Carbohydrate Microarrays: Methods and Protocols (Methods in Molecular Biology, v808)

Methods

in

Molecular Biology™

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK



For further volumes: http://www.springer.com/series/7651

Carbohydrate Microarrays Methods and Protocols Edited by

Yann Chevolot Université de Lyon, Institut des Nanotechnologies de Lyon – INL, UMR CNRS 5270, Site Ecole Centrale de Lyon, Ecully, France

Editor Yann Chevolot Université de Lyon Institut des Nanotechnologies de Lyon – INL – UMR CNRS 5270 Site Ecole Centrale de Lyon Ecully, France [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-372-1 e-ISBN 978-1-61779-373-8 DOI 10.1007/978-1-61779-373-8 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011938687 © Springer Science+Business Media, LLC 2012 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 d ­ issimilar 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. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface Carbohydrate microarrays have emerged in the late 1990s and early twenty-first century as a key technology for the deciphering of the glycospace. A carbohydrate microarray is a multiplex technology, where tens to hundreds of carbohydrate–protein interactions can be probed in parallel. Multiplexing is made possible, thanks to the immobilization of various carbohydrates at a discrete location on the surface of a material or on coded microspheres for example. This book, dedicated to carbohydrate microarray, is aimed to give the reader the theoretical and experimental clues for the fabrication and implementation of carbohydrate microarrays. The fabrication of carbohydrate microarrays requires (1) to obtain the carbohydrate probes (monosaccharides, oligosaccharides, polysaccharides, glycoconjugates, or glycoclusters), (2) to immobilize these probes, (3) to implement the protocols for biological/biochemical interaction with the desired target. Carbohydrate probes can be obtained from natural sources, by synthesis or are commercially available. The synthesis of oligosaccharides is beyond the scope of this book and requires specific skills. Nevertheless, an overview chapter is dedicated to glycochemistry as we felt that notions in the field are essential in particular for the introduction of derivatives (glycosides) for their immobilization. Indeed, immobilization of carbohydrates can be performed without further modifications of the carbohydrate or may require their conjugation with proteins, DNA, lipids, or the introduction of functional groups (thiol, amine, maleimide…). Immobilization can be performed through weak interaction (van der Waals, ionic interactions), covalent bonding or biochemical reactions (streptavidin/biotin, DNA/DNA hybridization). These immobilizations can be performed directly on unmodified surfaces (thiol glycosides on gold surfaces for example) or on chemically prepared surfaces. With the first two chapters, we have given an overview on carbohydrate microarray, carbohydrate chemistry. Chapter 3 is dedicated to sialic acids which display special biological and chemical properties. The following chapters deal with experimental protocols that present immobilization of various carbohydrates (oligosaccharides, glycoproteins, glyco‑ lipids, and polysaccharides) on different substrates (gold, glass, polymers…). Protocols span from the “straight forward” adsorption of unmodified polysaccharides/glycoconjugates from natural sources on commercial substrates to the synthesis of glycosides and their subsequent immobilization on chemically modified surfaces. Some materials used in certain protocols, such as glycosides or chemically modified surfaces are commercially available. Similarly, we have tried to illustrate different detection techniques (Mass Spectrometry, Surface Plasmon Resonance, Dual Polarization Interferometry, Fluorescence…) and ­applications (from lectin/carbohydrates applications to serum antibodies profiling, ­serodiagnosis, enzymes, whole viruses…). Hopefully, the catalogue of protocols will span most of the reader needs/applications. Finally, I gratefully acknowledge Pr. John Walker and all the authors for their contributions which have made the realization of this book possible. Ecully, France

Yann Chevolot

v

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v xi

  1 Recent Advances and Future Challenges in Glycan Microarray Technology . . . . . . José L. de Paz and Peter H. Seeberger   2 Chemical Synthesis of Carbohydrates and Their Surface Immobilization: A Brief Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Daniel B. Werz   3 General Consideration on Sialic Acid Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . Hongzhi Cao and Xi Chen   4 Synthesis of Azido-Functionalized Carbohydrates for the Design of Glycoconjugates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Samy Cecioni, David Goyard, Jean-Pierre Praly, and Sébastien Vidal   5 Polypyrrole-Oligosaccharide Microarray for the Measurement of Biomolecular Interactions by Surface Plasmon Resonance Imaging . . . . . . . . . . Julia Bartoli, André Roget, and Thierry Livache   6 Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis . . . . . . . . Fang Cheng and Daniel M. Ratner   7 Carbohydrate Microarrays for Enzymatic Reactions and Quantification of Binding Affinities for Glycan–Protein Interactions . . . . . . . . . . . . . . . . . . . . . . . Myung-Ryul Lee, Sungjin Park, and Injae Shin   8 Neoglycolipid-Based Oligosaccharide Microarray System: Preparation of NGLs and Their Noncovalent Immobilization on Nitrocellulose-Coated Glass Slides for Microarray Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yan Liu, Robert A. Childs, Angelina S. Palma, Maria A. Campanero-Rhodes, Mark S. Stoll, Wengang Chai, and Ten Feizi   9 Preparation of a Mannose-6-Phosphate Glycan Microarray Through Fluorescent Derivatization, Phosphorylation, and Immobilization of Natural High-Mannose N-Glycans and Application in Ligand Identification of P-Type Lectins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Xuezheng Song, Jamie Heimburg-Molinaro, Nancy M. Dahms, David F. Smith, and Richard D. Cummings 10 Production of Fluorous-Based Microarrays with Uncharged Carbohydrates . . . . . . Sahana K. Nagappayya and Nicola L.B. Pohl 11 General Procedure for the Synthesis of Neoglycoproteins and Immobilization on Epoxide-Modified Glass Slides . . . . . . . . . . . . . . . . . . . . . . Yalong Zhang and Jeffrey C. Gildersleeve

1

vii

13 31

57

69 87

103

117

137

149

155

viii

Contents

12 Immobilization of Polyacrylamide-Based Glycoconjugates on Solid Phase in Immunosorbent Assays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxana E. Galanina, Alexander A. Chinarev, Nadezhda V. Shilova, Marina A. Sablina, and Nicolai V. Bovin 13 Surface Plasmon Resonance Imaging Analysis of Protein Binding to a Sialoside-Based Carbohydrate Microarray . . . . . . . . . . . . . . . . . . . . . . . . . . . . Matthew J. Linman, Hai Yu, Xi Chen, and Quan Cheng 14 Glycoarray by DNA-Directed Immobilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . François Morvan, Yann Chevolot, Jing Zhang, Albert Meyer, Sébastien Vidal, Jean-Pierre Praly, Jean-Jacques Vasseur, and Eliane Souteyrand 15 Fabrication of Carbohydrate Surfaces by Using Non-derivatised Oligosaccharides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jonathan Popplewell, Marcus Swann, Gavin Brown, and Bob Lauder 16 Polysaccharide Microarrays: Application to the Identification of Heparan Sulphate Mimetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Julien Dheur, Nabil Dendane, Rémi Desmet, Fatima Dahmani, Gauthier Goormachtigh, Jérome Vicogne, Véronique Fafeur, and Oleg Melnyk 17 Carbohydrate Antigen Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Denong Wang 18 Probing Virus–Glycan Interactions Using Glycan Microarrays . . . . . . . . . . . . . . . . Jamie Heimburg-Molinaro, Mary Tappert, Xuezheng Song, Yi Lasanajak, Gillian Air, David F. Smith, and Richard D. Cummings 19 MALDI-ToF MS Analysis of Glycosyltransferase Activities on Gold Surface Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicolas Laurent, Rose Haddoub, Josef Voglmeir, and Sabine L. Flitsch 20 Microarray Technology Using Glycans Extracted from Natural Sources for Serum Antibody Fluorescent Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Emanuela Lonardi, André M. Deelder, Manfred Wuhrer, and Crina I.A. Balog 21 Studying Modification of Aminoglycoside Antibiotics by Resistance-Causing Enzymes via Microarray . . . . . . . . . . . . . . . . . . . . . . . . . . . Matthew D. Disney 22 Microarray Method for the Rapid Detection of Glycosaminoglycan–Protein Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Claude J. Rogers and Linda C. Hsieh-Wilson 23 Neoglycolipid-Based “Designer” Oligosaccharide Microarrays to Define b-Glucan Ligands for Dectin-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Angelina S. Palma, Yibing Zhang, Robert A. Childs, Maria A. Campanero-Rhodes, Yan Liu, Ten Feizi, and Wengang Chai 24 Measurement of Antibodies to Pneumococcal Polysaccharides with Luminex xMAP Microsphere-Based Liquid Arrays . . . . . . . . . . . . . . . . . . . . . Jerry W. Pickering and Harry R. Hill

167

183 195

221

231

241 251

269

285

303

321

337

361

Contents

25 Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates . . . . . . . . . . . . Jean-Philippe Ebran, Nabil Dendane, and Oleg Melnyk 26 Photoimmobilization of Saccharides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gregory T. Carroll and Denong Wang 27 Microwave-Assisted Method for Fabrication of Carbohydrate Cluster Microarrays on 3-Dimensional Hydrazide–Dendrimer Substrate . . . . . . . . . . . . . . Xichun Zhou, Jian Zhang, and Denong Wang 28 Combinatorial Glycoarray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simon Rinaldi, Kathryn M. Brennan, and Hugh J. Willison Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

377 393

401 413 425

Contributors Gillian Air  •  Department of Biochemistry & Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA Crina I.A. Balog  •  Department of Parasitology, Leiden University Medical Center, Biomolecular Mass Spectrometry Unit, Center for Infectious Diseases, Leiden, The Netherlands Julia Bartoli  •  Institut de Biologie Structurale, UMR 5075 (CEA, CNRS, UJF), Grenoble, France Nicolaï V. Bovin  •  Shemyakin & Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation Kathryn M. Brennan  •  Division of Clinical Neurosciences, University of Glasgow, Glasgow Biomedical Research Centre, Glasgow, UK Gavin Brown  •  Division of Biomedical and Life Sciences, School of Health and Medicine, Lancaster University, Lancaster, UK Maria A. Campanero-Rhodes  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK; Department of Biological Physical Chemistry, Instituto de Química-Física “Rocasolano”, CSIC and CIBERES, Madrid, Spain Hongzhi Cao  •  National Glycoengineering Research Center, Shandong University, Jinan, Shandong, People’s Republic of China Gregory T. Carroll  •  Centre for Systems Chemistry, Stratingh Institute for Chemistry, University of Groningen, Groningen, The Netherlands Samy Cecioni  •  Laboratoire de Chimie Organique 2, Institut de Chimie et Biochimie Moléculaires et Supramoléculaires, Glycochimie, UMR5246, CNRS, Université Lyon 1, Villeurbanne, France Wengang Chai  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK Xi Chen  •  Department of Chemistry, University of California, Davis, CA, USA Fang Cheng  •  Department of Bioengineering, University of Washington, Seattle, WA, USA Quan Cheng  •  Department of Chemistry, University of California, Riverside, CA, USA Yann Chevolot  •  Université de Lyon, Institut des Nanotechnologies de Lyon – INL, UMR CNRS 5270, Site Ecole Centrale de Lyon, Ecully, France Robert A. Childs  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK Alexander A. Chinarev  •  Shemyakin & Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation Richard D. Cummings  •  Department of Biochemistry, O. Wayne Rollins Research Center, Emory University School of Medicine, Atlanta, GA, USA

xi

xii

Contributors

Fatima Dahmani  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Nancy M. Dahms  •  Medical College of Wisconsin, Milwaukee, WI, USA Jose L. de Paz  •  Instituto de Investigaciones Químicas, Centro de Investigaciones Científicas Isla de La Cartuja, CSIC and US, Sevilla, Spain André M. Deelder  •  Department of Parasitology, Leiden University Medical Center, Biomolecular Mass Spectrometry Unit, Center for Infectious Diseases, Leiden, The Netherlands Nabil Dendane  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Rémi Desmet  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Julien Dheur  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Matthew D. Disney  •  Department of Chemistry, The Scripps Research Institute, Scripps Florida, Jupiter, FL, USA Jean-Philippe Ebran  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Véronique Fafeur  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Ten Feizi  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK Sabine L. Flitsch  •  Manchester Interdisciplinary Biocentre & School of Chemistry, The University of Manchester, Manchester, UK Oxana E. Galanina  •  Shemyakin & Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation Jeffrey C. Gildersleeve  •  Chemical Biology Laboratory, Center for Cancer Research, National Cancer Institute, Frederick, MD, USA Gauthier Goormachtigh  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France David Goyard  •  Laboratoire de Chimie Organique 2, Institut de Chimie et Biochimie Moléculaires et Supramoléculaires, Glycochimie, UMR5246, CNRS, Université Lyon 1, Villeurbanne, France Rose Haddoub  •  Manchester Interdisciplinary Biocentre & School of Chemistry, The University of Manchester, Manchester, UK Jamie Heimburg-Molinaro  •  Department of Biochemistry, Glycomics Center, Emory University School of Medicine, Atlanta, GA, USA Harry R. Hill  •  ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT, USA Linda C. Hsieh-Wilson  •  Division of Chemistry and Chemical Engineering, California Institute of Technology and the Howard Hughes Medical Institute, Pasadena, CA, USA Yi Lasanajak  •  Department of Biochemistry, Glycomics Center, Emory University School of Medicine, Atlanta, GA, USA Bob Lauder  •  Division of Biomedical and Life Sciences, School of Health and Medicine, Lancaster University, Lancaster, UK

Contributors

xiii

Nicolas Laurent  •  Manchester Interdisciplinary Biocentre & School of Chemistry, The University of Manchester, Manchester, UK Myung-Ryul Lee  •  Department of Chemistry, Center for Biofunctional Molecules, Yonsei University, Seoul, South Korea Matthew J. Linman  •  Department of Chemistry, University of California, Riverside, CA, USA Yan Liu  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK Thierry Livache  •  CREAB, UMR SPRAM 5819 (CEA, CNRS, UJF), INAC CEA Grenoble, Grenoble, France Emanuela Lonardi  •  Department of Parasitology, Leiden University Medical Center, Biomolecular Mass Spectrometry Unit, Center for Infectious Diseases, Leiden, The Netherlands Oleg Melnyk  •  CNRS UMR 8161, Institut Pasteur de Lille, Université Lille Nord de France, IFR 142 Molecular and Cellular Medicine, Lille, France Albert Meyer  •  Institut des Biomolécules Max Mousseron, UMR 5247, CNRS Université Montpellier 1, Université Montpellier 2, Montpellier, France François Morvan  •  Institut des Biomolécules Max Mousseron, UMR 5247, CNRS Université Montpellier 1, Université Montpellier 2, Montpellier, France Sahana K. Nagappayya  •  Department of Chemistry, The Plant Sciences Institute, and the Interdepartmental Program in Microbiology, Iowa State University, Ames, IA, USA Angelina S. Palma  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK; Chemistry and Technology Network Laboratory (REQUIMTE), Department of Chemistry, Faculty of Science and Technology, New University of Lisbon, Caparica, Portugal Sungjin Park  •  Department of Chemistry, Center for Biofunctional Molecules, Yonsei University, Seoul, South Korea Jerry W. Pickering  •  ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT, USA Nicola L.B. Pohl  •  Wilkinson Professor of Interdisciplinary Engineering, Department of Chemistry, Department of Chemical and Biological Engineering, The Plant Sciences Institute, and the Interdepartmental Program in Microbiology, Iowa State University, Ames, IA, USA Jonathan Popplewell  •  Farfield Group Ltd, Voyager West Wing, Level 7, Manchester Airport, Manchester, UK Jean-Pierre Praly  •  Laboratoire de Chimie Organique 2, Institut de Chimie et Biochimie Moléculaires et Supramoléculaires, Glycochimie, UMR5246, CNRS, Université Lyon 1,Villeurbanne, France Daniel M. Ratner  •  Department of Bioengineering, University of Washington, Seattle, WA, USA Simon Rinaldi  •  SRI International Biosciences Division, Menlo Park, CA, USA

xiv

Contributors

Claude J. Rogers  •  Division of Chemistry and Chemical Engineering, California Institute of Technology and the Howard Hughes Medical Institute, Pasadena, CA, USA André Roget  •  CREAB, UMR SPRAM 5819 (CEA, CNRS, UJF), INAC CEA Grenoble, Grenoble, France Marina A. Sablina  •  Shemyakin & Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation Peter H. Seeberger  •  Max-Planck Institute of Colloids and Interfaces, Potsdam, Germany; Freie Universität Berlin, Berlin, Germany Nadezhda V. Shilova  •  Shemyakin & Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation Injae Shin  •  Department of Chemistry, Center for Biofunctional Molecules, Yonsei University, Seoul, South Korea David F. Smith  •  Department of Biochemistry, Glycomics Center, Emory University School of Medicine, Atlanta, GA, USA Xuezheng Song  •  Department of Biochemistry, Glycomics Center, Emory University School of Medicine, Atlanta, GA, USA Eliane Souteyrand  •  Université de Lyon, Institut des Nanotechnologies de Lyon – INL, UMR CNRS 5270, Site Ecole Centrale de Lyon, Ecully, France Mark S. Stoll  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK Marcus Swann  •  Farfield Group Ltd, Voyager West Wing, Level 7, Manchester Airport, Manchester, UK Mary Tappert  •  Department of Biochemistry & Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA Jean-Jacques Vasseur  •  Institut des Biomolécules Max Mousseron, UMR 5247, CNRS Université Montpellier 1, Université Montpellier 2, Montpellier, France Jérome Vicogne  •  CNRS UMR 8161, Université Lille Nord de France, Institut Pasteur de Lille, IFR 142 Molecular and Cellular Medicine, Lille, France Sébastien Vidal  •  Laboratoire de Chimie Organique 2, Institut de Chimie et Biochimie Moléculaires et Supramoléculaires, Glycochimie, UMR5246, CNRS, Université Lyon 1, Villeurbanne, France Josef Voglmeir  •  Manchester Interdisciplinary Biocentre & School of Chemistry, The University of Manchester, Manchester, UK Denong Wang  •  Tumor Glycomics Laboratory, Center for Cancer Research, SRI International Biosciences Division, Menlo Park, CA, USA Daniel B. Werz  •  Institut für Organische und Biomolekulare Chemie, Georg-August-Universität Göttingen, Göttingen, Germany Hugh J. Willison  •  College of Medicine, Veterinary Medicine and Life Sciences, Glasgow Biomedical Reseach Centre, Glasgow, UK Manfred Wuhrer  •  Department of Parasitology, Center for Infectious Diseases, Biomolecular Mass Spectrometry Unit, Leiden University Medical Center, Leiden, The Netherlands Hai Yu  •  Department of Chemistry, University of California, Davis, CA, USA Jian Zhang  •  ADA Technologies, Inc., Littleton, CO, USA

Contributors

Jing Zhang   •  Université de Lyon, Institut des Nanotechnologies de Lyon – INL, UMR CNRS 5270, site Ecole Centrale de Lyon, Ecully, France Yalong Zhang  •  Chemical Biology Laboratory, Center for Cancer Research, National Cancer Institute, Frederick, MD, USA Yibing Zhang  •  The Glycosciences Laboratory, Department of Medicine, Imperial College London, Hammersmith Campus, London, UK Xichun Zhou  •  Vitan-Biotech LLC, Littleton, CO, USA

xv

Chapter 1 Recent Advances and Future Challenges in Glycan Microarray Technology José L. de Paz and Peter H. Seeberger Abstract Glycan microarrays, carrying hundreds of different sugars on chip surfaces, have become a standard tool for the study of interactions of biomolecules with carbohydrates. The chip-based format offers important advantages, including the ability to screen in parallel several thousand binding events on a single slide, the minimal amount of sample required for one experiment, and the multivalent display of sugars on the chip that mimics the presentation of carbohydrates in nature. This chapter presents recent advances and future challenges in glycan microarray technology. We describe different immobilization and detection methods as well as applications in glycomics, drug discovery, and biomedicine. Key words: Microarrays, Carbohydrates, Glycomics, Glycobiology, Immobilization chemistry

1. Introduction Carbohydrate microarrays (glycoarrays) (1–6) consist of sugars that are bound, covalently or noncovalently, to a solid surface in a spatially defined and miniaturized fashion. In the past few years, glycoarrays have become a standard tool to screen large number of sugar–biomolecule interactions and investigate the role of carbohydrates in biological systems. The most important advantages of glycoarray technology over conventional approaches, such as enzyme-linked lectin assay, surface plasmon resonance (SPR), or isothermal titration calorimetry, are the ability to screen several thousand binding events on a single slide and the miniscule amounts of both analyte and ligand required for one experiment. Additionally, glycoarrays are ideal platforms to detect interactions that involve carbohydrates because the multivalent display of ligands on a surface overcomes the relative weakness of these interactions by mimicking cell–cell interfaces. Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_1, © Springer Science+Business Media, LLC 2012

1

2

J.L. de Paz and P.H. Seeberger

Glycoarrays have been predominantly applied to determine the sugar-binding profiles of proteins, nucleic acids, whole cells, and pathogens. These experiments provide valuable information for the design of carbohydrate-based drugs. Glycoarrays also constitute an attractive platform to test serum samples, opening the way for a wide variety of biomedical applications, such as the identification of novel sugar markers related to early-stage diseases and the detection and diagnosis of pathogenic infections. In this chapter, we focus on recent advances, current limitations, and opportunities for improvement in glycoarray technology. In particular, we first present an overview of the immobilization methods that have been used for glycoarray fabrication. Then, we discuss the different detection methods that have been introduced to analyze binding events on carbohydrate microarrays. Finally, we describe glycoarray applications in glycomics, drug discovery, and biomedical research.

2. Glycoarray Fabrication To prepare glycoarrays, pure carbohydrates have to be attached to appropriate chip surfaces. A variety of different immobilization methods have been developed over the past few years, involving both covalent and noncovalent attachment (7). Glycans used for immobilization can be either chemically synthesized or isolated from natural sources. A standard size microscope glass slide is the most widely used solid surface. The sugars are printed onto the slide using automated arraying robots. The use of robotic equipment designed for DNA microarray preparation is common. A typical microarray spotter generates spots of approximately 200  mm in diameter by delivering 1 nL of sugar solutions. The carbohydrates are usually printed at different concentrations in several replicates. The array can contain up to 10,000 spots. Printing is conducted in a chamber with regulated humidity. After printing, slides are immediately placed in a humidity chamber and usually incubated for several hours. Arrays are then washed to remove the unbound carbohydrates from the surface. A typical printing protocol often includes a quenching step of remaining reactive groups on the slide. Microarrays are finally washed and dried by centrifugation, and can be stored for a long period of time. 2.1. Covalent Immobilization

Chemical synthesis can be designed to afford oligosaccharide probes with an orthogonal functional group that is suitable for a chemical reaction with a reactive solid support (Fig. 1). In this way, sugars are covalently coupled to the microarray surface. The choice of the covalent coupling chemistry mainly depends on the synthetic strategy of the carbohydrate. The coupling reaction should be fast,

1  Recent Advances and Future Challenges in Glycan…

3

a

O S

O SH

O

N

O O

O

N

b

O O Chemical synthesis

O

O NH 2

N O

O HN

O

c

O N

O N3

N N

Fig. 1. Covalent immobilization of synthetic sugars on chip surfaces: thiol-maleimide (a), amine-N-hydroxysuccinimide ester (b), and azide-alkyne (c) couplings.

specific, and high yielding. For instance, thiol-terminated oligosaccharides were prepared and attached to maleimide-derivatized glass slides (8). The selective reaction between thiol and maleimide groups has been widely employed to prepare various bioconjugates (9). Other ligation reactions used to prepare glycoarrays involved amide bond formation between an amine-containing sugar and an N-hydroxysuccinimide surface (10, 11), or a 1,3-dipolar cycloaddition between an azide and an alkyne (12). Epoxide-functionalized solid surfaces can be used to prepare glycan microarrays via immobilization with glycan-conjugated BSA (13). Due to recent advances in carbohydrate chemistry (14–16), including one-pot oligosaccharide synthesis (17) and the development of automated solidphase oligosaccharide synthesizers (18, 19), chemical synthesis has become the most popular sugar source to fabricate glycoarrays (20, 21). However, the synthesis of thousands of oligosaccharides for the construction of microarrays containing, ideally, the entire glycome of an organism is far from trivial. Synthetic libraries are still limited in number and structural diversity. Therefore, further development of more efficient synthetic methods is needed to expand the complexity and utility of glycoarrays containing synthetic probes.

4

J.L. de Paz and P.H. Seeberger

Fig. 2. Direct attachment of unmodified carbohydrates on hydrazide-containing chip surfaces.

Alternative approaches involving the use of naturally derived oligosaccharides should also be considered. Unprotected glycans can be directly derivatized by taking advantage of the reactivity of the masked aldehyde of the reducing end. Examples of procedures for the functionalization of free sugars include reductive amination (22), oxime bond formation (23), and glycosylamine synthesis (24). The modified sugars are then printed on appropriate slides. An even better method to prepare glycoarrays with isolated sugars is the direct attachment of free carbohydrates on aminooxyor hydrazide-containing surfaces that avoids prior modification (Fig. 2) (25, 26). For example, free reducing end sugars selectively bound to hydrazide groups on the chip surface to initially form the acyclic hydrazone, followed by conversion to cyclic, b-configured adduct. In contrast, acyclic structures were predominantly formed on aminooxy surfaces (25). 2.2. Noncovalent Immobilization

Noncovalent immobilization, mainly through ionic and hydrophobic interactions, has been also employed for the fabrication of glycoarrays. Noncovalent immobilization techniques include nonspecific adsorption of free glycans to underivatized solid surfaces. For example, polysaccharides have been passively adsorbed on nitrocellulose-coated glass slides (27). In these processes, the polysaccharides are site-nonspecifically attached to the chip. Polysaccharides with large contact areas are efficiently adsorbed on the surface, whereas small glycans are removed during washing steps. Therefore, this straightforward method is limited to high molecular weight sugars, leading to a random orientation of carbohydrates on the surface. Other noncovalent methods have been exploited for polysaccharide immobilization. Heparin polysaccharides have been directly attached to poly-l-lysine-coated slides via electrostatic interactions while maintaining their ability to interact with proteins (28). The methods described above are not suitable for the preparation of microarrays containing monosaccharides or low molecular weight oligosaccharides. Conjugation with a carrier molecule, such as a lipid chain, is required for noncovalent immobilization of smaller oligosaccharides onto nitrocellulose via hydrophobic interactions (29).

1  Recent Advances and Future Challenges in Glycan…

5

Another noncovalent immobilization method involves the attachment of fluorous-tagged glycans to fluoroalkylated slides via fluorousbased interactions (30). The lipid and fluorous tags retain attached glycans even after extensive washing and avoid random sugar orientation on the chip surface that can negatively influence binding assay reproducibility. Noncovalent attachment can also rely on the binding of biotinylated sugars to streptavidin-coated plates (31). A DNA-based glycan microarray has also been reported for the site-specific, noncovalent immobilization of sugars (32). The carbohydrates were conjugated with a DNA strand and immobilization was performed through hybridization using a glass slide coated with a complementary DNA sequence. Relative sugar densities were assessed by the introduction of a fluorescent probe on the DNA-glycoconjugate. This point is of particular interest since despite the impressive progress in the development of carbohydrate immobilization strategies, quantification of immobilized sugars on the array surface remains difficult. To tackle this problem, an alternative solution involved the attachment of a fluorescent linker to the sugar probe by reductive amination (22). This fluorescent linker contained an additional functional group that permitted the immobilization of the conjugate on the appropriate surface. The fluorescent label allowed quantification of immobilized sugar by comparing the fluorescence intensity before and after washing. A cleavable linker was also employed to estimate the yield of the immobilization step by using mass spectrometry (33, 34). However, these methods require more elaborate synthetic strategies to install the cleavable or fluorescent linker, thereby complicating glycoarray fabrication.

3. Detection of CarbohydrateMediated Interactions

Glycoarray technology has been applied most extensively to the high-throughput analysis of carbohydrate–protein interactions. The proteins are incubated on the microarrays to allow for binding to the exposed carbohydrates before unbound proteins are washed away from the surface. In the next step, bound proteins are detected (Fig.  3). Fluorescence-based methods are the most widely used because of high sensitivity and availability of highresolution fluorescence scanner. When fluorescent-labelled proteins are used, microarrays are directly read out and the fluorescence intensities indicate the amount of ligand bound to the chip. It should be noted that protein labelling with fluorescent tags may cause protein denaturation or modification of the sugar-binding domain. Therefore, an alternative approach involves the use of an antibody that specifically recognizes proteins bound to glycoarrays (Fig. 3). The antibody can be directly detected if it contains a fluorescent tag.

6

J.L. de Paz and P.H. Seeberger fluorescently labelled antibody

a protein

fluorescence scanner glass surface

b label-free SPR experiment gold chip

Fig. 3. Detection of carbohydrate–protein interactions. (a) Detection of bound protein by using a specific fluorescently labelled antibody; (b) Detection of bound protein by direct SPR experiment.

A typical sandwich procedure involving fluorescently labelled secondary antibodies has also been extensively used. However, specific antibodies can only be applied in some cases due to their limited availability. The introduction of label-free detection methods, such as mass spectrometry or SPR, is convenient since labelling of sugar interaction partners becomes obsolete (Fig. 3) (35). SPR affords kinetic parameters associated with the interaction, measuring association and dissociation constants on the array. For this technique, a gold surface is required as array support (36). Thiol-terminated oligosaccharides can be attached to the gold surface via Au–S linkage formation. The development of multichannel SPR instruments allows for the simultaneous analysis of hundreds of sugar spots (37–39). Another label-free detection method is MALDI-TOF mass spectrometry (40) that is particularly well suited to monitor carbohydrate transformations as a result of enzymatic reactions (41, 42). The specificity of enzymes can be studied by probing a panel of immobilized potential substrates. Moreover, the diversity of sugars on the array can be increased by enzymatic synthesis. Label free MALDI-TOF analysis is crucial for glycoarray applications involving enzymes. Alternatively, monitoring of on-chip enzymatic products with fluorescent lectins is very limited by the availability of such lectins. MALDI-TOF detection has also been employed for the on-chip chemical synthesis of oligosaccharides (43). A hydroxyphenyl-functionalized self-assembled monolayer of alkanethiols on gold was glycosylated using glycosyl trichloroacetimidates. Selective removal of a temporary protecting group unmasked a hydroxyl group of the sugar, allowing their chemical

1  Recent Advances and Future Challenges in Glycan…

7

glycosylation with another trichloroacetimidate. Thus, an oligosaccharide chain was elongated on the chip surface, observing the synthetic intermediates by mass spectrometry. This technique is however limitated by the inability to determine the configuration of the anomeric linkage. Besides proteins, glycoarrays have been applied to screen the interactions of carbohydrates with nucleic acids (44), whole cells (45, 46) and pathogens such as viruses (10) or bacteria. Binding of RNA to glycoarrays can be measured using either fluorescently labelled RNA or by staining the bound RNA with dyes such as SYBR Green. Cells can be stained with cell permeable nucleic acid dyes prior to incubation. Viruses can be detected by incubating antibodies against proteins present on the virion (10). Most binding experiments with glycoarrays have been performed at one concentration of protein to afford qualitative information about the carbohydrate binding profile. Recently, carbohydrate chips have been used to quantify glycan–protein interactions, determining dissociation constants (Kd) values (47, 48). A series of protein concentrations were incubated on the chip surface, and protein concentration was plotted against fluorescence intensity for each immobilized sugar. These curves were analyzed as Langmuir isotherms to afford Kd values that were similar to those obtained in SPR experiments. IC50 values of soluble inhibitors for protein binding to carbohydrates attached to the chip surface have also been determined (9, 49, 50). For this purpose, glycoarrays were incubated with a series of mixtures containing the protein and different concentrations of the soluble inhibitor. This competition experiment yielded the IC50 values by plotting the inhibitor concentration against the fluorescence intensity.

4. Applications of Carbohydrate Microarrays

Glycoarrays have been predominantly used for the study of the binding properties of a variety of sugar-binding partners such as proteins, antibodies, cells, and viruses. For example, microarrays of heparin oligosaccharides have been employed to establish the binding profiles of several proteins such as growth factors and chemokines (51, 52). Heparin is a highly sulfated, linear polymer that belongs to the glycosaminoglycan (GAG) family and participates in a plethora of biological processes, such as inflammation, angiogenesis, and blood coagulation, by interaction with many proteins. Heparin is formed by disaccharide repeating units of d-glucosamine and either l-iduronic or d-glucuronic acid carrying sulfate groups at different positions of the chain. The sulfate pattern and uronic acid distribution are controlled by a complex biosynthetic pathway that leads to heparin chains with a high level of structural diversity. This heterogeneity

8

J.L. de Paz and P.H. Seeberger

of heparin, particularly the sulfation patterns, is responsible for its specific interaction with a wide variety of proteins. However, little is known about the exact structural requirements of heparin– protein interactions. Microarrays constitute an ideal platform for the establishment of structure–activity relationships for heparin sequences (53). A series of amine-terminated heparin oligosaccharides of varying length and sulfation pattern were chemically synthesized and immobilized on N-hydroxysuccinimide-activated glass surfaces by amide bond formation (11). These arrays were incubated with several chemokines and growth factors, demonstrating that specific sulfation motifs are needed for binding. In addition, heparin microarrays allowed for the identification of potential low molecular weight inhibitors of GAG–protein interactions such as monosaccharides containing an artificial sulfate distribution that is not found in nature. These arrays give valuable data to understand the biological role of GAG–protein interactions, providing a first step for new strategies to modulate GAG-mediated physiological processes. Incubation of glycoarrays with sera samples opens the way to more medically relevant applications, including identification of carbohydrate cancer markers and specific identification of pathogen infection. The identification of markers in early-stage cancers could lead to improved therapies and survival rates of patients. Cancer cells often display carbohydrates with different structures than those observed in normal cells (54). Therefore, cancer-associated carbohydrates can be considered as cancer markers and glycoarrays constitute an ideal platform for diagnosis. For example, Globo H is a hexasaccharide that is highly expressed on various types of cancers, especially breast, prostate, and ovarian cancers. A microarray containing the Globo H hexasaccharide and several structural analogs (55) was employed to screen sera samples of breast cancer patients and healthy individuals (55, 56). These studies revealed significant differences in antibody levels of cancer and healthy patients. Thus, glycoarrays are valuable tools for cancer diagnosis. Furthermore, the fabrication and use of more complex glycoarrays, with a higher number of carbohydrate structures, will allow the identification of novel carbohydrate markers related to other disease states, increasing the utility of this platform in cancer research. Another important application of glycoarrays is the diagnosis of infection diseases. Most pathogens contain specific polysaccharides on their cell membrane. After infection, protective antibodies are produced against these capsular polysaccharides. Glyco arrays are employed to analyze serum samples, detecting specific antibodies in infected patients to facilitate diagnosis. Moreover, microarray experiments give important information to design carbohydratebased vaccines and discover novel pathogen biomarkers. Salmonella bacteria cause a variety of diseases in humans. The severity of the

1  Recent Advances and Future Challenges in Glycan…

9

disease depends on the particular Salmonella strain. A glycoarray containing a panel of Salmonella-related carbohydrate antigens was employed to analyze human sera from salmonellosis and healthy patients (57). Specific antibodies against bacterial infection were detected, and different Salmonella subtypes could also be distinguished. Glycosylphosphatidylinositol (GPI) glycolipids are present at the plasma membrane of Plasmodium falciparum, a parasite that causes malaria in humans. GPI has emerged as an important toxin in malaria disease and people living in malariaendemic regions often produce high levels of anti-GPI antibodies. Synthetic GPI glycoarrays (58) served to establish the binding specificities of anti-GPI antibodies, correlate antibody levels and protection from severe malaria and aid the design of efficient carbohydrate-based antitoxin vaccine. A different application of glycoarrays is based on the fact that viruses and bacteria use carbohydrates on the surface of human cells as initial recognition and attachment sites. Hemagglutinin proteins are present on the surface of influenza viruses and mediate the attachment to the host cell by binding to sialylated carbohydrates. The precise structure of these cell surface sialic acid epitopes depends on the host species and anatomical location. For instance, a 2,6-linked sialic acid structures are preferentially found on human epithelial cells in the upper respiratory tract whereas a 2,3-linked sialic acid units are found in the respiratory tract of birds. Hemagglutinins from different influenza variants vary in their sialic acid binding profile. Human influenza hemagglutinins preferentially recognize a 2,6-linked sialic acid residues and avian hemagglutinins are specific for a-2,3 linkages. Glycoarrays bearing many sialylated oligosaccharides were employed to determine the binding preferences of several influenza strains (59), including a highly pathogenic Vietnamese H5N1 avian virus (60) and two pandemic influenza A (H1N1) 2009 viruses (61). Microarray technology can be used to identify influenza virus subtypes from infected serum samples by analyzing their fine sugar binding specificities. Fast glycoarray tests might detect influenza strains in early stages of epidemic infection and, more importantly, identify changes in carbohydrate binding profiles suggesting dangerous virus mutations. Certain Escherichia coli strains cause urinary tract infections. Interestingly, these harmful strains strongly bind to mannosecontaining oligosaccharides on the host cell. In contrast, E. coli mutants exhibiting a reduced affinity to mannose do not cause infections. Glycoarrays were employed to study the binding profiles of E. coli bacteria (46). Different strains were distinguished, paving the way for developing detection tests for pathogens. Moreover, the chip format allowed for the fast screening for anti-adhesion compounds inhibiting carbohydrate–cell interactions that can be considered as therapeutics against pathogen infection.

10

J.L. de Paz and P.H. Seeberger

Acknowledgments We thank the Spanish Research Council (CSIC) (Grant 200880I041), the Spanish Ministry of Science and Innovation (Grant CTQ200907168), Junta de Andalucía (Grant P07-FQM-02969, “Incentivo a Proyecto Internacional”), and the European Union (FEDER support and Marie Curie Reintegration Grant) for financial support. Generous financial support from the Max-Planck Society is gratefully acknowledged. References 1. Laurent, N., Voglmeir, J., and Flitsch, S. L. (2008) Glycoarrays - tools for determining protein-carbohydrate interactions and glycoenzyme specificity. Chem. Commun., 4400–4412. 2. Horlacher, T., and Seeberger, P. H. (2008) Carbohydrate arrays as tools for research and diagnostics. Chem. Soc. Rev. 37, 1414–1422. 3. Park, S., Lee, M. R., and Shin, I. (2008) Carbohydrate microarrays as powerful tools in studies of carbohydrate-mediated biological processes. Chem. Commun., 4389–4399. 4. Wu, C. Y., Liang, P. H., and Wong, C. H. (2009) New development of glycan arrays. Org. Biomol. Chem. 7, 2247–2254. 5. de Paz, J. L., Horlacher, T., and Seeberger, P. H. (2006) Oligosaccharide microarrays to map interactions of carbohydrates in biological systems. Methods Enzymol. 415, 269–292. 6. de Paz, J. L., and Seeberger, P. H. (2006) Recent advances in carbohydrate microarrays. QSAR Comb. Sci. 25, 1027–1032. 7. Shin, I., Park, S., and Lee, M. R. (2005) Carbohydrate microarrays: An advanced technology for functional studies of glycans. Chem. Eur. J. 11, 2894–2901. 8. Adams, E. W., Ratner, D. M., Bokesch, H. R., McMahon, J. B., O’Keefe, B. R., and Seeberger, P. H. (2004) Oligosaccharide and glycoprotein microarrays as tools in HIV glycobiology: Glycan-dependent gp120/protein interactions. Chem. Biol. 11, 875–881. 9. Park, S., Lee, M. R., Pyo, S. J., and Shin, I. (2004) Carbohydrate chips for studying highthroughput carbohydrate-protein interactions. J. Am. Chem. Soc. 126, 4812–4819. 10. Blixt, O., Head, S., Mondala, T., Scanlan, C., Huflejt, M. E., Alvarez, R., Bryan, M. C., Fazio, F., Calarese, D., Stevens, J., Razi, N., Stevens, D. J., Skehel, J. J., van Die, I., Burton, D. R., Wilson, I. A., Cummings, R., Bovin, N., Wong, C. H., and Paulson, J. C. (2004) Printed

covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc. Natl. Acad. Sci. U.S.A. 101, 17033–17038. 11. Noti, C., de Paz, J. L., Polito, L., and Seeberger, P. H. (2006) Preparation and use of microarrays containing synthetic heparin oligosaccharides for the rapid analysis of heparin-protein interactions. Chem. Eur. J. 12, 8664–8686. 12. Calarese, D. A., Lee, H. K., Huang, C. Y., Best, M. D., Astronomo, R. D., Stanfield, R. L., Katinger, H., Burton, D. R., Wong, C. H., and Wilson, I. A. (2005) Dissection of the carbohydrate specificity of the broadly neutralizinganti-HIV-1 antibody 2 G12. Proc. Natl. Acad. Sci. U.S.A. 102, 13372–13377. 13. Manimala, J. C., Li, Z., Jain, A., VedBrat, S., and Gildersleeve, J. C. (2005) Carbohydrate array analysis of anti-Tn antibodies and lectins reveals unexpected specificities: implications for diagnostic and vaccine development. ChemBioChem 6, 2229–2241. 14. Seeberger, P. H., and Werz, D. B. (2007) Synthesis and medical applications of oligosaccharides. Nature 446, 1046–1051. 15. Zhu, X. M., and Schmidt, R. R. (2009) New Principles for Glycoside-Bond Formation. Angew. Chem. Int. Ed. 48, 1900–1934. 16. Boltje, T. J., Buskas, T., and Boons, G. J. (2009) Opportunities and challenges in synthetic oligosaccharide and glycoconjugate research. Nature Chem. 1, 611–622. 17. Zhang, Z. Y., Ollmann, I. R., Ye, X. S., Wischnat, R., Baasov, T., and Wong, C. H. (1999) Programmable one-pot oligosaccharide synthesis. J. Am. Chem. Soc. 121, 734–753. 18. Plante, O. J., Palmacci, E. R., and Seeberger, P. H. (2001) Automated solid-phase synthesis of oligosaccharides. Science 291, 1523–1527. 19. Seeberger, P. H. (2008) Automated oligosaccharide synthesis. Chem. Soc. Rev. 37, 19–28.

1  Recent Advances and Future Challenges in Glycan… 20. Seeberger, P. H., and Werz, D. B. (2005) Automated synthesis of oligosaccharides as a basis for drug discovery. Nat. Rev. Drug Discov. 4, 751–763. 21. Werz, D. B., and Seeberger, P. H. (2005) Carbohydrates as the next frontier in pharmaceutical research. Chem. Eur. J. 11, 3194–3206. 22. Xia, B. Y., Kawar, Z. S., Ju, T. Z., Alvarez, R. A., Sachdev, G. P., and Cummings, R. D. (2005) Versatile fluorescent derivatization of glycans for glycomic analysis. Nat. Methods 2, 845–850. 23. Liu, Y., Feizi, T., Carnpanero-Rhodes, M. A., Childs, R. A., Zhang, Y. N., Muiioy, B., Evans, P. G., Osborn, H. M. I., Otto, D., Crocker, P. R., and Chai, W. C. (2007) Neoglycolipid probes prepared via oxime ligation for microarray analysis of oligosaccharide-protein interactions. Chem. Biol. 14, 847–859. 24. Brun, M. A., Disney, M. D., and Seeberger, P. H. (2006) Miniaturization of microwave-assisted carbohydrate functionalization to create oligosaccharide microarrays. ChemBioChem 7, 421–424. 25. Lee, M., and Shin, I. (2005) Facile preparation of carbohydrate microarrays by site-specific, covalent immobilization of unmodified carbohydrates on hydrazide-coated glass slides. Org. Lett. 7, 4269–4272. 26. Zhi, Z. L., Powell, A. K., and Turnbull, J. E. (2006) Fabrication of carbohydrate microarrays on gold surfaces: Direct attachment of nonderivatized oligosaccharides to hydrazide-modified self-assembled monolayers. Anal. Chem. 78, 4786–4793. 27. Wang, D. N., Liu, S. Y., Trummer, B. J., Deng, C., and Wang, A. L. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–281. 28. Shipp, E. L., and Hsieh-Wilson, L. C. (2007) Profiling the sulfation specificities of glycosaminoglycan interactions with growth factors and chemotactic proteins using microarrays. Chem. Biol. 14, 195–208. 29. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. G. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nat. Biotechnol. 20, 1011–1017. 30. Ko, K. S., Jaipuri, F. A., and Pohl, N. L. (2005) Fluorous-based carbohydrate microarrays. J. Am. Chem. Soc. 127, 13162–13163. 31. Alvarez, R. A., and Blixt, O. (2006) Identification of ligand specificities for glycan-binding proteins using glycan arrays. Methods Enzymol. 415, 292–310.

11

32. Chevolot, Y., Bouillon, C., Vidal, S., Morvan, F., Meyer, A., Cloarec, J. P., Jochum, A., Praly, J. P., Vasseur, J. J., and Souteyrand, E. (2007) DNA-based carbohydrate biochips: a platform for surface glyco-engineering. Angew. Chem. Int. Ed. 46, 2398–2402. 33. Bryan, M. C., Fazio, F., Lee, H. K., Huang, C. Y., Chang, A., Best, M. D., Calarese, D. A., Blixt, C., Paulson, J. C., Burton, D., Wilson, I. A., and Wong, C. H. (2004) Covalent display of oligosaccharide arrays in microtiter plates. J. Am. Chem. Soc. 126, 8640–8641. 34. Lee, J. C., Wit, C. Y., Apon, J. V., Siuzdak, G., and Wong, C. H. (2006) Reactivity-based onepot synthesis of the tumor-associated antigen N3 minor octasaccharide for the development of a photocleavable DIOS-MS sugar array. Angew. Chem. Int. Ed. 45, 2753–2757. 35. Song, E.-H., and Pohl, N. L. B. (2009) Carbohydrate arrays: recent developments in fabrication and detection methods with applications. Curr. Opin. Chem. Biol. 13, 626–632. 36. Ratner, D. M., Adams, E. W., Su, J., O’Keefe, B. R., Mrksich, M., and Seeberger, P. H. (2004) Probing protein-carbohydrate interactions with microarrays of synthetic oligosaccharides. ChemBioChem 5, 379–382. 37. Smith, E. A., Thomas, W. D., Kiessling, L. L., and Corn, R. M. (2003) Surface plasmon resonance imaging studies of protein-carbohydrate interactions. J. Am. Chem. Soc. 125, 6140–6148. 38. Karamanska, R., Clarke, J., Blixt, O., MacRae, J. I., Zhang, J. Q. Q., Crocker, P. R., Laurent, N., Wright, A., Flitsch, S. L., Russell, D. A., and Field, R. A. (2008) Surface plasmon resonance imaging for real-time, label-free analysis of protein interactions with carbohydrate microarrays. Glycoconj. J. 25, 69–74. 39. de Boer, A. R., Hokke, C. H., Deelder, A. M., and Wuhrer, M. (2008) Serum antibody screening by surface plasmon resonance using a natural glycan microarray. Glycoconj. J. 25, 75–84. 40. Su, J., and Mrksich, M. (2002) Using mass spectrometry to characterize self-assembled monolayers presenting peptides, proteins, and carbohydrates. Angew. Chem. Int. Ed. 41, 4715–4718. 41. Laurent, N., Voglmeir, J., Wright, A., Blackburn, J., Pham, N. T., Wong, S. C. C., Gaskell, S. J., and Flitsch, S. L. (2008) Enzymatic glycosylation of peptide arrays on gold surfaces. ChemBioChem 9, 883–887. 42. Zhi, Z. L., Laurent, N., Powel, A. K., Karamanska, R., Fais, M., Voglmeir, J., Wright, A., Blackburn, J. M., Crocker, P. R., Russell, D. A., Flitsch, S., Field, R. A., and Turnbull, J. E. (2008) A versatile gold surface approach for

12

J.L. de Paz and P.H. Seeberger

fabrication and interrogation of glycoarrays. ChemBioChem 9, 1568–1575. 43. Ban, L., and Mrksich, M. (2008) On-chip synthesis and label-free assays of oligosaccharide arrays. Angew. Chem. Int. Ed. 47, 3396–3399. 44. Disney, M. D., and Seeberger, P. H. (2004) Aminoglycoside microarrays to explore interactions of antibiotics with RNAs and proteins. Chem. Eur. J. 10, 3308–3314. 45. Nimrichter, L., Gargir, A., Gortler, M., Altstock, R. T., Shtevi, A., Weisshaus, O., Fire, E., Dotan, N., and Schnaar, R. L. (2004) Intact cell adhesion to glycan microarrays. Glycobiology 14, 197–203. 46. Disney, M. D., and Seeberger, P. H. (2004) The use of carbohydrate microarrays to study carbohydrate-cell interactions and to detect pathogens. Chem. Biol. 11, 1701–1707. 47. Liang, P. H., Wang, S. K., and Wong, C. H. (2007) Quantitative analysis of carbohydrateprotein interactions using glycan microarrays: Determination of surface and solution dissociation constants. J. Am. Chem. Soc. 129, 11177–11184. 48. Park, S., and Shin, I. (2007) Carbohydrate microarrays for assaying galactosyltransferase activity. Org. Lett. 9, 1675–1678. 49. de Paz, J. L., Noti, C., Bohm, F., Werner, S., and Seeberger, P. H. (2007) Potentiation of fibroblast growth factor activity by synthetic heparin oligosaccharide glycodendrimers. Chem. Biol. 14, 879–887. 50. Houseman, B. T., and Mrksich, M. (2002) Carbohydrate arrays for the evaluation of protein binding and enzymatic modification. Chem. Biol. 9, 443–454. 51. de Paz, J. L., Noti, C., and Seeberger, P. H. (2006) Microarrays of synthetic heparin oligosaccharides. J. Am. Chem. Soc. 128, 2766–2767. 52. de Paz, L. L., Moseman, E. A., Noti, C., Polito, L., von Andrian, U. H., and Seeberger, P. H. (2007) Profiting heparin-chemokine interactions using synthetic tools. ACS Chem. Biol. 2, 735–744. 53. de Paz, J. L., and Seeberger, P. H. (2008) Deciphering the glycosaminoglycan code with

the help of microarrays. Mol. BioSyst. 4, 707–711. 54. Dube, D. H., and Bertozzi, C. R. (2005) Glycans in cancer and inflammation. Potential for therapeutics and diagnostics. Nat. Rev. Drug Discov. 4, 477–488. 55. Huang, C. Y., Thayer, D. A., Chang, A. Y., Best, M. D., Hoffmann, J., Head, S., and Wong, C. H. (2006) Carbohydrate microarray for profiling the antibodies interacting with Globo H tumor antigen. Proc. Natl. Acad. Sci. U.S.A. 103, 15–20. 56. Wang, C. C., Huang, Y. L., Ren, C. T., Lin, C. W., Hung, J. T., Yu, J. C., Yu, A. L., Wu, C. Y., and Wong, C. H. (2008) Glycan microarray of Globo H and related structures for quantitative analysis of breast cancer. Proc. Natl. Acad. Sci. U.S.A. 105, 11661–11666. 57. Blixt, O., Hoffmann, J., Svenson, S., and Norberg, T. (2008) Pathogen specific carbohydrate antigen microarrays: a chip for detection of Salmonella O-antigen specific antibodies. Glycoconj. J. 25, 27–36. 58. Kamena, F., Tamborrini, M., Liu, X. Y., Kwon, Y. U., Thompson, F., Pluschke, G., and Seeberger, P. H. (2008) Synthetic GPI array to study antitoxic malaria response. Nat. Chem. Biol. 4, 238–240. 59. Stevens, J., Blixt, O., Glaser, L., Taubenberger, J. K., Palese, P., Paulson, J. C., and Wilson, I. A. (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities. J. Mol. Biol. 355, 1143–1155. 60. Stevens, J., Blixt, O., Tumpey, T. M., Taubenberger, J. K., Paulson, J. C., and Wilson, I. A. (2006) Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science 312, 404–410. 61. Childs, R. A., Palma, A. S., Wharton, S., Matrosovich, T., Liu, Y., Chai, W. G., Campanero-Rhodes, M. A., Zhang, Y. B., Eickmann, M., Kiso, M., Hay, A., Matrosovich, M., and Feizi, T. (2009) Receptor-binding specificity of pandemic influenza A (H1N1) 2009 virus determined by carbohydrate microarray. Nat. Biotechnol. 27, 797–799.

Chapter 2 Chemical Synthesis of Carbohydrates and Their Surface Immobilization: A Brief Introduction Daniel B. Werz Abstract For all carbohydrate microarrays, two important prerequisites are necessary: the carbohydrate of interest has to be obtained either by isolation from natural sources, enzymatic or chemical synthesis; an immobilization of the carbohydrate at the surface of the chip has to be achieved. This chapter provides a very brief overview of the chemical synthesis of carbohydrates (creation of building blocks, assembly, and deprotection) and of immobilization techniques. Numerous methods are known to construct oligosaccharides by chemical methods. A typical monosaccharide building block, used in oligosaccharide assembly, is equipped with different protecting groups that mask the hydroxyl and amine groups. In general, a good leaving group at the anomeric center that can easily be activated is mandatory; especially trichloroacetimidates, phosphates, and thioethers have been widely used for the creation of glycosidic bonds. After the complete assembly of the oligosaccharide, a global deprotection of all permanent protecting groups affords the desired target structure with free hydroxyl groups. Linkers, which were introduced during the synthesis, must often be modified at the end to create appropriate functionalities for surface immobilization. Key words: Glycosylation, Protecting group, Deprotection, Building block, Chemical synthesis, Surface immobilization, Microarrays, Thiols

1. Introduction The chemical synthesis of complex oligosaccharides has been a synthetically difficult and rather time-consuming endeavor (1). In contrast to peptides and oligonucleotides, carbohydrates are often branched and several hydroxyl or amino groups at each building block lead to a dazzling variety of structures, which can be obtained by a small set of subunits. Even more complicated is the creation of a new stereogenic center resulting from the formation of a glycosidic bond during the combination of two monosaccharides.

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_2, © Springer Science+Business Media, LLC 2012

13

14

D.B. Werz

Table 1 Number of different oligomers for nucleotides, peptides, and carbohydrates Numbers of different oligomers Oligomer size

Nucleotides

Peptides

Carbohydrates

1

4

20

20

2

16

400

1,360

3

64

8,000

126,080

4

256

160,000

13,495,040

5

1,024

3,200,000

1,569,745,920

6

4,096

64,000,000

192,780,943,360

The numbers for the carbohydrates are based on the “10 mammalian” monosaccharide units. Branching is also taken into account

A recent statistical analysis regarding the diversity of mammalian carbohydrate structures – based on the “10 mammalian monosaccharides” without considering any further attachments – reveals astounding numbers, which dwarf the complexity of both peptides and oligonucleotides (Table  1) (2). Although the number of structural combinations encountered in nature is much smaller than the theoretical ones, it is obvious that synthetic carbohydrate chemists have to face a variety of challenges during the assembly of an oligosaccharide by chemical means.

2. Chemical Synthesis of Carbohydrates

A carefully designed synthetic plan is necessary before one starts with the total synthesis of the desired target structure. Such a plan includes the glycosylation strategy (type of anomeric leaving groups) and some kind of temporary protecting groups (3, 4). After that, mono- or disaccharide building blocks have to be prepared. With these compounds in hand, the assembly process can be started. Besides well-known solution-phase techniques, an automated solid-phase approach has also been developed (5, 6). To gain access to the native structure, a deprotection is required. Before the carbohydrate of interest is attached to the surface of a microarray other required steps include the functionalization of the linker, purification steps (such as dialysis or the use of HPLC), and quality control (mass spectrometry and NMR analyses). A simplified scheme of the different steps from the desired target structure to surface immobilization is depicted in Fig. 1.

2  Chemical Synthesis of Carbohydrates…

15

Fig. 1. Required steps from defining the carbohydrate target structure until the immobilization of the synthesized oligosaccharide.

2.1. Building Blocks

A typical monosaccharide building block used in oligosaccharide assembly is equipped with different protecting groups that mask the hydroxyl and amine groups. Two different kinds of protecting groups, permanent and temporary ones, need to be distinguished. The former mask hydroxyl moieties have to be unveiled at the end of the synthesis whereas the latter mark sites of further glycosylation. Permanent protecting groups are commonly ethers (mostly Bn ethers), which can be cleaved by reduction at the end of the synthesis, but rather stable esters (such as Piv or even Bz esters) are also used. Temporary protecting groups are functional groups that are easy to cleave (but should be stable during the glycosylation reaction) such as trityl ethers, fluorenylmethoxycarbonyl (Fmoc) groups, and several silyl groups, while acetates (Ac) are often employed as temporary protecting groups. However, whether a protecting group is used as permanent or temporary is often depend strongly on the glycosylation strategy (3). Besides controlling regioselectivity by orthogonal protecting groups to account for possible branching of the carbohydrate chain, the stereochemistry

16

D.B. Werz

at the anomeric carbon must be controlled during the formation of the glycosidic bond. Placement of participating protective groups at the C-2 hydroxyl or amine groups ensures the formation of trans-glycosidic linkages. In contrast, nonparticipating groups and low temperatures during the glycosylation lead to the preferential installation of cis-glycosides. Furthermore, an anomeric leaving group that can be easily activated to induce the formation of the glycosidic linkage is needed (3, 7). A large variety of anomeric leaving groups are known and several of them are widely utitized in complex oligosaccharide synthesis. Figure  2 depicts a selection of such anomeric leaving groups (1–7) with their respective activation procedures to perform a glycosylation reaction (3). The oldest anomeric leaving groups used for the chemical synthesis of saccharides are glycosyl bromides 1 that are activated by silver salts. Today, the most commonly used anomeric leaving groups are phosphates 2, trichloroacetimidates 3, and thioethers 4 (3). To date, the synthesis of building blocks is the most timeconsuming process of oligosaccharide synthesis. Rather than synthesizing each building block separately integrated synthetic paths that grant access to several building blocks from a common precursor are desirable.

Fig. 2. Various anomeric leaving groups in 1–7 and their respective activation modes.

2  Chemical Synthesis of Carbohydrates…

17

Commonly, the differently protected and functionalized monosaccharides are accessed from naturally occurring sugar starting materials through a series of protection–deprotection maneuvers. Such a process establishes the desired protecting group pattern and typically requires 6–20 steps depending on the sugar, the protecting group ensemble, and the anomeric leaving group. A plethora of different methods to chemically distinguish the different hydroxyls already exists. The reactivity of the anomeric hydroxyl group that is part of a hemiacetal differs significantly from that of the other hydroxyls. The selective removal of acetate esters in the anomeric position and the readily occurring substitution of esters by alcohols such as n-pentenol or p-methoxybenzyl alcohol in the presence of Lewis acids are well established (8, 9). The C-6 hydroxyl also exhibits special reactivity as the only primary hydroxyl group of the sugar moiety. In addition, the C-6 hydroxyl is sterically exposed and increased reactivity often leads to high regioselectivity (9). Another very popular method for the selective protection of four or five hydroxyls relies on rendering two of them silent. cisHydroxyls easily form 5-membered rings, so-called isopropylidenes (9), upon treatment with acetone dimethylacetal in the presence of a Lewis acid (Fig. 3). Benzaldehyde and benzaldehyde dimethylacetal are prone to form 6-membered rings (benzylidene acetals, 10) involving the C-4 and the C-6 hydroxyl. An elegant method to lock the C-1 and C-2 hydroxyls is the formation of an orthoester 11 (Fig.  3). Usually orthoesters are synthesized by nucleophilic

Fig. 3. Methods used in building block synthesis to render two hydroxyls “silent”.

18

D.B. Werz

Fig. 4.  Synthesis of glycosyl phosphate 17 starting from d-glucose (13).

attack of an alcohol at the oxocarbenium ion that can be generated by a glycosyl bromide with the neighboring ester group in the presence of a nonnucleophilic base (9). This method commonly yields two stereoisomeric products. The elimination of two hydroxyls to form a 1,2 double bond is yet another method to reduce the number of hydroxyl groups to be distinguished. The remaining hydroxyls of these 1,2-glycals 12 are more easily distinguished. Finally, selective oxidation procedures allow the reinstallation of the eliminated hydroxyls at a later stage (9). For example, Fig.  4 shows the generation of a glucosyl phosphate with Fmoc as a temporary protecting group in position 4. d-Glucose is used as starting material. After peracetylation and anomeric substitution of acetate by bromide, Zn-mediated reduction eliminates two hydroxyl groups of the glucose 13 to afford the glucal 14. Protection with permanent protecting groups (Bn) in positions 4 and 6 mediated by a tin reagent and further protection with FmocCl as temporary protecting group yielded 16. Epoxidation with dimethyl dioxirane (DMDO) is followed by opening of the 1,2-anhydrosugar with dibutylphosphate. Protection of the ensuing C2 hydroxyl group with a participating group (e.g. pivaloyl chloride [PivCl)] produced the desired glucosyl phosphate 17 in excellent yield (10, 11). 2.2. Glycosylation Reaction and Assembly

Definitely it would go beyond the scope of this brief chapter to discuss in detail all the parameters which affect the glycosylation reaction (3, 4, 12). However, a short overview of the mechanistic picture of this important reaction seems to be appropriate. Nevertheless, detailed information about glycosylation mechanisms is mostly fragmented. Commonly, the glycosylation reaction involves nucleophilic displacement at the anomeric carbon. Due to the endocyclic oxygen next to the anomeric center a positive charge

2  Chemical Synthesis of Carbohydrates…

19

Fig. 5. Simplified mechanistic picture of a glycosylation reaction either with non-participating group (Case A) or with participating group (Case B) in position 2.

is strongly stabilized. Thus, after activation of 18 one may consider a glycosyl cation 19 (oxocarbenium ion) as intermediate if no further stabilising substituents are in reach (Fig. 5, Case A). A nucleophile such as a hydroxyl group can attack this species either from the top or from the bottom face; a mixture of trans- and cis-glycosides 20 and 21 results. In the case of a participating group in position 2, such a carbocation might also be stabilized by neighboring group participation leading to an acyloxonium ion as major intermediate (Fig. 5, Case B). Such an acyloxonium ion 23 shields either the top or the bottom face (depending on the stereochemistry at C-2). As a result, the attack of the nucleophilic hydroxyl group can only take place from the nonshielded side leading to transglycoside 24 as major isomer. Using this model, it is easy to understand why participating groups must not be used for the formation of cis-glycosides. For the creation of cis-glycosides, one has to rely, in most cases, on thermodynamic considerations. In the case of glucose, galactose, and fucose, as well as in their corresponding amines and acids, the a product is thermodynamically preferred due to the anomeric effect. Low reaction temperatures favor their formation. However, a complete a selectivity is often difficult to

20

D.B. Werz

achieve. In the case of mannose, the cis-mannosides are the b product. Therefore, the creation of b-mannosides requires other approaches that have been described in the literature (12, 13). In addition to the glycosylation reaction, the second important reaction in oligosaccharide assembly is the highly selective removal of protecting groups (9). For branched oligosaccharides, careful synthetic planning with respect to the temporary protecting group ensemble is required. Permanent protecting groups stay unaffected until the very final steps of the synthesis to procure the completely deprotected carbohydrate. The cleavage of temporary protecting groups after each glycosylation reaction is essential to liberate hydroxyls that are needed as nucleophiles for the next glycosylation step. Silyl protecting groups (e.g. in 25) are cleaved by fluoride sources or strong acids whereas fluorenylmethoxycarbonyl (Fmoc) groups (e.g. in 27) are cleaved by weak bases such as piperidine. Stronger bases such as sodium methoxide are required for the clevage of acetate (Ac) groups (as shown for 29) whereas levulinoate (Lev) esters in 31 need only hydrazine to be cleaved. An overview of four deprotection methods is provided in Fig. 6. Caution has to be taken by employing two temporary protecting groups next to each other. In some instances, a migration of the remaining protecting group might occur when one of the two is removed (especially in the presence of bases) (8). 2.3. Global Deprotection

Once the oligosaccharide is assembled by solution- or solid-phase techniques, the deprotection of all permanent protecting groups (global deprotection) is a challenging endeavour. Two approaches dominate: The first is a single-step procedure with sodium, carried out in liquid ammonia, often referred as Birch reduction (8). Benzyl ether groups as well as esters and carbonates are removed. The fact that olefinic moieties (except allyl) stay unaffected is the major advantage since terminal CC double bonds are often used as a chemical handle, as a future point of attachment to the surface of a microarray or to a carrier protein (8, 14). Thus it should not be attacked during the deprotection sequence. Besides olefinic moieties, the reducing end hemiacetal of an oligosaccharide is also maintainable during Birch debenzylation (15). Sometimes the commonly employed trichloroacetyl (TCA) amino protecting group causes problems resulting in a tremendous reduction of yield. To circumvent these difficulties, the transformation of the TCA group into an acetyl group by a radical initiated reduction with Bu3SnH is advantageous. To facilitate the further purification process, the unprotected sugar, obtained by Birch reduction, is frequently peracetylated using acetic anhydride and pyridine (9, 10). The experimental setup for the Birch reduction using sodium in liquid ammonia is relatively high. Minor impurities in the starting material may cause a serious decrease in yield. A major advantage compared to hydrogenolysis procedures is avoidance of heavy

2  Chemical Synthesis of Carbohydrates…

21

Fig. 6. Four examples of temporary protecting groups (TBS, Fmoc, Ac, and Lev) and methods for their selective removal.

22

D.B. Werz

metals. Traces of such metals may cause problems in biological experiments employing the synthetic carbohydrates. The other protocol requires two steps. In a first step, ester groups have to be removed by base, before hydrogenolysis cleaves the benzyl (Bn) ethers (9). In the latter case, palladium on charcoal or Pd(OH)2 on charcoal (Pearlman’s catalyst) are widely used as catalysts. The reaction itself takes place under an atmosphere of hydrogen. Commonly the reaction proceeds well and is finished over night. If problems with the deprotection of several benzyl ethers (e.g. sterically very hindered Bn groups) occur, hydrogen pressure up to 100 bar may be applied. It is evident that the conditions employed for hydrogenolysis also reduce carbon–carbon double bonds. If one needs an alkene as a handle, the Birch reduction is the method of choice for global deprotection. 2.4. Case Study of Blood Group A Antigen

To illustrate the process of carbohydrate assembly, the blood group A antigen 34 with a thiol linker is chosen as an example (16). This glycan is a branched tetrasaccharide consisting of galactosamine, fucose, galactose, and glucosamine (Fig. 7). For the installation of a chemical handle being necessary for surface immobilization, a pentenyl group was attached to the reducing end of the sugar. Such an olefinic moiety can be transformed easily at the end of the total synthesis into a thiol unit. Figure 7 shows the retrosynthetic analysis into five building blocks 36–40. Pivaloyl (Piv) and acetate (Ac) esters are employed as permanent protecting groups for hydroxyls benzyl (Bn) ethers. The amine moieties are protected in the case of galactosamine 36 as azide (N3), in the case of glucosamine 40 as trichloroacetamide (TCANH). In the latter case a participating group, namely TCA, is necessary to achieve b selectivity whereas nonparticipating nitrogen functionality, namely the azide, is necessary to achieve a selectivity. For the branching galactose 37, two orthogonal temporary protecting groups are required. Fmoc in position 2 ensures, with its ability for neighboring groupparticipation, b selectivity and can be cleaved later on by the weak base piperidine without affecting the levulinoate in position 3. In the case of the fucose building block 39, it is again important to have a nonparticipating group in position 2 to afford mainly the a product. For the anomeric leaving groups, a mixed strategy consisting of one trichloroacetimidate and three phosphates was chosen. This selection was based on arguments with respect to the ease of building block synthesis. Figure  8 summarizes the assembly and deprotection process (16). The pentenyl functionalized monosaccharide 41 and galactosyl phosphate 37 were reacted in the presence of the Lewis acid TMSOTf, as activating agent, to produce a disaccharide whose Fmoc group was immediately deprotected to afford 42. Further fucosylation and removal of levulinoate in the presence of hydrazine yielded trisaccharide 43. To achieve high a selectivity,

2  Chemical Synthesis of Carbohydrates…

23

Fig. 7. Retrosynthesis of the A antigen 34 with a thiol linker, building blocks 36–40.

the last glycosylation was carried out at low temperatures. The completely protected tetrasaccharide 35 was subjected to Birch reduction. Sodium in liquid ammonia removed all permanent protecting groups and transformed the azide moiety as well as the

24

D.B. Werz

Fig. 8. Assembly of the tetrasaccharidic A antigen 34 with a thiol linker.

trichloroacetamide into “naked” amines, thus achieving global deprotection. The reaction with acetic anhydride and pyridine converted all hydroxyls to the corresponding acetates and all amines to the respective acetamido groups. This product is much less polar than the compound directly obtained after Birch reduction and can be more easily purified. A radical-initiated addition of thioacetic acid introduced the sulfur moiety (16). In the final step all acetates are cleaved to generate the native structure 34. Dialysis might be a good choice to get rid of salts, which might have accumulated during the last steps of the synthesis. Lyophilisation yields the carbohydrate ready for attachment to microarrays.

2  Chemical Synthesis of Carbohydrates…

25

3. Immobilization Techniques Microarrays in the “chip” format, prepared by attachment of biopolymers to a surface in a spatially discrete pattern, have enabled a low-cost and high-throughput methodology for screening interactions involving these molecules (17–19). The most important advantages compared to classical methods are that microarrays allow for several thousand binding events to be screened in parallel, hence the experiment requires only miniscule amounts of both analyte and ligand. Thus, binding profiles and lead structures can be readily examined. In addition, carbohydrate microarrays are ideal to detect interactions that involve carbohydrates since the multivalent display of ligands on a surface (cluster effect) overcomes the relative weakness of these interactions by mimicking cell–cell interfaces. However, an important question to be tackled is the way in which a carbohydrate becomes immobilized on the surface. This book provides and discusses a large variety of methods in detail. However, in the following paragraphs I would like to provide an overview of some selected techniques from the viewpoint of an organic chemist. Older methods for the preparation of carbohydrate microarrays consist of nitrocellulose-coated slides (in the case of noncovalent immobilization of microbial polysaccharides) (20) or self-assembled monolayers modified by Diels–Alder mediated coupling of cyclopentadiene-derived oligosaccharides (21). Unfortunately, the former method requires large polysaccharides or lipid modified sugars for the noncovalent interaction. The latter requires the preparation of oligosaccharides bearing the sensitive cyclopentadiene moiety. Today noncovalent immobilization is mainly based on fluorine–fluorine interactions (22) or DNA-directed immobilization (23). In the first case, the surface is coated by perfluorinated hydrocarbons and the carbohydrate to be attached bears a perfluorinated alkyl chain. In the latter case, the specific pairing of DNA bases due to the different number of hydrogen bridges is the key to immobilization; one single strand of DNA is attached to the carbohydrate moiety whereas the complementary one is attached to the surface of the microarray. For covalent immobilization (24–26) very good results were obtained by utilizing maleimide functionalization of glass slides and the immobilization of the oligosaccharides with thiol-containing linkers 44. As pointed out in the Subheading 2.4, thiols are obtained in a facile way by reacting double bonds with thiolacetic acid under radical conditions (either initiated by light or radical starters such as AIBN). Sodium methoxide or even potassium carbonate cleaves the thioacetate to generate the free thiol. In a Michaeltype addition the thiol moiety attacks the maleimide functionality. A stable covalent bond is formed to afford modified surfaces of type 45. Care has to be taken when thiol-containing substrates are

26

D.B. Werz

stored for a longer time. Oxygen from the air leads to the formation of disulfide linkages that no longer react with maleimides. Therefore, a weak reducing agent is often added to generate the thiols out of the respective disulfides. Another possible reaction partner is iodoacetyl groups, which are immobilized on a surface. Due to the high nucleophilicity of the sulfur, a simple nucleophilic substitution leads to immobilized saccharides of type 46 (Fig. 9). Also, other moieties such as linkers functionalized with azides are utilized for surface immobilization. In a Huisgen 1,3-dipolar cycloaddition, a five-membered trizole is formed with an alkyne under the catalytic influence of copper. This chemistry is often referred to today as “click-chemistry” (27). Of course, the carbohydrate part can bare an alkyne unit, while the surface bares the respective azide. Also, primary amines, which react under conditions of a reductive amination with aldehydes or vice versa to secondary amines, are utilized. However, the generation of aldehydes – either as part of a linker attached to the carbohydrate moiety or as part of the surface – is difficult, and decomposition may occur readily. If amines are used as reactive group for immobilization, one should use an activated amide as counterpart. Both functionalities are rather stable, and high yields are obtained. If one likes to use thiols instead of amines, but due to synthetic considerations amines have

Fig. 9. Two possibilities to attach thiol-functionalized saccharides to surfaces.

2  Chemical Synthesis of Carbohydrates…

27

Fig. 10. Transformation of an amine-functionalized linker 47 into a thiol-functionalized linker 49 by using Traut’s reagent 48.

been created, Traut’s reagent 48 will be able to convert amines 47 into corresponding thiols 49 (Fig. 10) (28). By this procedure, the length of the linker is increased by five heavy atoms. In general, every simple reaction that is easy to handle and promises a high reaction yield can be used for the immobilization of carbohydrates. The bonds formed have to be stable and should not be cleaved easily by hydrolysis. Of course, the corresponding functional groups have to be installed on both reaction partners and the oligosaccharide as well as the modified surface of the microarray. Further considerations that should be taken into account are the kind of linker, its length, and the point of attachment. Polyethylenglycol (PEG) linkers reveal a higher degree of water solubility than common alkyl chains. Linkers that are too short might cause problems when large proteins interact with the carbohydrate. As point of attachment, in most cases, the anomeric center is chosen to mimic the native structure; however, of course, other functionalization sites are also possible.

4. Conclusion Carbohydrate synthesis is still a rather time-consuming endeavour; however, major efforts have been done during the last decades to fasten this process such as the use of integrated paths to create building blocks, improved glycosylation strategies and even the introduction of automated methods for the assembly of oligosaccharides. The right choice of temporary and permanent protecting groups, participating and nonparticipating ones, is the prerequisite for a successful stereoselective assembly. As anomeric leaving groups especially trichloroacetimidates, phosphates, and thioethers have become prominent. In a final step, global deprotection has to be performed, either by saponification and hydrogenolysis or by Birch reduction. For surface immobilization, handles such as pentenyl chains can be easily converted at the end of the synthesis into thiols.

28

D.B. Werz

Acknowledgments The author would like to thank the Deutsche Forschungsgemeinschaft (Emmy Noether Fellowship) and the Fonds der Chemischen Industrie (Liebig Fellowship) for financial support. References 1. Seeberger, P. H., and Werz, D. B. (2007) Synthesis and medical applications of oligosaccharides. Nature 446, 1046–1051. 2. Werz, D. B., Ranzinger, R., Herget, S., Adibekian, A., von der Lieth, C.-W., and Seeberger, P. H. (2007) Exploring the Structural Diversity of Mammalian Carbohydrates (“Glycospace”) by Statistical Databank Analysis. ACS Chem. Biol. 2, 685–691. 3. Demchenko, A. V. (Ed.) (2008) Handbook of Chemical Glycosylation, Wiley-VCH, Weinheim. 4. Zhu, X., and Schmidt, R. R. (2009) New principles for glycoside-bond formation. Angew. Chem. Int. Ed. 48, 1900–1934. 5. Plante, O. J., Palmacci, E. R., and Seeberger, P. H. (2001) Automated solid-phase synthesis of oligosaccharides. Science 291, 1523–1527. 6. Seeberger, P. H., and Werz, D. B. (2005) Automated synthesis of oligosaccharides as basis for drug discovery. Nature Reviews Drug Discovery 4, 751–763. 7. Schmidt, R. R., Castro-Palomino, J. C., and Retz, O. (1999) New aspects of glycoside bond formation. Pure Appl. Chem. 71, 729–744. 8. Werz, D. B., and Seeberger, P. H. (2005) Total Synthesis of Antigen Bacillus Anthracis Tetrasaccharide – Creation of an Anthrax Vaccine Candidate. Angew. Chem. Int. Ed. 44, 6315–6318. 9. Lindhorst, T. K. (2007) Essentials of Carbohydrate Chemistry and Biochemistry, 3rd edition, Wiley-VCH, Weinheim. 10. Werz, D. B., Castagner, B., and Seeberger, P. H. (2007) Automated Synthesis of the TumorAssociated Carbohydrate Antigens Gb-3 and Globo-H: Incorporation of a-Galactosidic Linkages. J. Am. Chem. Soc. 129, 2770–2771. 11. Love, K. R., and Seeberger, P. H. (2004) Automated solid-phase synthesis of protected tumor-associated antigen and blood group determinant oligosaccharides. Angew. Chem. Int. Ed. 43, 602–605. 12. Crich, D. (2010) Mechanism of a Chemical Glycosylation Reaction. Acc. Chem. Res. 43, 1144–1153.

13. Crich, D. (2007) Stereocontrolled glycosylation: recent advances: b-D-rhamnosides and b-D-mannans. ACS Symposium Series 960, 60–72. 14. Tamborrini, M., Werz, D. B., Frey, J., Pluschke, G., and Seeberger, P. H. (2006) AntiCarbohydrate Antibodies for the Detection of Anthrax Spores. Angew. Chem. Int. Ed. 45, 6581–6582. 15. Dudkin, V. Y., Miller, J. S., and Danishefsky, S. J. (2004) Chemical Synthesis of Normal and Transformed PSA Glycopeptides. J. Am. Chem. Soc. 126, 736–738. 16. Horlacher, T., Oberli, M. A., Werz, D. B., Kröck, L., Bufali, S., Mishra, R., Sobek, J., Simons, K., Hirashima, M., Niki, T., and Seeberger, P. H. (2010) Determination of Carbohydrate-Binding Preferences of Human Galectins with Carbohydrate Microarrays. ChemBioChem 11, 1563–1573. 17. Barnes-Seemann, D., Park, S. B., Koehler, A. N., and Schreiber, S. L. (2003) Expanding the functional group compatibility of small molecule microarrays: Discovery of novel calmodulin ligands. Angew. Chem. Int. Ed. 42, 2376–2379. 18. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specifity assignments of carbohydrate-protein interactions. Nat. Biotechnol. 20, 1011–1017. 19. Hergenrother, P. J., Depew, K. M., and Schreiber, S. L. (2000) Small Molecule Microarrays: Covalent Attachment and Screening of AlcoholContaining Small Molecules on Glass Slides. J. Am. Chem. Soc. 122, 7849–7850. 20. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells, Nat. Biotechnol. 20, 275–281. 21. Houseman, B. T., and Mrksich, M. (2002) Carbohydrate arrays for the evaluation of protein binding and enzymatic modification, Chem. Biol. 9, 443–454. 22. Ko, K.-S., Jaipuri, F. A., and Pohl, N. L. (2005) Fluorous-Based Carbohydrate Microarrays. J. Am. Chem. Soc. 127, 13162–13163.

2  Chemical Synthesis of Carbohydrates… 2 3. Zhang, J., Pourceau, G., Meyer, A., Vidal, S., Praly, J.-P., Souteyrand, E., Vasseur, J.-J., Morvan, F., and Chevolot, Y. (2009) Specific recognition of lectins by oligonucleotide glycoconjugates and sorting on a DNA microarray. Chem. Commun. 2009, 44, 6795–6797. 24. Love, K. R., and Seeberger, P. H. (2002) Carbohydrate Arrays as Tools for Glycomics. Angew. Chem. Int. Ed. 41, 3583–3586. 25. Kiessling, L. L., and Cairo, C. W. (2002) Hitting the sweet spot, Nat. Biotechnol. 20, 234–235.

29

26. Disney, M. D., and Magnet, S., Blanchard, J. S., and Seeberger, P. H. (2004) Aminoglycoside Microarrays to Study Antibiotic Resistance. Angew. Chem. Int. Ed. 43, 1591–1594. 27. Kolb, H. C., Finn, M. G., and Sharpless, K. B. (2001) Click chemistry: diverse chemical function from a few good reactions. Angew. Chem. Int. Ed. 40, 2004–2021. 28. Tada, T., Mano, K., Yoshida, E., Tanaka, N., and Kunugi, S. (2002) SH-group introduction to the N-terminal of subtilisin and preparation of immobilized and dimeric enzymes. Bull. Chem. Soc. Jp. 75, 2247–2251.

Chapter 3 General Consideration on Sialic Acid Chemistry Hongzhi Cao and Xi Chen Abstract Sialic acids, also known as neuraminic acids, are a family of negatively charged a-keto acids with a nine-carbon backbone. These unique sugars have been found at the termini of many glycan chains of vertebrate cell surface, which play pivotal roles in mediating or modulating a variety of physiological and pathological processes. This brief review covers general approaches for synthesizing sialic acid containing structures. Recently developed synthetic methods along with structural diversities and biological functions of sialic acid are discussed. Key words: Sialic acid, Sialoside, Glycosylation, Chemoenzymatic synthesis, Sialyltransferase, Sialylation

1. Sialic Acids: A Unique Family of Monosaccharide 1.1. Structure Diversity of Sialic Acid

Sialic acids constitute a unique family of 2-keto-3-deoxynonulosonic acids. These nine-carbon monosaccharides have a tertiary anomeric center at C-2 position linked to an anomeric carboxylate, a deoxygenated C-3, a glycerol side chain at C-6 and different substituents at C-5. Three basic forms of sialic acids (Fig. 1) are N-acetylneuraminic acid (Neu5Ac, N-acetyl-5-amino-3,5-dideoxy-d-glycero-d-galacto2-nonulosonic acid, 1), N-glycolylneuraminic acid (Neu5Gc, having a hydroxyl group in place of one of the hydrogens at the N-acetyl group of Neu5Ac, 2), and deaminoneuraminic acid (KDN, 3-deoxy-d-glycero-d-galacto-2-nonulosonic acid, 3) (1–3). Further modifications of these three basic sialic acid forms, including O-acetylation, O-methylation, O-lactylation, O-sulfation, and O-phosphorylation at the hydroxyl groups at C-4, C-8, and/ or C-9, changes on the group at C-5, and the formation of

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_3, © Springer Science+Business Media, LLC 2012

31

32

H. Cao and X. Chen

HO

O

OH 9

8

N H

6

7 5

OH

OH

OH

O 4

CO2-

2

OH

HO

1

3

O

OH

OH

HO

N H HO

CO2-

O

OH

HO

CO2-

O

HO HO

Neu5Gc,2

Neu5Ac,1

OH

OH

KDN,3

Fig. 1. Three basic forms of naturally occurring sialic acids.

HO HO

OH

CO2-

OH O

AcHN

O

OH O

HO

HO

HO

OH

HO

HO O

O AcHN

HO

CO2-

OH

HO

CO 2-

OH

HO

O

AcHN O

O

(4Siaα2-6Galα1-)n

O

HO

HO

O

O HO

CO 2-

OH O O

OH

OR

CO2-

OH

AcHN

AcHN

n

OH

n

(9Siaα2-9Siaα2-)n HO

O HO

CO2O

HO

OH

O

Siaα2-6GalNAcαOR

OH

n

(8Siaα2-8Siaα2-)n

O

AcHN

OH

HO

HO

CO 2-

OH HO

OR

O AcHN

O

HO

AcHN

O

HO

HO

OH

CO2-

OH

O

HO

OH

Siaα2-6GalβOR

Siaα2-3GalβOR

O AcHN

O

AcHN

OR

CO 2-

OH

O

OH

O HO

(4Siaα2-6Glcα1-)n

OH n

Fig. 2. Common sialic acid containing structures with different sialyl linkages.

intramolecular lactone or lactam leads to the identification of more than 50 natural occurring sialic acids (2, 3). These structurally distinct sialic acids are the key components of many important glycolipids and glycoproteins presented on vertebrate cell surface. In vertebrates, sialic acids are typically presented in (Siaa2-8Sia)n homopolymers or as the outermost carbohydrate units of glycoconjugates, such as in Siaa2-3GalOR, Siaa2-6GalOR, Siaa2-6GalNAcOR structures (Fig. 2). In bacteria, sialic acids have been found in extracellular capsular polysaccharides and lipopolysaccharides (LPS) in homopolymers of sialic acid with a2-8-, a2-9-, or alternating a2-8/9-linkages (4, 5) as terminal residues in the side chains of capsular polysaccharide heteropolymers (2, 3, 6, 7), as terminal residues in the glycans of bacterial lipooligosaccharides (LOS) of various Gram-negative bacteria (8), and as internal residues in heteropolymers of capsular polysaccharides of Neisseria meningitidis (9). 1.2. General Biological Significance of Sialic Acid

As the terminal monosaccharides located at the outermost position of cell surface glycans, sialic acids are the key recognition sites of many biomolecules. Moreover, the structural diversity of sialic acids also help to fine-tune the pivotal roles that sialic acids play in

3  General Consideration on Sialic Acid Chemistry

33

biological and pathological processes, such as cell–cell interaction (10, 11), inflammation (12), fertilization (13), infection (14), differentiation (15), malignancy (16–18), and cell signaling (19, 20). They are also highly regulated during embryonic development (21–23). In bacteria, they are believed to be important virulence factors (24) and used by the bacteria to mimic sialylated host cell surface carbohydrate structures to evade detection and attack by the host’s immune defense mechanisms (3, 8, 25–27). Such examples include a2-8-linked (8Siaa2-8Siaa2-)n capsular polysialic acid (also called colominic acid) expressed by N. meningitidis serogroup B (NmB) (28) and Escherichia coli K1 (29), which mimics the structure of human neural cell adhesion molecules (NCAMs) (30–32). Other examples include sialylated oligosaccharides or polysaccharides produced at surface of pathogenic bacteria such as Campylobacter jejuni, Group B Streptococcus, P. haemolytica A2 (33). The sialic acid-based recognition processes are believed to be closely related to the fine structures of the sialic acids, the underlying glycan structures, and the types of the sialyl linkages. Malignant cells often display surface glycosylation patterns differing from normal cells. The overexpression of tumor-associated carbohydrate antigens (TACAs) is commonly observed on different types of malignant cells (34–36). Sialic acid residue is part of many TACAs and is believed to play essential role in different stages of cancer progress (37). These sialic acid containing TACAs (Fig. 3), such as sialyl Lewis x (sLex), sialy Lewis a (sLea), gangliosides (GM3, GM2, GM1, GD2, GD3, fucosyl GM1), sialyl Tn and sialyl T, and so on, have been the targets of studies for developing carbohydrate-based vaccines (38–40). Several carbohydrate-based

OH OH

HO

AcHN

O

CO2O

OH

OH

HO

O

O OH

O

O

OH O

HO

OR NHAc

OH OH

AcHN

CO 2-

OH NHAc O OR O O O HO OH OH O OH HO OH

O

O

OH

OH

HOOH Sialyl Lewis x (sLex) OH OH

HO

CO2O

AcHN

OH

O

O HO OH

OH

O HO

Sialyl Lewis a (sLea) OH O OH

HO

OR AcHN

OH O

HO

HO O

O O

HO OH

HO OH O HO

AcHN

OH O

OH

O

O OH

OH

O

O HO OH

HO

AcHN

OH -O C 2

CO 2-

HO O

OH O

HO

AcHN OR

O OH O

O OH HO

OH O

OH Fucosyl GM1

HO AcHN OR

CO2-

OH O OH

O AcHN

OH OH

HO

CO 2-

OH

AcHN

O

O

OH

O HO Sialyl Tn (sTn) AcHN OR

Sialyl T (sT)

GM3 HO

OH OH

OH

OH OH

CO2O

OH

O AcHN n

Polysialic acid (PSA)

Fig. 3. Representative sialic acid containing tumor-associated carbohydrate antigens (TACAs).

CO2O

OH

OR

34

H. Cao and X. Chen

vaccines are currently in clinic trails (38). The isolation of carbohydrates from natural source is a tedious process and often yields impurities in the carbohydrate preparations. Therefore, synthetic carbohydrates have been attractive targets for the development of carbohydratebased vaccine. Obtaining structurally defined complex carbohydrates in sufficient amount has been a big challenge (41). Recent advances in chemical, enzymatic, and chemoenzymatic synthesis provide efficient methods to overcome the challenge. Neu5Ac (1), the most abundant form of sialic acid, is synthesized from N-acetylmanosamine (ManNAc, 4) in both eukaryotes and prokaryotes though its biosynthetic pathways differ in bacteria and vertebrates. The synthesis of Neu5Ac from ManNAc in eukaryotes occurs in the cytosol with three consecutive steps as shown in route a in Fig. 4, including the formation of ManNAc-6-phosphate (5) from ManNAc (4) catalyzed by an N-acetylmannosamine kinase (EC2.7.1.60), formation of Neu5Ac-9-phosphate (6) catalyzed by an N-acetylneuraminic acid 9-phosphate synthase (EC2.5.1.57), and dephospholation catalyzed by an N-acetylneuraminic acid 9-phosphate phosphatase (EC3.1.3.29) (42). In bacteria, Neu5Ac is synthesized directly through the reaction of ManNAc with phosphoenolpyruvate (PEP) catalyzed by a Neu5Ac synthase (route b in Fig. 4). Sialic acid aldolase (lyase) catalyzing the reversible cleavage of Neu5Ac for the formation of ManNAc and pyruvate can be used in vitro for the inexpensive production of Neu5Ac from ManNAc and excess amount of pyruvate (28, 43–46). The production of activated sugar nucleotide donor cytidine 5¢-monophosphate sialic acid (CMP-sialic acid) required by sialyltransferases (47, 48) is catalyzed by CMP-sialic acid synthetases (CSSs, EC 2.7.7.43) in both eukaryotes and bacteria. CSSs catalyze the formation of CMP-Neu5Ac and pyrophosphate from Neu5Ac and CTP in the presence of a metal cation such as Mg2+ or Mn2+ (3, 42, 49). The biosynthetic pathway of CMP-KDN in bacteria is believed to follow a similar route as CMP-Neu5Ac though it is not

1.3. Biosynthetic Pathway of Sialic Acid and CMP-Sialic Acid

-2

O3PO NHAc O HO OH HO ManNAc-6-P, 5

ADP

ATP

ManNAc kinase

NHAc HO O HO HO ManNAc, 4

PEP Neu5Ac Pyruvate synt hase or Pi

PEP Neu5Ac-9-P synthase Pi -2O PO 3

OH

AcHN

route (a)

OH

OH O

HO Neu5Ac-9-P, 6

CO2-

route (b)

Neu5Ac-9-P HO phosphatase H2O

Pi

OH

OH

AcHN

OH

OH O

HO Neu5Ac, 1

CO2-

Pyruvate Sialic acid aldolase ( lyase)

CTP CSS

PPi

HO

HO

OCMP

OH

AcHN

O

HO CMP-Neu5Ac, 7

Fig. 4. Biosynthetic routes of Neu5Ac and CMP-Neu5Ac in eukaryotes (route (a)) and bacteria (route (b)).

CO2-

3  General Consideration on Sialic Acid Chemistry

35

as well elucidated (1, 3, 48, 50, 51). CMP-Neu5Gc, which has not been reported in bacterial sources (3), is obtained from CMPNeu5Ac by oxidative conversion of the N-acetyl group of Neu5Ac to the N-glycolyl residue in Neu5Gc catalyzed by a CMP-Neu5Ac hydroxylase (52). In addition to the biosynthesis discussed earlier, Neu5Ac, Neu5Gc, and KDN deliberated from glycoconjugates in catabolic pathways can be recycled via salvage pathways to form the activated sialic acid donors for sialyltransferases (53, 54). Sialyltransferases are the key enzymes that catalyze the transfer of sialic acid residue from CMP-sialic acid to suitable acceptors for the formation of sialyl linkages. Different sialyl linkages are usually formed by different sialyltransferases. Eukaryotic sialyltransferases are usually more specific than their bacterial counterparts in the formation of sialyl linkages and recognizing suitable acceptor substrates, though all are relatively flexible in using CMP-sialic acids containing different sialic acid forms. Due to their high activity, high solubility, and suitability to be expressed in large amount in E. coli system, bacterial sialyltransferases have been cloned from different species and expressed in E. coli for high efficient preparation of complex sialosides (1). Biosynthetically, sialic acid modifications including O-acetylation, O-methylation, O-lactylation, and O-sulfation, usually take place after the formation of sialic bonds. These modifications thus belong to carbohydrate postglycosylational modifications (55). Nevertheless, due to the promiscuity of bacterial sialyltransferases and other bacterial enzymes involved in the CMP-sialic acid synthesis, naturally occurring sialic acid modifications can be introduced to monosaccharide stage by chemical or enzymatic methods, which can be used by multiple enzymes for the synthesis of CMP-sialic acids and sialosides with different sialic acid forms (56–58).

2. Difficulties in Chemical Formation of Sialoside Bond

Synthesis of structurally defined complex carbohydrates poses great challenges due to the inherent structural complexity of the carbohydrates such as the presence of multiple hydroxyl groups of similar reactivity and varied glycosidic linkages in the molecule. Among different chemical glycosylation processes, sialylation is considered as one of the most difficult reactions due to the hindered tertiary anomeric center and the lack of a neighboring participating group in sialic acids (59, 60). Many factors affect the outcome of chemical sialylation, including the protecting group patterns of sialyl donor and glycosyl acceptor, the leaving group of sialyl donor, the promoter, reaction solvent, temperature, and so on. The most significant challenge of chemical sialylation comes from the inherited unique structural feature of sialic acid.

36

H. Cao and X. Chen

Neu5Ac is the most abundant sialic acid form in nature. So far, chemical sialylation has been mainly focused on the synthesis of Neu5Ac-containing structures. General chemical sialylation methods involve an activated sialic acid donor (8) with a proper leaving group (LG) (Fig.  5) and an acceptor (R¢OH) bearing a single free hydroxyl group. The electron-withdrawing carboxylate at the C-2 anomeric carbon destabilizes the formation of the oxocarbenium ion intermediate (9 and 10). Furthermore, the tertiary anomeric center at C-2 of sialic acid donor prevents an easy nucleophilic attacking by the hydroxyl group in the acceptor and makes sialylation a slow process. Therefore, the competing elimination process is usually quite significant to generate a considerable amount of 2,3-dehydro sialic acid byproduct, 12 (Fig. 5). Due to the lack of a hydroxyl group at C-3, the neighboring carbon of C-2 anomeric center of sialic acid, functional protecting group cannot be easily introduced at C-3 to regulate the stereoselectivity of the sialylation process for the formation of desired a-sialyl linkages found in nature. In addition, the desired a-sialyl linked products are thermodynamically less stable than their b-anomers due to anomeric effect (60). As a consequence, many early sialylation methods generally lead to low yields due to poor stereoselectivity and the generation of the elimination byproduct. It is worth to mention that commonly used tricholoracetimidate glycosyl donor (Schmidt donor) is not suitable for sialylation due to the unique structural feature of sialic acid (61). Other than the challenges described earlier for general chemical sialylation, chemical sialylation of a sialic acid to another sialic acid for the formation of a2-8-linked disialyl motif Neu5Aca2-8Neu5 AcaOR has additional complications due to decreased nucleophilic activity of the C-8 hydroxyl group in sialic acid acceptor caused by its intramolecular hydrogen bonding network with C-5 acetamido group, C-1 carboxyl, or C-2 oxygen (Fig. 6) (59).

PO

LG

OP O

AcHN PO

CO2R

Activation

PO

PO

OP

O

AcHN PO

8

CO2R

OP O

AcHN

OR'

PO

PO

OP

O

CO2R

AcHN

El im in at io n

PO

10

+

PO

PO

-H

PO

9

CO2R

R' Gl O yc os H yl at io n

PO

PO

PO

OP

O

AcHN PO

11

Fig. 5. A general chemical sialylation approach and related elimination process.

12

CO2R

3  General Consideration on Sialic Acid Chemistry

PO

O

O

H

OR

OP

PO O

AcHN PO

OR'

O

H

O

R' H O OP PO

OP O

AcHN PO

37

CO2R

AcHN

CO 2R O

OR'

PO

Fig. 6. Intramolecular hydrogen bonding network of C-8 hydroxyl group of Neu5Ac acceptor.

3. The Influence of Different Protecting Groups on Reactivity

The reactivity and stereoselectivity of chemical sialylation can be greatly affected by the protecting groups on sialyl donors. The stereoselective formation of a-sialyl linkages has been improved by introducing different protecting groups at C-1, C-3, and C-5 of sialyl donors.

3.1. C-1 Modification

Several auxiliary protecting groups with potential neighboring participating effect have been introduced to the C-1 carboxyl group in sialyl donors to overcome the low stereoselectivity and low yield of sialylation. For example, several decades ago, Ratcliff and coworkers noticed that benzyl ester protection of carboxyl in Neu5Ac gave better yields and stereoselectivity than commonly used methyl ester protection under Koenigs–Knorr glycosylation conditions (60). Recently, N,N-dimethylglycolamide was developed as a new C-1 auxiliary group in sialyl donors by the Gin group (62). This C-1 auxiliary neighboring participating group can stabilize the oxocarbenium ion intermediate to form a bicyclic intermediate and to improve a-selectivity for the formation of both a2-3- and a2-6-sialyl linkages. In another method, Wong and coworkers converted the C-1 carboxyl into a hydroxymethylene and found this modification can boost the reactivity of sialic acid donor and improve glycosylation yields significantly though the undesired b-sialosides were the predominated products (63).

3.2. C-3 Modification

Using a C-3 directing auxiliary group (Y in sialyl donor 13, Fig. 7) was one of the most common approaches for chemical sialylation a decade ago, and it was usually classified to “indirect glycosylation” methods (59, 60). The C-3 auxiliary group was introduced as a neighboring participating group to stabilize the positive charge of oxocarbenium ion intermediate by forming a bicyclic intermediate 14 and to prevent competing 2,3-elimination reaction. Several directing groups have been introduced and successfully employed to the synthesis of a broad library of sialosids (59). The C-3 hydroxyl group was initially introduced by Goto and coworkers for the synthesis of a2-9- and a2-8-linked Neu5Ac dimers under Koenigs–Knorr glycosylation conditions (64). They found that an O-acetyl-protected hydroxyl group or bromine at C-3 can provide

38

H. Cao and X. Chen

AcO

OAc OAc

Y O

AcHN

AcO 14

CO 2Me

Sialylation AcO

OAc OAc

AcO

AcO 15

R'OH

AcO

AcO

LG Y

OAc OAc AcO

Ph

O S

CO2Me O

AcHN

13

Y = OAc, Br, SPh, SePh,

OR' Y

Removal of Y

CO2Me O

AcHN

O

AcHN

Activation OAc OAc

CO2Me

OR'

16

S or

Fig. 7. Schematic illustration of using C-3 auxiliary groups to improve sialylation stereoselectivity.

better stereoselectivity compared to the same types of sialyl donors with no C-3 directing group. Ogawa et al. introduced C-3 phenylthiol and phenylselenyl groups as the neighboring participating groups (65). The phenylthiol group was shown to be a superior directing group due to its increased size and polarizability compared to oxygenbased C-3 auxiliaries. Furthermore, it can be easily removed after glycosylation by one-step treatment with Bu3SnH and AIBN under radical reaction conditions (65). Other C-3 auxiliary groups have also been employed to the chemical synthesis, such as 2,4-dimethylbenzenesulfenyl (66) and phenoxythiocarbonyloxy (67) groups, which have similar properties as the phenylthiol directing group. The auxiliary participating group strategy can dramatically increase the desired a-sialoside products and minimize the elimination byproduct. However, the disadvantage is obvious due to the required additional steps for adding the C-3 auxiliary group during the synthesis of sialyl donor and removal of the group after glycosylation. Currently, this “indirect strategy” is rarely used in chemical sialylation reactions. 3.3. C-5 Modification

In the process of looking for other positions for adding auxiliary groups to help chemical sialylation, Boons and Demchenko found that adding an additional acetyl group at the nitrogen in the N-acetyl group at C-5 in Neu5Ac (17) can significantly improve the reactivity of both sialyl donor and acceptor compared to the corresponding mono-N-acetylated sialyl donor and acceptor (68). This second N-acetyl group can be easily introduced by acetylation to mono-N-acetylated sialyl donor or acceptor and can be conveniently removed under Zemplén conditions after the glycosylation. Encouraged by these results, many other functional groups, such as azido (18) (69), N-trifluoroacetyl (NHTFA, 19) (70), N-trichloroethoxycarbonyl (NHTroc, 20) (71), N-t-butoxycarbonylacetamido (NAcBoc, 21)

3  General Consideration on Sialic Acid Chemistry

AcO

OAc OAc

Ac2N

AcO

CO2Me O

AcO

LG

CO2Me O

AcO

OAc OAc N3

17

OAc OAc

AcBocN

AcO

21

LG

AcO

CO2Me O

AcO

LG

AcO

22

LG

CO2Me O

TFAHN

CO2Me AcO O

HO

OAc OAc

18

OAc OAc

PhthN

AcO

O

23

CO 2Me O

TrocHN AcO

CO2Me O

HN

OAc OAc

19

OAc OAc

O

LG

AcO

LG

O

LG

20 CO2Me

O

O

HN OP

39

OP OP

LG

24

Fig. 8. Examples of C-5 modified sialyl donors.

(72), phthalimido (NPhth, 22) (73), 5-N,4-O-oxazolidinone (23) (74), and 5-N,7-O-oxazinone (24) (75) groups (Fig. 8) have been introduced to the C-5 position to increase the reactivity of sialyl donor or acceptor. The C-5 modification strategies along with the novel sialyl donor developed in the last decades have been demonstrated to be applicable as general approaches for chemical sialylation (76).

4. Different Chemical Sialylation Strategies

4.1. 2-Halo Sialyl Donors: The Koenigs– Knorr Protocol

Different sialyl donors have been developed. Several common and effective ones are discussed later with the corresponding suitable protecting groups. The first example of chemical sialylation was reported by Meindl et al. in 1965 (77). The Koenigs–Knorr glycosylation method was employed using 2-halo sialic acid derivatives as the donors and a silver salt such as Ag2CO3 (classical Koenigs–Knorr) or Hg(CN)2/ HgBr2 (Helferich modification) as the promoter (Fig.  9) (77). The 2-chloro derivative of Neu5Ac 25 has reasonable stability and is suitable for glycosylation with a number of highly reactive alcohols, such as primary alcohols, phenols, and primary hydroxyl groups of sugars in good yields and reasonable stereoselectivity (78, 79). This easily obtainable sialyl donor is still one of the popular sialyl donors for the synthesis of simple sialosides under Koenigs– Knorr conditions. In comparison, the 2-bromide and 2-fluoride derivatives of Neu5Ac have only limited applications due to the low stability and the tendency in forming nonnatural b-linkage, respectively (59). Nevertheless, the method is not ideal for sialylation of secondary or hindered hydroxyl groups on sugar acceptors, which favor the elimination pathway to give the 2,3-dehydro-Neu5Ac elimination byproduct in significant amount. Moreover, the application of this method is limited by the cost and the toxicity concern of using stoichiometric amount of heavy metal salts as the promoters.

40

H. Cao and X. Chen

AcO

OAc OAc

Cl O

AcHN AcO

CO2Me

+

O

25

O

Ag2CO3

OO

67%, α only

O

26

OAc OAc AcO AcHN AcO

OAc OAc CO2Me O O AcHN O AcO

AcO

OH

O

Cl O

CO 2Me

27

BnO OH Hg(CN)2/HgBr2 O + OBn BnO 84%, α/β = 3/4 OBn

25

O OO

OAc OAc CO2Me O O AcHN BnO AcO

AcO

28

29

BnO

O

OBn OBn

Fig. 9. Sialyl chloride as a donor for sialylation under Koenigs–Knorr glycosylation conditions.

AcO

OAc OAc

SMe O

RHN AcO

30

CO 2Me

AcO

OAc OAc

SEt O

RHN AcO

31

CO 2Me

AcO

OAc OAc

SPh O

RHN AcO

32

AcO CO 2Me

OAc OAc RHN

CO 2Me O

AcO 33 S

S OEt

Fig. 10. Representative 2-thio sialyl donors.

4.2. 2-Thio Sialyl Donors

Using 2-thio derivatives of sialic acid as sialyl donors is a great improvement in chemical sialylation during the last decade. The 2-thio derivatives of Neu5Ac are usually thio-alkyl (methyl 30, ethyl 31) or thio-aryl (phenyl 32, p-methylphenyl, etc.) sialosides, and sialyl xanthates (ethoxydithiocarbonate 33) (Fig. 10) (59, 80). These 2-thio derivatives of Neu5Ac can be readily prepared and are compatible with many protecting group manipulation conditions. The thioglycoside donors can be activated under mild conditions in the presence of stoichiometric amount of a thiophilic promoter such as N-iodosuccinimide/trifluoromethanesulfonic acid (NIS/TfOH). It can also be easily transformed to other glycosyl donors. Due to these advantages, 2-thio derived sialyl donors provide a great platform to accommodate many other protecting group modification strategies for the synthesis of a broad spectrum of sialyloligosaccharides and sialylglycoconjugates. Some representative examples are shown in the following section. As shown in Fig. 11, Boons and Demchenko used N-acetylacetamido (di-N-acetyl) group in a 2-thiomethyl sialyl donor to greatly shorten the reaction time from more than 2 h with monoN-acetyl 2-thiomethyl sialyl donor to a few minutes with increased yield (68). Solvent effect plays an important role in the stereoselective formation of desired natural a-sialosides using thioglycoside donors (59, 81). As shown in Fig. 12, thioglycoside donor (34) forms the oxocarbenium ion intermediate (39) upon activation by a suitable promoter. The acetonitrile as a solvent can attack this oxocarbenium ion intermediate to generate both a- (40) and

3  General Consideration on Sialic Acid Chemistry OAc OAc

AcO

AcO

34

OAc OAc

AcO

Ac2N

CO 2Me + HO

O

AcHN

35 HO

SMe O

AcO

HO

SMe

CO2Me + HO

37

35

OBz O

NIS, TfOH, MS 3A, CH3CN, -40oC

OTE OH

OBz O

AcO

OAc MeO 2C OAc O

AcHN

2-6 h, 61%

AcO NIS, TfOH, MS 3A, CH3CN, -40oC

OTE

5 min, 72%

OH

HO O

41 OBz O

OTE

OH

36

OAc MeO 2C HO OAc O O Ac2N AcO 38

AcO

OBz O

OTE OH

Fig. 11. Sialylation using di-N-acetyl thiosialoside donor.

AcO

OAc OAc

O

AcHN

Solvent CO 2Me participation

AcO

OAc OAc O

AcHN AcO

39

N

C CH 3

CO2Me N

OAc OAc

AcO

AcO

Activation AcO

OAc OAc

SMe O

AcHN AcO

AcO

CO2Me

O

AcHN

AcO 40

N CO2Me

41

ROH

Sialylation OAc CO2Me OAc

AcHN

O

OR

AcO 42

34

Fig. 12. a-Stereoselective sialylation via solvent participation.

b-nitrilium (41) intermediates. The predominance of the kinetically controlled b-nitrilium ion intermediate leads to a-selective sialylation via an SN2 mechanism. Many protecting group modification strategies have been developed in recent years for thiosialoside donors, such as N-trifluoroacetyl (NHTFA) (70), azido (69), N-trichloroethoxycarbonyl (NHTroc) (71), N-t-butoxycarbonylacetamido (NAcBoc) (72), and phthalimido (NPhth) (73) groups at C-5 position of sialyl donors. A higher a-stereoselectivity was observed by Wong and coworkers with a C-5 azido derivative of thiosialoside donor (45) compared to its C-5 N-acetyl derivative (34) (69). As shown in Fig. 13, the desired a-stereoselectivity was dramatically improved from 1:1.25 to 10:1 (a/b) by changing the C-5 acetamido group to C-5 azido under same glycosylation conditions (69). Unfortunately, the higher stereoselectivity controlled by the C-5 azido group of sialyl donor was associated with its deactivating effect leading to low yields. Ando et  al. developed novel 1,5-lactam sialic acid acceptors (47, 51, Fig.  14) with a thiophenylsialoside donor (48) for the synthesis of the carbohydrate components of gangliosides HLG-2 (50) (82) and Hp-s6 (53) (83) oligosaccharides (82, 83). These lactam acceptors render high reactivity at C4- and C8-hydroxyl groups to high yields production of a2-4- and a2-8-linked disialyl intermediates, respectively. Tanaka et al. developed a 5-N,4-O-carbonyl-protecting group for Neu5Ac and found this protecting group is superior for both

42

H. Cao and X. Chen

OAc OAc SMe AcO CO2Me O AcHN

O +

O

AcO

N3 AcO

o

NIS, TfOH, MS 3A, CH3CN, -40 C OTBDPS

45%, α/β = 1:1.25

OAc OAc

AcHN AcO

CO2Me O

N3

34 OAc OAc AcO

OH O

AcO

STol O

CO2Me

45

+

O

44 AcO

OH O N3

O

O

43 O

O O

OTBDPS

NIS, TfOH, MS 3A, CH 3CN, -40o C 53%, α/β = 10:1

43

OTBDPS

N3

OAc MeO C 2 OAc O N3 O AcO O 46

O

O

OTBDPS

N3

Fig. 13. Sialylation using C-5 N-acetyl versus C-5 azido thiosialoside donors.

Fig. 14. 1,5-Lactam sialyl acceptors/donors for the synthesis of gangliosides HLG-2 and Hp-s6.

Fig. 15. 5-N,4-O-Carbonyl protected sialyl donor and acceptor for the synthesis of oligosialic acids.

sialyl donor (54) and acceptor (55) (Fig. 15) in selective formation of a2-8-disialyl sequence in good yields (74). The efficiency of the approach was demonstrated in the synthesis of a2-8-linked oligosialic acid (57) (74), a2-9-linked oligosialic acid (84), and complex ganglioside GP1c (85). 4.3. 2-Phosphite Sialyl Donors

Sialyl phosphite donors (Fig. 16) were independently developed by Wong (58) (86) and Schmidt (61) (87) in 1992. These 2-phosphites derivatives of sialic acid can be synthesized from readily available 2-hydroxyl sialic acid derivatives. They can be activated at low temperature by catalytic amount of many common Lewis acid promoters, such as trimethylsilyl trifluoromethanesulfonate (TMSOTf). Since the sialyl phosphite donors are synthesized

3  General Consideration on Sialic Acid Chemistry

OAc OAc AcO AcHN

OP(OBn)2 O CO2Me

+

BnO BnO

AcO

OAc OAc

OP(OEt) 2 O

AcHN

+

CO2Me

TMSOTf, CH3CN, -42oC

BnO OMe 59

58 (Wong's Donor) AcO

AcO

OH O

BnO BnO

AcO

61 (Schmidt's Donor)

OH O BnO

59

OAc OAc

O BnO BnO

AcO

80%, α/β = 6/1

60 OAc OAc AcO

TMSOTf, CH3CN, -40oC

OMe

CO2Me O

AcHN

70%, α/β = 4/1

AcHN

43

O BnO

OMe

CO2Me O

AcO 60

O BnO BnO

O BnO

OMe

Fig. 16. Sialylation using 2-phosphite sialyl donors.

OAc MeO2C O OAc O O P OBu AcN OBu O O 62

AcO

63, CH2Cl 2, 4A MS, TMSOTf -78oC to -40oC

64, NIS

OAc MeO2C OAc O O AcN HO O O O BzO O 65 BzO BnO (80%, single isomer)

AcO

-40oC

OBn O

O(CH2)5N3 NPhth

Ph OAc MeO2C O OAc O O P OBu AcN OBu O O 62

66, CH2Cl 2, 4A MS, TMSOTf

AcO

HO BzO

-78oC to -40oC

64, NIS

O OAc MeO 2C O OAc O O O O AcN BzO BnO O 67 O (79%, single isomer )

AcO

-40o C Ph

OH O

STol

BzO 63

HO BnO

OBn O

O(CH2)5N 3 NPhth 64

O O HO

O

OBn O

O(CH2)5N3 NPhth

STol

BzO 66

Fig. 17. One-pot glycosylation using a 5-N-acetyl-5-N,4-O-carbonyl 2-phosphate sialyl donor.

predominately as b-anomers and activated at low temperature, the glycosylation usually leads to predominate a-sialoside either through direct SN2 mechanism or solvent participating pathway (Fig. 16) (86, 87). Most recently, Wong and coworkers demonstrated the 2-phosphate derivative of Neu5Ac containing a 5-N-acetyl-5-N,4O-carbonyl protecting group (62) is an ideal sialyl donor for the synthesis of complex sialosides (88). As shown in Fig.  17, the 2-phosphate sialyl donor (62) can be used in a programmable one-pot glycosylation protocol developed in the Wong group to stereoselectively produce a-sialyl products in good yields (88). 4.4. 2-Trifluoroace timidate Sialyl Donors

It was well known that the Schmidt type sialyl donors with a tricholoroacetimidate as the leaving group are not suitable for sialylation (61). Quite interestingly, a trifluoroacetimidate sialyl donor developed by Yu and coworkers (68 in Fig. 18) by substituting the trichloroacetimidate in the Schmidt donor with trifluoroacetimidate has been successfully applied to direct sialylation with impressive yields and a-stereoselectivity (Fig.  18a) (89, 90). Fukase and coworkers

44

H. Cao and X. Chen OAc MeO2C O OAc O O P OBu AcN OBu O O 62

AcO

63, CH2Cl2, 4A MS, TMSOTf -78oC to -40oC

64, NIS

OAc MeO2C OAc O O AcN HO O O O BzO O 65 BzO BnO (80%, single isomer)

AcO

-40oC

OBn O

O(CH2)5N3 NPhth

Ph OAc MeO2C O OAc AcO O O P OBu AcN OBu O O 62 HO BzO

66, CH2Cl2, 4A MS, TMSOTf

64, NIS

-78oC to -40oC

O OAc MeO 2C O OAc O O O O AcN BzO BnO O 67 O (79%, single isomer)

-40oC Ph

OH O

STol

BzO

HO BnO

O O

OBn O

O(CH2)5N 3 NPhth

63

OBn O

AcO

HO

O

O(CH2)5N3 NPhth

STol

BzO 66

64

Fig. 18. Sialylation using trifuloroacetimidate sialyl donors.

a AcO

OAc OAc

OH O

AcHN AcO

CO 2Me

+

BnO BnO

72

b AcO

OAc OAc

AcHN AcO

OH O

O 74

O

NMe2 O

+

OH O

BnOOMe 59

AcO

OH O

O

AcHN AcO

74

O

NMe2 O

66%, α/β = 2:1

OAc OAc

CO2Me O

AcHN AcO 73

26

Ar 2SO, Tf 2O

O

BnO BnO

OAc O OAc AcO O OH O AcHN O (p-NO C H )(Ph)SO, Tf O, CH Cl 2 6 4 2 2 2 O AcO OO 70%, α/β = 4:1

c OAc OAc

(p-NO 2C6H4)(Ph)SO, Tf2O, CH2Cl2

AcO

O BnO OMe

Ar S OTf O Ar AcO O O NMe2 AcHN O AcO O 76

AcO

OAc OAc AcO

O

O

OO

TfO O O

AcHN

NMe2 O O

75

OAc OAc

O

O

77

NHMe2 O O

Fig. 19. Dehydrative sialylation approaches using C-2 hemiketal Neu5Ac derivative as a sialyl donor.

found the “fixed dipole effect” of a C-5 N-phthaloyl protecting group on trifluoroacetimidate donor (69) provided predominant a-sialosides under the similar glycosylation conditions (Fig. 18b) (91). 4.5. C-2 Hydroxyl Derivatives of Neu5Ac as Donor for Dehydrative Sialylation

Recently, Gin and coworkers expanded their dehydrative glycosylation protocol to chemical sialylation (92, 93). As shown in Fig.  19, anomeric hemiketal Neu5Ac derivative with C-1 N,Ndimethylglycolamide ester auxiliary group (74) (Fig. 19b) produced higher a-selectivity compared to its counterpart without the C-1 participating group (72) (Fig.  19a). The higher a-stereoselecitive was contributed by the C-1 N,N-dimethylglycolamide ester auxiliary group as it can attack the anomeric center to form b-oriented bicyclic intermediate (77) due to anomeric effect (Fig. 19c).

3  General Consideration on Sialic Acid Chemistry

5. Enzymatic Sialylation With or Without CMPSia Regeneration

As described earlier, chemical sialylation is a time-consuming process, which involves tedious protecting group manipulation. So far, only a small library of sialosides has been prepared through chemical synthesis. Furthermore, the sialic acid form in most of these sialosides is restricted to Neu5Ac, the most abundant sialic acid form. Sialosides can be effectively synthesized by sialyltransferasecatalyzed enzymatic reactions. In these processes, CMP-sialic acid is used by sialyltransferases as an activated nucleotide donor. Galactose (Gal) or N-acetylgalactosamine (GalNAc)-terminating structures are common sialyltransferase acceptors for producing monosialylated products. Using different sialyltransferases, a2-3-, a2-6-, and a2-8-linked sialosides (Siaa2-3GalbOR, Siaa26GalbOR, and Siaa2-6GalNAcaOR) can be synthesized (Fig. 20a–c). For the enzymatic synthesis of Siaa2-8Sia-containing structures, an a2-3-linked monosialylated compound serves as an acceptor for a2-8-sialyltransferase for the synthesis of a2-8-disialyl

OH HO

OH O

GalβOR

OR

OH

a

HO

b HO

O

AcHN

OH

CO2-

HO

HO

c

CMP-Neu5Ac, 7

OR

OH

d HO

HO

O

O

OH O

HO

Siaα2-6GalβOR HO OR

HO

CO 2-

OH O

AcHN

O

HO

Siaα2-6GalNAcαOR

OH

HO

O

HO

AcHN

OR

OR

α-linked sialoside α2-8SiaT

OR

OH

CO2O

OR

OH CO 2-

HO

OH

AcHN

OH OH O

Siaα2-3GalβOR

OH

AcHN

GalNAcαOR α2-6SiaT OH

HO

OH O

AcHN

O

O

AcHN

α2-6SiaT

OCMP

OH

CO2-

OH HO

OH OH O

GalβOR

HO

HO

α2-3SiaT

OH

45

HO

CO 2-

OH HO AcHN

O HO

O AcHN

OH

Siaα2-8Siaα HO

Fig. 20. Sialyltransferase-catalyzed formation of four common sialyl linkages.

