., and Hoppe, T. (2009) Think locally: control of ubiquitin-dependent protein degradation in neurons. EMBO Rep 10, 44–50. 4. Hershko, A., and Ciechanover, A. (1998) The ubiquitin system. Annu Rev Biochem 67, 425–79.

5. Pickart, C. M. (2004) Back to the future with ubiquitin. Cell 116, 181–90. 6. Kim, I., and Rao, H. (2006) What’s Ub chain linkage got to do with it? Sci STKE 330, 18. 7. Kaiser, P., and Wohlschlegel, J. (2005) Identification of Ubiquitination Sites and Determination of Ubiquitin-Chain Architectures by Mass Spectrometry. Methods Enzymol 399C, 266–77. 8. Peng, J., Schwartz, D., Elias, J. E., Thoreen, C. C., Cheng, D., Marsischky, G., Roelofs, J., Finley, D., and Gygi, S. P. (2003) A proteomics approach to understanding protein ubiquitination. Nat Biotechnol 21, 921–6.

Global Analysis of Ubiquitination 9. Nijman, S. M., Luna-Vargas, M. P., Velds, A., Brummelkamp, T. R., Dirac, A. M., Sixma, T. K., and Bernards, R. (2005) A genomic and functional inventory of deubiquitinating enzymes. Cell 123, 773–86. 10. Xu, P., Duong, D. M., Seyfried, N. T., Cheng, D., Xie, Y., Robert, J., Rush, J., Hochstrasser, M., Finley, D., and Peng, J. (2009) Quantitative proteomics reveals the function of unconventional ubiquitin chains in proteasomal degradation. Cell 137, 133–45. 11. Meierhofer, D., Wang, X., Huang, L., and Kaiser, P. (2008) Quantitative analysis of global ubiquitination in HeLa cells by mass spectrometry. J Proteome Res 7, 4566–76. 12. Mayor, T., Graumann, J., Bryan, J., MacCoss, M. J., and Deshaies, R. J. (2007) Quantitative profiling of ubiquitylated proteins reveals proteasome substrates and the substrate repertoire influenced by the Rpn10 receptor pathway. Mol Cell Proteomics 6, 1885–95. 13. Bennett, E. J., Shaler, T. A., Woodman, B., Ryu, K. Y., Zaitseva, T. S., Becker, C. H., Bates, G. P., Schulman, H., and Kopito, R. R. (2007) Global changes to the ubiquitin system in Huntington’s disease. Nature 448, 704–8. 14. Tagwerker, C., Flick, K., Cui, M., Guerrero, C., Dou, Y., Auer, B., Baldi, P., Huang, L., and Kaiser, P. (2006) A tandem affinity tag for twostep purification under fully denaturing conditions: application in ubiquitin profiling and protein complex identification combined with in vivocross-linking. Mol Cell Proteomics 5, 737–48. 15. Kaiser, P., and Huang, L. (2005) Global approaches to understanding ubiquitination. Genome Biol 6, 233. 16. Guerrero, C., Tagwerker, C., Kaiser, P., and Huang, L. (2006) An integrated mass spectrometry-based proteomic approach:

209

quantitative analysis of tandem affinity-purified in vivo cross-linked protein complexes (QTAX) to decipher the 26 S proteasome-interacting network. Mol Cell Proteomics 5, 366–78. 17. Cronan, J. E., Jr. (1990) Biotination of proteins in vivo. A post-translational modification to label, purify, and study proteins. J Biol Chem 265, 10327–33. 18. Tagwerker, C., Zhang, H., Wang, X., Larsen, L. S., Lathrop, R. H., Hatfield, G. W., Auer, B., Huang, L., and Kaiser, P. (2006) HB tag modules for PCR-based gene tagging and tandem affinity purification in Saccharomyces cerevisiae. Yeast 23, 623–32. 19. Savage, D., Mattson, G., Desai, S., Niedlander, G., Morgensen, S., and Conklin, E. (1994) Avidin-Biotin Chemistry: A Handbook, 2nd ed., Pierce Chemical, Rockford. 20. Ong, S. E., Kratchmarova, I., and Mann, M. (2003) Properties of 13C-substituted arginine in stable isotope labeling by amino acids in cell culture (SILAC). J Proteome Res 2, 173–81. 21. Ong, S. E., Foster, L. J., and Mann, M. (2003) Mass spectrometric-based approaches in quantitative proteomics. Methods 29, 124–30. 22. 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, 376–86. 23. Mann, M. (2006) Functional and quantitative proteomics using SILAC. Nat Rev Mol Cell Biol 7, 952–8. 24. Blagoev, B., Ong, S. E., Kratchmarova, I., and Mann, M. (2004) Temporal analysis of phosphotyrosine-dependent signaling networks by quantitative proteomics. Nat Biotechnol 22, 1139–45.

wwwwwwwwwwwwwwwwww

Chapter 16 High-Throughput High-Content Functional Image Analysis of Neuronal Proteins Implicated in Parkinson’s Disease Eva Blaas and Ronald E. van Kesteren Abstract Parkinson’s disease (PD) is characterized by the progressive loss of dopamine neurons. Here, we describe how to use human SH-SY5Y neuroblastoma cells as an in vitro cell model to study the effects of candidate PD susceptibility genes on dopamine neuron differentiation and viability. This cell model can be used to study the effects of siRNA-induced gene knockdown, either alone or in combination with PD-related cellular stressors such as MPP+, and is compatible with high-throughput cellular screening platforms such as the Cellomics ArrayScan VTI HCS Reader. Key words: Parkinson’s disease, Neurodegeneration, Cellular screening, High-content analysis, SH-SY5Y, 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine

1. Introduction Recent genomics and proteomics studies implicate many gene products as the target molecules for the expression/maintenance of various neurodegenerative or psychiatric disorders. These disorders are complex and involve many cellular mechanisms. There is a need for high-throughput high-content assays to systematically assess the functional contribution of these genes to known pathogenic mechanisms. Automatic imaging systems have been developed for this purpose and applied for the analysis of many cell types. The Thermo Scientific Cellomics® ArrayScan® VTI is a modular high-content screening instrument designed for high-throughput fluorescence imaging and quantitative analysis of fixed and live cells. Cell cultures on, among others, a 96-well format can be analyzed automatically. An average of single-cell morphology is


211

212

Blaas and van Kesteren

obtained by scanning the cells at different locations of each well, where each well is systematically divided into fields. Automatic focussing of the images is acquired by the irradiation of a laser from a source next to the microscope lens to the bottom of the 96-well plate. Up to five different filters can be used to differentiate five different cellular components within one cell. We routinely use XF93-Hoechst, XF93-TRITC, XF93-Cy5, and XF93-FITC for localization of nuclei, mitochondria, neurites, and F-actin cytoskeleton, respectively. While scanning, the acquired images are simultaneously analyzed by Cellomics® vHCS™ software to generate statistically robust and biologically relevant parameters such as the percentage of cell death, changes in cell morphology, mitochondrial activity, length and branch points of the neurites, cytoskeleton integrity, etc. Different algorithms have been designed by Cellomics® to answer specific biological questions. Because of its high-throughput and high-content nature, Cellomics® has been employed for target identification, lead optimization, secondary screening, toxicity studies, and cell biology research (www.cellomics.com). In this chapter, we describe the use of Cellomics® to examine the interaction of environmental insult and susceptibility genes in a cellular model of Parkinson’s disease (PD). PD is the second most prevalent neurodegenerative disorder with a lifetime risk of developing the disease of approximately 1.5% (1). PD is characterized by the progressive loss of dopamine neurons in the substantia nigra pars compacta (SNpc). These neurons normally regulate motivated movements by releasing dopamine in the striatum and modulating the activity of striatal interneurons. The loss of nigrostriatal dopamine projections, thus, accounts for the typical Parkinsonian symptoms such as akinesia and bradykinesia. The reasons for the selective vulnerability of SNpc dopamine neurons in PD are not well understood. One neuropathological hallmark of affected neurons is the presence of intraneuronal cytoplasmic inclusions named Lewy bodies. Lewy bodies contain aggregates of the protein a-synuclein surrounded by a halo of ubiquitin (2, 3). Indeed, mutations in the a-synuclein gene (SNCA or PARK1/4) and the ubiquitin carboxyl-terminal esterase L1 gene (UCHL1 or PARK5) both cause autosomal dominant monogenic forms of PD (4), indicating that dysregulation of protein folding and proteasomal degradation are an important factor in the pathogenesis of PD. Other factors that contribute to the selective vulnerability of SNpc dopamine neurons are oxidative stress and mitochondrial dysfunction (5). This became particularly clear when the neurotoxic effects of 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine (MPTP) were first discovered. MPTP was in the 1970s accidentally produced during the incorrect synthesis of the opioid drug desmethylprodine (MPPP), and people using MPTP-contaminated desmethylprodine got acute and

High-Throughput High-Content Functional Image Analysis of Neuronal Proteins

213

permanent Parkinsionian symptoms. MPTP is metabolized in the brain into 1-ethyl-4-phenyl-pyridinium (MPP+), which is subsequently taken up by dopamine neurons via the dopamine transporter and blocks mitochondrial complex I activity (6, 7). Several gene mutations causing autosomal recessive forms of PD confirm an important role for oxidative stress and mitochondrial dysfunction in PD (e.g., PINK1/PARK6 and DJ-1/PARK7) (8). Despite the many insights that genetic forms of PD have provided in the pathogenic mechanisms underlying the disease, most PD cases are sporadic and are probably caused by environmental factors or by genetic variations that either modify the above-mentioned protein misfolding or oxidative stress pathways or change the vulnerability of dopamine neurons in some other way. Recent studies in genetic model organisms and on human postmortem brain tissue have started to reveal large numbers of additional candidate PD susceptibility genes (9–11), and there currently is a need for high-throughput cellular assays that can systematically assess the functional contribution of these genes to known PD pathogenic mechanisms or to dopamine neuron viability in general. Here, we describe a high-content cellular assay that makes use of MPP+ induced toxicity in human SH-SY5Y neuroblastoma cells. We describe a retinoic acid-induced differentiation procedure that allows SH-SY5Y cells to acquire a mature dopamine neuron-like phenotype. During differentiation, cells can be treated with a chronic low dose of MPP+ to induce PD-like cellular stress. We next show how to transfect SH-SY5Y cells with siRNAs to knock down candidate PD susceptibility genes of interest. We use siRNA-induced gene knockdown to study possible interactions of these genes with MPP+ treatment or with known PD susceptibility genes. Finally, we describe a multiplexed highcontent cellular analysis that assesses the effects of gene knockdown and/or MPP+ treatment on several cellular stress parameters, including cell viability, mitochondrial activity, F-actin content, and neuronal differentiation. This assay is compatible with highthroughput cellular screening platforms such as the Cellomics ArrayScan VTI HCS Reader.

2. Materials 2.1. Cell Plating and Transfection

1. Greiner flat bottom 96-well plates. 2. Poly-l-lysine (0.1 mg/ml, Sigma, store at 4°C). 3. Growth factor reduced (GFR) Basement Membrane Matrigel Matrix, Phenol Red-free (10 mg/ml, BD Biosciences, store at −20°C). 4. Phosphate-buffered saline (PBS, GIBCO, store at 4°C).

214


5. Advanced DMEM/F12 culture medium (GIBCO, store at 4°C). 6. Trypsin (2.5 mg/ml, GIBCO, store at −20°C). 7. Fetal calf serum (FCS, GIBCO, store at −20°C). 8. Sterile water (Baxter). 9. 5× siRNA buffer (Dharmacon, store at 4°C). Prepare a working solution of 1× siRNA buffer by diluting in sterile water. 10. Dharmafect 3 siRNA transfection reagent (Dharmacon, store at 4°C). 11. siRNA (Dharmacon, store at −20°C). Avoid extensive (>5 times) freezing–thawing of siRNA solutions. 12. Penicillin–Streptomycin (Pen/Strep) GIBCO, store at −20°C).

solution

(100×,

13. All-trans retinoic acid (RA, Sigma). Prepare a 0.1 mM stock solution of RA in DMSO (cell culture tested, Sigma), aliquot, and store at −20°C. 14. 1-Methyl-4-phenyl-pyridinium iodide (MPP+, Sigma). A stock solution of 0.1 M MPP+ in sterile water should be freshly prepared on the day of use and kept in the dark. 2.2. Immunohistochemistry

1. 10× PBS. Dissolve 80.0 g NaCl, 2.0 g KCl, 17.8 g Na2HPO4·2H2O, and 2.4 g KH2PO4 in 800 ml H2O. Adjust pH to 7.4 and adjust volume to 1,000 ml with H2O. Prepare a working solution of 1× PBS by diluting in H2O. 2. 4% Paraformaldeyhde (PFA). Add 40 g PFA to 800 ml H2O, add a few drops 10 M NaOH, and let dissolve at 60°C (do not boil). Add 10 g saccharose and 100 ml 10× PBS, adjust pH to 7.4, and adjust volume to 1,000 ml with H2O. 3. Wash buffer. Add 2.5 g gelatin and 5 ml Triton X-100 to 800 ml H2O, dissolve by stirring at 55°C, add 100 ml 10×PBS, adjust to pH 7.4, and adjust volume to 1,000 ml with H2O. Wash buffer should be filter-cleaned using a 0.2-mm bacterial filter. 4. Sterile distilled water (Baxter). 5. Goat serum (Sigma). 6. Anti-bIII-tubulin mouse primary antibody (Sigma). 7. Cy5-conjugated (Sigma).

anti-mouse

IgG

secondary

antibody

8. Mitotracker (Molecular probes). Prepare a 1 mM sock solution in DMSO, aliquot, and store at −20°C. 9. Alexa Fluor® 488-conjugated Phalloidin (Molecular Probes). Prepare a 200 U/ml stock solution in methanol and store at −20°C.


215

10. Hoechst 33258 (10 mg/ml, Molecular Probes). Prepare a working solution by diluting 1:20,000 in H2O and store in the dark at 4°C. 2.3. High-Content Analysis

1. Cellomics ArrayScan VTI HCS Reader (Thermo Scientific). 2. Cellomics vHCS:Scan, vHCS:ToolBox and vHCS:View software (Thermo Scientific). 3. Cellomics Neuronal Profiling V3.5 BioApplication software (Thermo Scientific). 4. Cellomics Compartmental Analysis V3 BioApplication software (Thermo Scientific). 5. Basic statistical analysis software (Microsoft Excel or SPSS). 6. Clustering and visualization software, e.g., MultiExperiment Viewer (MeV) from TM4 (www.tm4.org/mev.html).

3. Methods 3.1. Culturing SH-SY5Y Cells 3.1.1. Day 0: Preparing Plates and Culturing SH-SY5Y Cells (See Note 1)

At this moment in time, you will need a T75 culture flask with SH-SY5Y cells grown to near-confluence to start with the experiment. 1. Allow Matrigel to thaw on ice for 2 h (see Note 2). 2. Coat 96-well plate with PLL (100 ml/well) for 1 hour at RT and wash once with PBS and once with advanced DMEM/ F12 (see Note 3). 3. Place plate and medium on ice. 4. Prepare a 1:10 dilution of the Matrigel stock solution in icecold advanced DMEM/F12. 5. Add 50 ml diluted Matrigel to each well and incubate the plate for 1 h at room temperature. 6. Wash 2× with advanced DMEM/F12 and keep the plate at 37°C until use. 7. Prewarm advanced DMEM/F12, FCS, PBS, and trypsin. 8. Remove culture medium from the T75 flask containing your SH-SY5Y cells. 9. Wash the cells once with PBS. 10. Trypsinize the cells using 700 ml trypsin for ~2 min at 37°C. 11. Add 7.3 ml of advanced DMEM/F12/0.5% FCS (see Note 4) to the trypsinized cells and dissociate the cells from each other by shaking gently. 12. Complete the dissociation by pipetting up and down 40× using a flamed Pasteur glass pipet (see Note 5).

216


13. Pipet 10 ml of the cell suspension on each side of a cell count chamber window and count the cells. 14. Dissociate the cells once again by pipetting up and down 10× using a flamed Pasteur glass pipet, and dilute the cells to a concentration of 1.5 × 105/ml in of advanced DMEM/ F12/0.5% FCS. 15. Add 100 ml cell suspension to each well (1.5 × 104 cells/well) and allow cells to attach overnight. 3.1.2. Day 1: SH-SY5Y Cell Transfection and Differentiation

1. Dilute siRNA stock solutions in 1× siRNA buffer to obtain 1 mM working solutions (see Note 6). 2. Mix 0.15 ml Dharmafect 3 (per well) with 10 ml serum-free advanced DMEM/F12 (solution 1). 3. Mix 5 ml 1 mM siRNA solution (per well) with 5 ml serum-free advanced DMEM/F12 (solution 2). 4. Incubate for 5 min at room temperature. 5. Mix solutions 1 and 2 together (total volume of 20 ml per well) and incubate for 20 min at room temperature (see Note 7). 6. Add 80 ml prewarmed advanced DMEM/F12/0.5% FCS (without Pen/Strep, see Note 8). 7. Remove culture medium from the cells, replace with the siRNA medium (100 ml/well; final siRNA concentration: 50 nM), and incubate for 4 h at 37°C. 8. Prepare differentiation medium by diluting RA stock solution 1:100 (final concentration 1 mM) in advanved DMEM/0.5% FCS/1% Pen/Strep. 9. After 4 h, remove siRNA medium and replace with 200 ml differentiation medium (advanced DMEM/F12/0.5% FCS containing 1 mM RA, see Note 9).

3.1.3. Days 3 and 6: Replace Differentiation Medium and Treat Cells with MPP+

1. Prepare fresh differentiation medium by diluting RA stock solution 1:100 (final concentration 1 mM) in advanved DMEM/0.5% FCS/1% Pen/Strep. 2. Prepare a fresh stock solution of 0.1 M MPP+ by adding a sufficient amount of sterile water to an aliquot of MPP+ powder (see Note 10). 3. Dilute the MPP+ stock solution in RA differentiation medium (see Note 11). 4. Remove medium from the cells and replace with 200 ml fresh differentiation medium containing RA and MPP+. 5. Repeat steps 1–4 at day 6.


3.2. Staining of SH-SY5Y Cells 3.2.1. Day 8: Fixating and Staining of Cells

217

Immunocytochemistry for b-tubulin: 1. Remove 100 ml of the culture medium and add an equal amount (of the remaining volume) of fixative (4% PFA) to each well. Incubate for 10 min at room temperature (see Note 12). 2. Remove the fixation medium and replace with 100 ml of fixative. Incubate for 20 min at room temperature. 3. Remove fixative and wash 2× with PBS. 4. Block for 30 min with 2% goat serum in wash buffer. 5. Add 50 ml primary antibody solution (anti-b-tubulin, 1:600 in wash buffer/2% goat serum) and incubate overnight at 4°C (see Note 13). 6. Wash 2× with PBS. 7. Add 50 ml secondary antibody solution (Cy5-conjugated antimouse IgG, 1:400 in PBS/2% goat serum) and incubate for 1.5–2 h at room temperature. 8. Wash 2× for 5 min with 100 ml PBS. Mitotracker staining for mitochondria 1. Thaw the Mitotracker stock solution at room temperature for ~20 min before use. 2. Prepare a 100 nM working solution by diluting the stock solution 1:10,000 in PBS. 3. Incubate cells in 100 ml Mitotracker solution (100 nM) at room temperature for 15 min (see Note 14). 4. Wash 2× for 5 min with 100 ml PBS. Phalloidin staining F-actin 1. Thaw the Phalloidin stock solution at room temperature for ~20 min before use. 2. Prepare a 3 U/50 ml working solution in wash buffer/2% goat serum (see Note 15). 3. Incubate cells in 50 ml Phalloidin at room temperature for 45 min. 4. Wash 2× for 5 min with 100 ml PBS. Hoechst staining of nuclei 1. Incubate cells with 100 ml Hoechst 33258 (1:20,000 diluted) at room temperature for 15 min (see Note 16). 2. Wash cells 2× for 5 min with 100 ml sterile water. 3. Add 200 ml sterile water to each well.

218


3.3. High-Content Analysis

A brief description of how to analyze your cells using the Cellomics ArrayScan VTI HCS Reader is provided. For a detailed description, we refer to the Cellomics vHCS:Scan and vHCS:View user guides, and the Compartmental Analysis and Neuronal Profiling BioApplication user guides. 1. Start up the ArrayScan VTI HCS Reader instrument and PC. 2. Start the Cellomics vHCS:Scan software and select the Create Protocol View (Fig. 1). 3. From the Image Acquisition menu select the 20× objective. 4. From the Scan Limits menu select the number of fields/cells you want to scan. 5. From the Assay menu select the Compartmental Analysis protocol. 6. From the No. of Channels menu select 4 and set channel 1 as the Focus Channel.

Fig. 1. The Create Protocol View in vHCS:Scan. The numbers on the left-hand side refer to the steps in the high-content analysis protocol (Sect. 3.3). The three highlighted icons in the menu bar indicate how to switch between Create Protocol View, Protocol Interactive View, and Scan Plate View.


219

7. Specify the channel settings in the Channel Specific Parameters menu: ●●

Channel 1: XF-93 Hoechst.

●●

Channel 2: XF-93 TRITC (for Mitotracker).

●●

Channel 3: XF-93 FITC (for Phalloidin).

●●

Channel 4: XF-93 Cy5 (for b-tubulin).

8. For each channel, select the objects to be displayed from the Image Display Options menu. 9. Switch to the Protocol Interactive View (Fig. 2). 10. Click the Load Plate button in the menu bar on the top and insert your plate. 11. Select a random well from the 96-well plate animation on the right-hand side and click the Autofocus button. 12. If the Hoechst image appears in focus, click Acquire Image Set. 13. Check if all channels are in focus. If necessary, adjust the focal plane in the individual channels by clicking on the DZ button on the top left and changing the Z-offset. 14. Make sure that the box Automatically Acquire Image after Z-offset Change is marked.

Fig. 2. The Protocol Interactive View in vHCS:Scan. The numbers on the left-hand side refer to the steps in the highcontent analysis protocol (Sect. 3.3).

220


Fig. 3. The Scan Plate View in vHCS:Scan. The numbers on the left-hand side refer to the steps in the high-content analysis protocol (Sect. 3.3).

15. Adjust the exposure time per channel and make sure that each exposure is between 20 and 80% as indicated in the Exposure bar on the top left. 16. Switch to the Scan Plate View (Fig. 3). 17. Provide a Plate Name and a Plate ID for your plate. 18. Click the top 96-well plate icon in the menu bar and select the wells you want to scan. 19. Click the green arrow in menu bar to start the scan. 20. To perform a neurite outgrowth analysis on the same images, open the plate in vHCS:Scan within the Cellomics vHCS:ToolBox and run the Neuronal Profiling BioApplication using the Hoechst channel and the Cy5 channel (for b-tubulin). 3.4. Data Analysis

1. After the scan has been completed, start the Cellomics vHCS:View software by clicking the vHCS:View icon in the menu bar. 2. The most recent scans will automatically appear in the list; older scans can be listed by using the Find Plates icon from the menu bar. 3. Double-clicking on a plate in the plate list will open the Well Detail View for that plate (Fig. 4).


221

Fig. 4. The Well Detail View in vHCS:View. The numbers on the left-hand side refer to the steps in the data analysis protocol(Sect. 3.4).

4. In Well Detail View, you will see a list with all measured parameters on the left hand side, the well averages for each individual parameter on the right side (graph and spreadsheet). 5. Click on the parameters in the list to switch between parameters. 6. To export individual parameter data to Excel, right-click on the spreadsheet and select Export Current Spreadsheet to Excel. 7. To explore the data in more detail, right-click on the spreadsheet and select View Cell Details or View Images (Fig. 5). 8. To export all parameter data at once, go back to plate list in vHCS:View, right-click the plate of interest, and select Export Plate/Images. 9. Demark the box Export Images and click Export to export the data for further analysis in Excel or SPSS. 10. To further explore and visualize the data, read normalized logtransformed parameter values into TM4 MultiExperiment Viewer, perform hierarchical clustering of parameters and samples, and visualize the clustered data in heatmaps (Fig. 6).

222


Fig. 5. Examples of SH-SY5Y cell images produced and analyzed with the ArrayScan VTI HCS Reader. (A) Overlay image of SH-SY5Y cells stained with Hoechst, anti-b-tubulin, Mitotracker, and Phalloidin. (B) Hoechst image alone, before (B) and after (B¢) analysis. (C) Idem for Mitotracker staining alone. (D) Idem for Phalloidin staining alone. (E) Idem for antib-tubulin staining alone.

4. Notes 1. Experiments are performed in 96-well plates. In order to eliminate plate edge effects due to nonuniform evaporation of the culture medium, only the inner 60 wells of the plate are used for cell culture, and the outer wells are filled with 300 ml culture medium only. 2. Matrigel basement membrane matrix is a solubilized basement membrane preparation extracted from the EngelbrethHolm-Swarm (EHS) mouse sarcoma, a tumor rich in extracellular matrix (ECM) proteins. Its major components are laminin, collagen-IV, heparan sulphate proteoglycans, entactin, and nidogen. It is effective for the attachment and differentiation of many cell types. SH-SY5Y cells have the natural tendency to aggregate, and this is prevented by the use of Matrigel. Matrigel has to be kept on ice at all times. At room temperature, matrigel readily polymerizes to form an insoluble matrix.


223

Fig. 6. Heatmap showing the log-fold changes of 41 selected Cellomics parameters (y-axis) for five gene knockdown conditions (genes A–E and siControl, x-axis), either with or without MPP+ treatment. All knockdown measurements are in triplicate, except for the siControl/MPP+ condition. Cellomics parameters were normalized against the average siControl/ no MPP+ condition, and then hierarchically clustered based on the observed normalized values across different experimental conditions. Note that similar parameters cluster together, allowing robust detection of phenotypic changes based on multiple parameters. Also note that MPP+ treatment has characteristic effects on the phenotype of the cells and that some genes are able to specifically modify some of these effects when their expression is knocked down.

3. PLL is a synthetic compound that enhances cell adhesion and protein absorption by altering surface charges on the culture substrate. 4. The presence of serum in the medium will inactivate the trypsin. 5. SH-SY5Y cells tend to form aggragates, and only extensive pipetting will completely dissociate the cells from each other. 6. siRNA powder is resuspended in 1× siRNA buffer upon arrival. Control siRNAs (siGlo and siControl) are resuspended to a 20 mM concentration. siGlo is a nonfunctional (will not incorporate into the RISC complex), nontargeting (has no sequence alignment with any human, rat, or mouse mRNA) siRNA with a red fluorescent dye attached. siGlo is used to determine transfection efficiencies and to provide a negative control for the target siRNAs. siControl is a functional (will incorporate into the RISC complex), nontargeting siRNA. Other negative controls used are untreated cells and mock transfected cells without siRNA (Dharmafect 3 only).

224


The target siRNAs consist of four independent siRNAs per target and are resuspended to a 5 mM concentration. 7. This allows the liposomes and the siRNAs to form complexes. 8. Pen/Strep is omitted from the transfection medium because it may interfere with the transfection and reduce transfection efficiencies. 9. RA is used as a differentiation agent and prevents the cells from aggregating. RA is sensitive to light, so RA containing solutions and media should be kept in the dark. 10. MPP+ is a metabolite of MPTP. MPP+ toxic to dopamine neurons and causes Parkinson-like symptoms in humans and other primates. MPP+ solutions have to be freshly prepared on the day of use and should be kept in the dark. Surgical gloves and a 3 M mask (FFP2) are required as safety precautions when working with MPP+. All materials and solutions should be handled in a laminar flow cabinet and should be immediately disposed of in a closed container. 11. Ideally, a dose–response relationship for MPP+-induced toxicity should first be determined. We normally use MPP+ 0.01 mM (subtoxic) and 0.5 mM (toxic). 12. Cells should be fixed in a laminar flow cabinet, and all materials and solutions should be immediately disposed of in a closed container. 13. Anti-b-tubulin is used to stain neurites and analyze neuronal morphology. 14. Mitotracker is used to stain active mitochondria. 15. Phalloidin is used to stain F-actin. 16. Hoechst is used to stain DNA and analyze nuclear morphology.

