273 104 6MB
English Pages 373 [388] Year 2011
METHODS
IN
MOLECULAR BIOLOGY™
Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK
For other titles published in this series, go to www.springer.com/series/7651
wwwwwwwwwwwwwwww
Protein Microarray for Disease Analysis Methods and Protocols Edited by
Catherine J. Wu Division of Hematologic Neoplasia, Department of Medical Oncology, Cancer Vaccine Center, Dana-Farber Cancer Institute, Boston, MA, USA
Editor Catherine J. Wu Division of Hematologic Neoplasia Department of Medical Oncology Cancer Vaccine Center Dana-Farber Cancer Institute Boston, MA 02115 USA [email protected]
ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-042-3 e-ISBN 978-1-61779-043-0 DOI 10.1007/978-1-61779-043-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011921931 © 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 Protein microarrays are a rapidly growing segment of proteomics that enable highthroughput discovery-driven research through direct measurement of the molecular endpoints of various physiological and pathological states. The human genome has some 30,000 protein-coding genes, while the human proteome is estimated to have at least 90,000 proteins. By now, protein microarrays have been used for identifying protein– protein interactions, discovering disease biomarkers, identifying DNA-binding specificity by protein variants, and for characterization of the humoral immune response. In this volume, we provide concise descriptions of the methodologies to fabricate microarrays for comprehensive analysis of proteins or the response to proteins that can be used to dissect human disease. These methodologies are the toolbox for revolutionizing drug development and cell-level biochemical understanding of human disease processes. Three general categories of arrays have been developed, which we describe in detail in this volume. The first and most commonly used are the protein-detecting analytical microarrays, described in Part I. Conventionally, the design of these arrays is based on the principle of a sandwich immunoassay. Thus, these capture protein on an array surface from biologic samples and quantify presence of those specific analytes using a detection reagent. Arrays may be coated with antigen-specific antibodies to detect specific proteins from body fluids (Chap. 1), whose identity can be confirmed using label-free detection based on mass spectrometry (Chap. 2). An alternative to detection on solid phase uses newly available bead-based strategies (Chap. 3). Antibody-based detection can be also implemented in a high-throughput fashion on reverse-phase protein arrays. Here, cell lysates are printed to a solid support, followed by quantitative immunodetection, as described in Chap. 4. These general designs have been further modified by other investigators to optimize exploration of specific biologic problems. For example, aptamer (Chap. 5) and recombinant lectin (Chap. 6) arrays have been successfully developed. A second category of protein microarray is antigen microarrays that seek to detect antigen-specific antibody from biologic samples (primarily serum and plasma), covered in Part II. Here, arrays are coated with tens to thousands of proteins in order to detect specific reactive antibodies. These have proven valuable for biomarker discovery and detection. Many possible formats of antigen expression on microarrays are now available. Both commercial high-density protein microarrays that express recombinant protein for serum profiling, as well as technology for custom production of arrays to express a tailored collection of proteins, are now available (Chap. 7). Technology to synthesize comprehensive arrays of peptides has also been established (Chap. 8). Finally, high-throughput protein fractionation strategies have been developed that enable array spotting of antigens in their native format (Chap. 9). Production and isolation of proteins can be cost- and laborintensive. As an alternative, programmable arrays, in which cDNA-containing plasmids are spotted on solid support and protein is freshly translated in situ, offer a versatile solution to the problem of recombinant protein production (Chap. 10).
v
vi
Preface
The final category of protein microarray is protein function microarrays to interrogate direct biochemical and physical interactions among biomolecules (Part III). These include profiling of protein–protein, protein–lipid, protein–DNA/RNA, and small molecule interactions. In Chap. 11, we provide protocols for high-throughput mammalian-based detection of protein–protein interactions, operating on the principle of two-hybrid screening techniques. Programmable arrays have been also developed for this purpose (Chap. 12). Among the many specific applications of protein function arrays are the detection of kinase– substrates interactions (Chap. 13) and the characterization of posttranslational modifications that can serve important regulatory functions in eukaryotic cells (Chap. 14). In most cases, discovery by protein microarray screening requires validation of candidate targets, in order to focus subsequent biologic studies. Part IV of this volume offers two separate approaches to candidate target validation. Both require independent production of the protein analyte to confirm specific reactivity. Both the generation of protein microarrays and the implementation of validation steps have been greatly accelerated by the recent availability of large insect and mammalian proteome libraries. Within these libraries, numerous open reading frames have been cloned and deposited in vector formats that are amenable to protein expression (Part V). The two final sections of the volume are devoted to signal detection strategies (Part VI) as well as data analysis techniques (Part VII). The most conventional and widely used methods are based on fluorometric or colorimetric methods (Chap. 18), while newer label-free detection systems, such as using FRET (Chap. 19) or surface plasmon resonance (SPR) (Chap. 20), will likely be increasingly employed in the future. Validated software for analysis of protein microarrays is only developing now and is obviously critically important for data analysis (Chap. 21). Finally, knowledge of the publicly available databases that are relevant to proteomics studies can enable more efficient data analysis (Chap. 22). We hope that this volume provides a solid framework for understanding how protein microarray technology is developing and how it can be applied to transform our analysis of human disease. I am grateful to all the authors for their outstanding contributions to this edition. Boston, MA
Catherine J. Wu
Acknowledgments I want to thank my family for their support for all my academic endeavors. I want to also acknowledge the excellent assistance from Diana Ng in preparing this volume.
vii
wwwwwwwwwwwwwwww
Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PART I
PROTEIN-DETECTING ANALYTICAL MICROARRAYS
1 Detecting and Quantifying Multiple Proteins in Clinical Samples in High-Throughput Using Antibody Microarrays . . . . . . . . . . . . . . . . . . . . . . . . Tanya Knickerbocker and Gavin MacBeath 2 Analysis of Serum Protein Glycosylation with Antibody–Lectin Microarray for High-Throughput Biomarker Screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chen Li and David M. Lubman 3 Antibody Suspension Bead Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jochen M. Schwenk and Peter Nilsson 4 Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies . . . . . . . . . Víctor J. Cid, Ekkehard Kauffmann, and María Molina 5 Identification and Optimization of DNA Aptamer Binding Regions Using DNA Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicholas O. Fischer and Theodore M. Tarasow 6 Recombinant Lectin Microarrays for Glycomic Analysis. . . . . . . . . . . . . . . . . . . . . Daniel C. Propheter, Ku-Lung Hsu, and Lara K. Mahal
PART II
v xi
3
15 29 37
57 67
ANTIGEN MICROARRAYS FOR IMMUNOPROFILING
7 Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection . . . . . . 81 Persis P. Wadia, Bita Sahaf, and David B. Miklos 8 SPOT Synthesis as a Tool to Study Protein–Protein Interactions . . . . . . . . . . . . . . 105 Dirk F.H. Winkler, Heiko Andresen, and Kai Hilpert 9 Native Antigen Fractionation Protein Microarrays for Biomarker Discovery . . . . . 129 Robert J. Caiazzo, Jr., Dennis J. O’Rourke, Timothy J. Barder, Bryce P. Nelson, and Brian C.-S. Liu 10 Immunoprofiling Using NAPPA Protein Microarrays . . . . . . . . . . . . . . . . . . . . . . 149 Sahar Sibani and Joshua LaBaer
PART III
PROTEIN FUNCTION MICROARRAYS
11 High-Throughput Mammalian Two-Hybrid Screening for Protein–Protein Interactions Using Transfected Cell Arrays (CAPPIA). . . . . . . . . . . . . . . . . . . . . . 165 Andrea Fiebitz and Dominique Vanhecke 12 Protein–Protein Interactions: An Application of Tus-Ter Mediated Protein Microarray System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 Kalavathy Sitaraman and Deb K. Chatterjee 13 Kinase Substrate Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Michael G. Smith, Jason Ptacek, and Michael Snyder
ix
x
Contents
14 A Functional Protein Microarray Approach to Characterizing Posttranslational Modifications on Lysine Residues . . . . . . . . . . . . . . . . . . . . . . . . 213 Jun Seop Jeong, Hee-Sool Rho, and Heng Zhu
PART IV
STRATEGIES FOR VALIDATION OF CANDIDATE TARGETS
15 Multiplexed Detection of Antibodies Using Programmable Bead Arrays . . . . . . . . 227 Karen S. Anderson 16 A Coprecipitation-Based Validation Methodology for Interactions Identified Using Protein Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Ovidiu Marina, Jonathan S. Duke-Cohan, and Catherine J. Wu
PART V
GENERATION OF PROTEOMIC LIBRARIES
17 Development of Expression-Ready Constructs for Generation of Proteomic Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Charles Yu, Kenneth H. Wan, Ann S. Hammonds, Mark Stapleton, Joseph W. Carlson, and Susan E. Celniker
PART VI
DETECTION METHODS
18 Reverse Phase Protein Microarrays: Fluorometric and Colorimetric Detection. . . . 275 Rosa I. Gallagher, Alessandra Silvestri, Emanuel F. Petricoin III, Lance A. Liotta, and Virginia Espina 19 Förster Resonance Energy Transfer Methods for Quantification of Protein–Protein Interactions on Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 Michael Schäferling and Stefan Nagl 20 Label-Free Detection with Surface Plasmon Resonance Imaging . . . . . . . . . . . . . . 321 Christopher Lausted, Zhiyuan Hu, and Leroy Hood
PART VII
DATA ANALYSIS TECHNIQUES FOR PROTEIN FUNCTION MICROARRAYS
21 Data Processing and Analysis for Protein Microarrays . . . . . . . . . . . . . . . . . . . . . . 337 David S. DeLuca, Ovidiu Marina, Surajit Ray, Guang Lan Zhang, Catherine J. Wu, and Vladimir Brusic 22 Database Resources for Proteomics-Based Analysis of Cancer . . . . . . . . . . . . . . . . 349 Guang Lan Zhang, David S. DeLuca, and Vladimir Brusic Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
Contributors KAREN S. ANDERSON • Cancer Vaccine Center, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, USA HEIKO ANDRESEN • Karlsruhe Institute of Technology, Karlsruhe, Germany TIMOTHY J. BARDER • Eprogen, Darien, IL, USA VLADIMIR BRUSIC • Cancer Vaccine Center, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA ROBERT J. CAIAZZO, JR. • Molecular Urology Laboratory, Division of Urology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA JOSEPH W. CARLSON • Department of Genome Dynamics, Lawrence Berkeley National Laboratory, Berkeley, CA, USA SUSAN E. CELNIKER • Department of Genome Dynamics, Lawrence Berkeley National Laboratory, Berkeley, CA, USA DEB K. CHATTERJEE • Protein Expression Laboratory, Advanced Technology Program, SAIC-Frederick, Inc., NCI-Frederick, Frederick, MD, USA VÍCTOR J. CID • Departamento de Microbiología II, Facultad de Farmacia, Universidad Complutense de Madrid, Madrid, Spain DAVID S. DELUCA • Cancer Vaccine Center, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA JONATHAN S. DUKE-COHAN • Immunobiology Laboratory, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA VIRGINIA ESPINA • Center for Applied Proteomics and Molecular Medicine, George Mason University, Manassas, VA, USA ANDREA FIEBITZ • Campus Benjamin Franklin, Charité, Berlin, Germany NICHOLAS O. FISCHER • Physical and Life Sciences Directorate, Lawrence Livermore National Laboratory, Livermore, CA, USA ROSA I. GALLAGHER • George Mason University, Manassas, VA, USA ANN S. HAMMONDS • Department of Genome Dynamics, Lawrence Berkeley National Laboratory, Berkeley, CA, USA KAI HILPERT • Karlsruhe Institute of Technology, Karlsruhe, Germany LEROY HOOD • Institute for Systems Biology, Seattle, WA, USA KU-LUNG HSU • Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, TX, USA ZHIYUAN HU • Institute for Systems Biology, Seattle, WA, USA JUN SEOP JEONG • Department of Pharmacology and Molecular Sciences, High Throughput Biology Center, Johns Hopkins School of Medicine, Baltimore, MD, USA EKKEHARD KAUFFMANN • Zeptosens – A Division of Bayer (Schweiz) AG-, Witterswil, Switzerland TANYA KNICKERBOCKER • Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA JOSHUA LABAER • Virginia G. Piper Center for Personalized Medicine, Biodesign Institute, Arizona State University, Tempe, AZ, USA LANCE A. LIOTTA • George Mason University, Manassas, VA, USA xi
xii
Contributors
CHRISTOPHER LAUSTED • Institute for Systems Biology, Seattle, WA, USA CHEN LI • Department of Chemistry, The University of Michigan, Ann Arbor, MI, USA BRIAN C.-S. LIU • Molecular Urology Laboratory, Division of Urology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA DAVID M. LUBMAN • Department of Chemistry, Comprehensive Cancer Center, The University of Michigan, Ann Arbor, MI, USA; Department of Surgery, The University of Michigan Medical Center, Ann Arbor, MI, USA GAVIN MACBEATH • Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA LARA K. MAHAL • Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, TX, USA; Department of Chemistry, New York University, New York, NY, USA OVIDIU MARINA • Department of Radiation Oncology, William Beaumont Hospital, Royal Oak, MI, USA DAVID B. MIKLOS • Department of Medicine, Blood and Marrow Transplantation Division, Stanford University, Stanford, CA, USA MARÍA MOLINA • Departamento de Microbiología II, Facultad de Farmacia, Universidad Complutense de Madrid, Madrid, Spain STEFAN NAGL • Institute of Analytical Chemistry, University of Leipzig, Leipzig, Germany BRYCE P. NELSON • Gentel Biosciences, Inc., Madison, WI, USA PETER NILSSON • Science for Life Laboratory, Department of Proteomics, School of Biotechnology, KTH – Royal Institute of Technology, 10691 Stockholm, Sweden DENNIS J. O’ROURKE • Molecular Urology Laboratory, Division of Urology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA EMANUEL F. PETRICOIN III • George Mason University, Manassas, VA, USA DANIEL C. PROPHETER • Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, TX, USA JASON PTACEK • The Baxter Laboratory for Stem Cell Biology, Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA, USA SURAJIT RAY • Department of Mathematics and Statistics, Boston University, Boston, MA, USA HEE-SOOL RHO • Department of Pharmacology and Molecular Sciences, High Throughput Biology Center, Johns Hopkins School of Medicine, Baltimore, MD, USA BITA SAHAF • Department of Medicine, Blood and Marrow Transplantation Division, Stanford University, Stanford, CA, USA MICHAEL SCHÄFERLING • Institute of Analytical Chemistry, Chemo- and Biosensors, University of Regensburg, Regensburg, Germany JOCHEN M. SCHWENK • Science for Life Laboratory, Department of Proteomics, School of Biotechnology, KTH – Royal Institute of Technology, 10691 Stockholm, Sweden SAHAR SIBANI • Virginia G. Piper Center for Personalized Medicine, Biodesign Institute, Arizona State University, Tempe, AZ, USA
Contributors
xiii
ALESSANDRA SILVESTRI • George Mason University, Manassas, VA, USA; CRO-IRCCS, National Cancer Institute, Aviano, Italy KALAVATHY SITARAMAN • Protein Expression Laboratory, Advanced Technology Program, SAIC-Frederick, Inc., NCI-Frederick, Frederick, MD, USA MICHAEL G. SMITH • Illumina, Inc., San Diego, CA, USA MICHAEL SNYDER • Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA MARK STAPLETON • NuGEN Technologies, Inc., San Carlos, CA, USA THEODORE M. TARASOW • Tethys Bioscience, Inc., Emeryville, CA, USA DOMINIQUE VANHECKE • Center for Biomedicine, University Basel, Basel, Switzerland PERSIS P. WADIA • Department of Medicine, Blood and Marrow Transplantation Division, Stanford University, Stanford, CA, USA KENNETH H. WAN • Department of Genome Dynamics, Lawrence Berkeley National Laboratory, Berkeley, CA, USA DIRK F.H. WINKLER • Peptide Facility, Kinexus Bioinformatics Corporation, Vancouver, BC, Canada CATHERINE J. WU • Division of Hematologic Neoplasia, Department of Medical Oncology, Cancer Vaccine Center, Dana-Farber Cancer Institute, Boston, MA, USA CHARLES YU • Department of Genome Dynamics, Lawrence Berkeley National Laboratory, Berkeley, CA, USA GUANG LAN ZHANG • Cancer Vaccine Center, Department of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, USA HENG ZHU • Departments of Pharmacology and Molecular Sciences and Oncology, High Throughput Biology Center, Johns Hopkins School of Medicine, Baltimore, MD, USA
wwwwwwwww
Part I Protein-Detecting Analytical Microarrays
wwwwwwwwwwwwwwww
Chapter 1 Detecting and Quantifying Multiple Proteins in Clinical Samples in High-Throughput Using Antibody Microarrays Tanya Knickerbocker and Gavin MacBeath Abstract Many diagnostic and prognostic tests performed in the clinic today rely on the sensitive detection and quantification of a single protein, usually by means of an immunoassay. Even in the case of monogenic diseases, however, single markers are often insufficient to provide highly reliable predictions of disease onset, and the accuracy of these predictions only decreases for polygenic diseases and for very early detection or prediction. Recent studies have shown that predictive reliability increases dramatically when multiple markers are analyzed simultaneously. Antibody microarrays provide a powerful way to quantify the abundance of many different proteins simultaneously in a variety of sample types, including serum, urine, and tissue explants. Because the assay is highly miniaturized, very little sample is required and the assay can be performed in high-throughput. Using antibody microarrays, we have been able to identify prognostic markers of early mortality in patients with end-stage renal disease and have built multivariate models based on these markers. We anticipate that antibody microarrays will prove similarly useful in other discovery-based efforts and may ultimately enjoy routine use in clinical labs. Key words: Antibody microarray, Prognosis, Diagnosis, ELISA, Sandwich immunoassay, Highthroughput
1. Introduction Although some diseases can be accurately diagnosed by detecting a single mutation in a gene or by observing elevated serum levels of a single protein marker, most disease states are much more complex. For example, conditions such as high blood pressure, heart disease, or renal failure have both a genetic and environmental component and even diseases such as cancer, which are largely genetic in origin, are often difficult to diagnose using a simple, univariate test. Several recent studies have shown that the accuracy of cancer diagnoses can be enhanced substantially using multivariate approaches based on gene expression profiles (1–3). In addition, Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_1, © Springer Science+Business Media, LLC 2011
3
4
Knickerbocker and MacBeath
multivariate signatures based on DNA polymorphisms (4) or protein levels (5, 6) are proving useful in predicting how patients respond to targeted therapies. To usher in this era of personalized medicine, we need tools that can accurately, sensitively, and simultaneously measure the levels of many different proteins in a variety of clinical samples (serum, urine, and tissue explants). In addition, to enable the discovery of new diagnostic or prognostic signatures, we need methods that are relatively inexpensive and are compatible with high-throughput investigations. Antibody microarrays offer all of these features. They mimic an enzyme-linked immunosorbant assay (ELISA), but in a miniaturized and multiplexed format (Fig. 1). In a typical antibody microarray experiment, a panel of “capture antibodies” is spotted at high spatial density onto a solid support, typically a chemically derivatized glass substrate (Fig. 1a). A clinical sample (e.g., serum) is then applied to the array, and the immobilized antibodies capture
a
b
Cocktail of detection antibodies
Clinical sample (e.g., serum)
Cytokines
1. Incubate 2. Wash 3. Add detection antibodies
Glass substrate
Capture antibody
Antibody microarray
d
c
4. Incubate 5. Wash 6. Add labeled secondary antibody
7. Incubate 8. Wash 9. Scan for fluorescence
Fig. 1. Detecting and quantifying multiple proteins in clinical samples using antibody microarrays. (a) Capture antibodies are spotted at high spatial density onto a chemically derivatized glass substrate, where they become immobilized. When a clinical sample (e.g., serum) is applied to the array, each immobilized antibody captures its cognate antigen. (b) After a brief washing step, a cocktail of detection antibodies is applied to the array. Each detection antibody recognizes and binds to its cognate antigen. (c) After a brief washing step, the arrays are incubated with a labeled secondary antibody, which recognizes and binds to all of the detection antibodies. For convenience, the secondary antibody is best labeled with a bright fluorophore, such as PBLX-3. (d) After a final washing step, the arrays are dried and scanned for fluorescence.
Detecting and Quantifying Multiple Proteins
5
their cognate antigens. After a brief washing step, the captured proteins are detected by applying a cocktail of “detection antibodies” (Fig. 1b). To visualize and quantify the detection antibodies, the arrays are again washed and probed with a labeled secondary antibody (Fig. 1c). In a standard ELISA, highly sensitive detection is achieved using an enzyme label, such as horseradish peroxidase, which amplifies the signal by catalytically converting a soluble substrate into a chromophoric product. In an antibody microarray experiment, the final signal must be localized to each spot. A variety of strategies have been developed to achieve highly sensitive detection in a spatially localized fashion. For example, the process of rolling circle replication has been exploited to achieve enzymemediated signal amplification (7, 8). This method enables the detection of many proteins at concentrations as low as 1 pg/mL. We have found, however, that equally sensitive detection can be achieved in a more straightforward fashion without enzymemediated signal amplification using a secondary antibody that has been coupled directly to an extremely bright fluorophore (9). (PBXL-3, a phycobilisome protein complex isolated from red algae and cyanobacteria.) The biggest limitation of antibody microarrays, as well as other multiplexed technologies such as the Luminex® bead-based immunoassay, is the availability of suitable antibodies. Sandwichstyle immunoassays require two highly specific antibodies that recognize distinct, nonoverlapping epitopes on their target proteins. For this reason, most studies using antibody microarray technology have focused on cytokines, chemokines, and other frequently studied serum protein for which high quality, matched pairs of antibodies are commercially available (10). To date, antibody microarrays have been used to discover multivariate signatures for diagnostic purposes. For example, antibody microarrays were recently used to detect differential glycosylation patterns on a variety of serum proteins, which may prove useful for the early detection of pancreatic cancer (11). Similarly, antibody microarrays directed at a large panel of cluster of differentiation (CD) antigens on leukemias and lymphomas from peripheral blood and bone marrow aspirates showed high levels of consistency with diagnoses obtained using conventional clinical and laboratory criteria (12). In our own lab, we have used antibody microarrays to identify prognostic markers of early mortality in patients with endstage renal disease (ESRD) (9). This study serves as an example for how antibody microarrays can be used for discovery purposes. Approximately, 10% of patients with ESRD die within the first 3–4 months of initiating hemodialysis and, to date, no single marker has been found that accurately predicts outcome. We set out to develop a multivariate model that predicts which patients
6
Knickerbocker and MacBeath
are most at risk of dying within the first 15 weeks of initiating treatment. To do this, we collected serum samples from 468 patients initiating dialysis (13). We then assembled a panel of 14 matched pairs of antibodies directed at cytokines and other serum proteins that had previously been associated with ESRD, hypertension, or diabetes (14). To facilitate the rapid and accurate measurement of all 14 proteins in all 468 patient samples, we developed a high-throughput assay in which the capture antibodies were microarrayed in individual wells of 96-well microtiter plates (Fig. 2a). Serum samples were applied to each array and the captured cytokines were detected using a cocktail of biotinylated detection antibodies. The detection antibodies were subsequently visualized and quantified using PBXL-3-labeled streptavidin. Using this simple procedure, we were able to achieve exquisite sensitivity: most cytokines could be detected at a concentration of 1 pg/mL (Fig. 2b). The absolute concentration of each cytokine in each sample was determined by relating the fluorescence intensity of the microarray spots to a standard curve, generated for each cytokine in a multiplexed fashion using one column of each microtiter plate (Fig. 2a, b). For redundancy, each array contained five replicate spots of the capture antibodies and every sample was analyzed on two arrays. Overall, the average coefficient of variation was 6.6% for replicate spots within an array and 11% for replicate samples on separate arrays. Using these microarrays, cytokine levels were measured in all 468 patient samples (Fig. 2c). To develop a multivariate prognostic test, we started by building linear, additive models using logistic regression (9). To avoid overfitting and to construct a model that incorporates only as many variables as are necessary, we adopted the following strategy. If n is the number of variables in the model, we started with n = 1 and, in an incremental fashion, performed an exhaustive search for the best n-variable model. We continued to increment n until no n-variable model could be found in which all of the parameters were statistically significant (P < 0.05 for each cytokine). Based on this criterion, the best model was obtained using three cytokines: angiogenin (Ang), interleukin-12 (IL-12), and vascular cell adhesion molecule-1 (VCAM-1). We then refined our efforts by building generalized additive models (15). As anticipated, the nonparametric models picked up fine features in the relationship between death risk and each cytokine (Fig. 2d). We found that high levels of IL-12 and Ang are associated with low risk of early mortality, whereas increased levels of VCAM-1 are associated with increased risk of death. Interestingly, the three molecular markers are produced by and act on different cell populations. This may explain why a simple additive model is sufficient to capture their associations with early mortality.
a
b
4.5
Log10 { fluorescence }
Detecting and Quantifying Multiple Proteins
4.0
Ang EGF Fet-A ICAM IL-12 IL-1a IL-8 MIP-1b RANTES TNFb TNFR2 TNFR1 VCAM-1 VEGF
3.5 3.0 2.5 2.0 1.5 1.0 0.5
7
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Log10 { [cytokine] (pg/mL) }
c Outcome
0 −2 −4 200
500 1000 2000 [Ang] (pg/ml)
1
log-odds of death
2
log-odds of death
d
log-odds of death
Ang EGF Fet-A ICAM IL-12 IL-1a IL-8 MIP-1b RANTES TNF-b TNFR2 TNFR1 VCAM-1 VEGF
0 −1 −2 1 10 50 [IL-12] (pg/ml)
2 0 −2
100 1000 10000 [VCAM-1] (pg/ml)
Fig. 2. Serum cytokine levels measured using antibody microarrays. (a) 14 anticytokine capture antibodies were spotted in quintuplicate in each well of a 96-well microtiter plate. Serum samples were applied to each well in columns 1–11 and twofold serial dilutions of a mixture of the 14 cognate cytokines were applied to the wells in column 12. (b) Standard curves generated from the purified cytokines in column 12 of the microtiter plate. (c) Serum cytokine levels of 468 patients initiating hemodialysis. For visualization only, each cytokine was normalized relative to its mean over all the samples and the patients were ordered according to the first principle component of the cytokine profiles. The outcome of each patient is shown at the top (red died with 15 weeks of initiating dialysis; black survived more than 15 weeks). (d) Model built using the cytokine levels that represent the best three-variable model. The solid red lines are the mean of 100 bootstrap samples and the dashed black lines show the variance.
Cytokines acting on the same cell often exhibit synergistic or antagonistic effects (16), but IL-12, Ang, and VCAM-1 are, to a first approximation, independent. We also found in this study that molecular markers are not uniformly prognostic, but instead vary in their value depending on a combination of clinical variables (age, diastolic blood pressure, serum albumin, and method of vascular access) (9). This may explain why previous reports aiming to
8
Knickerbocker and MacBeath
identify prognostic markers without taking into account clinical variables were either conflicting or showed that markers have marginal prognostic value. Just as treatments are now being tailored to specific subsets of patients, our results show that prognosis can also benefit from a personalized approach. We anticipate that antibody microarray technology will play an increasingly important role in biomarker discovery and may ultimately be used on a routine basis in a clinical setting for the purposes of diagnosis, prognosis, patient selection in clinical trials, and theragnostics.
2. Materials 2.1. Buffers (see Note 1)
1. 10× HBS: 100 mM HEPES, 100 mM NaCl, 0.4% NaN3, pH 7.4. 2. Cy3-BSA: Bovine serum albumin (BSA) can be labeled according to manufacturer’s protocol (Amersham CyDye™ Antibody Labeling Kit, Piscataway, NJ). The Cy3-labeled BSA may be stored at 4°C wrapped in aluminum foil for approximately 3 months. 3. Printing Buffer: 1× HBS, 20% glycerol and 0.005 mg/mL Cy3-labeled BSA. This buffer should be freshly prepared from stock (1) for each print. 4. Dilution Buffer: The dilution buffer should be made so that the final concentration of the solution in each well after the addition of both the dilution buffer and the sample is 10 mM in HEPES, 10 mM in NaCl, 0.04% NaN3, pH 7.4. The actual concentrations of reagents in the dilution buffer will vary with the amount of sample added to each well. 5. Wash buffer I: 1× HBS with 1% BSA (w/v). 6. Wash buffer II: 1× HBS with 0.1% Tween-20.
2.2. Additional Equipment and Reagents
1. Aldehyde-displaying glass substrates (112.5 × 74.5 × 1 mm) (Erie Scientific Company, Portsmouth, NH). 2. A piezoelectric or contact microarrayer. 3. Bottomless 96-well microtiter plates (Greiner BioOne, Kremsmünster, Austria). 4. Silicone gaskets (Grace Bio-Labs, Bend, OR). 5. Streptavidin-conjugated Columbia, MD).
PBXL-3
6. A small-orbit orbital shaker. 7. A microarray plate scanner.
(Martek
Biosciences,
Detecting and Quantifying Multiple Proteins
9
3. Methods 3.1. Printing and Storage of the Microarray Plates
1. Reconstitute all monoclonal capture antibodies according to manufacturer’s instructions and then dilute to a final concentration of 0.5 mg/mL in Printing Buffer. 2. Microarray the antibodies on 112.5 × 74.5 × 1 mm aldehydedisplaying glass substrates using a piezoelectric or contact microarrayer (see Notes 2–6). Ninety-six identical microarrays should be fabricated in a 12 × 8 pattern on the glass, with an interarray pitch of 9 mm (to match the spacing of a 96-well microtiter plate). Each array should consist of a regular pattern of spots, with a center-to-center spacing determined empirically for each arrayer. A pitch of 250–350 mm is typical. Between three and five, spots should be printed for each antibody to provide redundant measurements. 3. Attach the glass to the bottom of a bottomless 96-well microtiter plate using an intervening silicone gasket (see Note 7). 4. Seal the arrays with foil and store at −80°C for at least 4 h, but no longer than 6 weeks (see Notes 8 and 9).
3.2. Preparation of the Mixing Plate
1. Prepare a serial dilution series of recombinant antigen in the first row of a low-binding, 96-well microtiter plate. Use Dilution Buffer as the diluent (see Note 10). (a) For samples with a variable composition and low protein concentration (such as urine), add 15% fetal bovine serum (FBS) to both the standard curve and the samples (see Note 11). (b) For samples such as serum or tissue culture supernatant, FBS may be added to the standard curves to ensure a complex environment similar to that of the samples. (c) Appropriate standard curve concentrations will vary with each antibody, but we typically use a 12-point, twofold serial dilution series ranging from 1 ng/mL of each antigen down to 0.5 pg/mL. For particularly abundant antigens, up to 200 ng/mL may be appropriate although many biologically significant antigens are present at lower concentrations in body fluids and cell culture supernatants. 2. In the remainder of the plate, dilute the clinical samples using Dilution Buffer. (a) For most biological samples, start with a 1:1 or 1:4 dilution, depending on the sample volume available. (b) Additional sample dilution(s) may be required for samples with antigens present at high concentrations. All antigen concentrations should be within the range of the standard curve.
10
Knickerbocker and MacBeath
3.3. Preparation of the Microarray Plate
1. In this section, perform all incubations at 4°C on a small-orbit orbital shaker. The shaker should be set at the maximum possible speed so that it does not cause cross-contamination (~400 rpm). Following each incubation, decant the solution by inverting the plate and shaking by hand. 2. Remove the microarray plate from the −80°C freezer and immediately add 300 mL of Wash Buffer I to each well. Incubate for 5 min (see Note 12) then decant the wash solution by inverting the plate and shaking by hand. 3. Repeat twice for a total of three washes. 4. Add 300 mL of Wash Buffer I to each well and incubate for an additional 1 h to block any remaining aldehydes. 5. Remove Wash Buffer I by decanting; then transfer at least 40 mL from each well of the mixing plate to the corresponding well of the microarray plate. Cover the microarray plate with a foil seal and incubate for up to 24 h (see Note 13). 6. Wash the plate three times for 5 min each with Wash Buffer I. 7. After decanting the final wash, add 40 mL of a mixture of biotinylated detection antibodies (0.5 mg/mL in Wash Buffer I) to each well and incubate for 1 h. 8. Decant the solution then wash the plate three times for 5 min each with Wash Buffer I. 9. Add 100 mL of a 4-mg/ml solution of streptavidin-conjugated PBXL-3, prepared in Wash Buffer I, to each well. Incubate for 1 h in the dark. From this point on, minimize exposure to light. 10. After decanting the PBXL-3 solution, wash the plate two times for 5 min each with Wash Buffer I. 11. Wash the plate once with Wash Buffer II. 12. Rinse the plate twice with ddH2O. 13. Centrifuge upside down for 1 min at 1,000 × g to remove residual water.
3.4. Scanning, Image Analysis, and Data Analysis
1. Scan the microarray plates using a scanner that accommodates microtiter plates (e.g., an LS400 scanner, Tecan, Salzburg, Austria) (see Note 14). 2. Using microarray analysis software, quantify the intensity of each spot. Do not use local background correction. Instead, generate a row of phantom spots within each well and subtract the mean intensity of the phantom spots from each microarray spot. 3. To generate the standard curve, plot the log of the mean fluorescence intensity of replicate microarray spots as a function of the log of the cytokine concentration. This should yield a straight line.
Detecting and Quantifying Multiple Proteins
11
4. Relate the mean intensity of replicate spots for each antibody and each clinical sample back to the standard curve to obtain values for the concentration of each cytokine in each clinical sample. 5. Calculate the mean concentration of replicate measurements of each sample (replicate arrays).
4. Notes 1. Unless otherwise noted, all buffers may be stored at room temperature for up to 1 year or until visible signs of contamination appear. 2. Not all antibodies that work for Western blots and other techniques will work on antibody microarrays. Be sure to validate each pair of antibodies using purified antigens. Mix all antigens together and use detection antibodies one at a time and in combination to ensure that detection antibodies do not cross-react with any of the other analytes under investigation. R & D Systems (Minneapolis, MN) is an excellent source of matched pairs of antibodies and their cognate antigens, particularly for the study of cytokines and chemokines. 3. In general, monoclonal antibodies are used as the capture antibody while biotinylated polyclonal antibodies are used for detection. If no monoclonal antibody is available, two polyclonal antibodies may be used. Ideally, these two polyclonal antibodies will have been raised against distinct and nonoverlapping epitopes. 4. When preparing the source plate, mix the antibody in Printing Buffer in an eppendorf tube and then transfer it to the source plate (microtiter plate). This ensures adequate mixing of the solution and increases reproducibility between wells when multiple pins are used to print the same sample. 5. In general, we have found aldehyde-displaying glass surfaces to be more robust and reproducible than epoxide- or aminedisplaying glass, nitrocellulose-coated glass, or hydragels. 6. Pay close attention to the liquid level in the source plate while fabricating microarrays. Even when using a 384-well microtiter plate as a source plate, substantial evaporation can occur during extended print runs. To minimize evaporation, use a cooling block set to between 4 and 10°C, if available, and set the relative humidity at 70–80%. If these options are not available, use an aluminum foil seal with a small hole over each well to allow tip/pin access. Check the liquid level after each print run (or more often, if necessary) and add ddH2O
12
Knickerbocker and MacBeath
as needed to ensure that the antibody concentration remains constant throughout the print run(s). 7. After printing or after assembly of the microarray plate, the arrays may sit at room temperature for several hours without appreciable loss of antibody reactivity. 8. To protect the arrays from freezer burn, they should be sealed, either in a plastic bag or with a foil cover. If using a cover, be sure that it will remain sealed when stored at −80°C. If using a cover that projects into the wells (such as a rubber Storage Mat), attach the cover to the bottomless 96-well microtiter plate before attaching the glass substrate. This ensures that the glass substrate is not pushed off the silicone gasket when the cover is applied to the wells (due to positive pressure). 9. Arrays can be stored for up to 1 year at −80°C, although some loss in activity will occur. Plates that have been stored for several months are best used for assay development purposes. For data collection, plates should be stored for no longer than 6 weeks. 10. For samples with low concentrations of total protein (e.g., urine), preblock the mixing plate with BSA to minimize protein loss. To do this, add enough Dilution Buffer containing 1% BSA (w/v) to completely fill each well, incubate for 1 h at room temperature, and then decant the blocking solution. For samples with very low protein content, all plastics should be rinsed with Dilution Buffer containing 1% BSA (w/v) before contacting the samples. 11. For samples with low concentrations of total protein, add 15% FBS (v/v) to the samples to minimize loss of target proteins. Test each new pair of antibodies to ensure than they do not cross-react with bovine proteins. 12. When removing antibody microarrays from the −80°C freezer, be sure to add Blocking Buffer immediately (within seconds). The buffer usually freezes when added to the wells and then thaws within minutes. Allowing the arrays to warm up, even slightly, results in poor spot morphology (“comet tails,” “coffee rings,” etc.). 13. The best length of time to incubate the antibody microarrays with the samples should be determined empirically. It depends on the concentration of antigens in the sample, the affinities of the capture antibodies for their antigens, and the efficiency of agitation. Incubation times generally range from 1 to 24 h. 14. Antibody microarrays should be scanned at several scanner settings (PMT voltages). For each antigen, use the scan with the highest possible setting that does not include any saturated pixels.
Detecting and Quantifying Multiple Proteins
13
Acknowledgment We thank Ravi Thadhani for directing ArMORR (Accelerated Mortality on Renal Replacement), a prospective study of ESRD patients, and Jiunn-Ren Chen for data analysis and interpretation. This work was supported by awards from the WM Keck Foundation and the Arnold and Mabel Beckman Foundation, and by grants from the National Institutes of Health (DK071674 and DK068465). T.K. is the recipient of an Eli Lilly Graduate Student Fellowship. References 1. Alizadeh AA, Eisen MB, Davis RE et al (2000) Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403:503–511 2. Liang Y, Diehn M, Watson N et al (2005) Gene expression profiling reveals molecularly and clinically distinct subtypes of glioblastoma multiforme. Proc Natl Acad Sci U S A 102:5814–5819 3. Ramaswamy S, Tamayo P, Rifkin R et al (2001) Multiclass cancer diagnosis using tumor gene expression signatures. Proc Natl Acad Sci U S A 98:15149–15154 4. Zhou SF, Di YM, Chan E et al (2008) Clinical pharmacogenetics and potential application in personalized medicine. Curr Drug Metab 9:738–784 5. Duffy MJ, Crown J (2008) A personalized approach to cancer treatment: how biomarkers can help. Clin Chem 54:1770–1779 6. Hanash S (2003) Disease proteomics. Nature 422:226–232 7. Schweitzer B, Roberts S, Grimwade B et al (2002) Multiplexed protein profiling on microarrays by rolling-circle amplification. Nat Biotechnol 20:359–365 8. Shao W, Zhou Z, Laroche I et al (2003) Optimization of rolling-circle amplified protein microarrays for multiplexed protein profiling. J Biomed Biotechnol 5:299–307
9. Knickerbocker T, Chen JR, Thadhani R, MacBeath G (2007) An integrated approach to prognosis using protein microarrays and nonparametric methods. Mol Syst Biol 3(123):1–8 10. MacBeath G (2002) Protein microarrays and proteomics. Nat Genet 32:526–532 11. Li C, Simeone DM, Brenner DE et al (2009) Pancreatic cancer serum detection using a lectin/glyco-antibody array method. J Proteome Res 8:483–492 12. Belov L, Mulligan SP, Barber N et al (2006) Analysis of human leukaemias and lymphomas using extensive immunophenotypes from an antibody microarray. Br J Haematol 135: 184–197 13. Thadhani R, Tonelli M (2006) Cohort studies: marching forward. Clin J Am Soc Nephrol 1:1117–1123 14. USRSD (2005) National Institutes of Health. National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda 15. Buja A, Hastie T, Tibshirani R (1989) Linear smoothers and additive models (with discussion). Ann Statist 17:453–555 16. Natarajan M, Lin KM, Hsueh RC et al (2006) A global analysis of cross-talk in a mammalian cellular signalling network. Nat Cell Biol 8: 571–580
wwwwwwwwwwwwwwww
Chapter 2 Analysis of Serum Protein Glycosylation with Antibody–Lectin Microarray for High-Throughput Biomarker Screening Chen Li and David M. Lubman Abstract The complexity of carbohydrate structures and their derivatives makes the study of the glycome a challenging subset of proteomic research. The microarray platform has become an essential tool to characterize glycan structure and to study glycosylation-related biological interactions, by using probes as a means to interrogate the spotted or captured glycosylated molecules on the arrays. The highthroughput and reproducible nature of microarray platforms have been highlighted by their extensive applications in the field of biomarker validation, where a large number of samples must be analyzed multiple times. This chapter presents an antibody–lectin microarray approach, which allows the efficient, multiplexed study of the glycosylation of multiple individual proteins from complex mixtures with both fluorescence labeling detection and label-free detection based on mass spectrometry. Key words: Microarray, Antibody, Glycoprotein, Biomarker, Serum, Lectin, MALDI, Mass spectrometry
1. Introduction Glycosylation is the most commonly occurring posttranslational modification on proteins involved in numerous biological processes, such as protein–protein interactions, protein folding, immune recognition, cell adhesion, and intercellular signaling. The function of glycoproteins is highly dependent on their carbohydrate structure. The alteration on the glycans is associated with multiple biological events and has been reported in a variety of diseases, especially cancer (1–4). In the search for effective glycosylated biomarkers for targeted diseases, there has been a great deal of effort invested in profiling and characterization of
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_2, © Springer Science+Business Media, LLC 2011
15
16
Li and Lubman
glycoproteins in complex samples. Cell lines, tissue, and other types of biofluids have been studied by mass spectrometry, fractionation techniques, and microarrays (5–9). Although a microarray assay does not usually provide in-depth structural information on the glycans compared to mass spectrometry, it is able to identify and quantify numerous glycosylation patterns and simultaneously analyze hundreds of samples in a high-throughput manner with excellent reproducibility (9–13). We herein describe an antibody–glycoprotein sandwich assay for high-throughput glycoprotein biomarker screening, where a fluorescent lectin and MALDI-MS are used to quantitatively measure glycosylation levels and identify analytes captured on the antibody arrays, respectively. The scheme for this procedure is illustrated in Fig. 1. The antibodies are first printed on nitrocellulose coated glass slides to generate identical arrays. Printed slides are processed to chemically block the glycans on the antibodies, which are otherwise reactive with lectins used for detection (10). After properly diluted human serum samples are deposited onto the separated antibody arrays, the captured antigens are probed with different lectins with a wide spectrum of binding specificity. The binding of the lectin is measured with a secondary fluorescent dye through a biotin–streptavidin reaction. To verify the effectiveness of previously discovered glycoprotein biomarkers, hundreds of serum samples collected from patients with different disease states are examined in parallel with healthy controls for altered glycosylation patterns. The technical error and bias in the analysis is minimized in several ways, including introducing a control slide to assess spatial variation on a slide and balancing samples from different groups on each slide to reduce experimental bias. MALDI-MS detection has only recently been used to detect peptides on a modified gold surface
Fig. 1. Experimental scheme using lectin and MALDI-MS detection with antibody microarray to analyze glycosylation of serum glycoproteins.