CO 2-

OH O

OR

46

H. Cao and X. Chen OPO32O

CO 2-

(PEP)

CO 2-

PK HO

HO

CO2- OH

OH O

AcHN HO

ATP

O

OH

O

HO

OH NHAc

79 HO HO

CMP

O O HO

78 2Pi

OH O

OPO32-

CDP

α2-6SiaT

OH

OH

NMK

O

O OH HO

ADP

CO 2-

PK

CMP-Neu5Ac, 7

CTP

O CO 2-

OH NHAc

PPase

PPi

CMP-Neu5Ac HO synthetase HO OH

OH O

AcHN HO

CO2-

1

Fig. 21. Sialyltransferase-catalyzed sialylation with CMP-sialic acid regeneration.

products (Fig.  20d). Polysialic acids can also be synthesized by polysialyltransferases using CMP-Neu5Ac as a donor substrate and sialic acid containing acceptors (94). To avoid the use of stoichiometric amount of expensive CMPsialic acid and to minimize product inhibition by CMP formed in the sialyltransferase reaction, Wong and coworkers developed an in situ CMP-Neu5Ac regeneration system (Fig.  21) for sialyltransferase-catalyzed synthesis of sialosides (95). In this system, CMPNeu5Ac can be regenerated from sialyltransferase byproduct CMP via a nucleoside monophophate kinase (NMK) in the presence of ATP, a pyruvate kinase (PK), phosphoenolpyruvate (PEP), a CMP-Neu5Ac synthetase and Neu5Ac. An a2-6-sialyltransferase (a2-6SiaT)-catalyzed reaction successfully produced a2-6-linked sialoside product (79). The pyrophosphate byproduct of the CMP-Neu5Ac synthetase reaction was decomposed to inorganic phosphate by pyrophosphatase (PPase) and the ADP byproduct of the nucleoside monophophate kinase (NMK) was recycled back to ATP by the pyruvate kinase (PK) (Fig. 21) (95). In this system, only Neu5Ac, sialyltransferase acceptor LacNAc (78) and phosphoenolpyruvate (PEP) (2 equivalent), needed to be used in stoichiometric amounts, thus decreased the cost for the synthesis of sialosides. However, the number (five) of required enzymes in the system and the limited availability of mammalian a2-6-sialyltransferase prevented this method for common application in largescale synthesis of sialosides.

3  General Consideration on Sialic Acid Chemistry

6. Chemoenzymatic Sialylation in Generating Diverse Sialosides

47

Recently, Chen and coworkers developed a highly efficient one-pot three-enzyme approach for the synthesis of diverse sialosides with different naturally occurring and nonnatural sialosides, various sialyl linkages, and diverse underlying glycans (56–58, 96). Many of the sialosides produced have a propyl azide aglycone at the reducing end of the glycans, which can be conveniently converted to an amino group for direct coupling to N-hydroxysuccinamide (NHS)or epoxy-activated slides for glycan microarray studies (97–99). Alternatively, the amino group can also be used to conjugate to biotin or other biomolecules (100–103) for bioassays. As shown in Fig.  22, chemical or enzymatic modification of N-acetylmannosamine (ManNAc) or mannose (Man) produced diverse sialic acid precursors, which can be coupled with pyruvate catalyzed by an E. coli (104) or a Pasteurella multocida (105) sialic acid aldolase for the formation of diverse naturally occurring and nonnatural sialic acid forms. These sialic acid forms can then be activated by a bacterial CMP-sialic acid synthetase (104) and transferred by different sialyltransferases to diverse glycans or glycoconjugates terminated with a Gal, GalNAc, or a sialic acid for the formation of a2-3-, a2-6-, or a2-8-linked sialosides (56–58, 106, 107). Efficient large-scale synthesis of diverse sialoside libraries has been successfully achieved due to the availability of highly active promiscuous bacterial sialoside biosynthetic enzymes in large amounts. The chemoenzymatic methods have been successfully

Fig. 22. One-pot three-enzyme system for chemoenzymatic synthesis of sialoside libraries.

48

H. Cao and X. Chen

applied for the synthesis of sialosides with naturally occurring labile sialic acid modification such as O-acetylation and O-lactylation in addition to O-methylation and nonnatural modification at different positions of sialic acid, most of which have not been achieved by chemical or pure enzymatic synthesis. Using the one-pot three-enzyme chemoenzymatic method described earlier, a library of a2-3-linked sialosides containing naturally occurring or nonnatural sialic acid modifications were prepared in preparative scale. Three enzymes used were E. coli sialic acid aldolase, N. meningitidis CMP-sialic acid synthetase (NmCSS), and a multifunctional a2-3-sialyltransferase from P. multocida (PmST1) (56). Some representative a2-3-linked sialosides synthesized are listed in Fig. 23 (56, 106–108). Following a similar approach, a library of a2-6-linked sialosides were also prepared in preparative scales by substituting the sialyltransferase from PmST1 to a multifunctional Photobacterium damsela a2-6-sialyltransfearse (Pd2,6ST) (57, 109, 110). Some representative a2-6-linked sialosides synthesized are listed in Fig. 24 (57, 106–109).

Fig. 23. Representative a2-3-linked sialosides prepared using the one-pot multienzyme chemoenzymatic approach described in Fig. 22.

3  General Consideration on Sialic Acid Chemistry

49

Fig. 24. Representative a2-6-linked sialosides prepared using the one-pot multienzyme chemoenzymatic approach described in Fig. 22.

The one-pot three-enzyme approach was also used in sequential for a two-step one-pot multienzyme synthesis of complex GD3 oligosaccharides. In the first step, a2-3-linked sialosides were synthesized using a one-pot three-enzyme system containing PmST1 as the sialyltransferase. In the second step, a2-8-linked sialosides were synthesized from a2-3-linked sialosides as acceptors using a one-pot three-enzyme system containing a multifunctional C. jejuni a2-8-sialyltransferase (CstII). These disialyl glycans are key components for developing ganglioside-based carbohydrate vaccines (58). A modified multiple enzyme approach was also developed by the Chen group for the synthesis of CMP-sialic acids and sialosides containing a 3-fluoro-sialic acid residue (111). These fluorinated derivatives are invaluable mechanistic probes for protein crystal structural studies of sialic acid processing enzymes (Fig. 25).

50

H. Cao and X. Chen

Fig. 25. Enzymatic synthesis of fluorinated sialic acid derivatives as mechanistic probes for sialic acid processing enzymes.

7. Conditions for Preventing De-O-Acetylation or O-Acetyl Migration

O-Acetylation is the most common postglycosylational sialic acid modification (3, 55). This modification has great impact on sialic acid-dependent recognition processes. For example, 9-O-acetylation of the sialic acids on host cell surface is necessary for influenza C virus binding and subsequent invasion (112, 113) but prevents the attachment of malaria parasites (114) and influenza A and B viruses (115, 116). In the biosynthetic pathway, the acetyl groups are usually introduced to the C-7 or C-9 hydroxyl groups of sialic acid moiety after the formation of sialic linkages. Using the chemoenzymatic method described earlier, 9-O-acetylated sialic acids can be either synthesized from sialic acid by chemical selective acetylation (117) or from 6-O-acetylated sialic acid precursors (6-O-acetylated ManNAc, ManNGc, or Man) followed by enzymatic aldol-addition reaction (96, 118). Other O-acetylated sialic acids are generally obtained from the hydrolysis products of naturally occurring O-acetylated sialic acid containing structures by treated with sialidases (119).

3  General Consideration on Sialic Acid Chemistry

51

The acetyl groups are labile under both acidic and basic conditions. For example, the C-7 O-acetyl groups can spontaneously migrate to the C-8 and C-9 hydroxyl groups under physiological conditions (3, 119). The deacetylated sialic acids or its containing structures are the common byproducts generated during the synthesis or purification processes. To minimize de-O-acetylation or O-acetyl migration during the enzymatic synthesis, a suitable buffer is generally employed to maintain the reaction mediate at near neutral conditions (pH 7.5), and excess amount of enzymes are added to shorten the reaction time (57, 58, 96). A buffer in a pH range of 3.0–6.0 is frequently used to prevent de-O-acetylation or O-acetyl migration during the process of purification and analysis (120). To minimize de-O-acetylation or O-acetyl migration during storage, purified samples should be dried and kept at −20°C or −80°C.

8. Prospective and Conclusion The emerging glycan microarray platform provides a high-throughput approach for studying structure–activity relationship (SAR) of sialic acid containing structures. The synthesis of structurally defined diverse sialosides is a bottleneck step for these studies. Recently developed chemical and enzymatic synthetic approaches allow the access to some sialosides. Taking advantages of high expression level in E. coli expression system, high activity, and substrate promiscuity of bacterial sialoside biosynthetic enzymes, onepot multi-enzyme chemoenzymatic approach opens up a new avenue to quickly access many natural occurring and nonnatural sialosides in sufficient amount for diverse glycan microarray-based SAR studies.

Acknowledgments We are grateful for financial supports from Shandong University (to H.C.), the National Science Foundation of China (No. 20902087 to H.C.), the Natural Science Foundation of Shandong Province (SDNSF, No. ZR2010BM018 to H.C.), the University of CaliforniaDavis (to X.C.), the National Institutes of Health (R01GM076360 and U01CA128442 to X.C.), the National Science Foundation (CAREER Award 0548235 to X.C.), Alfred P. Sloan Foundation (to X.C.), and the Camille & Henry Dreyfus Foundation (to X.C.). X.C. is an Alfred P. Sloan Research Fellow, a Camille Dreyfus Teacher-Scholar, and a UC-Davis Chancellor’s Fellow.

52

H. Cao and X. Chen

References 1. Chen, X., and Varki, A. (2010) Advances in the biology and chemistry of sialic acids. ACS Chem. Biol. 5, 163–176. 2. Schauer, R. (2000) Achievements and challenges of sialic acid research. Glycoconj. J. 17, 485–499. 3. Angata, T., and Varki, A. (2002) Chemical diversity in the sialic acids and related alphaketo acids: An evolutionary perspective. Chem. Rev. 102, 439–469. 4. Kiefel, M. J., and von Itzstein, M. (2002) Recent advances in the synthesis of sialic acid derivatives and sialylmimetics as biological probes. Chem. Rev. 102, 471–490. 5. Bhattacharjee, A. K., Jennings, H. J., Kenny, C. P., Martin, A., and Smith, I. C. (1975) Structural determination of the sialic acid polysaccharide antigens of Neisseria meningitidis serogroups B and C with carbon 13 nuclear magnetic resonance. J. Biol. Chem. 250, 1926–1932. 6. Paoletti, L. C., and Kasper, D. L. (2003) Glycoconjugate vaccines to prevent group B streptococcal infections. Expert Opin. Biol. Ther. 3, 975–984. 7. Lewis, A. L., Cao, H., Patel, S. K., Diaz, S., Ryan, W., Carlin, A. F., Thon, V., Lewis, W. G., Varki, A., Chen, X., and Nizet, V. (2007) NeuA sialic acid O-acetylesterase activity modulates O-acetylation of capsular polysaccharide in group B Streptococcus. J. Biol. Chem. 282, 27562–27571. 8. Moran, A. P., Prendergast, M. M., and Appelmelk, B. J. (1996) Molecular mimicry of host structures by bacterial lipopolysaccharides and its contribution to disease. FEMS Immunol. Med. Microbiol. 16, 105–115. 9. Bhattacharjee, A. K., Jennings, H. J., Kenny, C. P., Martin, A., and Smith, I. C. (1976) Structural determination of the polysaccharide antigens of Neisseria meningitidis serogroups Y, W-135, and BO1. Can. J. Biochem. 54, 1–8. 10. Crocker, P. R. (2002) Siglecs: sialic-acid-binding immunoglobulin-like lectins in cell-cell interactions and signaling. Curr. Opin. Struct. Biol. 12, 609–615. 11. Bruses, J. L., and Rutishauser, U. (2001) Roles, regulation, and mechanism of polysialic acid function during neural development. Biochimie 83, 635–643. 12. Simanek, E. E., McGarvey, G. J., Jablonowski, J. A., and Wong, C. H. (1998) Selectincarbohydrate interactions: From natural

ligands to designed mimics. Chem. Rev. 98, 833–862. 13. Dell, A., Morris, H. R., Easton, R. L., Patankar, M., and Clark, G. F. (1999) The glycobiology of gametes and fertilization. Biochim. Biophys. Acta 1473, 196–205. 14. Choung U., Kim, W. L., Williams, M. A., Liu, H., Zhang, L., Swaminathan, S., Bischofberger, N., Chen, M. S., Mendel, D. B., Tai, C. Y., Laver, W. G., and Stevens, R. C. (1997) Influenza neuraminidase inhibitors possessing a novel hydrophobic interaction in the enzyme active site: design, synthesis, and structural analysis of carbocyclic sialic acid analogues with potent anti-influenza activity. J. Am. Chem. Soc. 119, 681–690. 15. Eckhardt, M., Bukalo, O., Chazal, G., Wang, L., Goridis, C., Schachner, M., GerardySchahn, R., Cremer, H., and Dityatev, A. (2000) Mice deficient in the polysialyltransferase ST8SiaIV/PST-1 allow discrimination of the roles of neural cell adhesion molecule protein and polysialic acid in neural development and synaptic plasticity. J. Neurosci. 20, 5234–5244. 16. Daniel, L., Trouillas, J., Renaud, W., Chevallier, P., Gouvernet, J., Rougon, G., and Figarella-Branger, D. (2000) Polysialylatedneural cell adhesion molecule expression in rat pituitary transplantable tumors (spontaneous mammotropic transplantable tumor in WistarFurth rats) is related to growth rate and malignancy. Cancer Res. 60, 80–85. 17. Takano, R., Muchmore, E., and Dennis, J. W. (1994) Sialylation and malignant potential in tumour cell glycosylation mutants. Glycobiology 4, 665–674. 18. Malykh, Y. N., Schauer, R., and Shaw, L. (2001) N-Glycolylneuraminic acid in human tumours. Biochimie 83, 623–634. 19. Kelm, S., and Schauer, R. (1997) Sialic acids in molecular and cellular interactions. Int. Rev. Cytol. 175, 137–240. 20. Munday, J., Floyd, H., and Crocker, P. R. (1999) Sialic acid binding receptors (siglecs) expressed by macrophages, J. Leukoc. Biol. 66, 705–711. 21. Varki, A. (1993) Biological roles of oligosaccharides: all of the theories are correct, Glycobiology 3, 97–130. 22. Reuter, G., and Gabius, H. J. (1996) Sialic acids structure-analysis-metabolism-occurrence-recognition. Biol. Chem. Hoppe Seyler. 377, 325–342.

3  General Consideration on Sialic Acid Chemistry 23. Schwarzkopf, M., Knobeloch, K. P., Rohde, E., Hinderlich, S., Wiechens, N., Lucka, L., Horak, I., Reutter, W., and Horstkorte, R. (2002) Sialylation is essential for early development in mice. Proc. Natl. Acad. Sci. USA 99, 5267–5270. 24. Jann, B., and Jann, K. (1990) Structure and biosynthesis of the capsular antigens of Escherichia coli. Curr. Top. Microbiol. Immunol. 150, 19–42. 25. Mandrell, R. E., and Apicella, M. A. (1993) Lipo-oligosaccharides (LOS) of mucosal pathogens: molecular mimicry and host-modification of LOS, Immunobiology 187, 382–402. 26. Jennings, H. J. (1990) Capsular polysaccharides as vaccine candidates, Curr. Top. Microbiol. Immunol. 150, 97–127. 27. Vogel, U., Claus, H., Heinze, G., and Frosch, M. (1999) Role of lipopolysaccharide sialylation in serum resistance of serogroup B and C meningococcal disease isolates. Infect. Immun. 67, 954–957. 28. Masson, L., and Holbein, B. E. (1983) Physiology of sialic acid capsular polysaccharide synthesis in serogroup B Neisseria meningitidis, J. Bacteriol. 154, 728–736. 29. Barry, G. T. (1959) Detection of sialic acid in various Escherichia coli strains and in other species of bacteria. Nature 183, 117–118. 30. Frosch, M., Weisgerber, C., and Meyer, T. F. (1989) Molecular characterization and expression in Escherichia coli of the gene complex encoding the polysaccharide capsule of Neisseria meningitidis group B. Proc. Natl. Acad. Sci. USA 86, 1669–1673. 31. Ganguli, S., Zapata, G., Wallis, T., Reid, C., Boulnois, G., Vann, W. F., and Roberts, I. S. (1994) Molecular cloning and analysis of genes for sialic acid synthesis in Neisseria meningitidis group B and purification of the meningococcal CMP-NeuNAc synthetase enzyme. J. Bacteriol. 176, 4583–4589. 32. Muhlenhoff, M., Eckhardt, M., and GerardySchahn, R. (1998) Polysialic acid: threedimensional structure, biosynthesis and function. Curr. Opin. Struct. Biol. 8, 558–564. 33. Adlam C, K. J., Mugridge A, Williams JM, Lindon JC. (1987) Production of colominic acid by Pasteurella haemolytica serotype A2 organisms. FEBS Microbiol. lett. 42, 23–25. 34. Singhal, A., and Hakomori, S. (1990) Molecular-changes in carbohydrate antigens associated with cancer. Bioessays 12, 223–230. 35. Hakomori, S. (1996) Tumor malignancy defined by aberrant glycosylation and

53

sphingo(glyco)lipid metabolism. Cancer Res. 56, 5309–5318. 36. Kannagi, R., Izawa, M., Koike, T., Miyazaki, K., and Kimura, N. (2004) Carbohydratemediated cell adhesion in cancer metastasis and angiogenesis. Cancer Sci. 95, 377–384. 37. Varki, A. (2008) Sialic acids in human health and disease. Trends Mol. Med. 14, 351–360. 38. Astronomo, R. D., and Burton, D. R. (2010) Carbohydrate vaccines: developing sweet solutions to sticky situations? Nature Rev. Drug Discov. 9, 308–324. 39. Galonic, D. P., and Gin, D. Y. (2007) Chemical glycosylation in the synthesis of glycoconjugate antitumour vaccines. Nature 446, 1000–1007. 40. Hecht, M. L., Stallforth, P., Silva, D. V., Adibekian, A., and Seeberger, P. H. (2009) Recent advances in carbohydrate-based vaccines. Curr. Opin. Chem. Biol. 13, 354–359. 41. Boltje, T. J., Buskas, T., and Boons, G. J. (2009) Opportunities and challenges in synthetic oligosaccharide and glycoconjugate research. Nat. Chem. 1, 611–622. 42. Corfield, A. R. S. (1982) Sialic acids: chemistry, metabolism and function. Schauer, R. ed. Springer-Verlag, New York 10, 195–261. 43. Vann, W. F., Tavarez, J. J., Crowley, J., Vimr, E., and Silver, R. P. (1997) Purification and characterization of the Escherichia coli K1 neuB gene product N-acetylneuraminic acid synthetase. Glycobiology 7, 697–701. 44. Comb, D. G., and Roseman, S. (1960) The sialic acids. I. The structure and enzymatic synthesis of N-acetylneuraminic acid. J. Biol. Chem. 235, 2529–2537. 45. Rodriguez-Aparicio, L. B., Ferrero, M. A., and Reglero, A. (1995) N-Acetyl-Dneuraminic acid synthesis in Escherichia coli K1 occurs through condensation of N-acetylD-mannosamine and pyruvate. Biochem. J. 308 ( Pt 2), 501–505. 46. Warren, L., and Blacklow, R. S. (1962) Biosynthesis of N-acetyl-neuraminic acid and cytidine-5’-monophospho-N-acetylneuraminic acid in Neisseria meningitidis. Biochem. Biophys. Res. Commun. 7, 433–438. 47. Hinderlich, S., Stasche, R., Zeitler, R., and Reutter, W. (1997) A bifunctional enzyme catalyzes the first two steps in N-acetylneuraminic acid biosynthesis of rat liver. Purification and characterization of UDP-N-acetylglucosamine 2-epimerase/Nacetylmannosamine kinase. J. Biol. Chem. 272, 24313–24318. 48. Angata, T., Nakata, D., Matsuda, T., Kitajima, K., and Troy, F. A., 2nd. (1999) Biosynthesis

54

H. Cao and X. Chen

of KDN (2-keto-3-deoxy-D-glycero-Dgalacto-nononic acid). Identification and characterization of a KDN-9-phosphate synthetase activity from trout testis. J. Biol. Chem. 274, 22949–22956. 49. Kean, E. L. (1991) Sialic acid activation. Glycobiology 1, 441–447. 50. Angata, T., Kitazume, S., Terada, T., Kitajima, K., Inoue, S., Troy, F. A., 2nd, and Inoue, Y. (1994) Identification, characterization, and developmental expression of a novel alpha 2 → 8-KDN-transferase which terminates elongation of alpha 2 → 8-linked oligo-polysialic acid chain synthesis in trout egg polysialoglycoproteins. Glycoconj. J. 11, 493–499. 51. Terada, T., Kitazume, S., Kitajima, K., Inoue, S., Ito, F., Troy, F. A., and Inoue, Y. (1993) Synthesis of CMP-deaminoneuraminic acid (CMP-KDN) using the CTP:CMP-3deoxynonulosonate cytidylyltransferase from rainbow trout testis. Identification and characterization of a CMP-KDN synthetase. J. Biol. Chem. 268, 2640–2648. 52. Kawano, T., Koyama, S., Takematsu, H., Kozutsumi, Y., Kawasaki, H., Kawashima, S., Kawasaki, T., and Suzuki, A. (1995) Molecular cloning of cytidine monophospho-N-acetylneuraminic acid hydroxylase. Regulation of species- and tissue-specific expression of N-glycolylneuraminic acid. J. Biol. Chem. 270, 16458–16463. 53. Hirschberg, C. B., Goodman, S. R., and Green, C. (1976) Sialic acid uptake by fibroblasts. Biochemistry 15, 3591–3599. 54. Oetke, C., Hinderlich, S., Brossmer, R., Reutter, W., Pawlita, M., and Keppler, O. T. (2001) Evidence for efficient uptake and incorporation of sialic acid by eukaryotic cells. Eur. J. Biochem. 268, 4553–4561. 55. Yu, H., and Chen, X. (2007) Carbohydrate post-glycosylational modifications. Org. Biomol. Chem. 5, 865–872. 56. Yu, H., Chokhawala, H., Karpel, R., Wu, B. Y., Zhang, J. B., Zhang, Y. X., Jia, Q., and Chen, X. (2005) A multifunctional Pasteurella multocida sialyltransferase: A powerful tool for the synthesis of sialoside libraries. J. Am. Chem. Soc. 127, 17618–17619. 57. Yu, H., Huang, S. S., Chokhawala, H., Sun, M. C., Zheng, H. J., and Chen, X. (2006) Highly efficient chemoenzymatic synthesis of naturally occurring and non-natural alpha2,6-linked sialosides: A P. damsela alpha-2,6sialyltransferase with extremely flexible donor-substrate specificity. Angew. Chem. Int. Ed. 45, 3938–3944. 58. Yu, H., Cheng, J. S., Ding, L., Khedri, Z., Chen, Y., Chin, S., Lau, K., Tiwari, V. K., and

Chen, X. (2009) Chemoenzymatic synthesis of GD3 oligosaccharides and other disialyl glycans containing natural and non-natural sialic acids. J. Am. Chem. Soc. 131, 18467–18477. 59. Boons, G. J., and Demchenko, A. V. (2000) Recent advances in O-sialylation. Chem. Rev. 100, 4539–4565. 60. Halcomb, R. L., and Chappell, M. D. (2002) Recent developments in technology for glycosylation with sialic acid, J. Carbohydr. Chem. 21, 723–768. 61. Schmidt, R. R., and Kinzy, W. (1994) Anomeric-oxygen activation for glycoside synthesis-the trichloroacetimidate method. Adv. Carbohydr. Chem. Biochem. 50, 21–123. 62. Haberman, J. M., and Gin, D. Y. (2001) A new C(1)-auxiliary for anomeric stereocontrol in the synthesis of alpha-sialyl glycosides. Org. Lett. 3, 1665–1668. 63. Ye, X. S., Huang, X. F., and Wong, C. H. (2001) Conversion of the carboxy group of sialic acid donors to a protected hydroxymethyl group yields an efficient reagent for the synthesis of the unnatural beta-linkage. Chem. Commun., 974–975. 64. Okamoto, K., Kondo, T., and Goto, T. (1986) An effective synthesis of alpha-glycosides of N-acetylneuraminic acid by use of 2-betahalo-3-beta-hydroxyl-4,7,8,9-tetra-O-acetylneuraminic acid methyl-ester. Tetrahedron Lett. 27, 5233–5236. 65. Ito, Y., and Ogawa, T. (1987) An efficient approach to stereoselective glycosylation of N-acetylneuraminic acid-Use of phenylselenyl group as a stereocontrolling auxilliary. Tetrahedron Lett. 28, 6221–6224. 66. Martichonok, V., and Whitesides, G. M. (1996) A practical method for the synthesis of sialyl alpha-glycosides. J. Am. Chem. Soc. 118, 8187–8191. 67. CastroPalomino, J. C., Tsvetkov, Y. E., Schneider, R., and Schmidt, R. R. (1997) 8-O-Sialylation of neuraminic acid acceptor reactivity and anomeric stereocontrol. Tetrahedron Lett. 38, 6837–6840. 68. Demchenko, A. V., and Boons, G. J. (1999) A novel direct glycosylation approach for the synthesis of dimers of N-acetylneuraminic acid. Chemistry-Eur. J. 5, 1278–1283. 69. Yu, C. S., Niikura, K., Lin, C. C., and Wong, C. H. (2001) The thioglycoside and glycosyl phosphite of 5-azido sialic acid: Excellent donors for the alpha-glycosylation of primary hydroxy groups. Angew. Chem. Int. Ed. 40, 2900–2903. 70. De Meo, C., Demchenko, A. V., and Boons, G. J. (2001) A stereoselective approach for the synthesis of alpha-sialosides. J. Org. Chem. 66, 5490–5497.

3  General Consideration on Sialic Acid Chemistry 71. Ren, C. T., Chen, C. S., and Wu, S. H. (2002) Synthesis of a sialic acid dimer derivative, 2¢alpha-O-benzyl Neu5Ac-alpha-(2 → 5)Ne u5Gc. J. Org. Chem. 67, 1376–1379. 72. Sherman, A. A., Yudina, O. N., Shashkov, A. S., Menshov, V. M., and Nifant’ev, N. E. (2001) Synthesis of Neu5Ac- and Neu5Gcalpha-(2 → 6’)-lactosamine 3-aminopropyl glycosides. Carbohydr. Res. 330, 445–458. 73. Fujita, S., Numata, M., Sugimoto, M., Tomita, K., and Ogawa, T. (1994) Total synthesis of a modified ganglioside, de-N-acetyl GM(2). Carbohydr. Res. 263, 181–196. 74. Tanaka, H., Nishiura, Y., and Takahashi, T. (2006) Stereoselective synthesis of oligo-alpha-(2,8)-sialic acids. J. Am. Chem. Soc. 128, 7124–7125. 75. Crich, D., and Wu, B. L. (2008) Imposing the trans/gauche conformation on a sialic acid donor with a 5-N,7-O-oxazinanone group: effect on glycosylation stereo selectivity. Tetrahedron 64, 2042–2047. 76. De Meo, C., and Priyadarshani, U. (2008) C-5 modifications in N-acetyl-neuraminic acid: scope and limitations. Carbohydr. Res. 343, 1540–1552. 77. Meindl, P., and Tuppy, H. (1965) Uber synthetische ketoside der N-acetyl-Dneuraminsaure. 1. Darstellung einer reihe durch neuraminidase spaltbarer ketoside. Monatsh. Chem. 96, 802–815. 78. Paulsen, H., and Tietz, H. (1982) Oligosaccharide building-blocks. 43. Synthesis oftrisaccharidemoietiesfrom N-acetylneuraminic acid and N-acetyllactosamine. Angew. Chem. Int. Ed. 21, 927–928. 79. Ogawa, T., and Sugimoto, M. (1985) Synthetic studies on cell-surface glycans. 31. Synthesis of alpha-Neu5Acp-(2-)3)-D-Gal and alpha-Neu5Acp-(2-)3)-beta-D-Galp-(1-) 4)-D-Glc, Carbohydr. Res. 135, C5–C9. 80. Ress, D. K., and Linhardt, R. J. (2004) Sialic acid donors: Chemical synthesis and glycosylation. Curr. Org. Syn. 1, 31–46. 81. Schmidt, R. R., and Rucker, E. (1980) Stereoselective glycosidations of uronic-acids. Tetrahedron Lett. 21, 1421–1424. 82. Iwayama, Y., Ando, H., Ishida, H., and Kiso, M. (2009) A first total synthesis of ganglioside HLG-2. Chem-Eur. J. 15, 4637–4648. 83. Ando, H., Koike, Y., Koizumi, S., Ishida, H., and Kiso, M. (2005) 1,5-Lactamized sialyl acceptors for various disialoside syntheses: Novel method for the synthesis of glycan portions of Hp-S6 and HLG-2 gangliosides. Angew. Chem. Int. Ed. 44, 6759–6763. 84. Tanaka, H., Nishiura, Y., and Takahashi, T. (2009) Stereoselective synthesis of alpha(2,9)

55

di- to tetrasialic acids, using a 5,4-N,Ocarbonyl protected thiosialoside. J. Org. Chem. 74, 4383–4386. 85. Tanaka, H., Nishiura, Y., and Takahashi, T. (2008) An efficient convergent synthesis of GP1c ganglioside epitope. J. Am. Chem. Soc. 130, 1724–7125. 86. Kondo, H., Ichikawa, Y., and Wong, C. H. (1992) Beta-sialyl phosphite and phosphoramidite-synthesis of CMP-sialic acid and sialyl oligosaccharides. J. Am. Chem. Soc. 114, 8748–8750. 87. Martin, T. J., and Schmidt, R. R. (1992) Eficient sialylation with phophite as leaving group. Tetrahedron Lett. 33, 6123–6126. 88. Hsu, C. H., Chu, K. C., Lin, Y. S., Han, J. L., Peng, Y. S., Ren, C. T., Wu, C. Y., and Wong, C. H. Highly alpha-selective sialyl phosphate donors for efficient preparation of natural sialosides. Chemistry-Eur. J. 16, 1754–1760. 89. Cai, S. T., and Yu, B. (2003) Efficient sialylation with phenyltrifluoroacetimidates as leaving groups. Org. Lett. 5, 3827–3830. 90. Yu, B. A., and Sun, J. S. (2010) Glycosylation with glycosyl N-phenyltrifluoroacetimidates (PTFAI) and a perspective of the future development of new glycosylation methods. Chem. Commun. 46, 4668–4679. 91. Tanaka, K., Goi, T., and Fukase, K. (2005) Highly efficient sialylation towards alpha(2– 3)- and alpha(2–6)-Neu5Ac-Gal synthesis: Significant ‘fixed dipole effect’ of N-phthalyl group on alpha-selectivity. Synlett, 2958–2962. 92. Garcia, B. A., Poole, J. L., and Gin, D. Y. (1997) Direct glycosylations with 1-hydroxy glycosyl donors using trifluoromethanesulfonic anhydride and diphenyl sulfoxide. J. Am. Chem. Soc. 119, 7597–7598. 93. Haberman, J. M., and Gin, D. Y. (2003) Dehydrative sialylation with C2-hemiketal sialyl donors. Org. Lett. 5, 2539–2541. 94. Ferrero, M. A., and Aparicio, L. R. Biosynthesis and production of polysialic acids in bacteria. Appl. Microbiol. Biotechnol. 86, 1621–1635. 95. Ichikawa, Y., Liu, J. L. C., Shen, G. J., and Wong, C. H. (1991) A highly efficient multienzyme system for the one-step synthesis of a sialyl trisaccharide-In situ generation of sialic acid and N-acetyllactosamine coupled with regeneration of UDP-glucose, UDP-galactose, and CMP-sialic acid. J. Am. Chem. Soc. 113, 6300–6302. 96. Yu, H., Chokhawala, H. A., Huang, S. S., and Chen, X. (2006) One-pot three-enzyme chemoenzymatic approach to the synthesis of sialosides containing natural and non-natural functionalities. Nat. Protoc. 1, 2485–2492.

56

H. Cao and X. Chen

97. Oyelaran, O., and Gildersleeve, J. C. (2009) Glycan arrays: recent advances and future challenges. Curr. Opin. Chem. Biol. 13, 406–413. 98. Liu, Y., Palma, A. S., and Feizi, T. (2009) Carbohydrate microarrays: key developments in glycobiology. Biol. Chem. 390, 647–656. 99. Hsu, K. L., and Mahal, L. K. (2009) Sweet tasting chips: microarray-based analysis of glycans. Curr. Opin. Chem. Biol. 13, 427–432. 100. Yu, H., Chokhawala, H. A., Varki, A., and Chen, X. (2007) Efficient chemoenzymatic synthesis of biotinylated human serum albuminsialoglycoside conjugates containing O-acetylated sialic acids. Org. Biomol. Chem. 5, 2458–2463. 101. Chokhawala, H. A., Huang, S. S., Lau, K., Yu, H., Cheng, J. S., Thon, V., Hurtado-Ziola, N., Guerrero, J. A., Varki, A., and Chen, X. (2008) Combinatorial chemoenzymatic synthesis and high-through put screening of sialosides. ACS Chem. Biol. 3, 567–576. 102. Linman, M. J., Taylor, J. D., Yu, H., Chen, X., and Cheng, Q. (2008) Surface plasmon resonance study of protein-carbohydrate interactions using biotinylated sialosides. Anal. Chem. 80, 4007–4013. 103. Linman, M. J., Yu, H., Chen, X., and Cheng, Q. (2009) Fabrication and characterization of a sialoside-based carbohydrate microarray biointerface for protein binding analysis with surface plasmon resonance imaging. ACS Appl. Mater. Interfaces 1, 1755–1762. 104. Yu, H., Karpel, R., and Chen, X. (2004) Chemoenzymatic synthesis of CMP-sialic acid derivatives by a one-pot two-enzyme system: comparison of substrate flexibility of three microbial CMP-sialic acid synthetases. Bioorg. Med. Chem. 12, 6427–6435. 105. Li, Y. H., Yu, H., Cao, H. Z., Lau, K., Muthana, S., Tiwari, V. K., Son, B., and Chen, X. (2008) Pasteurella multocida sialic acid aldolase: a promising biocatalyst. Appl. Microbiol. Biotechnol. 79, 963–970. 106. Chokhawala, H. A., Yu, H., and Chen, X. (2007) High-throughput substrate specificity studies of sialidases by using chemoenzymatically synthesized sialoside libraries, Chembiochem 8, 194–201. 107. Cao, H. Z., Li, Y. H., Lau, K., Muthana, S., Yu, H., Cheng, J. S., Chokhawala, H. A., Sugiarto, G., Zhang, L., and Chen, X. (2009) Sialidase substrate specificity studies using chemoenzymatically synthesized sialosides containing C5-modified sialic acids. Org. Biomol. Chem. 7, 5137–5145. 108. Cao, H. Z., Muthana, S., Li, Y. H., Cheng, J. S., and Chen, X. (2009) Parallel chemoenzymatic synthesis of sialosides containing a

C5-diversified sialic acid. Bioorg. Med. Chem. Lett. 19, 5869–5871. 109. Cheng, J. S., Huang, S. S., Yu, H., Li, Y. H., Lau, K., and Chen, X. Trans-sialidase activity of Photobacterium damsela alpha2,6-sialyltransferase and its application in the synthesis of sialosides. Glycobiology 20, 260–268. 110. Sun, M. C., Li, Y. H., Chokhawala, H. A., Henning, R., and Chen, X. (2008) N-Terminal 112 amino acid residues are not required for the sialyltransferase activity of Photobacterium damsela alpha2,6-sialyltransferase. Biotechnol. Lett. 30, 671–676. 111. Chokhawala, H. A., Cao, H. Z., Yu, H., and Chen, X. (2007) Enzymatic synthesis of fluorinated mechanistic probes for sialidases and sialyltransferases. J. Am. Chem. Soc. 129, 10630–10631. 112. Rogers, G. N., Herrler, G., Paulson, J. C., and Klenk, H. D. (1986) Influenza C virus uses 9-O-acetyl-N-acetylneuraminic acid as a high affinity receptor determinant for attachment to cells. J. Biol. Chem. 261, 5947–5951. 113. Herrler, G., Rott, R., Klenk, H. D., Muller, H. P., Shukla, A. K., and Schauer, R. (1985) The receptor-destroying enzyme of influenza C virus is neuraminate-O-acetylesterase. Embo J. 4, 1503–1506. 114. Schauer R, S. H., Pommerencke J, Iwersen M, Kohla G. (2001) Metaolism and role of O-acetylated sialic acids. In Molecular immunology of complex carbohydrates 2, edited by Wu AM, Plenum, New York, 2001., 325–342. 115. Gottschalk, A. (1957) Neuraminidase: the specific enzyme of influenza virus and Vibrio cholerae. Biochim. Biophys. Acta 23, 645–646. 116. Klenk, E., Faillard, H., and Lempfrid, H. (1955) Enzymatic effect of the influenza virus. Hoppe Seylers Z Physiol. Chem. 301, 235–246. 117. Ogura, H., Furuhata, K., Sato, S., Anazawa, K., Itoh, M., and Shitori, Y. (1987) Studies on sialic acids. 6. Synthesis of 9-O-acetyl-sialic acid and 4-O-acetyl-sialic acid. Carbohydr. Res. 167, 77–86. 118. Liu, J. L. C., Shen, G. J., Ichikawa, Y., Rutan, J. F., Zapata, G., Vann, W. F., and Wong, C. H. (1992) Overexproduction of CMO-sialic acid synthesis. J. Am. Chem. Soc. 114, 3901–3910. 119. Lewis, A. L., Nizet, V., and Varki, A. (2004) Discovery and characterization of sialic acid O-acetylation in group B Streptococcus. Proc. Natl. Acad. Sci. USA 101, 11123–11128. 120. Varki, A., and Diaz, S. (1984) The release and purification of sialic acids from glycoconjugates: methods to minimize the loss and migration of O-acetyl groups. Anal. Biochem. 137, 236–247.

Chapter 4 Synthesis of Azido-Functionalized Carbohydrates for the Design of Glycoconjugates Samy Cecioni, David Goyard, Jean-Pierre Praly, and Sébastien Vidal Abstract As carbohydrates play a major role in numerous biological processes through their interactions with lectins and also appear as one of the most crucial post-translational modifications of proteins, chemists have developed several approaches for the design of glycoconjugates based on a series of conjugation methodologies. The recent development of copper(I)-catalyzed azide-alkyne cycloaddition (CuACC) paved the way to a novel conjugation strategy in which azido-functionalized carbohydrate derivatives can be readily connected to alkyne-functionalized (bio)molecules. This so-called “click chemistry” methodology has now found numerous applications both in chemistry and biology. The azido moiety can be introduced either directly at the anomeric carbon of the carbohydrate derivative, or attached to a spacer arm. We describe here the syntheses of 2,3,4,6-tetra-O-acetyl-b-d-glucopyranosyl azide as well as 1-azido-3,6-dioxaoct-8-yl 2,3,4, 6-tetra-O-acetyl-b-d-galactopyranoside and 1-azido-3,6-dioxaoct-8-yl 2,3,6,2¢,3¢,4¢,6¢-hepta-O-acetyl-bd-lactoside. These molecules can then be used in the construction of glycoconjugates to find applications in chemical biology. Key words: Carbohydrates, Click chemistry, 1,3-Dipolar cycloaddition, Azide, Alkyne, Glycoconju­ gates, Glycosylation, Triazole

1. Introduction Biological processes occur mainly through receptor–ligand interactions or chemical reactions catalyzed by enzymes. Among these processes, lectin–carbohydrate interactions are involved in both normal and pathological phenomena such as cell–cell communication, viral infection, cancer metastasis or even fecundation (1, 2). The design of high affinity ligands of lectins is therefore a promising strategy for the treatment of some pathological disorders involving

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_4, © Springer Science+Business Media, LLC 2012

57

58

S. Cecioni et al.

such receptor–ligand interactions. In this respect, chemists have designed several types of multivalent glycoconjugates (3–7) mimicking the natural multivalency associated with such interactions, a phenomenon usually termed the “cluster effect” (8, 9). Multivalent glycoconjugates are macromolecules displaying multiple copies of a carbohydrate moiety connected to a central core scaffold. Chemists have studied the possible strategies to connect carbohydrate derivatives to its scaffold. The chemical reaction selected as the conjugation technique must provide the desired glycoconjugates in high yields and short reaction times, with high chemo- and regioselectivities and must be usually compatible with aqueous media due to the properties of most biomolecules. The recent development of Cu(I)-catalyzed azide-alkyne cycloaddition (CuACC) (10–12) provides a very reliable and powerful technique for the conjugation of azides and alkynes, which are highly stable in biological media and also orthogonal to most functional groups present in biomolecules. This approach has found applications for chemical ligation of carbohydrate probes to synthetic multivalent scaffolds (13). The synthesis of glycoconjugates through CuAAC can be achieved from azido-functionalized carbohydrates bearing the azido moiety attached either directly at the anomeric center or connected through a spacer arm (Fig. 1).

Fig. 1. Schematic representation for the conjugation of azido-functionalized carbohydrates to alkynylated scaffolds (grey ball). “Click” refers to the conjugation through copper(I)-catalyzed azide-alkyne cycloaddition (CuACC).

4  Synthesis of Azido-Functionalized Carbohydrates for the Design…

59

2. Materials 2.1. General methods

Compounds 1 to 16 were from commercial sources and used without further purification. 1. N,N-Dimethylformamide (DMF, >99%, Sigma-Aldrich). 2. Ethyl acetate (EtOAc, technical, SDS). 3. Petroleum ether (PE, technical, SDS). 4. Dimethylsulfoxide (DMSO, 99.7%, Acros Organics). 5. 2-[2-(2-chloroethoxy)ethoxy]ethanol (97%, Acros Organics). 6. CH2Cl2 (Acros Organics). 7. Tin(IV) chloride: 1 M SnCl4 solution in CH2Cl2. 8. Silver trifluoroacetate (98%, Acros Organics). 9. Sodium azide (99%, Sigma-Aldrich). 10. Tetra-n-butyl ammonium iodide (98%, Sigma-Aldrich). 11. 2,3,4,6-tetra-O-acetyl-a-d-glucopyranosyl bromide 1 (99%, Sigma-Aldrich). 12. 1,2,3,4,6-penta-O-acetyl-b-d-galactopyranose 3 (99%, Carbo­ synth) (14). 13. Diethyl ether (technical grade, Sigma-Aldrich). 14. Magnesium sulfate (MgSO4, technical grade, VWR). 15. Sodium sulfate (Na2SO4, technical grade, Laurylab). 16. Sodium hydrogenocarbonate (NaHCO3, technical grade, Laurylab). 17. Sodium chloride (NaCl, technical grade, Laurylab). 18. 1,2,3,6,2¢,3¢,4¢,6¢-Octa-O-acetyl-d-lactose 5 was prepared from previously reported procedures (15). 19. Thin-layer chromatography plates: (TLC) was carried out on aluminum sheets coated with silica gel 60F254 (Merck). 20. TLC charring (9:1 v/v).

solution:

10%

H2SO4

in

EtOH/H2O

21. Geduran® silica gel Si 60 (40–63 mm) purchased from Merck (Darmstadt, Germany). 2.2. Analytical Techniques

1. Saturated aqueous NaHCO3: The solution was prepared using an excess of solid (NaHCO3, >100 g/L) in water in order to obtain a saturated aqueous solution meaning that some solid was still present at the bottom of the bottle even after vigorous shaking. 2. Saturated aqueous NaCl (Brine): The solution was prepared using an excess of solid (NaCl, >360 g/L) in water in order

60

S. Cecioni et al.

to obtain a saturated aqueous solution meaning that some solid was still present at the bottom of the bottle even after vigorous shaking. 3. Calcium hydride: CaH2. 4. Dry Dichloromethane (CH2Cl2, Carlo Erba): Dichloromethane (CH2Cl2, Carlo Erba) was distilled over CaH2 under argon atmosphere. 5. Proton (1H) and carbon (13C) NMR spectra were recorded at 298  K using a Bruker Avance DRX300 spectrometer at 300  MHz with the residual solvent as the internal standard (CHCl3 at 7.26 ppm for 1H and 77.16 for 13C). 6. ESI mass spectra were recorded in the positive mode using a Thermo Finnigan LCQ spectrometer. 7. Optical rotations were measured using a Perkin Elmer polari­ meter and values are given in 10−1 deg/cm2/g.

3. Methods The azido-functionalized carbohydrates can be conjugated to the desired (bio)molecules as their acetylated precursors, which can then be unmasked to the corresponding hydroxylated derivatives under standard conditions (e.g. NaOH/H2O, MeOH/Et3N/ H2O, NaOMe/MeOH). This final deprotection step can be avoided by using the hydroxylated azido-functionalized carbohydrates but the purification of the intermediates is sometimes troublesome. Acetylated precursors are usually preferred for most of the synthetic organic multivalent glycoconjugates since their purification can be readily achieved by silica gel column chromatography, while bioorganic glycoconjugates (e.g. proteins, peptides, DNA) would be conjugated with hydroxylated carbohydrate probes. Water used for liquid/liquid extractions was de-ionized and bacteria-free by treatment with a EASYpureRoDI purification system. All reactions were performed under argon. Solutions in organic solvents were dried with anhydrous Na2SO4 and concentrated under reduced pressure at 30–35°C. Compounds were visualized by UV light (l = 254 nm) and/or by charring with 10% H2SO4 in EtOH/H2O (9:1 v/v). Purification was performed by flash-chromatography with Geduran® silica gel Si 60 (40–63 mm) purchased from Merck (Darmstadt, Germany). 3.1. Synthesis of Glycosyl Azides

When subjected to CuAAC reaction, glycosyl azides displaying an azide moiety directly at the anomeric center of carbohydrates will afford N-heterocyclic derivatives, which could present interesting properties and possible resistance to carbohydrate processing

4  Synthesis of Azido-Functionalized Carbohydrates for the Design… OAc O

AcO AcO 1

AcO

Br

a 81 %

AcO AcO

OAc O 2

61

N3

OAc

Scheme 1. Synthesis of the azido-functionalized glucose derivative 2. Reagents and conditions: (a) NaN3, DMSO, r.t., 3 h.

enzymes (e.g. glycosidases, glycosyltransferases). This strategy could therefore prove beneficial for the design of enzyme-resistant glycoconjugates, which might not be eliminated too quickly from living systems (e.g. cells, bacteria, viruses, animals). The synthesis is usually achieved from the peracetylated brominated carbohydrates available from commercial sources. Never­ theless, glycosyl bromides are water and temperature sensitive and may decompose to the corresponding hemiacetal. The reactions must therefore be performed under strictly anhydrous conditions. The anomeric bromine atom is displaced by the azide anion via a SN2 mechanism (i.e. with inversion of configuration at the anomeric carbon) to obtain the desired azido-functionalized carbohydrates (Scheme 1). In addition, 1,1-diazides can be readily prepared from the bis-halogenated carbohydrate precursor (16). 3.1.1. Synthesis of the Glucose Derivative 2

1. Introduce sodium azide (Caution: see Note 1) (1.58  g, 24.2 mmol) in a 100-mL round-bottom flask. Close the flask with a rubber septum and inert with argon. 2. Add DMSO (40 mL) and stir the suspension at room temperature until complete dissolution of the solid. 3. Open the flask and introduce quickly peracetylated glucosyl bromide 1 (5 g, 12.2 mmol) and close the flask again with the rubber stopper then inert with argon. 4. Stir the reaction at room temperature for 3 h when TLC (eluent: Et2O/PE 1:1) shows complete conversion of the starting material 1 (Rf = 0.66) and the formation of the desired azide 2 as a more polar product (Rf = 0.49). 5. Dilute the solution with Et2O (150 mL) and wash with H2O (100 mL). Extract the aqueous layer with Et2O (2 × 50 mL). Combine organic layers and wash with H2O (4 × 150 mL), dry over MgSO4, filter the solid, wash with Et2O (2 × 15 mL) and concentrate the filtrate under reduced pressure on a rotatory evaporator (Buchi, 30°C, 20 mmHg). 6. Dissolve the crude material (beige solid) in a minimum of Et2O and precipitate with petroleum ether. Filter the solid to obtain a first crop. Concentrate the filtrate and precipitate again more solid off the solution as previously from Et2O and petroleum ether.

62

S. Cecioni et al.

Fig. 2. Proton NMR spectrum (1H, 300 MHz) of compound 2 in CDCl3 recorded at 25°C.

7. Both precipitations afford the product 2 as an analytically pure white powder (3.12 g, 9.83 mmol, 81%). Analytical data for 2,3,4,6-tetra-O-acetyl-b-d-glucopyranosyl azide (2, see Ref. 17 and Figs. 2 and 3): Rf = 0.49 eluent: Et2O/PE 1:1; M.p. = 125.6–126.5°C; [a]D = −29° (c = 2, CHCl3); 1H NMR (CDCl3, 300 MHz) d 5.22 (pdd, 1H, J = 9.4 Hz, H-3), 5.10 (pdd, 1H, J = 9.7 Hz, H-4), 4.96 (pdd, 1H, J = 9.2 Hz, H-2), 4.65 (d, 1H, J = 8.8 Hz, H-1), 4.28 (dd, 1H, J = 12.5, 4.8 Hz, H-6), 4.17 (dd, 1H, J = 12.5, 2.3 Hz, H-6¢), 3.79 (ddd, 1H, J = 9.9, 4.7, 2.4 Hz, H-5), 2.10, 2.08, 2.03, 2.01 (4s, 4 × 3 H, 4 × CH3CO); 13C NMR (CDCl3, 75 MHz) d 170.5, 170.0, 169.3, 169.1 (4s, 4 × CH3CO), 87.8 (C-1), 73.9 (C-5), 72.5 (C-3), 70.6 (C-2), 67.8 (C-4), 61.6 (C-6), 20.6, 20.4, 20.4, 20.4 (4s, 4 × CH3CO). 3.2. Synthesis of w-AzidoTriethyleneglycol Glycosides

Glycosylation reactions (18) usually require the activation of the anomeric carbon atom with a proper leaving group but the synthesis of such activated derivatives is adding up to three synthetic steps and the corresponding products are not always stable enough to be stored under simple conditions. Peracetylated carbohydrates are easily prepared from the corresponding native carbohydrates or can be also purchased from most chemical suppliers. They are very

4  Synthesis of Azido-Functionalized Carbohydrates for the Design…

63

Fig. 3. Carbon NMR spectrum (13C, 75 MHz) of compound 2 in CDCl3 recorded at 25°C. AcO AcO

OAc O 3

AcO AcO

OAc O

O OAc AcO 5

OAc OAc OAc O

a,b 60%

AcO AcO

a,b OAc 62% OAc

OAc O O OAc

AcO AcO

OAc O

O

O

N3

4

O OAc AcO 6

OAc O OAc

O

O

O

N3

Scheme 2. Syntheses of the azido-functionalized galactose and lactose derivatives. Reagents and conditions: (a) SnCl4, CF3CO2Ag, H(OCH2CH2)3Cl, CH2Cl2, r.t., 2–3 h, (b) NaN3, nBu4NI, DMF, 70°C, 16 h. Yields are indicated for the complete two-step process.

stable and can be stored for longer periods of time. The glycosylation from peracetylated carbohydrates usually requires a Lewis acid as a promoter of the reaction (e.g. boron trifluoride etherate, titanium(IV) chloride). We have recently revisited a glycosylation protocol (19) involving tin(IV) chloride (SnCl4) and silver trifluoroacetate (CF3CO2Ag) for the large-scale synthesis (typically 5  g and up to 20  g) of w-azido-triethyleneglycol glycosides and obtained reproducibly the desired 1,2-trans glycosides in short reaction times (typically 2–3 h) and high yields (Scheme 2). When a mixture of anomeric acetates is used, the b-derivative will be consumed in a few hours (~2–3 h) while the a-anomer will remain

64

S. Cecioni et al.

unchanged in the reaction mixture over a long period of time (>24 h). The b-configured sugar peracetates will therefore always be preferred to their a-isomers due to the greater reactivity of the b-anomers. 3.2.1. Synthesis of the Galactose Derivative 4

1. Introduce the peracetylated galactose 3 (5 g, 12.8 mmol) and silver trifluoroacetate (4.24 g, 19.2 mmol) in a 250-mL roundbottom flask. Close the flask with a rubber septum and inert with argon. The reaction flask will be kept away from light by wrapping into aluminum foil in order to prevent the silver salts from decomposing slowly. 2. Add freshly distilled dichloromethane (120  mL) and 2-[2-(2-chloroethoxy)ethoxy]ethanol (2.80 mL, 19.2 mmol) dropwise to obtain a solution. Keep an argon atmosphere over the reaction. 3. Add SnCl4 (1 M in CH2Cl2, 38.4 mL, 38.4 mmol) dropwise within ~30 min at 0°C (ice bath) to the stirred solution. Stir the solution under argon at room temperature. 4. Monitor the reaction completion by TLC (eluent: PE-EtOAc 1:1, Rf = 0.60 for 3 and 0.34 for 4). Disappearance of the starting material was observed within 1–3 h, occasionally with the mixture turning to a pale pink color. 5. Add a solution of saturated aqueous NaHCO3 (Solution 1, 100 mL) in order to adjust pH above 8 (pH paper) and stir the solution vigorously for 20 min (see Note 2). 6. Dilute the resulting biphasic system with 500 mL of water (see Note 3) and extract the aqueous layer with CH2Cl2 (3 × 150 mL). 7. Combine the organic layers and wash (see Note 4) with saturated aqueous NaHCO3 (Solution 1, 150  mL), water (3 × 150  mL), brine (Solution 2, 2 × 150  mL) and dry over sodium sulfate (Na2SO4). 8. Filter the solid and wash with CH2Cl2 (2 × 50 mL). Evaporate the solvent on a rotatory evaporator (Buchi, 30°C, 20 mmHg) to obtain a crude pale yellow gum and transfer the gum into a 250-mL round bottom flask by dissolving in CH2Cl2 and ­re-evaporating the solvent (see Note 5). 9. Add sodium azide (Caution: see Note 1) (4.16 g, 64.0 mmol) and tetra-n-butyl ammonium iodide (4.73  g, 12.8  mmol). Close the flask with a rubber septum and inert with argon. 10. Add DMF (150 mL) and stir the mixture at 70°C under argon for 16 h. 11. Cool the reaction mixture to room temperature, filter the solids off and wash with EtOAc (3 × 100 mL).

4  Synthesis of Azido-Functionalized Carbohydrates for the Design…

65

12. Dilute the filtrate with EtOAc to obtain a total volume of 1 L. Wash the organic layer (see Note 6) with saturated aqueous NaHCO3 (Solution 1, 3 × 300 mL), water (3 × 500 mL), brine (Solution 2, 400 mL) and dry over sodium sulfate (Na2SO4). 13. Filter the solid and wash with EtOAc (2 × 50 mL). Evaporate the solvent on a rotatory evaporator (Buchi, 30°C, 20 mmHg) to obtain a crude yellow to orange gum. 14. Purify the product by silica gel column chromatography (eluent: PE-Et2O 1:4, internal diameter = 45  mm; length = 250– 300 mm) to obtain 4 (3.88 g, 60% yield) as a pale yellow gum (see Note 7). Analytical data for 1-azido-3,6-dioxaoct-8-yl 2,3,4,6-tetra-Oacetyl-b-d-galactopyranoside (4, see Refs. 20–22 and Fig.  4): Rf = 0.34 (1:1 PE-EtOAc). 1H NMR (CDCl3) d 1.99, 2.04, 2.07, 2.15 (4s, 4 × 3H, 4 × CH3CO), 3.40 (t, 2H, J = 5.0 Hz, CH2N3), 3.60– 3.82 (m, 9H, OCH2), 3.88–4.02 (m, 2H, OCH2, H-5), 4.08–4.20 (m, 1H, H-6, H-6¢), 4.59 (d, 1H, J1,2 = 7.9 Hz, H-1), 5.02 (dd, 1H, J3,4 = 3.4 Hz, J3,2 = 10.5 Hz, H-3), 5.21 (dd, 1H, J2,1 = 7.9 Hz, J2,3 = 10.5 Hz, H-2), 5.39 (dd, 1H, J4,5 = 0.7 Hz, J4,3 = 3.4 Hz, H-4). 13 C NMR (CDCl3) d 20.3, 20.4, 20.4, 20.5 (4s, 4 × CH3CO), 50.4 (CH2N3), 61.0 (C-6), 66.8 (C-4), 68.5 (C-2), 68.8, 69.8, 70.1 (3s, 3 × CH2O), 70.4 (C-3), 70.4, 70.5 (2  s, 2 × CH2O), 70.6 (C-5), 101.1 (C-1), 169.2, 169.9, 170.0, 170.1 (4s, 4 × CH3CO).

2200 2100 2000 1900 1800 1700 1600 1500 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 0 .0

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 −0.5 f1 (ppm)

Fig. 4. Proton NMR spectrum (1H, 300 MHz) of compound 4 in CDCl3 recorded at 25°C. Traces of EtOAc were observed in the spectrum (triplet at 1.25 ppm).

66

S. Cecioni et al. 2800 2600 2400 2200 2000 1800 1600 1400 1200 1000 800 600 400 200 0 7.5

7.0

6.5

6.0

5.5

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0

f1 (ppm)

Fig. 5. Proton NMR spectrum (1H, 300 MHz) of compound 6 in CDCl3 recorded at 25°C. Traces of EtOAc were observed in the spectrum (triplet at 1.25 ppm).

3.2.2. Synthesis of the Lactose Derivative 6

The synthesis of the lactose derivative 6 can be performed under the same conditions as presented for compound 4. The only difference is in the purification over silica gel where a gradient was used from 7:3 to 2:3 PE-EtOAc with 10% increase every 400  mL. Compound 6 was obtained as a pale yellow gum in 62% yield from 5 g of peracetylated lactose 7. Analytical data for 1-azido-3,6-dioxaoct-8-yl 2,3,6,2¢,3¢,4¢, 6¢-hepta-O-acetyl-b-d-lactoside (6, see Ref. 23 and Fig. 5): Rf = 0.26 (1:1 PE–EtOAc). 1H NMR (CDCl3) d 1.94, 2.02, 2.04, 2.10, 2.13 (5 s, 21H, CH3CO), 3.37 (t, 2H, J = 6.0 Hz, CH2N3), 3.57–4.15 (m, 16H), 4.42–4.48 (m, 2H), 4.59 (d, 1H, J = 7.9  Hz), 4.87 (dd, 1H, J = 8.7  Hz, J = 8.9  Hz), 4.93 (dd, 1H, J = 2.8  Hz, J = 8.7 Hz), 5.09 (dd, 1H, J = 8.0 Hz, J = 11.5 Hz), 5.17 (t, 1H, J = 11.5 Hz), 5.31–5.33 (m, 1H).

4. Notes 1. CAUTION: Sodium azide, when inhaled, is highly toxic and cases of death have been reported (MSDS J.T. Baker). Precautions must be taken when weighing the material such as using a powder mask and a teflon spatula (metallic spatula may cause explosion). The azidation reaction was performed behind

4  Synthesis of Azido-Functionalized Carbohydrates for the Design…

67

a plastic shield due to the potential explosion. DMF is used as a polar solvent favoring the reaction but also to maintain a basic pH (>8) of the solution. Hydrazoic acid (HN3) may be formed in acidic pH, which may explode and/or, when inhaled, may cause intoxication, damage of the central nervous system and blood pressure effects. DMSO can be used for azidation procedures but halogenated solvents (CHCl3 and CH2Cl2) must be avoided as they may lead to unstable azide derivatives prone to explosive decomposition. 2. When performing the reaction on a larger scale (typically > 10 g), it is better to transfer the solution in a 1 L erlenmeyer containing 500  mL of saturated aqueous NaHCO3 (Solution 1). Special care should be taken with the formation of foam during the neutralization of tin(IV) chloride with NaHCO3. 3. The solid present in the suspension may be filtered through a bed of Celite for small scale syntheses (>1 g). 4. The water layers should be disposed with proper care considering the presence of tin and silver salt. 5. The removal of trace amounts of CH2Cl2 should be considered seriously to avoid the formation explosive species with sodium azide. 6. The aqueous layers must not be in contact with acidic solutions to avoid the formation of hydrazoic acid (see Note 1). 7. TLC analyses did not show a significant difference between the polarities of the chlorinated precursors and the azido compounds.

Acknowledgments The authors thank the University Claude Bernard Lyon 1 and the CNRS for financial support. S.C. thanks the Région Rhône-Alpes (Cluster de Recherche Chimie) for a PhD stipend and D.G. thanks the Agence Nationale de la Recherche (ANR) for a PhD stipend. References 1. Dwek, R. A. (1996) Glycobiology: Toward Understanding the Function of Sugars. Chem. Rev. 96, 683–720. 2. Varki, A., Cummings, R., Esko, J., Freeze, H., Hart, G. W., and Marth, J. Essentials of glycobiology, Cold Spring Harbor Laboratory Press: New York, 1999. 3. Flitsch, S. L. (2000) Chemical and enzymatic synthesis of glycopolymers. Curr. Opin. Chem. Biol. 4, 619–625.

4. Cloninger, M. J. (2002) Biological applications of dendrimers. Curr. Opin. Chem. Biol. 6, 742–748. 5. de la Fuente, J. M., and Penadés, S. (2006) Glyconanoparticles: Types, synthesis and applications in glycoscience, biomedicine and material science. Biochim. Biophys. Acta. 1760, 636–651. 6. Imberty, A., Chabre, Y. M., and Roy, R. (2008) Glycomimetics and Glycodendrimers as High

68

S. Cecioni et al.

Affinity Microbial Anti-adhesins. Chem. Eur. J. 14, 7490–7499. 7. Chabre, Y. M., and Roy, R. R. (2010) Design and Creativity in Synthesis of Multivalent Neoglycoconjugates. Adv. Carbohydr. Chem. Biochem. 63, 165–393. 8. Mammen, M., Choi, S.-K., and Whitesides, G. M. (1998) Polyvalent Interactions in Biological Systems: Implications for Design and Use of Multivalent Ligands and Inhibitors. Angew. Chem. Int. Ed. 37, 2754–2794. 9. Lundquist, J. J., and Toone, E. J. (2002) The Cluster Glycoside Effect. Chem. Rev. 102, 555–578. 10. Rostovtsev, V. V., Green, L. G., Fokin, V. V., and Sharpless, K. B. (2002) A Stepwise Huisgen Cycloaddition Process: Copper(I)-Catalyzed Regioselective “Ligation” of Azides and Terminal Alkynes. Angew. Chem. Int. Ed. 41, 2596–2599. 11. Tornøe, C.W., Christensen, C. and Meldal, M. (2002) Peptidotriazoles on Solid Phase: [1,2,3]-Triazoles by Regiospecific Copper(I)Catalyzed 1,3-Dipolar Cycloadditions of Terminal Alkynes to Azides. J. Org. Chem. 67, 3057–3064. 12. Meldal, M., and Tornøe, C. W. (2008) Cu-Catalyzed Azide-Alkyne Cycloaddition. Chem. Rev. 108, 2952–3015. 13. Dondoni, A. (2007) Triazole: the Keystone in Glycosylated Molecular Architectures Cons­ tructed by a Click Reaction. Chem. Asian J. 2, 700–708. 14. Wolfrom, M. L., and Thompson, A. (1963) in Methods in Carbohydrate Chemistry, Academic Press: New York, vol. 2, pp. 211–215. 15. Westerlind, U., Hagback, P., Tidbäck, B., Wiik, L., Blixt, O., Razib, N., and Norberg, T. (2005) Synthesis of deoxy and acylamino derivatives of lactose and use of these for probing the active site of Neisseria meningitidis

N-acetylglucosaminyltransferase. Carbohydr. Res. 340, 221–233. 16. Praly, J.-P., Péquery, F., Di Stefano, C., and Descotes, G. (1996) Synthesis of protected glycopyranosylidene 1,1-diazides. Synthesis, 577–579. 17. Tropper, F. D., Anderson, F. O., Braun, S., and Roy, R. R. (1992) Phase Transfer Catalysis as a General and Stereoselective Entry into Glycosyl Azides from Glycosyl Halides. Synthesis, 618–620. 18. Fugedi, P. Glycosylation methods In The organic chemistry of carbohydrates. Levy, D. E., and Fugedi, P. Eds, CRC Taylor and Francis: New York, 2006, Chapter 4, pp. 89–151. 19. Xue, J. L., Cecioni, S., He, L., Vidal, S., and Praly, J.-P. (2009) Variations on the SnCl4 and CF3CO2Ag-promoted glycosidation of sugar acetates: a direct, versatile and apparently simple method with either a or b stereocontrol. Carbohydr. Res. 344, 1646–1653. 20. Szurmai, Z., Szabo, L., and Liptak, A. (1989) Diethylene and triethylene glycol spacers for the preparation of neoglycoproteins. Acta Chim. Hung. 126, 259–269. 21. Chevolot, Y., Bouillon, C., Vidal, S., Morvan, F., Meyer, A., Cloarec, J.-P., et  al. (2007) DNA-Based Carbohydrate Biochips: A Platform for Surface Glyco-Engineering. Angew. Chem. Int. Ed. 46, 2398–2402. 22. Cecioni, S., Lalor, R., Blanchard, B., Praly, J.-P., Imberty, A., Matthews, S.E., and Vidal, S. (2009) Achieving High Affinity towards a Bacterial Lectin through Multivalent Topological Isomers of Calix[4]arene Glycoconjugates. Chem. Eur. J. 15, 13232–13240. 23. Li, J., Zacharek, S., Chen, X., Wang, J., Zhang, W., Janczuk, A., and Wang, P. G. (1999) Bacteria targeted by human natural antibodies using a-gal conjugated receptor-specific glycopolymers. Bioorg. Med. Chem. 7, 1549–1558.

Chapter 5 Polypyrrole-Oligosaccharide Microarray for the Measurement of Biomolecular Interactions by Surface Plasmon Resonance Imaging Julia Bartoli, André Roget, and Thierry Livache Abstract The polypyrrole approach initially developed for the construction of DNA chips, has been extended to other biochemical compounds such as proteins and more recently oligosaccharides. The copolymerization of a pyrrole monomer with a biomolecule bearing a pyrrole group by an electrochemical process allows a very fast coupling of the biomolecule (probe) to a gold layer used as a working electrode. Fluorescencebased detection is the reference method to detect interactions on biochips; however an alternative label free method, could be more convenient for rapid screening of biointeractions. Surface Plasmon Resonance (SPRi) is a typical label-free method for real time detection of the binding of biological molecules onto functionalized surfaces. This surface sensitive optical method is based upon evanescent wave sensing on a thin metal layer. The SPR approach described herein is performed in an imaging geometry that allows simultaneous monitoring of biorecognition reactions occurring on an array of immobilized probes (chip). In a SPR imaging experiment, local changes in the reflectivity are recorded with a CCD camera and are exploited to monitor up to 100 different biological reactions occurring onto the molecules linked to the polypyrrole matrix. This method will be applied to oligosaccharide recognition. Key words: Polypyrrole, Oligosaccharide, Array, Biochip, Surface plasmon resonance

1. Introduction Recently, there has been an increasing interest for the development of microarray devices adapted to the high-throughput analysis of biological events. The presentation of a large set of probes (DNA, proteins …) in the format of an array provides a mean to simultaneously monitor multiple binding events, with a compatible detection method. A number of techniques have been developed to

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_5, © Springer Science+Business Media, LLC 2012

69

70

J. Bartoli et al.

immobilize these probes on a high variety of supports. More recently, in the wake of DNA and proteins, the study of oligosaccharides structures and protein-oligosaccharide interactions became of special interest, and the development of carbohydrate chips is an emerging field. The understanding of the different roles of oligosaccharides is in an early stage as compared with our knowledge of DNA and proteins. If DNA and protein syntheses are well developed, oligosaccharide synthesis is much more difficult to perform and remains challenging. The development of DNA and protein chips, including on chip synthesis (1, 2) has been particularly important, but only a few strategies are described to prepare oligosaccharides chips. Although these approaches are of general interest for these biomolecules, the problem is not so easy to solve. A few approaches of in situ synthesis are described but suffer from the extreme complexity of the chemistry (3). A number of methods involve either the chemical synthesis in solution (4, 5) the enzymatic synthesis (6) or the extraction of the biomolecule and the conjugation to a chemical moiety allowing the immobilization on a great variety of support (glass, plastic, gold…). A comparison between different supports was done by Angenendt showing high signal uniformity and reproducibility of most plain glass and plastic slides (7). An efficient method, namely electrospotting, has been developed to covalently immobilize DNA sequences or proteins on gold surfaces using a simplified electrochemical process (8). This process allows the grafting of biomolecules on a nonstructured conducting layer (homogenous gold film). Initially developed for DNA chips, this approach has been adapted for oligosaccharides but necessitates a functionalization step. A number of functions are available on the oligosaccharide backbone and are a potential source of anchoring sites but as for DNA and proteins it is necessary that the chemical coupling doesn’t involve a functional domain of the molecule. For this reason it is more convenient to modify the oligosaccharide reducing extremity, although there is a lack of reactivity in this position (see Note 1). We describe in this chapter an approach combining the functionalization of oligosaccharides in two steps, the fast grafting of these oligosaccharides on a microarray, via an electrodeposition process and a polypyrrole film, and the detection of oligosaccharide–protein interactions by Surface Plasmon Resonance imaging (SPRi). SPRi is a label-free, surface sensitive technique, which allows the real-time measurement of biological binding events occurring onto a gold surface bearing covalently attached probes. An application example is given, using glycosaminoglycans (GAG) or GAG fragments as models of oligosaccharide. For that purpose, heparin and heparin derived octasaccharide were modified by a pyrrole moiety on their reducing extremity and immobilized on a gold surface through electrocopolymerization. The interactions with proteins (SDF-1alpha and IFN-gamma) on the surface are monitored without labelling in real time by SPRi.

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

71

2. Materials 2.1. Synthesis of the NHS Activated Pyrrole: N-Hydro­xy succinimidyl-6(Pyrrole-yl)-Caproate (NHS-Pyrrole)

1. 2,5-Dimethoxytetrahydrofuran, mixture of cis- and trans-isomers 99% (Acros organics). 2. 6-Aminocaproic acid 99 + % (Acros organics). 3. 1,4-Dioxan, synthesis grade. 4. Acetic acid, purex analytical grade. 5. N-hydroxysuccinimide 97% (Aldrich). 6. N,N¢-Dicyclohexylcarbodiimide 99% (Acros organics). 7. N,N-Dimethylformamide (DMF) for analysis. 8. Dimethyl sulfoxide, analytical reagent. 9. Dichloromethane (CH2Cl2) purex for analysis. 10. Ethyl Alcohol anhydrous (EtOH), for analysis. 11. Silica gel PF254 containing CaSO4 for preparative layer chromatography (MERCK). 12. Chloroform-d, 99.9 atoms % D, stabilized with 0.5% wt silver foil.

2.2. Two Steps Preparation of Pyrrolylated Oligosaccharides

1. 15 kDa Heparin (Sigma-Adrich). 2. Heparin fragment octasaccharide (dp8) obtained by heparinase I depolymerization (9). 3. Adipic dihydrazide (Merck). 4. First coupling buffer: 100 mM CH3COONa, adjusted at pH 5 with acetic acid. Store at 4°C. 5. NHS-Pyrrole, synthesized in the laboratory. 6. Second coupling buffer: 50% Phosphate buffer saline (PBS, Euromedex)/50% Dimethyl sulfoxide (DMSO, Sigma-Aldrich) v/v. 7. Dialysis membrane MWCO = 1,000 Da (Standard RC Dialysis Tubing, Pre-treated, Spectra Por 7, Spectrum labs). 8. SpeedVac Concentrator system. 9. Freeze-dryer.

2.3. Quantification of the Coupling Reaction: Tnbsa Method

1. TNBSA solution: 5% (w/v) Picrylsulfonic acid in H2O (P229710 mL, Sigma-Aldrich). The solution has to be stored at 4°C, and to be diluted at 1:10 in buffer 1, just before use. 2. Buffer 1 (reacting buffer): 100  mM Na2CO3/NaHCO3, pH 9.6 (the solution can be stored several months at 4°C). 3. Buffer 2 (stopping buffer): 98.5% 100  mM NaH2PO4/1,5% 100 mM Na2SO4 (each solution has to be stored at –20°C, and mixed before use).

72

J. Bartoli et al.

4. 10  mM stock solution of adipic dihydrazide in buffer 1, for calibration, prepared extemporaneously. 5. 96-well plate (Nunc). 6. Victor 1420 Multilabel Counter. 2.4. Oligosaccharides Immobilization on SPR Prisms: Electrospotting

1. 1  M Pyrrole (Tokyo Kasei, Japan) in anhydrous acetronitrile 99.8% (Sigma-Aldrich). Store at −20°C in a dark bottle to ­minimize pyrrole oxidation. 2. Pyrrole-oligosaccharide conjugates (prepared as in Sub­ heading 3.2), at different concentrations. 3. Electrocopolymerization buffer: 50  mM NaH2PO4, 50  mM NaCl, 10% (p/v) glycerol adjusted at pH 6.8 with a NaOH solution. Store at 4°C (−20°C for a conservation longer than 2 weeks). 4. Glass prisms (n = 1.717 at l = 633 nm) covered by 50 nm-thick gold layer (Genoptics, Orsay, France). 5. Pre-treatment solution: mixture of 70% H2SO4 and 30% H2O2 (v/v) (prepared just before use). Care must be taken as this reaction is highly exothermic and reacts strongly with organic compounds. 6. 96-well plate (Nunc, conical, #249946). 7. Microarrayer for the immobilization the oligosaccharides on the chip (Genomic Solutions), with an acquisition card U12 (LabJack), and a software developed from Labview. 8. Ceramic needle with an inox extremity (X-Tend Pin, 350 mm external diameter/150 mm internal diameter, Genoptics, Orsay, France), containing a platinum wire (diameter 200 mm).

2.5. SPRi Interactions Monitoring

1. “SPRi lab”: SPR imaging system (Genoptics, Orsay, France) and Software Genovision (Genoptics) and Labview. The apparatus is placed in an air oven, under controlled temperature (25°C). 2. 7.2 mL PEEK flow cell. 3. Syringe pump (Cavro/Tecan XL3000) and software developed from LabView. 4. Degasser (Alltech). 5. Injection valve with 500  mL injection loop and tubing in PEEK. 6. Injection and Washing buffer: 10 mM HEPES, 150 mM NaCl, 0,005% Tween 20, pH 7.4. 7. Saturation buffer: Bovine Serum Albumin (BSA, SigmaAldrich) 1% (w/v) in washing buffer.

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

73

8. Regeneration buffers: (a) 1 M NaCl (Sigma-Aldrich) in washing buffer. (b) 1% (w/v) Sodium Dodecyl Sulphate (SDS) (Euromedex) in water. (c) 2 M Guanidine Hydrochloride (Sigma-Aldrich) in washing buffer. 9. All the buffers are filtrated on a 0.22 mm membrane, degassed, aliquoted (50  mL, 15  mL, or 1  mL) and stored at –20°C, except for the SDS buffer, stored at room temperature (RT) (see Note 2). 10. Stromal cell-derived factor-1a (SDF-1a) and interferon-g (IFNg) obtained from H. Lortat-Jacob, prepared as previously described (9, 10).

3. Methods 3.1. Synthesis of the NHS Activated Pyrrole: N-Hydro­xysuccinimidyl-6(Pyrrol-yl)-Caproate (NHS-Pyrrole)

1. Heat under reflux for 4  h a mixture of 2,5-Dimetho­ xytetrahydrofuran (490  mmol, 64.85  mL), 6-Aminocaproic acid (430 mmol, 56.33 g), acetic acid (430 mL) and 1,4-Dioxan (570 mL) and stir it at room temperature overnight (11). 2. Remove the volatiles under reduced pressure in a rotavapor; dissolve the residue in ethanol and coevaporate it (2 × 100 mL) to eliminate acetic acid. 3. The product 1, 6-pyrrolyl caproic acid, is obtained with a yield of 86% after chromatographic purification on silica gel (500 g) column; CH2Cl2/EtOH as eluents. Start the elution by 500 mL CH2Cl2, then 300 mL 98% CH2Cl2/2% EtOH, then 400 mL 95%/5% EtOH, product 1 goes out the column for a gradient of 90%/10% EtOH. This product 1 looks like a brownish oil. 4. M.S. product 1 (m/z) = 182.1 (M+) 5. 1H-RMN product 1 (200 MHz; CDCl3/TMS) d (ppm): 1.72 (m, 6H, –CH2–(CH2)3– CH2–); 2.34 (t, 2H, –CH2–CO2H); 3.87 (t, 2H, –CH2–N); 6.13 (dd, 2H, 3-H and 4-H pyrrole); 6.64 (dd, 2H, 2-H and 5-H pyrrole). 6. Mix and stir overnight, at room temperature, product 1 (144  mmol, 26.05  g), N-Hydroxysuccinimide (144  mmol, 16.56 g), N,N  ¢-Dicyclohexylcarbodiimide (159 mmol, 32.75 g) and DMF (1,500 ml). 7. Filter the mixture to eliminate N,N¢-Dicyclohexylurea. 8. Remove the volatiles under reduced pressure. Product 2, N-Hydroxysuccinimidyl-6-(pyrrole-yl)-caproate, is used without any other purification.

74

J. Bartoli et al.

9. 1H-RMN product 2 (200 MHz; CDCl3/TMS) d (ppm): 1.72 (m, 6H, –CH2–(CH2)3– CH2–); 2.34 (t, 2H, –CH2–CO); 2.88 (tt, 4H, –CH2–CH2–: NHS); 3.87 (t, 2H, –CH2–N); 6.13 (dd, 2H, 3-H and 4-H pyrrole); 6.64 (dd, 2H, 2-H and 5-H: pyrrole). 3.2. Two Steps Preparation of Pyrrolylated Oligosaccharides

In a first step, the oligosaccharide is coupled with adipic dihydrazide, and after purification, the glycosylhydrazide is finally functionalized with NHS-Pyrrole (Fig. 1).

3.2.1. Functionalization of Oligosaccharides with Adipic Dihydrazide

The reaction between the aldehyde function of the oligosaccharide reducing end with the hydrazide function of adipic dihydrazide, in an acidic medium, is the limiting step. Hydrazides are attractive for coupling reactions with oligosaccharides as they retain their nucleophilicity in acidic aqueous media, and the acylhydrazone which is formed is comparatively stable. Equilibration to the tautomeric ringclosed glycosylhydrazide is favoured, which is an advantage since the native form of the reducing end monosaccharide is preserved (12). 1. Prepare a 500  mM adipic dihydrazide stock solution, in the CH3COONa coupling buffer. 2. Prepare the reacting solution by mixing 10 mL of 100 mg/mL oligosaccharide (solubilised in water) with 50 mL of 500 mM adipic dihydrazide and complete to 500 mL final volume with

Fig. 1. Preparation of modified oligosaccharide bearing a pyrrole group and a chemical linker, in two steps.

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

75

coupling buffer. These conditions lead to a 2  mg/mL final concentration of oligosaccharide and a 50 mM final concentration of adipic dihydrazide. 3. Prepare negative controls by omitting adipic dihydrazide. 4. Incubate 72 h, at 56°C, after which the samples can be stored at 4°C before purification. 5. Purify the samples by dialysis against water (the membranes have to be rinsed in water before use), in order to eliminate remaining adipic dihydrazide and salts. For a 500 mL sample, this is usually performed against 600 mL of water. For an optimum purification, the dialysis is done for 24 h and includes five changes of water. After that, samples (that can be stored at 4°C) are recovered in pre-weighed microtubes, concentrated (SpeedVac) and then freeze-dried. 6. Weigh the dry sample accurately to determine the total quantity of recovered oligosaccharide. 7. Solubilise the samples in PBS at 20 mg/mL. It can be stored at 4°C. 8. The determination of the linkage yield is performed as described in Subheading 3.3. 3.2.2. Functionalization with N-Hydroxysuccinimidyl-6(Pyrrole-yl)-Caproate (NHS-Pyrrole)

In a second step, the glycosylhydrazides react with NHS-Pyrrole, by an addition-elimination reaction between the activated ester of NHS-Pyrrole and the free NH2 function of the adipic dihydrazide arm. During this reaction, NHS is regenerated. 1. The reaction has to be performed in a 50% PBS/50% DMSO v/v buffer, in order to favour the solubility of NHS-Pyrrole without precipitating the oligosaccharide. 2. Prepare a 50  mM NHS-Pyrrole stock solution in DMSO. It can be stored at –20°C for several weeks. 3. Prepare the reacting solution by mixing 50 mL of 20 mg/mL oligosaccharide functionalized with adipic dihydrazide (solubilised in PBS), 30 mL of DMSO and 20 mL of 50 mM NHSPyrrole. The final concentrations are 10  mg/mL for the oligosaccharide and 10  mM for the NHS-Pyrrole, in a final volume of 100 mL. 4. Prepare negative controls by performing the reaction with oligosaccharide which has not been coupled with adipic dihydrazide. 5. Incubate 2 h at RT, in the dark to minimize pyrrole oxidation. 6. Complete to a 500 mL final volume with reacting buffer. 7. Purify the samples by dialysis against the 50% PBS/50% DMSO v/v buffer, during 4 h, in order to eliminate remaining NHS-Pyrrole. For a 500 mL sample, we use 125 mL of buffer.

76

J. Bartoli et al.

Then, dialyse against water, to eliminate PBS and DMSO. Use 600 mL of water for a 500 mL sample. For an optimum purification, the dialysis is performed during 24 h and the water is replaced five times. After that, samples are recovered in preweighed microtubes (it can be stored at 4°C from now), concentrated (Speed-Vac) and then freeze-dried. 8. Weigh the dry sample accurately to determine the total quantity of recovered oligosaccharide. 9. Solubilize the samples in water at 20 mg/mL. It can be stored at 4°C. The determination of the linkage yield is performed as described in Subheading 3.3. 3.3. Determination of the Linkage Yield

The TNBSA method is a colorimetric method allowing the quantification of primary amines and hydrazines. In a first step, it enables to determine the quantity of oligosaccharide coupled with adipic dihydrazide, which presents a free hydrazine function. In a second step, the method allows the estimation of pyrrolylated oligosaccharides content, by measuring the diminution of the hydrazine concentration (see Note 3). 1. Prepare a series of adipic dihydrazide (di-NH2 molecule) standards in the buffer 1: the solutions at 0–7.8–15.6–31.25– 62.5–125–250–500  mM are prepared by serial dilutions of a 10 mM adipic dihydrazide stock solution in water. The standard solutions have to be prepared just before use. 2. In a 96-well plate, dispense 10 mL of standard or sample (either oligosaccharide coupled with adipic dihydrazide, or coupled with pyrrole, or negative controls), with 40  mL of buffer 1 (dilution: 5). 3. Add in each well 10 mL of the TNBSA solution prepared by dilution at 1:10 of the stock solution in buffer 1, and incubate for 4 min. 4. Stop the colorimetric reaction by adding 100 mL of buffer 2 in each well. 5. Read the optical density at l = 450 nm with the Victor 1420 Multilabel Counter.

3.4. Oligosaccharides Immobilization on SPR Prisms: Electrospotting

All solutions are prepared extemporaneously. 1. Prepare a 20  mM pyrrole solution by diluting 10  mL of the 1  M stock solution with 490  mL of electrocopolymerization buffer. 2. Prepare the pyrrolylated oligosaccharide solutions at 3 different concentrations (for example, 100, 20, and 2 mM of pyrrolylated oligosaccharide, or 500, 100, and 10  mM of total oligosaccharide) by diluting the 20  mg/mL oligosaccharide

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

77

samples in the 20 mM pyrrole solution (see Note 4). The molar concentration of the pyrrolylated oligosaccharide solutions has been estimated as described in Subheading 3.3. The final concentration of free pyrrole must be 20 mM (see Note 5), in a 30 mL final volume (see Note 6). 3. Transfer the spotting solution in a 96-well plate. Be careful to ensure a good homogenization of the solution without introducing air bubbles. 4. The prism has to be pre-treated with a mixture of 70% H2SO4 and 30% H2O2 (v/v): incubate the prism in 20 mL of the solution, under ventilation, during 10 min. Care must be taken as this reaction is highly exothermic and reacts strongly with organic compounds. Rinse it thoroughly with water and dry it with an argon jet, before utilization. 5. Insert the plate in the microarrayer to carry out the spotting on the chip. It mainly consists in a ceramic needle (150  mm internal diameter) containing a platinum wire which is filled with the solution (6 nL) containing the pyrrolylated oligosaccharide to be grafted and which can move to a precise location on the chip. A 2.4 V electrical pulse for 0.1 s between the needle (counter electrode) and the gold surface of the chip (working electrode) induces the synthesis of the polypyrrole film and its deposition on the gold surface (Fig.  2). In-between each sample deposition and at the end of the cycle, the program includes a procedure where the needle is rinsed with water and subsequently dried.

Fig. 2. General scheme of oligosaccharides addressing on a glass prism coated with gold. The different pyrrolylated oligosaccharides and pyrrole monomer solutions are in the 96-well plate. The electrospotting is carried out on the gold surface via the needle containing the solution to be copolymerized.

78

J. Bartoli et al.

Kinetic Curves

Plasmon Curves

Buffer

Probe + target Probe

0.5

0.45 0.40

0.3

DR

0.2 0.1

Reflectivity

Reflectivity

0.4

0.0

Protein injection

0.35 0.30

DR

0.25 0.20

40

45 50 55 60 Incidence angle (deg.)

0.00

0

5

10

15

20

Time (min.)

Measurement angle

Fig. 3. Scheme of the Surface Plasmon Resonance Imaging (SPRi) system. The imaging system enables the visualisation in real time of molecular interactions on each plot. A typical reflectivity curve versus incidence angle is represented. The choice of the measurement angle (maximum of the slope) allows kinetic monitoring with optimal amplitude.

6. When all the spots are synthesized, disconnect the prism from the microarrayer, rinse with water and dry it perfectly with argon. In these conditions, the prism can be stored up to 6 months at 4°C. 3.5. SPRi Interactions Monitoring

1. The optical setup is described elsewhere (13) and constructed by Genoptics (Fig.  3). Briefly, the prism is inserted in the apparatus and connected to a flow cell inside the instrument which is hermetically sealed over the prism. A light source (650  nm) illuminates the prism and a 12-bit CCD camera monitors the changes of the reflectivity caused by the ligand– probe interactions, as grey-level contrasts. During interaction experiments, images are recorded at fixed intervals of time (0.2 s) and are analyzed on a PC computer with Genovision Software (Genoptics). The SPRi technique allows the obtention of sensorgrams, corresponding to the association and dissociation phases of the oligosaccharide-protein complexes (Fig.  4). Finally, the raw data are treated with an Excel program.

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

79

5.0

4.0

2.0

[Py-Heparin] =100 µM [Py-Heparin] =50 µM [Py-Heparin] =10 µM Control : pyrrole Control : gold

Regeneration

tion

3.0

Asso cia

% Reflectivity

Dissociation

1.0 Injection

0.0

-1.0

0

2

4

7

9

11

14

16

18

20

22

24

Time (min)

Fig. 4. Detection of ligand binding by SPRi: sensorgram obtained on five spots (Heparin at three concentrations of total oligosaccharide, and controls) upon injection of SDF-1 at 400 nM.

2. Insert the chip into the SPRi reader and start the apparatus. Work with freshly defrosted buffers. Be sure that they are at room temperature before starting. The flow rate of running solutions within the cell is 70 mL/min. Before starting kinetic measurements, one has to determine the working incidence angle, by drawing plasmon curves: the optimum angle is located at the inflexion point of a plasmon curve (see Fig. 3 and Note 9). The standard procedure for a SPRi experiment includes the following steps: Prism saturation (at the start) 1. Saturate the prism by injecting saturation buffer containing 1% BSA during 7 min. 2. Rinse the prism with washing buffer during 10 min. 3. Inject regeneration buffer containing 1 M NaCl during 7 min to remove remaining BSA. 4. Rinse with washing buffer and wait until the base line is stable. The chip is ready for sample injections. Sample injections 1. Inject the sample diluted in washing buffer during 7 min. 2. Rinse with washing buffer for at least 10  min to remove unbound molecules and observe the oligosaccharide-protein complex dissociation.

80

J. Bartoli et al.

3. Regenerate with 1 M NaCl during 7 min. If spots are properly regenerated, one can inject another sample, after a washing step. If proteins remain on some spots, one can use a denaturing buffer containing 1% SDS or 2 M Guanidine Hydrochloride. In this case, it is recommended to repeat the saturation step before the following injection. 4. Wash during 10 min. Ending procedure 1. At the end of the experiment, stop the running buffer, remove the prism and wash the system (tubing, flow cell, injection loop) with 1% SDS in water and then rinse with water. Rinse the prism with water and dry it perfectly with argon, store it at 4°C. 3.6. Application to Heparin–Cytokine Interactions

The biological application is carried out using glycosaminoglycans (GAG) as a model of oligosaccharides. GAGs are negatively charged polysaccharides that have been shown to bind and directly regulate the bioactivity of a variety of proteins, such as enzymes, growth factors and cytokines (14). Among them, Heparin (Hp) and Heparan Sulfate (HS) are complex and highly sulphated polymers. In order to illustrate the approach described above, we studied interactions between Heparin and a Heparin fragment of 8 monosaccharides (dp8) with stromal cell-derived factor-1a (SDF1-a) and interferon-g (IFN-g) (for another illustration, see ref. 15). The chemokine SDF1-1a is the natural ligand for CXC chemokine receptor 4 (CXCR4). This chemokine inhibits cells infection by human immunodeficiency virus (HIV) and interacts specifically with Hp and HS (16). IFN-g is a T cell secreted cytokine, centrally involved in the immune response, which is regulated upon binding to GAGs (17, 18). The binding domain of HS for the cytokine, contains two N-sulfated regions (8 monosaccharides) linked to each other by an internal N-acetylated domain (around 30 monosaccharide units) (19). Three parallel functionalizations were performed for each sample (A, B, C) in order to determine the reproducibility of the coupling reaction. Thus, the yields of the oligosaccharides pyrrolylation, determined as described in Subheading 3.3, were in the range of 15–25%, and were very similar for the same species. After functionalization, the pyrrolylated oligosaccharides were grafted at four different concentrations onto the prism by electropolymerization, and the chip was used to measure oligosaccharide–cytokine interactions by SPR imaging. The concentrations referred in this example are expressed as concentrations of total oligosaccharide. Free heparin or dp8, which have not been modified, were also grafted (see Fig. 5).

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

81

Fig. 5. Design of the 70 spots chip: copolymerizations were performed with a mix of 20 mM free pyrrole and various concentrations of oligosaccharide probes. Each condition was realized in duplicate. Six polypyrrole (PPy) spots were also immobilized, as negative controls spots.

The two proteins were successively injected at different concentrations in the cell. Each run was followed by a NaCl regeneration step. Representative association and dissociation curves obtained during cytokine injection are shown in Fig.  6. As anticipated, grafted oligosaccharide spots recognize the protein proportionally to their grafting concentration, and the interactions are oligosaccharide specific, with regard to the weak binding of the proteins upon gold surface and polypyrrole spots. As shown in Fig. 7 with an IFN-g injection, the signals obtained for pyrrolylated oligosaccharides resulting from different coupling reactions (A, B or C) are quite similar, which underlines the relative reproducibility of the functionalization reactions. Moreover, the non modified Heparin does not properly bind the cytokine, as the oligosaccharide cannot be grafted on the support. The signal intensity is also directly linked to the protein concentration injected, as shown in Fig. 8. By increasing the number of concentrations tested, equilibrium dissociation constant (Kd) and kinetics parameters (rate constants) can be obtained for each probe analyzed, allowing the direct comparison of molecular interactions for an important number of compounds. This approach can be applied for multiple studies of synthetic and natural oligosaccharides properties, making it a powerful tool in the glycobiology field.

82

J. Bartoli et al. Injection of IFNγ at 100 nM

Injection of SDF-1α at 400 nM

Injection of IFNγ at 100 nM on Heparin spots 4.5

10.0

4.0

9.0

3.5

8.0 7.0

% Reflectivity

% Reflectivity

11.0

6.0 5.0 4.0 3.0

3.0 2.5 2.0 1.5 1.0

2.0 1.0 0.0

Injection of SDF-1α at 400 nM on Heparin spots

0.5 0

2

4

6 8 Time (min)

10

12

14

0.0

0

2

4

6 Time (min)

8

10

12

[Py-Heparin] =100 µM

[Py-Heparin] =10 µM

Control : pyrrole

[Py-Heparin] =100 µM

[Py-Heparin] =10 µM

Control : pyrrole

[Py-Heparin] =50 µM

[Py-Heparin] =1 µM

Control : gold

[Py-Heparin] =50 µM

[Py-Heparin] =1 µM

Control : gold

Injection of SDF-1α at 400 nM on dp8 spots

Injection of IFNγ at 100 nM on dp8 spots

4.5

1.4

4.0

1.2 1.0

3.0

% Reflectivity

% Reflectivity

3.5

2.5 2.0 1.5 1.0

0.8 0.6 0.4 0.2

0.5 0

2

4

6 8 Time (min)

10

12

14

0.0

0

2

4

6 Time (min)

8

[Py-dp8] = 1 mM

[Py-dp8] = 100 µM

Control : pyrrole

[Py-dp8] = 1 mM

[Py-dp8] = 100 µM

[Py-dp8] = 500 µM

[Py-dp8] = 10 µM

Control : gold

[Py-dp8] = 500 µM

[Py-dp8] = 10 µM

10

12

Control : pyrrole Control : gold

Fig. 6. Injection of the cytokines on Heparin or dp8 spots immobilized at different concentrations: association and dissociation curves obtained for each oligosaccharide spot, and controls. At the top: differential images of the prism, at the end of the protein injection. The biological interactions cause reflectivity increment for the concerned spots.

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

83

Injection of IFN-γ at 100 nM on different Heparin spots

11.0 10.0 9.0

% Reflectivity

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 0

2

4

5

8

10

Time (min)

12

14

Py-Heparin A

Py-Heparin c

Control : pyrrole

Py-Heparin B

Free Heparin B

Control : gold

Fig. 7. Association and dissociation curves of IFNg injected at 100 nM with different heparin spots, grafted at the higher concentration (100 mM).

Injection of SDF-1α at three concentrations on Hp spots

4.5 4.0

% Reflectivity

% Reflectivity

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

0

2

4

6

8

10

12

Py-Hep Pyrrole

[SDF-1α] = 250 nM Py-Hep Pyrrole

11.0 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 0

2

4

6

8

10

12

14

Time (min)

Time (min) [SDF-1α] = 400 nM

Injection of IFN-γ at two concentrations on Hp spots

[SDF-1α] = 100 nM Py-Hep Pyrrole

[IFNγ] = 100 nM Py-Hep

[IFNγ] = 50 nM Py-Hep

Pyrrole

Pyrrole

Fig. 8. Association and dissociation curves of IFNg and SDF-1a injected at several concentrations with heparin spots, grafted at the higher concentration (100 mM).

4. Notes 1. The reaction rate is limited by the cycle opening in acidic medium (one can assume a 1% opened form presenting the aldehyde reactive function). It is important to notice that some oligosaccharides are not very stable in acidic medium. That’s why, different changes in the buffer composition, pH and reaction length should be performed to determine optimum conditions,

84

J. Bartoli et al.

leading to an efficient coupling and avoiding oligosaccharide degradation. 2. It is necessary to work with the same stock of buffers from an experiment to another, in order to minimize results variations due to the buffer composition. That is why it is recommended to prepare an important stock of buffers, to aliquote in 50-mL sterile tubes (for washing and regeneration buffer) and in 1.5  ml-microtubes (for saturation buffer), and to store it at –20°C. Thus, one can defrost the needed quantity of buffer for an experiment and store it temporarily at 4°C. The saturation buffer containing BSA cannot be stored more than one day at 4°C, otherwise bacteria contamination can occur. 3. The TNBSA method allows one to determine the molar ratio of oligosaccharides functionalized against the total amount of oligosaccharides. The latter is determined by weighing the dry sample after the freeze-dryer step. The molar concentration of oligosaccharide coupled with adipic dihydrazide is calculated as follows: (a) The neat optical density (OD) of the oligosaccharide bearing a primary amine (NH2) group is obtained by subtraction of the control OD (unmodified oligosaccharide) to the modified sample. (b) The OD of the adipic dihydrazide standards enable to draw a calibration curve and to establish a mathematical relation between the OD and the NH2 group concentration, taking into consideration that adipic dihydrazide standards contain two NH2 groups. This curve allows the determination of the functionalized oligosaccharide concentration from its neat OD. After the second coupling of oligosaccharides with NHS-Pyrrole, the method enables with the same approach, to determine the remaining oligosaccharides bearing a NH2 group. By subtracting the number of mole of oligosaccharide functionalized with adipic dihydrazide, introduced for the NHS-Pyrrole coupling reaction, one can calculate the number of mole of pyrrolylated oligosaccharides. 4. Different final pyrrolylated oligosaccharide concentrations (200 mM to 200 nM) have to be prepared in order to determine the concentration leading to the best reactivity with the tested ligand. Moreover, the coupling yield (between 10 and 50% usually) can differ from one sample to another, so it is recommended to work with a large range of concentrations. 5. Final 20mM free pyrrole concentration is essential for electropolymerisation processus. Depending on the wanted oligosaccharides concentration, corresponding volume needed may impose to prepare different free pyrrole solutions higher than 20mM.

5  Polypyrrole-Oligosaccharide Microarray for the Measurement…

85

6. The volume taken by the needle is in the nL range, so the 30  mL final volume is not crucial. However, we recommend preparing 15–30 mL solutions, since a significant evaporation can occur if the samples remain for a long time in the plate before being electrodeposited. 7. It is recommended to realize the spots at least in duplicates, to ensure the reliability of the results. These spots have to be spotted perpendicularly to the flow direction, to avoid potential ligand depletion between the first and second spot. 8. Control spots of polypyrrole and gold, as negative samples have to be performed, to quantify the background level. 9. Because of the considerable number of spots created on the gold surface and the possible variability of the used oligosaccharides, one have to realize a compromise by choosing the most suitable working angle for a majority of spots. 10. In order to verify the stability of the chip upon the several protein injections and SDS regenerations, it is better to inject a ligand of one grafted oligosaccharide, at the beginning of an experiment and then at regular intervals. The interaction intensities have to be very similar. The injection of the ligand at the beginning and at the end can be a sufficient control.

Acknowledgments We thanks H. Lortat-Jacob’s group at the “Institut de Biologie Structurale” – Grenoble for the preparation of recombinant IFNg and SDF-1a, and the “CREAB” group for their technical support with the microarrayer and SRPi apparatus. This work has been partially granted by the Carbinfec project from FUI and Lyonbiopôle. References 1. Fodor SP, Read JL, Pirrung MC, Stryer L, Lu AT, Solas D (1991) Light-directed, spatially addressable parallel chemical synthesis. Science 251(4995): 767–73 2. R. Frank (1992) Spot synthesis of peptides on membrane supports. Tetrahedron 48:9217–32 3. Ban L, Mrksich M (2008) On-chip synthesis and label-free assays of oligosaccharide arrays. Angew Chem Int Ed Engl 47(18):3396–9 4. Kiessling LL, Splain RA (2010) Chemical approaches to glycobiology. Annu Rev Biochem 79:619–53 5. Lepenies B, Yin J, Seeberger PH (2010) Applications of synthetic carbohydrates to

chemical biology. Curr Opin Chem Biol 14(3):404–11 6. Weijers CA, Franssen MC, Visser GM (2008) Glycosyltransferase-catalyzed synthesis of bioactive oligosaccharides. Biotechnol Adv 26(5):436–56 7. Angenendt P, Glökler J, Murphy D, Lehrach H, Cahill DJ (2002) Toward optimized antibody microarrays: a comparison of current microarray support materials. Anal Biochem 309(2):253–60 8. Livache T, Guedon P, Brakha C, Roget A, Levy Y and Bidan G (2001) Polypyrrole electrospotting for the construction of oligonucleotide arrays compatible with a surface plasmon

86

J. Bartoli et al.

r­ esonance hybridization detection. Synth. Met 121 (2–3):1443–1444 9. Sarrazin S, Bonnaffé D, Lubineau A, LortatJacob H (2005) Heparan sulfate mimicry: a synthetic glycoconjugate that recognizes the heparin binding domain of interferon-gamma inhibits the cytokine activity. J Biol. Chem 280:37558–64 10. Laguri C, Sadir R, Rueda P, Baleux F, Gans P, Arenzana-Seisdedos F and Lortat-Jacob H (2007) The novel CXCL12g isoform encodes an unstructured cationic domain which regulates bioactivity and interaction with both glycosaminoglycans and CXCR4. PLoS One 2, e1110 11. Jirkowski I. and Baudy R (1981) A facile large scale preparation of 1H-pyrrole-1-ethanamine and syntheses of substituted pyrrolo[1,2-a] pyrazines and hydro derivatives thereof. Synthesis 481–483 12. Emiliano Gemma, Odile Meyer, Dušan Uhrín and Alison N. Hulme (2008) Enabling methodology for the end functionalisation of glycosaminoglycan oligosaccharides. Mol. BioSyst 4:481–95 13. Guedon P, Livache T, Martin F, Lesbre F, Roget A, Bidan G, Levy Y (2000) Characterization and optimization of a realtime, parallel, label-free, polypyrrole-based

DNA sensor by surface plasmon resonance imaging. Anal Chem 72:6003–9 14. Capila I, Linhardt RJ (2002) Heparin-protein interactions. Angew Chem Int Ed Engl 41:391–412 15. Mercey E, Sadir R, Maillart E, Roget A, Baleux F, Lortat-Jacob H, Livache T (2008) Polypyrrole Oligosaccharide Array and Surface Plasmon Resonance Imaging for the Measurement of Glycosaminoglycan Binding Interactions. Anal Chem 80:3476–82 16. Sadir R, Baleux F, Grosdidier A, Imberty A, Lortat-Jacob H (2001) Characterization of the stromal cell-derived factor-1alpha-heparin complex. J Biol Chem 276:8288–96 17. Sadir R, Forest E, Lortat-Jacob H (1998) The heparan sulfate binding sequence of interferongamma increased the on rate of the interferongamma-interferon-gamma receptor complex formation. J Biol Chem 273:10919–25 18. Lortat-Jacob H, Brisson C, Guerret S, Morel G (1996) Non-receptor-mediated tissue localization of human interferon-gamma: role of heparan sulfate/heparin-like molecules. Cytokine 8:557–66 19. Lortat-Jacob H, Turnbull JE, Grimaud JA (1995) Molecular organization of the interferon-gamma-binding domain in heparan sulfate. Biochem J 310:497–505

Chapter 6 Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis Fang Cheng and Daniel M. Ratner Abstract Over the past few decades, carbohydrates (glycans) have received growing attention for their many roles in biological systems, including pathogenesis, receptor-ligand interactions, and cell signaling. To unravel the biology of this important category of biomolecules, a host of new tools have been developed for glycomics investigation. At the forefront is the carbohydrate microarray, developed to immobilize functional glycans on a solid substrate to rapidly screen a variety of potential binding partners (carbohydrates, proteins, nucleic acids, cells, and viruses). The essential role played by surface modification on glycan microarray performance requires new methods to rigorously characterize glycan surface chemistries. Due to their highly reproducible nature and well-studied properties, self-assembled monolayers (SAMs) on gold are powerful models for presenting glycans on a solid substrate, engineering biomimetic microenvironments and exploring the bioactivity of immobilized carbohydrates via surface plasmon resonance (SPR). However, it can be challenging to prepare high quality glycosylated SAMs (glyco-SAMs) that retain their biological function following surface immobilization. Herein, a selection of versatile methods for the preparation of glyco-SAMs using natural and chemically modified glycans is described. This chapter will highlight the following three immobilization techniques: (1) direct self assembly using thiolated glycosides onto gold, (2) tethering aminated glycosides onto ­amine-reactive SAMs, and (3) conjugating natural glycan onto divinyl sulfone-activated SAMs. Key words: Surface plasmon resonance, Self-assembled monolayer, Glyco-SAM, Thiolated glycoside, Divinyl sulfone

1. Introduction Surface-immobilized carbohydrates can be used to explore glycanmediated recognition and binding events, e.g. pathogen and toxin detection, (1, 2) cell adhesion (3), and characterizing the carbohydrate-specificity of novel carbohydrate-binding proteins (lectins) (4). However, displaying bioactive mono- and oligosaccharides on

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_6, © Springer Science+Business Media, LLC 2012

87

88

F. Cheng and D.M. Ratner

a surface often requires significant chemical derivation prior to immobilization. Functional groups, such as thiols (5), amines (6), fluorocarbons (7), or azides (8), must typically be introduced into a chemically modified glycan to be paired with surfaces bearing compatible chemistries (maleimide, amine-reactive activated ester, hydrophobic, or alkyne). These linking strategies represent a highly rational and engineered approach to surface modification, but can also become a significant synthetic burden to the user. Carbohydrate synthesis frequently requires multiple synthetic steps, complex protecting group manipulations, and deprotection strategies; (9) introducing linkers for surface immobilization only increases the difficulty of said synthesis. Due to the lack of a universal synthetic strategy for producing linker-modified glycans, the following describes the preparation of glyco-SAMs based upon three complementary immobilization methods: thiolated glycosides on gold, aminated glycans on amine reactive SAMs, and free reducing sugar on vinyl sulfone-activated SAMs (Scheme 1). Whether the glycans are prepared synthetically to include a thiolated linker, functionalized via simple amination chemistries, or isolated from natural sources, these techniques permit the facile preparation of bioactive monolayers presenting ­glycans for biological assays, biosensing or surface analysis applications.

2. Materials 2.1. Preparation of Au Substrates

Titanium (2  nm) and gold (45  nm) films sequentially deposited onto cleaned silicon wafers (Silicon Valley Microelectronics, San Jose, CA) and SF-10 glass (SCHOTT Glass Technology, Duryea, PA) for surface analysis and SPRi, respectively. Metal films prepared via electron beam evaporation at the Washington Technology Center (Seattle, WA) (see Note 1).

2.2. Preparation of Solutions

1. Thiolated mannoside: 4–10  mM thiolated mannoside (5) in ultrapure water. Store in single-use aliquots at −20°C and dilute at 0.1–1 mM prior to daily use (see Note 2). 2. 11-Mercaptoundecanol: 10  mM 11-Mercaptoundecanol powder (Sigma-Aldrich, St. Louis, MO) in ethanol (200 proof, USP) (Decon Labs, King of Prussia, PA). Store in single-use aliquots at −20°C and dilute at 1 mM prior to daily use (see Note 3). 3. HEPES buffer (10×): 0.1  M HEPES, 1.5  M NaCl, 10  mM Ca2+ and Mn2+, pH 7.4. Store at room temperature. 4. PBS buffer (10×): 27  mM KCl, 1.37  M NaCl, 0.1  M phosphate, pH 7.4. Store at room temperature.

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis

89

a HO

HO

O

O O

HO Au

O

O O

SH

EtOH, H2O or EtOH:H2O

b

Au

HO

O O

O O

O

HO

OR

O

NH

O

NH2

O

O

O

O

O PBS or Carbonate

O

S

S

S

O

S

Au

Au HO

R = NHS or TFP

O

c

R O

OH

S

OH

OH DVS

5 S Au

5 pH 11

S

O O OH

O

5 S

Au

S

5 pH 10

S

S

O

O

5 S

5

Au R = -OH, -NH2 Linker-NH2 or Linker-SH

Scheme 1. Glycan immobilization on to gold or modified gold substrates. (a) Direct self-assembly using thiolated glycosides on to a gold surface. (b) Tethering aminated glycans onto amine-reactive surfaces. (c) Conjugating natural or ­modified glycans onto hydroxyl terminated SAMs via divinyl sulfone chemistry.

5. Carbonate buffer: 0.5 M Na2CO3, pH 10 or pH 11. Store at room temperature. 6. Surface plasmon resonance imaging (SPRi) running buffers: Filter and degass via lab vacuum (4 torr) for at least 30 min prior to daily use (see Note 4). 7. BSA-T buffer (10×): 10  mg/mL Bovine serum albumin ­powder (Sigma-Aldrich, St. Louis, MO), 0.05% Tween-20

90

F. Cheng and D.M. Ratner

(Sigma-Aldrich, St. Louis, MO), pH 7.4. Filter and store at −20°C and dilute in 1× PBS or 1× HEPES for daily use. 8. Con A-HRP (10×): 2,500 ng/mL Con A horseradish peroxidase conjugate (EY Laboratories, San Mateo, CA), 10× BSA-T HEPES, pH 7.4. Store at −20°C, and dilute in 1× HEPES for daily use. 9. Con A (10×): 10 mg/mL Concanavalin A (MP Biomedicals, Solon, OH), 10× HEPES, pH 7.4. Store at −20°C, and dilute in 1× HEPES for daily use. 10. RCA120-HRP (10×): 2,500 ng/mL Ricin horseradish peroxidase conjugate (EY Laboratories, San Mateo, CA), 1× BSA-T PBS, pH 7.4. Store at −20°C, dialyze overnight against 1× PBS, and dilute in 1× PBS for daily use. 11. RNase A: 10  mg RNase A (New England Biolabs, Ipswich, MA), 10× PBS, pH 7.4. Store at −20°C. 12. RNase B: 10  mg RNase B (New England Biolabs, Ipswich, MA), 10× PBS, pH 7.4. Store at −20°C. 13. Divinyl sulfone (DVS): 10% v/v divinyl sulfone (Sigma-Aldrich, St. Louis, MO), 0.5 M sodium carbonate buffer, pH 11. Mix vigorously prior to daily use. 14. Urea: 8 M urea (Sigma-Aldrich, St. Louis, MO). Sonicate and store at room temperature. 15. Glycine: 10 mM glycine (Sigma-Aldrich, St. Louis, MO), pH 2. Filter and store at room temperature. 2.3. XPS Analysis

Sample X-ray photoelectron spectroscopy (XPS) composition data was acquired on a Kratos AXIS Ultra DLD instrument equipped with a monochromatic Al-Ka X-ray source (hn = 1,486.6 eV). All the samples were grounded and data were obtained at 0° take-off angle in the hybrid mode. The take-off angle is defined as the angle between the sample surface normal and the axis of the XPS analyzer lens. Compositional survey and detailed scans (N 1s, O 1s, and S 2p) were acquired using a pass energy of 80  eV. Highresolution scans (C 1s) were acquired using a pass energy of 20 eV. Data analysis was performed with the CasaXPS software (Casa Software Ltd.).

2.4. ToF-SIMS Analysis

Sample time-of-flight secondary ion mass spectrometry (ToFSIMS) spectra was acquired using an ION-TOF TOF.SIMS 5–100 system (ION-TOF GmbH, Münster, Germany). Positive spectra were recorded by rastering a pulsed 25 keV Bi3+ primary ion source over a (100 × 100) mm2 area, while maintaining the total ion dose below the static limit (i.e. 1012  ions/cm2). The mass resolution (m/Dm) for the positive ion spectra was typically 6,000 at m/z = 27 for all spectra, which were mass calibrated using the CH3+, C2H3+, C3H7+, and AuC2H4+ peaks.

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis

2.5. SPRi

91

Sample carbohydrate-mediated protein binding was performed on a surface plasmon resonance imaging system (SPRi), SPRimagerII (GWC Technologies). The SPRimagerII was operated at room temperature using a standard flow cell and a peristaltic pump (BioRad-EconoPump) at 100  mL/min. All surfaces were passivated with BSA-T for 30 min and equilibrated in an appropriate running buffer prior to protein binding. Data acquisition consisted of the averaging of 30 images over a short duration to create an average image. The SPR signal is measured via the average pixel intensity, which can be converted to normalized percentages change in reflectivity according to GWC protocol. Urea (8 M) or glycine (10  mM, pH 2.0) was used to strip bound protein and regenerate the array surfaces. For sensorgram acquisition, a 500 mm × 500 mm region-of-interest (ROI) was selected. For visual clarity, the contrast and brightness of SPR difference images were adjusted by ImageJ software (U.S. National Institutes of Health, Bethesda, MD).

3. Methods Preparing high quality self-assembled monolayers (SAMs) bearing carbohydrates is a prerequisite for exploring the bioactivity and surface chemistry of surface resident glycans. Ensuring the quality of glyco-SAMs is complicated by the cleanness of the substrate, purity of reagents, intrinsic instability of SAMs, and adventitious contamination originating from self-assembly and storage. Therefore, to produce reliable and reproducible glyco-SAMs for analysis, it is necessary to verify surface quality, track surface contaminants and their orgins, establish SAM stability, and examine the compositional change following surface treatment. XPS (10) and ToF-SIMS (11) are routinely utilized to obtain surface information at a molecular level, e.g. atomic composition, the chemical environment of elements at the interface, and surface species relevant to the immobilized biomolecules (Figs. 1 and 2) (12, 13). Ultimately, it is necessary to determine whether glycan presented on the surface retain their bioactivity. This is best accomplished via a biological assay using a natural binding partner, such as a lectin. Exploiting the sensitivity of the enzyme-linked lectin assay, ELLA (14) is a common and low cost method that derives its origins from the enzyme-linked immunosorbent assay (ELISA). As a label-free method, SPR is routinely employed to directly probe carbohydrate–protein interactions at the interface, providing accurate structure/function information on binding specificity and kinetics (Fig. 3) (15).

92

F. Cheng and D.M. Ratner

a

b

c

Fig. 1. XPS characterization of a glyco-SAMs formed using direct self-assembly of thiolated mannosides: (a) XPS survey scan, (b) detail sulfur 2p scan, and (c) high resolution carbon 1 s scan, indicating the high quality of SAMs bearing mannose on Au. (Note: sulfur 2p3/2 at 162.0 eV and C 1 s at 288.0 eV are assigned to bound thiol and mannose acetal species, respectively).

3.1. Preparing Glyco-SAMs 3.1.1. Direct Self-Assembly Using Thiolated Glycosides

Molecular self-assembly of thiols onto gold is arguably one of the most well studied methods for surface modification (16). Thiolated biomolecules, including carbohydrates, can be used to produce highly uniform and reproducible SAMs and are ideal as model surfaces for surface analysis and bioassays. In addition to the ease with which thiol-based SAMs can be assembled onto gold surfaces, their composition can be controlled through mixed self-assembly (10, 15). While no single chemistry has been described for the generalized

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis a

b 1.5x105

9.0x104

C2H5O

+ +

1.0x105

Positive ion counts

Positive ion counts

93

5.0x104

C3H5O2

6.0x104

3.0x104

0.0

0.0 44.5

45.0

m/z

45.5

46.0

72.5

73.0

m/z

73.5

74.0

Fig. 2. Representative positive ion ToF-SIMS of a glyco-SAM composed of a thiolated mannoside assembled on to gold. Fragments associated with both the linker chemistries and immobilized glycan can be observed by this method; (a) oligo(ethylene) glycol fragment C2H5O+ (m/z 45) and (b) pyranose (sugar) fragment C3H5O2+ (m/z 73).

synthesis of thiolated glycosides for surface modification, a wide variety of published methods describe the synthesis of thiol-bearing linkers for the assembly of modified carbohydrates onto gold surfaces or maleimide-activated substrates (5, 15, 17, 18). 1. Self-assembly of thiolated glycosides to form a uniform glycoSAM is achieved by immersing fresh gold substrate in a 0.1–1 mM solution of the thiolated glycan for 2–12 h at ambient temperature (incubations longer than 2 h do not significantly affect the quality of the glyco-SAM). Assembly may be performed from thiol solutions in water, ethanol or an ethanolwater mix, in accordance with the solubility of the thiolated glycoside. During self-assembly, the sample should be stored in the dark, and kept under an inert atmosphere, if possible (see Notes 2 and 3). 2. Following self-assembly, the modified gold substrate must be rinsed thoroughly. This can be achieved by dipping the ­sample into a vigorously stirring bath of fresh ultrapure water ­(50–100 ml) for 1 min (see Note 5). 3. The rinsed glyco-SAMs should be gently dried by a stream of argon and stored in the dark under an inert atmosphere (see Note 6). 3.1.2. Tethering Aminated Glycans onto Sams Bearing Activated Esters

While thiol-modified glycosides are ideal for direct self-assembly onto gold, the synthetic complexity of their preparation often precludes their use by all but the most highly specialized synthetic laboratories. As a less synthetically burdensome alternative,

94

F. Cheng and D.M. Ratner

a

b

c

d

Fig. 3. SPRi reflectivity image and the corresponding sensorgrams on a spotted glyco-SAM array composed of thiolated mannose, galactose and the glycoprotein RNase B. The bioactivity of immobilized glycans were verified via lectin binding with Con A (500 nM). The array includes (a) thiolated mannose, (b) the mannosylated glycoprotein RNase B and (c) thiolated galactose.

a­ minated glycans can easily be prepared by a variety of simple methods from natural isolated reducing sugars (4, 19–21). Glycans bearing a nucleophilic amine are ideal partners for amine reactive chemistries. Utilizing modified surfaces bearing activated ester functionality, it is possible to covalently couple aminated glycans directly to a surface via an amide bond. This provides a facile route for the synthesis of stable glycosylated surfaces, while minimizing the synthetic complexity of the precursors. 1. Amine-reactive surfaces are prepared by assembling SAMs bearing activated esters, e.g. hydroxysuccinimide or tetrafluorophenyl (11, 22). To prepare the activated surface, fresh Au

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis

95

substrates should be immersed in 1 mM ethanolic solution of the amine reactive disulfide for 2  h at room temperature to form the precursor SAMs for bioconjugation. 2. After SAM formation, rinse the surfaces in a stirring ethanol bath (50 ml) for 1 min (see Note 5). 3. Gently dry the cleaned surfaces by a stream of argon and store in the dark under an inert atmosphere (see Note 6). 4. The amine-reactive surfaces can be modified by printing or covering the surface with 0.1–1 mM aminated glycan (23) (pH 7.4) for hydroxysuccinimide ester or pH 10 carbonate for tetrafluorophenyl ester for 24 h at room temperature. 5. To quench unreacted active ester functionality, immerse the glycosylated substrate in BSA-T for 30 min, thoroughly rinse in a stirring water bath for 1 min (see Note 5), and dry by a stream of argon prior to storage in the dark under an inert atmosphere. 3.1.3. Conjugating Isolated Glycans onto HydroxylTerminated Sams via Divinyl Sulfone Chemistries

One of the ultimate goals for carbohydrate surface modification is a universal method for the covalent conjugation of glycans that does not require chemical synthesis of the precursor sugar species. In certain cases, the difficultly of carbohydrate synthesis, or lack of structural identification, prohibits synthetic access to the prerequisite glycans. In other instances, it is desirable to avoid repeated synthesis of glycosylated targets that have been prepared previously for alternative array surface chemistries (e.g. thiolated vs. aminated glycans). As an alternative, divinyl sulfone (DVS) represents a unique method for the covalent conjugation of thiolated-, aminated-, and unmodified sugars and proteins through the modification of a hydroxyl-terminated SAM or polymer. Upon DVS activation, hydroxyl-bearing surfaces display the highly electrophilic vinyl sulfone group, which will react with glycans bearing amines, thiols and hydroxyls under slightly basic conditions via a Michael addition (1,4-addition) (24–26). 1. Hydroxyl-terminated SAMs can be prepared by immersing fresh Au substrates in 1 mM 11-mercaptoundecanol for 2 h at room temperature to form a uniform SAM. 2. Dip surfaces in a stirring ethanol bath (50 ml) for 1 min and dry under a stream of argon for 1 min (see Note 5). 3. Immerse the cleaned hydroxyl-bearing surface in a solution of DVS (10% v/v, 0.5 M carbonate buffer, pH 11) with vigorous shaking at room temperature for 1 h. 4. The DVS-treated surface should be thoroughly rinsed with a stream of water, dried under a stream of argon, and stored in the dark at 4°C.