Acknowledgments This work is supported by TI Pharma (project T5-207: Parkinson and Alzheimer disease: from dysregulated human brain targets towards novel therapeutics). References 1. Lees, A. J., Hardy, J., and Revesz, T. (2009) Parkinson’s disease. Lancet 373, 2055–66. 2. Harrower, T. P., Michell, A. W., and Barker, R. A. (2005) Lewy bodies in parkinson’s disease: Protectors or perpetrators? Exp Neurol 195, 1–6.

3. Ross, C. A., and Poirier, M. A. (2004) Protein aggregation and neurodegenerative disease. Nat Med 10 Suppl, S10–7. 4. Lee, F. J., and Liu, F. (2008) Genetic factors involved in the pathogenesis of parkinson’s disease. Brain Res Rev 58, 354–64.

High-Throughput High-Content Functional Image Analysis of Neuronal Proteins 5. Abou-Sleiman, P. M., Muqit, M. M., and Wood, N. W. (2006) Expanding insights of mitochondrial dysfunction in parkinson’s disease. Nat Rev Neurosci 7, 207–19. 6. Nicklas, W. J., Youngster, S. K., Kindt, M. V., and Heikkila, R. E. (1987) Mptp, mpp+ and mitochondrial function. Life Sci 40, 721–9. 7. Richardson, J. R., Caudle, W. M., Guillot, T. S., Watson, J. L., Nakamaru-Ogiso, E., Seo, B. B., Sherer, T. B., Greenamyre, J. T., Yagi, T., Matsuno-Yagi, A., and Miller, G. W. (2007) Obligatory role for complex i inhibition in the dopaminergic neurotoxicity of 1-methyl-4phenyl-1,2,3,6-tetrahydropyridine (mptp). Toxicol Sci 95, 196–204. 8. Fitzgerald, J. C., and Plun-Favreau, H. (2008) Emerging pathways in genetic parkinson’s disease: Autosomal-recessive genes in parkinson’s disease--a common pathway? FEBS J 275, 5758–66.

225

9. Bossers, K., Meerhoff, G., Balesar, R., van Dongen, J. W., Kruse, C. G., Swaab, D. F., and Verhaagen, J. (2009) Analysis of gene expression in parkinson’s disease: Possible involvement of neurotrophic support and axon guidance in dopaminergic cell death. Brain Pathol 19, 91–107. 10. Hamamichi, S., Rivas, R. N., Knight, A. L., Cao, S., Caldwell, K. A., and Caldwell, G. A. (2008) Hypothesis-based rnai screening identifies neuroprotective genes in a parkinson’s disease model. Proc Natl Acad Sci USA 105, 728–33. 11. Simunovic, F., Yi, M., Wang, Y., Macey, L., Brown, L. T., Krichevsky, A. M., Andersen, S. L., Stephens, R. M., Benes, F. M., and Sonntag, K. C. (2009) Gene expression profiling of substantia nigra dopamine neurons: Further insights into parkinson’s disease pathology. Brain 132, 1795–809.

wwwwwwwwwwwwwwwwww

Part IV Analysis of Specific Brain Tissues and Fluid

wwwwwwwwwwwwwwwwww

Chapter 17 Neuropeptidomics of the Mammalian Brain Fang Xie, Elena V. Romanova, and Jonathan V. Sweedler Abstract A suite of bioactive peptides orchestrates a variety of cellular interactions in the mammalian brain. A new bioanalytical strategy, neuropeptidomics, has evolved from the quest to characterize these important signaling peptides (SPs). The goal of a neuropeptidomics experiment is to characterize the peptides present in an intact brain, brain region, or individual neuron. To succeed, a neuropeptidomics measurement needs to deal with the large dynamic range and low abundance of some neuropeptides in a background of peptides from postmortem degradation of ubiquitous proteins. Core components of a successful neuropeptidomics study include effective tissue sampling, sensitive and robust peptide characterization, and comprehensive data analysis and interpretation. Mass spectrometry (MS) has become the central analytical approach for high-throughput, high-confidence characterization of the brain peptidome because of its capability to detect, identify, and quantify known and unknown peptides. Robust fractionation techniques, such as two-dimensional liquid chromatography, are commonly used in conjunction with MS to enhance investigation of the peptidome. Identification and characterization of peptides are more complex when neuropeptide prohormone genes have not been annotated. This chapter outlines techniques and describes protocols for three different experimental designs that combine MS with liquid chromatography, each aimed at high-throughput discovery of peptides in brain tissue. Further, we describe the currently available bioinformatics tools for automatic query of the experimental data against existing protein databases, as well as manual retrieval of structural information from raw MS data. Key words: Neuropeptidome, Hormone, Neuropeptide, Bioinformatics, Liquid chromatography, Mass spectrometry

1. Introduction Signaling peptides (SPs), which include many different types of endogenous peptides such as neuropeptides and peptide hormones, are involved in the regulation of various biological functions and behaviors (1, 2). Identifying the cell–cell SPs in tissues and cells provides a basis for understanding numerous physiological processes and biological phenomena; however, the chemical analysis


229

230

Xie, Romanova, and Sweedler

of SPs is challenging because of the inherent structural and chemical complexity of the central nervous system, especially in mammals. Neuropeptides are produced via a series of enzymatic processing steps from protein precursors known as prohormones. A single prohormone can encode multiple bioactive peptides, either replicates of an individual peptide, or single copies of different peptides, or both. Mass spectrometry (MS)-based platforms are well suited for the discovery and unambiguous characterization of bioactive peptides, as nearly all posttranslational modifications (PTMs) result in characteristic mass shifts. Peptidomics, a term introduced by several groups in 2001 (3–5), is a field of study that aims to simultaneously identify the peptide complement of cells, tissues, organs, or an entire organism. Neuropeptidomics refers to the detailed analysis of endogenous peptides from the brain or other neuronal tissues and can be considered a subfield of peptidomics. A typical peptidomics analysis of SPs usually consists of four major steps (see Fig. 1): (1) extraction of SPs from cells, tissues, or releasates, (2) separation by liquid chromatography (LC) or capillary electrophoresis (CE), (3) detection by MS, and finally (4) identification based on the MS-derived data with the assistance of bioinformatics tools (1, 2). This chapter outlines the most common and effective protocols based on the neuropeptidomics studies published in the literature (6–10). Sample preparation is of utmost importance in neuropeptidomics. No matter the technical figures of merit of the instrumentation used to characterize the peptides, if the methods used to sacrifice the animal and isolate the tissue are not done well, the measurements will not succeed. While both direct tissues and tissue homogenates have been used in neuropeptidomics studies, this chapter focuses on the latter strategy. The most common approach for peptide extraction is based on the use of organic solvents and strong acids in combination with mechanical disruption

Fig. 1. Workflow of a typical peptidomics analysis.

Neuropeptidomics of the Mammalian Brain

231

of the tissue (6–8, 11, 12). It is essential to minimize the postmortem degradation of proteins and peptides, especially for mammalian brain samples. These proteolytic events can be delayed or minimized by the use of protease inhibitor cocktails (12), instant freezing (9), heat inactivation of snap-frozen samples (9), or focused microwave irradiation (7). Additional sample preparation steps, such as desalting, can be helpful in the case of high salt physiological samples. To reduce the complexity of the tissue homogenate, twodimensional (2D) LC, at times of different scales in each dimension, is usually employed before the MS analysis. Both two successive reversed-phase liquid chromatography (RPLC) analyses of different methods (9) and strong cation exchange (SCX) followed by RPLC (8) have been used to separate neuropeptides in tissue extracts; the two dimensions can be coupled either online or offline. As an example, two representative 2D LC setups are presented in this chapter: microbore RPLC offline coupled with nanoscale RPLC, and SCX online coupled with RPLC (see Fig. 2). The offline setup has the advantage of allowing large amounts of initial sample materials, as well as analysis of resulting fractions, by multiple MS platforms. Various LC/MS platforms have been used in neuropeptidomic studies. Depending on the ion source, LC can be coupled to MS either online (electrospray ionization (ESI) source) or offline (matrix-assisted laser desorption/ionization (MALDI) source). Mass analyzers commonly used in neuropeptidomics include ion trap (IT) (6, 10), Fourier transform ion cyclotron resonance (FTICR) (6, 9, 10), hybrid quadrupole orthogonal acceleration time-of-flight (Q-TOF) (7–9), and TOF/TOF (6, 10). Many systems are capable of alternating MS and tandem MS (MS/MS) scans automatically in a

Fig. 2. Schematic diagram of the online SCX-RPLC setup. Adapted with permission from ref. (2). © Wiley-VCH Verlag GmbH & Co. KGaA.

232


ata-dependent manner, while individual peaks have to be chosen d manually for fragmentation with MALDI-TOF/TOF MS. Here, we describe three LC/MS platforms that have been applied to neuropeptidomics: SCX-RPLC online coupled with ESI-Q-TOF MS, RPLC online coupled with ESI-IT MS, and RPLC offline coupled with MALDI-TOF/TOF MS. Bioinformatics tools play a vital role in the processing of MS-derived data; in general, they serve two important purposes – as a tool for the discovery/prediction of putative prohormone genes in newly sequenced genomes and as a tool for the identification of SPs to confirm/validate the predicted prohormones and peptides. The availability of genomic, transcriptomic, and proteomic information for a growing number of species allows in silico discovery of prohormones in newly sequenced species. Sequences of known prohormones from other species can be used to survey a nascent genome to find prohormone genes that were not previously known (13). The construction of a neuropeptide prohormone database is the first step in peptide identification. The most likely set of putative SPs can be predicted from annotated prohormones using several binary logistic and expert system tools such as the NeuroPred discovery tool (14, 15). A library of predicted peptides can aid in the follow-up MS analyses. The next step is to identify peptides via database searches based on the MS/MS data; a variety of bioinformatics platforms, such as Mascot (16), SEQUEST (17), X!Tandem (18), Peaks Studio (19), and Phenyx (20), have been developed to assist with this task. The algorithms of these platforms model the actions of exogenous proteolytic enzymes and search experimentally observed masses of precursor and fragment ions against predicted ones, yielding a probability score relating to whether the observed mass match may be random. When searching for endogenous peptides instead of digested proteins, “no enzyme” is chosen in the enzyme option menu on the search parameter setup page of most bioinformatics platforms. Different scoring algorithms result in different probability scores from each platform, and careful attention must be given to interpreting the data and comparing the outputs from the various platforms utilized. All platforms are user-friendly and similar in format; the steps for using each are outlined in this chapter. For the MS/MS mass spectra that cannot be assigned via automatic protein/peptide database searches, de novo analyses of fragmentation patterns and mass shifts characteristic of ion series from N- and/or C-termini are performed; these serve to generate sequence tags for unassigned MS/MS mass spectra of putative peptides. Obtained sequence tags can then be used to survey the sequenced genome of that particular species (which can be the first step in the discovery of novel prohormones) or to probe for prohormones in other species using homology searches.


233

Fig. 3. Nomenclature of fragment ions generated in tandem mass spectrometry.

This often can help identify a portion of the peptide or prohormone, but regions surrounding the active peptide, including signal sequence, must be reconstructed from the genomic sequence. Gene features, such as start and stop codons, can be determined using a variety of bioinformatics programs and Web applications, which eventually leads to complete reconstruction of the gene. This approach is one of the primary ways to discover SPs in species that have no sequenced genome. This chapter illustrates the major steps of de novo sequencing based on collision-induced dissociation (CID) fragmentation spectra as an example; MS/MS spectra obtained via other fragmentation mechanisms (see Fig. 3 and Table 1) can be interpreted similarly.

2. Materials The products used are listed below. Comparable products from other suppliers should also be effective. 2.1. Peptide Extraction

1. Methanol (MeOH; HPLC grade; Fisher Scientific, Fair Lawn, NJ, USA). 2. Acetone, HPLC grade (Fisher Scientific). 3. Deionized water (H2O) was prepared by a Milli-Q filtration system (Millipore, Bedford, MA, USA). 4. Formic acid (FA; Fluka, Milwaukee, WI, USA). 5. Hydrochloric acid (HCl; Fisher Scientific). 6. Acetonitrile (CH3CN; HPLC grade; Fisher Scientific).

234


Table 1 Fragment ion series of different types of MS instrument. Adapted from the Matrix Science Web site (http://www.matrixscience.com/help/search_field_help.html) ESI-Q-TOF ESI-IT ESI-IT ESI-FTICR ESI-FTICR MALDI-TOF/ CID CID ETD CID ECD TOF CID a ions

X

a-NH3 if fragment includes RKNQ

X

a-H2O if fragment includes STED

X

b ions

X

X

X

X

b-NH3 if fragment includes RKNQ

X

X

X

X

b-H2O if fragment includes STED

X

X

X

X

c ions

X X

X

y ions

X

X

X

X

y-NH3 if fragment includes RKNQ

X

X

X

X

y-H2O if fragment includes STED

X

X

X

X

z+H ions

X

X

z+2H ions

X

X

X

d or d¢ ionsa

X

v ions

X

w or w¢ ionsa

X

a Isoleucine and threonine are doubly substituted at the beta carbon so that side chain loss can give rise to two different ion structures. These pairs are designated d and d¢ or w and w¢

7. Trifluoroacetic acid (TFA; HPLC grade; Pierce Biotechnology, Inc., Rockford, IL, USA). 8. Microcon YM-10 unit (10 kDa molecular weight cutoff, Millipore). 2.2. Desalting

1. C18 ZipTip (Millipore), C18 SpinColumn (Pierce), or SepPak C18 cartridge (Waters, Milford, MA, USA). 2. CH3CN, H2O, TFA (see Sect. 2.1).

2.3. Fractionation

1. Microbore HPLC system with a reversed-phase C18 column (e.g., Vydac C18 MS column, 150 × 2.1 mm i.d., 5 mm particle size, 300 Å pores, Vydac, Hesperia, CA, USA). 2. UV detector. 3. Fraction collector. 4. MeOH (see Sect. 2.1).


235

5. HPLC solvents: Solvent A: 95% H2O/5% CH3CN/0.1% FA/0.01% TFA; Solvent B: 95% CH3CN/5% H2O/0.1% FA/0.01% TFA (v/v/v/v, see Sect. 2.1). 6. Reconstitution solution: 95% H2O/5% CH3CN/0.01% TFA (v/v/v, see Sect. 2.1). 2.4. LC-MS Analysis 2.4.1. SCX-RPLC Online Coupled with ESI-Q-TOF MS

1. Ammonium acetate (Sigma-Aldrich, St. Louis, MO, USA). 2. H2O (see Sect. 2.1). 3. RPLC solvents A and B (see Sect. 2.3). 4. Ultimate high-pressure Netherlands).

LC

pump

(LC

Packings,

5. Switchos column switching device (LC Packings). 6. Famos autosampler (LC Packings). 7. SCX column (e.g., 15 mm × 500 mm i.d., LC Packings). 8. C18 precolumn (e.g., m-guard column MGU-30 C18, LC Packings). 9. Analytical C18 column (e.g., PepMap C18, 150 mm × 75 mm i.d., 3 mm particle size, 100 Å pores, LC Packings). 10. Q-TOF mass spectrometer (Micromass, UK). 2.4.2. RPLC Online Coupled with ESI-IT MS

1. Capillary LC system (cap LC, Micromass) with a reversedphase C18 column (e.g., PepMap C18, 150 mm × 300 mm i.d., 3 mm particle size, 100 Å pores, LC Packings). 2. RPLC solvents should be different from those used in the first dimension to achieve better separation efficiency. If solvents A and B in the first dimension are as listed under Sect. 2.3, solvents A and B in the second dimension can be 98% H2O/2% MeOH/0.01% heptafluorobutyric acid (HFBA, Thermo Scientific, Rockford, IL, USA) and 98% MeOH/2% H2O/0.01% HFBA, respectively. 3. ESI-IT mass spectrometer (e.g., HCT Ultra-PTM Discovery system, Bruker Daltonics, Billerica, MA, USA).

2.4.3. RPLC Offline Coupled with MALDI-TOF/ TOF MS

1. Capillary LC system with a reversed-phase C18 column (see Sect. 2.4.2). 2. RPLC solvents (see Sect. 2.4.2). 3. a-Cyano-4-hydroxycinnamic acid (CHCA; Fluka) 2,5-dihydroxybenzoic acid (DHB; Sigma-Aldrich).

or

4. Acetone, CH3CN, H2O, TFA (see Sect. 2.1). 5. Automated fraction collector/spot picker (e.g., ProteineerFC, Bruker Daltonics). 6. MALDI-TOF/TOF mass spectrometer (e.g., Ultraflex II, Bruker Daltonics).

236


2.5. Data Analysis

1. Software or program for data format conversion.

2.5.1. Peptide Identification with the Assistance of Bioinformatic Tools

2. A bioinformatics platform for MS/MS data interpretation and database search (e.g., Mascot (www.matrixscience.com), X!Tandem (http://www.thegpm.org/TANDEM/), SEQUEST (http://www.thermo.com), Peaks Studio (Bioinformatics Solutions Inc., Waterloo, ON, Canada), and Phenyx (Geneva Bioinformatics (GeneBio) SA, Geneva, Switzerland)).

2.5.2. Manual De Novo Sequencing and Subsequent BLAST Search

1. Basic Local Alignment Search Tool (BLAST; see Note 1) algorithms (13) from either the Baylor College of Medicine Web site (http://www.hgsc.bcm.tmc.edu/blast.hgsc) or the National Center for Biotechnology Information (NCBI) Web site (http://blast.ncbi.nlm.nih.gov/Blast.cgi) or the standalone BLAST provided by NCBI (ftp://ftp.ncbi.nih.gov/ blast/executables).

3. Methods 3.1. Peptide Extraction

1. Homogenize the tissue sample in acidified methanol (90:9:1 MeOH–H2O–FA, v/v/v) or acetone (40:6:1 acetone–H2O– HCl, v/v/v) on a bed of ice (see Note 2). 2. Spin the homogenate for 20 min at 14,000 rpm at 4°C in a microcentrifuge tube to pellet any solid material; transfer the supernatant to a new microcentrifuge tube. If using acidified methanol for peptide extraction, skip to step 6; if using acidified acetone, follow steps 3–6. 3. Remove the acetone from the supernatant and preconcentrate the sample using a vacuum concentrator until the peptides are nearly dry. 4. Reconstitute the peptides in a reasonably small volume of a solution containing 95% H2O/5% CH3CN/0.01% TFA (v/v/v) in a microcentrifuge tube and vortex well. 5. Spin the reconstituted sample for 10 min at 14,000 rpm at 4°C to pellet any solid material. If any pellet is visible, transfer the supernatant to a new microcentrifuge tube. 6. Filter the supernatant through a Microcon YM-10 unit to remove large proteins.

3.2. Desalting (See Note 3)

1. Activate the sorbent with 50% H2O/50% CH3CN (v/v). 2. Equilibrate the sorbent with 95% H2O/5% CH3CN/0.01% TFA (v/v/v). 3. Load the sample. 4. Wash with 95% H2O/5% CH3CN/0.01% TFA (v/v/v).


237

5. Elute the peptides with 30% H2O/70% CH3CN/0.01% TFA (v/v/v). 6. Remove the CH3CN in a vacuum concentrator and reconstitute the peptides in a solution containing 95% H2O/5% CH3CN/0.01% TFA (v/v/v). The volume of the reconstituted sample should be compatible with the size of the injection loop and the scale of the LC instrument being used to fractionate the sample. 3.3. Fractionation

1. Wash the injection loop multiple times with MeOH followed by H2O. 2. Load the sample onto the injection loop. 3. Separate the peptides on a RP C18 column with a gradient using solvent A and B. Gradient can start with 5% of B for 5 min to allow peptides to move from the injection loop to the column and bind to the solid phase, ramp up to 80% of B in the next 60 min to elute peptides off the column, stay at 80% of B for another 5 min to elute any hydrophobic compounds, and ramp down to 5% of B in the next 10 min and maintain isocratic conditions to equilibrate the column. 4. Monitor the separation with UV detection set at 214 nm. 5. Collect the eluent with a fraction collector either manually or automatically. 6. Remove the CH3CN from each fraction in a vacuum concentrator and reconstitute the peptides in a solution containing 95% H2O/5% CH3CN/0.01% TFA (v/v/v). The volume of the reconstituted sample should be compatible with the size of the injection loop and the scale of the LC instrument being used to analyze the fractions (for example, 5–10 mL for a capillary LC/MS platform).

3.4. LC-MS Analysis

1. Switch the RP analytical column offline.

3.4.1. SCX-RPLC Online Coupled with ESI-Q-TOF MS

2. Switch the SCX column online with the RP trap column. 3. Load the sample. Most peptides are trapped on the SCX column, and those that do not interact with the SCX column are trapped on the RP trap column. 4. Switch the SCX column offline. 5. Rinse the RP trap column for 5 min to remove the salts. 6. Switch the RP trap column online with the RP analytical column. 7. Elute the peptides from the RP trap column and separate them on the RP analytical column using a linear gradient from 5 to 95% of solvent B.

238


8. Elute the peptides from the SCX column by injecting 20 mL of 20 mM ammonium acetate solution. The peptides are trapped on the RP trap column. 9. Repeat steps 1–7 nine more times, but incorporate a change in step 3; instead of loading the sample, 20 mL of successive concentrations (20, 50, 100, 200, 400, 600, 800, 1,000, and 2,000 mM) of ammonium acetate are loaded in step 3 to elute the peptides stepwise from the SCX column. 10. Interface the LC eluent online with the ESI source. 11. Perform MS scans in the range of 300–1,500 m/z. 12. Employ the data-dependent acquisition method to trigger the MS/MS scan. The most intense ion(s) in the MS scan are selected as parent ions for fragmentation, with the dynamic exclusion set to two spectra within 1 min. Preference is given to ions with two or more charges. The singly charged ions can be excluded, if desired. The collision energy can be ramped for the most efficient and reproducible MS/MS fragmentation. The MS/MS scans are typically performed in the range of 50–2,000 m/z. 3.4.2. RPLC Online Coupled with ESI-IT MS

1. Load 5 mL of a reconstituted fraction sample onto the injection loop. 2. Separate the peptides on a RP C18 column with a gradient using solvent A and B. The gradient should be determined based on the elution profile in the first dimension. 3. Interface the LC eluent online with the ion source on an ESI-IT mass spectrometer. 4. The following steps are mostly the same as described in steps 13 and 14 in Sect. 3.4.1. With some IT mass spectrometers, electron transfer dissociation (ETD) is also optional for peptide fragmentation in addition to CID and can be performed alternatively with CID on each parent ion.

3.4.3. RPLC Offline Coupled with MALDI-TOF/ TOF MS

1. The LC separation is the same as described in Sect. 3.4.2, except that the eluent is directed to a MALDI target plate and collected automatically at a rate of 1 min/spot. 2. Add 1 mL of a matrix solution (10 mg/mL CHCA in 50% acetone/50% H2O/0.01% TFA (v/v/v) or DHB in 50% CH3CN/50% H2O/0.01% TFA (v/v/v)) to each spot, before the eluent is completely dry. If the sample is not from LC, add 1 mL of the sample, along with 1 mL of the matrix solution, to a fresh spot on the target plate. 3. Allow the spots to air-dry. 4. Load the target into the mass spectrometer.


239

5. Calibrate the instrument with a standard peptide mixture in the positive-ion reflectron mode within a range of 550–6,000 m/z (or any m/z range of interest). 6. Acquire mass spectra of each sample spot by shooting the laser at random portions on the spot and sum the spectra from all laser shots to obtain a comprehensive peptide profile of each spot. Adjust the laser power during acquisition to gain the optimal mass resolution and sensitivity. 7. Change the instrument to MS/MS mode, and acquire MS/MS spectra on the peaks of interest. Usually, higher laser power is required in MS/MS analysis, and multiple fragmentation spectra should be summed to obtain a good spectrum. 3.5. Data Analysis 3.5.1. Peptide Identification with the Assistance of Bioinformatic Tools

1. Process and convert the MS/MS data to a universal format, such as the Mascot generic file format (.mgf ) or the peak list format (.pkl). 2. Import the converted data file into a bioinformatics platform. 3. Construct the prohormone database in FASTA format and import it into the bioinformatics platform. 4. Set parameters in the bioinformatics platform, such as mass tolerance (see Note 4) and PTMs (see Note 5), for automatic de novo sequencing and database search. 5. Execute the automatic de novo sequencing and the subsequent database search on the MS/MS data file. 6. Verify all the obtained peptide identities manually for accurate ion series, reasonable cleavage sites, and PTM identification (see Note 6).

3.5.2. Manual De Novo Sequencing and Subsequent BLAST Search

1. Examine the low-mass region for the b2-ion. The b2-/a2-ion pair, separated by 28 Da, serves as a guide for the b2-ion (see Fig. 3). 2. Examine the high-mass region for the first identifiable y-ion (ideally yn-1). The list of possible amino-acid combinations derived from the b2-ion limits the possible residues to consider. 3. Extend the y-ion series toward the low-mass region, and/or extend the b-ion series toward the high-mass region, depending on which ion series is predominant in the spectra. As a y-ion is identified, the low-mass region is examined to identify the corresponding b-ion, and vice versa. 4. Examine the ions that seem to be from the neutral losses, and check if the loss is compatible with the deduced amino-acid residue (see Note 7).

240


5. Search all the obtained sequence tags against various genomic information databases using BLAST algorithms. We find that at least five consecutive amino acids in the query sequence are needed to obtain reasonable hits, especially when searching against nucleotide databases. 6. Reconstruct the complete precursor by examining the nucleotides surrounding the active sequence for the presence of various gene features such as translation start sites, exons, and introns, as well as termination signals.

4. Notes 1. BLAST is a set of similarity search programs. Both the query and the targeted database can be nucleotide, translated nucleotide, or protein. The choice of BLAST program is based on the purpose of the search. For example, “tblastn” is used to search the translated nucleotide database using a protein query. The “BLAST Program Selection Guide” provided on the NCBI Web site (http://www.ncbi.nlm.nih.gov/blast/ producttable.shtml) can help you select the right service for a particular search. 2. We find that the amount of solvent is crucial for effective peptide recovery and recommend using at least a 10× v/v excess of solvent relative to the estimated volume of tissue. 3. There are a number of commercially available products for effective desalting at different scales, many of which rely on the use of C18 resin, a common adsorbent in RP chromatography. Depending on the sample size, desalting can be performed with a C18 ZipTip (Millipore), C18 SpinColumn (Pierce), or a Sep-Pak C18 cartridge (Waters; available in various loading capacities and volumes). The major steps are essentially the same and include activation and equilibration of the desalting sorbent, loading the sample and binding of peptides to the sorbent, washing out the salt and other components that do not bind to the sorbent, and finally eluting peptides off the sorbent. Manufacturer’s instructions provide the most effective protocols and helpful guidelines for each recommended product. The elution solvents may vary but usually contain organic modifiers such as CH3CN or MeOH with acidic additives such as FA and/or TFA. 4. Mass tolerance depends on the instrument used for data acquisition. For example, the mass tolerance can be set at 1,000 ppm (precursor ions) and 1,000 ppm (fragment ions) for IT MS, while the same is set at 10 ppm (precursor ions) and 40 ppm (fragment ions) for Q-TOF MS.