Analysis of Serum Protein Glycosylation
17
coated with antibodies (14). An on-slide digestion method, developed in our previous work (15), exploited the utility of MALDI-MS to identify antibody-captured proteins. The whole digestion, including automatic trypsin spotting and incubation, requires less than 10 min. While the antibody–lectin sandwich microarray provides a means to measure glycosylation changes on specific proteins captured from complex samples using lectin probes in a high-throughput array format, fluorescence-based detection provides limited structural information and cannot distinguish some glycoforms that have similar affinity with lectins, such as (GlcNAc)2(Man)8 and (GlcNAc)2(Man)9. Therefore, the MALDI-MS detection of the tryptic products of the captured protein on the antibody array serves as a complementary technique to verify the identity of the target of the antibody and a means to monitor the nonspecific binding so as to optimize the dilution fold for the experiment. As such, mass spectrometry is a powerful alternative to fluorescent detection, as it confirms the identity of the captured analyte and detects any undesired binding.
2. Materials 2.1. Antibody–Lectin Microarray with Fluorescence Detection
1. Monoclonal antibodies, for serum amyloid P component (SAP; Abcam), Alpha-1-beta glycoprotein (A1BG; Abnova), Antithrombin III (Abcam).
2.1.1. Printing
3. Nova nitrocellulose slides (GraceBio), PATH nitrocellulose slides (Gentel).
2. Nanoplotter 2.0 (GeSiM).
4. Printing buffer: 30% phosphate buffering saline (PBS), concentration of antibody diluted by water to 0.3 mg/mL. 5. 96-Well sample plate (BioRad). 2.1.2. Antibody Blocking
1. Washing buffer: PBS-T 0.1 (0.1% Tween-20). 2. Coupling buffer: 0.02 M sodium acetate, pH 5.5. 3. Oxidation buffer: 0.2 M sodium peroidate in coupling buffer. 4. 4-(4-N-maleimidophenyl)butyric acid hydrazide hydrochloride (MPBH) (Pierce), Cys–Gly (Sigma).
2.1.3. Hybridization of Slides
1. Blocking buffer: 1% w/v BSA in PBS-T 0.5 (0.5% Tween-20). 2. Sample buffer: 0.1% Brij-45, 0.1% Tween-20 in PBS. 3. Primary detection solution: For Aleuria aurentia (AAL), Maackia amurensis (MAL), Lens culinaris agglutinin (LCA),
18
Li and Lubman
and Sambuccus Nigra (SNA) – 10 mg/mL biotinylated lectin solution in PBS-T 0.1; and for Concanavalin A (ConA), 1 mg/mL lectin in PBS-T 0.1. All biotinylated lectins were purchased from Vector Laboratories (Burlingame, CA). 4. Washing buffer: PBS-T 0.1 (0.1% Tween-20). 5. Secondary detection solution: 1:1,000 solution of 1 mg/mL Streptavidin conjugated to Alexafluor555 (Invitrogen) in PBS-T 0.1. 6. Speedvac. 7. SIMplex Multiplexing system (Gentel). 2.1.4. Slide Scanning
1. Axon 4000A scanner (Molecular Devices, Sunnyvale, CA).
2.1.5. On-Slide Digestion
1. Sequencing grade modified trypsin (Sigma). 2. Acetonitrile (ACN). 3. Ammonia bicarbonate. 4. Oven. 5. Nanoplotter 2.0 (GeSiM). 6. Wetted paper box.
2.1.6. MALDI-QIT-TOF
1. MALDI-QIT-TOF (Shimadzu Biotech, Manchester, UK). 2. Trifluoroacetic acid (TFA). 3. 2,5-Dihydroxybenzoic acid (DHB), prepare 10 mg/mL solution in 50% ACN, 0.1% TFA. 4. Stainless steel plate adaptor.
3. Methods 3.1. Antibody Array Printing
The number of antibody arrays that can be printed on each slide is determined by the size of the arrays. The most popular format involves 16 coated pads on a standard 1 × 3 in. slide. Each pad is able to contain more than 9 × 9 spots with 0.6 mm spacing. For MALDI-MS detection, the sensitivity is much lower than fluorescence. Therefore to generate a spectrum with good S/N, additional sample needs to be printed on each spot. 1. Antibodies are diluted to 0.5 mg/mL in printing buffer and transferred to a 96-well sample plate. 2. Edit the spot layout in the NanoPlotter program to produce a 2 × 7 format of identical arrays with a 9 mm row and column distance from each other. The spacing between the spots is 0.6 mm. Each antibody is printed in triplicate. For MALDI-MS detection, the spacing between the spots is 1.5 mm.
Analysis of Serum Protein Glycosylation
19
3. Antibody solution is spotted onto nine ultrathin nitrocellulose coated slides. The first slide is discarded because of high variation of printing; the other eight are used for the experiment. Each spotting event results in 500 pL of sample being deposited and is programmed to occur 5 times/spot to ensure that 2.5 nL is being spotted per sample. The spot diameter is around 250 mm. For MALDI-MS detection, the amount of antibody on each of the spots in the antibody array is increased from 5 to 100 droplets. The spot diameter is around 700 mm (see Note 1). 3.2. Antibody–Lectin Array with Fluorescence Detection 3.2.1. Antibody Array Blocking
The IgG antibodies are usually glycosylated (15). The antibody glycans are reactive to detection lectins, thus need to be modified. To prevent the reaction between the antibody glycan and lectin, the antibodies on the slides are chemically derivatized with a modified method described in the previous work of Haab (10). 1. The printed slides are dried at room temperature overnight before gently being washed with PBS-T 0.1 and incubated in coupling buffer with 0.1% Tween 20 for 10 min. The slides are washed again with coupling buffer without Tween 20 before oxidation (see Note 2). 2. The slides are incubated in freshly made oxidation solution at 4°C in the dark. After 3 h the slides are removed from the oxidizing solution and rinsed with coupling buffer with 0.1% Tween 20 until the white precipitation disappears. The washing usually takes 30–60 min (see Note 3). 3. The slides are immersed in fresh 1 mM MPBH (in coupling buffer) at room temperature for 2 h to derivatize the carbonyl groups, then incubated with 1 mM Cys-Gly (in PBS-T 0.1) at 4°C overnight to stabilize the −SH group on MPBH. The slides are subsequently blocked with blocking buffer for 1 h and dried by spinning the slide at 1,000 rpm in a centrifuge (see Note 4).
3.2.2. Optimizing Conditions
Before screening a large number of samples, the optimum concentration of the serum is determined by a serial dilution test. In the dilution test, serum is diluted with sample buffer by 2–600 folds and incubated with different blocks of antibody arrays on a single slide (details of the experiment in Subheading 3.2.4). The signal is detected by lectin SNA (or any lectin) and plotted in Fig. 2. The figure depicts how the intensity of the signal changes for three antibodies (against Serum Amyloid P component, A1BG, and Antithrombin III) with decreasing dilution fold. A rising trend was noted from the 600× dilution to the 50× dilution for the three glycoproteins shown. In the 50× dilution to the 20× dilution, the signal was relatively unchanged
20
Li and Lubman
Fig. 2. Saturation curve showing how the antibodies (against serum Amyloid C component, A1BG, Antithrombin III) respond to different dilution of serum with SNA lectin detection. X-axis shows fold serum dilution before hybridization on the antibody array. The y-axis is the intensity of the signal. Reprinted with permission from Li et al. (15).
except for Antithrombin III, where the signal increased 20% from the 50× dilution to the 20×. The signal remained the same from the 20× dilution until it reached the 5× dilution, where a saturation of the signal has occurred. A decrease of signal for all three glycoproteins from the 5× dilution to the 2× dilution of serum sample can be seen in Fig. 3, likely due to competing nonspecific binding on the antibodies. The result of the dilution test demonstrates that the antibodies were saturated by their target protein at 20× dilution or above in the process of hybridization. Below 50× dilution, the antibodies were not completely occupied so the signal decreased with additional dilution. The nonlinear relation between the concentration of the serum and the intensity of the signal could be attributed to various factors that may affect the antibody–antigen reaction, including accessibility of the antibodies, diffusion rate, and solubility of the antigen in the hybridization buffer. Nonspecific binding on the antibodies was also considered as a possibility, but was further investigated and excluded by on-target digestion and MALDI-MS analysis. To analyze the difference of the glycosylation on potential biomarker proteins, protein expression levels must be normalized. Under saturation conditions, the amount of target biomarkers captured on the antibody spots was equal to the capacity of the printed antibody which should be the same in all the replicate blocks. As a result, protein assay is no longer needed and the intensity of the signal on the microarray directly represents the level of glycosylation.
Analysis of Serum Protein Glycosylation
21
a G1 G2 G4 G1 G3 G4 G2 G3 C1 C1 G1 G2 G4 G1 G3 G4 G3 G4 G1 G2 C1 C2 G3 G4 G1 G2 G3 B Slide 1
G2 G3 G4 G1 C1 C2 G2 G3 G4 G1 G2 B Slide 2
G1 G2 G3 G4 C1 C2 G1 G2 G3 G4 G1 B Slide 3
G4 G1 G2 G3 C1 C2 G4 G1 G2 G3 G4 B Slide 4
C1 C1 C1 C1 C1 C1 C1 C1 C1 C1 C1 C1 Slide 5
G3 G4 G1 G2 C1 C2 G3 G4 G1 G2 G3 B Slide 6
G2 G3 G4 G1 C1 C2 G2 G3 G4 G1 G2 B Slide 7
G1 G2 G3 G4 C1 C2 G1 G2 G3 G4 G1 B Slide 8
b
Fig. 3. Parallel processing of 77 samples on eight slides. (a) Sample arrangement on eight slides. G1, G2, G3, and G4 are four different groups of samples. Control samples are C1 and C2. B is blank. (b) A picture of SIMplex multiwell device.
3.2.3. Experimental Design
In the high-throughput biomarker screening, we usually parallel print and process eight slides which contain 112 identical blocks of antibody array. To minimize the technical error and bias on these blocks, serum samples are arranged to balance different disease/healthy groups and reference blocks are also introduced to adjust to signals of different blocks and slides. We provide an example of how to arrange samples on slides to minimize experimental biases in Fig. 3. 1. The slides are labeled from 1 to 8 in their printing order (see Note 2). 2. Slide 5 is used as a control slide; all the blocks on the control slide are incubated with a control serum sample C1. 3. Block 7 and block 8 on each slide except slide 5 are used as control blocks; they are incubated with control samples C1 and C2, respectively. 4. Block 14 is used as blank and incubated with sample buffer only. 5. The other 77 blocks are incubated with 19 samples from each of the four disease groups and 1 extra sample from a random group in a designated order to balance the number of samples from each group on any particular block (Fig. 3a).
22
Li and Lubman
3.2.4. Hybridization of Slides
1. The slides are placed into the SIMplex (Gentel) Multiplexing device which has 16 wells for each slide (the bottom two wells are not used) to separate the antibody arrays and prevent cross contamination between adjacent wells (Fig. 3b). 2. Serum samples are aliquoted into a volume of 10 mL in each vial and diluted 10× with 90 mL sample buffer. Diluted samples are added into the wells of the SIMplex Multiplexing device and incubated for 1 h with gentle shaking at room temperature. The wells must be sealed to prevent evaporation of samples (see Notes 5 and 6). 3. After completion of serum hybridization, slides are rinsed with PBS-T 0.1 three times to remove unbound proteins. The slides are incubated with biotinylated lectin solution in a plastic box with gentle shaking for an hour at room temperature. 4. The slides are washed 3 times with PBS-T 0.1 and incubated with secondary detection solution with gentle shaking for an hour at room temperature. 5. The slides are again washed 3 times with PBS-T 0.1, dried by centrifuge and kept at 4°C before scanning.
3.2.5. Slide Scanning
1. The dried slides are scanned with an Axon 4000A scanner. 2. Alexa555 labeled slides are scanned in the green channel (wavelength 545 nm). The photomultiplier tube (PMT) gain should be adjusted to obtain the best S/N without saturation. The size of the pixel of the image is 10 mm. 3. The program Genepix Pro 6.0 is used to extract the numerical data.
3.2.6. Data Analysis
The nonbiological variation between blocks on the same slide is termed as on-slide variation. This variation is mainly generated by antibody printing and slide scanning and its feature is that every slide follows the same pattern (i.e., the blocks at the top of the slides are brighter than the bottom ones). The blocks on the control slide incubated with the same control sample are thus used to estimate the on-slide variation and calculate adjustment index for all the blocks. The slide-to-slide variation is considered as specific changes of the signal that effect all the blocks on a single slide. This variation is estimated by control blocks on each of the slides. The data is adjusted by a second index calculated by comparing the signal of the control blocks to exclude the slide-to-slide variation. An example of assaying glycosylation expression of A1BG is shown in Fig. 4. Lectin SNA is used to probe the sialic acid present at the termini of the glycans of this protein. As shown in this figure, the mean value of the cancer samples is significantly higher than the other three groups (p < 0.05).
Analysis of Serum Protein Glycosylation
23
Fig. 4. Distribution of sialylation levels detected by lectin SNA on A1BG. The spots present the signal of the glycan on captured antigen for individual samples from different classes. The long and short lines give the mean value and the standard error of the mean, respectively.
1. A threshold of signal-to-background ratio is set at 3 and spots that are under this threshold are excluded. 2. The background-subtracted median of the intensity for the triplicates of each antibody is averaged and taken as a single data point into analysis. 3. On-slide variation index for antibody 1 in block 1 equals to the average signal of antibody 1 over all the blocks on the control slide divided by the signal of antibody 1 in block 1. I Ab1.B1 = Avg Ab1.CS / SAb1.B1 . 4. Slide-to-slide variation index for antibody 1 on slide 1 is calculated as follows: AvgAb1.S1 is the average signal of antibody 1 on slide 1. AvgAb1.AS is the average signal of antibody 1 on all the slides. I Ab1.S1 S = Avg Ab1.A / Avg Ab1.S1 . 5. The final adjusted signal is calculated by the following formula: SAb1.B1.S1 is the raw signal, SAb1.B1.S1.ad is the adjusted signal. SAb1.B1.S1.ad = SAb1.B1.S1 * I Ab1.B1 * I Ab1.S1 . 6. For each antibody the signal can be normalized to one for easy comparison, SAb1.B1.S1.n is the normalized signal. SAb1.B1.S1.n = SAb1.B1.S1.ad /Avg Ab1.
24
Li and Lubman
3.3. On-Slide Digestion and MALDI Sample Preparation
Nonspecific binding on antibodies may occur when the microarray is exposed to a concentrated and complex protein mixture such as serum. A commonly used method to study the specificity of an antibody is to digest and identify the protein released from antibody-conjugated medium, whereas eluting the captured protein is not very efficient and the procedure includes four or more steps. Thin layers of a conductive metal oxide and nitrocellulose make the surface of PATH slide perfect for MALDI. We developed an on-slide digestion and MALDI sample preparation protocol using the NanoPlotter to precisely spot enzyme and matrix to antibody arrays on the slide after the serum hybridization. Antibody arrays exposed to differently diluted sera are analyzed by this method to see if nonspecific binding occurs. Trypsin spotted on the antibody array usually simultaneously digests both the captured protein and the antibody; hence the tryptic peptides of the antibody must be excluded from the mass spectra for us to choose the peaks of interest. In an example, we prepared three identical spots of SAP antibody in separated blocks, which were then incubated with sample buffer (as control), 10× diluted serum, and 2× diluted serum and subjected to on-slide digestion and MALDI-MS. The MALDI-MS spectra of the three spots are shown in Fig. 5. The peaks that appear in the spectrum of the control spot are considered to be peptides of the antibody. The three highest peaks between 1,150 and 1,250 were identified by MS/MS as peptides from the Fc region of mouse IgG. In spectrum b where the antibody spot was hybridized with 10× diluted serum, the peaks at 1,166 and 1,407 m/z, are identified by MS/MS as the peptides digested from the target antigen, and the peak 993 matches the mass of a tryptic peptide of SAP. In the spectrum c there are two additional peaks. One of these was identified as human albumin, while the other one could not be identified or matched with a peptide mass of the target antigen. The additional peaks indicate that nonspecific binding might have occurred to the antibody spot. The serum was further diluted to assess the detection limit of the MALDI-MS technique. At 500× dilution (data not shown), the peak at 1,166 m/z disappeared while the 1,407 m/z still showed a signal-to-noise ratio of 2–3. Thus, the 500× dilution is considered as the detection limit of SAP, which is present in human serum with a concentration of around 30 mg/mL (16). The introduction of mass spectrometry based label-free detection has the potential to further characterize the glycan structure. However, due to the presence of the tryptic peptides of the antibody and the lack of a glycopeptide enrichment step, only a limited number of the nonglycosylated peptides of the antigen could be seen in the spectra. To improve the MALDI-MS detection of the targeted antigen and its glycopeptides, we are searching for other chemical strategies to block the tryptic digestion of the antibody and enrichment methods to selectively ionize the glycopeptides.
Analysis of Serum Protein Glycosylation
25
Fig. 5. The MALDI-MS spectra generated on the microarray spots of Amyloid p component antibody after on-target digestion. The peaks identified as Amyloid p component were marked with bold arrows where the extra peaks appearing in (c) were marked with regular arrows. (a) Control spot, without incubation of serum; (b) incubated with10× diluted serum; (c) incubated with 2× diluted serum. Reprinted with permission from Li et al. (15).
1. Antibody slide is printed and hybridized with diluted serum as described above. 2. Trypsin is diluted with 50 mM ammonium bicarbonate in 20% ACN and kept on ice before use. 3. Keep the humidity of the Nanoplotter chamber higher than 70% (use a humidifier or lay a wet paper towel on the deck).
26
Li and Lubman
4. In the program, set the same spot layout on the slide, print 100 droplets (0.5 nL per droplet) of trypsin on each spot (see Note 7). 5. Move the printed slide to a wet paper box and incubate them in an oven at 37°C for 5 min. Make sure the trypsin solution does not dry out on the spots. 6. Take the slide out from the oven, print the DHB solution on the slide with the same spot layout (50 droplets per spot). 3.4. MALDI-MS
1. Tape the slide onto a stainless steel MALDI plate adaptor, insert it into the MALDI-MS instrument. 2. Mass spectrometric analysis of the microarray slides was performed using the Axima quadrupole ion trap-TOF. Acquisition and data processing were controlled by Launchpad software (Kratos, Manchester, UK). A pulsed N2 laser light (337 nm) with a pulse rate of 5 Hz was used for ionization. Each profile resulted from two laser shots. Argon was used as the collision gas for CID and helium was used for cooling the trapped ions. 3. TOF was externally calibrated using 500 fmol/mL of bradykinin fragment 1–7 (757.40 m/z), angiotensin II (1046.54 m/z), P14R (1533.86 m/z), and ACTH (2465.20 m/z) (SigmaAldrich). The mass accuracy of the measurement under these conditions was 50 ppm. 4. The power of the laser is set at 80 to ionize the spots on the microarray. The focus of the laser can be moved from spot to spot manually under the camera or by using the Raster function to set up an automatic scan for all the spots.
4. Notes 1. When the pin on the Nanoplotter is in poor condition or the instrument is not set up correctly, the quality of antibody printing may fluctuate or gradually worsen as the printing continues. Sticky components, such as glycerol, in the antibody printing solution may also cause unstable printing. A simple test can be done in advance to assess the performance of the pin. Print 1,000 spots with a random antibody on a transparent slide. Observe the residue after the spots are dried. If the residues are in an intact round shape and their sizes and colors do not vary significantly, then the printing is acceptable, otherwise the printer needs to be checked or the printing solution must be changed.
Analysis of Serum Protein Glycosylation
27
2. Many types of chemicals can contaminate the nitrocellulose coating on the slide, resulting in increased background. The slides should not be labeled with any kind of marker. A disposable plastic box is a very good container for slide washing. 3. In the glycan blocking procedure, after the antibody is oxidized by NaIO4, white precipitation forms on the slides. This precipitation must be completely washed away before moving on to the next step. 4. Blocked slides should not be kept in solution for too long, while dried ones can be stored at 4°C for a long period of time. 5. Serum sample must be aliquoted immediately upon arrival and stored at −80°C. Serum frozen and thawed more than twice should not be used. When the sample set consists of multiple groups, all the samples must be in the same frozen and thaw cycle for bias-free comparison. 6. All the incubation should be done with gentle shaking to prevent uneven binding. 7. The higher number of droplets of antibody solution printed on the slides does not result in a higher density of antibody on the spot because the coating of the PATH slides is so thin that a few droplets are able to saturate the surface. The concern for the minimum amount of antibody solution printed on each spot is position variation, i.e., repeated printings on the same spot do not perfectly overlap. Printing 100 droplets of antibody solution produces a larger spot size which guarantees a certain area of overlap between the antibody spot and the printing of trypsin and matrix.
Acknowledgements Our work on microarray development described herein has been supported in part under grants from the National Cancer Institute under grant NCI R21 12441, R01 CA106402. This work has also received partial support from the National Institutes of Health under R01GM49500. We would like to thank Dr. Brian Haab and Dr. Chen Songming of the Van Andel Institute for sharing with us the procedures of preparing the antibody arrays. We would also like to thank Stephanie Laurinec, Jes Pedroza, and Missy Tuck for collection of the samples used in this work.
28
Li and Lubman
References 1. Rudd PM, Elliott T, Cresswell P, Wilson IA, Dwek RA (2001) Glycosylation and the immune system. Science 291:2370–2376 2. Kobata A, Amano J (2005) Altered glycosylation of proteins produced by malignant cells, and application for the diagnosis and immunotherapy of tumours. Immunol Cell Biol 83:429–439 3. Gessner P, Riedl S, Quentmaier A et al (1993) (1993) Enhanced activity of cmp-newac-galbeta-1–4glcnac-alpha-2, 6-sialyltransferase in metastasizing human colorectal tumor-tissue and serum of tumor patients. Cancer Lett 75:143–149 4. Gorelik E, Galili U, Raz A (2001) On the role of cell surface carbohydrates and their binding proteins (lectins) in tumor metastasis. Cancer Metastasis Rev 20:245–277 5. Zhao J, Simeone DM, Heidt D, Anderson MA, Lubman DM (2006) Comparative serum glycoproteomics using lectin selected sialic acid glycoproteins with mass spectrometric analysis: application to pancreatic cancer serum. J Proteome Res 5:1792–1802 6. Ressom HW, Varghese RS, Goldman L et al (2008) Analysis of MALDI-TOF mass spectrometry data for discovery of peptide and glycan biomarkers of hepatocellular carcinoma. J Proteome Res 7:603–610 7. An HJ, Peavy TR, Hedrick JL et al (2003) (2003) Determination of N-glycosylation sites and site heterogeneity in glycoproteins. Anal Chem 75:5628–5637 8. Block TM, Comunale MA, Lowman M et al (2005) Use of targeted glycoproteomics to identify serum glycoproteins that correlate with liver cancer in woodchucks and humans. Proc Natl Acad Sci U S A 102:779–784
9. Patwa TH, Zhao J, Anderson MA, Simone DM et al (2006) Screening of glycosylation patterns in serum using natural glycoprotein microarrays and multi-lectin fluorescence detection. Anal Chem 78:6411–6421 10. Chen SM, LaRoche T, Hamelinck D et al (2007) Multiplexed analysis of glycan variation on native proteins captured by antibody microarrays. Nat Methods 5:437–444 11. Zhao J, Patwa TH, Qiu WL et al (2007) Glycoprotein microarray with multi-lectin detection: unique lectin binding patterns as tools for classifying normal, chronic pancreatitis, and pancreatic cancer sera. J Proteome Res 5:1864–1874 12. Wu YM, Nowack DD, Omenn GS et al (2008) Mucin glycosylation is altered by pro-inflammatory signaling in pancreatic-cancer cells. Pancreas 37:502 13. Yue TT, Goldstein IJ, Hollingsworth MA et al (2009) The prevalence and nature of glycan alterations on specific proteins in pancreatic cancer patients revealed using antibody-lectin sandwich arrays. Mol Cell Proteomics 7:1697–1707 14. Evans-Nguyen KM, Tao SC, Zhu H et al (2008) Protein arrays on patterned porous gold substrates interrogated with mass spectrometry: detection of peptides in plasma. Anal Chem 5:1448–1458 15. Li C, Simeone DM, Brenner DE et al (2009) Pancreatic cancer serum detection using a lectin/glyco-antibody array method. J Proteome Res 8:483–492 16. Nyboa M, Olsenb H, Jeuneb B et al (1998) Increased plasma concentration of serum amyloid P component in centenarians with impaired cognitive performance. Dement Geriatr Cogn 9:126–129
Chapter 3 Antibody Suspension Bead Arrays Jochen M. Schwenk and Peter Nilsson Abstract Alongside the increasing availability of affinity reagents, antibody microarrays have been developed to become a powerful tool to screen for target proteins in complex samples. Besides multiplexed sandwich immunoassays, the application of directly applying labeled sample onto arrays with immobilized capture reagents offers an approach to facilitate a systematic, high-throughput analysis of body fluids such as serum or plasma. An alternative to commonly used planar arrays has become available in form of a system based on color-coded beads for the creation of antibody arrays in suspension. The assay procedure offers an uncomplicated option to screen larger numbers of serum or plasma samples with variable sets of capture reagents. In addition, the established procedure of whole sample biotinylation circumvents the purification steps, which are generally required to remove excess labeling substance. We have shown that this assay system allows detecting proteins down into lower pico-molar and higher picogram per milliliter levels with dynamic ranges over three orders of magnitude. Presently, this workflow enables the profiling of 384 clinical samples for up to 100 proteins per assay. Key words: Suspension bead array, Antibody array, Serum, Plasma, Labeling
1. Introduction The exploration of the human proteome is one of the major challenges of the postgenomics era, focusing on a better understanding of disease-related processes (1). Recent developments of miniaturized and parallelized technology platforms now offer affinity-based alternatives to widely used mass spectrometric analysis. Among these methods, various protein microarrays have been implemented into proteomic profiling approaches demonstrating their applicability in high-throughput screening for marker proteins in patient samples (2). Two alternative formats have been developed; reverse-phase microarrays, where large numbers of lysates from tissues and cells or serum samples are spotted onto array surfaces for the parallel analysis of a single
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_3, © Springer Science+Business Media, LLC 2011
29
30
Schwenk and Nilsson
parameter, and the forward-phase setting, such as multiplexed sandwich immunoassays or antibody arrays, which both utilize immobilized capture reagents to analyze many parameters (3). While dedicated robotic devices, which arrange molecules on microscopic slides with functionalized surfaces, are needed produce planar protein microarrays, alternative platforms have been employed for a parallelized and miniaturized analysis. One of these is based on a flow cytometeric system that currently allows to determine the identity of up to 500 color-coded micrometer sized beads in cooccurrence to protein interaction dependent reporter fluorescence (4). Arrays are thereby created in suspension by mixing beads with different codes, denoted here as bead IDs, and immobilized capturing reagents. This platform has recently been utilized to adapt the concept of antibody arrays from previously described planar arrays (5). The described workflow, summarized in Fig. 1, offers a microtiter plate-based alternative to methods based on planar microarrays for the analysis of labeled serum and plasma protein profiling and can be used for highly multiplexing in both the dimension of parameters measured per sample as well as samples studied per analysis. An example of a protein profile obtained from this approach is given in Fig. 2. Here, intensity levels over more that two orders of magnitude and a low intensity variability of £20% are observed.
2. Materials 2.1. Bead Coupling
1. Beads: MagPlex or MicroPlex microspheres (Luminex Corp). 2. Activation buffer (1×): 100 mM Monobasic Sodium Phosphate (Sigma), pH 6.2, stored at +4°C for up to 3 months and at −20°C for long term. 3. EDC solution: 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC, Pierce), aliquoted in screwcapped tubes and stored at +4°C. Dissolve in activation buffer to 50 mg/ml directly prior usage. 4. S-NHS solution: 50 mg/ml Sulfo-N-Hydroxysuccinimide (NHS, Pierce), prepared as aliquots in screw-capped tubes and stored at −20°C. Dissolve in activation buffer to final concentration directly prior usage. 5. Coupling buffer: 100 mM 2-(N-morpholino)ethanesulfonic acid (MES) pH 5.0, stored at +4°C for up to 3 months and at −20°C for long term. 6. Wash buffer: 0.05% (v/v) Tween20 in 1× PBS pH 7.4 (PBST). 7. Antibody detection solution: 0.25 mg/ml R-Phycoerythrin modified antispecies antibodies (e.g., Jackson), diluted to this concentration in PBST (see Note 1).
Antibody Suspension Bead Arrays
31
Fig. 1. Workflow overview.
2.2. Sample Labeling
1. Sample dilution buffer: 1× PBS pH 7.4. 2. Labeling solution: 10 mg/ml Sulfo-N-Hydroxysuccinimidepolyethylene oxide biotin (NHS-PEO4-Biotin, Pierce), dissolved in dimethyl sulfoxide (DMSO, Sigma) directly before use. 3. Stop solution: 1 M Tris–HCl pH 8.0, stored at +4°C and added cold.
2.3. Assay Procedure
1. Assay buffer (1×): 0.1% (w/v) casein, 0.5% (w/v) polyvinylalcohol, and 0.8% (w/v) polyvinylpyrrolidone (all Sigma), prepared in PBST and stored at +4°C for up to 3 months and at −20°C for the long term. Supplement before use with 0.5 mg/ml rabbit IgG (Bethyl).
32
Schwenk and Nilsson
MFI [AU]
10000 1000 100 10
-65 Ab
-60 Ab
-55 Ab
-50 Ab
-45 Ab
Ab
-40
-35 Ab
-30 Ab
-25 Ab
-20 Ab
-15 Ab
-10 Ab
-05 Ab
Ab
-01
1
Antibodies
Fig. 2. Intensity profile of a plasma sample. A bead mixture composed of 68 antibodies was employed to determine intensity levels for the targeted proteins in a plasma sample. Such profiles typically cover intensity range over more than two orders of magnitude (50–20,000 AU). Standard deviations of £20% can be commonly obtained from replicates.
2. Stop solution (4×): 4% paraformaldehyde (PFA) solution, to store at +4°C. Dilute 1:4 in PBS prior to usage. 3. Detection solution: R-Phycoerythrin modified streptavidin (Invitrogen) diluted to 0.5 mg/ml in PBST directly before use and protected from light.
3. Methods 3.1. Bead Coupling
In the following, a method for antibody coupling is described, for which magnetic and nonmagnetic beads can be utilized. The main difference between these two bead types is the handling of the beads during an exchange of surrounding liquid solution. For coupling quantities not exceeding the amount of positions found in bench top microcentrifuges, we suggest using microcentrifuge tubes or tubes with filter inserts to pellet the beads via centrifugation, while magnetic beads can additionally be manipulated by magnetic forces without centrifugation. For more than 24 couplings in parallel, microtiter plate based protocols are preferred. Hereby, proteins can be immobilized on nonmagnetic beads in filter bottomed microtiter plates (Millipore) with a filter pore sizes below bead diameter and vacuum devices (Millipore) accommodate these plates to remove liquid. For magnetic bead coupling in plates, dedicated plate magnets are available (LifeSept, Dexter Magnetic Technologies) to facilitate bead sedimentation and fixation. 1. Prepare antibodies at the desired concentration (e.g., 3 mg or a solution with antibody concentration of 30 mg/ml per 1 × 106 beads) in coupling buffer (see Note 2).
Antibody Suspension Bead Arrays
33
2. The beads are to be distributed in desired portions (e.g., 80 ml = 1 × 106 beads) into the wells of a half-area plate and the beads are washed with 3× 100 ml activation buffer (see Note 3). 3. Prepare fresh solutions of NHS and EDC, both at 50 mg/ml in activation buffer. Prepare 0.5 mg of each chemical per bead ID and coupling, and add 10 ml NHS, 10 ml EDC, and 80 ml activation buffer to each bead ID. 4. Incubate 20 min under continuous, gentle shaking, and wash thereafter with 3× 100 ml coupling buffer. 5. Continue without interruption (see Note 4) by adding the antibody solution to the activated beads and incubate for 2 h under continuous, gentle shaking. 6. The beads are washed 3× with 100 ml wash buffer. 7. The beads are then recovered from the wells into microcentrifuge tubes with 3× 100 ml wash buffer. The liquid is removed and 100 ml storage buffer is added prior to the bead storage at +4°C in the dark for at least 1 h. 3.2. Bead Mixture Preparation
The yield of antibodies immobilized on beads should be judged after the coupling. To allow a balanced and economic amount of beads to be applied and counted during the measurements, equal numbers of beads should be combined in a bead mixture. To facilitate this, the beads can be counted and an initial bead concentration can be determined which allows calculating the required volumes to be added in a common stock solution. During this bead counting procedure, the rate of antibody immobilization can be additionally approximated via fluorescently labeled antispecies specific antibodies. 1. The tubes with antibody-coupled beads are to be vortexed and sonicated for 5 min. 2. Each bead solution is diluted 1/100 in antibody detection solution (see Notes 1 and 5) in a microtiter plate. 3. The plates are incubated for 20 min and measured. 4. The number of counts per bead ID is multiplied by a correction factor of 3.3 for a 1/100 dilution to obtain a first estimation of beads per microliter storage solution. From this number the volumes of beads in storage solution can be calculated which are to be applied into the bead mixture. The required number of beads supplied should be adjusted for each assay procedure and be based on the quantity of beads being counted by the instruments. We suggest to always obtain ³32 counts per bead ID. 5. After each measurement and for the preparation of new bead mixtures, the count average is to be calculated for each bead ID and new volumes can be determined. We suggest adjusting
34
Schwenk and Nilsson
these volumes to a theoretical bead count, which is 20% above the estimate: For 100 beads to be counted from the new bead mixture, the previously obtained volumes should be calculated for 120 beads per assay and bead ID. 3.3. Sample Labeling
1. The serum or plasma samples are to be thawed according to the preferred protocol (see Note 6). 2. The samples are vortexed and centrifuged for 10 min at 10,000 × g to pellet insoluble components. 3. A previously designed plate layout, in which samples should be located randomly, is followed by transfer of 30 ml of serum/ plasma into the respective wells of a PCR plate, which is then sealed and centrifuged for 2 min at 1,500 × g. 4. As an option, the samples are incubated for 30 min at elevated temperatures such as 56°C (see Note 7) followed by 15 min at 20°C using in a thermo cycler. Using the heated lid function of the cycler helps to prevent the samples to evaporate into the lid/seal. 5. Transfer 3 ml into a second PCR plate containing 27 ml PBS, seal the plate, vortex, and centrifuge for 2 min at 1,500 × g. 6. Add 2.5 ml of NHS-Biotin to each well (see Note 8), then seal the plate, vortex and centrifuge for 2 min at 1,500 × g, and incubate for 2 h at 4°C under continuous shaking in a microtiter plate mixer. 7. Add 25 ml of 1 M Tris–HCl pH 8.0 to each well, seal the plate, vortex, and centrifuge for 2 min at 1,500 × g. 8. Store the plates at −20°C until usage or use directly.
3.4. Assay Procedure
1. The labeled samples are thawed and diluted 1/50 in assay buffer, which had been prepared in a PCR plate. Seal the plate, vortex, and centrifuge for 2 min at 1,500 × g. 2. The samples incubated for 30 min. As an option, the samples are treated for 30 min at elevated temperatures such as 56°C (see Note 7), followed by 15 min at 20°C using the heated lid function of the thermo cycler. Thereafter, the plate is vortexed and centrifuged for 2 min at 1,500 × g. 3. The previously prepared bead mixture is distributed into the wells of a half-area plate and protected from light. Then 45 ml of the diluted, labeled samples are added to the wells (see Note 5) and incubated at 23°C over night under continuous shaking on a microtiter plate mixer. 4. The plates are then washed 3× with 75 ml wash buffer, incubated with stop solution for 10 min and washed 1× with 75 ml wash buffer again.
Antibody Suspension Bead Arrays
35
5. R-PE labeled streptavidin is then added to each well at 0.5 mg/ml and 30 ml and the plates are incubated for 20 min under continuous shaking. 6. The plates are then finally washed 3× with 75 ml wash buffer and 100 ml of wash buffer are added before the plates are measured with the Luminex instrumentation. 7. Set the instrumentation setting according to the bead IDs included in the mixture and count at least 50 beads per bead ID. We suggest using the “median fluorescence intensity” to further process your data. An example of a plasma protein profile is shown in Fig. 2.
4. Notes 1. Other fluorescent dyes than R-Phycoerythrin such as Alexa546, Alexa532, or Cy3 can be utilized as well, but Luminex Corp. has indicated that lower reporter signal intensities are to be observed. 2. Employ solutions of purified proteins and avoid other stabilizing proteins, Tris or other amine-based buffers as they reduce the coupling efficiency. 3. At all times, try to minimize the light exposure, especially to direct sunlight, as the internal fluorescence of the beads as well as reporter fluorophores could be bleached. During incubation, protect the plates with an opaque cover or place plate into a light-tight box. 4. Do not interrupt the activation process after dissolving EDC and NHS, as these active substances are susceptible to hydrolysis resulting in a loss in activity. 5. When combining beads with solutions for counting and assay procedure, always distribute small volume bead solution (e.g., 5 ml) into the well first, then add larger volume buffer portion (e.g., 45 ml) to allow an instant distribution of the beads. 6. We have found that thawing overnight at +4°C was most practical if a larger number of samples were to be processed. Otherwise, place tube(s) into a 42°C water bath until a minor fraction of ice was still visible. 7. We have observed that heat treatment of labeled samples in combination with the applied multiplexed assay procedure affects antibody performance (5). This can lead to improved protein detectability by changing the accessibility of the epitopes in a complex sample solution but should be tested and balanced with the tendency of proteins to precipitate at higher temperatures.
36
Schwenk and Nilsson
8. Do not interrupt the experimental flow after dissolving NHS-Biotin, as this active substance is susceptible to hydrolysis resulting in a loss in activity. Add NHS-Biotin to the side of each well using single- or multichannel dispensers so that the labeling reactions for all samples are started contemporaneously.
Acknowledgments We like to thank the entire staff of the Human Proteome Resource (HPR) initiative for their tremendous efforts within the Human Protein Atlas project. This work is supported by the PRONOVA project (VINNOVA, Swedish Governmental Agency for Innovation Systems), and by grants from the Knut and Alice Wallenberg Foundation. References 1. Hanash S (2003) Disease proteomics. Nature 422:226–232 2. Templin MF, Stoll D, Schwenk JM, Pötz O, Kramer S, Joos TO (2003) Protein microarrays: promising tools for proteomic research. Proteomics 3:2155–2166 3. Kingsmore SF (2006) Multiplexed protein measurement: technologies and applications of protein and antibody arrays. Nat Rev Drug Discov 5:310–320
4. Fulton RJ, McDade RL, Smith PL, Kienker LJ, Kettman JR Jr (1997) Advanced multiplexed analysis with the FlowMetrix system. Clin Chem 43:1749–1756 5. Schwenk JM, Gry M, Rimini R, Uhlen M, Nilsson P (2008) Antibody suspension bead arrays within serum proteomics. J Proteome Res 7:3168–3179
Chapter 4 Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies Víctor J. Cid, Ekkehard Kauffmann, and María Molina Abstract Infection of cells and tissues by pathogenic microorganisms often involves severe reprogramming of host cell signaling. Typically, invasive microorganisms manipulate host cellular pathways seeking advantage for replication and survival within the host, or to evade the immune response. Understanding such subversion of the host cell by intracellular pathogens at a molecular level is the key to possible preventive and therapeutic interventions on infectious diseases. Reverse Protein Arrays (RPAs) have been exploited in other fields, especially in molecular oncology. However, this technology has not been applied yet to the study of infectious diseases. Coupling classic in vitro infection techniques used by cellular microbiologists to proteomic approaches such as RPA analysis should provide a wealth of information about how host cell pathways are manipulated by pathogens. The increasing availability of antibodies specific for phosphorylated epitopes in signaling proteins allows monitoring global changes in phosphorylation through the infection process by utilizing RPA analyses. In our lab, we have shown the potential of RPA technology in this field by devising a microarray consisting of lysates from cell cultures infected by Salmonella typhimurium. The protocols used are described here. Key words: Reverse protein arrays, Cell lysate arrays, In vitro infection, Virulence factors, Salmonella, Type III secretion, Host cell signaling, Protein kinase, Protein phosphorylation
1. Introduction Reverse protein arrays (RPAs) technology is an antibody-based proteomic approach based on high-throughput dot blots performed on cell lysates printed to a solid support, followed by quantitative immunodetection. RPA assays have been used to monitor cell signaling in different contexts, especially in the field of oncology (1–4). Nevertheless, its application to the study of host–pathogen interactions, as proposed here, has not been exploited to date. Obligate intracellular parasites such as viruses
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_4, © Springer Science+Business Media, LLC 2011
37
38
Cid, Kauffmann, and Molina
and some bacteria (typically chlamydiae and rickettsiae), as well as facultative intracellular pathogens, such as invasive bacteria and fungi, are able to subvert host cell signaling to promote their internalization into target cells, their intracellular survival, or their replication (5). Interaction of infectious agents with host cell pathways is also responsible for cell and tissue damage. Some pathogens have been also described to modulate signaling events aimed to evade the immune response. The pathogens achieve these goals by expressing specific virulence factors that directly interfere with the function of cell signaling proteins. One paradigmatic example is the injection into the host cell cytoplasm of bacterial proteins (“effectors”) by specialized secretory systems (for example, type III and type IV secretion systems) (6–8). This is a common mechanism for both plant and animal bacterial pathogens. Such effectors are regulators of cytoskeletal components, such as actin and tubulin, regulators of GTPases, kinases, or phosphatases of proteins or phosphoinositides, or regulators of ubiquitin ligases (9, 10). Thus, they are able to reprogram the host cell to generate a comfortable environment for their colonization. RPA technology opens the possibility of analyzing by immunoblotting the presence of particular proteins or their posttranslational modifications, such as phosphorylation, in cell lines after exposure to a particular pathogen in a given variety of experimental conditions. To set up the working conditions and demonstrate the potential of this technique, we have recently used RPAs to assess the involvement of Salmonella typhimurium type III secretion system (T3SS) effectors (11). These proteins are translocated from the bacterium to the host-cell cytoplasm by the T3SS encoded by the Salmonella pathogenicity island I (SPI-1), which is specifically involved in remodeling actin and promoting bacterial internalization during infection (12). A general scheme on the experimental design followed in our approach, which could be applied to other host–pathogen systems, is presented in Fig. 1. Specifically, we infected in vitro HeLa cells, widely used as a model
Fig. 1. General scheme of RPA hybridization coupled to in vitro infection as a tool to monitor signaling events through host–pathogen interaction. Alternative possibilities for experimental design are noted. The desired pathogen is grown in a variety of experimental conditions, as desired. The choice of different mutants or isolates will provide information on the contribution of mutated factors or particular isolates to discrete signaling events in the host model. Choice of the host cell line for in vitro infection (epithelial cells vs. lymphoid cell lines; primary vs. immortal cultures) should also be determined by the nature of the biological question under study. At the desired times after infection, control and problem cells are detached from culture plates and a collection of lysates is prepared, calibrated, and four different dilutions of each lysate are spotted in duplicate in a format of six arrays per chip. Zeptosens ZeptoMARK chips and the corresponding analytical technology were used in our assays (see Note 12). Hybridization of these chips with a collection of prevalidated antibodies specific to particular proteins or their posttranslational modifications (i.e., phosphorylation) considered as markers for the activation of signaling pathways, and subsequent quantification of the data, will reveal a wealth of information on how such pathways respond to contact with the pathogen.
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies EXPERIMENTAL DESIGN - Mutants in virulence-related genes - Different clinical isolates - Various infection times - Different infection doses - Pre-treatment with antibiotics etc.
Pathogen strains
in vitro infection EXPERIMENTAL DESIGN - Immortal lines - Primary cultures - siRNA directed silencing etc.