96

F. Cheng and D.M. Ratner

5. Functionalize the DVS activated surface by printing or covering with a solution of free reducing sugar (0.5–20% w/v, 0.5 M carbonate buffer, pH 10) at ambient temperature for 16–24 h. Thiolated and aminated glycosides should be conjugated at 0.1–1 mM using pH 7.4 PBS. 6. Quench the printed/functionalized surface for 30  m with BSA-T, thoroughly rinse with water, dry under a stream of argon for 1 min and store in the dark at 4°C. 3.2. CarbohydrateMediated Protein Binding 3.2.1. Enzyme-Linked Lectin Assay

ELLA is a popular qualitative probe for interrogating carbohydratemodified surfaces based on a modified enzyme-linked immunosorbent assay (ELISA). Utilizing commercial peroxidase or alkaline phosphatase-conjugated lectins, ELLAs provide users with an accurate method to compare the bioactivity of carbohydrates on various glycosylated substrates. This method is particularly attractive for general use, as it requires little to no specialized equipment for array fabrication or analysis. The following ELLA-based assay for the bioavailability of glyco-SAMs is based upon the quantification of HRP-modified lectins via a commercial fluorogenic peroxidase substrate. 1. The QuantaBlu fluorogenic peroxidase substrate kit (Thermo Fisher Scientific, Rockford, IL) must be equilibrated to room temperature for 30 min and the QuantaBlu working solution prepared in accordance with the manufacturer’s specifications. 2. Prior to ELLA, glyco-SAMs must be immersed in the appropriate working buffer (PBS or HEPES, depending on the lectin) and incubated for 30 min at room temperature to rehydrate the glycosylated surface. Small gold-coated chips can be placed at an angle in a 24-well plate and fully immersed by 1.5  ml buffer. Following incubation, the buffer can be removed by aspiration (see Note 7). 3. To passivate the glyco-SAM substrates against nonspecific protein interactions, gently add BSA-T PBS or HEPES (1.5 ml) to each well and incubate for 1 h at room temperature. Remove the BSA-T buffer and rinse both sides of the glyco-SAM ­sample three times with BSA-T (see Note 7). 4. Carefully remove the glycosylated substrate from the 24-well plate using forceps and clean by flushing three times with PBS or HEPES working buffer on both sides of the chip. Place cleaned chips in a fresh 24-well plate, resting flat on the surface. 5. Add the lectin-HRP solution (250 ng/mL, 1 ml) to each well, and incubate on a shaking platform at room temperature for 1 h (see Note 8). 6. Following incubation, remove the lectin-HRP solution by aspiration and carefully rinse the chip three times with BSA-T

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis

97

on both the gold-coated and unfunctionalized side to remove any unbound lectin-HRP. 7. Carefully remove the glyco-SAM substrates from the 24-well plate using clean forceps and place flat on the surface of a fresh 24-well plate. 8. Add the QuantaBlu working solution (1.0  mL) to each well and incubate on shaking platform at room temperature, in the dark, for 30 min. 9. Add the QuantaBlu stop solution (100 mL) to each well and mix briefly. 10. Transfer the fully developed QuantBlu solution (50 mL) to a 96-well plate and measure the Relative Fluorescence Units (RFU) (325 nm ex, and 420 nm em), in accordance with the manufacturer’s specifications. 3.2.2. SPRi

Surface plasmon resonance imaging (SPRi) is a powerful method for interrogating complex arrays of immobilized glycans in a highthroughput and label-free fashion. The versatility of SPRi makes it ideal for validating glyco-SAM bioactivity and optimizing surface modification based on the response to protein binding. Glycan arrays for SPRi can be easily prepared using any of the aforementioned glyco-SAM chemistries, including direct assembly of thiolated glycans on a gold surface, coupling of aminated glycans to an activated ester SAM or conjugation of glycan (regardless of source) to a DVS-activated substrate on gold. Arrays can be printed using microcontact printing, (27) pin-based printing, (28) inkjet printing, (4) or prepared by hand (10). The following protocol is based upon a silicone gasket cut to form individual wells and does not require specialized array printing instrumentation. 1. A 1 cm × 1 cm sheet of culture-well silicone (Grace Bio-Labs, Bend, OR) can be punched using a flat needle (φ 2 mm) to create an array of wells that can be overlaid on a gold surface or activated SAM substrate by gently pressing the pre-punched sheet onto the SPRi chip. 2. Spot ~2  mL of modified or unmodified glycan solution (in accordance with the specific immobilization chemistries) into the silicone well on the SPRi chip and incubate in a 100% relative humidity chamber overnight (see Note 9). 3. Following overnight incubation, remove the SPR chip from the humidity chamber and flush the array spots with water in a bottom-to-top sequence and gently remove the silicone mask. Following removal of the mask, the surface should be briefly rinsed with water (see Note 10). 4. Dry the SPR chip in a stream of argon and immerse the chip in BSA-T on a shaking platform at room temperature for 1 h to

98

F. Cheng and D.M. Ratner

passivate unfunctionalized areas on the array surface. The BSA blocking step will make it easier to identify glycan-modified spots on the array by increasing the SPRi contrast between the printed and unprinted regions on the chip. This will subsequently facilitate identification of printed spots during regionof-interest (ROI) selection on the SPR instrument. 5. Prior to installation in the SPRi instrument, the functionalized SPR must be rinsed in a stirring water bath, dried in a stream of argon, and positioned in the instrument according to the manufacturer’s recommendations (see Note 11). 6. Running buffer, protein dilutions, and protein stripping solutions (8.0 M urea or 10 mM glycine, pH 2.0) can be applied to the SPR chip to measure SPR response according to a preestablished binding recipe. Care must be taken to avoid caustic solutions or oxidative damage to glyco-SAMs. However, if treated gently, chips can be used in repeated binding and regeneration cycles (see Notes 12 and 13). 7. Following an SPR experiment, the regenerated, rinsed and dried glyco-SAMs can be saved and stored at 4°C for additional SPR experiments or surface analysis (see Note 14).

4. Notes 1. High quality surfaces require clean gold substrates. To ensure the purity of the gold surface, storage conditions must be carefully controlled – in particular, fresh gold substrates must be stored in a polydimethylsiloxane (PDMS)-free environment. Gold can be easily contaminated by PDMS, sulfur, and iodine, even following only brief exposure to the environment in a chemical hood or contaminated lab bench. Gold substrates should be packed/unpacked in a contaminant-free laminar flow hood and stored under an inert atmosphere in the dark. 2. Unless stated otherwise, all aqueous solutions need to be prepared in ultrapure water (18.2  MΩ  cm). In the event where disulfide formation needs to be avoided, 0.5 stoichiometric equivalents of tris(2-carboxyethyl)phosphine (TCEP) can be added to the aqueous thiol solution. This amount of TCEP has not been observed to foul SAMs, according to XPS analysis. 3. Unless otherwise stated, all ethanolic solutions are prepared using 200 proof ethanol (USP test) established to be free from copper contamination. The existence of copper in ethanol affects SAM formation and packing density. To quickly ascertain the purity of the ethanol in question, elemental composition of a hydrophilic SAM prepared from the ethanol can be performed by XPS to identify the presence of copper; for this application SAMs bearing oligo(ethylene) glycol on Au can be used.

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis

99

4. Glass vials with occasional magnetic stirring are recommended for degassing. Glassware should be checked for scratches and cracks to mitigate the implosion hazard. 5. It is essential to thoroughly rinse the surface of unbound molecular species, while minimizing exposure to heat, light, and oxygen. A stirring water bath, rather than sonication, should be used to remove free thiol/disulfide from the surface. The water bath avoids unnecessary heating of the sample and potential damage to the glyco-SAM. 6. Unless stated otherwise, all glyco-SAMs must be stored at room temperature in a dark and oxygen-free environment. Petri dishes in glass jars backfilled with argon and sealed with parafilm can reduce airborne contamination and oxidation of the thiol-based SAM. Alternatively, samples can be stored in a vacuum desiccator. However, care must be exercised to avoid the introduction of contaminating species when venting the desiccator. 7. It is essential that the entire gold-coated chip be exposed to the incubation and blocking buffers. By positioning the chips at an angle, propped against the side-wall of the 24-well plate, it is possible to ensure that both the coated and uncoated side are blocked by BSA-T to reduce nonspecific adsorption lectin HRP conjugates to the substrate (bottom side of the chip) and the gold-coated top. Blocking both sides of the chip has been observed to significantly reduce background lectin-binding to the substrate. 8. Lectin-HRP stored at −20°C gradually loses its peroxidase activity, particularly after a number of freeze-thaw cycles. To verify the HRP activity, a small amount of lectin-HRP should be added to the QuantaBlu working solution and mixed briefly. In the event that the fluorescent product is not observed, the lectin-HRP stock should be discarded in favor of a fresh solution. 9. The specific concentration of glycans used for surface modification (array fabrication) is a function of the immobilization chemistry. On DVS-activated surfaces, Unmodified reducing sugars are employed as 0.5–20% w/v solutions, pH 10, and aminated and thiolated glycosides are used at 0.1–1 mM, pH 10 and 0.1–1 mM, pH 7.4, respectively. On bare Au and amine-reactive surfaces thiolated glycosides and aminated glycosides are used at 0.1–1 mM. 10. In the event that glycoprotein or protein spots (e.g. RNase B and RNase A) are used in the SPRi array, particular care must be exercised to avoid contamination of the surface during rinsing. Protein spots should be placed on the edge of the array, and the surface rinsed so that unbound protein flows off the chip, and not across the patterned array.

100

F. Cheng and D.M. Ratner

11. This rinsing and drying step is essential to ensure the optical quality of the chip by removing excess BSA-T blocking solution and buffer salts. If the chip is not sufficiently rinsed, remaining protein or salts will interfere with refractive index matching to the SPR prism. 12. All buffers and solutions should be degassed prior to introduction to the SPR instrument to reduce the appearance of bubbles during binding experiments. 13. During SPR binding experiments, it is critical to avoid the introduction of air bubbles prior to rinsing the surface with water. Air bubbles have been found empirically to damage glyco-SAMs, as observed by an altered SPRi response. In addition, if bubbles are passed over a surface following the specific capture of proteins, surface-bound proteins can become resistant to stripping by urea or glycine, and the glyco-SAM cannot be completely regenerated. 14. To preserve the quality of the glyco-SAMs, the SPR chip should be thoroughly stripped of bound protein by 8.0  M urea or 10 mM glycine (pH 2) and rinsed with water to remove buffering salts. Following removal from the SPR instrument, the SPR chip (glyco-SAM) must be rinsed with water, dried in a stream of argon, and stored in the dark at 4°C under an inert atmosphere.

Acknowledgements The authors gratefully acknowledge support from the Washington Research Foundation, the University of Washington Royalty Research Fund, and the Department of Bioengineering, and NESAC/BIO (NIH Grant P41 EB002027). References 1. Shi, J., Yang, T., Kataoka, S., Zhang, Y., Diaz, A. J., and Cremer, P. S. (2007) GM1 clustering inhibits cholera toxin binding in supported phospholipid membranes. J. Am. Chem. Soc. 129, 5954–5961. 2. Disney, M., and Seeberger, P. (2004) The use of carbohydrate microarrays to study carbohydrate­cell interactions and to detect pathogens. Chemistry & Biology 11, 1701–1707. 3. Esaki, K., Yokota, S., Egusa, S., Okutani, Y., Ogawa, Y., Kitaoka, T., Goto, M., and Wariishi, H. (2009) Preparation of lactose-modified cellulose films by a nonaqueous enzymatic reac-

tion and their biofunctional characteristics as a scaffold for cell culture. Biomacromolecules 10, 1265–1269. 4. Song, X., Lasanajak, Y., Xia, B., Smith, D., and Cummings, R. (2009) Fluorescent glycosylamides produced by microscale derivatization of free glycans for natural glycan microarrays. ACS Chem. Biol. 4, 741–750. 5. Ratner, D., Adams, E., Su, J., O’Keefe, B., Mrksich, M., and Seeberger, P. (2004) Probing protein-carbohydrate interactions with microarrays of synthetic oligosaccharides. Chembiochem 5, 379–383.

6  Glycosylated Self-Assembled Monolayers for Arrays and Surface Analysis 6. Song, X., Xia, B., Stowell, S. R., Lasanajak, Y., Smith, D. F., and Cummings, R. D. (2009) Novel fluorescent glycan microarray strategy reveals ligands for galectins. Chemistry & Biology 16, 36–47. 7. Ko, K., Jaipuri, F., and Pohl, N. (2005) Fluorous-based carbohydrate microarrays. J. Am. Chem. Soc. 127, 13162–13163. 8. Zhang, Y., Luo, S., Tang, Y., Yu, L., Hou, K.-Y., Cheng, J.-P., Zeng, X., and Wang, P. G. (2006) Carbohydrate-protein interactions by “clicked” carbohydrate self-assembled monolayers. Anal. Chem. 78, 2001–2008. 9. Boons, G., and Lindhorst, T. (1998) Carbohydrate chemistry, Blackie Academic & Professional. 10. Dhayal, M., and Ratner, D. A. (2009) XPS and SPR Analysis of Glycoarray Surface Density. Langmuir 25, 2181–2187. 11. Cheng, F., Gamble, L. J., Grainger, D. W., and Castner, D. G. (2007) X-ray photoelectron spectroscopy, time-of-flight secondary ion mass spectrometry, and principal component analysis of the hydrolysis, regeneration, and reactivity of N-hydroxysuccinimide-containing organic thin films. Anal. Chem. 79, 8781–8788. 12. Castner, D., Hinds, K., and Grainger, D. (1996) X-ray photoelectron spectroscopy sulfur 2p study of organic thiol and disulfide binding interactions with gold surfaces. Langmuir 12, 5083–5086. 13. Lee, C. Y., Canavan, H. E., Gamble, L. J., and Castner, D. G. (2005) Evidence of impurities in thiolated single-stranded DNA oligomers and their effect on DNA self-assembly on gold. Langmuir 21, 5134–5141. 14. Shankar, S. P., Chen, I. I., Keselowsky, B. G., García, A. J., and Babensee, J. E. (2010) Profiles of carbohydrate ligands associated with adsorbed proteins on self-assembled monolayers of defined chemistries. J. Biomed. Mater. Res. A 92, 1329–1342. 15. Smith, E., Thomas, W., Kiessling, L., and Corn, R. M. (2003) Surface plasmon resonance imaging studies of protein-carbohydrate interactions. J. Am. Chem. Soc. 125, 6140–6148. 16. Ulman, A. (1996) Formation and structure of self-assembled monolayers. Chem. Rev. 96, 1533–1554. 17. Ratner, D. M., Adams, E. W., Disney, M. D., and Seeberger, P. H. (2004) Tools for glycomics: mapping interactions of carbohydrates in biological systems. Chembiochem 5, 1375–83.

101

18. Zhi, Z., Laurent, N., Powell, A., Karamanska, R., Fais, M., Voglmeir, J., Wright, A., Blackburn, J., Crocker, P., and Russell, D. (2008) A versatile gold surface approach for fabrication and interrogation of glycoarrays. Chembiochem 9, 1568–1575. 19. Gildersleeve, J., Oyelaran, O., Simpson, J., and Allred, B. (2008) Improved procedure for direct coupling of carbohydrates to proteins via reductive amination. Bioconjugate Chem. 19, 1485–1490. 20. Lubineau, A., AugÈ, J., and Drouillat, B. (1995) Improved synthesis of glycosylamines and a straightforward preparation of N-acylglycosylamines as carbohydrate-based detergents. Carbohyd. res. 266, 211–219. 21. Rodriguez, E., Marcaurelle, L., and Bertozzi, C. (1998) Aminooxy-, hydrazide-, and thiosemicarbazide-functionalized saccharides: versatile reagents for glycoconjugate synthesis. J. Org. Chem 63, 7134–7135. 22. Lockett, M., Phillips, M., Jarecki, J., Peelen, D., and Smith, L. (2008) A tetrafluorophenyl activated ester self-assembled monolayer for the immobilization of amine-modified oligonucleotides. Langmuir 24, 69–75. 23. Martinez-Avila, O., Hijazi, K., Marradi, M., Clavel, C., Campion, C., Kelly, C., and PenadÈs, S. (2009) Gold manno-glyconanoparticles: multivalent systems to block HIV-1 gp120 binding to the Lectin DC-SIGN. Chem-Eur. J. 15, 9874–9888. 24. Fornstedt, N., and Porath, J. (1975) Characterization studies on a new lectin found in seeds of Vicia ervilia. FEBS letters 57, 187. 25. Cheng, F., Shang, J., and Ratner, D. M. (2011) A versatile method for functionalizing surfaces with bioactive glycans. Bioconjugate Chem. 22, 50–57. 26. Ding, Z., Long, C., Hayashi, Y., Bulmus, E., Hoffman, A., and Stayton, P. (1999) Temperature control of biotin binding and  release with A streptavidin-poly (N-isopropylacrylamide) site-specific conjugate. Bioconjugate Chem. 10, 395–400. 27. Michel, O., and Ravoo, B. (2008) Carbohydrate microarrays by microcontact click chemistry. Langmuir 24, 12116–12118. 28. Bolles, K., Cheng, F., Burk-Rafel, J., Dubey, M., and Ratner, D. M. (2010) Imaging analysis of carbohydrate-modified surfaces using ToF-SIMS and SPRi. Materials 3, 3948–3964.

Chapter 7 Carbohydrate Microarrays for Enzymatic Reactions and Quantification of Binding Affinities for Glycan–Protein Interactions Myung-Ryul Lee, Sungjin Park, and Injae Shin Abstract Glycans are involved in a variety of physiological and pathological processes through interactions with proteins. Thus, the molecular basis of glycan–protein interactions provides valuable information on understanding biological phenomena and exploiting more effective carbohydrate-based therapeutic agents and diagnostic tools. Carbohydrate microarray technology has become a powerful tool for evaluating glycanmediated biological events in a high-throughput manner. This technology is mostly applied for rapid analysis of glycans–protein interactions in the field of functional glycomics. In order to expand application areas of glycan microarrays, we have used carbohydrate microarrays for measurement of binding affinities between glycans and proteins and profiling of glycosyltransferase activities. The glycan microarrays used for these studies are constructed by immobilizing maleimide or hydrazide-conjugated glycans on the thiol or hydrazide-derivatized glass slides, respectively. This protocol describes the fabrication of carbohydrate microarrays and their applications to enzymatic reactions and determination of quantitative binding affinities. Key words: Binding affinity, Carbohydrates, Chemoselectivity, Enzymatic activity, Enzymatic glycosylation, High-throughput analysis, Immobilization, Lectins, Microarrays

1. Introduction Cell surface is highly decorated with glycans which are present in the form of glycoconjugates such as glycolipids, glycoproteins, and proteoglycans. These glycans serve as important recognition markers for a wide variety of physiological and pathological processes through interactions with proteins. Glycan–protein interactions

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_7, © Springer Science+Business Media, LLC 2012

103

104

M.-R. Lee et al.

are involved in cell communication, cell adhesion, and immune responses (1–4). Interestingly, these biomolecular interactions also play a role in deleterious disease processes including tumor metastasis, inflammation, and bacterial and viral infection (5–8). Therefore, the molecular basis of glycan–protein interactions sheds light on understanding of biological systems as well as development of potent therapeutic agents. Owing to the significance of glycan–protein interactions for biological research and biomedical applications, a number of analytic methods have been exploited to probe these interactions. For example, hemagglutination inhibition assay (9), enzyme-linked lectin assay (10), surface plasmon resonance (11) and isothemal titration calorimetry (12) have been widely utilized to evaluate glycan–protein interactions and to determine their binding affinities. Although these traditional approaches have been successfully employed for understanding of the details of glycan–protein recognition events, they are not suitable for high-throughput analysis of glycan–protein interactions since they are labor-intensive and require large amounts of samples. To overcome this shortcoming, carbohydrate microarray-based technology has been developed for rapid and quantitative analysis of glycan–protein interactions. Since the seminal technology of carbohydrate microarrays was developed by us and others in 2002, this microarray technology has become a powerful tool for studies of glycan-mediated biological events (13–25). Carbohydrate microarrays facilitate fast, quantitative, and simultaneous analysis of a large number of biomolecular interactions using a tiny amount of samples. To prepare glycan microarrays, various immobilization strategies of glycans on the solid surface, which rely on chemical ligation reactions, have been reported (26–33). The constructed carbohydrate microarrays have been mostly applied for rapid analysis of carbohydrate-protein interactions. Using this microarray technology, the glycan binding properties of various proteins have been investigated. In addition, glycan microarrays have also been utilized to detect pathogens for diagnosis of diseases. Furthermore, the microarrays have been applied to determine quantitative binding affinities between glycans and proteins (19, 33) as well as to investigate enzymatic glycosylations and glycosyltransferase activities (17, 19). In this chapter, we provide protocols for applications of carbohydrate microarrays, that are constructed by immobilization of maleimide or hydrazide-conjugated carbohydrates on thiol or epoxide-functionalized glass slides, respectively, to measuring quantitative binding affinities (IC50 and Kd values) (see Fig. 1). We also describe their applications to synthesis of sialyl Lex and profiling of glycosyltransferase activities.

7

X

X

Thiol or epoxide coated glass slide

O

Carbohydrate Microarrays for Enzymatic Reactions…

105

a tether Y

Maleimide or hydrazide conjugated glycans

b Carbohydrate microarray

Fig. 1. Construction of glycan microarrays by respective immobilization of maleimide or hydrazide-conjugated glycans on the thiol or epoxide derivatized glass slide and their applications to (a) determination of binding affinities between glycans and proteins and (b) profiling of glycosyltransferase activities.

2. Materials 2.1. Fabrication of Carbohydrate Microarrays

1. Maleimide (13, 17) and hydrazide-linked carbohydrates (18, 19, 25) are not commercially available and thus should be prepared according to the known procedure. 2. Buffers for dissolving maleimide-linked glycans; phosphate buffered saline (PBS, pH 6.8) containing 40–50% (v/v) glycerol (see Note 1). 3. Buffer for dissolving hydrazide-linked glycans; 100 mM sodium phosphate buffer (pH 5.0) containing 40–50% (v/v) glycerol. 4. Thiol and epoxide-derivatized glass slides are purchased from commercial suppliers such as TeleChem International, Inc. and Schott Nexterion or prepared according to the known procedure (see Note 2) (13, 17–19). 5. 1% N-Ethylmaleimide (Sigma-Aldrich) in water to quench the unreacted thiol groups after immobilization of maleimidelinked glycans on the thiol-coated glass slide. 6. 1–3% 2-Aminoethanol in 10 mM NaHCO3 (pH 8.3) to quench the unreacted epoxide groups after immobilization of hydrazide-linked glycans on the epoxide-coated glass slide. 7. PBS (pH 7.4) containing 0.1% Tween 20 (Sigma-Aldrich). 8. A plastic film (thickness: 0.1–0.2 mm, B.S. Inc.) that is coated by adhesive at one side. 9. MicroSys 5100 microarrayer (Cartisian Technologies) fitted with Stealth Micro Spotting pins (TeleChem International, Inc.).

106

M.-R. Lee et al.

2.2. Applications of Carbohydrate Microarrays

1. PBS (pH 7.4) containing 0.1% Tween 20 and 1% bovine serum albumin (BSA). BSA should be added to the buffer prior to use.

2.2.1. Enzymatic Reactions on Glycan Microarrays

2. For enzymatic galactosylation: a solution of β-1,4-galactosyltransferase (1 or 23 mU) (Calbiochem), 10 mM MnCl2 and UDP-Gal (0.1 or 1 mM) in 50 mM HEPES buffer (pH 7.5). 3. For enzymatic sialylation: a solution of α-2,3-sialyltransferase (1 mU) (Calbiochem), 5 mM MnCl2, alkaline phosphatase (20 μU) and 0.1 mM CMP-NeuNAc in 100 mM HEPES buffer (pH 7.0). 4. For enzymatic fucosylation: a solution of α-1,3-fucosyltransferase (1 mU) (Calbiochem), 15 mM MnCl2, alkaline phosphatase (20 μU), and 0.1 mM GDP-Fuc in 50 mM MES buffer (pH 6.0). 5. Mouse anti-sialyl Lex antibody (5–10 μg/ml) (Calbiochem) in 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20. Tween 20 should be added to the solution prior to use. 6. 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20 for washing unbound anti-sialyl Lex antibody. 7. Goat anti-antibody (10 μg/ml) (Calbiochem) in 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20. Tween 20 should be added to the solution prior to use. 8. Cy3-labeled Ricinus communis agglutinin I (Cy3-RCA120, 10–20 μg/ml) in PBS (pH 7.4). 9. 35 mM SDS in PBS buffer (pH 7.4) for wash of the slide after treatment of β-1,4-galactosyltransferase for profiling β-1,4galactosyltransferase activity. 10. Cartisian AxSys software (Cartisian Technologies). 11. Temperature and humanity-controlled incubator (Daihan Scientific, Korea). 12. ArrayWorx™ microarray scanner (Applied Precisions, USA). 13. ImaGene 6.1 software and origin Pro7.0 software.

2.2.2. Determination of Binding Affinities Between Carbohydrates and Proteins

1. PBS (pH 7.4) containing 0.1% Tween 20 and 1% BSA. 2. A series of mixture of a fluorophore-labeled protein and a soluble inhibitor in PBS (pH 7.4) containing 0.1% Tween 20. 3. A solution of 0.1 nM–1 μM concentrations for Cy3-RCA120 (10–20 μg/ml) in PBS (pH 7.4) containing 0.1% Tween 20. 4. Cartisian AxSys software (Cartisian Technologies).

7

Carbohydrate Microarrays for Enzymatic Reactions…

107

5. Temperature and humanity-controlled incubator (Daihan Scientific, Korea). 6. ArrayWorx™ microarray scanner (Applied Precisions, USA). 7. ImaGene 6.1 software and origin Pro7.0 software.

3. Methods Glycan microarrays are constructed by printing very small quantities (usually 1 nl) of maleimide or hydrazide-conjugated glycans on the thiol or epoxide-derivatized glass slides using a microarrayer. Buffer solutions used for glycan immobilization should contain 40–50% glycerol to prevent undesired evaporation of nanodroplets during spotting and immobilization steps. For applications of glycan microarrays to enzymatic reactions, they are incubated with glycosyltransferases followed by treatment with fluorescent dyelabeled proteins to detect the transferred glycans on the microarrays. IC50 values of soluble inhibitors are measured by incubating glycan microarrays with the fluorescent dye-labeled protein in the presence of various concentrations of an inhibitor and then quantitating fluorescence intensity of bound protein with a fluorescence scanner. Dissociation constants (Kd values) for glycan–protein interactions are determined by incubating glycan microarrays with various concentrations of fluorescent dye-labeled protein and then quantitating fluorescence intensity of the bound protein with a fluorescence scanner. 3.1. Fabrication of Glycan Microarrays 3.1.1. Immobilization of Maleimide-Conjugated Carbohydrates on the Thiol-Derivatized Glass Slide

1. Dissolve maleimide-conjugated glycans used for immobilization at 0.1–1.0 mM in PBS (pH 6.8) containing 40–50% glycerol. 2. Transfer 5–10 μl of the solution into a 384-well microplate. The V-shaped 384-well microplate is recommended because a small amount of solutions can be loaded into the V-shaped well. The microplate containing the solutions can be stored at −70°C for a month (see Note 3). 3. Cut parts of a plastic film (thickness: 0.1–0.2 mm) that is coated by adhesive at one side with a knife (see Fig. 2). Attach the compartmentalized plastic film to the thiol-derivatized glass slide. 4. Print 1 nl of maleimide-conjugated glycan solutions from a 384-well microplate in predetermined places on the thiolcoated glass slide by using a robotic printing microarrayer (see Note 4).

108

M.-R. Lee et al. A cut plastic film

Derivatized glass slide Attach a plastic film to a glass slide Print glycans

Fig. 2. Scheme for the attachment of a plastic film to the glass slide.

5. After completion of printing, leave the slide in the print chamber (60% humidity) for 3 h at room temperature. The slide can be left overnight at room temperature in the print chamber. 6. After washing the slide with 30 ml of deionized water (5 min × 2), dry the slide by purging with Ar gas. 7. Drop 20–30 μl of a solution of 1% N-ethylmaleimide in water on each block of the slide by using a micropipette and incubate in the print chamber (60% humidity) for 1 h at room temperature. Unreacted thiols on the slide are removed at this step (see Note 5). 8. After washing the slide with 30 ml of deionized water (5 min × 2), dry the slide by purging with Ar gas. The slide can be stored at room temperature in a desiccator for several weeks. However, for best results, prepare the slide freshly prior to use. 3.1.2. Immobilization of Hydrazide-Conjugated Glycans on the EpoxideDerivatized Glass Slide

1. Prepare 100 μl of 0.1–1.0 mM hydrazide-conjugated glycan solutions in 100 mM sodium phosphate buffer (pH 5.0) containing 40–50% glycerol (see Note 1). 2. Transfer 5–10 μl of the solution into a 384-well microplate. 3. Cut parts of a plastic film (thickness: 0.1–0.2 mm) that is coated by adhesive at one side with a knife (see Fig. 2). Attach the compartmentalized plastic film to the epoxide-derivatized glass slide. 4. Print 1 nl of hydrazide-conjugated glycan solutions from a 384-well microplate in predetermined places on the epoxidecoated glass slide (see Note 4). 5. After completion of printing, leave the slide in the print chamber (60% humidity) for 3 h at room temperature. The slide can be left overnight at room temperature in the print chamber. 6. After washing the slide with 30 ml of deionized water (5 min × 2), dry the slide by purging with Ar gas.

7

Carbohydrate Microarrays for Enzymatic Reactions…

109

7. Drop 20–30 μl of a solution of 1–3% 2-aminoethanol or glycine in 10 mM sodium bicarbonate (pH 8.3) on each block of the slide by using a micropipette and incubate in the print chamber (60% humidity) for 1 h at room temperature. Unreacted epoxides on the slide are removed at this step. 8. After washing the slide with 30 ml of deionized water (5 min × 2), dry the slide by purging with Ar gas. The slide can be stored at room temperature in a desiccator for several weeks. However, for best results, prepare the slide freshly prior to use. 3.2. Applications of Carbohydrate Microarrays for Enzymatic Reactions 3.2.1. Enzymatic Glycosylations to Prepare Sialyl Le x from GlcnacImmobilized Glass Slides

1. Prepare the carbohydrate microarray immobilized by GlcNAc according to the procedure described in Subheading 3.1. 2. Drop 20–30 μl of a solution of PBS (pH 7.4) containing 1% BSA and 0.1% Tween 20 on each block of the slide by using a micropipette and incubate in the print chamber (60% humidity) for 0.5–1 h at room temperature (see Note 6). 3. Rinse the slide with PBS (pH 7.4) containing 0.1% Tween 20 to remove protein solution and wash with 30 ml of the same buffer with gentle shaking (10 min × 3). Rinse the slide with deionized water. 4. For galactosylation of GlcNAc, drop 15 μl of a solution of β-1,4-galactosyltransferase (GalT, 23 mU), MnCl2 (10 mM) and UDP-Gal (0.1 mM) in HEPES buffer (50 mM, pH 7.5) on each block of the BSA-pretreated slide by using a micropipette. Incubate the slide in a temperature-controlled humidity chamber (80% humidity) for 15 h at 37°C. 5. Wash the slide with PBS (pH 7.4) containing 0.1% Tween 20 with gentle shaking (15 min × 3) and rinse with water. Dry the slide by purging with Ar gas. 6. Incubate the GalT-treated slide with 15 μl of a solution of α-2,3-sialyltransferase (SialT, 1 mU), MnCl2 (5 mM), alkaline phosphatase (20 μU) and CMP-NeuNAc (0.1 mM) in HEPES buffer (100 mM, pH 7.0) in a temperature-controlled humidity chamber (80%) for 15 h at 37°C for sialylation of LacNAc. 7. Wash the slide with PBS (pH 7.4) containing 0.1% Tween 20 with gentle shaking (15 min × 3) and rinse with water. Dry the slide by purging with Ar gas. 8. Incubate the GalT and SialT-treated slide with 15 μl of a solution of α-1,3-fucosyltransferase (FucT, 1 mU), MnCl2 (15 mM), alkaline phosphatase (20 μU) and GDP-Fuc (0.1 mM) in MES buffer (50 mM, pH 6.0) in a temperaturecontrolled humidity chamber (80%) for 15 h at 37°C for fucosylation of NeuNAcα2,6LacNAc. 9. Wash the slide with PBS (pH 7.4) containing 0.1% Tween 20 with gentle shaking (15 min × 3) and rinse with water. Dry the slide by purging with Ar gas.

110

M.-R. Lee et al.

10. To examine successful glycosylation, incubate the slide with PBS (pH 7.4) containing 0.1% Tween 20 and 1% BSA for 30 min at room temperature and then treat with mouse antisialyl Lex antibody (5–10 μg/ml) in 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20 for 1 h at room temperature. 11. Wash the slide with 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20 (10 min × 3) and rinse with water. Dry the slide by purging with Ar gas. 12. Incubate the antibody-treated slide with Cy5-labeled goat anti-antibody (10 μg/ml) in 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20 for 1.5 h at room temperature (see Note 7). 13. Remove the compartmentalized plastic film from the glass slide and wash the slide with 10 mM sodium phosphate (pH 7.2) containing 500 mM NaCl and 0.02% Tween 20 (10 min × 3). Rinse the slide with water and dry by purging with Ar gas. 14. Scan the slide using a microarray scanner fitted with the appropriate filters. Process fluorescence data using ImaGene 6.1 software and analyze data using Origin Pro 7.0 or other software (see Fig. 3). 1. Prepare the carbohydrate microarray containing various carbohydrates according to the procedure described in Subheading 3.1.

3.2.2. Profiling of b-1,4Galactosyltransferase Activity

2. Drop 20–30 μl of a solution of PBS (pH 7.4) containing 1% BSA and 0.1% Tween 20 on each block of the slide by using a OH OH HO O O HO OH

NHAc HO HO

O OH

NH

β-1,4-GalT

NHAc O OH

NH

UDP-Gal

GlcNAc

LacNAc α-2,3-SialT

OH OH

CMP-NeuAc OH

HO2C HO

O

O H3C OH OH O O O O OH

HO OH AcHN OH

NHAc O OH

anti-sialyl Lex Cy5-anti-antibody HO2C

NH

α-1,3-FucT GDP-Fuc

Sialyl Lex

HO

O

OH OH HO O O O OH

NHAc O OH

NH

HO OH AcHN OH

NeuAcα2,3Galβ1,4βGlcNAc

Fig. 3. Enzymatic synthesis of sialyl Lex from GlcNAc immobilized on the glass surface by using three glycosyltransferases. Treatment of microarrays, which are incubated with three glycosyltransferases, with anti-sialyl Lex antibody and then Cy5-anti-antibody shows the successful preparation of the tetrasaccharide from GlcNAc.

7

Carbohydrate Microarrays for Enzymatic Reactions…

111

micropipette and incubate in the print chamber (60% humidity) for 0.5–1 h at room temperature (see Note 6). 3. Rinse the slide with PBS (pH 7.4) containing 0.1% Tween 20 to remove protein solution and wash with the same buffer with gentle shaking (10 min × 3). Rinse the slide with deionized water. 4. Drop 20–30 μl of a solution of β-1,4-galactosyltransferase (1 mU), MnCl2 (10 mM), and UDP-Gal (1 mM) in HEPES buffer (50 mM, pH 7.5) on each block and incubate the slide in a temperature-controlled humidity chamber (80% humidity) at 37°C for 3 h. 5. Rinse the slide with HEPES buffer to remove enzyme solutions and wash with PBS buffer (pH 7.4) containing 0.1% Tween 20 with gentle shaking (10 min × 3). 6. Further wash the slide with 35 mM SDS in PBS (pH 7.4) under sonication at 60°C–65°C for 5 min (see Note 8) and then rinse with deionized water. Dry the slide by purging with Ar gas. 7. To detect galactose transferred by GalT, drop 20–30 μl of a solution of Cy3-RCA120 in PBS (pH 7.4) containing 0.1% Tween 20 on each block of the slide by using a micropipette. Incubate the slide in the print chamber (60% humidity) for 0.5–1 h at room temperature (see Note 7). 8. Rinse the slide with PBS (pH 7.4) to remove protein solutions. Remove the compartmentalized plastic film from the glass slide and wash with PBS (pH 7.4) containing 0.1% Tween 20 with gentle shaking (10 min × 3). Rinse the slide with deionized water. 9. Dry the slide by purging with Ar gas. The slide can be stored in a desiccator at room temperature for several weeks. 10. Scan the slide using a microarray scanner fitted with the appropriate filters. Process fluorescence data using ImaGene 6.1 software and analyze data using Origin Pro 7.0 or other software (see Fig. 4). 3.3. Measurements of Quantitative Binding Affinities 3.3.1. Determination of IC50 Values

1. Prepare the carbohydrate microarray containing α-GlcNAc and α-GalNAc according to the procedure described in Subheading 3.1. 2. Drop 10 μl of a solution of PBS (pH 7.4) containing 1% BSA and 0.1% Tween 20 on each block of the slide by using a micropipette and incubate in the print chamber (60% humidity) for 0.5–1 h at room temperature (see Note 6). 3. Wash the slide with PBS (pH 7.4) containing 0.1% Tween 20 with gentle shaking (15 min × 3) and rinse with water. Dry the slide by purging with Ar gas.

112

M.-R. Lee et al. 3

4

5

6

7

8

9

10

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

11

12

13

14

15

16

17

18

19

20

6

8

18

20

35 30 25

GalT, UDP-Gal

20

Mn2+, 3 h

15 10 5 0 2

4

6

8

10

12 14 16

18

20

Fluorescence intensity (x1000)

2

Fluorescence intensity (x1000)

1

20 16 12 8 4 0

2

4

10

12 14 16

α- and β-GlcNAc → α- and β-LacNAc

Fig. 4. Applications of carbohydrate microarrays for profiling of β-1,4-galactosyltranserase activity (1: Fuc-α, 2: Fuc-β, 3: Gal-α, 4: Gal-β, 5: GalNAc-α, 6: Glc-α, 7: Glc-β, 8: GlcNAc-α, 9: GlcNAc-β, 10: Man-α, 11: Xyl-α, 12: Xyl-β, 13: Glcβ1,4Glc-β, 14: Galβ1,4Glc-β, 15: Glcα1,4Glc-β, 16: Galβ1,4GlcNAc-β, 17: Galβ1,6Man-α, 18: Glcβ1,6Man-α, 19: Manα1,6Man-α, 20: Manα1,6Manα1,6Man-α). Carbohydrate microarrays were incubated with 1 mU β-1,4-GalT in the presence of 10 mM MnCl2 and 1 mM UDP-Gal for 3 h at 37°C. To detect transferred Gal, the microarrays were probed with Cy3-RCA120. Fluorescence images of RCA120-treated slides before (left ) and after (right ) treatment with enzyme. α- and β-GlcNAc were converted to the corresponding LacNAc by this enzyme. Data are the average ± S.D. of triplicate determinants.

b 30000

WGA Fluorescence intensity

WGA Fluorescence intensity

a 25000 20000 15000 10000 5000 0

IC 50= 3.5 mM 1.5

2.0 2.5 3.0 3.5 4.0 4.5. 5.0 Log[α-GlcNAc-OMe] (mM)

45000 35000 25000 15000 5000 0

IC 50= 8.9 mM 2.0

2.5 3.0 3.5 4.0 4.5. 5.0 5.5 Log[α-GlcNAc-OMe] (mM)

Fig. 5. Determination of IC50 values of methyl N-acetyl-α-glucosaminide (α-GlcNAc-OMe) to inhibit 50% of WGA (wheat germ agglutinin) binding to (a) α-GlcNAc and (b) α-GalNAc on the carbohydrate microarray.

4. Incubate the carbohydrate microarray with a series of mixtures of a dye-labeled protein and a soluble inhibitor in PBS (pH 7.4) containing 0.1% Tween 20 for 1 h in the print chamber (60% humidity) at room temperature (see Note 7). 5. Wash the slide with PBS (pH 7.4) containing 0.1% Tween 20 with gentle shaking for 15 min three times and rinse with water. 6. Dry the slide by purging with Ar gas and measure fluorescence intensities using a fluorescence scanner. 7. Determine IC50 values of inhibitors from the obtained graph by finding points where concavity changes (see Fig. 5). In other

7

Carbohydrate Microarrays for Enzymatic Reactions…

113

words, IC50 values are obtained by calculating concentrations where d2f(x)/dx2 = 0 (f(x); fluorescence intensity and x; concentration of an inhibitor). 1. Prepare the carbohydrate microarray containing α- and β-LacNAc according to the procedure described in Subheading 3.1 and then each block is treated with BSA solution.

3.3.2. Determination of Dissociation Constants (Kd Values) for Protein– Carbohydrate Interactions

2. Drop 20–30 μl of a solution of 0.1 nM–1 μM concentrations of Cy3-labeled RCA120 (20 μg/ml) in PBS (pH 7.4) containing 0.1% Tween 20 on each block of the slide by using a micropipette. 3. Incubate the slide in the print chamber (60% humidity) for 1 h at room temperature. 4. Rinse the slide with PBS (pH 7.4) containing 0.1% Tween 20 to remove protein solution and wash with wash buffer with gentle shaking (3 min × 3). Rinse the slides with deionized water.

[P]0/FI (10-12)

5. Scan the slide using a microarray scanner fitted with the appropriate filters. Process fluorescence data using Imagene 6.1 software and analyze data using Origin Pro 7.0. Kd values are determined by using equation [P]o/FI = Kd/Flmax + [P]o/Flmax (Fl; fluorescence intensity, Flmax; maximum fluorescence intensity and [P]0; concentration of protein) (Fig. 6).

14 12 10 8 6 4 2 0

Kd = 3.4 x 10-8 (LacNAc-α)

[P]o

(Kd + [P]o) = FI Flmax Fl: fluorescence intensity Flmax: maximum fluorescence intensity [P]0: concentration of protein

Kd = 3.3 x 10-8 (LacNAc-β) 0

200

400

600

800

1000

[P]0 (nM) RCA120 (nM) 1000

500

250

3.7

1.8

125

62.5

31.5

15.6

0.45

0.23

0.12

LacNAcα LacNAcβ

7.5

0.9

Fig. 6. Applications of carbohydrate microarrays to determination of dissociation constants (Kd) for RCA120-surface-linked LacNAc interactions (filled square : LacNAc-β, filled circle: LacNAc-α, FI: fluorescence intensity). Kd values are determined by using equation [P]o/FI = Kd/Flmax + [P]o/Flmax (19).

114

M.-R. Lee et al.

4. Notes 1. Glycerol was added into sample solutions to prevent the unwanted evaporation of nanodroplets during spotting and immobilization. Evaporation of solutions causes uneven, inefficient immobilization of glycans. 2. Quality of the modified glass slides is critical for reproducible results. Therefore, it is highly recommended that those with little experience should purchase the modified glass slides from commercial suppliers. 3. Do not repeat thawing and freezing of the sample solutions many times. Repetitive thawing and freezing will result in decomposition of modified glycans in solutions. 4. To prevent contamination of slides, the print chamber and printing pins should be cleaned. 5. Capping of unreacted thiol groups with N-ethylmaleimide prevents oxidative disulfide-bond formation between thiols on the surface and cysteine residues of proteins during incubation with labeled proteins. N-Methylmaleimide can be used instead of N-ethylmaleimide without any problem. 6. One of the most serious problems for microarray experiments is the nonspecific interaction of the probing proteins with derivatized surfaces, leading to high background fluorescence. BSA treatment of microarrays considerably attenuates the nonspecific interaction. Hydrophilic surfaces modified with poly(ethylene glycol) greatly suppress the nonspecific interaction even without treatment with BSA. 7. Tween 20 should be added to protein solutions prior to use since Tween 20 results in the decrease of protein activity during storage. 8. Omission of sonication of the enzyme-treated slide in a hot SDS solution results in a high background fluorescence.

Acknowledgments This work was supported by grants of the National Creative Research Initiative and WCU programs.

7

Carbohydrate Microarrays for Enzymatic Reactions…

115

References 1. Bertozzi, C. R., and Kiessling, L. L. (2001) Chemical glycobiology. Science 291, 2357–2364. 2. Varki, A. (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3, 97–130. 3. Roth, J. (2002) Protein N-glycosylation along the secretory pathway: relationship to organelle topography and function, protein quality control, and cell interactions. Chem. Rev. 102, 285–303. 4. Park, S., Lee, M.-R., and Shin, I. (2008) Chemical tools for functional studies of glycans. Chem. Soc. Rev. 37, 1579–1591. 5. Smith, A. E., and Helenius, A. (2004) How viruses enter animal cells. Science 304, 237–242. 6. Fuster, M. M., and Esko, J. D. (2005) The sweet and sour of cancer: glycans as novel therapeutic targets. Nat. Rev. Cancer 5, 526–542. 7. Lasky, L. A. (1992) Selectins: interpreters of cell-specific carbohydrate information during inflammation. Science 258, 964–969. 8. Scanlan, C.N., Offer, J., Zitzmann, N., and Dwek, R.A. (2007) Exploiting the defensive sugars of HIV-1 for drug and vaccine design. Nat. Insight 446, 1038–1045. 9. Lis, H., and Sharon, N. (1972) Lectins: cellagglutinating and sugar-specific proteins. Science 177, 949–959. 10. McCoy Jr, J. P., Varani, J., and Goldstein, I. J. (1983) Enzyme-linked lectin assay (ELLA): use of alkaline phosphatase-conjugated Griffonia simplicifolia B4 isolectin for the detection of alpha-D-galactopyranosyl end groups. Anal. Biochem. 130, 437–444. 11. Duverger, E., Frison, N., Roche, A. C., and Monsingny, M. (2003) Carbohydrate-lectin interactions assessed by surface plasmon resonance. Biocheimie 85, 167–179. 12. Dam, T. K., and Brewer, C. F. (2002) Thermodynamic studies of lectin-carbohydrate interactions by isothermal titration calorimetry. Chem. Rev. 102, 387–430. 13. Park, S., and Shin, I. (2002) Fabrication of carbohydrate chips for studying proteincarbohydrate interactions. Angew. Chem. Int. Ed. 41, 3180–3182. 14. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–281. 15. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

microarrays for high-throughput detection and specificity assignments of carbohydrateprotein interactions. Nat. Biotechnol. 20, 1011–1017. Houseman, B. T., and Mrksich, M. (2002) Carbohydrate arrays for the evaluation of protein binding and enzymatic modification. Chem. Biol. 9, 443–454. Park, S., Lee, M.-R., Pyo, S. J., and Shin, I. (2004) Carbohydrate chips for studying highthroughput carbohydrate-protein interactions. J. Am. Chem. Soc. 126, 4812–4819. Lee, M.-R., and Shin, I. (2005) Fabrication of chemical microarrays by efficient immobilization of hydrazide-linked substances on epoxide-coated glass surfaces. Angew. Chem. Int. Ed. 44, 2881–2884. Park, S., and Shin, I. (2007) Carbohydrate microarrays for assaying galactosyltransferase activity. Org. Lett. 9, 1675–1678. Lee, M.-R., and Shin, I. (2005) Facile Preparation of Carbohydrate Microarrays by Site-Specific, Covalent Immobilization of Unmodified Carbohydrates on HydrazideCoated Glass Slides. Org. Lett. 7, 4269–4272. Blixt, O. et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc. Natl. Acad. Sci. USA 101, 17033–17038. de Paz, J. L., Noti, C., and Seeberger, P. H. (2006) Microarrays of synthetic heparin oligosaccharides. J. Am. Chem. Soc. 128, 2766–2767. Disney, M. D., and Seeberger, P. H. (2004) The use of carbohydrate microarrays to study carbohydrate-cell interactions and to detect pathogens. Chem. Biol. 11, 1701–1707. Park, S., Lee, M.-R., and Shin, I. (2009) Construction of carbohydrate microarrays by using one-step, direct immobilizations of diverse unmodified glycans on solid surfaces. Bioconjugate Chem. 20, 155–162. Park, S., Lee, M.-R., and Shin, I. (2007) Fabrication of carbohydrate chips and their use to probe protein-carbohydrate interactions. Nat. Protoc. 2, 2747–2758. Shin, I., Cho, J. W., and Boo, D. W. (2004) Carbohydrate arrays for functional studies of carbohydrates. Comb. Chem. High Throughput Screening 7, 565–574. Shin, I., Park, S., and Lee, M.-R. (2005) Carbohydrate microarrays: an advanced technology for functional studies of glycans. Chem. Eur. J. 11, 2894–2901.

116

M.-R. Lee et al.

28. Shin, I., Tae, J., and Park, S. (2007) Carbohydrate microarray technology for functional glycomics. Curr. Chem. Biol. 1, 187–199. 29. Shin, I. (2006) Carbohydrate microarrays for high-throughput analysis of carbohydrate-protein interactions. In Protein-Carbohydrate Interactions in Infectious Diseases. (Ed. Carole A. Bewley) 221–46 (RSC Publishing, Cambridge, UK). 30. Park, S., Lee, M.-R., and Shin, I. (2008) Carbohydrate microarrays as powerful tools in studies of carbohydrate-mediated biological processes. Chem. Commun. 4389–4399.

31. Horlacher, T., and Seeberger, P. H. (2008) Carbohydrate arrays as tools for research and diagnostics. Chem. Soc. Rev. 37, 1414–1422. 32. Laurent, N., Voglmeir, J., and Flitsch, S. L. (2008) Glycoarrays-tools for determining protein-carbohydrate interactions and glycoenzyme specificity. Chem. Commun. 4400–4412. 33. Liang, P.-H., Wang, S.-K., and Wong, C.-H. (2007) Quantitative analysis of carbohydrateprotein interactions using glycan microarrays: determination of surface and solution dissociation constants. J. Am. Chem. Soc. 128, 11177–11184.

Chapter 8 Neoglycolipid-Based Oligosaccharide Microarray System: Preparation of NGLs and Their Noncovalent Immobilization on Nitrocellulose-Coated Glass Slides for Microarray Analyses Yan Liu, Robert A. Childs, Angelina S. Palma, Maria A. Campanero-Rhodes, Mark S. Stoll, Wengang Chai, and Ten Feizi Abstract Carbohydrate microarrays, since their advent in 2002, are revolutionizing studies of the molecular basis of protein–carbohydrate interactions both in endogenous recognition systems and pathogen–host interactions. We have developed a unique carbohydrate microarray system based on the neoglycolipid (NGL) technology, a well-validated microscale approach for generating lipid-tagged oligosaccharide probes for use in carbohydrate recognition studies. This chapter provides an overview of the principles and key features of the NGL-based oligosaccharide microarrays, and describes in detail the basic techniques – from the preparation of NGL probes to the generation of microarrays using robotic arraying hardware, as well as a general protocol for probing the microarrays with carbohydrate-binding proteins. Key  words: Neoglycolipids, Microarrays, Nitrocellulose, Reductive amination, Noncovalent immobilization

1. Introduction 1.1. Neoglycolipid Technology: A Unique and Well-Validated Microscale Approach for Carbohydrate Ligand Discovery

The neoglycolipid (NGL) technology was originally introduced by Feizi and colleagues in 1985 as a novel approach to the study of the antigenicities and receptor functions of carbohydrate chains of glycoproteins (1). It was designed to address the need for microscale analysis and presentation of oligosaccharides in a multivalent form for studying carbohydrate–protein interactions which are generally weak. The technology involves conjugating oligosaccharides by reductive amination to an aminolipid, 1,2-dihexadecyl-sn-glycero3-phosphoethanolamine (DHPE). A library of NGLs from diverse

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_8, © Springer Science+Business Media, LLC 2012

117

118

Y. Liu et al.

oligosaccharide sequences has thus been synthesized: O-glycans or parts thereof (1, 2), N-glycans (3), oligosaccharide fragments of glycosaminoglycans (4, 5) and polysaccharides (5–9), as well as synthetic oligosaccharides (10, 11). The lipid tag confers amphipathic properties and enables efficient immobilization of NGLs in an oriented and clustered display on solid matrices, such as plastic wells, silica-gel thin layer chromatography (TLC) plates or nitrocellulose membranes, for probing with various carbohydraterecognizing systems. Moreover, there is the special advantage that highly heterogeneous mixtures of oligosaccharides when converted into NGLs can be probed for binding after being resolved on highperformance TLC (HPTLC) plates, and the components bound on the plate can be sequenced in situ by mass spectrometry (Fig. 1).

Fig. 1. Schematic presentation of the NGL-based oligosaccharide microarray platform.

8  Neoglycolipid-Based Oligosaccharide Microarray System…

119

This approach has proven to be a powerful means of identifying hitherto unsuspected oligosaccharides; metaphorically speaking, finding the needle in a haystack (12–14). Highlights during the early stages of application of the technology include the finding of sulfated rather than the familiar sialylated ligands for the selectins among highly heterogeneous oligosaccharides in an epithelial mucin (15) and the discovery of yeast type O-mannosyl glycans in the brain O-glycome (2), amounting to 30% of O-glycans in the brain glycome (16), and now known to be affected in congenital neuromuscular diseases. Other than ring-opening of the monosaccharide residues at the reducing ends, oligosaccharides remain intact in DHPE-derived NGLs (Fig. 2). Among the further developments of methodology are (1) the generation of fluorescent NGLs by conjugating oligosaccharides, via reductive amination, to a fluorescence-labeled lipid, the anthracene-modified DHPE (ADHP) (Fig.  2); these can be detected by UV or fluorescence after TLC and during HPLC with subpicomole detection sensitivity (17); and (2) the creation of a new generation of NGL probes prepared by oxime ligation (18, 19) (Fig.  2), of which a significant proportion of the core monosaccharides are ring-closed so that diverse short oligosaccharides are efficiently presented for direct binding assays. In addition, oligosaccharides that are hard to derivatize by reductive amination, e.g., fragments of glycosaminoglycans and fungal/bacterial

Fig. 2. Principles of the preparation of NGL probes from reducing oligosaccharides by reductive amination and oxime ligation. Lipid reagents ADHP (14, 16) and AOPE (17, 18) are prepared from DHPE.

120

Y. Liu et al.

polysaccharides, are efficiently derivatized to oxime-linked NGLs ((19) and unpublished results). The scope of the NGL-based oligosaccharide library has thus been further broadened. 1.2. NGL-Based Oligosaccharide Microarrays: Current State of the Art

The NGL technology was adapted to generate the first microarray system for sequence-defined oligosaccharides in 2002 (5). With the use of high precision robotic arraying hardware and imaging, the technology is now the basis of an advanced carbohydrate microarray system currently encompassing more than 700 lipidlinked oligosaccharide probes that are printed onto nitrocellulosecoated glass slides at low femtomole levels (Fig. 1). The repertoire of probes include NGLs derived from diverse oligosaccharide sequences of mammalian-type glycoproteins (N- and O-linked glycans), glycolipids, glycosaminoglycans, and polysaccharides of bacterial, fungal, and plant origins. Natural and synthetic glycolipids are also arrayed in parallel. The repertoire is continually expanding in number and structural diversity. A special feature of the NGLbased oligosaccharide microarray system is that the lipid-linked oligosaccharide probes are noncovalently immobilized on solid matrices and thus have potential lateral mobility; this is an advantage compared with most carbohydrate microarray systems in which oligosaccharides are covalently immobilized (by chemical reactions) and thus “fixed” on the array surfaces; this noncovalent presentation mimics to some extent the arrangement of clustered oligosaccharide structures at the cell surface. The NGL-based microarrays are providing crucial information on specificities of a variety of carbohydrate recognition systems operating endogenously and in pathogen–host interactions (6, 20). Among recent contributions are (1) the assignment of the ligands for a key receptor of the innate immune system against fungal pathogens, Dectin-1 (9); (2) the demonstration of sulfation as a modulator of carbohydrate recognition by sialic acid-binding receptors of the immune system, known as siglecs (21); (3) the discovery of the N-glycolyl analog of ganglioside GM1 (which humans do not synthesize) as the preferred host-cell receptor for Simian virus 40 (SV40) (22), which may explain the high susceptibility of simians to this pathogen; (4) the clinching of the ligand for a novel protein in the endoplasmic reticulum, malectin (23, 24), a candidate new player in the early steps of protein N-glycosylation, folding and quality control; (5) the elucidation of carbohydrate-binding specificities of several key surface-adhesive proteins of Toxoplasma gondii and other apicomplexan parasites (25–28), thus providing clues to host/tissue tropisms and pathobiology of infections by these parasites; and (6) the demonstration of a distinctive receptor-binding profile of the pandemic, swineorigin, influenza A (H1N1) 2009 (H1N1pdm) compared with that of seasonal human H1N1 influenza virus (29) which becomes more pronounced with hemagglutinin D222G mutants of

8  Neoglycolipid-Based Oligosaccharide Microarray System…

121

H1N1pdm viruses isolated from cases of fatal infection (30), thus uncovering a potential mechanism linking the receptor-binding specificity to severity of disease. 1.3. Purpose and Scope of This Chapter

In this chapter, we provide guidance for biochemists and molecular biologists who wish to generate and apply NGL-based oligosaccharide microarrays for studies of protein–carbohydrate interactions. We focus on the basic techniques for preparing, purifying, and quantifying DHPE-derived NGLs from reducing oligosaccharides by reductive amination (DH-NGLs) for the purpose of generating microarrays. The preparation of NGLs from reduced oligosaccharides and the preparation of fluorescent and oxime-linked NGLs (Fig. 2), including the step-by-step chemical syntheses of the lipid reagents (ADHP and AOPE in Fig.  2) from DHPE, have been described in detail elsewhere (14, 19) and can be readily adapted for the generation of NGL microarrays described in this chapter. Here, we provide protocols for preparing NGLs for arraying onto nitrocellulose-coated glass slides. Procedures for microarray analyses of plant lectins and antibodies are also given as a guide to carry out binding analyses on nitrocellulose-based surface. Procedures of analyses of recombinant proteins (Fc-tagged and His-tagged) are described in Chapter 23 (by Palma et al.) of this book series, which describes the application of the NGL-based microarray technology in oligosaccharide ligand discovery for Dectin-1.

2. Materials 2.1. Preparation of NGLs from Reducing Oligosaccharides

1. A wide range of free reducing oligosaccharides can be ­purchased from companies, including Dextra Laboratories (Reading, UK), GlycoTech (Gaithersburg, USA), Megazyme (Wicklow, Ireland), and Elicityl (Crolles, France). Stock solutions of oligosaccharides can be prepared in water, e.g., at 1 mg/ml, and can be stored in well-sealed sample vials at −20°C for several years. 2. Screw-top glass vials (1.1 ml tapered, 2 and 4 ml) with screw caps and silicone/PTFE seals are available from Chromacol (Herts, UK). 3. All organic solvents are of HPLC grade. Unless otherwise stated, deionised water is used. Compositions of mixed solvents are by volume. 4. DHPE stock solution: 8 nmol/ml of DHPE (1,2-dihexadecylsn-glycero-3-phosphoethanolamine, available from SigmaAldrich) in chloroform/methanol 1:1. This solution is made in a 4-ml screw-cap glass vial and is stable at −20°C (well sealed) at least for 1 year.

122

Y. Liu et al.

5. Sodium cyanoborohydride solution: 10  mg/ml sodium cyanoborohydride (Sigma-Aldrich) in methanol. This solution should be prepared freshly just before the conjugation reaction. 6. For incubation of reaction vials, a heating block (Dri-Block® heater) from Techne (Staffordshire, UK) is used. For evaporation of samples under a nitrogen stream, a Reacti-Therm Heating Module equipped with a Reacti-Vap Evaporator from Thermo Scientific (Rockford, USA) is used. 7. Aluminum-backed silica gel HPTLC plates (Merck, Darmstadt, Germany), semiautomatic TLC sample applicator Linomat 4 (Camag, Muttenz, Switzerland), glass TLC developing tank and sprayer (available from Sigma-Aldrich or Camag) are used for HPTLC analysis. 8. Primulin stock solution: 1 mg/ml of primulin (Sigma-Aldrich; referred to as primuline or Direct Yellow 59 in the catalogue) in acetone/water 1:9. This solution can be stored at 4°C (in dark) at least for 1 year. 9. Primulin spraying solution: 1/100 dilution of the primulin stock solution (above) in acetone/water 4:1. This solution can be stored at ambient temperature (in dark) for 2 weeks. 10. Orcinol spraying solution is prepared as follows: Dissolve 900 mg of orcinol monohydrate (Sigma-Aldrich) in 25 ml of water, and add 375 ml of ethanol. Cool on ice. Gradually add 50  ml of concentrated sulfuric acid (95–98%) with stirring, maintaining the temperature below 10°C. This solution can be stored at 4°C (in dark) and is stable for at least 1 year. 2.2. Purification of NGL Products

1. C18 cartridges (Sep-Pak Vac 3  cm3, 200  mg) and silica ­cartridges (Sep-Pak Vac 3 cm3, 500 mg) from Waters (Milford, USA). 2. Ammonium acetate solution: 0.2 M in water. 3. Materials and equipment used for semipreparative HPTLC purification are as described above in Subheading 2.1. 4. Empty solid-phase extraction (SPE) tubes fitted with frits can be purchased from Sigma-Aldrich and Agela Technologies (Newark, USA).

2.3. Quantification of NGLs

1. Many of the materials and equipment used for NGL quantitation on TLC plates are as described above in Subheading 2.1, but the following are required in addition. 2. Phosphate-buffered saline (PBS) solution: 10 mM phosphate buffer, pH 7.4, 150 mM sodium chloride (prepare freshly from concentrated PBS buffer or PBS tablet).

8  Neoglycolipid-Based Oligosaccharide Microarray System…

123

3. A flying-spot scanning densitometer, e.g., Shimadzu CS-9000 dual wavelength densitometer or Camag Scanner (Muttenz, Switzerland). 2.4. Arraying of NGLs on NitrocelluloseCoated Glass Slides 2.4.1. Preparation of Sample Solutions of NGL (or Glycolipid) Probes for Robotically Arraying

1. Snap-cap polypropylene microcentrifuge tubes (0.5 ml) can be purchased from Alpha Laboratories (Hampshire, UK). 2. Phosphatidylcholine (from egg yolk, ³99% pure) and cholesterol (³99%) can be purchased from Sigma-Aldrich as lyophilized powder. These are made into 200  pmol/ml stock solutions in methanol in 4-ml glass screw-cap glass vials (Chromacol), which can be stored at −20°C (well sealed) over 6 months. 3. The 5 pmol/ml working solution of NGLs and natural glycolipids are prepared as described below in Subheading 3.1–3.3. 4. Cy3 stock solution: 10  mg/ml Cy3 mono NHS ester (GE Healthcare) in HPLC grade water. This solution can be stored at 4°C in the dark for at least 2 years. 5. Ultrasonic water bath (Branson 2510, Danbury, USA).

2.4.2. Generating Microarrays Using Robotic Arrayer and Slide Handling

1. Plastic 384-well plates (Bio-Rad Microseal PCR plates 384well CLR, v2.0); polyolefin sealing film StarSeal for 384-well plates (Starlab, Milton Keynes, UK); Sigma benchtop centrifuge with swing out rotor for microtiter plates (DJB Labcare Ltd, Newport Pagnell, UK). 2. Noncontact microarraying system: We are using a Piezorray instrument (PerkinElmer LAS, Beaconsfield, UK) housed in an environmental enclosure located within a clean laboratory environment. 3. Nitrocellulose-coated single or multiple pad slides are available from commercial sources, e.g., FAST slide from Whatman (Kent, UK) and Nexterion® Slide NC from Schott (Elmsford, USA). 4. Microarray scanner: We are using ProScanArray microarray scanner equipped with red (633 nm) and green (543 nm) lasers and ScanArray Express software (PerkinElmer LAS, Beaconsfield, UK).

2.5. Microarray Analyses of CarbohydrateBinding Proteins

1. FAST slide 16-well incubation chamber with silicon gasket and FAST frame multislide plate are available from Whatman (Kent, UK). 2. Hepes-buffered saline (HBS): 5 mM Hepes, pH 7.4, 150 mM NaCl, 2  mM CaCl2. Other buffers (detergent free) may be used as appropriate for different proteins. 3. Blocking solutions: 3% (w/v) bovine serum albumin (BSA; protease free, Sigma-Aldrich) in HBS.

124

Y. Liu et al.

4. Carbohydrate-binding proteins and detection antibodies: A  wide range of biotinylated plant lectins can be purchased from Vector Laboratories (Peterborough, UK) and EY Laboratories (San Mateo, USA). Biotinylated secondary antibodies (e.g., against human, mouse, rat, rabbit) are available from Sigma-Aldrich, Vector Laboratories and many other suppliers. 5. Alexa Fluor 647-labeled streptavidin is from Molecular Probes (Invitrogen, Carlsbad, USA) available as 1  mg/ml solution. This can be stored as 20 ml aliquots at −20°C (in dark) and is stable for several years.

3. Methods 3.1. Procedure for Preparation of NGLs from Reducing Oligosaccharides

1. Dry down 100 nmol of oligosaccharides in a 1.1-ml tapered screw-cap glass vial either by lyophilization or under a stream of nitrogen gas. Add 1 ml of water to dissolve the oligosaccharide (see Note 1). 2. Add 100 ml of DHPE stock solution and 20 ml of freshly prepared sodium cyanoborohydride (10  mg/ml in methanol) to the vial. Seal the vial and heat at 60°C for 16 h. 3. Dilute the reaction mixture with chloroform/methanol/water 25:25:8 to give a concentration of 1  nmol starting oligosaccharide/5 ml, for HPTLC analysis and for storage. 4. Add a suitable solvent (see Note 2) to the TLC tank and allow vapor to equilibrate (>30 min, at ambient temperature). Cut a TLC plate along the aluminum-backed side to the desired size (e.g., 10 cm long and 5–10 cm wide). 5. Apply NGL solution (equivalent to 1 nmol of starting oligosaccharide) to the HPTLC plate using a nitrogen-assisted TLC applicator, at 15 mm from the bottom edge, as a 4 mm band, and allow 15 mm free at both edges. Carefully place the plate into the tank, with minimum disturbance to the equilibrated vapor layer and develop the chromatogram to 5 mm below the top edge. Remove the plate from the tank and air-dry the plate. 6. NGLs and excess lipids can be viewed and image recorded under the UV lamp after staining with the primulin reagent (see Note 3). The staining can be carried out by spraying the HPTLC plate with the primulin reagent until the surface appears slightly wet, and then allowing the surface to dry before viewing under longwave UV light (365 nm). 7. The same HPTLC plate (after primulin staining) can be stained with orcinol reagent for visualization of hexoses (in both

8  Neoglycolipid-Based Oligosaccharide Microarray System…

125

Fig. 3. HPTLC analyses of conjugation mixtures of DHPE-NGLs derived from lactose, 3¢ and 6¢ sialyllactose and their quantitation on an HPTLC plate. (a) Images of the chromatogram of the NGL conjugation mixtures of lactose (lane 1), 3¢ sialyllactose (lane 2 ) and 6¢ sialyllactose (lane 3 ). The NGL products can be visualized with both primulin (left ) and orcinol (right ) staining. Judging from the results in orcinol staining, the conjugation yields are almost 100%, as no reducing oligosaccharides are detected near the origin of the HPTLC plate. Asterisk, NGL derived from a minor component (lactose) present in the starting material, a synthetic compound of 3¢ sialyllactose. (b) Chromatography of the purified NGL products (lane 5, lactose; lane 6, 3¢ sialyllactose; lane 7, 6¢ sialyllactose) on an HPTLC plate and primulin staining for quantitation by densitometry. Malto5-DH (in a reaction mixture, see Note 8) is used as a standard: 500 pmol (lane 1), 250 pmol (lane 2 ); 100 pmol (lane 3 ); 50 pmol (lane 4 ). TLC development solvent system is chloroform/methanol/water 60:35:8 (for both panels).

unconjugated oligosaccharides and the NGL products). The staining can be carried out by spraying with orcinol reagent until the plate appears slightly wet. The plate is then heated in a vented oven at 105°C for 2–5 min until the violet color given by hexoses is maximal (see Note 4). An HPTLC chromatogram of conjugation mixtures of lactose, 3¢- and 6¢-sialyllactose is shown in Fig. 3a. 3.2. Purification of NGL Reaction Products

The NGL conjugation mixtures contain an excess of aminolipid DHPE and salts, such as the reducing agent cyanoborohydride, and possibly some unconjugated oligosaccharides. A two-step procedure, namely, desalting and further purification, is described here for obtaining purified NGL products. Depending on the results of HPTLC analyses of an NGL conjugation mixture, one of the two methods (Option A or B, as described below in Subheadings 3.2.2 and 3.2.3) can be used for the further purification step. In general, if the NGL is derived from an oligosaccharide longer than trisaccharide and appears to be a single band in HPTLC analysis (e.g., lane 3 in Fig. 3a), Option A (see Subheading 3.2.2) can be used; whereas for NGLs of mono- and disaccharides and those giving multiple primulin and orcinol positive bands in HPTLC (e.g., lane 2 in Fig.  3a), Option B (see Subheading  3.2.3) should be used.

126

Y. Liu et al.

Mass spectrometry analysis (see Note 5) should be performed to corroborate the molecular masses of the purified NGLs before quantitation and arraying. 3.2.1. Step 1: Desalting by C18 Cartridge

1. Prepare a C18 cartridge by sequential washing with 2  ml of chloroform/methanol 1:1, 2  ml of methanol and 2  ml of water. 2. Dry the NGL conjugation mixture under a nitrogen stream. Dissolve the dried mixture in a minimal amount of chloroform/methanol/water 15:70:30 (e.g., 100 ml), and carefully apply the sample solution to the prewashed C18 cartridge. Run the column to take up the applied solution. 3. Wash the column sequentially with 2 × 1 ml water followed by 1 ml of methanol, and collect the washings as fractions 1–3. 4. Elute the NGLs and excess lipid with 4 × 500  ml of chloroform/methanol/water 60:35:8. Combine the eluents and evaporate to dryness under a nitrogen stream (see Note 6).

3.2.2. Step 2 – Option A: Further Purification of NGL by Silica Cartridge

1. Wash the cartridge sequentially with 4 ml of methanol, 4 ml of water, 6 ml of 0.2 M ammonium acetate, 12 ml of water, 4 ml of methanol, and 6 ml of chloroform. 2. Dissolve the dried NGL mixture (<100 nmol of starting sugar) with up to 200 ml of chloroform/methanol/water 130:50:9. Apply the solution onto the prewashed cartridge and collect the fall through as fraction 1. 3. Wash the column sequentially with 4 × 300 ml of chloroform/ methanol/water 130:50:9 and collect the washings as fractions 2–5. 4. Elute the cartridge with 4 × 300 ml of chloroform/methanol/ water 60:35:8 followed by 4 × 300  ml of chloroform/methanol/water 25:25:8 and collect these as fractions 6–13. 5. Apply aliquots (1–2 ml) of each fraction for HPTLC analysis to identify, by primulin and orcinol staining, fractions containing the NGLs. 6. Pool the NGL fractions in a glass vial (1.1 or 2 ml with screw cap) and dry under a nitrogen stream.

3.2.3. Step 2 – Option B: Further Purification of NGL by Semipreparative HPTLC

1. Apply NGL reaction mixture as a long band to an aluminumbacked HPTLC plate using the nitrogen-assisted TLC applicator. Try not to overload the plate, e.g., as a guide do not apply more than 15 nmol NGL per cm. 2. Develop the plate using a suitable solvent (see Note 2), and air-dry the plate. 3. To locate the positions of migrating NGL bands, carefully cut out both vertical edges of the plate (see Note 7). The band

8  Neoglycolipid-Based Oligosaccharide Microarray System…

127

position can be identified and lightly pencil marked on the ­silica surface after staining the cut edges with primulin and orcinol. 4. After localizing the NGL bands, cut out the required strips of plate with a scalpel (on the aluminum-backed side of the plate). Carefully loosen and scrape off the silica gel from the TLC strips and collect the gel in a glass vial. 5. Pack the scraped silica gel as a mini-column in an empty SPE tube fitted with frits and elute the NGLs with 3 × 500  ml of chloroform/methanol/water 25:25:8. Evaporate the solution under a stream of nitrogen gas to obtain the purified NGLs (see Note 8). 3.3. Quantitation of NGL Probes Based on the Lipid Tag

Quantitation of NGLs can be carried out on HPTLC plates using a densitometer after staining with primulin reagent by spraying as described above (see Subheading  3.1, step 6). To obtain better quantitation results, some precautions need to be taken and these are described below. 1. As a standard, DH-NGL of maltopentaose is prepared (see Note 9) and applied as 4  mm bands on an HPTLC plate at four levels (500  pmol, 250  pmol, 100  pmol, and 50  pmol) using a nitrogen-assisted TLC applicator. 2. Dissolve the dried purified NGL, to be quantified, in chloroform/methanol/water 25:25:8 to have a stock solution with an estimated concentration in the range of 50–100 pmol/ml. Up to six NGLs can be applied and quantified on the same plate (10 cm long and 10 cm wide); an example of quantifying three NGLs on a plate is shown in Fig. 3b. 3. Carefully place the plate into a TLC tank equilibrated (for at least 1  h) with solvent chloroform/methanol/water 60:35:8 and develop the chromatogram to 5 mm below the top edge. The plate is removed from the tank and air-dried. 4. Slowly lower the plate into a PBS solution, being careful to avoid trapping air bubbles. Leave to soak for 10 min. 5. Transfer the plate without drying to the primulin working solution (stock primulin solution/PBS 1:100, v/v). Leave to soak for 20 min in the dark. 6. Transfer the plate to a fresh PBS solution without drying, and soak for 10 min in the dark. Drain and air-dry the plate. 7. NGLs can be quantified on TLC plates by a TLC densitometer. The densitometer is used in linear reflectance fluorescence mode with excitation at 370 nm and detection through a filter with a cut-off at 460 nm. 8. Dilute the quantified NGLs in chloroform/methanol/water 25:25:8–5  pmol/ml and these are kept in glass sample vials

128

Y. Liu et al.

with screw cap as working solutions. The vials should be well sealed and stored at −20°C (see Subheading 3.4.1 below and Note 10). 3.4. Arraying of NGLs on NitrocelluloseCoated Glass Slides

3.4.1. Preparation of Sample Solutions of NGL (or Glycolipid) Probes for Robotically Arraying

Purified and quantified NGLs in organic solvents (chloroform/ methanol/water, 25:25:8) can be directly arrayed, using a nitrogenassisted TLC applicator, onto nitrocellulose membranes to produce macroarrays for binding assays as described (5). An example is given in Chapter 23 of this book series. We describe below, the procedures in current use in our laboratory for preparing robotically generated microarrays of NGL and glycolipid probes in the screening format (each probe spotted at two levels, 2 and 5 fmol/spot in duplicate), using a noncontact robotic arrayer (see Note 11). 1. Add to a 0.5 ml microtube the two carrier lipids, phosphatidylcholine and cholesterol (5  ml each of the stock solutions at 200 pmol/ml), followed by 10 ml (for the 2 fmol level) or 30 ml (for the 5  fmol level) of the 5  pmol/ml working solution of NGL (or glycolipid) (see Note 12). 2. Many samples can be prepared at one time; the tubes are left uncapped and samples are allowed to evaporate to complete dryness at 37°C in an incubator for ~16 h. 3. Prepare Cy3 water (20 ng/ml, 26 fmol/ml) from the 10 mg/ ml Cy3 stock solution by 1:500 dilution in water (HPLC grade). 4. Reconstitute each dried mixture in 10 ml of Cy3 water, vortex (approximately 5 s), and centrifuge briefly (see Note 13). 5. Sonicate the tubes for 15 min in a sonic water bath at 30°C, and centrifuge at 8,000´ g (approximately 30 s) in a bench microcentrifuge.      

6. The sample solutions are now ready to be transferred, according to the planned microarray layout, into the “source plate” (e.g., a 384-well plate) for arraying. Place a piece of sealing film on top of the plate and centrifuge the plate at 300 × g for 3  min. The samples in the source plate are now ready for arraying. 3.4.2. Generating Microarray Slides Using “Piezorray” and Slide Handling

1. Piezorray is housed in an environmental enclosure located within a clean laboratory environment. Air supply to the laboratory is passed through an HEPA filter and the room has a slight positive pressure relative to the exterior. The environmental temperature is controlled in the range 19°C–20°C and the humidity within the enclosure is maintained in the range 40–50% using, if necessary, pumped humidified air.

8  Neoglycolipid-Based Oligosaccharide Microarray System…

129

2. The Piezorray source plate holder is temperature-controlled by circulating water chilled to a temperature just above the dew point to avoid condensation on the metal plate holder. Operation of the Piezorray is through manufacturer’s supplied software run from a dedicated PC. 3. Carefully remove the sealing film on the source plate and fit the plate on the chilled metal source plate holder. Start the preconstructed array program (see Note 14). 4. After arraying, carefully remove the microarray slides in order and store in the original plastic slide container. Slides can be stored at 19°C–20°C with no discernible loss of performance over a period of 1 year or more. 5. The quality of microarrays can be monitored by evaluation of the spot image obtained from scanning each slide for Cy3 dye (543 nm laser) using a microarray scanner. Scan image files are stored and used later in conjunction with Alexa Fluor 647 scan files (see Subheading 3.5) to locate spots at the time of quantitation of carbohydrate binding (Fig. 4).

Fig. 4. The 16-pad nitrocellulose-coated glass slides in current use for generating NGL-based microarrays and their scan images. Sixty four lipid-linked oligosaccharide probes were printed in duplicate at two levels (256 arrayed spots) in each pad with Cy3 dye included in the diluent as a marker. Binding is detected with Alexa Fluor 647-labeled streptavidin.

130

Y. Liu et al.

3.5. Protocols for Microarray Analyses of Carbohydrate-Binding Proteins

The carbohydrate-binding assays are most commonly performed with soluble proteins, such as plant lectins (19), anti-carbohydrate antibodies (31–33) and recombinant proteins with suitable tags (e.g., IgG-Fc) (9, 21) or hexa-His (23, 25). Whole viruses (29, 30) or virus-like particles (22) may also be analyzed, either after suitable labeling (e.g., biotinylation) (22) or using nonblocking antibodies for detection of binding (29, 30). It is generally necessary to have the carbohydrate-binding domains of the soluble proteins in oligomeric or multivalent state (see Note 15). Here, we elaborate on protocols for microarray analyses of biotinylated plant lectins and anti-carbohydrate antibodies (these do not require oligomerization) as a general guide giving the steps involved in a binding assay using microarrays of the lipid-linked probes on the 16-pad nitrocellulose-coated glass slides. Analyses of IgG-Fc- and hexa-His-tagged are described Chapter 23 of this book series. 1. Place a 16 pad silicon gasket over the slide and insert the assembly into an FAST Frame (see Note 16). 2. Wet the pads by pipetting onto the surface HBS buffer. Leave for 1 min and flick to remove solution; briefly blot the inverted slide/frame. 3. Add 140 ml of blocking solution (see Note 17) using an 8-channel multipipette to each pad and place the slide on a horizontal platform. Incubate at ambient temperature for 60 min. 4. Remove the blocker solution by flicking the frame and wash the wells by pipetting HBS buffer (flick to remove solution). Excess fluid is removed by briefly blotting the inverted frame. 5. Add the solutions of biotinylated plant lectins or anti-carbohydrate antibodies (diluted in blocking solution to a desired concentration as recommended by the provider or determined by performing optimization experiments) to the pads (50–100 ml of each), and incubate at ambient temperature for 90  min. Remove overlay materials by flicking or by careful introduction of pipette tips into the well to aspirate solution (see Note 18). 6. Wash pads by pipetting 140 ml HBS buffer into each well and flick or aspirate (see Note 19); three wash steps over 5  min should be performed. 7. For anti-carbohydrate antibodies, add appropriate biotinylated secondary antibody (diluted in blocking solution to the desired concentration recommended by the provider). Incubate at ambient temperature for 60  min and wash as in step 6. For biotinylated plant lectins, omit this step. 8. Add 100 ml of Alexa Fluor 647-labeled streptavidin (1 mg/ml in blocking solution). Incubate at ambient temperature (in dark) for 40 min. Remove by flicking or aspirating.

8  Neoglycolipid-Based Oligosaccharide Microarray System…

131

9. Wash pads by pipetting HBS buffer as in step 6. 10. Carefully remove the slide/gasket from the frame and wash the slide with HBS buffer, followed by purified water (approximately 5 s). Drain the slide and dry in dark at ambient temperature or at 37°C. 11. Scan the slide for Alexa Fluor 647 dye (633 nm laser) using the microarray scanner. Image files of the scanned slides are stored and used in conjunction with Cy3 scan files (to locate spots) for quantitation with the scanner software (see Note 20).

4. Notes 1. The water is added to assist the dissolution of the oligosaccharides in the chloroform/methanol-based reaction mixture. However, as an inhibitory effect of water has been found for reductive amination (34), the amount of water should be strictly kept at less than 5%. For pentasaccharides or shorter oligosaccharides, 1  ml of water (or no water) is preferred; whereas for longer chain oligosaccharides (longer than hexasaccharides) or highly acidic oligosaccharides, 5 ml of water in maximum can be added to 100 nmol of oligosaccharide. Brief sonication (10–15 s) is advised to ensure the dissolution. 2. Different TLC development solvent systems should be used for oligosaccharides of different sizes: Chloroform/methanol/ water 130:50:9, for NGLs of mono- to trisaccharides; 60:35:8, for NGLs of tri- to pentasaccharides; 55:45:10, for NGLs of penta- to octasaccharides; 105:100:28, for NGLs of longer oligosaccharides. 3. Primulin staining is required for visualization of DH-NGLs and the oxime-linked AO-NGLs under longwave UV light; whereas the fluorescent NGLs derived from ADHP can be viewed directly under a longwave UV light (365 nm) without primulin staining, and shortwave UV light (254  nm) gives more sensitive detection (14, 17). 4. As the orcinol reagent contains concentrated sulfuric acid, the spraying process should be carried out in a ventilated TLC spray hood (available from CAMAG) or an acid-resistant fume hood. For color development of the HPTLC plate, a TLC heating plate (CAMAG) set up at 105°C in a fume hood can be used instead of the oven. It should be noted that orcinol staining is for visualization of hexose-containing oligosaccharides (e.g., those having glucose, galactose, mannose, or fucose residues). It is not applicable to the visualization of nonhexose oligosaccharides, e.g., oligosaccharides derived from chitin,

132

Y. Liu et al.

polysialic acid, and glycosaminoglycans. Resorcinol reagent (14) can be used for specific staining of sialic acid containing oligosaccharides on TLC plates. 5. Molecular masses of the NGLs can be determined by matrixassisted laser desorption/ionization mass spectrometry as ­previously described (14). 6. Careful elution, namely collection of fractions of smaller volume (e.g., 200 ml) or using shallower gradient of polar solvent compositions (e.g., by including intermediate eluting solvents, such as chloroform/methanol/water 50:70:30 and 25:25:8 before 60:35:8), may achieve separation of the NGL products from both excess lipid and salt. HPTLC analysis of the fractions would be required to identify the fractions of NGLs and lipids. 7. Cutting should be carefully performed on the aluminumbacked side of the plate on a clean surface using a scalpel. For a 100 × 100 mm HPTLC plate with a 70 mm wide band applied in the middle, a 17 mm wide strip can be cut out on each vertical side for subsequent TLC staining procedures. For the preparative TLC purification of ADHP-derived fluorescent NGLs, the NGL band can be visualized directly under UV and thus there is no need to cut out the edges of the plate. 8. It may be necessary to remove the residual silica gel contained in the NGL solution when the preparative TLC method is used for purification: A small amount of chloroform/methanol/ water (e.g., 100 ml) is added for dissolution of the dried NGL in the 1.1 tapered glass vial. The undissolved silica residue can thus be separated after centrifugation and transfer of the NGL solution to a new sample vial. This reextraction process is normally repeated at least three times to minimize the loss of NGL. 9. The maltopentaose-DH standard can be prepared essentially as described in Subheading 3.1 using anhydrous maltopentaose (>95% pure, Sigma-Aldrich) as the starting material. To ensure a complete conversion of maltopentaose to its NGL, the reaction is finished by evaporation to dryness of the reaction mixture at 60°C (over approximately 1 h) by loosening the screw cap of the reaction vial. The standard maltopentaose-DH solution is made by adding 2 ml of chloroform/methanol/water 25:25:8 to the dried mixture without purification. HPTLC analysis (as in Subheading 3.1) can be carried out to corroborate the full conversion of the maltopentaose to NGL. 10. NGLs prepared by oxime-ligation, and natural and synthetic glycolipids are quantified in the same way using maltopentaoseDH as the standard. ADHP-derived fluorescent NGLs may be quantified by densitometry on HPTLC plates and by spectrophotometry in solution (14).

8  Neoglycolipid-Based Oligosaccharide Microarray System…

133

11. The use of contact arrayers has not proven ideal for generating microarrays of lipid-linked oligosaccharides into nitrocellulosecoated glass slides since damage may occur to the delicate nitrocellulose membrane of the slide. 12. Probes should be prepared under clean laboratory conditions; workers should wear disposable powder free gloves and a face mask. NGL working solutions stored at −20°C need to be warmed up to ambient temperature before pipetting. Seal the vials immediately after the pipetting is finished to minimize the evaporation of the organic solvents. 13. Dried lipid/NGL mixture should be reconstituted on the day of arraying. 14. As different laboratories may have different arraying operation systems, we do not go into technical and programming details of using Piezorray. The arraying time for one source plate is set to be no longer than 4 h in order to minimize any evaporation of the samples in the plate. Multiple source plates can be used, if required, to complete the arraying process. 15. Whereas proteins that are multivalent or have sufficiently high avidities for carbohydrate ligands are used without further treatment, those comprising single or double carbohydratebinding domains often require “artificial” oligomerization, e.g., by precomplexing with nonblocking antibodies in order to elicit detectable binding signals. As an example, a recombinant IgG Fc chimera is usually precomplexed with biotinylated anti-IgG in a 1:3 (w/w) ratio, and the complex is allowed to stand at ambient temperature for approximately 45 min before dilution for the binding experiments. 16. The gasket/slide combination should be a tight fit in the FAST frame (not to allow leaking from one pad to another); be careful that the slide does not break when assembling. 17. Optimization of blocking conditions is often required when analyzing a new protein. Frequently used blocking solutions include (a) 3% BSA in HBS and (b) blocker Casein in TBS, 1% (w/v) (Pierce, Rockford, USA) with addition of 1% (w/v) BSA. It should be noted that different batches (lots) of the same product (e.g., the Casein blocker) have been found to have differing blocking strengths (supplier may not provide an explanation). A variety of commercial blocking solutions may also be used, but the solutions containing detergents, which are likely to interfere the immobilization of lipid-linked oligosaccharide probes, should be avoided. 18. Introduction of pipette tips into the well to aspirate solution must be done very carefully (try to position the tips at a corner), as it may cause irreparable damage/scratching to the nitrocellulose surface. Aspiration is obviously necessary when

134

Y. Liu et al.

analyzing infectious agents. Care should be taken when handling infectious agents to ensure all safety precautions are in place and a risk assessment has been undertaken. If necessary, the binding assay should be performed inside a Class 1 microbiological safety cabinet and all liquid and solid waste disposed of appropriately according to local rules. It may be necessary to “fix” (using, for example, 4% paraformaldehyde) the slide before it can be handled in the open laboratory and put through the slide scanner; we typically include a fixation step after the incubation of influenza viruses. 19. It is important not to let the nitrocellulose dry once the protein being investigated has been added. 20. We use PerkinElmer ScanArray Express software for quantifying the bound fluorescence in the array spots, and a dedicated in-house-designed software suite for storing, retrieving, and displaying carbohydrate microarray data (35).

Acknowledgments We gratefully acknowledge collaborations of other members (past and present) in the Glycosciences Laboratory, in particular Alex Lawson, Yibing Zhang, and Colin Herbert. The Glycosciences Laboratory acknowledges with gratitude collaborators over the years with whom our microarray probes were studied. We also acknowledge UK Research Councils’ Basic Technology Initiative “Glycoarrays” (GRS/79268) and follow on EPSRC Translational Grant (EP/G037604/1), MRC and Wellcome Trust and NCI Alliance of Glycobiologists for Detection of Cancer and Cancer Risk (U01 CA128416) for financial support. A.S.P. is a fellow of the Fundação para a Ciência e Tecnologia (SFRH/ BPD/26515/2006, Portugal) and M.A.C. of the Ministerio de Ciencia e Innovación (JCI-2007-123-160, Spain). References 1. Tang PW, Gooi HC, Hardy M, Lee YC, Feizi T (1985) Novel approach to the study of the antigenicities and receptor functions of carbohydrate chains of glycoproteins. Biochem Biophys Res Commun 132:474–480 2. Yuen C-T, Chai W, Loveless RW, Lawson AM, Margolis RU, Feizi T (1997) Brain contains HNK-1 immunoreactive O-glycans of the sulfoglucuronyl lactosamine series that terminate in 2-linked or 2,6-linked hexose (mannose). J Biol Chem 272:8924–8931 3. Mizuochi T, Loveless RW, Lawson AM, Chai W, Lachmann PJ, Childs RA, Thiel S, Feizi T

(1989) A library of oligosaccharide probes (neoglycolipids) from N-glycosylated proteins reveals that conglutinin binds to certain complex type as well as high-mannose type oligosaccharide chains. J Biol Chem 264: 13834–13839 4. Leteux C, Chai W, Nagai K, Herbert CG, Lawson AM, and Feizi T (2001) 10E4 Antigen of scrapie lesions contains an unusual non­ sulfated heparan motif. J Biol Chem 276:12539–12545 5. Fukui S, Feizi T, Galustian C, Lawson AM, Chai W (2002) Oligosaccharide microarrays

8  Neoglycolipid-Based Oligosaccharide Microarray System… for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nat Biotechnol 20:1011–1017 6. Feizi T, Chai W (2004) Oligosaccharide microarrays to decipher the glyco code. Nat Rev Mol Cell Biol 5:582–588 7. Galustian C, Park CG, Chai W, Kiso M, Bruening SA, Kang YS, Steinman RM, Feizi T (2004) High and low affinity carbohydrate ligands revealed for murine SIGN-R1 by carbohydrate array and cell binding approaches, and differing specificities for SIGN-R3 and langerin. Int Immunol 16:853–866 8. Reddy ST, Chai W, Childs RA, Page JD, Feizi T, Dahms NM (2004) Identification of a Low Affinity Mannose 6-Phosphate-binding Site in Domain 5 of the Cation-independent Mannose 6-Phosphate Receptor. J Biol Chem 279:38658–38667 9. Palma AS, Feizi T, Zhang Y, Stoll MS, Lawson AM, Diaz-Rodríguez E, CampaneroRhodes AS, Costa J, Brown GD, Chai W (2006) Ligands for the beta-glucan receptor, Dectin-1, assigned using ‘designer’ microarrays of oligosaccharide probes (neoglycolipids) generated from glucan polysaccharides. J Biol Chem 281:5771–5779 10. Komba S, Galustian C, Ishida H, Feizi T, Kannagi R, Kiso M (1999) The first total synthesis of 6-sulfo-de-N-acetylsialyl Lewisx ganglioside: A superior ligand for human L-selectin. Angew Chem Int Ed Engl 38: 1131–1133 11. Green PJ, Tamatani T, Watanabe T, Miyasaka M, Hasegawa A, Kiso M, Stoll MS, Feizi T (1992) High affinity binding of the leucocyte adhesion molecule L-selectin to 3¢-sulphatedLea and -Lex oligosaccharides and the predominance of sulphate in this interaction demonstrated by binding studies with a series of lipid-linked oligosaccharides. Biochem Biophys Res Commun 188:244–251 12. Feizi T, Stoll MS, Yuen C-T, Chai W, Lawson AM (1994) Neoglycolipids: probes of oligosaccharide structure, antigenicity and function. Methods Enzymol 230:484–519 13. Feizi T, Childs RA (1994) Neoglycolipids: probes in structure/function assignments to oligosaccharides. Methods Enzymol 242: 205–217 14. Chai W, Stoll MS, Galustian C, Lawson AM, Feizi T (2003) Neoglycolipid technology deciphering information content of glycome. Methods Enzymol 362:160–195 15. Yuen C-T, Lawson AM, Chai W, Larkin M, Stoll MS, Stuart AC, Sullivan FX, Ahern TJ, Feizi T (1992) Novel sulfated ligands for the cell adhesion molecule E-selectin revealed by the neoglycolipid technology among O-linked

135

oligosaccharides on an ovarian cystadenoma glycoprotein. Biochemistry 31:9126–9131 16. Chai W, Yuen CT, Kogelberg H, Carruthers RA, Margolis RU, Feizi T, Lawson AM (1999) High prevalence of 2-mono- and 2,6-di-substituted Manol-terminating sequences among O-glycans released from brain glycopeptides by reductive alkaline hydrolysis. Eur J Biochem 263:879–888 17. Stoll MS, Feizi T, Loveless RW, Chai W, Lawson AM, Yuen CT (2000) Fluorescent neoglycolipids. Improved probes for oligosaccharide ligand discovery. Eur J Biochem 267:1795–1804 18. Liu Y, Chai W, Childs RA, Feizi T (2006) Preparation of Neoglycolipids with Ring-Closed Cores via Chemoselective Oxime-Ligation for Microarray Analysis of Carbohydrate-Protein Interactions. Methods Enzymol 415C:326–340 19. Liu Y, Feizi T, Campanero-Rhodes MA, Childs RA, Zhang Y, Mulloy B, Evans PG, Osborn HM, Otto D, Crocker PR, Chai W (2007) Neoglycolipid probes prepared via oxime ligation for microarray analysis of oligosaccharideprotein interactions. Chem Biol 14:847–859 20. Liu Y, Palma AS, Feizi T (2009) Carbohydrate microarrays: key developments in glycobiology. Biol Chem 390:647–656 21. Campanero-Rhodes MA, Childs RA, Kiso M, Komba S, Le Narvor C, Warren J, Otto D, Crocker PR, Feizi T (2006) Carbohydrate microarrays reveal sulphation as a modulator of siglec binding. Biochem Biophys Res Comm 344:1141–1146 22. Campanero-Rhodes MA, Smith A, Chai W, Sonnino S, Mauri L, Childs RA, Zhang Y, Ewers H, Helenius A, Imberty A, Feizi T (2007) N-glycolyl GM1 ganglioside as a receptor for simian virus 40. J Virol 81: 12846–12858 23. Schallus T, Jaeckh C, Feher K, Palma AS, Liu Y, Simpson JC, Mackeen M, Stier G, Gibson TJ, Feizi T, Pieler T, Muhle-Goll C (2008) Malectin - A Novel Carbohydrate-binding Protein of the Endoplasmic Reticulum and a Candidate Player in the Early Steps of Protein N-glycosylation. Mol Biol Cell 19:3404–3414 24. Palma AS, Liu Y, Muhle-Goll C, Butters TD, Zhang Y, Childs R, Chai W, Feizi T (2010) Multifaceted approaches including neoglycolipid oligosaccharide microarrays to ligand discovery for malectin. Methods Enzymol 478:265–286 25. Blumenschein TMA, Friedrich N, Childs RA, Saouros S, Carpenter EP, Campanero-Rhodes MA, Simpson P, Chai W, Koutroukides T, Blackman MJ, Feizi T, Soldati-Favre D, Matthews S (2007) Atomic resolution insight into host cell recognition by Toxoplasma gondii. EMBO J 26:2808–2820

136

Y. Liu et al.

26. Garnett JA, Liu Y, Leon E, Allman SA, Friedrich N, Saouros S, Curry S, Soldati-Favre D, Davis BG, Feizi T, Matthews S (2009) Detailed insights from microarray and crystallographic studies into carbohydrate recognition by microneme protein 1 (MIC1) of Toxoplasma gondii. Protein Sci 18:1935–1947 27. Allman SA, Jensen HH, Vijayakrishnan B, Garnett JA, Leon E, Liu Y, Anthony DC, Sibson NR, Feizi T, Matthews S, Davis BG (2009) Potent Fluoro-oligosaccharide Probes of Adhesion in Toxoplasmosis. Chembiochem 10:2522–2529 28. Friedrich N, Santos JM, Liu Y, Palma AS, Leon E, Saouros S, Kiso M, Blackman MJ, Matthews S, Feizi T, Soldati-Favre D (2010) Members of a novel protein family containing MAR domains act as sialic acid-binding lectins during host cell invasion by apicomplexan parasites. J Biol Chem 285:2064–2076 29. Childs RA, Palma AS, Wharton S, Matrosovich T, Liu Y, Chai W, Campanero-Rhodes MA, Zhang Y, Eickmann M, Kiso M, Hay A, Matrosovich M, Feizi T (2009) Receptorbinding specificity of pandemic influenza A (H1N1) 2009 virus determined by carbohydrate microarray. Nat Biotechnol 27:797–799 30. Liu Y, Childs RA, Matrosovich T, Wharton S, Palma AS, Chai W, Daniels R, Gregory V, Uhlendorff J, Kiso M, Klenk HD, Hay A, Feizi T, Matrosovich M (2010) Altered receptor specificity and cell tropism of D222G hemagglutinin mutants isolated from fatal cases of pandemic A(H1N1) 2009 influenza virus. J Virol 84:12069–12074

31. Capodicasa C, Chiani P, Bromuro C, De BF, Catellani M, Palma AS, Liu Y, Feizi T, Cassone A, Benvenuto E, Torosantucci A (2011) Plant production of anti-beta-glucan antibodies for immunotherapy of fungal infections in humans. Plant Biotechnol J. 9:776–787 32. Torosantucci A, Chiani P, Bromuro C, De Bernardis F, Palma AS, Liu Y, Mignogna G, Maras B, Colone M, Stringaro A, Zamboni S, Feizi T, Cassone A (2009) Protection by antibeta-glucan antibodies is associated with restricted beta-1,3 glucan binding specificity and inhibition of fungal growth and adherence. PLoS ONE 4:e5392 33. Dunlop C, Bonomelli C, Mansab F, Vasiljevic S, Doores KJ, Wormald MR, Palma AS, Feizi T, Harvey DJ, Dwek RA, Crispin M, Scanlan CN (2010) Polysaccharide mimicry of the epitope of the broadly neutralising anti-HIV antibody, 2G12, induces enhanced antibody responses to self oligomannose glycans. Glycobiology 20:812–823 34. Stoll MS, Mizuochi T, Childs RA, Feizi T (1988) Improved procedure for the construction of neoglycolipids having antigenic and lectin-binding activities from reducing oligosaccharides. Biochem J 256:661–664 35. Stoll, Feizi T (2009) Software tools for storing, processing and displaying carbohydrate microarray data. In: Kettner C (ed) Proceeding of the Beilstein Symposium on Glyco-Bioinformatics, 4–8 October, 2009, Potsdam, Germany. pp 123–140 (available online at http://www.beilsteininstitut.de/en/symposia/overview/procee dings/2009-glycobioin­formatics/)

Chapter 9 Preparation of a Mannose-6-Phosphate Glycan Microarray Through Fluorescent Derivatization, Phosphorylation, and Immobilization of Natural High-Mannose N-Glycans and Application in Ligand Identification of P-Type Lectins Xuezheng Song, Jamie Heimburg-Molinaro, Nancy M. Dahms, David F. Smith, and Richard D. Cummings Abstract Glycan microarrays prepared by immobilization of amino-functionalized glycans on NHS-activated glass slides have been successfully used to study protein–glycan interactions. Fluorescently tagged glycans with an amino functional group can be prepared from natural glycans released from glycoproteins. These tagged glycans can be enzymatically modified with various glycosyltransferases, phosphotransferases, sulfotransferases, etc., to quickly expand the size and diversity of the tagged glycan libraries (TGLs). The TGLs, presented in the format of microarrays, provide a convenient platform for identifying the glycan ligands of glycan-binding proteins (GBPs). The chapter provides the background to prepare a defined glycan microarray and uses as an example glycans generated as phosphodiesters and phosphomonoesters of high-mannose type N-glycans. The method describes the preparation of high-mannose type glycan-AEAB conjugates (GAEABs), the purification of their phosphodiesters, and the subsequent mild acid hydrolysis to obtain corresponding phosphomonoesters. These GAEABs are covalently printed as a phosphorylated glycan microarray and used for analysis of the glycan ligand specificities of P-type lectins, such as the mannose-6phosphate receptors (Man-6-P receptors or MPRs). Key words: Glycan microarray, Glycan derivatization, Phosphorylation, Mannose-6-phosphate, Man-6-P receptor, P-type lectin

1. Introduction Historically, the identification and characterization of protein–glycan interactions were determined using hapten inhibition of agglutination assays or hapten inhibition of binding detected in ELISA-like formats.

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_9, © Springer Science+Business Media, LLC 2012

137

138

X. Song et al.

These assays were sequential (one glycan hapten at a time), labor-intensive, and required large amounts of samples, which were difficult to obtain. The development of glycan microarrays (1–4), which is an extension of gene microarray technologies for genomics and protein microarray technologies of proteomics, has transformed the study of protein–glycan interactions into a microscale assay that can analyze hundreds of glycans simultaneously for their ability to bind proteins. The repertoire of glycans encompasses compounds that are either chemo/enzymatically synthesized (5, 6) or isolated from natural sources (7–9) and are able to be covalently (5, 10, 11) or non-covalently (12, 13) immobilized to an activated glass surface. This method can be used to test the binding specificity of any potential glycan binding protein (GBP) that can be detected fluorescently, including lectins, antibodies, viruses, and receptors. The high-mannose type phosphorylated N-glycan array (14) is an example of a defined glycan array generated from naturally occurring glycans. In this example, glycans released by PNGase F digestion of bovine ribonuclease B (RNaseB) and soybean agglutinin (SBA) are purified and labeled with a bifunctional fluorescent linker; subsequently, the isomeric structures are separated and phosphorylated using recombinant form of the UDPGlcNAc:glycoprotein N-acetylglucosamine-1-phosphotransferase (GlcNAc-phosphotransferase) (15). The resulting phosphodiesters are separated by HPLC, purified, structurally identified, and subsequently converted to their corresponding phosphomonoesters by mild acid hydrolysis as described previously (14). Thus, a library of fluorescently labeled N-linked high-mannose type N-glycans and their phosphodiesters and phosphomonoesters was available for printing a phosphorylated N-glycan array. Prior to this array, there were no formats available for studying large panels of phosphomonoesters and phosphodiesters of high-mannose N-glycans, due to the lack of methods for their chemical synthesis. The availability of a large panel of phosphomonoesters and phosphodiesters of high-mannose N-glycans permitted the investigation of the detailed specificity of P-type lectins, notably the cation-dependent and cation-independent mannose-6-phosphate receptors (Man-6-P receptors or MPRs), which are critical for the intracellular delivery of hydrolytic enzymes to the lysosome (16–19). Although the specificity of MPRs for phosphomonoesters and phosphodiesters had been observed (20), the array described here provides a platform for studying these important interactions (14, 21) and is available as a general platform for studies on P-type lectins.

9  Preparation of a Mannose-6-Phosphate Glycan Microarray…

139

2. Materials 2.1. Preparation of Functional Fluorescent High-Mannose Glycan Conjugates

1. Ribonuclease B (RNaseB) (Sigma-Aldrich). 2. SBA (prepared as described previously) (22). 3. PNGase F (New England Biolabs), including glycoprotein denaturing buffer (10×), G7 reaction buffer (10×), and NP-40 (10%). 4. 2-(N-aminoethyl)amino benzamide (AEAB) (9). 5. C18 Sep-Pak column (Millipore, 500 mg and 2 g). 6. Carbograph SPE columns (Alltech, 150  mg, 300  mg and 1 g). 7. Conjugation solvent: 7/3 (v/v) dimethyl sulfoxide (DMSO)/ acetic acid (AcOH). 8. Methanol. 9. Sodium 95%).

cyanoborohydride

(NaCNBH3)

(Sigma-Aldrich,

10. Trifluoroacetic acid (TFA) (Fisher Scientific, HPLC grade). 11. MilliQ water (dH2O). 12. Analytical HPLC system with UV and Fluorescence detection (Shimadzu). 13. Semipreparative normal-phase (250 × 9.2 mm, Agilent).

(NP)

HPLC

column

14. Hypercarb HPLC column (PGC-HPLC column, 150 × 4.6 mm, Thermo Scientific). 15. Hypercarb guard column (20 × 4 mm, Thermo Scientific). 16. Centri-vap (Labconco) at room temperature. 17. Sodium azide (NaN3). 18. Centrifuge filter (Nylon 0.2 mm, Costar). 19. MALDI-TOF/TOF (Bruker UltraflexII). 20. Matrix solution: 5 mg/mL 2,5-dihydroxybenzoic acid in 50% acetonitrile with 0.1% TFA. 21. Solution A: 50% acetonitrile in 0.1% TFA. 22. HPLC solvent A: Acetonitrile (Fisher Scientific, HPLC grade). 23. HPLC solvent B: 0.25 M ammonium acetate, pH 4.5. 24. HPLC solvent C: water. 25. HPLC solvent D: 1% TFA in water. 26. Normal phase (NP) HPLC linear gradient: 0 min: 80% A, 4% B, 16% C; 90 min: 10% A, 50% B, 40% C.

140

X. Song et al.

2.2. Synthesis of GlcNAc-6-PhosphoGAEABs and 6-Phospho-GAEABs

1. Fluorescent conjugates of purified isomers of high-mannose glycans were converted to their corresponding phosphodiesters using a recombinant GlcNAc-phosphotransferase (20) and kindly provided by Stuart Kornfeld and Marielle Boonen (Washington University, St. Louis, MO). 2. Hydrochloric acid (HCl). 3. Acetonitrile. 4. Trifluoroacetic acid. 5. LNnT-AEAB (prepared as described previously from LNnT) (9). 6. Hypercarb HPLC column (PGC-HPLC column, 150 × 4.6 mm, Thermo Scientific). 7. Hypercarb guard column (20 × 4 mm, Thermo Scientific). 8. Phosphorylation buffer: 0.1 M Tris-Hcl, pH 7.5, 0.1 M MgCl2, and 0.1 M MnCl2. 9. Hypercarb (PGC) HPLC linear gradient: 0 min: 15% A, 75% C, 10% D; 30 min: 45% A, 45% C, 10% D.

2.3. Printing Microarrays

1. MALDI-TOF/TOF (Bruker UltraflexII). 2. Piezorray noncontact printer (Perkin-Elmer). 3. ProScanArray Scanner (Perkin-Elmer). 4. Printing buffer (2×): 600 mM sodium phosphate buffer, pH 8.5. 5. NHS-activated America).

microarray

glass

slides

(Schott

North

6. Biotin-hydrazine (Sigma-Aldrich). 7. Blocking buffer: 50  mM ethanolamine in 50 mm sodium borate So Tris buffer, pH 9.0. 8. 384-well plate, V-shape (Biorad). 2.4. Interrogation of Phosphorylated Glycan Microarray with CD-MPR and CI-MPRs

1. CI-MPR and CD-MPR and antibodies for their detection were kindly provided by Nancy Dahms (Medical College of Wisconsin, Milwaukee, WI). 2. Assay buffer: 50  mM imidazole, pH 6.5, 150  mM NaCl, 10 mM MnCl2. 3. Binding buffer: Assay buffer containing 1% bovine serum albumin (BSA) and 0.05% Tween-20. 4. 16-Chamber adapter (Grace Biolabs). 5. Coplin jars. 6. Biotinylated lectins Concanavalin A (ConA) and Ricinus communis agglutinin-I (RCA-I) (Vector labs). 7. Cyanine5 (Cy5)-streptavidin (Zymed).

9  Preparation of a Mannose-6-Phosphate Glycan Microarray…

141

8. Rabbit polyclonal antibodies specific for CD-MPR (B3.5) and CI-MPR (B14.5) (Generated in rabbits immunized with corresponding MPRs purified from bovine liver). 9. Alexa Fluor 488-Goat anti-rabbit IgG (Invitrogen).

3. Methods 3.1. Preparation of Functional Fluorescent High-Mannose Glycan Conjugates from RNaseB

1. Dissolve RNaseB (100 mg) in 5 mL 1× glycoprotein denaturing buffer and boil for 10 min at 100°C. To the cooled solution, add 1  mL of 10% NP40, 1  mL of 10× G7 buffer and 1  mL of 0.2% sodium azide solution. Then add PNGase F (30 mL, 500,000 units/mL) and 2 mL water. Digest the glycoprotein at 37°C for 3  days in a stoppered tube to prevent loss of water (see Note 1). 2. Precondition C18 Sep-pak cartridges by washing with 1 column volume (cv) of methanol followed by 2 cv of dH2O; Precondition carbograph cartridges by washing with 1 column volume (cv) Solution A followed by 3 cv dH2O. 3. The PNGase F digested solution is boiled for 5  min, passed through a 2  g C18 Sep-Pak cartridge and washed with dH2O (see Note 2). The flow-through and 2 cv water wash are pooled and applied on a 1 g carbograph cartridge, which is subsequently washed with 6 cv dH2O. Free glycans are then eluted from the carbograph cartridge with 3 cv of Solution A. The eluent is dried by Centri-vap prior to fluorescent derivatization. 4. For fluorescent derivatization, prepare a solution of AEAB by adding 88 mg AEAB to 1 mL conjugation solvent; vortex the mixture for 10 min. This solution can be stored at −20°C for future use. Prepare fresh sodium cyanoborohydride solution by adding 64  mg NaCNBH3 to 1  mL conjugation solvent; vortex for 2 min. 5. To the dried free glycans, add 500 mL AEAB solution and an equal volume of NaCNBH3 solution. Vortex the mixture for 1 min and incubate at 65°C for 2 h. Add acetonitrile (10 mL) to precipitate the glycans. Cool the mixture at −20°C for 30 min. 6. Centrifuge the mixture at 5,000 × g for 5  min. Discard the supernatant. Dry the pellet in a Centri-vap for 5  min and reconstitute it in 500 mL dH2O. The solution is then centrifuge-filtered and ready for normal phase HPLC purification. 7. A linear gradient (0 min: 80% A, 4% B, 16% C; 90 min: 10% A, 50% B, 40% C) is applied for the semipreparative NP-HPLC separation of GAEABs from RNaseB. Five well-separated peaks are collected, which corresponds to Man5-AEAB, Man6-AEAB, Man7-AEAB (isomers), Man8-AEAB (isomers), and Man9AEAB, respectively. Fractions are dried in a Centri-vap.

142

X. Song et al.

8. Resuspend the dried fractions in dH2O and pass them individually through a preconditioned carbograph (use 1 g carbograph for ~10 mg GAEABs) for desalting as described in step 3. The eluent is lyophilized. Quantify these fractions based on their UV 330 nm absorption and fluorescence (Ex 330 nm/Em 420 nm); they are then ready for enzymatic phosphorylation. 9. Characterize the desalted fractions in step 8 using MALDITOF. Spot 0.5  mL matrix solution and then 0.5  mL sample. Run MALDI-TOF using reflective positive mode. 3.2. Preparation of GlcNAc-6-PhosphoGAEABs and 6-Phospho-GAEABs

1. Incubate the purified GAEABs (0.25–0.5 mmol) separately at 37°C with 0.5 mg of purified GlcNAc-phosphotransferase (20), and 1 mmol of UDP-GlcNAc in 20 mL of phosphorylation buffer. The reactions are stopped by freezing and lyophilizing after 48 h incubation. 2. The crude products are reconstituted into 100 mL dH2O and centrifuged. The clear supernatant is injected into porous graphitized carbon (PGC)-HPLC for separation (see Note 3). Apply a linear gradient (0 min: 15% A, 75% C, 10% D; 30 min: 45% A, 45% C, 10% D) (Fig. 1). The mono- and diphosphodiesters elute later than the starting GAEABs. Collect individual fluorescent peaks, which are then analyzed by MALDI-TOF and dried in Centri-vap. Structures are assigned according to the molecular weights and the HPLC behavior. 3. A portion of each fraction of phosphodiester is converted to its corresponding phosphomonoester by mild acid hydrolysis (0.01  M HCl, 1  h, 100°C). After heating in 10  mM HCl (100 mL) at 100°C for 1 h, the solution is cooled down and injected into PGC-HPLC for purification (see Note 4). 4. The AEAB derivatives of the phosphomonoesters, as well as the starting nonphosphorylated and phosphodiesters are quantified based on their UV absorbance relative to LNnT-AEAB standards and reconstituted to 200  mM in dH2O and stored frozen until use (see Note 5).

3.3. Printing Microarray of High-Mannose Glycan AEAB Derivatives and Corresponding Phosphomonoesters and Phosphodiesters

1. We use a noncontact Piezorray printer (Piezorray, PerkinElmer) to print the microarrays. 2. For each sample, 5 mL of the 200 mM solution described above is mixed with 5 mL 2× printing buffer and loaded into a 384well source plate with V-shape wells. The plate is fitted onto a metal adaptor for controlling the temperature during printing and set up securely in the printer. 3. According to the printer software, input the sample names, IDs, and positions. The printing pattern is programmed using the software according to manufacturer’s instructions, including

9  Preparation of a Mannose-6-Phosphate Glycan Microarray…

143

Fig. 1. Purification of phosphodiesters of AEAB-derivatized glycans (GAEABs) by HPLC on PGC-HPLC. The GAEABs prepared from Man5-9 released from RNase B are converted to phosphodiesters by GlcNAc phosphotransferase as described in the text. The starting material (upper profiles) and the corresponding products of the phosphotransferase reactions (lower profiles) were applied to PGC-HPLC, and eluted with a gradient of acetonitrile and dH2O.

subarray number and sample replicate number. In this case, a microarray includes 14 identical subarrays. Each subarray incorporates eight GAEABs, eight phosphodiesters and eight phosphomonoesters along with two control glycans in six replicates. The structures and chart IDs are shown in Fig. 2. The printer software generates a GAL file based on the input, which is used by the Scanner software in the image processing step. 4. Bring NHS-activated slides from −20°C to room temperature in vacuum desiccators before opening the package. Then open the package, and set the desired number of NHS-slides on the printing platform. 5. Tune the four tips of the printer so that ideal morphology and size dispersion are reached according to manufacturer’s instructions. Start the printing process. 6. Perform a pin washing step to avoid carry over contamination from different samples. Pins are washed by pushing out sample with two cycles of water washes. The outside of the pins are washed with flowing water. Pins are then submerged in 1% Tween 20 in printing buffer and then sonicated. Two more full round of wash cycles with water are performed again.

144

X. Song et al.

Fig. 2. The chart IDs of the individual glycans printed on the microarray.