241

5. Common modifications include C-terminal amidation, formation of N-terminal pyroglutamic acid from N-terminal Glu and Gln, and disulfide bonding. 6. The C-terminal amide group is known to be derived from a Gly residue preceding the cleavage sites in the precursor. If the software assigned a C-terminally amidated peptide without a Gly residue, followed by the prohormone sequence, this assignment would not be correct. In addition, a minimum of three consecutive ion (b- and y-ion) matches is required to be a true-positive match. 7. Water loss (−18 Da) is expected for Ser, Thr, Asp, and Glu; ammonium loss (−17 Da) is expected for Asn, Gln, Lys, and Arg. A loss of 48 Da may be observed for Met, or −64 Da if it is oxidized. Table 1 provides a comprehensive list of expected fragment ions from each MS platform.

Acknowledgments The project described was supported by Award Number P30 DA018310 from the National Institute on Drug Abuse (NIDA), and by Award No. NS031609 from the National Institute of Neurological Disorders and Stroke (NINDS). The content is solely the responsibility of the authors and does not necessarily represent the official views of NIDA, NINDS, or the National Institutes of Health. References 1. Li, L., Sweedler, J. V. (2008) Peptides in the brain: mass spectrometry-based measurement approaches and challenges. Ann. Rev. Anal. Chem. 1, 451–483. 2. Boonen, K., Landuyt, B., Baggerman, G., Husson, S. J., Huybrechts, J., Schoofs, L. (2008) Peptidomics: the integrated approach of MS, hyphenated techniques and bioinformatics for neuropeptide analysis. J. Sep. Sci. 31, 427–445. 3. Clynen, E., Baggerman, G., Veelaert, D., Cerstiaens, A., Van der Horst, D., Harthoorn, L., et al. (2001) Peptidomics of the pars intercerebralis-corpus cardiacum complex of the migratory locust, Locusta migratoria. Eur. J. Biochem. 268, 1929–1939. 4. Schrader, M., Schulz-Knappe, P. (2001) Peptidomics technologies for human body fluids. Trends Biotechnol. 19, S55–60.

5. Verhaert, P., Uttenweiler-Joseph, S., de Vries, M., Loboda, A., Ens, W., Standing, K. G. (2001) Matrix-assisted laser desorption/ionization quadrupole time-of-flight mass spectrometry: an elegant tool for peptidomics. Proteomics 1, 118–131. 6. Bora, A., Annangudi, S. P., Millet, L. J., Rubakhin, S. S., Forbes, A. J., Kelleher, N. L., et al. (2008) Neuropeptidomics of the supraoptic rat nucleus. J. Proteome Res. 7, 4992–5003. 7. Che, F. Y., Lim, J., Pan, H., Biswas, R., Fricker, L. D. (2005) Quantitative neuropeptidomics of microwave-irradiated mouse brain and pituitary. Mol. Cell Proteomics 4, 1391–1405. 8. Boonen, K., Baggerman, G., D’Hertog, W., Husson, S. J., Overbergh, L., Mathieu, C., et al. (2007) Neuropeptides of the islets of Langerhans: a peptidomics study. Gen. Comp. Endocrinol. 152, 231–241.

242


9. Dowell, J. A., Heyden, W. V., Li, L. (2006) Rat neuropeptidomics by LC-MS/MS and MALDI-FTMS: Enhanced dissection and extraction techniques coupled with 2D RP-RP HPLC. J. Proteome Res. 5, 3368–3375. 10. Zhang, X., Scalf, M., Berggren, T. W., Westphall, M. S., Smith, L. M. (2006) Identification of mammalian cell lines using MALDI-TOF and LC-ESI-MS/MS mass spectrometry. J. Am. Soc. Mass Spectrom. 17, 490–499. 11. Canas, B., Pineiro, C., Calvo, E., Lopez-Ferrer, D., Gallardo, J. M. (2007) Trends in sample preparation for classical and second generation proteomics. J. Chromatogr. A 1153, 235–258. 12. Garden, R. W., Shippy, S. A., Li, L., Moroz, T. P., Sweedler, J. V. (1998) Proteolytic processing of the Aplysia egg-laying hormone prohormone. Proc. Natl. Acad. Sci. USA. 95, 3972–3977. 13. Altschul, S. F., Lipman, D. J. (1990) Protein database searches for multiple alignments. Proc. Natl. Acad. Sci. USA. 87, 5509–5513. 14. Amare, A., Hummon, A. B., Southey, B. R., Zimmerman, T. A., Rodriguez-Zas, S. L., Sweedler, J. V. (2006) Bridging neuropeptidomics and genomics with bioinformatics: prediction of mammalian neuropeptide prohormone processing. J. Proteome Res. 5, 1162–1167. 15. Southey, B. R., Amare, A., Zimmerman, T. A., Rodriguez-Zas, S. L., Sweedler, J. V. (2006)

NeuroPred: a tool to predict cleavage sites in neuropeptide precursors and provide the masses of the resulting peptides. Nucleic Acids Res. 34, W267–272. 16. Perkins, D. N., Pappin, D. J., Creasy, D. M., Cottrell, J. S. (1999) Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20, 3551–3567. 17. Eng, J. K., McCormack, A. L., Yates, J. R. (1994) An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J. Am. Soc. Mass Spectrom. 5, 976–989. 18. Craig, R., Beavis, R. C. (2004) TANDEM: matching proteins with tandem mass spectra. Bioinformatics 20, 1466–1467. 19. Ma, B., Zhang, K., Hendrie, C., Liang, C., Li, M., Doherty-Kirby, A., et al. (2003) PEAKS: powerful software for peptide de novo sequencing by tandem mass spectrometry. Rapid Commun. Mass Spectrom. 17, 2337–2342. 20. Colinge, J., Masselot, A., Cusin, I., Mahe, E., Niknejad, A., Argoud-Puy, G., et al. (2004) High-performance peptide identification by tandem mass spectrometry allows reliable automatic data processing in proteomics. Proteomics 4, 1977–1984.

Chapter 18 In-Depth Analysis of the Cerebrospinal Fluid Proteome and Biomarker Discovery: Abundant Protein Depletion Sample Pretreatment Method Silvina A. Fratantoni and Connie R. Jimenez Abstract Cerebrospinal fluid (CSF) contains peptides and proteins important for brain physiology and potentially also relevant to brain pathology. Therefore, CSF provides an attractive source for biomarker discovery in brain and neurological diseases. CSF proteomics provides an analytical challenge as ~80% of proteins originate from serum, and removal of these major proteins is necessary to study brain-derived proteins that are present at low concentrations. In this book chapter, we describe a CSF sample pretreatment method to allow for in-depth and comparative analysis of the CSF proteome. To this end, we provide a protocol for batch-mode abundant protein depletion using the multiple affinity removal system MARS cartridge (Agilent) as well as for subsequent protein concentration using ultrafiltration with a 3 kDa molecular weight cutoff (MWCO) spin filter (Millipore). In addition, our experience with this depletion method coupled to 1D page/LC-MS/MS as well as other depletion and protein fractionation methods for CSF analysis is discussed. Key words: CSF, Affinity depletion, IgY spin filter, MARS cartridge, MWCO filter, Sensitivity, Reproducibility, Biomarker discovery

1. Introduction Discovery of protein biomarkers in cerebrospinal fluid (CSF) will be facilitated by the application of robust, quantitative proteomics strategies to high quality patient samples. In order to use CSF for biomarker discovery, reliable sample collection and processing methods must be developed and implemented (1, 2). Previously, we assessed the impact of pre-analytical variables and established guidelines for CSF sample collection (3). Here, we describe an


243

244

Fratantoni and Jimenez

abundant protein depletion method for CSF proteomics that is reproducible and provides acceptable throughput and depth of analysis. We focus on the use of batch-mode depletion using the multiaffinity-removal-system (MARS) spin filter (Agilent) that has been designed to remove the top 14 abundant plasma proteins, i.e., albumin, IgG, antitrypsin, IgA, transferrin, haptoglobulin, fibrinogen, a2-macroglobulin, a1-acid glycoprotein, IgM, apolipoprotein AI, apolipoprotein AII, complement C3, and transthyretin (4). Inclusion of an abundant protein depletion step in the CSF proteomic workflow is important, as it greatly improves the depth of analysis of the CSF proteome as compared to more simple methods that analyze whole CSF or that only employ ultrafiltration for CSF protein concentration (5, 6). Using 1D page/LC-MS/MS on a LTQ-FTMS platform, we typically identify 260–290 proteins in ~30 ml whole CSF, whereas when we apply abundant protein depletion to 0.5 ml CSF in spin filters, we identify on average 600–670 CSF proteins in a single sample. Importantly, in the depleted CSF fractions, we can detect hundreds of CSF proteins, including many brain-derived proteins that are not detected in whole or concentrated CSF. Perhaps even more important, the spin-filter-based depletion in our hands turns out to be reproducible as assessed by the protein band patterns in a SDS-PAGE gel (Fig. 1) and analysis of technical replicates in conjunction with shot-gun sequencing and spectral count-based protein quantitation (7, see Chap. 21). Using the abundant protein depletion method described for the MARS cartridge, we have performed many (>30) depletions using a single column and found that the performance is very stable. We chose to use the spin filter format for abundant protein depletion, as opposed to a LC column format, as the spin filters are easy to use and compatible with relatively small CSF volumes (0.1–0.5 ml CSF). Table 1 shows the relation between sample volume depleted CSF and the number of identified proteins. Prior to settling for the MARS spin filter from Agilent for our in-depth CSF proteomics workflow, we have also evaluated IgY-based spin filters from Genway for CSF analysis. While both spin filters showed comparable reproducibility of protein identification (71 and 74%) and quantification (17 and 18% CV on spectral counts), the total number of identified CSF proteins was ~9% higher with the Agilent spin filters with 767 protein identifications in the technical triplicate analysis with the Agilent filter and 703 CSF proteins with the Genway filter (7), therefore providing a more sensitive analysis. Recently, we have successfully employed the described abundant protein depletion workflow coupled to 1D page/LC-MS/ MS to the analysis of 20 CSF patient samples in the context the

In-Depth Analysis of the Cerebrospinal Fluid Proteome and Biomarker Discovery

245

Fig. 1. Overview of the workflow for in-depth CSF proteomics including the abundant protein depletion and concentration steps prior to proteomics analysis by 1D gel electrophoresis and nanoLC-MS/MS. The Coomassie protein band patterns obtained for the depleted CSF fractions (bottom-left ) and the bound CSF fractions (bottom-right ) are shown.

Table 1 Relation between input volume CSF and the number of identified CSF proteins and albumin peptides Number of proteins in A (number o f ALB peptides)

Number of proteins in B (number of ALB peptides)

Number of proteins in C (number of ALB peptides)

Number of proteins in D (number of ALB peptides)

90 ml CSF analyzed

1,000 ml CSF analyzed

500 ml CSF analyzed

500 ml CSF analyzed

227 (156)

469 (32)

402 (34)

419 (16)

Four proteomics analyses were performed with 90, 500 and 1,000 ml of CSF. Increasing the input volume from 500 to 1,000 ml only results in slightly more protein identifications (469 versus 402 and 419) ALB albumin

246


EU FP6 cNeuPRO project for biomarker discovery in Alzheimer’s Disease. The obtained dataset contained 1,079 CSF proteins including several candidate biomarkers for mild cognitive impairment and AD (Fratantoni and Jimenez, unpublished data). The most promising ones include brain plasma-membrane-derived proteins, neuroplasticity and neurodevelopment-related proteins, underscoring the sensitivity of the approach. In this book chapter, we describe the steps for CSF abundant protein depletion and subsequent protein concentration that are needed prior to analysis by mass spectrometry-based proteomics. Figure 1 shows an overview of the whole workflow, including a gel view of the depleted and bound CSF protein fractions as obtained in a triplicate analysis of a pooled CSF sample.

2. Materials 2.1. Depletion of CSF

1. Particulate-free CSF (see Notes 1 and 2). 2. Multiaffinity removal spin cartridge Human 14 (cat. no 5188–6560, Agilent Technologies) (see Note 3). 3. Buffer A (cat. no 5188–5987, Agilent Technologies). 4. Buffer B (cat. no 5188–5988, Agilent Technologies). 5. Luer-lock adapters for spin cartridge (cat. no 5188–5249, Agilent Technologies). 6. Luer-lock plastic syringes 5 ml (cat. no 5188–5250, Agilent Technologies). 7. Eppendorf safe-lock micro test tubes 1.5 ml (cat. No 0030120.086, Eppendorf). 8. Polypropylene tubes Greiner 15 and 50 ml (cat. No 210216, Greiner Bio-one). 9. Buffer Tris HCl 1 M, pH 8. 10. NaOH 5 M. 11. Optional: Syringe-driven filter unit 0.22 mm (if CSF needs to be filtered to remove debris or particles).

2.2. Concentration of Depleted Fraction

1. Amicon ultra-4 centrifugal device 3 kDa (cat. no UFC800396, Millipore) (see Note 4). 2. Buffer 25 mM ammonium bicarbonate NH4HCO3 (prepared fresh).


247

3. Methods 3.1. Batch-Mode Abundant Protein Depletion from CSF Using Spin Cartridge 3.1.1. General Preparation

All the steps and centrifugations are done at room temperature (RT). 1. Label two 5 ml luer-lock syringes and luer-lock adapters with letters A and B, referring to buffers A and B. 2. Put ~45 ml buffer A and B in two 50 ml tubes to be able to easily fill the syringes with each buffer. 3. Let desired volume CSF thaw on ice. The maximum volume of raw CSF for 1 depletion round is 250–330 ml. See Note 2 for depleting larger volumes and see Note 5 for variations of the protocol.

3.1.2. MARS Cartridge Preparation

1. Take the cartridge out of the refrigerator and allow 10 min to equilibrate to RT. 2. Remove cartridge cap and plug. Attach luer-lock adapter (labelled A) to spin cartridge, draw 4 ml of Buffer A into syringe, and attach it to luer-lock on spin cartridge. 3. Dispense buffer A through spin cartridge to prepare resin and to remove any trapped air bubbles. 4. Remove the luer-lock adapter and with a pipette remove excess buffer A from top of the spin cartridge. The spin filter is now ready for use.

3.1.3. CSF Depletion

1. Place spin cartridge in an eppendorf tube labelled “flowthrough” F1. Add 250 ml of the CSF to spin cartridge and place the cartridge cap loose. NB: this is the first part of the depleted CSF fraction. 2. Centrifuge for 1 min at 100 × g. Put the cartridge in a new eppendorf tube (F2) and leave for 5 min on the bench. Place the eppendorf tube F1 on ice. 3. Elution step 1: add 400 ml of buffer A to the cartridge. Centrifuge for 2.5 min at 100 × g. Pool together the flowthrough fraction (F2) with the F1. Place the mix of F1 and F2 on ice. 4. Elution step 2: put the cartridge in a new eppendorf tube (F3) and add 400 ml of buffer A to the cartridge. Centrifuge for 2.5 min at 100 × g. The flow-through fraction (F3) can be pooled together with the F1 and F2. This combined depleted CSF fraction (~1 ml) may be processed further (see below) or store at −80°C. See Note 2 for depleting larger volumes.

3.1.4. MARS Cartridge Re-Equilibration

In order to reuse the MARS cartridge, first the bound proteins are stripped with buffer B and then the column is washed and reequilibrated with buffer A.

248


1. Attach luer-lock adapter (labelled B) to the cartridge top. 2. Fill the plastic luer-lock syringe (labelled B) with 5 ml of buffer B and attach to the spin cartridge via de luer-lock adapter. 3. Place the spin cartridge in a 15-ml tube labelled “Bound fraction” and slowly push the buffer B through the spin cartridge so that the bound proteins will elute from the column. The rate should be ~1 ml/min done at a steady rate. 4. If the bound fraction is going to be analyzed, neutralize the proteins with 1/10 of buffer 1 M Tris HCl, pH 8, e.g., 500 ml of neutralization buffer for a 5 ml bound fraction. Check pH and if necessary add some drops of NaOH 5 M until pH is neutral. Store the bound fraction at −80°C. 5. Remove buffer B syringe and luer-lock adapter B. 6. Attach luer-lock adapter A. 7. Fill the 5-ml syringe labelled A with 5 ml of buffer A. 8. Reequilibrate the spin cartridge by slowly pushing buffer A to the resin bed. The rate should be ~1 ml/min done at a steady rate. Leave a small volume of buffer A on top of the frit. 9. Recap both ends of the spin cartridge tightly and cover well both ends with parafilm. Store spin cartridge in the fridge in a vertical position. 3.2. Protein Concentration Using Ultrafiltration 3.2.1. Concentration of the Depleted CSF Fraction

1. Take the depleted CSF protein fraction(s) from the freezer and let it thaw on ice. 2. Add 1 ml of milli-Q water to each Amicon filter (one per sample). 3. Centrifuge in a swinging-bucket centrifuge capable of handling 17 × 124 mm 15-ml conical-bottomed tubes. Spin at 3,000 × g for 10 min at RT. Make sure that the filter is wet and remove the excess of water. 4. When the sample is to be concentrated to a small volume (e.g., 30 ml for loading into minigel), we recommend the following: 5. Add the depleted CSF fraction (the pooled flowthrough of ~1 ml) to the prewet Amicon filter. 6. Spin at 3,000 × g for 10 min and repeat spin cycles until the volume is around 250 ml. 7. When the volume is around 250 ml, add 1 ml of 25 mM ammonium bicarbonate buffer and repeat spin cycle until the volume is again around 250 ml. 8. Transfer this volume to an eppendorf tube and place in the SpeedVac for the time necessary until the protein is concentrated to 20 ml.


249

9. Stay in close proximity of the SpeedVac and check several times to make sure that the protein is still dissolved in buffer. To avoid sample losses, do not let the sample go dry! If the volume is less than 20 ml, add enough buffer (25 mM ammonium bicarbonate) to 20 ml. 10. Prepare the following mix: 20 ml of depleted fraction, 7.5 ml of SDS sample buffer (4×), 3 ml of reducing agent (10×). 11. Cook at 100°C for 5 min. 12. Use the samples tfor 1D page or freeze for storage (−20°C). For best results, use Invitrogen Nupage Bis Tris gradient gel, 1.5 mm × 10 wells, for 1D pape. Proceed with protein identification and quantiation by LC-MS/MS (see Chaps. 12 and 21). 3.2.2. Concentration of the CSF Bound fraction

1. Take the protein fractions from the freezer and let them thaw on ice. 2. Add 1 ml of milli-Q water to each Amicon filter (one per sample). 3. Centrifuge in a swinging-bucket centrifuge capable of handling 17 × 124 mm 15-ml conical-bottomed tubes. Spin at 3,000 × g for 10 min at RT and check that the filter is wet and remove the excess of water. 4. If the sample is to be concentrated to a small volume (e.g., 30 ml for loading into minigel), we recommend the following: 5. Add 4 ml of the bound CSF fraction to the prewet Amicon filter. 6. Spin at 3,000 × g for 10 min. (a) 7 Add the rest of the fraction (1.5 ml) to the same Amicon filter. 7. Repeat spin cycles until the volume is around 250 ml. 8. When the volume is around 250 ml, add 1 ml of 25 mM ammonium bicarbonate buffer and repeat spin cycle until the volume is again around 250 ml. 9. Transfer this volume to an eppendorf tube and place in the SpeedVac for the time necessary until the protein is concentrated to 110–120 ml. 10. Prepare the following mix: 20 ml of bound fraction, 7.5 ml of sample buffer (4×), 3 ml of reducing agent (10×) 11. Cook at 100°C for 5 min. 12. Use the samples for 1D page or freeze for storage (−20°C). For best results, use Invitrogen Nupage Bis Tris gradient gel, 1.5 mm × 10 wells, for 1D pape. Proceed with protein identification and quantiation by LC-MS/MS (see Chaps. 12 and 21).

250


4. Notes 1. CSF to be used for proteomics analysis has to be treated with extreme consideration from the moment is taken from the patient. Avoid blood contamination by throwing away the first ml of CSF that is collected. CSF should be centrifuged and stored at −80°C within 2 h. Typically, one aliquot is used for counting the number of cells/volume fraction. The rest is centrifuged to remove cells and particles. One small aliquot is kept for protein concentration and the rest has to be stored at −80°C in aliquots of 500 ml to minimize thaw–freeze cycles. Before depleting precious CSF samples for a comparative analysis, we advice to “practice” the protocol one or two times to make sure that the fine details of the protocol are successfully applied. For this practice, a pool of “nonpreciousCSF” (e.g., leftover of diagnostics) may be used. 2. Make sure the CSF is clear and does not contain particulates before applying it to the column. Centrifuge CSF prior to loading to the cartridge. Alternatively, CSF can also be filtered through a 0.22-mm filter unit, but this step may increase protein losses. For CSF volumes larger than 250 ml, use 2–3 rounds of depletions, e.g., for 500 ml, 2 rounds of 250 ml each; for 1 ml, 3 rounds of 333 ml each. The elution steps 1 and 2 remain unchanged. CSF can be applied without any pretreatment (our preferred method, see Table 1). Another possibility is to concentrate the CSF using 3 kDa MWCO filter and diluting it with Agilent buffer A prior to loading onto the cartridge. 3. Never allow the resin bed to run dry. The frits also should remain moist. After using the spin filter, put some parafilm around it to prevent the cartridge to lose liquid or moisture. Never push air through the spin cartridge. If there is a bubble in the liquid above the frits, remove it with a yellow tip. Do not freeze the spin cartridge. Keep a detailed record of the times that it has been used. After multiple depletions, the upper frit on top of the depletion resin may clog and air may be trapped inside. This problem may be avoided by making sure that particulate-free CSF is used. When this problem still occurs, the upper frit needs to be replaced. After frit replacement, new CSF samples can be depleted without loss of depletion performance. Follow instructions from the company to change the frits. 4. The Amicon filters are extremely easy to use. As with any filtration, particulates should be avoided and removed by centrifugation. To prevent protein losses, the sample volume in the filter should not be less than 200 ml. Use a yellow tip to remove every single drop from the bottom of the filter.


251

5. Possible variations in CSF depletion protocol: We tested different protocols with the same CSF sample (a pool from different patients, 400 mg/ml): (A) 300 ml of intact CSF was concentrated in a 5 kDa Amicon filter. When the volume was 200 ml, we added 1 ml of 25 mM ammonium bicarbonate buffer. We concentrated again to 150 ml. This volume was then concentrated to 80 ml in a SpeedVac. A 24 ml aliquot corresponding to 90 ml of whole CSF was applied to the gel, followed by analysis using nanoLC-MS/MS. (B) 1 ml of CSF was concentrated to 100 ml in a 3 kDa Amicon filter. We added 100 ml of buffer A. The whole sample of 200 ml was applied in one time to the column. (C) 500 ml of CSF was diluted with 500 ml of buffer A (default Agilent protocol). The 1 ml of diluted CSF was applied in three successive rounds of 333 ml. (D) 500 ml of CSF was added in two successive rounds of 250 ml each (our preferred protocol). From B–D, The whole depleted fraction was analyzed by nanoLC-MS/MS. See Table 1 for the summary of the number of identified CSF proteins and the number of identified albumin peptides for each condition. The table shows that the number of identified proteins in the depleted fractions is higher than in crude CSF for which the maximum amount was loaded onto the minigel. Dilution of CSF with Agilent buffer A (condition C) did not improve the number of identified CSF proteins whereas depletion seemed to work less well as deduced from the number of detected albumin peptides (in brackets). Therefore, our preferred protocol which is also the fastest, includes application of intact CSF to the cartridge. Doubling the starting amount of CSF for depletion (to 1 ml of CSF, condition B) only increased the number of identified CSF proteins by ~10%.

Acknowledgments SF is supported by the EU FP6 cNeuPRO project. References 1. Zhang, J., Goodlett, D.R. and Montine, T.J. (2005) Proteomic biomarker discovery in cerebrospinal fluid for neurodegenerative diseases. J Alzheimers Dis. 8, 377–386.

2. Westman-Brinkmalm, A., Ruetschi, U., Portelius, E., Andreasson, U., Brinkmalm, G., Karlsson, G., et al. (2009) Proteomics/peptidomics tools to find CSF biomarkers for

252


neurodegenerative diseases. Front Biosci. 14, 1793–1806. 3. Jiménez, C.R., Koel-Simmelink, M., Pham, T.V., Van der Voort, L. and Teunissen, C.E. (2007) Endogeneous peptide profiling of cerebrospinal fluid by MALDI-TOF mass spectrometry: Optimization of magnetic bead-based peptide capture and analysis of preanalytical variables. Proteomics Clin. Appl. 1, 1385–1392. 4. Zolotarjova, N., Mrozinski, P., Chen, H. and Martosella, J. (2007) Combination of affinity depletion of abundant proteins and reversedphase fractionation in proteomic analysis of human plasma/serum. J Chromatogr. A. 1189, 332–338. 5. Ogata, Y., Charlesworth, M.C. and Muddiman, D.C. (2005) Evaluation of

rotein depletion methods for the analysis of p total-, phospho- and glycoproteins in lumbar cerebrospinal fluid. J Proteome Res. 4, 837–845. 6. Shores, K.S. and Knapp, D.R. (2007) Assessment approach for evaluating high abundance protein depletion methods for cerebrospinal fluid (CSF) proteomic analysis. J Proteome Res. 6, 3739–3751. 7. Fratantoni, S.A., Piersma, S.R. and Jimenez, C.R. (2010) Comparison of the performance of two affinity depletion spin filters for quantitative proteomics of cerebrospinal fluid: Evaluation of sensitivity and reproducibility of CSF analysis using GeLC-MS/MS and spectral counting. Proteomics Clin. Appl. 4, 613–617.

Chapter 19 Lipidomics of the Nervous System: Phospholipidomics by Liquid Chromatography Coupled to Mass Spectrometry or Tandem Mass Spectrometry Su Chen Abstract Phospholipidomics by high-performance liquid chromatography coupled to electrospray mass spectrometry or tandem mass spectrometry, including sample preparation, instrumental analyses, and data interpretation, is described in the chapter. Key words: Lipidomics, Phospholipidomics, Phospholipids, Ether phospholipids, Liquid chromatography/mass spectrometry, Liquid chromatography/tandem mass spectrometry

1. Introduction 1.1. Neural Membrane Lipids

The basic structure of neural membranes is a series of recurrent unities of phospholipid–cholesterol–protein complexes. The function of the external (cellular) and internal (subcellular) membrane system depends on their composition and on the integrity of their phospholipid structures, and the membrane proteins act as receptors for biologically active substances. Because phospholipids belong to the amphipathic molecules with a water-soluble and a fat-soluble component, the hydrophilic groups are arranged at the outer and inner side of membranes toward the surrounding medium; the lipophilic groups, by contrast, face each other at inner side of the bilayer configuration. Phosphatidylserine (PS), phosphatidylethanolamine (PE), and plasmalogen phosphatidylethanolamine (pPE) are usually located in the inner monolayer, whereas phosphatidylcholine (PC) and sphingomyelin (SM) are located in the outer layer (1, 2).


253

254

Chen

A phospholipid class, such as PS, consists of a mixture of molecular species. Their structural diversity is due to (i) a variety of fatty-acid chains esterified at the sn-1 and sn-2 positions of the glycerol backbone, (ii) locations of the double bond(s) (between 1 and 6) within unsaturated fatty-acid chains, which are usually located at the sn-2 position, with a number of carbon atoms (between 16 and 22), and (iii) a polar head carried at the sn-3 position (Fig. 1). Ether phospholipids (3) are usually present in brain phospholipids, mainly pPE as the major components, existing together with diacyl molecular species carrying the same polar-head group.