Host cell line
Lysate collection representing all infection conditions Array printing Hybridization with a collection of validated antibodies Antibody 1 Antibody 2
... Antibody n
Comparative quantitative readout of - protein levels - postranslational modifications (phosphorylation...) Data analysis
39
40
Cid, Kauffmann, and Molina
for bacterial infection on epithelial cells (13), with either wild-type Salmonella or different mutants impaired in particular aspects of T3SS function, including a mutant unable to assemble the T3SS (invA), as well as combinations of particular mutations in wellknown T3SS effectors, such as SopE and SopB. SopE and its paralog SopE2 are activators of the small GTPase Cdc42 that elicits signals aimed to recruit actin to the area of bacterial contact to promote membrane ruffling and eventual internalization of the bacterium (14, 15). Besides, SopE and SopE2 have been involved in the activation of mitogen-activated protein kinase (MAPK)mediated signaling pathways in the host cell (16). SopB, also known as SigD, is a phosphatidylinositol-phosphate phosphatase involved in different steps of the formation of the Salmonellacontaining vacuole (17–19) that also interacts with Cdc42 (20) and specifically triggers activation of protein kinase B (Akt) in the host cell (21). The method detailed here proved useful to confirm previously described signaling events that depend on SopB and SopE effectors, to detect novel changes in phosphorylation, and to assess the contribution of those particular bacterial effectors to such changes (11). We believe that the same approach could be used for other Salmonella effectors (more than 30 T3SS-secreted proteins have been detected, many of them of yet unknown function), other invasive bacteria (Shigella, Listeria, Mycobacterium, Chlamydia, etc.) or fungi (Candida, Cryptococcus), or even noninvasive pathogens that are known to severely subvert host cell signaling (enteropathogenic E. coli, etc.).
2. Materials 2.1. Preparation of Lysates 2.1.1. In Vitro Infection
1. HeLa cells (human adenocarcinoma cervix epithelial cell line, obtained from the American Type Culture Collection CCL-2, Manassas, VA) (see Note 1). 2. Salmonella strain SL1344 (22) and genetically manipulated derivatives (11). 3. Growth medium for HeLa: RPMI 1640 (Biological Industries, Israel) supplemented with 10% fetal calf serum, penicillin (100 units/mL), streptomycin (100 mg/mL), and 2 mM L-glutamine. 4. Growth medium for bacteria: Luria Bertani (LB) broth: 10 g/L bacto tryptone, 5 g/L bacto yeast extract, and 10 g/L NaCl, sterilized by autoclaving. If plasmid maintenance was required, LB was supplemented with 12 mg/mL chloramphenicol. 5. P100 plates (10 cm tissue culture dishes, Cellstar).
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies 2.1.2. Cell Lysis
41
1. Phosphate-buffered saline (PBS) (Fluka). 2. Cell lysis buffer CLB1 (Zeptosens, Switzerland) (see Note 2).
2.2. Sample Preparation for Spotting 2.2.1. Determination of Protein Concentration in Lysates
1. Cell lysis buffer CLB1 (Zeptosens, Switzerland). 2. Phosphate-buffered saline (Fluka). 3. Coomassie Plus – The Better Bradford assay reagent (Thermo Fisher). 4. Bovine serum albumin (BSA) ampoules 2 mg/mL (Thermo Fisher). 5. 96-Well microtiter plates, flat bottom, suitable for optical readout (Greiner bio-one, Germany). 6. Microtiter plate reader suitable for absorbance measurement at 595 nm (e.g., SpectraMax Plus, Molecular Devices, CA). 7. Low volume liquid handling robot, 96 channel (e.g., Zephyr, Twister II, Caliper Life Sciences, MA).
2.2.2. Preparation of Spotting Microplate
1. ZeptoMARK CSBL1 – Spotting Buffer (Zeptosens, Switzerland). 2. ZeptoMARK CLB1 Lysis Buffer (Zeptosens, Switzerland). 3. Dilution buffer: 1 part CLB1, 9 parts CSBL1. 4. Fluorophore conjugated with albumin from bovine serum (BSA) [e.g., Cy5 (GE Healthcare)]. 5. Alexa Fluor 647 (Invitrogen), HiLyte Fluor 647 (Anaspec), or Dylight 649 (Thermo Fisher). 6. ZeptoMARK Spotting Buffer for References CSBR1, (Zeptosens, Switzerland). 7. ZeptoMARK Reference Dilution Buffer RDB1 (Zeptosens, Switzerland). 8. PBS tablets (Fluka). 9. Sodium azide (NaN3) (Sigma). 10. Microcentrifuge filtration unit, 1.5 mL tube 0.22 mm (e.g., Millipore UFC30GV00). 11. Low volume liquid handling robot, 96 channel (e.g., Zephyr in combination with Twister II, Caliper Life Sciences, MA).
2.3. Microarray Spotting
1. Noncontact microarray spotter, modified NanoPlotter (Zeptosens/GeSiM, Switzerland).
2.4. Chip Blocking
1. ZeptoFOG ultrasonic nebulizer blocking station (Zeptosens, Switzerland). 2. ZeptoCHIP blocking racks, including lids and handles (Zeptosens, Switzerland). 3. ZeptoMARK blocking buffer BB1 (Zeptosens, Switzerland) (see Note 3).
42
Cid, Kauffmann, and Molina
2.5. Reverse Array Assay
1. ZeptoCARRIER chip holder with fluidic cells (Zeptosens, Switzerland). 2. Eight-channel aspiration tool (e.g., VacuSafe, IBS Integra Biosciences, Switzerland). 3. Specific primary antibodies, directed against the proteins and posttranslationally modified proteins of interest (see Subheading 3.8). 4. Antispecies secondary antibodies (full IgGs or Fab fragments), carrying fluorescence labels in the red spectral range, e.g., Cy5 (GE Healthcare), Alexa Fluor 647 (Invitrogen), HiLyte Fluor 647 (Anaspec), or Dylight 649 (Thermo Fisher).
2.6. Readout and Data Analysis
1. ZeptoREADER planar waveguide fluorescence reader for ZeptoMARK chips, assembled in microplate-footprint ZeptoCARRIERs (Zeptosens, Switzerland). 2. ZeptoVIEW 3 reverse array analysis software (Zeptosens, Switzerland).
3. Methods The choice of a particular set of different infection conditions will depend on the experimental problem to tackle. Thus, RPA experimental design is the key to obtain a maximum amount of valuable and comprehensive information. Among the diverse RPA printing and analysis systems available in the market, we have used the technology developed by Zeptosens. As shown in Fig. 2, the standard configuration, this system allows working simultaneously with multiples of 32 samples. It should be considered that controls of noninfected cells must be included for each of the conditions tested. Thus, for example, if a collection of 30 mutants lacking particular virulence genes is available for the pathogen under study, they should be compared with a wild-type control sample at a fixed time point after infection, as well as with a noninfected cells sample to complete the array. Also, RPA analysis is a powerful tool to investigate the timing of signaling dynamics. So an alternative experimental design could be to study eight time points after infection for only two given mutants or strains of interest as compared to the same points of a wild-type or control strain and the equivalent set of samples from control uninfected cells. The method presented is based on our experience with Salmonella typhimurium infection of HeLa cells, but we believe it could be adapted to any host–pathogen system of in vitro infection. Experiments on cultures of fibroblasts or lymphoid cells might reveal novel aspects of the modulation of signaling by the pathogen, especially for bacteria that are able to survive within professional phagocytes, such as Salmonella typhi or Mycobacterium
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies
D
Zeptosens
E
F
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Buffer Standards
relative concentration 0.25
1
reference
Ref
0.5
C
Ref
0.75
B
Ref
1.0
A
Ref
reference
6 Arrays /Chip
43
Fig. 2. Layout of ZeptoMARK chip with six arrays. Each array accommodates four columns of reference spots and duplicate spots of four dilutions of 32 lysate samples. Control samples for the secondary antibody are spotted at the right bottom of each array. Reference spotting solution containing a BSA-fluorophore conjugate is spotted as columns of reference spots on each array.
tuberculosis. Different host cell lines, even primary cultures, or the influence of the presence of compounds or particular environmental conditions for both the pathogen and the host could also be compared (see Fig. 1). The most important prerequisite of this method is the quality of the antibodies selected for the experiment. Any antibody should be very specific for the epitope to be detected because any cross-reaction with other proteins in the lysates will give rise to background noise and failure to detect variations of the target epitope in the samples. All antibodies used must be previously validated for both their specificity and their sensitivity (see Subheading 3.8). To evaluate posttranslational modifications, such as phosphorylation, an antiphosphoprotein and antiprotein pair of antibodies must be used. 3.1. Preparation of Lysates 3.1.1. In Vitro Infection (see Note 4)
1. To prepare HeLa cells for the infection, grow them in a humidified 5% CO2 tissue culture incubator at 37°C for 24 h on P100 plates. 2. Before infection, wash HeLa cells twice with the same growth medium lacking any antibiotic supplement, and add fresh medium. 3. For the preparation of Salmonella inocula, grow bacterial cells overnight at 37°C in LB medium in a shaking orbital incubator (200 rpm). 4. Use 150 mL of these saturated cultures to inoculate 5 mL of the same fresh medium and incubate further until an OD600 = 0.5 (see Note 5).
44
Cid, Kauffmann, and Molina
5. At the time of infection (t0), add 25–30 mL of the bacterial inoculum per milliliter of RPMI 1640 medium. This should yield a multiplicity of infection (MOI or bacteria:cell ratio) of about 80:1 (see Note 6). In our experiments, to study bacterial invasion, infected cells were incubated for 10, 30, or 60 min (time points t10, t30 and t60) (see Note 7). 6. If it is intended to study modulation of host cell signaling by internalized bacteria in the absence of further invasion, it is necessary to remove extracellular bacteria. In this case, HeLa cells incubated for 60 min should be washed twice with fresh medium, then medium supplemented with 100 mg/mL gentamycin should be added and incubated for 1 h to kill extracellular bacteria. Then, incubate cells in medium with a lower gentamycin concentration (10 mg/mL) for the desired extra time to obtain further time points. In our experiments, we incubated cells for one extra hour (tis; “is” stands for “intracellular survival”). Intracellular survival can be monitored by counting colony-forming units (CFUs) on LB agar plates. 3.1.2. Cell Lysis (see Note 8)
1. Wash HeLa cells once with PBS. 2. Add CLB1 lysis buffer (0.2 mL per 10 cm-diameter culture dish), scrape and transfer into 1.5-mL reaction tubes. 3. Incubate 30 min at room temperature. 4. Centrifuge cell lysates for 5 min at 15,000 × g in order to remove debris. 5. Collect supernatants and freeze in liquid nitrogen. Cell lysate samples should be stored frozen at −20 or −80°C. They should be thawed only immediately prior to use.
3.2. Sample Preparation for Spotting
3.2.1. Determination of Protein Concentration in Lysates
Samples might be prepared for spotting by manual pipetting. However, in routine application of RPA, a liquid handling robot for protein quantification and the subsequent dilution steps will be beneficial in various ways. The sample preparation will be faster, avoiding degradation of samples and evaporation. Additionally, it will become more reproducible and the risk of erroneously exchanging samples is greatly reduced. The protocols for protein quantification and the preparation of the spotting described here can efficiently be integrated on a low-volume 96-channel Zephyr liquid handler (Caliper Life Science). An attached Twister II plate handler provides the interface to the microplate reader and serves as microplate storage. 1. Freshly prepare the BSA solutions shown in Table 1 (final CLB1 concentration 5%) to elaborate a standard curve. 2. Equilibrate the Coomassie solution and PBS buffer, and lysate samples to room temperature. After equilibration, gently shake Coomassie solution preventing generation of foam.
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies
45
Table 1 BSA standard solutions BSA (2 mg/mL ampoules) (mL)
CLB1 (mL)
PBS (mL)
Final BSA-concentration (mg/mL)
0
50
950
0
25
50
925
0.05
50
50
900
0.1
75
50
875
0.15
100
50
850
0.2
150
50
800
0.3
200
50
750
0.4
300
50
650
0.6
3. Vortex the lysate samples and centrifuge them at 10,000 × g for 1 min. Collect supernatants. 4. Dilute each sample 1:20 with PBS: Add 2.5 mL of the sample to 47.5 mL PBS at room temperature and mix thoroughly. 5. For Bradford Assay Preparation, transfer two times 10 mL of each of the diluted lysate samples and of each BSA standard solution (see Table 1) into duplicate wells of an empty 96well plate. 6. Add 240 mL Coomassie solution to each well. Work quickly, e.g., by using a multichannel pipette. 7. Place the 96-well plate on an appropriate plate mixing device for careful mixing of the reagents without generating foam or air bubbles. Incubate the solutions at room temperature. 8. Mix the plate again for 30 s right before measuring the absorbance at 595 nm in the MTP-Reader. Measure absorbance exactly 10 min after addition of the Coomassie solution. Since the development of the assay signal is time-dependent, it is advisable to include a calibration curve in every microtiter plate. 9. Generate a calibration curve by plotting the measured absorbance values versus the BSA concentrations of the standards and perform a linear regression. 10. Calculate back the protein concentrations in the lysates from the measured absorbance values, taking into account the dilution factor of 20.
46
Cid, Kauffmann, and Molina
3.2.2. Preparation of Reference Spotting Solution
For referencing of the fluorescence intensities on the array to the excitation light intensity, a reference spotting solution (RSS) containing a BSA-fluorophore conjugate is spotted as columns of reference spots on each array (see Fig. 2). RSS is prepared as follows: 1. Prepare a stock solution of the BSA-fluorophore in PBS and 0.1% sodium azide at a concentration of 5 mg/mL. It can be aliquoted and stored at −20°C for at least 1 year. 2. Thaw 10 mL aliquot of the stock solution. Centrifuge solution for 20 s at 10,000 × g and use the supernatant only. 3. Thaw a 500 mL aliquot of buffer RDB1, vortex well. 4. Thaw a 500 mL aliquot of buffer CSBR1, vortex well. 5. Prepare solution A by adding 2 mL of stock solution to 198 mL of RDB1 (dilution 1:100). This dilution may be varied in order to obtain the desired reference spot intensity of approximately 30,000 counts in an emission image with 10 s integration time. 6. Prepare solution B by adding 2 mL of Solution A to 198 mL RDB1 (dilution 1:100). 7. Prepare solution C by pipetting 180 mL of CSBR1 and 15 mL RDB1 into a microcentrifuge filter unit and filter it at 10,000 × g. 8. Add 5 mL of Solution B to filtered Solution C and mix well. The reference solution is ready for use.
3.2.3. Preparation of Spotting Microplate
The spotting solutions for the samples are prepared in 384-well polypropylene microwell plates which are used as source plates by the spotting robot. The well position of each solution has to be adapted according to the transfer scheme of the spotting robot and the desired spotting layout. 20 mL of spotting solution per well in a 384-well microplate is sufficient for the NanoPlotter described in this protocol. It is advisable to spot a series of at least four dilutions of each sample to allow for verification of linearity of dose response in the assays. Prepare the spotting solutions as follows: 1. Normalize samples to 2 mg/mL total protein concentration using buffer CLB1. 2. Dilute normalized samples tenfold with buffer CSBL1 (e.g., 20 mL sample + 180 mL CSBL1). This solution (0.2 mg/mL total protein concentration) is the first dilution spotted. 3. For the second spotting solution, dilute the first spotting solution with the dilution buffer (1 part CLB1, 9 parts CSBL1) to 0.15 mg/mL total protein concentration (e.g., 60 mL first spotting solution + 20 mL buffer).
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies
47
4. For the third spotting solution, dilute the first spotting solution with the dilution buffer (1 part CLB1, 9 parts CSBL1) to 0.1 mg/mL total protein concentration (e.g., 40 mL first spotting solution + 40 mL buffer). 5. For the forth spotting solution, dilute the first spotting solution with the dilution buffer (1 part CLB1, 9 parts CSBL1) to 0.05 mg/mL total protein concentration (e.g., 20 mL first spotting solution + 60 mL buffer). 3.3. Microarray Spotting
In the spotting process, droplets of 400 pL of each lysate solution, the reference solution, and control samples are deposited as arrays on ZeptoMARK hydrophobic chips. The best reproducibility and robustness of the spotting process have been reached using piezoelectric dispensing systems as, for example, provided on the NanoPlotter (Zeptosens/GeSiM, Germany). The array layout described here accommodates 32 lysate samples in four dilutions as duplicate spots, plus a negative control (spotting buffer) and a positive control [rabbit IgG and mouse IgG (a mix of mouse IgG1, IgG2a and IgG2b)], both as duplicate spots (see Fig. 2). 1. Start up the NanoPlotter and load/create a suitable spotting program (see Note 9). 2. Place source plate(s) on the spotter. The plate holder should be chilled to a temperature at which no evaporation of spotting solution can be measured over 12 h. At 21°C/50% relative humidity, the temperature of the coolant will typically be at 14°C (noncondensing conditions). 3. Place the chips on the slide deck. ZeptoMARK chips bind the proteins in the lysate droplets by hydrophobic interaction. It is advisable to spot more chips than at least needed for the envisioned number of antibodies (see Note 10). 4. Flush the piezo-electric dispensers (NanoJets, GeSiM) at least 5 min using the backfill water system. 5. Check the performance of the piezo jets with one of the samples using the stroboscope camera which is installed on the spotter. A single droplet with stable position should be generated. To optimize droplet formation, tune pulse width and voltage for the piezo jets in an alternating manner. Once optimized, pulse parameters are rather stable over the lifetime of a piezo jet. 6. Load the spotting program and start the spotting run. 7. After spotting, dry chips for 1 h at 37°C.
3.4. Chip Blocking
1. Place the ZeptoFOG blocking station in a laboratory hood. 2. Equilibrate blocking buffer BB1 to room temperature and filter it through a 0.5-mm cellulose acetate syringe filter into the blocking station.
48
Cid, Kauffmann, and Molina
3. Place upto two blocking racks with ZeptoMARK chips in the blocking station. The carrier must be tilted in a way that the spotted side of the chips is facing upward. The chips should be tilted towards the air inlet. 4. Close the ZeptoFOG blocking station with the lid. Connect the air inlet to the membrane pump. 5. Switch on the ultrasound generator. Fog is generated now. After 30 s, switch on the membrane pump. Nebulized blocking buffer will now rise. If optimally adjusted, the dense blocking fog is distributed homogeneously in the ZeptoFOG blocking station and a weak stream of fog leaves the system through the outlet. 6. Block the chips for 30 min. The blocking buffer should not warm up above 40°C. 7. Switch off membrane pump and ultrasound generator. Install lids and handles to the racks. Move to first wash bath filled with ca. 1.5 L water immediately. 8. Rinse the blocking rack holding the chips thoroughly in a sequence of four water baths with 1.5 L of water. 9. To dry the blocked chips, place the blocking racks (without lids and handles) on top of a 96-well plate in a microtiter plate centrifuge. The spotted side of the chips should face the center of rotation. Spin dry the chips at a maximum of 197 × g for 3 min applying maximum centrifuge acceleration. 10. Thorough cleaning of equipment is essential for low number of particle artifacts on blocked chips. Rinse the ZeptoFOG blocking station intensively with water and clean the blocking racks with water in an ultrasonic bath. 11. Store the blocked ZeptoMARK chips at 4°C until use. 3.5. Reverse Protein Array Assay
Six ZeptoCHIPs at a time are assembled in microplate-format ZeptoCARRIERs with fluidic cells. The micro flow system allows addressing each of the six arrays of a chip individually with 50 mL of antibody solution. The detection of the specific proteins follows a two-step sequential assay: 1. Prime all arrays with 100 mL of ZeptoMARK assay buffer CAB1, aspirate buffer again with the 8-channel aspiration tool. 2. Inject 50 mL of primary antibody solution ZeptoMARK assay buffer CAB1 or CAB2 in each cell. 3. Incubate for 20 h at 25°C in the dark. 4. Wash three times with ZeptoMARK buffer CAB1. 5. Inject 80 mL of appropriate fluorescent-labeled secondary antibody solution (at 500-fold dilution in buffer CAB1) in each cell.
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies
49
6. Incubate for 2.5 h at 25°C in the dark. 7. Wash the arrays again three times with ZeptoMARK buffer CAB1. After the last wash, leave the buffer in the cells for readout. In this two-step assay, the secondary antibody itself might generate unspecific signal on the spots (“Blank Signal”). Separate arrays, incubated with buffer only in the first part of the assay, and secondary antibody solution in the second part of the assay should therefore be prepared. The ZeptoMARK chips are imaged in a ZeptoREADER using planar excitation (see Note 11 and Fig. 4).
3.6. Readout and Data Analysis
1. Acquire fluorescence images of the arrays in the ZeptoREADER, exciting at lex = 635 nm and collecting fluorescence at lem = 670 nm, with exposure times of 1, 5 and 10 s. These are typical exposure times for measurement of medium to low abundant signaling pathway markers. 2. Determine the mean net fluorescence intensities of all spots (Image analysis module in ZeptoVIEW 3). Typically, a spot diameter setting of 100 mm will be optimal. The background should be determined for each spot individually and subtracted from the sample spot signal leading to the net signal. 3. Reference the mean net intensities of the sample spot to the excitation light intensity, represented by the reference spots (spotted fluorescence-labeled BSA, see Fig. 3). This results in
a
b
1.5
1.5
RFI
2
RFI
2
1
0.5
0 0
1
0.5
0 0.25
0.5
0.75
relative dilution of total protein
1
0
0.25
0.5
0.75
1
relative dilution of total protein
Fig. 3. (a) Dilution plot of two lysate samples with differing mean RFI under ideal assay conditions. Normalization with the mean RFI results in lines with slope equal to 1. (b) Dilution plot of a lysate sample under assay conditions with significant unspecific binding. The slope of the normalized dilution plot differs from 1. Note that the normalized dilution plot is only used for quality control. The mean RFI for each lysate sample is calculated from the nonnormalized data.
50
Cid, Kauffmann, and Molina
“referenced fluorescence intensity” (RFI) for each spot (Result module of ZeptoVIEW 3). 4. Calculate the mean RFI for each sample with the corresponding standard deviation based on an error-weighted linear fit through the RFI values of the eight sample spots (duplicate spots of the four sample dilutions) (23). The center among all spotted sample dilutions should be used as mean RFI value (Reverse array module of ZeptoVIEW 3). 5. Subtract the mean RFI of the blank assays with the corresponding secondary antibody and assay buffer from the mean RFI values of each marker assay. 3.7. Data Interpretation and Quality Control
RPAs allow quantifying the relative amounts of each (posttranslationally modified) marker protein in the set of lysate samples of an experiment. However, a number of quality control metrics have to be fulfilled. The immunoassay has to be performed in the linear range, i.e., excess of antibody. This can be verified using a dilution plot, in which the RFI values of single spots of a lysate sample are plotted against the sample dilution (see Fig. 4a). The signal of the dilution series has to correlate linearly with the spotted sample concentration. The correlation coefficient of a linear regression of RFI value with sample dilution can be used as quality parameter. The contribution of unspecific signal to the RFI should be small, i.e., when extrapolated to infinite dilution, the signal should approach zero. For an entire sample set, this can efficiently be verified using the slope of the dilution plot after normalization with the mean RFI, bNORM. In an ideal assay, bNORM equals 1. The effect of unspecific signal is exemplified in Fig. 3b.
planar waveguide principle
laser beam
planar waveguide excitation
confocal excitation
CCD camera
Fig. 4. Schematic of the planar waveguide excitation and image acquisition principle incorporated in the ZeptoREADER. Unlike confocal excitation, the evanescent field of the waveguide excites only surface-confined fluorophores and the direction of excitation is perpendicular to the direction of detection. This results in significant reduction of the background.
Reverse Protein Arrays Applied to Host–Pathogen Interaction Studies
51
9 30 min post-infection data
Ratio phosphoprotein/protein (nomalized to non-infected cells)
8 WT invA sopE/E2 sopB sopE/E2/B
4
3
2
1
0
Akt1 GSK-3b FoxO1 p70 S6K p42/44 S473 S9 T421 S256 (ERK) S424 T202 Y204
p38 T180 Y182
C-Jun S73
IkBa S32
CREB S133
PLCg1 Y783
Fig. 5. Evaluation of the activation of several signaling pathways after infection of HeLa cells with wild-type or mutants of Salmonella typhimurium lacking different subsets of the invasion-related proteins invA, sopE, sopE2, and sopB, as noted. Data represented in ordinates are the ratio of the signal obtained from the analysis of the phospho-epitopes indicated relative to that of the nonphosphorylated total protein. These data were previously normalized with respect to those obtained for the noninfected control. Thus, values above 1 imply an increase in phosphorylation in the conditions studied over background, whereas values below 1 imply a decrease in phosphorylation for that particular marker. Original raw data and further information for these experiments can be found in ref. (11).
The signal and the immunoassay should not reach saturation. On the other hand, spots at high sample dilution should only be taken into account if they provide measurable signal above the background with a signal-to-noise ratio above 3. Both events can be verified by determining the curvature of the dilution plot. In host–pathogen interaction studies, the mean RFI values obtained for a particular lysate of infected cells with each antibody should be normalized to the mean RFI value of the corresponding noninfected control. To assess differences in posttranslational modifications, such as phosphorylation, on particular signaling proteins, a phosphorylation ratio should be calculated from normalized mean RFI data obtained using the antiphosphoproteinspecific antibody and the antiprotein antibody for each particular lysate. A graphic representation of the phosphorylation ratio for different signaling proteins after 30 min of infection with different mutant strains of Salmonella on HeLa cells is provided in Fig. 5. 3.8. Antibody Validation
Like in a classical dot blot, the primary antibodies used on RPAs have to be highly specific. As only specific antibodies enable relative quantification of markers, significant effort is required to ensure the antibody validation before they are applied (24).
52
Cid, Kauffmann, and Molina
Antibodies for total, unmodified proteins should be tested for specificity in Western blots from non-overexpressing cell lines. Antibodies against modified proteins can be validated for specificity in a Western blot of broadly stimulated cultured cells. For example, an antibody against the phosphorylated form of AKT at position serine 473 could be tested on cultured cells stimulated with the unspecific Ser/Thr phosphatase inhibitor calyculin. Additionally, the antibodies have to be validated for linearity of dose–response in the reverse array experiment. The antibody concentration and buffer should be optimized for signal intensity and for the quality parameters described in Subheading 3.7. Typically, the optimum concentration for a reverse array is about double the concentration used in a Western blot.
4. Notes 1. It is important to obtain mammalian cell lines from a reliable source and to keep good frozen stocks and records. Low passage number (50 mL PBS-MT on oscillating platform for 3 min. Repeat for a total of 5 washes. If desired, final wash can be conducted in PBS-M to remove residual Tween-20. 7. To dry slide, place in a clean 50 mL conical tube and centrifuge 5 min at 2,000 × g. 8. To store the slide prior to analysis, place in a clean conical tube purged with N2. 3.4. Array Scanning and Data Analysis (Depends on Array Manufacturer)
1. Aptamer array scanning: Array scanning and feature extraction details will depend on the array vendor and scanner model. For the Agilent arrays, slides were scanned using the Agilent DNA microarray scanner at a 5 m resolution. In cases where feature saturation was observed, the slides were rescanned using the extended dynamic range (XDR) feature. The fluorescence intensity at each feature was measured using Agilent’s Feature Extraction Software and compiled into a database for subsequent analysis. 2. Data analysis: Microsoft Access (or similar database software) can be used to maintain the complete array database and facilitate sorting and filtering of the array data (see Note 5). Microsoft Excel can be used for relatively rapid graphical analysis of the sequence–function relationship when the fluorescence intensity of the truncations are graphed sequentially (see Fig. 3). M-Fold online software (13) enables correlation of aptamer structure and function by predicting thermodynamic parameters of oligonucleotide secondary structure (see Note 10). By conducting the analysis to display all possible structures for each sequence, even those that are thermodynamically unfavorable, the interplay between aptamer structure can be mapped to the function (i.e., fluorescence binding data) (see Fig. 4) (see Note 11).
Identification and Optimization of DNA Aptamer Binding Regions
63
Fluorescence Intensity (a.u.)
Consensus Sequence 10000
1000 5’ truncations 3’ truncations
100 5’
3’
T T T AT CCGT T CC T CC T AGT GG
Group 1 binding structure
T A T T T
C
C G T GA
C T C T C A
1
2
Region 3
4
5
6 lowest energy structure
1.0 0.5
no 2° structure present
0.0
4000 3500 3000 2500 2000 1500 1000 500 0
Group 2 non-binding structure 1
Group 3 non-binding structure 2 ... N
G C T T A T CT C T A T A A T C G A G C C G T G A T G G G C N ... A T
-0
-5
-10
-15
-20
-25
-30
-35
-40
Fluorescence Intensity (a.u.)
C G GGGT G T A TC A T T A A GC AT GC AT GC T A AT G C T A T ... N C G N ... T T A A T T A C C G G A C A T G G T A C G T A T C A C G G A T C G G A C G G C ... N A T N ...
Fraction of lowest free energy
Fig. 3. Interrogation of aptamer sequence and function elucidates the minimized functional binding domains. Serial truncations from both 5¢ and 3¢ ends of the aptamer sequence are superimposed. A disruption of target binding, observed as a decrease in fluorescence intensity, occurs when key nucleotides are removed. By superimposing the 5¢ and 3¢ serial truncations of the parent aptamer (filled and open circle, respectively), the minimally required IgE binding sequence is represented by those truncations exhibiting nominal fluorescence. In the case of the IgE clone (D-12.0), the experimentally determined binding region corresponds to the IgE consensus binding sequence TTTATCCGTTCCTCCTAGTGG. Figure 3 was modified from Fischer et al. (10).
-45
Serial 5’ truncations
Fig. 4. The free energy of observed secondary structures (IgE clone D-66.0) can be correlated to observed fluctuations in fluorescence intensity upon IgE binding, providing insight into the relationship between structure and function. Three dominant structural groups for the truncation set were predicted by Mfold (left ): (1) a stable stem:loops structure presenting the complete binding sequence in the loop, (2) a stem–loop structure, whereby the binding sequence is involved in both stem and loop formation, (3) a collection of short, random secondary structures. Aptamer bases required for IgE binding are highlighted. The corresponding DG values of these folding groups are plotted (top) in relation to overall fluorescence intensity observed at each 5¢ truncate (bottom). Six unique regions correlating free energy of the competing structural groups with fluorescence intensity are identified: (1) increased accessibility to binding domain by removal of 5¢, nonbinding bases, (2) higher propensity for the correct consensus fold due to lower stability of Group 2, (3) stem disruption in Group 1, increased stability of Group 2, and introduction of Group 3 compete with Group 1 formation, (4) Group 2 is completely abrogated, making Group 1 the most favorable secondary structure, (5) low stability in Group 1 stem limits proper folding, but binding of IgE target can readily induce proper folding, (6) Group 1 fold is no longer possible as bases required for stem formation and IgE binding are removed. Mfold structures were calculated using experimental parameters (25°C, 137mM NaCl, 1 mM MgCl2). Error bars are 1 SD of triplicate data points and all DG values are in kcal/mol. Figure 4 was modified from Fischer et al. (10).
64
Fischer and Tarasow
4. Notes 1. Typically, the standard oligonucleotide length on DNA arrays is ca. 60 nt. Customized arrays featuring longer oligonucleotides (up to ca. 100 nt) can be prepared, although at a higher cost. Ideally, aptamer constant/primer regions should be included in the design of the aptamer arrays, as these regions may be involved in secondary or tertiary structures required for binding. However, testing the aptamer variable region may suffice in some cases. 2. The maximal solubility of casein is 1%, and requires heating of the casein solution to 60°C for ca. 30 min. The remaining insoluble material is pelleted using a tabletop centrifuge (4,000 × g for 1 min) and passed through a 0.22 mm syringe filter. 3. Alternatively, more comprehensive truncation sets can be tested, such as single base truncations from both 5¢ and 3¢ and/or single base 5¢ truncations with two-base 3¢ truncations. Other groups have used similar approaches to probe mutational impact on and optimization of target binding. 4. In our original aptamer array design, the poly-T spacers were maintained at ten bases for all serial truncations. However, based on our observations (10) and those of others (11), longer spacers may provide the target protein great accessibility to the aptamer binding region. In particular, we observed an erosion of signal in all 3¢ truncation series, which we attribute to a decrease in the target protein’s accessibility to the binding sequence due to accessibility factors. As the 3¢ truncations remove a nucleotide from the base of the aptamer, the aptamer binding region is “pulled” toward the array surface. To ameliorate this effect, we suggest compensating the removal of each base from the aptamer by adding a base to the poly-T spacer, in effect maintaining the position of the binding region constant with respect to the array surface. This is reflected in Fig. 1b. 5. The Text and Data Function tools found in Microsoft Excel can be used to rapidly generate serial truncations, from both 3¢ and 5¢ ends. Also, tools exist to stitch together multiple text strings, facilitating the incorporation of the necessary poly-T spacer sequences into the aptamer array design. The RIGHT/LEFT tools allow one to separately prepare pairings of truncated sequences and appropriately sized poly-T spacers RIGHT(text [number_chars]) = provides indicated number of characters starting from right of the text.
Identification and Optimization of DNA Aptamer Binding Regions
65
LEFT(text [number_chars]) = provides indicated number of characters starting from the left of the text. The CONCATENATE tool allows stitching the truncated aptamer sequence to the poly-T spacer. CONCATENATE(text1, [text2], [text3], …) = joins text strings in a single cell. 6. When naming individual sequences, sufficient identifying information should be included in separate columns. This will facilitate any downstream sorting and filtering of fluorescence intensity data, providing for more agile analysis. 7. The 55 mL sample solution volume is optimized for an 8-chambered gaskets slide. From this point on, avoid letting the slide dry completely. 8. We found that stationary incubation works well for our aptamer arrays. When using stationary incubation, it is necessary to use caution during slide/gasket assembly to avoid the formation of air bubbles. Any trapped air will prevent contact between the array surface and the target solution, resulting in uneven target binding. Arrays designed with triplicate features, however, may provide sufficient redundancy to salvage data collected from a subarray compromised by an air bubble. Conversely, the standard incubation of DNA arrays is facilitated by the presence of a single, large mixing bubble within the chamber. Vertical rotation using a hybridization oven will ensure proper mixing. 9. The aptamer arrays were incubated with a range of IgE concentrations from 0.1 to >300 nM. IgE binding was observed at 0.1 nM IgE with a substantial signal-to-noise ratio, suggesting that significantly less protein is required. This may prove useful for protein targets that are in limited supply. 10. The Mfold server (version 3.2) can be accessed at http://mfold. bioinfo.rpi.edu/cgi-bin/dna-form1.cgi. The Quikfold server, which facilitates simultaneous analysis of multiple sequences, can be found at http://dinamelt.bioinfo.rpi.edu/quikfold.php. 11. For Mfold analysis, ensure that conditions during target incubation are represented (e.g., 37 mM Na+, 1 mM Mg++, 25°C).
Acknowledgments This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 with support from Lawrence Livermore National Laboratory (LLNLJRNL-415696).
66
Fischer and Tarasow
References 1. Proske D, Blank M, Buhmann R, Resch A (2005) Aptamers – basic research, drug development, and clinical applications. Appl Microbiol Biotechnol 69:367–374 2. Green LS, Jellinek D, Jenison R, Ostman A, Heldin CH, Janjic N (1996) Inhibitory DNA ligands to platelet-derived growth factor B-chain. Biochemistry 35:14413–14424 3. Nimjee SM, Rusconi CP, Sullenger BA (2005) Aptamers: an emerging class of therapeutics. Annu Rev Med 56:555–583 4. Tuerk C, Gold L (1990) Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase. Science 249:505–510 5. Ellington AD, Szostak JW (1990) In vitro selection of RNA molecules that bind specific ligands. Nature 346:818–822 6. Berezovski M, Musheev M, Drabovich A, Krylov SN (2006) Non-SELEX selection of aptamers. J Am Chem Soc 128:1410–1411 7. Knight CG, Platt M, Rowe W, Wedge DC, Khan F, Day PJ, McShea A, Knowles J, Kell DB (2009) Array-based evolution of DNA aptamers allows modelling of an explicit
8.
9.
10.
11.
12.
13.
sequence-fitness landscape. Nucleic Acids Res 37:e6 Jayasena SD (1999) Aptamers: an emerging class of molecules that rival antibodies in diagnostics. Clin Chem 45:1628–1650 Shangguan D, Tang ZW, Mallikaratchy P, Xiao ZY, Tan WH (2007) Optimization and modifications of aptamers selected from liver cancer cell lines. Chembiochem 8:603–606 Fischer NO, Tok JBH, Tarasow TM (2008) Massively parallel interrogation of aptamer sequence, structure and function. PLoS ONE 3:e2720 Katilius E, Flores C, Woodbury NW (2007) Exploring the sequence space of a DNA aptamer using microarrays. Nucleic Acids Res 35:7626–7635 Wiegand TW, Williams PB, Dreskin SC, Jouvin MH, Kinet JP, Tasset D (1996) Highaffinity oligonucleotide ligands to human IgE inhibit binding to Fc epsilon receptor I. J Immunol 157:221–230 Zuker M (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 31:3406–3415
Chapter 6 Recombinant Lectin Microarrays for Glycomic Analysis Daniel C. Propheter, Ku-Lung Hsu, and Lara K. Mahal Abstract The cell surface is covered with a myriad of carbohydrates that form a complex matrix of oligosaccharides. Carbohydrate recognition plays critical roles in pathogenesis, trafficking, and differentiation. Lectin microarray technology presents a novel platform for the high-throughput analysis of these structurally diverse biopolymers. One drawback of this technology has been limitations imposed by the commercially available plant lectins used in the array. Not only are a majority of these plant-derived proteins glycosylated, which can complicate glycomic analysis, but they also differ in activity and availability. Our lab has recently introduced recombinant lectins to enhance the stability and scope of our lectin panel. Herein, we provide a detailed procedure for the expression of bacterially-derived lectins and their application to a recombinant lectin microarray. Key words: Recombinant lectin, Microarray, Glycomics, Glycosylation
1. Introduction Glycomics, the high-throughput analysis of carbohydrates, is an arduous task given the inherent structural and chemical properties of glycans (1, 2). The carbohydrates of the mammalian glycome are complex, consisting of linear and branched polymers of structural isomers. The development of lectin microarray technology has helped to address these issues by using carbohydratebinding proteins (lectins), which can discriminate glycan structures and linkages, to analyze the glycome in a high-throughput manner (3, 4). Although this method has proven quite useful, the number of commercially available lectins, most of which are plant-derived, limits the level of structural resolution observed with the microarray. In addition, plant lectins often show lotto-lot differences in activity and availability due to their purification from natural sources. Recombinant lectins have helped to address these issues by taking advantage of bacterial expression
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_6, © Springer Science+Business Media, LLC 2011
67
68
Propheter, Hsu, and Mahal
systems, which allow for more stringent quality control and systematic purification techniques. Since bacteria require lectins for host-pathogen interactions and few bacterial lectins have been identified, bacterial genomes may represent a voluminous source of lectins (5). Furthermore, bacterial lectins are not known to be glycosylated (6). Therefore, our lab has recently cloned and purified a small set of recombinant lectins, and analyzed their glycanbinding profiles (7). These lectins are a nice complement to the commercially available lectin panel and show great promise as a scaffold for creating an expansive and diverse set of glycan-binding proteins via directed evolution.
2. Materials 2.1. Cloning and Purification of Recombinant DNA
1. Purified genomic DNA (may be purchased or isolated from natural source). 2. pET-41 Ek/LIC vector kit (Novagen, San Diego, CA). Kit contains: 1 mg pET-41 Ek/LIC vector, 10 mL pET-41 Ek/ LIC control vector, 25 units of T4 DNA polymerase (LIC qualified), 50 mL 10× T4 DNA polymerase buffer, 100 mM DTT, 50 mL 25 mM EDTA, 40 mL 25 mM dATP, 1.5 mL nuclease-free H2O, 22 × 50 mL NovaBlue Competent Cells (DH5a), 0.2 mL BL21(DE3) competent cells, 0.2 mL BL21(DE3)pLysS cells, 5 × 2 mL SOC media, and 10 mL test plasmid. 3. MJ Mini Personal Thermal Cycler (Bio-Rad, Hercules, CA). 4. Purified PCR insert: 0.2 pmol (after PCR). 5. Taq polymerase and dNTP solutions (Novagen). 6. Electrocompetent DH5a (Invitrogen, Carlsbad, CA). 7. Agarose, ethidium bromide, and kanamycin sulfate (Thermo Fisher Scientific, Rockford, IL). 8. Agarose gel-running equipment (Bio-Rad). 9. QIAprep spin miniprep kit (Qiagen, Valenica, CA). 10. Luria Broth (LB): 10 g tryptone, 5 g yeast extract, 10 g NaCl, 1 L of H2O (autoclave). 11. LB-Agar kanamycin plates: 4 g tryptone, 2 g yeast extract, 4 g NaCl, 6 g BactoAgar, 400 mL H2O (autoclave). After autoclaving the LB-agar mixture, let cool until warm to the touch, then add kanamycin sulfate (30 mg/mL final concentration). Pour solution into 100 × 15 mm petri dishes and let stand at room temperature for 20 min. Store at 4°C for up to 3 months. 12. 50× TAE buffer: 2.0 M Tris-base, 0.1 M ethylenediamine tetraacetic acid disodium dehydrate, 57.1 mL glacial acetic acid, 1 L ddi H2O, pH 8.5.
Recombinant Lectin Microarrays for Glycomic Analysis
2.2. Protein Expression and Purification
69
1. Electrocompetent BL21(DE3) cells (Novagen). 2. Terrific broth (TB): 12 g tryptone, 24 g yeast extract, 4 mL glycerol, 900 mL total ddi H2O (autoclave). 3. Avanti J-E Centrifuge (Beckman Coulter, Fullerton, CA): Equipped with rotors (JS-5.3 and JA-25.50). 4. Smartspec 3000 (Bio-Rad). 5. IC600 Incubator (Yamato Scientific, Chuo-ku, Tokyo). 6. Micropulser (Bio-Rad). 7. Rotary shaker and incubator (ATR, Laurel, MD). 8. Phosphate Buffer Saline (PBS): 100 mM sodium phosphate, 150 mM sodium chloride, pH 7.4. 9. Lysis buffer: PBS + 0.2% Triton X-100. 10. Aqueous 1,000× protease inhibitor cocktail: 3 mg leupeptin, 5 mg antipain, 12.5 mg pefabloc, 25 mg benzamidine, 50 mg trypsin inhibitor, 2.5 mL aprotin. Store aliquots at −20°C (Thermo Fisher Scientific). 11. Dimethylsulfoxide (DMSO) 1,000× protease inhibitor cocktail: 10 mg chymostatin, 5 mg pepstatin, 2 mL DMSO. Store aliquots at −20°C (components from Thermo Fisher Scientific, Rockford, IL). 12. DNAse (New England Biolabs, Ipswich, MA). 13. BioLogic LP low-pressure gradient chromatography with fraction collected (Bio-Rad). 14. GSTrap HP, 1 mL glutathione column (Amersham, Piscataway, NJ). 15. Slide-A-Lyzer Dialysis Cassette, 3,500 MW cut-off, 0.1–0.5 mL capacity. 16. DC Protein Assay (Bio-Rad). 17. BioTek Synergy HT plate reader (Bio-Tek, Winooski, VT). 18. Lactose, reduced glutathione, Bovine Serum Albumin (BSA) fraction V, and chicken egg white lysozyme (Thermo Fisher Scientific).