7. After the printing is finished, the slides are set in a box in a 55°C water bath without direct water contact for 1  h. The water bath serves as a high humidity chamber. 8. The slides are washed with 0.05% Tween-20 in PBS, then dH2O, and then blocked with blocking buffer for 1 h, washed again with 0.05% Tween-20 in PBS and subsequently dried by centrifugation and stored desiccated at −20°C until use. 3.4. Interrogation of Phosphorylated Glycan Microarray with CD-MPR and CI-MPRs

1. The binding assay on the microarray includes binding, washing, and scanning steps. The proteins interact and specifically bind to their ligands printed on the microarray during the binding step. Nonspecific bound proteins and excess reagents are washed away in the washing step. The fluorescence-based detection is carried out using the microarray scanner in the last step to give a fluorescence image. In this experiment, biotinylated ConA and RCA I are used as controls to validate printing of the microarray, since they bind to specific types of glycans on the array. RCA-I binds to terminal Galb1, 4-GlcNAc, therefore it should only bind to LNnT-AEAB and NA2-AEAB on the array; ConA binds to the trimannosyl core of N-glycans, therefore it should bind to all glycans printed on the array except LNnT-AEAB, which is not an N-glycan. 2. The bound lectins are detected by a secondary incubation with Cy5-streptavidin. The CD-MPR and CI-MPRs, on the other hand, are detected with specific rabbit polyclonal antibodies followed by Alexa Fluor488-labeled goat anti-rabbit IgG.

9  Preparation of a Mannose-6-Phosphate Glycan Microarray…

145

3. Remove slide(s) from desiccators and label slide as necessary on the top of the slide. 4. Do not write/touch on the slide where the glycans are printed. 5. Hydrate the slide by placing in a glass Coplin staining jar containing 100 mL of assay buffer for 5 min (see Note 6). 6. For multipanel experiments on a single slide, affix the multichamber adaptor on the slide to separate a single slide into 14 chambers sealed from each other during the assay. This allows for simultaneous multiple assays on up to 14 identical subarrays on the same slide. Apply sample (for example, biotinylated lectin, 50–100 mL) into each chamber and incubate the slide on a shaker at 60 rpm for 1 h. If some subarrays on one slide are not used, save the slide at −20°C for use at a later time. 7. For evaluation of printing efficiency, dissolve biotinylated lectins (ConA and RCA-I) in binding buffer (assay buffer containing 1% BSA and 0.05% Tween-20) at appropriate concentrations. Add the lectin solutions (50–100 mL) to separate chambers and incubate for 1 h at room temperature. Wash the chambers three times with the following buffers: (a) assay buffer containing 0.05% Tween-20 and (b) assay buffer. Detect bound lectins by a secondary incubation with 1 mg/mL streptavidin labeled with Cy5 and subsequently washed as described above. To remove salts prior to drying, wash the chambers three times with dH2O. 8. For assay with MPRs, apply proteins at the indicated concentrations in 70 mL of binding buffer to glycan arrays separated by the 16-chamber adapter, incubated for 1  h at room temperature, and processed as described above. After washing, apply 70 mL of 1:250 dilutions of rabbit polyclonal antibodies specific for CD-MPR and CI-MPR in binding buffer to appropriate subarrays and incubate for 1 h at room temperature. To detect MPRs, wash the subarrays as described, and add 70 mL of goat anti-rabbit IgG labeled with Alexa Fluor488 to each subarray and incubate for 1 h at room temperature. After washing, air-dry the slides. 9. Scan slide with a fluorescent scanner at the appropriate wavelengths. For Alexa Fluor488-labeled secondary antibodies, scan with 488 nm lamp, and to detect Cy5-streptavidin, scan with a 633 nm lamp. Save each image separately. 10. Process the image with Scanarray quantitation software to generate an excel worksheet as the raw data. Histograms with X-axis representing glycan chart IDs and Y-axis representing relative fluorescent units are generated for each protein sample. The binding of plant lectins and MPRs are shown in Figs. 3 and 4, respectively.

146

X. Song et al.

Fig. 3. Binding of plant lectins to the phosphorylated high-mannose glycan microarray. The glycan microarray was printed as described in the text and the individual glycans are identified by their glycan number as indicated in Fig. 2. (a) Biotinylated ConA (0.5 mg/mL) detected with 5 mg/mL of Cy5-labeled streptavidin; (b) Biotinylated RCA-I (0.1 mg/mL) detected with 5 mg/mL of Cy5-labeled streptavidin. Error bars indicate ±1 standard deviation.

4. Notes 1. In this protocol, PNGase F digestion is applied directly to denatured RNaseB. For other glycoproteins, especially when larger amounts of glycoprotein are needed, the glycoproteins can be reduced and alkylated, then trypsinized. The dialyzed tryptic peptides are digested with PNGase F, after which the C18-carbograph purification procedure can be applied. This procedure increases the efficiency of PNGase F. 2. The capacity of C18 Sep-pak and carbograph varies toward various glycans. Typically, for 100  mg glycoprotein starting material, a 2 g C18 cartridge and a 1 g carbograph cartridge are used to obtain the free reducing glycans. 3. PGC-HPLC (Hypercarb) is a well-suited reverse phase system for the purification of phosphomonoesters and phosphodiesters. It has a wide pH range (0–14) and high capacity. The fractions collected can be dried and directly used for printing.

9  Preparation of a Mannose-6-Phosphate Glycan Microarray…

147

Fig. 4. Binding of CD-MPR and CI-MPR to the phosphorylated high-mannose glycan microarray. (a) CD-MPR (50 mg/mL) detected with rabbit antibody B3.5 (1:250) and Alexa Fluor488-labeled goat anti-rabbit IgG (5  mg/mL); (b) CI-MPR (50  mg/mL) detected with rabbit antibody B14.5 (1:250) and Alexa Fluor488-labeled goat anti-rabbit IgG (5 mg/mL).

4. Phosphodiesters are relatively unstable toward acidic conditions, therefore need to be handled more carefully. However, the phosphomonoesters are fairly stable. 5. Extreme caution should be used to avoid contamination of buffers or glassware with alkaline phosphatase that might come into contact with phosphorylated glycans. 6. We have also used Tris buffers and MES buffers at pH 6.5 which is the pH optimum for ligand binding by the MPRs. Binding of the CD-MPR is influenced by divalent cations, with highest binding affinities observed in the presence of MnCl2, whereas the binding affinity of the CI-MPR is not significantly affected by divalent cations. References 1. de Paz JL, and Seeberger PH. (2006) Recent advances in carbohydrate microarrays. QSAR & Combinatorial Science 25, 1027–1032. 2. Feizi T, and Chai W. (2004) Oligosaccharide microarrays to decipher the glyco code. Nat Rev Mol Cell Biol 5, 582–588.

3. Hirabayashi J. (2003) Oligosaccharide microarrays for glycomics. Trends Biotechnol 21, 141–143. 4. Park S, Lee MR, and Shin I. (2008) Carbohydrate microarrays as powerful tools in studies of carbohydrate-mediated biological processes. Chem Commun (Camb), 4389–4399.

148

X. Song et al.

5. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, Alvarez R, Bryan MC, Fazio F, Calarese D, Stevens J, Razi N, Stevens DJ, Skehel JJ, van Die I, Burton DR, Wilson IA, Cummings R, Bovin N, Wong CH, and Paulson JC. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101, 17033–17038. 6. Ratner DM, Adams EW, Su J, O’Keefe BR, Mrksich M, and Seeberger PH. (2004) Probing protein-carbohydrate interactions with microarrays of synthetic oligosaccharides. Chembiochem 5, 379–382. 7. de Boer AR, Hokke CH, Deelder AM, and Wuhrer M. (2007) General microarray technique for immobilization and screening of natural glycans. Anal Chem 79, 8107–13. 8. Song X, Lasanajak Y, Xia B, Smith DF, and Cummings RD. (2009) Fluorescent glycosylamides produced by microscale derivatization of free glycans for natural glycan microarrays. ACS Chem Biol 4, 741–750. 9. Song X, Xia B, Stowell SR, Lasanajak Y, Smith DF, and Cummings RD. (2009) Novel fluorescent glycan microarray strategy reveals ligands for galectins. Chem Biol 16, 36–47. 10. Bohorov O, Andersson-Sand H, Hoffmann J, and Blixt O. (2006) Arraying glycomics: a novel bi-functional spacer for one-step microscale derivatization of free reducing glycans. Glycobiology 16, 21C-27C. 11. Park S, Lee MR, and Shin I. (2009) Construction of carbohydrate microarrays by using one-step, direct immobilizations of diverse unmodified glycans on solid surfaces. Bioconjug Chem 20, 155–162. 12. Feizi T, Fazio F, Chai W, and Wong CH. (2003) Carbohydrate microarrays – a new set of technologies at the frontiers of glycomics. Curr Opin Struct Biol 13, 637–645. 13. Liu Y, Feizi T, Campanero-Rhodes MA, Childs RA, Zhang Y, Mulloy B, Evans PG, Osborn HM, Otto D, Crocker PR, and Chai W. (2007) Neoglycolipid probes prepared via oxime ligation

for microarray analysis of oligosaccharide-protein interactions. Chem Biol 14, 847–859. 14. Song X, Lasanajak Y, Olson LJ, Boonen M, Dahms NM, Kornfeld S, Cummings RD, and Smith DF. (2009) Glycan microarray analysis of P-type lectins reveals distinct phospho-mannose glycan recognition. J Biol Chem. 284, 35201–35214. 15. Kudo M, and Canfield WM. (2006) Structural requirements for efficient processing and activation of recombinant human UDP-Nacetylglucosamine:lysosomal-enzyme-Nacetylglucosamine-1-phosphotran sferase. J Biol Chem 281, 11761–11768. 16. Dahms NM, and Hancock MK. (2002) P-type lectins. Biochim Biophys Acta 1572, 317–340. 17. Dahms NM, Olson LJ, and Kim JJ. (2008) Strategies for carbohydrate recognition by the mannose 6-phosphate receptors. Glycobiology 18, 664–678. 18. Ghosh P, Dahms NM, and Kornfeld S. (2003) Mannose 6-phosphate receptors: new twists in the tale. Nat Rev Mol Cell Biol 4, 202–212. 19. Kornfeld S, and Mellman I. (1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5, 483–525. 20. Chavez CA, Bohnsack RN, Kudo M, Gotschall RR, Canfield WM, and Dahms NM. (2007) Domain 5 of the cation-independent mannose 6-phosphate receptor preferentially binds phosphodiesters (mannose 6-phosphate N-acetylglucosamine ester). Biochemistry 46, 12604–12617. 21. Bohnsack RN, Song X, Olson LJ, Kudo M, Gotschall RR, Canfield WM, Cummings RD, Smith DF, and Dahms NM. (2009) Cationindependent mannose 6-phosphate receptor: A composite of distinct phosphomannosyl binding sites. J Biol Chem. 284, 35215–35226. 22. Gordon JA, Blumberg S, Lis H, and Sharon N. (1972) Purification of soybean agglutinin by affinity chromatography on sepharose-N-epsilonaminocaproyl-beta-D-galactopyranosylamine. FEBS Lett 24, 193–196.

Chapter 10 Production of Fluorous-Based Microarrays with Uncharged Carbohydrates Sahana K. Nagappayya and Nicola L.B. Pohl Abstract Fluorous-based carbohydrate microarrays provide an alternative to traditional covalent microarray ­platforms for probing protein–carbohydrate-binding interactions. The most studied plant lectin, concanavalin A  (ConA), is known to bind to terminally a-linked mannose. In the studies presented, the binding of ConA with a-mannose is analyzed using a microarray formed on a fluorous-coated glass slide with the sugar containing a fluorous tag at the anomeric position. Key words: Fluorous, Microarray, Carbohydrate, Lectin, FITC-ConA

1. Introduction Carbohydrate microarrays are a facile method to qualitatively and quantitatively study carbohydrate binding with various binding partners like proteins and antibodies (1). Given the difficulty in obtaining large amounts of pure oligosaccharides from either synthesis or isolation from natural sources, microarray platforms are a particularly good method to study the interaction of these saccharides with other molecules as only a milligram of sugar can be used to form dozens of microarray spots. Whereas most microarray platforms require covalent attachment of the analyte to the slide, an alternate noncovalent approach relies on the hydrophobic, or “fluorophobic” interactions of a fluorocarbontagged compound with a perfluoroalkylsilane-coated glass slide surface. Such fluorous-based microarrays have seen use for probing carbohydrates as well as noncarbohydrate molecules (2, 3). In the case of carbohydrates, a fluorous-tagged carbohydrate – obtained through direct reaction of the reducing end of an unmodified sugar

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_10, © Springer Science+Business Media, LLC 2012

149

150

S.K. Nagappayya and N.L.B. Pohl

Fig. 1. Schematic of work process for fluorous-based carbohydrate microarray production and screening.

with a tag (4) or through synthesis in which the fluorous tag aids the purification steps throughout (5) is spotted onto the fluorous glass slide using a standard robot used for DNA arraying (Fig.  1). The fluorous tail of the carbohydrate aligns itself onto the glass in a humidified chamber and then the slide is incubated with a protein of interest, for example the jack bean lectin concanavalin A (ConA) labeled with fluorescein isothiocyanate (FITC) (5, 6). After the slide is washed to remove nonbinding proteins, the binding of FITCConA with a-mannose is then observed under a standard fluorescence slide scanner (5, 6).

2. Materials 2.1. Printing on Fluorous-Coated Glass Slide

1. Fluorous-tagged carbohydrates are synthesized (7). The fluorous tail for each carbohydrate contains a four-carbon alkyl spacer at the anomeric position linking 3-(perfluorooctyl) propanol tag to the sugar. The solutions of the carbohydrate for printing are prepared in 40% methanol, 40% DMSO, 20% water (5) (see Note 1). These solutions are placed in a 384well plate in a pattern defined in the arrayer.

10  Production of Fluorous-Based Microarrays with Uncharged Carbohydrates

151

2. Fluorous-coated glass slides coated with perfluoroocty­lethylsilyl chains are commercially available from Fluorous Technologies Inc (Pittsburgh, PA). 3. 0.1–0.2 mM solution of FITC-ConA in 1× phosphate-­buffered saline (PBS) (5) (see Note 2). 4. The spotting is carried out using a robotic spotter with pin ­lifting technology (Cartesian PixSys 5500 Arrayer, Cartesian Technologies, Inc., Irvine, CA) greater than 60% relative humidity. 2.2. Microarray Incubation, Washing, and Screening

1. 1 mM CaCl2 and 1 mM MnCl2 solution in deionized water. 2. For incubation, 10 mL 1 mM CaCl2, 10 mL 1 mM MnCl2 and 200  mL of FITC-ConA solution are added to 780  mL of 1× PBS buffer set at pH 7.2 (see Note 2). 3. PC500 CoverWell incubation chamber (Grace Biolabs, Bend, OR) to incubate the array. 4. For washing, 1× PBS buffer containing 1% BSA solution.

2.3. Scanning of the Glass Slide

1. Fluorous-coated glass slides are scanned using general scanning on a ProScan Array HT (Perkin Elmer Life and Analytical Sciences, Shelter, CT) set at 488 nm, the common wavelength to detect the FITC label at 10 mm resolution.

3. Methods Fluorous-tagged carbohydrates are dissolved in 40% methanol, 40% DMSO, and 20% water. This solvent combination was chosen for the unique role of each solvent. Methanol is used to dissolve fully deprotected fluorous-tagged carbohydrates; DMSO is used to decrease the speed of solvent evaporation allowing enough time for the fluorous tail to orient itself on the glass slide. The role of water is to facilitate the hydrophobic fluorous–fluorous interaction. After printing, the glass slide is allowed to dry in a humidified chamber to further enforce the orientation of the fluorous tags toward the slides and the hydrophilic sugars on the surface. (Note that this process does not likely form a monolayer, though.) The slide is then placed in a PC500 CoverWell incubation chamber. For the detection of protein–carbohydrate binding, FITC-ConA in PBS buffer is applied to the printed glass slide and the slide is incubated for an hour. The carbohydrate that recognizes the protein ConA will noncovalently bind to it. Washing is carried out using PBS buffer and deionized water to get rid of unbound protein; the glass slide is dried and stored in a dark chamber and scanned for the protein–carbohydrate interaction.

152

S.K. Nagappayya and N.L.B. Pohl

3.1. Carbohydrate Spotting

1. Fluorous-tagged carbohydrates are synthesized. The fluorous tail for each carbohydrate contains a perfluorooctyl propane (or ethane, which is less reliably made) tag attached to the anomeric position of the sugar separated by a four-carbon alkyl linker (6, 7). 2. Completely deprotected fluorous-tagged carbohydrate is dissolved in 40% methanol, 40% DMSO, and 20% water. The solution of the carbohydrate is prepared in methanol and diluted with DMSO and water to make the concentration of the carbohydrate from 125 to 450 mM (5). 3. 30  mL of the fluorous-tagged carbohydrate is deposited in a 384-well plate using a micropipette. 4. Using the robotic spotter with pin lifting technology the carbohydrate solution is spotted onto the 12 × 2 cm2 fluorinated glass slide coated with perfluorooctylethylsilyl chains at a relative humidity of 60%. The spotting pin is dipped into the carbohydrate solution for one second, and then the robotic pin takes 25  ms to spot each spot on the slide at a distance of 400  mm apart (5, 6). The robotic spotter pin-head is programmed to be washed with pure distilled water after each spotting. 5. The glass slide is allowed to dry in a humidified chamber for 18–24 h after spotting.

3.2. Incubation and Washing

1. The printed glass slide is carefully placed in a PC500 CoverWell incubation chamber. 2. 200 mL solution of FITC-ConA (1 mM) incubation solution is applied onto the printed glass slide. This glass slide is incubated in the incubation chamber for 1 h (see Note 3) (4). 3. The incubated glass slide is washed once with 1× PBS buffer containing 0.01% Tween-20 and then with deionized water two to three times. Air is blown gently over this slide to let it dry and then it is placed in a dark chamber for ~2 h to let it completely dry (see Note 4).

3.3. Scanning

1. The washed glass slide is stored in a dark chamber and then scanned using a General Scanning ProScan Array HT set at 488 nm (see Note 5).

4. Notes 1. After preparation all carbohydrate solutions are stored at 4°C. 2. The PBS buffer is prepared by diluting the commercially available 10× PBS with deionized water. The pH of the PBS buffer is set at 7.26 and checked each time using a pH meter.

10  Production of Fluorous-Based Microarrays with Uncharged Carbohydrates

153

3. The FITC-ConA solution will decompose if exposed to light or allowed to stand at room temperature for more than 4 h. This solution is prepared in vials covered with aluminum foil immediately before carrying out the arrays. 4. The number of washings of the glass slides with PBS buffer and deionized water can greatly affect the outcome of the fluorescent scan. A few initial washings as a trial have to be carried out to find a suitable washing protocol for the carbohydrateprotein to be studied and both positive and negative control compounds should be spotted and tested in the same well. Generally, each compound is also spotted 8–12 times to ascertain the consistency of the particular microarray spotter. (Such multiple spotting is particularly important for quantitative binding experiments.) 5. This general protocol can be adapted to various noncharged mono- and oligosaccharides tagged with a variety of fluorous tails to find their binding with other carbohydrate-binding proteins, such as lectins and antibodies.

Acknowledgments We are grateful to Iowa State University, the US National Science Foundation (MCB-0349139 CAREER grant and CHE-0911123) and the US National Institutes of Health (R41 GM075436-01, R42 GM075436, and R42 GM075436-03  S1) with Fluorous Technologies, Inc. for their support of our microarray development work. References 1. Pohl NL, (2008) Fluorous tags catching on microarrays. Angew Chem Int Ed 47:3868–3870 2. Nicholson RL, Ladlow ML, Spring DL (2007) Fluorous tagged small molecule microarrays. Chem Commun 3906–3908 3. Vegas AJ, Bradner JE, Tang W, MvPherson OM, Greenberg EF, Koehler AN, Schreiber SL (2007) Fuorous-based small-molecule microarrays for the discovery of histone deacetyase inhibitors. Angew Chem Int Ed 46: 7960–7964 4. Chen G, Pohl NL (2007) Synthesis of fluorous tags for incorporation of reducing sugars in

quantitative microarray platform. Org. Lett 10: 785–788 5. Jaipuri FA, Collet B Y M, Pohl NL (2008) Synthesis and quantitative evaluation of glycero-D-mannoheptose binding to Concanavalin A by fluorous-tag assistance. Angew Chem Int Ed 47:1708–1710 6. Ko K-S, Jaipuri FA, Pohl NL (2005) Fluorousbased carbohydrate microarrays. J Am Chem Soc 127: 12162–13163 7. Mamidyala SK, Ko K-S, Jaipuri FA, Park G, Pohl NL (2006) Noncovalent fluorous interactions for the synthesis of carbohydrate microarrays. J Flourine Chem. 127: 571–579

Chapter 11 General Procedure for the Synthesis of Neoglycoproteins and Immobilization on Epoxide-Modified Glass Slides Yalong Zhang and Jeffrey C. Gildersleeve Abstract Neoglycoproteins, such as BSA-glycosides, contain carbohydrates covalently attached to a protein carrier via nonnaturally occurring linkages. These conjugates have been used for decades to study carbohydrate–protein interactions and are frequently used as immunogens to raise antibodies to carbohydrate antigens. In fact, neoglycoproteins have been used extensively as vaccine antigens and several have obtained FDA approval. More recently, neoglycoproteins have been used in the construction of glycan arrays to produce “neoglycoprotein microarrays.” In this chapter, two methods for preparing neoglycoproteins are described along with methods to immobilize these conjugates on epoxide-coated glass microscope slides to produce arrays. Key words: BSA-glycosides, Neoglycoprotein, Glycan array, Lectin inhibitors, Cancer biomarker

1. Introduction Carbohydrate–protein interactions play fundamental roles in many biological processes, such as fertilization, inflammation, viral and bacterial infection, and cancer metastasis (1). Therefore, there have been significant efforts to identify and characterize carbohydrate– protein interactions and to develop agonists/antagonists of these interactions. For these objectives, neoglycoproteins have been extremely useful (2, 3). Neoglycoproteins contain one or more copies of a carbohydrate covalently attached to a carrier protein, such as albumin. They are distinct from natural glycoproteins in that the linkage between the carbohydrate and protein is nonnatural. More importantly, neoglycoproteins are normally much more homogeneous than natural glycoproteins, which are usually produced as complex mixtures of glycoforms having different glycan chains attached to the protein core and/or have varying degrees of

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_11, © Springer Science+Business Media, LLC 2012

155

156

Y. Zhang and J.C. Gildersleeve

Fig. 1. Coupling via reductive amination.

glycosylation site occupancy. Neoglycoproteins can be used in many common assays developed for proteins/glycoproteins, such as ELISAs and Western blots. They have also been used extensively as multivalent inhibitors of carbohydrate–protein interactions and as vaccine antigens. Many methods have been developed for conjugating carbohydrates to proteins to produce neoglycoproteins. Polysaccharides and oligosaccharides that contain a free reducing end can be coupled directly to lysine residues via reductive amination (see Fig. 1) (4, 5). The process results in ring opening of the reducing-end monosaccharide residue, but this is often acceptable, especially for longer glycan chains. Alternatively, carbohydrates can be synthesized or derivatized with a linker to facilitate conjugation. For example, linkers containing a free carboxylic acid can be coupled to free amines on proteins via formation of an activated ester, such as an N-hydroxysuccinimide ester (see Fig. 2) (6) or p-nitrophenyl esters (7). Linkers containing a free thiol can be conjugated to maleimide-derivatized proteins or dehydroalanine-containing proteins (8). Other conjugation methods include copper(I)-catalyzed azide–alkyne cycloaddition (9) and conjugation of two amines with a squaric decyl ester (10). Typically, the resulting product contains a distribution of conjugates, with varying attachment sites and conjugation ratios. For example, coupling to lysines might produce a product with an average conju­gation ratio of 20 carbohydrates per molecule of protein. However, there is a significant proportion of conjugates with 18, 19, 21, or 22 carbohydrates per molecule of protein, and the exact lysines that are modified vary from one molecule of protein to another. Nonetheless, they each have only one type of carbohydrate structure attached. Methods for sitespecific attachment of glycan chains to proteins have also been developed (11–14). Glycan arrays have recently been developed as high-throughput tools for studying carbohydrate–protein interactions (15–17).

11  General Procedure for the Synthesis of Neoglycoproteins…

157

Fig. 2. Coupling via EDC activation. (a) Examples of sugar acids used for coupling to albumin and (b) reaction sequence for coupling.

Glycan arrays contain many different carbohydrates immobilized on a solid support in a spatially defined arrangement. They can be constructed in many ways, such as immobilization of linkermodified carbohydrates, lipid-modified carbohydrates, and neoglycoproteins (15–18). Our group has focused on printing neoglycoproteins on epoxide-coated glass microscope slides to produce neoglycoprotein microarrays (6, 19, 20). Since many neoglycoproteins are commercially available and others are readily prepared from oligosaccharides, this approach provides a convenient method to produce glycan arrays. In addition, the neoglycoproteins used on the array surface can be used in solution as agonists/ antagonists of carbohydrate–protein interactions and can be used in many other assays, enabling one to readily transition from array results to other applications. In this chapter, we describe two methods for preparing and characterizing neoglycoproteins. The first method involves reductive amination of oligosaccharides containing a free reducing end. The second involves activation of glycans that contain a linker with a free carboxylic acid. In addition, we describe methods to immobilize neoglycoproteins on epoxide-coated glass microscope slides to prepare glycan arrays, along with quality control checks to verify effective printing.

158

Y. Zhang and J.C. Gildersleeve

2. Materials 2.1. Solutions for Reductive Amination Method

1. BSA solution (10×): 150 mg/mL bovine serum albumin (BSA, essentially protease free, ³98%, Sigma, St. Louis, MO). 2. Borate buffer (10×): 400  mM sodium tetraborate (Sigma), adjusted with 1 M NaOH to pH 8.5. Store at room temperature. If a precipitate forms, heat at 70°C until the solution is clear. 3. Sodium sulfate solution (6×): Approximately 3 M sodium sulfate (Sigma) was prepared at a temperature of 50°C to produce a saturated solution. Solutions are stored at room temperature but are reheated to 50°C prior to use. 4. Oligosaccharide solutions: 20  mM oligosaccharides in water. Solutions were stored frozen at −20°C until used. 5. Sodium cyanoborohydride solution (10×): 3 M sodium cyanoborohydride (Sigma). Prepare a fresh solution prior to reactions. 6. Dialysis buffer: 6 mM NaCl (Sigma) solution. 7. Dialysis units (MWCO 10 000, Slide-A-Lyzer MINI Dialysis Unit, Pierce, Rockford, IL). 8. Applied Biosystem Voyager-DE Pro MALDI-time of flight (TOF) mass spectrometry (International Equipment Trading Ltd., Vernon Hills, IL) for evaluating the content of glycans on BSA. 9. MALDI matrix solution: Sinapinic acid (3,5-dimethoxy-4hydroxycinnamic acid, Sigma) as matrix for all experiments. Matrix was prepared as a saturated solution in 50% acetonitrile/ H2O (0.1% TFA) (v/v).

2.2. Solutions for Active Ester Method

1. 1-Ethyl-3-(3-dimethyllaminopropyl)carbodiimide hydrochloride (EDC; Anaspec, Fremont, CA) was dissolved in water and then diluted with an equal volume of DMF to give a final concentration of 300 mM. Prepare a fresh solution prior to reactions. 2. 300  mM N-Hydroxysuccinimide (NHS; Aldrich, St. Louis, MO). Prepare a fresh solution prior to reactions. 3. 4  mg/mL BSA in sodium borate buffer (10  mM sodium borate, 90 mM NaCl, pH 8.0). 4. Glycan solutions: Glycans containing a linker with a free carboxylic acid (sugar acids) were dissolved in water at a final concentration of 150 mM. Solutions were stored frozen at −20°C until used. Sugar acids are typically obtained by chemical synthesis. For example, glycopeptides that contain a free C-terminal carboxylic acid can be prepared via solid-phase peptide synthesis, deprotection, and purification.

11  General Procedure for the Synthesis of Neoglycoproteins…

2.3. Carbohydrate Microarray Fabrication

159

1. Printing buffer: 1× PBS, 2.5% glycerol, 0.006% Triton-X 100. 2. PBS buffer (20×): 2.74  M NaCl, 60  mM KCl, 200  mM Na2HPO4, 36  mM KH2PO4, pH 7.4. Store at room temperature. 3. Fluorophore-labeled BSA (Invitrogen Corporation, Carlsbad, CA) and BSA-glycosides were diluted with printing buffer to a concentration of 125 mg/mL. 4. ArrayIt SuperEpoxy 2 glass slides (TeleChem International, Inc., Sunnyvale, CA). 5. PBST0.05 washing buffer: 1× PBS, 0.05% tween 20. 6. Robotic microarrayer: MicroGrid II 600/610 (Genomic Solutions, BioRobotics, Ann Arbor, MI). 7. ArrayIt stealth pins: SMP2 pins (uptake volume 0.25  mL, delivery volume 0.5  nL) were purchased from TeleChem International, Inc. (Sunnyvale, CA). 8. Sample plates: 384-well V-bottom sample plates (Genetix, San Jose, CA). 9. 16-well slide modules (Grace Bio-Labs, Bend, OR). 10. Genepix 4000A microarray scanner (Molecular Devices Corporation, Union City, CA).

3. Methods 3.1. Preparation of BSA-Glycosides by Reductive Amination Method

1. Mix BSA (2  mL, 150  mg/mL) with sodium borate (5.5 mL, 400 mM, pH 8.5), sodium sulfate (3.7 mL, saturated at 50°C), oligosaccharide (3.3 mL, 20 mM solution for 15 equiv), and water (5.3  mL) in a 200-mL PCR tube. Add sodium cyanoborohydride (2.2 mL, 3 M) and cap the tube (see Notes 1 and 2). 2. Incubate the solution in a PCR thermal cycler at 56°C for 96 h with a heated lid (see Note 3). 3. Dilute the reaction with H2O (100 mL), transfer to a 500-mL dialysis tube, and dialyze against H2O (2.5 L) containing 6 mM NaCl three times. 4. Dilute the dialyzed BSA conjugates to a final concentration of 1.0 mg/mL and store at −20°C.

3.2. Preparation of BSA-Glycosides by Active Ester Method

1. Mix sugar acid, NHS (300 mM), and then EDC (300 mM) in an Eppendorf tube at a volume ratio of 2:1:1 and incubate at room temperature with occasional gentle mixing to form the NHS ester in situ (see Note 4). 2. After 1 h, add the solution containing the NHS ester to a solution of BSA (4  mg/mL in sodium borate buffer, pH 8.0). Precool the solution to 4°C.

160

Y. Zhang and J.C. Gildersleeve

3. After 15  min, warm the solution to room temperature and incubate for another 1 h. 4. Dialyze BSA-conjugates against H2O (2.5 L) containing 6 mM NaCl three times using dialysis units. 3.3. MALDI MS Analysis of BSA-Glycoside

1. BSA conjugates are analyzed by MALDI-TOF MS and the extent of conjugation is determined by subtracting the average mass of BSA (66431) from the average mass of the conjugate and then dividing by the molecular weight of the sugar acid minus 18 (see Notes 5 and 6). 2. The instrument is operated in linear mode under positive condition with the accelerating voltage of 25 kV, guide wire 0.15%, and grid voltage 91.5%. 3. A nitrogen laser should be used at 337 nm with 250 laser shots averaged per spectrum. 4. Sample preparation is a modified “dried droplet” procedure, whereby 0.3 mL of samples are spotted on a MALDI sample plate followed by 0.3 mL of matrix. The mixture is then allowed to air dry prior to analysis. 5. Data analysis is carried out using Data Explorer software resident on Voyager mass spectrometer and used BSA as a standard for external calibration.

3.4. Carbohydrate Microarray Fabrication

1. Transfer 15 mL of fluorophore-labeled BSA and neoglycoprotein solution (125 mg/mL) to wells of a 384-well V-bottom sample plates (see Notes 7 and 8). 2. Print carbohydrate microarrays on epoxide-modified glass slides using a robotic microarrayer fitted with SMP2 Microspotting Pins (see Note 9). 3. The humidity level should be maintained at ~50% with an ultrasonic humidifier. 4. Each BSA-glycoside component should be printed in duplicate (see Figs. 3 and 4). The diameter of each spot is around 90 mm and the pitch is 200 mm (see Note 10). 5. Printed slides are allowed to dry at room temperature for an additional few hours (keeping the total exposure time less than 36 h) and then stored at −20°C until use (see Notes 11 and 12).

3.5. Quality Assessment of Microarray Slides

Most slides have a complete array grid. Toward the end of a print run, pins may occasionally run out of liquid resulting in missing spots. In addition, pins may occasionally get clogged, resulting in missing spots. Tolerance for missing spots depends on the particular application. For highest quality data, only slides containing a complete grid should be used. For less-critical experiments, such as

11  General Procedure for the Synthesis of Neoglycoproteins…

161

Fig. 3. Microarray printing. Hollow pins are dipped into wells of a 384-well plate containing print solutions. The filled pins are then moved to the slides and lightly tapped on the glass to print array spots.

Fig. 4. Immobilization of neoglycoproteins on epoxide-coated slides.

optimizing reaction parameters, missing spots are acceptable. Below, we describe two procedures for quality assessment of microarray slides. 3.5.1. Physical Inspection of Printed Arrays

1. Inspect each printed slide under a microscope at 10× magnification (see Fig. 5). 2. Note the slide, block, column, and row number for any missing spots. 3. Evaporation can complicate the analysis. Inspect slides immediately after printing.

3.5.2. Evaluate Binding with One or More Lectins

1. Fit slides with 16-well slide modules. 2. Block slides by addition of 3% BSA/PBS (200 mL per well) and incubate at room temperature for 2.0 h. 3. After removing block solution, incubate slides with one or more lectins (diluted in 1% BSA/PBS; 100  mL per well) at room temperature for 2  h. Lectins are typically assayed at a concentration of 1–10 mg/mL (see Note 13).

162

Y. Zhang and J.C. Gildersleeve

Fig. 5. View of microarray spots after printing. A magnified view of liquid spots after printing showing uniform spots with consistent spacing and no missing spots.

4. After washing 3× with PBST0.05 (200 mL per well), incubate slides with Cy3–streptavidin (2  mg/mL in 1% BSA/PBS; 100 mL per well) at room temperature for 2 h. 5. After washing 3× with PBST0.05 (200 mL per well), remove slides from the well module and immerse in PBST0.05 for 15 min. 6. Place the slide in a 50-mL conical tube and centrifuge for 5 min at 200 × g to dry the slide. 7. Scan slides using a Genepix 4000A microarray scanner at PMT voltage settings, where no saturated pixels are obtained. Image analysis is carried out with Genepix Pro 6.0 analysis software from the same company. Spots are defined as circular features with maximum diameter of 100 mm. The background-corrected median feature intensity, F532median-B532, should be used for data processing. 8. Binding data is compared with results from previous batches of slides and with published binding data for each lectin.

4. Notes Reductive amination 1. The reaction conditions have been optimized for BSA. Other proteins may be more or less tolerant of the high salt conditions and high temperature. 2. Reactions occasionally result in precipitation, especially at higher temperatures.

11  General Procedure for the Synthesis of Neoglycoproteins…

163

3. Evaporation and condensation of liquid on the lid of the reaction tube can be a significant problem, especially for small scale reactions. The use of a PCR tube and a PCR thermal cycler with a heated lid minimizes this problem. EDC/NHS coupling 4. EDC should be added last to the reaction mixture. MALDI MS analysis 5. The ratio of sugar acid to protein can be modulated to produce conjugates with varying numbers of carbohydrates. For the active ester method, at a ratio of 30:1, we typically get about 15 oligosaccharides per molecule of BSA. At 7:1, the final ratio is typically around 4/BSA. 6. For the reductive amination method, at a starting ratio of 9:1, the product typically has about 4-5 oligosaccharides per molecule of BSA. At a ratio of 35:1, the final conjugate should have around 15 oligosaccharides per molecule of BSA. Printing 7. We typically print Cy5–BSA as the first component and Cy3– BSA toward the end of the grid. These fluorophore–BSA conjugates simplify alignment of the array grid when analyzing experiments and serve as quality control checks for the print procedure. For example, carryover due to insufficient washing of pins is easily detected. 8. Print solutions in the source plate evaporate over time. Water should be added to compensate for evaporation. For example, when a source plate is in use for 4–5 h, we would add about 2 mL of water to each well. The exact amount depends on the total time, humidity, temperature, and arrayer. 9. Clean pins thoroughly to avoid carryover from one print component to the next. The optimal conditions should be determined experimentally. For our recommended print solutions and a MicroGrid II microarrayer, we recommend at least two complete wash cycles with a 5-mm submission height, a 3-s dwell time in each of two wash baths, and an 8-s drain time (where the water in the pins is removed by vacuum). 10. The optimal printing conditions are dependent on the type of microarrayer. For a MicroGridII fitted with SMP2 pins, our preferred conditions are listed in Tables 1 and 2. 11. Keep the total time for the printing process below 36  h to avoid protein degradation. 12. Cy5 is sensitive to degradation and photobleaching. Slides containing Cy5–BSA or Cy5-labeled proteins should be stored in the dark at −20°C.

164

Y. Zhang and J.C. Gildersleeve

Table 1 Prespotting settings (to remove excess liquid on the pins) Parameter

Description

Recommended value

Number of prespots

Spotting times of a freshly loaded pin

6

Delay before spotting

Time of suspended pins above slide

0.5 s

Target height

Distance of pin into contact with slide

1.5 mm

Speed

Speed of pin into contact with slide

4 mm/s

Surface height

Thickness of slide

1 mm

Dwell time

Time tool is held in place at its target height

0.250 s

Multiple strikes

Number of tapping for each spot

1

Pitch

Distance between the centers of adjacent spots

0.45 mm

Table 2 Printing parameters Parameter

Description

Recommended value

Number of slides

High-print-quality slides

12–15

Delay before spotting

Time of suspended pins above slide

0 s

Target height

Distance of pin into contact with slide

1 mm

Speed

Speed of pin into contact with slide

4 mm/s

Surface height

Thickness of slide

1 mm

Dwell time

Time tool is held in place at its target height

0 s

Pitch

Distance between the center of spots

0.2 mm

Quality control checks 13. The choice of lectin(s) for quality control assessment depends on the composition of the array. We frequently use wheat germ agglutinin (WGA), ricinus communis agglutinin (RCA120), bauhinia purpurea agglutinin (BPA), and concanavalin A (ConA).

Acknowledgments This research was supported by the Intramural Research Program of the NIH, NCI.

11  General Procedure for the Synthesis of Neoglycoproteins…

165

References 1. Dwek, R. A. (1996) Glycobiology: Toward Understanding the Function of Sugars, Chem. Rev.96(2), 683–720. 2. Roy, R. (1996) Syntheses and some applications of chemically defined multivalent glycoconjugates, Curr. Opin. Struc. Biol. 6, 692–702. 3. Stowell, C. P., and Lee, Y. C. (1980) Neoglyco­ proteins: the preparation and application of synthetic glycoproteins, Adv. Carbohyd. Chem. Bi. 37, 225–281. 4. Gildersleeve, J. C., Oyelaran, O., Simpson, J. T., and Allred, B. (2008) Improved Procedure for Direct Coupling of Carbohydrates to Proteins via Reductive Amination, Bioconjugate Chem.19(2), 1485–1490. 5. Clausen, M., and Madsen, R. (2004) Synthesis of oligogalacturonates conjugated to BSA, Carbohyd. Res. 339, 2159–2169. 6. Manimala, J. C., Roach, T. A., Li, Z., and Gildersleeve, J. C. (2006) High-throughput carbohydrate microarray analysis of 24 lectins, Angew. Chem. Int. Edit. 45, 3607–3610. 7. Wu, X., Ling, C.-C., and Bundle, D. R. (2004) A New Homobifunctional p-Nitro Phenyl Ester Coupling Reagent for the Preparation of Neoglycoproteins, Org. Lett. 6, 4407–4410. 8. Wang, J., Schiller, S. M., and Schultz, P. G. (2007) A biosynthetic route to dehydroalaninecontaining proteins, Angew. Chem. Int. Edit. 46, 6849–6851. 9. van Kasteren, S. I., Kramer, H. B., Jensen, H. H., Campbell, S. J., Kirkpatrick, J., Oldham, N. J., Anthony, D. C., and Davis, B. G. (2007) Expanding the diversity of chemical protein modification allows post-translational mimicry, Nature 446, 1105–1109. 10. Karelin, A. A., Tsvetkov, Y. E., Paulovicova, L., Bystricky, S., Paulovicova, E., and Nifantiev, N. E. (2009) Synthesis of a heptasaccharide fragment of the mannan from Candida guilliermondii cell wall and its conjugate with BSA, Carbohydr. Res. 344, 29–35.

11. Liu, H., Wang, L., Brock, A., Wong, C.-H., and Schultz, P. G. (2003) A method for the generation of glycoprotein mimetics, J. Am. Chem. Soc. 125, 1702–1703. 12. van Kasteren, S. I., Kramer, H. B., Gamblin, D. P., and Davis, B. G. (2007) Site-selective glycosylation of proteins: creating synthetic glycoproteins, Nat. Protoc. 2, 3185–3194. 13. Zhang, Y., Bhatt, V. S., Sun, G., Wang, P. G., and Palmer, A. F. (2008) Site-Selective Glycosylation of Hemoglobin on Cys b93, Bioconjugate Chem. 19, 2221–2230. 14. Chen, R., and Tolbert, T. J. (2010) Study of On-Resin Convergent Synthesis of N-Linked Glycopeptides Containing a Large High Mannose N-Linked Oligosaccharide, J. Am. Chem. Soc. 132, 3211–3216. 15. Bernardes, G. J. L., Castagner, B., and Seeberger, P. H. (2009) Combined Approaches to the Synthesis and Study of Glycoproteins, ACS Chem.Biol. 4, 703–713. 16. Liang, C.-H., and Wu, C.-Y. (2009) Glycan array: a powerful tool for glycomics studies, Expert Rev. Proteomics 6, 631–645. 17. Oyelaran, O., and Gildersleeve, J. C. (2009) Glycan arrays: recent advances and future challenges, Curr. Opin. Chem. Biol. 13, 406–413. 18. Zhao, J., Patwa, T. H., Lubman, D. M., and Simeone, D. M. (2008) Protein biomarkers in cancer: natural glycoprotein microarray approaches, Curr. Opin. Mol. Ther. 10, 602–610. 19. Zhang, Y., Li, Q., Rodriguez, L. G., and Gildersleeve, J. C. (2010) An Array-Based Method To Identify Multivalent Inhibitors, J. Am. Chem. Soc. 132, 9653–9662. 20. Oyelaran, O., Li, Q., Farnsworth, D., and Gildersleeve, J. C. (2009) Microarrays with Varying Carbohydrate Density Reveal Distinct Subpopulations of Serum Antibodies, J. Proteome Res. 8, 3529–3538.

Chapter 12 Immobilization of Polyacrylamide-Based Glycoconjugates on Solid Phase in Immunosorbent Assays Oxana E. Galanina, Alexander A. Chinarev, Nadezhda V. Shilova, Marina A. Sablina, and Nicolai V. Bovin Abstract Our experience in coating of solid surfaces with glycans, mainly for obtaining routine glycoarrays based on immunological plates, is summarized. Three polystyrene coating techniques are described: direct physical adsorption, covalent binding, and immobilization using the biotin tag. Protocols for studies on anticarbohydrate antibodies are considered, with special emphasis on the application niches of different immobilization techniques as related to the specificity of each method of glycan-binding protein assay, as well as the problems of background binding and quantitative estimation of the results. Key words: Glycoarrays, ELISA, Glycopolymers, Immobilization techniques, Physical adsorption, Covalent immobilization, Biotin–streptavidin bridge, Polyacrylamide

Abbreviations Atri AP Btri Biot BSA ELISA Glyc GPC Ig HRPO m.w. PAA PBS pNPA pNSA Str TLC

A trisaccharide, GalNAca1-3(Fuca1-2)GalbAlkaline phosphatase B trisaccharide, Gala1-3(Fuca1-2)GalBiotin Bovine serum albumin Enzyme-linked immunosorbent assay Glycoside residue Gel-permeation Anti-mouse antibodies Horse radish peroxidase Molecular weight Poly(acrylamide) Phosphate saline buffer poly(4-nitrophenylacrylate) poly(N-oxisuccinimidylacrylate) Streptavidin Thin-layer chromatography

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_12, © Springer Science+Business Media, LLC 2012

167

168

O.E. Galanina et al.

1. Introduction Glycans constitute a complex and diverse family of biomolecules involved in a broad range of physiological (cell development and differentiation, cell recognition and cell-to-cell adhesion, cell motility, etc.) and pathological (malignant transformations, autoimmune diseases, host–pathogen interactions, etc.) processes, many important characteristics of which remain poorly studied (1–3). For profiling glycan-binding proteins and studying interactions involving them, enzyme-linked immunosorbent assay (ELISA) and related assays based on the classical multiwell plates are generally used. Although many researchers note that ELISA is not sensitive enough and time consuming, its major advantages are simplicity and flexibility of the procedure. This allows the method to be adjusted to a specific research task, e.g., selection of the direct binding or inhibition conditions; choice of the optimal detection method (colorimetric or fluorescent); instant variation of glycan combinations, coating concentrations, and surface density; and choice between the monovalent and multivalent types of their presentation. ELISA requires at least one order of magnitude lower concentrations of the analyzed proteins to be than, e.g., the printed glycan array (PGA) and suspension (Bioplex) array. It is noteworthy that the assay does not require sophisticated equipment. Glycan ELISA has its own application niche; it is optimal for miniformat glycan profiling when 5–15 glycans are necessary for characterization of a glycan-binding protein. In practice, ELISA has proved to be a good complementary “partner” for PGA, when PGA is used as a screening tool, followed by ELISA for accurate verification of the results. Apart from research purposes, ELISA can be easily adapted for medical applications. The assay can be performed manually or, if necessary, robotized (4–7). Typically, interactions involving glycans are weak; this impels researchers to use carbohydrates in the multivalent form. In particular, glycoconjugates based on linear polyacrylamides (PAAs) with attached-side carbohydrate groups (Glyc) solve this problem. The polymer backbone plays two roles: first, a flexible linker between Glyc residues facilitating their optimal arrangement for multivalent binding to an antibody or lectin irrespective of the distance between their carbohydrate-binding domains (8, 9); second, a long spacer between glycans and the support. On the contrary, monomeric glycans arrayed on a solid surface are often hardly capable of multivalent binding with glycan-recognizing proteins because of suboptimal density of Glyc residues and length of their bonds with the surface; i.e., this type of glycan arrays can be only technically regarded as multivalent (10–12).

12  Immobilization of Polyacrylamide-Based Glycoconjugates…

169

Fig. 1. The use of glycopolymers in ELISA. A glycan from the library of w-aminoalkyl ­glycosides is quantitatively coupled to fully activated polyacrylic acid prepared by radical polymerization of nitrophenylacrylate or N-acryloyl succinimide. There are three ways of further use of the resultant glycopolymers containing active ester groups. First, they can be covalently immobilized on aminated surfaces (including polystyrene) through the ­formation of amide bonds. Second, they can be physically adsorbed in the form of ­Glyc–PAA. Third, they can be coated onto a streptavidin-modified surface in the form of a biotinylated glycopolymer.

In this paper, we describe a set of techniques using various types of PAA-based glycoconjugates for coating polystyrene ELISA plates; the general strategy of the glycopolymer use in ELISA is shown in Fig. 1.

2. Materials 2.1. Carbohydrates

Atri-O(CH2)3NH2 and Btri-O(CH2)3NH2 were obtained from Lectinity, Inc. (Russia).

2.2. Biotin

Biot-NH(CH2)6NH2 was obtained from Lectinity, Inc. (Russia).

170

O.E. Galanina et al.

2.3. Activated Polyacrylic Acids

Poly(4-nitrophenylacrylates), pNPA, biot1-pNPA, and poly (N-oxisuccinimidyl acrylate), pNSA, were prepared as described before (11, 13).

2.4. Other Commercial Reagents and Solvents

Regular chemicals were obtained from Fluka (USA), Aldrich (USA), and Merck (Germany).

2.5. Buffers

1. PBS: 137  mM NaCl, 2.7  mM KCl, 10.0  mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. 2. ELISA-coating buffer: 15  mM Na2CO3, 35  mM NaHCO3, pH 9.6. 3. ELISA washing buffer: 0.1% Tween-20 in PBS. 4. ELISA blocking buffer: 3% BSA in PBS. 5. Colorimetric solution: 0.1 M sodium phosphate, 0.1 M citric acid buffer, 0.04% O-phenylenediamine, pH 4.7. 6. Fluorescent-revealing solution: 10−4  M 4-methylumbellyferyl phosphate disodium salt in the coating buffer.

2.6. Antibodies

1. Mouse monoclonal antibodies B8 (against Btri) and A16 (against Atri) were obtained from All-Russian Hematology Research Center (Moscow, Russia). 2. Streptavidin–horseradish peroxidase (Str–HRPO) conjugate was the product of GE Healthcare (UK). 3. Anti-mouse IgG + IgM (H + L)–alkaline phosphatase (Ig–AP) conjugate was the product of AP Biotech. Inc. (UK).

2.7. 96-Well Plates

1. Reacti-Bind TM Streptavidin coated High Binding Capacity black plates (#15503) were purchased from ThermoScientific (USA). 2. MaxiSorp transparent immunoplates (#439454) were purchased from NUNC (Denmark). 3. NH2-modules were obtained from Costar (USA).

2.8. Chromatography

1. Thin layer chromatography (TLC) aluminum sheets covered with silicagel 60 (Merck, Germany). 2. Sephadex LH-20 gel (Pharmacia BioTech, Austria). 3. Eluent A for 10:1:1:1:0.3.

TLC:

EtOH/BuOH/Py/H2O/AcOH,

4. Eluent B for TLC: MeOH/1 M Py·AcOH, 3:1. 5. Eluent C for gel-permeating chromatography (GPC): MeCN/ H2O, 1:1. 6. Charring reagent for TLC plates: 7% H3PO4. 2.9. Equipment

Victor2 multilabel counter (PerkinElmer, USA).

12  Immobilization of Polyacrylamide-Based Glycoconjugates…

171

3. Methods Three approaches to immobilization of Glyc–PAA are considered here: physical (noncovalent) adsorption, chemical (covalent) immobilization, and immobilization via biotin–Str bridge. Physical adsorption due to Van der Waals interactions between glycopolymer molecules and the plate surface is the simplest and the least sophisticated method. The adsorption efficiency depends on the m.w. of the glycopolymer; the number of contacts and interaction strength between glycopolymer molecules and the surface increase with an increase in the m.w. (13). Two glycoconjugates differing in m.w., Glyc–PAA30 and Glyc– PAA2000 (see Notes 1 and 2), were synthesized (the protocols were developed for blood group A trisaccharide derivatives, Atri–PAA) as shown in Fig. 2a and coated onto the plates. The coating procedure included incubation of the glycoconjugate solutions in wells and subsequent washing off of nonadsorbed material. It was found that only as little as about 2% of the added low-m.w. glycopolymer is adsorbed, whereas this value for high-m.w. ­glycopolymer is ~20% (11). The difference in the efficiency of immobilization for low- and high-m.w. Glyc–PAAs is illustrated in Fig.  3. Addition of equal amounts of Glyc into wells yields a stronger optical signal in the case of Glyc–PAA2000; the deference between glycoconjugate signals is greater at a low coating concentration (<50 pmol Glyc per well). The maximum adsorption for the high- and low-m.w. glycoconjugates is observed, respectively, at 50 and 200 pmol of Glyc residue per well. Thus, the use of the high-m.w. glycoconjugate allows reducing the amount of carbohydrates required for coating by a factor of 10–20. The difference observed in mAb binding with glycopolymers is explained as follows. The average size of coiled glycopolymer molecules adsorbed on polystyrene is known to be ~100 Å for 30-kDa conjugates and ~500 Å for 2,000-kDa ones (11), whereas the distance between Fab fragments of antibodies varies from 100 to 150 Å for IgG and may be even more for IgM molecules (14). Thus, in the case of Atri–PAA30, binding of an antibody with several neighboring glycopolymer molecules on the surface is much more probable compared to an antibody with only one glycopolymer (Fig. 2b). An increase in the density of Atri–PAA30 molecules on the surface leads to an increase in the detected signal. On the contrary, the size of Atri–PAA2000 molecules allows them to bind one or several antibodies simultaneously: a signal is detectable even at a low density of Atri–PAA2000 (Fig. 2c). Chemical immobilization of glycopolymers on the surface of aminated polystyrene is accomplished via formation of covalent

172

O.E. Galanina et al.

Fig. 2. Polyacrylamide glycoconjugates for physical adsorption on polystyrene plates. (a) Synthesis: One of activated polyacrylic acids differing in m.w. (pNPA or pNSA) is bound with Glyc-O(CH2)3NH2 and then treated with ethanolamine to quench the remaining ­activated ester groups. The average sizes of molecules and distances for (b) low- and (c) high-m.w. glycoconjugates on the surface are shown.

amide bonds (Fig. 4). Activated polyacrylic acids with pendant Glyc residues, Glyc–pNPA30 and Glyc–pNSA2000 (see Notes 1 and  2), are used without isolation. The coating procedure consists of ­consecutive incubations of plates in a glycopolymer solution (the coupling stage) and then in an aqueous ammonia solution (the quenching stage), which is followed by washing off of the unbound

12  Immobilization of Polyacrylamide-Based Glycoconjugates…

173

Fig. 3. Physical adsorption of (a) high- and (b) low-m.w. glycoconjugates (Atri–PAA2000 and Atri–PAA30, respectively) on 96-well, transparent polystyrene plates. The efficacy of adsorption was demonstrated by colorimetric detection using anti-Atri mAbs.

Fig. 4. Coating of an aminated surface with glycopolymers.

174

O.E. Galanina et al.

material. As a result, the plates are coated with PAA glycoconjugates differing in m.w. (30 and 2,000 kDa). A stronger binding is observed for the high-m.w. glycoconjugate; the sorption is maximum at a concentration of 75 pmol of Glyc per well, whereas in the case of the low-m.w. glycoconjugate, a concentration of 200 pmol per well is required (Fig. 5). A similar immobilization protocol allows glycopolymers to be attached to other aminated surfaces, e.g., coronary angioplasty stents made of stainless steel coated with nanodiamonds and then aminated plasmachemically over the diamond layer (Plasmachem GmbH, Germany). To solve the problem of smooth muscle cell overgrowth and to make the stent surface attractive for selectinbearing endothelial cells, the stent surface is modified with carbohydrates capable of mediating endothelial cell adhesion (15). The efficacy of chemical coating of the stents with SiaLea–pNPA30 was proved by ELISA with anti-SiaLea mAbs (Fig. 6). A convenient method of glycopolymer immobilization on plates is based on the use of biotin–Str bridge, when a glycopolymer bearing a biotin tag is anchored to a Str-coated plate. We used two approaches to the synthesis of biotinylated glycopolymers. According to the first one, a biotin derivative containing an amino group in the linker (biot-NH(CH2)6NH2) and an w-aminoalkyl glycoside (Glyc-O(CH2)3NH2) are consecutively coupled to pNPA. Finally, the remaining active ester groups in the polymer are quenched using ethanolamine (Fig. 7a). The resultant glycopolymer contains several side-pendant Glyc and biotin residues (see Notes 2 and 3). In the second method, a biotin residue is introduced into the ­polymer scaffold as a single end group (Fig. 7b) with a fragment of

Fig. 5. Chemical immobilization of (a) high- and (b) low-m.w. glycoconjugates (Btri–PAA2000 and Btri–PAA30, respectively) on aminated polystyrene plates. The efficacy of coating was demonstrated using anti-Btri mAbs.

12  Immobilization of Polyacrylamide-Based Glycoconjugates…

175

Fig. 6. Photograph of a SiaLea-modified stent placed into a well of a nonadsorbing plate (as a vessel) and stained with ­anti-SiaLea mAbs; the color is developed due to HRPOlabeled secondary antibodies.

Fig. 7. Biotinylated glycopolyacrylamides for coating onto Str plates: Synthesis of the ­glycopolymers with (a) side- and (b) end-pendant biotin residues. (c) Anchoring of glycopolymers with one and several pendant biotins on the surface and their presentation for interaction with antibodies.

176

O.E. Galanina et al.

biotinylated initiator of polymerization. The resultant activated polyacrylic acid, pNPA–biot1, interacts with Glyc-O(CH2)3NH2, which is followed by treatment with ethanolamine. The resultant glycopolymer, Glyc–PAA30–biot1, contains several side-pendant Glyc residues and the only one end biotin (see Note 2). The immobilization procedure includes incubation of the biotinylated glycopolymers, Glyc–PAA30–biot5 and Glyc–PAA30– biot1, with Str plates and washing off of nonadsorbed material; the procedure has been shown to give a quantitative yield (16). The monobiotinylated glycopolymer Glyc–PAA30–biot1 slightly better bound mAbs compared to an analogue bearing several-side biotins, Glyc–PAA30–biot5 (Fig.  8). The adsorption was maximal for the monobiotinylated conjugate at a concentration of 70  pmol/well, whereas for the pauci–biot analog it was maximal at 100 pmol/well. As the glycopolymers have the same m.w., the observed difference is likely to have resulted from differences in the anchoring on the surface and, hence, in the presentation of the glycan for mAbs (16). In immobilized molecules of Glyc–PAA30–biot1, the single end biotin is hidden inside the streptavidin matrix. The monobiotinylated glycopolymer is recommended for plate coating when additional biotin residues are expected to affect the results of the assay, i.e., when hydrophobicity of extra biotin residues could critically elevate the background signal of the assay or when (strept)avidin-based reagents are used not only for immobilization, but also at one of the next stages, as illustrated in Fig. 9. Finally, it should be noted that immobilization through Str–biotin bridge has appeared to be a less reagentconsuming method.

Fig. 8. Binding of anti-Btri mAbs by glycopolymers with (a) end- and (b) side-pendant biotin residues (Btri–PAA30–biot1 and Btri–PAA30–biot5) immobilized on an Str plate; fluorimetric detection.

12  Immobilization of Polyacrylamide-Based Glycoconjugates…

177

Fig. 9. Demonstration of the lack of available biot residues in the composition of Str ­complexed with end-biotinylated glycopolymer. Str plates are coated with a glycopolymer with side- or end-pendant biotin residues, Glyc–PAA30–biot5 (open bars) and Glyc–PAA30– biot1 (hatched bars), and developed with Str–AP.

In summary, three approaches can be used for coating immunological plates with glycans in the form of Glyc–PAA; in all cases, the immobilized molecules are practically the same. What is the reason for selecting one of the methods for design of a specific assay? Physical adsorption has proved to be the simplest protocol; however, it has a serious drawback, namely, a high consumption of carbohydrates to be coated. Therefore, it is impracticable when only a minute quantity of glycan is available or the cost of the assay is the critical point. Biotin-mediated coating eliminates this disadvantage and, in addition, is methodologically simple. One important characteristic of biot–Str immobilization should be noted. In the case of physical adsorption or the aforementioned covalent binding, the amount of glycoconjugate immobilized on the plate surface is not strictly proportional to their amount added. In contrast, biotinylated glycoconjugate can be immobilized quantitatively when the biot-to-Str ratio is <1 (in assays using the same plates as in this study, at concentrations of biotin of 0–50  pmol Glyc/well). Therefore, immobilization via biot–Str bridge allows a researcher to know the absolute amount of the coated material, which is necessary, e.g., when a correct affinity constant is calculated. Limitations of the biot–Str technique are the high cost of commercially available plates and, sometimes, a low stability of the streptavidin monolayer. Chemical immobilization is undoubtedly the most sophisticated and time-consuming technique. It could be recommended for nonroutine assays, when the use of other methods may lead to desorption because of unusual conditions, or the plate should be regenerated for repeated assays without loss of the coating glycoconjugate.

178

O.E. Galanina et al.

3.1. Preparation of Polyacrylamide-Based Glycoconjugates 3.1.1. Synthesis of Atri–PAA30 for Physical Adsorption on Polystyrene Plates (See Note 4)

1. Add a mixture of 3.2 mg (5.5 mmol) of Atri-O(CH2)3NH2 and 1.5 mL (11 mmol) of NEt3 in 100 mL of DMSO to a solution of 4.7 mg (27.5 mmol) of pNPA in 400 mL of DMSO. 2. Keep the mixture for 12 h at 40°C performing control for the reaction by TLC; eluent A: Rf 0.35 for Atri-O(CH2)3NH2, Rf 0 for Atri–PAA30 (see Subheading 3.2). 3. Add an excess of ethanolamine (90 mL) to the reaction mixture and keep it for 24 h at 40°C. 4. Purify the product by GPC (see Subheading 3.3). 5. Yield 85–95%; white solid.

3.1.2. Synthesis of Atri–PAA2000 for Physical Adsorption on Polystyrene Plates

1. Add a mixture of 3.2 mg (5.5 mmol) of Atri-O(CH2)3NH2 and 1.5 mL (11 mmol) of NEt3 dissolved in 100 mL of DMSO to a solution of 4.1  mg (27.5  mmol) of pNSA in 400  mL of DMSO. 2. Keep the mixture for 12 h at 40°C performing control for the reaction by TLC; eluent A: Rf 0.35 for Atri-O(CH2)3NH2, Rf 0 for Atri–PAA2000 (see Subheading 3.2). 3. Add an excess of ethanolamine (90 mL) to the reaction mixture and keep it for 24 h at 40°C. 4. Purify the product by GPC (see Subheading 3.3). 5. Yield 85–95%; white solid.

3.1.3. Synthesis of Btri–pNPA30 for Chemical Immobilization on Aminated Polystyrene Plates

1. Add a mixture of 1.4 mg (2.6 mmol) of Btri-O(CH2)3NH2 and 1.5 mL (11 mmol) of NEt3 dissolved in 100 mL of DMSO to a solution of 1.9  mg (12.8  mmol) of pNPA in 400  mL of DMSO. 2. Keep the mixture for 12 h at 40°C performing control for the reaction by TLC; eluent A: Rf 0.3 for Btri-O(CH2)3NH2, Rf 0 for Btri–pNPA30 (see Subheading 3.2). 3. Dilute 100 mL of Btri–pNPA30 reaction mixture with 900 mL of DMSO containing 0.2 mL of NEt3 (to concentration 0.94 nmol Btri/mL) and use the obtained solution for coating of aminated polystyrene plates (see Subheading 3.4.2).

3.1.4. Synthesis of Btri–pNSA2000 for Chemical Immobilization on Aminated Polystyrene Plates

1. Add a mixture of 1.4 mg (2.6 mmol) of Btri-O(CH2)3NH2 and 1.5 mL (11 mmol) of NEt3 dissolved in 100 mL of DMSO to a solution of 1.7  mg (12.8  mmol) of pNSA in 400  mL of DMSO. 2. Keep the mixture for 12 h at 40°C performing control for the reaction by TLC; eluent A: Rf 0.3 for Btri-O(CH2)3NH2, Rf 0 for Btri–pNSA2000 (see Subheading 3.2). 3. Dilute 100 mL of Btri–pNSA2000 reaction mixture with 900 mL DMSO containing 0.2 mL of NEt3 (to concentration 0.94 nmol Btri/mL) and use the obtained solution for coating of aminated polystyrene plates (see Subheading 3.4.2).

12  Immobilization of Polyacrylamide-Based Glycoconjugates… 3.1.5. Synthesis of Btri–PAA30–biot5 for Immobilization on Str Plates

179

1. Add a mixture of 500 mg (1.3 mmol) of biot-NH(CH2)6NH2 and 0.4 mL (2.6 mmol) of NEt3 dissolved in 100 mL of DMSO to a solution of 4.7  mg (27.5  mmol) of pNPA in 300  mL of DMSO. 2. Add a mixture of 3 mg (5.5 mmol) of Btri-O(CH2)3NH2 and 1.5 mL (11 mmol) of NEt3 in 100 mL of DMSO to the obtained solution and keep it for 24 h at 40°C; absence of the free aminoligands in the reaction mixture was confirmed by TLC; eluent A: Rf 0.3 for Btri-O(CH2)3NH2, Rf 0 for Btri–PAA30–biot5; eluent B: Rf 0.54 for biot-NH(CH2)6NH2; Rf 0 for Btri–PAA30– biot5 (see Subheading 3.2). 3. Add an excess of ethanolamine (90 mL) to the reaction mixture and keep it for 24 h at 40°C. 4. Purify the product by GPC (see Subheading 3.3). 5. Yield 85–95%; white solid.

3.1.6. Synthesis of Btri–PAA30–biot1 for Immobilization on Str Plates

1. Add a mixture of 3 mg (5.5 mmol) of Btri-O(CH2)3NH2 and 1.5 mL (11 mmol) of NEt3 dissolved in 100 mL of DMSO to a solution of 4.7 mg (27.5 mmol) of pNPA–biot1 in 300 mL of DMSO. 2. Keep the mixture for 12 h at 40°C performing control for the reaction by TLC; eluent A: Rf 0.3 for Btri-O(CH2)3NH2, Rf 0 for Btri–PAA30–biot1; eluent B: Rf 0.54 for biot-NH(CH2)6NH2; Rf 0 for Btri–PAA30–biot1 (see Subheading 3.2). 3. Add an excess of ethanolamine (90 mL) to the reaction mixture and keep it for 24 h at 40°C. 4. Purify the product by GPC (see Subheading 3.3). 5. Yield 85–95%; white solid.

3.2. Thin-Layer Chromatography

1. Apply 5 mL of reaction mixture on a TLC plate and develop the plate using appropriate eluent (see Subheading 3.1.1–3.1.6). 2. Spray the charring reagent and heat the plate for 10  min at 150°C to visualize spots.

3.3. Preparative GPC

1. Apply reaction mixture on a glass column (75 × 1.5 cm) packed with Sephadex LH-20. 2. Elute the column with the eluent C. 3. Collect the first 50 mL of eluate and evaporate it to dryness. 4. Dissolve the residue in water and freeze dry the obtained solution.

3.4. Plate Coating with Glycopolymers 3.4.1. Physical Adsorption

1. Add serially twofold diluted Atri–PAA solution in coating buffer (concentration 0.02–200 mg/mL or 2–2 × 105 pmol Atri per well) to the plate (100 mL per well); keep the plate for 1 h at 37°C (see Notes 5, 6, and 7).

180

O.E. Galanina et al.

2. Wash the plate (hereinafter, the washing buffer, 200  mL per well). 3. Block the plate (the blocking buffer, 200  mL per well) and wash. 3.4.2. Chemical Immobilization

1. Add serially twofold diluted Btri–pNPA30 or Btri–pNSA2000 solution in DMSO (concentration 0.02–200 mg/mL or 2–188 pmol Btri per well) to the aminated plate (100 mL per well) and incubate the plate for 24 h at room temperature. 2. Add 3 M aqueous ammonia to the plate (10 mL per well) and incubate for 24 h at room temperature. 3. Wash the plate (hereinafter, the washing buffer, 200  mL per well). 4. Block the plate (the blocking buffer, 200  mL per well) and wash (See Note 7).

3.4.3. Coating via Biotin–Str Bridge

1. Rinse streptavidin-coated, black, 96-well plate (see Note 7) twice with PBS. 2. Add serial twofold dilution of coating solution in PBS (containing 0.002–80  mg/mL biotinylated glycoconjugate or, respectively, 2–9 × 104 pmol Btri per well) to the plate (100 mL per well); keep the plate for 1 h at 37°C. 3. Wash the plate (hereinafter, the washing buffer, 200  mL per well). 4. Block the plate (the blocking buffer, 200  mL per well) and wash (See Note 7).

3.5. ELISA 3.5.1. ELISA with Colorimetric Detection

1. Add A16 mAbs (1:4,000 in PBS containing 0.3% BSA) to the 96-well transparent plate (100 mL per well). 2. Incubate the plate for 1 h at 37°C and wash with the washing buffer. 3. Add Ig–Biot (1:2,000 in PBS containing 0.3% BSA) to the plate (100 mL per well), incubate the plate for 1 h at 37°C, and wash with the washing buffer. 4. Add Str-HRPO (1:2,000 in PBS containing 0.3% BSA) to the plate (100 mL per well), incubate the plate for 1 h at 37°C, and wash with the washing buffer. 5. Just before use, add 0.03% H2O2 to a colorimetric solution and apply the obtained solution to the plate. Incubate the plate for 30 min at room temperature; stop the reaction by addition of 50 ml 1 M aqueous H2SO4. Read an absorbance at 492 nm in Victor2 counter. 6. Measure optical density (492  nm) by Victor2 counter. Carry out each assay in duplicate, and omit mAbs to obtain blank reaction.

12  Immobilization of Polyacrylamide-Based Glycoconjugates… 3.5.2. ELISA with Fluorimetric Detection

181

1. Add B8 mAbs (1:100 in PBS containing 0.3% BSA) to the 96-well black plate (100 mL per well). 2. Incubate the plate for 1 h at 37°C and wash. 3. Add Ig–AP (1:5,000 in PBS containing 0.3% BSA) to the plate (100  mL per well); incubate the plate for 1  h at 37°C and wash. 4. Add fluorescent-revealing solution to the plate (100  mL per well); incubate the plate for 30 min at room temperature. 5. Measure fluorescence intensity (355/460 nm) by Victor2 multilabel counter (PerkinElmer, USA). Carry out each assay in duplicate, and perform blank reaction by omitting the mAbs. Subtract the blank reading from the final fluorescence to provide the corrected fluorescence intensity values.

4. Notes 1. The meanings of apparent m.w. (in kDa) are shown in the glycopolymer notations as the upper index. Degree of polymerization in the case of high-m.w. activated polyacrylic acid is reproduced with high accuracy; it could be monitored by analytical GPC. For details, see ref. 16. 2. Molar fraction of the monomer units substituted with Glyc is 20 mol% for all the glycopolymers considered here; according to our previous experience, this is the optimal loading of polymer scaffolds affording to better binding with carbohydraterecognizing antibodies (8, 9). 3. Molar fraction of the monomer units substituted with biotin in Glyc–PAA30–biot5 is 5 mol%, which corresponds to five biotin residues per molecule of the glycopolymer. 4. 30-kDa Glyc–PAAs can be available via home pages of CFG (http://www.functionalglycomics.org/), Lectinity Corp. (http://www.lectinity.com/), or GlycoTech Corp. (http:// www.glycotech.com/). 5. Physical immobilization of Glyc–PAA is carried out at 37°C for 1 h or alternatively at 4°C for 12 h. These two regimens give rise to the same efficiency. 6. Coating buffer pH basically does not affect the efficiency of Glyc–PAA immobilization in the pH interval from 5.0 to 9.6. Salts-free water as a solvent gives twofold less rate of coating. 7. Serially diluted coating antigen and constant concentration of antibodies is recommended as a protocol optimization stage. Then, when the optimal antigen coating has been selected, the antibodies in test are serially diluted.

182

O.E. Galanina et al.

8. Storage and use of Str-coated plates should be in strict accordance with manufacturer’s manual in order to avoid loss of capacity. Otherwise, plates with similar quality and capacity could be prepared in-house by coating Str onto NUNC Maxisorb polystyrene. 384-well avidin-coated plates could be used as well (6).

Acknowledgments This study was supported by the Program of the Presidium of the RAS “Molecular and Cell Biology” (FP6 Project SP5B-CT2007-044512) and the RFBR grant #10-04-01693-a. References 1. Varki, A., Cummings, R.D., Esko, J.D., et  al. (2009) Essentials of Glycobiology, 2nd ed. CSH Press, New York. 2. Fukuda, M., Hindsgaul, O. (2003) Introduction to Glycobiology. Oxford University Press, New York. 3. Brockhausen, I., Kuhns, W. (1997) Glycoproteins and Human Disease. Springer, Heidelberg, Germany. 4. Katrlik, J., Švitel, J., Gemeiner, P., Kožár, J., Tkac, J. (2010) Glycan and lectin microarrays for glycomics and medicinal applications. Med Res Rev 30: 394–418. 5. Liang, P.H., Wu, C.Y., Greenberg, W.A., Wong, C.H. (2008) Glycan arrays: Biological and medical applications. Curr Opin Chem Biol 12: 86–92. 6. Alvarez, R.A., Blixt, O. (2006) Identification of ligand specificities for glycan-binding proteins using glycan arrays. Meth Enzym 415: 292–410. 7. Larsen, K., Thygesen, M.B., Guillaumie, F., Willats.,W.G.T., Jensen, K.J. (2006) Solidphase chemical tools for glycobiology. Carb Res 341: 1209–1234. 8. Bovin, N.V. (2003) Neoglycoconjugates as probes in glycobiology. In Chemical Probes in Biology (M. P. Schneider Ed.), Kluwer Academic Publishers, Dordrecht, Germany, pp. 207–225.

9. Bovin, N.V. (1998) Polyacrylamide-based ­glycoconjugates as tools in glycobiology. Glycoconj J 15: 431–446. 10. Galanina, O.E., Mecklenburg, M., Nifant’ev, N.E., et  al. (2003) GlycoChip: multiarray for the study of carbohydrate-binding proteins. Lab Chip 3: 260–265. 11. Shilova, N.V., Galanina, O.E., Pochechueva, T.V., et al. (2005) High molecular weight neoglycoconjugates for solid phase assays. Glycoconj J 22: 43–51. 12. Dyukova, V.I., Shilova, N.V., Galanina, O.E., et al. (2006) Design of carbohydrate multiarrays. Biochem Biophys Acta 1760: 603–609. 13. Esser, P. (2010) Principles in Adsorption to Polystyrene. ThermoScientific Bulletin 6a. http://www.nuncbrand.com/en/frame. aspx?ID=579. Accessed August 2010. 14. Stryer, L. (1995) Biochemistry, 4th ed. W. H. Freeman and Company, New York, pp. 361–390. 15. Berg, E.L., Robinson, M.K., Mansson, O., Butcher, E.C., Magnani, J.L. (1991) A carbohydrate domain common to both sialyl Le(a) and sialyl Le(x) is recognized by the endothelial cell leukocyte adhesion molecule ELAM-1. Biol Chem 266: 14869–72. 16. Chinarev, A.A., Galanina, O.E., Bovin, N.V. (2010) Biotinylated multivalent conjugates for surface coating. In: Functional Glycomics (J. Li Ed.), Meth Mol Biol 600, Humana Press, 67–78.

Chapter 13 Surface Plasmon Resonance Imaging Analysis of Protein Binding to a Sialoside-Based Carbohydrate Microarray Matthew J. Linman, Hai Yu, Xi Chen, and Quan Cheng Abstract Monitoring multiple biological interactions in a multiplexed array format has numerous advantages. However, converting well-developed surface chemistry for spectroscopic measurements to array-based, high-throughput screening is not a trivial process and often proves to be the bottleneck in method development. This chapter reports the fabrication and characterization of a new carbohydrate microarray with synthetic sialosides for surface plasmon resonance imaging analysis of lectin–carbohydrate interactions. Contact printing of functional sialosides on neutravidin-coated surfaces was carried out and the properties of the resulting elements were characterized by fluorescence microscopy. Sambucus nigra agglutinin (SNA) was used for testing on four different carbohydrate-functionalized surfaces and differential binding was analyzed. Multiplexed detection of SNA/biotinylated sialoside interactions on arrays up to 400 elements has been performed with good data correlation, demonstrating the effectiveness of the biotin–neutravidin-based biointerface to control probe orientation for reproducible and efficient protein binding to carbohydrates. Key words: Carbohydrate microarray, Surface plasmon resonance imaging, Biotinylated sialosides, Biointerface, Biosensor

1. Introduction Carbohydrates are key components in cell surface interactions with proteins, responsible for recognition, cell adhesion, and cell-to-cell signaling (1). Since they are structurally complex, there has been considerable effort in synthesizing carbohydrate analogs and investigating their structure–function properties. Monitoring multiple carbohydrate interactions is challenging but highly desirable for gaining affinity information in a timely and cost-effective manner, particularly for pharmaceutical research. This has led to the construction of carbohydrate microarrays for multiplexed detection of protein–carbohydrate interactions as early as in 2002 (2–5). Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_13, © Springer Science+Business Media, LLC 2012

183

184

M.J. Linman et al.

In the work presented in this chapter, the carbohydrate microarrays are created by contact printing. Generally, methods for array fabrication can be classified as either contact or noncontact printing. The former utilizes pin-type arrayers that transfer a defined volume of sample by directly touching the surface of the substrate, whereas the latter dispenses the sample droplets onto the substrate without a direct contact to the surface. We use contact printing with solid pins as there are a number of advantages to this technology, including easy deposition of viscous solutions (6), simple design that enables reproducible and efficient printing (7), and simple cleaning procedure for the pins (7). In this chapter, we employ the use of a pin and ring spotter known as the GMS 417 arrayer, which uses a large ring to hold the sample and the pass of a pin through the ring deposits a droplet (8). In the last portion of this chapter, biomolecular interaction using this carbohydrate microarray is described in detail. Biointeraction studies desire proper orientation of the ligands and probe density for functional binding. It is well-known that surface chemistry employed to immobilize the biomolecules of choice is of great importance. This is especially true when immobilizing carbohydrates in an array format. Among various covalent coupling approaches that are typically used, one main problem is the lack of control with regards to the orientation of the probe glycan relative to the surface (9). In addition, how this lack of control affects detection performance with new detection schemes, such as label-free methods, has not been investigated as the majority of detection is fluorescence based. Label-free detection with microarray has gained considerable acceptance in the past few years (10, 11) and there was a recent report that uses SPR and XPS to examine a self-assembled carbohydrate monolayer for array analysis (11). But in general, the effects of carbohydrate immobilization on microarray performance and reliability, especially that involving protein-based surface chemistries, have not been extensively studied. In this chapter, we describe in detail the fabrication of a carbohydrate microarray using contact printing with biotinylated sialosides. This array has been further characterized for its effectiveness in the study of interactions with lectins by surface plasmon resonance imaging (SPRi). SPRi allows multiplex detection of biological interactions with real-time kinetic analysis possible (10). We recently demonstrated an SPR spectroscopic analysis with surface chemistry based on synthetic sialosides for the detection of lectins (12). In this chapter, we use a homebuilt SPRi system (13) along with a commercial contact printing arrayer (the pin-and-ring arrayer) to reveal the morphological connection to biotin–neutravidin chemistry in the immobilization of carbohydrates and its impact on performance (14). The arrayer is employed to create upward of 400 individual array elements for demonstrating proof of principle

13  Surface Plasmon Resonance Imaging Analysis of Protein Binding…

185

for high-density fabrication with quality. The biointerface design described herein should be applicable to many biological systems for high-throughput bioanalysis.

2. Materials 2.1. Synthesis of Biotinylated Sialosides

1. Hexaethylene glycol (Sigma–Aldrich).

2.1.1. Reagents

4. KI (Fisher Scientific).

2. Ag2O (Sigma–Aldrich). 3. p-Toluenesulfonyl chloride (Sigma–Aldrich). 5. Celite (Fisher Scientific). 6. Sodium azide (Sigma–Aldrich). 7. TMSOTf (Sigma–Aldrich). 8. Biotin-NHS (Sigma–Aldrich). 9. Sodium methoxide (Fisher Scientific). 10. DOWEX HCR-W2 (H+) resin (Fisher Scientific). 11. 10% palladium on charcoal (Fisher Scientific). 12. 4 Å powdered molecular (Fisher Scientific). 13. Et3N (Fisher Scientific). 14. Hexane (Fisher Scientific). 15. 95% Ethanol (Fisher Scientific). 16. DMF (Fisher Scientific, cat. no. AC34843-0010). 17. CH2Cl2 (Fisher Scientific). 18. Ethyl acetate (Fisher Scientific). 19. Methanol (Fisher Scientific). 20. ManNAc (Sigma–Aldrich). 21. Mannose (Sigma–Aldrich). 22. Sodium pyruvate (Fisher Scientific). 23. Cytidine 5¢-triphosphate disodium salt (CTP; Sigma–Aldrich). 24. MgCl2⋅6H2O (EMD). 25. p-Anisaldehyde (Sigma–Aldrich). 26. H2SO4 (EM science). 27. Acetic acid (Fisher Scientific). 28. Tris base (Fisher Scientific).

2.1.2. Equipment

1. Buchi Rotary evaporator (Fisher Scientific). 2. C25KC incubator shaker (Fisher Scientific). 3. Model 2110 fraction collector (Bio-Rad).

186

M.J. Linman et al.

4. Sorvall legend T/RT benchtop centrifuge (Fisher Scientific). 5. Sterile 50-mL centrifuge tubes (Biologix Research Company). 6. Bio-Gel P-2 Gel, fine (Bio-Rad). 7. Thin-layer silica gel plates (Sorbent technologies). 8. Silica gel (Sorbent technologies). 2.1.3. Enzymes

1. Enzymes used in this protocol, including sialic acid aldolase from Escherichia coli K-12, CMP-sialic acid synthetase from Neisseria meningitidis (NmCSS), sialyltransferase from Pasteu­ rella multocida (PmST1), a2–6-sialyltransferase from Photo­ bacterium damsela (Pd2,6ST), are recombinant proteins cloned, expressed, and purified in Prof. Xi Chen’s laboratory (15–18). 2. p-Anisaldehyde sugar stain is prepared by adding p-anisaldehyde (25  mL) to ice-chilled methanol (425  mL). Concentrated H2SO4 (50 mL) was then cautiously added dropwise to the methanol solution incubated in an ice bath during a 60-min period with vigorous stirring without splashing. The staining solution prepared was stored at −20°C before use.

2.2. Preparation for Biotinylated Sialoside Surface

1. Running buffer: 20 mM phosphate-buffered saline (PBS), pH 7.4, 150 mM NaCl. 2. Prepare Sambucus nigra agglutinin (SNA) and Fluorescein tagged-SNA (FL-SNA) (Vector Laboratories, Burlingame, CA) in 20 mM PBS, pH 7.4, 150 mM NaCl at stock concentrations of 1 mg/mL and dilute thereafter for various assays. 3. Prepare Biotinylated bovine serum albumin (b-BSA), Neutravidin, Avidin, and Streptavidin (Pierce Biotechnology, Rockford, IL) in PBS in stock solutions of 1 mg/mL. 4. Prepare BSA (Sigma–Aldrich, St. Louis, MO) in a stock solution of 10  mg/mL to use as a control/blocking reagent to prevent nonspecific binding. 5. Prepare a piranha solution, H2SO4:H2O2 (70/30% w/v) by warming the solution in a large glass dish at 70°C for 30 min. This solution is used to clean the glass microscope slides.

2.3. Array Fabrication and SPR Imaging Characterization

1. Clean BK-7 Glass Slides (Corning, NY) coated with about 50 nm of gold via e-beam evaporation are the sensing substrates. 2. The sensing substrates are spotted with various carbohydrate probes via a 96-well EIA/RIA Plate (Corning, NY) in the GMS 417 Microarrayer (Genetic Microsystems) contact printer. 3. All substrates are analyzed for biomolecular binding with a homebuilt SPRi system (13). Attached to this imaging system is a homebuilt flow cell with area ~2  cm2 as defined by the polycarbonate O-ring.

13  Surface Plasmon Resonance Imaging Analysis of Protein Binding…

187

4. 60-mL Syringe with Luer-Lock tip (BD Biosciences, Sparks, MD) is used to store buffer solution which is delivered to the surface at a specific flow rate via a syringe pump (KD ScientificModel 100, Holston, MA). 5. Various analyte injections are made with a 6-Port Medium Pressure Injection Valve, Model V-451, PFA Tubing. The whole microfluidic system also combines 0.040″ diameter, fingertight PEEK fittings (F-120), flangeless fittings, and ferrules (Upchurch Scientific, Oak Harbor, WA).

3. Methods The preparation of biotinylated sialosides and fabrication of the array itself are both equally important steps. Presented herein, we cover the synthesis of the carbohydrates themselves and also the fabrication of the carbohydrate microarray. Finally, the SPRi experimental process is detailed. 3.1. Synthesis of Biotinylated Sialosides

1. Dissolve hexaethylene glycol 1 (25 g, 90 mmol) in anhydrous CH2Cl2 (300 mL) in a 1,000-mL round-bottom flask and add Ag2O (33.0 g, 135 mmol). Cool the flask to 0°C, stir the mixture for 15 min, and then follow with the addition of p-toluenesulfonyl chloride (20.0 g, 100 mmol) and KI (3.2 g, 20 mmol). Stir the reaction mixture at room temperature overnight. Remove the precipitated silver salts by filtration through Celite and wash the Celite with EtOAc. Evaporate the combined solution in a rotary evaporator and purify the compound by a silica gel column using a mixed solution, CH2Cl2:MeOH = 20:1, as an eluent providing the product hexaethylene glycol monotosylate 2 as a colorless oil (Fig. 1). 2. Add sodium azide (11 g, 169 mmol) to a solution of hexaethylene glycol monotosylate 2 (32  g, 73.3  mmol) in DMF (150 mL) in a 500-mL round-bottom flask. Stir the reaction mixture at 60°C for 12  h. Cool to room temperature, and remove the DMF by rotary evaporation at 40°C. Purify the residue by flash chromatography using a silica gel column and CH2Cl2:MeOH = 10:1 as an eluent to produce pure hexaethylene glycol azide 3 as a colorless oil. 3. Dissolve lactosyl trichloroacetimidate 4 (4.7  g, 6.02  mmol) (prepare by chemical synthesis as described (18) and store at −20°C before use) in dry CH2Cl2 (60 mL) containing 4 Å powdered molecular sieves (1 g) in a 250-mL round-bottom flask under argon atmosphere. Add hexaethylene glycol azide 3 (1.85 mL, 6.02 mmol) using a syringe and cool the reaction mixture to −30°C and stir for 30  min. Added TMSOTf

188

M.J. Linman et al.

Fig. 1. Chemoenzymatic synthesis of biotinylated sialosides 10–13. Aldolase recombinant Escherichia coli K-12 sialic acid aldolase, NmCSS recombinant Neisseria meningitidis CMP-sialic acid synthetase, PmST1 recombinant multifunctional Pasteurella multocida a2–3-sialyltransferase, Pd2,6ST recombinant multifunctional Photobacterium damsela a2–6sialyltransferase.

(3.0 mmol, 0.5 mL) dropwise for a duration of 10 min. Allow the reaction mixture to warm up to room temperature and stir it overnight. Then, add Et3N (1.0 mL) to quench the reaction. Filter and concentrate the reaction mixture over Celite. Finally, purify the crude material by a silica gel column using hexane:acetone = 1.5:1 as an eluent to produce peracetylated lactoside 5 as a white solid. 4. Dissolve lactoside 5 (2.0  g, 2.16  mmol) in dry methanol (30 mL) containing a catalytic amount of sodium methoxide. Stir the resulting mixture at room temperature for 12  h and neutralize it with DOWEX HCR-W2 (H+) resin. Filtration and concentration afforded deacetylated lactoside 6 as a white solid. 5. Dissolve compound 6 (1.1 g, 1.74 mmol) in water (15 mL) and stir the mixture under hydrogen atmosphere in the presence of 10% palladium on charcoal for 2  h. Next, filter and concentrate the solution to afford the reduced intermediate. Dissolve the residue (0.48 g, 1.08 mmol) in anhydrous DMF (15  mL) and biotin-NHS (0.41  g, 1.19  mmol) was added. After stirring overnight, concentrate and purify the reaction

13  Surface Plasmon Resonance Imaging Analysis of Protein Binding…

189

mixture by flash chromatography to produce biotintylated lactoside 7 as a white solid. 6. Chemoenzymatic sialylation of biotintylated lactoside 7 is achieved using a highly efficient and convenient one-pot, threeenzyme system established in the Chen laboratory (16–18). Dissolve N-Acetylmannosamine (ManNAc) 8 or mannose 9 (50–100 mg, 1.5 equiv), biotintylated lactoside 7 (1.0 equiv), sodium pyruvate (7.5 equiv), and CTP (1.5 equiv) in Tris–HCl buffer (5 mL, 100 mM, pH = 8.5) containing 20 mM of MgCl2 in a 50-mL centrifuge tube. After adding E. coli K-12 sialic acid aldolase (2.0 mg), N. meningitidis CMP-sialic acid synthetase (NmCSS) (0.8 mg), and PmST1 (1.2 mg for preparing a2–3 linked sialosides) or Pd2,6ST (0.5 mg for preparing a2–6 linked sialosides), the total volume of the reaction mixture should be brought to 10  mL by adding water. Incubate the reaction solution at 37°C with agitation at 140 rpm for 2–4 h. Monitor the reaction by TLC analysis (EtOAc:MeOH:H2O:HOAc = 4: 2:1:0.1, v/v) and stained with a p-anisaldehyde sugar stain. When the product formation reaches its optimum, add 10 mL of 95% EtOH to precipitate the enzymes in the reaction mixture. Then, remove the precipitate by centrifugation. Concentrate the supernatant by rotary evaporation and purify it by a Bio-Gel P-2 gel filtration column to produce pure sialoside products 10–13 (Fig. 1). The structures of sialoside products are characterized by 1H and 13C NMR as well as mass spectrometry. 3.2. Preparation of Gold Slides

1. Before array fabrication, immerse the sensor substrates for 10 min in a piranha solution (caution – see Note 1) and then rinse with both ethanol and DI water. After drying with N2 gas, the sensor substrates should be ready for use. 2. SPR gold chips with a 2-nm chromium adhesion layer and 51  nm of gold are deposited by e-beam evaporation onto cleaned BK-7 glass slides and can be subsequently used in all assays (see Note 2). E-beam evaporation of gold is based on a standardized procedure that should be available on any campus with a Class 100 clean room.

3.3. Surface Preparation with Biotinylated Sialosides

1. All SPR and SPRi experiments use a 20  mM PBS (pH 7.4, 150 mM NaCl) as both a running buffer and dilution buffer (see Note 3). 2. Place the buffer in a 60-mL syringe and pump at a rate of ~8  mL/h across the sensor surface via a homebuilt flow cell (area ca. 2  cm2) connected through to the syringe pump via PFA tubing and fittings/ferrules noted above. 3. The substrates for the arrays are generated by injecting 100 ml of 0.5 mg/mL biotin-BSA through a glass syringe via a six-port, medium-pressure valve. Anchor this valve to a ring stand and connect it to the flow cell.

190

M.J. Linman et al.

4. Allow the sensor chips to incubate for 30 min (see Note 4). 5. Rinse the surface copiously with DI water to avoid salt buildup from the buffer. 6. Add neutravidin (0.5 mg/mL) to the substrate via injection, followed by incubation and rinsing as previously mentioned. 3.4. Array Fabrication Using the GMS 417 Microarrayer

1. Take the functionalized sensor chip from the flow cell on the SPR imager and dry it under a gentle stream of nitrogen gas. 2. Place the cleaned sensor chip in the GMS 417 Arrayer by clipping it into the arrayer platform. The GMS 417 Arrayer uses the pin-and-ring type of contact printing technology, which delivers highly reproducible dots (7). Set the parameters on the pattern such that the arrayed spots are approximately 150 mm in diameter and spot spacing (center to center) adjusted anywhere from 300 to 375 mm. 3. Next, fill the wells of a plastic microtiter plate (A1 to D4–16 spots total) with 50  mL of 2.0  mg/mL of each biotinylated sialoside and clamp into the well plate holder. 4. Fill up the DI water reservoir on the left side of the arrayer. This water is used to clean the pins before and after deposition (see Note 5). 5. Open the “GMS 417” software program and for the first set of arrays, create four small 2 × 2 arrays containing the four biotinylated sialosides. To do this, set the dot spacing at 300 mm and with all four arrays set them to be separated by 1 mm. 6. Start the arrayer run and the arrayer should spot the substrate with the four different biotinylated sialosides. Afterward, the pins should be automatically rinsed in the DI water reservoir which continuously is replaced from a large bottle of fresh DI water with contaminated water taken to a waste reservoir via a vacuum pump. 7. To create 400-spot, high-density arrays, perform the same steps as 1 through 5, but in this case fill all the spots in the microtiter plate and then run the program “test.” This creates a high-density, 400-spot array (see Note 6).

3.5. Setup of SPR Imaging Experiment for Carbohydrate Microarray

1. After creating the carbohydrate microarrays, take the arrayed chips from the GMS 417 microarrayer and clamp it carefully onto the flow cell. 2. Run PBS buffer across the surface of the arrayed substrate to establish a pre-lectin deposition baseline for about 15 min. 3. Once the signal is stable, rotate the polarizer to 90° to take an s-polarized image for background subtraction and then subsequently rotate back to its original position for p-polarized image acquisition (see Note 7).

13  Surface Plasmon Resonance Imaging Analysis of Protein Binding…

191

Fig. 2. Processing SPR images. (a) Image after spotting 1.0  mg/mL Neu5Aca2–6-LHEB carbohydrate on the surface; (b) 500 mg/mL SNA lectin; (c) difference image.

4. An image of the array under p-polarization with carbohydrate only is shown in Fig. 2b with buffer flowing. 5. While continuing to run PBS buffer across the sample surface to achieve a stable baseline, add 500 mg/mL SNA (or fluoresceintagged SNA) onto the substrate via flow injection. The images of the substrate should look similar to Fig. 2a. 6. The image before SNA binding (Fig. 2b) was subtracted from the image after binding (Fig. 2c) to obtain difference images (Fig. 2c). 7. For confirmation of binding events, FL-SNA was deposited on the carbohydrate-functionalized surface in the same manner and images of the arrayed substrate were obtained with a confocal fluorescence microscope (Fig. 3). 8. Fluorescence microscopy experiments can be carried out on a Zeiss LSM 510 confocal laser scanning microscope (CLSM) with 488-nm argon laser excitation, CCD camera, and 527-nmlong pass emission filter with a 10× dipping objective. The carbohydrate arrays are fabricated in the same manner as normal SNA arrays as noted above. 9. After FL-SNA is deposited, place the substrate in a Petri dish and cover with aluminum foil until analysis to avoid photodegradation. 10. Using the 10× dipping objective on the confocal fluorescence microscope, snapshot fluorescent images (like those shown in Fig. 3) can be taken of the arrayed substrates and presented without further optimization in an effort to confirm the specificity of lectin–carbohydrate binding.

192

M.J. Linman et al.

Fig. 3. Fluorescence image of Neu5Aca2–6-LHEB array with SNA interaction on the sensor surface.

Fig. 4.  SPR difference image of a high-density, carbohydrate microarray of (a) Neu5Aca2–6-LHEB interacting with 500 mg/ mL SNA; (b) a zoomed up SPR image of a 6 × 5 arrayed area; and (c) the corresponding reflection intensity profile for the array elements.

3.6. SPR Imaging Analysis

1. Analyze all SPR images using the freeware software ImageJ (available at http://rsbweb.nih.gov/ij/). Firstly, process all the images as 8-bit images with a max signal of 256. 2. Using the difference images, line profiles can be created by drawing a rectangular box around the region of interest. A sample result from this process is shown in Figs. 4.

13  Surface Plasmon Resonance Imaging Analysis of Protein Binding…

193

4. Notes 1. Piranha solution is an extremely strong oxidant and should be handled very carefully. When cleaning slides, always perform this step in a chemical fume hood. 2. After e-beam evaporation of gold, there is a simple method to identify the gold-laden side of the slide. Look down the short axis of the slide and if it lies flat, that is the gold side. If there is a “ledge,” then that is the non-gold side. 3. When making the buffer, we generally make 1 L; this makes the calculations easier and allows for enough buffer for a whole set of experiments. After making the buffer, it is recommended to filter the buffer via vacuum filtration and store it in a plastic container. Before use, we also recommend degassing the buffer with a vacuum hose in a sonicating bath to remove all air bubbles. This extra step helps prevent any air bubble in the microfluidics that could ruin experiments. 4. Biotin-BSA directly binds to gold. But to get high surface coverage to prevent further nonspecific adsorption, at least a 30-min incubation at this step is recommended. 5. When cleaning the pins on the GMS 417 microarrayer, fill up the waste reservoir to DI water. This reservoir should be connected to a pump that takes this dirty DI water to a waste container while constantly replenishing the waste reservoir water from a fresh DI water-holding container. This method always keeps the pins clean and prevents cross-contamination during the spotting process. 6. This “test” program on the GMS 417 microarrayer is an automated program, where the loop of the ring is completely filled and then the solution is spotted 400 times. This assay is generally used to figure out how much solution one would need to put in a given well to allow for 400 spots to be placed in adequate coverage. We found that with 100 ml or so in a given well on the microplate making a 400-spot array with uniform spot density was very reliable. 7. On conventional gold substrates, s-polarized light does not excite surface plasmons; so this s-polarize image serves as a great tool for background subtraction.

Acknowledgments The authors would like to acknowledge the financial supports from NSF grant CHE-0719224 (to QC) and NIH grant R01GM076360

194

M.J. Linman et al.

(to XC). MJL would like to acknowledge the support from the American Chemical Society, Division of Analytical Chemistry Fellowship, sponsored by Procter & Gamble. References 1. Hakomori, S. (2004) Carbohydrate-tocarbohydrate interaction, through glycosynapse, as a basis of cell recognition and membrane organization. Glycoconj. J. 21, 125–37. 2. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nat. Biotechnol. 20, 1011–17. 3. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–81. 4. Park, S. J., and Shin, I. J. (2002) Fabrication of carbohydrate chips for studying protein-carbohydrate interactions. Angew. Chem.-Int. Edit. 41, 3180–82. 5. Houseman, B. T., and Mrksich, M. (2002) Carbohydrate arrays for the evaluation of protein binding and enzymatic modification. Chem. Biol. 9, 443–54. 6. Weibel, C. (2002) “The spotting accelerator™”, customizable head assembly for advanced microarraying. JALA 7, 89–94. 7. Barbulovic-Nad, I., Lucente, M., Sun, Y., Zhang, M. J., Wheeler, A. R., and Bussmann, M. (2006) Bio-microarray fabrication techniques – A review. Crit. Rev. Biotechnol. 26, 237–59. 8. Mace, M. L., Montagu, J., Rose, S. D., and McGuinness, G. (2000) in “Microarray Biochip Technology”, pp. 39–64, Eaton Publishing, Natick, MA. 9. Liang, P. H., Wu, C. Y., Greenberg, W. A., and Wong, C. H. (2008) Glycan arrays: biological and medical applications. Curr. Opin. Chem. Biol. 12, 86–92. 10. Chen, Y. L., Nguyen, A., Niu, L. F., and Corn, R. M. (2009) Fabrication of DNA Microarrays with Poly(L-glutamic acid) Monolayers on Gold Substrates for SPR Imaging Measurements. Langmuir 25, 5054–60.

11. Dhayal, M., and Ratner, D. A. (2009) XPS and SPR analysis of glycoarray surface density. Langmuir 25, 2181–87. 12. Linman, M. J., Taylor, J. D., Yu, H., Chen, X., and Cheng, Q. (2008) Surface plasmon resonance study of protein–carbohydrate interactions using biotinylated sialosides. Anal. Chem. 80, 4007–13. 13. Wilkop, T., Wang, Z. Z., and Cheng, Q. (2004) Analysis of u-contact printed protein patterns by SPR imaging with a LED light source. Langmuir 20, 11141–48. 14. Linman, M. J., Yu, H., Chen, X., and Cheng, Q. (2009) Fabrication and characterization of a sialoside-based carbohydrate microarray biointerface for protein binding analysis with surface plasmon resonance imaging. ACS Appl. Mater. Interfaces 1, 1755–62. 15. Yu, H., Yu, H., Karpel, R., and Chen, X. (2004) Chemoenzymatic synthesis of CMP-sialic acid derivatives by a one-pot two-enzyme system: comparison of substrate flexibility of three microbial CMP-sialic acid synthetases. Bioorg Med Chem 12, 6427–35. 16. Yu, H., Huang, S., Chokhawala, H., Sun, M., Zheng, H., and Chen, X. (2006) Highly efficient chemoenzymatic synthesis of naturally occurring and non-natural alpha-2,6-linked sialosides: a P. damsela alpha-2,6-sialyltransferase with extremely flexible donor-substrate specificity. Angew Chem Int Ed Engl 45, 3938–44. 17. Yu, H., Chokhawala, H. A., Huang, S., and Chen, X. (2006) One-pot three-enzyme chemoenzymatic approach to the synthesis of sialosides containing natural and non-natural functionalities. Nat Protoc 1, 2485–92. 18. Yu, H., Chokhawala, H., Karpel, R., Wu, B., Zhang, J., Zhang, Y., Jia, Q., and Chen, X. (2005) A multifunctional Pasteurella multocida sialyltransferase: a powerful tool for the synthesis of sialoside libraries. J Am Chem Soc 127, 17618–9.

Chapter 14 Glycoarray by DNA-Directed Immobilization François Morvan, Yann Chevolot, Jing Zhang, Albert Meyer, Sébastien Vidal, Jean-Pierre Praly, Jean-Jacques Vasseur, and Eliane Souteyrand Abstract Glycoarrays have become a powerful platform to investigate the interactions of many biological events involving carbohydrates. The carbohydrates immobilization on the surface of the substrates is a key step of glycoarray fabrication. Plenty of strategies have been applied to the immobilization process. Herein a protocol for the synthesis of oligonucleotide glycomimetic conjugates is proposed. The resulting molecules are immobilized by hybridization on a DNA microarray (DNA-directed immobilization; DDI). DDI has been proved to be a very efficient and site-selective. This protocol provides detailed procedures for the preparation of fluorescent oligonucleotide trigalactosylmimetic conjugates and for the preparation of carbohydrate microarrays by DDI on glass slides. Key words: Glycoarrays, Click chemistry, Glycomimetics, DNA-directed immobilization

1. Introduction In recent years, considerable attention has been paid to the study of the interactions involving carbohydrates (1–5). Glycoarrays are very powerful investigation tools, which can realize high-throughput profiling and quantitative analysis by using minute amount of materials (6–18). The surface immobilization of oligosaccharides is one of the crucial steps during the processing of glycoarray fabrication. To set up an efficient glycoarray, many important factors should be considered during the immobilization step, such as the distance between each saccharide (surface density), the distance from the saccharide to the surface of the substrate, the orientation and spatial structure of the saccharide, and so on. Diverse methods of immobilization have been reported (19), which can be divided

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_14, © Springer Science+Business Media, LLC 2012

195

196

F. Morvan et al.

a

Cy5

b

Cy5 Lectins

Oligosaccharides Anchoring sequences

Lectins

Oligosaccharides Cy3 Cy3 Oligonucleotides

Fig. 1. Schematic map of recognition methods. (a) “On-chip” recognition. (b) “In-solution” recognition. Reproduced with permission from Wiley-VCH Verlag GmbH & Co. KGaA (36).

into three groups: (1) physisorption (14, 17, 20, 21), (2) chemical covalent grafting (22–33) and (3) specific biological-based interactions (34–38). The DNA-directed immobilization (DDI) belongs to the third group. The principle of DDI is to immobilize biomolecules coupled with a single-stranded DNA moiety, via its specific hybridization with its complementary sequence immobilized on a solid support (Fig.  1a). DDI was first introduced into protein microarray to study biological events involving proteins and peptides (39, 40), and more recently it has been applied to the fabrication of glycoarray (35–38). The DDI of glycoconjugates has been proved to be a very efficient, site-selective. The DNA platform is reusable (36). Furthermore, a fluorescent tag can be introduced at the free end of the carbohydrate–DNA conjugate for surface immobilization quality control (36). Finally this approach offers the possibility to perform carbohydrate–protein interaction in solution and to address the resulting complex on the surface according to the DNA sequence reference (36, 37) (Fig. 1b). In the following a protocol for the synthesis of fluorescent oligonucleotide galactosylmimetic conjugates is given. Next, the fabrication of the DNA anchoring platform is described: It includes microfabrication of microwells by wet etching on glass slides, surfaces functionalization and covalent immobilization of oligonucleotide. Next the immobilization of fluorescent oligonucleotide carbohydrate conjugates by DDI (Fig.  1a) is described. Finally, interaction of immobilized glycoconjugates with Ricinus communis Agglutinin 120 (RCA 120) is described. The interaction of the fluorescent oligonucleotide carbohydrate conjugate with RCA120 in the solution followed by the immobilization of the resulting complex on the DNA anchoring platform is also described (Fig. 1b).

14  Glycoarray by DNA-Directed Immobilization

197

2. Materials 2.1. Equipment

Rotary evaporator. Thin layer chromatography (TLC) from Macherey-Nagel (Precoated TLC-sheets ALUGRAM SIL G UV254. Silica gel (0.04–0.06 nm). Vacuum desiccator with phosphorous pentoxide (P2O5). Magnetic stirrer. UV–Vis spectrophotometer. Microwave synthesizer with 0.2- to 0.5-mL vials, caps, and crimper. ABI 394 DNA synthesizer with synthesis columns. SpeedVac vacuum system. MALDI-TOF mass spectrometer. Reverse-phase HPLC. Reverse-phase C18 (5 mM) column (150 × 4.6 mm; for analyses). Reverse-phase C18 (15  mM) column (300 × 7.8  mm; for purifications). 822 sputtering system from Material Research Corporation. MJB3 Mask aligner from Zeiss. Nanodrop™ 2000 rim Thermo Scientific. Genepix personal 4100 A fluorescent scanner.

2.2. Solvents and Buffers for Synthesis of Glycoconjugates 2.2.1. Dry Solvents

All the solvents are stored at room temperature if not specified. Dry solvents were obtained as described later. Otherwise synthesis grade solvents were used as received from the manufacturers. 1. Dry Dimethylformamide (DMF) was used. 2. Dry pyridine was dried by distillation over CaH2. 3. Dry Methylene chloride (CH2Cl2) was obtained by distillation over P2O5. 4. Dry and extra dry acetonitrile (CH3CN) was obtained from Biosolve. 5. Dry Triethylamine (Et3N) was obtained by distillation over CaH2. 6. Pyridine Synthesis Grade. 7. Methylene chloride (CH2Cl2). 8. Acetonitrile (CH3CN). 9. Triethylamine (Et3N). 10. Cyclohexane.

198

F. Morvan et al.

11. Ethyl acetate (EtOAc). 12. Methanol (MeOH). 13. Toluene. 14. Acetone. 15. Concentrated aqueous ammonia solution (NH4OH) was stored at 4°C. 16. Sodium bicarbonate saturated aqueous solution. 17. Sodium chloride solution (Brine) saturated aqueous. 18. 0.05 M triethylammonium acetate buffer, (pH 7) (TEAAc). 19. Sulfuric acid (H2SO4) stain solution (EtOH/H2SO4 90:10 v/v). 20. Vanillin stain solution was fabricated as follow: 5 g of vanillin were added in 25 mL of H2SO4, 225 mL of MeOH/H2O, 1:1, v/v. 21. Water was 18.2 MW. 2.3. Reagents for the Synthesis of Carbohydrates DNA Conjugates

In this section, reagents for the synthesis of DNA conjugates are listed. It comprises reagent for the synthesis of azido glycoside, H-phosphonate monoester linkers, DNA (H-phosphonate or phosphoramidate) cycles, and reagents for amidative oxidation and clik chemistry. Reagents were stored according to the manufacturer’s recommendations. Reagents recommended for automated solid-phase oligonucleotide synthesis. All these reagents are commercially available. (1-Dimethoxytrityloxy-propanediol-3-succinoyl)-long chain alkylamino-CPG from Glen Research or could be prepared according to Seela and Kaiser (41). 1. 2-[2-(2-Chloroethoxy)ethoxy]ethanol. 2. 1,2,3,4,6-Penta-O-acetyl-b-d-galactopyranose. 3. Tin(IV) chloride (SnCl4) as a 1 M solution in CH2Cl2. 4. Silver trifluoroacetate. 5. Sodium azide (NaN3) (see Note 1). 6. Tetra n-butylammonium iodide (nBu4NI). 7. Sodium sulfate (Na2SO4). 8. 4,4’-O,O’-Dimethoxytrityl chloride. 9. 1,4-cyclohexanedimethanol (cis- and trans-mixture). 10. Tetraethyleneglycol. 11. Diphenylphosphite. 12. Propargyl amine. 13. Carbon Tetrachloride. 14. Dry pyridine acetonitrile solution 1:1 v/v.

14  Glycoarray by DNA-Directed Immobilization

199

15. Pivaloyl chloride (0.2 M) in dry pyridine acetonitrile solution 1:1 v/v. 16. 3% dichloroacetic acid (DCA) in dry CH2Cl2 was used as deblocking solution in H-phosphonate cycle. 17. Standard 2-cyanoethyl deoxyribonucleoside phosphoramidites. 18. Cy3 phosphoramidite from General Electric Healthcare. 19. 5-Benzylthio-1H-tetrazole (BMT) in solution in dry acetonitrile (0.3 M) was used as activator in the phosphoramidate cycle synthesis of DNA. 20. Oxidation solution: 0.1 M iodine in 70/10/20 (v/v/v) THF/ pyridine/water was used as oxidant Reagents recommended for automated solid-phase oligonucleotide synthesis. All these reagents are commercially available. 21. (1-Dimethoxytrityloxy-propanediol-3-succinoyl)-long chain alkylamino-CPG from Glen Research or could be prepared according to Seela and Kaiser (41). 22. Acetic anhydride in dry pyridine/THF 10/10/80 (v/v/v) was used as Cap A solution in the phosphoramidate cycle synthesis of DNA. 23. N-methylimidazole (10%) in dry THF was used as Cap B solution in the phosphoramidate cycle synthesis of DNA. 24. Dichloroacetic acid 3% (DCA) in dry CH2Cl2 as deblocking solution in the phosphoramidate cycle synthesis of DNA. 25. Copper sulfate (CuSO4 or CuSO4⋅5H2O). 26. Sodium ascorbate. 27. 1-Azido-3,6-dioxaoct-8-yl 2,3,4,6-tetra-O-acetyl-b-d-gala­ cto­pyranoside. 2.4. Fabrication of DNA Anchoring Platform

The glass slides (Borosilicate, Nexterion D, Schott GMBH, Germany) were used as the immobilization substrates. Microreactors are etched on the flat glass slides by means of photolithography and wet etching (42, 43). Each glass slide featured 52 round microreactors (2 mm diameter, 65–100 mm deep with less than 1% variation in depth for one etching lot) with a volume near 1 mL. Next the glass slides were silanized (35–38, 42, 44, 45) leading to ester functional bearing glass slides. After acidolysis of the ester function and activation of the corresponding carboxylic acid, amino-modified oligonucleotides were covalently grafted on the surfaces. Eventually, the DNA-carbohydrate conjugates could be immobilized by hybridization on the surface. Solvents Tetrahydrofuran (purum grade, THF) and Dichloromethan (DCM) (ACS reagent grade) (Riedel de Haen, Seelze, Germany) were dried and stored over molecular sieves when required.

200

F. Morvan et al.

Deionized water 18.2 MW was fabricated by a PURELAB ultra from ELGA (see Note 12). 1. TFD4 detergent (Franklab SA, Billancourt, France). 2. Piranha solution: 96% Sulphuric acid (Riedel de Haen, Puriss, Seelze, Germany), 7 volumes is mixed with 35% hydrogen peroxide (Fluka, Puriss, Steinheim, Germany), 3 volumes (see Note 13). 3. SPR 220-4.5 photoresist (Rohm Haas electronic materials, Lucerne, CH). 4. Photolithographic mask were laser-printed plastic foils. 5. MF 26A developer (Rohm Haas electronic Materials, Lucerne, CH) was used. 6. Chromium Etchant for the selective etching of chromium (Merck, Darmstadt, Germany). 7. Glass etching solution: Buffered oxide etchant (abbreviated BOE, 7/1, Hydro fluoridric acid: ammonium fluoride, Honeywell), 37% hydrochloric acid (HCl, Riedel de Haen, Seelze, Germany) and water (18.2  MW) were mixed in a 1/2/2, v/v ratio. The etching solution is freshly prepared before use. 8. Ethanol and Acetone from Riedel de Haen were used or the removing of SPR 220-4.5 Photoresist. 9. Dry nitrogen (quality 5.0, Linde, Toulouse, France) was used. 10. Pentane (Riedel de Haen, Seelze, Germany). 11. Tetra Hydro furan (THF) was purchased from (quality purum (>99%), Riedel de Haen, Seelze, Germany). 12. Tert-butyl-11-(dimethylamino) silylundecanoate TDMU was synthesized according to Dugas et al. (45). 13. Solution for the conversion of the ester function into the corresponding carboxylic acid: Glacial formic acid from Riedel de Haen (>99%, Seelze, Germany). 14. Carboxylic acid activating solution: 0.1-M N-hydroxysuccinimide (Fluka, Steiheim, Germany) and 0.1-M di (isopropyl) carbodiimide (Fluka, Steinheim, Germany) were dissolved in dry THF. Activated ester glass slides should be used as soon as possible after activation. 15. 10× Phosphate buffer saline (abbreviated PBS 10×): 0.1-M phosphate, 1.38-M NaCl, and 0.027-M KCl. The 10× PBS was prepared by dissolving PBS powder (Sigma, Steinheim, Germany) in 100  mL of Dionised water (18.2  MW). The pH was adjusted to 8.5 by addition of 15-M sodium hydroxide (Sigma, Steinheim, Germany). The solution was stored at 4°C.

14  Glycoarray by DNA-Directed Immobilization

201

16. 1× Phosphate buffer saline (abbreviated PBS 1×): 0.01-M phosphate, 0.138-M NaCl and 0.0027-M KCl. The solution was fabricated by dissolution of Phosphate buffer saline powder (Sigma, Steinheim, Germany) in 1,000 mL of Deionized Water. PBS 1× was stored at 4°C. 17. Amino-modified oligonucleotides solution: 100  mM Aminomodified oligonucleotides (Eurogentec) in PBS 1×. The mother solution was stored at −20°C. Allicuots were further diluted to 25 mM in PBS 10× pH 8.5 prior immobilization on activated ester glass slides. 18. 0.1% (w/v) sodium dodecyl sulfate (SDS, Sigma Steinheim, Germany) washing solution: 0.1% of SDS in DI water; the solution was stored in at RT. 19. Surface blocking solution: 4% Bovine Serum Albumin (BSA, Sigma, Steinheim) in PBS 1×. It was stored at 4°C. 20. Washing solution: 0.05% Tween20 (Roth, Karlsruhe, Germany) in PBS 1×. The solution was stored at 4°C. 21. 1-mM Glycoconjugates hybridization solution: 100-mM glycoconjugates in PBS 1× concentration. The mother solution was allicuoted and stored at −20°C. Allicuots were further diluted at 1 mM in PBS 1× pH 7.4 prior hybridization on glass slides (on-chip approach). 22. Sodium Saline Citrate (SSC) 2× with 0.1% Sodium Dodecyl Sulfate (SDS, Sigma Steinheim, Germany) and SSC 2× were fabricated by dilution of Saline Sodium Citrate 20× (named here after SSC 20×) powdered blend (Sigma, Steinheim, Germany). The SSC 20× was diluted with DI water (18.2 MW) four times to give at SSC 2×. Sodium Dodecyl Sulfate (SDS; Sigma, Steinhiem) was eventually added to give SSC 2×, SDS 0.1%. Solutions were stored at 4°C. 2.5. Biological Recognition

1. 1× Phosphate buffer saline: see Subheading 2.4, item 16. 2. R. communis Agglutinin 120 (RCA120,Sigma, Steiheim, Germany) was stored at 4°C. 3. Cy5 Ab Labeling Kit (Amersham Biosciences, GE Healthcare, Buckinghamshire, UK) was used for the labeling of RCA120. Store at 4°C. Avoid light exposure. 4. Cy5 labeled RCA 120 incubation solution: Cy5 labeled RCA120 (at the desired concentration), 10 mM CaCl2 (Sigma, Steinheim, Germany) and 2% BSA (Sigma, Steiheim) in PBS 1×. The solution was stored at 4°C. 5. Washing solution: 0.02% v/v Tween 20 in PBS 1×. The solution was stored at 4°C.

202

F. Morvan et al.

3. Methods 3.1. Synthesis of 5 ¢-Cy3-DNA-3 ¢Trigalacto­sylmimetics

This protocol describes the method to synthesize 5¢-fluorescent oligonucleotides conjugated with a glycomimetic at the 3¢ end. The glycomimetic is constituted of a scaffold bearing three galactose residues with two different linkers between each galactose residue (Fig. 2). We describe (1) the preparation of galactose azide derivative; (2) the H-phosphonate monoester linkers; (3) the introduction of the H-phosphonate diester linkages by H-phosphonate chemistry; (4) their oxidation into phosphoramidate linkage by carbon tetrachloride in presence of propargylamine allowing the preparation of the alkyne scaffold; (5) the conjugation by click chemistry with carbohydrate azide derivative under copper (I) catalysis by 1,3-dipolar cycloaddition affording the solid-supported fully acetylated glycomimetic; (6) the elongation and the labeling of the oligonucleotide by phosphoramidite chemistry; and (7) the deprotection, purification and characterization.

3.1.1. Synthesis of 1-Azido-3,6-Dioxaoct-8-yl 2,3,4,6-Tetra-O-Acetyl-bd-Galactopyranoside

1. Introduce the peracetylated galactose 1 (5 g, 12.8 mmol) and silver trifluoroacetate (4.24 g, 19.2 mmol) in a 250-mL roundbottom flask. Close the flask with a rubber septum and inert with Argon (Scheme 1). 2. Add dropwise 2-[2-(2-chloroethoxy)ethoxy]ethanol (2.80 mL, 19.2 mmol) and freshly distilled dichloromethane (120 mL) to obtain a solution. Keep an argon atmosphere on the reaction. 3. Add SnCl4 (1 M in CH2Cl2, 38.4 mL, 38.4 mmol) dropwise within ~30 min at 0°C (ice bath) to the stirred solution. Stir the solution under argon at room temperature. DNA anchoring sequence

Cy3 DNA

Linker-O

Fluorescent label HO HO

OH O

O P O NH

N N N O

OH

OH 3 Conjugation on solid-support through amidative oxidation

3

Selective chemical ligation through "click chemistry"

Interaction with lectins

Fig. 2. Structure of the 5¢-fluorescent oligonucleotide 3¢-tri-galactosylmimetic. Reproduced with permission from Wiley-VCH Verlag GmbH & Co. KGaA (36).