O

O P

O

+

N

O

O

O

H

O

H

+

O

N

P O

O O

O

R1

H O

−

H

Phosphatidylcholine (PC)

R2

O H

R1

O

−

Phosphatidylethanolamine (PE)

R2 O

OH

HO HO HO

O

O O

P O −O

OH

R2 O

+

H3N −O C O

O

H O

P O −

R1 Phosphatidylinositol (PI)

O H O

O

O

O

R1

H O

Phosphatidylserine (PS)

R2 O

O P O O O

+

N

−

HO

R1 H HN

O−

HO

Sphingomyelin (SM)

R2 O

O S O O O

HO

H

OH

HO O

H Sulfatides (ST)

R1 H HN

R2 H O HO

Fig. 1. Chemical structures of the common phospholipids, sphingomyelin, and sulfatides present in the brain tissues.

Lipidomics of the Nervous System

255

Ether phospholipids have the two predominant types of ether bond linkage in the molecules. One form is represented by the plasmalogen (1-alk-1¢-enyl fatty chain(s) linked to the sn-1 position of the glycerol backbone). The other form is an alkyl ether phospholipid species that contains 1-O-alkyl fatty chain(s) linked to the sn-1 position of the glycerol backbone. A lysophospholipid species (4, 5) contains only one fatty chain (an acyl chain or an ether fatty chain such as 1-O-alkyl and 1-alk-1¢-enyl) esterified at the sn-1 position (major species) or the sn-2 position (the minor and an acyl chain only) species. Sphingomyelin consists of a sphingosine backbone linked to a N-acyl derivative of fatty acids and a phosphocholine polar-head group. As similar to sphingomyelin, a sulfatide species consists of a sphingosine backbone that links to the N-acyl derivative of a fatty acid and a sulfatide galactosyl polar-head group in the molecule (Fig. 1). 1.2. Phospholipidomics by Liquid Chromatography/Mass Spectrometry

Conventional strategies for analyzing phospholipid classes and their molecular species require many steps. The first step is to extract a total lipid fraction from a biological sample (for example, the brain tissues), followed by isolating individual phospholipid class from other lipids by either silica thin-layer chromatography (TLC) or normal-phase high-performance liquid chromatography (HPLC). Subsequent steps include the removal of the phospholipid head group, derivatization of the sn-3 position of glycerol backbone, and then separation by normal- and reversed-phase HPLC. Individual phospholipids (separated by TLC or/and normal-phase HPLC) or molecular species (separated by reversedphase HPLC) are collected from liquid chromatography, then saponified and esterified to produce fatty-acid methyl esters and dimethylacetals that are analyzed by gas chromatography. It is clear to see that these prior separations and manipulations are labor-intensive and time-consuming, resulting in suffering from reduced recovery and selective losses of certain phospholipid molecular species up to 50%. The analysis of unprocessed total lipid extract by HPLC coupled to electrospray ionization mass spectrometry (LC/ESI-MS) or tandem mass spectrometry (LC/ESI-MS/MS) bypasses many of these analytical problems because it (i) requires minute amounts of biological samples, (ii) separates phospholipids or molecular species and then employs a “soft” ionization procedure that produces mostly singly charged molecular ions of “chromatography-isolated phospholipids”, (iii) produces reproducible mass spectra with much less overlapping molecular ions of phospholipid species, (iv) has the advantage of analyzing bioactive minor phospholipid species, such as lysoPS and LysoPE, and is available for the automation. Thus, LC/ESI-MS and LC/ESI-MS/MS methods provide unparalleled speed and precision in qualitatively and quantitatively analyzing phospholipid molecular species in lipid mixtures.

256

Chen

2. Materials 2.1. Materials Used for Extraction of the Total Lipid Fraction from the Brain Sample

1. HPLC-grade methanol and chloroform

2.2. Materials Used for HPLC Performance and Internal Standards

1. A silica HPLC column (3 mm; 2.0 × 150 mm or 5 mm; 4.6 × 250 mm)

2. Ammonium acetate 3. HPLC-grade water 4. 0.1% Ammonium acetate in water

2. HPLC-grade chloroform and methanol 3. Ammonium hydroxide 4. HPLC-grade water 5. Internal standards for phospholipid assay – 14:0–14:0 PC (1,2-ditetradecanoyl-glycero-sn-3-phosphocholine); 14:0–14:0 PE (1,2-ditetradecanoyl-glycero-sn-3-phosphoethanolamine); 14:0–14:0 PS (1,2-ditetradecanoyl-glycero-sn-3-phosphoserine); or phospholipids containing the fatty-acid chains with the odd carbon number 6. Internal standard for lysophospholipid assay – 14:0 lysoPC (1-tetradecanoyl-2-hydroxy-glycero-sn-3-phosphocholine); 14:0 lysoPE (1-tetradecanoyl-2-hydroxy-glycero-sn-3-phosphoethanolamine); 14:0 lysoPS (1-tetradecanoyl-2-hydroxyglycero-sn-3-phosphoserine), or phospholipids containing the fatty-acid chains with the odd carbon number

2.3. Liquid Chromatography Coupled to Mass Spectrometers (Fig. 2)

1. HPLC coupled to a single quadruple mass analyzer (a mass spectrometer), (LC/ESI-MS) 2. HPLC coupled to a triple quadruple mass analyzer (a tandem mass spectrometer; Fig. 2), (LC/ESI-MS/MS) 3. Other LC/MS and LC/MS/MS systems

3. Methods 3.1. A Protocol for Initially Designing an Analysis (Fig. 3)

Before phospholipidomics of the brain tissues by LC/ESI-MS or/and LC/ESI-MS/MS, choosing a right analytical approach to reach the goal is the first important step. In the qualitative analysis of phospholipids, LC/ESI-MS has demonstrated its power in the identification of various known molecular species present in the brain tissues. For structurally elucidating novel phospholipid molecule(s) or for structurally determining the molecular species extracted from a brain sample at the first time, LC/ESI-MS/MS shows its unique advantage over any other methods in providing


257

Phospholipidomics by LC Coupled to ESI-MS Qualitative Analyses or Quantitative Assay

Identifying Known Species or Elucidating New Species

LC/ESI-MS

LC/ESI-MS/MS

[M+H]+ [M−H]− Positive-ion Negative-ion

Positive-ion Negative-ion Same as Same as (LC/ESI-MS) (LC/ESI-MS)

PC, LysoPC, PS, LysoPS, Sphingomyelin, PE, LysoPE, (SM) PI, Sulfatide (ST) (Detectable but less sensitive for the lipids below)

[M-15]PS, LysoPS PE, LysoPE PC, LysoPC, SM,

Structurally informative fragments

Concentration Assay1 and Percentage Analyses2

LC/ESI-MS

LC/ESI-MS/MS

(ESI-SIR-M (ESI-MS)

(Neutral loss scan, precursor ion scan)

[M+H]+ PC, LysoPC,SM, PE, LysoPE PS, LysoPS [M-H]PI, LysoPE, PE, LysoPE PS,LysoPS PG, PA,

Positive neutral loss scan of 141 Da (PE and LysoPE) Positive precursor ion scan of 184 Da (PC, LysoPC and SM) Positive neutral loss scan of 185 Da (PS and LysoPS) Negative precursor ion scan of 241 Da (PI and LysoPI) Negative precursor ion scan of 153 Da (All kind of phospholipid species)

Fig. 2. A diagram of a LC/ESI-MS/MS system.

Fig. 3. Methods for phospholipidomics by LC/ESI-MS and LC/ESI-MS/MS.

detailed structural information of phospholipid species at the lower nanogram levels. At concentrations of phospholipid species ranging from about 2 to 200 ng/single species (detected in a quadruple mass analyzer) or about 1–1,000 ng/single species

258

Chen

(analyzed in a triple quadruple mass analyzer system), there is a linear relationship between the lipid concentration and detectable signal responding. The accurate assay (concentration profiles of phospholipid species) and percentage analyses of phospholipid molecular species in the brain in the presence of internal standard(s) can be carried out using the two methods. Choosing an effective ionization mode to produce abundant ESI-generated phospholipid molecular ions is another important factor to be considered. Usually, the positive-ion mode is effective in producing abundant protonated molecules ([M+H]+) of PC, lysophosphatidylcholine (LysoPC) and SM species, as well as PE and PS. However, molecular ions of PS, PE, and their lyso derivatives can be well ionized to produce abundant deprotonated molecules ([M−H]−) under the negative-ion mode (6–10). But molecular ions of phosphatidylinositol (PI), sulfatides (ST), and other acidic lipids such as gangliosides can be generated only under the negative-ion mode (11–14). Structurally designing the composition of phospholipid molecular species based on both the protonated and deprotonated molecules and structurally informative product ions in mass spectra, which are generated by MS and collision-induced dissociation MS/MS of protonated and deprotonated precursors, is a challenging task. Fragments of phospholipid molecular species, including the characterization of (i) the polar head, (ii) the composition and location of the two fatty-acid chains, and (iii) the sn-1 fatty chain linkage in molecular species, can be produced at nanogram levels of phospholipid samples. 3.2. Sample Preparation for Qualitative Analyses

1. Brain tissue (wet material about 100 mg) is rinsed in 0.1% ammonium acetate in water (400 mL), and homogenized in the solution in a homogenizer on ice. 2. Aliquots (200 mL) of the homogenate solution are placed into a weighted glass tube (20 mL) with a Teflon cap, and then 3.0 mL of methanol is added to the tube. After 1 min of vortex, 6 mL of chloroform is added, and the mixture is incubated for 1 h at room temperature in a shaker. After adding 2.5 mL of water, two phases are formed. The sample is centrifuged at 3,000 × g for 5 min. The lower phase (chloroform) is taken off with a glass Pasteur pipette, and the upper phase is reextracted with 2 mL of chloroform, followed by vortex for another 2–3 min and then centrifugation at 3,000 × g for 5 min. After mixing the combined two lower-phase solutions, the sample is filtered with a 0.2-mm polytetrafluoroethylene syringe filter and is then dried. 3. After drying under the nitrogen gas, the weighted tube containing dried lipid extract is weighted again. A solution containing about 200 ng/mL of the lipids in a mixture of chloroform– methanol (8:2; v/v) is then prepared for further LC/ESI-MS


259

or LC/ESI-MS/MS analyses. 10–20 mL of the sample (approximately 2–4 mg/injected lipids) is introduced into the HPLC system. 3.3. Sample Preparation for Quantitative Assay

1. Brain tissue (wet material about 100 mg) is rinsed in 0.1% ammonium acetate in water (400 mL), and homogenized in the solution in a homogenizer on ice. 2. Aliquots (100 mL) of the homogenate solution, alone with the internal standard(s) (the concentration of the internal standard(s) is based on the standard curves; the internal standard is not added for the percentage analyses of the phospholipid species in this step; see below), are placed into a weighted glass tube (20 mL) with a Teflon cap and 1.5 mL of methanol is added to the tube. After 1 min of vortex, 3 mL of chloroform is added, and the mixture is incubated for 1 h at room temperature in a shaker. After adding 1.25 mL of water, two phases are formed. The sample is then centrifuged at 3,000 × g for 5 min. The lower phase (chloroform) is taken off with a glass Pasteur pipette, and the upper phase is reextracted with 2 mL of chloroform, followed by vortex for another 2–3 min and centrifugation at 3,000 × g for 5 min. After mixing the two lower-phase fractions, the sample is filtered with a 0.2-mm polytetrafluoroethylene syringe filter and is dried under the nitrogen gas. 3. The weighted tube containing both the dried lipid extract and standard is weighted again and then a solution containing about 100 ng/mL of the lipids, along with 200 pg/mL of the internal standard used in a mixture of chloroform–methanol (8:2; v/v), is prepared for further LC/ESI-MS or LC/ ESI-MS/MS analyses. Injection volumes range from 10 to 20 mL (about 1–2 mg/total injected lipids; and 20–40 ng/ total injected internal standard(s)).

3.4. Normal-phase LC/ESI-MS

1. LC analyses are performed on a normal-phase HPLC column (HILIC silica; 3 mm, 2.1 × 150 mm). The flow rate is at 0.35 mL/min. The mass spectrometer ion source and desolvation temperature is set at 80 and 250°C, respectively. The lipids are eluted with a linear gradient of 100% solvent A (chloroform–methanol–30% ammonium hydroxide, 80:19.5:0.5; v/v) to 100% solvent B (chloroform–methanol– water–30% ammonium hydroxide, 60:34:5.5:0.5; v/v) for 15 min, then in 100% solvent B for other 15 min. Choline phospholipids (PC, SM, and LysoPC) prefer to be analyzed under the positive-ion mode (a mass scan range from 300 to 1,100 Da); analyses of ethanolamine, serine, and inositol phospholipids can be carried out under the negative-ion mode (a mass scan range from 200 to 1,100 Da). The normal mass scan and selected-ion recording (SIR) mode of ESI-MS can be used in the data acquisition.

260

Chen

Fig. 4. A HILIC silica HPLC/ESI-MS total ion current chromatography of the major phospholiopid classes; PG, PE, PI, PC, PS, and LysoPC are analyzed under the negative-ion mode (a); choline phospholipids (PC, SPH and LysoPC) are detected with the positive-ion mode (b). The PC profile usually shows two peaks, corresponding to polyunsaturated PC species and other species (a); SPH and LysoPC profiles show two peaks (b) as well.

2. The eluting order of the major phospholipid classes using the silica HPLC method is as follows: triglycerides and free fatty acids (1–6 min in retention time), PG, PE, PI, PC, PS, and LysoPC (Fig. 4a in the negative-ion mode), and PE, PC, SM, and LysoPC (the positive-ion mode in Fig. 4b). The PC profile of LC/ESI-MS usually shows two peaks, corresponding to polyunsaturated PC species and other species (Fig. 4b); the SPH profile of LC/ESI-MS shows two peaks as well, corresponding to long-chain sphingosine-based species and shortchain sphingosine-based species (Fig. 4b), and the LysoPC profile shows two peaks (Fig. 4b). 3.5. Normal-Phase LC/ESI-MS/MS

1. LC analyses were performed on a normal-phase HPLC column (silica; 5 mm, 4.6 × 250 mm; a 1/3 split of the flow into the mass spectrometer). The flow rate is 1.0 mL/min. The mass spectrometer ion source and desolvation temperature is set at 80 and 250°C, respectively. The lipids are eluted with a linear gradient of 100% solvent A (chloroform–methanol–30%


261

ammonium hydroxide, 80:19.5:0.5; v/v) to 100% solvent B (chloroform–methanol–water–30% ammonium hydroxide, 60:34:5.5:0.5; v/v) for 15 min, then in 100% solvent B for other 25 min (5, 15). Mass spectra of ethanolamine and choline phospholipid molecular species can be obtained by the positive neutral loss scan of 141 Da (for PE, pPE, and LysoPE), the positive precursor ion scan of 184 Da (for PC, LysoPC, and SM). Collision energy used is −28 V (N2) for ethanolamine phospholipids, and −30 V(N2) for choline phospholipids. Serine phospholipid species can be detected using either the positive neutral loss scan of 185 Da (−30 V(N2)) or the negative neutral loss scan of 87 Da (+30 V(N2)). All phospholipid species can be analyzed by the negative precursor ion scan of 153 Da under the collision energy at +50 V(N2). Mass ranges of the scan are from 300 to 1,100 Da for the positiveion mode, and from 200 to 1,100 Da for the negative-ion mode, respectively. The precursor scan, neutral loss scan, and molecular-reaction-monitor (MRM) scan modes of ESI-MS/ MS can be used in the data acquisition (16–18). 2. The eluting order of the major lipids with the above-mentioned silica normal-phase HPLC method is the same as that of normal-phase LC/ESI-MS, shown in Fig. 4. 3.6. Reverse-phase LC/ESI-MS/MS and Direct Sample Introduction/ ESI-MS/MS

1. Prior sample analyses by LC/ESI-MS, a phospholipid class should be purified by silica-TLC or normal-phase liquid chromatography. LC analyses are performed on a reverse-phase HPLC column (C18; 3 mm, 3.0 × 250 mm). Acetonitrile– methanol (40:60; v/v) containing 5 mM ammonium formate or acetate, or 5 mM ammonium chloride, can be used as the mobile phase. The flow rate is 1 mL/min. About 300 mL of the mobile phase is introduced into the LC/MS system after the 1/3 split. It takes about 30–40 min to accomplish an analysis. Under the negative-ion mode ionization, [M+COOH]− or [M+CH3COOH]− or [M+Cl]− ions of choline phospholipids are generated for further MS/MS analyses, resulting in producing carboxylate anions of the fatty-acid chains. When a mixture of acetonitrile–methanol/10 mM lithium acetate (40:60:0.3; v/v) is used, [M+Li]+ ions of phospholipid species can be generated, which are particularly useful for further MS/MS analyses of ether phospholipid isomers, such as the structural determination of 1-O-alkyl(16:0)/acyl PC, 1-O-alk-1¢-enyl/acyl PC (or 1-O-alkyl(16:1n−1)/acyl PC and 1-O-alkyl(18:1n−9)/acyl PC) (or 1-O-alk-9¢-enyl(16:1)/acyl PC). 2. The major disadvantages of using reverse-phase LC/ESI-MS and LC/ESI-MS/MS for phospholipidomics are as follows: time-consuming and selective losses of the minor phospholipid species present in biological sample(s).

262

Chen

3. Analyses of phospholipid molecular species can be introduced directly into the mass spectrometer without prior chromatography separation (called “shotgun lipidomics”). The purified phospholipid or crude lipid extract can be dissolved in chloroform–methanol (1:1; v/v) containing 5 mM ammonium formate or ammonium chloride for producing [M+COOH]− and [M+Cl]−, as well as in chloroform–methanol (1:1; v/v) containing 5 mM lithium and sodium acetate for generating [M+Li]+ or [M+Na]+, followed by the MS/MS analyses of the adducts. When analyzing PE, PS, ST, and other acidic phospholipids, the sample can be dissolved in chloroform– methanol (1:1; v/v) containing 0.1% of ammonium hydroxide for producing [M−H]− ions. The infusion flow rate is about 5 mL/min. Mass spectra are averaged over 20–50 scans. 3.7. Structural Designation Based on Molecular Ions and Fragmentation

The most challenging work in phospholipidomics is the structural determination of phospholipid species analyzed by mass spectrometry, to further understand the relationship between structures and functions of the lipids in the nervous system. The structural information of phospholipid species generated by ESI-MS and ESI-MS/MS includes: 1. Molecular ion of a molecular species, mainly represented by the protonated ([M+H]+) and deprotonated ([M−H]−) molecules, as well as [M+Cl]− adduct ion (generated by ESI-MS). 2. The characterization of a polar-head linked to the sn-3 position of the glycerol backbone (generated by ESI-MS and ESI-MS/MS). 3. The composition of the two fatty-acid chains and their locations esterified at the sn-1 or sn-2 position of the glycerol backbone (preferring to use the negative ion ESI-MS/MS) (Fig. 5). 4. The identification of the sn-1-fatty chain linkage, such as the sn-1-acyl fatty chain linkage, the sn-1-O-alkyl fatty chain linkage, and the sn-1-O-alk-1¢-enyl fatty chain linkage (generated by ESI-MS/MS of [M+Li]+ precursors) (Fig. 6). 5. Tables 1– 4 exhibit ESI-MS- and ESI-MS/MS-generated abundant ions and fragments, which are very useful in phospholipidomics, including the structural identification of PC and lysoPC species (Table 1), PE (Table 2), PS and PI (Table 3), and ether phospholipid species (Table 4). The following reading materials may help in structurally elucidating various phospholipid molecular species, such as for choline phospholipids (1–5), aminophospholipids (6–10), inositol phospholipids (11), and other phospholipids (12–15), as well as references on LC/MS and LC/MS/MS applications to phospholipidomics (16–22).


a

c

263

327(22:6)

100

m/z 719 [M-H-87]-

Relative Intensity (%)

255(16:0) [M – H]– 762.6 -H

452 0 200

b

300

400

500

600

700

d

100

Relative Intensity (%)

327(22:6) [M – H]– 748.6 438

0

200

300

400

-H

500

600

700

M/Z Fig. 5. Representative product ion spectra of (a) 1–16:0/2–22:6 PS (m/z 806.6 [M−H]−) and (b) 1-O-alkyl(18:0)/2–22:6 PS and 1–17:0/2–22:6 PS (m/z 820.6 [M−H]−), as well as product ion mass spectra of (c) 1–16:0/2–22:6 PE (m/z 762.6 [M−H]−) and (d) 1-O-alkyl(16:0)/2–22:6 PE (m/z 748.6 [M−H]−), obtained by LC/negative ion ESI-MS/MS. Please pay an attention on (b), a MS-generated molecular ion of phospholipid species sometimes corresponds to the two different molecular species that have the identical molecular weights. It is clear to see that a precursor ion at m/z 820.6 relates to 1-O-alkyl(18:0)/2–22:6 PS and 1–17:0/2–22:6 (evidence by an ion at m/z 269 due to a 17:0 fatty acid chain esterified at the sn-1 position), respectively. This example clearly indicates the importance in structurally elucidating phospholipid molecular species based on ESI-MS/MS-generated fragments (see Fig. 6).

3.8. Quantitative Assay and Percentage Analyses

1. The introduction of lipids at the lower concentration to match the linear responding range is of importance. In the percentage analyses, the experiment can be performed with or without the internal standard. A process for the effective selection of a full-scan mass spectrum of a phospholipid class obtained by LC/ESI-MS is carried out for the following reasons: (i) to make a selected mass chromatography by inputting a mass of protonated or deprotonated molecules of a polyunsaturated phospholipid species (for example, a mass at m/z 834 [M−H]− due to 18:0/22:6 PS that is usually eluted earlier), (ii) to get another selected mass chromatography by inputting a mass of a protonated or deprotonated molecule of a saturated phospholipid species (for example, 760 [M−H]− due to 16:0/18:1 PS that is usually eluted later), and (iii) to select the scan numbers between the rising point of the first peak and the

264

Chen

Fig. 6. Representative product ion spectra of (a) 1-O-alkyl(16:0)/2–22:6 PC, (b) 1-O-alkyl(16:1n−9)/2–22:6 PC (or 1-O-alk-9¢enyl(16:1)/2–22:6 PC), and (c) 1-O-alk-1¢enyl(16:1)/2–22:6 PC, generated by the MS/MS analyses of [M+Li]+ precursors of the PCs using the direct sample introduction. The interpretation of ESI-MS/MS-generated fragments of the three ether PC species is shown in Table 4. The example clearly shows the power of MS/MS in the structural determination of 1-O-alkyl(16:1n−9)/2–22:6 PC (1-O-alk-9¢enyl(16:1)/2–22:6 PC), an ether linkage phosphatidylcholine species containing a novel sn-1-O-alkyl fatty chain.

falling point of the last peak, to get a full scan mass spectrum of phospholipid molecular species for the percentage analyses (Fig. 7). 2. The percentage of each molecular species can be calculated as following: %=

Aor I [(peak)(m / z )] , å Aor å I (peaks)

(1)

where A(peak) or I (peak) refer to the peak area (A) or the peak intensity (I) of selected mass at m/z, and ∑A (peaks) or ∑I (peaks) refer to the sum of all peak areas or all peak intensities when internal standards are not used; %=

Aor I [peak(m / z ) / IS(peak)(m / z )] , A å or å I [(peaks)(m /z ) / IS(peak)(m /z )]

(2)

where A (peak) or I (peak) refer to the peak area (A) or the peak intensity (I) of selected mass at m/z, IS (peak; area or

734.6

758.6

760.6

780.6

782.6

788.6

810.6

834.6

496.5

522.5

524.5

544.5

568.5

16:0/16:0 PC

16:0/18:2 PC

16:0/18:1 PC

16:0/20:5 PC

16:0/20:4 PC

18:0/18:1 PC

18:0/20:4 PC

18:0/22:6 PC

16:0 LysoPC

18:1 LysoPC

18:0 LysoPC

20:4 LysoPC

22:6 LysoPC

542.5

528.5

508.5

506.5

480.5

818.6

794.6

772.6

766.6

764.6

744.6

742.6

718.6

[M−15]−

–

–

–

–

–

868.7

844.7

822.7

816.7

814.7

794.7

792.7

768.7

[M+Cl]−

–

–

–

–

–

840.6

816.6

794.6

788.6

786.6

766.6

764.6

740.6

[M+Li]+

184

184

184

184

184

184

184

184

184

184

184

184

184b

[Polar-head]+

327

303

283

281

255

283

283

283

255

255

255

255c

255

[COOR1]−

–

–

–

–

–

327

303

281

303

301

281

279d

255

[COOR2]−

[5,21, 22]f

[3]e

[16,18,22,23]f

[1–5]e

References

a n:m/N:M (for example 16:0–18:2), where n is the total number of carbon in the sn-1 position and m is the total number of double bonds in fatty-acid chain at the sn-1 position; N is the total number of carbon in the sn-2 position and M is the total number of double bonds in fatty-acid chain at the sn-2 position. See Fig. 1 for structures b m/z 184 is due to phosphocholine structure c A weak ion d An abundant ion (generated by MS/MS of [M−15]− or [M + Cl]− of PC species) e References on the gas-phase ion chemistry of choline phospholipids f References of LC/MS and LC/MS/MS applications on phospholipidomics

[M+H]+

Molecular speciesa

Table 1 MS-generated abundant ions and fragments of the major rat brain PC and LysoPC for phospholipidomics

Lipidomics of the Nervous System 265

690.6

718.6

724.6

752.6

750.6

750.6

734.6

740.6

768.6

764.6

792.6

794.6

p16:0/18:1 PEb

p18:0/18:1 PE

p16:0/20:4 PE

p18:0/20:4 PE

p18:1/20:4 PE

p16:0/22:5 PE

18:0/18:1 PE

16:0/20:4 PE

18:0/20:4 PE

16:0/22:6 PE

18:0/22:6 PE

18:0/22:5 PE

792.6

790.6

762.6

766.6

738.6

732.6

748.6

748.6

750.6

722.6

716.6

688.6

[M−H]−

141

141

141

141

141

141

141

141

141

141

141

141c

[Polar-head]+

482

480

452

480

452

480

436

462

464

436

464

436

[M−H–R2CHC=O]−

283

283

255

283

255

283d

–

–

–

–

–

–

[COOR1]−

329

327

327

303

303

281e

329

303

303

303

281

279

[COOR2]−

[16–22]g

[6–10]f

References

a

n:m/N:M (for example 16:0/20:4), where n is the total number of carbon in the sn-1 position and m is the total number of double bonds in fatty-acid chain at the sn-1 position; N is the total number of carbon in the sn-2 position and M is the total number of double bonds in fatty-acid chain at the sn-2 position b p16:0/18:1: p means plasmalogen (1-alk-1¢-enyl or plasmenyl) c m/z 141 is due to phosphoethanolamine structure d A weak ion e An abundant ion f References on the gas-phase ion chemistry of ethanolamine phospholipids g References of LC/MS and LC/MS/MS applications on phospholipidomics

[M+H]+

Molcuar speciesa

Table 2 MS-generated abundant ions and fragments of the major rat hippocampus PE for phospholipidomics

266 Chen

762.6

788.6

790.6

812.6

818.6

836.6

–

–

16:0/18:1 PS

18:1/18:1 PS

18:0/18:1 PS

18:0/20:4 PS

18:0/20:1 PS

18:0/22:6 PS

16:0/20:4 PI

18:0/20:4 PI

885.6

857.6

834.6

816.6

810.6

788.6

786.6

760.6

[M–H]−

747

729

723

701

699

673

[M−H-87]−

241

241

420

420

420

420

418

392

[M−H–87R2COOH]−

283

255

283

283

283

283

281

c

255b

[COOR1]−

303

303b

327

309

303

281

281

281c

[COOR2]−

[11]d

[16–22]e

[6–10]d

References

a

n:m/N:M (for example 18:0–18:1), where n is the total number of carbon in the sn-1 position and m is the total number of double bonds in fatty-acid chain at the sn-1 position; N is the total number of carbon in the sn-2 position and M is the total number of double bonds in fatty-acid chain at the sn-2 position b An abundant ion c A weak ion d References on the gas-phase ion chemistry of serine phospholipids; e References of LC/MS and LC/MS/MS applications on phospholipidomics