2.3. ELISA Activity Assay
1. Sodium azide and o-phenylenediamine hydrochloride (OPD) (Thermo Fisher Scientific). 2. Grenier 96-well Microlon microtiter plates (Grenier Bio One, Monroe, NC). 3. ELISA wash buffer: PBS + 0.05% Tween-20. 4. ELISA blocking buffer: PBS + 5% BSA. 5. PBST++: PBST + 1% BSA, 1 mM CaCl2, 1 mM MgCl2. 6. Anti-His6-horseradish peroxidase conjugated antibody (antiHis6-HRP) (Novus Biologicals, Littleton, CO).
70
Propheter, Hsu, and Mahal
7. OPD solution: 0.4 mg/mL OPD, 0.004% H2O2, 0.1 M phosphate/citrate buffer, pH 5.0. 8. Quenching reagent: 2.5 M H2SO4. 2.4. Recombinant Lectin Microarray
1. Print buffer: PBS, 0.5 mg/mL BSA, 1 mM monosaccharide (see Table 1). 2. Nexterion H slides (Schott North America, Elmsford, NY). 3. Contact microarray printer such as the SpotBot2 (ArrayIt Corp., Sunnyvale, CA) or the Microgrid II (DigiLab, Inc., Holliston, MA). 4. SMP3 pins (ArrayIt Corp.). 5. 16-frame FAST frame hybridization chamber (Schleicher and Schuell, Keene, NH). 6. Genepix 4100A slide scanner and Genepix Pro 5.1 software (Molecular Devices Corporation, Union City, CA). 7. Slide blocking buffer: 50 mM ethanolamine in 50 mM sodium borate, pH 8.0. 8. PBST+: PBS + 0.005% Tween 20, 1 mM CaCl2, 1 mM MgCl2. 9. Slide wash buffer: PBST (PBS + 0.005% Tween 20). 10. Slide spinner (Labnet International, Edison, NJ). 11. Coplin jars. 12. 384-well plates (Whatman, Piscataway, NJ). 13. NHS-Cy3 or Cy5 (GE Healthcare Life Sciences, Piscataway, NJ).
Table 1 Lectin print list Lectin
Source
Sugar in print
Specificity
GafD
F17 fimbriae (Escherichia coli)
GlcNAc
b-GlcNAc
PA-IL
Nonfimbriae (Pseudomonas aeruginosa)
Galactose
Galactose
PA-IIL
Nonfimbriae (P. aeruginosa)
Fucose
Fucose/mannose
PapGII
P-pili (E. coli)
Galactose
GbO4
PapGIII
P-pili (E. coli)
Galactose
GbO5
RS-IIL
Nonfimbriae (Rhizoctonia solanacearum)
Mannose
Mannose/fucose
GafD-m
F17 fimbriae (E. coli)
GlcNAc
b-GlcNAc (80% reduction in binding)
Recombinant Lectin Microarrays for Glycomic Analysis
71
3. Methods Glycan recognition of cell surface carbohydrates using lectin microarrays has proven to be a valuable tool for glycomics (8). By utilizing a simple methodology to clone and express bacterial lectins from microbial genomes, we can expand the detection capabilities of the lectin microarray technology (Fig. 1). The pET-41 Ek/LIC vector system enables both rapid cloning and the incorporation of two affinity tags: A hexa-histidine (His6) and an N-terminal glutathione-S-transferase (GST), for dual modes of purification and detection. The recombinant lectins are overexpressed in BL21 Escherichia coli cells, and desired cells are selected against kanamycin sulfate. Expression and purification use standard methods. Once the recombinant protein is purified, the lectin activity (defined as the signal-to-noise ratio, S/N) is tested against glycoprotein standards. The ELISA assay presented here can be used to test lectin activity against multiple glycoproteins, thereby enabling a wide range of binding profiles to be analyzed. Once the recombinant protein has displayed significant activity (S/N >5), the lectin can be utilized in the microarray. In brief, the recombinant lectin microarray is fabricated by spotting the bacterial lectins onto an N-hydoxysuccinimide-, (NHS-), activated glass slide, and immobilization is achieved through amine-coupling of side-chain lysines.
Fig. 1. Schematic representation of the cloning, expression, and use of bacteria-derived recombinant lectins. The desired lectin is cloned out of the microbial genome and amplified by PCR. The gene is then annealed into the pET-41Ek/LIC vector and expressed into E. coli. The expressed protein is purified, and analyzed for activity using ELISA and microarray techniques. Adapted from Hsu et al. (7).
72
Propheter, Hsu, and Mahal
Fig. 2. Recombinant lectin microarray screening of tumor cell lines (ACHN, and Sk-MEL-5). Samples were prepared and analyzed as previously described (8). In brief, cell membranes were sonicated, and the resulting micellae isolated and labeled with NHS-Cy5. The samples were then incubated with the recombinant microarray (10 mg in 100 mL of buffer) and the arrays processed as previously described. A clear differential pattern can be observed between ACHN and Sk-Mel-5, which is described in the text.
A list of the recombinant bacterial lectins cloned, purified, and added to the lectin microarrays to date is given in Table 1. These lectins come from a variety of bacterial sources and include both adhesins from pili (GafD, PapGII, and PapGIII) and secreted lectins (PA-IL, PA-IIL, and RS-IIL). Incubation of the printed slides with fluorescently labeled samples provides a discernable pattern that gives insight into the extent of glycosylation of a given sample (9). Using this technology, we have shown that even the small panel of recombinant bacterial lectins utilized to date (Table 1) can distinguish tumor cell lines in the NCI-60 panel (Fig. 2). Clear differences can be observed between the renal cell carcinoma ACHN, which shows fucosylation (PA-IIL in the absence of RS-IIL), the presence of terminal b-N-acetyl-d-glucosamine (GafD) and galactosylation (PA-IL), and Sk-Mel-5, which shows an absence of both the terminal GlcNAc and galactose epitopes. Although one can obtain differences with this small of a lectin panel you cannot obtain a comprehensive snapshot of the glycome. However, the inclusion of these lectins in a larger lectin microarray format allows for a far more detailed analysis than is presented herein. 3.1. Cloning and Purification of Recombinant DNA
1. Identify microbial lectin via BLAST, the literature or other sources. 2. Prepare primers flanking the lectin encoding region (see Note 1). 3. Prepare PCRs as follows: 1× reaction buffer, 400 mM dNTP solution, 1 mM 5′ primer, 1 mM 3′ primer, 2.5 units of Taq
Recombinant Lectin Microarrays for Glycomic Analysis
73
Polymerase, and template DNA. Dilute to final volume of 50 mL with ddi H2O. (see Note 2). 4. Place the PCR tubes into the thermocycler and run on the following conditions: 95°C for 10 min, 95°C for 30 s, 45°C for 30 s (Tm), and 72°C for 1 min. Let the reaction go for 40 cycles, and cool PCRs to 4°C (see Note 3). 5. Prepare a 2% w/v agarose gel solution in 1× TAE buffer and heat until mixture is miscible (see Note 4). Add EtBr to a final concentration of 0.5 mg/mL. Pour gel into cast and allow it to solidify. Next, add 5 mL of DNA ladders and 2 mL of PCR mixture. Run gel on 90 V for 45 min, then visualize under UV irradiation. 6. To anneal the PCR insert, first determine the amount of PCR product required for the T4 treatment by using the following formula: (number of base pairs in insert) × 650 × 0.2 pmol = n pg PCR insert. 7. In a sterile 1.5 mL microfuge tube, add the amount of purified PCR product calculated in step 6 (n pg), 2 mL 10× T4 DNA polymerase buffer, 2 mL 25 mM dATP, 1 mL 100 mM DTT, and 0.4 uL 2.5 units/uL T4 DNA polymerase. Add enough ddi H2O to have 20 mL of total volume (see Note 5). 8. Mix the components by flicking the tube and then incubate at room temperature for 30 min. 9. Inactivate the enzyme by incubating at 75°C for 20 min. 10. To anneal into pET-41 Ek/LIC vector, mix 1 mL of the vector with 2 mL of the treated PCR insert in a sterile 1.5 mL microfuge tube and incubate for 5 min at room temperature. 11. Add 1 mL of 25 mM EDTA to the reaction mix, and incubate at room temperature for 5 min (see Note 6). 12. Transform competent DH5a with 1 mL of the annealing reaction, add 1 mL SOC media and allow cells to recover for 1 h, shaking at 250 rpm at 37°C (see Note 7). 13. After 1 h, plate the transformed cells on LB-Agar plates (see Note 8). Allow the cells to grow overnight (~12 h) at 37°C. 14. Pick single colonies and grow in 5 mL of LB (~15 h) with 30 mg/mL kanamycin on a rotary shaker and incubator at 250 rpm at 37 C. 15. After 15 h, take the optical density of the colonies at 600 nm (OD600). Pick the best growing colony and inoculate in 25 mL of LB with kanamycin (30 mg/mL) and place on rotary shaker at 250 rpm, 37°C, for ~15 h. 16. After overnight culture, take OD600, and purify plasmid DNA via Qiagen Miniprep Kit and Instructions. Once DNA is isolated, check the DNA sequence.
74
Propheter, Hsu, and Mahal
3.2. Protein Expression and Purification
1. Transform electrocompetent BL21(DE3) cells with recombinant DNA. Transform 1–5 mg of DNA into a 50 mL aliquot of BL21 cells (see Note 9). 2. Upon electroporation, promptly add 1 mL of LB and grow on a rotary shaker for 1 h at 250 rpm and 37°C. Next, plate cells onto LB-Agar plates and incubate at ~15 h at 37°C. 3. Next, pick single colonies and grow each in 5 mL of LB with kanamycin (30 mg/mL) for 15 h at 250 rpm and 37°C. 4. Take OD600 of colonies, choose a colony with an average rate of growth, and inoculate 5 mL culture into 25 mL culture (see Note 10). Grow the culture to an OD600 of 0.7–1.0, then induce the culture with 1% w/v lactose and grow for 3 h at 250 rpm and 37°C (see Note 11). 5. After 3 h, transfer culture into centrifuge tubes and pellet cells at 6,000 × g, 4°C, 15 min (see Note 12). Discard the supernatant. 6. Resuspend pellet in 1 mL of lysis buffer and dilute 1,000× DMSO and aqueous protease inhibitor cocktails to 1× in lysis buffer (see Note 13). Then add approximately 1 mg/mL lysate of chicken egg white lysozyme and mix at 4°C for 30 min (see Note 14). 7. Next, immediately add DNAse, 5 mg/mL of lysate final concentration, and incubate further at 4°C for 10 min (see Note 15). 8. Centrifuge the samples in the appropriate tubes at 30,000 × g for 30 min at 4°C. Keep the supernatant. 9. Purify the lysate using the BioLogic LP low-pressure gradient chromatography system (or similar system). Load the supernatant onto an equilibrated glutathione column at a flow rate of 0.5 mL/min (see Note 16). Wash the column with ~10 column volumes of PBS at a rate of 1 mL/min. Elute lectin with 10 mM of reduced, free acid glutathione in PBS collecting 1 mL fractions at a rate of 1 mL/min (see Note 17). 10. Monitor lectin purification by 10% SDS-PAGE analysis (see Note 18). Pool fractions containing lectin and dialyze against PBS at 4°C. Aliquot, flash freeze, and store at −80°C (see Note 19).
3.3. ELISA Activity Assay
1. Dilute glycoprotein to a final concentration of 1–10 mg/mL in PBS containing 0.1% NaN3. Take a 96-well plate and coat each well with 100 mL of glycoprotein and incubate for ~12 h at 4°C. 2. Wash each well with wash buffer 5×. Next, add ELISA blocking buffer to each well and incubate at room temperature for 1 h (see Note 20). 3. After blocking, wash each well with wash buffer 5×. Next, dilute lectin into PBST++ (see Note 21). Add 50 mL of each
Recombinant Lectin Microarrays for Glycomic Analysis
75
dilution to each well (see Note 22). Incubate samples at room temperature for 1 h. 4. After incubation, wash each well with wash buffer 5×, then dilute anti-His6-HRP to an optimized dilution in the wash buffer + 1% BSA (see Note 23). Next, add 50 mL of anti-His6HRP solution into each well and incubate at room temperature for 1 h. 5. Wash each well with wash buffer 5×, and freshly prepare OPD reagent buffer. Add 100 mL to each well immediately after the last wash. Incubate at room temperature for 30 min, add 50 mL of stopping reagent to each well, and read on BioTek Plate Reader at 492 nm wavelength (see Note 24). 3.4. Recombinant Lectin Microarray
1. Prepare samples as previously described (8). 2. Dilute lectins to 1 mg/mL in print buffer and 1 mM monosaccharide as specified (see Table 1). 3. Print lectins onto Nexterion H slides using the SpotBot personal microarray with an SMP3 pin. Maintain cold plate at 8°C and internal humidity at 50–60%. 4. Print 5 spots per lectin, to ensure spot quality, on a 16-subarray format (see Note 25). 5. Upon completion of the print, slides are allowed to warm to room temperature in the SpotBot arrayer for 1 h while maintaining humidity control. Slides are then placed into blocking buffer inside a Coplin jar for 1 h at room temperature. 6. After blocking, wash slides with PBST 3× for 3 min, rinse once with PBS, and dry using the slide spinner. 7. Affix a 16-well subarray FAST frame to the slide and incubate with appropriate fluorescent sample for 2 h at room temperature (see Note 26). 8. After incubation, aspirate sample from the subarrays and wash 5× with PBST, once with PBS, and then dry using a slide spinner (see Note 27). 9. Scan slides using the Genepix 4100A scanner at the appropriate wavelength. Extract data using Genepix Pro 5.1 software and analyze using Microsoft Excel and/or Graphpad Prism 4.0.
4. Notes 1. The designed primers must have the overlapping LIC extensions. The forward primer must begin with 5¢ GAC GAC GAC AAG A 3¢ and the reverse primer must begin with 5¢ GAG GAG AAG CCC GG 3¢.
76
Propheter, Hsu, and Mahal
2. You will have to titrate the amounts of template DNA used in the first PCR to obtain the desired results. 3. The melting temperature (Tm) and number of cycles may need to be changed in order to obtain the desired results. 4. Depending on the size of your DNA, you may need to augment the percent of agarose used. 5. pET-41 Ek/LIC vector kit has a control vector which should be used to gauge efficiency of system. Also, it is important to include as a negative control plasmid with no insert to evaluate the selection. T4 DNA polymerase from Novagen is specifically designed for these ligation-independent cloning reactions. Nuclease-free or ddi H2O (i.e., from a purification system) may be used. 6. The T4 treated insert can be stored at −20°C for up to 3 months. 7. SOC may be substituted for LB in this recovery although efficiency may be reduced. Transform the DNA using a Micropulser (Bio-Rad), following the Bio-Rad electroporation protocol (found at http://www.bio-rad.com). 8. For the best results, plate two dilutions of sample. Also, the negative control should be plated to ensure the integrity of the kanamycin. 9. Following the Bio-Rad electroporation protocol referred to earlier (see Note 4). 10. Colony-dependent variations in protein expression arise, so be sure to test ~3 to 5 colonies. Take the best expressing colony and move on to next step. 11. The cultures can be easily scaled up to a 4 L culture 12. This pellet can be stored at −80°C for an indefinite amount of time. 13. Typically we add 4 mL of lysis buffer per 100 mL of culture. If using a 1,000× protease inhibitor cocktail, simply add 1 mL/mL of culture. 14. Be sure to keep all reagents and solutions on ice. 15. If lysate is very viscous, the DNA can be sheared by drawing the suspension through an 18-gauge needle several times. Keep a small aliquot (~50 mL) of the crude lysate for SDS-PAGE gel analysis. 16. We have found that taking the following steps ensures that the column maintains integrity through multiple experiments: First, flush the system with ddi H2O and inspect to make sure no clogs or air are in the system. Second, flow filtered PBS for 5 min at a rate of 1 mL/min. Then load supernatant and follow the protocol. After eluting the column, wash the column
Recombinant Lectin Microarrays for Glycomic Analysis
77
with 5 column volumes of 70% ethanol, followed by 5 column volumes of PBS, and then 1 column volume of 20% ethanol. Store column at 4°C indefinitely. 17. Be sure to keep an aliquot (~50 mL) of the lysate after purification for SDS-PAGE gel analysis. 18. You may need to use a different % acrylamide gel depending on the molecular weight of the desired lectin. 19. Aliquots may be stored for up to 6 months. 20. Be sure to keep the 96-well plate covered to prevent contamination. 21. We recommend serial dilutions for the first ELISA to obtain a larger, more informative dataset. 22. For negative control wells, simply add PBST++ with no lectin. 23. We recommend testing serial dilutions of anti-His6-HRP on initial work to optimize the working dilution. 24. Read ELISA plates also at 620 nm as the reference wavelength. For data analysis, subtract the readings at 620 nm from the 492 nm data set to obtain the true values. The 620 nm value is a background value taken to correct for any imperfections in the sample plate. 25. You can print in a 24-well format and/or limit the number of spots to three, based on previous protocols. Print spots are typically 100 mm. 26. Be sure to keep slides unexposed to light, which affects fluorescence. 27. Be sure to keep slides in the dark. Slides can be kept for long term storage at −20°C. References 1. Mahal LK (2008) Glycomics: towards bioinformatic approaches to understanding glycosylation. Anticancer Agents Med Chem 8: 37–51 2. Hirabayashi J (2008) Concept, strategy and realization of lectin-based glycan profiling. J Biochem 144(2):139–147 3. Pilobello KT, Krishnamoorthy L, Slawek D, Mahal LK (2005) Development of a lectin microarray for the rapid analysis of protein glycopatterns. Chembiochem 6:985–9 4. Hirabayashi J (2004) Lectin-based structural glycomics: glycoproteomics and glycan profiling. Glycoconj J 21(1):35–40 5. Sharon N (2006) Carbohydrates as future anti-adhesion drugs for infectious diseases. Biochim Biophys Acta 1760:527–37
6. Dodson KW, Pinker JS, Rose T, Magnusson G, Hultgren SJ, Waksman G (2001) Structural basis on the interaction of the pyelonephritic E. coli adhesion to its human kidney receptor. Cell 105:733–43 7. Hsu KL, Gildersleeve JC, Mahal LK (2008) A simple strategy for the creation of a recombinant lectin microarray. Mol Biosyst 4:654–62 8. Pilobello KT, Slawek DE, Mahal LK (2007) A ratiometric lectin microarray approach to analysis of the dynamic mammalian glycome. Proc Natl Acad Sci USA 104:11534–9 9. Krishnamoorthy L, Bess JW Jr, Preston AB, Nagashima K, Mahal LK (2009) HIV-1 and microvesicles from T-cells share a common glycome, arguing for a common origin. Nat Chem Biol 5(4):244–250
wwwwwwwwwwwwwwww
Part II Antigen Microarrays for Immunoprofiling
wwwwwwwwwwwwwwww
Chapter 7 Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection Persis P. Wadia, Bita Sahaf, and David B. Miklos Abstract Recombinant antigen arrays represent a new frontier in parallel analysis of multiple immune response profiles requiring only minute blood samples. In this article, we review the benefits and pitfalls of recombinant antigen microarrays developed for multiplexed antibody quantification. In particular, we describe the development of antigen arrays presenting a set of Y chromosome-encoded antigens, called H-Y antigens. These H-Y antigens are immunologically recognized as minor histocompatibility antigens (mHA) following allogeneic blood and organ transplantation. Clinically relevant B-cell responses against H-Y antigens have been demonstrated in male patients receiving female hematopoietic stem cell grafts and are associated with chronic graft versus host development. This chapter discusses our recombinant antigen microarray methods to measure these clinically relevant allo-antibodies. Key words: H-Y proteins, Antibodies, Plasma, Recombinant antigen microarrays, Minor histocompatibility antigens
1. Introduction Identifying, understanding, and confirming complex multicellular processes, such as immunity, require a systems biology approach to integrate each component’s function and regulation within the network. Traditionally, genes and proteins were discovered and characterized in isolation as individual molecules. However, the development of DNA microarrays facilitated multiplexed gene expression pattern analysis in a variety of genomes spanning bacteria (1–3) to human (4, 5). In immunology, gene expression profiling has determined important lymphocyte gene regulation pathways and their linked biological functions (6–9). However, a more complete understanding of adaptive immune responses requires systematic target screening of proteomes isolated from Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_7, © Springer Science+Business Media, LLC 2011
81
82
Wadia, Sahaf, and Miklos
bacteria, viruses, or humans. Antibody secretion marks effective B lymphocyte immune responses, and historically, these antibodies have identified specific targets ex vivo via western blot, immunoprecipitation, or Enzyme-Linked Immunosorbant Assay (ELISA). While Western blot and immunoprecipitation provides qualitative antigen identification from complex lysates, determining the specific protein or responsible gene often requires numerous subsequent biochemical fractionations and sequencing reactions. ELISA quantifies antibody against specific antigens, but their single antigen design consumes precious samples and resources. In contrast, protein microarrays enable high-density presentation of thousands of spatially isolated candidate antigens. Following antibody incubation, specific antigen binding is detected with fluorochrome conjugation. In fact, differential flurochrome conjugation of multiple samples enables multiplexed detection using the same antigen microarray. Ideally, these protein microarrays contain highly-purified antigens (see Note 1) that maintain native protein structure and include posttranslational modifications (see Note 2). In this chapter, we discuss two critical considerations for the generation of recombinant antigen microarrays: (1) the format of the antigens to be printed (Subheading 1.1) and (2) optimization of printing the recombinant protein on printing substrates (Subheading 1.2). We will discuss commercially available microarrays followed by a detailed description of our approach to optimizing the generation of microarrays to express custom antigens (H-Y antigens) for the detection of allo-antibodies (Subheadings 1.3 and 1.4). 1.1. Considerations in Expression of Recombinant Antigens for Protein Microarrays
Posttranslational modifications vary by organisms used for recombinant protein expression. The various organisms used to produce proteins include: Escherichia coli, yeast (10), CHO cells (11), or baculovirus in insect cells (12, 13), and are listed in Fig. 1. The scientific need to preserve posttranslational modifications determines expression system requirements and is also offset by expression efficiency. Modifications such as phosphorylations, acylations, glycosylations, and carboxylations demand a eukaryotic expression system since prokaryotic expression, such as through E. coli, lacks the necessary posttranslational machinery. However, the disadvantage of decreased protein yield through eukaryotic expression is overcome by the decreased antigen requirements for the protein microarray. Nonetheless, bacterial expression will suffice for many recombinant antigen expression needs and remains ubiquitously available, inexpensive, and fast. A significant disadvantage of bacterial expression is the frequent development of inclusion bodies necessitating protein denaturation with subsequent renaturation attempts. Yeast systems and baculovirus-infected insect cells represent reasonable compromises providing proteins in large amounts with eukaryotic modifications.
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
Bacterial cells
Yeast
Insect cells
83
Mammalian cells
Transform cells with gene (with/without tag) Induce expression of protein Small scale purification of expressed protein: select the most robust protein expression colony/culture Large scale purification of expressed protein
Secreted protein obtained in the supernatant of cells Native Nickel affinity chromatography
Nonsecreted protein obtained as inclusion bodies Denatured Nickel affinity chromatography
Renaturation of purified protein
Quantify and concentrate expressed protein
Print expressed protein of interest
Fig. 1. Flow-sheet for protein expression and purification. A schematic flow-sheet of choosing an expression system and purifying the proteins is detailed in the figure.
However, these systems are more laborious and expensive than prokaryotic systems. Figure 1 presents a schema for the steps involved in antigen purification for recombinant antigen microarrays after the appropriate expression system is chosen (to be discussed in detail in the Subheading 3). The incorporation of epitope tags (GST, V5 or 6xHis tags) for detection and/or isolation of expressed antigens provides a major advantage for recombinant microarray development. Expression plasmids inserting open reading frames (ORF) in frame following N-terminal tags usually provide high-yield protein expression and an affinity tag for protein purification. C-terminal epitope tag recognition indicates the entire ORF has been expressed (see Note 3). One example of a commercially available high-density protein microarray that prints proteins expressed in the baculoviral expression system are Protoarrays™ marketed by Invitrogen (Carlsbad, CA). More than 9,000 human proteins with N-terminal GST epitopes expressed in baculovirus-infected insect cells are affinitypurified and printed in duplicate on nitrocellulose-coated slides. An advantage of using commercial microarrays is that there are
84
Wadia, Sahaf, and Miklos
Fig. 2. Representative figure of a subarray from a commercially available protein microarray with controls and antigen hits. Commercially available protein microarrays contain 48 subarrays with 9,000 proteins printed in duplicate (Protoarray, version 5.0). A representative subarray is shown with negative controls such as Buffer, GST tags in different concentrations, and empty spots. The subarray also contains positive controls, such as anti-human IgG and human IgG, each printed in four concentrations. We use human IgG3 (second highest concentration) and we aim to obtain an MFI of 55,000–60,000 while scanning to normalize our arrays. Alexa 647 is printed in various positions, but fixed positions, across subarrays to help distinguish subarrays while gridding the spotted antigens.
numerous controls printed on each subarray, and once a target has been identified, the protein can be purchased for further analysis or ELISA development analysis. A representative subarray with negative and positive controls is shown in Fig. 2. Negative controls include buffer, empty spots, and GST tags printed in different concentrations and positive controls include human IgG as well as anti-human IgG printed in four different concentrations. Currently, cost prevents wide use of proteome microarrays, but increased content and decreased cost are expected. Our laboratory has extensive experience in using the commercially available protein microarrays from Invitrogen. We used
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
85
Fig. 3. Donor, Pretransplant, and 12-month Posttransplant proteins detected serologically using commercially available protein microarrays. The same representative subarray (Protoarray version 3.0) is shown across three different slides which were processed using donor serum, pre, and posttransplant plasma. One of the two targets identified statistically is shown in the figure (CHAF1b). This was chosen because CHAF1b was absent in the donor and pretransplant plasma sample, but was recognized by antibodies in the posttransplant plasma sample.
protein microarrays to serologically identify Nucleolar and Spindle Associated Protein 1 (NuSAP1) and Chromatin Assembly Factor 1, subunit B (p60) (CHAF1b) as targets of new antibody responses that developed after allogeneic hematopoietic cell transplantation (HCT; Fig. 3). Western blots and ELISA validated their postHCT recognition and enabled ELISA testing of 120 other alloHCT patients with various malignancies. CHAF1b-specific antibodies were predominantly detected in AML patients, whereas NuSAP1-specific antibodies were exclusively detected in AML patients 1 year posttransplant (p < 0.0001). Expression profiles and RT-PCR showed that NuSAP1 was predominately expressed in the bone marrow CD34+CD90+ hematopoietic stem cells, leukemic cell lines, and B lymphoblasts as compared to other tissues or cells. Thus, NuSAP1 is recognized as an immunogenic antigen in 65% AML patients following allogeneic HCT and suggests a tumor antigen role. In conclusion, though protein microarrays is a nascent technology, clinically important tumor antigens can be identified as new antibody targets after allogeneic HCT using high-density protein microarrays (14). As with each new technology, there are advantages and disadvantages. The primary disadvantages of commercially available protein microarrays are that (1) there are new versions available with little bioinformatics support to compare across different versions and (2) thus far they are not normalized for protein concentration. Marina et al. (15) have recently published a concentration-dependent analysis (CDA) method to normalize for the concentration of the spotted antigen on commercially available protein microarrays. Their method is complementary to other commonly employed analyses and demonstrated experimental validation of 92% of hits identified by the intersection of
86
Wadia, Sahaf, and Miklos
CDA with other tools. However, these disadvantages are offset by the many advantages of using recombinant protein microarray antibody detection technology; these include the ability to simultaneously screen (1) thousands of antigens, (2) antigens that reflect varied tissue organs, and (3) thousands of antigens using only small amounts of sample, in a multiplexed, quantitative, reproducible, and rapid fashion. Thus, combining the advantages of antibody-binding specificity and antigen microarray technology provides global humoral immune assessment. 1.2. Considerations in Printing of Recombinant Antigens in Custom Protein Microarrays
Printing substrates vary in their array-binding efficiency, antigen orientation, and epitope availability. Printing proteins by absorption (nitrocellulose and PVDF-coated glass slides have high protein-binding capacity) (16) is a popular method since no further antigen modification is necessary and the bound proteins often retain the functional active-binding sites. However, antigen retention varies and is weaker for absorption printing as compared to affinity or covalent cross-linking of proteins on the array. Some printing substrates enable nonspecific covalent binding such as polylysine or aldehyde-coated glass slides (17, 18). In either absorption or covalent cross-linking methods, proteins will bind to the array in often unpredictable orientations. Substrates linking antigens via affinity-printing can both improve printed substrate retention and maintain antigen orientation. 6xHis epitope-tagged proteins bound onto nickel-coated glass slides are an example of affinity-printing (19–21). Affinity-printing recombinant proteins promise improved antigen retention and consistent orientation. We have used epitope-tag affinity purification of custom proteins and then directly printed the antigens onto nitrocellulosecoated microarrays. We have found that the critical variables that need to be considered to optimize custom printing of recombinant antigen microarrays are: (1) solubility of proteins (see Note 4); (2) print buffer conditions (our H-Y antigens are printed in their 250 mM imidazole elution buffer); and (3) step gradient versus a linear gradient elution step. For example, in our own experience, we have used affinity chromatography in which 6xHis epitopes specifically bind Nickel-NTA resins and are subsequently eluted by increasing concentrations of imidazole. We have found that a linear elution gradient with linearly increasing imidazole concentrations results in broad elution peaks which lead to diluted proteins in the eluate. Thus, instead, we use a step gradient of imidazole concentrations (low to high concentration of imidazole) to obtain sharp elution peaks. This results in smaller aliquots of tagged proteins, but with high protein purity. Antigen printing remains unpredictable, but depends upon printing pin size and type, print buffer, and microarray surfaces. Because heterogeneous antigens yield a spectrum of array binding,
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
87
strategies to confirm and normalize retained antigens are important. Basically, printing pins vary by size and style (solid pins vs. quill pins). Solid pins deposit the antigens while touching the coating surface of the slide, while quill pins dispense repeating aliquots via capillary action, limiting the need for pin reloading (see Note 5). Solid and quill pins vary by their spot sizes with the spot diameter ranging from 62.5 to 600 mm. In our experience, we have used a solid pin (300 mm spot diameter with 360 mm spacing between spots) and have also used a quill pin (100 mm spot diameter with 120 mm spacing between spots; also used for printing the H-Y arrays). The spot sizes can be varied to accommodate more antigens per subarray if needed. The amount of antigen printed, i.e., protein concentrations, should be uniform intra- and interarray for each antigen (see Note 6). Spots printed should also be of the same size and shape. Irregular spots may make further analysis nonreproducible (see Note 7). Also, if many spots are printed from one-time uptake of antigen in the pin, care should be taken to account for evaporation, leading to concentrated antigen spots printed in a later batch as compared to the initial batch of slides (see Note 8). After antigen printing and processing antigen microarrays, bioinformatics for these antigen microarrays printing and detection as well as final analyses need to be standardized. Bioinformatics needs to take into account the local background versus the spot background. Once a spot is gridded, either the software can contour to the shape of the spot or can remain a fixed circle. If the spot remains a fixed circle, the mean fluorescence intensity (MFI) of the spot is determined by the average of the pixels within the enclosed fixed spot (see Note 9). 1.3. Allogeneic Antibodies Against H-Y Antigens Develop After Sex-Mismatched Transplantation
Our group focuses on identifying novel minor histocompatibility antigens (mHA) responsible for graft-versus-leukemic (GVL) effects and graft-versus-host disease (GVHD) after allogeneic HCT. Historically, allogeneic immune responses have been characterized as alloreactive T-cell clones where T-cells play key roles in posttransplant alloimmune responses, but T cell epitope determination remains laborious cell culture-dependent and HLA restricted (22). Our studies have shown that allogeneic B-cell responses against mHA influence clinical outcome (23). Specifically, allogeneic antibodies (allo-Ab) against mHA encoded on the Y chromosome, called H-Y antigens, develop in patients undergoing sex-mismatched HCT in association with both chronic GVHD development and persistent disease remission (24). Female lymphocytes develop devoid of the Y-chromosome and remain naïve to H-Y antigens. After these female lymphocytes are transplanted into male recipients, they recognize these H-Y antigens as foreign and mount a coordinated T- and B-cell immune response (25). Our studies have demonstrated that allogeneic antibodies develop against five H-Y antigens, namely,
88
Wadia, Sahaf, and Miklos
EIF1AY, RPS4Y, DDX3Y, ZFY, and UTY, but not their 91–99% identical X-homologues EIF1AX, RPS4X, DDX3X, ZFX, and UTX, respectively (24). We developed an ELISA panel to characterize the frequency, intensity, and specificity of these H-Y allo-Abs (23). In order to further multiplex and more accurately quantify H-Y and H-X antibodies, we developed a custom H-Y recombinant antigen microarray. The microarray print and development method are described in this chapter. 1.4. H-Y Antigen Microarray Development
Our H-Y microarray presents five human H-Y and H-X proteins that include C-terminal 6xHis and V-5 epitopes printed in quadruplicate on nitrocellulose-coated glass slides. All proteins are expressed in E. coli and isolated using nickel affinity chromatography. Nickel-bound 6-his tagged proteins are eluted by imidazole competition (24). In addition to the custom proteins, positive and negative proteins are also printed on the slide. The positive control proteins printed are human IgG in three different concentrations (IgG1 < IgG2 < IgG3), which should give similar readings across different processing time points because printed IgG recognition depends on uniform secondary antibody application, processing, and flurochrome detection (Fig. 4). To assess and ensure the quality of the batch of slides printed, the first two, middle two, and last two slides are probed with anti-V5 antibody, which will detect all antigens (Fig. 4). In addition to the standard negative controls such as bovine serum albumin (BSA), printing buffer, and blank spots, we have selected an HIV protein (p24 subunit). HIV-p24 antigen is expressed and purified from E. coli in a similar manner to our custom proteins. Since our patients tested negative for HIV, the fluorescent reading obtained for HIV-p24 antigen is considered background nonspecific reactivity to E. coli proteins and HIV-p24 measurements are subtracted as background. In short, the methodology consists of blocking the slide with 1% BSA followed by addition of a plasma sample. After application of the secondary antibody, the slides are washed, dried, and scanned. We have observed agreement between duplicate spots within a slide (R2 = 0.96). In our studies, we identify allo-ab targets 1 year after transplantation. Thus, when a male patient undergoes hematopoietic cell transplantation (HCT) with a female donor, the male pretransplant plasma fails to detect H-Y antigens, but 1 year following HCT, 50% F→M HCT patients develop allo-Ab against at least one of the H-Y antigens (23, 24). Thus, the pretransplant plasma serves as its own internal control and one can monitor patients longitudinally for the development of an immune response against each of the H-Y antigens.
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
a
Anti-human IgG detection without patient sera
b
Anti-V5 detection alone
c
89
Anti-human IgG detection with patient sera
IgG1 IgG2 IgG3
Optimize printing and scanner detection using three different concentrations of IgG
Normalize recombinant antigen printing using V5 tag on proteins
Normalize inter-slide variation by normalizing IgG detection and then quantify specific IgG response to antigens
Fig. 4. Schematic representation of H-Y Microarray Detection. A schematic representation of an H-Y slide is shown in the figure. The H-Y slide in (a) is probed with anti-human IgG and thus only the three different concentrations of IgG are identified. This detection helps to visualize and optimize printing across various subarrays. (b) Is an H-Y slide probed with anti-V5 for antigen detection to normalize for antigens in a batch inter-slide and intra-slide (across subarrays). (c) Shows that when an H-Y slide is probed with patient plasma, all the IgG spots and a few H-Y antigens are identified as hits. Using the fluorescence intensity units of IgG one can normalize for processing and scanning the arrays in one batch.
Demonstrating this concept schematically, Fig. 4 shows three slides printed with different (H-Y) antigens and three different concentrations of IgG printed in two subarrays. Panel A shows recognition of only IgG spots when the slide is probed with antihuman IgG Ab. These spots are preferentially printed at the beginning of each subarray to visualize correct printing orientation and optimize printing. Panel B shows a slide probed with only anti-V5 Ab that recognizes all the H-Y antigens tagged with V5 epitope tag, but not the IgG spots. This data helps normalize antigen printing across batches and subarrays. Thus, when patient sera/plasma is applied, all IgG spots are recognized in addition to patient-specific H-Y antigens (Panel C). Figure 5 represents H-Y arrays that were printed and probed with anti-human IgG (Panel A) and anti-V5 Ab (Panel B). As mentioned, only IgG spots or all H-Y antigens are recognized. When a male patient plasma/sera with a female donor HCT is applied on the slide, specifically this patient recognizes DDX3Y, ZFY, and UTY (Panel C),
90
Wadia, Sahaf, and Miklos
Fig. 5. Representative H-Y Microarray Detection. (a) Shows a slide probed with anti-human IgG where three different concentrations of IgG (IgG1 < IgG2 < IgG3) in each subarray are identified. (b) Shows an H-Y slide probed with anti-V5 identifying all the H-Y antigens tagged with V5, and as expected, no immunoglobulins are detected. For our custom H-Y arrays, we also print infectious agent antigens such as VZV, Pneumococcus, and EBV and negative controls such as printing buffer and BSA, and also have some empty spots. Since these infectious antigens are not tagged with V5 (except for EBV ), these antigens and negative controls will not be recognized by the anti-V5 antibody. (c) Shows detection of all three IgG concentrations in each subarray along with DDX3Y, UTY, and ZFY as targets on the H-Y array when probed with sera from a male patient collected 1 year after undergoing HCT from a female donor. Detection of these same proteins was absent pretransplant (d).
whereas detection of these same proteins was absent pretransplant (Panel D). In addition to the H-Y antigens, viral antigens are also spotted on the slide and EBV is recognized by antibodies in the pretransplant and posttransplant plasma irrespective of the donor (Panel C and D).
2. Materials 2.1. Preparation of Sera/Plasma
1. Venous blood collection tubes (BD Vacutainer®) or (BD Vacutainer serum tubes). 2. Cryovials (Fisher Scientific). 3. Centrifuge (Beckman Coulter Inc, Allegra 6KR centrifuge).
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
91
4. 1× PBST pH 7.4 (1× PBS [Gibco, Invitrogen] + 0.1% Tween 20 [American Bioanalytical]). 5. 1 M EDTA (VWR International). 2.2. Recombinant Antigens for Printing
1. Anti-human IgG Alexa647 conjugate (Molecular probes, Invitrogen). 2. Purified Human IgG (Sigma). 3. H-Y proteins ORFs cloned into E. coli.
2.3. Bacterial Cell Culture and Purification of Recombinant Protein Expression
1. 2XYT broth (EMD Chemicals, Inc). 2. Ampicillin (Amp; BD Diagnostics). 3. IPTG (EMD Chemicals, Inc). 4. B-PER Bacterial Protein Extraction Reagent (Pierce Chemicals). 5. Imidazole (Sigma; 10 mM Imidazole). 6. Glycerol (EMD Chemicals). 7. Monobasic sodium phosphate (MP Biomedicals, LLC). 8. Dibasic sodium phosphate (Sigma). 9. Urea (Sigma). 10. NuPAGE® Sample reducing agent (10×; Invitrogen). 11. NuPAGE® MOPS SDS running buffer (20×; Invitrogen). 12. NuPAGE® LDS 4× LDS sample buffer (Invitrogen). 13. NuPAGE® Novex 4–12% Bis-Tris gel 1.0 mm, 12 well (Invitrogen). 14. Anti-V5-HRP antibody (Invitrogen). 15. TBS buffer (Invitrogen). 16. 1.5 ml Eppendorf tubes (Eppendorf). 17. Eppendorf 5415D centrifuge (Eppendorf). 18. BD Falcon™ disposable centrifuge tubes, polypropylene, conical bottom (BD Biosciences). 19. mm glass beads (BioSpec Products, Inc). 20. Milk powder. 21. RNAse A (Qiagen). 22. Beckman centrifuge (Beckman Rotor No. SW28). 23. Sorvall RC-5 centrifuge (Thermo Scientific). 24. Dri-Block® heaters (Techne, Ltd). 25. VWR® Vortex Mixer (VWR). 26. DU® 800 UV/Vis spectrophotometer (Beckman Coulter). 27. Sonicator (Fischer Scientific, Model 100). 28. Dounce Homogenizer.
92
Wadia, Sahaf, and Miklos
29. FPLC chromatographic system (GE Healthcare, Model AKTA). 30. Denatured Nickel affinity chromatography Lysis Buffer (500 ml): Make stock 10× Buffer B (500 ml): 5 M NaCl (146.1 g), 200 mM dibasic Na phosphate (26.81 g). Bring volume to 500 ml (may need heat). Add 6 M Guanidine HCl (285 g) to 50 ml of 10× Buffer B, pH 7.8. Bring volume to 500 ml (heat on stir plate to get guanidine into solution, adjust pH with 10 N NaOH). 31. Wash Buffer pH = 6.0 (500 ml): Make stock 10× Buffer A (500 ml): 5 M NaCl (146.1 g), 200 mM monobasic Na phosphate (13.7 g). Bring volume to 500 ml (may need heat). Make stock 10× Buffer B (500 ml): 5 M NaCl (146.1 g), 200 mM dibasic Na phosphate (26.81 g). Bring volume to 500 ml (may need heat). Add 36.9 ml of 10× Buffer A and 13.1 ml 10× Buffer B with 100 ml of 20% Glycerol and adjust the pH to 6.0. 32. 6 M Urea Buffer (500 ml): Make stock 10× Buffer A (500 ml): 5 M NaCl (146.1 g), 200 mM monobasic Na phosphate (13.7 g). Bring volume to 500 ml (may need heat). Make stock 10× Buffer B (500 ml): 5 M NaCl (146.1 g), 200 mM dibasic Na phosphate (26.81 g). Bring volume to 500 ml (may need heat). Add 36.9 ml of 10× Buffer A and 13.1 ml 10× Buffer B; to the mix, add 100 ml of 20% Glycerol and 180 g urea (6 M final concentration). Adjust the pH to 6.0. 33. Imidazole Elution Buffer (500 ml): Make stock 10× Buffer A (500 ml): 5 M NaCl (146.1 g), 200 mM monobasic Na phosphate (13.7 g). Bring volume to 500 ml (may need heat). Add 50 ml of 10× Buffer A to 100 ml of 20% Glycerol and Imidazole (34.04 g) (final concentration 1 M). Adjust the pH to 6.0. 34. Native Nickel affinity chromatography Lysis Buffer (500 ml): Make a stock of 10× Buffer A (500 ml): 5 M NaCl (146.1 g), 200 mM monobasic Na phosphate (13.7 g). Bring volume to 500 ml (may need heat). Make 10× Buffer B (500 ml): 5 M NaCl (146.1 g), 200 mM dibasic Na phosphate (26.81 g). Bring volume to 500 ml (may need heat). Add 50 mM Imidazole (1.7 g) to12 ml of 10× Buffer A and 38 ml of 10× Buffer B. Bring volumes to 500 ml and confirm pH 7.8.