14  Glycoarray by DNA-Directed Immobilization AcO

OAc O

AcO 1

OAc

a, b OAc

60 %

AcO AcO

OAc O OAc

O

O

O

203

N3

2

Scheme 1. Synthesis of azido-functionalized galactose derivative. Reagents and conditions: (a) SnCl4, CF3CO2Ag, H(OCH2CH2)3Cl, CH2Cl2, rt, 2–3 h, (b) NaN3, nBu4NI, DMF, 70°C, 16 h.

4. Follow the reaction by TLC (eluent: PE-EtOAc 1:1, Rf = 0.60 for 1 and 0.34 for 2). Disappearance of the starting material was observed within 1–3 h, occasionally with the mixture turning to a pale pink color. 5. Add a solution of saturated aqueous NaHCO3 (100  mL) to adjust pH above 8 and stir the solution vigorously for 20 min. 6. Dilute the resulting biphasic system with 500 mL of water (see Note 2) and extract the aqueous layer with CH2Cl2 (3 × 150 mL). 7. Combine the organic layers and wash with saturated aqueous NaHCO3 (150 mL), water (3 × 150 mL), brine (2 × 150 mL) and dry over sodium sulfate (Na2SO4). 8. Filter the solid on cotton and wash with CH2Cl2 (2 × 50 mL). Evaporate the solvent on a rotary evaporator to obtain a crude pale yellow gum and transfer the gum in a 250-mL roundbottom flask (dissolve and re-evaporate the solvent). 9. Add sodium azide (see Note 1) (4.16 g, 64.0 mmol) and tetran-butyl ammonium iodide (4.73  g, 12.8  mmol). Close the flask with a rubber septum and inert with Argon. 10. Add anhydrous DMF (150 mL) and stir the mixture at 70°C under argon for 16 h. 11. Cool the reaction mixture to room temperature, filter the solids off on a frit and wash with EtOAc (3 × 100 mL). 12. Dilute the filtrate with EtOAc to obtain a total volume of 1 L. Wash the organic layer with saturated aqueous NaHCO3 (3 × 300  mL), water (3 × 500  mL), brine (400  mL), and dry over sodium sulfate (Na2SO4). 13. Filter on cotton the solid and wash with EtOAc (2 × 50 mL). Evaporate the solvent on a rotary evaporator to obtain a crude yellow to orange gum. 14. Purify the product by silica gel column chromatography (eluent: PE-EtOAc 1:1, internal diameter = 45  mm; length =  250–300 mm) to obtain 2 (3.88 g, 60% yield) as a pale yellow gum (see Note 3). 1-Azido-3,6-dioxaoct-8-yl 2,3,4,6-tetra-O-acetyl-b-d-galactopyranoside 2, (36, 46, 47): Rf = 0.34 (1:1 PE-EtOAc). 1H NMR

204

F. Morvan et al.

(CDCl3) d 1.99, 2.04, 2.07, 2.15 (4s, 4 × 3H, 4 × CH3CO), 3.40 (t, 2H, J = 5.0 Hz, CH2N3), 3.60–3.82 (m, 9H, OCH2), 3.88–4.02 (m, 2H, OCH2, H-5), 4.08–4.20 (m, 1H, H-6, H-6¢), 4.59 (d, 1H, J1,2 = 7.9 Hz, H-1), 5.02 (dd, 1H, J3,4 = 3.4 Hz, J3,2 = 10.5 Hz, H-3), 5.21 (dd, 1H, J2,1 = 7.9 Hz, J2,3 = 10.5 Hz, H-2), 5.39 (dd, 1H, J4,5 = 0.7 Hz, J4,3 = 3.4 Hz, H-4). 13C NMR (CDCl3) d 20.3, 20.4, 20.4, 20.5 (4s, 4 × CH3CO), 50.4 (CH2N3), 61.0 (C-6), 66.8 (C-4), 68.5 (C-2), 68.8, 69.8, 70.1 (3s, 3 × CH2O), 70.4 (C-3), 70.4, 70.5 (2s, 2 × CH2O), 70.6 (C-5), 101.1 (C-1), 169.2, 169.9, 170.0, 170.1 (4s, 4 × CH3CO). 3.1.2. Preparation of H-Phosphonate Monoester Linkers

3.1.3. Preparation of [4-(Dimethoxytritylo­ xymethyl)cyclohexyl] methanol 4a

This protocol describes the method to prepare the H-phosphonate used to introduce the H-phosphonate diester linkages which are oxidizer into phosphoramidate linkage by carbon tetrachloride in presence of progargyl amine allowing the preparation of the alkyne scaffold. 1. In a 250-mL round-bottom flask, coevaporate 2.88  g (20 mmol) of 1,4-cyclohexanedimethanol (mixture of cis and trans) 3a with anhydrous pyridine twice, each time with 25 mL of dry pyridine. Dissolve the syrup in 25 mL dry pyridine. 2. While stirring under argon atmosphere, add 5.08 g (15 mmol) of 4,4¢-dimethoxytrityl chloride in three portions over 30 min. 3. Stir the solution for 3 h at room temperature. Monitor by TLC (see Note 4) and stain with sulfuric acid solution. Rf = 0.70 (CH2Cl2/MeOH, 95:5, v/v). 4. Evaporate the solvent and dissolve the syrup in 200  mL of ethylacetate and wash twice with 50 mL of saturated aqueous NaHCO3 and twice with 50  mL of brine using a separatory funnel. 5. Dry the organic layer by adding about 10  g of anhydrous sodium sulfate, filter on cotton in a funnel, concentrate on a rotary evaporator under reduced pressure and coevaporate with toluene to remove pyridine trace. Yellow oil is obtained. 6. Dissolve the crude in a minimal volume of CH2Cl2 and apply on a 5-cm diameter chromatography column containing 80 g of silica gel (0.04–0.06  nm) equilibrated with 200  mL of CH2Cl2 containing 0.5% of Et3N (see Note 5). 7. Gradually increase the concentration of methanol up to 5% in CH2Cl2 containing 0.5% Et3N. 8. Control the purity of fraction by TLC and combine those containing the pure compound. 9. Evaporate to dryness on a vacuum evaporator to obtain pale yellow oil.

14  Glycoarray by DNA-Directed Immobilization

205

10. Characterize 4a by 1H, 13C-NMR and mass spectrometry. [4-(Dimethoxytrityloxymethyl)cyclohexyl]methanol 4a: 4.02 g, 45% yield. TLC Rf = 0.70 (CH2Cl2/MeOH, 95:5, v/v). 1H NMR (CDCl3, 400 MHz): d 1.02–1.93 (4 m, 11H), 2.94–3.03 (m, 2H), 3.47–3.51 (m, 2H), 3.85 (s, 6H), 6.85–7.50 (m, 13H). 13C NMR (CDCl3, 100 MHz): d 25.5, 26.0, 29.1, 29.7, 36.2, 38.1, 38.8, 40.7, 55.2, 66.0, 66.2, 68.6, 68.7, 113.0, 113.2, 126.6, 127.7, 127.8, 129.2, 130.1, 136.1, 136.8, 145.5, 158.4, 158.7. HRFAB (positive mode, nitrobenzyl alcohol) m/z: calculated for C29H34O4 [M]+ 446.2457, found 446.2435. 3.1.4. Preparation of 11-(4,4¢-Dimetho­ xytrityloxy)-3,6,9Trioxaundecanol 2b

3.1.5. Preparation of Triethylammonium [4-(Dimetho­ xytrityloxymethyl) cyclohexyl]methyl Hydrogen Phosphonate 5a

The compound 4b, initially described by Salo et al. (48), is prepared according to the same protocol used to synthesize 4a starting from 3.88 g (20 mmol) of tetraethyleneglycol 3b and affording 5.61 g of 4b. 11-(4,4¢-Dimethoxytrityloxy)-3,6,9-trioxaundecanol 2b: 75% yield TLC (CH2Cl2/MeOH: 95:5  v/v); Rf2b = 0.7, 1H NMR (CDCl3): d 3.6–3.7 (14H, m), 3.6 (6H, s) 3.73 (2H, m), 6.8–7.5 (13H, m). 1. In a 100-mL round-bottom flask, coevaporate 1 g (2.24 mmol) of 2a three times, each with 10 mL dry pyridine. Dissolve in 10 mL dry pyridine. 2. Under magnetic stirring and argon atmosphere, add 3  mL (15.7 mmol) of diphenylphosphite. 3. Monitor by TLC. (CH2Cl2/MeOH/NEt3: 90:2:8, v/v/v): Rf3a = 0.15. 4. After 45 min cool the solution to 5°C with an ice bath and add 5 mL of water: Et3N 1:1 v/v. 5. After 15  min stirring, added 100  mL of NaHCO3 saturated aqueous solution and extract with 100 mL CH2Cl2 three times, each with 100 mL CH2Cl2 using a separatory funnel. 6. Dry the organic layer by adding about 10  g of anhydrous sodium sulfate, filter on cotton in a funnel, concentrate on a rotary evaporator under reduced pressure and coevaporate with toluene to remove pyridine trace. 7. Dissolve the crude in a minimal volume of CH2Cl2 and apply on a 5-cm diameter chromatography column containing 50 g of silica gel (0.04–0.06  nm) equilibrated with 100  mL of CH2Cl2 containing 2% of Et3N (see Note 5). 8. Gradually increase the concentration of methanol up to 8% in CH2Cl2 containing 2% Et3N. 9. Control the purity of fraction by TLC and combine those containing the pure compound.

206

F. Morvan et al.

10. Evaporate to dryness on a vacuum evaporator to obtain as colorless oil. 11. Characterize 5a by 1H, 13C, 31P-NMR and mass spectrometry. Triethylammonium [4-(dimethoxytrityloxymethyl)cyclohexyl] methyl hydrogen phosphonate 5a: 780  mg, 57% yield Rf = 0.15 (CH2Cl2/MeOH/Et3N 89:3:8, v/v/v). 1H NMR (CDCl3, 400 MHz): d 0.82–1.73 (m, 19H), 2.94 (m, 6H), 2.75–2.84 (m, 2H), 3.51–3.58 (m, 2H), 3.63, 3.65 (2s, 6H), 6.68–7.34 (m, 13H). 13C NMR (CDCl3, 100  MHz): d 8.5, 25.7, 25.9, 29.2, 29.6, 36.0, 36.4, 38.6, 38.9, 45.3, 55.1, 55.2, 66.4, 66.6, 68.6, 69.2, 112.9, 113.1, 126.5, 127.0, 127.6, 127.78, 127.83, 128.2, 129.2, 130.0, 130.1, 136.67, 136.70, 145.4, 158.3, 158.6. 31P NMR (CD3CN, 80 MHz): d 7.32, 7.60 (2s, P). HRFAB (negative mode, nitrobenzyl alcohol) m/z: calculated for C29H34O6P1 [M–Et3NH]– 509.2096, found 509.2092. 3.1.6. Preparation of Triethylammonium [11-(4,4¢-dimetho­ xytrityloxy)-3,6,9trioxaundecanol] Hydrogen Phosphonate 5b

Compound 5b is synthesized applying the same protocol than for 5a described earlier starting from 5.32  g of 4b (1.7  mmol) and affording 5.56 g of 5b as colorless oil. Triethylammonium [11-(4,4¢-dimethoxytrityloxy)-3,6,9-trioxaundecanol] hydrogen phosphonate 5b 78% Rf = 0.20 (CH2Cl2/ MeOH/Et3N 89:3:8, v/v/v). 1H NMR (CDCl3, 400  MHz): d 1.27 (t, 9H, J = 7.3  Hz), 2.97–3.03 (m, 6H), 3.22 (t, 2H, J = 5.2  Hz), 3.61–3.66 (m, 12H), 3.75 (s, 6H) 3.97–4.05 (m, 2H), 6.79–7.47 (m, 13H), 6.79 (d, 1H, 1JH-P = 628  Hz). 13C NMR (CDCl3, 100  MHz): d 8.5, 45.4, 55.2, 63.2, 70.3, 70.5, 70.6, 70.7, 70.9, 85.9, 113.1, 126.7, 127.8, 129.2, 130.1, 136.3, 145.1, 158.4, 158.6. 31P NMR (CD3CN, 80  MHz): d 6.8. HRFABMS (negative mode, nitrobenzyl alcohol) m/z: calculated for C29H36O9P1 [M-Et3NH+]– 559.2097, found 559.2067.

3.1.7. Synthesis of Solid-Supported Tri-(DMCH or TEG) H-Phosphonate Diester 7a-b

This protocol describes the automated synthesis on a DNA synthesizer (ABI 394) of the H-phosphonate diester scaffold according to the H-phosphonate chemistry starting from commercially available propanediol solid support (Schemes 2 and 3). 1. Prepare a 200 mM solution of pivaloyl chloride in dry CH3CN pyridine (1:1, v/v) (306 mL in 12.1 mL). Place the bottle in position #11 using the change bottle procedure. Place the dry CH3CN pyridine (1:1, v/v) in position #10. 2. Prepare a 60  mM solution of H-phosphonate monoester 5a and 5b in dry CH3CN pyridine (1:1, v/v) and place in position #5 and #6 using the change bottle procedure. 3. Program the modified elongation cycle for ABI 394 DNA synthesizer (Table 1) (see Note 6). 4. Enter the sequence of the desired pseudo-oligonucleotides: 5¢555T-3¢ and 5¢-666T-3¢. 5 and 6 refer to the position of the

14  Glycoarray by DNA-Directed Immobilization HO

DMTr O

H

3a

HO

H

OH

H

OH

O

O

DMTr O

3 OH

3b

H

O O P O H

1)

4a

DMTr-Cl dry pyridine

H

DMTr O

H

5a

2) H2O Et3N

3 OH

DMTr O

4b

O 5b

207

O O P O Et3NH H

O O P O3 H Et3NH

Scheme 2. Preparation of DMTr-protected H-phosphonate monoesters 5a-b.

O DMTrO

O

3% DCA CH2Cl 2

O HO

DMTr-O-Linker-O

5a-b

O

O P O Et3NH+ H DMTr-O-Linker-O

PivCl / CH 3CN / Pyridine

6

O P O H 3

O O

7a Linker DMCH 7b Linker TEG

Repeated twice

Scheme 3. Elongation of H-phosphonate diester scaffold by H-phosphonate chemistry.

Table 1 Elongation cycle of H-phosphonate building blocks for an ABI 394 DNA synthesizer Step

Reagents and solvents

Time (s)

1

Wash

CH3CN

10

2

Wash

CH2Cl2

35

3

Detritylation

Deblocking solution

43

4

Wash

CH3CN

20

5

Wash

Dry pyridine CH3CN (1:1, v/v)

6

Coupling

0.060-M H-phosphonates in dry pyridine CH3CN (1:1, v/v)

2.5

7

Coupling

0.2-M pivaloyl chloride in dry pyridine CH3CN (1:1, v/v)

2.5

8 9

Steps 6 and 7 repeated 5 times Wash

CH3CN

15

H-phosphonate monoester 5a and 5b respectively, and T refers to the propanediol solid support 6. 5. Pack two synthesis DNA columns with 1  mmol of (1-dimethoxytrityloxy-propanediol-3-succinoyl)-long chain alkylamino-CPG. Place it on the synthesizer. 6. Run the synthesis trityl ON to prepare 7a and 7b exhibiting three DMCH and TEG linker, respectively.

208

F. Morvan et al.

7. Flush the column with dry nitrogen for 3 min to remove trace of solvent. 3.1.8. Synthesis of Solid-Supported Tri-(DMCH and TEG) N-Propargyl Phosphoramidate Diester 8a-b

This protocol describes the amidative oxidation of H-phosphonate diester linkages into the corresponding phosphoramidate diester using carbon tetrachloride and propagyrlamine in pyridine allowing the introduction of the triple bounds affording 8a-b (Scheme 4). 1. Prepare a solution containing 100  mL of propargylamine in 900 mL of CCl4/C5H5N (1:1, v/v). 2. Treat, back and forth using two syringes, the column containing 7a and 7b with the propargylamine solution for 30 min. 3. Remove the solution and wash with 1 mL pyridine and three times with 2 mL acetonitrile. 4. Flush with nitrogen for 3 min. This protocol describes the 1,3-dipolar cycloaddition, under Cu(I) catalysis, between the solid-supported three propargyl scaffolds 8a-b and tetracetyl galactose triethylenglycol azide 7 under microwaves activation to speed up the reaction. The copper (I) is generated in situ by reduction of copper sulfate by sodium ascorbate (Scheme 5).

3.1.9. “Click Reaction” Cu (I)-Catalyzed 1,3-Dipolar Cycloaddition

1. Transfer carefully the beads bearing the alkyne scaffold 8a-b (1 mmol) from the synthesis column to a microwave vial containing a mini magnetic barrel. 2. Prepare a 40 mM solution of CuSO4 in degassed water (6.4 mg CuSO4 in 1.0 mL) (see Note 7).

DMTr-O-Linker-O

CCl4 / Pyridine

O

O P O H 3

DMTr-O-Linker-O

O

7a Linker DMCH 7b Linker TEG

O P O NH

H2N

O O 3

8a Linker DMCH 8b Linker TEG

Scheme 4. Amidative oxidation of H-phosphonate diester scaffold introducing the propargyl group. AcO

DMTr-O-Linker-O

O P O NH

O O 3 8a Linker DMCH 8b Linker TEG

OAc O

AcO 2

OAc

O 3

N3

CuSO4 / Na Ascorbate MeOH / H2O Microwaves (60 °C, 30 min)

Scheme 5. Click reaction for the synthesis of the galactosylmimetics.

DMTr-O-Linker-O

AcO AcO

OAc O OAc

O P O NH

N N N O

3

O O 3

9a Linker DMCH 9b Linker TEG

14  Glycoarray by DNA-Directed Immobilization

209

3. Freshly prepare a 50  mM solution of sodium ascorbate in degassed water (5 mg in 0.5 mL) (see Note 7). 4. Prepare a 100  mM solution of tetracetyl-galactose triethyleneglycol azide derivative 2 in methanol (10.1 mg in 200 mL). 5. Add to 8a-b, 50 mL of degassed water and 100 mL of solution of 2 prepare on step 4. 6. Premix 10 mL of CuSO4 solution and 40 mL of ascorbate solution and transfer the brown mixture to the microwave vial. 7. Flush the vial with argon and close it with an aluminum seal. 8. Place the vial in a microwave synthesizer Initiator™ from Biotage set at 100 W for 30 min at 60°C with a 30 s premixing time. 9. Filter off the beads using a cone and cotton and wash with 500  mL water:methanol (1:1, v/v), 500  mL methanol and 500 mL diethyl ether. 10. Transfer carefully the beads in a synthesis column and dried the column in a dessicator over P2O5 for 2 h under vacuum (see Note 8). 3.1.10. Elongation of Oligonucleotide and Fluorescent Labeling

Since the hydroxyls of galactose are protected by an acetyl group, the elongation of the oligonucleotide and its 5¢ end labeling is performed by standard phosphoramidite chemistry (see Note 6) affording after a last treatment with ammonia the 5¢-Cy3oligonucleotide 3¢-tri-galactose conjugates (10a-b). The conjugates are purified by reverse phase HPLC and used for the immobilization on DNA chips (Scheme 6). 1. Place the synthesis column containing the glycomimetic 10a-b on a DNA synthesizer. 2. Type the sequence and run the synthesis according to standard phosphoramidite chemistry (Table 2). 3. Transfer the beads into a sealed HPLC vial, add 3 mL of concentrated aqueous ammonia solution (NH4OH), and place the vial in a dry bath for 5 h at 65°C. 4. Withdraw the supernatant and evaporate the ammonia using a SpeedVac vacuum system. 5. Dissolve the residue in 300  mL MilliQ water and purify the conjugates on a C18 column by HPLC using a linear gradient of acetonitrile 12–28% in 25  min at 2  mL/min in TEAAc 50 mM pH7 buffer affording pure 5¢-Cy3-oligonucleotide trigalactosylmimetics 10a-b. 6. Pool the fractions containing the pure conjugate and evaporate the solvent. 7. Coevaporate with 3 mL MilliQ water ten times to remove the triethylammoniun acetate salt (see Note 9).

O

O DMTr-O-Linker-O P O NH n AcO AcO

OAc O

N N N O

O

9a Linker DMCH 9b Linker TEG

3

OAc

Automated oligonucleotide synthesis by Phosphoramidite chemistry

and coupling of Cy3-phosphoramidite

NH4OH O O Cy3 O P O CTG CCT CTG GGC TCA OP O O O

10a Linker DMCH 10b Linker TEG

HO

OH O

HO

Linker-O

O P O NH 3

OH

N N N O

3

OH

Scheme 6. Elongation and labeling of the galactosylmimetics.

Table 2 Standard oligonucleotide elongation cycle for an ABI 394 DNA synthesizer Step

Reagents and solvents

Time (s)

Wash

CH3CN

10

Wash

CH2Cl2

35

Detritylation

Deblocking solution

43

Wash

CH3CN

20

Coupling

0.075-M commercial amidites + 0.3-M BMT in CH3CN

Wait

6 20

Wash

CH3CN

15

Capping

Cap A and Cap B solution

10

Wait

15

Wash

CH3CN

Oxidation

Oxidation solution

Wait Wash

10 8 13

CH3CN

20

14  Glycoarray by DNA-Directed Immobilization

211

8. Dissolve the pure conjugates into 300  mL MilliQ water and quantify their amount by UV and visible spectroscopy (see Note 10). 9. Characterize the conjugates by MALDI-TOF mass spectrometry using hydroxyl pinacolic acid (HPA) as matrix with 10% of ammonium citrate (see Note 11). 10. Freeze the solution and lyophilize the conjugates. They could be stored at −20°C for several months. 3.2. Fabrication of the Anchoring Platform 3.2.1. Fabrication of the Microreactors (42, 43)

Technology process of the microreactors fabrication was derived from the protocol of Mazurczyk et  al. (42). The process flow is shown in Fig. 3. It is a three-step process that comprises the physical vapor deposition of a chromium protective layer, photolithography, and chemical ecthing of the glass substrate with fluoridric acid based etchant. Deposition of the chromium layer 1. Wash successively glass slides with TDF4 detergent solution, a fresh Piranha mixture (10 min) (see Note 13). 2. Rinse with DI water (18.2 MW). Glass substrate Chromium deposition Glass substrate Spin-casted photoresist Glass substrate Photolithography Glass substrate BOE/HCl etching

Glass substrate Removal of the photoresist Glass substrate Removal of the Chromium Glass substrate

Fig. 3. Technology process flow of microreactors fabrication. (1) The deposition of a chromium layer. (2) A photolithographic step. (3) Opening of the chromium. (4) Glass etching. (5) Removing of the protective layers.

212

F. Morvan et al.

3. Dry by centrifugation at 250 × g for 3 min. 4. A 150  nm chromium layer was deposited using magnetron sputtering (MRC 822 system). The system was operated at a RF power of 5  kW, reflected power was 2  W, turret voltage 2.6 kV. The Argon flux was set to 50 sccm and the working pressure was 2.6 × 10−3 Torr. 5. Spin-cast SPR 220 4.5 photoresist at 4,000  rpm for 30  s (30,000 rpm/s) resulting in a 4 mm thick layer. 6. Perform the baking at 115°C for 1 min 30 s on a hot plate. 7. Photolithography was carried out with a Karl Suss (CH) MJB3 Mask Aligner. A 22 s illumination was performed. 8. Immerse the slides in MF26 A and shake gently for 1 min. 9. Rinse in running DI water for 5 min and dry under a dry nitrogen flux. 10. Postbake at 115°C for 2 min (see Note 13). 11. Open the chromium windows with chromium etchant (Merck). 12. Rinse for 15 min in running DI water. 13. Etch the glass substrate using the wet etching solution (1/2/2 of BOE: HCl: H2O, v/v/v) at room temperature for 1 h and 15 min. No mixing (see note 13). Depth of the microwell can be monitor by removing slides from the etching buffer, by rinsing thoroughly the slide, and by measuring the depth with a mechanical profiler (Alfa-step 500 from KLA Tencor). If further etching is required, the slide should be reimmersed in the etching solution until the desired depth is obtained. 14. Rinse for 15 min in running water. 15. Remove the remaining photoresist with acetone. 16. Rinse with ethanol and water. 17. Remove the remaining chromium with chromium etchant (Merck, Darmstadt, Germany) Rinse in DI water for 5 minutes. 18. Depth and surface roughness was measured using a mechanical profiler (Alfa-step 500 from KLA Tencor). 3.2.2. Silianisation of the Glass Slides (45, 49)

1. The following protocol was developed for 20 standard microscope glass slides. 2. Wash the glass slides with fresh Piranha solution for 20  min (see Note 13). 3. Rinse in deionized water four times 10 min. 4. Dry by centrifugation at 250 × g for 3 min. 5. Place the slide in a sealed reactor. 6. Heat under dry nitrogen for 2 h at 150°C. 7. Add dry pentane (250 mL) to completely immerse the glass slides. 8. Add 300 mL of tert-butyl-11-(dimethylamino) silylundecanoate.

14  Glycoarray by DNA-Directed Immobilization

213

9. Incubate at room temperature under dry nitrogen for 2 h. 10. Evaporate the pentane under reduce pressure. By connecting a vacuum pump to the sealed reactor. Do not allow air to enter in the reactor. 11. Heat the slides at 150°C overnight. 12. Wash in THF 10 min under sonication. 13. Rinse with DI water. The resulting glass slide is stable over 6 months to a year. 3.2.3. Conversion of the Tert-Butyl Ester into Carboxylic Function

1. Immersed the silanized glass slides in formic acid for 7  h at RT.

3.2.4. Activation of Carboxylic Functions

1. After removal of the tert-butyl protecting group, the resulting carboxylic function should be activated for amine coupling.

2. Wash successively 10 min in THF (Sonication) and 10 min in water sonication.

2. Immerse the glass slides overnight at RT in the activating solution (NHS 0.1 M and DIC 0.1 M in THF) under gentle stirring. 3. Rinse successively in THF 10 min and dichloromethane 10 min under sonication. 3.2.5. Immobilization of Single-Stranded DNA (Amino-Modified Oligonucleotides (36))

1. Pipette down the desired volume of the amino oligonucleotides in PBS (25 mM) in each microreactor see note 14, 15 and 16. 2. Allow to react with the carboxylic activated glass slides overnight at RT in a H2O vapor saturated chamber. 3. Allow the solution to slowly evaporate overnight at RT room temperature. 4. Wash the slides with SDS 0.1% at 70°C for 30 min. 5. Rinsed with DI water.

3.2.6. Blocking

1. Immerse the slide bearing covalently grafted DNA in the Blocking solution (4% BSA) for 2 h at 37°C. 2. Wash with 0.05% PBS 1× Tween 20 for 3 × 3 min followed by PBS 1× (pH 7.4) three times, and finally rinsed with DI water and dried by centrifugation.

3.2.7. Hybridization of the Glycoconjugates

1. Pipette down in each well the Glycoconjugates solution (at the desired concentration) see note 14, 15 and 16. 2. Incubate for 3 h at RT in a H2O vapor saturated chamber. 3. Wash successively the slide in SSC 2 × 0.1% SDS at 51°C for 1 min followed by SSC 2× at RT for an additional 5 min. 4. Rinse quickly with DI water and dry as soon as possible by centrifugation.

214

F. Morvan et al.

3.3. Biological Recognition (36) 3.3.1. Lectin Labeling

3.3.2. “On-Chip” Recognition (See Fig. 1a)

1. Label R. communis Agglutinin 120 (RCA120) lectin (Sigma) by strictly following the manufacturer protocol (Cy5 Ab Labeling Kit Amersham Biosciences) see note 17. 2. Estimate Protein concentration and the dye to protein ratio by optical density reading the absorbance at 280 and 650 nm with Nanodrop™. Lectin concentration was estimated to be 4 mM bearing an average of four dyes per protein. 1. Dilute Cy5 labeled RCA120 to the desired concentration in PBS 1× (pH 7.4), CaCl2 (final concentration1  mg/mL) and 20% BSA (final concentration 2%) see note 17. 2. Pipette down the lectin solution at the bottom of each microreactors of the slides. 3. Incubate at 37°C in a H2O vapor saturated chamber for 2 h. 4. Wash in PBS 1× (pH 7.4) and Tween 20 (0.02%) for 5 min and dry by centrifugation at 250 × g.

3.3.3. “In-Solution” Recognition (See Fig. 1b)

1. Recognition solution: one glycoconjugate (1 mM final concentration), Cy5 labeled RCA 120 (at desired concentration 0.02–2 mM) were diluted in PBS and Tween 20 (0.02%). 2. Add BSA 20% to reach 2% BSA final concentration. 3. Add CaCl2 (1 mg/mL final concentration). 4. Pipette down 0.5–0.8 mL of each solution in each well of the slide. 5. Incubate 2 h at 37°C in a water saturated chamber. 6. Wash with PBS 1× (pH 7.4) and Tween 20 (0.02%) for 5 min and dried by centrifugation at 250 × g.

3.4. Fluorescence Scanning

Slides were scanned with the Microarray scanner GenePix 4100A software package (Axon Instruments, Sunnyvale, USA) at excitation wavelengths of 532 and 635 nm. The fluorescence signal of each conjugate was determined as the average of the mean fluorescence signal of corresponding spots.

4. Notes 1. CAUTION: Sodium azide, when inhaled, is highly toxic and may cause death (MSDS J.T. Baker). Precautions must be taken when weighing the material such as using a powder mask and a teflon spatula (metallic spatula may cause explosion). The azidation reaction was performed behind a plastic shield due to the potential explosion. DMF is used as a polar solvent favoring the

14  Glycoarray by DNA-Directed Immobilization

215

reaction but also to maintain a basic pH (>8) of the solution. In acidic pH hydrazoic acid (HN3) may be formed, which may explode and/or, when inhaled, may cause intoxication, damage of the central nervous system and blood pressure effects. 2. The solid present in the suspension may be filtered through a bed of Celite for small scale syntheses (>1 g). 3. TLC analyses did not show a significant difference between the polarities of the chlorinated precursors and the azido compounds. 4. Before spotting the reaction mixture, TLC plates must be neutralized by dipping in a solution of 99:1 (v/v) CH2Cl2/Et3N and air-dried, to avoid degradation on the plate (due to the acidity of the silica). Before the elution, the spot of the reaction mixture must be air-dried to remove pyridine and/or Et3N, which could modify the retention factor. After the elution, and before the revelation with stain solution, direct visualization using 254-nm UV lamp can reveal the starting and/or the final compounds. Staining the TLC plate in the stain solution (vanillin or sulfuric acid) reveals compounds. The plate is slowly heated to remove solvent, dipped in the stain solution, washed with water, stamped with absorbent paper to remove excess of water and then slowly heated to dry the plate. 5. It is recommended to equilibrate the chromatography first with an elution solvent containing Et3N to avoid degradation due to the acidity of silica. Derivatives bearing a dimethoxytrityl group are sensitive to acidic treatments so any trace of acid must be neutralized. 6. The H-phoshonate elongation cycle displaying only two steps detritylation and coupling with washes in-between is described in Table 1. The standard phosphoramidite elongation cycle is described in Table 2. 7. To prepare these solutions, water must be degassed to avoid degradation of the oligonucleotide due to the oxygen in presence of copper. Furthermore, it is better to prepare sodium ascorbate solution just before the reaction to allow a good reduction of copper (II) into copper (I). The copper sulfate solution can be prepared in advance and stored several months at −20°C. When both solutions are mixed together the color turns brown and opaque. 8. It is compulsory to have the solid-supported glycomimetic 8a-b as dry as possible since trace of water will react with the nucleoside phosphoramidites leading to less efficient elongation of the oligonucleotide. 9. Traces of TEAAc appear as white film in the round-bottom flask.

216

F. Morvan et al.

10. The extinction coefficient of the oligonucleotide at 260 nm is calculated as the sum of the extinction coefficients of the constitutive nucleosides: dA 15,400; dG 11,500; dC 7,400; and dT 8,700 L/mol/cm. Amount of glycomimetic is also determined at 550  nm taking the extinction coefficient of Cy3 150,000 L/mol/cm. 11. Mass spectrometric data obtained using MALDI-TOF (m/z calcd/found): 9a 6905.08/6906.40, 9b 7054.81/7056.45. 12. Unless stated otherwise, all solutions should be prepared by using DI water (18.2 MW). 13. CAUTION: It is better to be careful with some solutions or agents that maybe harmful for the operator, for example, the Piranha solution (see Subheadings 2.2 and 2.3) is a hazardous solution, the operator need to be protected against spatters caused by the mixing of H2SO4 into H2O2. Buffer oxide etchant is very harmful. Correct protection is required for its manipulation. 14. Try not to touch the surface of the slide (after Silianisation of the glass slides), especially during the process of placing the samples into the microreactors. 15. Adding each sample to the corresponding microreactors with a desired volume, be careful not to overfill, excess fluid may cause cross contamination. 16. The solution of ssDNA and glycoconjugates which stored at −20°C need to be heated at 51°C for 25 min before used due to non specific adsorption on the vial walls.. 17. The lectins are very sensitive, they should be put in an ice cube box (4°C) when they are been used, and then they need to be replaced to the fridge (4°C) quickly.

Acknowledgments The authors thank the financial support from the CNRS, Ecole Centrale de Lyon, Université Claude Bernard Lyon 1, and Université Montpellier 2, “Interface Physique Chimie Biologie: soutien à la prise de risque,” ANR-08-BLAN-0114-01 and Lyon Biopole, Région Rhône–Alpes programme Cible 2010 and the Chinese Scientific Council for the award of a research studentship and Scholarship. The NanoLyon Platefrom is acknowledged for technical support.

14  Glycoarray by DNA-Directed Immobilization

217

References 1. Varki, A. (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3: 93–130. 2. Varki, A., Cummings, R., Esko, J., Freeze, H., Hart, G. W., and Marth, J. (1999) Essentials of glycobiology, Cold Spring Harbor Laboratory Press, New York. 3. Sharon, N., and Lis, H. (2004) History of lectins: from hemagglutinins to biological recognition molecules. Glycobiology 14: 53R-62R. 4. Werz, D., B., and Seeberger, P., H. (2005) Carbohydrates as the Next Frontier in Pharmaceutical Research. Chemistry 11: 3194–3206. 5. Pohl, N., L., and Kiessling, L., L. (1999) Scope of multivalent ligand function: lactose-bearing neoglycopolymers by ring-opening metathesis polymerization. Synthesis 1515–1519. 6. Ratner, D., M., Adams, E., W., Disney, M., D., and Seeberger, P., H. (2004) Tools for Glycomics: Mapping Interactions of Carbohydrates in Biological Systems. ChemBioChem 5: 1375–1383. 7. Ratner, D., M., Adams, E., W., Su, J., O’Keefe, B., R., Mrksich, M., and Seeberger, P., H. (2004) Probing Protein-Carbohydrate Interactions with Microarrays of Synthetic Oligosaccharides. ChemBioChem 5: 379–383. 8. Disney, M., D., and Seeberger, P., H. (2004) The Use of Carbohydrate Microarrays to Study Carbohydrate-Cell Interactions and to Detect Pathogens. Chemistry and Biology 11: 1701–1707. 9. de Paz, J. L., Horlacher, T., and Seeberger, P. H. (2006) Oligosaccharide microarrays to map interactions of carbohydrates in biological systems. Methods in Enzymology 415: 269–292. 10. Uttamchandani, M., Neo, J. L., Zhung Ong, B., N., and Moochhala, S. (2008) Applications of microarrays in pathogen detection and biodefence. Trends in Biotechnology 27: 53–61. 11. Parthasarathy, N., DeShazer, D., Peacock, S. J., Wuthiekanun, V., England, M. J., Norris, S. L., and Waag, D. M. (2008) Application of Polysaccharide Microarray Technology for the Serodiagnosis of Burkholderia pseudomallei Infection (Melioidosis) in Humans. Journal of Carbohydrate Chemistry 27: 32–40. 12. Liang, P.-H., Wu, C.-Y., A, G. W., and Wong, C.-H. (2008) Glycan arrays: biological and medical applications. Current Opinion in Chemical Biology 12: 86–92.

13. Laurent, N., Voglmeir, J., and Flitsch, S. L. (2008) Glycoarrays—tools for determining protein–carbohydrate interactions and glycoenzyme specificity. Chem. Commun. 4400–4412. 14. Wang, D., Liu, S. y., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nature Biotechnology 20: 275–281. 15. Park, S., Lee, M. R., Pyo, S. J., and Shin, I. (2004) Carbohydrate Chips for Studying HighThroughput Carbohydrate-Protein Inter­ actions. Journal of the American Chemical Society 126: 4812–4819. 16. Wang, D. (2003) Carbohydrate microarrays. Proteomics 3: 2167–2175. 17. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nature Biotechnology 20: 1011–1017. 18. Feizi, T., Fazio, F., Chai, W., and Wong, C.-H. (2003) Carbohydrate microarrays – a new set of technologies at the frontiers of glycomics. Current Opinion in Structural Biology 13: 637–645. 19. Larsen, K., Thygesen, M. B., Guillaumie, F., Willats, W. G. T., and Jensen, K. J. (2006) Solid-phase chemical tools for glycobiology. Carbohydrate Research 341: 1209–1234. 20. Palma, A., S., Feizi, T., Zhang, Y., Stoll, M., S., Lawson, A., M., Diaz-Rodrıguez, E., Campanero-Rhodes, M., A., Costa, J., Gordon, S., Brown, G., D., and Chai, W. (2006) Ligands for the b-Glucan Receptor, Dectin-1, Assigned Using “Designer” Microarrays of Oligosaccharide Probe (Neoglycolipids) Generated from Glucan Polysaccharides. Journal of Biological Chemistry 281: 5771–5779. 21. Jaipuri, F., A., Collet, B., Y., M., and Pohl, N., L. (2008) Synthesis and Quantitative Evaluation of Glycero-d-manno-heptose Binding to Concanavalin A by Fluorous-Tag Assistance. Angewandte Chemie, International Edition in English 47: 1707 –1710. 22. Blixt, O., Head, S., Mondala, T., Scanlan, C., Huflejt, M. E., Alvarez, R., Bryan, M. C., Fazio, F., Calarese, D., Stevens, J., Razi, N., Stevens, D. J., Skehel, J. J., Van Die, I., Burton, D. R., Wilson, I. A., Cummings, R., Bovin, N., Wong, C.-H., and Paulson, C. (2004) Printed

218

F. Morvan et al.

covalent glycan array for ligand profiling of diverse glycan binding proteins. Proceedings of the National Academy of Sciences of the United States of America 10: 17033–17038. 23. Park, S., and Shin, I. (2007) Carbohydrate Microarrays for Assaying Galactosyltransferase Activity. Organic Letters 9: 1675–1678. 24. Seo, J. H., Adachi, K., Lee, B. K., Kang, D. G., Kim, Y. K., Kim, K. R., Lee, H. Y., Kawai, T., and Cha, H. J. (2007) Facile and Rapid Direct Gold Surface Immobilization with Controlled Orientation for Carbohydrates. Bioconjugate Chemistry 18: 2197–2201. 25. Smith, E. A., Thomas, W. D., Kiessling, L. L., and Corn, R. M. (2003) Surface Plasmon Resonance Imaging Studies of ProteinCarbohydrate Interactions. Journal of the American Chemical Society 125: 6140–6148. 26. Chevolot, Y., Bucher, O., Léonard, D., Mathieu, H.-J., and Sigrist, H. (1999) Synthesis and Characterization of a Photoactivatable Glycoaryldiazirine for Surface Glycoengineering. Bioconjugate Chemistry 10: 169–175. 27. Chevolot, Y., Martins, J., Milosevic, N., Leonard, D., Zeng, S., Malissard, M., Berger, E., G., Maier, P., Mathieu, H., J., Crout, D., H., and Sigrist, H. (2001) Immobilisation on Polystyrene of Diazirine Derivatives of MonoandDisaccharides: Biological Activities of Modified Surfaces. Bioorganic and Medicinal Chemistry 9: 2943–2953. 28. Leonard, D., Chevolot, Y., Bucher, O., Haenni, W., Sigrist, H., and Mathieu, H. J. (1998) ToFSIMS and XPS study of photoactivatable reagents designed for surface glycoengineering Part 2. N-[m-(3-(trifluoromethyl)diazirine-3-yl) phenyl]-4-(−3-thio(−1 -D-galactopyrannosyl)maleimidyl)butyramide (MAD-Gal) on diamond. Surface and Interface Analysis 26: 793–799. 29. Léonard, D., Chevolot, Y., Bucher, O., Sigrist, H., and Mathieu, H.-J. (1998) ToF-SIMS and XPS study of photoactivatable reagents designed for surface glycoengineering. Part I. N-(m-(3-(trifluoromethyl)diazirine-3-yl) phenyl)-4-maleimido-butyramide (mad) on silicon, silicon nitride and diamond. Surface and Interface Analysis 26: 783–792. 30. Léonard, D., Chevolot, Y., Heger, F., Martins, J., Crout, D., H., G., Sigrist, H., and Mathieu, H., J. (2001) ToF-SIMS and XPS Study of Photoactivatable Reagents Designed for Surface Glycoengineering: Part III: 5-carboxamidopentyl N-[m-[3-(trifluoromethyl) diazirin-3-yl] phenyl b-D-galactopyranosyl]-(1->4)-1-thiob-D-glucopyranoside (Lactose aryl diazirine) on Diamond. Surface and Interface Analysis 31: 457–464.

31. Manimala, J., C., Li, Z., Jain, A., VedBrat, S., and Gildersleeve, J., C. (2005) Carbohydrate Array Analysis of Anti-Tn Antibodies and Lectins Reveals Unexpected Specificities: Implications for Diagnostic and Vaccine Development. ChemBioChem 6: 2229–2241. 32. Manimala, J., C., Roach, T., A., Li, Z., and Gildersleeve, J., C. (2006) High-Throughput Carbohydrate Microarray Analysis of 24 Lectins. Angewandte Chemie, International Edition in English 45: 3607–3610. 33. Manimala, J. C., Roach, T. A., Li, Z., and Gildersleeve, J. C. (2001) High-throughput carbohydrate microarray profiling of 27 antibodies demonstrates widespread specificity problems. Glycobiology 17: 17C–23C. 34. Bochner, B., S., Alvarez, R., A., Mehta, P., Bovin, N., V., Blixt, O., White, J., R., and Schnaar, R., L. (2005) Glycan Array Screening Reveals a Candidate Ligand for Siglec-8. Journal of Biological Chemistry 280: 4307–4312. 35. Zhang, J., Pourceau, G., Meyer, A., Vidal, S., Praly, J. P., Souteyrand, E., Vasseur, J. J., Morvan, F., and Y., C. (2009) DNA-directed immobilisation of glycomimetics for glycoarrays application: Comparison with covalent immobilisation, and development of an on-chip IC(50) measurement assay. Biosensors and Bioelectronics 24: 2515–2521. 36. Chevolot, Y., Bouillon, C., Vidal, S., Morvan, F., Meyer, A., Cloarec, J.-P., Jochum, A., Praly, J.-P., Vasseur, J.-J., and Souteyrand, E. (2007) DNA-Based Carbohydrate Biochips: A Platform for Surface Glyco-Engineering. Angewandte Chemie, International Edition in English 46: 2398–2402. 37. Zhang, J., Pourceau, G., Meyer, A., Vidal, S., Praly, J. P., Souteyrand, E., Vasseur, J. J., Morvan, F., and Chevolot, Y. (2009) Specific recognition of lectins by oligonucleotide glycoconjugates and sorting on a DNA microarray. Chem. Commun. 28: 6795–6797. 38. Moni, L., Pourceau, G., Zhang, J., Meyer, A., Vidal, S., Souteyrand, E., Dondoni, A., Morvan, F., Chevolot, Y., Vasseur, J. J., and Marra, A. (2009) Design of Triazole-Tethered Glycoclusters Exhibiting Three Different Spatial Arrangements and Comparative Study of their Affinities towards PA-IL and RCA 120 by Using a DNA-Based Glycoarray. ChemBioChem 10: 1369–1378. 39. Niemeyer, C. M., Boldt, L., Ceyhan, B., and Blohm, D. (1999) DNA-Directed Immobilization: Efficient, Reversible, and SiteSelective Surface Binding of Proteins by Means of Covalent DNA-Streptavidin Conjugates. Analytical Chemistry 268: 54–63.

14  Glycoarray by DNA-Directed Immobilization 40. Wacker, R., Schröder, H., and Niemeyer, C. M. (2004) Performance of antibody microarrays fabricated by either DNA-directed immobilization, direct spotting, or streptavidin–biotin attachment: a comparative study. Analytical Biochemistry 330: 281–287. 41. Seela, F., and Kaiser, K. (1987) Oligodeoxy­ ribonucleotides Containing 1,3-Propanediol as Nucleoside Substitute. Nucleic Acids Research 15: 3113–3129. 42. Mazurczyk, R., El Khoury, G., Dugasb, V., Hannes, B., Laurenceau, E., Cabrera, M., Krawczyk, S., Souteyrand, E., Cloarec, J. P., and Chevolot, Y. (2008) Low-cost, fast prototyping method of fabrication of the microreactor devices in soda-lime glass. Sensors and Actuators, B: Chemical Sensors and Materials 128: 552–559. 43. Vieillard, J., Mazurczyk, R., Morin, C., Hannes, B., Chevolot, Y., DesbèneP-L., and Krawczyk, S. (2007) Application of microfluidic chip with integrated optics for electrophoretic separations of proteins. Journal of Chromatography B: Biomedical Applications 845: 218–225. 44. Morvan, F., Meyer, A., Jochum, A., Sabin, C., Chevolot, Y., Imberty, A., Praly, J.-P., Vasseur, J.-J., Souteyrand, E., and Vidal, S. (2007) Fucosylated Pentaerythrityl Phosphodiester Oligomers (PePOs): Automated Synthesis of DNA-Based Glycoclusters and Binding to

219

Pseudomonas aeruginosa Lectin (PA-IIL). Bioconjugate Chemistry 18: 1637–1643. 45. Dugas, V., Depret, G., Chevalier, Y., Nesme, X., and Souteyrand, E. (2004) Immobilization of single-stranded DNA fragments to solid surfaces and their repeatable specific hybridization:covalent binding or adsorption. Sensors and Actuators, B: Chemical Sensors and Materials 112–121. 46. Szurmai, Z., Szabo, L., and Liptak, A. (1989) Diethylene and Triethylene Glycol Spacers for the Preparation of Neoglycoproteins. Acta Chimica Hungarica-Models in Chemistry 126: 259–269. 47. Cecioni, S., Lalor, R., Blanchard, B., Praly, J. P., Imberty, A., Matthews, S. E., and Vidal, S. (2009) Achieving High Affinity towards a Bacterial Lectin through Multivalent Topological Isomers of Calix[4]arene Glycoconjugates. Chemistry-a European Journal 15: 13232–13240. 48. Salo, H., Virta, P., Hakala, H., Prakash, T. P., Kawasaki, A. M., Manoharan, M., and Lonnberg, H. (1999) Aminooxy functionalized oligonucleotides: Preparation, on-support derivatization, and postsynthetic attachment to polymer support. Bioconjugate Chemistry 10: 815–823. 49. Dugas, V., and Chevalier, Y. (2003) Surface hydroxylation and silane grafting on fumed and thermal silica. Journal of Colloid and Interface Science 264: 354–361.

Chapter 15 Fabrication of Carbohydrate Surfaces by Using Non-derivatised Oligosaccharides Jonathan Popplewell, Marcus Swann, Gavin Brown, and Bob Lauder Abstract Surface-based tools, such as microarrays and optical biosensors, are being increasingly applied to the analysis of carbohydrate–protein interactions. A key to these developments is the presentation of the carbohydrate to the protein target. Dual polarisation interferometry (DPI) is a surface-based technique that permits the real-time measurement of the changes in thickness, refractive index, and mass of adsorbates 100-nm thick or less on the surface of a functionalised waveguide. DPI has been used to design and characterise a surface on which the orientation and density of the immobilised carbohydrates are suitable for studying their interactions with proteins and where non-specific binding is reduced to less than 5% of total binding. A thiol-functionalised surface was derivatised with a heterobifunctional cross-linker to yield a hydrazide surface. This was treated with oligosaccharides, derived from keratan sulphate, chondroitin sulphate, and heparin that possess a reducing end. To block the unreacted hydrazide groups, the surface was treated with an aldehyde-functionalised PEG, and the surfaces were then challenged with a variety of proteins. Key  words: Biosensors, Glycosaminoglycans, Growth factors, Heparin, Lactoferrin – dual polarisation interferometry

1. Introduction Surface-based approaches for elucidation of carbohydrate functionality are attractive since they afford sensitivity and the potential for the high-throughput analysis of interactions. A key challenge in fabricating carbohydrate surfaces is the establishment of reliable and reproducible chemistries for the immobilisation of the oligosaccharides onto a solid substrate while retaining their functionality. A number of different strategies have been employed to immobilise carbohydrates on surfaces, including biotinylation of the sugars (1), formation of mercury sugar adducts (2), and

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_15, © Springer Science+Business Media, LLC 2012

221

222

J. Popplewell et al.

addition of maleimide functionality (3). We have developed an oligosaccharide-immobilisation strategy for orientated coupling of oligosaccharides. This strategy incorporates the direct immobilisation of oligosaccharides through their reducing ends onto a hydrazide-functionalised surface and the blocking of unreacted hydrazides with an aldehyde-functionalised poly(ethylene glycol).

2. Materials 2.1. Formation of CarbohydrateFunctionalised Surface

1. Phosphate-buffered saline (PBS) is prepared by dissolving one tablet in 200  ml of deionised water (see Note 1) to yield 0.01 M phosphate buffer, 0.0027 M potassium chloride, and 0.137  M sodium chloride, which is adjusted to pH 7.4 if required. Buffer is degassed by sonicating under vacuum for 20 min. PBS buffer is freshly prepared and degassed on the day of oligosaccharide surface formation. 2. 10  mM 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid, N-(2-Hydroxyethyl)piperazine-N¢-(2-ethanesulfonic acid) (HEPES), also containing 150 mM NaCl, 2 mM calcium, and magnesium and manganese ions is adjusted to pH 7.4 if required and degassed as for PBS. 3. Waveguide chips modified with an alkoxysilylthiol layer as supplied by Farfield Group Ltd (Manchester, UK). Calibrated with 80% (w/w) ethanol solution. 4. N-(b-maleimidopropionic acid)hydrazide, trifluoroacetic acid salt (BMPH) (Pierce Cramlington UK) is dissolved in running buffer (PBS or HEPES depending on which surface is to be manufactured at 5  mg/ml). Solution is prepared less than 5 min before perfusion over the surface. 5. Heparin (degree of polymerisation 4–20, dp 4–20) supplied as 10  mg/ml stock solution in water and kept frozen at −20°. Size-defined fractions of sodium heparin oligosaccharides are prepared by partial nitrous acid digestion of the parental polysaccharide followed by Superdex 30 chromatography. The size and expected composition of the selected oligosaccharide fragments are confirmed by comparison with authentic heparin standards on Superdex 30 and by PAGE, and by digestion with heparinases. For covalent attachment to the surface, stock heparin solution is allowed to thaw and diluted to 2 mg/ml with PBS adjusted to pH 5 (see Note 2). 2  mg/ml solutions are prepared less than 5  min before perfusion over the BMPHfunctionalised sensor surface. 6. Keratan sulphate (KS) and chondroitin sulphate (CS) oligosaccharides are extracted from bovine cartilage following papain

15  Fabrication of Carbohydrate Surfaces…

223

digestion, alkali treated and separated. Both KS and CS are supplied as lyophilised powder and dissolved to 2 mg/ml in HEPES buffer at pH 5, with solutions being made up less than  5  min before perfusion over the BMPH-functionalised surface. 7. The heterofunctional polyethylene glycol CH3O–PEG–NH– CO–C4H8–CHO (PEG–CHO) (Rapp Polymere GmbH, Tubingen Germany). The material is supplied as 1  g, and is aliquoted into 20-mg portions and frozen. Each 20-mg aliquot is dissolved in PBS (for the heparin surface) or HEPES buffer (for the KS or CS surfaces) just prior to perfusion over the sensor surface. 2.2. Protein Challenges to Heparin Surface

1. Bovine serum albumin (BSA) is dissolved in buffer at 1 mg/ml and stored at 4°C between experiments. 2. Bovine lactoferrin (Lf) is dissolved in buffer to form a stock solution of 0.1 mg/ml serially diluted to required concentrations with PBS or HEPES and stored at 4°C between experiments. 3. Human recombinant basic fibroblast growth factor (FGF-2) is stored frozen and diluted to required concentration on the day of use. 4. 2 M NaCl is prepared in water and degassed before use as a surface regenerant.

2.3. Protein Challenges to KS or CS Carbohydrate Surface

1. Wheat germ agglutinin (WGA) from Triticum vulgaris (MW = 36  kDa, main specificity N-acetyl-b-d-glucosamine) is  dissolved in HEPES buffer and stored at 4°C between experiments. 2. Maackia amurensis lectin (MAA, MW = 130  kDa, which specifically recognises sialic acid residues) is dissolved in HEPES buffer and stored at 4°C between experiments. 3. Lactoferrin from bovine colostrum (bLf) is dissolved in HEPES buffer and stored at 4°C between experiments. 4. Con A from Canavalia ensiformis (MW = 104 kDa, main specificity a-mannose) is dissolved in HEPES buffer and stored at 4°C between experiments. 5. HEPES running buffer is modified to contain 250 mM EDTA and adjusted to pH 8 for use as the surface regenerant.

2.4. Dual Polarisation Interferometry

Dual polarisation interferometry (DPI) permits the direct measurement of changes in thickness, refractive index (RI), and mass of materials 100-nm thick on the surface of a glass waveguide in real time. The dual-slab waveguide is illuminated with laser light of alternating polarisations at one end, and as the light exits the two waveguides at the other end, they interfere to produce an interference

224

J. Popplewell et al.

pattern. The waveguide structure is integrated with a fluidic system permitting the continuous flow of material over the top waveguide, and as material is added to or removed from the waveguide, the interference pattern moves, and these changes in the position of the interference pattern can be “resolved” into changes in thickness, RI, and mass of the material on the waveguide (4). Thus, if a protein is added to a heparin surface, the change in measured thickness will reflect the orientation the protein layer adopts on the surface. In addition, there is a surface coverage value (in ng/mm2) and a refractive index measurement that reflect the density of protein packing on the oligosaccharide surface.

3. Methods 3.1. Calibration of Thiol Waveguide

1. All DPI measurements are performed on an AnaLight® Bio200 equipped with a 632.8-nm laser. The instrument is a dualchannel system, with each 2-ml cell maintained at 20°C ± 0.002°C unless otherwise stated. 2. A new thiol-functionalised waveguide is inserted into the instrument, and freshly prepared, freshly degassed buffer ­perfused over the sensor surface until a stable baseline is obtained. 3. Calibration. Once the baseline sensor response is stable, a calibration of the upper waveguide surface takes place. This procedure involves sequentially injecting two different materials of known refractive index and measuring the phase change in radians upon transition from material one to material two. Material one is ethanol/water (80%, w/w) and is allowed to flow over the sensor for 90 s, followed by the running buffer. The baseline in PBS is retained. See Note 3. The second injection is water and is injected for 90 s; thus, we can determine the refractive index of the buffer solution (given that the refractive index of 80% ethanol and pure water is known). The instrument response can be used to calculate the thickness and refractive index of the waveguide and the refractive index of the buffer solution (bulk refractive index) to be used in subsequent layer calculations.

3.2. Oligosaccharide Surface Formation

In the following section, the formation of two different groups of surfaces is described. Subheading 3.2.1 describes the formation of keratan sulphate and chondroitin sulphate surfaces, which are challenged with lectins. The KS and CS interactions with their respective lectins require divalent metal ions; thus, the surface formation and running buffer for the CS/KS surfaces are performed with HEPES buffer and the metal ions as detailed in Subheading 2.1.

15  Fabrication of Carbohydrate Surfaces…

225

The surfaces formed with heparin from tetrasaccharide up to dp-20 and their subsequent interactions with proteins described here do not require divalent metal ions to be present; thus, the running buffer deployed is PBS. 3.2.1. In Situ Formation of the Oligosaccharide Surfaces Keratan Sulphate and Chondroitin Sulphate

1. For formation and use of the KS or CS surfaces, the running buffer used is HEPES (10 mM) containing 150 mM NaCl and the metal cations Mg, Mn, and Ca (all at 2 mM). 2. Following calibration steps as detailed in Subheading  3.1,­ step 1, the flow cell temperature is then set to 30 C (see Note 4) and the flow rate is reduced to 10 ml/min. 3. Upon thermal equilibration, BMPH (5  mg/ml dissolved in buffer just prior to injection) is perfused over the thiol sensor surface for 15 min. 4. HEPES buffer is then used to rinse the sensor for a further 10 min at a flow rate of 50 ml/m (see Note 5). 5. The oligosaccharide of choice is dissolved in running buffer (2 mg/ml) adjusted to pH 5 and perfused over only one of the two flow cells at a rate of 2 ml/min for 100 min. 6. Oligosaccharides that were not bound after 100 min are removed by washing the surface with running buffer at 50  ml/min for 10 min.

3.2.2. In Situ Formation of Oligosaccharide Heparin Surfaces

1. For formation and use of the heparin surfaces, the running buffer used is PBS. 2. Following calibration steps as detailed in Subheading  3.1, step 1, the flow cell temperature is then set to 30°C (see Note 4) and the flow rate is reduced to 10 ml/min. 3. Upon thermal equilibration, BMPH (5  mg/ml dissolved in buffer just prior to injection) is perfused over the thiol sensor surface for 15 min. 4. Running buffer is used to rinse the sensor for a further 10 min at 50 ml/m. 5. Next, the polymer of heparin is diluted from its stock 10 mg/ml solution in water and dissolved in PBS adjusted to pH 5 and injected into only one of the two flow cells at a rate of 2 ml/min for 100 min (Figs. 1 and 2). 6. Oligosaccharides that are not bound after 100 min are removed by washing the surface with PBS at 50 ml/min for 10 min.

3.3. Blocking the Unreacted Hydrazide Groups and Creation of a Control Flow Cell

Each of the glycochips described here is composed of two parts, one flow cell containing the BMPH linker, an oligosaccharide, and the PEG blocker, while the control flow cell is identical except for the absence of oligosaccharide, thereby creating side-by-side active (carbohydrate containing) and control surfaces.

226

J. Popplewell et al.

8

0.8

7

0.7

6

0.6

5

0.5

4

0.4

3

0.3

2

0.2

Thickness Surface Coverage

1 0

Density / gcm-3

Surface Coverage/ng mm-2 & Thickness/nm

Fig. 1. Complete coupling reactions for BMPH with a reducing sugar. Reproduced with permission from Wiley-VCH Verlag GmbH & Co. KGaA (5).

0.1

Density

0

10

20

30

40

50

60 70 80 Time / minute

90

100 110 120 130

0

Fig. 2. Changes in surface coverage and dimensions of BMPH and heparin as the oligosaccharide surface is formed. The BMPH linker was added at approximately t = 3 min, and following PBS washing the heparin is added at 21 min, and at 70 min the surface is washed with PBS. Reproduced with permission from Wiley-VCH Verlag GmbH & Co. KGaA (5).

15  Fabrication of Carbohydrate Surfaces… 0

7

-0.05

-3

5 -0.1 4 -0.15 3 PEG binding to BMPH & DP10 surface

-0.2

PEG binding to BMPH surface

2

Change in density /g cm

Change in mass /ng mm

-2

6

Density change to BMPH & DP10 surface Density change to BMPH surface

1

0 130

227

140

150

160

170

180

190

200

210

-0.25

220

230

-0.3 240

Time /min

Fig. 3. Changes in surface coverage on PEG–CHO binding. The increase in surface coverage was determined for the non-heparin-treated surface (black line) and compared to that of the BMPH surface already treated with DP-10 (grey line). The dense linker layer undergoes a greater decrease in density (right-hand axis) upon binding CHO–PEG compared to the less-dense BMPH–carbohydrate surface. Reproduced with permission from Wiley-VCH Verlag GmbH & Co. KGaA (5).

1. Unreacted hydrazide groups are blocked by the addition of  CH3O–(EG)17–NH–CO–C4H8–CHO or “PEG–CHO” (20 mg/ml) dissolved in PBS at pH 7 (see Note 6). 2. The PEG–CHO is perfused over both flow cells at a rate of 2 ml/min for a total of 100 min. 3. PEG–CHO that was not bound after 100 min is removed by washing the surface with running buffer at 50  ml/min for 10 min. 4. PEG–CHO is perfused over both flow cells at a rate of 2 ml/min a second time to ensure complete blocking of the hydrazide surfaces. 5. Again, PEG–CHO that is not bound after 100 min is removed by washing the surface with running buffer at 50 ml/min for 10 min (Fig. 3). 6. For protein challenges described in Subheading 3.4, the flow cell temperature is decreased to 20°C; see Note 4. 3.4. Protein Challenges 3.4.1. Protein Challenges to KS or CS Surface

All the proteins injected over the KS chip are dissolved in HEPES/ NaCl buffer with divalent metal ions (2 mM). Lectins, such as Con A, WGA, and MAA, bind carbohydrates through a network of hydrogen bonds, hydrophobic interactions, van der Waals interactions, and metal ion co-ordinations. Metal ions, such as Mg2+ and Ca2+, can assist in the positioning of the amino acid residues to interact with the carbohydrates, but with few conformational

228

J. Popplewell et al.

changes taking place upon binding in either protein or carbohydrate. Thus, the divalent metal ions allow the protein to adopt the required conformation to bind to carbohydrates and are not known to alter the conformation of the carbohydrates. 1. In all cases, 150 ml solutions of different proteins are injected over the surfaces at a flow rate of 50 ml/min and then allowed to rinse for 10 min. For the KS surfaces, the following proteins are tested: (a) WGA from Triticum vulgaris (MW = 36 kDa, main specificity N-acetyl-b-D glucosamine) (b) MAA (MW = 130 kDa, which specifically recognises sialic acid residues) (c) Lactoferrin from bovine colostrum which binds to any oligosaccharide 2. In addition, the following control (non-binding) proteins are added: (a) Con A from Canavalia ensiformis (MW = 104  kDa, main specificity a-mannose) (b) BSA 3.4.2. Protein Challenges to Heparin Surfaces

FGF2 (nM affinity) and bLf (nM affinity) are perfused over the heparin surfaces to attain kinetic and equilibrium binding constants.

4. Notes 1. Unless stated otherwise, all solutions should be prepared in water that has a resistivity of 18.2 MW cm and this standard is referred to as “water” in this text. 2. Initial experiments used acetate buffer at pH 5, but significant upward drift was observed with acetate as the running buffer, presumably as acetate deposited on the alkyl thiol surface. While adjusting PBS to pH 5 is outside the buffering range of PBS, no problems with this were observed. 3. In order for the chip calibration to be accurate, the baseline must return to the same position as before the ethanol injection. It is not uncommon for the baseline after the ethanol injection to be a few percent lower than before. This is usually the removal of trace surface contamination and is usually achieved with one 50-ml injection. The transition from buffer to 80% EtOH is normally about 50 radians, and it should be possible for the baseline to return to 0.1 radians of the preethanol injection after two ethanol injections. If after more

15  Fabrication of Carbohydrate Surfaces…

229

than two injections the baseline continues to decrease, it is likely some system contamination is present and a thorough cleaning of the instrument is recommended. 4. It is recommended that the instrument temperature to be changed in a ramped procedure. In other words, adjust the temperature from 20 to 30°C in steps of 2°C to allow thermal expansion of the flow cell to be achieved in a controlled fashion. 5. Loop washing: In order to prevent residual BMPH in the sample loading loops reacting with the oligosaccharides as they are loaded onto the sample loops, it is recommended that users wash the 200-ml loops with a total of 4 ml of running buffer before the oligosaccharide is loaded onto the sample loops for injection onto the thiol sensor surface. 6. Initial tests with PEG–CHO were done with PEG–CHO dissolved in PBS adjusted to pH 5, as for the oligosaccharide, but we observed that the dissociation of PEG–CHO from the surface was significantly faster if added at pH 5; thus, we used pH 7 buffer for this step.

Acknowledgements The authors thank the Biotechnology and Biological Sciences Research Council, the Human Frontiers Science Programme, the RCUK “Glycochips” programme, and the North West Cancer Research Fund for financial support. References 1. M. Delehedde, M. Lyon, J. T. Gallagher, P. S. Rudland, D. G. Fernig, Biochem.J. 2002, 366, 235–244. Fibroblast growth factor-2 binds to small heparin-derived oligosaccharides and stimulates a sustained phosphorylation of p42/44 mitogen-activated protein kinase and proliferation of rat mammary fibroblasts 2. M. A. Skidmore, S. J. Patey, N. T. Thanh, D. G. Fernig, J. E. Turnbull, E. A.Yates, Chem. Commun. 2004, 2700–2701. Attachment of glycosaminoglycan oligosaccharides to thiolderivatised gold surfaces 3. S. Park, I. Shin, Angew. Chem. 2002, 114, 3312–3314; Angew. Chem. Int.Ed. 2002, 41,

3180–3182. Fabrication of Carbohydrate Chips for Studying Protein–Carbohydrate Interactions 4. G. H. Cross, A. A. Reeves, S. Brand, J. F. Popplewell, L. L. Peel, M. J.Swann, N. J. Freeman, Biosens. Bioelectron. 2003, 19, 383–390. A new quantitative optical biosensor for protein characterisation 5. J Popplewell, M Swann, Y Ahmed, D. Fernig, J. Turnbull, ChemBioChem 2009, 10, 12181226. Fabrication of Carbohydrate Surfaces by Using Nonderivatised Oligosaccharides, and their Application to Measuring the Assembly of Sugar–Protein Complexes

Chapter 16 Polysaccharide Microarrays: Application to the Identification of Heparan Sulphate Mimetics Julien Dheur, Nabil Dendane, Rémi Desmet, Fatima Dahmani, Gauthier Goormachtigh, Jérome Vicogne, Véronique Fafeur, and Oleg Melnyk Abstract The interaction of polysaccharides with proteins modulates or triggers many biological effects. In particular, heparan sulphate proteoglycans (HSPGs) have multiple regulatory interactions with growth factors, enzymes, enzyme inhibitors, and some components of the extracellular matrix. The important role played by HSPGs has motivated the synthesis and selection of HSPG mimetics for modulating the biological activity of HS-binding proteins. We present hereinafter an efficient polysaccharide microarray method that allows the screening of HS-mimetic libraries towards HS-binding growth factors, a major class of polypeptides whose inhibition or potentiation is of high medical interest. Key words: Polysaccharide microarrays, Heparan sulphate, Growth factor, Fluorescence

1. Introduction The interaction of polysaccharides with proteins modulates or triggers many biological processes. As a consequence, the design of novel tools for studying polysaccharide–protein interactions is an essential field of research for deciphering biological mechanisms or identifying molecules of therapeutical interest (1). In particular, heparan sulphate proteoglycans (HSPGs) are often used by pathogens, such as HIV (2), Listeria monocytogenes (3), or Mycobacterium tuberculosis (4) for interacting and infecting host cells. HSPGs regulate also the biological activity of growth factors, enzymes, enzyme inhibitors, and some components of the extracellular matrix. The important role played by HSPGs has motivated the synthesis and

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_16, © Springer Science+Business Media, LLC 2012

231

232

J. Dheur et al.

selection of HSPG mimetics for potentiating or inhibiting the biological activity of HS-binding proteins (5). For example, HIV entry in vitro can be inhibited by dextran sulphate (6). HS mimetics are also studied for their capacity to modulate the biological activity of growth factors. For example, potentiation of plateletderived growth factor-BB (PDGF-BB) biological activity by synthetic sulphated polysaccharides is regarded as a promising strategy for improving wound-healing processes (7). In another application, polysaccharide potentiation of human bone morphogenetic protein-2 (BMP-2), a heparin-binding osteoinductive growth factor, has been shown to stimulate bone repair (8). Alternately, inhibition to modulate the biological activity of growth factors HS mimetics is explored for inhibiting tumour growth and angiogenesis (9). We describe hereinafter an efficient polysaccharide microarray method that allows the screening of HS-mimetic libraries towards HS-binding growth factors, such as human hepatocyte growth factor (hHGF). Polysaccharides were microarrayed on semicarbazide microscope glass slides and incubated with hHGF. Detection was performed using tetramethylrhodamine-labelled proteins and a standard microarray fluorescence scanner.

2. Materials 2.1. SemicarbazideFunctionalised Glass Slides Preparation

1. The experiment requires an efficient fume hood (see Note 1). 2. Microscope glass slides (Gold Seal Rod.90° 25.4 × 76.2 ×  1.0 mm). 3. Glass staining dishes, a Teflon® staining dish, and a Wheaton glass rack for 20 slides (75 × 25 mm). 4. A Teflon® pot (1 L, ref 213–0242, VWR). 5. Glass desiccator and vacuum pump. 6. Oven. 7. Sulphuric acid 96% solution in water. 8. 35% hydrogen peroxide (H2O2) solution in water. 9. Deionised water (see Note 2). 10. Methanol (MeOH, Carlo Erba). 11. 3-Aminopropyltrimethoxysilane. 12. Triphosgene (highly toxic, Lancaster). 13. N,N,N-diisopropylethylamine. 14. 1,2-Dichloroethane (Janssen). 15. 9-Fluorenylmethoxycarbonyl-NHNH2 (Fmoc-NHNH2, see Note 3). 16. N,N-dimethylformamide (DMF, Carlo Erba).

16  Polysaccharide Microarrays: Application to the Identification…

233

17. Ethanol (EtOH, Carlo Erba). 18. Piperidine (ACROS Organics). 19. 1,8-Diazabicyclo[5,4,0]undec-7-ene (DBU, Sigma–Aldrich). 20. Goniometer (GBX, Romans-sur-Isère, France). 2.2. Printing

1. Semicarbazide-functionalised glass slides. 2. HS-mimetic polysaccharides. 3. Heparin (15 kDa, Sigma–Aldrich). 4. Phosphate-buffered saline (PBS): Prepare 0.01  M solution (0.138 M NaCl, 0.0027 M KCl, pH 7.4) by dissolving 1 foil pouch (ref. P3813, Sigma–Aldrich) in 1 L of deionised water. 5. V-bottomed 384-Well Microarray Plate, cylindrical wells (ABGene, Surrey, UK, AB-1055). 6. BCA1 Microarrayer (Perkin Elmer, MA, USA) (see Note 4).

2.3. Incubations

1. PBS prepared as above (b4). 2. Bovine Serum Albumin Fraction V (BSA, Roche). 3. Tween® 20 (Sigma–Aldrich). 4. Absolute ethanol (Carlo Erba). 5. Recombinant hHGF (ref. 100–39, Peprotech Inc., NJ, USA) (see Note 5). 6. Biotinylated anti-human HGF antibody (ref. BAF294, affinity purified, R&D Systems, Inc., MN, USA). 7. Tetramethylrhodamine-labelled streptavidin (ref. 016-020084, Jackson ImmunoResearch Europe Ltd., UK). 8. Microplate microarray hardware 96 wells (ref. MMH4 × 24, Arrayit Corp. CA, USA) equipped with a silicone gasket (ref. GMMH4X24, Arrayit Corp. CA, USA). 9. Standard incubator/agitator for 96-well microtiter plates (ref. PST-60HL-4, BioSan, MI, USA). 10. Standard 96-well microtiter plate washing station (wellwash AC, Thermo Electron, Finland). 11. Fluorescence microarray laser scanner (LS Reloaded, Tecan Group Ltd., Switzerland) and image analysis software (ArrayPro® Analyzer, Switzerland).

3. Methods Methods allowing the identification of compounds modulating the biological activity of HS-binding growth factor are of great therapeutical interest. Microarrays have attracted much attention during

234

J. Dheur et al.

the last decade as useful tools for screening libraries of biological or synthetic polysaccharides (10–12). The preparation of polysaccharide microarrays is faced with the difficulty to immobilise polysaccharides while maintaining their binding properties. We have reported that semicarbazide slides are useful substrates for preparing polysaccharide microarrays and for selecting PDGF-BB binders (13). Interestingly, the best binders were found to potentiate PDGF-BB-induced proliferation of human dermal fibroblasts. The microarray method used is illustrated hereinafter with hHGF and a collection of synthetic HS mimetics. hHGF is a heparin-binding polypeptide of approximately 90 kDa that stimulates cell proliferation, survival, motility, and morphogenesis through activation of its receptor, the c-Met tyrosine kinase (14, 15). Aberrant signalling of this complex is a feature of many tumours and appears to contribute to the growth, invasiveness, and metastasis of both carcinomas and sarcomas. Previous studies with PDGF-BB showed that amine slides, a substrate often used for DNA or protein microarray printing, display high signal-to-noise ratios for neutral polysaccharides but poor results for anionic polysaccharides (13). Ammonium groups present on amine slides interact presumably with anionic groups (sulphate, carboxylate) within the polysaccharides, thereby limiting their access to target proteins. The superiority of semicarbazide over amine slides in HS-mimetic binding studies might be due to the low pKa of semicarbazide group (pKa 3.6), which makes the surface of semicarbazide slides uncharged at pH 7. In another study, the stability of protein microarrays prepared on both substrates was compared. Microarrays prepared on semicarbazide slides were stable for months in an antibody-binding assay, whereas the signal strength for microarrays prepared on amine slides varied significantly in less than 1 month of ageing (16, 17). These data highlight, if necessary, the importance of surface chemistry for microarray preparation. We present first the preparation of semicarbazide slides, the substrate for microarray printing. Our strategy to functionalise microscope glass slides is a multistep procedure that involves first a silanization step with 3-aminopropyltrimethoxysilane to form an amine layer. Then, amino groups are converted into isocyanate groups with triphosgene. Finally, reaction of isocyanate groups with an Fmoc-protected hydrazine derivative and removal of Fmoc group in basic medium furnishes semicarbazide slides ready for polysaccharide printing. The data presented in Fig. 1 correspond to a microarray prepared with a synthetic 40–60-kDa, HS-mimetic, polysaccharide library (compounds labelled P1 to P23) (see Note 6). Microarrays were incubated with hHGF. Detection of the captured growth factor was performed using biotinylated anti-hHGF antibody followed by tetramethylrhodamine-labelled streptavidin incubation. Heparin was used as positive control, whereas polysaccharide P23 was used as a negative control. This experiment identified

16  Polysaccharide Microarrays: Application to the Identification…

235

5000 I / A.U.

4500

Rho

4000

Streptavidin Biotin

3500

Anti-hHGF -

3000 hHGF

2500 Polysaccharide

2000

Semicarbazide glass slide

1500 1000 500

4 P1 5 P1 6 P1 7 P1 8 P1 9 P2 0 P2 1 P2 2 P2 3

P1

2

1

3 P1

P1

0

P1

P1

P8

P9

P7

P5

P6

P4

P3

P2

H

ep

ar

in P1

0

Fig. 1. Quantification of the hHGF-binding experiment. Polysaccharide solutions (1 mg/mL in PBS) were printed using a non-contact piezoelectric microarrayer (three drops, 300 pL each, diameter 200 mm). Compounds were printed in triplicate with a distance of 500 mm between centres of adjacent spots. hHGF was incubated at a concentration of 2.5 mg/mL for 2 h at room temperature. Detection of captured hHGF was performed using a biotinylated anti-hHGF antibody (1 mg/mL) and tetramethylrhodamine-labelled streptavidin (1 mg/mL). Incubations were performed in triplicate. Detection was performed at 532 nm. The data correspond to the median and interquartile range of the fluorescence values expressed in arbitrary units (A.U.) from which the background of the slide was subtracted (mean background = 163.3 ± 18.6 A.U.).

polysaccharide P17 as the best binder within the polysaccharide library. Furthermore, P17 but not P23 was found to inhibit the binding of hHGF to MET-extracellular domain using an Alphascreen® in-house assay (see Note 7) (18). Finally, P17 inhibited hHGF-induced MET phosphorylation in a cell-based assay (data not shown), confirming the interest of the microarray method described here for the selection of polysaccharides of biological interest. Furthermore, this microarray method can be applied to the analysis of multiple growth factors at varying concentrations. Since they generate a large amount of data, we have developed a Web-based platform, called PASE, which provides an integrated framework to manage and to statistically analyse the data (19). 3.1. SemicarbazideFunctionalised Glass Slides Preparation

1. Prepare the microscope slides by using glass rack slides. The slides are placed back to back so that one rack can contain up to 40 slides. Place the slide rack in the Teflon® staining dish equipped with a magnetic bar. 2. Preparation of piranha solution: In the Teflon® pot, add carefully within 1 min 250 mL of concentrated sulphuric acid to 250 mL of hydrogen peroxide under stirring. The reaction is highly exothermic.

236

J. Dheur et al.

3. Once prepared, clean and activate the microscope slides by immersing the substrates overnight in the piranha solution under stirring. Discard piranha solution in a dedicated waste container and wash with water (3 × 3  min) and methanol (3 × 3 min) (see Note 8). 4. Introduce the amine functionalities by treatment with 15 mL of 3-aminopropyltrimethoxysilane, 460 mL of methanol, and 24.5 mL of deionised water. Place the slide rack in a staining dish containing this solution and place it for 30  min in a sonicator. 5. Wash the surfaces with methanol, water (twice), and methanol and then place the slide rack for 15 min in an oven at 110°C (see Note 9). 6. The treatment of the slides with 500 mL of a 0.1 M triphosgene/0.8 M diisopropylethylamine solution in 1,2-dichloroethane for 2  h under sonication leads to the formation of isocyanate groups on the glass surface. 7. Isocyanate groups are reacted immediately with Fmoc-NHNH2 by immersing the slides in 500 mL of a 22 mM Fmoc-NHNH2 solution in DMF containing 1% of ethanol under sonication for 2 h (see Note 10). 8. Deprotection of the semicarbazide groups is performed with 500 mL of a 0.2% piperidine and 2% DBU solution in DMF for 30  min. Finally, the slides are washed with DMF, water (twice), and methanol and dried in desiccator under reduced pressure overnight. 9. Characterise the slide surface by contact angle measurements (see Note 11). 10. Store the slides at room temperature in a dust-proof container (see Note 12). 3.2. Printing

1. Prepare probe solutions by dissolving compounds P1-23 and heparin at 1 mg/mL in PBS and place 25 mL of each solution in the V-bottomed, 384-well, microtitre plate. 2. Print the mixtures (three drops, 300 pL each) in triplicate on semicarbazide glass slides (see Note 13). 3. Until use, printed slides are stored at 4°C.

3.3. Incubations

1. Prepare PBS-A buffer by dissolving 0.05% Tween®20 in PBS. 2. Prepare PBS-B buffer by dissolving 0.1% w/v BSA in PBS. 3. Insert the printed glass slides in the microplate microarray hardware 96 wells. 4. Use three microarrays per experiment. The microarrays are first saturated for 5 min under shaking using the incubator/ agitator at room temperature in PBS-B (100 mL per well).

16  Polysaccharide Microarrays: Application to the Identification…

237

5. Dissolve hHGF in PBS-B at a final concentration of 1 mg/mL. 6. Wash the microarrays three times with PBS-A using the 96-well microtitre plate washing station and incubate immediately with hHGF solution (50  mL per well, 2  h at room temperature under shaking) (see Note 14). 7. In the meantime, dissolve biotinylated anti-hHGF antibody in PBS-B (final concentration 1 mg/mL). 8. The microarrays are washed three times with PBS-A using the washing station and incubated immediately with the biotinylated anti-hHGF antibody solution at 1  mg/mL (50  mL per well, 2 h at room temperature under shaking). 9. During incubation with the biotinylated anti-human HGF antibody, dilute the stock solution of tetramethylrhodamine– streptavidin in PBS-B (final concentration 1 mg/mL). 10. The microarrays are washed three times with PBS-A and incubated immediately with the tetramethylrhodamine–streptavidin solution (50  mL per well, 2  h at room temperature under shaking). 11. Wash three times with PBS-A and three times with water. 12. Remove the slides from the microplate hardware and wash the slides with absolute ethanol. Dry the slides by flushing with nitrogen. 13. Analyse with the microarray scanner at 532 nm.

4. Notes 1. The piranha solution is a strong oxidant. It reacts violently with organic materials. It can cause severe skin burns. It must be handled with extreme care in a well-ventilated fume hood while wearing appropriate chemical safety protection. 2. Deionised water has a resistivity of 18 MW cm. 3. Fmoc-NHNH2 can be purchased from various suppliers or easily prepared from Fmoc-chloroformate and hydrazine hydrate as described in ref. 20. 4. The experiment described here was performed using a noncontact piezoelectric microarrayer (BCA1, Perkin Elmer). Typically, three drops (300  pL each) were printed per spot. The quill pins printing technology has been tested as well using a Qarray2 contact microarrayer (Genetix, New Milton, UK). However, this printing method led to heterogeneous spot morphologies and poor binding results with growth factors. 5. This recombinant protein corresponds to an immature form of hHGF, i.e. to a single-chain unprocessed polypeptide chain.

238

J. Dheur et al.

Mature HGF is a heterodimeric protein composed of a and b chains linked by a disulphide bond. The high-affinity HGF binding site for MET is present in the pro and mature HGF forms. Conversion of pro-HGF into mature HGF can be done by incubating with diluted foetal calf serum. 6. This polysaccharide microarray technology is compatible with a large range of polysaccharide chain lengths. For example, we could immobilise efficiently 15–100-kDa dextran molecules on semicarbazide slides. However, an increase in the chain length resulted in an increase in the signal strength (13). In the work presented here, the molecular weights of P1-23 are similar. 7. Alphascreen® technology (Amplified Light Proximity Homo­ genous Assay, Perkin Elmer©) allows the quantitative measurement of molecular interaction between protein and protein, protein and peptide, or protein and polysaccharide. When Alphascreen beads are in close proximity, a 680-nm laser activates donor beads that produce oxygen singlets. This shortlasting oxygen singlet activates acceptor beads on a 200-nm range that readily emit 520–620-nm photons measured by a dedicated plate reader. Beads recognition with targets required specific antibodies and/or appropriate labelling (biotinylation, polyhistidine cluster, Fc fragment, …). Many conjugated beads (proteine A, steptavidin, NiNTA, …) are commercially available and allow a large number of combinations. In our in-house assays performed into 384-well plates, we routinely use donor steptavidin beads (anti-HGF biotinylated mouse antibodies binders) and acceptor ProtA beads (Fc–MET binders) to measure hHGF–MET interaction. This assay was used to determine polysaccharides inhibiting HGF–MET interaction. 8. The piranha solution must not be in contact with organic solvents or other aqueous solutions (see also Note 1). A dedicated waste container must be used. 9. This thermal treatment allows the formation of additional cross links within the amine layer and between the amine layer and the silicon oxide substrate. 10. The presence of ethanol during the reaction of Fmoc-NHNH2 with isocyanate surface groups is crucial. Ethanol acts presumably as a catalyst by forming hydrogen bonds with isocyanate groups, thereby amplifying the electrophilicity of its carbonyl. The catalytic effect of alcohol on urethane formation on surfaces has been discussed by Kulik et al. who showed the catalytic effect of methanol vapour on the chemisorption of butyl isocyanate with surface-exposed hydroxyl groups (21). A more detailed discussion of this catalytic effect in solution can be found in ref. 22.

16  Polysaccharide Microarrays: Application to the Identification…

239

11. Contact angle measurements were performed at 20°C with water, diiodomethane, and formamide as reference liquids (at least nine drops of 1 mL for each liquid). Typical contact angles are 36.3° ± 0.5 for water, 30.9° ± 0.2 for diiodomethane, and 15.6° ± 0.9 for formamide; see ref. 16. Measurements were performed 10 s after deposition. 12. Semicarbazide groups are chemically stable for months using this storage conditions. Nevertheless, we observed an increase of contact angle values over time, indicating a progressive pollution of the glass slide surface by organic pollutants present in air. Consequently, we recommend using the slides rapidly (within 2–3 weeks) after preparation. 13. Twenty-four patterns can be printed per slides. The position of these patterns must be carefully controlled to allow the correct positioning of the patterns in the microplate microarray hardware 96-well. 14. It is of prime importance to avoid the drying of the microarrays during the incubations; otherwise, high background levels are obtained. For the automated washing, standard washing protocols are usually satisfactory.

Acknowledgements We gratefully acknowledge financial support from CNRS, Université de Lille Nord de France, Institut Pasteur de Lille, IFR 142, Région Nord Pas de Calais, the European Community (FEDER), and Cancéropôle Nord-Ouest. This research was performed using the Chemistry Systems Biology platform (http://csb.ibl.fr). References 1. Oppenheimer, S. B., Alvarez, M., and Nnoli, J. (2008) Carbohydrate-based experimental therapeutics for cancer, HIV/AIDS and other diseases. Acta Histochem. 110, 6–13. 2. Ugolini, S., Mondor, I., and Sattentau, Q. J. (1999) HIV-1 attachment: another look. Trends Microbiol. 7, 144–9. 3. Pizarro-Cerda, J., Sousa, S., and Cossart, P. (2004) Exploitation of host cell cytoskeleton and signalling during Listeria monocytogenes entry into mammalian cells. C. R. Biol. 327, 523–31. 4. Menozzi, F. D., Bischoff, R., Fort, E., Brennan, M. J., and Locht, C. (1998) Molecular characterization of the mycobacterial heparin-binding hemagglutinin, a mycobacterial adhesin. Proc. Natl. Acad. Sci. U.S.A 95, 12625–30.

5. Kovensky, J. (2009) Sulfated oligosaccharides: new targets for drug development? Curr. Med. Chem. 16, 2338–44. 6. Baba, M., Pauwels, R., Balzarini, J., Arnout, J., Desmyter, J., and De Clercq, E. (1988) Mechanism of inhibitory effect of dextran sulfate and heparin on replication of human immunodeficiency virus in  vitro. Proc. Natl. Acad. Sci. U S A 85, 6132–6. 7. Vercoutter-Edouart, A. S., Dubreucq, G., Vanhoecke, B., Rigaut, C., Renaux, F., DahriCorreia, L., Lemoine, J., Bracke, M., Michalski, J. C., and Correia, J. (2008) Enhancement of PDGF-BB mitogenic activity on human dermal fibroblasts by biospecific dextran derivatives. Biomaterials 29, 2280–92.

240

J. Dheur et al.

8. Degat, M. C., Dubreucq, G., Meunier, A., Dahri-Correia, L., Sedel, L., Petite, H., and Logeart-Avramoglou, D. (2009) Enhancement of the biological activity of BMP-2 by synthetic dextran derivatives. J. Biomed. Mater. Res. A 88, 174–83. 9. Bagheri-Yarmand, R., Kourbali, Y., Rath, A. M., Vassy, R., Martin, A., Jozefonvicz, J., Soria, C., Lu, H., and Crepin, M. (1999) Carboxymethyl benzylamide dextran blocks angiogenesis of MDA-MB435 breast carcinoma xenografted in fat pad and its lung metastases in nude mice. Cancer Res. 59, 507–10. 10. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–81. 11. Wang, D. (2003) Carbohydrate microarrays. Proteomics 3, 2167–75. 12. de Paz, J. L., Moseman, E. A., Noti, C., Polito, L., von Andrian, U. H., and Seeberger, P. H. (2007) Profiling heparin-chemokine interactions using synthetic tools. ACS Chem. Biol. 2, 735–44. 13. Carion, O., Lefebvre, J., Dubreucq, G., DahriCorreia, L., Correia, J., and Melnyk, O. (2006) Polysaccharide microarrays for polysaccharideplatelet-derived-growth-factor interaction studies. Chembiochem. 7, 817–26. 14. Accornero, P., Pavone, L. M., and Baratta, M. (2010) The Scatter Factor Signaling Pathways as Therapeutic Associated Target in Cancer Therapy. Curr. Med. Chem. 17, 2699–712. 15. Canadas, I., Rojo, F., Arumi-Uria, M., Rovira, A., Albanell, J., and Arriola, E. (2010) C-MET as a new therapeutic target for the development

of novel anticancer drugs. Clin. Transl. Oncol. 12, 253–60. 16. Duburcq, X., Olivier, C., Desmet, R., Halasa, M., Carion, O., Grandidier, B., Heim, T., Stievenard, D., Auriault, C., and Melnyk, O. (2004) Polypeptide semicarbazide glass slide microarrays: characterization and comparison with amine slides in serodetection studies. Bioconjugate Chem. 15, 317–25. 17. Duburcq, X., Olivier, C., Malingue, F., Desmet, R., Bouzidi, A., Zhou, F., Auriault, C., GrasMasse, H., and Melnyk, O. (2004) Peptideprotein microarrays for the simultaneous detection of pathogen infections. Bioconjugate Chem. 15, 307–16. 18. Eglen, R. M., Reisine, T., Roby, P., Rouleau, N., Illy, C., Bosse, R., and Bielefeld, M. (2008) The use of AlphaScreen technology in HTS: current status. Curr. Chem. Genomics 1, 2–10. 19. Pamelard, F., Even, G., Apostol, C., Preda, C., Dhaenens, C., Fafeur, V., Desmet, R., and Melnyk, O. (2009) PASE: A Web-Based Platform for Peptide/Protein Microarray Experiments. Methods Mol. Biol. 570, 413–30. 20. Zhang, R. E., Cao, Y. L., and Hearn, M. W. (1991) Synthesis and application of Fmochydrazine for the quantitative determination of saccharides by reversed-phase high-performance liquid chromatography in the low and subpicomole range. Anal Biochem. 195, 160–7. 21. Kulik, N. V., Negievich, L. A., Kurgan, N. P., Belitskaya, G. F., and Kachan, A. A. (1972) Mechanism of the combined reaction of butyl isocyanate and methanol with aerosil. Theor. Exp. Chem. 6, 48–52. 22. Arnold, R. G., Nelson, J. A., and Verbanc, J. J. (1957) Recent Advances In Isocyanate Chemistry. Chem. Rev. 57, 47–76.

Chapter 17 Carbohydrate Antigen Microarrays Denong Wang Abstract This chapter describes one of my laboratory’s working protocols for carbohydrate-based microarrays. Using a standard microarray spotter, we print carbohydrate antigens directly on the nitrocellulose-coated bioarray substrates. Because these substrates support noncovalent immobilization of many spotted antigens, in general no chemical modification of the antigen is needed for microarray production. Thus, this bioarray platform is technically simple and applicable for high-throughput construction of carbohydrate antigen microarrays. A number of nitrocellulose-coated glass slides with different technical characteristics are commercially available. Given the structural diversity of carbohydrate antigens, examining each antigen preparation to determine the efficacy of its immobilization in a given type of substrate and the surface display of the desired glycoepitopes in a microarray assay is essential. Key words: Carbohydrate microarrays, Glycoconjugate, Glycoprotein, Glycolipid, Glycoepitope, Glycomics, Nitrocellulose, Polysaccharide

1. Introduction Although recognition of carbohydrate antigens began in the early twentieth century with the study of microbial polysaccharides (1), the scope of investigation into these antigens has since been substantially extended (2). Conceptually, carbohydrate antigens are carbohydrate-containing macromolecules that can evoke and react with carbohydrate-specific antibodies. In terms of their structural characteristics, carbohydrate antigens are polysaccharides with solely carbohydrate moieties and various forms of glycoconjugates. The latter include natural glycoproteins and glycolipids of living cells, as well as synthetic antigens produced by coupling monoand oligosaccharides to a larger carrier molecule.

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_17, © Springer Science+Business Media, LLC 2012

241

242

D. Wang

Carbohydrate moieties are unsurpassed in generating structural diversity and are prominent in surface displays. They may be specifically or selectively recognizable by either soluble or membranebound cellular proteins. In the higher eukaryotic species, the expression of cellular glycoconjugates, and especially their complex carbohydrate structures, is frequently cell type or tissue specific. In microbes, many sugar chains, including those displayed on the surface of a microbial cell and those secreted outside it, have been recognized as “signatures” of specific pathogens. Exploring the biological information content of sugar chains is a current focus of glycomics research. A number of experimental approaches have been developed to construct carbohydrate-based microarrays to facilitate the exploration of sugar chain diversity and its biomedical significance (3–13). These carbohydrate microarrays are all solid-phase binding assays for carbohydrates and their interaction with other biological molecules. In spite of their technological differences, they share a number of common characteristics and technical advantages. First, they have the capacity to display a large panel of carbohydrates in a limited chip space. Second, the amount needed to spot each carbohydrate is drastically smaller than that required for a conventional molecular or immunological assay. Third, the microarray-based assays have higher detection sensitivity than most conventional molecular and immunological assays; this increased sensitivity is due to the fact that the binding of a molecule in solution phase to an immobilized microspot of ligand in the solid phase minimally reduces the molar concentration of the molecule in solution (14). Therefore, it is much easier to have a binding equilibrium take place in a microarray assay and result in a high sensitivity. This chapter presents the method of nitrocellulose-based immobilization of carbohydrate-containing macromolecules. This method is suitable for the high-throughput construction of carbohydrate antigen microarrays (2, 3, 6, 10, 15), and thus is readily applicable for the large-scale immunological characterization of carbohydrate antigens and anticarbohydrate antibodies and the interaction of carbohydrates and other receptors.

2. Materials 2.1. Apparatus

1. Microspotting: Cartesian Technologies, Irvine, CA).

PixSys

5500C

(Cartesian

2. Bioarray Substrates: FAST Slides (Schleicher & Schuell, Keene, NH). 3. Scanner: ScanArray 5000A Microarray Scanner (PerkinElmer, Torrance, CA).

17  Carbohydrate Antigen Microarrays

2.2. Software

243

1. Array Design: CloneTracker (Biodiscovery, Marina del Rey, CA). 2. Array Printing: Cartesian AxSys (Cartesian Technologies, Irvine, CA). 3. Array Scanning: ScanArray Express (PerkinElmer, Torrance, CA). 4. Array Data-processing and Statistic Analysis: JMP-Genomics (SAS Institute Inc., Cary, NC).

2.3. Reagents and Buffers

1. Species-specific anti-immunoglobulin antibodies and their fluorescent conjugates, Cy3, Cy5, or FITC (Sigma, St. Louis, MO; BD-Pharmingen, San Diego, CA). 2. Dilution buffer: Saline (0.9% NaCl). 3. Rinsing solution: 1× phosphate-buffered saline (PBS), pH 7.4 with 0.05% (vol/vol) Tween 20. 4. Blocking solution: 1% bovine serum albumin (BSA) (wt/vol) in 1× PBS with 0.05% (wt/vol) NaN3.

3. Methods The key steps in our carbohydrate microarray applications are: 1. Designing and constructing sugar arrays. 2. Microspotting molecules onto bioarray substrates. 3. Immunostaining arrays. 4. Microarray scanning and data analysis. 5. Validating microarray data using conventional immunological assays. 6. Identifying carbohydrate-based biomarkers using carbohydrate microarrays. Each of these steps is described below. 3.1. Designing and Constructing Carbohydrate Antigen Arrays

The full surface of a microscope slide can be used to construct a “repertory” carbohydrate microarray with ~20,000 spots capacity for biomarker discovery. A multichamber, subarray system can also be used to construct customized, carbohydrate microarrays for defined purposes (16). For example, each glass slide is separated into 8 subarrays with 488 microspots spotted per subarray; each spot is approximately 200 mm, and the spots are situated at 300-mm intervals, center to center. A single slide of this design is suitable for eight microarray assays. Our laboratory has been using the latter approach more often in our research and clinical applications.

244

D. Wang

Repeats and dilutions: We usually print carbohydrate antigens at an initial concentration of 0.1–0.5 mg/ml. The amount of antigen or antibody solution printed on the chip substrate is in the range of 0.5–1.0 nL per microspot. The carbohydrate antigens are further diluted at 1:3, 1:9, and 1:27. A given concentration of each preparation is repeated at least three times. Antibody isotype standard curves: Antibodies of IgG, IgA, and IgM isotypes of corresponding species are diluted in 1× PBS, pH 7.4, and printed at given concentrations to serve as standard curves in the microarray format. Usually, they are applied at an initial concentration of 0.1 mg/ml and further diluted at 1:5, 1:25, 1:125, and 1:625. This design allows antibody signals that are captured by spotted carbohydrate antigens to be quantified by comparing the signal levels with the standard Ig curves of corresponding antibody isotypes. In addition, such standard curves are useful for microarray data normalization and cross-chip scaling of microarray detection. 3.2. Printing Carbohydrate Arrays onto Nitrocellulose Slides

Using Cartesian Technologies’ PixSys 5500C, a high-precision robot designed for cDNA microarrays, carbohydrates of various complexities (see Note 1) are picked up by quill pins dipped into carbohydrate solutions and printed onto bioarray substrates. The complementary Cartesian AxSys software is used to instruct movement of pins about the dispense platform and the printing process. Using the nitrocellulose-coated glass slides as bioarray substrates, carbohydrate antigens are immobilized to the nitrocellulose by physical–chemical adsorption involving noncovalent interactions, such as H bonding, ionic, and hydrophobic interactions with the substrate. The specific steps are the following. 1. Prepare samples of carbohydrate antigens in 0.9% NaCl at 0.1–0.5 mg/ml and transfer them in 96-well plates (see Notes 2 and 3). 2. Place the 96-well plates containing the samples and standards on the Cartesian arrayer robot for printing. 3. Adjust the print program so that the carbohydrate antigens and antibodies are printed at spot sizes of ~150  mm and at ­375-mm intervals, center to center. 4. Spot each antigen and antibody as triplet replicates in parallel. 5. Air dry the printed carbohydrate microarrays and store at room temperature overnight (see Note 5).

3.3. Immunostaining Carbohydrate Microarrays

The staining procedure for carbohydrate microarrays is basically identical to the routine procedure for immunohistology. Immunostaining steps of carbohydrate arrays are listed below (see Notes 4, 6, and 7). 1. Rinse printed microarray slides with 1× PBS, pH 7.4, and 0.05% Tween 20 for 5 min.

17  Carbohydrate Antigen Microarrays

245

2. Block slides with 1% BSA in PBS containing 0.05% NaN3 at room temperature for 30 min. 3. Stain each subarray with 50 ml of test sample, which is diluted in 1% BSA PBS containing 0.05% NaN3 and 0.05% Tween 20. 4. Incubate the slide in a humidified chamber at room temperature for 60 min. 5. Wash slides five times with 1× PBS, pH 7.4, and 0.05% Tween 20. 6. Stain slides with 50  ml of titrated secondary antibodies. Antihuman (or other species) IgG, IgM, or IgA antibodies with distinct fluorescent tags, Cy3, Cy5, or FITC, are mixed and then applied on the chips. 7. Incubate the slide in a humidified chamber with light protection at room temperature for 30 min. 8. Wash slides five times. 9. Place slide in a 50-ml falcon centrifuge tube and spin at 1,000 rpm for 2 min to remove washing buffer. 10. Cover slides in a histology slide box to prevent fluorescent quenching of signal by lights. 3.4. Microarray Scanning and Data Analysis

We scan microarray with ScanArray5000A Microarray Scanner (PerkinElmer Life Science) following the manufactory user ­manual. Fluorescence intensity values for each array spot and its background were calculated using ScanArray Express software. A  positive staining result is considered if the mean fluorescent intensity value of microspot is significantly higher than the mean background of the identically stained microarray with the same fluorescent color. Further microarray data analysis requires specialized software package and guidance of bioinformatics and statistical experts. It  is  important to conceptually understand the functions of the bioinformatics tools in use in order to correctly interpret the results. Our general approach is to process carbohydrate array datasets using JMP-Genomics software from SAS Institute. In brief, antigenspecific antibody reactivity is shown as microarray scores, which are the log2 transformed microarray values (mean and/or mean minus background). Then, we utilized an antigen-by-antigen ANOVA model to obtain statistically significant differences. Data from triplicate spots for each antigen were included in the ANOVA model for that antigen. A cutoff to detect significant differences is determined by applying a multiple testing correction to statistical results from the ANOVA model.

3.5. Validation and Further Investigation of Microarray Observations

A microarray finding may require further validation by other experimental approaches. We usually confirm our results by at least one of the alternative immunoassays, such as ELISA, Dot blot, Western Blot, flow cytometry, and immunohistology. However, the epitopes

246

D. Wang

or antigenic determinants displayed by a carbohydrate antigen in a specific biarray substrate may not be necessarily identical to those that are displayed by other assay systems. Thus, there is possibility that a “chip hit” is not reproduced by other assays. In such circumstances, one may conduct multiple carbohydrate array assays to confirm the initial microarray observation. 3.6. Identification of CarbohydrateBased Biomarkers Using Carbohydrate Microarrays

Our laboratory is interested in carbohydrate-based biomarkers, especially the immunogenic sugar moieties that play key roles in the host recognition of complex carbohydrates and immune responses to carbohydrate antigens. To facilitate these investigations, we explore the use of multiple platforms of carbohydrate microarrays. These include the carbohydrate antigen microarrays described here, a method for photogeneration of oligosaccharide arrays (11, 17), and methods for constructing carbohydrate cluster microarrays (13). Investigations using these technologies have led  to the discovery of autoimmunogenic sugar moieties of ­SARS-CoV (10), a highly potent immunological target of Bacillus anthracis exosporium (In this Issue, Proteomics, 7(2), pNA, and 11) and a number of glycan markers of human tumors (18). Conceptually, we take the advantage that the immune systems of many animal species are able to recognize subtle changes of sugar moieties displayed by cells or by soluble antigens. Technically, we make the use of complementary methods to diversify and extend the repertoires of glycoepitopes in bioarray substrates. It is important to note that carbohydrate microarrays constructed by various methods may differ in their technical features and suitability for a given practical application. For example, they may differ in the detection specificity. The carbohydrate antigen microarray discussed here would be antigen specific but not epitope specific, if the native carbohydrate antigens were spotted. This is owing to the fact that many carbohydrate antigens, such as polysaccharides, glycoproteins, and glycolipids, display more than one antigenic determinant, including glyco- and nonglycoepitopes. Examining the finer details of the binding properties would require the use of microarrays of pure saccharide sequences. The mono- and oligosaccharide array-based binding assays, such as sister chapters in this book, may be employed, in combination with saccharide competition assays, to decipher precise saccharide components of a specific antigenic determinant or glycoepitope (6, 11, 13).

17  Carbohydrate Antigen Microarrays

247

4. Notes 1. Antigen preparations suitable for this high-throughput biochip platform: Carbohydrate antigens of multiple structural configurations, including polysaccharides, natural glycoconjugates, oligosaccharide–protein, and oligosaccharide–lipid conjugates, are applicable for platform of carbohydrate antigen microarrays (3, 6, 10). In addition to printing carbohydrate microarrays, this platform is also applicable for producing protein microarrays (15). As with carbohydrate microarrays, there is no need to chemically conjugate a protein for its surface immobilization. However, it is recommended that each antigen preparation be tested on chip substrate for the efficacy of immobilization and expression of antigenic determinants. 2. Preservation of polysaccharides and glycoconjugates: Purified polysaccharides are generally stored as dried powder at room temperature. They can also be preserved in saline solutions (0.9% NaCl) containing a droplet of chloroform and stored at 4°C for a long period of time. Glycoconjugates are structurally diverse. Conditions for their storage are likely variable although it is practical to store them as smaller aliquots in Revco freezers (−80°C) to avoid multiple freezing and thawing of given preparation. 3. Printing of samples: Before loading sample solutions onto the arrayer, it is important to spin the solution in an Eppendorf centrifuge at maximum speed (at least 15,000 × g) for at least 15 min. Before and after each arraying experiment, it is recommended to examine and clean the printing pins following a routine pin cleaning protocol recommended by the corresponding printing pin manufacturers. A test slide is usually implemented to optimize quality of printing. The water supply of the Cartesian arrayer should constantly be checked during the arraying experiment to ensure adequate flow to the wash chamber. 4. Examination of the presence of samples on array and their antigenic structures: It is essential to examine whether proteins, synthetic peptides, and carbohydrates are successfully “printed” and whether desired epitopes or antigenic determinants are preserved on the chips. The printed microarrays can be incubated with antibodies, receptors, or lectins known to react with the printed substance. The reaction is detected either by conjugating directly a fluorochrome to the “detector” or by a second-step staining procedure. 5. Storage of printed carbohydrate and protein microarrays: The arrays are usually air dried and stored at room temperature. For long-term preservation, the chips can be sealed in a vacuumdried plastic bag with desiccant and stored at −20°C.

248

D. Wang

6. Staining considerations: After the last wash between each staining step, it is important to completely withdraw the wash buffer from inside reaction chambers. Otherwise, the remaining buffer may lower antibodies’ concentration to be analyzed. 7. Biosafety procedures: When working with chemicals, suitable protective wear, such as lab coat and disposable gloves, are advised. When human serum specimens are involved, experiments must be conducted in accordance to the standard biosafety procedures instituted by CDC and WHO. 8. Scanning and data collection: Training with the technical experts of PerkinElmer is necessary before performing microarray scanning and data collection using the ScanArray Express software package.

Acknowledgments Development of the carbohydrate microarrays described in this chapter was partially supported by NIH grants R21AI45326 and U01CA128416 to D. Wang. References 1. Dochez, A. R., and Avery, O. T. (1917) The elaboration of specific soluble substance by pneumococcus during growth, J. Exp. Med. 26, 477–493. 2. Wang, D. (2004) Carbohydrate Antigens, In: Encyclopedia of Molecular Cell Biology and Molecular Medicine, (ed. Robert A. Meyers) II, 277–301. 3. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells, Nat Biotechnol 20, 275–281. 4. Willats, W. G., Rasmussen, S. E., Kristensen, T., Mikkelsen, J. D., and Knox, J. P. (2002) Sugar-coated microarrays: A novel slide surface for the high-throughput analysis of glycans, Proteomics 2, 1666–1671. 5. Fazio, F., Bryan, M. C., Blixt, O., Paulson, J. C., and Wong, C. H. (2002) Synthesis of sugar arrays in microtiter plate, J Am Chem Soc 124, 14397–14402. 6. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrateprotein interactions, Nat Biotechnol 20, 1011–1017.

7. Houseman, B. T., and Mrksich, M. (2002) Carbohydrate arrays for the evaluation of protein binding and enzymatic modification, Chem Biol 9, 443–454. 8. Park, S., and Shin, I. (2002) Fabrication of carbohydrate chips for studying protein-carbohydrate interactions, Angew Chem Int Ed Engl 41, 3180–3182. 9. Adams, E. W., Ratner, D. M., Bokesch, H. R., McMahon, J. B., O’Keefe, B. R., and Seeberger, P. H. (2004) Oligosaccharide and glycoprotein microarrays as tools in HIV glycobiology; glycan-dependent gp120/protein interactions, Chem Biol 11, 875–881. 10. Wang, D., and Lu, J. (2004) Glycan arrays lead to the discovery of autoimmunogenic activity of SARS-CoV, Physiol Genomics 18, 245–248. 11. Wang, D., Carroll, G. T., Turro, N. J., Koberstein, J. T., Kovac, P., Saksena, R., Adamo, R., Herzenberg, L. A., Herzenberg, L. A., and Steinman, L. (2007) Photogenerated glycan arrays identify immunogenic sugar moieties of Bacillus anthracis exosporium, Proteomics 7, 180–184. 12. Zhou, X., and Zhou, J. (2006) Oligosaccharide microarrays fabricated on aminooxyacetyl functionalized glass surface for characterization of carbohydrate-protein interaction, Biosensors & Bioelectronics 21, 1451–1458.

17  Carbohydrate Antigen Microarrays 13. Zhou, X., Turchi, C., and Wang, D. (2009) Carbohydrate cluster microarrays fabricated on three-dimensional dendrimeric platforms for functional glycomics exploration, Journal of proteome research 8, 5031–5040. 14. Ekins, R., Chu, F., and Biggart, E. (1990) Multispot, multianalyte, immunoassay, Ann Biol Clin 48, 655–666. 15. Wang, D. (2003) Carbohydrate microarrays, Proteomics 3, 2167–2175. 16. Wang, R., Liu, S., Shah, D., and Wang, D. (2005) A practical protocol for carbohydrate

249

microarrays, Methods in Molecular Biology (Totowa, NJ, United States) 310, 241–252. 17. Carroll, G. T., Wang, D., Turro, N. J., and Koberstein, J. T. (2006) Photochemical micropatterning of carbohydrates on a surface, Langmuir 22, 2899–2905. 18. Newsom-Davis, T. E., Wang, D., Steinman, L., Chen, P. F., Wang, L. X., Simon, A. K., and Screaton, G. R. (2009) Enhanced immune recognition of cryptic glycan markers in human tumors, Cancer research 69, 2018–2025.

Chapter 18 Probing Virus–Glycan Interactions Using Glycan Microarrays Jamie Heimburg-Molinaro, Mary Tappert, Xuezheng Song, Yi Lasanajak, Gillian Air, David F. Smith, and Richard D. Cummings Abstract Glycan microarrays are surfaces that contain immobilized oligosaccharides or glycoconjugates and have proven useful in probing the interactions between glycan-binding proteins (GBPs) and individual glycans. Such glycan microarrays have been especially important in studying virus–glycan interactions, as most viruses express one or more GBPs important for pathogenesis. For studying interactions of influenza viruses with glycans, we describe protocols for fluorescent labeling of virus, addition of virus to a glycan microarray, analysis of a glycan microarray slide experiment, and interpretation of data. Key words: Glycomics, Glycan microarrays, Influenza virus, Fluorescence detection, Glycan-binding motif

1. Introduction It is well established that the extracellular matrix and cell surface of epithelial cells are covered with glycans. These glycans are critical to physiological processes, such as cell growth, cell adhesion, and cell–cell signaling, and also act as receptors for many microbes, such as bacteria, parasites, and viruses (1). Pathogens utilize the glycan receptors for initial binding events, followed by cell infection. Influenza viruses are a well-studied example of a group of viruses that interact with glycans on the cell surface. Influenza viruses express a glycan-binding protein called hemagglutinin, which specifically binds to sialic acid-containing glycans to ultimately allow endocytosis and fusion of the virus with the host cell (2–4). There are many variants of hemagglutinin proteins in influenza viruses, which convey specificity for different sialic acids and

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_18, © Springer Science+Business Media, LLC 2012

251

252

J. Heimburg-Molinaro et al.

therefore different tissue tropisms for individual viruses (2). Human influenza viruses bind sialic acid-linked a2-6 to the next sugar, while avian viruses bind a2-3 sialic acid. Interestingly, human parainfluenza viruses all bind a2-3 sialic acid. An excellent method for analyzing the specificity of viruses and viral hemagglutinins is the glycan microarray. The Consortium for Functional Glycomics (CFG) has produced a large mammalian glycan microarray, which has been widely used to study the glycan interactions of many pathogens, notably viruses (5). This has yielded an abundance of information on virus–glycan-binding specificities (https://www.functionalglycomics.org/static/consortium/news.shtml). The most straightforward technique for evaluating virus binding to the microarray is by direct fluorescent labeling of the virus. This allows the virus to be incubated with the glycan microarray in one step followed by the quantification of binding and data analysis to obtain the glycan specificity of the virus (6–8). The example of influenza virus will be discussed below to illustrate the protocol for virus labeling, assaying on the glycan microarray, and data analysis. The glycan microarray is a powerful tool (9–11) for screening viruses for their glycan-binding properties and generating hypotheses based on the detected interactions.

2. Materials 2.1. Fluorescent Labeling of Virus

1. Prepared Influenza virus (for example, virus isolated from MDCK cells, purified by sucrose gradient centrifugation, checked for purity by SDS-PAGE, quantified by hemagglutination (HA) assay and/or total viral protein) (BioRad Protein Assay or quantitative SDS-PAGE) (6, 7). (a) Virus is grown in LLC-MK2 cells (hPIVs) or MDCK (influenza) in infection medium [DMEM/Ham’s F12 (Gibco) supplemented with 1% ITS + (BD) and 0.1% gentamicin (Sigma)]. For influenza, 0.5 mg/mL trypsin (TPCK treated, Worthington) is added to infection medium. For hPIV1 and -2, add 1  mg/mL trypsin to the infection medium (see Note 1). (b) Virus is purified by a low speed spin first to remove cell debris, then by pelleting virus from the clarified medium at 52,112 × g for 2 h (Beckman L-80 ultracentrifuge, SW-28 rotor), resuspending overnight in calcium–magnesium saline (50  mL per pellet), then centrifugation through a 10–60% sucrose gradient (hPIV) or 10–40% (influenza) (Beckman TL-100 ultracentrifuge, TLS-55 rotor, gradient

18  Probing Virus–Glycan Interactions Using Glycan Microarrays

253

at 105,000 × g for 15  min then final pelleting of band at 214,200 × g for 1 h). (c) Virus concentrations average ~0.5  mg/mL total viral ­protein. Corresponding HA titer, which is a more relevant measure of viral activity, varies among viral species. A “good” preparation of most influenza or hPIV1 is around 64,000 HAU/mL; a “good” preparation of hPIV2 or -3 is around 12,800 HAU/mL. 2. Calcium/magnesium saline for resuspension of virus (after pelleting to removing sucrose or other buffer) and for dialysis: 0.15  M NaCl, 0.25  mM CaCl2, and 0.8  mM MgCl2 (Fisher Scientific). 3. 0.15 M Sodium chloride. 4. 1 M Sodium bicarbonate, pH 9.0. 5. Fluorescent dye with reactive group (for example, Alexa Fluor 488 succinimidyl ester, Molecular Probes). 6. Slide-A-Lyzer Mini Dialysis Units (7,000 MWCO) (Pierce/ Thermo Fisher Scientific). 7. 9% SDS gel. 2.2. Assay of Virus Binding to the CFG Glycan Microarray

1. Glycan microarray-printed slides (CFG) (Do not touch the printed area). 2. Cover slips (Fisher scientific). 3. Humidified Slide processing chambers (Fisher Scientific) (see Note 2). 4. 100-mL Coplin jars for washing slides. 5. MilliQ water (dH2O). 6. Cyanine5 (Cy5)-labeled Streptavidin (ZYMED). 7. TSM buffer: 20 mM Tris–HCl, pH 7.4, 150 mM NaCl, 2 mM CaCl2, and 2 mM MgCl2. 8. TSM Wash Buffer (TSMW): TSM Buffer + 0.05% Tween 20. 9. TSM-Binding Buffer (TSMBB): TSM buffer + 0.05% Tween 20 + 1% BSA (see Note 3).

2.3. Data Analysis of Microarray Binding

1. Processed glycan microarray slide. 2. ProScanArray Scanner equipped with multiple lasers for detecting fluorophores (Perkin Elmer). 3. Image analysis software (Imagene software (BioDiscovery) or Perkin Elmer ScanArray Express). 4. Data analysis software (Microsoft excel).

254

J. Heimburg-Molinaro et al.

3. Methods 3.1. Fluorescent Labeling of Virus

1. To 100 mL of virus in calcium–magnesium saline, add 10 mL of 1 M sodium bicarbonate pH 9.0 (see Note 4). 2. For 100 mL virus (for example, at 1.25 × 104 HAU/mL), prepare the Alexa Fluor 488 labeling reagent. Typically, ~70 mg of Alexa Fluor 488 succinimidyl ester should be used. This is equivalent to ~0.005 mg dye per HAU (see Note 5). 3. Resuspend the Alexa Fluor 488 succinimidyl ester in 25  mL dH2O and add to the tube of virus (see Note 6). 4. Incubate samples for 1 h at room temperature in the dark or wrapped in foil on stirrer, using a 5 mm × 2 mm mini Spinbar (Bel-Art Products, 371210010) to stir the sample inside a microcentrifuge tube. 5. Transfer the entire labeling mixture to a Slide-A-Lyzer Mini Dialysis Unit (7,000 MWCO). 6. Dialyze against three changes of calcium/magnesium saline at 4°C (change after 1 h, overnight, then for an additional 1 h). Keep samples in the dark. 7. Run samples on a 9% SDS gel. Look at the gel under UV light; only the bands corresponding to the known molecular weights of surface proteins should fluoresce. The gel can be subsequently Coomassie stained to confirm band identity if necessary.

3.2. Assay of Virus Binding to Glycan Microarray

1. Prepare TSM, TSMW, and TSMBB Buffers (as described above). 2. Prepare sample(s): (a) Prepare 100  mL of virus sample by diluting the fluorescently labeled virus in TSMBB to an appropriate final concentration required for the analysis (e.g. 1:20 dilution of above-prepared virus) (see Notes 7 and 8). (b) Prepare 100 mL of Cy5-labeled streptavidin at 0.5 mg/mL final concentration in TSMBB. 3. Hydrate glycan microarray slides in 100  mL of TSMW in a Coplin Jar for 5  min and drain excess buffer from slide by briefly touching corner of slide to a paper towel. 4. Lay the slide flat and add 70 mL of the Cy5-streptavidin to the slide (see Notes 9 and 10). 5. Slowly place cover slip on the slide, avoiding the formation of bubbles in the sample under the cover slip. If necessary, remove any bubbles by gently tapping the cover slip with a pipette tip or slowly lifting one side of the cover slip. Make sure that the

18  Probing Virus–Glycan Interactions Using Glycan Microarrays

255

cover slip is properly positioned over the glycan microarray printed in the designated area on the slide. 6. Incubate slide in a humidified slide-processing chamber in the dark for 1 h at RT or other appropriate time and temperature depending on the experimental design. 7. When the incubation is complete, remove cover slip by gently allowing it to slip off the slide either directly into the biohazard trash or into a Coplin Jar filled with wash buffer (TSMW) (see Note 11). 8. Wash the slide by gently dipping four times into 100  mL of each of the following buffers in Coplin Jars: (a) TSMW (b) TSM 9. Immediately after the TSM wash, drain excess buffer from slide by briefly touching corner of slide to a paper towel and lay the slide flat and add 70 mL of the virus preparation. 10. Slowly place cover slip on slide, trying to avoid the formation of bubbles in the sample under the cover slip. Remove any bubbles by gently tapping the cover slip with a pipette tip if necessary, or slowly lifting one side of the cover slip. Make sure that the cover slip is properly positioned over the glycan microarray printed in the designated area on the slide. 11. Incubate slide in a humidified slide-processing chamber in the dark for 1 h at 4°C (see Note 12). 12. After 1 h incubation, remove cover slip by gently allowing it to slip off into the glass trash/biohazard trash. 13. Wash the slide by gently dipping four times into 100  mL of each of the following buffers in Coplin Jars: (a) TSMW (b) TSM (c) dH2O 14. Spin slide in slide centrifuge for ~15 s or remove dH2O under a gentle stream of nitrogen. Wipe bottom (non-printed) side of the slide with a Kimwipe. 3.3. Data Analysis of Microarray Binding 3.3.1. Scanning Slides

1. Turn on Scan Array Express or other scanner equipped with appropriate lasers for analysis of the selected fluorescent tag(s). 2. Allow lasers at least 15 min to warm up before scanning. 3. Configure scanner by selecting the appropriate laser setting for each fluorophore used; i.e. Alexa Fluor 488, Cyanine5 (Cy5), Cyanine3 (Cy3), etc. (a) Check PMT and power of lasers: Standard PMT­ (% gain) = 70, Laser power = 90%.