[M+H]+

Molecular speciesa

Table 3 MS-generated abundant ions and fragments of the major rat striatal PS and PI species for phospholipidomics

Lipidomics of the Nervous System 267

798.6

796.6

796.6

1-O-alkyl(16:0)/22:6

1-O-alkyl(16:1n−9)/22:6b

1-O-alk-1¢enyl/22:6

607

607

609

[M+Li-189]+

462

462

464

[M+Li–R2COOLi]+

–

468

–

[M+Li–R2COOH]+

–

364

–

[M+Li–R2COOH-104]

279

–

–

[M+Li-189-R2COOH]+

a

n:m/N:M (for example 1-O-alkyl(16:0)/22:6, where n is the total number of carbon of the alkyl-chain in the sn-1 position and m is the total number of double bonds in fattyacid chain at the sn-1 position; N is the total number of carbon in the sn-2 position and M is the total number of double bonds in fatty-acid chain at the sn-2 position. Please see the references 1, 3, and 5 on structural characterization and fragmentation processes of ether linkage phosphatidylcholine molecular species b A marine ether phosphatidylcholine species having a novel 1-O-alkyl(16:1n−9) fatty chain structure

[M+Li]+

Molecular speciesa

Table 4 MS-generated informative ions and fragments of natural ether linkage PC species for phospholipidomics

268 Chen


269

Fig. 7. An example of phospholipidomics of the brain tissue by LC/ESI-MS (the percentage analysis of the rat striatal PS molecular species before and after treatment by exogenous DHA phospholipid species); (a) a striatal PS species profile of the 21-month-old rat (control-1), (c) a striatal PS species profile of the 21-month-old rat after treated with highly pure DHA PS, (d) a striatal PS species profile of the 21-month-old rat after treated with highly pure DHA PE, (e) a striatal PS species profile of the 21-month-old rat after treated with highly pure DHA monomethyl PE (PMME), and (b) a striatal PS species profile of a 3-month-old rat (control-2). It is clear to see that phospholipidomics by LC/negative ion ESI/MS can provide the striatal PS species profiles of the 21-month-old rats, before and after treatment with the highly pure DHA phospholipids, based on the percentages analyses.

intensity) is due to an internal standard used, and ∑A (peaks) or ∑I (peaks) refer to the sum of all peak areas or all peak intensities when the internal standard(s) are used. 3. Based on the percentage of each molecular species, the distribution profiles of the fatty acids in a phospholipid class can be calculated directly as following (9):

Fatty acids distribution % = [( A or B ) / ( A + B )] ´ D,

(3)

where A is the molecular weight of the sn-1 fatty acid in the molecular species, B is the molecular weight of the sn-2 fatty acid in the molecular species, and D is the percentage of the molecular species (obtained by ESI-MS or ESI-MS/MS)

270

Chen

4. Notes 1. Alternative normal-phase and reverse-phase HPLC methods are available as well (16–22). 2. PS and PC molecular species can be overlapped in the silica HPLC/MS ion chromatography. It can be “separated,” however, if PC species can be detected under the positive-ion mode, whereas PS species is analyzed under the negative-ion mode (5, 6). 3. It should be stated here again that in the quantitative assay and percentage analyses of phospholipid species, the concentration of the lipids introduced into the LC/MS system should be less than 4 mg/total injected lipids, to acquire signals with the linear responding to the phospholipid concentration. The preexperiments are needed (4). 4. To wash a HPLC injector or an autosampler on the LC/MS or LC/MS/MS instruments during analyses, a mixture of chloroform–methanol–water (45/45/10; v/v) is recommended. 5. When the lipid samples need to be analyzed in the next few days (within 4 days), the lipid samples must be stored at −20°C after analyses. A blank sample (methanol can be used) should be introduced into the system for every three sample injections, especially in quantitative assay and percentage analyses, to avoid the memory effect. 6. The percentages of the fatty acids in a phospholipid class usually can be obtained by gas chromatographic analyses of the fatty-acid methyl esters. However, the error of the percentages of the major fatty acids in a phospholipid class analyzed by the method mentioned in the Sect. 3.8.3 in the present chapter and by gas chromatography is within 15% (9). References 1. Hsu FF, Turk J. (2003) Electrospray ionization/tandem quadrupole mass spectrometric studies on phosphatidylcholine: The fragmentation processes. J. Am. Soc. Mass Spectrom. 14:352–63. 2. Chen S, Curcuruto O, et al. Identification of phospholipid molecular species containing two fatty acyl chains differing by 2D by negativeion fast atom bombardment with mass-analyzed ion kinetic energy analysis. Rapid Communi. Mass Spectrom. 1992;6:454–58. 3. Hsu FF, Turk J, et al. Characterization of alkylacyl, alk-1-enylacyl and lysosubclasses of glycerophosphocholine by tandem quadrupole

mass spectrometry with electrospray ionization. J. Mass Spectrom. 2003;38:752–63. 4. Han X, Gross, R.W. Structural determination of lysophospholipid regioisomers by electrospray ionization tandem mass spectrometry. J. Am. Chem. Soc. 1996;118:451–57. 5. Chen S, Li, K.W. Mass spectrometric identification of molecular species of phosphatidylcholine and Lysophosphatidylcholine extracted from shark liver. J. Agric. Food Chem. 2007;55:9670–77. 6. Chen S. Tandem mass spectrometric approach for determining structure of molecular species of aminophospholipids. Lipids, 1997;32:85–100.

Lipidomics of the Nervous System 7. Hsu FF, Turk J. Charge-remote and chargedriven fragmentation processes in diacyl glycerophosphoethanolamine upon low-energy collisional activation: a mechanistic proposal. J. Am. Soc. Mass Spectrom. 2000;11:892–99. 8. Chen S, Carvey PM, Li KW. Characterization of the molecular species of phosphatidylethanolamine from kidney of fresh water snail lymneae stagnalis by mass spectrometry. Rapid Commun. Mass Spectrom. 1999;13:2416–23. 9. Chen S, Li KW. Comparison of molecular species of various transphosphatidylated phosphatidylserine (PS) with bovine cortex PS by mass spectrometry. Chem. Phys. Lipids 2008;152:46–56. 10. Hsu FF, Turk J. Studies on phosphatidylserine by tandem quadrupole and multiple stage quadrupole ion-trap mass spectrometry with electrospray ionization: structural characterization and the fragmentation processes. J. Am. Soc. Mass Spectrom. 2005;16: 1510–22. 11. Hsu FF, Turk J. Characterization of phosphatidylinositol, phosphatidylinositol-4-phosphate, and phosphatidylinosiyol-4,5-bisphosphate by by electrospray ionization tandem mass spectrometry: a mechanistic study. J. Am. Soc. Mass Spectrom. 2000;11:986–99. 12. Hsu FF, Turk J. Structural determination of sphingomyelin by tandem mass spectrometry with electrospray ionization. J. Am. Soc. Mass Spectrom. 2000;11:437–49. 13. Hsu FF, Turk J. Studies on sulfatides by quadrupole ion-trap mass spectrometry with electrospray ionization: structural characterization and the fragmentation processes that include an unusual internal galactose residue loss and the classical charge-remote fragmentation. Structural determination of sphingomyelin by tandem mass spectrometry with electrospray ionization. J. Am. Soc. Mass Spectrom. 2004;15:536–46. 14. Hsu FF, Turk J. Charge-driven fragmenta tion processes in diacyl glycerophosphatidic acid upon low-energy collisional activation: a

271

mechanistic proposal. J. Am. Soc. Mass Spectrom. 2000;11: 797–803. 15. Chen S, Subbaiah PV. Phospholipid and fatty acid specificity of endothelial lipase: potential role of the enzyme in the delivery of docosahexaenoic acid (DHA) to tissues. Biochim. Biophys. Acta 2007;1771:1319–28. 16. Guan Z. Discovering novel brain lipids by liquid chromatography/tandem mass spectrometry. J. Chromatogr. B. 2009;877:2814–21. 17. Brugger B, Erben G, et al. Quantitative analysis of biological membrane lipids at the low picomole level by nano-electrospray ionization tandem mass specteometry. Proc. Natl. Acad. Sci. USA. 1997;94:2339–44. 18. Kakela R, Somerharju P, Tyynela J. Analysis of phospholipid molecular species in brains from patients with infantile and juvenile neuronalceroid lipofuscinosis using liquid chromatography-electrospray ionization mass spectrometry. J. Neurochem. 2003;84:1051–65. 19. Gao F, Tian X, et al. Analysis of phospholipid species in rat peritoneal surface layer by liquid chromatography/electrospray ionization iontrap mass spectrometry. Biochim. Biophys. Acta 2006;1761:667–76. 20. Pang LQ, Liang QL, et al. Simultaneous determination and quantification of seven major phospholipid classes in human blood using normal-phase liquid chromatography coupled with electrospray mass spectrometry and the application in diabetes nephropathy. J. Chromatogr. B. 2008;869:118–25. 21. Pruzanski W, Lambeau L, et al. Differential hydrolysis of molecular species of lipoprotein phosphatidylcholine by group IIA, V and X secretory phospholipase A2, Biochim. Biophys. Acta 2005;1736:38–50. 22. Grandois JL, Marchioni E, et al. Investigation of natural phosphatidylcholine sources: separation and identification by liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS2) of molecular species. J. Agric. Food Chem. 2009;57:6014–20.

wwwwwwwwwwwwwwwwww

Part V Bioinformatics

wwwwwwwwwwwwwwwwww

Chapter 20 Bioinformatics Procedures for Analysis of Quantitative Proteomics Experiments Using iTRAQ Pim van Nierop and Maarten Loos Abstract The combined use of liquid chromatography followed by tandem mass spectrometry (LC-MS-MS) in proteomics research has proven to be a valuable asset in the success of this field of science. Advances in LC-MS-MS technology have allowed researchers to identify an increasing number of proteins from complex biological preparations in a high-throughput fashion. Moreover, techniques based on the labeling of peptides with stable isotopes have made it possible to determine relative differences in abundance of proteins between biological samples. As has been the case for microarray technology, the newly emerging field of quantitative proteomics is associated with the development of novel bioinformatics and statistical approaches that, within the boundaries of particular aspects and limitations of the technique, allow us to ask biological questions and derive meaningful answers (1–3). In this chapter, we describe a protocol of an integrated bioinformatics workflow that deals with the identification of proteins and the relative quantification using iTRAQ labeling in complex proteomics experiments that also involves comparison of quantitative data obtained in separate LC-MS-MS runs. Key words: Proteomics, iTRAQ, Bioinformatics, Protocol, Mascot server, Sequence clustering

1. Introduction 1.1. Protein Identification with Tandem Mass Spectrometry

A typical LC-MS-MS experiment starts with the isolation of proteins from a biological sample that subsequently is enzymatically digested by a protease (such as trypsin) into peptide fragments. Prior to analysis in the mass spectrometer, the complexity of peptides of the protein digest is reduced by one or two steps of liquid chromatography. In the mass spectrometer, the mass of the peptide, also referred to as precursor mass, is determined in the socalled MS1 stage of analysis. Selected peptides with a sufficient


275

276

van Nierop and Loos

MS1 signal are subsequently selected for the MS2 stage. In the MS2 stage, a peptide that was selected based on its precursor mass is fragmented by collision with an inert gas, referred to as Collision Induced Dissociation (CID), which preferentially breaks the peptide at its peptide bonds. The respective masses of these fragments are then determined as a MS2 spectrum. The combined information of the precursor mass and the MS2 spectrum is used for the identification of the peptide sequence. A single LC-MS-MS experiment can result in the recording of tens of thousands of MS1/MS2 recordings. An important first step in proteomics data analysis is the mapping of these mass spectra to the amino acid sequence of the precursor peptide. Using specialized software tools, such as Mascot server (see for instance (4)) or SEQUEST (5), peptide sequence candidates are selected from a predefined list of peptides that match the MS1 precursor mass. Subsequently, these software packages return the probability that a particular amino acid sequence belongs to a particular MS1 precursor mass and its respective MS2 spectrum while taking into account mass imprecision of the mass spectrometry hardware. In the second stage of interpretation, the annotation data of individual spectra, that each can be associated with multiple amino acid sequences at different probabilities, are combined in order to predict a collection of protein species that constituted the original biological sample in a so-called protein report. Although this procedure might seem trivial at first glance, it certainly is not. First, although a single MS2 spectrum most likely has only a single most amino acid sequence match that is most probable, it cannot safely be assumed that this indeed is the correct peptide assignment for this spectrum. Second, certain amino acid sequences are shared between different proteins making it difficult to interpret the finding of such a sequence as evidence of one protein or the other. When generating a protein report, software packages combine the probability of peptide annotations, while implementing the assumption that the solution with smallest number of proteins that explains the presence of all spectra must be the best one (also referred to as Occam’s principle). The protein report, therefore, will be a better-or-worse approximation of the true biological situation. With software tools, such as Mascot server and SEQUEST, MS2 spectra are annotated by determining their degree of similarity to peptide sequences in a predefined list. In most cases, this predefined list of peptide sequences is derived from a collection of protein sequences that was digested in silico at predicted proteolytic sites of the protease used in the experiment. One consequence of this approach is that the correct peptide sequence cannot be assigned to a spectrum if it is not present in the predefined list of proteins. The selection of proteins used for spectrum annotation, therefore, is a determining factor in the outcome

Bioinformatics Procedures for Analysis of Quantitative Proteomics Experiments

277

of the experiment. In most cases, protein collections are derived from public protein databases of which UniProtKB/Swissprot (6), UniprotKB (6), International Protein Index (IPI) (7), and NCBInr (8) are frequently used. Although all mentioned databases contain protein sequences, there are fundamental differences between their protein compositions that impact the final proteomics results. In particular, the method of coping with sequence redundancy of a protein database is to be considered from the perspective of downstream proteomics analysis. In its most strict definition, sequence redundancy refers to the situation where multiple entries in the protein database have an identical amino acid sequence. From a biological standpoint, however, redundancy can have a multitude of different interpretations that can be distinguished based on their interpretation of the relevant biologically entity. UniProtKB/Swissprot (Swissprot in short) is a manually curated database that aims for high-quality protein information and a very low level of sequence redundancy (9). High-quality data is achieved through expert curation of published literature. As a consequence, protein sequences that are predicted from DNA sequences are not included in Swissprot. To remove sequence redundancy, all proteins derived from a single gene are convoluted into a single entry; this means that amino acid variations derived from alternative splicing and from nonsynonymous polymorphisms are not included. UniprotKB (Uniprot Knowledge Base) is a combination of the UniProtKB/Swissprot and UniProtKB/TrEMBL databases. UniProtKB/TrEMBL (TrEMBL in short) consists of a non-curated collection of proteins predicted from the DNA sequences submitted to the EMBLBank, GenBank, and DDBJ nucleotide databases. Protein fragments, isoforms, variants and so on, encoded by the same gene, are stored in TrEMBL as separate entries. In this way, UniprotKB combines highly curated protein information (Swissprot) with protein information that is less annotated and more redundant. The IPI database provides a collection of proteins that includes known proteins and protein predictions based on protein prediction algorithms (10). IPI contains protein information from UniprotKB and other protein databases but does not contain information on fragments and polymorphic sequences. The interesting aspect of IPI is its incorporation of state-of-theart genome annotation information in its prediction of novel protein sequences. As a consequence of the efforts to keep pace with these proceedings, the IPI database is only available for a limited number of species, including human, rat, and mouse. NCBInr is the nonredundant protein database of the National Center for Biotechnology Information. NCBInr contains nonidentical protein sequences translated from all nucleotide sequences in the GenBank database but also contains Swissprot besides some other protein databases. As such, NCBInr contains some well-curated

278

van Nierop and Loos

protein information, as well as automatically generated protein sequences that include fragments, protein isoforms, and variants as a result of polymorphisms and strain differences. Whether a database is suited for the annotation of peptide spectra depends on the biological question and the mass precision of the mass spectrometer. As mentioned, if a desired correct peptide sequence is not present in the database, it is not possible to annotate a spectrum to this sequence and the spectrum will acquire no or, even worse, wrong peptide annotations. To prevent such problems, databases should be selected that contain all required sequences. For instance, if proteins sequences from predicted gene products should be identified, the Swissprot database is not a sensible choice. From this it may seem that the best solution would be to search the collective information contained in all databases. However, the analysis of a larger number of sequences results in a lower confidence score generated by annotation algorithms. The reason behind this is that for every spectrum only peptide sequences with a predicted mass close to the precursor mass are statistically being evaluated for their similarity to the MS2 spectrum. Because the number of statistical tests performed affects the number of false positive results one expects to find (multiple testing problem), a penalty is applied on the significances of individual findings to reduce this effect (multiple testing correction). The effect of searching many proteins at once would be, on average, a proportional increase in the number of evaluations for each spectrum and an according decrease in confidence. A good database for spectrum annotation, therefore, contains all required protein sequences, but as little as additional sequences as possible. The delicate balance of selecting the right database comes down to this example: using the Swissprot database will yield a more significant hit if it contains the correct sequence, as compared to the same sequence being present in for instance NCBInr, purely because the database size negatively influences annotation confidence. One important aspect of the proteomics workflow described in this protocol approaches this problem with the integration of different database search results. The negative impact of database size on the annotation confidence, however, is counterbalanced somewhat by advances in mass spectrometer technology that has resulted in instruments with greater mass resolution and precision (see for instance (11)). The greater mass precision of modern instruments allowed the spectrum annotation software to restrict the evaluation of candidate peptides to a more limited set of peptides with a predicted mass closer to the precursor mass. The mass precision of the instrument used, therefore, is another determining factor in the successful analysis of large sequence databases. The confidence of a peptide match to a mass spectrum as produced by annotation software is, some way or another, derived


279

from the probability that this particular match is a result of chance (null hypothesis). The smaller this probability, the larger the confidence is that the match between peptide sequence and the spectrum can be trusted. The logical consequence of such a scoring system is also that for each spectrum annotation there is a (smaller or larger) chance that its occurrence is in fact a false positive. In a typical LC-MS-MS experiment that generates thousands of spectrum annotations and a considerable number of false positive hits may be expected in the results. The precise algorithm by which particular annotation software assesses the probability of correct vs. incorrect matches is based on a probabilistic model that represents the heart of the software and reflects its assumptions on how information in mass spectra relates to peptide sequences. It is the difference in these algorithms that underlies the fact that the analysis of a single LC-MS-MS database with different software tools results in shared but also distinct peptide identifications (12). To get more accurate estimation of the extent of false positive hits in an experiment that is not directly derived from probabilistic model of the software, an approach to empirically estimate the false discovery can be devised. This empirical estimation makes use of decoy database that has the same amino acid content and peptide mass composition as the real database that was used for spectrum annotation, but in which the sequence of all amino acids has been randomized. Besides annotation in the real database, each spectrum that is matched against the decoy database, which, by the nature of the decoy database, will only yield confident matches purely by chance. By comparing the number of confident matches in the real and the decoy database a measure of the number of false positives can be derived that does not depend on a probabilistic model. The relation between the number of true vs. false positive hits is mostly expressed as a false discovery rate, which is defined as the percentage of the number of true hits divided by the number of false hits. Many proteomics journals nowadays require the false discovery rate to be represented as part of the data for publication (see for instance (13)). An additional possibility of determining the false discovery rate is that a confidence threshold for spectrum annotations can be determined that corresponds to a desired fraction of false positive hits (normally 1 or 5%). The protocol will elaborate on a way to perform the determination of this threshold. 1.2. Protein Quantification with Tandem Mass Spectrometry

Several techniques presented in the last years have enabled the quantitative comparison of proteins between biological samples in LC-MS-MS experiments. One of these techniques is “Isobaric tags for Relative and Absolute Quantification” (iTRAQ™; Applied Biosystems) (see Chaps. 10 and 11 for the application of iTRAQ). iTRAQ consists of chemical compounds that via covalent linkage to the N-termini and side chains of Lysine residues can be used to

280

van Nierop and Loos

label peptides from a particular biological sample (14). Currently, two versions of iTRAQ are available (Applied Biosystems); iTRAQ-4plex that consists of a set of four different labels, and iTRAQ-8plex that consists of a set of eight labels. Each iTRAQ label consists of two different functional moieties (Fig. 1a). The reporter element has a different mass for each iTRAQ label. For the iTRAQ-4plex set the respective reporter elements are 114, 115, 116, and 117 Da, whereas for the iTRAQ-8plex set they are 113, 114, 115, 116, 117, 118, 119, and 121 Da in mass. The mass of the reporter element is counterbalanced by the balancer element so that all iTRAQ labels in a respective iTRAQ set are isobaric, i.e., have the same mass. Collision-induced disintegration of an iTRAQ-labeled peptide prior to MS2 analysis not only breaks the peptide bonds but also breaks the link between the iTRAQ reporter and balancer elements. As a result, the reporter element is no longer attached to the peptide and it will be detectable as a separate peak the MS2 spectrum that corresponds to the specific mass of the reporter (Fig. 1b). Since different labels in an iTRAQ label set are isobaric, peptides labeled with different labels from a particular iTRAQ set behave indistinguishably LC

Fig. 1. Isobaric tags for Relative and Absolute Quantification (iTRAQ) technology. (a) Schematic representation of reporter and balancer groups of iTRAQ-8plex. Note that the combined reporter and balancer groups of each label are of equal mass. (b) Example spectrum with reporter on peaks (see inset ).


281

separation and MS1 analysis. This fact allows for combining peptides from different biological sources that were each labeled with another iTRAQ label prior to LC-MS-MS. A particular peptide that is present in some of the samples will appear as a single ion in the MS1 spectrum. Upon CID and MS2 analysis, however, the relative amount of peptides in the biological samples can be deduced from the relative peak intensities of the respective reporter ions. Because the reporter ions in an MS2 spectrum are all derived from the same precursor ion/peptide, iTRAQ allows for quantification of relative differences in abundance of a particular peptide without having to deal with the uncertainties of matching different peaks in the MS1 spectrum as is the case with other stable isotope-labeling techniques such as SILAC (15). A limitation of iTRAQ is posed, however, by the maximum number of samples that can be compared in a single LC-MS-MS run, which is limited by the number of different labels in an iTRAQ set. Procedures that allow for the comparison of iTRAQ relative quantification data between samples in different LC-MS-MS runs will be treated in this chapter. 1.3. Aim of This Chapter

The sections above provide a brief introduction on several aspects of proteomics data analysis that somehow should be integrated into a quantitative proteomics bioinformatics workflow. This protocol aims to provide a complete description of all bioinformatics procedures of a proteomics experiment that uses iTRAQ for protein quantification. The difficulty with a very detailed and specific description of such a bioinformatics workflow is that its applicability relies heavily on the particular hardware and software setup of the proteomics laboratory. For instance, the software used to extract the MS2 peaklists from the mass spectrometer is dependent on the brand of instrument because the hardware vendor generally provides the software. In addition, the type of spectrum annotation software determines the in- and output format of spectra, the way scores and their significances are represented, and the way decoy databases are incorporated into the search results. Lastly, different programming languages such as Java, Perl, Python, and C++ can be used for the final implementation of the workflow. Given the large number of possible combinations of hardware and software, it is impossible to provide the specific implementation of the workflow that applies to every experimental setup. Therefore, this protocol provides specifics on the hardware and software setup that we have in our laboratory, which consists of a 4000-series MALDI-TOF-TOF instrument by Applied Biosystems and the popular Mascot Server software by Matrix Science for spectrum annotation. Because this protocol emphasizes the theoretical considerations that underlie each specific procedural step we hope, however, that this protocol provides a useful guide for all researchers working in quantitative

282

van Nierop and Loos

proteomics but with different hard and software tools. Given sufficient informatics skills, there are no limitations that prevent implementation of this workflow in another hardware/software environment.

2. Materials 1. 4000-series Applied Biosystems MALDI-TOF-TOF instrument. The protocol assumes data from an iTRAQ quantitative proteomics experiment are present on the instrument database. 2. Server running spectrum Mascot Server software. Mascot Server version must be v2.2 or higher because older versions do not allow for decoy database searches. The server should have recent versions of the NCBInr and Swissprot databases available for spectrum annotation. Guides on the installation and update procedures of protein databases in Mascot Server can be found at the Matrix Science website (16). 3. Installation of the Perl programming language. This language is present by default on most standard Unix, Linux, and Mac OSX operating systems. For Microsoft Windows operating systems, an ActivePerl distribution can be obtained at the ActiveState website (17). 4. Mascot Parser package. This package can be obtained free of charge at the Matrix Science website (16). Make sure to select the particular version that applies to your operating system. The Mascot Parser API documentation is included in the downloaded package or can be obtained separately. 5. Installation of TS2Mascot software. This utility retrieves peaklists from the ABI 4000-series database. TS2Mascot can be obtained free of charge at the Matrix Science website (16) (only available for Microsoft Windows operating system). 6. Installation of Mascot Daemon software. This utility submits peaklists to Mascot Server in a batch-wise fashion. TS2Mascot can be obtained free of charge at the Matrix Science website (16) (only available for Microsoft Windows operating system). 7. fdr_table.pl script. This Perl script is used to determine the relation between score thresholds and the false discovery rate. This script as well as installation and usage instructions can be obtained free of charge at the Matrix Science website (16).


283

8. Installation of the NCBI BLAST package. This package can be downloaded from the National Center of Biotechnology Information FTP site (18). From the BLAST package only the Blastclust program will be used. 9. Installation of BioPerl. This package provides a rich collection of molecular biology-related utilities that can be used in Perl (19). From BioPerl, we will use the Bio::Index::Fasta module. 10. Average level programming skills in Java, Perl, or C++. The implementation discussed in this protocol makes use of the Mascot Parser software, which is only available for these languages.

3. Protocol 3.1. Spectrum Annotation and Protein Identification 3.1.1. Retrieval MS-MS Peaklists from the ABI 4000-Series MALDI Oracle Database

3.1.2. Optional: Filtering of Peaklists

This stage can be performed using the TS2Mascot utility. This utility exports peaklists in mascot generic format (.mgf) for each selected MALDI plate. All peaklists from a single LC-MS-MS run (iTRAQ set) are combined into a single peaklist file by text file concatenation. In the .mgf files for each spectrum, a unique identifier named Peak_List_Id is contained in the line beginning with TITLE. The value of Peak_List_Id should be used as spectrum reference throughout this protocol. It can be extracted from the TITLE line by regular expression matching. Prior to annotation, spectra can be filtered to include only those MS2 peaks with the largest surface area, and to exclude iTRAQ reporter ion peaks that carry no information for spectrum annotation. The goal of spectrum filtering is to reduce the complexity of the MS2 spectrum to the extent that peaks relevant for spectrum annotation are used for annotation, whereas noise peaks are not included. Besides reducing the amount of computation time for the annotation of each spectrum, the removal of noise has the additional effect of improving the annotation scores. In our workflow, we use the 80 peaks with largest surface area for spectrum annotation. It should be noted, however, that spectrum filtering is not essential for obtaining significant annotation results. In addition, since spectrum filtering reduces the complexity of the spectrum, it should be used with caution not to remove informative peaks. Filtered peaklists are saved in .mgf format where the Peak_List_Id in the TITLE line should be present for spectrum identification. A Perl script is provided that performs the task of peak filtering as described above (supplemental data S1).