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
93
35. Wash Buffer pH = 6.0 (500 ml): Make stock 10× Buffer A (500 ml): 5 M NaCl (146.1 g), 200 mM monobasic Na phosphate (13.7 g). Bring volume to 500 ml (may need heat). Make stock 10× Buffer B (500 ml): 5 M NaCl (146.1 g), 200 mM dibasic Na phosphate (26.81 g). Bring volume to 500 ml (may need heat). Add 36.9 ml of 10× Buffer A and 13.1 ml 10× Buffer B and adjust the pH to 6.0. 36. Imidazole Elution Buffer (500 ml): Make stock 10× Buffer A (500 ml): 5 M NaCl (146.1 g), 200 mM monobasic Na phosphate (13.7 g). Bring volume to 500 ml (may need heat). Add 50 ml of 10× Buffer A to Imidazole – 34.04 g (final concentration 1 M). Adjust pH to 6.0. 2.4. Printing the Purified Recombinant Proteins to Obtain Protein Microarrays
1. Slides: Precoated glass slides with nitrocellulose FAST Slide-1 Pad (Whatman, Inc.). 2. Proteins to be printed (1 mg/ml). 3. 384-well amplification plates (Nunc). 4. 1× PBS (Gibco, Invitrogen). 5. Contact protein printer (Bio-Rad, Model. ChipWriter Pro). 6. Stealth Micro spotting prints (Telechem International). 7. Stealth Microarray printhead for 32 pins (Telechem International).
2.5. Probing and Developing H-Y Recombinant Protein Microarrays
1. Four-well trays (quadriPERM four-chamber culture dish) (Greiner ISC Express). 2. 10× PBS (Gibco, Invitrogen). 3. Blocking Buffer: 3% Bovine Serum Albumin (BSA; Sigma), 1× PBS, 0.1% Tween 20. 4. Wash buffer: 1× PBS, 0.1% Tween 20. This buffer is made fresh. 5. Anti-human IgG Alexa647 conjugate (Molecular probes, Invitrogen) (diluted 1:1,000 in 1× PBST). 6. Anti-V5 FITC (Invitrogen) (diluted 1:100 in 1× PBST). 7. Eppendorf centrifuge (Fisher Scientific). 8. Lab rotator (Lab-Line Instruments).
2.6. Data Analysis
1. GenePix 4000B microarray scanner (Molecular Devices Corporation). 2. GenePix Pro software (Molecular Devices Corporation).
94
Wadia, Sahaf, and Miklos
3. Methods 3.1. Preparation of Plasma Samples
1. Once blood from either patients/control subjects is obtained using the sera collection tube or plasma collection tube, the tube is inverted gently 5 times. 2. Centrifuge the tubes for 10 min at 1,430 × g at 4°C. 3. Transfer all sera/plasma obtained in 1 ml aliquots and add 0.1 mM EDTA (final concentration) in each aliquot to avoid fibrin formation. 4. Store aliquots at −80°C until further use. 5. When ready to use sera/plasma, thaw the samples on ice and vortex the samples for 5 s. 6. Centrifuge samples at 16,100 × g for 10 min at 4°C and use sera/plasma from the top of the tube. 7. For protein microarrays, dilute the plasma samples (1:150) by taking 5 ml from the top of the centrifuged plasma tube (avoid dipping tips in the centrifuged plasma tubes below half the volume) and add this to 750 ml of 1× PBST.
3.2. Recombinant Antigen Design
1. 11 H-Y antigens were selected for our H-Y recombinant protein array, namely, EIF1AY, RPS4Y, ZFY, DDX3Y, and UTY, along with their X-homologues, EIF1AX, RPS4X, ZFX, DDX3X, UTX, respectively, and HIV-p24. 2. Full-length cDNA for each gene was reversed transcribed from male peripheral blood mononuclear cells and polymerase chain reaction (PCR)-amplified with primers derived from GenBank sequences (23). 3. Each gene was Topo-cloned (Invitrogen, Carlsbad, CA) and expressed with C-terminal V5 epitope tag and 6 histidine residues in E. coli (pET-Dest42) and female-derived 293 cell line (pcDNA-Dest40). 4. Protein HIV-p24 is encoded by the second open reading frame (ORF) proteolytically processed from the Gag-pol polypeptide in vivo during HIV infection. The HIV-p24 was expressed in E. coli (pET-Dest42) and was purified in similar fashion. Since all patients/donors were previously screened for antibody to HIV-p24 and were known to be negative, reactivity with recombinant HIV-p24 was used as a negative control on our protein microarray and was subtracted from each patient’s H-Y protein measurement after probing the array with patient/donor sera. 5. These proteins are engineered such that they are tagged with 6-Histidine tag (6xHis). After cell lysis, the proteins are purified by fast protein liquid chromatography (FPLC) using a
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
95
nickel agarose column. In the FPLC column, the purification steps typically are binding, washing, and elution. The nickel agarose column only binds proteins tagged with 6xHis tag. These bound proteins are washed and eluted using imidazole. Imidazole being structurally similar to histidine competes with the 6xHis-tagged proteins for available nickel sites. By stepwise increasing the concentrations of imidazole, the proteins get displaced from the column, thus giving a pure recombinant protein solution. 6. H-Y proteins are also C-terminally tagged with a V5 epitope tag. The V5 epitope tag is a short series of amino acids (GKPIPNPLLGLDST) that is not usually cross-reactive with mammalian sera. The tag facilitates detection of proteins in cell lysates or to detect eluted protein purity after FPLC for recombinant proteins using western blots. 3.3. Bacterial Cell Culture and Purification of Recombinant Protein Expression
Day 1: 1. Streak an LB-Ampicillin (LB-Amp) plate with cells containing the vector for the corresponding H-Y protein of interest. 2. Incubate the plates for 18 h or overnight at 37°C for colony formation. Day 2: 1. Select five distinct colonies from LB-Amp plates. 2. Inoculate each colony in 2 ml cultures of 2XYT liquid broth containing 50 ml/ml Ampicillin (2XYT + Amp; see Note 10). 3. Incubate inoculated cultures for 18 h or overnight at 37°C with constant agitation (rotating shaker). Day 3: Perform a miniinduction protocol to choose the clone expressing the maximum protein from the five inoculated culture tubes: 1. Inoculate 100 ml of the incubated culture into 1 ml of fresh media (2XYT + Amp). Preserve the rest of the inoculated cultures from Day 2 at 4°C. 2. Incubate the cultures at 37°C for 1 h. 3. Induce expression of protein in cultures for 1–2 h using 1 mM IPTG (10 ml from a 100 mM IPTG stock in 1 ml media). 4. Harvest cells by pelleting in a 1.5-ml Eppendorf at 16,100 × g for 4 min. Discard the supernatant. 5. Prepare samples for an SDS-PAGE. {Add water (65 ml/ sample), 4× LDS buffer (25 ml/sample), and 10× reducing agent (10 ml/sample), along with ten 1.0 mm glass beads to lyse the cells}.
96
Wadia, Sahaf, and Miklos
6. Vortex continuously at high speed for 1 min, then place samples in a heat block at 80°C for 10 min, and again vortex for another 1 min. 7. Centrifuge tubes for 10 s at 3,060 × g and add 15 ml of prepared samples in each lane of a 4–12% Bis-Tris Gel (see Note 11). 8. Proceed with SDS-PAGE (4–12% Bis-Tris Gel in MOPS running buffer). 9. Western Blot the proteins to determine the most robust protein expressor clone. 10. Detect protein using Anti-V5 HRP antibody (1:5,000 in 5% milk in 1× TBS) and visualizing the blots with ECL. 11. The clone with the maximum level of protein expression is used for 4 L scale-up preparation on day 4. 12. Add 200 ml of the selected high protein expression clone to each four 100 ml cultures (2XYT + Amp). 13. Incubate for 18 h or overnight at 37°C. Day 4: A. Proteins obtained under Denatured Condition Protocol 1. Add each of the saturated 100 ml cultures into 1 L of fresh media (2XYT + Amp). 2. Cells are incubated at 37°C and optical density (OD) is checked after 1 h periodically until OD reading of 0.7–0.8 is achieved (see Note 12) 3. Induce each flask with a final concentration of IPTG as 1 mM. 4. Grow cultures for the next 2 h. 5. Centrifuge cultures at 2,050 × g for 20 min at 4°C in 225 ml conical BD Falcon centrifuge tubes. 6. After obtaining cell pellets, discard the supernatants and continue working on ice to prevent protein degradation (see Note 13). 7. Wash pellets with increasing concentrations of imidazole in 80 ml lysis buffer (BPER) and 20 ml of DNAse 8. From the 80 ml resuspended pellet, add 40 ml into a homogenizer. 9. Dounce (15 strokes with a tight pestle) and then split the 40 ml into two centrifuge tubes with 20 ml each. Repeat with the other remaining 40 ml of resuspended pellets. 10. Centrifuge at 31,780 × g for 15 min in Sorvall centrifuge. 11. Discard the supernatant and keep inclusion body pellets (protein is in pellets).
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
97
12. Resuspend in 80 ml 50/50 BPER and water with same concentration of imidazole as in step 7, and homogenize as above. 13. Centrifuge at 16,100 × g for 15 min. 14. Discard the supernatant and keep inclusion body pellets. 15. Resuspend in 40 ml 1:10 BPER:water with same concentration of imidazole as in step 7, homogenize, and centrifuge again (two tubes with 20 ml each). 16. Resuspend and combine pellets for final spin in 15 ml Denatured Nickel affinity chromatography 6 M Guanine HCl lysis buffer at 37°C to improve solubilization. 17. Dounce homogenize (15 strokes) the pellets and transfer the homogenized mixture to ultracentrifuge tubes. 18. Centrifuge tubes in high speed centrifuge, SW28 swinging bucket rotor, at 141,000 × g for 30 min. 19. Collect supernatant to run on an FPLC column. Day 4: B. Proteins obtained under Native Condition Protocol 1. Follow steps 1–6 as described in section Day 4: A: Proteins obtained under Denatured Condition Protocol. 2. Suspend all pellets in a total of 30 ml of native lysis buffer. 3. Add lysozyme to a final concentration of 1 mg/ml and incubate on ice for 30 min. 4. Sonicate on ice. 5. Add RNAse A (1:1,000 dilution) and DNAse I (1:5,000 dilution) and incubate on ice for 10–15 min. 6. Centrifuge in Sorvall centrifuge at 100,000 × g for 5 min. 7. Using the supernatant, run it through a high speed centrifuge, SW28 swinging bucket rotor at 141,000 × g for 30 min. 8. Collect supernatant to run on an FPLC column and extraction is performed using a native nickel extraction protocol. Day 5: 1. HIV-p24, EIF1AY, and EIF1AX proteins follow native nickel affinity chromatography protocol and all other H-Y proteins require denatured nickel affinity chromatography processing when synthesized in E. coli. All proteins were solubilized in their respective buffers and subsequently purified by nickel affinity chromatography. 2. Apply supernatants to a 10-ml Nickel affinity column that has been equilibrated in 6 M Guanidine lysis buffer or native nickel affinity chromatography lysis buffer based on the protein extraction protocol.
98
Wadia, Sahaf, and Miklos
3. For proteins using the denatured protocol, the lysis buffer is then exchanged with 6 M Urea solution over an eight-column volume gradient. For proteins using the native affinity nickel chromatography protocol, the 6 M Urea step is absent. (Instead of imidazole, a reduction in pH can also be used as an alternative to elute proteins). 4. The 6 M urea solution (denatured protocol) or the native lysis buffer (native affinity method) is washed with a wash buffer (pH = 6.0). 5. The proteins are eluted using an Imidazole elution buffer (containing 500 mM sodium chloride, 20 mM sodium phosphate, pH 6.0, and 20% glycerol) with increasing step gradients of Imidazole. 6. We use FPLC automation to run this protocol overnight and collect 48 fractions of eluted proteins of various concentrations. 7. The use of 6xHis tags and nickel affinity columns allows onestep protein purification, under either native or denaturing conditions, from dilute solutions and crude lysates, and also does not depend on the 3-D structure of the protein or the 6xHis tag. Day 6: 1. The fractions collected are tested using SDS-PAGE followed by Western blotting and proteins are detected using Anti-V5 HRP antibody. 2. The highest concentration fractions are reserved for microarray printing at 1 mg/ml. 3.4. Printing the Purified Recombinant Proteins to Obtain Protein Microarrays
1. A 384-well plate with purified proteins is prepared where 10 ml of each protein (1 mg/ml) is placed in wells separated by a single column and two rows. The moisture control in the printing chamber must be optimized to minimize solvent evaporation preventing antigen concentration flux. 2. 100 slides are placed in the correct orientation on the protein printer platform and immobilized using vacuum. 3. The printer is programmed to load 0.2 ml in the solid pin attached to the printer from the 384-well plate to spot proteins. 10 ml protein/well in the 384-well plate prints 100 slides. 4. The first two, middle two, and the last two slides after the print run are tested and processed using anti-V5 FITC to ensure all the proteins are printed and in the correct concentrations.
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
3.5. Probing and Developing H-Y Recombinant Protein Microarrays 3.5.1. Probing with Anti-V5 to Detect Printed Protein Spots
99
1. All steps are carried out at room temperature. 2. Add 5 ml of wash buffer in one chamber of the four-well tray. 3. Prewet the H-Y protein microarray for 5 min at room temperature with gentle agitation. 4. Remove wash buffer by aspiration either using vacuum suction or with the aid of a pipette. 5. Add 5 ml of blocking buffer and incubate for 2 h at room temperature with gentle agitation. 6. Remove the block buffer by aspiration. 7. Dilute FITC-labeled anti-V5 antibody 1:1000 (v/v) in 1× PBST buffer. 8. Add 5 ml of the diluted antibody on the array without touching the array. 9. Incubate at room temperature for 2 h without any agitation. 10. Place the slide in sterile 50 ml conical tube containing 25 ml wash buffer. 11. Wash the slide for 10 min at room temperature with gentle agitation with wash buffer. 12. Discard the wash buffer and repeat two more times. 13. Place the protein microarray in a sterile, dry, 50 ml conical tube and centrifuge the tube at 228 × g for 10 min at room temperature.
3.5.2. Probing with Human Serum/Plasma
1. All steps are carried out at room temperature. 2. Add 5 ml of wash buffer in one chamber of the four-well tray. 3. Prewet the H-Y protein microarray for 5 min at room temperature with gentle agitation (see Note 14). 4. Remove wash buffer by aspiration either using vacuum suction or with the aid of a pipette. 5. Add 5 ml of blocking buffer and incubate for 2 h at room temperature with gentle agitation. 6. Remove the block buffer by aspiration. 7. Dilute plasma 1:150 (v/v) in 1× PBST buffer as described in Subheading 3.1. 8. Add 150 ml of the diluted plasma on the array without touching the array. 9. Using forceps, place a Lifter Slip on the array to cover the membrane area, ensuring that there are no air bubbles trapped between the membrane and lifter slip. 10. Incubate at room temperature for 2 h without any agitation.
100
Wadia, Sahaf, and Miklos
11. Place the slide in sterile 50 ml conical tube containing 25 ml wash buffer. 12. Carefully remove the lifter slip after it detaches from the slide in the conical tube, making sure not to touch the surface of the protein microarray, and discard the lifter slip. 13. Wash the slide for 10 min at room temperature with gentle agitation. 14. Discard the wash buffer and repeat two more times. 15. Remove the H-Y protein microarray and place it in a new four-well tray chamber. 16. Add 5 ml of the secondary detection antibody diluted in wash buffer in the chamber without touching the surface of the slide. 17. Incubate for 1 h at room temperature without any agitation. 18. Place the slide in a sterile 50 ml conical tube containing 25 ml wash buffer. 19. Wash the slide for 10 min at room temperature with gentle agitation. 20. Discard the wash buffer and repeat two more times. 21. Place the protein microarray in a sterile, dry, 50 ml conical tube and centrifuge the tubes at 228 × g for 10 min at room temperature. 22. Ensure that the slides are completely dry (there should be no translucent areas) and place the slides in a slide container box in the dark at room temperature until they are read. Ideally, the slides should be read within 24 h of probing, processing, and drying the slide. 3.6. Data Analysis
1. Place the slides in the microarray slide holder of the scanner with the side containing the protein printed spots facing the laser source. 2. Scan the slides using wavelength 488 for slides probed using anti-V5 FITC or wavelength 647 for slides probed with patient sera/donors. 3. Photomultiplier tube (PMT) gain should be 600 U, laser power should be 100%, the pixel size and focus position should be 10 and 0 mm, respectively. 4. In order to insure maximum and reproducible dynamic range and since the maximum reading value for the reader is 60,000 fluorescent units, the highest IgG concentration is set at 55,000 signal by adjusting the PMT voltage. In this way, one can avoid being off-scale for readings and should calibrate the reader on daily basis, ensuring the reproducibility of the assay between readings.
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
101
5. Using GenePix Pro software (Molecular Devices Corporation) package, a gal file is created indicating the position of each protein spotted on the array (see Note 15). 6. The protein microarray is scanned at the selected wavelength, and after gridding the obtained tiff image file with the gal file, the results for MFI readings for each spot are obtained and results of patients and controls can be compared. For each spot, a local background around the diameter of the spot is calculated and may be subtracted from the MFI of the spot accounting for the background. This background will reflect and account for the unwanted background shifts that may occur in different location on the nitrocellulose slide. 7. The Genepix reading of the mean fluorescent intensity (MFI) value for each spot is created in the GenePix Result file or GPR file. In this file the fluorescence intensity of each spot is presented and annotated in table format. Median, mean standard error, local background for each spot is also tabulated and could be used for control and calculation purposes.
4. Notes 1. Protein purity is essential to avoid false positive results. 2. Most genes are polymorphic encoding single nucleotide polymorphisms and splice variations resulting in multiple isoforms. For custom arrays, using other techniques, one can determine the isoform of the antigen to be printed on the array. Alternatively, all isoforms obtained can be printed on the array. However, one may not necessarily know the isoform printed on a commercial array. 3. Care needs to be taken to remove the native stop codon from the ORF in order for the plasmid-encoded C-terminus tag to be incorporated in the protein. 4. The antigens to be printed should be in the right buffers with optimal pH and the buffers containing preservative agents like azide or stabilizing agents like glycerol should be avoided. Maintenance of native protein conformation requires nondenaturing isolation. 5. In our experience, a quill pin is preferred because while printing, solid pins can damage the coating surface of the slide, or if inadequate printing material is present in the pin, the spots will be doughnut-shaped when the slide is processed. 6. Ideally, protein concentration should be uniform. Protein stability should be assessed to determine shelf storage conditions. Shelf life of antigens printed (stability) can be monitored
102
Wadia, Sahaf, and Miklos
using epitope detection, e.g., using anti-V5 antibody against the V5 tag bound to each antigen. 7. If irregular spots are obtained, each operator may grid the spot differently leading to various and nonreproducible results. For example, if the spots are “comet” shaped, one can grid the whole spot or part of the spot and leave the tail of the comet out and these could lead to significantly different results for the same printed antigen spot. 8. In order to avoid concentration through evaporation of antigens from the pins while printing, ensure the printing takes place in a defined humidity environment. 9. If the intensity of the processed antigen spot is unevenly distributed within the circle, the MFI of the spot will be low as compared to the MFI when analysis is done using software that fits the contour of the processed spot, and hence, we prefer using software which takes the shape of the processed spot. 10. Do not add Ampicillin until after autoclaving. Add Ampicillin when the temperature of the media has reached room temperature. 11. Take less viscous surface of supernatant and add about 15 ml of sample in a gel lane. If the sample is too viscous (won’t sit at the bottom of the well), vortex and centrifuge again. 12. Absorbance is measured at 595 nm with the blank being fresh 2XYT + Amp media. 13. Keep pellets on ice and/or at 4°C at all times until the lysis buffer step. 14. Make sure that there are no white patches of nitrocellulose surface seen and the slide is completely wet, else it will cause white patches of uneven staining for further steps. 15. A file linking the spot position with its identity is created and usually is called a gal file (GenePix Array List). The software to create this file is usually provided with the GenePix scanner, but alternatively one can create a similar file as a tab delimited file in Microsoft Excel.
Acknowledgments This work was supported by NIH R21 HL084318-01A1 and P01 CA049605. We would like to thank Mrs. Fang Wu for her help in processing the H-Y slides. We would also like to thank Dr. John Coller, Director of the Stanford Protein Array Core Facility, for his advice and printing of the H-Y recombinant antigen arrays.
Recombinant Antigen Microarrays for Serum/Plasma Antibody Detection
103
References 1. Spence RP, Wright V, Ala-Aldeen DA, Turner DP, Wooldridge KG, James R (2008) Validation of virulence and epidemiology DNA microarray for identification and characterization of Staphylococcus aureus isolates. J Clin Microbiol 46:1620–1627 2. Jin D, Qi H, Chen S, Zeng T, Liu Q, Wang S (2008) Simultaneous detection of six human diarrheal pathogens by using DNA microarray combined with tyramide signal amplification. J Microbiol Methods 75:365–368 3. Jandu N, Ho NK, Donato KA, Karmali MA, Mascarenhas M, Duffy SP, Tailor C, Sherman PM (2009) Enterohemorrhagic Escherichia coli O157:H7 gene expression profiling in response to growth in the presence of host epithelia. PLoS One 4:e4889 4. Mengual L, Burset M, Ars E, Lozano JJ, Villavicencio H, Ribal MJ, Alcaraz A (2009) DNA microarray expression profiling of bladder cancer allows identification of noninvasive diagnostic markers. J Urol 182:741–748 5. Li L, Wadia P, Chen R, Kambham N, Naesens M, Sigdel TK, Miklos DB, Sarwal MM, Butte AJ (2009) Identifying compartment-specific non-HLA targets after renal transplantation by integrating transcriptome and “antibodyome” measures. Proc Natl Acad Sci U S A 106: 4148–4153 6. Sakuishi K, Oki S, Araki M, Porcelli SA, Miyake S, Yamamura T (2007) Invariant NKT cells biased for IL-5 production act as crucial regulators of inflammation. J Immunol 179:3452–3462 7. Imamichi T, Yang J, Huang DW, Brann TW, Fullmer BA, Adelsberger JW, Lempicki RA, Baseler MW, Lane HC (2008) IL-27, a novel anti-HIV cytokine, activates multiple interferon-inducible genes in macrophages. Aids 22:39–45 8. Hwang SS, Kim YU, Lee W, Lee GR (2009) Differential expression of nuclear receptors in T helper cells. J Microbiol Biotechnol 19:208–214 9. De Vos J, Hose D, Reme T, Tarte K, Moreaux J, Mahtouk K, Jourdan M, Goldschmidt H, Rossi JF, Cremer FW, Klein B (2006) Microarray-based understanding of normal and malignant plasma cells. Immunol Rev 210:86–104 10. Yamamoto R, Sakamoto T, Nishi S, Sakai M, Morinaga T, Tamaoki T (1990) Expression of human alpha-fetoprotein in yeast. Life Sci 46:1679–1686 11. Carruthers AM, Warner AJ, Michel AD, Feniuk W, Humphrey PP (1999) Activation
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
of adenylate cyclase by human recombinant sst5 receptors expressed in CHO-K1 cells and involvement of Galphas proteins. Br J Pharmacol 126:1221–1229 Predki PF, Mattoon D, Bangham R, Schweitzer B, Michaud G (2005) Protein microarrays: a new tool for profiling antibody cross-reactivity. Hum Antibodies 14:7–15 Mattoon D, Michaud G, Merkel J, Schweitzer B (2005) Biomarker discovery using protein microarray technology platforms: antibodyantigen complex profiling. Expert Rev Proteomics 2:879–889 Wadia PP, Coram M, Armstrong RJ, Mindrinos M, Butte AJ, Miklos DB (2010) Antibodies specifically target AML antigen NuSAP1 after allogeneic bone marrow transplantation. Blood 115(10):2077–2087 Marina O, Biernacki MA, Brusic V, Wu CJ (2008) A concentration-dependent analysis method for high density protein microarrays. J Proteome Res 7:2059–2068 Stillman BA, Tonkinson JL (2000) FAST slides: a novel surface for microarrays. Biotechniques 29:630–635 MacBeath G, Schreiber SL (2000) Printing proteins as microarrays for high-throughput function determination. Science 289: 1760–1763 Angenendt P, Glokler J, Murphy D, Lehrach H, Cahill DJ (2002) Toward optimized antibody microarrays: a comparison of current microarray support materials. Anal Biochem 309:253–260 Kusnezov W, Pulli T, Witt O, Hoheisel JO (2005) Solid Supports for protien microarrays and related devices. Jones and Bartlett, Sudbury, pp 247–283 Paborsky LR, Dunn KE, Gibbs CS, Dougherty JP (1996) A nickel chelate microtiter plate assay for six histidine-containing proteins. Anal Biochem 234:60–65 Zhu H, Bilgin M, Bangham R, Hall D, Casamayor A, Bertone P, Lan N, Jansen R, Bidlingmaier S, Houfek T, Mitchell T, Miller P, Dean RA, Gerstein M, Snyder M (2001) Global analysis of protein activities using proteome chips. Science 293:2101–2105 Goulmy E (1996) Human minor histocompatibility antigens. Curr Opin Immunol 8:75–81 Miklos DB, Kim HT, Zorn E, Hochberg EP, Guo L, Mattes-Ritz A, Viatte S, Soiffer RJ, Antin JH, Ritz J (2004) Antibody response to DBY minor histocompatibility antigen is
104
Wadia, Sahaf, and Miklos
induced after allogeneic stem cell transplantation and in healthy female donors. Blood 103: 353–359 24. Miklos DB, Kim HT, Miller KH, Guo L, Zorn E, Lee SJ, Hochberg EP, Wu CJ, Alyea EP, Cutler C, Ho V, Soiffer RJ, Antin JH, Ritz J (2005) Antibody responses to H-Y minor histocompatibility antigens correlate
with chronic graft-versus-host disease and disease remission. Blood 105:2973–2978 25. Zorn E, Miklos DB, Floyd BH, Mattes-Ritz A, Guo L, Soiffer RJ, Antin JH, Ritz J (2004) Minor histocompatibility antigen DBY elicits a coordinated B and T cell response after allogeneic stem cell transplantation. J Exp Med 199:1133–1142
Chapter 8 SPOT Synthesis as a Tool to Study Protein–Protein Interactions Dirk F.H. Winkler, Heiko Andresen, and Kai Hilpert Abstract Peptide arrays are a widely used tool in proteomic research for investigations of drug development and molecular interactions including protein–protein or protein–peptide interactions. Most peptide synthesis techniques are able to simultaneously synthesize only up to a few hundred single peptides. Using the SPOT™ technique, it is possible to synthesize and screen in parallel up to 8,000 peptides or peptide mixtures. In addition, such peptides can be released from the membrane and transferred onto peptide microarrays. Here we present protocols for the peptides synthesis on cellulose including the preparation of different cellulose membranes and easy-to-use detection methods on these peptide macroarrays. In addition, a protocol to produce and screen peptide microarray using the SPOT technology is provided. Key words: Spot synthesis, Peptide array, Screening, Cellulose membranes, Protein–protein interaction, Protein–peptide interaction, Microarray
1. Introduction Last year we are celebrating the 20th anniversary of the first presentation of the SPOT™ method by the team of Ronald Frank in 1990 (1). Since that presentation, the SPOT synthesis has proven to be a powerful tool for the study of protein–protein interactions (2–4). This method is inexpensive, easy to perform, and can be established in virtually any laboratory. SPOT synthesis is a special type of parallel solid-phase peptide synthesis on planar, porous surfaces – most commonly on cellulose membranes. The activated coupling solutions are positionally addressed and delivered in small drops to distinct points on the membrane forming a pattern of small spots. Over several coupling cycles, peptides are built upon these spots (5, 6). Using the automated SPOT synthesis,
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_8, © Springer Science+Business Media, LLC 2011
105
106
Winkler, Andresen, and Hilpert
it is possible to synthesize and screen up to 8,000 peptides, peptide mixtures, or other organic compounds on membranes of the size of a 96-well plate (8 × 12 cm) up to about letter size (19 × 28 cm) (7–10). The screening of peptide libraries synthesized using the SPOT technique is usually carried out while the peptides are bound to the membrane (11–13). The detection of bound proteins is similar to a Western Blot. In order to test the peptides in solution, it is also possible to release the peptides from the membrane. These free peptides can be further used either directly for types of screening (14, 15) or for the preparation of peptide microarrays (16, 17). SPOT synthesis protocols presented in this publication can be carried out manually or using automation. Manual SPOT synthesis is most convenient for rather relatively small numbers of peptides (up to 100) and large pipetting volumes (>0.5 ml). For more complex libraries, it is recommended to perform the synthesis semi- or fully-automated (18). Here we describe the standard procedures for SPOT synthesis of peptides. Protocols for the synthesis of modified peptides, for example, cyclic peptides or such with side-chain modifications, are described elsewhere in the literature (6). The following protocols are applicable not only to the use of natural amino acid building blocks, but can also be adopted for the use of unnatural amino acids and several other organic building blocks, e.g., peptide nucleic acid (PNA) monomers and peptoidic elements (19–22). The SPOT method is not restricted to the production of single membrane-bound peptide macroarrays, but can also provide soluble peptides for the production of multiple identical copies of peptide microarrays. In fact, the peptide amount yielded from a single spot of a peptide macroarray is sufficient to produce several hundred identical peptide spots in microarray formats. This is particularly attractive when many different samples have to be screened against the same set of peptides. For this reason, we also include protocols for the release of the peptides from the cellulose membranes, their attachment to chemically modified glass surfaces as well as for the analysis of protein–protein interactions with the corresponding peptide microarrays.
2. Materials Solvents necessary: 1. N,N¢-Dimethylformamide (DMF; toxic, flammable). 2. Methanol (MeOH; toxic, flammable). 3. Ethanol (EtOH; flammable). 4. N-Methylpyrrolidone (NMP).
SPOT Synthesis as a Tool to Study Protein–Protein Interactions
107
5. Diethylether (DEE; highly flammable). 6. Dichloromethane (methylene chloride, DCM; toxic, cancer hazard). 7. 1,4-Dioxane (flammable, suspected carcinogenic). 8. Dimethylsulphoxide (DMSO). The quality of solvents for the required washing steps should be of at least ACS. Solvents for dissolving reagents for the synthesis must be amine- and water-free. Due to possible decomposition under the influence of light, organic solvents, with the exception of MeOH and EtOH, should be stored in the dark. If not noted different elsewhere, the used water is always distilled/deionized. 2.1. Preparation of Cellulose Membranes and SPOT Synthesis of Macroarrays
1. Membranes are prepared from filter paper Whatman 50 or Whatman 540 (Whatman) (20, 23). Several ready-to-use cellulose membranes are commercially available (e.g., from AIMS Scientific, Intavis, or Sigma-Genosys). 2. Amino-acid ester-linked membranes: diisopropylcarbodiimide (DIPC, DIC; Fluka; very toxic), N-Methylimidazole (NMI; Sigma; flammable, corrosive), and Fmoc-b-alanine (EMD Biosciences) or Fmoc-glycine (GL Biochem) (see Note 1). 3. Amino-alkyl ether-linked membranes: 70% perchloric acid (Alfa Aesar; oxidizing, corrosive), epibromohydrine (Fluka; toxic), 1,3-diaminopropane (Alfa Aesar; toxic, corrosive, flammable), 4,7,10-trioxa-1,13-tridecanediamine (Fluka; corrosive), and sodium methylate (sodium methoxide; Fluka; highly flammable, corrosive). 4. Staining solution: 0.002% bromophenol blue (BPB; Sigma) in MeOH (20 mg in 1 l). 5. Coupling reagents: DIC and N-hydroxybenzotriazole (HOBt; EMD Biosciences; flammable). Coupling reagents are only necessary when no preactivated amino acid derivatives are used (see Note 2). 6. Nonpreactivated amino acids with protection groups according to the Fmoc strategy (24, 25) (EMD Biosciences and GL Biochem); preactivated amino acid derivatives with protection groups according to the Fmoc strategy, e.g., pentafluorophenyl esters (OPfp ester; EMD Biosciences and Bachem) (26) (see Note 2). 7. Piperidine (Sigma; toxic, highly flammable). 8. Capping solution: 2% acetic anhydride (Sigma; flammable, corrosive) in DMF. For second consecutive treatment, 2% ethyl-diisopropylamine (DIPEA, DIEA; Sigma; corrosive) can be added to deprotonate amino groups and buffer the acetic acid generated during reaction.
108
Winkler, Andresen, and Hilpert
9. Deprotection solution A: Trifluoroacetic acid (TFA; VWR; corrosive) (v/v) containing 5% dist. water (v/v), 3% triisopropylsilane or triisobutylsilane (TIPS or TIBS; Fluka; irritant) (v/v), 1% phenol (Sigma; toxic, corrosive) (w/v) (Important! see Note 3). 10. Deprotection solution B: 60% TFA (v/v), 3% TIPS or TIBS (v/v), 2% dist. water (v/v), 1% phenol (w/v), 34% DCM (v/v) (Important! see Note 3). 2.2. Peptide Modifications for the Preparation of Microarrays and Cleavage of Peptides from the Cellulose Membrane
1. N1-(9-Fluorenylmethoxycarbonyl)-1,13-diamino-4,7, 10-trioxatridecane-succinamic acid (Fmoc-TTDS-OH; Iris Biotech).
2.3. Preparation of Microarrays
1. Spotting buffer: 0.1 M sodium acetate buffer pH 5.0 containing 10 mM sodium cyanoborohydride (very toxic!) and 10% (v/v) glycerol (see Note 4).
2. Coupling reagents: DIC, HOBt, and NMI. 3. 4-(Fmoc-hydrazino)-benzoic acid (Fmoc-HBA; Bachem). 4. Ammonia gas (Air Liquide; irritant, corrosive).
2. Microscope glass slides with aldehyde surface coating (Schott Nexterion®) or epoxide coating (Corning Life Sciences) (see Note 5). 3. Blocking Buffer: 0.5% (v/v) Nonidet P40 (Sigma-Aldrich), 5% (w/v) skim milk in 1 M Tris–HCl pH 8.5. The skim milk should be freshly added. Homogenization is facilitated by warming the buffer to 60°C and application of ultrasound. 4. Petri dishes with 145 mm diameter (Greiner Bio-One), parafilm, filter paper. 5. QuadriPERM® 4-compartment cell culture plates (Greiner Bio-One). 6. Nitrogen gas. 2.4. Detection Methods
1. 50 mM Tris-buffered saline (TBS), pH 8.0. 2. 50 mM TBS with 0.2% (v/v) Tween (TBS-T). 3. Blocking buffer: 5% (w/v) Casein or skim milk (Sigma) and 4% (w/v) Sucrose (Sigma) in TBS-T (see Note 6). 4. Probing solution (protein, antibody): Depending on the estimated affinity, 0.1–10 mg/ml protein/antibody in blocking buffer. For a lower estimated affinity, a higher concentration of the protein should be used (see Note 7). For protein mixtures (blood, plasma, cell extracts, etc.), an estimation of the concentration of the target protein is necessary.
SPOT Synthesis as a Tool to Study Protein–Protein Interactions 2.4.1. Detection of Bound HRP-Labeled Protein Using Chemiluminescence
109
1. Horseradish peroxidase (HRP, POD)-labeled or unlabeled antibody against the probed peptides/protein: depending on the estimated affinity 0.1–1.0 mg/ml (in blocking buffer) (see Notes 7 and 8). 2. 10 ml mixture for chemiluminescence detection (see Note 9): 1 ml 1 M Tris–HCl pH 8.5; 22 ml 80 mM p-coumaric acid in DMF; 50 ml 250 mM luminol in DMF; 3 ml 30% H2O2; 9 ml water; Mix the p-coumaric acid and the luminol with the corresponding amount of Tris–HCl. Add to this mixture the corresponding amount of water. Activate this solution immediately before use by mixing with hydrogen peroxide.
2.4.2. Detection of Bound HRP-Labeled Protein Using Staining
2.4.3. Detection of Bound AP-Labeled Protein Using Staining
1. For the usage of protein solutions see step 1 in Subheading 2.4.1. 2. Mixture for staining (10 ml): 5 mg 4-chloro-1-naphthol dissolved in 1.7 ml methanol, 2.5 ml 200 mM Tris–HCl pH 7.4 (24.2 g/l), 100 mg NaCl, 5.8 ml H2O, 5 ml 30% H2O2. First, dissolve the NaCl in the above amount of water and Tris buffer (pH 7.4), followed by adding the methanolic chloronaphthol solution. Shortly before use, mix this solution with hydrogen peroxide. 1. Alkaline phosphatase (AP)-labeled or unlabeled antibody against the probed protein: depending on the estimated affinity 0.1–1.0 mg/ml (in blocking buffer) (see Notes 7 and 8). 2. NBT stock solution: 0.5% (w/v) nitrotetrazolium blue chloride (NBT; Sigma) in 70% aqueous DMF; this solution can be stored cold (6,000 × g for 15 min at 4°C. 8. Remove the supernatant from the cell debris and refrigerate at 2–8°C until needed. For long term storage (>1 week), keep frozen at −20°C. 9. Equilibrate a PD-10 Column with 25 cc of ProteoSep Start Buffer (SB) (see Note 5).
Native Antigen Fractionation Protein Microarrays
135
10. Add the 2.5 cc of the centrifuged sample to the PD-10 column and discard the eluent. 11. Elute and collect the lysis proteins with 3.5 cc of SB and dilute to 5.0 cc with SB. 3.1.2. Protein Fractionation: First Dimension – Chromatofocusing
1. Prepare the working buffers for the first dimension as follows: (a) Start buffer (SB) i. Warm to room temperature and sonicate for 5 min. ii. Using a calibrated pH meter, adjust the pH to 8.5 ± 0.1 using either a saturated solution of iminodiacetic acid (IDA) if the buffer is too basic or 1 M NH4OH if the buffer is too acidic (see Note 6). iii. Store in refrigerator (4–8°C) until needed. (b) Eluent buffer 1. Warm to room temperature and sonicate for 5 min. 2. Using a calibrated pH meter, adjust the pH to 4.0 ± 0.1 using either a saturated solution of IDA if the buffer is too basic or 1 M NH4OH if the buffer is too acidic. 3. Store in refrigerator (4–8°C) until needed. 2. Set the operating conditions for the HPLC system as follows (these settings are for an HPLC system with a single pump, injector, UV, and pH detectors): Column: HPCF 1D Column. Flow rate: 0.2 mL/min. Detection: UV 280 nm. Temperature: Ambient. Mobile phase: Start buffer: pH 8.5 ± 0.1. EB: pH 4.0 ± 0.1. 3. First dimension chromatofocusing procedure: (a) If using a Beckman ProteomeLab PF2D instrument, inject 5.0 mL of the PD-10 exchanged lysed sample using the default 2D fraction collection method preprogrammed into the PF2D computer. Loading 1–5 mg of total protein on the CF column is recommended (If using an AKTA Purifier or equivalent pH controlled HPLC instrument, before injection, flush the HPCF column with 100% HPLC grade water for 10CV (1CV = 0.87 mL), then equilibrate column with 30 CV of SB. After equilibration, inject 5.0 mL of the PD-10 exchanged lysed sample. Loading 1–5 mg of total protein on the CF column is recommended.).
136
Caiazzo et al.
(b) After injection, wash the CF column with 100% start buffer at a flow rate of 0.2 mL/min while monitoring the baseline absorbance. Collect this eluent wash as a single fraction since all proteins with pI’s >8.5 will elute from the column during this wash step. After the UV absorbance values return to baseline, stop the flow of Start Buffer eluent. This wash step will take ~10–30 min (see Note 7). (c) Before performing the pH gradient, flush the tubing and pump head with 100% EB (the flush volume depends on the system configuration) in order to facilitate starting of the pH gradient. Initiate start of 100% EB at 0.2 mL/ min for 95 min to perform the pH gradient. At ~70 min the pH of the eluent should be pH 4.0 ± 0.1. Collect fractions every 0.3 pH units using a fraction collector controlled by pH change. (d) After completing the EB run, wash the HPCF column with 10 column volumes of a 1.0 M sodium chloride 30% n-PrOH solution (filtered through 0.45 mm membrane filter). Collect the first 4 mL of this eluent as a fraction for analysis with the second dimension HPRP column. (e) Transfer the HPCF pH fractions from the fraction collector to appropriate vials for use with the HPRP HPLC autosampler for the second dimension HPRP analysis. Fractions should be stored at 2–8°C if the second dimension analysis will be delayed for more than ~8–10 h. (f) After the salt wash, wash the CF column with 10 column volumes of HPLC grade water. 3.1.3. Protein Fractionation: Second Dimension – Reversed Phase HPLC
1. If using a Beckman ProteomeLab PF2D instrument, inject 250 mL of each CF fraction collected from the first dimension fractions using the default 2D fraction collection method preprogrammed into the PF2D computer. Collect second dimension fractions in the 650 mL 96-well plates, collecting the effluent from the HPRP column with retention times between 10 and 22 min. If using another type of HPLC, use the following operating conditions (these setting are for an HPLC system with low dead volume mixing and 5,000 psi operation): Column: HPRP 2D Column Mobile phase: A: 0.1% TFA in water (HPLC grade) B: 0.08% TFA in acetonitrile (HPLC grade) Gradient: (a) 100% A for 2 min (b) 0–100% B in 30 min.
Native Antigen Fractionation Protein Microarrays
137
(c) Hold at 100% B for 5 min. (d) 100% B to 100% A in 2 min. (e) 100% A for 5 min. Flow rate: 0.75 mL/min Detection: UV 214 nm Temperature: 50°C Injection vol.: 250 mL injections are recommended for each HPCF fraction collected in the first dimension. Less can be injected if desired, but 250 mL is the recommended minimum injection volume for detecting low level expressed proteins present using standard HPLC equipment. 2. Save the Raw UV absorbance data or each HPRP analysis of the CF fractions for protein mapping and data analysis using the ProteoVue Software (Eprogen). 3.1.4. Microarray Printing and Blocking
The 96-well plates containing the second dimension fractions were used to fabricate the microarrays (see Note 8). 1. Add 50 mL of 40% glycerol in PBS to each well already containing a water/acetonitrile/TFA mixture. 2. Evaporate the plates using a SpeedVac centrifuge until only 20 mL of the glycerol remains. 3. Add 30 mL of PBS to each well to make a 40% glycerol/PBS print buffer and transfer into two sets of 384 well plates for printing. 4. Using a solid pin QArray2 (Genetix, Ltd.) print the arrays on a suitable support surface, in this case the arrays were printed on thin layer nitrocellulose PATH and Apix slides (Gentel Biosciences, Inc.) (see Note 9). 5. To allow the slides to dry to completion and prevent spot migrations, the printed slides should be set aside for at least 72 h prior to blocking them. To block the slides, immerse them in 1× Gentel Block Buffer for 1 h. Afterward, the slides can be air dried and stored in a sealed slide box under desiccating conditions until use.