256

J. Heimburg-Molinaro et al.

Fig. 1. Image of a glycan microarray slide with Alexa Fluor 488-labeled virus bound to specific glycans. Glycans are printed in replicates of 6, and a pattern of binding to the glycans can be seen. The fluorescence detected in the image is then quantified using the quantitation software described, generating data that is shown in Figs. 2 and 3.

4. Select scan protocol that was configured in step 3. 5. Place slide in scanner with the microarray facing up and the barcode entering the scanner last. 6. Select scan and then Run scan protocol-press OK. Scan takes ~5 min at each excitation wavelength. 7. Each image for each wavelength scanned should be saved separately as a TIF file (see Fig. 1). 3.3.2. Analyzing Slide Images (Using Imagene Software) (See Note 13)

1. Open Imagene software (or another appropriate software if Imagene is not available). 2. Load appropriate TIF image(s). 3. Load appropriate grid file. 4. Load appropriate gene ID file. 5. Align grid. (a) Click Selection: adjust metagrid-click and drag grid to align with biotin spots on the microarray. (b) Click Selection: adjust subgrid-adjust individual grids using biotin spots. (c) Auto adjust spots: click Auto: auto adjust all spots – be sure to be clicked off of any individual grid.

18  Probing Virus–Glycan Interactions Using Glycan Microarrays

257

(d) Manually adjust spots: click Selection: adjust spots – adjust individual spots so that the entire spot fits inside the circle, and tighten the spots if needed by dragging the lower right corner of the circle to the correct size. If circles are around background spots that have appeared on the slide, the circle can be moved off of the spot. 6. Click Measure: make measurements. This process measures intensity of spots reporting the results as average relative fluorescence units (RFU) of the n = 6 spots on the CFG microarray. The results are in the form of a text file, which is used up upload the data onto the CFG website where it is converted to an interactive Bar Chart linked to several databases. 7. Save files. 3.3.3. Analyzing Data

1. The text file is also opened in an Excel macro that is produced for each glycan microarray format. For the CFG glycan microarray, a unique Excel Macro is produced for each version of the glycan microarray. 2. Once the values are calculated, the results of processing by the Excel Macro are presented in an Excel spreadsheet. Here the averages of RFU of binding of virus to individual replicates of the printed glycans are presented in tables, which list the glycan structures with associated RFU values sorted in order of appearance on the microarray in one column and in descending order of Average RFU values in another column. A histogram of RFU plotted vs. glycan number is also generated. See Figs. 2 and 3. 3. Each successive version of the CFG glycan microarray contained an increasing number of glycans; i.e. v4.2 of the microarray consists of 511 glycans in replicates of 6. The average RFU value from the 6 replicates is provided in the tables, along with the standard deviation, and the standard error of the mean. The coefficient of variation is also included as a %CV and is determined by multiplying the Standard Deviation/Mean by 100. The highest and lowest RFU values from each set of 6 replicates are removed so the mean is of 4 values rather than 6. This eliminates some of the “outliers” that contain a single very high or low point. Thus, points with high %CV should be considered suspect. 4. Data from glycan microarrays, composed of hundreds of printed glycans, are quite complex. In the case of most influenza or parainfluenza viruses, the data are somewhat simplified because most viruses bind to glycans possessing sialic acid at their non-reducing ends. Thus, for v4.2 of the CFG glycan microarray, there are 123 glycans containing sialic acid in various linkages on a variety of structures that need to be evaluated.

258

J. Heimburg-Molinaro et al.

Fig. 2. Alexa Fluor 488-labeled human parainfluenza virus 1 (hPIV1), strain C-35 (51,200 HAU/ml) diluted 1:10, binding to v3.2 of the CFG glycan microarray (http://www.functionalglycomics.org/static/consortium/resources/resourcecoreh8. shtml). Histogram presents the glycan number (x-axis) vs. the relative fluorescence units (RFU) (y-axis) of binding to each glycan. The highest binding glycans contain terminal epitopes of sialic acid-linked a2-3 to galactose.

Fig. 3. Alexa Fluor 488-labeled human influenza A virus, strain A/Oklahoma/447/08 H1N1 (61,147 HAU/ml) diluted 1:20, binding to v4.0 of the CFG glycan microarray (http://www.functionalglycomics.org/static/consortium/resources/ resourcecoreh8.shtml). Histogram presents the glycan number (x-axis) vs. the relative fluorescence units (RFU) (y-axis) of binding to each glycan. The highest binding glycans contain terminal epitopes of sialic acid-linked a2-6 to galactose.

18  Probing Virus–Glycan Interactions Using Glycan Microarrays

259

Most viruses differentiate between sialic acid-linked a2-3 or a2-6 to a terminal, non-reducing galactose, but may differ ­significantly with respect to the underlying glycan structure. Interpretation of the data is currently done by manual inspection of the bound glycan structures. One approach to determine relative binding strengths is by ordering the glycans based on RFU bound. For this data to be valid, one needs to be sure the data are in a linear range. Determining the linear range can be accomplished by doing analyses at multiple concentrations of virus; i.e. from a high dilution to a low dilution using the same slide (to conserve slides) until a constant RFU is obtained for the highest binding glycans. Data from the curve showing about half-maximal binding of the highest binding glycan will be in a linear range. The highest binding glycans under these conditions may be considered the strongest binding glycans. To analyze specificity of binding, it is important to inspect not only what glycans the virus binds, but also what related structures are not bound. We use the following steps for these analyses: (a) The binding assay is carried out at multiple concentrations by preparing a series of two- or fivefold serial dilutions of the stock suspension of labeled virus. Using a single slide, a binding assay is carried out according to the protocols outlined above at the highest dilution of virus. After the slide is analyzed, it is rehydrated in the wash buffer and the next dilution (higher concentration) of virus is applied to the slide. The slide is processed and increasing concentrations are similarly assayed until the RFU of the highest binding glycan(s) stop increasing, which indicates the slide has been saturated. (b) To select the top binding glycans in this analysis, the data at each dilution of virus is normalized by determining the percentile of binding of each glycan relative to the strongest bound glycan. This analysis is performed on two or more assays at non-saturating concentrations of virus to be sure the data are in a linear range. Then the average percentile ranking of each glycan is determined by averaging the Rank of each glycan at each dilution of virus. To order the ranked glycans, the data are sorted according to the Average Rank in decreasing order. This can all be done using an excel Spreadsheet. (c) As an example, the data in Table 1 show the results of ranking the glycans bound by labeled human parainfluenza virus 1, strain C35 (51,200 HAU/ml). The analyses were done at dilutions of 1:100, 1:40, 1:20, and 1:10. Notice that the RFU bound of strongest binding glycan (#234) increased with increasing concentration of virus until a

Structure

Neu5Aca2-3Galb14GlcNAcb1-3Galb14GlcNAcb1-3Galb14GlcNAcb-Sp0

Neu5Aca2-3(6OSO3) Galb1-4GlcNAcb-Sp8

Neu5Aca2-3Galb14GlcNAcb1-3Galb13GlcNAcb-Sp0

Neu5Aca2-3Galb14GlcNAcb1-2Mana13(Neu5Aca2-3Galb14GlcNAcb1-2Mana1-6) Manb1-4GlcNAcb14GlcNAcb-Sp12

Neu5Aca2-3Galb13(Neu5Aca2-3Galb1-4) GlcNAcb-Sp8

Neu5Aca2-3Galb1-4 (Fuca1-3)GlcNAcb-Sp0

Glycan no.

234

44

282

142

218

229

11,149 13

2

5

10,794

  9,476

2

1

4

12,967

14,742

13,910

76

64

73

88

100

  94

10,761

10,328

11,329

11,224

13,359

13,706

%CV Rank RFU

7

3

6

7

6

7

79

75

83

82

97

100

9

6,256 11

6,846 20

7,841 13

8,191

9,433 15

11,692 14

54

59

67

70

81

100

4,846   15

5,009   16

4,252   14

5,188   17

3,875   18

5,722   20

85

88

74

91

68

100

72

74

75

81

82

100

%CV Rank Avg. rank

Max RFU = 5,721

1:100 dilution

%CV Rank RFU

Max RFU = 11,691

1:40 dilution

%CV Rank RFU

Max RFU = 13,705

Max RFU = 14,742 RFU

1:20 dilution

1:10 dilution

Table 1 Binding of hPIV1, strain C-35 (51,200 HAU/ml) to v3.2 of the CFG glycan array is determined over four dilutions and the Rank of each glycan is based on its RFU divided by the maximal RFU at each dilution calculated as a percentile

260 J. Heimburg-Molinaro et al.

Neu5Aca2-3Galb1-4 (Fuca1-3)GlcNAcb 1-3Galb1-4(Fuca1-3) GlcNAcb1-3Galb14(Fuca1-3)GlcNAcb-Sp0

Neu5Aca2-3Galb1-4 (Fuca1-3)GlcNAcb13Galb-Sp8

Neu5Aca2-3Galb14Glcb-Sp8

NeuAca2-3Galb1-4Glc NAcb1-3GalNAc-Sp14

Neu5Aca2-3Galb1-3 (Neu5Aca2-3Galb1-4Glc NAcb1-6)GalNAca-Sp14

Neu5Aca2-3Galb1-4 (Fuca1-3)GlcNAcb-Sp8

Neu5Gca2-3Galb1-4 (Fuca1-3)GlcNAcb-Sp0

Galb1-3(Neu5Aca23Galb1-4GlcNacb1-6) GalNAca-Sp14

Neu5Aca2-3(GalNAcb1-4) Galb1-4GlcNAcb-Sp0

Neu5Aca2-3(GalNAcb1-4) Galb1-4Glcb-Sp0

Neu5Aca2-3Galb1-4 (Fuca1-3)GlcNAcb13Galb1-4GlcNAcb-Sp8

228

231

239

371

312

230

256

274

208

210

232

6

7

9

1

7

  8,106

  7,126 7

4

  8,006 12

  7,399

  9,048

  9,877

  8,415

  9,414 10

  7,588

55

48

54

50

61

67

57

64

51

90

13,339

5

80

11,866 14

6

9

13

5,266

5,583

6,306

9

3

9

5,370 22

7,887 15

7,733

7,490 15

9,143 12

8,491

12,521

10,529 14

38

41

46

39

58

56

55

67

62

91

77

13

6

30

8

9

24

4,477

3,430

3,694

7

13

11

3,605 34

4,627

5,514

5,904

5,095

5,175 12

6,561

7,212

38

29

32

31

40

47

50

44

44

56

62

1,685   15

2,073   32

2,036   52

2,520   49

1,990   29

2,696   39

3,482   32

3,724   45

4,124   15

3,403   12

3,953   9

29

36

36

44

35

47

61

65

72

59

69

(continued)

35

35

38

38

44

50

55

58

59

69

69

18  Probing Virus–Glycan Interactions Using Glycan Microarrays 261

Structure

Neu5Aca2-3(GalNAcb1-4) Galb1-4GlcNAcb-Sp8

Neu5Aca2-3Galb14GlcNAcb-Sp8

Neu5Aca2-3(6OSO3)Galb14(Fuca1-3)GlcNAcb-Sp8

Neu5Aca2-6Galb1-4Glcb-Sp0

Neu5Aca2-3Galb14GlcNAcb-Sp0

Neu5Aca2-8Neu5Aca23Galb1-4Glcb-Sp0

Neu5Aca2-3Galb13GlcNAcb1-3Galb14GlcNAcb-Sp0

Neu5Aca2-3Galb14GlcNAcb1-3Galb14GlcNAcb-Sp0

Glycan no.

209

236

207

247

235

251

223

237

Table 1 (continued)

6

11,642

4

  79

  41

8

6,003

  47   26

6

  54

  63

  69

  35

3,905 11

6,963

7,915 10

9,258

10,202 13

5,224 13

3,638

2,997

3,960

4,322

5,234

5,149

5,908

2,080

%CV Rank RFU

  7

  11

  11

  14

  23

  16

  28

  15

27

22

29

32

38

38

43

15

2,635

2,817

1,890

4,235

2,898

4,049

4,148

2,119

12

4

20

11

19

7

26

13

  23

  24

  16

  36

  25

  35

  35

  18

446

900

1,107

1,367

1,797

1,456

1,405

4,056

54

32

14

15

46

7

14

39

  8

  16

  19

  24

  31

  25

  25

  71

  19

  21

  21

  31

  31

  33

  34

  35

%CV Rank Avg. rank

Max RFU = 5,721

1:100 dilution

%CV Rank RFU

Max RFU = 11,691

1:40 dilution

%CV Rank RFU

Max RFU = 13,705

Max RFU = 14,742 RFU

1:20 dilution

1:10 dilution

262 J. Heimburg-Molinaro et al.

Neu5Aca2-3Galb14Glcb-Sp0

Neu5Aca2-3Galb13GlcNAcb-Sp8

Neu5Aca2-6Galb14GlcNAcb1-3Galb14GlcNAcb-Sp0

Neu5Aca2-3Galb1-4(6OSO3) GlcNAcb-Sp8

Neu5Aca2-3Galb1-4 (Fuca1-3)GlcNAcb13GalNAca-Sp14

Neu5Aca2-8Neu5Aca28Neu5Aca2-3 (GalNAcb1-4)Galb14Glcb-Sp0

Neu5Aca2-8Neu5Aca28Neu5Aca2-8Neu5Aca23(GalNAcb1-4) Galb1-4Glcb-Sp0

Neu5Gca2-3Galb14GlcNAcb-Sp0

238

225

246

226

373

203

202

257

3

5

8

  10

1,401 17

  9

  11

1,586 12

1,345 45

  10

  35

  24

  27

  58

  58

1,547 30

5,223 22

3,540

3,955 29

8,567

8,538

  994

1,101

1,238

1,643

1,666

1,862

2,115

2,381

3,437

  25

  9

  27

  36

  17

  14

  38

  11

  5

 7

 8

 9

12

12

14

15

17

25

546

309

476

617

1,458

1,954

1,563

2,833

3,271

64

61

49

67

6

3

13

35

6

  5

  3

  4

  5

  12

  17

  13

  24

  28

34

50

81

78

60

73

10

98

96

79

5 114

118

29

310

442

50

75

  0

  2

  0

  2

  1

  5

  8

  1

  1

  4

  4

  4

  6

  8

  12

  12

  14

  18

The Average Rank is determined from at least two dilutions in the linear range of binding. In this case, the Average Rank is calculated from the 1:100, 1:40, and 1:20 dilution data. The glycans are sorted with the highest binding glycan structures (highest Average Rank) at the top of the list

Neu5Aca2-3Galb14GlcNAcb1-2Mana13(Neu5Aca2-6Galb14GlcNAcb1-2Mana1-6) Manb1-4GlcNAcb14GlcNAcb-Sp12

315

18  Probing Virus–Glycan Interactions Using Glycan Microarrays 263

264

J. Heimburg-Molinaro et al.

dilution of 1:10 was reached. Thus the linear range of virus concentration was represented by the lower 3 dilutions. (d) By calculating the average rank of the binding of each glycan in the linear range (the lower 3 dilutions), an average rank was obtained and the glycans were ordered according to relative binding from highest to lowest ranking. All rankings below 4 percentile were considered weak- or nonbinders. 5. The data from the ranked and ordered glycans determined from concentration-dependent binding assays is used to manually determine a motif of binding. (a) Tally up the appearance of particular sugars and linkages in the first, second, third, and/or beyond positions in the binding glycans (those binding above 4 percentile, as described in 4d). For example, 34 glycans are identified as “bound” in Table 1; 30 of them have Neu5Aca2-3 as the sialic acid attached to the Gal, 2 have Neu5Aca2-6, and 2 have Neu5Gca2-3. (b) Look for the minimum number of sugars in the chains bound (in Table 1, three). (c) Look for common substituents (in Table  1, e.g. fucose, sulfate, and GalNAc) and the sugars to which they are attached. (d) Look for numbers of branches on the structures. (e) Take all of the above and draw the sequence of the generic bound structure (from Table  1: (±Neu5Aca2-8) Neu5Aca2-3(±6OSO3)Galb1-4(±Fuc, ±GalNAc)GlcNAc; possibility of Neu5Aca2-6 or Neu5Gca2-3 binding). (f) Take the generic bound structure and look at the nonbinding glycans, especially those that are in the general glycan category. If nonbinding structures that otherwise fit the generic bound motif are found, refine binding motif according to characteristics of these structures (in Table 1, two binding structures have Neu5Aca2-6 linkages, but hPIV1 would not be defined as generally binding to Neu5Aca2-6 due to the large number of these structures that are not bound.)

4. Notes 1. hPIV3 does not require trypsin (12, 13). Four 175-cm tissue culture flasks (about 200 mL medium) will generally yield sufficient virus for purification and glycan microarray analysis.

18  Probing Virus–Glycan Interactions Using Glycan Microarrays

265

2. A homemade system for a humidified chamber can be made using a lidded Petri Dish (150  mm in diameter), with wet paper towels in the bottom of the chamber and two parallel glass rods or cut plastic pipette on top of the moist paper towels as a “slide rack” to hold the slide off of the paper towels for incubation. 3. If buffers have been prepared ahead of time, they should be stored at 4°C and brought to room temperature before use. 4. The purified virus should be used at about 1.25 × 105 HAU/ mL. This corresponds to a dilution of 1:100 that gives an HA titer of 64 in the standard HA assay (14). An SDS-PAGE gel should be run prior to use to ensure the virus preparation has only virus proteins. 5. The amount of dye per amount of virus should be optimized for each virus to ensure sufficient labeling with no loss in biological activity. For viruses without hemagglutination or neuraminidase activity, amount of dye per amount of virus can be determined relative to total viral protein. Labeling is sufficient when the surface proteins are visible under UV light on an SDS gel. 6. To prepare and store the Alexa Fluor 488 reagent, dissolve 1 mg Alexa Fluor 488 in 1 mL 0.15 M saline and make 25 mL aliquots. Then use a speed-vac to dry the samples and store at −20°C until use. Reconstitute in 25 mL dH2O. 7. The buffers used for dilution can vary based on the sample. The glycan microarrays are very stable and can be used with any buffer or ionic strength in the pH range of 4–8.5. Since some glycans may be O-acetylated, buffers above pH 8 should be avoided when such glycans are being monitored. For example, some viruses are not stable in buffers containing calcium, magnesium, Tween, etc. In assays for endogenous viral neuraminidase activity, a sodium acetate based buffer at pH 5.5 is used with incubations at 37°C. 8. An appropriate final concentration should be determined experimentally for each virus preparation. This is accomplished empirically by preparing several dilutions; i.e. a twofold serial dilution down to 1:128 and applying an aliquot of the lowest dilution to the microarray, incubating for binding as described below and processing to determine amount of virus bound. Repeat this process on the same slide (to conserve glycan microarrays) until a constant value is reached for the strongest binding glycan. The dilution producing approximately half maximal binding will represent a concentration that is in a linear range of virus binding and is suitable for analysis. 9. Cy5-streptavidin is added to the slide and washed away before virus is added to visualize the biotin spots that are printed on

266

J. Heimburg-Molinaro et al.

the glycan microarray, to locate the subarrays of glycans and perform the grid alignment, which is necessary for data analysis. If possible, the Cy5-streptavidin may be mixed in with the sample and incubated together on the microarray. Although the Cy5 and Alexa Fluor 488 do not interfere with each other, we apply the streptavidin prior to the virus to avoid any question about streptavidin interacting with the virus. 10. The Cy5 label on the streptavidin is one of several possible fluorophores. If the virus is labeled with Alexa Fluor 488, for example, two separate images are obtained, one scanned at 488 nm for binding of Alexa Fluor 488 and another for binding of Cy5 at an excitation wavelength of 633 nm. 11. The cover slip should be carefully removed to avoid damage to the glycan microarray printed surface. Harsh manipulations that would allow the edge of the cover slip to scratch the microarray surface must be avoided. An alternative to doing the analysis under a cover slip, which conserves sample, is to use a hydrophobic pen to surround the surface of the glass slide around the area of the printed glycan microarray prior to step 3. After the resulting hydrophobic barrier has dried, up to 1 mL of sample can be applied and retained within the barrier. Incubations can be carried out as described without a cover slip, but larger sample volumes are required. After incubation, the sample can be drained off and the slide can be washed with a gentle stream of buffer from a plastic squeeze bottle prior to washing as described in step 13. 12. The incubation for influenza virus on the glycan microarray is typically performed at 4°C to eliminate the possibility of interference by the activity of the viral neuraminidase. However, if this is not a concern, or if the neuraminidase activity is to be studied, the experiment can be performed at room temperature or 37°C. 13. The image analysis is individual to each scanner and software. Different systems may have different methods for performing these steps, and the instructions for individual scanners and software should be followed. References 1. Varki, A., Cummings, R. D., Esko, J. D., Freeze, H. H., Stanley, P., Bertozzi, C. R., Hart, G. W., and Etzler, M. E. (2009) Essentials of Glycobiology, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 2. Nicholls, J. M., Chan, R. W. Y., Russell, R. J., Air, G. M., and Peiris, J. S. M. (2008) Evolving

complexities of influenza virus and its receptors, Trends Microbiol 16, 149–157. 3. Stevens, J., Blixt, O., Glaser, L., Taubenberger, J. K., Palese, P., Paulson, J. C., and Wilson, I. A. (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities, J Mol Biol 355, 1143–1155.

18  Probing Virus–Glycan Interactions Using Glycan Microarrays 4. Viswanathan, K., Chandrasekaran, A., Srinivasan, A., Raman, R., Sasisekharan, V., and Sasisekharan, R. (2010) Glycans as receptors for influenza pathogenesis, Glycoconj J 27, 561–570. 5. Blixt, O., Head, S., Mondala, T., Scanlan, C., Huflejt, M. E., Alvarez, R., Bryan, M. C., Fazio, F., Calarese, D., Stevens, J., Razi, N., Stevens, D. J., Skehel, J. J., van Die, I., Burton, D. R., Wilson, I. A., Cummings, R., Bovin, N., Wong, C. H., and Paulson, J. C. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins, Proc Natl Acad Sci U S A 101, 17033–17038. 6. Amonsen, M., Smith, D. F., Cummings, R. D., and Air, G. M. (2007) Human parainfluenza viruses hPIV1 and hPIV3 bind oligosaccharides with alpha2-3-linked sialic acids that are distinct from those bound by H5 avian influenza virus hemagglutinin, J Virol 81, 8341–8345. 7. Kumari, K., Gulati, S., Smith, D. F., Gulati, U., Cummings, R. D., and Air, G. M. (2007) Receptor binding specificity of recent human H3N2 influenza viruses, Virol J 4, 42. 8. Gulati, S., Smith, D. F., and Air, G. M. (2009) Deletions of neuraminidase and resistance to oseltamivir may be a consequence of restricted

267

receptor specificity in recent H3N2 influenza viruses, Virol J 6, 22. 9. Lepenies, B., and Seeberger, P. H. (2010) The promise of glycomics, glycan arrays and carbohydrate-based vaccines, Immunopharmacol Immunotoxicol 32, 196–207. 10. Liang, C. H., and Wu, C. Y. (2009) Glycan array: a powerful tool for glycomics studies, Expert Rev Proteomics 6, 631–645. 11. Disney, M. D., and Seeberger, P. H. (2004) The use of carbohydrate microarrays to study carbohydrate-cell interactions and to detect pathogens, Chem Biol 11, 1701–1707. 12. Liu, C., Eichelberger, M. C., Compans, R. W., and Air, G. M. (1995) Influenza type A virus neuraminidase does not play a role in viral entry, replication, assembly, or budding, J Virol 69, 1099–1106. 13. Henrickson, K. J. (1995) Human Parainfluenza Viruses, In Diagnostic Procedures for Viral, Rickettsial, and Chlamydial Infections (Lennette, E. H., Lennette, D. A., and Lennette, E. T., Eds.), pp 481–494, American Public Health Association, Washington, DC. 14. WHO Manual on Animal Influenza Diagnosis and Surveillance, World Health Organization.

Chapter 19 MALDI-ToF MS Analysis of Glycosyltransferase Activities on Gold Surface Arrays Nicolas Laurent, Rose Haddoub, Josef Voglmeir, and Sabine L. Flitsch Abstract Glycan-processing enzymes such as glycosyltransferases and glycosidases are responsible for the makeup of the glycome. The definition of their substrate specificities is, therefore, a central task in glycomics. In addition, these enzymes are themselves useful synthetic tools for the generation of complex carbohydrate structures as an alternative to tedious chemical synthesis. There has been great interest in using microarrays for studying these glycoenzymes because it allows the specificity of the enzyme to be probed against a panel of immobilized potential substrates, and also expands the repertoire of sugar arrays available for further carbohydrate–protein interaction studies. In particular, self-assembled monolayers (SAMs) of alkanethiols on gold surfaces have proven to be a valuable platform for such studies due to their robustness and their biocompatible, well-defined structure. Furthermore, a direct observation of the change in mass of immobilized substrates due to enzymatic processing is possible through label-free MALDI-ToF MS (matrix-assisted laser desorption/ionization time-of-flight mass spectrometry) technique. In this chapter, we describe the preparation of SAMs-coated gold surface arrays presenting carbohydrate or (glyco)peptide substrates, either pre-formed or directly synthesized on-chip, and MALDI-ToF MS analysis of glycosyltransferase activities on these immobilized substrates. Key words: Carbohydrate arrays, (Glyco)peptide arrays, Glycosyltransferases, Self-assembled monolayers, MALDI-ToF MS, SPOT synthesis

1. Introduction Over recent years, the need for miniaturization and automated high-throughput screening platforms has led to the development of several carbohydrate arrays, which required only very small amounts of materials (as opposed to solution-phase assays) and are amenable to detailed analysis of their implications in biological processes (1). First generations of glycan arrays used fluorescence detection of carbohydrate–protein interactions (2). More recently,

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_19, © Springer Science+Business Media, LLC 2012

269

270

N. Laurent et al.

label-free analytical methods such as surface plasmon resonance (SPR) (3) and MALDI-ToF MS (4) on self-assembled monolayers (SAMs)-coated gold surfaces have been introduced as versatile techniques for monitoring protein binding and enzyme activities on immobilized substrates. In particular, MALDI-ToF MS offers great flexibility as it does not require the use of labelled probes, such as fluorescent markers or radiolabels. It is, therefore, possible to monitor any transformation occurring on a surface through direct observation of the change in mass due to enzymatic reaction and also allows multiplexing. Recently, the robustness of SAMcoated gold platform has been further demonstrated by the development of an on-chip synthesis method of peptide and glycopeptide arrays (so-called “SPOT synthesis”) (5) suitable for probing glycosyltransferase substrate specificities in array format. This chapter presents first the preparation of a 64-well SAM-coated gold surface presenting a terminal N-hydroxysuccinimidyl (NHS)-activated ester suitable for substrate immobilization (i.e. 2-aminoethyl glycosides (6) or lysine-containing peptides (7)). A detailed protocol for on-chip SPOT synthesis of peptide and glycopeptide arrays is then given. Finally, MALDI-ToF MS analysis of glycosyltransferase activities on such a surface is illustrated with the recombinant bovine b1,4-galactosyltransferase (b1,4-GalT) and human polypeptide GalNAc-transferase (ppGalNAcT) enzymes.

2. Materials 2.1. Preparation of NHS-Activated SAMs-Coated Gold Slides

1. Disposable 64-well gold surface (Applied Biosystems). 2. Concentrated sulphuric acid. 3. Hydrogen peroxide 33%. 4. Deionized water. 5. Carboxylic acid-terminated alkanethiol (HS(CH2)17(OCH2 CH2)6OCH2COOH, Prochimia, Poland, stored at −20°C). 6. Tri(ethyleneglycol)-terminated alkanethiol (HS(CH2)17(OCH2 CH2)3OH, Prochimia, Poland, stored at −20°C). 7. Dimethyl sulfoxide (DMSO, anhydrous). 8. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC). 9. N-hydroxysuccinimide (NHS). 10. N,N-dimethylformamide (DMF, peptide synthesis grade, ROMIL). 11. Ethanol (Anhydrous). 12. Nitrogen cylinder.

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities…

2.2. Immobilization of Synthetic 2-Aminoethyl Glycosides and Lysine-Terminated Peptides

271

1. Freshly prepared NHS-activated SAMs-coated gold surface (see Subheading 3.1). 2. 100 mM Sodium phosphate-buffered saline pH 8 (PBS). 3. 2-Aminoethyl glycosides (see Note 1). 4. Lysine-terminated synthetic peptides. 5. 2 M Sodium hydroxide solution (NaOH). 6. Ethanol (Anhydrous). 7. Deionized water. 8. Nitrogen cylinder.

2.3. SPOT Synthesis of Peptides and Glycopeptides on Gold Surface Arrays

1. Freshly prepared NHS-activated SAMs-coated gold surface (see Subheading 3.1). 2. N-Fmoc diaminobutane (Aldrich). 3. N-Fmoc amino acids (Novabiochem). 4. Benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate (PyBOP, Novabiochem). 5. N,N-diisopropylethylamine (DIPEA). 6. 2,4,6-Trimethylpyridine (TMP). 7. N-Fmoc per-O-acetyl glycosylamino acid (see Note 2). 8. Piperidine. 9. Trifluoroacetic acid (TFA). 10. N,N-dimethylformamide (DMF, peptide synthesis grade). 11. Dichloromethane (Anhydrous). 12. Sodium. 13. Methanol (Anhydrous). 14. Acetic anhydride. 15. Ethanol (Anhydrous). 16. Deionized water. 17. M HCl in water. 18. 100 mM Sodium phosphate-buffered saline pH 8 (PBS). 19. Nitrogen cylinder.

2.4. Expression of Recombinant Glycosyltransferases 2.4.1. Bovine Galactosyltransferase I (b1,4GalT) in Escherichia Coli

1. E. coli strain BL21 (DE3) containing the b1,4GalT encoding plasmid (template DNA kindly provided by Dr. Dubravko Rendic, Vienna). 2. LB broth medium (Aldrich). 3. Kanamycin: 50 mg/mL stock solution in water. 4. Isopropyl b-d-1-thiogalactopyranoside (IPTG): 1  M stock solution in water. 5. Phenylmethylsulfonyl fluoride (PMSF): 100  mg/mL stock solution in 2-propanol.

272

N. Laurent et al.

6. Lysis buffer: 50  mM Tris–HCl, 500  mM NaCl, 1% Triton X100, pH 8. 7. Binding buffer: 50 mM Tris–HCl, 500 mM NaCl, 6 M Urea, 10 mM imidazole, pH 8. 8. Elution buffer: 50 mM Tris–HCl, 500 mM NaCl, 6 M Urea, 250 mM imidazole, pH 8. 9. Refolding buffer: 50 mM Tris–HCl, 500 mM NaCl, 1% protease inhibitor cocktail (Aldrich), 1% Triton X100, 8 mM reduced glutathione, 1 mM oxidized glutathione, pH 8. 10. Sonicator Soniprep 150 (MSE). 11. Syringe filter, 0.45 mm pore size (Sartorius). 12. Glycerol. 13. Nickel-chelation column Akta Explorer 100 (GE Healthcare). 14. HiTrap Chelating HP 1 mL FPLC-column (GE Healthcare). 15. Centrifugal concentrators (Vivaspin 20, 30  kDa MWCO, Aldrich). 2.4.2. Human ppGalNAcT2 and T13 in Pichia pastoris

1. P. pastoris colony containing the ppGalNAcT2 or T13 gene (template DNA from Geneservice Ltd, UK). 2. 10× YNB medium: 34  g/L YNB without (NH4)2SO4 w/o Amino acids, 100 g/L (NH4)2SO4. 3. BMGY medium: 20 g/L bacteriological peptone, 10 g/L yeast extract, 100 mL/L potassium phosphate buffer (1 M, pH 6), 100 mL/L YNB medium, 0.4 mg/L biotin, 10 g/L glycerol, 100 mg/L Zeocin. 4. BMMY medium: 20  g/L bacteriological peptone, 10  g/L yeast extract, 100  mL/L potassium phosphate buffer (1  M, pH 6.0), 100 mL/L YNB medium, 0.4 mg/L biotin, 10 g/L methanol. 5. Exchange buffer: 40  mM AMPD (2-Amino-2-methyl-1, 3-propandiol), pH 7.4 (store at +4°C). 6. Methanol. 7. Syringe filter, 0.45 mm pore size (Sartorius). 8. Glycerol. 9. Centrifugal concentrators Vivaspin 20, 30  kDa MWCO (Aldrich).

2.5. Determination of Enzyme-Specific Activity

1. Recombinant enzyme, concentrated (see Subheading 3.4).

2.5.1. Bovine b1,4-GalT

3. Enzyme substrate: 10 mM GlcNAc.

2. 400 mM AMPD buffer (2-amino-2-methyl-1,3-propanediol), pH 7.4 (store at +4°C). 4. 5 mM Uridyldiphosphate-Galactose (UDP-Gal) in water (store at +4°C).

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities…

273

5. 1 M MnCl2 in water (store at +4°C). 6. Deionized water. 7. 1 mM Ethylenediaminetetraacetic acid (EDTA). 8. Reference product: 1 mM Galb1,4GlcNAc (Aldrich). 9. 1100 Series HPLC (Agilent) equipped with photodiode array. 10. Aminex-87H HPLC column (Bio-Rad). 11. 5 mM sulphuric acid in water. 2.5.2. Human ppGalNAcT2 and ppGalNAcT13

1. Recombinant enzyme, concentrated (see Subheading 3.4). 2. 400 mM AMPD buffer (2-amino-2-methyl-1,3-propanediol), pH 7.4 (store at +4°C). 3. Enzyme substrate: 10  mM GAGAPGPTPGPAGAGAK synthetic peptide. 4. 5  mM Uridyldiphosphate-GalNAc (UDP-GalNac) in water (store at +4°C). 5. 1 M MnCl2 in water (store at +4°C). 6. Deionized water. 7. 1 mM Ethylenediaminetetraacetic acid (EDTA). 8. Reference product: PGPAGAGAK.

1 

mM

GAGAPGPT(aGalNAc)

9. 1100 Series HPLC (Agilent) equipped with photodiode array. 10. ODS Hypersil 5 mm column (0.4 × 25 cm, Phenomenex). 11. Solvent A: 0.1% trifluoroacetic in water (HPLC grade, ROMIL). 12. Solvent B: 0.1% trifluoroacetic in acetonitrile (HPLC grade, ROMIL). 2.6. Enzymatic Reaction on Gold Surface-Bound Substrates

1. Substrate arrays (as prepared according to Subheadings 3.2 or 3.3). 2. 400 mM AMPD buffer (2-amino-2-methyl-1,3-propanediol), pH 7.4 (store at +4°C). 3. 5  mM Uridyldiphosphate-Gal/GalNAc (UDP-Gal/GalNAc) in water (store at +4°C). 4. Recombinant glycosyltransferase, concentrated (see Sub­ heading 3.4). 5. 1 M Manganese chloride (MnCl2) in water (store at +4°C). 6. Deionized water. 7. Ethanol (anhydrous). 8. Nitrogen cylinder.

274

N. Laurent et al.

2.7. MALDI-ToF MS Analysis

1. Voyager-DE STR MALDI-ToF mass spectrometer, 337  nm nitrogen laser (Applied Biosystems). 2. Sample holder adaptor for disposable gold slide (Applied Biosystems). 3. 2,4,6-Trihydroxyacetophenone (THAP, MALDI-ToF MS grade, Aldrich). 4. Acetone (Anhydrous).

3. Methods 3.1. Preparation of NHS-Activated SAMs-Coated Gold Slides

1. Prepare a 5:1 (v/v) solution of concentrated sulphuric acid and hydrogen peroxide (caution: strong oxidizing agent). 2. Dip the gold plate into the above solution for 10  min, then thoroughly rinse with deionized water, ethanol and dry under a stream of nitrogen. 3. Prepare stock solution of carboxylic acid-terminated and tri(ethyleneglycol)-terminated alkanethiols at 0.2  mg/mL in DMSO. Mix the two solutions at a 1:1 ratio (see Note 3). 4. Spot 0.4 mL of the above solution of alkanethiols into each well of the gold slide. Store the plate overnight in the dark. 5. Rinse the plate with DMSO, ethanol and dry under a stream of nitrogen (see Note 4). 6. Dissolve 38 mg of EDC in 1 mL of dry DMF, and then add 6 mg of NHS. Immediately spot 0.4 mL of the resulting solution into each well of the gold slide. Leave the plate standing for 1 h, rinse with DMF, ethanol and dry under a stream of nitrogen.

3.2. Immobilization of Synthetic 2-Aminoethyl Glycosides and Lysine-Terminated Peptides

1. Prepare 50 mM solutions of aminoethyl glycosides or lysineterminated peptides in 100  mM PBS pH 8 (see Note 5). If necessary, adjust the pH to 8 with a 2 M solution of NaOH (see Note 6). 2. Spot 0.4 mL of each solution into corresponding wells of the NHS-activated gold slide as prepared in Subheading 3.1. Leave the plate standing overnight in a close chamber with soaked paper (to prevent evaporation of the spots), then rinse with PBS, deionized water, ethanol and dry under a stream of nitrogen. An example of MALDI-ToF MS spectrum of N-acetyl-dglucosamine (GlcNAc) immobilized according to Subheading 3.2 is given in Fig. 1. Table 1 shows the results of array screening of immobilized mono- and disaccharides against the bovine b1.4GalT

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities… 100

80

% Intensity

70

1

2

60

40 30

10 0 800

S – C17 – EG6 – COOH S – C17 – EG3 - OH

3 S – C17 – EG6 – CONH – C2 - GlcNAc S – C17 – EG3 - OH

50

20

229.5

S – C17 – EG3 - OH 1 S – C17 – EG3 - OH

3

90

275

2

4 S – C17 – EG6 – CONH – C2 - GlcNAc S – C17 – EG6 – CONH – C2 - GlcNAc

4 1340

1880

Mass (m/z)

2420

2960

0 3500

Fig. 1. MALDI-ToF MS analysis of immobilized 2-aminoethyl-2-acetamido-2-deoxy-b-d-glucopyranoside (GlcNAc) on gold surface. Structures of the identified species are indicated on the right panel. (1) m/z 861, (2) m/z 1,051, (3) m/z 1,298, and (4) m/z 1,734 for the sodium adduct of the species 1, 2, 3, and 4, respectively. It is to be noted that alkanethiols are generally observed as symmetrical or mixed disulphides under MALDI-ToF MS analysis conditions, and therefore several species corresponding to a single immobilized probe can be detected.

Table 1 MALDI-ToF MS analysis of 2-aminoethylglycosides immobilized on a gold surface array and detection of enzymatic transformation by the bovine galactosyltransferase I (b1,4GalT) SPOT

Sugar

Detected m/z

+b1,4GalT treatment

1

b-d-Glc

1,256

1,256

2

b-d-Gal

1,256

1,256

3

a-d-Man

1,256

1,256

4

b-d-GlcNAc

1,298

1,460

5

b-d-Xyl

1,226

1,226

6

a-d-Xyl

1,226

1,226

7

a-d-Glc(1,4)Glc

1,419

1,419

8

b-d-GlcNAc(1,4)Glc

1,460

1,622

9

b-d-Gal(1,4)Glc

1,419

1,419

For clarity, only m/z ratios corresponding to the sodium adduct of the mixed disulphide formed by the sugar-terminated alkanethiol and the tri(ethyleneglycol) alkanethiol are indicated. Glycoside substrates of the enzyme (i.e. spots 4 and 8, highlighted) are detected by the shift in m/z ratio of +162 corresponding to the addition of a terminal galactose

276

N. Laurent et al.

Table 2 Comparison of the substrate specificities of two human isoforms of the ppGalNAcT family on a panel of peptides Peptide

Detected m/z

+ ppGalNAcT2 treatment

+ ppGalNAcT13 treatment

1

GAGAPGPTPGPAGAGK

2,295

2,498 (+1)

2,498 (+1)

2

GAGAPGPSPGPAGAGK

2,281

2,484 (+1)

2,281 (0) 2,484 (+1)

3

GAPGPTPGPAGK

2,039

2,242 (+1)

2,242 (+1)

4

AcPTPGPAGK

1,799

2,018 (+1)

1,799 (0) 2,018** (+1)

5

GAHGSTAPPAGK

2,043

2,246 (+1)

2,246 (+1)

6

GAHVTSAPAGK

2,085

2,085 (0) 2,288 (+1)

2,288 (+1)

7

GAAPDTRPAAGK

2,144

2,144 (0)

2,144 (0)

8

YSPTSPSKR

2,033

2,236 (+1)

2,236* (+1)

9

PTTDSTTPAPTTK

2,350

2,756 (+2) 2,959 (+3) 3,162 (+4)

2,350 (0) 2,756 (+2) 2,959 (+3) 3,162 (+4)

2,246* (0)

2,246* (0)

SPOT

10

DDDSIEGSGGRK

**

*

2,246*

*

Peptides were immobilized on a gold slide through their C-terminal lysine residue following the procedure described in Subheading 3.2. After incubation with either ppGalNAcT2 or ppGalNAcT3 according to Subheading 3.5, MALDIToF MS analysis was performed to determine enzyme substrate preferences. For clarity, only m/z ratios corresponding to the sodium adduct of the mixed disulphide formed by the (glyco)peptide-terminated alkanethiol and the tri(ethyleneglycol) alkanethiol are indicated except for *proton adduct and **potassium adduct of the species was then main signal detected. MALDI-ToF MS analysis of the array after incubation with the enzyme allowed observing the number of GalNAc residues added onto the substrate (indicated in brackets, each GalNAc added corresponding to a shift in m/z ratio of +203)

enzyme (see Subheading  3.6 for details of enzymatic reaction on gold-bound substrates). Screening of human polypeptide N-acetylgalactosaminyltransferase enzymes (ppGalNAcT) against a panel of immobilized peptides is illustrated in Table  2. Figure  2 represents the MALDI-ToF MS spectra of a multiplexing experiment (see figure caption for experimental details). 3.3. SPOT Synthesis of Peptides and Glycopeptides on Gold Surface Arrays

The following section describes the on-chip synthesis of (glyco) peptide arrays using a SPOT synthesis methodology (illustrated in Fig. 3) on amino-functionalized SAM-coated gold surfaces. This procedure is an alternative to the immobilization of pre-synthesized peptide libraries described in Subheading 3.2 and is more amenable to high-throughput assay and automation. The synthesis uses

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities…

a

2123.9345

100 90

2125.2414

100

80

90

70

80

60

ppGalNAcT2

70 60

50 40

50 40 30

30 2046.0915 2047.1949

20 10

1800

277

2147.2306 2127.2353

20

2146.1061

10

2025.0862

2040

2280

2125.9439

2146.9257

2247.9919

2091.9221

1800

m/z

αGalNAc

2145.9039

2040

2280

m/z

b 2088.3103

100

90

80

80

70

70

60

2089.3671

50 40

2067.3834

30

20

2110.2899

2088.1366

αGalNAc

2065.0867

20

2146.4111

10

2087.0779

60

2066.2754

30

1800

ppGalNAcT2

2125.3550

50 40

2124.1335

100

90

2126.1520 2290.2399

10

2040

2280

m/z

1800

2040

2280

m/z

Fig. 2. Multiplex analysis of ppGalNAcT2 substrate preference. Peptides 5, 6, 7 (see Table 2) derived from the Muc1 tandemrepeat region were co-immobilized on SAMs-coated gold plate within a single spot. Following incubation with the enzyme in the presence of UDP-GalNAc, MALDI-ToF MS analysis is performed to detect enzymatic glycosylation. As shown in (a), peptide 5 (white circle) is fully converted into its glycopeptide product by the enzyme, whereas peptide 7 (grey lozenge) is not a substrate for the enzyme. In the second experiment (b), peptide 7 is co-immobilized with peptide 6 (black circle). A partial glycosylation of 6 is then detected upon enzymatic treatment. This preferential order of glycosylation of Muc1 by ppGalNAcT2 (i.e. 5 > 6 > 7) is in good agreement with data obtained using a one-spot one-peptide experiment as in Table 2 and in previously published solution-phase assays (8).

1 R = OH i

i: EDC, NHS, DMF, 1h, rt ; ii: H2N(CH2)4NHFmoc, DMF, 16h, rt ; iii: 20% piperidine, DMF, 10 min, rt; iv: a) SPOT synthesis (Fmoc amino acid, PyBOP, DIPEA, DMF then 20% piperidine, DMF), b) 50% TFA, DCM

O 2R=O N

ii

O 3 R = HN

NHFmoc

iii 4 R = HN

NH2

iv

O 5 R = HN

N H

R

H N R

NH2 O

n

Fig. 3. Strategy for SPOT synthesis of (glyco)peptide substrates on a 64-well gold slide. Starting from mixed SAMs of carboxylic acid-terminated and tri(ethyleneglycol)-terminated alkanethiols (as prepared in Subheading 3.1), introduction of a 1,4-diaminobutane linker suitable for covalent attachment of Fmoc-amino acid is performed by (i) EDC/NHS activation of the carboxylic acid, (ii) coupling of N-Fmoc diaminobutane followed by (iii) N-Fmoc cleavage. Repeating cycles of Fmocamino acid coupling, N-Fmoc cleavage and final TFA-mediated side-chain deprotection under standard peptide synthesis conditions (iv) afforded the peptide library suitable for probing glycosyltransferase activities.

278

N. Laurent et al.

commercially available N-Fmoc amino acids (further protected by acid-labile protecting groups on their side-chain residues). Glycopeptide synthesis requires synthetic N-Fmoc glycosylated amino acid building blocks carrying O-acetyl protecting groups on the sugar moiety (see Note 2). 1. Prepare a 100  mM solution of N-Fmoc diaminobutane in DMF. 2. Spot the above solution into each well of the freshly prepared NHS-activated SAMs-coated gold surface (see Subheading 3.1). 3. Leave the plate standing overnight in the dark, then rinse with DMF, ethanol and dry under a stream of nitrogen. 4. Dip the slide into 10 mL of a 20% piperidine solution in DMF. 5. Leave the plate standing for 10  min, then rinse with DMF, ethanol and dry under a stream of nitrogen. 6. Prepare a fresh solution of activating reagent consisting of 52 mg of PyBOP and 36 mL of DIPEA in 0.5 mL dry DMF (solution A). Note that for the coupling of N-Fmoc per-Oacetyl glycosylamino acid, an activating reagent consisting of 52  mg of PyBOP and 14  mL of 2,4,6-trimethylpyridine in 1 mL DMF (solution B) is preferred above the PyBOP/DIPEA mixture (see Note 7). 7. Prepare stock solution of N-Fmoc amino acid (200  mg in 2.5 mL DMF) (see Note 8). 8. Mix 100  mL of stock solution of N-Fmoc amino acid and 100  mL of activating reagent (solution A). In the case of N-Fmoc per-O-acetyl glycosylamino acid, dissolve 10 mmol of compound in 126 mL of activating reagent B (see Note 9). 9. Spot 0.4 mL of activated amino acid mix into the corresponding wells of the gold slide. 10. Place the slide into a sealed container and leave for standing at 37°C for 1 h. 11. Rinse with DMF, ethanol and dry under a stream of nitrogen (see Note 10). 12. Optional: a capping step of eventual unreacted amine can be performed at this point by dipping the slide into pure acetic anhydride and allowing standing for 10 min, followed by rinsing with ethanol and drying under a stream of nitrogen. 13. Dip the slide into 10 mL of 20% piperidine solution in DMF. 14. Leave the plate standing for 10  min, then rinse with DMF, ethanol and dry under a stream of nitrogen. 15. Repeat steps 8–13 until completion of the peptide synthesis. 16. Carry out side-chain deprotection by dipping the slide into 10 mL of a 50% TFA solution in dichloromethane, in a sealed container to avoid evaporation. Leave the plate standing for

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities…

279

+1 position : AHGVTxAPA 2

-1 position : AHGxTSAPA

1,8 1,6

S P R D I G M H F Y Q V E T N A

Relative intensity

1,4 1,2 1 0,8 0,6 0,4 0,2 0

/min

Fig. 4. Influence of the neighbouring amino acid on the enzymatic glycosylation of peptides by ppGalNAcT2. An array of peptides derived from the Muc1 sequence AHGVTSAPA containing varied amino acid at the ±1 position relative to the threonine glycosylation site (underlined) was prepared by SPOT synthesis. This array was then used to probe the activity of ppGalNAcT2. As each pair of peptide/glycopeptide differs only by a GalNAc unit, relative conversion for each peptide can be determined from the MALDI-ToF MS signals of the glycopeptide product and the parenting peptide substrate (r = SI(glycopeptide)/(SI(glycopeptide) + SI(peptide)), where r is the relative conversion, SI(glycopeptide) is the sum of the MS intensities of the glycopeptides product, and SI(peptide) is the sum of the MS intensities of the peptide substrate. Values are normalized at 1 for the unmodified AHGVTSAPA substrate. The results obtained showed a strong influence of the amino acid at the −1 position: bulky and/or charged amino acid reduced the extent of glycosylation, whereas a proline residue resulted in a strong enhancement of the ppGalNAcT2 activity. Noteworthy, these results are in good agreement with previously reported solution-phase studies (10). Furthermore, when the original serine at the +1 position was replaced by a threonine, an MS signal corresponding to a glycopeptide bearing two GalNAc residues was detected at the +1 position (indicated by a star on the above graph).

4 h, then rinse successively with ethanol, PBS pH 8, deionized water, ethanol and dry under a stream of nitrogen. 17. For glycopeptides containing O-acetylated sugars, de-­O-­acetylation is performed by dipping the slide into a solution prepared with 17 mg of sodium in 10 mL of methanol. Leave the plate standing for 30 min, then rinse with ethanol, 0.1 M HCl, deionized water, ethanol and dry under a stream of nitrogen. Example of enzyme activity assay using the recombinant ppGalNAcT2 and an array of peptide generated by SPOT synthesis according to Subheading 3.3 is illustrated in Fig. 4 (see caption for experimental details). 3.4. Expression of Recombinant Glycosyltrans ferases

Recombinant glycosyltransferases were produced in different expression systems. The bovine galactosyltransferase I was expressed in soluble form in E.coli strain BL21 (DE3), based on a method described by Boeggeman et  al. (9). The ppGalNAcT2 and T13

280

N. Laurent et al.

require posttranslational modifications to fold correctly and were, therefore, produced in eukaryotic host P. pastoris. 3.4.1. Bovine Galactosyl transferase I (b1,4GalT) in E. coli

1. Incubate E. coli strain BL21 (DE3) containing the galactosyltransferase encoding plasmid for 16  h at 37°C in 7  mL LB medium containing 50 mg/mL kanamycin with shaking. 2. Transfer 1 mL of the overnight culture into a 2 L Erlenmeyer flask filled with 400 mL of LB medium. Grow at 37°C with shaking until reaching an OD600 value of 0.5. 3. Induction of the culture is performed using IPTG (final concentration 1 mM) for 6 h at 30°C. 4. Harvest the cells by centrifugation (5,000 × g, 10  min, 4  C) and resuspend pellet in 40  mL Lysis buffer; Add 400  mL of PMSF solution. 5. Lyse cells by sonication (30 min with 10 s on-/10 s off-cycles, 15 mm amplitude). 6. After centrifugation (30,000 × g, 30 min, 4°C), discard supernatant and wash debris again in 40  mL of Lysis buffer, Centrifuge again, and resuspend the washed debris in 20 mL of Binding buffer. 7. Dissolve inclusion bodies in the debris by gentle shacking on a laboratory rocker for 16 h at 4°C. 8. Centrifuge the debris suspension (30,000 × g, 30  min, 4°C) and remove particles with a filter syringe (0.45 mm pore size). 9. Apply filtrate on a Nickel-chelation column. Wash column with 25 mL Binding buffer to remove unbound proteins from the column. 10. Elute dissolved inclusion bodies with 10 mL of Elution buffer from the Nickel-chelation column. 11. Add the 10 mL from the elution fraction dropwise to 100 mL of Refolding buffer, and stir for 16 h at 4°C. 12. Concentrate samples 100-fold using centrifugal concentrators at 4°C. 13. Add 100 mL of glycerol to the concentrates and store 20 mL aliquots at −80°C.

3.4.2. Human ppGalNAcT2 and T13 in P. pastoris

1. Transfer a P. pastoris colony containing ppGalNAcT2 or T13 integrated into the genome into a 250  mL Erlenmeyer flask containing 50 mL BMGY medium, and incubate for 24 h at 30°C in a shaking incubator at 250 rpm. 2. Measure OD600 (normally between 8 and 20). To obtain a 400  mL final volume with an OD600 value of 1, transfer the necessary amount (i.e. 27 mL if OD600 was 15) into a falcon

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities…

281

tube, and remove medium by centrifugation (1,500 × g, 5 min, 4°C). Wash the pellet with BMMY medium by centrifugation (1,500 × g, 5  min, 4°C), discard the supernatant and finally resuspend the pellet in 400 mL of BMMY medium. 3. Transfer culture in a 2 L Erlenmeyer flask. 4. Incubate for 96 h at 16°C in the shaking incubator at 250 rpm and add 4 mL of methanol to the culture every 24 h. 5. Spin off cells by centrifugation (5,000 × g, 10 min, 4°C) and filter the supernatant with a particle filter (0.45 mm pore size). 6. Concentrate the filtrate 200-fold using centrifugal concentrators at 4°C to a final volume of 2 mL. Add 20 mL Exchange buffer and concentrate again down to 2 mL. 7. Store samples after the addition of 200 mL glycerol in 20 mL aliquots at −80°C. 3.5. Determination of Enzyme-Specific Activity

1. Prepare 20  mL of enzyme reaction mixture containing 2  mL AMPD buffer, 2 mL enzyme substrate, 4 mL UDP-Gal, 0.2 mL MnCl2, 7.8 mL enzyme concentrate, and 4 mL water.

3.5.1. Bovine b1,4-GalT

2. Incubate at 37°C for 30 min, and then stop the reaction by adding 0.5 mL EDTA solution. 3. Analyse reaction mixture by HPLC on aminex 87H column using 5 mM sulphuric acid isocratic eluent (0.6 mL/min flow rate) at 40°C. Elution is monitored at 206 nm. 4. Determine relative conversion rate using enzyme substrate (retention time 11.8  min) and reference product (retention time 9.1 min) as standards. Conversion rate Cv = (peak area of glycosylated product)/[(peak area of glycosylated product) + (peak area of starting material)]. 5. Calculate the enzyme-specific activity (mU/mL). 1 mU/mL corresponds to 1 nmol of substrate converted per minute and per mL of enzyme concentrate in the assay mixture. Specific activity (mU/mL) a = (Cv × ns)/(t × Ve), where Cv is the relative conversion rate (as determined in step 4), ns the amount of substrate in the reaction mixture in nmol (here ns = 1  mM × 20  mL = 20  nmol), t the reaction time in min (t = 30 min), and Ve the volume of enzyme concentrate in mL (7.8 × 10−3 mL). Here, a = 85.47 × Cv. b1,4-GalT-specific activity was found 22.5 mU/mL.

3.5.2. Human ppGalNAcT2 and ppGalNAcT13

1. Prepare 20  mL of enzyme reaction mixture containing 2  mL AMPD buffer, 2  mL enzyme substrate, 4  mL UDP-GalNAc, 0.2 mL MnCl2, 0.8 mL enzyme concentrate, and 11 mL water. 2. Incubate at 37°C for 120 min, and then stop the reaction by adding 0.5 mL EDTA solution.

282

N. Laurent et al.

3. Analyse reaction mixture by HPLC on ODS Hypersil column using a linear gradient of solvent A containing 10–15% of solvent B from 0 to 9 min at a 1 mL/min flow rate. Elution is monitored at 214 nm. 4. Determine relative conversion rate using enzyme substrate (retention time 7.6  min) and reference product (retention time 6.8 min) as standards. Conversion rate Cv = (peak area of glycosylated product)/[(peak area of glycosylated product) + (peak area of starting material)]. 5. Calculate the enzyme-specific activity (mU/mL). 1 mU/mL corresponds to 1 nmol of substrate converted per minute and per mL of enzyme concentrate in the assay mixture. Specific activity (mU/mL) a = (Cv × ns)/(t × Ve), where Cv is the relative conversion rate (as determined in step 4), ns the amount of substrate in the reaction mixture in nmol (here ns = 1  mM  × 20 mL = 20 nmol), t the reaction time in min (t = 120 min), and Ve the volume of enzyme concentrate in mL (0.8 × 10−3 mL). Here, a = 208,3 × Cv. Specific activities of 41  mU/mL and 4.5  mU/mL were found for the enzymes ppGalNAcT2 and ppGalNAcT13, respectively. 3.6. Enzymatic Reaction on Gold Surface-Bound Substrates

1. Prepare 100 mL of enzyme reaction mixture containing 10 mL AMPD-buffer, 20  mL UDP-Gal (for b1,4GalT reaction) or UDP-GalNAc (for ppGalNAcT reaction), 1 mL MnCl2, 39 mL H2O, and 30 mL enzyme concentrate. 2. Transfer 0.5 mL of solution onto each spot of the surface bound substrates. Enclose the slide into a sealed petri dish containing soaked paper to prevent evaporation. Incubate for 16 h at 37°C in static incubator. 3. Rinse with deionized water, ethanol and dry under a stream of nitrogen. 4. If MALDI-ToF MS analysis shows incomplete conversion, the enzymatic reaction can be repeated.

3.7. MALDI-ToF MS Analysis

1. Prepare a 10 mg/mL solution of THAP in acetone. 2. Spot 0.4 mL of the solution into each well of the gold slide. 3. Mount the gold slide into the sample holder. 4. Load the target in to the MALDI-ToF MS. 5. Acquire spectra using the following parameters: reflector or linear ToF, positive ion mode, 20  kV accelerating voltage, 200 ns (for reflector mode) or 225 ns (for linear mode) extraction delay, 100 acquisitions/spectrum (see Note 11).

19  MALDI-ToF MS Analysis of Glycosyltransferase Activities…

283

4. Notes 1. 2-Aminoethyl glycosides can be prepared according to a number of well-established procedures. For example, N-carbobenzoxyethanolamine is glycosylated using the appropriate per-O-acetylated glycosyl donor (i.e. halide or acetate). Subsequent deprotection of the sugar head-group under Zemplen conditions followed by hydrogenolysis of the N-Cbz protecting group give the desired glycoside suitable for immobilization. 2. N-Fmoc per-O-acetyl glycosylamino acids are synthesized according to standard procedures, for example by glycosylation of N-Fmoc l-serine (or threonine) benzyl ester with the appropriate per-O-acetyl glycosyl donor (i.e. acetate, halide, or trichloroacetimidate) followed by the removal of the benzyl protecting group. 3. Ratio may be adjusted. However, we found that a 1:1 ratio was suitable for easy MALDI-ToF MS detection without compromising complete enzymatic reactions. Stock solutions of alkanethiols can be aliquoted and stored at −20°C under inert atmosphere for several months. 4. SAMs-coated gold slides can be stored at 4°C for several weeks in a sealed container flushed with nitrogen or argon. However, it is not recommended to store the NHS-activated SAMscoated slides. 5. Clean coupling were also obtained with more dilute solutions (i.e. 1 mM). For complete solubilization of peptides, PBS containing 6 M guanidinium hydrochloride may be used. 6. This is especially important for synthetic peptides which are usually isolated as trifluoroacetate salts. 7. It was found that using PyBOP/2,4,6-trimethylpyridine allows to avoid acetyl migration from the sugar moiety onto the terminal amine of the peptide during coupling step. 8. Most of the N-Fmoc amino acid stock solutions can be kept at +4°C between each coupling step and −20°C for longer-term storage. However, we recommend discarding solutions after 24 h and preparing fresh ones. 9. It is not recommended to store this solution. Prepare fresh one for each coupling step. 10. A second coupling can be performed at this point to ensure complete reaction by repeating the steps 8–11. However, most reactions were found complete after 1 h at 37°C, as judged by MALDI-ToF MS monitoring. 11. Reflector mode is suitable for the detection of simple glycosides bound on SAMs-coated gold slides. Linear mode is more

284

N. Laurent et al.

appropriate for the analysis of peptide and glycopeptides arrays. For enhanced sensitivity, a low mass gate can be set up at m/z 1,000 to suppress background ions detection, in particular the signal at m/z 861 corresponding to the sodium adduct of the symmetrical disulphide formed by the tri(ethyleneglycol)terminated alkanethiol. References 1. For recent reviews, see Laurent, N., Voglmeir, J., and Flitsch, S. L. (2008) Glycoarrays – tools for determining protein-carbohydrate interactions and glycoenzyme specificity. Chem. Commun. 4400–4412. Horlacher, T., and Seeberger, P. H. (2008) Carbohydrate arrays as tools for research and diagnostics. Chem. Soc. Rev. 37, 1414–1422. Liu, Y., Palma, A. S., and Feizi, T. (2009) Carbohydrate microarrays: key developments in glycobiology. Biol. Chem. 390, 647–656. Wu, C.-Y., Liang, P.-H., and Wong, C.-H. (2009) New development of glycan arrays. Org. Biomol. Chem. 7, 2247–2254. 2. Blixt, O., Head, S., Mondala, T., Scalan, C., Huflejt, M., Alvarez, R., et al. (2004) Printed covalent array for ligand profiling of diverse glycan binding proteins. Proc. Natl. Acad. Sci. USA 49, 17033–17038. 3. Karamanska, R., Clarke, J., Blixt, O., MacRae, J. I., Zhang, J. Q. et al. (2008) Surface plasmon resonance imaging for real time, label-free analysis of protein interactions with carbohydrate microarrays. Glycoconj. J. 25, 69–74. 4. Su, J., and Mrksich, M. (2002) Using mass spectrometry to characterize self-assembled monolayers presenting peptides, proteins, and carbohydrates. Angew. Chem. Int. Ed. 41, 4715–4718. 5. Laurent, N., Haddoub, R., Voglmeir, J., Wong, S. C. C., Gaskell, S. J., and Flitsch, S. L. (2008) SPOT synthesis of peptide arrays on self-assembled monolayers and their evaluation

as enzyme substrates. ChemBioChem. 9, 2592–2596. 6. Zhi, Z.-l., Laurent, N., Powell, A. K., Karamanska, R., Fais, M., Voglmeir, J., et  al. (2008) A versatile gold surface approach for fabrication and interrogation of glycoarrays. ChemBiochem 9, 1568–1575. 7. Laurent, N., Voglmeir, J., Wright, A., Blackburn, J. M., Pham, N. T., Wong, S. C. C., Gaskell, S. J., and Flitsch, S. L. (2008) Enzymatic glycosylation of peptide arrays on gold surfaces. ChemBioChem. 9, 883–887. 8. Wandall, H. H., Hassan, H., Mirgorodskaya, E., Kristensen, A. K. (1997) Substrate specificities of three members of the human UDP-Nacetyl-alpha-D-galactosamine: polypeptide alpha-N-acetylgalactosaminyltransferase family, ppGalNAc-T1, -T2 and -T3. J. Biol. Chem. 272, 23503–23514. 9. Boeggeman, E. E., Balaji, P. V., Sethi, N., Masibay, A. S., Qasba, P. K. (1993) Expression of deletion constructs of bovine beta-1, 4-galactosyltransferase in Escherichia coli: importance of Cys134 for its activity. Prot. Eng. 6, 779–785. 10. Gerken, T. A., Raman, J., Fritz, T. A., Jamison, O. (2006) Identification of common and unique peptide substrate preferences for the UDP-GalNAc: polypeptide alpha-N-acetylgalactosaminyltransferases T1 and T2 derived from oriented random peptide substrates. J.  Biol. Chem. 281, 32403–32416.

Chapter 20 Microarray Technology Using Glycans Extracted from Natural Sources for Serum Antibody Fluorescent Detection Emanuela Lonardi, André M. Deelder, Manfred Wuhrer, and Crina I.A. Balog Abstract Glycan microarray technology enables the screening of large numbers of glycan samples for glycan–protein interactions, based on the presentation of immobilized glycans in a discrete pattern on a solid support. Here we describe a glycan microarray approach employing glycans enzymatically released from proteins and lipids of in vitro cultured cells and of human and animal tissues, followed by the detection of serum antibody binding. This approach may be used to detect autoantibodies in cancer as well as in autoimmune diseases. Key words: Autoantibody, Epoxy-silane, Glycolipids, HILIC, Mass spectrometry, N-glycans, Shotgun microarray

1. Introduction Many biological events capitalize on complex coded information in the form of glycan decoration patterns (1): protein folding and intracellular trafficking (2), cell–cell interactions and cell–matrix interactions (3), differentiation and development (4, 5), diseases and tumor metastasis (6–8), host–pathogens interactions (9), and many aspects of the immune response (10). The immunogenic properties of cancer have been extensively studied: alterations in the level or the type of glycosylation lead, in the case of cancer onset, to variants which are not covered by tolerance, resulting in humoral autoimmunity toward tumorassociated antigens (TAAs). Antibodies against these TAAs can be used to monitor cancer progression and eventually response to

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_20, © Springer Science+Business Media, LLC 2012

285

286

E. Lonardi et al.

therapy (11, 12). However, the prediction power of one distinct antibody may be misleading, since antibodies to specific antigens may be absent in particular patients. Glycan microarrays facilitate the simultaneous screening of serum antibodies against a variety of antigens. Moreover, the advent of microarray technologies may enable the diagnosis of a disease based on its immunological signature (the combination of antibodies developed against diseasespecific antigens) (13). With the recent introduction of glycan microarrays, researchers may now screen in parallel a multitude of glycan samples for glycan–protein interactions, on the basis of the presentation of immobilized glycans in a discrete pattern on solid support. The technique has greatly benefited from chemical and enzymatic synthesis of well-known glycan structures (14), a field that has evolved dramatically in the last few years (15, 16). However, the synthesis of glycans is a difficult, costly, and tedious process. In the case of natural glycans microarrays, not only is carbohydrate synthesis bypassed, but also researchers can probe a representation of the natural glycome repertoire which is derived from the target of analysis (cells, tissues, and organisms). This approach, known as shotgun glycomics, allows the researchers to pinpoint functionally important glycans before starting their detailed structural analysis. For this purpose, the production of natural glycan libraries is essential, and various methods (17, 18) have been developed to represent the glycan repertoire of specific biological materials. In our method (19, 20) glycans from glycoproteins and glycolipids of natural sources are enzymatically released and subsequently labeled with a multifunctional fluorescent dye (2-aminobenzoic acid, 2AA, which is also called anthranilic acid) by reductive amination. Labeled glycans are purified by HPLC and preliminarily analyzed by mass spectrometry. The attachment of the label to the glycan results in the formation of a secondary aromatic amine which is amenable for reaction with an epoxy-activated surface. The presentation of the derivatized glycans is therefore oriented, while the core glycan structure is present in open-ring configuration due to the addition of the 2AA tag. Glycan microarrays are printed on epoxy-silane slides and probed with carbohydrate-binding proteins (CBPs), in the form of lectins, purified antibodies, autoantibodies from sera, and other CBPs. Bound CBPs are then detected using secondary antibodies with a fluorescent label and the signals are integrated and plotted against the sample number, which results in unique fingerprints (Fig. 1). Glycan HPLC fractions recognized by CBPs are then fully structurally characterized, but because they may still contain a mixture of glycans, those fractions can be further purified by second-dimension HPLC and probed by a secondgeneration microarray to identify with high confidence individual interaction partners of CBPs which are then structurally characterized by tandem mass spectrometry.

20  Microarray Technology Using Glycans Extracted…

287

Fig. 1. Workflow for the production of a natural glycan microarray using AA-labeled glycans. (a) Glycans are derivatized using 2AA. (b) The fluorescence of the label is used to monitor fractionation. (c) Glycan composition of each fraction is determined via mass spectrometry. (d) Fractions are immobilized on epoxide-activated glass slides. (e) The microarray is screened with carbohydrate-binding proteins and bound carbohydrate-binding proteins are detected using fluorescently labeled secondary antibodies. Light circle galactose, dark circle mannose, square N-acetylglucosamine, triangle fucose, diamond N-acetylneuraminic acid, AA 2-aminobenzoic acid. Adapted from (19) with permission.

2. Materials 2.1. Sample Preparation (2AA Labeling)

Prepare all solutions using high quality deionized water (milliQ) and analytical grade reagents (see Note 1). Some buffers include the organic solvent acetonitrile (ACN) which should be added after the pH of the buffer has been set. Prepare and store all reagents at room temperature (unless indicated otherwise).

288

E. Lonardi et al.

1. In vitro-cultured human cells, or biopsy materials (not fixated). 2. Methanol/water 10/90, (v/v). 3. Methanol/water 50/50, (v/v). 4. Chloroform/methanol/water 10/10/1, (v/v/v). 5. Reconstitution buffer: 50 mM sodium acetate (pH 5.0) containing 0.1% sodium taurodeoxycholate. Dilute 340  mg of sodium acetate (CH3COONa·3H2O) and 50  mg of sodium taurodeoxycholate in 40 mL of milliQ. Mix and adjust the pH with glacial acetic acid to pH 5. Make up with milliQ to a total volume of 50 mL. 6. Recombinant endoglycoceramidase II (rEGCase II from Rhodococcus spec.; Takara Bio Inc., Shiga, Japan). 7. 5× PBS buffer (0.189 M Na2HPO4·2H2O, 3.5 mM KH2PO4, 1.45 M NaCl): prepare 2 L of 5× PBS buffer (0.175 M, pH 7.3): dissolve 57.0 g Na2HPO4·2 H2O, 4.76 g KH2PO4, and 85.0 g NaCl in 1.5 L milliQ. Adjust pH (if necessary) using HCl. Make up with milliQ to a total volume of 2 L. 8. 1× PBS buffer: 200 mL of 5× PBS in 800 mL of milliQ. 9. SDS (sodium dodecyl sulfate). 10. 2-Mercaptoethanol. 11. NP-40. 12. Sodium azide. 13. PNGase F (peptide N-glycosidase F; Roche Diagnostics, Mannheim, Germany). 14. Water/ACN 40/60, (v/v). 15. Water/ACN 90/10, (v/v). 16. Water/ACN 50/50, (v/v). 17. 0.1% TFA: 0.1 trifluoroacetic acid (v/v) in milliQ. 18. 0.1% TFA (v/v) in water/ACN 50/50. 19. Labeling solution: 0.35 M 2AA, 15% (v/v) glacial acetic, and 85% DMSO (v/v). Prepare a glacial acetic acid/DMSO (15/85) solution by adding 375  ml of glacial acetic acid to 2,125  ml of DMSO. Add 120  mg of 2-aminobenzoic acid (2AA). Mix by vortexing until completely dissolved. 20. Reducing solution: 1  M 2-picoline borane in DMSO. Add 2.5  mL DMSO to 297.5  mg 2-picoline borane complex (Sigma-Aldrich, St. Louis, MO, USA). Mix by vortexing until completely dissolved. 21. 77% ACN, (v/v). 22. Sonicator probe (Branson Sonifier 250, Gemini BV, The Netherlands).

20  Microarray Technology Using Glycans Extracted…

289

23. C18-RP-cartridges (500  mg; J.T.Baker, Deventer, The Netherlands). 24. Self-packed RP-cartridges: 4 g of Chromabond C18 (MachereyNagel, Dueren, Germany) in 15 mL reservoirs equipped with frits (PART# 210315 and 211412, Alltech Associates, Inc., Deerfield, IL, USA). 25. Graphitized carbon solid phase extraction (SPE) cartridges (150 mg Carbograph; Alltech). 26. HILIC column: TSK amide 80, 3  mm, 150 × 4.6  mm inner diameter (Tosoh Bioscience, Stuttgart, Germany). 27. HILIC Solvent A: ACN/Solvent B, 80/20 (v/v). 28. HILIC Solvent B: 50 mM ammonium formate pH 4.4. Prepare by dissolving 3.15 g ammonium formate in 1 L of milliQ. Mix and adjust the pH with formic acid (see Note 2). 29. 96-Well plates 500 ml (ABgene, Epsom, Surrey, UK). 2.2. Plate Setup and Printing

1. Labeled glycans produced as described in Subheading  3.1. Stored at −20°C long term. 2. Suitable positive controls: (a) Fluorescently labeled secondary antibodies: goat antiHuman IgG (Fc specific), Cy3 conjugated (Sigma), and Goat Anti-Human IgM (m chain specific), AlexaFluor® 647 conjugated (Invitrogen, Carlsbad, CA, USA). (b) Blood group antigens (Dextra, Reading, UK). 3. Polystyrene 384-well plates, V-bottom (Genetix, New Milton, UK. cat# X6004) and lids. 4. Sealing aluminum film (Thermowell™ Sealing tape, Corning, Amsterdam, The Netherlands). 5. SCHOTT Nexterion® Spot Buffer, 2× concentrated (SCHOTT Benelux B.V., Tiel, The Netherlands). 6. DMSO. 7. SCHOTT Nexterion® Slide E slides, coated with epoxysilane layer. Stored at room temperature in sealed original packaging. 8. ArrayIt® Stealth Micro SMP3 Spotting Pins (ArrayIt, Sunnyvale, CA, USA). The pins can be stored directly in the printhead for short periods of time or in the manufacturers’ box. 9. Omnigrid 100 microarrayer (Genomic Solutions, Ann Arbor, MI, USA).

2.3. Humidification/ Immobilization

1. Large sealable container (such as acrylic-glass container, plastic food container, and desiccator). 2. Test-tube racks. 3. Saturated NaCl solution.

290

E. Lonardi et al.

2.4. Assay

1. Printed slides. 2. Press-to-Seal™ silicone sheets (Invitrogen) cut to size to make suitable gaskets. 3. Petri dish. 4. Well plate shaker. 5. 10% Tween® 20: For a 100  mL solution, dilute 10  mL Tween®20 with milliQ to a final volume of 100 mL. 6. PBST1: 0.01% Tween®20 in 1× PBS. For a 1 L solution, take 200 mL of 5× PBS, dilute to 800 mL with milliQ, add 1 mL 10% Tween®20, bring to volume with milliQ. 7. PBST5: 0.05% Tween®20 in 1× PBS. For 1  L solution, take 200 mL of 5× PBS, dilute to 800 mL with milliQ, add 5 mL 10% Tween®20, bring to volume with milliQ. 8. Blocking solution: 4% (w/v) BSA (bovine serum albumine, further purified Fraction V, g globulin-free, powder, stored at 4°C; Sigma) in PBS, 50 mM ethanolamine. 9. Dilution buffer: PBST1, 1% (w/v) BSA. Sera are diluted 1:50. Lectins are applied at a concentration of 10 mg/mL unless otherwise specified, fluorescently labeled secondary antibodies are diluted 1:1,000. 10. Fluorescently labeled secondary antibodies: goat anti-Human IgG (Fc specific), Cy3 conjugated (Sigma) and Goat AntiHuman IgM (m chain specific), AlexaFluor® 647 conjugated (Invitrogen).

2.5. Scanning and Analysis

1. G2565BA scanner (Agilent Technologies, Santa Clara, CA), with dual laser (532, 635 nm) and 78 slides autoloader. 2. Analysis program Genepix Pro 6.1.

3. Methods 3.1. Sample Preparation

3.1.1. Tissue Homogenization and Isolation of Proteins and Glycolipids

The protocol illustrates the preparation of glycolipid glycans (called in the following) and N-glycans as exemplified in the following for one piece of tumor tissue (colorectal carcinoma biopsy material of approximately 0.9 cm3) (see Note 3). l-glycans

1. Cut the tissue in small pieces and homogenize these in 4 mL of milliQ with a sonicator probe (21). 2. Add 7 mL of methanol and homogenize the sample by vortexing and subsequently place it in a sonicator water bath for about 15 min. 3. Add 13 mL of chloroform (see Note 4) and repeat the homogenization and sonication step.

20  Microarray Technology Using Glycans Extracted…

291

4. Centrifuge the sample at 15,000 × g for about 15  min. This step results in a phase separation. 5. Remove the upper phase (keep the upper phase) and replace it by methanol/water (50/50, v/v). 6. Vortex, sonicate, and centrifuge the sample at 15,000 × g for about 15 min. 7. Repeat steps 5 and 6 two more times. 8. Pool the three upper phases (contain the glycolipids, further purification described in Subheadings  3.1.2 and 3.1.3) (see Note 5). 9. Add 1 mL of methanol to the lower phase and the inter-phase (containing most of the proteins). 10. Homogenize the sample by vortexing and subsequently place it in a sonicator water bath for about 15 min. 11. Centrifuge the sample at 15,000 × g for about 15  min. This step results in the pelleting of the protein fraction. 12. Discard the supernatant. 13. Add 1 mL of methanol and perform steps 10–12. 14. Repeat step 13. 15. Dry the pellet under a gentle stream of nitrogen (contains the proteins and glycoproteins, further purification described in Subheading 3.1.4) (see Note 6). 3.1.2. Purification of Glycolipids by Reverse Phase Solid Phase Extraction

1. Pre-wash two C18-RP-cartridges with 5 mL of methanol. 2. Equilibrate the cartridges with 5  mL of methanol/water (10/90, v/v). 3. Apply the upper phase containing the glycolipids (see Subheading 3.1.1, item 8) on one cartridge. 4. Wash with 10  mL of milliQ (keep the flow-through and wash). 5. Apply the combined flow-through and wash on the second C18-RP-cartridge. 6. Wash the second cartridge with 10 mL of milliQ. 7. Elute the bound glycolipids from both cartridges with 5 mL of chloroform/methanol/water (10/10/1, v/v/v). 8. Dry the glycolipid sample under a stream of nitrogen (see Note 7). 9. Reconstitute the sample in reconstitution buffer.

3.1.3. Enzymatic Release and Purification of Lipid Glycans

1. Sonicate the reconstituted sample for 10 min and subsequently incubate it for 10 min at 50°C. Sonicate the sample again for 10 min.

292

E. Lonardi et al.

2. Add 8 ml of recombinant endoglycoceramidase II to the sample and incubate overnight at 37°C (see Note 8). 3. Add another 8 ml of rEGCase II to the sample and incubate overnight at 37°C. 4. Pre-wash a C18-RP-cartridge with 5 mL of methanol. 5. Equilibrate the cartridge with 5 mL of milliQ. 6. Apply the sample. 7. Wash with 7 mL of milliQ. 8. Combine the flow-through and wash (they contain the glycan moieties derived from purified glycolipids). 9. Dry the pooled flow-through and wash either by freeze-drying overnight or by vacuum centrifugation. 3.1.4. Enzymatic Release and Purification of N-Glycans

1. Add 10  mL of PBS to the pellet (see Subheading  3.1.1, item 15). 2. Homogenize the sample by vortexing and subsequently place it in a sonicator water bath for about 15 min. 3. Add SDS and 2-mercaptoethanol to the sample to final concentrations of 1.3% (w/v) and 0.5% (v/v), respectively. 4. Heat the sample at 95°C for 10 min. 5. Transfer the sample on ice and allow cooling down. 6. Add NP-40 and sodium azide to the sample to a final concentration of 1.5% and 0.1%, respectively. 7. Add 12 mU of PNGase F and incubate the sample overnight at 37°C (see Note 9). 8. Add another 12  mU of PNGase F and incubate the sample overnight at 37°C. 9. Pre-wash a self-packed RP-cartridge with 30 mL of ACN and 15 mL of water/ACN (40/60, v/v). 10. Equilibrate the RP-cartridge with 30 mL of milliQ. 11. Apply the sample (released N-glycans) to the column. 12. Wash the RP-cartridge with 10  mL of water/ACN (90/10, v/v) and subsequently with 20 mL of milliQ. 13. Combine the flow-through and wash containing the released N-glycans. 14. Pre-washed a carbon SPE cartridge with 5  mL of ACN and 5 mL of water/ACN (50/50, v/v). 15. Equilibrate the carbon cartridge with 0.1% TFA. 16. Apply the pool of glycan-containing flow-through and washes (see Subheading 3.1.4, item 13) to the carbon cartridge. 17. Perform a wash step using 20 mL of 0.1% TFA.

20  Microarray Technology Using Glycans Extracted…

293

18. Elute the N-glycans with 5 mL of 0.1% TFA (v/v) in water/ ACN 50/50. 19. Remove the organic solvent (ACN) by vacuum centrifugation and freeze-dry the samples. 3.1.5. Fluorescent Labeling of N- and l-Glycans

1. Resuspend the l- and N-glycans (see Subheading 3.1.3, item 9 and Subheading 3.1.4, item 19, respectively) in 200 ml milliQ. 2. Add 100 ml of a freshly prepared labeling solution and 100 ml of freshly prepared reducing solution (22) (see Note 10 and 11). 3. Mix the sample by shaking or vortexing for about 10 min and subsequently incubate it at 65°C for 2 h. 4. Allow the reaction mixture to cool down to room temperature.

3.1.6. Purification of Labeled N- and l-Glycans Using Graphitized Carbon SPE

1. Precondition two graphitized carbon SPE cartridges by adding 5 mL of ACN, 5 mL of 0.1% TFA in water/ACN (50/50), and equilibrate with 10 mL of 0.1% TFA. 2. Apply the samples and wash with 10 mL of 0.1% TFA. 3. Elute the glycans with 5 mL of 0.1% TFA (v/v) in water/ACN 50/50. 4. Remove the organic solvent (ACN) by vacuum centrifugation and freeze-dry the samples. 5. Redissolve the samples in 200 ml of 77% ACN for further analysis and fractionation. Part of the sample may be used for mass spectrometric analysis to investigate the glycan structures present in the samples (see Note 11). Mass spectrometric analysis is beyond of the scope of this chapter, but useful guidelines covering this part of the analysis are available (23–25).

3.1.7. HPLC Analysis and Fractionation of Glycans

1. Use less than 1% of the sample for an analytical run to estimate the total amount of sample per fraction on the basis of the fluorescent signal. Fractionate the samples by HILIC-HPLC (TSK amide column, 1 mL/min flow rate with fluorescence detection 360/420  nm). Equilibrate the column for 30  min with 3% HILIC solvent B, inject sample, and apply a linear gradient from 3% solvent B (0 min) to 43% solvent B (50 min) to elute the glycans (see Notes 12 and 13). 2. Perform a preparative run using the same gradient conditions as for the analytical run. Collect the fractions every 30 s in a 96-well plate (the fraction collection may be performed manually or using a fraction collector). 3. Speed-vac and/or freeze-dry extensively (several rounds) to remove ammonium formate (see Notes 14 and 15).

294

E. Lonardi et al.

4. Dissolve the fractions in 30 ml of milliQ. 5. Store at −20°C until use (see Note 16). Individual fractions may be analyzed by MALDI-ToF/ToF-MS in negative ion mode to characterize the features present in the fractions (see Notes 11 and 17). 3.2. Plate Setup Printing

It is assumed that the reader is familiar with working with a robotic microarray spotter and slide scanner. 1. Sample plates are set up by transferring 8 ml of each fraction (from point 6 in Subheading 3.1.7), to a 384-well plate. 10 ml Schott Nexterion® Spot Buffer 2× and 2 ml of DMSO are added (final volume 20 ml, 10% DMSO). 2. Fluorescently labeled secondary antibodies are diluted in Schott Nexterion® Spot Buffer 1× with 10% DMSO using dilution factors 1:50, 1:100, and 1:200 (final volume 20 ml). Blood group antigens are prepared at three different concentrations (125, 250, and 500 mM, final volume 20 ml) in the same buffer (see Note 18). 3. Sample plates may be stored at −20°C (see Note 19). 4. Open the slide box in a clean environment using gloves. Avoid direct contact with the printing surface. 5. Mark slides in the corner, on the back of the slide (not the print surface), avoiding marking behind the printed area, which would stall the autofocus procedure on the scanner later on. Marking the slides makes it easier to recognize the directionality of the array. After incubation, in fact, the printed spots are not visible anymore. 6. Set up the printing method. In our case, samples are printed in triplicate, using SMP3 stealth pins, which nominally print spots of 100 mm of diameter and have a delivery volume of 0.7 nL per spot (see Note 20). To avoid coalescing of the spots, minimum spot distance is kept above 240 mm and the inter-array spacing to 4,000 mm. Samples are loaded onto the pins during dipping time (3,000 ms), deposition of the samples onto the slide happens during contact time (25  ms). Re-dip is performed after 100 spots, or as determined during the test run (see Note 21). 7. Printing is performed at 20°C, 60% relative humidity.

3.3. Humidification/ Immobilization

After the samples have been deposited on the surface of the slides, it is important to allow time for the immobilization reaction (between epoxysilane and secondary amine of the labeled glycan) to reach completion. This is achieved by placing the slides in a humidity chamber, preferably overnight (see Note 22). After immobilization, slides can be removed from the humidity chamber and dried at room temperature in a fume hood for 1 h. After drying, slides can be stored in the manufacturer’s original boxes, in the dark, at room temperature, in a desiccator (see Note 23).

20  Microarray Technology Using Glycans Extracted…

295

Negative Control

Control Serum

Serum 2

Serum 1

Fig. 2. Scheme of a slide printed with four microarrays (top) and the corresponding silicon gasket (middle). The gasket is 1 mm thick and separates the microarrays into four individual wells, to which different sera and controls can be applied. Negative control = incubation with fluorescently labeled secondary antibodies only.

3.4. Assay

1. Dust particles eventually present on the slides are cleaned using compressed air. To define separate wells for the manual application of multiple serum samples onto the slide, appropriate silicone masks are applied (Fig.  2). Check that the silicone does not touch the array, then flip the slide onto non-fluff paper and press firmly so that the gasket adheres to the glass. Check that the slides are properly sealed by looking at the gasket from the non-printed side of the slide (see Note 24). 2. Flush-sealed slides with PBS, place them horizontally, and apply PBS to each well. Shake for 3 min. 3. Proceed to blocking: making sure the arrays never dry out, remove the PBS and apply the blocking solution. Incubate at room temperature with gentle shaking for 60 min. 4. Flush slides with PBS and PBST5, wash with PBS for 1  min while shaking. Extensive washing can improve the quality of the screening by lowering the background. 5. Dilute sera in dilution buffer and apply each serum to a different array (see Note 25). Incubate at room temperature for 60 min while gently swirling the slide. Prevent arrays from drying as well as cross-contamination of the samples by setting the rotating table to the appropriate RPM. Check at regular intervals of time to make sure no leakage occurs. In the case of screening with lectins: if the latter are fluorescently labeled, it is important to protect the incubating chamber from light. 6. Remove sera from the arrays, flush slides with PBST5, then PBS. Wash slides for 1  min with PBS. Extensive wash may reduce background problems.

296

E. Lonardi et al.

7. Add secondary antibodies and incubate at room temperature for 30 min. Because the secondary antibodies are coupled with a fluorescent dye it is recommended to protect the incubation chamber from light (see Note 26). 8. Flush slides with PBST5, then PBS, finally milliQ. Wash for 1 min. 9. Remove gaskets and dry the slides in an oil-free nitrogen stream or by centrifugation. 10. Protect the array from light, dust, and abrasion of the surface until ready for scanning. 3.5. Scanning and Analysis

We assume that the reader is familiar with generating an array list and operating appropriate image processing software. 1. After the arrays have been incubated, they can be immediately scanned or stored appropriately in the dark. When loading the slides onto the scanner, ensure the print surface is facing the laser source. Using the software provided with the scanner, set up the instrument to scan using appropriate settings. Settings can be tested and optimized, keeping in mind that each scan lowers the fluorescence of the signals. Slides are scanned using the appropriate laser channel. Glycan arrays can be scanned at different PMT and resolution, generating images with as many signals as possible within its dynamic range. We scan at fixed PMT (generally 100, but this can be lowered to 20 in the case of high numbers of saturated signals). 2. Images generated by the scanner software should be properly renamed and saved in TIFF format (or other format supported by the analysis software). 3. For post-acquisition analysis, the features (spots) on the array must have the correct name and ID assigned to, i.e., its coordinates on the array need to be linked to the file listing the content of each sample printed. Load image and array list into the analysis software, manually adjust the grid until it overlaps the features (see Note 27). 4. Spots are defined as irregular features with a diameter of 100 mm and are allowed to resize in the range 50–300 mm and the CPI (Composite Pixel Intensity) is set to 0. Features automatically flagged as bad/missing by the software are manually double checked. Flagged features are removed from further steps in the analysis. For each feature, we extract the total fluorescent intensity (TI) and average this value for each carbohydrate (each being printed in triplicate). The final set of values is plotted against the sample number (see Note 28). 5. Features that are positive in the control incubation (secondary antibodies only) are also removed from the analysis (Fig. 3)

20  Microarray Technology Using Glycans Extracted…

a

b

297

40000 35000 30000 25000 20000 15000 10000 5000 0

c

20000

1

51

101 151 201 251 301 351 401 451 501 551

1

51

101 151 201 251

15000

10000

5000

0 301 351 401 451 501

551

Glycan number

Fig. 3. Data extraction from array images. (a) Scanned image of an array incubated with serum. Each spot corresponds to one feature and contains immobilized glycans. Each carbohydrate sample is printed in triplicate (see white box). Luminous spots arise when autoantibodies present in the serum are recognized by fluorescently labeled secondary antibodies. The intensity of fluorescence is averaged along the triplicate and plotted against the glycan ID (b, black histogram). An incubation performed only with fluorescent secondary antibodies provides the negative control (gray histogram). Features which have intensity above cutoff in the negative control are subsequently removed from further analysis (c).

6. For the comparison of data from a series of incubations, the intensities of each array are rescaled: the averaged total intensity across each array is calculated (TImean) and one of the reference incubation (control serum) is used to calculate a rescaling factor, for each array: Irescaled = I × TImean/refTImean (see Notes 29 and 30).

4. Notes 1. The chemicals and reagents are readily available from a number of laboratory chemical suppliers. 2. Degassing of solvents (e.g., by vacuum, sonication, or helium sparging) is required for optimal performance of the HPLC

298

E. Lonardi et al.

systems. Ammonium formate buffer is semi-volatile so milliQ should be degassed before adding the buffer concentrate. 3. Adjust the volumes in proportion to the size of the biological sample used for glycan preparation. For in vitro-cultured cells, sonication will generally be sufficient for homogenization. 4. Chloroform is volatile and toxic. Work in a fume hood as much as possible. Heating of the sample should be avoided. Elevated temperatures as they may occur during extensive sonication may be avoided by applying a cooling bath. 5. Small glycolipids of approximately up to three monosaccharide residues are predominantly found in the lower phase. Therefore, the prepared glycolipid pools derived from the upper phases will be biased toward large glycolipid structures. 6. Extensive drying should be avoided, because it may be difficult to get the (glyco-)proteins from extensively dried pellets back into solution. 7. Optionally, glycolipid samples may be analyzed by mass spectrometry, preferably MALDI-MS. 8. The endoglycoceramidase may discriminate between different glycolipid synthetic series and is, e.g., known to release the glycan portions of globo-series glycosphingolipids with rather low efficacy. 9. PNGase F will not release core-3-fucosylated N-glycans which are commonly found in plants and many invertebrates. 10. 2-Aminobenzamide (2AB) label may be used instead of 2AA. This results in similar glycan microarray results (19). Also other reductive amination tags such as 2-aminopyridine are expected to be compatible with the epoxide immobilization used in this approach. 11. MALDI-ToF/ToF-MS provides the possibility of sequencing and structurally elucidating the individual 2AA-labeled glycans. The 2AA-labeled glycans may preferably be analyzed in negativeion mode to efficiently detect the negatively charged species (glycan structures containing sialic acids and sulfates). For mass spectrometric analysis, the organic content should be evaporated and negative-ion mode MALDI-ToF/ToF-MS can be performed after using the ZipTip protocol for sample desalting. The samples are eluted directly onto a MALDI target plate with 1–2  ml of 5  g/L 2,5-dihydroxybenzoic acid (DHB) in 50% ACN. 12. BEH amide/glycan column (Waters) may be used as an alternative. This column has become recently available and has an improved separation power when compared with Amide 80 (3 mm) from Tosoh Biosciences (26). While the BEH amide is designed for UPLC use, it can be run using a conventional

20  Microarray Technology Using Glycans Extracted…

299

HPLC with slightly prolonged run times and reduced flow rate resulting in excellent separation. 13. Use 2AA-labeled glycan standards to determine the fluorescent signal observed per picomole of glycan injected. Useful glycan standards may be chito-oligosaccharides such as chitohexaose (Dextra) and maltooligosaccharides such as maltotetraose, maltopentaose, maltohexaose, and maltoheptaose (Sigma). 14. Alternatively, one may chose for reverse phase-SPE or reverse phase HPLC purification of the labeled glycans of each HILIC fraction which is useful to remove residual ammonium formate completely. 15. Efficient removal of amines (ammonium formate) is critical for efficient immobilization of the glycans on epoxy-activated glass slides. In case immobilization efficacy is compromised, one may choose to reduce the concentration of ammonium formate used in the HILIC separation, or replace ammonium formate by, e.g., sodium formate. While the latter cannot be removed with vacuum centrifugation, it is expected not to interfere with printing. 16. For more extensive fractionation, second dimension chromatography such as RP and carbon may be completed. Using RP the neutral glycans are separated from the acid glycans. Glycanscontaining sialic acids have an increased retention on RP material. 17. The fractions may be further analyzed using LC-MS/MS to obtain additional structural information. 18. The efficacy of immobilization for blood group antigens and landing spots might slightly differ from 2AA-labeled glycans, as the former contain primary amine groups, while the latter rely on secondary aromatic amines for the reaction with the epoxy-activated surface. 19. To thaw the sample plates thoroughly, incubate them, without removing the Thermowell sealing tape, at 37°C while shaking, for 20 min at least. Spin sample plates to collect the entire sample to the bottom of the wells. Make sure that there is no precipitate. Repetitive freezing and thawing of the sample plate can cause a precipitate to appear at the bottom of the wells. This might be due to interaction between DMSO and printing buffer, and precipitation is less if DMSO content is kept at 10%. 20. Pins need to be examined before each run to ensure that they are clean and undamaged using a microscope. The presence of crystals on the tips of the pins is a clear indication that thorough cleaning is required. For the appropriate cleaning protocol, refer to the manufacturer’s instructions. Bent or dented pins should be discarded.

300

E. Lonardi et al.

21. To check the settings for the print method, set up a mock-plate containing only printing buffer (Schott Nexterion® Spot Buffer 1×, 10% DMSO) in 20  ml aliquots. Using the same pin and plate setup which is required by the print method, print 100 spots keeping the values for dipping time, printing time, speed of the print head. Observe the result under the microscope. Spots should appear uniform in size and equally spaced. If spots start fading towards the end of the test run, determine after how many spots this happens and calculate accordingly how often to re-dip the pins. This depends on the number of arrays per slide and on the number of spots per each array. 22. The preparation of a humidity chamber requires equilibration and should be done well in advance. Prepare a suitable amount of oversaturated NaCl solution and place it at the bottom of a sealable container, which is then kept tightly closed. Equilibration takes several hours, but the humidity chamber can be reused several times, provided it is kept sealed at all times (plastic wrap around it can help trapping the moisture). The relative humidity of the chamber will be ~75% at room temperature. Slides can be placed on suitable racks (to avoid contact with the solution at the bottom of the box), print side up, inside the chamber. 23. After the slides have been humidified and dried, the spots are still visible as small crystals left by the printing buffer. It is important to examine the slides at this point to check for missing spots, regular gridding, smears, and other major imperfections. Virgin slides (slides which have been printed, humidified, and dried but have not yet been screened) can also be scanned to ascertain the presence of all spots and to check the quality of the landing spots, if present. 24. For custom made arrays, silicone masks can be cut to size to efficiently seal the array slide providing at the same time enough space for pipetting. 25. The array can be used to screen for autoantibodies (low titer) as well as for anti-pathogen antibodies (high titer). Depending on the biological question, the dilution of the sera used for incubation needs to be adjusted accordingly. 26. Serum IgG and IgM antibodies do hardly appear to suffer from repeated freezing and thawing. 27. To help orientating the image and correctly overlapping the grid, it is recommended to use “landing spots.” Landing spots are well-characterized samples, which can include fluorophores (compatible with the laser source), fluorescently labeled antibodies or lectins. Our lab prints the same secondary antibodies used in the sandwich assay (anti-Human IgG, Cy3conjugated, and AlexaFluor 647-conjugated anti-Human IgM

20  Microarray Technology Using Glycans Extracted…

301

both developed in goat). We use a serial dilution which produces a characteristic pattern and is easily identified even in virgin slides. 28. Instead of the total intensity (TI), the background corrected median fluorescence can be calculated and processed in the same way, finally plotting the values against the sample number. 29. Alternatively, the intensity of the landing spots can be used to rescale the intensities of different batches before further analysis. When analyzing the sera from a cohort, it is essential to randomize the incubations, making sure to have patient sera and healthy control sera on the same slide, as well as a negative control (secondary antibody control). If the analysis is performed on multiple days, always include a control serum (unrelated to the cohort) for each day. 30. The fractions, which are positive in the array scanning, may be analyzed in detail using tandem mass spectrometry: negativeion mode MALDI-ToF/ToF-MS and liquid chromatography coupled to an electrospray ionization tandem mass spectrometer. When combined with common glycobiology knowledge, tandem mass spectrometric data may be sufficient to propose glycan structures. However, detailed structural information (such as linkage and branching) regarding unusual structures should be obtained using exoglycosidases in combination with mass spectrometry analysis.

Acknowledgements The authors thank Dr. C.H. Hokke, Dr. A.R. de Boer, U. Lambertz, Dr. A. van Diepen, and R. Curfs for their help in establishing the array platform. This project was financed by the Netherlands Genomics Initiative (Horizon Breakthrough Project 050-71-302). References 1. Varki A (1993) Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3:97–130 2. Hebert DN, Garman SC, Molinari M (2005) The glycan code of the endoplasmic reticulum: asparagine-linked carbohydrates as protein maturation and quality-control tags. Trends Cell Biol 15:364–370 3. Taylor ME, Drickamer K (2007) Paradigms for glycan-binding receptors in cell adhesion. Curr Opin Cell Biol 19:572–577

4. Endo T (2005) Glycans and glycan-binding proteins in brain: galectin-1-induced expression of neurotrophic factors in astrocytes. Curr Drug Targets 6:427–436 5. Thisse B, Thisse C (2005) Functions and regulations of fibroblast growth factor signaling during embryonic development. Dev Biol 287:390–402 6. Lau KS, Dennis JW (2008) N-Glycans in  cancer progression. Glycobiology 18: 750–760

302

E. Lonardi et al.

7. Moskal JR, Kroes RA, Dawson G (2009) The glycobiology of brain tumors: disease relevance and therapeutic potential. Expert Rev Neurother 9:1529–1545 8. Zhao YY, Takahashi M, Gu JG, et  al (2008) Functional roles of N-glycans in cell signaling and cell adhesion in cancer. Cancer Sci 99:1304–1310 9. Vasta GR (2009) Roles of galectins in infection. Nat Rev Microbiol 7:424–438 10. Sperandio M, Gleissner CA, Ley K (2009) Glycosylation in immune cell trafficking. Immunol Rev 230:97–113 11. Pedersen JW, Blixt O, Bennett EP, et al (2011) Seromic profiling of colorectal cancer patients with novel glycopeptide microarray. Int J Cancer 128:1860–1871 12. Lu H, Goodell V, Disis ML (2008) Humoral immunity directed against tumor-associated antigens as potential biomarkers for the early diagnosis of cancer. J Proteome Res 7:1388–1394 13. Lu H, Goodell V, Disis ML (2007) Targeting serum antibody for cancer diagnosis: a focus on colorectal cancer. Expert Opin Ther Targets 11:235–244 14. Blixt O, Head S, Mondala T, et  al (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101:17033–17038 15. Gruber K, Horlacher T, Castelli R, et al (2011) Cantilever Array Sensors Detect Specific Carbohydrate-Protein Interactions with Picomolar Sensitivity. ACS Nano 16. Weishaupt M, Eller S, Seeberger PH (2010) Solid phase synthesis of oligosaccharides. Methods Enzymol 478:463–484

17. Song X, Lasanajak Y, Xia B, et  al (2011) Shotgun glycomics: a microarray strategy for functional glycomics. Nat Methods 8:85–90 18. Liu Y, Feizi T, Campanero-Rhodes MA, et  al (2007) Neoglycolipid probes prepared via oxime ligation for microarray analysis of oligosaccharideprotein interactions. Chem Biol 14:847–859 19. de Boer AR, Hokke CH, Deelder AM, et  al (2007) General microarray technique for immobilization and screening of natural glycans. Anal Chem 79:8107–8113 20. de Boer AR, Hokke CH, Deelder AM, et  al (2008) Serum antibody screening by surface plasmon resonance using a natural glycan microarray. Glycoconj J 25:75–84 21. Potter VR, Elvehjem CA (2011) A modified method for the study of tissue oxidations. J Biol Chem 114:495–504 22. Ruhaak LR, Steenvoorden E, Koeleman CA, et al (2010) 2-picoline-borane: a non-toxic reducing agent for oligosaccharide labeling by reductive amination. Proteomics 10:2330–2336 2 3. Harvey DJ (2005) Fragmentation of negative ions from carbohydrates: part 3. Fragmentation of hybrid and complex N-linked glycans. J Am Soc Mass Spectrom 16:647–659 24. Harvey DJ (2005) Fragmentation of negative ions from carbohydrates: part 2. Fragmentation of high-mannose N-linked glycans. J Am Soc Mass Spectrom 16:631–646 25. Zaia J (2004) Mass spectrometry of oligosaccharides. Mass Spectrom Rev 23:161–227 26. Ahn J, Bones J, Yu YQ, et al (2010) Separation of 2-aminobenzamide labeled glycans using hydrophilic interaction chromatography ­columns packed with 1.7 microm sorbent. J Chromatogr B Analyt Technol Biomed Life Sci 878:403–408

Chapter 21 Studying Modification of Aminoglycoside Antibiotics by Resistance-Causing Enzymes via Microarray Matthew D. Disney Abstract Widespread bacterial resistance to antibiotics is a significant public health concern. To remain a step ahead of evolving bacteria, new methods to study resistance to antibacterials and to uncover novel antibiotics that evade resistance are urgently needed. Herein, microarray-based methods that have been developed to study aminoglycoside modification by resistance-causing enzymes are reviewed. These arrays can also be used to study the binding of aminoglycoside antibiotics to a mimic of their therapeutic target, the rRNA aminoacyl site (A-site), and how modification by resistance-causing enzymes affects their abilities to bind RNA. Key words: Antibiotics, Bacteria, Aminoglycosides, RNA, Antibiotic resistance, Glycoarray

1. Introduction The therapeutic target of aminoglycosides is the bacterial ribosome (1–5). Different sites within the ribosome bind aminoglycosides, for example, streptomycin (6) binds the S12 protein and affects ribosomal assembly, while kanamycin A, neomycin B, and amikacin target the aminoacyl site (A-site) and interfere with decoding (7–10). A variety of structural and biochemical studies have investigated how aminoglycosides affect decoding and bind to the ribosome. Early work to identify aminoglycoside binding sites in rRNA was completed by Moazed and Noller (7). This work was extended by Purohit and Stern who showed that aminoglycosides bind the A-site as part of the entire ribosomal and as a small rRNA fragment similarly (11). High-resolution NMR structural analysis of the aminoglycoside–A-site complex was completed by the Puglisi

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_21, © Springer Science+Business Media, LLC 2012

303

304

M.D. Disney

group (9, 12–14). These early structural studies showed that local conformational changes within the RNA target occurred upon aminoglycoside binding. These NMR studies served as a catalyst to understand the molecular recognition of the codon–anticodon helix by the 16S rRNA using chemical modification. The results of that study showed that the 30S ribosome is critical for the decoding process by interacting with the codon–anticodon complex (15). Crystallographic studies completed on ribosomal subunits or entire ribosomes in the presence of aminoglycosides provided near atomic resolution information on codon–anticodon interactions and the complex’s interaction with aminoglycosides (16–18). These studies served as a springboard for the Pilch and Hermann groups to develop fluorescence-based methods to study the effect that aminoglycosides have on A-site dynamics (19–22). Subsequent crystallographic studies by the Cate group on whole intact Escherichia coli ribosomes revealed that, in addition to the A-site, aminoglycosides bind helix 69 in 70S ribosomes, helping to explain how aminoglycosides inhibit ribosomal recycling (8). Bacterial resistance to aminoglycosides is not as complex as the mechanism involved in aminoglycoside inhibition of protein synthesis. The most common mechanism of resistance is due to enzymatic modification of the aminoglycoside. Modifications introduced into aminoglycosides that confer resistance include O-phosphorylation, O-nucleotidylation, and N-acetylation. Enzymes that confer these modifications are called aminoglycoside phosphotransferases (APH’s), aminoglycoside nucleotidyltransferase (ANT’s), and aminoglycoside acetyltransferases (AAC’s), respectively (3, 4, 23–25). Enzymatic modification is so extensive that many of the parent aminoglycoside antibacterials are no longer effective in a clinical setting. Figure 1 depicts some of the positions in the aminoglycoside ribostamycin that are modified by resistancecausing enzymes and the functional groups that are added by resistance-causing enzymes. Modified aminoglycosides have a greatly diminished affinity (>10-fold, Fig.  2) toward their therapeutic RNA target (26). Investigation of the binding pockets of RNA–aminoglycoside complexes shows that diminished binding can be due to removing functional groups from forming direct contacts with RNA (e.g., acylation of amino groups) or introducing steric bulk or charge– charge repulsion (e.g., a phosphate group lying near the negatively charged backbone of RNA). Based on these observations and the development of carbohydrate microarray technology (27–33), our group has developed a microarray approach that monitors aminoglycoside modification by resistance-causing enzymes and the effect of modification on rRNA A-site binding using site-specifically immobilized aminoglycoside substrates (34, 35). These studies, therefore, extend previous investigations that used nonspecifically immobilized aminoglycosides to only probe their binding to RNA (29) or resistance-causing

21  Studying Modification of Aminoglycoside Antibiotics…

305

Fig. 1. Types of enzymatic modification of aminoglycosides that confer resistance.

HO HO

6' 3'

OH O HO

Kd = 1.1 µM Kanamycin A HO H

6'

O

NH2 O NH2 O

H 2N O HO

O

Kd = 2.0 µM Tobramycin

NH2 O

NH2

OH 3''

NH2 O

3'

APH P γ ATP

O P O

-

6' 3'

HO H NH2

OH

APH NH2 OH

NH2 O OH O HO

Kd = 1300 µM

NH2 OH

OH

OH 3''

O

32

-

HO O

32

P γ ATP

6' 3'

NH2 O

NH2

OH 3''

O

NH2 OH

OH

NH2 O

H 2N O HO

NH2 O

NH2

OH 3''

O

NH2 OH

OH

Fig. 2. The structures of the aminoglycosides kanamycin A and tobramycin, their measured affinity to a mimic of the bacterial rRNA A-site (26) and the product of their modification by APH(3¢)-IIIa. Note that kanamycin A is modified by APH(3¢) because it contains a reactive hydroxyl group at the 3¢ position, whereas tobramycin contains a hydrogen atom and is thus not susceptible to APH(3¢) modification.

enzymes (27). The methods outlined here focus on the use of radioactive labeling to detect enzymatic modification and the impacts of modification on binding RNA. Most recently (mid-2010), our group developed a fluorescence-based microarray approach that can be used to detect modification of carbohydrates by acetyltransferases (36). Please see that manuscript for a description of those studies.

306

M.D. Disney

HO HO

6' 3'

NH2 O OH O HO

NH2 O

NH2 6' O HO HO 3' OH O HO

NH2

OH 3''

O

Cu(I), Vit. C TBTA

NH2 OH

NH2 O

NH2

OH 3''

O

N3

N

NH2 OH

N N

N

CuSO 4 Vit. C TBTA Glycoside

+ + + +

+ + +

+ + +

N

+ + + -

Array hybridized w/ A-site Fig. 3. Using the Huisgen dipolar cycloaddition reaction to immobilize azide-functionalized kanamycin A onto an alkynefunctionalized microarray surface.

The preparation of the agarose microarray surfaces is critical for these studies, as agarose provides a porous layer that allows for high ligand loading and supports high-density modification of sugars displayed on the array surface (as much as 80% of the total amount of immobilized aminoglycoside) (34, 35). Agarose array surfaces are also quite versatile in the number of chemically reactive handles that can be introduced onto the surface as reactions with amines, alkynes, and azides have all been reported (Fig. 3) (29, 36–41). Portions of this document have been adapted from Aminova and Disney (42), primarily the construction of appropriately functionalized agarose microarray surfaces. Microarray fabrication is included such that two stand alone chapters would be available. That methods paper reviews our group’s work on the development of two-dimensional combinatorial screening (2DCS) (37, 38) and people interested in those methods should refer to that manuscript.

2. Materials 2.1. Array Preparation and Spotting

This section is based on ref. 42, which describes the protocol for 2DCS. 1. 31.8  mM solution of NaCNBH3: Dissolve 0.2  g of sodium cyanoborohydride in 100 mL of 4:1 1× PBS:ethanol. Prior to applying this solution to the array surface, ensure that all of the NaCNBH3 is dissolved by stirring the solution for 2–3 min. It is important to prepare a fresh solution each time.

21  Studying Modification of Aminoglycoside Antibiotics…

307

2. 10× Phosphate-buffered saline (PBS): Dissolve 14.2  g of Na2HPO4, 2.45 g of KH2PO4, 81.8 g of NaCl, 1.86 g of KCl in 900 mL of nanopure water. Adjust the pH to 7.5 using 1 M NaOH or 1 M H3PO4. Add nanopure water to bring volume to 1 L. Store the solution at room temperature. 3. 0.2% Sodium dodecyl sulfate (SDS): Dissolve 2  g of SDS in 900 mL of nanopure water; add water to make 1 L of solution. Store the solution at room temperature. 4. 10× Phosphate buffer: Prepare the following solutions: 1 M K2HPO4 (dissolve 1.74 g of K2HPO4 in 10 mL of nanopure water) 1 M KH2PO4 (dissolve 1.36 g of KH2PO4 in 10 mL of nanopure water) 1 M Na2HPO4 (dissolve 1.42 g of Na2HPO4 in 10 mL of nanopure water) 1 M NaH2PO4 (dissolve 1.2 g of NaH2PO4 in 10 mL of nanopure water) Prepare a 100  mM potassium phosphate solution by mixing 940 mL of 1 M K2HPO4, 60 mL of 1 M KH2PO4 and 9 mL of nanopure water. Prepare a 100  mM sodium phosphate solution by mixing 932 mL of 1 M Na2HPO4, 68 mL of 1 M NaH2PO4, and 9 mL of nanopure water. Finally, mix equal volumes of the two 100  mM solutions together; this solution should have a pH of ~8. Store the solution at room temperature. 5. 1% Agarose solution: Dissolve 1 g of high melting agarose in nanopure water by heating the solution in a microwave or on a stir plate. Only use this solution once as repeated heating cycles will create a fragile microarray surface. 6. 0.02 M NaIO4 solution. Dissolve 2.14 g of NaIO4 in 500 mL of nanopure water. 7. 10% Ethylene glycol solution: Mix 20 mL of ethylene glycol with 180 mL of nanopure water. 8. 0.1  M NaHCO3: Dissolve 4.2  g of NaHCO3 in 500  mL of nanopure water. Adjust the pH to 8.5 using NaOH. 9. 10  mM 3-azidopropylamine: Prepare 3-azidopropylamine as previously described (37). Dissolve 0.1 g of 3-azidopropylamine in 100 mL of nanopure water. 10. 10  mM propargylamine: Mix 64  mL of propargylamine in 100 mL of nanopure water. 11. Tris [(1-benzyl-1H-1,2,3-triazol-4-yl) methyl] amine (TBTA) solution (43): Stock solutions of TBTA are to be made in a 4:1

308

M.D. Disney

mixture of 2-butanol:dimethylsulfoxide and should be stored at 4°C. TBTA is now commercially available from SigmaAldrich (Milwaukee, WI), however, the synthesis is straightforward, high-yielding, and inexpensive (43). 12. 5 mM (Tris (2-Carboxyethy) phosphine hydrochloride (TCEP) (10×): Dissolve 0.14 g of TCEP in 10 mL of nanopure water. 13. 10× Spotting Solution A: Mix 121 mg of Tris–HCl, 25 mg of CuSO4·5H2O, 1.8 mg of ascorbic acid, 1 mL of glycerol, add nanopure water to 10 mL of total volume, add enough of the above TBTA solution to bring the concentration of TBTA to 1 mM, and pH to 8.5. 14. 10× Spotting Solution B: Mix 1 mL of 10× phosphate buffer, 25 mg of CuSO4·5H2O, 1 mL of 5 mM TCEP, 1 mL of 1 mM TBTA, 1 mL of glycerol, and 6 mL of nanopure water. 15. Azide or Alkyne ligands: Prepare 1:5 serial dilutions of alkyne/ azide-displaying ligands from 5 mM to 1 mM in 1× Spotting Solution A or B, respectively. 2.2. Preparation of Cellular Lysates

1. E. coli strains: JM109(DE3) and BL21(DE3) pLysS. 2. Plasmids encoding resistance-causing enzymes: pETSACG1 (encodes APH(3¢)-IIIa); pET22b(+) (encodes ANT(2″)-Ia); pET22a (encodes AAC(3)-IV); or plasmid containing a resistancecausing enzyme of interest. 3. LB medium: Dissolve 10  g of bacto tryptone, 5  g of bacto yeast extract, 10 g of NaCl, 1 mL of 1 M NaOH in 900 mL of deionized water. Adjust the pH to 7.0 with 1  M NaOH. Sterilize by autoclaving on the liquid cycle at 121°C for 20 min. Store the solution at room temperature. 4. 1,000× Ampicilin solution: Dissolve 50  mg of ampicillin in 1 mL of water. 5. 1,000× Kanamycin A solution: Dissolve 10 mg of kanamycin A in 1 mL of water. 6. 1,000× Carbenicllin: Dissolve 50 mg of carbenicillin in 1 mL of water. 7. 1  M (500 or 1,000×) isopropyl-b-d-1-thiogalactopyranoside (IPTG): Dissolve 2.83 g of IPTG in a total volume of 10 mL by the addition of nanopure water. 8. APH(3¢)-IIIa (44) lysis buffer: Dissolve 606 mg of Tris–HCl, 1.1 g of NaCl, 1.6 mg of dithiolthreitol (DTT) to 100  mL. Adjust the pH to 8.0 and then add 17.4 mg of phenylmethylsulfonyl fluoride (PMSF; dissolved in ethanol). 9. APH(3¢)-IIIa (44) dialysis buffer: Dissolve 24.2 g of Tris–HCl, 11.9  g of KCl, 8.13  g of MgCl2·6H2O to 4  L of nanopure water; adjust the pH to 7.5.

21  Studying Modification of Aminoglycoside Antibiotics…

309

10. ANT(2″)-Ia (45) lysis buffer: Add 600 mg Tris–HCl, 101 mg of MgCl2·6H2O, and 35  mL of 2-mercaptoethanol, and nanaopure water to afford a solution with a total volume of 100 mL. Adjust the pH to 8.0. 11. ANT(2″)-Ia (45) dialysis buffer: Add 29 g of Tris–HCl, 32.5 g of MgCl2·6H2O, 85.6  g of NH4Cl, and 99  mg of DTT to 4 L of nanopure water and adjust the pH to 7.1. 12. AAC(3)-IV (46) lysis buffer: Add 300 mg of triethanolamine and 17 mg of PMSF (dissolved in ethanol) to 100 mL of nanopure water. Adjust the pH of the solution to 7.8. 13. AAC(3)-IV (46) dialysis buffer: Add 11.2 g of triethanolamine to 4 L of nanopure water and adjust the pH to 7.8. 14. BCA Protein Assay Kit (Pierce Biotechnologies/Thermo Fisher; catalog number 23227). 15. 4× SDS-PAGE Stacking Buffer: Dissolve 30.4 g of Tris–HCl base and 2.0 g of SDS in 500 mL of nanopure water. Adjust the pH to 6.8 with 1 M HCl. 16. 4× SDS-PAGE Resolving Buffer: Dissolve 91.0 g of Tris–HCl base and 2.0 g of SDS 500 mL of nanopure water. Adjust the pH to 8.8 with 1 M HCl. 17. 5× SDS-PAGE Electrophoresis Buffer: Dissolve 15.1  g of Tris–HCl base, 72.0 g of Glycine, and 5.0 g of SDS in 1 L of nanopure water. 18. 30% (w/v) Acrylamide/bis-acrylamide (19:1): Acrylamide solutions can be purchased from Sigma-Aldrich or prepared as follows: dissolve 28.5 g of acrylamide and 1.5 g of bis-acrylamide in 100 mL of nanopure water. (Caution: acrylamide is a known neurotoxin). 19. 4× Protein gel sample loading buffer: Mix 2.0 mL of 1 M Tris– HCl, 0.8 g of SDS, 4.0 mL of 10% glycerol, 0.4 mL of 14.7 M b-mercaptoethanol, 1.0  mL of 0.5  M EDTA, and 8  mg of bromophenol blue. 2.3. Modification and Hybridization of Microarrays

1. 1× ANT and APH assay buffer: Add 479  mg of HEPES, 22.3 mg of MgCl2·6H2O, and 16.4 mg of KCl to 10 mL of nanopure water. Adjust the pH to 7.5. 2. 1× AAC(3)-IV assay buffer: Add 1.916 g of HEPES to 40 mL of nanopure water and pH the solution to 7.5. 3. 1× RNA hybridization buffer: Add 45 mg of Na2HPO4, 2 mg of EDTA, and 407 mg of NaCl to 40 mL of nanopure water. Adjust the pH of the solution to 7.1. 4. Fluorescently labeled A-site oligonucleotide mimic: A fluorescently labeled oligonucleotide mimic of the bacterial A-site can be purchased from Dharmacon or Integrated DNA Technologies, Inc.