284

van Nierop and Loos

3.1.3. Spectrum Annotation Using Swissprot and NCBInr Databases

Filtered (or unfiltered if desired) peaklists can be submitted for spectrum annotation via the Mascot Daemon utility supplied by Matrix Science. In order to improve the chance that a spectrum can be annotated, separate Swissprot and NCBInr searches are conducted for each peaklist (but see note below). The parameters for the database search follow standard settings for the Mascot server and should be set appropriately. Since these parameters strongly depend on the particular instrument (e.g., mass error tolerance) and experiment (e.g., species) in question, these fall outside the scope of this protocol. Importantly, the “decoy database” search option should be selected. The result of a Mascot search is contained in a .dat file that can be found in the “data” directory of a standard Mascot Server installation. The name and the path of this .dat file can be obtained in the “Status” tab of the Mascot Daemon application.

3.1.4. FDR Analysis of Spectrum Annotation Results

In order to determine the number of false positive identification that are present in the analysis, we determine the appropriate score threshold (we use the homology threshold) using the fdr_ table.pl script (Fig. 2). This script is executed in order to derive an e-value cutoff that corresponds to an empirical FDR of 1% (or any other desired level). Using this e-value cutoff, spectra from both the Swissprot and NCBInr searches are identified to be included in subsequent stages of analysis (see next section).

3.1.5. Retrieval of Spectrum Annotation Results

Customized peptide summary reports of Mascot searches can be generated using the Mascot Parser API by Matrix Science. Mascot Parser presents views on the annotation data according to rules that specify how duplicate observations, e.g., multiple observations of a single peptide and multiple annotations of the same spectrum, are reflected in the report. The ability to precisely determine these rules and the ability to produce a custom report format are important reasons to use the Mascot Parser API. More on the definition of rules on the treatment of duplicate observations is available in the Mascot Parser documentation. In our workflow, we currently consider only the best scoring spectrum of each peptide sequence. This option is reflected by activating rules A to H in combination with the “MSRES_DUPE_REMOVE_” prefix in the Mascot Parser peptide summary options. Moreover, only the best scoring annotation of each spectrum is considered. By using the FDR e-value cutoff determined under Sect. 3.1.4 as parameter for report generation, only annotations are included that agree with the desired FDR level (Fig. 3). Using the API we can generate a list (e.g., tab-separated) that reflects the (1) spectrum identifier (Peak_List_Id), (2) peptide sequence, (3) accession number of protein match, (4) annotation score (ions score), and (5) homology threshold for each annotated spectrum.


285

a b

"FDR from target-decoy search for matches exceeding homology threshold" "(4117 matches with non-zero score from 5249 spectra)" "Calc. Expect." 0.005000 0.005186 0.005378 0.005578 0.005786 0.006001 0.006223 0.006455 0.006695 0.006943 0.007201 0.007469 0.007746 0.008034 0.008333 0.008642 0.008963 0.009296 0.009642 0.010000

"Actual FDR" 0.0086 0.0086 0.0086 0.0085 0.0085 0.0085 0.0085 0.0084 0.0084 0.0084 0.0095 0.0095 0.0094 0.0094 0.0099 0.0116 0.0120 0.0125 0.0130 0.0129

"Target" 1736 1739 1750 1755 1759 1761 1770 1776 1781 1784 1789 1798 1802 1807 1815 1818 1829 1840 1843 1854

"Decoy" 15 15 15 15 15 15 15 15 15 15 17 17 17 17 18 21 22 23 24 24

Fig. 2. (a) Command to use the fdr_table.pl script by Matrix Science. Path-and-name refers to the filename of the Mascot .dat file in the Mascot data directory, a refers to the minimum expectation value to be evaluated, b is the maximum expectation value to be evaluated, and c is the number of expectation values between the maximum and the minimum expectation value at which to determine the FDR. (b) Example result of fdr_table.pl. The column Calc. Expect. refers to the e-value (theoretical false positive level). The column Actual FDR refers to the empirical FDR level calculated as the number of Decoy hits divided by the number of Target hits. At the desired FDR level (e.g., 0.01), the corresponding e-value cutoff can be obtained. In this example the 1% FDR level corresponds to an e-value cutoff of 0.008333 (highlighted row ).

my $pepsummary = new msparser::ms_peptidesummary( $fileObj, $msparser::ms_mascotresults::MSRES_MUDPIT_PROTEIN_SCORE | $msparser::ms_mascotresults::MSRES_GROUP_PROTEINS, 0, 10000, "", $e-valueCutoff, 0 );

Fig. 3. Perl code for creation of a peptide summary object in Mascot Parser. Variable $fileObj represents the path to a Mascot result file (.dat file). The variable $e-valueCutoff refers to the e-value. By setting $e-valueCutoff to the e-value limit obtained with FDR analysis (Sect. 3.1.4) only annotations will be included that correspond to the desired FDR level. Please refer to the Mascot Parser documentation for additional information and example scripts.

286

van Nierop and Loos

3.1.6. Selection of Spectrum Annotations

In the previous sections, we have selected those spectrum annotations from each search that meet the desired FDR-criteria. Because each peaklist has been searched against both Swissprot and NCBInr, a number of spectra might be represented by two annotations that remain after FDR-selection. In this section the most appropriate annotation, Swissprot or NCBInr, is selected for each spectrum. For this we follow the following rule: if a spectrum is related to a significant Swissprot annotation this annotation is selected, else if a spectrum is related to a significant NCBInr annotation this annotation is selected. The rationale behind this rule is that when a spectrum can be positively annotated in the relatively small Swissprot database we preferably use this annotation, but when this is not the case a possible suitable annotation is used from the larger NCBInr database. For Mascot annotation results, significance of a spectrum annotation is assessed by comparing the ions score with the homology threshold; only annotations where the ion score exceeds the threshold value are considered significant. If the homology threshold is not defined for an annotation, the identity threshold should be used instead. Note: It is not correct to select the best scoring annotation from the Swissprot and NCBInr annotations. The reason for this is that this would be analogous to the simultaneous searching of both databases. The annotation scores and significance thresholds used for annotation do not properly describe to this annotation strategy since they were obtained in separate database searches. A correct execution of this strategy would be to combine the Swissprot and NCBInr databases and perform the spectrum annotation as if it were a single database. One has to take in mind however that this strategy is associated with an overall reduction in annotation significance due to the increased database size (see Sect. 1).

3.1.7. Retrieval of Full-Length Protein Sequences

The NCBInr database search is included to improve the coverage of annotation. Although the NCBInr database is nonredundant in the sense that it does not include duplicates of identical sequences, it may contain multiple entries that relate to a single protein isoform. To remove this sequence redundancy, a protein sequence clustering step is performed that groups highly related sequences into a single protein cluster. Each protein cluster then consists of one or more protein sequences derived from and representing a single protein isoform. It is important that for clustering the spectrum annotations of every LC-MS-MS run in the experiment are combined and treated as a whole. The clustering requires a file that contains the full-length protein sequences that belong to the spectrum annotations selected under Sect. 3.1.6. These sequences can be obtained via a number of ways. The most simple automated approach is to index the FASTA protein sequence file on the Mascot server using the Bio::Index::Fasta module of BioPerl. Bio::Index::Fasta allows to efficiently retrieve a sequence from a large FASTA text file using


287

the accession number as a key. For this to work, it is required that the Swissprot and NCBInr FASTA files are indexed using a regular expression that extracts the protein identifier from the title line of each FASTA sequence that corresponds to the protein identifiers reported by Mascot Server. For NCBInr the GenBank identifier or gi number serves as the protein identifier (Fig. 4a). For Swissprot, either the Swissprot identifier (e.g., KCC2B_HUMAN) or the Swissprot accession (e.g., Q13554) can be used as sequence identifiers (Fig. 4b). Beware that the format of the FASTA title lines may change without announcement requiring an update of the regular expressions in accordance with these changes. For the indexing of the FASTA files a Perl script is provided (supplemental data S2). Full-length protein sequences can be retrieved using the protein identifier using another script (supplemental data S3). Retrieved sequences are exported in a FASTA format file. The approach above provides fast access to the protein sequences. A slower alternative is to retrieve protein sequences using the Swissprot and NCBI web services. A Perl script is provided that incorporated this functionality by reading an input file with protein identifiers and writing a FASTA file with protein sequences (supplemental data S4 and S5). However, it is not encouraged to extract large number of sequences regularly using this approach since it can negatively impact the server performance for other users of the web services. 3.1.8. Removal of Sequence Redundancy with Blastclust

The clustering of protein sequences is performed using the NCBI Blastclust program that is part of the BLAST package. Blastclust performs pairwise alignments of sequences and regards sequences that share a predetermined amount of identical aligned amino acids as a single protein entry, referred to as a cluster. The method for linking proteins to existing clusters is the single linkage approach, which allows proteins to join an existing cluster based on the highest identity score obtained with the member sequences of this cluster. There are three scor-

a >gi|12643413|sp|Q13554.2|KCC2B_HUMAN protein kinase type

RecName:

Full=Calcium/calmodulin-dependent

II subunit beta; Short=CaM kinase II subunit beta; Short=CaMK-II subunit beta

b >sp|Q13554|KCC2B_HUMAN Calcium/calmodulin-dependent protein kinase type II subunit beta OS=Homo sapiens GN=CAMK2B PE=1 SV=2

Fig. 4. (a) Example of a FASTA record of a sequence from NCBInr. The gi number is indicated in bold font. The gi number in this record can be extracted using the “(gi\|[0–9]+)” regular expression. (b) Example of a FASTA record of a sequence from Swissprot. The Swissprot accession (indicated in bold font) can be selected using the “\|(\S+)\|\S+” regular expression. The Swissprot identifier (indicated in italic font) can be extracted using the “\|\S+\|(\S+)” regular expression. Whether the Swissprot identifier or accession should be used depends on the identifier of Swissprot sequences returned by the spectrum annotation software.

288

van Nierop and Loos

ing parameters that determine the outcome of the Blastclust algorithm. First, the identity threshold that determines the percentage of identical amino acids that is shared between two sequences in order to be grouped into a single cluster. Second, the coverage threshold that determines the portion of a sequence in which the identity threshold is considered. Third, the number of sequences that the coverage threshold should apply to. This value can be either “single sided” or “two sided” and determines whether fragments can be confidently grouped into a single cluster. For the Blastclust clustering, we use an identity threshold of 85%, a coverage threshold of 0.90 that is enforced single sided (Fig. 5a). The effect of this is that only sequences and subfragments that are very alike (85% identical over 90% over both sequences) are grouped into a single cluster. 3.1.9. Selection of Unique Spectrum Annotations

a b

The protein clusters produced by Blastclust allow to group spectrum annotations that all represent a single protein isoform but that together pointed to a number of different database identifiers. In this section we map the peptide sequences of spectrum annotations back to the newly defined protein clusters. This is performed by simple text matching of the peptide sequence onto each member protein sequence of a cluster. If a peptide sequence matches any of the member sequences, the respective spectrum is

blastclust -i input.fasta -o blastclustoutput.txt -L 0.85 -S 90 -b F 17367415 149043414 6978547 57358 O88778 17380501 9506497 P11505 1095168 190358918 13489067 56263 P15999 P10719 6978593 56188 32189355 6755588 61557085 P04775 P15146

9507135 Q9QUH6 P06687 P01830 158749559 P16086 P11442 16758008 Q63198 P21575 Q9QUL6 P31596 203055 1374715 P11275 P04797 Q05962 P60881

Q9QWN8 4417207 149056618 207308

Cluster1 Cluster2 Cluster3 Cluster4 ...

Fig. 5. (a) Command line command that submits the input.fasta FASTA file to the Blastclust program with an identity threshold of 85%, a coverage threshold of 0.90 that is enforced single sided. The result is written to a file named blastclustoutput.txt. (b) Example output of Blastclust. Every line represents a single protein cluster, and protein identifiers (gi numbers and Swissprot accession numbers) represent the protein members of the cluster.


289

regarded to represent that particular protein cluster. When a peptide sequence matches multiple protein clusters, this spectrum annotation is disregarded for further analysis. The exclusive inclusion of spectra that uniquely represent a single protein cluster is very stringent and leads to the elimination of a large number of good quality spectra. Since our final goal will be the relative quantification of the protein represented by the cluster, however, this stringency will eventually benefit the precision of the quantification because any spectrum wrongfully attributed to a protein likely contributes erroneous relative iTRAQ label areas. Bottom line is, that although not so relevant for correct protein identification (see next section), the inclusion of redundant spectra may have a strong negative impact on the quantification. The result of Blastclust is a text file where each line with protein identifiers represents a single cluster (Fig. 5b). Before the text matching of peptide sequences of significant annotations on each member sequence can be performed, each line of the Blastclust output file should be read, the protein identifiers should be extracted, and the sequences should be retrieved (see Sect. 3.1.7). Next, spectrum annotations are related to protein clusters by text matching of the peptide sequence and each protein sequence in the cluster. Text matching is performed using modified patterns that combine the peptide sequence with information of cleavage site of the proteolytic enzyme used for protein digestions. For instance, with trypsin that cleaves after every Arg of Lys residue, an R or K residue/character precedes every peptide sequence in a protein with the exception of extreme N-terminal peptides. We found that this approach greatly increased the specificity of peptide assignments. The results of peptide matching are exported while keeping note of the LC-MS-MS run of each spectrum. For downstream analysis, only spectrum annotations are included that could be matched to a single protein cluster. 3.1.10. Identification of Proteins

Each unique spectrum annotation associated with a protein cluster can be regarded as evidence for the presence of the protein represented by that cluster. Because the (small) chance exists that any spectrum annotation is wrong (remember, by using 1% FDR level there are ~1% false spectrum annotations), a single spectrum match to a protein cluster is not regarded as sufficient evidence for the presence of a protein. Instead, two or more peptides that have a different amino acid sequence are required in order to regard this protein as being identified in the experiment. The requirement of two or more peptides for protein identification is enforced at the level of each LC-MS-MS run where different peptides may be identified between LC-MS-MS runs. Only proteins identified in each LC-MS-MS run are included in the downstream analysis. The output of Sect. 3.1.9 (list of unique spectrum annotations) is used in this section by simply counting the

290

van Nierop and Loos

umber of unique peptides from each LC-MS-MS run associated n with protein clusters. Alternative approach: in Sects. 3.1.3, 3.1.4, and 3.1.6 two separate Swissprot and NCBInr searches were preformed for each spectrum. Because each spectrum might be represented by two annotations, this stage was followed by selection of the most appropriate annotation for each spectrum. The use of the Mascot Daemon utility makes it possible to integrate this logic into a standard Mascot search. By defining of an NCBInr follow-up task to a Mascot Daemon Swissprot search task, spectra that could not be significantly matched in the Swissprot database are subsequently searched against NCBInr. With this approach, Mascot produces separate result files for Swissprot and NCBInr searches in which each spectrum is presented once instead of twice. 3.2. Quantification of Identified Proteins 3.2.1. Extraction of iTRAQ Peak Areas

3.2.2. Correction of iTRAQ Peak Areas for Isotope Impurities

Peaks derived of iTRAQ reporter ions are present in the lower mass ranges of the MS2 spectrum. The surface areas of these peaks are extracted from spectra in the unfiltered peaklist that has been obtained in Sect. 3.1.1. The selection of the appropriate peak area of a particular reporter is performed in two stages. First, the peak in the MS2 spectrum is identified with an m/z that most closely matches the predicted m/z of the reporter ion. Second, the mass deviation (delta) between the candidate peak from the spectrum and the predicted m/z of the reporter ion (Table 1) is determined. When the delta falls within the mass accuracy of the mass spectrometer used in the analysis, the surface area peak is accepted as quantitative value for the respective iTRAQ reporter. If this last criterion is not met, the iTRAQ reporter is considered not detected in this spectrum and its quantitative value is set to zero. In order to confidently retrieve iTRAQ reporter ions from the MS2 spectrum, it is important that the mass spectrometer is calibrated well and the mass accuracy is known. Although the peak at 120 m/z does not represent one of the iTRAQ-8plex labels, its surface area is extracted since it will be used during correction for isotope impurities of iTRAQ reagents (see next section). All quantitative values of reporter ions are retrieved for later use while keeping a reference to the spectrum (the Peak_List_Id). An example Perl script is included that extracts iTRAQ peak areas from a .mgf file (supplemental data S6). Individual iTRAQ labels in an iTRAQ set are distinguished in the MS2 spectrum by their known difference in mass. Isotope impurities during the synthesis of iTRAQ labels, however, cause (slight) contamination of iTRAQ labels with reporter elements that are of incorrect mass. For this reason iTRAQ reporter ion peak surface areas are in part determined by peaks of neighboring iTRAQ reporter ion peaks. In order to correctly interpret iTRAQ quantification data, a correction is applied that adjusts extracted iTRAQ


291

Table 1 m/z Value of iTRAQ reporter ions iTRAQ label

m/z Value

iTRAQ-4plex 114

114.1

iTRAQ-4plex 115

115.1

iTRAQ-4plex 116

116.1

iTRAQ-4plex 117

117.1

iTRAQ-8plex 113

113.1

iTRAQ-8plex 114

114.1

iTRAQ-8plex 115

115.1

iTRAQ-8plex 116

116.1

iTRAQ-8plex 117

117.1

iTRAQ-8plex 118

118.1

iTRAQ-8plex 119

119.1

iTRAQ-8plex 120

120.1

iTRAQ-8plex 121

121.1

Comment

Not an iTRAQ label: value used for isotope correction

The values are rounded to one decimal place

reporter ion peak surface areas based on the known amount of isotopic impurities of the iTRAQ labels. The correction is a standard element of any iTRAQ workflow. The correction for a particular iTRAQ reporter ion is applied by subtracting surface areas of other iTRAQ reporter ion, each corrected by its individual isotope correction factor that is provided by the manufacturer. For ease of calculation, a table is provided with converted iTRAQ8plex correction factors (Table 2). This correction procedure is a relatively accurate and convenient approximation of the theoretically most appropriate correction procedure. An example calculation is provided in Fig. 6. 3.2.3. Normalization of iTRAQ Peak Areas

Following labeling with different iTRAQ label reagents, biological samples are pooled and submitted to LC-MS-MS. Ideally, iTRAQ reporter ion surface areas reflect differences in abundance between proteins in the different biological samples. In reality, differences between iTRAQ labels are biased, however, by smaller or larger differences between the amount of peptide material of each sample that contributed to the peptide mixture. To compensate for differences in the input amount, a normalization step is performed. This normalization is performed under the critical

292

van Nierop and Loos

Table 2 Correction factors for iTRAQ-8plex for isotope impurities Correction factors 113

114

116

117

118

119

120

121

1

−0.009 0

0

0

0

0

0

0

114

−0.069

1

0

0

0

0

0

0

115

0

−0.059 1

−0.028 0

0

0

0

0

116

0

0

−0.049

1

−0.038 0

0

0

0

117

0

0

0

−0.039 1

0

0

0

118

0

0

0

0

−0.029 1

−0.057 0

0

119

0

0

0

0

0

−0.019

1

0

−0.003

121

0

0

0

0

0

0

0

−0.084 1

Seed label 113

115

−0.019

−0.047

In order to correct the surface area of an iTRAQ-8plex label, select the correction factors in the respective seed label row. The approximation of the corrected surface area is then obtained by subtracting the relative contribution of neighboring iTRAQ surface areas. The correction factors in this table are derived from the correction factors provided by the manufacturer to allow the calculation as described above. A corresponding table for iTRAQ-4plex could not be derived since the correction factors of this iTRAQ type are determined separately for each batch that is produced

iTRAQ-8plex surface area label 113 114 115 116 117 118 119 120 121

20 30 10 70 50 90 20 5 70

Correction factors 113 surface area * CF label (CF) 1 −0.009 0 0 0 0 0 0 0

Corrected 113 surface area =

20 −0.27 0 0 0 0 0 0 0 19.73

+

Fig. 6. Example calculation of iTRAQ 113 reporter ion surface area corrected for isotope impurities. Correction factors for the iTRAQ 113 reporter ion are selected (highlighted row Table 2). The corrected surface area of the 113 label is calculated by summation of the products each iTRAQ reporter ion surface area in the spectrum and their respective correction factor.

assumption that biologically relevant differences in abundance of proteins between different samples relate only to a negligible minority of all protein present in the samples. Under this assumption, the combined amount of iTRAQ surface area of all spectra between samples (read: iTRAQ labels) must be equal. Normalization is performed by transformation of iTRAQ surface areas so that the total amount of each iTRAQ label is equal.


293

Extracted iTRAQ values are log2-transformed and for each iTRAQ label the average iTRAQ surface area is calculated. The value of each iTRAQ surface area is then normalized by subtraction of the average iTRAQ surface area of the corresponding iTRAQ label. Please note that because after log-transformation subtraction is the equivalent of division on untransformed data, in effect every iTRAQ value is expressed relative to the average surface area of that particular iTRAQ label. 3.2.4. Standardization of iTRAQ Peak Areas

In case an experiment involves replication to gain statistical power, somehow the results of multiple LC-MS-MS runs (iTRAQ label sets) should be analyzed together. A critical difference between iTRAQ quantification experiments and most other quantitative analyses (i.e., microarray) is the relative nature of the quantification. Within one LC-MS-MS run, the area under the iTRAQ signature peaks can be compared to each other to derive meaningful biological comparisons (example Run1, log2 transformed 4plex iTRAQ values of mutant vs. wild-type mice could be: wt1 = 10, ko1 = 11, wt2 = 10.5, ko2 = 11). However, in a subsequent experiment, the same peptide will have a different LC-MS-MS elution profile, which will result in different areas under the iTRAQ signature peaks (example Run2, the log2 values could now be wt3 = 14, ko3 = 15, wt4 = 14.5, ko4 = 15). Three methods can be used to deal with this scaling problem. First, within each LC-MS-MS run log-ratios of control vs. experimental conditions could be calculated (ko1-wt1, ko2-wt2, ko3-wt3, ko4-wt4) and these ratios could be pooled (i.e., 1, 0.5, 1, 0.5) and for instance tested together in a one-sample t-test compared to the value of 0 (i.e., no regulation). However, using this method eight samples are convoluted into four values, resulting in a P-value of 0.014 for the example data. Second, the log2 values could be compared using Analysis of Variance (ANOVA) or independent samples t-test without considering the scaling problem (testing wt1, wt2, wt3, wt4 vs. ko1, ko2, ko3, ko4). However, the variation within groups due to the scaling problem will result in nonsignificance (P = 0.663). Third, a two-way ANOVA can be performed with mutation and LC-MS-MS run (scaling effect) as factors. The resulting P-value for the factor mutation is 0.013. Finally, the scaling effect can be mathematically removed by subtracting the average log2 iTRAQ peak area of each spectrum within each LC-MS-MS run (i.e., subtracting 10.625 in Exp1 and 14.625 in Exp2). It is important to note that this method only centers the log2 value around 0 (i.e., wt1 = −0.625, ko1 = 0.375, wt2 = −0.125, ko2 = 0.375, wt3 = −0.625, ko3 = 0.375, wt4 = −0.125, ko4 = 0.375), but does not alter the variation among the samples within an experiment. Subsequently, these standardized values, which are centered around zero, can be used

294

van Nierop and Loos

in a regular independent sample t-test, which will result in a P-value of 0.002. Although the advantage of the fourth standardization method above may be evident, it comes with one important limitation; the replicate LC-MS-MS runs should all have an identical setup of biological sample arrangement. All measurements are expressed relative to the mean of all samples in the run, and when the mean is very much different because of differences between biological sample types the downstream comparison is untrustworthy. 3.2.5. Integration of Peptide Level iTRAQ Values to the Protein Level

Up until this level all iTRAQ quantification data relate to individual spectrum annotations. Based on the result of Sect. 3.1 where we assigned spectra to particular protein clusters, it is possible to integrate the quantification data of different spectra to a single protein species. Two different criteria determine whether the iTRAQ data of a particular spectrum is to be included at the protein level. First criterion is that the peptide sequence of the respective spectrum annotation of that spectrum is unique for a single protein cluster (as determined under Sect. 3.1.9). The second criterion is that the largest surface area of any iTRAQ label reporter ion in a spectrum must exceed a certain threshold value, in order to reduce the effect of noise on the calculation of protein abundance. The numerical value of this threshold value depends on the machine type and its calibration settings and therefore has to be determined empirically on each experimental system. An acceptable value has to be extracted by evaluation of many iTRAQ reporter ion spectra and assessing the relationship between noise and identifiable reporter ion peaks. As mentioned above, proteins are only considered to be present when two or more peptides were matched to that particular protein. With respect to quantification, similar criteria apply. For single experiments, we currently use a criterion of three or more peptides (with iTRAQ quantification data above threshold) to consider the protein for quantification. Average protein abundance is then calculated as the average of the normalized and standardized iTRAQ values of the three or more peptides. In case the experiment is repeated two or more times, we use a criterion of two quantified peptides in each experiment. If the experiment is repeated multiple times, we typically use criteria of 1-2-2, 1-2-2-2, etc.

3.2.6. Calculate Relative Protein Abundance Differences

Using standardized iTRAQ reporter areas, it is now possible to determine the relative differences between protein clusters identified under Sect. 3.1. Since values were log-transformed in Sect. 3.2.3, protein amounts are expressed as log-values. The log-ratio of a protein of two experimental conditions is calculated by subtracting the average log-transformed-normalized-standardized protein values of the two conditions.

Bioinformatics Procedures for Analysis of Quantitative Proteomics Experiments 3.2.7. Assessment of Significance of Protein Abundance Differences

295

In Sect. 3.2.4, it was explained how to effectively assess the significance for one peptide. Following the same logic, significance can be assessed for log-transformed-normalized-standardized protein averages. In addition, since multiple testing is an issue in proteomics experiments in which many proteins are quantified, a separate assessment using the FDR is necessary. The FDR can be calculated on by permuting the log-transformed-normalized-standardized protein averages. Software packages such as Significance Analysis of Microarrays (SAM, (20)) can be used to do this.

Acknowledgments The authors would like to thank Roel van der Schors for his assistance and Matrix Science for correction of the manuscript.

Supplemental data ref

Script name

Description

S1

filter_mgf.pl

Filters each spectrum in .mgf file to contain largest peaks

S2

make_fasta_index.pl

Creates indexed library of FASTA sequences

S3

fasta_from_index.pl

Extracts sequences from indexed FASTA library

S4

fasta_from_webservice_ncbi.pl

Retrieves FASTA sequences from NCBI webservice

S5

fasta_from_webservice _uniprot.pl

Retrieves FASTA sequences from Uniprot webservice

S6

extract_itraq.pl

Extracts iTRAQ reporter ion peak areas from .mgf file

These scripts can be downloaded from http://neuroweb.cncr.vu.nl/Neuroproteomics Protocols/

References 1. Klychnikov, O.I., et al., Quantitative cortical synapse proteomics of a transgenic migraine mouse model with mutated Ca(V)2.1 calcium channels. Proteomics, 2010. 10(13): p. 2531–5. 2. Van den Oever, M.C., et al., Prefrontal cortex AMPA receptor plasticity is crucial for cue-induced relapse to heroin-seeking. Nat Neurosci, 2008. 11(9): p. 1053–8. 3. von Engelhardt, J., et al., CKAMP44: a brainspecific protein attenuating short-term synaptic

plasticity in the dentate gyrus. Science, 2010. 327(5972): p. 1518–22. 4. Perkins, D.N., et al., Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis, 1999. 20: p. 3551–67. 5. Yates, J.R., III, et al., Method to correlate tandem mass spectra of modified peptides to amino acid sequences in the protein database. Anal Chem, 1995. 67(8): p. 1426–36.