3.2. Sample Preparation and Microarray Experiments 3.2.1. Purification of IgG Using Melon™ Gel Kit
The following steps may be carried out several days prior to the rest of the microarray experiment. Purified IgG can be stored for up to 1 week at 4°C. If IgG is to be stored for longer than 1 week, aliquots should be placed in a −20°C freezer for storage until use; avoid repeated freeze/thaw cycles. 1. Equilibrate the Melon™ Gel IgG Purification Support and Purification Buffer to room temperature (~30 min) and swirl the bottle containing the Purification Support (do not vortex) to obtain an even suspension. To ensure proper gel
138
Caiazzo et al.
slurry dispensing, use a wide bore or cut pipette tip to dispense 500 mL of gel slurry into a Handee™ Mini-Spin Column placed in a microcentrifuge tube. Swirl the bottle of gel slurry before pipetting each sample to maintain an even gel suspension. 2. Centrifuge the uncapped column/tube assemblies for 1 min at 3,000 × g, then remove the spin columns and discard the flow-through. 3. Add 300 mL of Purification Buffer to the column, pulse the centrifuge for 10 s and discard the flow-through. Repeat this wash once. Place the bottom caps on the columns. 4. Add 50 mL of each serum sample diluted 1:10 in 1× Melon™ Gel Purification Buffer to a column. Cap the columns and incubate for 5 min at room temperature with end-over-end rotation. 5. Remove the bottom caps from the columns, loosen the top caps and insert the spin columns into fresh 2 mL collection tubes. Then, centrifuge for 1 min at 3,000 × g to collect the purified antibody in the collection tubes. 6. Set up a new column corresponding to each sample that has been purified, and repeat steps 2–5 in order to further purify the collected IgG using fresh Melon™ Gel (see Note 10). 7. Measure the concentration of IgG in each purified sample using Pierce’s BCA™ Protein Assay Reagent Kit. 8. Dilute each sample to 1 mg antibody/mL by adding the appropriate volume of 1× Melon™ Gel Purification Buffer (see Note 11). The final amount of IgG should be at least 200 mg per sample. 3.2.2. Labeling of IgG with Fluorescent Dyes
The following steps may be carried out several days prior to the rest of the microarray experiment. However, Subheadings 3.2.2 and 3.2.3 must be carried out together so that unbound dye is removed before storage of the labeled IgG. Labeled IgG can be stored for a maximum of 1 week at 4°C protected from light. If labeled IgG is to be stored for longer than 1 week, aliquots should be protected from light in a −20°C freezer until use. Repeated freeze/thaw cycles should be avoided. Once the Cy™ dyes are reconstituted, they must be used immediately. 1. Set up and label one 0.5 mL microfuge tube for each sample (four tubes total: A-Cy™3, A-Cy™5, B-Cy™3, B-Cy™5). 2. Transfer 100 mg of the appropriate purified antibody (1 mg/mL) from Subheading 3.1 to the corresponding tube prepared in step 1. 3. Tap the bottom of the Cy™ Dye reagent vials against a hard surface to ensure that there is no dye in the caps and reconstitute
Native Antigen Fractionation Protein Microarrays
139
one vial of Cy™3 dye and one vial of Cy™5 dye by adding 110 mL of Clontech Labeling/Extraction Buffer to each vial. 4. Vortex the two dye vials for 20 s and briefly centrifuge to collect the reconstituted dyes at the bottom. 5. Add 4 mL of Cy™3 to each of the corresponding tubes from step 2 (see Note 12). 6. Add 2 mL of Cy™5 to each of the corresponding tubes from step 2 (see Note 12). 7. Vortex the four microfuge tubes gently and briefly centrifuge the microfuge tubes to collect the samples at the bottom. 8. Incubate the tubes for 90 min at 4°C protected from light. Mix the samples every 20 min by inverting the microfuge tubes. 9. Add 4 mL of Clontech Blocking Buffer to each tube. 10. Incubate the tubes for 30 min at 4°C protected from light. Mix the samples every 10 min by gently vortexing the microfuge tubes. 11. Proceed immediately with the removal of unbound dye. 3.2.3. Removal of Unbound Dye with Desalting Columns
As long as you work quickly, desalting can be completed at room temperature. Otherwise, if you have access to a cold room, we suggest you complete the procedure at 4°C. 1. Set up and label one 1.5 mL microcentrifuge tube and one Protein Desalting Spin Column for each dye labeled sample (see Note 13). 2. Twist off the bottom of each column and loosen the caps before placing each one in its collection tube and centrifuge each column at 1,500 × g for 2 min to remove the storage buffer. Note the side of each column where the compacted resin is slanted upward, and be sure to place the columns in the centrifuge with this area of the column facing outward in all subsequent centrifugations. 3. Transfer each desalting column to a fresh 2 mL microcentrifuge tube. 4. Add 400 mL 1× Clontech Desalting Buffer to each of desalting columns. Then, centrifuge the columns for 2 min at 1,500 × g. Discard the flow-through. Repeat this step once. 5. Blot the bottom of the columns against a laboratory tissue to remove excess liquid, and place the columns into a fresh, labeled 1.5 mL microcentrifuge tube. 6. Carefully apply the Cy™3 and Cy™5 labeled samples (~105 mL) directly onto the center of the resin bed of the corresponding column. Allow the samples to pass into the columns.
140
Caiazzo et al.
7. Centrifuge the columns for 2 min at 1,500 × g to collect the desalted, labeled antibodies. 8. At this point, purified, labeled IgG samples can be stored for future use, or used in Subheading 3.2.4 of a current experiment. 3.2.4. Antibody Array Incubation with Patient IgG
These microarray experiments utilize a dye swap procedure (see Note 14), which necessitates the use of two microarray slides. One slide will be incubated with the “Mix 1” samples (IgG A-Cy™3 and IgG B-Cy™5) and the other slide will be incubated with the “Mix 2” samples (IgG A-Cy™5 and IgG B-Cy™3). Be sure to record the barcodes printed on the microarray slides so that they can be identified as either Mix 1 or Mix 2. 1. Label one 15 mL conical tube “Mix 1” and label one 15 mL conical tube “Mix 2”; then add 5 mL of Incubation Buffer (1× PBS, 0.05% Tween-20, 1% BSA) to each of these tubes. 2. Transfer the entire sample of IgG A-Cy™3 and IgG B-Cy™5 (200 mg total; from Subheading 3.2.3) to the Mix 1 tube from step 1. Store the tube at 4°C until needed. 3. Transfer the entire sample of IgG A-Cy™5 and IgG B-Cy™3 (200 mg total; from Subheading 3.2.3) to the Mix 2 tube from step 1. Store the tube at 4°C until needed. 4. Add the PATH fractionated lysate microarrays to two of the chambers of the incubation tray. Cover each slide with 5 mL of Wash Buffer, and rock the slides at room temperature for 1 min. Remove the buffer from the incubation chambers. Repeat this wash five times, for a total of 6 washes. 5. Add the contents of the tube labeled “Mix 1” to incubation chamber 1 and “Mix 2” to incubation chamber 2. 6. Incubate the fractionated protein array slides at room temperature for 1 h with gentle rocking. Every 15 min, use a pipette tip to lift one end of the slide while gently rocking the incubation tray. 7. Add 5 mL of Wash Buffer to each wash chamber, transfer the slides to their respective wash chambers, and incubate at room temperature for 1 min with gentle rocking. 8. Remove the buffer from the wash chambers. 9. Repeat steps 7–8 five times, for a total of 6 washes. 10. Add 5 mL of PBS (1×) to each wash chamber and incubate at room temperature for 1 min with gentle rocking. 11. Remove the PBS from the wash chambers. 12. Add 5 mL of ultrapure water to each wash chamber and incubate at room temperature for 1 min with gentle rocking. 13. Remove the water from the wash chambers.
Native Antigen Fractionation Protein Microarrays
141
14. Dry the slides. It is important to remove as much moisture as possible from the surface of the slides before the liquid evaporates passively. (a) Using scissors or a knife, puncture a small, round hole in the bottom of the slide container. This will facilitate the removal of excess liquid from the slides during centrifugation. (b) Using gloved hands and touching only the edges of the slides, hold the slides so that the excess liquid drips toward the bottom of the array slides (the area containing the manufacturer’s label/barcode) and gently touch this edge to a clean Kimwipe several times. (c) Carefully place the slides in the empty slide container with the ends containing the manufacturer’s label/barcode at the bottom of the vial. Do not touch the array surface. (d) Cap the vial and centrifuge the slides at ~1,000 × g for 25 min at room temperature. 15. Proceed immediately with microarray scanning (Subheading 3.3.1). 3.3. Results and Analysis 3.3.1. Microarray Scanning and Quantitation (see Note 15)
Antibody microarray slides should be scanned using a laser scanner, such as the Axon GenePix 4000B or the Perkin Elmer ScanArray 4000, according to the manufacturer’s specifications. The scanner must be able to measure fluorescence in the ranges of the Cy™3 and Cy™5 fluorophores. 1. Turn on the scanner and allow the lasers to warm up. The lasers on the Perkin Elmer ScanArray 4000 require 15 min to warm-up prior to scanning your arrays. 2. Run a quick/preview scan of the entire slide in order to determine the area containing the arrayed features. 3. Create a scan protocol on the computer attached to the microarray scanner (see Note 16). (a) Set the protocol to scan for Cy™3 and Cy™5 fluorescence. (b) Determine the area containing the arrayed protein fractions and select this as the portion of the array to be scanned. (c) If possible, select an area on the array that contains no arrayed features to be scanned for background intensity. (d) Set the laser powers and PMT (photomultiplier) Gains so that the signal is high enough without being saturated (see Note 17). We suggest the following settings for scanning with the Perkin Elmer ScanArray 4000 Cy3: PMT = 62%; laser power = 90% Cy5: PMT = 50%; laser power = 90%
142
Caiazzo et al.
4. Insert the first slide, containing the Mix 1 autoantibodies, into the scanner. For the ScanArray 4000, the microarray surface is inserted face up. 5. Begin the scan and make sure that the signal is sufficiently high, but not saturated. You may need to readjust the laser and PMT Gain settings or rerun an Automatic Sensitivity Calibration if the image is saturated or the signal is too low. 6. Save the Cy™3 and Cy™5 images as separate TIFF files. The single-file TIFF format is the most useful for quantifying data from the images. 7. Scan the second microarray slide, containing the Mix 2 autoantibodies, using the same settings used to scan the first slide and save the images in the same manner. 8. Obtain the GAL file that corresponds to your fractionated lysate microarrays. This file should be available from the institution/company that printed the microarrays. If you have printed your own slides, you will need to create a GAL file; information on creating your own GAL file can be found at: http://www.moleculardevices.com/pages/software/gn_gal_ examples.html 9. Using the GenePix Pro software, use the “Alt Y” command to open the GAL file from step 8 and open the TIFF files corresponding to your first slide (Mix 1) using the “Alt O” command. 10. Automatically align the grid with the array features using “F8.” Use the Zoom-In feature to ensure the proper alignment of the grid with the features of your array. You can adjust the fit of the Grid to the entire array or to individual array features using the tools on the left-hand side of the screen (see Note 18). 11. Carry out an Automatic Analysis using the “Alt A” command (see Note 19) and save your data as a GPR (GenePix Results) file using the “Alt U” command (see Note 20). 12. Data can be exported to Excel using the “Ctrl A” command to select all data, followed by the “Ctrl C” command to copy all data to the clipboard. 13. Data from your first slide can now be pasted into a blank Excel sheet with the “Ctrl V” command. 14. Repeat steps 9–13 using the TIFF images from your second slide (see Note 21). Note, the same GAL file should be used to extract data from the second slide since slides provided as a pair always have the same lot number. Examples of autoantibody profiling of fractionated lysate microarrays are shown in Figs. 1 and 2. Figure 1 shows a scan
Native Antigen Fractionation Protein Microarrays
143
Fig. 1. An example of results obtained from a fractionated lysate microarray experiment where the dye-swap method was employed. On the left is a composite image of the “Mix 1” scans, with the IC/PBS IgG labeled with Cy™3 (green) and the control IgG labeled with Cy™5 (red ). Shown in comparison (on the right ) is an image of a fractionated lysate microarray incubated with unlabeled IC/PBS IgG; this slide was visualized using the APiX chromogenic detection system.
Fig. 2. Results obtained from a fractionated cancer lysate microarray where the dye-swap method was used. Autoantibody profile of an ovarian cancer patient was compared with an age-matched normal healthy control. Antibodies were purified from the sera and used as probes against a cancer cell lysate generated from a cell line (also can be generated from autologous tumor) that was fractionated and spotted on the array as outlined in this chapter. In this “Mix 1” scan, autoantibodies from an ovarian cancer patient were labeled with Cy™3 (green) and autoantibodies from a control sample were labeled with Cy™5 (red ). This figure shows a slide with 12 subarrays of 8 × 12 spots, with each spot representing one of the 960 fractions that were separated by our 2D fractionation protocol. In this slide, each fraction is spotted once (can be spotted in duplicates or triplicates depending on the user). Each fraction (thus each spot) contains approximately 3–5 proteins per fraction. As shown in this figure, there are 57 fractions containing antigens recognized by ovarian sera alone (green), 50 fractions that were recognized in common between cancer and control IgG (yellow ), and only two fractions that were recognized by the control patient sample (red ). The red dots at the corner of each subarray are Cy™5-labeled streptavidin orientation markers and do not represent any autoantibody reactivity.
144
Caiazzo et al.
image from a fractionated lysate microarray experiment where the dye-swap method was employed. In this experiment, test IgG was purified from the serum of an IC/PBS patient, while control IgG was purified from the sera of an age-matched normal female; IgGs were differentially labeled with fluorescent dyes. The spots that appear green represent Cy™3 reactivity, and the red spots represent the Cy™5 reactivity. Shown in comparison is an image of a fractionated lysate microarray visualized using the APiX chromogenic detection system. Figure 2 shows a scan image from a cancer fractionated lysate microarray experiment with dye-swap on an ovarian cancer sample. In this experiment, IgG was purified from a patient with ovarian cancer, while control IgG was purified from an age-matched normal female control. IgGs were differentially labeled with fluorescent dyes. 3.3.2. Biostatistical Data Analysis
Biostatistical analyses are needed to appropriately normalize, analyze, and interpret the vast amounts of data obtained from microarray studies. There are many analysis tools to extract reliable information from microarray data (commercial software packages as well as programs provided on the web at no cost to investigators). Regardless of the tools that are employed, the fractionated lysate array data must be normalized and transformed before reasonable data comparisons can be made (see Note 22). This protocol will examine analysis using the bioinformatics software package Acuity 4.0. 1. Import each of the microarrays slides (GPR files) into Acuity by clicking the “Import Microarrays” link on the Common Tasks menu located on the left hand side of the screen. 2. Once all of the microarrays have been imported, the microarrays can be normalized using the “Normalization Wizard” link on the Common Tasks menu. Select all of your microarrays located in the Project Tree menu and open the Normalization Wizard. Here researchers can normalize their data using a variety of methods, though we recommend performing a ratio-based normalization. 3. Create a data set from the normalized microarrays by selecting them in the Project Tree menu and clicking on the “Create and Open Dataset” link on the Common Tasks menu. Once a dataset has been created, researchers can manipulate their data in a number of ways, including sorting columns, removing rows, removing columns, and finding specific values. Additionally, researchers can now perform advanced statistical testing and clustering to further examine their data (see Note 23).
Native Antigen Fractionation Protein Microarrays
145
4. Notes 1. While this protocol specifically uses the T24 human bladder cancer cell line as the source of cellular lysate for protein fractionation and microarray construction, researchers should feel free to use alternate cell lines or tissue samples as the source of their cellular lysate. 2. The incubation buffer is simply the wash buffer (1× PBS + 0.05% Tween-20) with 1% BSA added. While the wash buffer can be made beforehand and stored at room temperature on the bench, we recommend making the incubation buffer fresh the day of the microarray experiments. 3. We use the rectangular quadriPerm cell culture containers as incubation trays. These containers are split into four equally sized chambers and are large enough to hold standard sized microscope slides. Each of the chambers is roughly 3.5 in. long, 1.25 in. wide, and 0.5 in. deep; these dimensions allow the microarray slides to be incubated in 5 mL of buffer. The quadriPerm containers are available from Sigma-Aldrich, though researchers should feel free to use any container of similar dimensions for the slide incubations. 4. At this point, we froze and shipped the cell pellet on dry ice to another facility. It may be necessary for researchers to send their cell or tissue samples to outside facilities, such as Eprogen, for protein fractionation. 5. For ProteoSep Start Buffer (SB) preparation, please look at Subheading 3.1.2, step 1(a). 6. If needed, IDA is available from Sigma-Aldrich. 7. Fractions from the first dimension chromatofocusing should be collected using the 2.0 mL 96-well plates. 8. Extra fractions can be stored at −80°C in the 96-well plates following fractionation for downstream MS analysis. 9. Positive controls (1–100 mg/mL Human IgG) and negative controls (print buffer only) were also printed along with the second dimension fractions. 10. The IgG purification procedure is repeated in order to ensure exclusive isolation of IgG from patient sera. We have found that with only one purification run, high-abundance proteins other than IgG may be present in the eluted solution. 11. In order to label the purified antibodies with fluorescent dyes, the antibodies must be in a buffer free of primary amines; the use of buffers that contain primary amino groups, such as
146
Caiazzo et al.
TRIS and glycine, will inhibit the labeling reaction. The Melon™ Gel Purification Buffer used in Subheading 3.2.1 is compatible with all of the Cy™ labeling reagents used in Subheading 3.2.2. 12. GE Healthcare supplies the Cy™3 and Cy™5 dyes as monofunctional N-hydroxysuccinimide (NHS)-esters in dried premeasured amounts. The NHS-ester is a functional group that cross-links to primary amines. This reaction releases the NHS group and produces a covalent amide bond that links the dye to the amine. Due to the unique affinity ratio of each dye to the antibody that it is labeling, different volumes of the two dyes are used in the labeling reactions. We recommend that researchers conduct an estimation of the final dye/ protein ratio on their own according to the manufacturer’s instructions. 13. The Protein Desalting Spin Columns are used to remove free dye. These columns contain a desalting resin and molecular weight cutoff. They perform well in desalting small sample volumes (30–120 mL), providing excellent protein recovery and ³95% retention of small molecules and salts (90% pure. 5. Ensure that the kinase is active in any one of the number of alternative assays. 6. Make sure that the protein kinase is soluble and active in the buffers used for probing the microarray (see recipes in section 2. Materials); if they are not then you may need to revise your buffer selection. 7. Since many of the proteins that are spotted on the microarray will readily bind ATP, including kinases which possess autophosphorylation ability, a negative control in which radiolabeled ATP is applied to an array in the absence of kinase is essential for proper interpretation of the results. 8. Since each purified kinase will have different activity, a titration of kinase concentration is strongly recommended to determine the optimal working concentration for your kinase.
210
Smith, Ptacek, and Snyder
9. At this point, the data is ready to be analyzed. For each feature spotted onto the microarray, a .gpr file will return a range of data points, including the signal and background values that will be used to determine your final list of substrates, or hitlist. 10. In GenePix, signal and background values are returned as the median of the pixels within a feature or the mean of the pixels within a feature. The median is the preferred metric as it is less susceptible to outliers in the data that can be seen if dust or other contamination ends up on the array during scanning. 11. Background is determined for each feature as a local value, typically 2 or 3 pixels from the features limits. This method of determining signal and background intensities has its limitations, particularly if the slide has uneven signals across the slide. This can occur due to uneven probing or inadequate mixing during the incubation. More problematic can be individual spotted kinases whose signals are so strong that they bleed into neighboring features causing an increased background in the neighboring regions such that genuine signals can be mistakenly subtracted out. Likewise, this is the reason for using lower energy [g33P]ATP and not [g32P]ATP. 12. One method, called ProCAT, to account for these irregularities was developed by Zhu et al. (10). ProCAT employs a multi-step approach to affect background correction, signal normalization, positive spot identification, feature crossreactivity, signal quality inspection, and protein amount normalization. For background correction, rather than assigning the background values as the signal surrounding a particular feature, ProCAT determines the background for the eight features surrounding the feature of interest (using a 3 × 3 window) and assigns the background as the median value from all nine. In this way, local variations can be diminished, and signals can be accentuated. 13. Signal normalization is performed by using a sliding window across the slide within which median signals and median absolute deviations are calculated. These values are then used to correct each feature’s signal such that uneven signal distributions across the slide are reduced. 14. Similarly, positive spot identification also uses a sliding window approach. The default setting analyzes a 9 × 9 window, and those features with signals greater than 2 standard deviations above the local mean are assigned as a hit. Two different filters are then employed. 15. The feature cross-reactivity filter determines which features are deemed positive hits in negative control experiments and removes them from the final hitlist. This removes chip features that bind ATP or are autophosphorylated.
Kinase Substrate Interactions
211
16. The signal quality filter examines the reactivity of duplicate spots, and those features with wide variations between duplicate spots are removed from analysis. This filter clearly assumes that the protein microarray that is used for these experiments contains each feature printed in duplicate. 17. Lastly, signals are adjusted by the amount of protein that is spotted in each feature. This can be a very powerful filter in that often the brightest signals are returned from those features spotted at the highest concentration. However, accurate assessment of the actual spotted amount is often difficult to ascertain. Nonetheless, relative protein abundance can be used to adjust the signals. 18. Alternatives to ProCAT exist that identify kinase substrates by their signal-to-noise ratios both within arrays and by comparison with negative control arrays. Like ProCAT, a series of metrics are generated and features are filtered based on their rank within these filters. 19. The first metric is the signal-to-noise ratio, or Z-factor (11). Signals and standard deviations for each feature are compared to the signals and standard deviations for negative controls. If the signals are high and deviations are low relative to the negative controls, then the Z-factor value will approach a value of 1. Low signals and high deviations will drive Z-factor values toward zero. As a rule of thumb, features with Z-factor values above 0.5 can be considered as hits. 20. Z − factor = 1 − [3 × (s (feature of interest) + s (negative controls))]/ | m(feature of interest) − m(negative controls) | . 21. Another metric is termed the Z-score, which is the number of standard deviations, a features’ signal is from the mean of the entire array. A feature with a Z-score above 3 can be considered a positive. 22. Z − score = [Signal(feature of interest) − m(all features)] / s (all features). 23. Lastly, coefficients of variation (CV) are determined for each pair of duplicate spots. Only those features with low interspot CV will be considered for a final hitlist. 24. Filtering the data against values obtained with a negative control can help reduce false positive calls. Any feature that has a Z-factor score greater than 0.5 or a Z-score value greater than 3 on a negative control slide can be removed from analysis as they represent those proteins that either bind ATP directly or are kinases with robust autophosphorylation capacity.
212
Smith, Ptacek, and Snyder
25. The absolute values of these cut-off values can be adjusted based on the length of the hitlists returned; higher stringency will reduce hitlist length and lower stringency will elongate the candidate substrate list. References 1. Gershon D (2003) Proteomics technologies: probing the proteome. Nature 424:581–587 2. MacBeath G, Schreiber SL (2000) Printing proteins as microarrays for high-throughput function determination. Science 289:1760–1763 3. Phizicky E, Bastiaens PI, Zhu H, Snyder M, Fields S (2003) Protein analysis on a proteomic scale. Nature 422:208–215 4. Zhu H, Bilgin M, Bangham R, Hall D, Casamayor A, Bertone P, Lan N, Jansen R, Bidlingmaier S, Houfek T, Mitchell T, Miller P, Dean RA, Gerstein M, Snyder M (2001) Global analysis of protein activities using proteome chips. Science 293:2101–2105 5. Ptacek J, Devgan G, Michaud G, Zhu H, Zhu X, Fasolo J, Guo H, Jona G, Breitkreutz A, Sopko R, McCartney RR, Schmidt MC, Rachidi N, Lee SJ, Mah AS, Meng L, Stark MJ, Stern DF, De Virgilio C, Tyers M, Andrews B, Gerstein M, Schweitzer B, Predki PF, Snyder M (2005) Global analysis of protein phosphorylation in yeast. Nature 438:679–684 6. Gelperin DM, White MA, Wilkinson ML, Kon Y, Kung LA, Wise KJ, Lopez-Hoyo N, Jiang L, Piccirillo S, Yu H, Gerstein M,
7.
8.
9.
10.
11.
Dumont ME, Phizicky EM, Snyder M, Grayhack EJ (2005) Biochemical and genetic analysis of the yeast proteome with a movable ORF collection. Genes Dev 19:2816–2826 Schweitzer B, Predki P, Snyder M (2003) Microarrays to characterize protein interactions on a whole-proteome scale. Proteomics 11:2190–2199 Haab B (2003) Methods and applications of antibody microarrays in cancer research. Proteomics 11:2116–2122 Manning G, Whyte DB, Martiniez R, Hunter T, Sudarsanam S (2002) The protein kinase complement of the human genome. Science 298:1912–1934 Zhu X, Gerstein M, Snyder M (2006) ProCAT: a data analysis approach for protein microarrays. Genome Biol 7:R110 Zhang JH, Chung TD, Oldenburg KR (2000) Confirmation of primary active substances from high throughput screening of chemical and biological populations: a statistical approach and practical considerations. J Comb Chem 3:258–265
Chapter 14 A Functional Protein Microarray Approach to Characterizing Posttranslational Modifications on Lysine Residues Jun Seop Jeong, Hee-Sool Rho, and Heng Zhu Abstract Functional protein microarrays offer a versatile platform to address diverse biological questions. Printing individually purified proteins in a spatially addressable format makes it straightforward to investigating binary interactions. To connect substrates to their upstream modifying enzymes, such as kinases, ubiqutin (Ub) ligases, SUMOylation E3 ligases, and acetyltransferases, is an especially daunting task using traditional methodologies. In recent years, regulation via various types of posttranslational modifications (PTMs) on lysine residues is emerging as an important mechanism(s) underlining diverse biological processes. Our group has been developing and applying functional protein microarrays constructed for different model organisms to globally identify enzyme–substrate interactions with a focus on lysine PTMs. In particular, we have characterized the pleiotropic functions of a ubiquitin E3 ligase, Rsp5, via identification of its downstream substrates using a yeast proteome chip. Also, we have identified nonhistone substrates of the acetyltransferase NuA4 complex in yeast, and revealed that reversible acetylation on a metabolic enzyme affects a glucose metabolism and contributes to life span. In this chapter, we will provide detailed protocols for the investigation of ubiquitylation and acetylation. These protocols are generally applicable for different organisms. Key words: Protein microarray, Posttranslational modification, Ubiquitylation, Ubiquitin E3 ligase, Rsp5, Acetylation, The NuA4 complex, Lysine residues
1. Introduction Functional protein microarrays by definition are constructed by spotting down hundreds of thousands of individually purified proteins in high density on a solid surface (1–4). Unlike the mass spectrometry technologies, in which en masse affinity-purified proteins are subjected to analysis, functional protein microarrays provide information about direct biochemical and physical interactions among biomolecules. For example, protein microarrays
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0_14, © Springer Science+Business Media, LLC 2011
213
214
Jeong, Rho, and Zhu
have been used to profile protein–protein, –lipid, –DNA, –RNA, and –small molecule interactions (4–8). In addition to these binding assays, protein microarrays offer a unique platform to profile covalent modifications. Posttranslational modification (PTM) is an important regulatory mechanism in eukaryotes. While phosphorylation on serine, threonine, and tyrosine residues is well characterized (9, 10), modifications on lysine residues such as ubiquitylation, acetylation, mono-, di- and tri-methylation, SUMOylation, and neddylation are now emerging as important PTMs that are involved in many aspects of cellular functions. Our group is focusing on these mutually exclusive lysine PTMs to better understand the physiological roles and interplay of different signaling pathways. For instance, we and our colleagues set out to identify potential substrates of the yeast NuA4 acetyltransferase complex via covalent acetylation reactions on a yeast proteome chip that contains >5,800 yeast proteins in full-length (11). We identified and validated many nonhistone substrates and by performing in-depth studies on one substrate Pck1, a PEP carboxylase kinase in the cytosol, we found that the acetylation status of Pck1 regulates its enzymatic activity and contributes to longer life span for yeast under nutrient-deprived conditions. This surprising discovery illustrates the power of unbiased, global screening using functional protein microarrays in addressing important biological questions. Ubiquitylation mediates a diverse array of biological functions (Fig. 1). The attachment of ubiquitin (Ub) and poly-Ub chains to a substrate is mediated by a sequential action of E1-activating enzyme, E2-conjugating enzyme, and E3 ligase (12). E3 ligases are broadly classified into RING (really interesting new gene) and HECT (homologous to E6-associated protein C-terminus) domain-containing ligases. There is a variation in RING domains as well (13). There are seven lysine residues in Ub
Fig. 1. Physiological roles of mono- and poly-ubiquitylation in cells. Ubiquitylation of a protein results in the addition of single ubiquitin or multiple ubiquitins as a chain. Ubiquitin chains formed on individual lysine residues (K48, K63, or K29) on a ubiquitin mediate different cellular functions such as proteasome-mediated degradation, DNA repair, and intracellular trafficking.
A Functional Protein Microarray Approach
215
and all of them are shown to form poly-Ub chains in vivo (14). Recent findings indicate that the terminal amine is also a site for ubiquitylation, totaling eight modification sites (12). Among these modifications, the most well-characterized example is the poly-Ub chain formed via K48 linkage, which mediates proteasome-dependent degradation. The K63 chains are involved in DNA repair and intracellular trafficking. Recently, we and another group separately characterized a HECT domain E3 ligase, Rsp5, in yeast using the protein microarray approach (15, 16). Rsp5 belongs to the Nedd4 E3 family. Unlike many RING domain E3 ligases, HECT domain E3 ligases form an intermediate covalent bond with Ub on Cys residues. Rsp5 is highly conserved from yeast to humans, and defects in the human ortholog, Nedd4, can cause congenital disorder. In our study, we predicted previously unknown function of Rsp5 based on the known function of in vivo validated substrates and applied various in vivo assays to confirm the prediction. For example, we predicted that Rsp5 may function in the DNA-damage response via the RNR complex because one of the essential components of the RNR complex, Rnr2, was validated as a bona fide substrate of Rsp5. To demonstrate Rsp5’s new role in DNA damage response, we showed that only upon low dose of hydroxyurea treatment, which is known to target the RNR complex in yeast, did the Rsp5 temperature sensitive strain show growth defect at semipermissive temperature. Further characterization showed that in fact Rnr2 subcellular localization is dependent on the Rsp5 ubiquitylation (15). Construction of functional protein microarrays at high density poses a challenge in every step. Protein samples are usually prepared in a buffer containing ~30% glycerol to prevent evaporation and to ensure protein stability. To fabricate high-density arrays, printing is performed in under a low relative humidity of ~30%. We found that ambient humidity affects printing quality. We are currently using NanoPrinter LM210 (ArrayIt, Inc), which utilizes an enclosed chamber to precisely control the relative humidity inside, a near ideal situation to avoid spotmerging problems on glass. We also found that pilot studies are always helpful to determine printing buffers to achieve high density. Some previous works described general procedures for generation of protein samples and printing (4, 17). Here, we describe detailed protocols for establishing ubiquitylation and acetylation reactions on protein chips. The surface chemistry for printing has to be decided by pilot studies. In our hands, we found noticeable differences in performance for particular applications. For general protein–protein interaction and antibody-based assays, the FullMoon glass surface (FullMoon BioSystems) appears to be the most widely applicable. For HAT (histone acetyltransferase) assay
216
Jeong, Rho, and Zhu
Fig. 2. Acetylation assays on protein chips. X-ray film image of radioisotope detection and comparison of signal intensity and specificity affected by different surface chemistries. An overview of detection of HAT substrate by X-ray film. Landmarks such as histone proteins are visible in individual blocks (top panel). FAST (nitrocellulose-coated), Ni-NTA (tethered with nickel), and FullMoon slides (unknown surface chemistry) were compared for different HAT enzymes and substrate specificity (lower panel). Histone H3 and H4 were used as positive controls, whereas BSA as a negative control H4 is a preferred substrate of NuA4 while H3 is preferred by a different HAT enzyme, SAGA. The substrate specificity is illustrated by the intensity of the signals.
with radio-isotopes, we found that the FAST surface (Whatman) resulted in the highest signal compared to FullMoon, Ni/Cu surface, hydrogel, PVDF, and others (Fig. 2). And again, pilot studies with known biochemical conditions should determine the appropriate surfaces for real experiments.
2. Materials 2.1. Equipment
1. Bench-top centrifuge (Thermo Multifuge 3SR+ centrifuge).
Scientific:
HERAEUS
2. Microarray printer (BioRad: VersArrayer ChipWriter™ Pro System or ArrayIt Corporation: NanoPrinter LM210). 3. Microarray scanner (Molecular Devices: GenePixPro 4000B). 2.2. Protein Microarray Printing
1. 384-well plate (Whatman: 7701-5101). 2. Full Moon protein slides (Bar-coded) (FULL MOON BioSystems: PRT 50B).
A Functional Protein Microarray Approach
217
3. FAST slides (Whatman: 10486111A, Nitrocellulose Coated Slides). 4. Printing pins (ArrayIt Corporation: SMP3). 2.3. Ubiquitylation Reaction
The Ub E1 and E2 enzymes, ubiquitin and its modified forms are available commercially (Boston Biochem, Inc.). In most cases, the E3 ligases have to be purified in a recombinant form in labs following standard protocols. Where applicable, catalog numbers of some ubiquitin products from Boston Biochem are listed in the brackets below. 1. Lifterslip (Thermo Scientific: 25× 60I-2-4789). 2. E1 enzyme: UBE1 (U-100). 3. E2 enzyme: Ubc4 or UbcH5. 4. Ubiquitins (wild-type ubiquitin: U-100H, FLAG-ubiquitin: U-120, myc-ubiquitin: U-115). 5. 4-Well dish (Nalgen Nunc International: 267061). 6. Reaction Buffer: 25 mM Tris–HCl at pH 7.6 containing 50 mM NaCl, 10 mM MgCl2, 4 mM ATP, and 0.5 mM DTT (ATP and DTT are added just before use). 7. Antibodies: rabbit anti-GST (Millipore: AB3282), mouse monoclonal anti-FLAG (Sigma: F7425), goat anti-mouse antibody (Alexafluor 647) (Invitrogen: A21235), goat antirabbit antibody (Alexafluor 555) (Invitrogen: A21429) (see Note 1).
2.4. Protein Acetylation Reaction
1. SuperBlock blocking buffer in TBS (Thermo Scientific: 37535). 2. Reaction buffer (5×): 250 mM Tris–HCl, pH 7.5; 25% glycerol, 0.5 mM EDTA, 250 mM KCl, 5 mM DTT (200×), 0.1 nM Trichostatin A (TSA) (83×), 5 mM PMSF (40×), 5 mM nicotinamide (200×). DTT, TSA, nicotinamide, and PMSF are added just before use. For purified enzymes, HDAC inhibitors (TSA and nicotinamide) may be omitted. 3.
C-Acetyl CoA (Amersham; CFA729; 50 mCi/mL): 0.5 mL.
14
4. Kodak BioMax MR Film (8701302).
3. Methods 3.1. Printing Protein Microarrays
Printing protein microarrays is quite challenging even though similar equipments to the DNA microarray printing is used in general. Protein storage buffer contains glycerol and some amount of detergents, but both may work against good spot morphology.
218
Jeong, Rho, and Zhu
For example, excess amount of detergents may lower the surface tension of droplets deposited on the glass surface, resulting in spread-out and merged features. Glycerol may absorb ambient water, which may create merged features as well in high-density printing (2). Thus, formulation of solutions for printing requires that several aspects be considered. To reduce the merged features, printing is done at lower humidity (~30%). We have used both the VersArrayer (BioRad) and NanoPrinter™ (ArrayIt) microarrayers to print protein microarrays. The NanoPrinter has a builtin humidity monitor and dehumidifier. Printing pins are an integral part of the printing process. We use SMP3 or 946MP2 pins from ArrayIt for routine printing. 1. Adjust the relative humidity to 30% by turning on dehumidifier or other dehumidifying equipment. 2. Meanwhile, set a printing program. The most important factor in high-density printing is spacing between features. For yeast printing, we use SMP3 pins from ArrayIt with column and row spacing 256 and 240 mm, respectively. After setting the parameters for printing, try a dry run without loading the pins to make sure that the program runs as expected. 3. Load pins into the print head, and make sure that the pins move freely by gravity. As a final check for the printing program and pins, load a 30% glycerol plate, and print a couple of slides. Make sure that each pin leaves glycerol spots on the blotting or preprinted slides. 4. Load slides into each slot in the microarray printer, and wait until humidity drops to 30% before starting printing. 5. After the printing is done, let the slides sit or cure in the printing station for at least 8 h to overnight so that protein binding is maximal. After the curing, the protein microarray is ready to use. Cured protein microarrays are stored in a −80°C deep freezer. 3.2. Ubiquitylation Reaction on a Protein Microarray
In general, the same reaction mixture for an in vitro reaction is prepared for the reaction on a chip. To reduce the amount of reagents, the reaction is routinely performed using lifterslip. To cover the entire printed area, we use 26 × 60 mm lifterslip. The space between the lifterslip and glass surface can hold 100 mL of volume. We prepare 100–120 mL of reaction mix and distribute the reaction mixture evenly on the printed area. The reaction setup is schematically shown in Fig. 3. The amount of E1, E2, and E3 enzymes used may be determined in vitro before embarking on the chip experiment using known substrates. Identifying physiological E2 and E3 pairs is not trivial. If the physiological E2–E3 pair is known, you may follow the protocols in published literature. However, if the pairing is not known, you may set up
A Functional Protein Microarray Approach
219
Fig. 3. Design of ubiquitylation reactions to identify substrates of the E3 ligase, Rsp5. Right : A mixture of ubiquitin E1/E2/E3 and ubiquitin with ATP in a ubiquitylation reaction buffer is incubated on a yeast proteome chip blocked with 1% BSA. After reactions, the chip is washed under denaturing condition to remove any nonspecific binding signals. To detect ubiquitylated substrates on the chip, anti-Ub antibodies are used. As a negative control, a reaction mixture without adding the E3 ligase is incubated on the chip, which goes through the same procedure. To identify E3 ligase-dependent hits, signals obtained from the E3 ligase reactions are compared with those from the negative control. GST signals, which serve as a surrogate of protein concentration on the chips, are used to further normalize the signals. This step, however, might not be necessary for data analysis because we found that it did not improve the results significantly.
reactions with a panel of E2 enzymes with known substrates. E2 enzymes are important players in determining chain length and types. In general, E1 and E2 enzymes are robust and work in different ranges of concentration. As a starting point, 100 nM of E1 and E2 would be appropriate. For multiple reactions, we routinely use 4-well dishes. 1. Take out chips from deep freezer and plunge directly into TBST at room temperature (RT). 2. Replace the TBST with TBST containing 5% skim milk or 3% BSA for blocking (see Note 2). 3. Incubate the slide on an orbital shaker for 1 h at RT. 4. Wash the slide with TBST 3 times. 5. Equilibrate the slide with 1× reaction buffer without ATP by changing the buffer 3 times. 6. Meanwhile, prepare the reaction mixture without ATP by adding a predetermined amount of E1, E2, E3, and Ub. Keep the reaction mixture on ice. Right before the reaction, add ATP and mix well.
220
Jeong, Rho, and Zhu
7. Take out the slide from the buffer using a pair of forceps, and remove excessive buffer by gently tapping the corner of the slide on lint free paper. It is important not to dry the slide completely or this will result in a smeared final scanned image. Apply the reaction mixture containing all the components on the slide without touching the surface. Gently place the lifterslip on the slide. Make sure no bubble is trapped. If you find bubbles, gentle tapping will help move the bubbles toward the outside of the printed area. 8. Incubate the reaction at RT for 1 h in a humidified chamber (homemade). 9. After the reaction, the lifterslip is removed by flooding with wash buffer. To remove noncovalently bound Ub, TBST with SDS (1%) may be used to wash the slides. 10. Subsequent washing is done by using TBST for 3 times, 5-min each (see Note 3). 11. Meanwhile, Antibody is prepared in TBST 5% milk (rabbit anti-GST and mouse anti-FLAG). 12. Incubate the slides with a mix with primary antibodies for 1 h with gentle shaking. 13. Wash with TBST for 3 times, 5 min each. 14. Incubate with a mix with secondary antibodies labeled with fluorophores (goat anti-rabbit Alexafluor 555 and goat antimouse Alexafluor 647). 15. Incubate for 1 h with gentle shaking. 16. Wash the slide with TBST for 10 min, 3 times. 17. Rinse the slides with distilled water (see Note 4). 18. The slides are spun dry by centrifugation at 712 × g for 3 min. 19. Scan the image. 3.3. Acetylation Assay on a Chip
Acetylated proteins are detected by radiolabeled 14C-Acetyl CoA as donor reagents. Although fluorescence-based detection is superior to radioactive-based detection, the quality of pananti-Ac-lysine antibodies is not good enough for the detection. One main advantage of using radioisotopes as the labeling reagent is that it allows detection of de novo, modifications by the enzyme in the reaction mixture. In this section, we describe radioactivebased method. 1. Take out slides from deep freezer, directly plunge into blocking solution as described above, and incubate for 1 h at RT with gentle rotation. 2. Briefly wash the slides with PBST.
A Functional Protein Microarray Approach
221
3. Equilibrate the slides by washing 2 times with 1× HAT buffer (without DTT, HDAC inhibitors or PMSF) at RT for 5 min each. 4. Meanwhile, assemble the reaction mixture. 5× HAT buffer: 20 mL C-Acetyl CoA (Amersham; CFA729; 50 mCi/mL): 0.5 mL
14
1 M Sodium butyrate: 5 mL 100 mM PMSF: 5 mL HAT: 20 mg (see Note 5) ddH2O to: 100 mL In the control reaction, HAT enzyme is left out. 5. Excess buffer is removed by gently tapping one side of the slide on a paper towel, and apply the reaction mixture evenly across the surface. 6. Using forceps, place lifterslip on the slide. Be careful not to introduce any bubbles. Incubate at 30°C for 1 h. 7. Wash the slide by flooding the slides 3 times with 50 mM NaHCO3–Na2CO3 (pH 9.3). 8. Rinse the slide with PBS. 9. Spin at 712 × g for 3 min to dry the slide. Make sure that slides are completely dried. 10. Place the slide in an X-ray cassette, and put X-ray film on the slides with a direct contact. Expose the film as needed. It usually takes 2 or more weeks. 11. Develop the film. The film is scanned using an office scanner, and the image is processed using Photoshop as follows. 12. Set a HP Scanjet 8300 scanner at the high sensitivity setting with a resolution of 4,800 dpi, and obtain chip images from the autoradiograph film. 13. Process the captured image of the autoradiograph film, and crop the image size as 6.4 × 4.4 cm and format as jpg, after magnification of image. 14. Open the image in adobe Photoshop and click the image in order to edit the image. 15. Edit the image with grayscale in 16-bit/channel mode. 16. Invert image (under Adjustments tab). 17. Compress the image to 53.3% with percentage in image size. 18. Save the image in TIFF format for further analysis. 19. Open the image in GenePixPro 6.0 and load the gal file containing each protein’s ID.
222
Jeong, Rho, and Zhu
20. Align the setting with gal file based on each spot from anti-GST profiling and positive control sets. 21. Adjust the size of the grid based on sizes or form of signals on chip. 22. Quantify the signal intensities using GenePixPro 6.0 by calculating the background and foreground signals on each spot.
4. Notes 1. The choice of detection antibody is up to the individual researchers. If a tagged antibody (FLAG- or HA-tagged Ub) is used, you will detect the Ub that are added during the reaction. However, if anti-Ub is used, signal will be from de novo as well as endogenous Ub. When choosing fluorescent dye, Cy5 or its equivalent Alexafluor 647 is preferred to Cy3 in general. Nitrocellulose-coated slides have an intrinsic autofluoresence when scanned in the Cy3 channel. Also, some protein samples emit autofluorescence at Cy3 channel, which makes it hard to discern specific and nonspecific interaction of probes. 2. Depending on the application, blocking can be done using 5% milk in TBST or PBST instead of BSA. In our hands, blocking with milk generates a better background to signal ratio. 3. During the washing, lift the sides of the slides alternatively with gentle shaking. Due to the flatness of the slides and chamber, small amount of previous samples remain in the space between the slides and bottom surface. By lifting the slide briefly, you can get rid of the reaction mix efficiently. 4. A brief rinse with distilled water is essential for a clean slide image. An incomplete rinse will result in smearing and watermarks on the scanned image due to residual salts. 5. For purified protein complexes from mammalian cells, it is advised to include HDAC inhibitors and protease inhibitors. If a purified enzyme is used, HDAC inhibitor and protease inhibitors can be omitted by adjusting the volume of each component as necessary. The amount of enzyme to be used is determined by an in vitro assay or pilot experiment with known substrate.