310

M.D. Disney

5. Phosphorimager. 6. Microarray scanner.

3. Methods Although the microarrays are generally robust, we prefer to use them within approximately 1 month of their construction. This is due to surface cracking that can occur if the array dries out, producing sub-optimal images. When spotting microarrays, we suggest using a freshly prepared spotting solution. For the majority of our array work, dialyzed cellular lysates were applied to the array surface to monitor modification. Partially or totally purified proteins can also be used, however, their preparation can be timeconsuming. 3.1. Preparation of Agarose Slides (42)

1. Prepare a 1% agarose solution (w/v) using nanopure water. Melt in a microwave on high for 2–3 min, swirling the solution every 20–30 s. 2. While the solution is hot, apply ~1.5 mL to the surface of a glass slide using a P-1000 pipette. Ensure that the solution is spread evenly over the slide surface. The solution can be spread over the surface by simply tracing the outside of the array with the pipette tip during application to the array surface. Allow the agarose to dry to a thin film overnight.

3.2. Functionalization of Agarose Slides (42)

1. Submerge the slides in 0.02 M NaIO4 for 30 min at room temperature, and then wash them with nanopure water for 30 min. 2. Submerge the slides in 10% (v/v) ethylene glycol for 1  h at room temperature to quench residual NaIO4. Wash with water for 1.5 h, changing the water every 20 min. Subheading  3.2, step 1 and 2 afford slides that display aldehydes. 3. (a) To construct microarrays of amine-displaying ligands, complete the following: ●●

●●

●●

Allow the slides to dry. Prepare spotting solutions as follows: small molecule at desired concentration (typically serially diluted from 5 mM to 1 mM), 0.1 M NaHCO3, and 10% glycerol. Spot 0.4  mL of the solutions onto aldehyde-agarose slides in duplicate. Incubate for 3  h at room temperature in a humidity chamber (box containing a saturated solution of NaCl). Wash the slides 3 × 10  min with 1× RNA hybridization buffer, followed by water, 2 × 10  min. Continue to Subheading 3.2, step 6.

21  Studying Modification of Aminoglycoside Antibiotics…

311

(b) For microarrays of alkyne-displaying ligands, submerge the slides in a solution of 0.1  M NaHCO3 and 10  mM 3-azidopropylamine for 3 h at room temperature. Continue to Subheading 3.2, step 6. (c) For microarrays of azide-displaying ligands, submerge the slides in a solution of 0.1 M NaHCO3 and 10 mM propargylamine. Continue to Subheading 3.2, step 6. For the series of aminoglycosides that we have tested in modification assays, all have been functionalized with an azide tag for surface immobilization (34, 36, 39). 4. Submerge slides in a solution of 31.8 mM NaCNBH3 for 3 min to reduce the imine formed on the microarray surface. 5. Wash the slides with 0.2% SDS, 3 × 15  min, and then with water, 2 × 15 min. 6. Dry the slides under a stream of air. 3.3. Immobilization of Azido- or AlkyneAminoglycosides on the Slide Surface (42)

In this section, spotting the aminoglycoside microarrays manually by delivering fixed volumes of solutions from a pipette is described. This method is feasible to produce microarrays with £100 features. The use of replicators or robotic arrayers can be used, if more features are desired. It should be noted that spot sizes should be ~500–1,000 mm in diameter and each spot should be separated by at least 2,000 mm for them to be observable using a phosphorimager. Therefore, ensure that the pin and spotting buffer combination produce features that are large enough to be observed. 1. Prepare spotting solutions as follows: (a) For azide-displaying small molecules: add the small molecule at the desired concentration (typically serially diluted from 5 mM to 1 mM) in 1× Spotting Solution A. Spot 0.4 mL of the solutions in duplicate onto alkyne-agarose slides. (b) For alkyne-displaying small molecules: add the small molecule at the desired concentration (typically serially diluted from 5 mM to 1 mM) in 1× Spotting Solution B. Spot 0.4 mL of the solutions in duplicate onto azide-agarose slides. 2. Incubate the slides for 3 h at room temperature in a humidity chamber (box containing a saturated solution of NaCl). Wash the slides 3 × 10  min with 1× RNA hybridization buffer, followed by water (2 × 10 min).

3.4. Preparation of Cellular Lysates 3.4.1. Preparation of APH(3 ¢)-IIIa Lysate

1. Transform E. coli JM109(DE3) with the pETSACG1 plasmid containing the APH(3¢)-IIIa resistance-causing gene (46) using standard protocols. 2. Grow a 1 L culture at 37°C in LB medium containing 100 mg/L of ampicillin (final concentration is 1× per Subheading 2.2, item 2). (As a negative control, grow a 1  L culture of JM109(DE3)

312

M.D. Disney

cells that have not been transformed. Do not use ampicillin in the LB medium. Follow steps 3–7 and 9 in Subheading 3.4.1). 3. Once the culture reaches an OD600 of around 0.5, add IPTG to a final concentration of 1  mM (final concentration is 1× for 1,000× stock in Subheading 2.2, item 5). Incubate the culture for an additional 3 h at 37°C. 4. Pellet the cell suspension by centrifugation (10 min at 5,000 × g) and wash with ice cold lysis buffer (see Subheading 2.2, item 6). 5. Resuspended the pellet in a minimal volume of lysis buffer and lyse the cells by sonication. 6. Pellet cellular debris by centrifugation (20 min at 10,000 × g) and then dialyze the supernatant against 4 L of APH(3¢) dialysis buffer (Subheading 2.2, item 7) for 24 h at 4°C. 7. Concentrate the lysate to 500 mL in using a centrifugal concentrator at 4°C. 8. Confirm the isolation of the resistance-causing enzyme by SDS-PAGE (47) and activity using a phosphocellulose capture assay (48). 9. Total protein content of the lysate can then be determined using a BCA Protein Assay Kit. 3.4.2. Preparation of ANT(2 ″)-Ia Lysate

1. Transform E. coli JM109(DE3) with pET 22b(+) plasmid containing the ANT(2″)-Ia resistance gene using standard protocols. 2. Grow a 1 L culture at 37°C in LB medium containing 50 mg/L of amplicillin (final concentration is 0.5× per Subheading 2.2, item 2) and 10 mg/L of kanamaycin A (final concentration is 1× per Subheading 2.2, item 3). (As a negative control, grow a 1 L culture of JM109(DE3) cells that have not been transformed. Do not use ampicillin or kanamycin A in the LB medium. Follow steps 3–7 and 9 in Subheading 3.4.2). 3. Once the culture reaches an OD600 of ~0.5, add IPTG to a final concentration of 0.5  mM (final concentration is 0.5× for 1,000× stock in Subheading 2.2, item 5). Incubate the culture for an additional 3 h at 37°C. 4. Pellet the cells by centrifugation (10  min at 5,000 × g) and wash with ice cold lysis buffer (see Subheading 2.2, item 8). 5. Resuspended the pellet in a minimal volume of lysis buffer and lyse the cells by sonication. 6. Pellet cellular debris by centrifugation (20 min at 10,000 × g) and dialyze the supernatant 4  L of ANT(2″) dialysis buffer (Subheading 2.2, item 9) for 24 h at 4°C. 7. Concentrate the lysate to 500 mL using a centrifugal concentrator.

21  Studying Modification of Aminoglycoside Antibiotics…

313

8. Confirm the isolation of the resistance-causing enzyme by SDS-PAGE (47) and activity using a phosphocellulose capture assay (48). 9. Total protein content of the lysate can then be determined using a BCA Protein Assay Kit. 3.4.3. Preparation of AAC(3)-IV Lysate

1. Transform E. coli BL21(DE3) pLysS with pET 22a plasmid containing the AAC(3)-IV resistance-causing gene using standard protocols. 2. Grow a 1 L culture at 37°C in LB medium containing 50 mg/L of carbenicillin (final concentration is 1× per Subheading 2.2, item 4) for 24 h. (As a negative control, grow a 1 L culture of JM109(DE3) cells that have not been transformed. Do not use carbenicillin in the LB medium. Follow steps 3–7 and 9 in Subheading 3.4.3). 3. Pellet the cells by centrifugation (10 min at 5,000 × g). 4. Resuspend the pellet in 25  mL of AAC(3)-IV lysis buffer (Subheading 2.2, item 10). 5. Lyse the cells by sonication. Add 50 U of DNase I and then stir the solution on ice for 30 min. 6. Pellet cellular debris by centrifugation (20 min at 10,000 × g) and dialyze the supernatant against 4 L of AAC(3)-IV dialysis buffer (Subheading 2.2, items 9 and 11). 7. Concentrate the lysate to 500  mL using a centrifugal concentrator. 8. Confirm the isolation of the resistance-causing enzyme by SDS-PAGE (47) and activity using a phosphocellulose capture assay (48). Activity can also be confirmed by using a spectrophotometric assay as previously described (46). 9. Total protein content of the lysate can then be determined using a BCA Protein Assay Kit.

3.5. Modification of Array-Displayed Aminoglycosides by ResistanceCausing Enzymes

This section describes the mechanics of completing a modification experiment on an array surface. There are two ways in which these experiments can be completed: using a silicon gasket that produces microwells or hybridizing the entire microarray surface. Gaskets that afford 50 microwells are available from Grace Bio Labs (Bend, OR). Each well can hold approximately 12 mL of solvent. When hybridizing the entire slide surface with a solution containing the resistance-causing enzyme, a hybridization chamber or a hydrophobic marker can be used. It is important that hybridization chambers cover the entire array surface. [A variety of chambers can be purchased from Sigma-Aldrich (Milwaukee, WI).] A hydrophobic marker, or a PAP pen, can be used to draw a rectangle on the diameter of an array and can be purchased from Abchem (San Francisco, CA).

314

M.D. Disney

PAP pens are much less expensive than hybridization chambers and allow customization of the array into various segments. 3.5.1. Modification of an Array Surface with 32 P by Using ANT(2 ″)-Ia and APH(3¢)-IIIa. Results Are Illustrated in Fig. 4

1. Pre-hybridize an air-dried microarray with 1× ANT and APH assay buffer. It is important to pre-hybridize arrays as application of a radioactive solution to a dry array leads to nonspecific deposition of radioactivity that is very difficult to remove. (a) For arrays with microwells, add 12 mL to each well, incubate for 5 min, and remove the buffer. Repeat four times. (b) For arrays that are affixed with a hybridization chamber or that use a PAP pen, add 1.5–2.0 mL of 1× ANT and APH assay buffer to the array surface and incubate for at least 5 min. Just prior to the addition of the resistance-causing enzyme solution, tip the array on its side to remove the buffer from the surface. 2. To assay aminoglycoside modification, add 2.3 nmol of phosphoenolpyruvate, 0.018  U of pyruvate kinase, 11.4  nmol of

Fig. 4.  A microarray that is used to probe modification of array immobilized kanamaycin A and tobramycin by APH(3¢)-IIIa. Radioactivity is only deposited where the reactive kanamcyin A is deposited on the array surface.

21  Studying Modification of Aminoglycoside Antibiotics…

315

ATP to 12  mL of 1× ANT, or APH assay buffer. For APH modification, add 0.2 OD260 of cell lysate and 500,000 CPM of (g-32P]ATP; for ANT modification, add 0.8 OD260 of ANT and 500,000 CPM of (a-32P]ATP. For control experiments, substitute the cell lysate with one that does not contain a resistancecausing enzyme. 3. Incubate the arrays for 12  h at 37°C and 100% humidity to prevent evaporation. 4. Remove the gasket from the surface and wash the slide by submerging it in a solution of 0.2% SDS (2× 15 min) and then in water for 30 min at 37°C. 5. Air dry the array surface, cover with plastic wrap, and expose in a phosphorimager cassette. 3.5.2. Modification of an Array Surface with 14 C by Using AAC(3)-IV

1. Pre-hybridize an air-dried microarray with 1× AAC(3)-IV assay buffer for at least 5 min at room temperature. (These experiments were only completed using arrays constructed with a PAP pen.) Just prior to the addition of the resistance-causing enzyme solution, tip the array on its side to remove the buffer from the surface. 2. To assay aminoglycoside modification, prepare 0.13  nmol 14 C-Acetyl coenzyme A (AcCoA), 5 nmol of unlabelled AcCoA, and 0.3  mg of total protein lysate in 240  mL of AAC(3)-IV assay buffer. For control experiments, add 0.3 mg of total cell lysate that does not contain a resistance-causing enzyme. 3. Incubate the arrays for 20  h at 37°C and 100% humidity to prevent evaporation. 4. Remove the gasket from the surface. Wash the slide by submerging it in a solution of 0.2% SDS (2× 15 min) and then in water for 30 min at 37°C. 5. Air dry the array surface, cover with plastic wrap, and expose in a phosphorimager cassette for at least 48 h. If signal is low or background is high, see Notes 1 & 2.

3.6. Hybridization of the Arrays with a Fluorescently Labeled Mimic of the Bacterial rRNA A-Site. Results Are Illustrated in Fig. 5

The activity of a variety of aminoglycoside resistance-causing enzymes has been probed on a microarray surface via these routes. In all but one case, the extent of array modification was high enough to observe a decrease in binding to an oligonucleotide mimic of the rRNA A-site. Interestingly, in the case in which decreased binding to the A-site was not observed, the aminoglycoside modifying enzyme used was not known to confer resistance to aminoglycosides in vivo (34). Thus, the extent of modification of arrayed ligands is critical to observe a decrease in ligand binding to biomolecules.

316

M.D. Disney

Fig. 5. Overall approach to using microarrays to study antibacterial resistance that is conferred by aminoglycoside modification. The image illustrates results of binding of array-immobilized substrates that have been modified by APH(3¢)-IIIa to a mimic of the bacterial rRNA A-site. The plot below the array image illustrates the decrease in affinity of the arrayed aminoglycoside upon modification by APH(3¢)-IIIa; the orange bar shows the decreased signal due to the modification of kanamaycin A by APH(3¢)-IIIa.

1. Pre-equilibrate the modified slide from Subheading  3.5.1 or 3.5.2 with 600 mL of 1× RNA hybridization buffer containing 200 mg/mL of BSA for at least 5 min at room temperature. 2. During the pre-hybridization, prepare a 600  mL solution of 1 mM of a fluorescently labeled RNA in 1× RNA hybridization buffer. Fold the RNA by heating at 95°C for 4 min and then cooling to room temperature. After cooling, add BSA to the sample to a final concentration of 100 mg/mL. 3. Pipette the solution containing the folded RNA onto the array surface and spread evenly using a custom-cut sheet of parafilm (same dimensions as the array surface). 4. Incubate the array for 40  min at room temperature in the dark. 5. Remove unbound RNA by delivering 10 × 1 mL aliquots of 1× RNA hybridization buffer containing 200 mg/mL of BSA to the surface. Remove any salt stuck to the surface by delivering 5 × 1 mL aliquots of water. 6. Image the slide using a microarray scanner. If signal is low or background is high, see Notes 1 & 2.

21  Studying Modification of Aminoglycoside Antibiotics…

317

4. Notes 1. If microarray features are faint or missing after incubation with resistance-causing enzymes, attempt one of the following: (a) Check that the spotting solution was prepared properly. (b) Check that the assay buffer was prepared properly. (c) Use a new stock of radioactively labeled substrate ([a-] or [g-32P] ATP or 14C-acetyl Coenzyme A). (d) Ensure that the total concentration of substrate added is greater than the Km. (e) Ensure that the resistance-causing enzyme was overexpressed and is active. ●●

●●

●●

Compare the protein content of cellular lysates in which the resistance-causing enzyme was and was not expressed by SDS-PAGE. The resistance-causing enzyme should be a significant percentage of total protein (>30%). If this is not the case, ensure that the IPTG solution was prepared correctly. If so, either induce the cells with a higher concentration of IPTG (2 mM) or increase the induction time. Optimization of induction can be assessed by SDS-PAGE. Test the activity of the cellular lysates using a ­phosphocellulose assay (48) or a spectrophotometric assay (46). Increase the amount of cellular lysate incubated with the slide surface. The amount of control cellular lysate should be increased to the same extent.

(f) If testing acetyltransferases, add a deacetylase inhibitor such as butyric acid. 2. If the slides have high background after incubating with resistance-causing enzymes or fluorescently labeled RNA, check the following: (a) Reactive groups were not quenched after the immobilization of aminoglycosides. NaBH3CN is hygroscopic; obtain a new stock if clumps are present. (b) Ensure that the slides were pre-hybridized prior to incubation with resistance-causing enzymes or fluorescently labeled RNA. Addition of 1% BSA may reduce the background. (c) Wash the slides in 0.2% SDS for 5  min with gentle agitation.

318

M.D. Disney

Acknowledgments I thank Olivia Barrett, Alexei Pushechnikov, Meilan Wu, Pavel Tsitovich, and Jon French for their contributions to this project. Funding for this work was provided by the Research Corporation through a Cottrell Scholar award, the Camille and Henry Dreyfus Foundation through a New Investigator in the Chemical Sciences Award and a Teacher-Scholar Award, by NYSTAR through a J.D. Watson Young Investigator award, and by the National Institutes of Health (R01-GM079235). References 1. Tenson, T., and Mankin, A. (2006) Antibiotics and the ribosome, Mol Microbiol 59, 1664–1677. 2. Gallego, J., and Varani, G. (2001) Targeting RNA with small-molecule drugs: therapeutic promise and chemical challenges, Acc Chem Res 34, 836–843. 3. Magnet, S., and Blanchard, J. S. (2005) Molecular insights into aminoglycoside action and resistance, Chem Rev 105, 477–498. 4. Wright, G. D., Berghuis, A. M., and Mobashery, S. (1998) Aminoglycoside antibiotics. Structures, functions, and resistance, Adv Exp Med Biol 456, 27–69. 5. Francois, B., Russell, R. J., Murray, J. B., Aboul-ela, F., Masquida, B., Vicens, Q., and Westhof, E. (2005) Crystal structures of complexes between aminoglycosides and decoding A site oligonucleotides: role of the number of rings and positive charges in the specific binding leading to miscoding, Nucleic Acids Res 33, 5677–5690. 6. Funatsu, G., and Wittmann, H. G. (1972) Ribosomal proteins. 33. Location of aminoacid replacements in protein S12 isolated from Escherichia coli mutants resistant to streptomycin, J Mol Biol 68, 547–550. 7. Moazed, D., and Noller, H. F. (1987) Interaction of Antibiotics With Functional Sites in 16 s Ribosomal-Rna, Nature 327, 389–394. 8. Borovinskaya, M. A., Pai, R. D., Zhang, W., Schuwirth, B. S., Holton, J. M., Hirokawa, G., Kaji, H., Kaji, A., and Cate, J. H. (2007) Structural basis for aminoglycoside inhibition of bacterial ribosome recycling, Nat Struct Mol Biol 14, 727–732. 9. Fourmy, D., Recht, M. I., Blanchard, S. C., and Puglisi, J. D. (1996) Structure of the A site of Escherichia coli 16  S ribosomal RNA complexed with an aminoglycoside antibiotic, Science 274, 1367–1371.

10. Wong, C. H., Hendrix, M., Priestley, E. S., and Greenberg, W. A. (1998) Specificity of aminoglycoside antibiotics for the A-site of the decoding region of ribosomal RNA, Chemistry & Biology 5, 397–406. 11. Purohit, P., and Stern, S. (1994) Interactions of a small RNA with antibiotic and RNA ligands of the 30 S subunit, Nature 370, 659–662. 12. Fourmy, D., Yoshizawa, S., and Puglisi, J. D. (1998) Paromomycin binding induces a local conformational change in the A-site of 16  S rRNA, J Mol Biol 277, 333–345. 13. Lynch, S. R., Gonzalez, R. L., and Puglisi, J. D. (2003) Comparison of X-ray crystal structure of the 30  S subunit-antibiotic complex with NMR structure of decoding site oligonucleotide-paromomycin complex, Structure (Camb) 11, 43–53. 14. Recht, M. I., Fourmy, D., Blanchard, S. C., Dahlquist, K. D., and Puglisi, J. D. (1996) RNA sequence determinants for aminoglycoside binding to an A-site rRNA model oligonucleotide, J Mol Biol 262, 421–436. 15. Yoshizawa, S., Fourmy, D., and Puglisi, J. D. (1999) Recognition of the codon-anticodon helix by ribosomal RNA, Science 285, 1722–1725. 16. Schlunzen, F., Zarivach, R., Harms, J., Bashan, A., Tocilj, A., Albrecht, R., Yonath, A., and Franceschi, F. (2001) Structural basis for the interaction of antibiotics with the peptidyl transferase centre in eubacteria, Nature 413, 814–821. 17. Yonath, A. (2003) Structural insight into functional aspects of ribosomal RNA targeting, Chembiochem 4, 1008–1017. 18. Carter, A. P., Clemons, W. M., Brodersen, D. E., Morgan-Warren, R. J., Wimberly, B. T., and Ramakrishnan, V. (2000) Functional insights from the structure of the 30 S ribosomal subunit and its interactions with antibiotics, Nature 407, 340–348.

21  Studying Modification of Aminoglycoside Antibiotics… 19. Barbieri, C. M., Kaul, M., and Pilch, D. S. (2007) Use of 2-aminopurine as a fluorescent tool for characterizing antibiotic recognition of the bacterial rRNA A-site, Tetrahedron 63, 3567–6574. 20. Kaul, M., Barbieri, C. M., and Pilch, D. S. (2004) Fluorescence-based approach for detecting and characterizing antibiotic-induced conformational changes in ribosomal RNA: comparing aminoglycoside binding to prokaryotic and eukaryotic ribosomal RNA sequences, J Am Chem Soc 126, 3447–3453. 21. Kaul, M., Barbieri, C. M., and Pilch, D. S. (2006) Aminoglycoside-induced reduction in nucleotide mobility at the ribosomal RNA a-site as a potentially key determinant of antibacterial activity, J Am Chem Soc 128, 1261–1271. 22. Shandrick, S., Zhao, Q., Han, Q., Ayida, B. K., Takahashi, M., Winters, G. C., Simonsen, K. B., Vourloumis, D., and Hermann, T. (2004) Monitoring molecular recognition of the ribosomal decoding site, Angew Chem Int Ed Engl 43, 3177–3182. 23. Pfister, P., Hobbie, S., Vicens, Q., Bottger, E. C., Westhof, E., Vicens, Q., and Westhof, E. (2003) The molecular basis for A-site mutations conferring aminoglycoside resistance: relationship between ribosomal susceptibility and X-ray crystal structures, Chembiochem 4, 1078–1088. 24. Vetting, M., Roderick, S. L., Hegde, S., Magnet, S., Blanchard, J. S., Vetting, M. W., Hegde, S. S., Javid-Majd, F., Dam, T. K., and  Brewer, C. F. (2003) What can structure tell us about in  vivo function? The case of aminoglycoside-resistance genes, Biochem Soc Trans 31, 520–522. 25. Wright, G. D. (2003) Mechanisms of resistance to antibiotics, Curr Opin Chem Biol 7, 563–569. 26. Llano-Sotelo, B., Azucena, E. F., Jr., Kotra, L. P., Mobashery, S., and Chow, C. S. (2002) Aminoglycosides modified by resistance enzymes display diminished binding to the bacterial ribosomal aminoacyl-tRNA site, Chem Biol 9, 455–463. 27. Disney, M. D., Magnet, S., Blanchard, J. S., and Seeberger, P. H. (2004) Aminoglycoside Microarrays To Study Antibiotic Resistance, Angew Chem Int Ed Engl 43, 1591–1594. 28. Disney, M. D., and Seeberger, P. H. (2004) The use of carbohydrate microarrays to study carbohydrate-cell interactions and to detect pathogens, Chem Biol 11, 1701–1707. 29. Disney, M. D., and Seeberger, P. H. (2004) Aminoglycoside microarrays to explore interactions of antibiotics with RNAs and proteins, Chemistry 10, 3308–3314.

319

30. Ratner, D. M., Adams, E. W., Disney, M. D., and Seeberger, P. H. (2004) Tools for ­glycomics: mapping interactions of carbohydrates in biological systems, Chembiochem 5, 1375–1383. 31. Blixt, O., Head, S., Mondala, T., Scanlan, C., Huflejt, M. E., Alvarez, R., Bryan, M. C., Fazio, F., Calarese, D., Stevens, J., Razi, N., Stevens, D. J., Skehel, J. J., van Die, I., Burton, D. R., Wilson, I. A., Cummings, R., Bovin, N., Wong, C. H., and Paulson, J. C. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins, Proc Natl Acad Sci U S A 101, 17033–17038. 32. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions, Nat Biotechnol 20, 1011–1017. 33. Park, S., Lee, M. R., and Shin, I. (2007) Fabrication of carbohydrate chips and their use to probe protein-carbohydrate interactions, Nat Protoc 2, 2747–2758. 34. Barrett, O. J., Pushechnikov, A., Wu, M., and Disney, M. D. (2008) Studying aminoglycoside modification by the acetyltransferase class of resistance-causing enzymes via microarray, Carbohydr Res 343, 2924–2931. 35. Disney, M. D., and Barrett, O. J. (2007) An aminoglycoside microarray platform for directly monitoring and studying antibiotic resistance, Biochemistry 46, 11223–11230. 36. Tsitovich, P. B., Pushechnikov, A., French, J. M., and Disney, M. D. (2010) A chemoenzymatic route to diversify aminolgycosides enables a microarray-based method to probe acetyltransferase activity, Chembiochem 11, 1656–1660. 37. Childs-Disney, J. L., Wu, M., Pushechnikov, A., Aminova, O., and Disney, M. D. (2007) A small molecule microarray platform to select RNA internal loop-ligand interactions, ACS Chem Biol 2, 745–754. 38. Disney, M. D., Labuda, L. P., Paul, D. J., Poplawski, S. G., Pushechnikov, A., Tran, T., Velagapudi, S. P., Wu, M., and Childs-Disney, J. L. (2008) Two-dimensional combinatorial screening identifies specific aminoglycosideRNA internal loop partners, J Am Chem Soc 130, 11185–11194. 39. Velagapudi, S. P., Seedhouse, S. J., and Disney, M. D. (2010) Structure-activity relationships through sequencing (StARTS) defines optimal and suboptimal RNA motif targets for small molecules, Angew Chem Int Ed Engl 49, 3816–3818. 40. Tran, T., and Disney, M. D. (2010) Twodimensional combinatorial screening of a bacterial rRNA A-site-like motif library: defining

320

M.D. Disney

privileged asymmetric internal loops that bind aminoglycosides, Biochemistry 49, 1833–1842. 41. Labuda, L. P., Pushechnikov, A., and Disney, M. D. (2009) Small molecule microarrays of RNA-focused peptoids help identify inhibitors of a pathogenic group I intron, ACS Chem Biol 4, 299–307. 42. Aminova, O., and Disney, M. D. (2010) A microarray-based method to perform nucleic acid selections, Methods Mol Biol, 669, 209–224. 43. Chan, T. R., Hilgraf, R., Sharpless, K. B., and Fokin, V. V. (2004) Polytriazoles as copper(I)stabilizing ligands in catalysis, Org Lett 6, 2853–2855. 44. McKay, G. A., Thompson, P. R., and Wright, G. D. (1994) Broad spectrum aminoglycoside phosphotransferase type III from Enterococcus: overexpression, purification,

and substrate specificity, Biochemistry 33, 6936–6944. 45. Wright, E., and Serpersu, E. H. (2004) Isolation of aminoglycoside nucleotidyltransferase(2″)-Ia from inclusion bodies as active, monomeric enzyme Protein Express Purif 35, 373–380. 46. Magalhaes, M. L., and Blanchard, J. S. (2005) The kinetic mechanism of AAC3-IV aminoglycoside acetyltransferase from Escherichia coli, Biochemistry 44, 16275–16283. 47. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, Vol. 1, 2, and 3, Cold Spring Harbor Laboratory Press, NY. 48. Wybenga-Groot, L. E., Draker, K., Wright, G. D., and Berghuis, A. M. (1999) Crystal structure of an aminoglycoside 6¢-N-acetyltransferase: defining the GCN5-related N-acetyltransferase superfamily fold, Structure 7, 497–507.

Chapter 22 Microarray Method for the Rapid Detection of Glycosaminoglycan–Protein Interactions Claude J. Rogers and Linda C. Hsieh-Wilson Abstract Glycosaminoglycans (GAGs) perform numerous vital functions within the body. As major components of the extracellular matrix, these polysaccharides participate in a diverse array of cell-signaling events. We have developed a simple microarray assay for the evaluation of protein binding to various GAG subclasses. In a single experiment, the binding to all members of the GAG family can be rapidly determined, giving insight into the relative specificity of the interactions and the importance of specific sulfation motifs. The arrays are facile to prepare from commercially available materials. Key words: Glycosaminoglycan, Chondroitin sulfate, Dermatan sulfate, Heparan sulfate, Heparin, Hyaluronic acid, Microarray, Growth factor, Glycosaminoglycan-binding protein

1. Introduction Glycosaminoglycans (GAGs) are a large family of linear polysaccharides that fulfill diverse functions in vivo, such as joint lubrication and movement (1), cell signaling and development, angiogenesis (2), axonal growth (3), viral invasion (4), spinal cord injury (5, 6), tumor progression (7, 8), metastasis (7, 9), and anti-coagulation (10, 11). GAGs are large (typically 10–100 kDa), highly charged, and heterogeneously sulfated molecules composed of repeating disaccharide units. Members of the GAG family vary subtly in stereochemistry, length, and sulfation pattern (Fig. 1a). For instance, chondroitin sulfate (CS), the most abundant GAG in the body is composed of the repeating disaccharide d-glucuronic acid (GlcA) and N-acetyl-d-galactosamine (GalNAc). CS is further classified by the sulfation pattern of its disaccharides, the most common of

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_22, © Springer Science+Business Media, LLC 2012

321

Fig. 1. (a) Structures of glycosaminoglycan family members. Indicated hydroxyl groups can either be protonated or sulfated (R = H or SO3−). Indicated amino groups can be either protonated, acetylated, or sulfated (R¢ = H, Ac, or SO3−). (b) The common CS sulfation patterns.

322 C.J. Rogers and L.C. Hsieh-Wilson

22  Microarray Method for the Rapid Detection…

323

which are termed CS-A, -C, -D, and -E (Fig.  1b). Dermatan sulfate (DS), also known as CS-B, differs from CS in the stereochemistry of the C-5 position of the uronic acid. Heparin and heparin sulfate (HS) are composed of d-glucosamine (GlcN) and either GlcA or its C-5 epimer l-iduronic acid (IdoA). The GlcN can either be N-sulfated, protonated, or acetylated. In general, HS has more GlcA and N-acetylated GlcN then heparin, and heparin has a much higher charge density and more N-sulfated GlcN than HS. Keratan sulfate (KS) is composed of d-galactose and N-acetyld-glucosamine (GlcNAc) and is the only GAG that does not contain uronic acid. Hyaluronic acid (HA), the only unsulfated GAG, is composed of GlcA and GlcNAc. The chemical diversity of GAGs is believed to have important functional consequences, enabling a large number of protein-binding motifs to be generated from a relatively simple scaffold (12, 13). For instance, HS is important for growth factor signaling, inflammation, and blood coagulation (10, 11, 14, 15), while CS has been shown to interact with various growth factors involved in stem cell proliferation, neurogenesis and gliogenesis, and is a major component of the glial scar, an inhibitory barrier that forms after spinal cord injury (6, 16). A major challenge in understanding GAG function has been the lack of high-throughput methods to identify protein–GAG interactions. While effective, methods such as affinity chromatography, electrophoretic mobility shift assays, competition experiments, mass spectrometry-based approaches, isothermal titration calorimetry, and surface plasmon resonance are frequently labor intensive and require significant quantities of carbohydrate and/or protein. Given the diverse structure of GAGs and the large number of potential protein-binding motifs, a high-throughput approach for the discovery and study of protein–GAG interactions is needed. Moreover, the highly anionic character and other structural similarities among GAGs necessitate a method to compare the relative affinities of proteins for different GAG family members and for different sulfation patterns within a GAG class. The recent development of GAG microarrays (17–21) has enabled many of these challenges to be addressed. Microarrays allow for the rapid, simultaneous detection of multiple protein– GAG binding events and require minimal amounts of carbohydrate and protein. Methodologies have been developed for studying the binding of growth factors, cell-surface receptors, and chemokines to sulfated variants of CS and HS (17, 19, 21) and for comparing the binding specificities of proteins across various GAG classes (18, 20). Microarrays have been constructed using chemically synthesized CS and HS oligosaccharides, which have the advantage of defined sulfation patterns (17, 19, 21), or from naturally occurring polysaccharides (18, 20). Here, we describe a microarray-based approach for the study of protein–GAG interactions that employs

324

C.J. Rogers and L.C. Hsieh-Wilson

commercially available sugars and simple adsorption to affix the sugars to the array surface. The microarrays are relatively inexpensive, easy to prepare, and enable the rapid evaluation of proteinbinding specificities across the entire GAG family in a single assay.

2. Materials 2.1. Slide Preparation

1. Microslides (25 × 75 × 1.0  mm, VWR; West Chester, PA, see Note 1). 2. Phosphate-buffered saline (PBS): Prepare 10× stock with 1.37  M NaCl, 27  mM KCl, 54  mM Na2HPO4, 18  mM KH2PO4, pH 7.4. Dilute 100 mL of 10× stock with 900 mL of water for use (see Note 2). 3. Glass staining dishes with removable racks (105 × 70 × 85 mm, Wheaton Science Products, Millville, NJ). 4. Poly-l-lysine solution: combine 80 mL of 0.1% (w/v) poly-llysine solution in H2O with 80  mL of PBS and 640  mL of water. 5. Etch solution: dissolve 150 g solid NaOH in 600 mL of water. Mix in 900 mL of 95% ethanol (see Note 3). 6. Slide box. Prior to use, blow compressed air into the box to remove any dust particles. 7. Chondroitin sulfate A, C, D, E (Seikagaku America; Fallmouth, MA) are dissolved at 500 mM in water and stored at 4°C (see Note 4). 8. Chondroitin sulfate B (known also as dermatan sulfate, SigmaAldrich; St. Louis, MO) is dissolved at 500 mM in water and stored at 4°C. 9. Hyaluronic acid (Sigma-Aldrich; St. Louis, MO) is dissolved at 500 mM in water and stored at 4°C. 10. Heparin polysaccharide VI (Neoparin; Alameda, CA) are dissolved at 500 mM in water and stored at 4°C. 11. Keratan sulfate (Seikagaku America; Fallmouth, MA) is dissolved at 500 mM in water and stored at 4°C. 12. High sample recovery 384-well plate and covers (Genetix; Boston, MA). 13. Microplate sealing film (VWR; West Chester, PA). 14. Lint-free paper: Bluesorb 750, 4 × 4, non-woven polyester/ cellulose (Berkshire; Surrey, UK). 15. 70% Ethanol solution: Dilute 37  mL of 95% ethanol with 13 mL of water. 16. Microgrid II (Biorobotics; Cambridge, UK) or other suitable arrayer.

22  Microarray Method for the Rapid Detection…

2.2. Carbazole Assay

325

1. Acid borate reagent: A solution of 0.80 g sodium tetraborate dissolved in 16.6 mL of water and 83.3 mL of sulfuric acid is stored at room temperature. 2. Carbazole reagent: 0.1% (w/v) carbazole is dissolved in 100% ethanol, protected from light and stored at 4°C. 3. Glucuronolactone standard: d-Glucuronic acid lactone (SigmaAldrich; St. Louis, MO) at 1 mg/mL in water, stored at 4°C.

2.3. Protein Binding Assay

1. Super Pap Pen (Research Products International Corp.; Mount Prospect, IL). 2. Blocking buffer: 3% (w/v) bovine serum albumin (BSA) in PBS. Alternatively, 10% (w/v) fetal bovine serum (FBS) in PBS can be used. In either case, filter through a 0.20 mm cellulose nitrate membrane and store at 4°C. 3. Protein dilution buffer: PBS supplemented with 1% (w/v) BSA, filtered through a 0.20 mm cellulose nitrate membrane and stored at 4°C. 4. Protein(s) of interest. For example: Recombinant human betaNGF (Peprotech; Rocky Hill, NJ). 5. Incubation box: DVA211 6-compartment plastic box (7 × 3.75 × 1.25 in., Durphy Packaging Co.; Ivyland, PA). 6. Primary antibody: Primary antibody/antibodies against protein of interest at 1 mg/mL. For example: Rabbit anti-human b-NGF (Peprotech; Rocky Hill, NJ). 7. Secondary antibody: Cy3- or Cy5-conjugated secondary antibody at 1 mg/mL (see Note 5). For example: Cy3 goat anti-rabbit IgG (Invitrogen; Carlsbad, CA). 8. GenePix 5000a scanner with GenePix 6.1 software (Affymetrix; Fremont, CA).

3. Methods The carbohydrate microarray methodology described herein exploits the high charge density of GAGs to affix the sugars to the array surface. Adsorption is simple and effective and allows GAGs to be used directly, without additional modification. As members of the GAG family can vary considerably in length, GAG concentration must be determined in terms of uronic acid concentration. It is necessary to normalize binding data with respect to the uronic acid concentration because longer GAG molecules may have more binding sites per mole. The carbohydrate microarray methodology is robust and provides reproducible and consistent results. The microarrays

326

C.J. Rogers and L.C. Hsieh-Wilson

are very sensitive and even weak protein–GAG interactions (e.g. KD > 10 mM) can be detected. Therefore, it is important to interpret results with caution and use independent methods to confirm that the observed binding is strong and physiologically relevant. Carbohydrate microarrays provide a powerful, rapid method to screen for novel protein–GAG interactions, but as with any method, they must be used in combination with other techniques. Furthermore, care should be taken when comparing the relative affinity of a given protein–GAG interaction to another based on the difference in fluorescence intensity between two microarrays. The difference could be due to number of factors and does not necessarily reflect a difference in affinity. A quantitative assay should be used to compare differences in affinity from protein to protein. 3.1. Preparation of Poly-l-Lysine-Coated Slides (See Note 1)

1. Place the microslides into the removable racks of the staining dishes. Examine each slide, checking for markings that cannot be removed by Kim-Wipes. Place 19 slides in each rack and place the rack into the empty dish (see Note 6). 2. Carefully pour approximately 200 mL of the etch solution into the dishes (see Note 7). Make sure that the slides are completely covered. Cover the dish with the lid and incubate the slides in the etch solution for 1 h. 3. Remove the etch solution (see Note 8), and rinse the slides in the dishes five or more times in approximately 200 mL of water for approximately 10  s, moving the rack in an up-and-down motion at a constant and consistent speed. It is critical that all of the etch solution is removed before continuing to the next step. 4. Pour approximately 200 mL of the poly-l-lysine solution into each dish, making sure the slides are covered. Place the dishes on an orbital shaker at a speed low enough that none of the poly-l-lysine solution will splash out. Incubate with shaking for 1 h. 5. Remove the poly-l-lysine solution and rinse the slides with water as described in Subheading 3.1, step 3 above. After the final rinse, leave the slides in water for the next step. 6. One dish at a time, remove the rack from the water and dry the rack and the slides under a stream of compressed air to remove most of the water. Then, without touching the surface, dry the slides individually under a stream of compressed air, making sure the slides are completely dry. Place the slides into the slide box (see Note 6). Once all the slides have been dried and transferred to the slide box, label and date the box, and place the box in a desiccator. Allow at least 2 weeks before printing the slides to ensure complete dryness.

22  Microarray Method for the Rapid Detection…

3.2. Preparation of Sugar Samples

327

1. Make a series of samples containing 0, 1, 3, 5, 7.5, and 10 mL of the glucuronolactone standard. Adjust the final volume to 50 mL with water. 2. For each GAG, prepare 1, 3, and 5  mL samples from the 500 mM stocks. Adjust the final volume to 50 mL with water. 3. For each sample (both the glucuronolactone and GAG dilutions, 33 samples in total), add 1 mL of acid borate reagent to a test tube, followed by the 50  mL samples prepared in Subheading 3.2, steps 1 and 2. Mix by vortexing, cover each tube with foil or parafilm, and place the samples in a boiling water bath for 10 min. 4. After cooling the samples to room temperature, add 50 mL of carbazole reagent, mix by vortexing, cover, and return the mixtures to the boiling water bath for an additional 15 min. 5. After cooling the samples, measure the absorbance at 530 nm. For the d-glucuronolactone standards and the GAG samples, plot volume of stock used versus absorbance, and determine the slope of the resulting curve using linear regression analysis. Determine the molarity of each sample by dividing the slope of the GAG dilution series by that of the d-glucuronolactone standard, then divide the quotient by the average molecular weight of the GAG.

3.3. Printing Slides

1. These instructions assume the use of a Microgrid II arrayer. If using another instrument to print arrays, follow the manufacturer’s instructions. It is critical that the arrays have multiple replicates of each concentration of GAG and that the spot morphology is consistent. Maintaining the dimensions of the array and the location of the GAGs within the array is less important. 2. Using the concentration of the GAG samples determined in Subheading 3.2, step 5, prepare 15 mL of 0.5, 1, 2, 5, 10, 15, 20 mM samples of each GAG in water from the standardized stocks. 3. Place the samples into the high sample recovery 384-well plate. Start filling the plate at well A1, and fill the remaining wells such that a minimal number of 4 × 4 grids are filled. For example, if 16 samples are used, fill the wells A1–D4, inclusive. For these arrays, with seven concentrations of nine GAGs, use the wells between A1 and P4, with one well empty as a blank. Cover the plate with the microplate sealing film and lid and store at 4°C until use. Record the location of each sample, including the blank, in an Excel spreadsheet. Export the file as a tab delimited text file. If using an operating system that uses endof-line characters (EOLs) different from the Windows operating system (such as Unix-based systems, including Mac OS X), change the EOLs to be Windows compatible (see Note 9).

328

C.J. Rogers and L.C. Hsieh-Wilson

4. Transfer the tab separated text file with Windows compatible EOLs to the computer that operates the Microgrid II Arrayer. Open the TAS application suite program on this computer to set up the print run. 5. From the file menu, select “New Microarray.” In the new window that appears, under the “Options” tab, under “Group:” select “2. MicroSpot (384 well)” and under “Tool:” select “4 × 4 configuration.” 6. Click the “Source” tab. Under “Microplate Group”: select “Generic,” under “Microplate Type”: select “384 well (low profile),” under “Number of Plates”: type 1. Confirm the “Number of Samples” matches the number of samples in the 384-well plate in Subheading 3.3, step 2. Under “Lid Removal” check “Replace lid immediately.” Select “Remove one lid at a time.” Under “Source action” select “dwell.” 7. Under the “Target” tab, under “Tool array definition” change the “size” to be 6 × 5, and the “pitch” to be 0.500 mm. 8. Under the “Format” section, select the “n” radio button and enter “10” for the number of replicates to print and edit the location of each replicate within the print block in the “Edit” window. After the layout is saved, the selected radio button will become “Custom.” Under “Adapter Plate and Slide Layout,” enter the number of slides to print in the “targets” field. 9. Enter the dimensions of the array. Press the “Slide layout” button. Make sure the option “Mirror vertical margins” option is unchecked. Enter 18.15 mm for the top margin, and 12.90 mm for the bottom margin (see Note 10). Check the “Mirror horizontal margins” options and type 3.40  mm for Left margin, 0.00  mm for x spacing, and 11.00  mm for y spacing. The resulting array will have two identical array regions per slide with dimensions as shown in Fig. 2. A representative array with sample GAG concentrations is depicted. The concentrations, GAGs and layout of the array can be tailored to the protein of interest. 10. Under the “Target action” tab, type 0s under “Delay before spotting,” 0.6 mm under “Target Height,” 0s under “Dwell time,” 1 under “Multiple strikes.” Make sure the “soft touch” option is checked and that the “Pre-spotting” option remains unchecked. 11. Close this window, when prompted save the method. 12. From the file menu of the TAS application suite program, select “Clone tracking wizard…” Click “Next” twice, then select “No, plates do not have barcodes,” and then click “Next” again. In the type of output file dialog window, select “I already know what file type I need” and choose “Axon GAL” from the pull-down menu. Select the “Import name and ID” option

22  Microarray Method for the Rapid Detection…

329

Fig. 2. (a) The dimensions of the array on the 25 × 75 mm slide. The gray boxes represent the array regions. (b) Detail of the array regions from (a). The array features 16 blocks with ten replicates of four concentrations of GAGs. The concentrations and GAGs are labeled within each block. (c) A detail of the layout of a block. This is the block in the third row from the top, second column from the left.

and type 80 mm under the field labeled “Typically the spots I am printing are.” Click “Next,” and then select the tab delimited text file with Windows compatible EOLs that was saved on the computer in Subheading 3.3, step 3. Check the “Tab” option below. 13. After clicking “Next,” the wizard will display the contents of the imported text file. Confirm that the imported file is correct and that there are no errors. Click “Next.” Check again that the file is correct, if so, press “Output file.” Give the file a name and select a location to save it. Click “Save,” then “Next.” Transfer this file to the computer that runs the GenePix Sanner. 14. Select 16 pins from the Microgrid II arrayer accessories. Make sure that the pins are not bent or damaged in any way. Submerge the tips of the pins in 15  mM KOH in water. After 5  min, remove the pins from this solution and sonicate the tips of the pins for 5 min, while submerged in 0.01% Tween-20 in water. Rinse the pins by submerging the tips in water and sonicating for 5 min. Replace the water and repeat two more times. Rinse the pins by dipping them in 95% ethanol in water and place them on the lint-free paper to dry (see Note 11). 15. Fill the large bottle supplied with the arrayer with water. Click the “Fill 6-litre reservoir” icon on the TAS application suite program. The progress can be monitored with the icon to the left of this button. Turn on the recycling water bath pump and wait for the coolant temperature to drop to 8°C (approximately 30 min). 16. Under the “Housekeeping” menu in the TAS application suite program, click “load/unload tray 1.” Clean tray with compressed air to make sure that it is dust free. Carefully place the

330

C.J. Rogers and L.C. Hsieh-Wilson

poly-l-lysine-coated slides onto the tray after checking that they are dust free. Remove any dust with a stream of compressed air, if needed. Continue loading slides into the remaining trays if necessary. Each tray must contain exactly 30 slides. If printing a number of slides that is not divisible by 30, use dust-free plain glass slides to fill the remaining slots in the tray. Nothing will be printed on these slides, but they are necessary to maintain the vacuum applied to the tray to keep the slides in place during printing. 17. Select the “Load tool” option under the “Housekeeping” menu. Load the clean and dry pins in the orientation shown by the wizard. 18. Next, select “Load biobank” from the “Housekeeping” menu. After removing the film and placing the cover on the 384-well sample plate, place into machine. 19. On the bottom panel of the chamber in the robot, there should be three reservoirs. Fill the left-hand reservoir with water and the middle with 70% ethanol in water. Lastly, to maintain humidity in the chamber, take three 384-well plate lids, and place a few paper towels into each lid, cutting them to fit as necessary. Fill the lids with water, making sure the paper towels are saturated. Place the lids on the bottom of the chamber. Close the chamber lid, and press the “GO” icon in the TAS application suite program. This will initiate printing. 20. When the printing is finished, unload the slides via the wizard in the “Housekeeping” menu. Transfer slides into a dust-free slide box. Label the top-right corner of the slide using a diamondtipped pen (see Note 6). Store arrays in a low-humidity, dustfree desiccator. 21. Unload pins via the wizard in the “Housekeeping” menu and repeat the cleaning procedure detailed in Subheading  3.3, step 13. 22. Remove the 384-well sample plate via the wizard in the “Housekeeping” menu. If sufficient volume remains, the plate can be resealed with film, covered and stored at −20°C for an additional print run. 23. Drain the reservoir and shut down the robot. 3.4. Protein Binding Assay

1. Using a hydrophobic marker, such as a PapPen, draw a perimeter around the printed region of the slide according to the dimensions for the array region given in Subheading  3.3, step 8. This perimeter reduces the amount of protein required to fully cover the array region. However, take care not to mark the slide too close to the printed region, leaving up to 0.5 cm of space when possible. This is important because the hydrophobic marker can prevent the protein from interacting with the carbohydrate spots near the edge of the array.

22  Microarray Method for the Rapid Detection…

331

2. Place the slide in the incubation box and cover the slide with 2.5 mL of blocking buffer at 37°C for 1 h with gentle rocking. This step is necessary to prevent nonspecific interactions between the proteins and the surface of the array. 3. Remove the blocking buffer and add the protein sample (0.5–2  mM in protein dilution buffer, see Note 12) to the printed region of the slide. Make sure that the slide does not dry out before adding the protein. Also, make sure that there are no water “bridges” over the hydrophobic pen markings. If so, carefully blot dry with a Kim-Wipe. Be sure to add sufficient volume to fully cover the region (100–200 mL). Incubate at room temperature for 1–3 h. 4. Wash the slide five times for 30 s each with 2.5 mL PBS with gentle rocking. 5. Incubate the slide in 2.5  mL of a 1:1,000 or appropriate dilution (see Note 13) of primary antibody in protein dilution buffer for 1 h at room temperature with gentle rocking. Alternatively, 100–200 mL of the antibody solution can be added to the array region as described in Subheading 3.4, step 3. 6. Remove the antibody solution and wash the slide five times for 30 s each with 2.5 mL PBS with gentle rocking. 7. Incubate the slide in 2.5 mL of a 1:5,000 or appropriate dilution (see Note 13) of secondary antibody in protein dilution buffer for 1 h at room temperature with gentle rocking. 8. Remove the antibody solution and wash the slide three times for 30 s each with 2.5 mL PBS and two times for 30 s each with 2.5 mL water with gentle rocking. 9. Immediately after the final wash step, dry the slide(s) under a gentle stream of air or nitrogen. This prevents water droplets from evaporating on the array, which could potentially obscure the signal. 10. (Optional) Add a droplet (~5 mL) of a fluorescence-specific mounting medium, such as VectaShield, to the printed area of the array. Carefully place a coverslip over the drop, taking care to avoid forming any bubbles, and seal the coverslip with nail polish. 3.5. Recording Data

1. These instructions are specific for the GenePix 5000a scanner using GenePix 6.1 software. They are easily adaptable to other microarray scanners. Follow the manufacturer’s instructions. It is critical that the dye on the secondary antibody is compatible with the filters on the scanner (see Note 5) and that the array is scanned using appropriate laser power and gain. 2. On the computer controlling the GenePix 5000a scanner, open the GenePix 6.1 software and wait for the scanner to initialize.

332

C.J. Rogers and L.C. Hsieh-Wilson

3. Place the slide into the GenePix 5000a scanner. Orient the slide such that the printed region of the slide is facing down and the top of the array is pointed into the scanner. If the slides were labeled according to Subheading 3.3, step 19, the label will be in the back left of the scanner. 4. In the GenePix 6.1 program, click the “Hardware Settings” icon. In the menu that appears, there will be two fields labeled “Select Wavelength.” Select one by clicking the checkbox on the left, and make sure the other is not checked. Under the pull-down menu to the right of the “Select Wavelength” field, select the wavelength used for the experiment (532 nm for Cy3 or 635  nm for Cy5). Under “PMT Gain”: enter the desired value for the gain. A reasonable place to start is 400. Under “Power (%)”: begin at a low percent power, such as 5–10%. Under “Filter”: select the corresponding filter for the wavelength (“Standard Green” for 532 nm/Cy3 or “Standard Red” for 635 nm/Cy5). Lastly, change “Pixel Size (mm)” to 5, “Lines to Average” to 1, and “Focus Position” to 0, if necessary. 5. Click the “Preview Scan” icon. This will take a quick scan of the array. Adjust the brightness and contrast in the Tools section if necessary. While it may not be possible to distinguish the signal from the background at this resolution, the preview scan is helpful for determining whether the PMT Gain or Power needs to be increased or decreased. If so, repeat the preview scan. 6. If the preview scan is acceptable, click the “View Scan Area” icon under the Tools panel. Using the mouse, highlight the array region of the slide. A white rectangle will appear. Resize or move the rectangle with the mouse if necessary. 7. When the array region of the slide is within the area delimited by the rectangle, click the “Data Scan” icon. This takes a highresolution image of the array region of the slide. After the scan has finished, click the “File…” icon and select “Save Images.” After saving, zoom into the array region to see the signal, which should look like small, ordered spots. Adjust the brightness and contrast as needed, or rescan the image after adjusting the PMT Gain and Power if necessary. The image should look similar to Fig. 3a, b. 8. If the data scan is acceptable, click the “View Blocks” icon. Then, under the “File…” menu, click “Load Array List…” Find the .gal file that was created in Subheading 3.3, steps 11 and 12 and click “Open.” A series of boxes with small circles inside will appear. Using the mouse, select all of the boxes and move them roughly into position (i.e., over the spots corresponding to protein bound to GAG). To more precisely position the blocks, select one block at a time, zoom into the

22  Microarray Method for the Rapid Detection…

333

Fig. 3. (a) A representative image of nerve growth factor (NGF) binding to a GAG microarray as visualized using a Cy3-conjugated secondary antibody against anti-NGF. (b) An expansion of a region of the microarray from (a). The columns, from left to right, in (b) are 2 mM CS-C; 15 mM CS-D; 1 mM CS-D; 10 mM CS-E; 2 mM CS-C; 15 mM CS-D; 1 mM CS-D; 10 mM CS-E. Each concentration is repeated five times down the column. (c) Quantification of the data in (a) for binding of NGF to various GAG subclasses.

region, and move the block such that the spots (i.e., the signal) are centered in the circles. When the ideal adjustment has been achieved, press F5. Repeat for the remaining blocks (see Note 14). 9. Click the “Analyze” icon. Under the “File…” menu, click “Save Results As…” Name the file and click “Save.” This .gpr file can be opened in Excel and analyzed, as in Fig. 3c.

4. Notes 1. Alternatively, we recommend the poly-l-lysine-coated slides from Erie Scientific, Portsmouth, NH. If pre-coated slides are used, ignore Subheading  3.1 and start the procedure at Subheading 3.2.

334

C.J. Rogers and L.C. Hsieh-Wilson

2. Throughout the text, “water” refers to water that has a resistivity of 18.2 MW cm and total organic content of less than five parts per billion. 3. Caution: Wear lab coat and safety glasses when preparing this solution. This solution becomes hot when the reagents are mixed. Ensure the solution is carefully vented if mixing in a sealed container. 4. Due to the heterogeneity of GAG samples in terms of chain length, degree of sulfation, and number and type of counter ions (both within a sample and between different GAGs), the molecular weights for each sample are only approximate. Prepare 500  mM samples based on the average molecular weight for each sample. To compare different GAGs to one another, we measure the average uronic acid concentration for each sample using the carbazole assay described in Subheading 3. 5. The choice of Cy3 and Cy5 dyes was based on the scanner wavelengths of the GenePix 5000a scanner. If using a different scanner, check the manufacturer’s specifications and use dyes compatible with the instrument’s filters. 6. Wear gloves whenever handling the slides. Make sure the slides are arranged in the rack such that both sides of the slide are exposed to solution. 7. Caution: Wear lab coat and safety glasses when handling the etch solution. 8. It is possible to reuse the etch solution. The solution is good for up to 1 month, although if discoloration is observed, the solution should be remade. 9. If using Mac OSX, it is possible to convert the tab separated file to be windows compatible by using the following command in Terminal: tr ‘\r’ ‘\n’ < inputfile > outputfile. 10. The asymmetric margins will help determine the proper orientation of the slide if necessary. 11. Handle the pins very carefully and only with tweezers. 12. When testing a protein with unknown affinity to GAGs, a good starting concentration is 2  mM, although less protein can be used if the sample is precious. Some proteins have very high affinity to GAGs and will saturate the signal when incubated at 2 mM, even when scanned at extremely low laser power. If this is the case, it is necessary to reduce the concentration of protein to obtain useful data. 13. When using unknown antibodies, a good starting dilution is 1:1,000 for the primary antibody and 1:5,000 for the secondary

22  Microarray Method for the Rapid Detection…

335

antibody. However, particularly strong antibodies may require a higher dilution, and weak antibodies may require a lower dilution. 14. If the program has difficulty adjusting the grid to the signal, right click on a box and select “Block Properties,” and adjust the diameter of the circles accordingly.

Acknowledgments We thank Dr. Igor Antoshechkin, Director of the Millard and Muriel Jacobs Genetics and Genomics Laboratory at the California Institute of Technology, and Dr. Jose Luis Riechmann for assistance with printing the microarrays and Joshua M. Brown, Russell E. Roberson, and Dr. Igor Antoshechkin for critically reading the manuscript. This work was supported by the National Institutes of Health (R01 GM093627). References 1. Laurent, T. C., Laurent, U. B., and Fraser, J. R. (1996) The structure and function of hyaluronan: an overview, Immunol. Cell Biol. 74, A1–A7. 2. Iozzo, R. V., and Antonio, J. D. S. (2001) Heparan sulfate proteoglycans: heavy hitters in the angiogenesis arena, J. Clin. Invest. 108, 349–355. 3. Holt, C. E., and Dickson, B. J. (2005) Sugar codes for axons?, Neuron 46, 169–172. 4. Thammawat, S., Sadlon, T. A., Hallsworth, P. G., and Gordon, D. L. (2008) Role of cellular glycosaminoglycans and charged regions of viral G protein in human metapneumovirus infection, J. Virol. 82, 11767–11774. 5. Galtrey, C. M., and Fawcett, J. W. (2007) The role of chondroitin sulfate proteoglycans in regeneration and plasticity in the central nervous system, Brain Res. Rev. 54, 1–18. 6. Sugahara, K., Mikami, T., Uyama, T., Mizuguchi, S., Nomura, K., and Kitagawa, H. (2003) Recent advances in the structural biology of chondroitin sulfate and dermatin sulfate, Curr. Opin. Struct. Biol. 13, 612–620. 7. Lui, D., Shriver, Z., Qi, Y., Venkataraman, G., and Sasisekharan, R. (2002) Dynamic regulation of tumor growth and metasis by heparan sulfate glycosaminoglycans, Semin. Thromb. Hemost. 28, 67–78. 8. Tímár, J., Lapis, K., Dudás, J., Sebestyén, A., Kopper, L., and Kovalszky, I. (2002) Proteo­ glycans and tumor progression: Janus-faced

molecules with contradictory functions in cancer, Semin. Cancer Biol. 12, 173–186. 9. Sanderson, R. D. (2001) Heparan sulfate proteoglycans in invasion and metastasis, Semin. Cell Dev. Biol. 12, 89–98. 10. Casu, B., Guerrini, M., and Torri, G. (2004) Structural and conformational aspects of the anticoagulant and anti-thrombotic activity of heparin and dermatan sulfate, Curr. Pharm. Des. 10, 939–949. 11. Fareed, J., Hoppensteadt, D. A., and Bick, R. L. (2000) An update on heparins at the beginning of the new millennium, Semin. Thromb. Hemost. 26, 5–18. 12. Capila, I., and Linhardt, R. J. (2002) Heparinprotein interactions, Angew. Chem. Int. Ed. 41, 391–412. 13. Whitelock, J. M., and Iozzo, R. V. (2005) Heparan sulfate: a complex polymer charged with biological activity, Chem. Rev. 105, 2745–2764. 14. Rubin, J. B., Choi, Y., and Segal, R. A. (2002) Cerebellar proteoglycans regulate sonic hedgehog responses during development, Development 129, 2223–2232. 15. Taylor, K. R., and Gallo, R. L. (2006) Glycosaminoglycans and their proteoglycans: host-associated molecular patterns for initiation and modulation of inflammation, FASEB J. 20, 9–22.

336

C.J. Rogers and L.C. Hsieh-Wilson

16. Sugahara, K., and Mikami, T. (2007) Chondroitin/ dermatan sulfate in the central nervous system, Curr. Opin. Struct. Biol. 17, 536–545. 17. de Paz, J. L., Noti, C., and Seeberger, P. H. (2006) Microarrays of synthetic heparin oligosaccharides, J. Am. Chem. Soc. 128, 2766–2767. 18. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions., Nat. Biotechnol. 20, 1011–1017. 19. Gama, C. I., Tully, S. E., Sotogaku, N., Clark, P. M., Rawatand, M., Vaidehi, N., Goddard, W.

A., Nishi, A., and Hsieh-Wilson, L. C. (2006) Sulfation patterns of glycosaminoglycans encode molecular recognition and activity, Nat. Chem. Biol. 2, 467–473. 20. Shipp, E. L., and Hsieh-Wilson, L. C. (2007) Profiling the sulfation specificities of glycosaminoglycan interactions with growth factors and chemotactic proteins using microarrays, Chem. Biol. 14, 195–208. 21. Tully, S. E., Rawat, M., and Hsieh-Wilson, L. C. (2006) Discovery of a TNF-alpha antagonist using chondroitin sulfate microarrays, J. Am. Chem. Soc. 128, 7740–7741.

Chapter 23 Neoglycolipid-Based “Designer” Oligosaccharide Microarrays to Define b-Glucan Ligands for Dectin-1 Angelina S. Palma, Yibing Zhang, Robert A. Childs, Maria A. Campanero-Rhodes, Yan Liu, Ten Feizi, and Wengang Chai Abstract In this chapter, we describe the key steps of the “designer” oligosaccharide microarray approach we followed to prove the carbohydrate binding activity and define the oligosaccharide ligands for Dectin-1, an atypical C-type lectin-like signaling receptor of the mammalian innate immune system with a key role in anti-fungal immunity. The term “designer” microarray, which we introduced in the course of the Dectin-1 study refers to a microarray of oligosaccharide probes generated from ligand-bearing glycoconjugates to reveal the oligosaccharide ligands they harbor, so that these can be isolated and characterized. Oligosaccharide probes were generated from two polysaccharides, one that was bound by Dectin-1 and known to be rich in b1,3glucose sequence and another that was not bound and was rich in b1,6-glucose sequence and served as a negative control. The approach involved: classic ELISA-type binding assays to select the polysaccharides; partial depolymerization of the polysaccharides by chemical hydrolysis; fractionation by size of the glucan oligosaccharides obtained and determination of their chain lengths by mass spectrometry; detection of Dectin-1 ligand-positive and ligand-negative oligosaccharides using the neoglycolipid (NGL) technology; methylation analysis of oligosaccharides to derive glucose linkage information, and incorporation of the newly generated glucan oligosaccharide probes into microarrays encompassing diverse mammalian-type and exogenous sequences for microarray analysis of Dectin-1. Key words: Designer microarrays, b-Glucan polysaccharides, Depolymerization of polysaccharides, b1,3- and b1,6-glucan oligosaccharides, Neoglycolipids

1. Introduction Dectin-1 was initially identified as a C-type lectin-like receptor that recognizes an unidentified ligand on T lymphocytes (1) and soon thereafter as a b-glucan receptor following a screen of a murine macrophage cDNA expression library for binding to zymosan, a b-glucan-rich extract of Saccharomyces cerevisiae (2). Subsequent Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_23, © Springer Science+Business Media, LLC 2012

337

338

A.S. Palma et al.

studies established that Dectin-1 is a signaling receptor of the innate immune system directed against fungal b-glucans and that it mediates phagocytosis and the inflammatory mediator production in innate immunity to fungal pathogens (3, 4). The fold of the lectin-like domain of Dectin-1 is related to those of the C-type (calcium-dependent) lectin family (5), but it lacks the calcium ligating elements that mediate carbohydrate binding in typical C-type lectins (1, 2, 6). Thus, it was important to establish that the calcium-independent binding to the polysaccharides (7) is mediated by carbohydrates and not by contaminating non-carbohydrate materials. We investigated the oligosaccharide ligands for Dectin-1 using the NGL-based oligosaccharide microarray technology, a unique approach for constructing microarrays of lipid-linked oligosaccharide probes from desired glycoconjugates (7–11). We generated designer microarrays of glucan oligosaccharides from polysaccharides and used these in conjunction with diverse, sequence-defined, predominantly mammalian-type, oligosaccharide probes. We showed that the oligosaccharide ligands for Dectin-1 are highly restricted; these are linear b1,3-linked glucose oligomers, the minimum chain length required for recognition being a 10- or 11-mer (Fig. 5), which is unusually long for a lectin (7). Thus, the oligosaccharide ligands assigned for Dectin-1 are exogenous and, so far, binding has not been observed to any of the 270 endogenous mammalian type, oligosaccharide probes investigated (7). In this chapter, we dwell on requirements for the assignment of the oligosaccharide ligands for Dectin-1. The general principles apply to the assignment of oligosaccharide ligands for other carbohydrate-recognition systems. These include: (1) Soluble form of Dectin-1, ideally a recombinant form with an immuno-detectable tag; (2) Dectin-1 ligand-positive and ligand-negative glucan polysaccharides; (3) Partial depolymerization of the polysaccharides and separation of the fragments into oligosaccharide fractions with defined chain lengths; (4) Assignments of oligosaccharide chain length and monosaccharide linkage position by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) and GC-MS methylation analyses, respectively; (5) Conversion of the oligosaccharides into NGLs by chemical derivatization with an aminolipid reagent; (6) Simple and quick (optional) exploratory experiments, namely, spotting of the glucan oligosaccharide NGL probes onto nitrocellulose membranes (macroarrays) to identify ligand-positive oligosaccharide fractions this can be omitted in favor of proceeding to microarrays; (7) Generation of microarrays of NGL probes that include, in addition to the glucan oligosaccharide NGLs, a diverse range of sequence-defined oligosaccharide probes, many of which are of the mammalian type, to examine broadly the specificity of oligosaccharide recognition.

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

339

2. Materials 2.1. Interactions of Dectin-1 with Soluble Polysaccharides

1. Recombinant soluble Dectin-1: For example, murine Dectin1-human IgG Fc-chimera (Fc-Dectin-1) (see Note 1). 2. Polysaccharides (see Note 2): (a) b1,3-Linked backbone with b1,6-branches glucans: Poly(1,6)-b-d-glucopyranosyl-(1,3)-b-d-glucopyranose (PGG) (MW 120–205  kDa) and neutral soluble glucan (NSG) (MW 10–20  kDa) both isolated from S. cerevisiae (12) obtained from Biothera; laminarin (Laminaria digitata, MW 7.7 kDa) (13) from Sigma; (b) b1,6-Linked backbone glucan: pustulan (Umbilicaria papullosa, MW 20 kDa) (14) from Calbiochem; (c) Mixed b1,3- and b1,4-linked backbone glucan: lichenan (Icelandic moss) (15) from Megazyme and barley glucan (MW 190 kDa) (16) from Sigma; (d) a1,6-Linked backbone glucan with minor (5%) a1,3branches: dextran (MW 500  kDa) (17) from Amersham Biosciences; (e) Mixed a1,6- and a1,4-linked backbone glucan: pullulan (Pullularia pullulans) (16) from Megazyme. (f   ) a1,6-Linked mannose: mannan (S. cerevisiae) (18) from Sigma. 3. Curdlan polysaccharide (b1,3-linked glucan) hydrolysate, obtained by acid hydrolysis, from Megazyme (see Note 3). 4. Sodium phosphate buffer (PBS):10 mM phosphate buffer, pH 7.4, containing 137 mM NaCl. 5. Polystyrene wells (96-well Technologies Ltd.).

Immulon-4

plates;

Dynex

6. Tris-buffered saline (TBS): 10 mM Tris–HCl buffer, pH 8.0, 150 mM NaCl. 7. Blocking solution: 1% (w/v) casein in TBS from Pierce (casein Pierce blocker solution). 8. Diluent solution: 0.1% (w/v) casein in TBS. 9. Biotinylated goat anti-human IgG heavy and light (H + L) chains (anti-IgG) (Vector Laboratories). 10. Streptavidin conjugated to horseradish peroxidase from Pierce. 11. Horseradish peroxidase substrate: O-phenylenediamine hydrochloride (Sigma) in 100 mM phosphate citrate buffer containing 0.2% (v/v) hydrogen peroxide (OPD solution). 12. H2SO4 (3 M). 13. Plate reader equipment, e.g., ELISA Dynatech MRX microplate reader (Thermo Lab systems).

340

A.S. Palma et al.

2.2. Preparation of Oligosaccharide Fractions with Defined Chain Lengths

1. Deionized water. 2. HCl (0.01 M) in deionized water (see Note 4). 3. NaOH (0.2 M) in deionized water (see Note 4). 4. Solvent n-propanol:water (8:3 by volume) (see Note 4). 5. Gel filtration columns: Sephadex G10 resin (Amersham Biosciences) packed into a column 1.6 × 30  cm, and Bio-Gel P4 or Bio-Gel P6 resins (Bio-Rad) packed into a column 1.6 × 90 cm, with an on-line refractive index detector and auto sample collector (e.g., Gilson Fraction Collector FC203). 6. Silica gel high-performance (HP) TLC plates (Silica gel 60) with aluminum-backed silica Gel 60 high-performance thin layer chromatography (HPTLC) plates (Merck). Glass developing tank (Sigma-Aldrich or Camag). 7. Orcinol (Aldrich) staining reagent (see Note 5). 8. Nitrogen-assisted TLC applicator (e.g., Linomat IV, Camag). 9. A flying-spot scanning densitometer, e.g., Shimadzu CS-9000 dual wavelength densitometer or Camag Scanner (Muttenz).

2.3. Determination of Oligosaccharide Chain Length by MALDI-MS

1. 2,4,6-Trihydroxyacetophenone in ethanol (HPLC grade) at a concentration of 46 mg/ml (Matrix solution A). 2. Ammonium citrate (98%) in water at a concentration of 22 mg/ml (Matrix Solution B). 3. MALDI mass spectrometer, e.g., Tof Spec-2E instrument (Micromass, Manchester, UK).

2.4. Conversion of Glucan Oligosaccharides into NGLs

1. Aminolipid reagent 1,2-dihexadecyl-sn-glycero-3-phosphoethanolamine (DHPE) from Sigma-Aldrich, or the fluorescent derivative of DHPE reagent N-aminoacetyl-N-(9anthracenylmethyl)-1,2-dihexadecyl-sn-glycero-3-phosphoethanolamine (ADHP) (see Note 6). 2. ADHP stock solution: 10  nmol/ml in chloroform/methanol 1:3 (v/v). 3. Tetrabutylammonium cyanoborohydride solution: 20  mg/ml in methanol (prepare freshly). 4. Solvents: HPLC grade.

2.5. Binding Assays of Dectin-1 to Macroarrays of Glucan-Derived NGL-Probes on Nitrocellulose

1. NGL solutions prepared at 50 pmol/ml in chloroform/methanol/water (25:25:8, by volume). 2. Nitrogen-assisted TLC applicator (e.g., Linomat IV, Camag). 3. Nitrocellulose membrane (Millipore). 4. Blocking solution: 1% (w/v) casein in TBS from Pierce (casein Pierce blocker solution). 5. Tris-buffered saline (TBS): 10  mM Tris buffer, pH 8.0, 150 mM NaCl.

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

341

6. Fc-Dectin-1 at 1 mg/ml in blocking solution. 7. Biotinylated goat anti-human IgG heavy and light (H + L) chains (anti-IgG) (Vector Laboratories) (10 mg/ml in blocking solution). 8. Streptavidin conjugated to horseradish peroxidase from Pierce (10 mg/ml). 9. Peroxidase substrate FAST-3,3-diaminobenzidine peroxidase substrate (FAST-DAB) (Sigma). 2.6. Linkage Assignment by Methylation Analysis

1. Dimethyl sulphoxide (DMSO): anhydrous (>99.9%) Aldrich. 2. NaOH pellets: analytical grade, BDH. 3. Iodomethane: 99.5%, Aldrich. 4. Trifluoroacetic acid (TFA) solution: 2  M in water, Sequanal Grade, Pierce. 5. Sodium borodeuteride (0.1  M)/NaOH (0.01  M) solution: prepared from sodium borodeuteride (98%, Aldrich) and sodium hydroxide (BDH). 6. Pyridine/acetic anhydride (1:1, by volume) solution; prepared from pyridine (anhydrous 99.8%, Aldrich) and acetic anhydride (analytical grade, Aldrich). 7. Solvents chloroform, H2O, and methanol: HPLC grade. 8. GC-MS instrument, e.g., ThermoQuest Trace mass spectrometer (Thermo Electron, Hemel Hempstead, UK).

2.7. Carbohydrate Microarray Analysis of Dectin-1

1. Nitrocellulose-coated glass slide, e.g., FAST slide (16-pad) Whatman, printed with probes of interest (array format could vary; see Subheading 3.4 of Chapter 8 of this series). 2. Fluorescence microarray slide scanner with image software (e.g., ProScanArray and ScanArray-Express software, PerkinElmer LAS). 3. Incubation chamber, e.g., FAST slide 16-pad silicon gasket (Whatman). 4. Multi-slide plate, e.g., FAST frame (Whatman). 5. Murine IgG Fc-Dectin-1 chimera (see Note 1). 6. Biotinylated goat anti-human IgG heavy and light (H + L) chains (anti-IgG) (Vector Laboratories). 7. Saline buffers: Tris-buffered saline (TBS): 10  mM Tris–HCl buffer, pH 8.0, 150 mM NaCl; or HBS (5 mM HEPES, pH 7.4, 150 mM NaCl). 8. Blocking solution: 0.1% (w/v) casein in TBS from Pierce (casein Pierce blocker solution) (see Note 7). 9. Alexa Fluor-647-labeled streptavidin (Molecular Probes) (see Note 8).

342

A.S. Palma et al.

3. Methods We describe below the methods that we followed to assign the Dectin-1 ligands; and we refer to some modifications and possible improvements of the procedures used. The principles of the NGL technology and the procedures in current use to generate the NGLbased microarrays used here is the focus of Chapter 8 of this series. Due to the high complementarities of the two chapters, the reader is advised to refer to sections in Chapter 8 where appropriate. 3.1. Interactions of Dectin-1 with Soluble Polysaccharides 3.1.1. Binding Assay

1. Polysaccharide solutions are prepared, in duplicate, at half-log dilutions in PBS buffer, typically at a concentration range from 0.03 to 10 mg/ml (see Note 9). 2. The polysaccharide solutions (50 ml) are added directly to the microwells and allowed to dry at 37°C for 16 h (see Note 10). 3. Each well is washed twice with 200 ml of TBS. 4. Nonspecific binding sites are blocked by incubating the wells with casein Pierce blocker solution (100 ml per well) for 1 h at ambient temperature. 5. Each well is washed twice with 200 ml of TBS. 6. Fc-Dectin-1 solution is made in diluent solution at 1 mg/ml, and 50  ml are added to each well. Incubation is for 1.5 h at ambient temperature (see Note 11). 7. Each well is then washed three times with 200 ml of TBS. 8. The biotinylated goat anti-human-IgG at 10 mg/ml in diluent solution is added to each well (50 ml per well) and incubation is for 1 h. 9. Each well is then washed three times with 200 ml of TBS. 10. The streptavidin conjugated to horseradish peroxidase at 5 mg/ ml in diluent solution is added to each well (50  ml per well, incubation 30 min). 11. Each well is washed three times with 200 ml of TBS. 12. The OPD solution is added to each well (50 ml per well) and color is developed for 5–10  min. The reaction is stopped by the addition of 3 M H2SO4, 50 ml per well. 13. The readout is at 490 nm using a plate reader equipment. An example of the results obtained is shown in Fig. 1a, b.

3.1.2. Inhibition Assay

1. Ligand-positive polysaccharide NSG (10  mg/ml in PBS) is added to each well (50 ml per well) and allowed to dry at 37°C for 16 h. 2. For inhibition assays, Fc-Dectin-1 has to be used at a nonsaturating concentration. To evaluate this, a binding curve of the Fc-Dectin-1 to immobilized NSG is carried out, prior to

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

343

Fig. 1. Interactions of recombinant murine Dectin-1 with polysaccharides. The polysaccharides PGG (a), NSG and mannan (b) were added and dried down in microwells at the indicated concentrations, and the binding of Fc-Dectin-1 (1 mg/ml) was assayed in the presence of 2 mM CaCl2 or 10 mM EDTA. Results are expressed as the means of duplicate wells with the range indicated by error bars. (c) Binding curve of Fc-Dectin-1 to immobilized NSG (applied to wells at 10 mg/ml); the arrow indicates the non-saturating point on the binding curve (at 0.5 mg/ml) used as a positive control in the inhibition experiments. (d) Inhibition of the binding of Fc-Dectin-1 to NSG by three of the nine polysaccharides tested. Eight glucan polysaccharides and mannan were examined as inhibitors of the Fc-Decin-1 binding at the final concentrations indicated. This research was originally published in Journal of Biological Chemistry? Reproduced with permission from the American Society for Biochemistry and Molecular Biology (7).

the inhibition assay, using a series of half-log dilutions of Fc-Dectin-1 prepared in duplicate, typically at a concentration range of 0.01–3 mg/ml (Fig. 1c) (see Note 11) in diluent solution. The assay procedure for detecting the binding follows steps 2–13 in Subheading 3.1.1 above. 3. Typically 12 wells with immobilized NSG are prepared for six dilutions of inhibitor in duplicate. A ligand-negative polysaccharide is also added to four wells (10 mg/ml per well) as negative control. 4. Follow steps 2–5 in Subheading 3.1.1 above. 5. During the blocking step (as in Subheading 3.1.1 above), the inhibitor solutions are prepared in diluent solution at half-log steps to give concentrations ranging from 0.03 to 100 mg/ml. 6. Dilutions of the inhibitors (25  ml per well) are added to the wells coated with ligand-positive polysaccharide. Also included are quadruplicate wells to which diluent solution rather than inhibitor is added (25 ml per well). These are the uninhibited,

344

A.S. Palma et al.

positive control wells. Diluent solution (25 ml per well) is also added to the wells with the immobilized ligand-negative polysaccharide (negative control wells). 7. The Fc-Dectin-1 at 1 mg/ml (double the non-saturating concentration of 0.5  mg/ml) (see Note 12) prepared in diluent solution is added to each well (25 ml) and incubation is for 2 h. 8. The procedure follows steps 7–13 in Subheading 3.1.1 above (see Note 13). An example of the results obtained is shown in Fig. 1d. 3.2. Preparation of Glucan Oligosaccharide Fractions with Defined Chain Lengths 3.2.1. Depolymerization of Polysaccharides and Fractionation of the Oligosaccharides

A series of glucan oligosaccharides of differing chain lengths can be prepared after partial depolymerization of the polysaccharides by acid hydrolysis as described below (see Note 14).

1. Partial depolymerization of NSG (100 mg in terms of hexose content) can be carried out by mild acid hydrolysis using 10  mM HCl (10  ml). The reaction mixture is incubated at 100°C for 2  h. For depolymerization of pustulan by acid hydrolysis (see Note 15), the polysaccharide (100  mg) is treated with 10 ml of 0.2 M HCl. The reaction mixture is incubated at 100°C for 8 h and the reaction is stopped by neutralization with aqueous NaOH solution. The estimation of hexose content can be determined by TLC-dot orcinol assay (see Subheading 3.2.2 below). 2. The NSG and pustulan hydrolysate are desalted using a short Sephadex G10 column (1.6 × 30  cm) eluted with deionized water at a flow rate of 20 ml/h, monitored on-line by refractive index. The oligosaccharides eluted in the void volume are collected and lyophilized. 3. Oligosaccharide fractions of NSG, pustulan and curdlan are obtained by gel filtration chromatography on a column (1.6 × 90 cm) of Bio-Gel P4 or Bio-Gel P6. 4. The Bio-Gel P4 or P6 columns are first equilibrated with deionized water and calibrated with dextran hydrolysate (see Note 16) by elution with deionized water at a flow rate of 15 ml/h. 5. The desalted oligosaccharide mixtures are reconstituted in deionized water. At this point, an estimation of the concentration (hexose/ml solution) can be made by TLC-dot orcinol assay (Subheading 3.2.2 below). 6. Typically 1–2 ml of clear solution (after high speed centrifugation, e.g. 10,000 × g for 5  min), the concentration (mg hexose/ml) varies according to solubility, are applied to the column and eluted under the same conditions as in step 5. The eluate is monitored on-line by refractive index and

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

345

Fig. 2. Gel filtration chromatography of polysaccharide hydrolysates. (a) Bio-Gel P4 profile of dextran hydrolysate, 4  mg in 1  ml water was applied with dextran polysaccharide, 0.1  mg, to mark the exclusion volume (Vo); (b) Bio-Gel P4 profile of NSG hydrolysate, 20 mg in 2 ml deionized water; (c) Bio-Gel P6 profile of curdlan hydrolysate 20 mg in 2 ml of water; (d) Bio-Gel P4 profile of pustulan hydrolysate (50 mg in 1 ml of water). The glucose units as determined by reference to the elution profile of dextran hydrolysate, or by MALDI-MS, are indicated for each peak. The asterisk indicates RI values that are off-scale.

c­ ollected as 1.5 ml fractions. Examples of gel filtration chromatograms of the polysaccharide hydrolysates are shown in Fig. 2 (see Note 17). 7. The fractions are pooled according to their predominant glucose units, 7 to 13mers, as determined by MALDI-MS (Subheading 3.3), and lyophilized (see Note 17). As representatives, MALDI-MS spectra of curdlan, NSG, and pustulan fractions containing the 11mer are shown in Fig. 3. 8. Quantitation of the oligosaccharide fractions, after their reconstitution with deionized water, is carried out by TLC-dot orcinol assay (Subheading 3.2.2 below).

346

A.S. Palma et al. Glc11 1823

a 100

NSG-F11 Glc10 1661

%

0 1000

1100 1200 1300

1400 1500 1600 1700 1800 1900 2000

Curdlan- F11 Glc10 1661

%

2100 2200 2300

Glc12 1985

0 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300

Pustulan- F11

%

m/z

Glc11 1823

c

100

m/z

Glc11 1823

b 100

Glc12 1985

Glc10 1661

Glc12 1985

0 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300

m/z

Fig. 3. MALDI-MS analysis of NSG, curdlan and pustulan hydrolysate fractions containing 11mers. The molecular masses of the sodiated molecules indicated that the major component in each fraction was the 11mer with 10mer and 12mer as minor components.

3.2.2. Quantitation of Glucan Oligosaccharides by Dot Orcinol Assay on TLC Plates

1. A 1  ml aliquot of each oligosaccharide solution and standard glucose solutions (at 0.05, 0.1, 0.2, 0.5, and 1  mg/ml) are spotted on a TLC plate using a Hamilton syringe. 2. The plate is air-dried and sprayed with the orcinol reagent until it appears slightly wet (see Note 5). The plate is heated, e.g., in a vented oven, at 105°C for about 5 min or until violet color develops (see Note 5). 3. The glucan oligosaccharides are quantified on TLC plates using a densitometer in the linear reflectance absorption mode at 550  nm. Test samples are measured against the glucose ­standards (see Note 18).

3.2.3. HPTLC Analysis of Glucan Oligosaccharides

1. The solvent n-propanol/water, 8:3, is added to the TLC tank and vapor is allowed to equilibrate (1 h, at ambient temperature). A TLC plate is cut to the desired size (e.g., 10 cm long and 5–10 cm wide) from the back using a scalpel blade. 2. Each oligosaccharide solution is applied (ideally 2 mg hexose) to the TLC plate as a 4 mm band using a TLC applicator at

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

347

Fig. 4. Thin layer chromatography of the glucan oligosaccharide fractions F7–F13 from NSG, curdlan, and pustulan and macroarray analyses of interactions of Fc-Dectin-1 with NGL probes derived from them. (a) HPTLC analysis of the oligosaccharide fractions in the solvent system: n-propanol/water, 8:3 by volume, developed twice. The arrow indicates positions of sample application. Glucose was revealed by staining with orcinol reagent. (b) Macroarrays, on nitrocellulose membranes, of fluorescent ADHP-NGL probes (50 pmol/spot), revealed under UV light and examined for Fc-Dectin-1 binding. This research was originally published in Journal of Biological Chemistry. Reproduced with permission from the American Society for Biochemistry and Molecular Biology (7).

15 mm from the bottom edge and 15 mm free at both edges. The plate is air dried. 3. The plate is carefully placed into the tank. The TLC is developed to 5 mm below the top edge (around 1.5 h). The plate is removed from the tank and dried with warm air. 4. The plate is then stained with orcinol reagent for the visualization of hexose. The staining is carried out by spraying with orcinol reagent until the plate appears slightly wet (see Note 5). Heat the plate in a vented oven at 105° for about 5 min or until the violet color given by hexose is developed (see Note 5). An example of the results obtained is shown in Fig. 4a (see Note 19). 3.3. Determination of Oligosaccharide Chain Length by MALDI-MS

1. Solutions of oligosaccharide fractions obtained from gel filtration chromatography are diluted in water, to an estimated concentration of 10–20 pmol/ml. 2. Sample solution (0.5 ml) is deposited on the MALDI sample target (see Note 20).

348

A.S. Palma et al.

3. Matrix solutions for MALDI-MS are freshly made by mixing Solutions A and B in a ratio of 4:1 (by volume) before sample application (0.5 ml) to the MALDI target. 4. After air-drying, 0.25  ml ethanol is added for sample re-crystallization. 5. Mass spectral data acquisition is carried out on a MALDI mass spectrometer in the positive-ion mode. 6. The chain lengths are deduced from the sodiated molecular ions (MNa+) of the main components. An example of the results obtained is shown in Fig. 3. 3.4. Conversion of the Glucan Oligosaccharides into NGLs

The procedures for the preparation of the NGL probes from freereducing oligosaccharides, and their purification and quantitation are described in Chapter 8 of this series by Liu et al. (Subheadings 3.1–3.3). Below are the stepwise procedures for preparing the fluorescent glucan oligosaccharide (ADHP-derived) NGLs used in the original Dectin-1 study (7). DHPE can also be used in the same procedure to yield non-fluorescent glucan oligosaccharide NGLs (see Note 21). 1. To the lyophilized oligosaccharide fraction (typically 100 nmol) are added 5 ml of water, 100 ml of ADHP solution (10 nmol/ml) in chloroform/methanol (1:3, by volume), and 20  ml of a freshly made solution of the reducing agent tetrabutylammonium cyanoborohydride (20 mg/ml in methanol). Seal tightly the reaction vial. 2. The mixture is incubated at 75°C for 72 h in a heating block. 3. For oligosaccharides higher than 9-mers, additional reagents (2.5 ml of water, 50 ml of ADHP solution, and 10 ml of reducing agent) are added after 24 h of incubation. 4. The NGL products are analyzed by HPTLC as described in Subheading 3.1 of Chapter 8 of this book series. ADHPderived NGLs can be visualized directly under longwave UV light without chemical staining. 5. The NGL products are purified and quantified as described in Subheadings 3.2 and 3.3 of Chapter 8 of this book series. Molecular masses of the NGLs are determined by MALDI-MS.

3.5. Binding Assays of Dectin-1 to Macroarrays of Glucan-Derived NGL-Probes

1. The NGL probes (1 ml of stock solution at 50 pmol/ml in chloroform/methanol/water, 25:25:8, by volume) are arrayed by a nitrogen-assisted TLC applicator as 2-mm bands onto nitrocellulose membranes (see Note 22). Two membranes are prepared: one to overlay with Fc-Dectin-1 and the other for the negative control, i.e., the detection antibody in the absence of Fc-Dectin-1.

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

349

2. The membranes are pre-wetted from underneath by first lying onto and then immersing into TBS, and placed onto Parafilm in a plastic box. 3. The casein Pierce blocking solution is added to cover all the surface area of the membranes, and incubation is for 1  h at ambient temperature. Cover the box with lid to prevent drying of the solution. 4. The membranes are briefly washed by immersing into TBS (a few seconds). 5. Fc-Dectin-1 solution is made in casein Pierce blocker solution at 1 mg/ml, and overlaid onto the membrane; for the antibody control the blocking buffer is added instead. Incubation is for 2 h at ambient temperature. For optimization, a range of concentrations may be tested. 6. The membranes are washed four times by immersing into TBS (2–5 min each wash). 7. The biotinylated goat anti-human-IgG at 10 mg/ml, in casein Pierce blocker solution, is added to each membrane and incubation is for 1 h at ambient temperature. 8. The membranes are washed with TBS (described in step 6 above). 9. The streptavidin conjugated to horseradish peroxidase, at 5  mg/ml in casein Pierce blocker solution, is added to the membranes and incubation is for 30 min. 10. The membranes are washed with TBS (described in step 6 above). 11. The FAST-DAB solution is added to the membranes and color is developed for 5–10 min. The reaction is stopped by rinsing the membrane with deionized water. An example of the results obtained is shown in Fig. 4b (see Note 23). 3.6. Linkage Assignment by Methylation Analysis

The monosaccharide linkage position analysis of the oligosaccharide fractions is based on GC-MS methylation analysis. This involves permethylation of all free hydroxyl groups of the oligosaccharides followed by the hydrolysis of the glycosidic linkages, reduction, and acetylation to give partially methylated alditol acetates (PMAAs), and finally GC-MS analysis of the PMAAs (19). 1. Permethylation is carried out using the NaOH method (20, 21). Oligosaccharides (approximately 10–20 nmol) are freezedried and dissolved in 75 ml DMSO with sonication for 15 min. To the solution is added, a suspension of NaOH (~80 mg/ml) in 75  ml DMSO, prepared by finely powdered NaOH (see Note 24) produced by crushing two pellets in a small beaker using a thick glass rod, followed by the addition of 50  ml iodomethane.

350

A.S. Palma et al.

2. The reaction mixture is incubated under N2 at 22°C for 15 min. This is extracted into chloroform (1 ml) and the chloroform layer washed six times with 4 ml H2O. 3. For hydrolysis, the permethylated oligosaccharides are incubated with 100 ml 2 M TFA for 1 h at 120°C. The mixture is cooled, evaporated under N2 and residual TFA removed by co-evaporation with methanol (3 × 100 ml aliquots) under N2. 4. The hydrolyzed products are reduced in 100  ml solution of sodium borodeuteride (0.1 M)/NaOH (0.01  M) at 4°C for 16 h. A drop of acetic acid is then added and the mixture evaporated under N2, followed by three co-evaporations with 200 ml methanol. The residue is dissolved in 150 ml pyridine/ acetic anhydride (1:1, by volume) under N2 with sonication and incubated for 1.5 h at 100°C. The reaction is evaporated to dryness under N2, extracted with chloroform (1  ml), and washed six times with 4 ml H2O. 5. The PMAAs thus obtained are analyzed by capillary GC-MS using a 15  m RTX-5 capillary column (Hewlett-Packard). The initial column temperature is 50°C programmed to 100°C at 25°C/min, to 220°C at 5°C/min, and to 310°C at 10°C/min. 6. The retention times on the GC and the fragmentation in MS of the PMAAs are used to deduce the monosaccharide types and linkage information of the oligosaccharide fractions. An example of the results obtained is shown in Table 1.

Table 1 Linkage determination by GC-MS methylation analysis Polysaccharides

Oligosaccharide fractions F11

NSG F11 subfractions

Residues

NSG

Curdlan

Pustulan

NSG

Curdlan

Pustulan

a

b

c

−3Glc1-

62

88

–

83

96

–

94

 7

89

−4Glc1-

19

 3

–

10

 2

–

–

76

–

−6Glc1-

 5

 3

98

 2

–

95

–

 8

 2

−3,6Glc1-

10

 5

 1

 4

–

–

 6

 2

 8

−4,6Glc1-

 4

–

–

–

 2

 4

–

 7

–

Percentage shown is based on total internally linked residues and “–” indicates less than 1. This research was originally published in Journal of Biological Chemistry. Palma AS, Feizi T, Zhang Y, Stoll MS, Lawson AM, Díaz-Rodríguez E, Campanero-Rhodes MA, Costa J, Gordon S, Brown GD, Chai W. Ligands for the beta-glucan receptor, Dectin-1, assigned using “designer” microarrays of oligosaccharide probes (neoglycolipids) generated from glucan polysaccharides. J Biol Chem. 2006; 281: 5771–5779. © the American Society for Biochemistry and Molecular Biology

23 

3.7. Carbohydrate Microarray Analysis of Dectin-1

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

351

The procedures for the generating NGL-based microarrays, the arraying and scanning hardware, and general protocols for the microarray analyses of carbohydrate-binding proteins are described in Chapter 8 of this series (see Subheadings 3.4 and 3.5). Below are the procedures for detecting binding by Dectin-1 to the microarrays. The results presented in Fig.  5 are from a recent analysis (A.S. Palma, M.A. Campanero-Rhodes, Y. Liu, R.A. Childs and colleagues, unpublished) using a microarray of 327 oligosaccharide probes, among them are glucan oligosaccharide sequences with b1,3- (5–13mers), b1,6- (7mer, 10mer, 11mer, and 13mer), b1,4- (5mer), a1,2- (2mer), a1,4(2–5mers), and a1,6-linkages (2–7mers), and in addition, more than 270 of mammalian-type (see Note 25). Binding is highly restricted, and detected only to the 11mer and 13mer of b1,3linked glucose, and not to the 7mer nor to any of the other probes in the microarray.

Fig. 5. Microarray analysis of murine Dectin-1. The Dectin-1 was analyzed as an IgG-Fc chimera at 5 mg/ml. Oligosaccharide probes (lipid-linked) were printed in duplicate on nitrocellulose-coated glass slides. The results are the means of fluorescence intensities with subtracted background of duplicate spots, printed at 7 fmol, with the range indicated by error bars. Binding was detected exclusively to b1,3-linked glucose sequences (11 and 13mers), in accordance with earlier data (7). Reproduced with permission from De Gruyter Publishing (11).

352

A.S. Palma et al.

1. The printed microarray slides are scanned for Cy3 fluorescence (e.g., 10  mm resolution, 100% laser power, 35% PMT gain) using a fluorescence microarray scanner before the binding experiment. 2. Incubation chamber (16-pad silicon gasket) is placed over the slide to be probed and the assembly inserted into the multislide plate. 3. The slide pads are prewetted with HBS buffer and blocked for 1 h with blocking solution (140 ml) and washed by pipetting HBS buffer (140 ml). The removal of excess solution can be done by flicking and briefly blotting the inverted frame. 4. The microarray analysis with Fc-Dectin-1 is performed with the protein diluted to 5 mg/ml in blocking solution (see Notes 7 and 26). The protein solution (100 ml) is added onto each pad and incubated for 1.5 h at ambient temperature. A negative control, i.e., the detection antibody in the absence of Fc-Dectin-1, is also included. For this the blocking solution is added at this step. The overlay materials are removed by flicking or by careful introduction of pipette tips into the well to aspirate solution. 5. The pads are then washed by pipetting 140 ml HBS buffer into each pad; three wash steps over 5 min should be performed. The removal of excess solution can be done by flicking and briefly blotting the inverted frame or by aspiration. 6. The biotinylated goat anti-human-IgG at 10 mg/ml in blocker solution is added onto each pad and incubation is for 1 h at ambient temperature. The removal of excess solution and washing are as in step 5. 7. Binding is detected using Alexa Fluor-647-labeled streptavidin at 1 mg/ml in blocking solution. Incubation is for 40 min in the dark. The removal of excess solution and washing are as in step 5. 8. The slide is removed from the multi-slide plate and from the 16-pad incubation chamber. An additional rinse with water (e.g., by spraying) is performed. 9. The slides are dried and kept in the dark before scanning. 10. The slides are scanned for Alexa Fluor-647 using a fluorescence microarray slide scanner (ProScanArray, Perkin-Elmer), e.g., at 10 mm resolution; 100% laser power or other convenient laser power, 35% PMT gain. 11. The Alexa Fluor-647 fluorescence intensity of each array spot is quantified using the scanner software (Perkin Elmer ScanArray Express software). The Cy3 image obtained before overlay is used for spot location. The results are saved as an Excel data sheet for further processing (see Note 27).

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

353

4. Notes 1. The recombinant Dectin-1 we describe here is a soluble chimeric protein containing the extracellular carbohydrate-recognition domain of the murine Dectin-1, fused to the Fc portion of human IgG (22). There are Dectin-1 constructs commercially available that can also be used, e.g., the histidine-taggedDectin-1 from R&D (human and murine Dectin-1 sequence) systems and from Sino Biological (murine sequence). 2. Glucan polysaccharides can be highly heterogeneous. Our selection was based on commercial availability, solubility in aqueous solution and reported content of predominant b- or a-linked sequences. 3. Curdlan polysaccharide, a linear b1,3-linked glucan, is not a soluble polysaccharide and therefore not tested for Dectin-1 binding. Only the acid hydrolysate was investigated. 4. To achieve a good separation of glucan oligosaccharides other solvent systems can also be tried, e.g. n-propanol:ethanol:water (7:1:2 by volume) or acetonitrile: water (4:1 by volume). 5. For the preparation of the orcinol reagent, dissolve 900 mg of orcinol in 25 ml of water. Add 375 ml of ethanol. Cool on ice. Gradually add 50 ml of concentrated sulfuric acid (18 M) with stirring, maintaining the temperature below 10°C. Great care should be taken in handling concentrated sulfuric acid. For spraying always use a ventilation TLC spray hood (available from Camag) or an acid-resistant fume hood and gloves. For color development of the HPTLC plate, a TLC heating plate (Camag) set up at 105°C in a fume hood can be used instead of the oven. The orcinol reagent is stable for at least 1 year when stored at 4°C in the dark. 6. ADHP was used in the original Dectin-1 studies (7). ADHP can be prepared from DHPE by incorporation of the fluorescent label anthracene, and the detailed procedures involving multiple steps of chemical modification have been described elsewhere (23, 24). These can be followed by readers with essential skills and experience in chemical synthesis. For the procedure described in Subheading  3.4, both DHPE and ADHP can be used. 7. The blocking solution to use in the binding experiment of a given protein always requires optimization. For Fc-Dectin-1 analysis, we found that casein from Pierce diluted to 0.1% in TBS gives a good signal-to-noise ratio, but this could vary also with the lot of casein that is used. For other commonly used blocker solutions, please see Note 17 in Chapter 8 of this series. For example, for the His-tagged Dectin-1 analysis, the blocker

354

A.S. Palma et al.

3% (w/v) bovine serum albumin (BSA) (from Sigma) in HBS gives a better signal-to-noise ratio than the casein-based blocker (see Note 26). 8. The Alexa Fluor-647-labeled streptavidin is stored at −20°C in small working aliquots of 20 ml and protected from light. 9. The concentrations of the polysaccharides for immobilization will depend on the strength of the binding signal detected and on the solubility of the individual polysaccharide. Higher concentrations could be tested if feasible and if necessary. 10. Neutral polysaccharides do not require any treatment of the plastic microwell plate for efficient immobilization, e.g., treatment with poly-l-lysine, which is required for efficient immobilization of acidic polysaccharides (25). However, the degree of immobilization will vary for the different polysaccharides and is dependent on their molecular sizes. Therefore, inhibition of binding assays is performed to exclude any effects of different degrees of immobilization. 11. Variables to investigate in the initial experiments are (a) binding signals using Fc-Dectin-1 noncomplexed or precomplexed with the biotinylated anti-IgG (Dectin-1:anti-IgG ratio 1:3 by weight), 1  h before overlay (7) and (b) calcium dependence using TBS in the presence of 2 mM CaCl2 or TBS in the presence of 10 mM EDTA for the blocking, incubation, and washing steps. We have found that precomplexation of the Fc-Dectin-1 with the detection antibodies did not enhance the binding signals and that the Dectin-1 binding was calciumindependent (7). 12. The concentration point chosen as non-saturating is the point half-way up on the binding curve (0.5 mg/ml in Fig. 1c). In the assay, Dectin-1 is used at double the concentration selected as non-saturating, taking into account the twofold dilution with the inhibitor. 13. The results are expressed as percentage of inhibition of binding as follows: percentage inhibition = ((OD no inhibitor − OD with inhibitor)/(OD no inhibitor − OD negative control)) × 100. 14. For the Dectin-1-binding studies, we focused on glucan oligosaccharide fractions containing 7mer to 13mer. The selection was based on our initial experiments, where b1,3-linked 7mer gave no inhibition of Fc-Dectin-1 binding to NSG, and the NGL of the b1,3-linked 6mer was not bound by Fc-Dectin-1. The requirement for long oligosaccharide chain lengths by Dectin-1 was corroborated by the inhibition of binding experiments using oligosaccharides in solution. The mechanism by which Dectin-1 recognizes glucan sequences is not yet elucidated, but the sizes of the oligosaccharides bound by Dectin-1 are unusually long for a lectin. The phenomenon is likely to

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

355

relate to a conformational motif displayed only by the long b1,3 oligosaccharides, as these oligosaccharides have the particularity to adopt an helical conformation. For a new investigation where the chain length requirement of a carbohydrate-binding protein is not known, we examine 2–13mers. The oligosaccharides used for generation of NGL probes were obtained from partial depolymerization of NSG (ligand-positive glucan with a b1,3-linked backbone and b1,6-branches) and pustulan (ligand-negative glucan with a b1,6linked backbone) and from oligosaccharide fragments of curdlan (glucan with a linear b1,3-linked backbone). 15. In our published study of Dectin-1 ligands, the pustulan oligosaccharides were prepared by acetolysis of the pustulan (7). Acid hydrolysis is selected here as the depolymerization procedure for pustulan upon our observation that the b-linked oligosaccharide series prepared by acetolysis of the parent polysaccharide contained some a-anomers. 16. Dextran hydrolysate can be prepared by acid hydrolysis of the dextran polysaccharide: Dextran (100  mg) is treated with 0.1 M HCl, at a concentration of 20 mg polysaccharide/ml, at 100°C for 4 h. 17. The fractions containing b1,3- and b1,6-glucan oligosaccharides longer than 9mer and 10mer did not resolve well on the gel filtration columns, particularly the b1,3-linked (Fig. 2). We have overcome this by collecting fractions of smaller volumes (e.g., of 0.75 ml) and by analyzing the main components of individual fractions by MALDI-MS before selection for pooling. To further improve the resolution and reduce heterogeneity, other methods of separation are performed as necessary. These can include repeated gel filtration chromatography of selected fractions and preparative HPTLC (7). Alternatively, oligosaccharides can be fractionated by HPLC after conversion into reversible aminopyridine derivatives (7, 26). 18. Stock solutions of the oligosaccharide fractions are prepared typically at 0.5–2 mg hexose/ml dependent on the solubility of glucan oligosaccharides in water, and stored at −20°C. Overall, solubility of fractions up to 8mer is good, but from 9mer onwards the solubility decreases and there is a tendency of the b1,3- and b1,6-linked oligosaccharides to precipitate at −20 or 4°C. We observed that if the fractions are warmed up to 37°C (for the b1,3-linked) or to 100°C (for the b1,6linked), a clear solution is obtained and maintained at ambient temperature. 19. To improve the resolution of the glucan oligosaccharides, the HPTLC plate can be developed twice. After the first development, the plate should be air-dried thoroughly before placing back

356

A.S. Palma et al.

into the TLC tank for the second development. Care should be taken when opening the cover of the tank and placing the plate, trying not to disturb the equilibrated solvent vapor. For NSG oligosaccharide fractions, we observed three subfractions with fast, intermediate, and slow migration, designated subfractions a, b, and c, respectively (Fig. 4a). Dectin-1 ligandpositive fractions were further separated into a, b, and c subfractions by preparative HPTLC (7). These were converted to NGLs and analyzed for Dectin-1 binding. Dectin-1 activity was observed on subfractions a and c, containing 1-3-linked glucose. Due to its relative homogeneity, we selected subfraction a for further studies (7). 20. Steps 2 and 3 (application of sample and matrix solution) can be reversed. 21. To improve the conjugation yields for NGLs of glucan oligosaccharides larger than heptasaccharide, the procedure involves modifications compared with that described in Subheading 3.1 of Chapter 8 of this series, e.g., in solvent composition, incubation temperature, and incubation time. The conjugation yields for oligosaccharides larger than decasaccharide are lower than for the shorter oligosaccharides and typically a yield of 10–30% is expected. We observed in our recent studies (27 and unpublished data) that higher conjugation yields can be obtained using the oxime-ligation method. Detailed procedures for preparing the aminooxy-functionalized lipid reagent (AOPE) and oxime-linked NGLs have been described elsewhere (27, 28) and are not within the scope of this chapter. 22. The spotting on the nitrocellulose has to be done with extreme care and at a slow speed, not to spread the sample too much and not to damage the nitrocellulose membrane with the organic solvent. Spotting more than 2 ml is not advisable. 23. The results in the left panel of Fig.  4b show the intensities, revealed by UV light, of fluorescence NGL probes spotted on nitrocellulose membranes before overlay. If using the non-fluorescent glucan oligosaccharide NGLs (derived from DHPE or AOPE), these cannot be directly visualized on the nitrocellulose membranes. For visualization, these NGLs need to be arrayed on a silica gel TLC plate and sprayed with primulin reagent for lipid staining (see Subheadings 3.1, step 6 of Chapter 8 of this series). 24. The operation has to be completed in a short period of time to avoid adsorption of moisture. 25. The microarray used for studies of Dectin-1 included a diverse range of mammalian type sequences, all lipid-linked: N-glycans

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

357

of high-mannose and of neutral and sialylated complex-type; O-glycans, blood group-related sequences (A, B, H, Lewisa, Lewisb, Lewisx, and Lewisy) on linear or branched backbones and their sialylated and/or sulfated analogs, gangliosides, and oligosaccharide fragments of glycosaminoglycans and polysialic acid. Also included were homo-oligomers of other monosaccharides (mannose, N-acetylglucosamine, xylose, and arabinose). 26. The microarray analysis can be performed with the commercially available his-tagged Dectin-1 (see Note 1). For the dimeric Fc-Dectin, precomplexation with the detection antibody did not enhance the binding signals, but for the Histagged monomeric constructs the precomplexation with the detection antibodies is required for detectable binding. The Dectin-1-His-antibody complexes can be prepared as described (29). Dectin-1-His is precomplexed with mouse monoclonal anti-polyhistidine (Ab1) and biotinylated anti-mouse IgG (Ab2) in a ratio of 1:3:3 (by weight). The Dectin-1–antibody complexes are prepared by preincubating Ab1 and Ab2 for 15  min at ambient temperature, followed by the addition of Dectin-1-His and incubation for a further 15 min. The Dectin1–antibody complexes are diluted in the blocking solution (HBS containing 3% (w/v) BSA (Sigma)), to give a final Dectin-1-His concentration of 5 mg/ml. 27. A dedicated software suite, designed in our laboratory, is used for storing, retrieving, and displaying carbohydrate microarray data. Further information of the latter has been described elsewhere (30).

Acknowledgments We gratefully acknowledge collaborations of our colleagues, past and present, in the Glycosciences Laboratory: Mark Stoll, Alex Lawson, and Colin Herbert; and we acknowledge Gordon Brown for the murine Dectin-1-Fc chimera. The Glycosciences Laboratory acknowledges with gratitude collaborators over the years with whom our microarray probes were studied. For grant support, we acknowledge the UK Medical Research Council, the UK Research Councils Basic Technology Grant (GR/S79268, “Glycoarrays”), Engineering and Physical Research Councils Translational Grant EP/G037604/1, and the NCI Alliance of Glycobiologists for Detection of Cancer and Cancer Risk (U01 CA128416). A.S.P. is a fellow of the Fundação para a Ciência e Tecnologia (SFRH/ BPD/26515/2006, Portugal).

358

A.S. Palma et al.

References 1. Ariizumi, K., Shen, G. L., Shikano, S., Xu, S., Ritter, R., III, Kumamoto, T., Edelbaum, D., Morita, A., Bergstresser, P. R., and Takashima, A. (2000) Identification of a novel, dendritic cell-associated molecule, dectin-1, by subtractive cDNA cloning, J. Biol. Chem. 275, 20157–20167. 2. Brown, G. D. and Gordon, S. (2001) Immune recognition. A new receptor for beta-glucans, Nature 413, 36–37. 3. Brown, G. D., Taylor, P. R., Reid, D. M., Willment, J. A., Williams, D. L., MartinezPomares, L., Wong, S. Y., and Gordon, S. (2002) Dectin-1 is a major beta-glucan receptor on macrophages, J. Exp. Med. 196, 407–412. 4. Brown, G. D., Herre, J., Williams, D. L., Willment, J. A., Marshall, A. S., and Gordon, S. (2003) Dectin-1 mediates the biological effects of beta-glucans, J. Exp. Med. 197, 1119–1124. 5. Weis, W. I., Taylor, M. E., and Drickamer, K. (1998) The C-type lectin superfamily in the immune system, Immunol. Rev. 163, 19–34. 6. Willment, J. A., Marshall, A. S., Reid, D. M., Williams, D. L., Wong, S. Y., Gordon, S., and Brown, G. D. (2005) The human beta-glucan receptor is widely expressed and functionally equivalent to murine Dectin-1 on primary cells, Eur. J. Immunol. 35, 1539–1547. 7. Palma, A. S., Feizi, T., Zhang, Y., Stoll, M. S., Lawson, A. M., Diaz-Rodríguez, E., Campanero-Rhodes, A. S., Costa, J., Brown, G. D., and Chai, W. (2006) Ligands for the beta-glucan receptor, Dectin-1, assigned using ‘designer’ microarrays of oligosaccharide probes (neoglycolipids) generated from glucan polysaccharides, J Biol. Chem. 281, 5771–5779. 8. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions, Nat. Biotechnol. 20, 1011–1017. 9. Feizi, T. and Chai, W. (2004) Oligosaccharide microarrays to decipher the glyco code, Nat. Rev. Mol. Cell Biol. 5, 582–588. 10. Schallus, T., Jaeckh, C., Feher, K., Palma, A. S., Liu, Y., Simpson, J. C., Mackeen, M., Stier, G., Gibson, T. J., Feizi, T., Pieler, T., and MuhleGoll, C. (2008) Malectin - A Novel Carbohydrate-binding Protein of the Endoplasmic Reticulum and a Candidate Player in the Early Steps of Protein N-glycosylation, Mol. Biol. Cell 19, 3404–3414.

11. Liu, Y., Palma, A. S., and Feizi, T. (2009) Carbohydrate microarrays: key developments in glycobiology, Biol. Chem. 390, 647–656. 12. Jamas, S., Easson, D. D. J., Ostroff, G. R., and Onderdonk, A. B. (1991) Glucans a novel class of macrophage activating immunomodulators, ACS Symp Ser 469, 44–51. 13. Black, W. A. P., Cornhill, W. J., Dewar, E. J., and Woodward, F. N. (1951) Manufacture of algal chemicals. III. Laboratory-scale isolation of laminarin from brown marine algae, J Appl. Chem. 501–517. 14. de la, C. J., Pintor-Toro, J. A., Benitez, T., and Llobell, A. (1995) Purification and characterization of an endo-beta-1,6-glucanase from Trichoderma harzianum that is related to its mycoparasitism, J. Bacteriol. 177, 1864–1871. 15. Chanda, N. B., Hirst, E. L., and Manners, D. J. (1957) A comparison of isoLichenan and Lichenan from Icelandic Moss (Cetrarai islandica), J. Chem Soc. 1951–1958. 16. McCleary, B. V. and Matheson, N. K. (1987) Enzymic analysis of polysaccharide structure, Advances in Carbohdydrate Chemistry and Biochemistry 44, 147–276. 17. de Belder, A. N. (1993) Dextran, in Industrial Gums (Whistler, R. L. and BeMiller, J., Eds.) 2nd ed., pp 399–426, Academic Press, New York. 18. Haworth, W. N., Hirst, E. L., and isherwood, F. A. (1937) Polysaccharides. Part XXIV. Yeast mannan, J Chem Soc 748–790. 19. Chai, W., Hounsell, E. F., Cashmore, G. C., Rosankiewicz, J. R., Bauer, C. J., Feeney, J., Feizi, T., and Lawson, A. M. (1992) Neutral oligosaccharides of bovine submaxillary mucin. A combined mass spectrometry and 1H-NMR study, Eur. J Biochem. 203, 257–268. 20. Ciucanu, I. and Kerek, F. (1984) A simple and rapid method for the permethylation of carbohydrates., Carbohydr. Res. 131, 209–217. 21. Hounsell, E. F., Lawson, A. M., Stoll, M. S., Kane, D. P., Cashmore, G. C., Carruthers, R. A., Feeney, J., and Feizi, T. (1989) Characterisation by mass spectrometry and 500-MHz proton nuclear magnetic resonance spectroscopy of penta- and hexasaccharide chains of human foetal gastrointestinal mucins (meconium glycoproteins), Eur. J. Biochem. 186, 597–610. 22. Graham, L. M., Tsoni, S. V., Willment, J. A., Williams, D. L., Taylor, P. R., Gordon, S., Dennehy, K., and Brown, G. D. (2006) Soluble Dectin-1 as a tool to detect beta-glucans, J. Immunol. Methods 314, 164–169.

23 

Neoglycolipid-Based “Designer” Oligosaccharide Microarrays…

23. Stoll, M. S., Feizi, T., Loveless, R. W., Chai, W., Lawson, A. M., and Yuen, C. T. (2000) Fluorescent neoglycolipids. Improved probes for oligosaccharide ligand discovery, Eur. J. Biochem. 267, 1795–1804. 24. Chai, W., Stoll, M. S., Galustian, C., Lawson, A. M., and Feizi, T. (2003) Neoglycolipid technology - deciphering information content of glycome, Methods Enzymol. 362, 160–195. 25. Leteux, C., Chai, W., Nagai, K., Herbert, C. G., Lawson, A. M., and and Feizi, T. (2001) 10E4 Antigen of scrapie lesions contains an unusual nonsulfated heparan motif., J. Biol. Chem. 276, 12539–12545. 26. Her, G. R., Santikarn, S., Reinhold, V. N., and Williams, J. C. (1987) Simplified approach to HPLC precolumn fluorescent labeling of carbohydrates:N-(2-pyridinyl)-glycosaylamines, J Carbohydr Chem 6, 129–139. 27. Liu, Y., Feizi, T., Campanero-Rhodes, M. A., Childs, R. A., Zhang, Y., Mulloy, B., Evans, P. G., Osborn, H. M., Otto, D., Crocker, P. R., and Chai, W. (2007) Neoglycolipid probes

359

­ repared via oxime ligation for microarray p ­analysis of oligosaccharide-protein interactions, Chem. Biol. 14, 847–859. 28. Liu, Y., Chai, W., Childs, R. A., and Feizi, T. (2006) Preparation of Neoglycolipids with Ring-Closed Cores via Chemoselective OximeLigation for Microarray Analysis of Carbohydrate-Protein Interactions, Methods Enzymol. 415C, 326–340. 29. Palma, A. S., Liu, Y., Muhle-Goll, C., Butters, T. D., Zhang, Y., Childs, R., Chai, W., and Feizi, T. (2010) Multifaceted approaches including neoglycolipid oligosaccharide microarrays to ligand discovery for malectin, Methods Enzymol. 478, 265–286. 30. Stoll, M. S. and Feizi, T. (2009) Software tools for storing, processing and displaying carbohydrate microarray data. Proceeding of the Beilstein Symposium on Glyco-Bioinformatics, 4–8 October, 2009, Potsdam, Germany, (Kettner, C., Ed.) pp 123–140, Beilstein Institute for the Advancement of Chemical Sciences, Frankfurt, Germany.