296

van Nierop and Loos

6. The UniProt Consortium, The Universal Protein Resource (UniProt) in 2010. Nucleic Acids Res, 2010. 38(Database issue): p. D142–8. 7. Kersey, P.J., et al., The International Protein Index: an integrated database for proteomics experiments. Proteomics, 2004. 4(7): p. 1985–8. 8. NCBI website. Available from: http://www. ncbi.nlm.nih.gov. 9. Swissprot detail page. Available from: http:// www.expasy.ch/sprot/sprot_details.html. 10. IPI FAQ. Available from: http://www.ebi. ac.uk/IPI/FAQs.html. 11. Hu, Q., et al., The Orbitrap: a new mass spectrometer. J Mass Spectrom, 2005. 40(4): p. 430–43. 12. Alves, G., et al., Enhancing peptide identification confidence by combining search methods. J Proteome Res, 2008. 7(8): p. 3102–13. 13. Bradshaw, R.A., et al., Reporting protein identification data: the next generation of guide-

lines. Mol Cell Proteomics, 2006. 5(5): p. 787–8. 14. Zieske, L.R., A perspective on the use of iTRAQ reagent technology for protein complex and profiling studies. J Exp Bot, 2006. 57(7): p. 1501–8. 15. Ong, S.E., et al., Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics, 2002. 1(5): p. 376–86. 16. Matrix Science homepage. Available from: www.matrixscience.com. 17. ActiveState home page. Available from: www. activestate.com. 18. NCBI FTP site. Available from: ftp.ncbi.nlm. nih.gov/blast/executables/LATEST/. 19. Bioperl home page. Available from: www. bioperl.org. 20. Tusher, V.G., R. Tibshirani, and G. Chu, Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA, 2001. 98(9): p. 5116–21.

Chapter 21 Statistical Analysis of Spectral Count Data Generated by Label-Free Tandem Mass Spectrometry-Based Proteomics Thang V. Pham and Connie R. Jimenez Abstract Label-free strategies for quantitative proteomics provide a versatile and economical alternative to labelingbased proteomics strategies. We have shown for different types of biological samples that spectral counting-based label-free quantitation is a promising avenue for biomarker discovery. Analyzing spectral count data generated from these studies is, however, not straightforward, as commonly used techniques for genomics data analysis are not suitable. In this book chapter, we describe three methods to analyze spectral count data, namely, cluster analysis, significance analysis of independent samples, and significance analysis of paired samples. For cluster analysis, we devise a novel distance measure between samples based on the Jeffrey divergence. This measure prevents highly abundant proteins from dominating others in contribution to the total sample difference. We employ the beta-binomial distribution for significance analysis of independent samples, which integrates both within-sample variation and between-sample variation into a single statistical model. Finally, the Mantel–Haenszel test is used for significance analysis of paired samples. We provide detailed illustrations of the steps involved in the analyses. Key words: Beta-binomial distribution, Biomarker discovery, Cluster analysis, Comparative analysis, Label-free tandem mass spectrometry-based proteomics, Spectral counting

1. Introduction Mass spectrometry (MS) is a powerful approach for identifying and quantifying hundreds of proteins in complex proteomes (1). This proteomics strategy is being improved both in the field of multidimensional protein separation and in high-resolution MS instrumentation. In addition, multiple analyzers for tandem mass spectrometry (MS/MS) can be used for accurate peptide identification. This chapter addresses issues generally involved in statistical data analysis for discovery of proteins with altered abundances and candidate biomarkers in tandem mass spectrometry-based proteomics.


297

298

Pham and Jimenez

We concentrate on label-free strategies for quantitative proteomics (2) as for analysis of multiple clinical samples, they provide a versatile and economical alternative to labeling-based proteomics strategies. In particular, spectral counting, the number of MS/MS events observed for a protein in the mass spectrometer, has been shown to correlate strongly with the protein’s abundance in a complex mixture (3). Spectral counting has been successfully used in a number of recent studies (4–10). In comparison to intensity-based quantification which requires complex signal processing, quantification using spectral counts is advantageous because the numbers are readily available after protein identification. Hence, spectral counting is the method of choice when robust signal processing tools are lacking. This chapter describes our approach to three common tasks in analyzing spectral count data: cluster analysis, significance analysis of independent samples, and significance analysis of paired samples. We use R as our computational platform (11). In addition, we have organized frequently used functions into a toolbox called oplSC. For each of the three tasks, we demonstrate a detailed analysis using the toolbox and R. The chapter is organized as follows. Section 2 describes the data preparation step and notions used in the rest of the chapter. Cluster analysis is presented in Sect. 3. Significance analyses of independent samples and paired samples are presented in Sections 4 and 5, respectively. Section 6 concludes the chapter.

2. Data Preparation and Notations

We begin with a data matrix of spectral count numbers. Such matrix can be obtained from the result of database search using a bioinformatics tool for data organization such as Scaffold (Proteome Software Inc., Portland, OR, USA). The columns of the matrix denote samples, and the rows denote proteins. Let X be an m × n data matrix of n samples and m proteins. Thus, Xij is the spectral count number of protein i in sample j . The running index i is used for proteins and j for samples in the rest of the chapter. For comparative studies, let y denote the group number, y j = 1,…, G where G is the number of groups. For example, for a case/control study ( G = 2 ), y j = 1 indicates that sample j belongs to group 1 (case), and y j = 2 indicates that sample j belongs to group 2 (control) (see Table 1). In R, the data matrix X can be conveniently stored in a table structure together with other details such as protein names and identifiers in additional columns. Data normalization is an important first step because of the variation in sample load. An effective normalization procedure is

299

Statistical Analysis of Spectral Count Data Generated

Table 1 Data matrix and notations of a two-group comparison experiment Group 1 ( n1 = 3 )

Group 2 ( n2 = 3 )

y1 = 1 …

y2 = 1 …

y3 = 1 …

y4 = 2 …

y5 = 2 …

y6 = 2 …

…

Xi1 …

Xi 2 …

Xi 3 …

Xi 4 …

Xi 5 …

Xi 6 …

Total counts

t1 =

t2

t3

t4

t5

t6

Sample group

… Protein i

∑X

1≤ i ≤ m

i1

to rescale the spectral count number in each sample so that all samples have an equal number of total counts. The average sample count z is

z=

∑

1≤ i ≤ m ;1≤ j ≤ n

X ij

n

.

(1)

The normalization factor s j for sample j is then

sj =

∑

z 1≤ i ≤ m

(2)

Xij

~

and the normalized value Xij for protein i in sample j is ~ Xij = s j X ij .

(3)

The raw fold change between two groups 1 and 2 defined as the ratio between the average abundances can be calculated as ~

fi =

(1/ n1 )∑ y j =1 Xij ~

(1/ n2 )∑ y j = 2 Xij

,

(4)

where n1 and n2 are the number of samples in group 1 and 2, respectively. One has to take into consideration the zero-count situation. One approach is to add a small number as a regularization term. Alternatively, one might choose to treat the zero cases separately. Box 1 shows an example for data preparation using an example dataset. Functions with the opl prefix are implemented in our toolbox. The function opl.fc calculates fold change values as in Eq. 4 with positive and negative sign denoting up and down regulation respectively. Zero cases are treated separately by assigning

300

Pham and Jimenez

2.1. Box 1 Data Preparation, Normalization, and Fold Change Calculation


301

the values −100 and +100 to these cases. If one wishes to use a regularization term for the zero cases, it is a matter of providing the function with an adjusted data matrix as input.

3. Cluster Analysis and Heatmap Visualization

Cluster analysis and heatmap visualization can be used to get a global overview of the data. There are two types of clustering: protein clustering and sample clustering (Figs. 1 and 2). The former is used to explore clusters of proteins exhibiting similar regulation pattern. The latter is used for grouping of samples and to detect

Fig. 1. A heatmap visualization and unsupervised clustering of biological samples and proteins. The number of protein clusters is set to 5.

302

Pham and Jimenez

Fig. 2. A heatmap visualization and supervised clustering using proteins with p-value less than 0.05 in the case/control comparison.

possible outlier. In combination with heatmap visualization, the two clustering analyses enable the detection of interesting patterns in the data. We make use of hierarchical clustering tools for both protein clustering and sample clustering because of its simplicity and its visualization capability. The hierarchical clustering approach starts by computing the distance between all pairs of objects (in our case, an object is either a protein or a sample). Initially, each object is assigned to its own cluster. Subsequently, object clusters are merged step-by-step until the last cluster is formed containing all objects. The final cluster is the root of a hierarchical clustering tree. The process of merging requires the calculation of distance between two object clusters. There are several methods to compute this distance based on the distances between the objects within the two clusters, including single linkage, complete linkage,


303

average linkage, and the Ward’s method. The reader should consult the R user’s manual for a discussion of the different approaches (11). Here, we concentrate on the distance measure between two individual objects, which is different for protein clustering and sample clustering. For protein clustering, to group proteins that regulate similarly, after total count normalization (and optionally, log transformation), the abundances are further normalized to zero mean and unit variance per protein (12). Subsequently, the Euclidean distance measure is employed. Specifically, the average abundance ui of protein i across all samples is

∼ 1 ∑ Xij n j

(5)

∼ 1 ( Xij − ui )2 . ∑ n j

(6)

ui = and the variance vi is

vi =

 ij is The zero mean and unit variance normalization X ∼

 ij = Xij − ui . X vi

(7)

The Euclidean distance between protein i1 and i2 is

1/ 2

di1i2

 i j − X  i j )2  . =  ∑ (X  1 2  j 

(8)

For sample clustering, let the detection rates of a protein i in the two samples j1 and j2 be λ ij1 = Xij1 / t j1 and λ ij2 = Xij2 / t j2 . We define the difference in protein i between the two samples as

di( 1

j ; j2 )

= λ ij1 ln

λ ij1 λ ij2

+ λ ij2 ln

λ ij2 λ ij1

.

(9)

This measure is the Jeffrey divergence, a symmetric variant of the Kullback–Leibler divergence, between two Poisson distributions with parameters λ ij1 and λ ij2 . This measure has the advantage that it prevents highly abundant proteins from dominating others in contribution to the total sample difference. The difference between two samples is defined as the sum of differences in individual proteins

m

d ( j1 ; j2 ) = ∑ di( j1 ; j2 ) . i =1

(10)

304

Pham and Jimenez

3.1. Box 2 Cluster Analysis and Heatmap Visualization

Box 2 presents two examples of cluster analysis using the data in Box 1. For unsupervised cluster analysis, one may specify a desired number of protein clusters. The function opl.cluster returns the cluster assignment of the proteins, which can be useful for subsequent analysis such as functional annotation. Note that it is important to use a complete data matrix for normalization purposes. Thus, for supervised clustering, it is required to specify the indices of interesting proteins via the index parameter rather than using a subset of rows of the data matrix as input.

4. Significance Analysis of Independent Samples

Significance analysis gives statistical information to aid in the determination of proteins that exhibit altered protein level between different conditions. We can broadly categorize current methods for significance analysis of spectral count data into ones that require replicates and ones that do not (13, 14). When there are more than one replicates for each sample group, the t-test (15) can be considered. The t-test assuming a normal distribution can be used with either the raw spectral count numbers or their transformed values (for examples, logarithm or square root transformation). The t-test however does not take into account withinsample variation, and hence, it can result in artificially large test statistics for proteins at a high abundance region.


305

The Fisher’s exact test and the G-test (15) are representative for the class of methods that do not require replicates. For each protein, a 2 × 2 contingency table is created. The tests are based on the assumption of a multinomial distribution and a hypergeometric distribution of the number of counts in the contingency table, respectively. When there are replicates, a suboptimal approach is to pool the data. In ref. (16), the authors use the beta-binomial distribution to test the significance of differential protein abundances expressed in spectral counts. The beta-binomial distribution is able to integrate both within-sample variation and between-sample variation into a single statistical model. Consider a spectral count number Xij of protein i in sample j . Since each protein is considered separately, for clarity, we simplify the notation as follows, x j = Xij . To model the within-sample variation, assume x j is distributed according a binomial distribution with success probability π j , 0 £ π j £ 1 ,

 tj  x t −x p ( x j | π j , t j ) =   π j j (1 − π j ) j j . x  j

(11)

Furthermore, for the between-sample variation, π j is modeled as a random variable from a beta distribution with real parameters a > 0 and b > 0

p(π j | α, b ) =

πaj −1 (1 − p j )b −1 B(a , b )

,

(12)

where B(·,·) is the beta function. The marginal distribution of x j is then the beta-binomial distribution (17)

 t j  B(a + x j , t j + b − x j ) . p( x j | a , b , t j ) =   B(a , b )  xj 

(13)

The likelihood ratio test can be used for significance analysis (15, 18). Let Lg be the maximal value of the log likelihood for each group, g = 1,…, G. Let L0 be the maximal value of the log likelihood when data from all groups are used. The statistic S to test the homogeneity across groups G   S = 2  − L0 + ∑ Lg  (14) g =1   is approximately χ distributed with 2 (G − 1) degrees of freedom. Experimental results show that the beta-binomial test performs favorably in comparison with other methods on several datasets in terms of both true detection rate and false-positive rate. In addition, it can be applied for experiments with one or more replicates, and for multiple condition comparisons (16). 2

306

Pham and Jimenez

4.1. Box 3 BetaBinomial Test for Significance Analysis

Box 3 presents a significance analysis of the data in Box 1 using the beta-binomial model.

5. Significance Analysis of Paired Samples

Consider the case where the samples taken from each patient before and after treatment. The data in this case are no longer independent. This experimental setting is typically dealt with using a paired statistical test such as the paired t-test (15). For spectral count data, for each patient we can construct a 2 × 2 table for each protein, resulting in K 2 × 2 tables for the dataset where K = n / 2 . This section briefly describes a statistical test called the Mantel–Haenszel test for significance analysis of paired samples.


307

Table 2 ~ A 2¥2 table of protein i for sample pair k ( j and j ) ~

Sample j

Sample j

Column sum

Protein i

x11k

x12k

x1+k

All other proteins

x21k

x22k

x2+k

Row sum

x+1k

x+2 k

x++k

For ease of presentation, we introduce a new set of notations. Table 2 illustrates a 2 × 2 table for sample pair k (consisting of ∼ sample j and j ) for protein i , k = 1,…, K . In relation to the previous notations, we have x11k = Xij , x12k = X ∼ , x21k = t j − x11k , ij and x22 k = t ∼ − x21k . In addition, x1+k and x2+k are row sum, x+1k and j x+2 k are column sum, and x++k is the total sum of the 2 × 2 table. As in the Fisher’s exact test, assuming conditional independence, the hypergeometric mean and variance of x11k are

x1+ k x+1k , x++ k

(15)

x1+ k x2 + k x+1k x+2 k . 2 x++ k ( x++ k − 1)

(16)

E ( x11k ) =

V ( x11k ) =

The Mantel–Haenszel statistic M 2 is

M2 =

 

1 ∑ x11k − ∑ E ( x11k ) − 2 

∑ V ( x11k )

2

.

(17)

Under the null hypothesis 2of conditional independence, this statistic has approximately a χ distribution with one degree of freedom. We use an existing R function to carry out the Mantel– Haenszel test. For convenience, we provide a wrapper in the toolbox that saves the user from manually creating the K 2 × 2 tables. See Box 4 for an example session with significance analysis of paired data. It is important to note that the concept fold change computed from the Mantel–Haenszel test is different from Eq. 4. It is the so-called common odds ratio of the K tables. For table k in Table 2, the odds ratio OR is

OR =

x11k x22 k . x21k x12 k

(18)

308

Pham and Jimenez

5.1. Box 4 The Mantel–Haenszel Test for Paired Data

Since typically the values x21k and x22k are considerably greater than x11k and x12k , the odds ratio OR approximates well the fold change in Eq. 4.

6. Concluding Remarks We discuss three tasks for analyzing spectral count data in tandem mass spectrometry-based proteomics, namely, cluster analysis and visualization, significance analysis of independent samples, and significance analysis of paired samples. Our future work is to extend the features of the oplSC toolbox to cover other experimental settings such as time course data, one-way ANOVA, multiway ANOVA. Currently, the beta-binomial test is capable of multigroup comparison and, hence, can be used for oneway ANOVA analysis. Finally, the oplSC toolbox and the datasets used in the illustrations can be obtained from the authors.

Acknowledgments This work is supported by the VUmc Cancer Center, Amsterdam. References 1. Aebersold R, Mann M (2003) Mass spectrometry-based proteomics. Nature, 422(6928):198–207. 2. Domon B, Aebersold R (2006) Mass spectrometry and protein analysis. Science, 312(5771):212–217. 3. Liu H, Sadygov RG, Yates JR III (2004) A model for random sampling and estimation of

relative protein abundance in shotgun proteomics. Analytical Chemistry, 76(14): 4193–4201. 4. Albrethsen J, Knol JC, Piersma SR, Pham TV, de Wit M, Mongera S, Carvalho B, Verheul HM, Fijneman RJ, Meijer GA, Jimenez CR (2010) Sub-nuclear proteomics in colorectal cancer: Identification of proteins enriched in


5.

6.

7.

8.

9.

10.

the nuclear matrix fraction and regulation in adenoma to carcinoma progression. Molecular and Cellular Proteomics, 9(5):988–1005. Dix MM, Simon GM, Cravatt BF (2008) Global mapping of the topography and magnitude of proteolytic events in apoptosis. Cell, 134(4), 679–691. Piersma SR, Fiedler U, Span S, Lingnau A, Pham TV, Hoffmann S, Kubbutat MHG, Jimenez CR (2010) Workflow comparison for in-depth, quantitative secretome proteomics for cancer biomarker discovery: Method evaluation, differential analysis and verification in serum. Journal of Proteome Research, 9(4): 1913–1922. Rajcevic U, Piersma SR, Bougnaud S, Pham TV, Enger P, Bjerkvig R, Jimenez CR, Niclou SP (2009) Enrichment of tumorigenic stemlike cells in biopsy spheroids from colorectal cancer. In Proceedings of the 8th Annual World Congress HUPO 2009, Toronto, Canada. Ramani AK, Li ZH, Hart GT, Carlson MW, Boutz DR, Marcotte EM (2008) A map of human protein interactions derived from coexpression of human mRNAs and their orthologs. Molecular Systems Biology, 4:180. Saydam O, Senol O, Schaaij-Visser TB, Pham TV, Piersma SR, Stemmer-Rachamimov AO, Wurdinger T, Peerdeman SM, Jimenez CR (2010) Comparative protein profiling reveals minichromosome maintenance (MCM) proteins as novel potential tumor markers for meningiomas. Journal of Proteome Research, 9(1):485–494. Zybailov B, Friso G, Kim J, Rudella A, Rodriguez VR, Asakura Y, Sun Q, van Wijk KJ (2009) Large scale comparative proteomics of a Chloroplast Clp protease mutant reveals folding stress, altered protein homeostasis, and feedback regulation of metabolism. Molecular & Cellular Proteomics, 8(8), 1789–1810.

309

11. R Development Core Team (2009) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-90005107-0, URL http://www.R-project.org. 12. Schmidt MW, Houseman A, Ivanov AR, Wolf DA (2007) Comparative proteomic and transcriptomic profiling of the fission yeast Schizosaccharomyces pombe. Molecular Systems Biology, 3:79. 13. Bantscheff M, Schirle M, Sweetman G, Rick J, Kuster B (2007) Quantitative mass spectrometry in proteomics: a critical review. Analy tical and Bioanalytical Chemistry, 389(4), 1017–1031. 14. Zhang B, VerBerkmoes NC, Langston MA, Uberbacher E, Hettich RL, Samatova NF (2006) Detecting differential and correlated protein expression in label-free shotgun proteomics. Journal of Proteome Research, 5(11), 2909–2918. 15. Sokal RR, Rohlf FJ (1995) Biometry: the principles and practice of statistics in biological research (3rd edition). W. H. Freeman: New York., Chapter 17. Analysis of frequencies, 685–793. 16. Pham TV, Piersma SR, Warmoes M, Jimenez CR (2010) On the beta binomial model for analysis of spectral count data in label-free tandem mass spectrometry-based proteomics. Bioinformatics, 26(3):363–369. 17. Skellam JG (1948) A probability distribution derived from the binomial distribution by regarding the probability of success as variable between the sets of trials. Journal of the Royal Statistical Society. Series B (Methodological), 10(2), 257–261. 18. Williams DA (1975) The analysis of binary responses from toxicological experiments involving reproduction and teratogenicity. Biometrics, 31(4), 949–952.

wwwwwwwwwwwwwwwwww

Index A Acetic acid.................................31, 100, 101, 109, 112, 120, 122, 123, 134, 161, 164, 186, 187 Acetone................................89, 98, 102, 147, 149, 150, 155, 157, 233, 235, 236, 238 Acetonitrile..................................6, 133, 135, 148, 149, 151, 160, 161, 163, 164, 184, 203, 233, 261 Acridine orange.......................................................... 36, 39 Acrylamide........................39, 84–88, 97, 99, 105, 120, 122, 161, 165, 166, 170, 173, 175, 177, 179, 184, 186 Active zone....................................................... 70, 115–123 Affinity chromatography................................................ 129 Agilent 3100 OFFGEL fractionator.............................. 148 Akinesia.......................................................................... 212 Alexa Fluor..................................................................... 214 Alkaline phosphatase.............................................. 171, 174 Alzheimer’s disease (AD)......................................... 28, 246 Amicon filter.......................................................... 248–251 e-aminocaproic acid.......................................................... 30 6-aminohexanoic acid................................................. 82, 84 Ammonium persulfate (APS).........................100, 105, 120, 121, 173, 178, 184, 186 Ammonium sulphate...................................................... 101 Anesthesia.............................................................21, 35, 74 Angiogenesis.......................................................... 130, 131 Anion exchange chromatography (SAX), 183 Antibiotics.............................................................. 202, 206 Antibody..................................32, 41, 72, 73, 75–77, 81–89, 116, 162, 170– 172, 174, 177–179, 203, 204, 206, 208, 214, 217 Antibody Super Shift Assay....................................... 81–89 Ascorbic acid...................................................120, 121, 123 Axon......................................................................63, 69, 70

B 16-BAC-SDS polyacrylamide gel electrophoresis................................. 118–121 Beckman Optima LE–80K.............................................. 31 Beckman 45Ti rotor................................................. 31, 134 Beckman VTi 50 vertical-type rotor................................ 31 BEEM-capsules......................................................... 33, 41 Benzyldimethyl-n-hexadecylammonium chloride (16-BAC).............................................. 118–123

Beta-binomial distribution..................................... 305, 306 Bioinformatics........................................ 230, 232, 233, 236, 239, 275–295, 298 Biomarker....................................................4, 129, 130, 297 Biomarker discovery........................................... 5, 243–251 BioPerl.................................................................... 283, 286 Biotin.......................................................129, 179, 200, 201 Biotinylation....................................................199, 200, 208 Bisacrylamide.................................... 84, 120, 178, 184, 186 2-bis(2-hydroxy-ethyl)amino]–2-hydroxymethyl)–1,3propanediol (Bis-Tris)..................................... 32 BLAST....................................................137, 236, 239, 240 Blastclust.................................................137, 283, 287–289 Blue Native....................................................81–89, 97, 171 Blue Native polyacrylamide gel electrophoresis (BN-PAGE)..................................81–83, 86–89 Bradford..........................................................132, 134, 185 Bradford assay.....................................................76, 99, 103 Bradykinesia................................................................... 212 Brain capsula externa...................................................... 21, 25 cerebellum..................................... 13, 14, 17, 22, 49, 58 corpus callosum...............................................20, 24, 25 cortex................................ 14, 17– 21, 23–25, 72, 76, 97 forebrain..............................................48, 52–56, 58, 59 hippocampus................................. 13, 14, 17–19, 22–24 medial prefrontal cortex (mPFC)...................14, 17, 20, 21, 24, 25 medulla......................................................13, 14, 17, 22 olfactory bulb...................................................13, 22, 24 pons...........................................................13, 14, 17, 22 prefrontal cortex........................................13, 14, 24, 25 striatum......................................... 13, 14, 17, 24, 25, 76 substantia nigra pars compacta (SNpc)....................... 23 thalamus..................................................................... 14 Bromophenol blue............................................ 98–100, 122 BSA.................. 58, 74, 77, 78, 150, 171, 178, 179, 204, 207

C CaCl2........................................................................ 51, 147 CaMKII..............................................................49, 57, 130 Capillary electrophoresis (CE)....................................... 230 Capillary reverse phase (RP)...................... 6, 133, 135, 137, 148, 151, 160, 161, 164, 235, 237, 238, 240

Ka Wan Li (ed.), Neuroproteomics, Neuromethods, vol. 57, DOI 10.1007/978-1-61779-111-6, © Springer Science+Business Media, LLC 2011

311

Neuroproteomics 312 Index

CAPSO.................................................................. 174, 178 Capsula externa.......................................................... 21, 25 Casein..............................................................184, 190–192 Cation exchange chromatography (SCX)....................... 6, 133, 135–137, 155, 183, 203, 206, 207, 231, 232, 235, 237, 238 Cellomics................................. 211–213, 215, 218, 220, 223 Cerebellum.......................................... 13, 14, 17, 22, 49, 58 Cerebrospinal fluid (CSF)...................................... 243–251 CHAPS...................................................................... 46, 98 Chemiluminescence............................................... 176, 177 Chloroform......................119, 121, 123, 256, 258–262, 270 Cholesterol..................................................................... 253 Cluster analysis................................................298, 301–304 Collision induced dissociation (CID).....................136, 164, 233, 234, 238, 276, 281 Colloidal Coomassie blue................................123, 162, 163 Comparative analysis...................................................... 250 Coomassie Brilliant Blue (CBB)....................81–83, 86, 96, 101, 102, 106, 107, 109, 120 Corpus callosum................................................... 20, 23–25 Cortex.................................14, 17–19, 21, 23–25, 72, 76, 97 Cryostat............................................................................ 26 a-cyano hydroxy-cinnamic acid (CHCA)..............133, 139, 148, 149, 235, 238 Cys-S-beta-propionamide.............................................. 165 Cytoskeleton............................................ 50, 57, 69, 70, 212

D 2D BAC-SDS gel electrophoresis...................... 6, 118–119 Decapitation..........................................................49, 52, 57 Dendrite..................................................................... 69, 70 Deoxycholate.................................................................... 50 Dephosphorylation......................................................... 181 Detergent-resistant membrane......................................... 49 Dextran..................................................................64, 65, 67 2D fluorescence difference gel electrophoresis............... 111 2D gel.............................................. 4, 95, 96, 102, 108–112 Dharmafect......................................................214, 216, 223 DHB................................................................188, 235, 238 2D IEF-SDS gel electrophoresis........................................ 6 Digitonin...............................................................82, 84, 85 Dithio-DL-threitol (DTT)...................... 30, 74, 89, 98, 99, 104, 111, 120, 169, 172, 177, 185, 203 DMEM/F12 culture medium........................................ 214 Dodecyl-b-D-maltoside (DDM)..........................82, 84, 85 Dopamine........................................................212, 213, 224 Douncer...................................................................... 73, 74 1D-PAGE/LC-ESI MS/MS............................................. 6 DTT..................................30, 74, 89, 98, 99, 104, 111, 169, 172, 177, 185, 203 Dulbeco’s Modified Eagle’s Medium (DMEM)................202, 204, 214–216 Duracryl............................................................99, 100, 112