Acknowledgments We thank Dr. Wendy Yap for critical comments and editing. This work is in part supported by the NIH.
A Functional Protein Microarray Approach
223
References 1. Chen CS, Korobkova E, Chen H, Zhu J, Jian X, Tao SC, He C, Zhu H (2008) A proteome chip approach reveals new DNA damage recognition activities in Escherichia coli. Nat Methods 5:69–74 2. Paul B, Michael S (2005) Advances in functional protein microarray technology. FEBS J 272:5400–5411 3. Tao SC, Chen CS, Zhu H (2007) Application of protein microarray technology. Comb Chem High Throughput Screen 10:706–718 4. Zhu H, Bilgin M, Bangham R, Hall D, Casamayor A, Bertone P, Lan N, Jansen R, Bidlingmaier S, Houfek T, Mitchell T, Miller P, Dean RA, Gerstein M, Snyder M (2001) Global analysis of protein activities using proteome chips. Science 293:2101–2105 5. Hu S, Xie Z, Onishi A, Yu X, Jiang L, Lin J, Rho H-S, Woodard C, Wang H, Jeong J-S, Long S, He X, Wade H, Blackshaw S, Qian J, Zhu H (2009) Profiling the human proteinDNA interactome reveals ERK2 as a transcriptional repressor of interferon signaling. Cell 139:610–622 6. MacBeath G, Schreiber SL (2000) Printing proteins as microarrays for high-throughput function determination. Science 289:1760–1763 7. Huang J, Zhu H, Haggarty SJ, Spring DR, Hwang H, Jin F, Snyder M, Schreiber SL (2004) Finding new components of the target of rapamycin (TOR) signaling network through chemical genetics and proteome chips. Proc Natl Acad Sci USA 101: 16594–16599 8. Zhu J, Gopinath K, Murali A, Yi G, Hayward SD, Zhu H, Kao C (2007) RNA-binding proteins that inhibit RNA virus infection. Proc Natl Acad Sci USA 104:3129–3134 9. Zhu H, Klemic JF, Chang S, Bertone P, Casamayor A, Klemic KG, Smith D, Gerstein M, Reed MA, Snyder M (2000) Analysis of
10.
11.
12.
13.
14.
15.
16.
17.
yeast protein kinases using protein chips. Nat Genet 26:283–289 Ptacek J, Devgan G, Michaud G, Zhu H, Zhu X, Fasolo J, Guo H, Jona G, Breitkreutz A, Sopko R, McCartney RR, Schmidt MC, Rachidi N, Lee SJ, Mah AS, Meng L, Stark MJ, Stern DF, De Virgilio C, Tyers M, Andrews B, Gerstein M, Schweitzer B, Predki PF, Snyder M (2005) Global analysis of protein phosphorylation in yeast. Nature 438:679–684 Lin Y-Y, Lu J-Y, Zhang J, Walter W, Dang W, Wan J, Tao S-C, Qian J, Zhao Y, Boeke JD, Berger SL, Zhu H (2009) Protein acetylation microarray reveals that NuA4 controls key metabolic target regulating gluconeogenesis. Cell 136:1073–1084 Pickart CM (2001) Mechanisms underlying ubiquitination. Annu Rev Biochem 70: 503–533 Fang S, Weissman AM (2004) A field guide to uniquitylation. Cell Mol Life Sci 61: 1546–1561 Xu P, Duong DM, Seyfried NT, Cheng D, Xie Y, Robert J, Rush J, Hochstrasser M, Finley D, Peng J (2009) Quantitative proteomics reveals the function of unconventional ubiquitin chains in proteasomal degradation. Cell 137:133–145 Lu J-Y, Lin Y-Y, Qian J, Tao S-C, Zhu J, Pickart C, Zhu H (2008) Functional dissection of a HECT ubiquitin E3 ligase. Mol Cell Proteomics 7:35–45 Gupta R, Kus B, Fladd C, Wasmuth J, Tonikian R, Sidhu S, Krogan NJ, Parkinson J, Rotin D (2007) Ubiquitination screen using protein microarrays for comprehensive identification of Rsp5 substrates in yeast. Mol Syst Biol 3:116 Fasolo J, Snyder M (2009) Protein microarrays. Methods Mol Biol 548:209–222
wwwwwwwwwwwwwwww
Part IV Strategies for Validation of Candidate Targets
wwwwwwwwwwwwwwww
Chapter 15 Multiplexed Detection of Antibodies Using Programmable Bead Arrays Karen S. Anderson Abstract The detection of antibodies in sera has broad applications for detection and monitoring of infectious diseases, autoimmunity, and cancer. Proteomic methods of antigen detection, such as protein microarrays, are excellent clinical discovery tools, but due to both cost and specialization of manufacture, these are limited to screening small numbers of sera. Downstream assays for biomarker validation studies require rapid, reproducible, multiplexed assays for the simultaneous screening of fewer (90% of the spots are flagged, do not subtract the negative control values for the spots. This is because even though the spots are flagged, each spot value is subtracted. Therefore, if the value is negative, it will be added to the antibody value rather than subtracted. 23. Draw the grids by positioning the mouse cursor near the leftmost spot in the first row. Click and drag the grid so the grid encompasses all the spots. The red grid lines should not touch the spots. The red grid line is used to calculate the local area background intensity around each spot. 24. Note that there are two different four-headed arrows (big/ black and small/white). Each arrow has a distinct function. The larger/black arrow is used to move the grid. The smaller white arrow is used to resize the grid. Precise spot analysis is best achieved using the same size grid for each antibody slide. Do not resize the grid once it is drawn unless every grid for each array is resized. If the grid is inadvertently resized, it is best to delete the grid and either redraw it or copy and paste the grid again. The Sypro Ruby Protein Blot-stained arrays will have a unique set of grids compared to the antibodystained arrays due to the differences in image size generated from the laser scanner or CCD camera. 25. Once the dimension of a grid is changed, it cannot be canceled or changed back. The program will only allow the image to be closed and reopened again. For this reason, it is important to save the analysis very often to avoid loosing grids that have been allocated correctly.
300
Gallagher et al.
References 1. Ekins R, Chu F, Biggart E (1990) Multispot, multianalyte, immunoassay. Ann Biol Clin (Paris) 48:655–666 2. Liotta LA, Espina V, Mehta AI, Calvert V, Rosenblatt K et al (2003) Protein microarrays: meeting analytical challenges for clinical applications. Cancer Cell 3:317–325 3. Haab BB (2001) Advances in protein microarray technology for protein expression and interaction profiling. Curr Opin Drug Discov Devel 4:116–123 4. Paweletz CP, Charboneau L, Bichsel VE, Simone NL, Chen T et al (2001) Reverse phase protein microarrays which capture disease progression show activation of prosurvival pathways at the cancer invasion front. Oncogene 20:1981–1989 5. Tonkinson JL, Stillman BA (2002) Nitrocellulose: a tried and true polymer finds utility as a post-genomic substrate. Front Biosci 7:c1–c12 6. Zhu H, Snyder M (2003) Protein chip technology. Curr Opin Chem Biol 7:55–63 7. MacBeath G, Schreiber SL (2000) Printing proteins as microarrays for high-throughput function determination. Science 289: 1760–1763 8. Espina V, Mehta AI, Winters ME, Calvert V, Wulfkuhle J et al (2003) Protein microarrays: molecular profiling technologies for clinical specimens. Proteomics 3:2091–2100 9. Espina V, Woodhouse EC, Wulfkuhle J, Asmussen HD, Petricoin EF III et al (2004) Protein microarray detection strategies: focus on direct detection technologies. J Immunol Methods 290:121–133 10. Celis JE, Gromov P (2003) Proteomics in translational cancer research: toward an integrated approach. Cancer Cell 3:9–15 11. Hunyady B, Krempels K, Harta G, Mezey E (1996) Immunohistochemical signal amplification by catalyzed reporter deposition and its application in double immunostaining. J Histochem Cytochem 44:1353–1362 12. Bobrow MN, Harris TD, Shaughnessy KJ, Litt GJ (1989) Catalyzed reporter deposition, a novel method of signal amplification. Application to immunoassays. J Immunol Methods 125:279–285 13. Bobrow MN, Shaughnessy KJ, Litt GJ (1991) Catalyzed reporter deposition, a novel method of signal amplification. II. Application to membrane immunoassays. J Immunol Methods 137:103–112
14. Bobrow MN, Litt GJ, Shaughnessy KJ, Mayer PC, Conlon J (1992) The use of catalyzed reporter deposition as a means of signal amplification in a variety of formats. J Immunol Methods 150:145–149 15. King G, Payne S, Walker F, Murray GI (1997) A highly sensitive detection method for immunohistochemistry using biotinylated tyramine. J Pathol 183:237–241 16. Wiese R (2003) Analysis of several fluorescent detector molecules for protein microarray use. Luminescence 18:25–30 17. Panchuk-Voloshina N, Haugland RP, BishopStewart J, Bhalgat MK, Millard PJ et al (1999) Alexa dyes, a series of new fluorescent dyes that yield exceptionally bright, photostable conjugates. J Histochem Cytochem 47: 1179–1188 18. Lesaicherre ML, Uttamchandani M, Chen GY, Yao SQ (2002) Antibody-based fluorescence detection of kinase activity on a peptide array. Bioorg Med Chem Lett 12:2085–2088 19. Ekins R, Chu F, Biggart E (1990) Fluorescence spectroscopy and its application to a new generation of high sensitivity, multi-microspot, multianalyte, immunoassay. Clin Chim Acta 194:91–114 20. Bacarese-Hamilton T, Mezzasoma L, Ingham C, Ardizzoni A, Rossi R et al (2002) Detection of allergen-specific IgE on microarrays by use of signal amplification techniques. Clin Chem 48:1367–1370 21. Templin MF, Stoll D, Schrenk M, Traub PC, Vohringer CF et al (2002) Protein microarray technology. Trends Biotechnol 20:160–166 22. VanMeter AJ, Rodriguez AS, Bowman ED, Jen J, Harris CC et al (2008) Laser capture microdissection and protein microarray analysis of human non-small cell lung cancer: differential epidermal growth factor receptor (EGPR) phosphorylation events associated with mutated EGFR compared with wild type. Mol Cell Proteomics 7:1902–1924 23. Gulmann C, Sheehan KM, Conroy RM, Wulfkuhle JD, Espina V et al (2009) Quantitative cell signalling analysis reveals down-regulation of MAPK pathway activation in colorectal cancer. J Pathol 218: 514–519 24. Gulmann C, Espina V, Petricoin E III, Longo DL, Santi M et al (2005) Proteomic analysis of apoptotic pathways reveals prognostic factors in follicular lymphoma. Clin Cancer Res 11:5847–5855
Reverse Phase Protein Microarrays 25. Petricoin EF III, Espina V, Araujo RP, Midura B, Yeung C et al (2007) Phosphoprotein pathway mapping: Akt/mammalian target of rapamycin activation is negatively associated with childhood rhabdomyosarcoma survival. Cancer Res 67:3431–3440 26. Wulfkuhle JD, Speer R, Pierobon M, Laird J, Espina V et al (2008) Multiplexed cell signaling analysis of human breast cancer applications for personalized therapy. J Proteome Res 7:1508–1517 27. Hsu SM, Soban E (1982) Color modification of diaminobenzidine (DAB) precipitation by metallic ions and its application for double immunohistochemistry. J Histochem Cytochem 30:1079–1082 28. Pawley JB (1995) Handbook of biological confocal microscopy. Plenum, New York 29. Berggren K, Steinberg TH, Lauber WM, Carroll JA, Lopez MF et al (1999) A luminescent ruthenium complex for ultrasensitive detection of proteins immobilized on membrane supports. Anal Biochem 276:129–143 30. Berggren KN, Schulenberg B, Lopez MF, Steinberg TH, Bogdanova A et al (2002) An improved formulation of SYPRO Ruby protein gel stain: comparison with the original formulation and with a ruthenium II tris (bathophenanthroline disulfonate) formulation. Proteomics 2:486–498 31. Mackintosh JA, Choi HY, Bae SH, Veal DA, Bell PJ et al (2003) A fluorescent natural product for ultra sensitive detection of proteins in one-dimensional and two-dimensional gel electrophoresis. Proteomics 3: 2273–2288 32. Fowler S (1996) Protein staining and immunodetection using colloidal gold. In: Walker J (ed) Protein protocols handbook. Humana, Totowa, NJ, pp 275–287 33. Switzer RC III, Merril CR, Shifrin S (1979) A highly sensitive silver stain for detecting proteins and peptides in polyacrylamide gels. Anal Biochem 98:231–237 34. Huels C, Muellner S, Meyer HE, Cahill DJ (2002) The impact of protein biochips and microarrays on the drug development process. Drug Discov Today 7:S119–S124 35. Grote T, Siwak DR, Fritsche HA, Joy C, Mills GB et al (2008) Validation of reverse phase protein array for practical screening of potential biomarkers in serum and plasma: accurate
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
301
detection of CA19-9 levels in pancreatic cancer. Proteomics 8:3051–3060 Korf U, Derdak S, Tresch A, Henjes F, Schumacher S et al (2008) Quantitative protein microarrays for time-resolved measurements of protein phosphorylation. Proteomics 8:4603–4612 Korf U, Wiemann S (2005) Protein microarrays as a discovery tool for studying protein-protein interactions. Expert Rev Proteomics 2:13–26 Espina V, Liotta LA, Petricoin EF III (2009) Reverse-phase protein microarrays for theranostics and patient tailored therapy. Methods Mol Biol 520:89–105 Zhou G, Li H, DeCamp D, Chen S, Shu H et al (2002) 2D differential in-gel electrophoresis for the identification of esophageal scans cell cancer-specific protein markers. Mol Cell Proteomics 1:117–124 Washburn MP, Ulaszek R, Deciu C, Schieltz DM, Yates JR III (2002) Analysis of quantitative proteomic data generated via multidimensional protein identification technology. Anal Chem 74:1650–1657 Gygi SP, Rist B, Gerber SA, Turecek F, Gelb MH et al (1999) Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat Biotechnol 17:994–999 Gorg A, Obermaier C, Boguth G, Harder A, Scheibe B et al (2000) The current state of twodimensional electrophoresis with immobilized pH gradients. Electrophoresis 21:1037–1053 Tibes R, Qiu Y, Lu Y, Hennessy B, Andreeff M et al (2006) Reverse phase protein array: validation of a novel proteomic technology and utility for analysis of primary leukemia specimens and hematopoietic stem cells. Mol Cancer Ther 5:2512–2521 Stanislaus R, Carey M, Deus HF, Coombes K, Hennessy BT et al (2008) RPPAML/RIMS: a metadata format and an information management system for reverse phase protein arrays. BMC Bioinformatics 9:555 Newland J, Jones J (1980) Fluorometry. In: Hicks R, Schenken JR, Steinrauf MA (eds) Laboratory instrumentation. Harper & Row, Hagerstown, MD, pp 61–66 VanMeter A, Signore M, Pierobon M, Espina V, Liotta LA et al (2007) Reverse-phase protein microarrays: application to biomarker discovery and translational medicine. Expert Rev Mol Diagn 7:625–633
wwwwwwwwwwwwwwww
Chapter 19 Förster Resonance Energy Transfer Methods for Quantification of Protein–Protein Interactions on Microarrays Michael Schäferling and Stefan Nagl Abstract Methods based on Förster (or fluorescence) resonance energy transfer (FRET) are widely used in various areas of bioanalysis and molecular biology, such as fluorescence microscopy, quantitative real-time polymerase chain reaction (PCR), immunoassays, or enzyme activity assays, just to name a few. In the last years, these techniques were successfully implemented to multiplex biomolecular screening on microarrays. In this review, some fundamental considerations and practical approaches are outlined and it is demonstrated how this very sensitive (and distance-dependent) method can be utilized for microarraybased high-throughput screening (HTS) with a focus on protein microarrays. The advantages and also the demands of this dual-label technique in miniaturized multiplexed formats are discussed with respect to its potential readout modes, such as intensity, dual wavelength, and time-resolved FRET detection. Key words: FRET, Protein microarray, Biomolecular interaction, Screening, Time-resolved fluorescence
1. The Basis of FRET Detection Energy transfer is a widespread phenomenon in nature and can be observed on many length scales from stellar to subatomic dimensions. Short-range energy transfer processes (typically 50,000 mutations genetics/CGP/cosmic
Human gene mutation A collection of germline mutations in www.hgmd.org/ database nuclear genes that are associated with human-inherited disease MutationView KMcancerDB
Human gene mutation database; The number of genes in the databases is less than what was claimed in the paper
http://mutview.dmb.med.keio. ac.jp/
Cancer gene census
A catalog of mutations in more than 400 cancer implicated genes
www.sanger.ac.uk/genetics/ CGP/Census/
IARC TP53 mutation database
TP53 gene variations identified in human populations and tumor samples
http://www-p53.iarc.fr/
CDKN2a database
The variants of CDKN2A recorded in human disease
https://biodesktop.uvm.edu/ perl/p16
Androgen receptor gene mutations database
374 Published mutations, most being http://androgendb.mcgill.ca/ point mutations identified in patients with androgen insensitivity syndrome
Breast cancer information core
A repository for all mutations and polymorphisms in genes related to breast cancer
http://research.nhgri.nih.gov/ bic/
GAC
A collection of gene mutations, loss of heterozygosity, and/or chromosome changes in tumors from humans, mice, or rats gathered from peerreviewed journals
www.niehs.nih.gov/research/ resources/databases/gac/
SNP500 cancer database
Information on over 3,400 SNPs in genes important in cancer
http://snp500cancer.nci.nih.gov/
dbSNP
A general catalog of genome variations
www.ncbi.nlm.nih.gov/projects/ SNP/
MedRefSNP
Information about 36,199 unique SNPs collected from the PubMed and OMIM databases
www.medclue.com/medrefsnp
Chromosomal abnormalities in cancer
A compilation of human cancers associated with chromosome aberrations
www.slh.wisc.edu/cytogenetics/ cancer/
Database Resources for Proteomics-Based Analysis of Cancer
355
release version 44 contains 101,138 mutations resulting from analysis of 13,501 genes and 419,018 tumors. Human Gene Mutation Database (HGMD) has a collection of germline mutations in nuclear genes in association with human-inherited disease (23). The database contains 93,347 mutation entries. The public version of the database is free for academic/nonprofit usage and a registration is required. MutationView, a human gene mutation database, contains mutation data for approximately 300 genes, most of which are involved in monogenic diseases. Among them, 42 genes are cancer-related and have been placed in a separate database named KMcancerDB (24). The Cancer Gene Census is a catalog of the genes for which mutations have been causally implicated in cancer (25). To date, more than 1% of all human genes are implicated via mutation in cancer. Of these cancerassociated genes, approximately 90% have somatic mutations, 20% bear germline mutations that predispose to cancer, and 10% show both somatic and germline mutations. In addition to these broad-scoped cancer mutation databases, there are more specialized databases which focus on specific cancer proteins and protein families. For example, P53 functions as a tumor suppressor through the regulation of the cell cycle. In humans, P53 is encoded by the Tp53 gene, which is the most frequently mutated gene in human cancer. The International Agency for Research on Cancer (IARC) Tp53 Mutation Database contains more than 10,000 entries of Tp53 somatic mutations, 144 entries of Tp53 germline mutations, and 13 entries of p53 polymorphisms (26). Another important cancer gene represented in a specialist database is cyclin-dependent kinase inhibitor 2A (CDKN2A or P16). Cdkn2a is a major melanoma predisposition gene. The CDKN2a Database provides information on germline and somatic variants of the Cdkn2a tumor suppressor gene recorded in human disease (27). Another specialist database, the Androgen Receptor Gene Mutations Database, shows associations between mutations in the androgen receptor (Ar) gene and male prostate cancer (28). Androgen regulated genes are critical for the development and maintenance of the male sexual phenotype. Some germline mutations in Ar are associated with the occurrence of breast cancer in males suffering from partial androgen insensitivity, while somatic mutations in the Ar are associated with metastatic prostate cancer. The database contains 605 entries of reported mutations and 70 AR-interacting proteins (28, 29). Breast Cancer Information Core at National Human Genome Research Institute facilitates the detection and characterization of breast cancer susceptibility genes. To use the database, membership is needed, which is open to all and can be obtained through online registration. The Genetic Alterations in Cancer (GAC) database is a webbased system for collecting and summarizing data reported in the
356
Zhang, DeLuca, and Brusic
publications for genetic alterations in various types of human and rodent tumors (30). The system also provides a tool for assessing groups of related genes for mutations, allelic loss, and homozygous deletions based on species and tumor site. It gives users a broader view of tumor development and facilitates comparative analysis of multiple genes in genetic pathways to cancer rather than focusing on single genes. 2.2.2. SNP Databases
SNPs are a class of genetic markers that aid in identifying individuals at high risk of developing certain cancers and developing tailored medication. For example, the Tp53 Arg/Pro polymorphism at codon 72 may be linked to an increased risk of lung cancer (31). Several SNP databases are listed in Table 2. SNP500 Cancer Database is a component of the National Cancer Institute Cancer Genome Anatomy Project. It provides sequence and genotype assay information for over 13,400 SNPs which are useful in mapping complex diseases, such as cancer (32). The database provides gene locations and >200 bp of surrounding annotated sequence (including nearby SNPs). Furthermore, frequency information and per subpopulation as well as calculation of Hardy-Weinberg equilibrium for each subpopulation are also provided. The dbSNP was set up at NCBI to serve as a central repository for genetic variation (33). MedRefSNP provides integrated information about SNPs collected from the PubMed and OMIM databases (34).
2.2.3. Databases Cataloguing other Genetic Abnormalities
Chromosomal abnormalities can be caused by mutations which change the number of chromosomes (numerical abnormalities) or change the structure of the chromosome (structural abnormalities). One of the important causes of cancer is gene translocations and gross gene deletions. The Chromosomal Abnormalities in Cancer website, hosted by the Wisconsin State Laboratory of Hygiene, provides information on several human cancers associated with chromosome aberrations.
2.3. Tumor Antigen Databases
Since the identification of MAGEA1 as a tumor antigen recognized by cytolytic T lymphocyte on human melanoma, the number of characterized tumor antigens has exponentially increased (35). The identification of tumor antigens remains a high priority in cancer research and is an essential component in developing immune-based strategies to combat cancer. The databases listed in Table 3 are useful resources for the study of immune responses against tumors. The term cancer-testis (CT) antigen was proposed by Scanlan and colleagues to encompass a heterogeneous groups of antigens, which show restricted expression in cancer and testis and restricted immunogenicity in cancer patients (36). CT antigens are ideal targets for cancer immunotherapy due to their restricted expression pattern.
Database Resources for Proteomics-Based Analysis of Cancer
357
Table 3 Online databases on human tumor antigens Database
Description
URL
CTdatabase
A repository of cancer-testis antigen data
http://www.cta.lncc.br/
CAPD
An analysis system for cancerrelated data
www.bioinf.uni-sb.de/CAP/
Cancer immunome database
A repertoire of antigens eliciting antibody responses in cancer patients
http://ludwig-sun5.unil.ch/ CancerImmunomeDB/
Cancer immunity peptide Four data tables containing 129 database tumor antigens with defined T-cell epitopes
http://www.cancerimmunity.org/ peptidedatabase/Tcellepitopes.htm
TANTIGEN
http://cvc.dfci.harvard.edu/tadb/
A human tumor T-cell antigen database
CT database provides information on CT antigens, including gene names and aliases, RefSeq accession numbers, genomic location, known splicing variants, gene duplications, and additional family members. It also provides gene expression at the mRNA level in normal and tumor tissues, manually curated data related to mRNA and protein expression, antigen-specific immune responses in cancer patients, and links to PubMed for relevant CT antigen articles (37). The Cancer-Associated Protein Database (CAPD) was built upon SEREX database, in which the sequences were obtained by screening cDNA expression libraries using serum from cancer patients as probes and sequencing individual reactive clones (38). The database also contains microarray, epigenetic, and immunostaining data. It aims to provide information covering all the gene products against which an immune response has been documented in cancer patients. The Cancer Immunome database is a continuation of the SEREX database in a more organized form. It is an access point of information about all of the gene products against which an immune response has been documented in cancer patients (39). The development of T-cell immunity against cancer has the potential to effective rejection and elimination of tumor cells, and hence, T cell-defined tumor antigens are a particular focus of several databases. The Cancer Immunity Peptide database provides four static data tables, containing 129 human tumor antigens with defined T-cell epitopes (40). Among them, 45 entries are tumor antigens resulting from mutations, 29 are shared tumorspecific antigens, 12 are differentiation antigens, and 43 are antigens overexpressed in tumors. For each tumor antigen, a link to
358
Zhang, DeLuca, and Brusic
GeneCards (14) and literature reference are provided. The database is relatively simple without any query function or analysis tool. A list of mouse and human tumor T-cell antigens were reported in (41). The Tumor T-Cell Antigen Database (TANTIGEN) is a data source and analysis platform for cancer vaccine target discovery focusing on human tumor-derived HLA ligands and T-cell epitopes. It contains 4,006 curated antigen entries representing 251 unique proteins. TANTIGEN also provides information on experimentally validated T-cell epitopes and HLA ligands, antigen isoforms, antigen sequence mutations, and tumor antigen classification. Analysis tools integrated in the database include search tool for querying the dataset, multiple sequence alignment of antigen isoforms, sequence similarity search using BLAST, visual display of T-cell epitopes/HLA ligands, and prediction of binding peptides of 15 HLA Class I and Class II alleles. TANTIGEN is the most comprehensive database on Tumor T-cell antigens so far. 2.4. Databases of Cancer-Associated Genes
This section provides descriptions of two databases that integrate multiple heterogeneous datasets, including molecular data, clinical data, and experimental data, together with computational analysis tools to advance the cancer research. The Cancer Genome Anatomy Project (CGAP) aims to improve detection, diagnosis, and treatment for cancer patients through the analysis of the gene expression profiles of normal, precancer, and cancer cells (42). Its website provides genomic data for humans and mice, including transcript sequence, gene expression patterns, SNPs, clone resources, and cytogenetic information. The Mitelman Database of Chromosome Aberrations in Cancer (http://cgap.nci.nih.gov/Chromosomes/Mitelman) is part of CGAP. It is one of the largest online catalogs of cytogenetic aberrations in cancer, containing 56,694 cases as of 2009. The database relates chromosomal abnormalities to tumor characteristics (43). The Mouse Tumor Biology Database (MTBD) supports the use of the mouse as a model system of hereditary and induced cancers (44). The database provides access to tumor names and classifications, tumor incidence and latency data in different strains of mice, tumor pathology reports and images, information on genetic factors that are associated with tumor biology, and the references associated with these data (Table 4).
2.5. Protein Interaction and Pathway Databases
Biological pathways are the blueprints of cellular actions and they describe the roles of genomic entities in various cellular mechanisms. Human PPI data are important for understanding molecular signaling networks and the functional roles of biomolecules. Approaches involving pathway and PPI data are useful for analyzing microarray data and for generating testable hypotheses. Much effort has been put into pathway studies and many pathway
Database Resources for Proteomics-Based Analysis of Cancer
359
Table 4 Databases of cancer-associated genes Database
Description
URL
Cancer genome anatomy project
Containing genomic data for humans and mice, including transcript sequence, gene expression patterns, SNPs, clone resources, and cytogenetic information
http://cgap.nci.nih.gov.ezp-prod1.hul. harvard.edu/
MTBD
Supports the use of the mouse as a model system of hereditary cancer
http://tumor.informatics.jax.org/ mtbwi/
databases have been developed and made available online. Table 5 contains a list of databases providing information on PPI and biological pathways. Pathguide is a meta-database which contains information about 302 biological pathway resources (45). They include databases on metabolic pathways, signaling pathways, transcription factor targets, gene regulatory networks, genetic interactions, protein–compound interactions, and PPIs. Pathguide serves as a starting point for biological pathway analysis. A recent paper reviewed the major databases of human pathways and discussed how to use the information for the reconstruction of signaling pathways (46). The KEGG pathway database is a collection of manually drawn pathway maps representing our knowledge on the molecular interaction and reaction networks involved in metabolism, genetic information processing, environmental information processing, cellular processes, and pathogenesis (47). The BioCarta website catalogs and summarizes classical pathways as well as newly suggested pathways information on more than 120,000 genes from multiple species, including human and mouse. Reactome is an expert-curated knowledgebase of human reactions and pathways. As of 2009, it hosts 2,975 human proteins, 2,907 reactions, and 4,455 literature citations (48). The Pathway Interaction Database (PID) hosts 100 human Pathways containing 6,298 interactions curated by domain experts from US National Cancer Institute and Nature Publishing Group. It also encompasses 329 human pathways containing 7,418 interactions imported from BioCarta and Reactome (49). The Human Pathway Database (HPD) combines heterogeneous human pathway data from PID, Reactome, BioCarta, KEGG, or indexed from the Protein Lounge Web sites (50). So far, HPD contains
360
Zhang, DeLuca, and Brusic
Table 5 Protein interaction and pathway database Database
Description
URL
Pathguide
A metadatabase providing an overview of web-accessible biological pathway and network databases
http://www.pathguide.org/
KEGG pathway database
A collection of manually drawn pathway maps
www.genome.jp/kegg/ pathway.html
BioCarta pathway
A collection of pathways for multiple species
http://www.biocarta.com/ genes/index.asp
Reactome
A curated resource for human pathway data
http://www.reactome.org/
PID
A collection of curated pathways related to human molecular signaling, regulatory events, and key cellular processes
http://pid.nci.nih.gov/
HPD
Providing combined view connecting human proteins, genes, RNAs, enzymes, signaling, metabolic reactions, and gene regulatory events
http://bio.informatics.iupui. edu/HPD
NetPath
A catalog of annotations for cancer and immune signaling pathways
www.netpath.org/
HAPPI
One of the most comprehensive public compilation of human protein interaction information
http://bio.informatics.iupui. edu/HAPPI/
HomoMINT
An inferred human network based on orthology mapping of protein interactions discovered in model organisms
http://mint.bio.uniroma2. it/HomoMINT
IntAct
An open-source, open data molecular interaction database and toolkit
www.ebi.ac.uk/intact
999 human pathways and more than 59,341 human molecular entities. A set of analysis tools is also provided in HPD to allow searching, managing, and studying human biological pathways. NetPath is a component of Human Protein Reference Database (HPRD), which is a centralized platform to visually depict and integrate information pertaining to domain architecture, posttranslational modifications, interaction networks, and disease association for each protein in the human proteome (51). NetPath has 20 annotated immune and cancer signaling pathways involving 1,682 molecules and 1,800 interactions. The HAPPI database integrates protein interaction data from multiple public databases, including HPRD, BIND, MINT, STRING, and OPHID. A measure of reliability (rank levels from 1 to 5) has been given to each entry in the database. As of 2008,
Database Resources for Proteomics-Based Analysis of Cancer
361
the database contains 142,956 nonredundant, medium to high-confidence level human protein interaction pairs among 10,592 human proteins (52). HomoMINT extends PPIs experimentally verified in models organisms to the orthologous proteins in human (53). As of 2009, the database has 2,4439 interactions of 8,041 proteins. The curated data can be analyzed in the context of the high-throughput data and displayed in graphics by a tool named MINT Viewer. Data in IntAct are obtained from the literature or from direct data depositions by expert curators. As of September 2009, it contains over 200,000 curated binary interaction entries. IntAct provides a two-tiered view of the interaction data: a simplified and tabular view and a specialized view providing the full annotation of interactions, interactors and their properties (54). Detailed review and evaluation of public human PPI databases can be found in (55, 56).
3. Discussion Here we have reviewed and summarized five groups of databases related to proteomics studies in cancer research. We have included databases containing gene/protein expression data produced by microarray studies, next-generation sequencing, and other high-throughput experiments, gene mutation and SNP databases, tumor antigen databases, databases of cancer-associated genes, and protein interaction and pathway databases. For more cancerrelated databases, refer to the 2009 database special issue of Nucleic Acids Research (www.oxfordjournals.org/nar/database/ subcat/8/33) (57). The relevant databases can be found in subheading “Cancer gene databases” under heading “Human Genes and Diseases.” An increasing use of databases is data mining (58). These applications involve systems that combine data from multiple specialized databases and analytic tools enabling detailed analysis. For example, BiomarkerDigger (http://biomarkerdigger.org) performs data analysis, searching, and metadata-gathering (59). When gathering metadata, it searches proteome DBs for PPI, Gene Ontology annotations, protein domains, human genetic disorders, and tissue expression profile information. These diverse sources are integrated into protein data sets that are accessed through a search function in BiomarkerDigger. The identification of a serological biomarker for hepatocellular carcinoma by comparison of plasma and tissue proteomic data sets from healthy volunteers and cancer patients was demonstrated by using this resource (59). The vast amount of information generated from cancer research has presented both a challenge and an opportunity for
362
Zhang, DeLuca, and Brusic
researchers worldwide. Researchers have invested lots of effort and time in cleaning, annotating, and organizing the data produced from proteomics studies and put them into specialized biological databases. Information retrieval from proteomics databases is the starting point in performing downstream bioinformatics analyses. Making use of these data enhances understanding of the disease and advance anticancer treatments.
Acknowledgments We thank Dr. Catherine J. Wu for thoughtful reviews of this manuscript. References 1. James P (1997) Protein identification in the post-genome era: the rapid rise of proteomics. Q Rev Biophys 30:279–331 2. Koomen JM, Haura EB, Bepler G, Sutphen R, Remily-Wood ER, Benson K, Hussein M, Hazlehurst LA, Yeatman TJ, Hildreth LT, Sellers TA, Jacobsen PB, Fenstermacher DA, Dalton WS (2008) Proteomic contributions to personalized cancer care. Mol Cell Proteomics 7:1780–1794 3. Gygi SP, Rochon Y, Franza BR, Aebersold R (1999) Correlation between protein and mRNA abundance in yeast. Mol Cell Biol 19:1720–1730 4. Ivanov SS, Chung AS, Yuan ZL, Guan YJ, Sachs KV, Reichner JS, Chin YE (2004) Antibodies immobilized as arrays to profile protein post-translational modifications in mammalian cells. Mol Cell Proteomics 3:788–795 5. Bamford S, Dawson E, Forbes S, Clements J, Pettett R, Dogan A, Flanagan A, Teague J, Futreal PA, Stratton MR, Wooster R (2004) The COSMIC (catalogue of somatic mutations in cancer) database and website. Br J Cancer 91:355–358 6. Kopf E, Zharhary D (2007) Antibody arrays – an emerging tool in cancer proteomics. Int J Biochem Cell Biol 39:1305–1317 7. Tao SC, Chen CS, Zhu H (2007) Applications of protein microarray technology. Comb Chem High Throughput Screen 10:706–718 8. Sherman BT, da Huang W, Tan Q, Guo Y, Bour S, Liu D, Stephens R, Baseler MW, Lane HC, Lempicki RA (2007) DAVID knowledgebase: a gene-centered database integrating
9.
10.
11.
12.
13.
14.
15.
heterogeneous gene annotation resources to facilitate high-throughput gene functional analysis. BMC Bioinformatics 8:426 da Huang W, Sherman BT, Tan Q, Kir J, Liu D, Bryant D, Guo Y, Stephens R, Baseler MW, Lane HC, Lempicki RA (2007) DAVID Bioinformatics Resources: expanded annotation database and novel algorithms to better extract biology from large gene lists. Nucleic Acids Res 35:W169–W175 Rajcevic U, Niclou SP, Jimenez CR (2009) Proteomics strategies for target identification and biomarker discovery in cancer. Front Biosci 14:3292–3303 Grantzdorffer I, Carl-McGrath S, Ebert MP, Rocken C (2008) Proteomics of pancreatic cancer. Pancreas 36:329–336 Sahin U, Tureci O, Schmitt H, Cochlovius B, Johannes T, Schmits R, Stenner F, Luo G, Schobert I, Pfreundschuh M (1995) Human neoplasms elicit multiple specific immune responses in the autologous host. Proc Natl Acad Sci USA 92:11810–11813 Boutet E, Lieberherr D, Tognolli M, Schneider M, Bairoch A (2007) UniProtKB/Swiss-Prot. Methods Mol Biol 406:89–112 Rebhan M, Chalifa-Caspi V, Prilusky J, Lancet D (1998) GeneCards: a novel functional genomics compendium with automated data mining and query reformulation support. Bioinformatics 14:656–664 Barrett T, Troup DB, Wilhite SE, Ledoux P, Rudnev D, Evangelista C, Kim IF, Soboleva A, Tomashevsky M, Marshall KA, Phillippy KH, Sherman PM, Muertter RN, Edgar R (2009) NCBI GEO: archive for high-throughput
Database Resources for Proteomics-Based Analysis of Cancer
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
functional genomic data. Nucleic Acids Res 37:D885–D890 Ikeo K, Ishi-i J, Tamura T, Gojobori T, Tateno Y (2003) CIBEX: center for information biology gene expression database. C R Biol 326:1079–1082 Rocca-Serra P, Brazma A, Parkinson H, Sarkans U, Shojatalab M, Contrino S, Vilo J, Abeygunawardena N, Mukherjee G, Holloway E, Kapushesky M, Kemmeren P, Lara GG, Oezcimen A, Sansone SA (2003) ArrayExpress: a public database of gene expression data at EBI. C R Biol 326:1075–1078 Parkinson H, Kapushesky M, Kolesnikov N, Rustici G, Shojatalab M, Abeygunawardena N, Berube H, Dylag M, Emam I, Farne A, Holloway E, Lukk M, Malone J, Mani R, Pilicheva E, Rayner TF, Rezwan F, Sharma A, Williams E, Bradley XZ, Adamusiak T, Brandizi M, Burdett T, Coulson R, Krestyaninova M, Kurnosov P, Maguire E, Neogi SG, Rocca-Serra P, Sansone SA, Sklyar N, Zhao M, Sarkans U, Brazma A (2009) ArrayExpress update – from an archive of functional genomics experiments to the atlas of gene expression. Nucleic Acids Res 37:D868–D872 Kato K, Yamashita R, Matoba R, Monden M, Noguchi S, Takagi T, Nakai K (2005) Cancer gene expression database (CGED): a database for gene expression profiling with accompanying clinical information of human cancer tissues. Nucleic Acids Res 33:D533–D536 Penkett CJ, Bahler J (2004) Navigating public microarray databases. Comp Funct Genomics 5:471–479 Zeng ZY, Xiong W, Zhou YH, Li XL, Li GY (2006) [Advances in high-density whole genome-wide single nucleotide polymorphism array in cancer research]. Ai Zheng 25:1454–1458 Forbes SA, Bhamra G, Bamford S, Dawson E, Kok C, Clements J, Menzies A, Teague JW, Futreal PA, Stratton MR (2008) The catalogue of somatic mutations in cancer (COSMIC). Curr Protoc Hum Genet. Chapter 10, Unit 10 11 Stenson PD, Mort M, Ball EV, Howells K, Phillips AD, Thomas NS, Cooper DN (2009) The human gene mutation database: 2008 update. Genome Med 1:13 Shimizu N, Ohtsubo M, Minoshima S (2007) MutationView/KMcancerDB: a database for cancer gene mutations. Cancer Sci 98:259–267 Futreal PA, Coin L, Marshall M, Down T, Hubbard T, Wooster R, Rahman N, Stratton MR (2004) A census of human cancer genes. Nat Rev Cancer 4:177–183
363
26. Hernandez-Boussard T, Rodriguez-Tome P, Montesano R, Hainaut P (1999) IARC p53 mutation database: a relational database to compile and analyze p53 mutations in human tumors and cell lines. International Agency for Research on Cancer. Hum Mutat 14:1–8 27. Murphy JA, Barrantes-Reynolds R, Kocherlakota R, Bond JP, Greenblatt MS (2004) The CDKN2A database: integrating allelic variants with evolution, structure, function, and disease association. Hum Mutat 24:296–304 28. Gottlieb B, Trifiro M, Lumbroso R, Pinsky L (1997) The androgen receptor gene mutations database. Nucleic Acids Res 25:158–162 29. Gottlieb B, Beitel LK, Wu JH, Trifiro M (2004) The androgen receptor gene mutations database (ARDB): 2004 update. Hum Mutat 23:527–533 30. Jackson MA, Lea I, Rashid A, Peddada SD, Dunnick JK (2006) Genetic alterations in cancer knowledge system: analysis of gene mutations in mouse and human liver and lung tumors. Toxicol Sci 90:400–418 31. Hu Y, McDermott MP, Ahrendt SA (2005) The p53 codon 72 proline allele is associated with p53 gene mutations in non-small cell lung cancer. Clin Cancer Res 11:2502–2509 32. Packer BR, Yeager M, Burdett L, Welch R, Beerman M, Qi L, Sicotte H, Staats B, Acharya M, Crenshaw A, Eckert A, Puri V, Gerhard DS, Chanock SJ (2006) SNP500Cancer: a public resource for sequence validation, assay development, and frequency analysis for genetic variation in candidate genes. Nucleic Acids Res 34:D617–D621 33. Smigielski EM, Sirotkin K, Ward M, Sherry ST (2000) dbSNP: a database of single nucleotide polymorphisms. Nucleic Acids Res 28:352–355 34. Rhee H, Lee JS (2009) MedRefSNP: a database of medically investigated SNPs. Hum Mutat 30:E460–E466 35. van der Bruggen P, Traversari C, Chomez P, Lurquin C, De Plaen E, Van den Eynde B, Knuth A, Boon T (1991) A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma. Science 254:1643–1647 36. Scanlan MJ, Gure AO, Jungbluth AA, Old LJ, Chen YT (2002) Cancer/testis antigens: an expanding family of targets for cancer immunotherapy. Immunol Rev 188:22–32 37. Almeida LG, Sakabe NJ, deOliveira AR, Silva MC, Mundstein AS, Cohen T, Chen YT, Chua R, Gurung S, Gnjatic S, Jungbluth AA, Caballero OL, Bairoch A, Kiesler E, White SL, Simpson AJ, Old LJ, Camargo AA,
364
38.
39. 40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
Zhang, DeLuca, and Brusic Vasconcelos AT (2009) CTdatabase: a knowledge-base of high-throughput and curated data on cancer-testis antigens. Nucleic Acids Res 37:D816–D819 Donnes P, Hoglund A, Sturm M, Comtesse N, Backes C, Meese E, Kohlbacher O, Lenhof HP (2004) Integrative analysis of cancer-related data using CAP. FASEB J 18:1465–1467 Jongeneel V (2001) Towards a cancer immunome database. Cancer Immun 1:3 Novellino L, Castelli C, Parmiani G (2005) A listing of human tumor antigens recognized by T cells. Cancer Immunol Immunother 54:187–207 Van den Eynde BJ, van der Bruggen P (1997) T cell defined tumor antigens. Curr Opin Immunol 9:684–693 Strausberg RL, Buetow KH, Greenhut SF, Grouse LH, Schaefer CF (2002) The cancer genome anatomy project: online resources to reveal the molecular signatures of cancer. Cancer Investig 20:1038–1050 Schaefer C, Grouse L, Buetow K, Strausberg RL (2001) A new cancer genome anatomy project web resource for the community. Cancer J 7:52–60 Krupke D, Naf D, Vincent M, Allio T, Mikaelian I, Sundberg J, Bult C, Eppig J (2005) The Mouse Tumor Biology Database: integrated access to mouse cancer biology data. Exp Lung Res 31:259–270 Bader GD, Cary MP, Sander C (2006) Pathguide: a pathway resource list. Nucleic Acids Res 34:D504–D506 Bauer-Mehren A, Furlong LI, Sanz F (2009) Pathway databases and tools for their exploitation: benefits, current limitations and challenges. Mol Syst Biol 5:290 Kanehisa M, Goto S, Furumichi M, Tanabe M Hirakawa, M (2010) KEGG for representation and analysis of molecular networks involving diseases and drugs, Nucleic Acids Res 38:D355–D360 Matthews L, Gopinath G, Gillespie M, Caudy M, Croft D, de Bono B, Garapati P, Hemish J, Hermjakob H, Jassal B, Kanapin A, Lewis S, Mahajan S, May B, Schmidt E, Vastrik I, Wu G, Birney E, Stein L, D’Eustachio P (2009) Reactome knowledgebase of human biological pathways and processes. Nucleic Acids Res 37:D619–D622 Schaefer CF, Anthony K, Krupa S, Buchoff J, Day M, Hannay T, Buetow KH (2009) PID: the Pathway Interaction Database. Nucleic Acids Res 37:D674–D679 Chowbina SR, Wu X, Zhang F, Li PM, Pandey R, Kasamsetty HN, Chen JY (2009) HPD: an
51.