E ECF........................................................................ 174, 177 EDTA.......................................30, 33, 74, 87, 185, 187, 203 Electro-blotting...............................................171, 174, 176 Electron microscopy BEEM-capsules................................................... 33, 41 epon...................................................................... 33, 41 glutaraldehyde...................................................... 32, 41 OsO4..................................................................... 33, 41 PIPES................................................................... 32–33 ultramicrotome........................................................... 41 Electron transfer dissociation (ETD)..................... 234, 238 Electrospray ionization (ESI).............................6, 129, 135, 159–167, 190, 231, 232, 234, 235, 237–238, 255–264, 269 EMBL-Bank.................................................................. 277 Endocytosis.................................................................... 116 Epitope..................................................................73, 88, 89 Epon........................................................................... 33, 41 ESI-IT mass spectrometer..................................... 235, 238 ESI MS/MS............................... 6, 129, 135, 167, 255–257, 259, 261–264, 269 Ethanol..............................33, 41, 74, 77, 98, 100–102, 108, 134, 139, 140, 148, 151 Ether phospholipids........................................255, 261, 262 Ethylenediaminetetraacetic acid (EDTA)................. 30, 33, 74, 87, 184, 185, 187, 203 1-ethyl–4-phenyl-pyridinium (MPP+).......................... 213, 214, 216, 223, 224 Expression proteomics.................................................... 4–7 Extracellular matrix (ECM)........................................... 222

F FDR........................................................284–286, 289, 295 FeCl3....................................................................... 185, 187 Ferrous sulfate.................................................120, 121, 123 Fetal bovine serum (FBS)................................184, 185, 202 Fetal calf serum (FCS)............................................. 32, 214 Forceps.......................................................14–17, 20, 22–24 Forebrain...................................................48, 52–56, 58, 59 Formaldehyde............................................31, 101, 108, 214 Fourier transform ion cyclotron resonance (FTICR).............................. 231, 234 Free flow electrophoresis (FFE) buffer for FFE acetic acid............................. 31, 100, 101, 109, 112, 120, 122, 123, 134, 161, 164, 185–187 azochromotrope (SPADNS)......................31, 36, 37 buffer for FFE...........................................29, 33, 39 formaldehyde........................................................ 31 glucuronic acid...................................................... 31 HCl...................................................................... 33 HEPES................................................................ 30


3-[[2-hydroxy–1,1-bis(hydroxymethyl)ethyl] amino]–1-propanesulfonic acid (TAPS)......... 32 2-hydroxyisobutyric acid (HIBA)......................... 31 hydroxypropylmethylcellulose (HPMC)..............................................32, 37, 43 imidazole.............................................................. 31 MOPS.................................................................. 31 morpholinoethanol............................................... 32 NaCl..................................................................... 32 NaOH.................................................................. 32 2-(N-morpholino)ethanesulfonic acid (MES)..................................................31, 32, 39 3-(N-morpholino)–2-hydroxypropanesulfonic acid (MOPSO)...................................................... 31 4-pyridineethanesulfonic acid (PES).................... 31 sodium acetate...................................................... 32 sucrose............................................................ 32, 39 TRIS..................................................................... 32 isotachophoresis (ITP)....................................29, 32, 36 zonal electrophoresis...................... 29, 30, 32–34, 37–39 Functional proteomics.................................................. 4, 96

Horseradish peroxidise............................193, 203, 206, 208 H2SO4......................................................................... 32, 65 Hybrid quadrupole orthogonal acceleration time-of-flight (Q-TOF).......................188, 231, 232, 235, 237–238, 240 Hydrogenperoxide.................................................. 123, 184 Hydrophilic interaction chromatography (HILIC)........................................183, 259, 260 3-[[2-hydroxy–1,1-bis(hydroxymethyl)ethyl]amino]–1propanesulfonic acid....................................... 32 2-hydroxyisobutyric acid.................................................. 31 Hydroxypropylmethylcellulose (HPMC).................. 31, 32, 37, 43

I

Gangliosides................................................................... 258 Gas chromatography.............................................. 255, 270 Gelatine.......................................................................... 179 GenBank................................................................ 277, 287 Genomics............................................................... 127, 211 Glioblastoma.................................................................. 130 Gliomas.................................................................. 130–132 Glucoside.......................................................................... 50 Glucuronic acid................................................................ 31 Glutaraldehyde........................................................... 32, 41 Glycerol...............................51, 55, 59, 84, 85, 99, 100, 119, 120, 148, 153, 172, 203, 254, 255, 262 Glycine......................................32, 39, 82, 85, 88, 110, 120, 121, 173, 174, 178, 198 GPS explorer.................................................................. 136 Gradient maker.....................................................83, 85, 86 Guanidinium.................................................................. 199

ICAT.............................................................................. 129 IEF/SDS-PAGE...........................................89, 96, 97, 117 Image analysis......................... 101, 107, 109–110, 211–224 Imidazole...........................................................31, 202, 205 Iminodiacetic acid.......................................................... 182 Immobilized metal affinity chromatography (IMAC).......................................7, 53, 181–195 Immunoblotting........................... 32, 39–41, 82, 83, 85–87, 169, 171, 174, 204, 206 Immunoglobulin....................................................72, 75, 76 Immunoprecipitation......................... 70–73, 75, 76, 78, 129 International Protein Index (IPI)................................... 277 Iodoacetamide (IAA)........................... 50, 65, 99, 104, 105, 111, 165, 166 IPG....................................... 83, 86, 96–100, 102–106, 111, 148, 151, 152, 154, 155, 157 IPG strips............. 96, 99, 100, 103–106, 111, 148, 151, 157 Iso-cyanate..................................................................... 110 ISODalt.................................................................. 100, 106 Isoelectric focusing (IEF)............... 6, 28, 82, 89, 95–97, 99, 102–104, 110, 116–118, 145–179 Isopropanol.......................120, 122, 123, 147, 148, 151, 203 Isotachophoresis (ITP)..............................28–37, 39, 43, 44 iTRAQ ............................................. 7, 127–141, 146–151, 153–157, 275–295

H

K

HB-tag....................................................199, 200, 203, 204 Heatmap..........................................................221, 301, 302 HEPES....................................30, 31, 35, 40, 41, 51, 52, 55, 74, 75, 132, 134 Heptafluorobutyric acid (HFBA)................................... 235 Hippocampus........................ 13, 14, 17–19, 22–24, 76, 266 Hoechst...........................212, 215, 217, 219, 220, 222, 224 Hoefer Ettan ISODalt electrophoresis tank................... 100 Homogenizer...................................... 52, 54, 55, 57, 65, 66, 87, 132, 133, 258, 259 Hormones............................................................... 229, 230

Kinase........................................................49, 130, 181, 287 Krebs-Ringer buffer......................................................... 75

G

L Label-free............................................ 6, 154, 195, 297–308 LC-MS/MS............................6, 7, 128–130, 146, 161, 164, 166, 184, 188–194, 207, 235, 237–239, 244, 245, 249, 251, 275, 276, 279, 281, 283, 286, 289–291, 293, 294 Linear ion trap........................................................ 159–167 Lipidomics.............................................................. 253–270


Liquid chromatography (LC) anion exchange chromatography (SAX)................... 183 capillary reverse phase (RP)..........................6, 133, 135, 137, 160, 164 cation exchange chromatography (SCX).............. 6, 155 hydrophilic interaction chromatography (HILIC)....................................................... 183 nano-LC............................ 129, 148, 153, 161, 245, 251 normal-phase.....................................255, 259–261, 270 Lithium acetate.............................................................. 261 Lowry............................................................................... 58 LTQ-Orbitrap........................................................ 161, 164 Luer-lock adapter................................................... 246–248 Luer-lock plastic syringe........................................ 246, 248 Lysophosphatidylcholine (LysoPC)........256, 258–262, 265

M MADLI MS/MS............................................................... 6 Magnetic beads..................................................73, 119, 182 MALDI matrix...............................................133, 135, 139 MALDI MS/MS....................................129, 137, 154, 188 MALDI-TOF........................ 116, 119, 127–141, 148, 149, 160, 188, 191, 192, 232, 235, 238–239, 281, 282 Mantel-Haenszel test............................................. 306–308 Mascot.....................................136, 140, 190, 193, 207, 232, 236, 239, 276, 281–287, 290 Mascot Daemon..............................................282, 284, 290 Mascot Parser......................................................... 281–285 Mass spectrometry collision induced dissociation (CID)................136, 164, 233, 238, 258, 276, 281 electron transfer dissociation (ETD)........................ 238 electrospray ionization (ESI)............................ 231, 255 ESI-IT mass spectrometer............................... 235, 238 ESI MS/MS......................... 6, 129, 135, 167, 255–257, 259–264, 269 Fourier transform ion cyclotron resonance (FTICR)............................................... 231, 234 hybrid quadrupole orthogonal acceleration time-of-flight (Q-TOF)...............188, 231, 232, 234, 235, 237–238, 240 linear ion trap................................................... 159–167 MALDI MS/MS..............................129, 137, 188, 282 MALDI-TOF.................. 116, 119, 127–141, 148, 149, 160, 188, 191, 192, 232, 234, 235, 238–239, 281, 282 matrix-assisted laser desorption/ionization (MALDI)......................................137, 188, 191 negative-ion mode.....................................258–261, 270 Orbitrap............................................................ 159–167 4800 proteomics analyzer......................................... 135 selected-ion recording (SIR)..................................... 259 Matrigel...........................................................213, 215, 222

Matrix-assisted laser desorption/ionization (MALDI).................... 115, 116, 119, 127– 141, 146, 148, 149, 153–154, 160, 188, 191, 192, 231–235, 238–283 Matrix Science......................... 234, 281, 282, 284, 285, 295 Medial prefrontal cortex (mPFC)........ 14, 17, 20, 21, 24, 25 Medulla...........................................................13, 14, 17, 22 b-Mercaptoethanol......................................................... 177 Methanol...........................85, 101, 119, 121, 123, 148, 151, 174, 176, 214, 233, 236, 256, 258–262, 270 1-methyl–4-phenyl–1,2,3,6-tetrahydropyridine (MPTP).........................................212, 213, 224 MgCl2......................................................................... 51, 74 Microcon.....................................................83, 85, 234, 236 Microcon YM–3............................................................... 85 Mineral oil.............................................................. 148, 152 Minigel system................................................................. 88 Mini-PROTEAN............................... 83, 88, 161, 174, 175 Mitochondria....................................... 6, 28, 30, 31, 34, 35, 37–40, 43, 47, 49, 70, 71, 81, 87–89, 131, 134, 139, 212, 213, 217, 224 Mitotracker..................................... 214, 217, 219, 222, 224 Molecular weight marker...84, 106, 162, 170, 172, 176, 179 Morpholinoethanol.................................................... 31, 32 mRNA.................................................................48, 71, 223 Multiple affinity removal system (MARS).........................................244, 247–248

N NaCl............... 32, 74, 75, 123, 174, 185, 187, 202, 203, 214 NaHCO3................................................... 33, 51, 52, 54, 56 Nano-LC.........................................................148, 153, 161 NaOH..................................31, 32, 187, 205, 214, 246, 248 National Center for Biotechnology Information (NCBI).................. 137, 140, 236, 240, 283, 287 NCBInr............................277, 278, 282, 284, 286, 287, 290 Negative-ion mode..........................................258–262, 270 Nervous system....................................... 3, 63, 64, 159, 181, 197, 199, 201, 253–270 N-ethylmaleimide (NEM)............................................... 50 Neurite.......................................................69, 212, 220, 224 Neuron......................................30, 49, 69, 70, 212, 213, 224 Neuropeptide.......................................................... 229–232 Neuropeptidomics.............................................. 7, 229–241 Neuroplasticity......................................................4, 13, 246 Neuroproteomics...............................................3–8, 14, 159 Neurosome....................................................................... 70 Neurotransmission...........................................3, 63, 71, 115 Neurotransmitter............................. 28, 63, 70, 71, 115, 116 Ni-NTA...........................................................185, 186, 195 Nitrilotriacetic acid (NTA)..................................... 182, 186 Nitrocellulose......................................................... 171, 178 Nitrotetrazolium Blue chloride........................................ 33

N-lauroyl sarcosinate (NLS)............................................ 50 2-(N-morpholino)ethanesulfonic acid.............................. 31 3-(N-morpholino)–2-hydroxypropanesulfonic acid......... 31 N,N’-methylenebisacrylamide........................................ 120 Non-fat dried milk..........................................171, 178, 203 Normal-phase HPLC..............................255, 259–261, 270 NP40.............................................................................. 178 N-tris(hydroxymethyl)methyl-glycine.............................. 82

O Oasis HLB cartridges............................................. 148, 151 OFFGEL............................................................... 145–157 Olfactory bulb.......................................................13, 22, 24 OPTIMA™ L–100 K.................................................... 132 Optiprep..........................................................30, 35, 39, 40 Orbitrap.................................................................. 159–167 Organellar proteomics.................................................. 4, 27 OsO4........................................................................... 33, 41

P Parafilm.............................................. 67, 98, 102, 248, 250 Parkinson’s disease (PD)......................................... 211–224 PDQuest................................................................ 101, 110 Peaks Studio........................................................... 232, 236 Penicillin G............................................................ 184, 185 Peptidomics.................................................................... 230 Perchloric acid................................................................ 123 Percoll..............................................................71, 73, 75, 76 Phalloidin........................................ 214, 217, 219, 222, 224 Phenyx.................................................................... 232, 236 Phosphatase.....................................................171, 181, 182 Phosphatase inhibitor................................ 50, 64, 65, 68, 73 Phosphate buffered saline (PBS)....................32, 41, 74, 77, 78, 147, 178, 185, 202, 204, 213–215, 217 Phosphatidylcholine (PC)..............................218, 253, 256, 258–262, 264, 265, 268, 270 Phosphatidylethanolamine (PE)....................253, 256, 258, 260–263, 266, 269 Phosphatidylinositol (PI)........................258, 260, 262, 267 Phosphatidylserine (PS)......................... 253, 254, 256, 258, 260, 262, 263, 267, 269, 270 Phospholipid lysophosphatidylcholine (LysoPC)...............................256, 258–262, 265 phosphatidylcholine (PC).........................253, 256, 258, 260–262, 264, 270 phosphatidylethanolamine (PE)...............253, 256, 258, 260–263, 269 phosphatidylinositol (PI)...........................258, 260, 262 phosphatidylserine (PS)............................253, 254, 256, 258, 260, 262, 263, 269, 270 plasmalogen phosphatidylethanolamine (pPE).............................................253, 254, 261

Neuroproteomics 315 Index Phospholipidomics................................................. 253–270 Phosphopeptides.................................7, 182–184, 187–195 Phosphoproteins......................................182–185, 191–195 Phosphoproteome...............................................5, 183, 192 Phosphoric acid...............................................101, 120, 190 Phosphorylation dephosphorylation.................................................... 181 immobilized metal affinity chromatography (IMAC).................7, 181, 182, 184, 190, 192–194 kinases...................................................................... 181 phosphatase........................................................ 64, 181 phosphopeptides.................... 7, 182, 184, 190, 192–194 phosphoproteins................................182, 184, 192, 193 phosphoserine........................................................... 190 phosphothreonine..................................................... 190 phosphotyrosine....................................................... 190 TiO2............................................................................. 7 Phosphoserine........................................................ 190, 256 Phosphothreonine.......................................................... 190 Phosphotyrosine............................................................. 190 pH paper......................................................................... 147 p-iodonitrotetrazorium violet (INT)................................ 49 PIPES....................................................................32, 33, 41 Plasmalogen................................................................... 255 Plasmalogen phosphatidylethanolamine (pPE).............................................253, 254, 261 Plasma membrane (PM)...................... 63–68, 81, 116, 118, 130, 131, 134, 139, 246 Plasticity...........................................................5, 13, 70, 71, 182, 197 Polyacrylamide gel.......................... 32, 48, 88, 89, 105–106, 116, 170, 173–176, 179 Polyallomer tubes....................................................... 31, 50 Polyethylene glycol (PEG)......................................... 64, 65 Poly-L-lysine (PLL)........................................213, 215, 223 Polymer............................................................... 63–68, 170 Polyvinylidene fluoride (PVDF) membrane.............. 39, 87, 171, 174, 176, 178 Pons.................................................................13, 14, 17, 22 Post synaptic density (PSD)....................... 5, 28, 32, 39, 43, 47–59, 70, 71, 129 Posttranslational modification (PTM) phosphorylation...................................................... 7, 64 sumoylation.............................................................. 5, 7 ubiquitination............................................5, 7, 198, 201 Potassium phosphate buffer...............................33, 120, 121 Potter-elvehjem.......................................................... 30, 35 Prefrontal cortex................................................... 13, 24–26 Probot..............................................................135, 148, 153 Protease inhibitor................................... 50, 73, 74, 87, 132, 133, 147, 185, 231 Proteasome..............................................197, 199, 201, 204 Protein A................................................. 72, 74, 77, 78, 171


Protein databases EMBL-Bank............................................................ 277 GenBank.......................................................... 277, 287 International Protein Index (IPI)............................. 277 NCBInr......................277, 278, 282, 284, 286, 287, 290 UniProtKB/Swissprot............................................... 277 UniProtKB/TrEMBL.............................................. 277 Protein G.................................................................... 74, 77 ProteinPilot.....................................................149, 153, 157 Protein Prospector.......................................................... 207 Proteins A/G.........................................................73, 75, 78 Proteome............................... 4, 5, 8, 13, 27, 28, 64, 97, 115, 127, 132, 145–157, 159–167, 171, 172, 182, 197–199, 201, 243–251, 267 Proteomics biomarker discovery...............................5, 243–245, 250 expression proteomics............................................... 4–7 functional proteomics........................................4, 28, 96 lipodomics.................................................................... 7 neuropeptidomics......................................................... 7 neuroproteomics........................... 5–8, 14, 96, 110, 111, 127–132, 140, 145, 146, 160, 171, 172, 211 organellar proteomics............................................. 4, 27 phosphoproteome......................................................... 5 proteome................................. 8, 27, 127, 132, 145, 146, 160, 171, 172, 244–245, 250, 267 shotgun proteomics...........................................6, 7, 146 4800 Proteomics analyzer............................................... 135 Proteomics methodology 2D BAC-SDS gel electrophoresis................................ 6 2D IEF-SDS gel electrophoresis.................................. 6 1D-PAGE/LC-ESI MS/MS............................... 6, 244 ESI MS/MS.............................. 129, 135, 261–264, 269 ICAT......................................................................... 129 iTRAQ.......................... 7, 127–132, 140, 146, 275–265 LC-MS/MS............................. 6, 7, 129, 164, 166, 184, 188–194, 207, 244, 245, 249 MADLI MS/MS......................................................... 6 offgel................................................................. 145, 146 stable isotope labeling with amino acids in cell culture (SILAC)................................... 7, 128, 129, 195, 200–202, 204, 281 4-pyridineethanesulfonic acid (PES)................................ 31 Pyronin Y........................................................................ 120

Q Q-TOF.....................188, 231, 232, 234, 235, 237–238, 240

R Raft............................................................................. 49, 50 Rapigest...........................................................132, 134, 139 Razor blade............................................................17, 24, 26 Retinoic acid (RA)..................................213, 214, 216, 224 RGS-His antibody..................................203, 204, 206, 208

Ribonuclease (RNase) inhibitor............................50, 73, 74 Rotofor........................................................................... 146

S Scanner................................................................... 101, 110 S-carboxyamidomethylcysteine...................................... 165 Scissors...................................................... 15, 21, 22, 35, 52 Search engines MASCOT.........................136, 140, 190, 193, 207, 232, 236, 239, 276, 281–287, 290 Peaks Studio..................................................... 232, 236 Phenyx.............................................................. 232, 236 ProteinPilot...............................................149, 153, 157 Protein Prospector.................................................... 207 SEQUEST........................................207, 232, 236, 276 X!Tandem......................................................... 232, 236 Selected-ion recording (SIR).......................................... 259 Sepharose.......... 73–75, 77, 78, 199, 200, 203, 205, 206, 230 Sep-Pak.................................................................. 234, 240 SEQUEST..............................................207, 232, 236, 276 Serva Blue G.............................................................. 84, 89 Shotgun proteomics.................................................6, 7, 146 SH-SY5Y cells........................ 154, 213, 215–217, 222, 223 Signaling peptides (SPs).................. 200, 229, 230, 232, 233 Significance analysis of microarrays (SAM)................... 295 Silver nitrate................................................................... 101 Silver stain.................................. 96, 97, 100–102, 106–109, 117, 122, 123 siRNA.............................................. 213, 214, 218, 223, 224 S-methyl methanethiosulfonate (MMTS)........................................139, 147, 153 Sodium acetate......................................................... 32, 262 Sodium carbonate................................................... 101, 108 Sodium dodecyl sulfate (SDS).................... 6, 33, 39, 41, 82, 86, 88, 89, 99, 100, 104, 106, 110, 118, 120–123, 148, 153, 155, 162, 170, 172–178, 185, 203, 249 sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE)......................... 39, 77, 78, 81–89, 95–100, 106–107, 116, 117, 120–123, 160–163, 166, 167, 171–179, 184–186, 192–194, 244 Sodium succinate.............................................................. 33 Sodiumthiosulphate........................................................ 101 Sonication..........................................................49, 147, 155 Spectral count..........................................165, 244, 297–308 SpeedVac................................. 186, 206, 207, 248, 249, 251 Sphingomyelin (SM)...............................253–255, 258–261 Spin cartridge..................................................246–248, 250 Spin column....................................................185, 186, 195 Spine........................................................................... 49, 70 Stable isotope labeling of amino acids in cell culture (SILAC)................... 7, 128, 129, 195, 200–202, 204, 207, 281 Staphylococcus aureus............................................................72


Streptavidin............................. 199, 200, 203, 205, 206, 208 Striatum............................13, 14, 17, 20, 21, 24–26, 76, 212 Strong cation exchange................ 6, 133, 136, 203, 207, 231 Substantia nigra pars compacta (SNpc).......................... 212 Sucrose..................................30–32, 35, 37, 39, 41–, 47–59, 71, 74, 81, 87, 116, 131–134, 139 Sulfatides (ST)........................................254, 255, 258, 262 Sumoylation................................................................... 5, 7 SV2................................................................................. 116 Swissprot........................................ 137, 140, 277, 278, 282, 284, 286–288, 290 Synapse..................................4–7, 13, 28, 39, 47–59, 70, 71, 95–112, 130, 159–167, 182 Synaptic plasma membrane (SPM)......................47, 48, 52, 55–58, 116 Synaptic vesicles......................................5, 28, 70, 115–123 Synaptoneurosomes.................................................... 69–78 Synaptosome................................ 28, 30–31, 34–37, 39, 43, 47–58, 70, 97, 116, 163, 165, 166 Synuclein........................................................................ 212 Sypro-ruby......................................... 96, 101, 102, 106–109

T TEMED....................................... 85, 86, 99, 100, 105, 120, 121, 173, 178, 184, 186 Thalamus.......................................................................... 14 Thin-layer chromatography (TLC)........................ 255, 261 Thiourea......................................................................... 147 TiO2................................................................................... 7 Trichloroacetic acid (TCA)...................................... 89, 155 Triethylammonium bicarbonate (TEABC)......................................147, 184–186 Trifluoroacetic acid (TFA).........................74, 75, 132–136, 139, 148, 149, 153, 161, 164, 184, 188, 193, 206, 207, 234–238, 240 Tris[2-carboxyethyl] phosphine (TCEP)............... 139, 147 Triton................ 33, 48, 49, 74, 75, 82, 85, 98, 123, 207, 217 Trypsin................ 6, 132, 134, 136, 147, 149–151, 155, 157, 160–161, 164, 165, 184, 186, 192, 193, 198, 201, 203, 206–208, 214, 215, 223, 275, 289 TS2Mascot............................................................. 282, 283

b-tubulin..........................................................217, 219, 220 Tumor.............................................. 127, 130–132, 140, 222 Tween–20............................................ 32, 41, 171, 178, 223 Two dimensional gel electrophoresis 16-BAC-SDS polyacrylamide gel electrophoresis................................. 118, 119 2D fluorescence difference gel electrophoresis......................................... 111 2D gel........................................ 4, 95, 96, 102, 108–112 IEF/SDS-PAGE...................................96, 97, 117, 118 Two-phase system...................................................... 63–68

U Ubiquitin......................................... 197–201, 204, 205, 212 Ubiquitin activating enzyme.......................................... 198 Ubiquitination.................................................5, 7, 197–208 Ubiquitin carboxyl-terminal esterase L1 gene (UCHL1/PARK5)............ 212 Ubiquitin carboxyl-terminal hydrolases (UCHs)........... 198 Ubiquitin conjugating enzymes...................................... 198 Ubiquitin ligases............................................................. 198 Ubiquitin-specific processing proteases.......................... 198 Ultracentrifuge Beckman Optima LE–80K........................................ 31 Beckman 45Ti rotor................................................... 31 Beckman VTi 50 vertical-type rotor........................... 31 OPTIMA™ L–100 K.............................................. 132 polyallomer tubes........................................................ 31 Ultramicrotome................................................................ 33 UniProtKB/Swissprot............................................ 153, 277 UniProtKB/TrEMBL.................................................... 277 Urea.....................................98, 99, 103, 110, 119, 120, 123, 147, 155, 199, 203, 204

X X!Tandem............................................................... 232, 236

Z ZipTip............................................. 147, 151, 188, 234, 240 Zonal electrophoresis........................................................ 28

Neuroproteomics (Neuromethods 57)

General Neurochemical Techniques (Neuromethods) (Neuromethods)

57)

57)

57)

Neuropsychology (Neuromethods)

Amines & Their Metabolites (NEUROMETHODS) (Neuromethods 2 Series I Neurochemistry)

In Vivo Neuromethods

2DArtist Issue 57 September 2010 issue 57

57 Chevy

2000 (57)

Calcium Measurement Methods (Neuromethods)

Patch-Clamp Analysis (Neuromethods)

Cell Culture Techniques (Neuromethods)

Neuropeptide Techniques (Neuromethods)

Electrophysiological Recording Techniques (Neuromethods, 54)

57 Chevy

Carbohydrates and Energy Metabolism (Neuromethods)

Controlled Genetic Manipulations (Neuromethods, v65)

Apoptosis: Techniques and Protocols (Neuromethods)

General Neurochemical Techniques (Neuromethods, 1)

Electrophysiological Recording Techniques (Neuromethods, 54)

Dumbarton Oaks Papers 57

Martin B-57 Canberra

57-мм. самоходное орудие СУ-57. Руководство службы

SF-Stories 57 v3

B-57 Canberra

Цифровой журнал «Компьютерра» № 57

Science Fiction Stories 57

Hiroshima Bugi: Atomu 57

Neuroproteomics (Neuromethods 57)

General Neurochemical Techniques (Neuromethods) (Neuromethods)

57)

57)

57)

Neuropsychology (Neuromethods)

Amines & Their Metabolites (NEUROMETHODS) (Neuromethods 2 Series I Neurochemistry)

In Vivo Neuromethods

2DArtist Issue 57 September 2010 issue 57

57 Chevy

2000 (57)

Calcium Measurement Methods (Neuromethods)

Patch-Clamp Analysis (Neuromethods)

Cell Culture Techniques (Neuromethods)

Neuropeptide Techniques (Neuromethods)

Electrophysiological Recording Techniques (Neuromethods, 54)

57 Chevy

Carbohydrates and Energy Metabolism (Neuromethods)

Controlled Genetic Manipulations (Neuromethods, v65)

Apoptosis: Techniques and Protocols (Neuromethods)

General Neurochemical Techniques (Neuromethods, 1)

Electrophysiological Recording Techniques (Neuromethods, 54)

Dumbarton Oaks Papers 57

Martin B-57 Canberra

57-мм. самоходное орудие СУ-57. Руководство службы

SF-Stories 57 v3

B-57 Canberra

Цифровой журнал «Компьютерра» № 57

Science Fiction Stories 57

Hiroshima Bugi: Atomu 57

Recommend Documents