52.
53.
54.
55.
56.
57.
58.
59.
online integrated human pathway database enabling systems biology studies. BMC Bioinformatics 10(Suppl 11):S5 Keshava Prasad TS, Goel R, Kandasamy K, Keerthikumar S, Kumar S, Mathivanan S, Telikicherla D, Raju R, Shafreen B, Venugopal A, Balakrishnan L, Marimuthu A, Banerjee S, Somanathan DS, Sebastian A, Rani S, Ray S, Harrys Kishore CJ, Kanth S, Ahmed M, Kashyap MK, Mohmood R, Ramachandra YL, Krishna V, Rahiman BA, Mohan S, Ranganathan P, Ramabadran S, Chaerkady R, Pandey A (2009) Human Protein Reference Database – 2009 update. Nucleic Acids Res 37:D767–D772 Chen JY, Mamidipalli S, Huan T (2009) HAPPI: an online database of comprehensive human annotated and predicted protein interactions. BMC Genomics 10(Suppl 1):S16 Persico M, Ceol A, Gavrila C, Hoffmann R, Florio A, Cesareni G (2005) HomoMINT: an inferred human network based on orthology mapping of protein interactions discovered in model organisms. BMC Bioinformatics 6(Suppl 4):S21 Aranda B, Achuthan P, Alam-Faruque Y, Armean I, Bridge A, Derow C, Feuermann M, Ghanbarian AT, Kerrien S, Khadake J, Kerssemakers J, Leroy C, Menden M, Michaut M, Montecchi-Palazzi L, Neuhauser SN, Orchard S, Perreau V, Roechert B, van Eijk K, Hermjakob H (2009) The IntAct molecular interaction database in 2010. Nucleic Acids Res 38(Database issue):D525–31 Lehne B, Schlitt T (2009) Protein-protein interaction databases: keeping up with growing interactomes. Hum Genomics 3:291–297 Mathivanan S, Periaswamy B, Gandhi TK, Kandasamy K, Suresh S, Mohmood R, Ramachandra YL, Pandey A (2006) An evaluation of human protein–protein interaction data in the public domain. BMC Bioinformatics 7(Suppl 5):S19 Galperin MY, Cochrane GR (2009) Nucleic acids research annual database issue and the NAR online molecular biology database collection in 2009. Nucleic Acids Res 37:D1–D4 Brusic V, Zeleznikow J (1999) Knowledge discovery and data mining in biological databases. Knowledge Eng Rev 14:257–277 Jeong SK, Kwon MS, Lee EY, Lee HJ, Cho SY, Kim H, Yoo JS, Omenn GS, Aebersold R, Hanash S, Paik YK (2009) BiomarkerDigger: a versatile disease proteome database and analysis platform for the identification of plasma cancer biomarkers. Proteomics 9:3729–3740
INDEX
A Absorption/covalent printing .......................................... 86 Accutase.... ........................................................171, 175, 178 Acetone..... ................................................................171, 173 Acetylation .............................................214–217, 220–222 Acuity 4.0..........................................................144, 146, 147 Adherent cells ........................... 53, 167, 171, 175, 178, 182 Affinity...... ...............17, 29, 57, 58, 71, 83, 86, 98, 108, 109, 116, 120, 122, 146, 149, 155–157, 159, 186, 190, 213, 250, 259, 276, 297, 322, 351 printing ................................................................ 83, 86 reagent. ............................... 57, 149, 186, 250, 276, 322 Allo-antibodies .................................................... 82, 87–88 Amino acid activation ..................................................112, 123, 124 coupling ............................................110, 113, 123, 124 pentafluorophenyl esters .................................. 107, 112 Ampicillin............ 91, 95, 102, 187, 258, 260, 261, 265, 270 Analytical methods ........................... 38, 316–317, 338, 341 Androgen receptor (AR) ...............................168–170, 176, 177, 354, 355 Angiogenin (Ang) ......................................................... 6, 7 Anti-beta galactosidase .................................................. 179 Antibodies.. ............3–12, 15–27, 29–39, 42, 43, 47–52, 69, 81–102, 108–110, 118, 119, 121, 122, 130, 138, 140–141, 143, 145, 146, 149, 150, 153, 155–157, 159, 179, 186, 189–190, 194–199, 202, 217, 219, 220, 222, 227–242, 245, 250, 275–277, 279–285, 293, 295–297, 299, 307, 308, 311, 312, 317, 321–327, 329–332, 338, 340–342, 344–345, 350, 357 autoantibody ......130, 131, 142–143, 150, 151, 155–156 profiling ............................................130, 155, 157–158 screening ...................................................155, 240, 245 validation ...................................... 51–52, 240, 250, 276 Antibody-based detection...........4–6, 10, 11, 16, 30, 32, 33, 81–102, 118, 119, 186, 202, 222, 227–238 Antigens.....................4–5, 9, 11, 12, 16, 20, 23, 24, 81–102, 129–147, 149–153, 155, 156, 158–160, 181, 186, 191, 196, 227, 228, 230, 234–237, 241, 243, 244, 281, 337, 350–352, 356–358, 361 isoform............................................................. 101, 359 Anti human IgG Alexa647 conjugate........................ 91, 93
Anti-phosphoprotein antibodies ............................... 43, 51 Aptamer characterization ................................................... 57–59 high-throughput optimization............................. 57–65 length minimization .................................................. 63 microarray ............................................................ 57–65 AR. See Androgen receptor Array format .......................................................17, 75, 166 Autoantibodies. ................130, 131, 142–143, 149–151, 153, 155–156, 159, 351 immunoglobulin G labeling of .......................... 133, 138–140, 143, 146, 189, 195, 198, 229–230, 235, 238 purification of .................... 132–133, 137–138, 144, 145, 243, 245, 324 profiling ............................................130, 142–143, 155 Auto-fluorescence................... 123, 171, 172, 222, 277, 297 Autoimmune disease ............................................. 130, 227 Autoimmunity ........................................130, 131, 227, 239 Automation ..............................98, 106, 114, 142, 147, 171, 282–283, 286, 288, 291, 296, 342, 344 AxioVision LE ...................................................... 172, 178
B Bacterial adhesins ............................................................ 72 Bacterial invasion............................................................. 44 Bait..................130, 166–170, 172–174, 177–180, 182–183, 240, 250, 275, 276 BCL2............................................................................. 150 Bead array..................................................... 29–36, 227–238 BIOCCD image reader ..........................167, 169, 171, 178 Bioinformatics ................................... 85, 87, 144, 146, 346, 350–352, 362 Biological pathway..........................................239, 358–360 Bioluminescence ............................................................ 304 Biomarker................................ 8, 15–27, 129–147, 152, 158, 228, 311, 337, 349–351, 361 Biomolecular interaction ........................202, 304, 311, 316 Biotin............................................. 16, 31, 34, 36, 116, 146, 186, 229, 235, 241–244, 250–252, 277, 281, 282, 314, 315 Biotinylation. ..............................6, 10, 11, 18, 22, 186, 236, 277, 279, 282, 313, 314
Catherine J. Wu (ed.), Protein Microarray for Disease Analysis: Methods and Protocols, Methods in Molecular Biology, vol. 723, DOI 10.1007/978-1-61779-043-0, © Springer Science+Business Media, LLC 2011
365
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 366 Index Bovine serum albumin (BSA).............. 8, 12, 17, 41, 43–46, 49, 59, 61, 69, 70, 75, 88, 90, 93, 122, 133, 140, 145, 204, 216, 219, 222, 229, 234, 235, 266, 279, 283, 284, 313–315, 332 Buffer blocking...............................12, 17, 19, 41, 47, 48, 52, 69, 70, 74, 75, 93, 99, 108, 109, 118, 119, 122, 132, 133, 139, 193, 196, 204, 205, 217, 228, 244, 248, 252 crossdown ........................................................ 109, 119 incubation ....................75, 109, 120, 121, 133, 140, 145 regeneration ......................................109, 120, 323, 328 spotting ......................................... 41, 47, 108, 118, 323 transfer.............................10, 11, 18, 118, 140, 244, 248 tris-buffered saline (TBS) ..........................91, 122, 204, 217, 244, 252 tween-tris buffered saline (T-TBS) ..................108, 122, 220, 244, 248, 279, 282, 283, 297
C Cancer breast... ............................................... 150, 158, 353–355 melanoma .................................................150, 355, 356 prostate .....................................................130, 150, 355 vaccine target ........................................................... 358 Candida..... ..................................................................... 40 CAPPIA. See Cell array protein–protein interaction assay Capping........................................... 30, 107, 113, 123, 124 Carbohydrates ......................................................15, 67, 71 Catalyzed signal amplification (CSA) ................... 277, 281 CDA. See Concentration dependent analysis cDNAs............... 94, 150, 154–156, 230, 258, 262, 270, 357 Cell array protein–protein interaction assay (CAPPIA) .................................. 165–183 Cell arrays... ........................................................... 165–183 Cell culture ...............................9, 38, 42–43, 52, 87, 91–93, 95–98, 108, 118, 120, 131, 134–135, 145, 168, 171, 176, 178, 181, 260, 262, 284, 323 293 Cell line ......................94, 169, 171, 175, 176, 181, 182 Cell lysate arrays .............................................130, 143, 322 Cell monolayer .......................................166, 167, 181, 182 Cell signaling ..........................37, 38, 40, 44, 155, 228, 229, 244, 279, 280, 284 Cellulose membrane b-alanine.................................................................. 107 amino-alkyl linked membranes................107, 110–112, 117, 123–124 amino functionalization ............................110, 123–124 CAPE membrane ............................................ 110, 111 cleavage.....................................................108, 115–118 esterified membranes ........................110–111, 123–124 preparation........................................107–108, 110–118 TFA-soluble membrane .......................................... 109 TOTD membrane ........................................... 110, 111
Charge-coupled device (CCD) camera .............50, 53, 178, 284, 285, 299, 311, 314, 316 Chemiluminescence....................... 109, 119, 122, 125, 229, 232, 244, 249, 252, 304, 321 Chemokine .................................................................. 5, 11 Chip blocking ...............................................41, 47–48, 219 Chlamydia...............................................................................40 Chronic lymphocytic leukemia .............................. 241, 242 CLAMP kinetic analysis software ................................. 330 Cloning...................................68, 71, 72, 76, 170, 172, 179, 187, 190–192, 231, 250, 251, 257, 258, 259, 261–263, 266, 270 Colorimetric assay ......................................................... 285 Competitive assay ...........................................311, 314–317 Concentration dependent analysis (CDA) ............... 85–86, 338–343, 345, 346 Conjugation ............................................... 8, 10, 18, 24, 41, 43, 46, 59, 69, 82, 91, 93, 110, 124, 155, 157, 179, 188, 199, 214, 230, 232, 244, 281, 285, 308, 309, 312, 313 Contact printing ............................................................ 181 Contact protein printer.................................................... 93 Co-precipitation .................................................... 239–254 COS7........ ........................................................171, 175, 176 Coupling cycle ................................ 105, 113–115, 117, 123 Cre. See Cre/loxp Cre/loxp.................................................................258, 269 CSA. See Catalyzed signal amplification Cy3.....................................8, 35, 59, 70, 134, 141, 157, 158, 188, 189, 193, 195–199, 222, 277, 310–312 Cy5...................................41, 42, 70, 72, 110, 121, 134, 141, 146, 190, 195–199, 222, 277, 310, 312, 313 Cytokine.................................................................. 4–7, 10, 11
D DAPI. See 4’,6-Diamidino–2-phenylindole dihydrochloride Dark quencher ....................................................... 309, 314 Data analysis.......................... 10–11, 22–23, 39, 42, 49–50, 62–63, 77, 93, 100–101, 134, 137, 144, 147, 153, 204, 207–209, 219, 279, 281, 288–296, 298, 324, 330–331, 337–346, 350, 351 Database..............................................................62, 349–362 DBD. See DNA binding domain Denatured nickel affinity chromatography ...........83, 92, 97 Deprotection .................................. 108, 111, 115–117, 124 Detection alkaline phosphatase (AP) ............................... 109, 118 antibody ......................3–13, 16–23, 30, 33, 43, 81–102, 118, 119, 195, 202, 222, 227–238, 277, 308, 322, 350 chemiluminescence ..........................................109, 119, 122, 125, 232 fluorescence ......................... 17–23, 109–110, 120–121, 171–172, 179, 220, 277, 285, 309, 311, 313
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 367 Index horse-radish peroxidase (HRP, POD) .................. 5, 69, 75, 77, 91, 96, 98, 109, 118, 119, 155, 157, 188, 228–230, 232, 244, 248, 251, 252, 276–277, 281, 283, 285 methods ........................5, 108–110, 118–121, 146, 276, 279–281, 285, 297, 313 staining ..................................... 109, 119, 168, 284, 322 X-ray film ........................................................ 119, 216 DGC. See Drosophila gene collection Diabetes..............................................................6, 150, 158 Diagnosis.... ................................................8, 311, 350, 358 4’,6-Diamidino–2-phenylindole dihydrochloride (DAPI) .................................171, 177, 178, 182 Diastolic blood pressure..................................................... 7 Differential expression analysis .............................. 340–342 DLI. See Donor lymphocyte infusion D-MEM. See Dulbecco’s Modified Eagle’s Medium DNA binding domain (DBD)........ 166, 170, 172, 175, 179 DNA method .................................................172, 173, 186 DNA microarray reader ......................................... 171, 178 DNA microarrays ........................ 57–65, 81, 154–156, 166, 167, 186, 198, 213–214, 217, 276, 311, 312, 321, 322, 338 DNA polymerase.................................. 68, 73, 76, 260, 270 Donor-acceptor pairs ............................................. 308–310 Donor lymphocyte infusion (DLI) .................241, 242, 244 Dose-dependence ...........................................168, 176, 177 Drosophila gene collection (DGC) ....................... 258, 263 Dulbecco’s Modified Eagle’s Medium (D-MEM) ............................171, 175, 323, 325
E EBNA. See Epstein–Barr virus nuclear antigen EC-buffer.... .......................................................... 171, 173 ECL. See Enhanced chemiluminiscence E.coli clones ................................................................... 259 Effectene transfection reagent ...............166, 167, 170–173, 180, 182 EGFP. See Enhanced green fluorescent protein ELISA. See Enzyme-linked immunosorbent assay Endo-Free Plasmid Maxi Kit ................................ 169, 172 End-stage renal disease (ESRD) ................................... 5–6 Energy transfer ...................................................... 303–317 Enhanced chemiluminiscence (ECL)......................96, 109, 119, 122, 125, 229, 232, 244, 249, 321 Enhanced green fluorescent protein (EGFP) ............... 167, 171–175, 178, 179, 310 Enteropathogenic Escherichia coli (EPEC) ...................... 40 Enzyme-linked immunosorbent assay (ELISA)........... 4, 5, 69–71, 74–75, 77, 82, 84, 85, 88, 151–152, 155, 228–237 EPEC. See Enteropathogenic Escherichia coli Epithelial cells .......................................................... 38–40, 52, 130
Epitope tags GST...... ............................................................... 83, 84 V5............ ........................................... 83, 88, 89, 94, 95 6xHis.... .............................. 83, 86, 88, 94–95, 269–270 Epoxide coating ..................................................... 108, 122 Epstein–Barr virus nuclear antigen (EBNA) .........153, 158, 236, 237, 242 ESRD. See End-stage renal disease Excited state .......................................................... 304–306 Expression clone collection............................................ 259 Expression profiling..........................................85, 330, 350 Expression-ready clone collection.......................... 257–271 Expression system coupled transcription-translation in vitro ................ 156
F FAST slides ..............................................93, 188, 194, 217 FBS. See Fetal bovine serum Fetal bovine serum (FBS) ............. 9, 12, 131, 171, 175, 323 Filter paper ......107, 108, 110–112, 114, 118, 120–122, 177 FLAG............. 186, 188, 191–192, 198, 217, 220, 229, 230, 231, 233, 237, 250, 291 FLAG-HA .................................................................... 222 Flow cell ................................. 136, 322, 324, 327–329, 332 Fluidic cells................................................................ 42, 48 Fluorescence ..................... 4, 6, 7, 10, 17–23, 30, 35, 42, 46, 49–50, 53, 61–63, 65, 77, 87, 89, 101, 109–110, 120–121, 123, 141, 147, 169, 171–172, 174, 175, 178–182, 197, 220, 277, 278, 285, 297, 304–307, 309, 311, 321, 322, 326, 344–345 lifetime......................................................312, 314–316 Fluorescent label ................................. 48, 59, 134, 199, 313 Fluorochrome ............................................................ 82, 88 Fluorometric assay ................................................. 275–299 Fluoromount-G..................................................... 171, 178 Fmoc building block ...........................................112, 116, 117 chemistry ................................................................. 216 removal of protecting-group ............................ 113, 114 Formaldehyde ........................................................ 171, 176 Förster distance .................................................... 306–307, 309, 310 Förster resonance energy transfer .......................... 303–317 FPLC chromatographic system ........................... 92, 94–95 Free peptides..................................................106, 110, 115, 117–118, 124 Fusion proteins (amino and carboxy terminal) ............. 258, 259, 269
G GAD65...... ...............................................................150, 158 GAL4........ .....................................................172, 175, 179 GAL (file format). See Gene array list Gal4-pZsGreen ..............................................172, 175, 179
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 368 Index GAPS II coated slides ........................................... 180, 181 Gateway system/technology .................. 179, 187, 190, 231, 250, 257, 259, 270 Gelatin.................................... 170, 172, 173, 178, 180, 332 Gel electrophoresis ................. 244, 259–261, 264, 278, 351 Gene array list (GAL) ................... 142, 207, 209, 221, 222, 261, 327, 330, 339, 341, 342 Gene expression.............3, 81, 278, 310, 351–353, 357–359 profile....................................................3, 352, 353, 358 GenePix..... ...........................70, 75, 93, 101, 102, 141, 142, 189, 190, 193, 195, 197, 204, 207–210, 216, 339, 341–344 GenePix Pro 6.0 .......................................22, 146, 221, 222 GenePix results (GPR) .......................... 101, 142, 144, 147, 209, 210, 342, 344 Gentamycin protection assay ........................................... 53 Glass slides .......................16, 71, 86, 88, 93, 108, 115, 118, 121, 150, 159, 166, 167, 173, 174, 176, 178, 180, 181, 201–202, 240, 290, 297, 315, 324, 327 aldehyde surface ........................................108, 122, 124 epoxide coating ........................................................ 108 Glutathione-S-transferase (GST) ............... 71, 83, 84, 155, 157, 186, 217, 219, 228, 230–237, 344–345 Glycan........................ 15–16, 19, 22–24, 27, 67, 68, 71, 322 Glycerol..........8, 26, 69, 91, 92, 98, 101, 108, 121, 131, 132, 137, 215, 217, 218, 244, 267, 312, 323 Glycomics... ............................................................... 67–77 Glycoprofiling ............................................................... 350 Glycoprotein....................................... 15–17, 19, 20, 71, 74 Glycosylation .................5, 15–27, 68, 72, 82, 240, 244–245 GPR (file format). See GenePix results Graft versus host disease (GVHD) .................................. 87 Graft versus leukemia (GVL) .......................................... 87 GST. See Glutathione-S-transferase GTPases........................................................................38, 40 GVHD. See Graft versus host disease GVL. See Graft versus leukemia
H HEK 293... .............................................171, 175, 176, 181 HEK 293 T ............................ 169, 171, 175, 176, 181, 182 HeLa cells 38, 40, 42–44, 51, 52, 171, 175, 176 HepG2...... ................................................171, 175, 176, 325 HepG2 cell line ..................................................... 323, 324 High-throughput............................. 3–12, 37, 67, 129, 201, 202, 204, 228, 231, 276, 321, 337, 350–353, 361 High-throughput screening (HTS) .......................... 15–27, 165–183, 229, 240, 244, 308, 312 HIV-p24...............................................................88, 94, 97 HLA ligand ................................................................... 358 Hormone-dependence ........................................... 169, 172 Host cell signalling ...............................................38, 40, 44 Host-pathogen interaction .................................. 37–53, 68
HTS. See High-throughput screening Human...... .......................16, 24, 29, 40, 59, 81–83, 88, 130, 131, 134, 145, 191, 201, 202, 215, 259, 278, 349, 351, 354, 355– 361 Human blood .......................16, 24, 99–100, 150, 230, 232, 235, 245 Human IgG................................. 84, 88, 91, 145, 153, 230, 235, 238 H-Y antigens/H-Y proteins DDX3X ......................................................... 87–88, 94 DDX3Y ......................................................... 87–90, 94 EIF1AX ....................................................87–88, 94, 97 EIF1AY ....................................................87–88, 94, 97 RPS4X........................................................... 87–88, 94 RPS4Y ........................................................... 87–88, 94 UTX.... .......................................................... 87–88, 94 UTY..................................................................87–90, 94 ZFX..... .......................................................... 87–88, 94 ZFY...................................................................87–90, 94 Hybridization .........................17, 20, 24, 25, 38, 39, 59, 65, 70, 146, 155, 157, 159, 166, 168, 175, 188, 189, 193–197, 308, 311, 312, 314, 353 Hydroxyflutamide (OH-Flu) ................................ 176, 177
I IA2, 150 IC/PBS. See Interstitial cystitis/painful bladder syndrome IgG purification.....................................132–133, 137–138, 145, 243, 245 IL–12. See Interleukin–12 Image acquisition..................................... 50, 209, 279, 280, 285–288, 298, 315 ImageJ.....................................................242, 249–250, 252 Imidazole... ................................. 86, 88, 91–93, 95–98, 107 Immunoassay ................................... 5, 30, 50, 51, 202, 275, 308, 316–317, 322 Immunoblotting .......................................38, 249, 250, 278 Immunoglobulin E .......................................................... 59 Immunoprecipitation ...............................82, 240, 241–243, 245–247, 249–250 Immunoprofile ...................................................... 149–159 Immunostaining .....................................279–284, 312, 357 Individualized therapy ................................................... 278 In situ activation .............................................112, 123, 156 InstrumentONE high-performance ...................... 171, 174 Interleukin–12 (IL–12) ................................................. 6–7 Interstitial cystitis/painful bladder syndrome (IC/PBS) ..............................130, 131, 143, 144 Intracellular parasites ................................................. 37–38 Invasive bacteria .................................................. 37–38, 40 In vitro infection .............................................38–44, 52–53 IPTG. See Isopropyl-b-D-thiogalactoside IRB approval ......................................................... 132, 151 Isopropyl-b-D-thiogalactoside (IPTG) ...............91, 95, 96
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 369 Index K Kinase........ ..........................38, 40, 199, 214, 308, 313, 355 Kinase-substrate interaction ...........................201–212, 337 Kinetics...................................................115, 243, 321, 322
L Label-free............................................................24, 321–332 Lab-Tek™ Chamber Slide™, 181 LacZ. See b-galactosidase Lanthanide .............................................309, 310, 314, 317 Large T antigen of SV40 ............................................... 181 LB-Ampicillin (LB-Amp) plates .................................... 95 LB-Amp plates. See LB-Ampicillin (LB-Amp) plates LBD. See Ligand binding domain Lectin........ ............................................... 15–27, 67–77, 322 microarray .................................................15–27, 67–77 l-glutamine ...............................................40, 171, 175, 323 Library.............................................. 168, 178, 270, 283, 312 Ligand binding domain (LBD) ...... 168, 169, 170, 176, 177 Linker biotin...........................................................................116 coupling ....................................................116–117, 122 hydrazinobenzoic acid (HBA) ......................... 116, 117 Lipid bilayer array.......................................................... 312 Lipid-DNA method .............................................. 172–173 Listeria....... ........................................................................40 Loxp. See Cre/loxp Luminex........................... 5, 30, 35, 228–230, 232–235, 237 Lysates......... ... 29, 37–41, 43–45, 47, 49, 51, 74, 76, 82, 95, 98, 130–137, 140, 142–145, 149, 156, 190, 192, 194, 228, 230, 231, 237, 242, 243, 245–247, 251, 278, 279, 284, 289, 322, 332, 350 Lysis buffer .............................41, 44, 52, 53, 69, 74, 76, 92, 96–98, 102, 131, 134, 345 Lysozyme..............................................................69, 74, 97
M Macroarrays ....................................106, 107–108, 110–115 MALDI. See Matrix-assisted laser desorption/ionization Mammalian cells ....................... 52, 83, 166–168, 175, 179, 180, 222, 239, 241, 242 Mammalian protein-protein interaction trap (MAPPIT) .................................................. 179 Mammalian two hybrid ......................................... 165–183 MAPK. See Mitogen-activated protein kinase MAPPIT. See Mammalian protein-protein interaction trap Mass spectrometry (MS) ............................. 16, 17, 24, 145, 147, 213, 351 Mastoparan.................................................................... 116 Matrix-assisted laser desorption/ionization (MALDI) ................................................ 24–26 Mean fluorescence intensity (MFI) .....................10, 32, 84, 87, 101, 102, 236
Medroxyprogesterone acetate (MPA) .................... 176, 177 Melanoma inhibitor of apoptosis (ML-IAP) ................ 150 Membrane b-alanine................................... 107, 110, 113, 121, 123 amino-alkyl linked ............ 107, 110–112, 117, 123–124 amino functionalization ................................... 110, 121 CAPE.. ............................................................ 110, 111 cleavage.....................................108, 115–118, 121–122 esterification .................................................... 110, 121 PEG....................................................................115, 124 TFA-soluble .............................................108, 109, 124 TOTD ............................................................. 110, 111 Methallothionein inducible promoter ........................... 257 MFI. See Mean fluorescence intensity Mfold....................................................................62, 63, 65 mHA. See Minor histocompatibility antigens Microarrays.............................4, 16, 29, 41, 58, 67, 81, 106, 129, 149, 166, 185, 201, 213, 227, 239, 275, 310, 321, 337, 350 analysis..........................10, 15–27, 30, 67–77, 106, 115, 134, 144, 147, 155, 199, 276, 280, 288–295, 337–346 forward phase microarrays ....................29–30, 129–130 printer 4 ................................................. 70, 93, 98, 118, 132, 216, 218, 324 printing .............................9, 12, 17, 86–87, 93, 98, 132, 137, 150, 152, 153, 188–189, 192, 193, 197, 216–218, 277, 326, 344 printing and blocking of ...........................137, 188, 193 quality control ...................................188–189, 193, 322 reverse phase microarrays............. 29–30, 129–130, 227, 275–299 scanner ..................................8, 10, 12, 62, 93, 100, 121, 134, 141, 216, 316 Micro flow system ........................................................... 48 Microtiter plates (MTP) ...........................6–12, 30, 32–34, 41, 45, 48, 69, 118 Minimal binding domain ................................................ 63 Minor histocompatibility antigens (mHA) ..................... 87 Mitogen-activated protein kinase (MAPK) .................... 40 pathways .................................................................... 40 ML-IAP. See Melanoma inhibitor of apoptosis MOI. See Multiplicity of infection Molecular interaction analysis ............................... 359, 360 Monolayer....................................... 166, 167, 178, 181, 182 MPA. See Medroxyprogesterone acetate MS. See Mass spectrometry MTP. See Microtiter plates Multiplex.......................... 4–6, 18, 22, 29–30, 58, 81, 82, 86, 88, 129, 201, 227–238, 276, 280, 287, 312, 316 Multiplexed assay .................................................... 35, 228 Multiplicity of infection (MOI) .......................... 44, 52–53 Multivariate approach ....................................................... 3 Mycobacterium .......................................................40, 42–43
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 370 Index N Nanoparticle ...........................................309, 312–313, 317 NAPPA. See Nucleic acid programmable protein array Native nickel affinity chromatography ...........83, 92, 97, 98 Nebulization .............................................................. 41, 48 NF-kB....... ............................................................ 172, 175 Non-contact microarray spotter....................................... 41 Non-contact piezo-dispensing system ........................... 171 Nonradiative decay ........................................................ 305 N-terminal domain (NTD) ............................168–170, 176 Nucleic acid programmable protein array (NAPPA) ............................................. 149–159
O Open reading frames (ORFs) ............................83, 94, 101, 172, 191, 192, 194, 239–240, 250, 257–259, 262, 263, 269, 270 Organic compounds .............................................. 105–106
P p53......................................................... 150, 158, 175, 191, 195, 197, 237, 355 pAD-SV40T ..................................................169, 172, 175 pAD-TRAF ...................................................169, 172, 175 Pathogen.........................................37–53, 67–68, 310–312 Pathogenesis .......................................................... 239, 359 pBD-NF-kB ..................................................169, 172, 175 pBD-p53...........................................................169, 172, 175 PBXL–3..................................................................5, 6, 8, 10 PC–3......... ....................................................... 171, 175–176 pcDNA4-EGFP.............................................172–174, 178 pcDNA4/HisMax TOPO ............................................. 172 pCMV-AD ........................................................... 170, 172 pCMV-BD .................................................................... 172 PCR. See Polymerase chain reaction PCR primer ............................................259, 262, 269, 270 Pellet.......... ..... 32, 34, 53, 64, 74, 76, 95–97, 102, 134, 145, 191, 192, 231, 235, 247, 268 Penicillin/streptomycin ..................... 40, 131, 171, 175, 323 Peptide array macroarray ................................................. 106, 114 microarray ..................................106, 109–110, 115, 118, 120–121 free peptide ....................... 106, 110, 115, 117–118, 124 immobilization ................................................ 116, 124 reconstitution ........................................................... 118 solution .....................................................115, 120, 124 synthesis .......................................................... 105–125 transfer..................................................................... 118 unprotected peptides................................................ 116 pGAL/lacZ ................................................................... 179
Phosphoprotein/protein ratio .......................................... 51 Phosphorylation ................................ 38–40, 43, 51, 52, 82, 214, 240, 278, 312, 313 status........................................................................ 312 Photo multiplier tubes (PMT) .................. 12, 22, 100, 141, 142, 146–147, 197, 286–287, 298, 316, 343 pIRES2-EGFP ............................................................. 172 Planar waveguide excitation ...................................... 50, 53 Plasma....... .........................30, 32, 34, 35, 59, 81–102, 108, 157, 159, 228, 241, 243, 361 Plasmid...... .......................40, 68, 73, 76, 83, 101, 150, 155, 166, 167, 169–170, 172–175, 179–181, 186, 187, 189–195, 198, 228, 229, 231, 241, 243, 246, 250–251, 265 PMT. See Photo multiplier tubes Poly-l-lysine ...........................................171, 173, 180, 181 Polymerase chain reaction (PCR) ..................34, 68, 71–73, 76, 85, 94, 168, 172, 186, 187, 190, 192, 245, 246, 250, 251, 258–267, 269, 270, 276, 309 Post-translational modifications glycosylation ...................................15–27, 82, 244–245 phosphorylations .........38–40, 43, 51, 82, 214, 240, 278 Pre-activated derivatives ........................................ 107, 112 Prey............ ..............166–170, 172–174, 177–180, 182, 183 Prey-reporter- (PR-) slides .....................168, 169, 182–183 Primer design ........................................................ 260, 263 Printing buffer ...................8, 9, 11, 17, 18, 88, 90, 197, 215 Printing substrates glycosylation ........................................................ 16, 72 nitrocellulose...................................................... 86, 278 Probing solution ............................................................ 108 ProCAT..... ............................................................ 210, 211 Prognosis... ........................................................................ 8 Prospector.. .............................................207, 341–342, 344 Protein....... .............................3, 15, 29, 38, 57, 67, 81, 106, 129, 149, 165, 185, 201, 213, 227, 239, 257, 275, 304, 321, 337, 349 arrays...............................37–53, 94, 140, 150, 166, 207, 227, 241, 311, 344, 350 binding ...........................64, 67, 86, 115, 119, 120, 132, 156, 179, 201, 214, 218, 235, 245, 321, 322 detection .......................................................... 186, 276 expression ...............................20, 53, 69, 74, 76, 82, 83, 91–93, 95–98, 171, 192, 193, 195, 198, 228, 231, 232, 237, 241–242, 250, 257–259, 270, 311, 322, 337, 340, 350–352, 357, 361 gel quantification ....................................... 44, 303–317 kinase.... .............................. 40, 199, 202, 209, 308, 337 kinase B ..................................................................... 40 labeled protein .............61, 109, 119, 122, 251, 252, 307 microarray .............29, 30, 82–86, 94, 98–100, 152–155, 185–199, 201–203, 208, 213–222, 215, 239, 241, 275–299, 311, 337–346, 350
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 371 Index antigen ....................................................... 129–147 blocking .................................. 12, 99, 108, 205, 324 DNA.........................57–65, 81, 166, 167, 171, 178, 186, 190, 191, 193, 194, 197, 213–214, 217, 276, 311, 312, 338 nucleic acid ................................................ 150, 155 printing .....................................86–87, 93, 216–218 replicates ...................................................6, 10, 152 reproducibility.......................................47, 156, 346 scanning ...................... 134, 141–144, 146, 207, 311 zone variation .....................................152–153, 155 phosphorylation .................. 38, 202, 207, 208, 278, 284 profiling ........... 30, 35, 53, 130, 185, 214, 322, 323, 350 protein interaction (PPI) .................105–125, 165–183, 185–199, 215, 240, 244, 276, 303–317, 312, 317, 337, 352, 358, 359, 361 slide...........................................................216, 293–295 solution ....................................... 95, 108, 109, 297, 332 synthesis .................................. 156, 186, 192, 195, 198, 241, 243, 245–246 translation ................................................................ 231 Protein kinase B (Akt) ..........................................40, 51, 52 Proteomics.. ........................29, 37, 129, 227, 228, 257–271, 278, 279, 349–362 ProtMAT... .....................................................341–343, 345 Protoarray™ .......... 83, 84, 85, 202, 207, 338, 341, 342, 344 PR-stable-bait assay ...................................................... 182 PR-trans-bait assay........................................................ 182
Q QuadriPERM® .........................93, 108, 118, 120, 145, 171, 176, 181, 182, 189, 205 Quantum yield....................................................... 306, 307
R R1881............ ......................................... 168, 169, 176, 177 Rabbit reticulocyte lysate .......................156, 190, 192–194, 228, 230, 237, 243 Radiolabel....................................... 202, 204, 208, 209, 220 Recombinant antigen arrays ............................ 79–102, 130 Recombinant lectins .................................................. 67–77 Recombinant protein expression..........................82, 91–93, 95–98, 228, 230 Recombination ...............................................179, 257, 259 Referenced fluorescence intensity (RFI) .................... 49–51 Referencing ..........................21, 41, 43, 46, 47, 49, 77, 153, 309, 316, 330, 346 Regeneration ...........109, 120, 125, 322, 323, 328, 330, 331 Regions of interest (ROI) ...............................291, 329–331 Reporter..... .. 30, 35, 166–169, 172–176, 179, 180, 182, 277 Reverse-phase protein arrays ......................................... 227 Reverse phase protein microarray .......................... 275–299 Reverse protein arrays (RPA) .....................37–53, 129–130
Reverse transfection.......................................169, 171, 172, 175–176, 178–181 RFI. See Referenced fluorescence intensity ROI. See Regions of interest RPA. See Reverse protein arrays
S Salmonella.. ..........................................38, 40, 42–43, 51–53 Salmonella-containing vacuole (SCV) ............................. 40 SAM. See Significance analysis of microarrays Sandwich style immunoassay............................................. 5 ScanArray Express ......................................................... 146 SciFlexArrayer ....................................................... 171, 174 sciFlexArrayer piezo–dispensing system S5................... 174 Screening.......................................................................15–27 SCV. See Salmonella-containing vacuole SDS-PAGE. See Sodium dodecyl sulfate polyacrylamide gel electrophoresis Selection marker ampicillin ......................................................... 258, 270 carbenicillin ..................................................... 260, 261 chloramphenicol .........................................40, 187, 260 SELEX. See Systematic evolution of ligands by exponential enrichment Sensorgram .....................................................330, 331, 332 Sequencing .........................18, 58, 60, 62–65, 82, 179, 186, 190–192, 239, 241, 250–252, 257–260, 262, 263, 267–270, 350–353, 356–359, 361 Serum............ 3–9, 15–27, 29–30, 34, 40, 81–102, 121, 130, 132, 134, 138, 144, 150, 154–159, 227, 228, 230, 232, 235, 240–243, 246, 279, 297, 322, 325, 332, 350, 357 screening ........................... 153, 155, 159, 240, 244, 245 screening study .........................................151, 158–159 Shigella....... ...................................................................... 40 Signal corrections .................................................. 178, 339 Signaling................. 15, 37–38, 40, 42–44, 49, 51, 155, 214, 228, 229, 244, 278–280, 284, 337, 358–360 Significance analysis of microarrays (SAM) ..................................338, 340, 342, 346 Single nucleotide polymorphisms (SNPs) .....101, 310, 311, 351–356, 354, 356, 358, 359, 361 Single-stranded DNA (ssDNA) ...........................58, 59, 62 SNPs. See Single nucleotide polymorphisms Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) ........................ 74, 76, 77, 95, 96, 98, 244, 246, 247, 252 Software.................... 10, 26, 42, 60–62, 70, 75, 87, 93, 101, 102, 121, 134, 137, 142, 144, 146, 147, 172, 197, 204, 207, 209, 249, 260, 262, 280, 285, 286, 288–296, 324, 325, 327, 328, 330, 331, 338, 339, 341–342, 344, 346 Solid phase peptide synthesis .........................105, 112, 116
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 372 Index Solvents dichloromethane (DCM) .................107, 108, 114, 115 dimethylformamide (DMF) ............106, 107, 109–114, 123, 125 dimethylsulphoxide (DMSO)........................31, 69, 74, 107, 113, 178, 323, 325 ethanol (EtOH) ........................... 76–77, 106, 107, 109, 111–113, 115, 119, 123, 173, 231, 262, 268, 270, 297 methanol (MeOH) .................. 106, 107, 109, 111–113, 115, 119, 120, 123, 244, 248, 280 N-methylpyrrolidone (NMP) ..................106, 112, 113, 115–117 SopB........... ............................................................... 40, 51 SopE.......................................................................... 40, 51 SopE2........................................................................ 40, 51 SPOT synthesis cellulose membranes .........................105–108, 110–115 coupling solutions ............................................ 105, 112 SPOT macroarray.............................106–108, 110–115 SPOT robot....................................................41, 44, 46 SPOT technology.............................105–106, 114, 115 Spotting..... ........... 17, 19, 43–47, 53, 71, 93, 108, 110, 113, 116–118, 124, 125, 153, 173, 174, 192, 201–202, 213, 240, 323–325 microplates ...............................................41, 44, 46–47 SPR. See Surface plasmon resonance ssDNA. See Single-stranded DNA Stable transfection ......................................................... 182 Staining bromophenol blue (BPB)......................................... 107 Statistical analysis false discovery rate ........................................... 158, 341 overfitting data .................................................... 6, 158 Steady-state fluorescence ............................................... 309 Stealth microarray printhead ........................................... 93 Stealth micro spotting prints ........................................... 93 Stock solution .............................33, 46, 109, 112, 178, 246, 247, 260 Storage amino-acid solutions................................................ 112 membranes .............................................................. 123 microarray slides ...................................................... 118 Streptavidin ...................... 6, 8, 10, 16, 18, 32, 35, 116, 143, 146, 186, 229, 235, 238, 242, 252, 277, 279–283, 285, 313–316 Streptavidin-HRP .......................... 244, 248, 252, 281, 283 Stripping......................................... 109, 120, 260, 279, 298 Sucrose.............................108, 171, 173, 176, 244, 258, 260 Supernatant ................ 9, 44–46, 53, 74, 76, 83, 95–97, 102, 123, 134, 233, 235, 240, 247, 268, 328, 330 Surface plasmon resonance (SPR) angle scan ................................................................ 329 applications .............................................................. 322 binding .....................................................321, 327–330 blocking ....................................................323, 324, 327
data analysis ..............................................324, 330–331 equipment ................................................................ 324 imaging instruments ........................................ 325, 327 materials ...................................................325, 326, 329 printing arrays.................................................. 324–327 region of interest (ROI) ................................... 329–330 sample preparation........................................... 322, 325 Suspension bead array ............................................... 29–36 SV40.......... .................................................................... 181 Systematic evolution of ligands by exponential enrichment (SELEX)........................................................ 58
T T-cell epitope....................................................87, 357, 358 T7 expression..........................................154, 156, 231, 250 T24 human bladder cancer cell line ........130, 131, 134, 145 Time-resolved fluorescence ....................309, 311, 313, 316 Tissue culture ..........................9, 40, 43, 134, 240, 259, 324 Transcription ................................. 154, 156, 166, 175, 179, 186, 187, 189, 191, 228, 231, 242–245, 250, 251, 278, 322, 359 Transcriptional activating domain (AD) ............... 166, 179 Transfection .................... 166, 167, 170–176, 178–182, 241 Transformation .......................73, 74, 76, 83, 144, 147, 190, 259, 261, 262, 266–267, 345–346 Translocated bacterial effectors.................................. 38, 40 Transplantation...............................................84–85, 87–88 Triple-transfection ......................................................... 180 Troglitazone (TGZ) ...............................323, 325, 328, 331 TSA. See Tyramide signal amplification Tumor antigen ...................85, 237, 351, 352, 356–358, 361 Tumor-specific mutation ....................................... 353–357 Two hybrid ............................................................ 165–183 Type III secretion ............................................................ 38 Type III secretion effector ............................................... 38 Type IV secretion ............................................................ 38 Tyramide signal amplification (TSA) .............155, 157, 217
U Ubiquitin (Ub) ...............................................214, 217–219 E3 ligase .............................................38, 214, 217, 219 Universal cloning system ............................................... 258 Urea.................................................91, 92, 98, 109, 123, 131
V Vascular cell adhesion molecule–1 (VCAM–1) ............. 6, 7 VCAM–1. See Vascular cell adhesion molecule–1 VECTABOND™ reagent......................171, 173, 180, 181 Vectors......... ............. 68, 71, 73, 76, 95, 169, 170, 172, 178, 179, 186, 187, 191, 192, 228, 229, 231, 234, 236, 237, 241–242, 250, 251, 257–259, 261–263, 266, 269, 270 Viral antigens .................................................. 90, 235–237 Virulence factor ....................................................38, 39, 42 VPL slides.. ........................................................... 173, 180
PROTEIN MICROARRAY FOR DISEASE ANALYSIS 373 Index W
Z
Western blotting ...................................... 11, 52, 82, 85, 95, 96, 98, 106, 240, 242, 244, 246–250, 252, 297
Z-factor.......................................................................... 211 Z-score...... .....................................................211, 338–342 ZsGreen..........................................................171, 178, 179