Chapter 24 Measurement of Antibodies to Pneumococcal Polysaccharides with Luminex xMAP Microsphere-Based Liquid Arrays Jerry W. Pickering and Harry R. Hill Abstract The 23 valent pneumococcal polysaccharide vaccine (PPV) is often used to assess an individual’s ability to produce antibodies to polysaccharides. The Luminex xMAP microsphere-based liquid array system, when applied to the determination of antibodies to pneumococcal polysaccharides (PnPs), allows for the antibody response to the 23 serotypes to be determined simultaneously. Multiplexing saves considerable time and expense over the traditional method of testing each PnPs serotype individually by the enzyme-linked immunosorbent assay (ELISA). This chapter describes methods for (a) conjugation of poly-l-lysine (PLL) to PnPs, (b) coupling of PnPs–PLL conjugates to Luminex microspheres, and (c) a multiplex Luminex assay for the measurement of serotype-specific IgG concentrations to pneumococcal serotypes. Key words: Pneumococcal, Polysaccharide, Luminex, Antibody, Immune deficiency, Multiplex

1. Introduction Type-specific IgG antibodies to the capsular polysaccharides of Streptococcus pneumoniae protect against invasive diseases by opsonizing the organism (1, 2). Type-specific anticapsular polysaccharide antibodies also protect against infection by preventing the acquisition and carriage of the pneumococci (3, 4). People with one of a variety of immunodeficiency disorders do not produce antibodies to polysaccharide antigens and, therefore, may experience chronic or recurring respiratory infections caused by S. pneumoniae and other encapsulated bacteria (5, 6). In addition, some people may not produce antibodies to polysaccharides even

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_24, © Springer Science+Business Media, LLC 2012

361

362

J.W. Pickering and H.R. Hill

though their serum immunoglobulin levels are normal and may have what is commonly referred to as specific polysaccharide antibody deficiency syndrome (7). The antibody response to vaccination with 23-valent PnPs vaccine is used to access a patient’s ability to produce antipolysaccharide antibodies. The 23-valent pneumococcal polysaccharide vaccine (PPV) (Pneumovax 23, Merck, West Point, NY and Pnu-Immune 23 Wyeth, Philadelphia, PA) contains serotypes 1, 2, 3, 4, 5, 6B, 7F, 8, 9N, 9V, 10A,11A, 12F, 14, 15B, 17F, 18C, 19A, 19F, 20, 22F, 23F, and 33F. A 7-valent-conjugated pneumococcal vaccine, Prevenar\Prevnar™ (Wyeth), was introduced in 2000 for infants and toddlers younger than 2 years and contains serotypes 4, 6B, 9V, 14, 18C, 19F, and 23F. Recently, Prevnar 13™ has received approval from the Food and Drug administration. Prevnar 13™ contains serotypes 1, 3, 5, 6A, 7F, and 19A in addition to those in Prevnar™. The antibody response to the 23 serotypes in the polysaccharide vaccine varies considerably among healthy individuals and an individual may respond differently or not at all to the different serotypes (8, 9). Although a number of criteria have been proposed, there is no established consensus for the interpretation of antibody responses to the polysaccharide antigens. Paris and Sorensen proposed that the antibody response to 12–14 different serotypes should be tested and that seven of the serotypes should be different from those in the conjugate vaccine (9). Children 2–5 years of age should respond to at least 50% of the serotypes tested and children older than 6 years should respond to 70% of the serotypes tested. An adequate response was defined as a fourfold increase in antibody titer between pre-immunization and post-immunization samples and a post-immunization antibody concentration of 1.3 mg/mL or greater. A standardized enzyme-linked immunosorbent assay (ELISA) for pneumococcal antibody testing was adopted by a World Health Organization (WHO) advisory committee and was used for clinical trials of the Prevnar™ and Prevnar 13™ pneumococcal conjugate vaccines (10, 11). The standardized ELISA uses pneumococcal C-polysaccharide (C-Ps) and PnPs 22F as absorbents to remove common antigens and the international reference standard, 89-SF (12). The ELISA, however, requires a separate assay for each serotype and testing of each patient sample by 12–14 different assays becomes a time consuming, laborious and expensive process. The Luminex xMAP technology is a multiplexing technology that allows all of the pneumococcal serotypes of interest to be tested simultaneously in a single assay. Another advantage of the Luminex system over ELISA is the much wider dynamic range which reduces the need for retesting. Also, liquid phase kinetics of the Luminex system reduces assay time. Unlike the ELISA, however, pneumococcal polysaccharides (PnPs) must be covalently

24  Measurement of Antibodies to Pneumococcal Polysaccharides…

363

attached to the carboxylated Luminex microspheres. Some chemical modification of the polysaccharides is required to covalently attach them to the microspheres. The chemical modifications of the PnPs structure must not destroy the type-specific epitopes. We first described a multiplex assay based on the Luminex xMAP technology for simultaneous determination of IgG concentrations to 14 pneumococcal serotypes (13). For this assay, we used cyanuric chloride to activate PnPs and attach poly-l-lysine (PLL). The PnPs–PLL conjugates were then coupled to the carboxylated Luminex microspheres (see Fig. 1 for reaction scheme). Lal et al. (14) and Elberse et al. (15) have also developed multiplexed Luminex NH2 Cl

PnPs OH

H

N

N

[N

+

+ Cl

H N

C C

Cl

O C

C

]n

O

O

N Cyanuric Chloride

C

PLL

OH

NH2

Cl

PnPs O

Sulfo-NHS EDC

N

N

NH

N

H [N

N

C

C

C

C

O

O

O

H

]n

C

+

O

SO3Na

N O

O NH2 Cl N

N PnPs O

N

NH

H [N

O

H C C

N

C

C

]n

O O HN

C

Fig. 1. Reaction scheme for coupling of pneumococcal polysaccharides (PnPs) to Luminex microspheres. PnPs are first conjugated to PLL with cyanuric chloride. The PnPs–PLL conjugates are then attached to microspheres via carboxyl groups that have been activated with EDC and Sulfo-NHS.

364

J.W. Pickering and H.R. Hill

assays for antibodies to 9 and 13 serotypes, respectively, using the PLL coupling method. In addition, two other groups have developed multiplexed Luminex assays for pneumococcal antibodies using sodium periodate and 4-(4,6-dimethoxy(1,3,5)triazin-2-yl)4-methyl-morpholinium) (DMTMM) to couple the polysaccharides to carboxylated microspheres (16, 17). Although the Luminex xMAP technology is not yet approved for testing of the efficacy of pneumococcal vaccines, it has been found to be useful for identifying individuals with low responses to the PPV (18).

2. Materials 2.1. Preparation of Pneumococcal– Poly-l-Lysine Conjugates

1. PnPs stock solutions: Dissolve polysaccharides (American Type Culture Collection, Manassas, VA) at 5 mg/mL in water (use Nanopure or equivalent purified water for all solutions). Store aliquots at −70°C. (see Notes 1 and 2). 2. PLL solution: 0.1% PLL hydrobromide (Sigma-Aldrich, St. Louis, Mo.). Store at 2–8°C. 3. NaOH/Phenolphthalein: 0.3 M NaOH and 0.001% phenolphthalein. Store at room temperature. 4. Cyanuric chloride. (Sigma-Aldrich). 5. Phosphate buffer: 0.2 M monobasic sodium phosphate. Store at room temperature. 6. Disposable PD-10 desalting columns and buffer reservoirs. (GE Healthcare Bio-Sciences, Piscataway, NJ). 7. Disposable transfer pipets, fine tip (Samco Scientific, San Fernando).

2.2. Coupling of PnPs–PLL Conjugates to Luminex Microspheres

1. MicroPlex® carboxylated microspheres (Luminex Corporation, Austin, TX) (see Note 3). 2. Coupling Buffer. 0.05  M MES (4-morpholineethanesulfonic acid) (Sigma-Aldrich). Adjust the pH to 5.0 ± 0.1 with 5N NaOH or 5N HCl. 3. EDC/Sulfo-NHS solution. Dissolve 20 mg of EDC [(1-Ethyl3-[3-dimethylaminopropyl] carbodiimide hydrochloride), Pierce/Thermo Fisher] and 20 mg of Sulfo-NHS [(N-­hydroxysulfosuccinimide), Pierce/Thermo Fisher] in 0.4  mL of Coupling Buffer just before use. 4. Blocking/storage (B/S) buffer. Add 0.2  mL of Proclin® 300 (Supelco, Bellefonte, PA) and 1 mL of yellow food dye (McCormick) to 1  L of StabliGuard® Immunoassay Stabilizer (BSA free) (SurModics Corporation, Eden Prairie, MN). Store at 2–8°C. 5. 1.5-mL Microcentrifuge tubes (USA Scientific, Ocala, FL).

24  Measurement of Antibodies to Pneumococcal Polysaccharides…

2.3. Multiplex Assay for Pneumococcal Antibodies

365

1. Standard and Control serum. Prepare a reference standard by pooling post-PPV sera from different individuals. Also, prepare serum pools with a range of antibody concentrations for the serotypes to be tested (see Note 4). 2. Pneumococcal Cell Wall Polysaccharide (CWPS) (C-Ps, Teichoic acid), and Pneumococcal CWPS MULTI (Staten Serum Institut, Copenhagen, Denmark). Dissolve at 10 mg/mL in purified water. 3. Absorbent/Diluent. 0.0015% bromocresol purple (SigmaAldrich), 5  mg/mL C-Ps, 5  mg/mL PnPs 22F in 1× Wash Buffer (see Notes 5–7). 4. 10× Wash Buffer. 1.37 M NaCl, 0.023 M KH2PO4, 0.077 M Na2HPO4, 0.031 M NaN3, pH 7.2 ± 0.2. 5. PE conjugate. 7.5 mg/mL R-phycoerythrin-labeled anti-human IgG, g chain specific (Southern Biotech, Birmingham, AL) (see Note 8), 0.15 M NaCl, 0.01 M NaH2PO4, 0.04 M Na2HPO4, 0.008 M NaN3, 0.1% (w/v) bovine serum albumin. 6. 96-Well filtration plates, 1.2-mm MultiScreen-BV (Millipore Corporation, Billerica, MA). 7. MultiScreen Vacuum manifold (Millipore Corporation).

3. Methods 3.1. Preparation of Pneumococcal–Poly-lLysine Conjugates

1. Bring PnPs stock solutions and PLL solution to room temperature. 2. Label 12 × 75-mm disposable glass test tubes with the PnPs serotypes to be conjugated and place in a rack. 3. Add 0.5 mL of 5 mg/mL polysaccharide stock solution of each serotype to be coupled to the appropriate tube. 4. Add 0.5  mL NaOH/phenolphthalein solution to each tube. Vortex briefly to mix. 5. Under a fume hood, add approximately 5 mg of solid cyanuric chloride to each tube (an accurate weigh of cyanuric chloride is not necessary). Vortex the tube vigorously for 10 s. 6. Immediately add 25 ml of PLL solution and vortex again vigorously for 10 s. 7. Complete step 4 in succession for each tube. Cover the tubes with parafilm and store at 2–8°C overnight. 8. Label a PD-10 desalting column (GE Healthcare) for each PnPs serotype conjugated above and prepare the column. (a) Remove the cap from the column. (b) Pour off the packing liquid into a waste container.

366

J.W. Pickering and H.R. Hill

(c) Cut the tip of the column at an angle. (d) Attach a reservoir to the column. (e) Place the column in a suitable column rack. 9. Equilibrate the column with 0.2 M phosphate buffer. (a) Fill the reservoir to the line with 25 mL of phosphate buffer. (b) Allow the 25 mL buffer to drip through the column. 10. Remove the reservoir and, using a thin tip disposable transfer pipette, add the contents of each of the PnPs–PLL tubes above to the top of the appropriate column. Allow the conjugate solution to pass into the column until no eluate drips from the column. Discard the eluate. 11. Add 1 mL of phosphate buffer to the column and allow the buffer to pass into the column. Discard the 1 mL of eluate. 12. Place the column over a 5-mL collection tube. Add 3.5 mL of phosphate buffer to the column and collect all of the 3.5 mL eluate containing the PnPs–PLL conjugates. 13. Cap the collection tube and invert to mix. 14. Aliquot, label and store the PnPs–PLL conjugates at −70°C. 3.2. Coupling of PnPs–PLL Conjugates to Luminex Microspheres

1. Remove PnPs–PLL conjugates from storage at −70°C and warm to room temperature. 2. Label 1.5-mL microcentrifuge tubes with the serotype to be coupled and the microsphere number (see Note 9). 3. Vortex MicroPlex® carboxylated microspheres vigorously to resuspend. 4. Dispense 1 mL of the microspheres to the appropriate 1.5-mL microcentrifuge tube (see Note 10). 5. Centrifuge the tubes at 8,000 × g for 1 min. 6. Remove and discard the supernatant from each tube using a fine tip transfer pipette. 7. Add 100 mL of the PnPs–PLL conjugate to the tube containing the pelleted microspheres and vortex to resuspend the microspheres. 8. Prepare EDC/Sulfo-NHS solution. (a) Accurately weigh 20  mg of EDC and 20  mg of SulfoNHS. (b) Place the EDC and Sulfo-NHS in a 12 × 75-mm glass test tube. (c) Add 400 mL of MES coupling buffer to the tube containing EDC and Sulfo-NHS. (d) Vortex briefly until the chemicals are dissolved.

24  Measurement of Antibodies to Pneumococcal Polysaccharides…

367

9. Immediately add 20 mL of the EDC/Sulfo-NHS solution to each tube containing the resuspended microspheres. 10. Vortex briefly to mix and incubate 2 h in the dark on a rotating shaker. 11. Wash the microspheres 2× with Blocking/Storage buffer. (a) Add 250 mL of Blocking/Storage buffer to each tube. (b) Vortex to resuspend the microspheres. (c) Centrifuge the tubes at 8,000 × g for 1 min. (d) Remove the supernatant with a fine tip transfer pipette. (e) Repeat steps (a)–(d). 12. Resuspend the microspheres in 1 mL of blocking storage buffer and store at 2–8°C (see Note 11). 3.3. Preparation of Master Mix

1. Prepare a pre-master mix by diluting 5 mL of each PLL–PnPs coupled microsphere with Blocking/Storage buffer to a final volume of 5 mL. 2. Test the pre-master mix in the final assay format (see Note 12). (a) Test each new lot against controls and a challenge panel of pre- and post-pneumococcal vaccine sera. Compare the results to a previous lot. If any of the coupled microspheres do not meet quality control (QC) specifications, couple that serotype again and retest the pre-master mix (see Note 13). (b) If all of the coupled microspheres meet QC specifications, prepare the final master mix. Combine the remaining 1 mL of each of the coupled microspheres and dilute the mixture of microspheres to 200  mL with Blocking/ Storage buffer (see Note 14).

3.4. Multiplexed Luminex Assay for Pneumococcal Antibodies

1. Allow all test sera and reagents to come to room temperature. 2. Prepare dilutions of standard. Prepare a 1:15 dilution of standard in Absorbent/Diluent. Prepare 7× threefold dilutions of the 1:15 dilution of standard. 3. Prepare 1:25 dilutions of controls and test samples. Add 10 mL of test sera and serum controls to 240  mL of Absorbent/ Diluent in a round bottom microtiter dilution plate. A color change from purple to blue will be apparent after the addition of serum to the Absorbent/Diluent (see Note 15). 4. Before beginning the assay set up the Luminex 100/200 instrument. Follow the instrument protocols for warming and priming the instrument. Set up a batch for the run. Load a protocol. The protocol should be set up in advance and should specify the microspheres to be counted, the serotype attached to each microsphere, and the values and location for each the

368

J.W. Pickering and H.R. Hill

standards. Place the standards in row 1 beginning with the lowest standard in well A. Specify four-parameter logistic curve fitting (non-weighted). Set the instrument to count 150 mL of each well and 100 of each numbered microsphere. Enter locations and identity of samples to be tested and their dilution factors (see Notes 16 and 17). 5. Transfer 100 ml of each dilution of standard, control and test sera to the appropriate wells of a 96-well filtration plate. Load the standard dilutions in column 1, B through H with the highest standard dilution in 1B. Place 100 ml of the Absorbent/ Diluent in well 1A as the zero standard. Add 100  ml of the dilutions of controls and test specimens to the appropriate wells of the filtration plate. Do not include any blank wells. The Luminex instruments count each well in succession from top to bottom beginning with well 1A. 6. Vortex the mixture of coupled microspheres vigorously for 10–15 to resuspend and pour the microsphere suspension into a reagent reservoir. 7. Using as multichannel pipette, transfer 100  ml of the microsphere suspension into each well of the filtration plate containing a dilution of standard, control or test serum. 8. Place the filtration plate securely on a rotating shaker and shake the plate on a setting of 700  rpm for 20  min (see Note 18). 9. Place the filtration plate on the vacuum manifold and apply vacuum. Insure that all the supernatant fluid has been drawn through the membrane. Wash the microspheres 1×. (a) Add 300 ml of 1× Wash Buffer to each well. (b) Allow the Wash Buffer to be drawn though the membrane. 10. Remove the filtration plate from the vacuum plate and add 100 ml of PE conjugate to each well containing microspheres. Repeat the incubation and wash steps 7 and 8 above. 11. Using an 8 or 12 channel pipette, add 200  ml of 1× Wash Buffer to each well containing microspheres. Pipette up and down several times to resuspend the microspheres and transfer the microsphere suspension from each well to the corresponding well of a new polystyrene round bottom microtiter plate (see Note 19). 12. Count the plate. Eject the XYP platform and load the plate into the instrument. Run the batch. 13. Run the cure fitting software to calculate the mg/mL of IgG for each serotype tested for the unknown samples.

24  Measurement of Antibodies to Pneumococcal Polysaccharides…

369

4. Notes 1. The Luminex xMAP technology allows for antibody concentrations to all 23 serotypes in the PPV to be measured simultaneously (see Fig. 2 and Table 1). Testing of all 23 serotypes, however, is not necessary to evaluate the antibody response to polysaccharides. Seven serotypes in the PPV and not the PCV13 and 7 serotypes in the PCV13 should be sufficient to evaluate and compare the antibody response to conjugated and unconjugated polysaccharides. Adding more serotypes to the assay increases the cost and labor to prepare the reagents and makes optimization of the assay more difficult.

Fig. 2. Titration of a pooled pneumococcal reference serum against 23 PnPs serotypes and C-Ps. The mean fluorescence intensity (MFI) for each dilution of the standard was determined simultaneously for 23 PnPs serotypes and C-Ps in a Luminex 100 analyzer. The sample diluent contained C-Ps and PnPs 25 as absorbents. The curves were constructed by plotting the MFI against the dilution factor for the standard.

0.15

<0.01

0.02

0.04

0.27

0.07

0.05

0.03

0.07

0.36

0.02

0.04

<0.01

0.02

0.05

0.02

2

3

4

5

6B

7F

8

9N

9V

11A

12F

14

15B

17F

18C

Pre

1

PnPs Serotype

Patient 1

0.15

0.15

1.51

<0.01

0.07

3.77

0.57

0.20

1.94

0.07

0.15

0.33

0.09

8.66

0.38

0.22

Post

0.03

0.03

0.06

<0.01

0.02

0.17

0.12

0.02

0.02

0.05

0.03

0.05

0.01

0.19

<0.01

0.07

Pre

Patient 2

4.52

0.11

0.10

0.80

0.10

0.24

4.59

0.17

0.16

0.16

0.26

0.30

1.93

0.21

<0.01

0.42

Post

0.03

0.11

0.04

<0.01

0.07

0.06

0.06

0.05

0.04

0.04

0.11

0.07

0.03

0.02

<0.01

0.16

Pre

Patient 3

5.62

0.32

0.08

2.62

0.22

0.14

5.00

0.20

0.61

0.12

1.56

0.26

0.92

0.10

0.04

0.48

Post

0.83

0.26

0.29

0.13

0.07

0.32

0.10

0.23

0.16

0.13

0.16

0.39

0.04

0.22

0.40

0.45

Pre

Patient 4

3.24

0.62

1.80

0.82

0.75

1.56

1.20

1.31

12.14

0.43

1.80

2.89

0.34

1.10

7.93

1.62

Post

Table 1 Immunoglobulin G concentrations for pre- and post-pneumococcal vaccine samples determined with a 23 valent Luminex assay

370 J.W. Pickering and H.R. Hill

<0.01

0.22

0.34

0.15

0.03

0.06

19A

19F

20

22F

23F

33F

0.14

0.15

9.52

0.59

0.33

0.01

Post

0.04

0.02

0.06

0.03

0.08

<0.01

Pre

Patient 2

0.09

0.33

0.29

0.22

2.58

0.23

Post

0.04

0.76

0.17

0.06

0.16

0.05

Pre

Patient 3

0.10

2.80

0.48

0.18

1.15

0.66

Post

0.04

0.13

0.25

0.07

0.71

4.17

Pre

Patient 4

0.54

1.62

1.74

1.87

4.65

39.40

Post

Serum samples were submitted to ARUP Laboratories for pneumococcal antibody testing and were de-identified according to protocols approved by the University of Utah Institutional review Board (no. 7275)

Pre

PnPs Serotype

Patient 1

24  Measurement of Antibodies to Pneumococcal Polysaccharides… 371

372

J.W. Pickering and H.R. Hill

2. PnPs dissolve very slowly. Add water to the lyophilized polysaccharides and store at refrigerated temperature until they dissolve completely before bringing to the final volume. 3. SeroMAP™ or MagPlex® carboxylated microspheres from Luminex Corporation could also be used for the assay. SeroMAP™ microspheres were introduced by Luminex Corporation to address a problem with nonspecific binding of some human sera to the MicroPlex® microspheres causing false-positive assay results. We have shown, however, that the SeroMAP™ microspheres reduced but did not eliminate the nonspecific binding problem (19). We also have shown that StabliGuard, which we use for the Blocking/Storage buffer, completely eliminated nonspecific binding to MicroPlex® and SeroMAP™ microspheres (19). We use MicroPlex® microspheres because they couple at a higher efficiency than the SeroMAP™ microspheres (20). MagPlex® microspheres might be the best choice if current or future plans include a magnetic washer. 4. Pooling samples from a large number of individuals is necessary to prepare a pool with sufficient volume and with high antibody concentrations to all of the serotypes. Immune globulin has also been used as a reference standard (15). The reference standard serum should be calibrated against an international reference standard. The current reference standard, 89SF, is nearly depleted, however, and will be replaced by a new reference standard, 007sp, which should be available by the end of 2010. The pneumococcal reference sera are available upon request from the Center for Biologicals Evaluation and Research (CBER), Food and Drug Administration. Check the Bacterial Respiratory Pathogen Reference Laboratory web site, www.vaccine.uab.edu, for contact information. 5. C-Ps and PnPs 22F are added to the Absorbent/Diluent to absorb out antibodies that are common to all the serotypes (12). PnPs 25 and PnPs 72 (available from ATCC) have been substituted for PnPs 22F when 22F was one of the test serotypes (21). A better option, however, would be to use CWPS MULTI from Staten Serum Institut as the absorbent (15). The common antigen of PnPs 22F, CWPS2, was recently isolated and characterized and CWPS MULTI from Staten Serum Institut contains both C-Ps and CWPS2 (22). 6. Tween 20, routinely used in ELISA protocols to reduce background, should not be added to the Absorbent/Diluent or wash buffer. We have shown that Tween 20 increases the nonspecific background in the pneumococcal antibody assay (19). 7. Bromocresol purple (BCP) is added as an indicator dye to monitor the addition of serum samples to the Absorbent/

24  Measurement of Antibodies to Pneumococcal Polysaccharides…

373

Diluent in our microtiter plate format. BCP binds serum albumin and changes color from purple to blue. BCP does not affect the assay and could be left out of the Absorbent/Diluent. 8. PE-labeled anti-human IgG can be purchased from a number of vendors, but aliquots of several lots of PE conjugate should be obtained from the vendor and tested in advance. A large supply of an acceptable lot can then be purchased. 9. Microcentrifuge tubes from USA Scientific should be used for processing of Luminex microspheres. With the microcentrifuge tubes from USA Scientific, microspheres do not adhere to the sides of the tubes as they do with microcentrifuge tubes from all other vendors we have tried. 10. We have found it convenient to couple 1 mL of microspheres at a time since the coupling is done in 1.5-mL microcentrifuge tubes. In our hands, the microspheres do not couple as efficiently when scaled up to larger volumes. Rather than coupling multiple serotypes in 1 day, we have found it best to couple multiple 1  mL vials of only one serotype at each coupling. The multiple coupled 1  mL vials of each serotype are tested individually and pooled into a large lot so that lot-to-lot variation is greatly reduced. Since the coupled microspheres are stable for at least 1 year at 4°C, up to a 1 year’s supply of coupled microspheres can be produced for each serotype. 11. We have found that by eliminating the wash step between the EDC/Sulfo NHS activation and coupling steps we achieve higher coupling efficiencies and more reproducible results. 12. It is much less time consuming to QC the coupled microspheres in a multiplexed format than individually. 13. Quality control criteria for acceptance of each new lot of coupled microspheres should be established. Most lots of coupled microspheres should be acceptable by the QC criteria but reject lots that are far outside of the established ranges or that have MFI values that are significantly lower than previous lots. The QC criteria may need to be adjusted with each new lot of purified polysaccharide from ATCC. 14. It becomes a very fruitless and time-consuming effort to try and adjust the coupled microspheres so that the instrument counts the same number of microspheres for each serotype. If the instrument is set to count a minimum 100 of each microsphere, it does not matter if more than 100 microspheres are counted for some serotypes. 15. Ideally, pre- and 1-month post-pneumococcal serum samples should be tested together in the same assay run to minimize any run-to-run or reagent lot-to-lot variation. The pre-sample should be stored frozen (−70°C) until the post-sample becomes available.

374

J.W. Pickering and H.R. Hill

16. We find it more convenient to run controls as samples rather than entering control values for 14 different serotypes into the protocol. 17. Luminex has licensed the xMAP technology to a number of other companies. Instruments and curve fitting software from other vendors could be used for the assay. However, assay results may differ among different software vendors and among different curve fitting models. 18. Although not recommended, the plates could be incubated without shaking but the incubation times must be increased from 20  min to 30 or 40  min to compensate for the slower reaction kinetics. It is not necessary to cover the plates during the incubations. We did not observe photo bleaching of the microspheres after 2 × 1 h incubations with plates not covered. 19. The microspheres can be counted directly in the filtration plate, but in our hands some wells occasionally wick dry and require recounting. We eliminate this problem by transferring the microspheres to a round bottom plate before counting. References 1. Musher, DM (1992) Infections caused by Streptococcus pneumoniae: clinical spectrum, pathogenesis, immunity, and treatment. Clin Infect Dis 14:801–807 2. Whitney CG, Farley MM, Hadler J, et al (2003) Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide c­onjugate vaccine. New Engl J Med 348:1737–1746 3. Dagan R, Givon-Lavi N, Zamir O et al (2002) Reduction of nasopharyngeal carriage of Streptococcus pneumoniae after administration of a 9-valent pneumococcal conjugate vaccine to toddlers attending day care centers. J Infect Dis 185:927–936 4. Goldblatt D, Hussain M, Andrews N et  al (2005) Antibody responses to nasopharyngeal carriage of Streptococcus pneumoniae in adults: a longitudinal household study. J Infect Dis 192:387–393 5. Picard C, Puel A, Bustamante J et  al (2003) Primary immunodeficiencies associated with pneumococcal disease. Curr Opin Allergy Clin Immunol 3:451–459 6. Rijkers GT, Sanders LA, Zegers B J (1993) Anti-capsular polysaccharide antibody deficiency states. Immunodeficiency 5:1–21 7. Cheng YK, Decker, PA, O’ByrneMM et  al (2006) Clinical and laboratory characteristics of 75 patients with specific polysaccharide

a­ ntibody deficiency syndrome. Ann Allergy Asthma Immunol 97:306–311 8. Go, ES, Ballas Z K (1996) Anti-pneumococcal antibody response in normal subjects: a metaanalysis. J Allergy Clin Immunol 98:205–215 9. Paris K, Sorensen RU (2007) Assessment and clinical interpretation of polysaccharide antibody responses. Ann Allergy Asthma Immunol 99:462–464 10. Bryant KA, Block SL, Baker SA et  al (2010) Safety and immunogenicity of a 13-valent pneumococcal conjugate vaccine. Pediatrics 125:866–875 11. Wernette CM, Frasch, CE, Madore D et  al (2003) Enzyme-linked immunosorbent assay for quantitation of human antibodies to pneumococcal polysaccharides. Clin Diagn Lab Immunol 10:514–519 12. Concepcion NF, Frasch CF (2001) Pneumococcal type 22  F polysaccharide absorption improves the specificity of a pneumococcal-polysaccharide enzyme-linked immunosorbent assay. Clin Diagn Lab Immunol 8:266–72 13. Pickering JW, Martins TB, Greer RW et  al (2002) A multiplexed fluorescent microsphere immunoassay for antibodies to pneumococcal capsular polysaccharides. Am J Clin Path 117:589–596 14. Lal G, Balmer P, Stanford E et  al (2005) Development and validation of a nonaplex assay

24  Measurement of Antibodies to Pneumococcal Polysaccharides… for the simultaneous quantitation of antibodies to nine Streptococcus pneumoniae serotypes. J Immunol Meth 296:135–147 15. Elberse KE, Tcherniaeva I, Berbers GA et  al (2010) Optimization and application of a ­multiplex bead-based assay to quantify serotypespecific IgG against Streptococcus pneumoniae polysaccharides: response to the booster vaccine after immunization with the pneumococcal 7-valent conjugate vaccine. Clin Vaccine Immunol 17:674–682 16. Biagini RE, Schlottmann SA, Sammons DL et al (2003) Method for simultaneous measurement of antibodies to 23 pneumococcal capsular polysaccharides. Clin Diagn Lab Immunol 10:744–750 17. Schlottmann, S. A., Jain, N., Chirmule, N. and Esser, M. T. (2006) A novel chemistry for conjugating pneumococcal polysaccharides to Luminex microspheres. J Immunol Methods 309:75–85 18. Borgers H, Moens L, Picard C et  al (2010) Laboratory diagnosis of specific antibody deficiency to pneumococcal capsular polysaccharide

375

antigens by multiplexed bead assay. Clin Immunol 134:198–205 19. Pickering JW, Larson MT, Martins TB et  al (2010) Elimination of false-positive results in a luminex assay for pneumococcal antibodies. Clin Vaccine Immunol 17:185–189 20. Waterboer T, Sehr P, Pawlita M (2005) Supression of non-specific binding in serological Luminex assays. J Immunol Methods 309:200–204 21. Marchese RD, Jain NT, Antonello J et al (2006) Enzyme-linked immunosorbent assay for measuring antibodies to pneumococcal polysaccharides for the PNEUMOVAX 23 vaccine: assay operating characteristics and correlation to the WHO international assay. Clin Vaccine Immunol 13:905–912 22. Skovsted IC, Kerrn MB, Sonne-Hansen J et al (2007) Purification and structure characterization of the active component in the pneumococcal 22  F polysaccharide capsule used for adsorption in pneumococcal enzyme-linked immunosorbent assays. Vaccine 25: 6490–6500

Chapter 25 Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates Jean-Philippe Ebran, Nabil Dendane, and Oleg Melnyk Abstract A method for the preparation of carbohydrate microarrays inside 96-well polystyrene microtiter plates is described. The key step in this strategy represents the synthesis of carbohydrate–dextran conjugates by copper (I)-catalyzed [3 + 2] cycloaddition between alkyne carbohydrate derivative and a specially designed azido dextran polymer. The conjugates adsorb efficiently on polystyrene surface and can be printed inside 96-well plates using a non-contact piezoelectric microarrayer. Model interactions with a selection of lectins (concanavalin A, wheat germ agglutinin, Erythrina Cristagalli) display the efficiency of the immobilization method, its reproducibility and the specificity of biomolecular interactions occurring at the polystyrene– water interface. Key words: Carbohydrate microarrays, Polystyrene, Dextran, Click chemistry, Azide, Alkyne, Lectin, Fluorescence

1. Introduction Carbohydrate microarrays have been developed for high-throughput analysis of large carbohydrate libraries to decipher carbohydrate– protein interactions involved in biological processes (1). In a first approach, surface chemistries and detection techniques previously developed for gene or protein microarrays have been adapted to this novel analytical tool. Therefore, carbohydrate biochips have been mainly prepared on microscope glass slides, a widely accepted format. One step forward application of carbohydrate microarrays in biology relies on the capacity to screen not only carbohydrate libraries but also large collections of biological samples. A solution to overcome this problem is to combine microarray and 96-well formats such as those used in enzyme-linked immunosorbent assays (ELISA).

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_25, © Springer Science+Business Media, LLC 2012

377

378

J.-P. Ebran et al.

Contrary to proteins, oligosaccharides or polysaccharides adsorb poorly on polystyrene surfaces. Physisorption of negatively charged carbohydrates on polystyrene can be promoted by chemically modifying plastic substrates, in particular by introducing positive charges (2, 3). However, negatively charged functional groups within the carbohydrate, involved in protein binding, may be masked by interacting with the solid substrate. Other methods employing the formation of a covalent bond between the probe and the substrate (4), a specific immobilization approach based, for example, on biotin–streptavidin interaction (5, 6) or preparation of conjugates binding to plastic through mainly interfacial forces have been described (6–9). The later strategy was selected in this work because carbohydrate conjugates can be characterized prior to the printing step. Since carbohydrate–protein interactions are usually less affine than other biomolecular ones, high-specific interactions within the carbohydrate microspot as well as low background signals are required for the efficiency of carbohydrate biochips (10). Therefore, an anionic carboxymethylated dextran polymer was selected as a template for carbohydrate immobilization to minimize nonspecific adsorption of proteins (11). To overcome the low propensity of negatively charged polysaccharides to physisorb on plastic substrates (2), dextran polymer was chemically modified with hydrophobic benzyl groups. To allow attachment of alkynefunctionalized carbohydrate probes using the chemoselective copper catalyzed 1,3-dipolar Huisgen cycloaddition reaction, the dextran polymer was further derivatized with azido moieties. The conjugates were printed inside 96-well polystyrene microtiter plates using a non-contact piezoelectric microarrayer. Model interactions with a selection of lectins (concanavalin A, wheat germ agglutinin, Erythrina Crystagalli) display the efficiency of the immobilization method, its reproducibility and the specificity of biomolecular interactions occurring at the polystyrene–water interface.

2. Materials 2.1. Polysaccharides Preparations

1. d-(+)-Glucose, anhydrous, 99 + %. 2. 4-Pentyn-1-ol 98%. 3. Sulphuric acid p.a. (H2SO4). 4. Silica gel 60 M (230–400 mesh ASTM). 5. Dichloromethane (CH2Cl2). 6. Methanol (MeOH). 7. Pyridine (puriss.; absolute; over molecular sieve). 8. Acetic anhydride (Ac2O).

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates

379

9. 4-Dimethylaminopyridine 99% (DMAP). 10. Petroleum ether 40–60°C (Pentane, GPR Rectapur). 11. Ethyl acetate (EtOAc). 12. 1H NMR spectra were performed on an Avance DR × 300 MHz Brücker spectrometer. 13. Chloroform Deuterated + 0.03% tetramethylsilane TMS(CDCl3). 14. a-Cyano-4-hydroxycinnaminic acid 99%. 15. MALDI TOF MS spectra were acquired in the linear mode using a PerSeptive Biosystems Voyager DE-STR instrument fitted with a 337  nm nitrogen laser (25  kV, grid 92%, delay time 750 ns). 16. Sodium methoxide (MeONa). 17. Folded filter paper, Qualitative (Dia/Size: 185 mm). 18. Dowex-50 W × 4 ion-exchange resin (100–200 Mesh, H+). 19. Deuterium oxide (D2O). 20. Dextran T40 from Leuconostoc ssp. (Mr ~ 40,000). 21. Deionized water (see Note 1). 22. Isopropyl alcohol (i-PrOH, 2-propanol). 23. Sodium hydroxide (NaOH), aqueous sodium hydroxide 1N solution. 24. Chloroacetic acid 99%. 25. Hydrochloric acid fuming 37%, aqueous hydrochloric acid 1N solution. 26. N-Cyclohexyl-N ¢-(2-morpholinoethyl)-carbodiimide methyl sulfonate 95%. 27. Benzylamine 99%. 28. Sodium chloride (NaCl), aqueous NaCl 2 M solution. 29. Molecular porous membrane tubing (MWCO: 3500, Spectra/ Por®, Spectrum Laboratories, Inc.) 30. Lyophilisator (Christ, ALPHA 2–4 LD plus). 31. Phosphate-buffered saline (PBS); 1 foil pouch prepares 1  L 0.01 M PBS (NaCl 0.138 M, KCl 0.0027 M) pH 7.4, 25°C. 32. L(+)-Ascorbic acid sodium salt (sodium ascorbate), freshly prepared sodium ascorbate 200  mg/mL solution in PBS 0.01 M. 33. Copper (II) sulfate pentahydrate ~99% (CuSO4·5H2O), CuSO4·5H2O 25 mg/mL solution in PBS 0.01 M. 34. Ethylenediaminetetraacetic disodium salt (EDTA-2Na), aqueous EDTA-2Na 0.5 M solution.

380

J.-P. Ebran et al.

2.2. Microspotting

1. Phosphate buffer saline (PBS): prepare 0.01  M solution (0.138 M NaCl, 0.0027 M KCl, pH 7.4) by dissolving 1 foil pouch in 1 L of deionized water. 2. V-bottomed 384-Well Microarray Plate, cylindrical wells (ABGene, Surrey, UK, AB-1055). 3. Nunc-ImmunoTM Plates MicroWell, Nunc.

MaxiSorp,

Polystyrene,

F-96

4. Non-contact piezoelectric microarrayer (BCA1 Perkin Elmer, MA, USA). 2.3. Incubations and Scanning

1. Bovine serum albumine fraction V (BSA). 2. Tween® 20. 3. Biotinylated concanavalin A (Con A), (Vector Laboratories, Inc., Burlingame, CA 94010, USA). 4. Biotinylated wheat germ agglutinin (WGA), Laboratories, Inc., Burlingame, CA 94010, USA).

(Vector

5. Biotinylated Erythrina Cristagalli lectin (ECL), (Vector Laboratories, Inc., Burlingame, CA 94010, USA). 6. Calcium (II) chloride anhydrous 99% (CaCl2). 7. Manganese (II) chloride tetrahydrate 99 + % (MnCl2). 8. Tetramethylrhodamine-labeled streptavidin (Jackson Immuno Research Europe Ltd., UK). 9. Standard incubator/agitator for 96-well microtiter plates (BioSan, MI, USA). 10. Standard 96-well microtiter plate washing station (wellwash AC, Thermo Electron, Finland). 11. Centrifuge, AllegraTM 25 R (Beckman Coulter, Germany). 12. Fluorescence microarray laser scanner (LS Reloaded, Tecan Group Ltd., Switzerland) and image analysis software (ArrayPro® Analyzer, Switzerland).

3. Methods Methods for preparation of carbohydrate microarrays consist generally in site-specific and covalent attachment of chemically modified carbohydrates to appropriately derivatized surfaces, essentially derivatized glass slides (12–14). Carbohydrate microarrays have also been conceived by physisorption of neoglycolipids to nitrocellulose-coated substrates (15) or of polysaccharides to surface-modified glass slides (16, 17). Carbohydrate polystyrene microtiter plate functionalization by in situ copper (I)-catalyzed [3 + 2] cycloaddition between oligosaccharide azides and lipid alkynes was also described (8).

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates

381

Polystyrene surface represents one of the most standard substrate for ELISA analysis. Nevertheless, its application to carbohydrate microarrays remains rare since carbohydrates adsorb poorly on hydrophobic surfaces. The novel carbohydrate microarray technique depicted herein uses standard polystyrene microtiter plates (Fig. 1) (see Note 2). It is illustrated with the synthesis of a a-glucosepolysaccharide conjugate (Fig.  2), its printing inside polystyrene microtiter wells using a non-contact piezoelectric microarrayer (see Note 3) and proof-of-concept binding studies with biotinylated lectins, i.e., biotinylated concanavalin A Con A, WGA, or biotinylated ECL. Captured lectins were detected with tetramethylrhodaminelabeled streptavidin (Fig. 1) and a fluorescence array scanner. The results are presented in Fig. 2. Synthetic a-glucose conjugate was synthesized in four steps from dextran T40, a a-glucose polymer (Fig. 2). The first step (Step 1) consists in introducing the carboxymethyl moiety to obtain carboxymethyldextran (CMD) (18). Then, amidification with benzylamine in the presence of a water-soluble carbodiimide CMC leads to the carboxymethyldextran benzylamide (CMDB) derivative (Step 2). Another amidification reaction with 3-azidopropylamine (see Note 4) affords carboxymethyldextran benzylamide azide (CMDBA) derivative (Step 3). Finally, the copper (I)-catalyzed cycloaddition (19, 20) between the dextran azide CMDBA and alkyne-derivative of a-glucose affords polysaccharide CMDB a-Glu (Step 4, see Note 5) with 90% conversion (see Note 6). Quantification of carbohydrate microarrays (Fig.  3), after Con A, ECL, and WGA lectins incubations, proved a CMDB a-Glu high specificity for Con A, while ECL and WGA lectins showed no interactions. The signal strength displayed by polysaccharide derivatives CMD, CMDB, and CMDBA were within the background of the microarray. Similarly, b-lactose or b-N-acetylglucosamine polysaccharide conjugates were synthesized and corresponding carbohydrate microarrays demonstrated high specificities, respectively, for ECL or WGA lectins (data not shown). 3.1. Preparation of Alkyne Derivative of Glucose

1. Synthesize a-2-(Pentynyl)-tetra-O-acetyl glucopyranoside d-(+)-Glucose (360.3 mg, 2.0 mmol) is suspended in 4-pentyn1-ol (1.01  g, 12.0  mmol, 6.0 equiv) and stirred at 65°C. H2SO4–SiO2 (see Note 7) (120 mg) is added and the mixture is stirred at 65°C for 10 h. After cooling to room temperature, the reaction mixture is directly purified by flash column chromatography on silica-gel (CH2Cl2/MeOH 100:0 to 90:10) to give a colorless oil as an anomeric mixture (a/b, ratio 1.6:1, 405.0 mg, 1.63 mmol, 79% yield). (see Note 8) To an ice-cold solution of the alkyne residue (405.0 mg, 1.63 mmol, 1.0 equiv) in dry pyridine (5 mL) is added under

O

HO O

O

Step 2

synthetic carbohydrate

Polysaccharide

O

HO OH OH

Step 1

HO O HO O

O

HO OH OH

Polysaccharide

O

HO OH OH

Fluorescence Scanner

Carbohydrate microarray

Polysaccharide

N3

Step 5

Step 3

O

HO O

Polysaccharide

O

HO OH OH

HO O

O

Step 4

HO OH OH

HO O

O

HO OH OH

Streptavidine*

Polysaccharide

Rhodamine-labelled Streptavidine

Biotinylated Lectins

HO O

HO OH OH

Fig. 1. Method used for the preparation of carbohydrate microarrays in polystyrene microtiter plates. Carbohydrates are covalently attached to a dextran template (Step 1), characterized and then microarrayed inside 96-well polystyrene plates by non-contact printing (Step 2). Microarrays were incubated with biotinylated lectins (Step 3) then with tetramethylrhodamine-labeled streptavidin (Step 4). The plate was scanned using a microarray fluorescence scanner at 532 nm (Step 5).

Polystyrene Surface

HO O

Linker

HO OH OH

Copper (I) Catalysis

OH

Linker

N3

Linker

O

Linker

HO

re pt St

e* av id in

HO HO

Linker Linker

Linker Linker

382 J.-P. Ebran et al.

Step 1

O

CMD

+

O

H N

HO HO

OH

BnHN

O CMDB α-Glu

O

O

n

N N

HO HO

OH

O BnHN

Step 4

OH O

O

O

O CMDB

Polysaccharide

O

CuSO4-5H2O (10 mol/%) Sodium Ascorbate (100 mol/%)

N

Step 2

CMC, BnNH2

OH O

n

Polysaccharide O

O

-

O O Na O

Na+ -O O

O

Polysaccharide O

O n

+

O

BnHN

O CMDBA

O

Polysaccharide

O

H N

Cl- +H3N

CMC

O O Na O

-

Step 3

O

O O- Na+ O O- Na+

Fig. 2. Synthetic strategy for the preparation of carbohydrate–dextran conjugates derivatives from dextran T40.

Dextran

NaOH ,ClCH2COOH

O O- Na+ O O- Na+

O n

N3

N3

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates 383

384

J.-P. Ebran et al. 14000

Fluorescence Intensity Arbitrary Unit A. U.

12000

10000 CMD 8000

CMDB CMDBA

6000

CMDB alpha-Glu 4000

2000

0 no lectin

ConA

WGA

ECL

Fig. 3. Quantification of the polysaccharide microarray. CMD, CMDB, CMDBA, and CMDB a-Glu were printed in triplicate at 10 mg/mL in polystyrene microtiter wells using a noncontact piezoelectric microarrayer. Carbohydrate microarrays were incubated in triplicate with biotinylated lectins in PBS-A buffer, i.e., 5 mg/mL biotinylated Con A, 20 mg/mL biotinylated WGA, or 20 mg/mL biotinylated ECL. MnCl2 and CaCl2 were added for a final concentration of 1 mM. The microarrays were illuminated with 5 mg/mL tetramethylrhodamine-labeled streptavidin. Detection was performed at 532  nm. The data correspond to median and interquartile range of fluorescence values expressed in arbitrary units (AU) from which well background was subtracted (mean background level = 570.7 ± 26.7 AU).

argon atmosphere acetic anhydride (1.33  g, 13.0  mmol, 8.0 equiv). After addition of a catalytic amount of DMAP (20.0 mg, 0.32 mmol, 0.1 equiv), the solution is stirred for 5 h at room temperature. The reaction mixture is quenched at 0°C by dropwise addition of MeOH (5  mL) and concentrated in vacuo. The crude product is purified by column chromatography on silica gel (pentane/EtOAc, 8:2) to afford the desired a-isomer as white crystals (190 mg, 0.46 mmol, 28% yield). 1 H RMN (300  MHz, CDCl3) d (ppm): 5.47 (dd, 1H, J = 10.0, 9.6  Hz), 5.06 (m, 2H), 4.87 (dd, 1H, J = 10.2, 3.7 Hz), 4.27 (dd, 1H, J = 12.3 , 4.3 Hz), 4.07 (m, 2H), 3.85 (dt, 1H, J = 9.9 , 5.9 Hz), 3.54 (dt, 1H, J = 9.9 , 6.1 Hz), 2.34 (dt, 2H, 1H, J = 6.8 , 2.7 Hz), 2.10 (s, 3H), 2.07 (s, 3H), 2.03 (s, 3H), 2.02 (s, 3H), 1.97 (t, 1H, J = 2.6 Hz), 1.82 (q, 2H, J = 6.5 Hz). 13 C RMN (75  MHz, CDCl3) d (ppm): 170.9, 170.4, 170.3, 169.8, 96.1, 83.3, 71.0, 70.4, 69.4, 68.9, 68.7, 67.4, 66.9, 62.1, 28.4, 20.9, 20.9, 20.8, 20.8, 15.2.

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates

385

MALDI-TOF (a-cyano-4-hydroxycinnaminic monoisotopic): calculated for C19H26O10 [M + Na]+ 437.5, found 437.1.

acid,

2. Synthesize Pent-4-ynyl-a-d-glucopyranoside To an ice-cold solution of a-2-(pentynyl)-tetra-O-acetyl glucopyranoside (180.0  mg, 0.434  mmol) in MeOH (3  mL) is added 1 M MeONa in MeOH (4 mL) and the reaction mixture is stirred at room temperature for 4 h. The reaction mixture is diluted with MeOH (5  mL) and neutralized with Dowex-50  W (H+) resin (pH 7) (see Note 9). The reaction mixture was filtered over one folded filter paper and evaporated to dryness under reduced pressure. The crude product is purified by column chromatography on silica gel (CH2Cl2/ MeOH; 9:1) to afford a colorless oil (106.9 mg, 0.434 mmol, 99% yield). 1 H RMN (300  MHz, D2O) d (ppm): 5.12 (d, 1H, J = 3.9 Hz), 4.08 (m, 2H), 4.00 (m, 1H), 3.92 (m, 2H), 3.82 (m, 1H), 3.76 (m, 1H), 3.62 (m, 1H), 2.55 (m, 3H), 2.03 (m, 2H). 13C RMN (75  MHz, D2O) d (ppm): 100.0, 87.0, 75.1, 73.6, 73.3, 71.6, 71.4, 68.2, 62.4, 29.3, 16.5. MALDI-TOF (a-cyano-4-hydroxycinnaminic acid, monoisotopic): calculated for C11H18O6. [M + Na]+ 269.1, found 269.0. 3.2. Preparation of Polysaccharide Template



1. Synthesize CMD (carboxymethyldextran) To a well-stirred solution of dextran T40 (2.00 g, 11.1 mmol, 1 equiv) dissolved in deionized water (6.4 mL) is added dropwise 2-propanol (36  mL) then sodium hydroxide (2.4  g, 60.0 mmol, 5.4 equiv) (see Note 10). The reaction mixture is heated under vigorous stirring for 1 h at 60°C. Monochloroacetic acid (3.0  g, 31.7  mmol, 2.8 equiv) is then added and the mixture is heated at 60°C for 16 h. The white gum is collected, precipitated three times with methanol and centrifuged. The white solid is dissolved in deionized water and lyophilizated to afford CMD (2.00 g). Typically this procedure gives substitution ratio (SR) from 1.2 to 1.4 in carboxymethyl moiety. 1 H RMN (300  MHz, D2O) d (ppm): 5.07 (s, ls, Hano), 4.88 (s, ls, Hano), 4.12–3.24 (m, ls). Substitution ratio (SR) of CMD is determined by applying the following equation SR =

(A / B ) - 6 , 2

where A is the proton integral corresponding to carboxymethyl and C2-6 protons generally between 3.00 and 4.50 ppm

386

J.-P. Ebran et al.

and B is the proton integral corresponding to anomeric protons typically between 5.30 and 4.70 ppm. 2. Synthesize CMDB (carboxymethyldextran benzylamide) Carboxymethyldextran CMD (513.3 mg, SR = 1,2) is dissolved in deionized water (10 mL). The pH of the solution is adjusted to 4.74 with an aqueous hydrochloric acid 1 N solution. N-Cyclohexyl-N¢-(2-morpholinoethyl)-carbodiimide methyl sulfonate CMC (931.85 mg, 2.2 mmol, 1.1 equiv) is added to the reaction mixture under vigorous stirring. The pH of the solution is adjusted again to 4.74 and benzylamine (428.6 mg, 4.0 mmol, 2 equiv) is added dropwise to the mixture. The reaction mixture is stirred at room temperature for 16 h. The purification of the dextran is realized by dialysis with a molecular porous membrane first in an aqueous NaCl 2 M solution for 3 days and in water for 3 days. After lyophilization, the CMBD is obtained as a white amorphous foam with a 0.49 substitution ratio SR in benzyl group. 1 H RMN (300 MHz, D2O) d (ppm): 7.21 (s, ls, Har), 5.09 (s, ls, Hano), 4.89 (s, ls, Hano), 4.53–3.39 (m, ls). Substitution ratio (SR) of CMDB is estimated by applying the following equation SR = (A / B ) / 5, where A is the proton integral corresponding to aromatic protons between 7.00 and 7.50 ppm and B is the proton integral corresponding to anomeric protons typically between 4.70 and 5.30 ppm. 3. Synthesize CMDBA (carboxymethyldextran benzylamide azide) To a stirred solution of CMDB (224.2  mg, SRCMD = 1.2, SRCMDB = 0.49) in deionized water (5 mL) beforehand adjusted to pH 4.74 with a hydrochloric acid 1 N aqueous solution, is added N-Cyclohexyl-N ¢-(2-morpholinoethyl)-carbodiimide methyl sulfonate CMC (372.7  mg, 0.88  mmol, 1.1 equiv). The pH of the solution is adjusted to 4.74 and 3-azidopropylamine hydrochloride (327.8  mg, 2.4  mmol, 3 equiv) is added to the mixture. The pH solution is then adjusted to 8.4 with a sodium hydroxide 1 N aqueous solution. The reaction mixture is stirred at room temperature for 17 h. Purification of the dextran is performed by dialysis with a molecular porous membrane first in an aqueous NaCl 2 M solution for 3 days and in water for 3 days. Lyophilization affords CMDBA as a white foam with a 0.38 substitution ratio SR. 1 H RMN (300 MHz, D2O) d (ppm): 7.23 (s, ls, Har), 5.05 (s, ls, Hano), 4.84 (s, ls, Hano), 4.33–3.26 (m, ls), 1.68 (s, ls, CH2). Substitution ratio (SR) of CMDBA is evaluated by employing the following equation

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates

387

SR = (A / B ) / 5, where A is the proton integral corresponding to CH2 of azidopropyl moiety between 1.50 and 2.00 ppm and B is the proton integral corresponding to anomeric protons typically between 4.70 and 5.30 ppm. 4. Synthesize CMDB a-Glu Sodium ascorbate (100  mL of freshly prepared 200  mg/mL solution in PBS 0.01 M, 0.05 mmol, 1 equiv) and copper (II) sulfate pentahydrate (100  mL of 25  mg/mL in PBS, 0.005 mmol, 0.1 equiv) are added to a stirred suspension of pent-4-ynyl-a-d-glucopyranoside (43.8 mg, 0.178 mmol, 1,8 equiv) and CMDBA (28.7  mg, SRCMD = 1.2, SRCMDB = 0.42, SRAzide = 0.38) in a 1:1 mixture of deionized water and PBS 0.1 M (1 mL). The reaction mixture is stirred at room temperature for 20 h. The mixture is quenched with an EDTA– 2Na 0.5 M aqueous solution (2 mL) (see Note 11) and directly purified by dialysis with a molecular porous membrane in an aqueous NaCl 2 M solution for 3 days then in water for 3 days. Lyophilization gives CMDB a-Glu as a white amorphous foam with a 0.34 substitution ratio in triazole. 1 H RMN (300 MHz, D2O) d (ppm): 7.63 (s, ls, Htriazole), 7.21 (s, ls, Har), 5.00 (s, ls, Hano), 4.90 (s, ls, Hano), 4.85 (s, ls, Hano), 4.23–3.22 (m, ls), 3.09 (s, ls, CH2), 2.64 (s, ls, CH2), 1.22 (s, ls, CH2), 1.19 (s, ls, CH2). Substitution ratio (SR) of CMDB a-Glu is determined with the following equation SR = A / B , where A is the proton integral corresponding to the triazole between 7.50 and 7.80 ppm and B is the proton integral corresponding to anomeric protons typically between 4.70 and 5.30 ppm. 3.3. Carbohydrate Microspotting

1. Prepare printing solutions of CMD, CMDB, CMDBA, and CMDB a-Glu at 10.0 mg/mL in PBS buffer A (pH 7.4) and place 25  mL of each solution in the V-bottomed 384-well microtiter plate. 2. Print the mixtures (three drops, 300 pL each, three replicates per polysaccharide, distance of 500  mm between centers of adjacent spots, spot size 210  mm in diameter) in microtiter plates using the BCA1 Microarrayer (see Note 12). 3. Printed plates are stored in aluminum foil at 4°C.

3.4. Incubations and Scanning

1. Prepare PBS-A buffer by dissolving 0.01% v/v Tween®20 in PBS. 2. Prepare PBS-B buffer by dissolving 3.0% w/v BSA in PBS-A. 3. Prepare PBS-C buffer by dissolving MnCl2 and CaCl2 in PBS-A for final concentrations of [MnCl2] = [CaCl2] = 1 mM.

388

J.-P. Ebran et al.

4. Three patterns are incubated per lectin. The wells are first saturated 15 min under shaking using the incubator/agitator at 25°C in PBS-B (100 ml per well). 5. Prepare biotinylated Con A solution in PBS-C buffer at 5 mg/mL. 6. Prepare biotinylated WGA solution in PBS-C buffer at 20 mg/mL. 7. Prepare biotinylated ECL solution in PBS-C buffer at 20 mg/mL. 8. After saturation time, microarrays are washed three times 5 min at 25°C with PBS-A (100 mL per well). 9. Microarrays are incubated at 25°C with lectin solutions (100 mL per well, 1 h under shaking). 10. In the meantime, prepare solution of tetramethylrhodamine– streptavidin in PBS-A (final concentration of 5 mg/mL). 11. Microarrays are washed 3 × 5 min at 25°C with PBS-A (100 mL per well) and immediately incubated at 25°C with tetramethylrhodamine–streptavidin solution (100 mL per well, 2 h, under shaking in darkness). 12. Microarrays are washed three times 5 min at 25°C with PBS-A (100 mL per well) followed by three times 5 min at 25°C with deionized water (100 mL per well). 13. The plate is spin-dried by centrifugation at 1120 × g for 10 min at 4°C. Finally the wells are flushed to dryness with a nitrogen stream. 14. Microchips are analyzed with the microarray scanner at l = 532 nm.

4. Notes 1. Deionized water has a resistivity of 18 MW cm. 2. The conjugates can also be used for performing standard ELISAs. Typically the conjugate is coated at 10.0 mg/mL for 16 h at room temperature. 3. The quill pin printing technology was unsuccessfully applied on polystyrene microtiter well surface using a Qarray2 contact microarrayer (TMC, Technical Manufacturing Corporation, Peabody, MA 01960, USA). Indeed no spots were adsorbed due to the high water contact angle (80°) on untreated polystyrene. 4. Few syntheses of the 3-azidopropylamine have been previously published (21–23). Preparation of 3-azidopropylamine remains,

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates

389

however, difficult due to its low boiling point (48–50°C at 15 mmHg). Therefore, synthesis of the hydrochloride salt by acidic extraction, followed by lyophilization is highly recommended. 5. Adsorption of CMDB, CMDBA, and CMDB a-Glu polysaccharides on polystyrene plate was characterized by contact angle measurements at 20°C with water as reference liquid (at least 10 drops of 1 mL). Typical experiment consists in soaking a PS plate into a 10  mg/mL polysaccharide solution in PBS buffer A for 18  h. Water contact angles on coated PS plates were typically in the range 45–50° and significantly lower than untreated PS (~80°). 6. Conversion was determined by 1H NMR analysis in D2O and by estimating ratio between newly formed triazole proton peak integral (7.63  ppm) and phenyl protons peak integral (7.16 ppm). 7. Preparation of H2SO4–SiO2 (24): To a slurry of silica gel (10 g, 200–400 mesh) in dry diethyl ether (50 mL) was added concentrated H2SO4 (3 mL) with stirring for 5 min. The solvent was evaporated under reduced pressure resulting in free flowing H2SO4–SiO2 which was then dried at 110°C for 3 h. 8. Ratio of a and b isomers was determined by 1H NMR analysis in D2O. 9. Dowex-50 W (H+) resin was beforehand washed with MeOH. 10. 2-Propanol needs to be added very slowly and under vigorous stirring. 11. The rule of EDTA–2Na 0.5  M solution is to remove most of copper traces from the polysaccharide which may interfere during lectin–carbohydrate interactions. Copper level was determined by ICP-AES atomic emission spectroscopy on a 25  mg sample. EDTA treatment decreases copper level from 210 to 33 ppm. 12. Pattern must be positioned preferably in the middle of each well to avoid any edge effects.

Acknowledgments We gratefully acknowledge financial support from CNRS, Université de Lille Nord de France, Institut Pasteur de Lille, IFR 142, Région Nord Pas de Calais, the European Community (FEDER), Endotis Pharma, Inc., and from Cancéropôle Nord-Ouest. This research was performed using the Chemistry Systems Biology platform (http://csb.ibl.fr).

390

J.-P. Ebran et al.

References 1. Love, K. R., and Seeberger, P. H. (2002) Carbohydrate arrays as tools for glycomics. Angew Chem Int Ed Engl 41, 3583–6, 3513. 2. Leinonen, M., and Frasch, C. E. (1982) Classspecific antibody response to group B Neisseria meningitidis capsular polysaccharide: use of polylysine precoating in an enzyme-linked immunosorbent assay. Infect Immun 38, 1203–7. 3. Marson, A., Robinson, D. E., Brookes, P. N., Mulloy, B., Wiles, M., Clark, S. J., Fielder, H. L., Collinson, L. J., Cain, S. A., Kielty, C. M., McArthur, S., Buttle, D. J., Short, R. D., Whittle, J. D., and Day, A. J. (2009) Development of a microtiter plate-based glycosaminoglycan array for the investigation of glycosaminoglycan-protein interactions. Glycobiology 19, 1537–46. 4. Zielen, S., Broker, M., Strnad, N., Schwenen, L., Schon, P., Gottwald, G., and Hofmann, D. (1996) Simple determination of polysaccharide specific antibodies by means of chemically modified ELISA plates. J Immunol Methods 193, 1–7. 5. Gehring, A. G., Albin, D. M., Reed, S. A., Tu, S. I., and Brewster, J. D. (2008) An antibody microarray, in multiwell plate format, for multiplex screening of foodborne pathogenic bacteria and biomolecules. Anal Bioanal Chem 391, 497–506. 6. Cherniak, R., Cheeseman, M. M., Reyes, G. H., Reiss, E., and Todaro, F. (1988) Enhanced binding of capsular polysaccharides of Cryptococcus neoformans to polystyrene microtitration plates for enzyme-linked immunosorbent assay. Diagn Clin Immunol 5, 344–8. 7. Gray, B. M. (1979) ELISA methodology for polysaccharide antigens: protein coupling of polysaccharides for adsorption to plastic tubes. J Immunol Methods 28, 187–92. 8. Fazio, F., Bryan, M. C., Blixt, O., Paulson, J. C., and Wong, C. H. (2002) Synthesis of sugar arrays in microtiter plate. J Am Chem Soc 124, 14397–402. 9. Satoh, A., Fukui, E., Yoshino, S., Shinoda, M., Kojima, K., and Matsumoto, I. (1999) Comparison of methods of immobilization to enzyme-linked immunosorbent assay plates for the detection of sugar chains. Anal Biochem 275, 231–5. 10. Dyukova, V. I., Shilova, N. V., Galanina, O. E., Rubina, A. Y., and Bovin, N. V. (2006) Design of carbohydrate multiarrays. Biochim Biophys Acta 1760, 603–9.

11. Graves, H. C. (1988) The effect of surface charge on non-specific binding of rabbit immunoglobulin G in solid-phase immunoassays. J Immunol Methods 111, 157–66. 12. Lee, M. R., and Shin, I. (2005) Facile preparation of carbohydrate microarrays by site-specific, covalent immobilization of unmodified carbohydrates on hydrazide-coated glass slides. Org Lett 7, 4269–72. 13. Shin, I., Park, S., and Lee, M. R. (2005) Carbohydrate microarrays: an advanced technology for functional studies of glycans. Chemistry 11, 2894–901. 14. Park, S., and Shin, I. (2002) Fabrication of carbohydrate chips for studying proteincarbohydrate interactions. Angew Chem Int Ed Engl 41, 3180–2. 15. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., and Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrateprotein interactions. Nat Biotechnol 20, 1011–7. 16. Wang, D., Liu, S., Trummer, B. J., Deng, C., and Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–81. 17. Carion, O., Lefebvre, J., Dubreucq, G., DahriCorreia, L., Correia, J., and Melnyk, O. (2006) Polysaccharide microarrays for polysaccharideplatelet-derived-growth-factor interaction studies. Chembiochem 7, 817–26. 18. Huynh, R., Chaubet, F., and Jozefonvicz, J. (2001) Anticoagulant properties of dextranmethylcarboxylate benzylamide sulfate (DMCBSu); a new generation of bioactive functionalized dextran. Carbohydr Res 332, 75–83. 19. Rostovtsev, V. V., Green, L. G., Fokin, V. V., and Sharpless, K. B. (2002) A stepwise huisgen cycloaddition process: copper(I)-catalyzed regioselective “ligation” of azides and terminal alkynes. Angew Chem Int Ed Engl 41, 2596–9. 20. Tornoe, C. W., Christensen, C., and Meldal, M. (2002) Peptidotriazoles on Solid Phase: [1,2,30-Triazoles by Regiospecific Copper(I)Catalyzed 1,3-Dipolar Cycloadditions of Terminal Alkynes to Azides. . J. Org. Chem. 67, 3057–3064. 21. Carboni, B., Benalil, A., and Vaultier, M. (1993) Aliphatic amino azides as key building blocks for efficient polyamine syntheses. J. Org. Chem. 58, 736–3741.

25  Carbohydrate Microarrays in 96-Well Polystyrene Microtiter Plates 22. Tamanini, E., Rigby, S. E., Motevalli, M., Todd, M. H., and Watkinson, M. (2009) Responsive metal complexes: a click-based “allosteric scorpionate” complex permits the detection of a biological recognition event by EPR/ENDOR spectroscopy. Chem. Eur. J. 15, 3720–8. 23. Wan, Q., Chen, J., Chen, G., and Danishefsky, S. L. (2006) A potentially valuable advance

391

in the synthesis of carbohydrate-based anticancer vaccines through extended cycloaddition chemistry. J. Org. Chem. 71, 8244–8249. 24. Roy, B., and Mukhopadhyay, B. (2007) Sulfuric acid immobilized on silica: an excellent catalyst for Fischer type glycosylation. Tetrahedron Lett. 48, 3783–3787.

Chapter 26 Photoimmobilization of Saccharides Gregory T. Carroll and Denong Wang Abstract Interest in understanding the biological information content of carbohydrates has led to the development of a variety of methodologies for preparing carbohydrate microarrays. A key challenge is to find a general method to make carbohydrate molecules adhere to a solid chip in a stable manner while preserving the biorecognition properties present when in their native biological environment. The complexity of carbohydrates makes chemical modification prior to surface deposition rigorous when large libraries of surface-immobilized sugars are desired. In this report, we review a versatile photochemical method to carbohydrate immobilization that does not require premodification of the carbohydrate. The method utilizes surfaces modified with photoactive carbonyls that can insert into C–H bonds upon photoexcitation. Key words: Anthrose, Bacillus anthracis, Carbohydrates, Glycans, Glycoepitopes, Microarrays, Oligosaccharides, Photocoupling, Phthalimide, Saccharides, Self-assembled monolayer

1. Introduction Many biotic surfaces present carbohydrate moieties that contribute to biological recognition processes. Important examples include fertilization, embryonic development, cell differentiation and cell– cell communication, blood coagulation, inflammation, chemotaxis, as well as host recognition and immune responses to microbial pathogens (1, 2). Recognition of the importance of understanding carbohydrate interactions has led to a variety of methods for immobilizing carbohydrates on surfaces for high-throughput studies (3–5). The methods generally differ in three characteristics: (1) the type of surface interaction (covalent vs. noncovalent); (2) whether or not a premodification of the carbohydrate is required; and (3)

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_26, © Springer Science+Business Media, LLC 2012

393

394

G.T. Carroll and D. Wang

the postulated orientational control (specific vs. nonspecific). Ideally, covalent immobilization without premodification is ­preferred. A carbohydrate immobilized through stable covalent bonds does not rinse off the surface when subjected to routine assay protocols and hence microarrays containing both low- and high-molecular-weight sugars can be steadfastly prepared. The ability to immobilize carbohydrates without premodification is an obvious advantage in terms of time, cost, and the suitability for use in biological laboratories. Site-specific immobilization is preferred; however, even methods that are not expected to occur in sitespecific manners most likely have enough carbohydrates presenting the proper epitopes required for protein recognition as evidenced by a number of reports (6–11). In this chapter, we review a photochemical method that allows unmodified carbohydrates to be immobilized on a photoactive chip (7). A key advantage of the method is that underivatized carbohydrates can be immobilized using the procedure. The surface presents phthalimide chromophores that can insert into C–H groups as depicted in Fig.  1. Upon absorption of a photon, the carbonyl can undergo a hydrogen abstraction reaction, generating radicals (12). The radicals can then recombine to form new carbon–carbon bonds. Although sugars could be immobilized on a modified surface composed of only phthalimide-derivatized silanes (Fig. 2) when delivered via spin coating, the preparation of mixed surfaces containing phthalimide and amine functionalities was found to be more suitable for delivery of carbohydrates via a robotic spotter (Fig. 3). A variety of sugars ranging from 2,000-kDa polysaccharides to monosaccharides can be immobilized using the photochemical approach outlined below and have been shown to retain their immunogenic properties in most cases. In addition to revealing known carbohydrate–protein interactions, such as the recognition of Conconavalin A by surface-bound mannose oligosaccharides, the surface was used to characterize the immunogenic properties of a tetrasaccharide found on the surface of the Bacillus anthracis exosporium (9), indicating that the methodology is truly applicable to illuminating biological phenomena.

O N R1 + R2 O

O

O R4 C H

hν

R4 N R1 + R2

C R3

R3 OH

N R1 OH

C R2 R3 R3

Fig. 1. The phthalimide chromophore can undergo a photochemical hydrogen abstraction reaction followed by recombination to form a covalent bond.

26  Photoimmobilization of Saccharides

Br O

N

O OH OH

N

O

O

O

O

395

N

O

O

N

OH OH

DMF + O

O Si O O

N K

toluene

O

O Si O O

O Si O

O

Si O

O

Si O O

Compound 1 SAM 1

Fig. 2. Preparation of a silane-modified phthalimide for self-assembly on glass.

Fig. 3. Photogenerated glycan arrays for rapid identification of pathogen-specific immunogenic sugar moieties. This diagram shows that a panel of mono-, oligo-, and polysaccharides were spotted on the photoreactive glass slides followed by UV irradiation to induce covalent coupling of the saccharides to the array substrate. To probe potential immunogenic sugar moieties, pathogen-specific antibodies were applied to react with these glycan arrays. A reproducible positive detection in this assay may lead to rapid recognition of pathogen-specific carbohydrate structures of immunological importance. Reproduced with permission from Wiley-VCH Verlag GmbH & Co. KGaA (9).

396

G.T. Carroll and D. Wang

2. Materials 1. 11-Bromoundecanetrimethoxysilane (Gelest). Moisture sensitive. Store at room temperature. 2. Potassium phthalimide (Aldrich). Photosensitive. Store at room temperature. 3. Anhydrous DMF (Aldrich). Store at room temperature. 4. Argon or nitrogen for performing reaction in dry environment and photochemical reaction in deoxygenated environment. 5. Chloroform. Store at room temperature. 6. Glass microscope slides. 7. Piranha solution: 7:3 (v/v) sulfuric acid (H2SO4):hydrogen peroxide (30% H2O2). CAUTION: Piranha is a strong oxidizer and may result in an explosion if not handled carefully. Exercise extreme caution in preparation and handling. Mixing H2SO4 with H2O2 is highly exothermic and can lead to violent boiling and splashing. Mixing piranha with organic compounds may cause an explosion. 8. Anhydrous toluene. Store at room temperature. 9. Aminopropyltrimethoxysilane (Gelest). 10. Methanol (solvent grade). 11. Carbohydrate solutions in saline (0.9% NaCl) at a given concentration (0.01–1.0 mg/ml). 12. 1× phosphate-buffered saline (PBS), pH = 7.4 (NaCl 137 mM, KCl 2.7 mM, Na2HPO4 8 mM, KH2PO4 2 mM). 13. 1× PBS with 0.05% (v/v) Tween 20. 14. Blocking solution: 1% (w/v) bovine serum albumin (BSA) in PBS containing 0.025% (w/v) NaN3. 15. Primary antibody solution: 1.0–10  mg/ml of antibody in 1% (w/v) BSA in PBS containing 0.05% (v/v) Tween 20 and 0.025% (w/v) NaN3. 16. Secondary staining reagents: Antihuman (or other species) IgG, IgM, or IgA antibodies with distinct fluorescent tags, Cy3, Cy5, or FITC, in 1% (w/v) BSA in PBS containing 0.05% (v/v) Tween 20 and 0.05% (w/v) NaN3.

3. Methods 3.1. Synthesis of a Photoactive PhthalimideTerminated Silane, 1, for Self-Assembly on a Glass Chip

1. Add 3.3  mmol of 11-bromoundecanetrimethoxysilane to an equimolar amount of potassium phthalimide in 60  ml anhydrous DMF. 2. Stir overnight at room temperature in an argon or nitrogen environment.

26  Photoimmobilization of Saccharides

397

3. Add 50 ml chloroform. 4. Transfer the solution to a separatory flask containing 50 ml of DI H2O. After mixing, separate the organic and aqueous layer. 5. Add 20  ml chloroform to the aqueous layer. Mix and separate. 6. Repeat step 5 with fresh chloroform. 7. Combine the organic layers and rinse with at least three 20 ml portions of DI H2O. 8. Remove the chloroform by rotoevaporation to obtain a pale yellow liquid. 9. If solution contains residual DMF, this can be removed with a vacuum pump; however, residual DMF does not prevent surface functionalization. 3.2. Preparation of Photoactive Surface for Carbohydrate Immobilization

1. Suitable substrates include glass, quartz, and silicon containing a layer of SiO2. Boil substrates in piranha (7:3 sulfuric acid:hydrogen peroxide) for 1 h. CAUTION: Piranha is highly explosive. Exercise extreme caution in preparation. Do not mix with organic material. Remove substrates and place in vial of H2O. Repeat twice using fresh H2O. Leave substrate in H2O until further use (see Note 1). 2. Prepare a solution for self-assembly containing a concentration of 1 mM 1 and 5 mM aminopropyltrimethoxysilane in anhydrous toluene (see Note 2). 3. Remove substrate from H2O, dry with argon or nitrogen, and immerse in self-assembly solution. Leave undisturbed for 12 h. 4. Upon removal from self-assembly solution, immediately immerse in fresh toluene. Rinse substrates by three cycles of sonication for 2 min each in the following solvents: toluene, toluene:methanol 1:1, and methanol (see Note 3).

3.3. Photocoupling Carbohydrates to Chip

1. Apply carbohydrate solutions to chip with a high-precision robot to form spots with diameters of approximately 150  mm containing approximately pg-to-ng quantities of carbohydrates. 2. Air dry substrate and place in quartz tube. Cap with rubber septum and purge with argon or nitrogen to remove oxygen. Irradiate with approximately 300-nm UV light. Irradiation for 1  h with a simple desktop lamp containing a 300-nm bulb (8 W) is suitable. 3. Rinse substrates with PBS (pH 7.4) and 0.05% Tween 20 five times, with 5 min of incubation during each washing step.

398

G.T. Carroll and D. Wang

3.4. Immunostaining of Carbohydrate Microarrays

The staining procedure for carbohydrate microarrays is basically identical to the routine procedure for immunohistology. Immunostaining steps of carbohydrate arrays are as follows. 1. Rinse printed microarray slides with 1× PBS, pH 7.4, and 0.05% Tween 20 for 5 min. 2. Block slides with 1% BSA in PBS containing 0.05% NaN3 at room temperature for 30 min. 3. Stain each subarray with 50  mL of test sample, which is diluted in 1% BSA PBS containing 0.05% NaN3 and 0.05% Tween 20. 4. Incubate the slide in a humidified chamber at room temperature for 60 min. 5. Wash slides five times with 1× PBS, pH 7.4, and 0.05% Tween 20. 6. Stain slides with 50  mL of titrated secondary antibodies. Antihuman (or other species) IgG, IgM, or IgA antibodies with distinct fluorescent tags, Cy3, Cy5, or FITC, are mixed and then applied on the chips. 7. Incubate the slide in a humidified chamber with light protection at room temperature for 30 min. 8. Wash slides five times with solution indicated in step 5. 9. Place the slide in a 50 ml FalconTM centrifuge tube and spin at 1,000  rpm (199.36 × g) for 2  min to remove the washing buffer. 10. Cover the slides in a histology slide box to prevent fluorescent quenching of signal by lights.

3.5. Probing Immunogenic Sugar Moieties Using Photogenerated Glycan Arrays

We have explored the use of photogenerated glycan arrays to probe the potential immunogenic sugar moieties expressed by the spores of Bacillus anthracis (9). Our rationale was that if B. anthracis spores expressed antigenic carbohydrate structures, then it would be possible for immunizing animals using the spore preparations to elicit antibodies specific for these structures. This assumption was made on the basis of the fact that the host immune system is able to recognize subtle changes in sugar structures, especially those that are exposed on the surfaces of microbial pathogens that are foreign components of the mammalian hosts. We characterized the precise carbohydrate structures that are recognized by the rabbit antisera using the photogenerated glycan arrays. A unique technical advantage of this method is the ability to produce epitope-specific glycan arrays using unmodified monoand oligosaccharides. We applied, therefore, this technology to display a large panel of saccharide structures, including synthetic fragments and derivatives of the anthrose-containing tetrasaccharide

26  Photoimmobilization of Saccharides

399

side chain of the B. anthracis exosporium and a number of control carbohydrate antigens, for an immunological characterization. A schematic overview of this biomarker identification strategy is shown in Fig. 3. This glycan array analysis confirmed that a tetrasaccharide of BclA glycoprotein bears a dominant antigenic determinant, which is composed of a terminal anthrose residue and three adjacent l-rhamnoses. The terminal trisaccharide unit is essential for the constitution of a highly specific antigenic determinant. Given the fact that this carbohydrate moiety is displayed on the outermost surfaces of B. anthracis spores and its expression is highly specific for the spore of B. anthracis, the anthrose-containing tetrasaccharide can be considered an important immunological target. Its applications may include identification of the presence of B. anthracis spores, surveillance and diagnosis of anthrax infection, and development of novel vaccines targeting the B. anthracis spore. Efforts must also be made to explore the biological role of this highly specific carbohydrate moiety of B. anthracis.

4. Notes 1. The quality of the surfaces obtained depends in part on cleanliness. Use the most pure water available. Ideally, water with a resistivity of 18.2 MW cm and total organic contaminant of less than 5 parts per billion should be used; however, it is still possible to prepare the surfaces after rinsing with water of lower quality. 2. Self-assembly of trimethoxysilanes can be enhanced by adding small amounts of water and acid. Such treatment was not used for the mixed surface described above but may possibly enhance the reaction times or surface coverage obtained. 3. If surface is allowed to dry upon removal from self-assembly solution, the aminosilane may polymerize leaving white deposits that are difficult to remove.

Acknowledgments We are grateful to Drs. Nicholas J. Turro and Jeffrey T. Koberstein of Columbia University for their contribution to the joint development of photogenerated glycan arrays. This work is currently supported by NIH grant U01CA128416 to D. Wang.

400

G.T. Carroll and D. Wang

References 1. Wang, D. (2006) Carbohydrate Antigens. In: Meyers, R.A. (ed.) Encyclopedia of Molecular Cell Biology and Molecular Medicine, 2nd Ed. Wiley-VCH, Weinheim. 2. Carroll, G. T.; Wang, D.; Turro, N. J.; Koberstein, J. T. (2009) Photogeneration of Carbohydrate Microarrays. In: Dill, K.; Liu, R.H.; and Grodzinsky, P. (ed.) Microarrays: Preparation, Detection Methods, Microfluidics, Data Analysis and Applications. Springer, New York. 3. Wang, D. (2003) Carbohydrate Microarrays. Proteomics 3, 2167–2175. 4. Carroll, G.; Wang, D.; Turro, N.; Koberstein, J. (2008) Photons to illuminate the universe of sugar diversity through bioarrays. Glycoconj. J. 25, 5–10. 5. Zhou, X.; Carroll, G. T.; Turchi, C.; Wang, D. (2008) Carbohyrate Microarrays as Essential Tools of Post-genomic Medicine. In: Garg, H.G.; Cowman, M.K.; and Hales, C.A. (ed.) Carbohydrate Chemistry, Biology and Medical Applications. Elsevier. 6. Wang, D.; Liu, S.; Trummer, B. J.; Deng, C.; Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–281.

7. Carroll, G. T.; Wang, D.; Turro, N. J.; Koberstein, J. T. (2006) Photochemical micropatterning of carbohydrates on a surface. Langmuir 22, 2899–2905. 8. Angeloni, S.; Ridet, J. L.; Kusy, N.; Gao, H.; Crevoisier, F.; Guinchard, S.; Kochhar, S.; Sigrist, H.; Sprenger, N. (2005) Glycoprofiling with micro-arrays of glycoconjugates and lectins. Glycobiology 15, 31–41. 9. Wang, D.; Carroll, G. T.; Turro, N. J.; Koberstein, J. T.; Kovac, P.; Saksena, R.; Adamo, R.; Herzenberg, L. A.; Herzenberg, L. A.; Steinman, L. (2007) Photogenerated glycan arrays identify immunogenic sugar moieties of Bacillus anthracis exosporium. Proteomics 7, 180–184. 10. Liu, L. H.; Dietsch, H.; Schurtenberger, P.; Yan, M. D. (2009) Photoinitiated Coupling of Unmodified Monosaccharides to Iron Oxide Nanoparticles for Sensing Proteins and Bacteria. Bioconjugate Chem. 20, 1349–1355. 11. Wang, D.; Lu, J. (2004) Glycan arrays lead to the discovery of autoimmunogenic activity of SARS-CoV. Physiol. Genomics 18, 245–248. 12. Turro, N. J. (1991) Modern Molecular Photochemistry. University Science Books, Sausalito, CA.

Chapter 27 Microwave-Assisted Method for Fabrication of Carbohydrate Cluster Microarrays on 3-Dimensional Hydrazide–Dendrimer Substrate Xichun Zhou, Jian Zhang, and Denong Wang Abstract We present here a method for construction of carbohydrate cluster microarrays. This technology utilizes a 3-dimensional poly(amidoamine) starburst dendrimer monolayer assembled on glass surface, which is functionalized with terminal hydrazide groups for site-specific coupling of carbohydrates without prior chemical derivatization. Microwave radiation energy is applied to accelerate carbohydrate coupling on chips. Since this bioarray platform is designed to present sugar chains in defined orientation and cluster configurations, it is particularly suitable for exploration of the structural and conformational diversities of glycoepitopes and their functional properties. Key words: Carbohydrate microarray, Glycan arrays, Glycan-binding protein, Glycomics, Microwaveassisted reaction, Oligosaccharides, Polysaccharides

1. Introduction Carbohydrate microarray, where multiple carbohydrates are arrayed on a substrate, is a new development at the frontiers of glycomics. Carbohydrate microarrays allow parallel screening of the interactions of carbohydrates with other biological macromolecules in a high-throughput manner. Each screening experiment requires only picomoles of material (1–4). Carbohydrate microarrays possess, thus, a wide variety of potential applications. These include, but are not limited to, rapid determination of the binding profile of carbohydrate-binding proteins, detection of specific antibodies for the diagnosis of diseases, characterization of carbohydrate–cell recognition events, and the high-throughput screening of inhibitors to prevent carbohydrate–protein interactions.

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_27, © Springer Science+Business Media, LLC 2012

401

402

X. Zhou et al.

Although polysaccharide- and glycoconjugate-based microarrays have been reported for years (5–7), there is so far no commercial substrate and method that researchers can directly use to fabricate the mono- and oligosaccharide-based glycan arrays. Several groups have made significant progress in developing novel approaches to fill up this gap (9–18). Mono- and oligosaccharides may be chemically modified with reactive tags, such as amine (8, 9), thiol (10), cyclopentadiene (11), lipid (12), fluorous (13) azide (14), photoactive (15), and biotin (16). These chemically “activated” carbohydrates can be immobilized onto substrates that were functionalized with compatible chemical groups. The chemical modifications of carbohydrates usually require multistep synthetic manipulations, which are often technically challenging and time consuming. Efforts have been also made to devise methods that allow for covalent immobilization of carbohydrates on substrates without prior chemical derivatization. Carroll et al. reported a method for the covalent patterning of unmodified oligosaccharides onto substrates functionalized with photoactive phthalimide chromophores (17, 18). Upon exposure to UV light, the photoactive aromatic carbonyls react with C–H groups of the sugars to form a covalent bond. The photocoupling procedure may target more than one CH- group on the sugar rings. This strategy has an advantage in improving saccharide-coupling efficiency by grapping a sugar chain with multiple attachment sites. However, it is unlikely suitable for production of the single-epitope glycan arrays since the coupled saccharides are presented in more than one configurations in a given microspot. We report here an alternative platform and technique for fabrication of carbohydrate microarrays through site-specific binding of unmodified carbohydrates. This technology presents sugar chains in defined orientation and cluster configurations. It is, thus, uniquely suitable for determination of the fine specificity of a carbohydrate–protein interaction.

2. Materials 2.1. Solutions for Preparing Hydrazide– Dendrimer-Coated Glass Slide

1. Microscope glass slides (Corning® 75 × 25 mm Plain Microscope Slides, Product #2947-75 × 25). 2. Piranha solution is prepared by adding 3 part H2SO4 into 1 part of H2O2 dropwisely before being used. 3. GPTS silanization solution: 3-Glycidyloxypropyl trimethoxysilane (GPTS) is dissolved into toluene to form 1 mM solution. 4. 10% (w/v) solution of PAMAM starburst dendrimers (Generation 4) from Sigma–Aldrich is diluted in methanol to form 1% dendrimer solution in methanol.

27  Microwave-Assisted Method for Fabrication of Carbohydrate Cluster…

403

5. Cross-linker solution: 1-ethyl-3-(3-dimethylaminopropylcarbodimide) (EDC) and N-hydroxy-succinimide (NHS) are dissolved into DMF to form 1 M solution and stored at 4°C. 6. Saturation solution of succinic anhydride in DMF. 7. Adipic acid dihydrazide (ADHZ) is dissolved at 10 mg/mL in DI water and stored at 4°C. 2.2. Reagents and Equipment for Carbohydrate Microarray Fabrication and Immunoassay

1. Sources of carbohydrates on the arrays: (1) Sigma–Aldrich (e.g., (Mana1-3-(Mana1-6)-Mana1-6)-(Mana1-3)-Manb14GlcNAcb1-4GlcNAc (MAN-5) and ((Mana1-6Mana1)23,6Mana1-6)-(Mana1-6Mana1-6Mana1-3)-Manb14GlcNAcb1-4GlcNAc (MAN-9), Mannan); (2) Dextra (e.g., Lewis Blood Group oligosaccharides) (see Note 1). 2. 384-well microplates (Corning, low-volume) (see Note 2). 3. Printing buffer is prepared by mixing 1 part of 50% formamide with 1 part of phosphate-buffered saline (PBS). 4. Carbohydrate probe solutions: Carbohydrate probes are dissolved into the printing buffer at 100 mM for monosaccharides and oligosaccharides, and 1  mg/mL for polysaccharides (see Note 3). 5. Automated arraying robots (Perkin Elmer or ArrayIt Spot®2). 6. Microwave oven for domestic usage from GE. 7. Blocking buffer: 1× PBS containing 0.05% Tween 20 (PBST) and 1% BSA. 8. Washing buffer: 1× PBS containing 0.05% Tween 20. 9. Assay buffer: 0.1× PBS (pH 7.4) containing 0.05% Tween 20 and 1 mM MnCl2 and CaCl2. 10. Biotinylated lectins Laboratories.

from

EY

Laboratories

or

Vector

11. Antibodies from abcam or Invitrogen. 12. Cy3–streptavidin (Cy3–SA) conjugate kit from Invitrogen. 13. Detection antibody of Cy3 labeled anti-Mouse (or antiHuman) IgG conjugates from Invitrogen. 14. Multi-well (1 × 16) microarray hybridization Cassettes (ArrayIt Inc.). 15. Fluorescence scanners with Cy3 and Cy5 channels (PerkinElmer or Alpha Innotech Inc.). 16. The fluorescence images are analyzed using the Image Acquisition and Analysis software (AlphaScan) or ScanArray Express software (PerkinElmer). 17. Statistical analyses are performed using SigmaPlot 5.0 (Jandel Scientific, San Rafael, CA) or by Microsoft Excel®.

404

X. Zhou et al.

3. Methods The immobilization of carbohydrates onto hydrazide substrate is based on the hydrazone formations between the highly reactive nucleophilic hydrazide and the carbonyl group at the reducing end of suitable carbohydrates via irreversible condensation. Thus, the monosaccharides and oligosaccharides immobilized on the novel substrate are in defined orientation. Reactions of hydrazide groups with free carbohydrates are known to be chemoselective and have been used for the synthesis of various glycoconjugates and glycopeptides. The efficiency of the method relies on the ease of formation and good stability of the schiff base linkage up to pH 9. It is known that cyclic adduct with b-configuration is produced predominantly in reactions of carbohydrates with hydrazide-containing compounds. This carbohydrate immobilization method requires only a few derivatization steps on slide surface treatments but without the requirement of prior chemical modification of carbohydrate probes, which makes it very applicable for preparing carbohydrate microarrays in individual laboratories. Microwaves (~0.3–300  GHz) lie between the infrared and radio frequency (RF) electromagnetic spectrum. In the past two decades, the use of microwave radiation has greatly increased in its application to accelerating reactions in synthetic organic chemistry and biochemistry. It is widely thought that microwaves accelerate chemical reactions by increased rate of mutarotation (the process of equilibration between unreactive closed-ring structures that proceeds via the reactive open-chain intermediate) above that achieved by equivalent thermal heating and by accelerated solvent heating which increases the rate of mutarotation. Microwave is clean, convenient, and effective energy for chemical reaction. The use of microwave energy to assist the immobilization of carbohydrates has not been reported before. To demonstrate the utilization of microwave radiation energy for rapid immobilization of carbohydrates spotted onto hydrazide–dendrimer substrates, representative carbohydrate probes, including Mannose (monosaccharides), Mannobiose (disaccharides), Sialyl Lewis X (tetrasaccharide), isomaltopentose (pentasacchride), and Mannan (polysaccharide) (see Table  1), are printed onto the hydrazide–dendrimer substrates. The affinities of most oligosaccharide–protein interactions are usually lower than the affinities from DNA–DNA or protein–protein interactions. This fact has pointed the requirement of forming multivalent binding between the immobilized carbohydrates and proteins for detecting their interactions. The high density of surface functional groups on PAMAM starburst dendrimers should help to maximize the immobilization capacity of carbohydrates and thus provide the multivalency (i.e., the cluster effect) required for detectable carbohydrate–protein interaction.

27  Microwave-Assisted Method for Fabrication of Carbohydrate Cluster…

405

Table 1 Glycans immobilized on hydrazide–dendrimer-coated slides for GBP interrogation No.

Carbohydrate

No.

Carbohydrate

1

d-Glc

34

Fuca1-2Gal

2

d-Gal

35

GlcNAcb1-4GlcNAcb1-4GlcNAc

3

L-Rha

36

Neu5Aca2-6Galb1-4Glc

4

GlcOMe

37

Neu5Aca2-6Galb1-4GlcNAc

5

ManNAc

38

SO3-3Galb1-4-(Fuc1-4)-GlcNAc

6

NeuNAc

39

Glca1-4Glca1-4Glc

7

Galb1-4Glc

40

Fuca1-2Galb1-4Glc

8

Galb1-3GlcNAc

41

Gala1-4Galb1-4GlcNAc

9

Glca1-4Glc

42

GalNAca1-3-(Fuc1-2)Gal

10

GlcNAcb1-4GlcNAc

43

Galb1-3GlcNAcb1-3Galb1-4Glc

11

Mana1-3Man

44

Gala1-3Galb1-4Galb1-3Gal

12

Fuca1-2Galb1-3(Fuca1-4)GlcNAc

45

Neu5Aca2-3Galb1-4-(Fuca1-3)GlcNAcb1-3Gal

13

Galb1-4(Fuca1-3) GlcNAcb1-3Galb1-4Glc

46

Galb1-3(Fuca1-4)GlcNAcb1-3Galb14(Fuca1-3)Glc

14

Glca1-4Glca1-4Glc 4Glca1-4Glc

47

(Mana1-3-(Mana1-6)-Mana1-6)- (Mana1-3)Manb1-4GlcNAcb1-4GlcNAc

15

Neu5Ac2-6(Galb1-3) GlcNAcb1-3Galb1-4Glc

48

(Mana1-6Man)n (Mannan)

16

GalNAca1-3(Fuca1-2) Galb1-4(Fuca1-3)Glc

49

(4GlcUAb1-3GalNAcb1)n (Chondroitin sulfate)

17

d-Man

50

Galb1-4Galb1-4Glc

18

d-Fuc

51

Neu5Aca2-3Galb1-4Glc

19

L-Fuc

52

Neu5Aca2-3Galb1-4GlcNAc

20

GalOMe

53

SO3-3Galb1-4-(Fuc1-3)GlcNAc

21

ManOMe

54

Galb1-4-(Fuc1-3)-GlcNAc

22

GlcNAc

55

Glcb1-4Glcb1-4Glc

23

Frucb2-1Glca

56

Galb1-3Galb1-3GlcNAc

24

Galb1-4GlcNAc

57

Fuca1-2Galb1-3GlcNAc

25

Galb1-3GalNAc

58

Gala1-3-(Fuc1-2)Gal

26

Glcb1-4Glc

59

Galb1-4GlcNAcb1-3Galb1-4Glc

27

Mana1-2Man

60

Glca1-4Glca1-4Glc4Glc (continued)

406

X. Zhou et al.

Table 1  (continued) No.

Carbohydrate

No.

Carbohydrate

28

Mana1-6Man

61

Neu5Aca2-3Galb1-3-(Fuca1-4)-GlcNAcb1-3Gal

29

Fuca1-2Galb1-4-(Fuca1-3) GlcNAc

62

Neu5Ac2-3Galb1-3(Neu5Ac2-6)GlcNAcb13Galb1-4Glc

30

Fuca1-2Galb1-3Galb14(Fuca1-3)Glc

63

((Mana1-6Mana1)2-3,6Mana1-6)-(Mana16Mana1-6Mana1-3)-Manb1-4GlcNAcb14GlcNAc

31

(Mana1-3-(Mana1-6)Mana1-6)(Mana1-3)-Man

64

(4GlcUAb1-4GlcNAc(6S)a1)n (Heparan sulfate)

32

Neu5Ac2-6Galb14GlcNAcb1-3Galb1-4Glc

65

E.coli LPS

33

(GlcNAcb1-2Mana1)213,6Man

66

Printing buffer

3.1. Preparation of HydrazideFunctionalized Dendrimer Substrates (Fig. 1)

1. Microscope glass slides are cleaned with hot piranha solution and then thoroughly rinsed with distilled water and HPLCpurified ethanol. 2. The cleaned slide substrates are immersed into a GPTS silanization for 30 min to form a monolayer on the glass surface with epoxy functional groups toward the outside (step 1, Fig. 1). 3. The GPTS-functionalized slides are then incubated overnight in PAMAM solution at room temperature under gentle shaking (step 2, Fig. 1). Excessive monomers physically adsorbed on slide surface are removed by washing with methanol and deionized water, and the dendrimer-coated slides are dried in a nitrogen stream. This can generate a PAMAM dendrimer monolayer on the glass slide surface (see Note 4). 4. The PAMAM dendrimer-coated glass slides are placed into a saturation solution of succinic anhydride in DMF for 3 h (step 3, Fig. 1). The slides are then carefully washed with pure DMF several times. 5. Subsequently, the carboxyl groups on the substrate surface are activated with 1  M cross-linker solution for 2  h and washed with pure DMF. 6. The activated slides are then immersed into a solution of 10  mg/mL aqueous solution of ADHZ for 2.5  h (step 4, Fig. 1) to produce the hydrazide surface groups. Finally, the slides are washed with water and dried under N2 blow (see Note 5).

27  Microwave-Assisted Method for Fabrication of Carbohydrate Cluster…

407

Fig. 1. Chemical process for the preparation of a 3-dimensional, hydrazide-functionalized glass substrate.

3.2. Fabrication of Carbohydrate Microarrays

1. The carbohydrate probe solutions are allocated to the wells of 384-well microplates. A given concentration of each preparation is repeated at least three times to support statistical analysis. 2. The 384-well microplate is then placed into the station of microarray printer. 3. 1  nL of each carbohydrate probe solution from a 384-well plate is printed onto the hydrazide-functionalized dendrimeric substrates by using the arraying robots in 60% relative humidity. 4. The printed carbohydrate microarray slides are transferred into microwave oven for providing microwave radiation energy for immobilization of spotted carbohydrates on the hydrazidefunctionalized, dendrimer-coated slides. 5. Set the energy power level of the microwave at 50%. 6. The microwave radiation dosage is controlled by the microwave time (4–15 min, see Note 6). 7. After treatment of the microarray with microwave radiation, the slide is immersed into washing buffer under gentle shaking for 2 min to remove unbound carbohydrates (see Note 7).

3.3. Detection of GBP Interactions with Carbohydrate Microarrays

1. Spotted substrates are immersed into the blocking buffer for 60  min at room temperature and then rinsed with washing buffer before immunoassay. 2. The immunoassays to probe carbohydrate–protein interactions usually apply indirect sandwich assay format. In brief, the carbohydrate microarrays are incubated with cocktail solution of individual or mixed biotinylated lectins (or other biotinylated

408

X. Zhou et al.

Fig. 2. The representative microarray immunoassay image obtained on the hydrazide–PAMAM dendrimer-coated substrate that is spotted with 100 mM of mono-, di-, oligo-, and polysaccharides (66 carbohydrates in Table 1) in printing buffer and treated under microwave for 10 min. The slide is probed with 1 mg/mL of biotin–WGA and 1 mg/mL of Cy3–streptavidin (Cy3–SA) for 1 h at room temperature. Each carbohydrate is printed three replicates on the array. The RFU intensity represents mean relative fluorescence units that have been subtracted by the local background intensity.

GBPs) and Cy3–streptavidin in assay buffer for 1 h. Following incubation, slides are washed twice in 50  mL of washing buffer for 10 min and then rinsed briefly (30 s) in 0.01× PBS and deionized water, respectively, to remove salts prior to spin dried. 3. Microarrays are scanned at 10-mm resolution with the scanning laser confocal fluorescence microscope. The emitted fluorescent signal is detected by a photomultiplier tube (PMT) at 570  nm for Cy3. For all microarray experiments, the laser power is 75% and the PMT gain is 65%. 4. Fluorescence intensity values for each array spot and its background are calculated using software of microarray scanners. The local background signal is automatically subtracted from the immunoassay signal of each separate spot and the mean signal intensity of each spot is used for data analysis. 5. Statistical analyses are performed using the software listed in Subheading 2.2, item 17. 6. A paradigm for studying carbohydrate–lectin interaction on hydrazide–dendrimer-coated substrate is shown in Fig.  2, where the wheat germ agglutinin (WGA) lectin specifically recognizes the N-acetylglycosamine (GlcNAc) containing carbohydrates displayed by the dendrimer substrate. These include a GlcNAc monosaccharide (No. 22), a GlcNAcb1-4GlcNAc

27  Microwave-Assisted Method for Fabrication of Carbohydrate Cluster…

409

disaccharide (diacetylchitobiose; No.10), and a GlcNAcb14GlcNAcb1-4GlcNAc trisaccharide (triacetylchitotriose; No. 35) that are immobilized on the hydrazide–dendrimer substrate. In addition, this microarray assay recognizes interaction of WGA with Mannan (No. 48) obtained from Saccharomyces cerevisiae, which demonstrates the possibility of the presence of GlcNAc or other WGA cross-reactive glycoepitopes in this Mannan preparation (Fig. 2).

4. Notes 1. Table 1 lists the carbohydrate probes collected from the indicated vendors. 2. The minimal volume required for the 384-well microplate is 5 mL. We usually applied 20 mL of the carbohydrate solution in the plate. 3.  (a) The solution containing 50% formamide and 50% 1× PBS has been identified as the optimal printing buffer for immobilizing monosaccharides, oligosaccharides, and polysaccharides due to the good solubility to carbohydrates and high volatile point. Additionally, it is also a good microwave-adsorbing solvent due to its high dielectric constant. (b) Glycerol, which is widely used as a printing buffer component for protein microarrays to avoid the evaporation of protein printing solution, is not a good component in printing buffer for fabrication of carbohydrate microarrays. This is probably due to the adherence of glycerol on the slide surface after microwave radiation, which might block the carbohydrate’s binding sites. 4. Comparative studies have suggested that the alkali solution is not required for coating dendrimer on the epoxy surface. However, in the case without using alkali, longer incubation time (24 h) is needed to ensure the dendrimer coating. 5. The hydrazide-functionalized slides can be stored at room temperature (20–25°C) in a sealed package and used prior to the expiration date (6 months). Once the package is opened, slides should be stored in a vacuum condition inside a desiccator and protected from light at room temperature. 6. Maximum fluorescent intensities were obtained for 10 min of microwave treatment on both the hydrazide–dendrimer slides, which indicated an effective immobilization of carbohydrate probes on the slide surfaces. We also tested the immobilization of carbohydrates on these slides by traditional heating method

410

X. Zhou et al.

through incubation of the printed slides in water bath for 16 h under 50°C. However, these slides showed much lower signals in lectin detection as compared to the slides under microwave for 10  min, which indicated that fewer carbohydrates were immobilized on the slide by traditional heating method. 7. Given the structural diversity of carbohydrate molecules, precaution must be made on the possible influence of carbohydrate structure on coupling efficiency via reducing end of carbohydrate molecules. A general order of the coupling activity of carbohydrates onto hydrazide substrate is acetylamino carbohydrate (e.g., GlcNAc, GalNAc) > hexose/pentose (e.g., Glc, Man, Fuc) > acidic carbohydrate (e.g., sialyl Lewis X).

Acknowledgments This work was supported by Grant Number 1R43 RR023763-01 and 2R44 RR023763-02 (X.C. Zhou) from the National Center for Research Resources Services (NCRR) and U01CA128416 (D. Wang) from National Cancer Institute (NCI), components of the National Institutes of Health (NIH). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NCRR or NCI or the National Institutes of Health. References 1. Paulson, J. C., Blixt, O., Collins, B. E. (2006) Sweet spots in functional glycomics. Nat. Chem. Biol. 2, 238–248. 2. Wang, D. (2003) Carbohydrate microarrays. Proteomics, 3, 2167–2175. 3. Feizi, T., Chai, W. (2004) Oligosaccharide microarrays to decipher the glyco-code. Nat. Rev. Mol. Cell Biol. 5, 582–588. 4. Park, S., Lee, M. R., Shin, I. (2008) Carbohydrate microarrays as powerful tools in studies of carbohydrate-mediated biological processes. Chem. Commun. 37, 4389–4399. 5. Wang, D., Liu, S., Trummer, B. J., Deng, C., Wang, A. (2002) Carbohydrate microarrays for the recognition of cross-reactive molecular markers of microbes and host cells. Nat. Biotechnol. 20, 275–281. 6. Wang, D., Lu, J. (2004) Glycan arrays lead to the discovery of autoimmunogenic activity of SARS-CoV. Physiol. Genomics 18, 245–248. 7. Willats,W. G. T., Rasmussen, S. E., Kristensen, T., Mikkelsen, J. D., Knox, J. P. (2002) Sugarcoated microarrays: a novel slide surface for the

high-throughput analysis of glycans. Proteomics. 2, 1666–1671. 8. Xia, B., Kawar, Z. S., Ju, T., Alvarez, R. A., Sachdev, G. P., Cummings, R. D. (2005) Versatile fluorescent derivatization of glycans for glycomic analysis. Nat. Methods 2, 845–850. 9. Blixt, O., Head, S., Mondala, T., Scanlan, C., Huflejt, M. E., Alvarez, R., Bryan, M. C., Fazio, F., Calarese, D., Stevens, J., Razi, N., Stevens, D. J., Skehel, J. J., van Die, I., Burton, D. R., Wilson, I. A., Cummings, R., Bovin, N., Wong, C. H., Paulson, J. C. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc. Natl. Acad. Sci. U.S.A. 101, 17033–17038. 10. Ratner, D. M., Adams, E. W., Su, J., O’Keefe, B. R., Mrksich, M., Seeberger, P. H. (2004) Probing Protein-Carbohydrate Interactions with Microarrays of Synthetic Oligosaccharides. ChemBioChem 5, 379–382. 11. Houseman, B. T., Mrksich, M. (2002) Carbohydrate Arrays for the Evaluation of Protein Binding and Enzyme Activity. Chem. Biol. 9, 443–454.

27  Microwave-Assisted Method for Fabrication of Carbohydrate Cluster… 12. Fukui, S., Feizi, T., Galustian, C., Lawson, A. M., Chai, W. (2002) Oligosaccharide microarrays for high-throughput detection and specificity assignments of carbohydrate-protein interactions. Nat. Biotechnol. 20, 1011–1017. 13. Pohl, N. L. (2008) Fluorous Tags Catching on Microarrays. Angew. Chem., Int. Ed. 47, 3868–3870. 14. Sun, X. L., Stabler, C. L., Cazalis, C. S., Chaikof, E. L. (2006) Carbohydrate and protein immobilization onto solid surfaces by sequential DielsAlder and azide-alkyne cycloadditions. Bioconjugate Chem. 17, 52–57. 15. Kohn, M., Wacker, R., Peters, C., Schroder, H., Soulere, L., Breinbauer, R., Niemeyer, C., Waldmann, H. (2003) Staudinger ligation: a new immobilization strategy for the preparation

411

of small-molecule arrays. Angew. Chem., Int. Ed. 42, 5830–5834. 16. Feizi, T., Fazio, F., Chai, W. Wong, C. H. (2003). Carbohydrate microarrays - a new set of technologies at the frontiers of glycomics. Curr Opin Struct Biol. 13, 637–645. 17. Carroll, G. T., Wang, D., Turro, N. J., Koberstein, J. T., Photochemical micropatterning of carbohydrates on a surface. Langmuir 2006, 22, (6), 2899–905. 18. Wang, D., Carroll, G. T., Turro, N. J., Koberstein, J. T., Kovac, P., Saksena, R., Adamo, R., Herzenberg, L. A., Herzenberg, L. A., Steinman, L., Photogenerated glycan arrays identify immunogenic sugar moieties of Bacillus anthracis exosporium. Proteomics 2007, 7, (2), 180–4.

Chapter 28 Combinatorial Glycoarray Simon Rinaldi, Kathryn M. Brennan, and Hugh J. Willison Abstract Glycolipid–protein interactions are increasingly recognised as critical to numerous and diverse biological processes, including immune recognition, cell–cell signalling, pathogen adherence, and virulence factor binding. Previously, such carbohydrate–lectin interactions have been assessed in vitro largely by assaying protein binding against purified preparations of single glycolipids. Recent observations show that certain disease-associated autoantibodies and other lectins bind only to complexes formed by two different gangliosides. However, investigating such 1:1 glycolipid complexes can prove technically arduous. To address this problem, we have developed a semi-automated system for assaying lectin binding to large numbers of glycolipid complexes simultaneously. This employs an automated thin-layer chromatography sampler. Single glycolipids and their heterodimeric complexes are prepared in microvials. The autosampler is then used to print reproducible arrays of glycolipid complexes onto polyvinylidene difluoride membranes affixed to glass slides. A printing density of 300 antigen spots per slide is achievable. Following overnight drying, these arrays can then be probed with the lectin(s) of interest. Detection of binding is by way of a horseradish peroxidase-linked secondary antibody driving a chemiluminescent reaction rendered on radiographic film. Image analysis software can then be used to measure signal intensity for quantification. Key words: Glycolipid, Ganglioside, Ganglioside complexes, Microarray, Glycoarray, Lectin, Immunoassay, Bacterial toxins

1. Introduction In the original fluid mosaic hypothesis describing the biological membrane, lipids and glycolipids were considered an inert solvent in which the functionally more important proteins were randomly distributed and free to move (1). More recent proposals suggest that combinations of oligosaccharide groups from different glycans might form a distinct epitope, termed a “clustered saccharide patch” (2). Much evidence now exists to suggest that cholesterol and glycosphingolipids form clusters, the so-called “lipid rafts”, that act as platforms for proteins and are involved in signal Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8_28, © Springer Science+Business Media, LLC 2012

413

414

S. Rinaldi et al.

transduction (3). Furthermore, protein–carbohydrate interactions are now known to be critically important in a number of diverse biological processes, including immune recognition, cell–cell signalling, pathogen adherence, and virulence factor binding (4–12). Trans carbohydrate–carbohydrate interactions (i.e. between two neighbouring cells) were initially described over 20 years ago having been shown to be required for compaction of the early mouse embryo (13, 14). There is now an increasing realisation that cis interactions (i.e. “side-to-side” lateral interaction within the plane of the membrane) are functionally relevant, as shown for the gangliosides GM2 and GM3 in the anti-metastasis factor CD82mediated control of cell motility (15). Antibodies in the sera of some patients suffering from the acute inflammatory neuropathy Guillain–Barré syndrome (GBS) bind only to heterodimeric complexes of gangliosides while failing to recognise either component molecule in isolation (16–19). This pattern of reactivity has subsequently been termed “complex enhanced” binding (20, 21). Furthermore, the ability of some other anti-ganglioside monoclonal antibodies to bind to their targets has previously been shown to be inhibited by the presence of other ganglioside species. In different melanoma cell lines, anti-GM3 antibodies were only able to bind when GM3 was the sole surface ganglioside. Even in cells expressing 50% GM3, if a long-chain ganglioside, such as GM2, was also present, anti-GM3 binding was abolished (22). A conceptually similar observation has now also been made with regard to antibodies associated with GBS (23, 24). Thus, certain anti-GM1 antibodies are unable to bind their target ligand in tissue and in solid-phase assays, such as the combinatorial glycoarray described here, when a second ganglioside, such as GD1a, is also present (“complex attenuated”). New experimental data, thus, provides the rationale for a reconsideration of lectin–glycolipid interactions to include this combinatorial aspect, and here we describe a method for screening lectins and antibodies against a large number of glycolipid complexes simultaneously. This semi-automated system is a synthesis of previously described methods (16, 25) and employs an automated thin-layer chromatography (TLC) sampler to print reproducible arrays of glycolipid complexes onto polyvinylidene difluoride (PVDF) membranes affixed to glass slides. In this way, some of the technical difficulties involved in working with the large numbers of distinct epitopes generated by the combinatorial approach can be overcome, enabling this property to be more efficiently investigated. The PVDF acts as a hydrophobic surface to which the lipid tails of the glycolipids attach following automatic dispensation by the TLC sampler. This leaves the oligosaccharide head groups free for interaction with the proteins of interest, and from this point on the arrays can be processed along similar lines to a standard ELISAbased assay or a Western dot blot.

28  Combinatorial Glycoarray

415

2. Materials 2.1. TLC Autosampler

2.2. Software

Array printing uses an ATS4 TLC Autosampler (Camag, Switzerland). For spotting small volumes of glycolipids on the array, the smaller 10-ml syringe is recommended. A spray (not contact) needle should be fitted. To allow slides to be held on the printing platform, a separate metal grid with rectangular slots to accommodate 12 slides needs to be produced in-house and fitted over the standard application platform (Fig.  1). The wash bottle should be filled with methanol. 1. Array design and programming: winCATS planer chromatography management software (Camag, Switzerland) plus a spreadsheet package. 2. Image analysis and spot quantification: TotalLab TL100 Array (Nonlinear Dynamics, UK), ImageQuant TL software (Amersham Biosciences, UK), or similar.

2.3. Slides and Membranes

1. Plain glass slides (VWR International, UK). 2. PVDF membrane (Sigma, UK). 3. UHU glue (UHU GmbH, Germany). 4. Hydrophobic barrier pen (Vector Laboratories, UK). 5. Diamond-tipped pen.

Fig. 1. The ATS4 autosampler and slide guide. (a) Dosing syringe and application needle. (b) Sample rack and microvials. (c) Application platform with bespoke slide guide loaded with 12 slides. (d) Wash and rinse bottles. (e) Close-up of slide guide and slides. (f) PVDF membrane affixed to glass slide. (g) Membrane outlined with hydrophobic pen.

416

S. Rinaldi et al.

2.4. Glycolipids and Reagents

1. Many gangliosides and glycolipids are available commercially (e.g. Sigma, UK, Accurate Chemical and Scientific, USA, Calbiochem, USA, Matreya, USA). Others may need to be extracted from natural sources and purified. 2. 1:1 chloroform:methanol (v/v): 20  ml chloroform, 20  ml methanol (>98% purity, HPTLC grade). 3. Stock glycoplipid solutions: 1 mg lyophilised glycolipid, 1 ml 1:1 chloroform:methanol (v/v). Store at −20°C. Stability unknown, but used in assay over 8 weeks without noticeable degradation in signal. 4. Working glycolipid solutions: 100 ml stock glycolipid solution, 900 ml methanol (>98% purity, HPTLC grade). Store at −20°C. Stability unknown, but used in assay over 8 weeks without noticeable degradation in signal. 5. Glycolipid complexes (A:B): 100 ml glycolipid “A”, 100 ml glycolipid “B”. Ganglioside and glycolipid complexes are created by mixing equal volumes of the working solution of each glycolipid in 300  ml capacity microsampling vials (Chromacol, UK) and sealing with a screw top cap with a rubber insert (Chromacol, UK). This allows puncture by the septum punch of the ATS4. The mixtures are sonicated for 3 min prior to use. Store at −80°C. Replace rubber insert caps with solid tops for storage to reduce evaporation. 6. 10× phosphate-buffered saline (PBS) stock: 80  g NaCl, 2  g KH2PO4, 29 g Na2HPO4×12H2O, 2 g KCl, 1,000 ml dH2O. Store at 4°C. Stable long term. 7. 1× PBS working solution: 100 ml 10× PBS, 900 ml dH2O. pH to 7.4. Store at 4°C. Stable long term. 8. 2% (w/v) bovine serum albumin (BSA)/PBS: 20  g BSA, 1,000 ml 1× PBS. Store at 4°C. Stable for 1–2 weeks. 9. 1% (w/v) BSA/PBS: 10 g BSA, 100 ml 1× PBS. Store at 4°C. Stable for 1–2 weeks. 10. Primary antibody/lectins working solution: 1% BSA/PBS, stock antibody/lectin. Dilute to required concentration(s) in just prior to use. A dilution series is recommended for previously untested reagents. 11. Secondary antibody working solution: 1% BSA/PBS, stock antibody. The appropriate horseradish peroxidase (HRP)conjugated secondary antibody should again be prepared just prior to use. Concentration should be adjusted to obtain the optimum signal-to-noise ratio. We found 1:30,000 to be optimal using anti-human IgG(H + L)–HRP (Dako), anti-human IgG(Fc specific)–HRP (Sigma), anti-mouse IgG–HRP (Sigma), and anti-mouse IgM–HRP (Sigma).

28  Combinatorial Glycoarray

417

12. Enhanced Chemiluminescence Plus (ECL+, Amersham/GE Healthcare, UK): 100  ml of solution B, 4  ml of solution A. Store component solutions at 4°C, allow to warm to room temperature prior to mixing. Prepare immediately before use and mix gently.

3. Methods The PVDF glycoarray is a powerful tool both because large number of protein–glycolipid interactions can be assayed simultaneously and because tailor-made grids can be designed and modified for the particular system under investigation. Furthermore, the protocol has the capacity to be expanded beyond that described here to incorporate multimeric (i.e. three or more different component) glycolipid antigens. The principles underlying the technique are simple. The glycolipid targets attach to the PVDF membrane via a hydrophobic interaction. These are then exposed to the lectin of interest, followed by an enzyme-linked secondary antibody. Binding is detected by way of an enzyme-driven chemiluminescent reaction rendered on radiographic film, and following digitisation binding intensity can be quantified using image analysis software. 3.1. Membranes

1. Cut sheets of PVDF into 20 × 25-mm squares (or as required for different array dimensions) using a scalpel. 2. Affix one membrane per slide using UHU glue (see Note 1). The position on the slide needs to be calculated to correlate with the location of the printed array and maintained for each slide (see Note 2). Make sure that there is no overhang of the membrane beyond the slide edge. 3. Outline each membrane with the hydrophobic barrier pen, taking care not to allow the hydrophobic liquid to stain the membrane. Label each slide using a diamond-tipped pen to write directly on the glass. 4. Allow to air dry for at least 10 min prior to further use.

3.2. Array Programs

1. The x and y locations, dimensions, volume, and type of lipid applied need to be specified for each spot. Our standard array utilises spots spaced 2 mm apart, of 0.4 × 0.4-mm dimensions, with 0.1  ml of 100  mg/ml glycolipid solution per spot (see Note 3). 2. The winCATS software only allows manual data entry for each spot. As such, the use of a non-specialised spreadsheet package is invaluable in planning and creating new programs. By use of formulae, fill series, and copy and paste functionality, new

418

S. Rinaldi et al.

programs can be much more easily and reliably produced before being copied and pasted into winCATS. Once the layout for one slide has been produced, simply altering all of the x and/or y co-ordinates by the appropriate amount covers a second slide with an identical layout. In our slide guide, equivalent spots on each consecutive slide in the either column are 31.4 mm apart in the x dimension, whereas the y co-ordinates of the two columns of six slides differ by 92 mm (see Note 4). 3. The maximum number of spots per programs is around 720. Therefore, larger programs need to be broken up into separate subprograms. 4. In winCATS, adjust the filling quality settings to avoid wasting large volumes of glycolipids (under Edit | Plate properties | Required filling quality). The fill-only required volume box should be ticked, and filling vacuum time reduced to 0 s (see Note 5). 3.3. Print Runs

1. Prior to each print run, ensure that the wash bottle has sufficient methanol to cover the duration of the programs and that the waste bottle has enough capacity. 2. Prior commencing the program, the microvials containing the glycolipids should be sonicated for 3 min (see Note 6). 3. Ensure that the platform table is locked in the “up” position to prevent the slides from moving between applications. 4. If separate subprograms are required to produce the array, do not remove the slides between each program. 5. Following completion of the print run, allow the membranes to air dry for a further 10 min before storing at 4°C overnight (see Note 7).

3.4. Assay

1. Block membranes by immersion in at least 100 ml/cm2 of 2% bovine serum albumin/phosphate buffer saline (BSA/PBS) for 1 h at 4°C. 2. Dilute serum samples, CSF, monoclonal antibodies, or other lectins of interest to the required concentration in 1% BSA/ PBS. Between 250  ml (for 20 × 20-mm arrays) and 1  ml (for 70 × 20-mm arrays) of dilute primary solution is required per membrane. 3. Remove the blocked slides, tip off the blocking solution, and briefly tap the edge of the slide on blue roll to dry further. Do not bring the PVDF membrane itself directly into contact with the blue roll or allow any portion of the membrane to dry out completely. 4. In a staining tray, apply the dilute primary lectin solution to each slide as required, cover loosely, and incubate for 2  h at

28  Combinatorial Glycoarray

419

4°C. Ensure that the full area of the membrane is covered and that enough solution is present such that the membranes do not dry out during incubation. 5. Tip off and dry gently as before, and then place the slides briefly back in a rack in the blocking solution. 6. Lectins requiring a secondary antibody now undergo a primary wash phase. Those which do not can enter the secondary wash phase directly (step 8 below). For the primary wash phase, transfer the rack form the blocking solution into at least 500  ml/cm2 of fresh 1% BSA/PBS. Place on a shaker set at 100 rpm for 15 min. Change the 1% BSA/PBS and shake for a further 15 min. 7. Again tap dry the slides, before applying the appropriate HRPlinked secondary antibody (diluted in 1% BSA/PBS, typically to 1 in 30k as above, using the same volume as for the primary), and incubate for 30 min at 4°C. 8. For the secondary wash phase, use two changes of 1% BSA and three changes of 1× PBS, again each of at least 500 ml/cm2. Tap dry the slides first. BSA washes are of 30-min duration, PBS 5 min, both on a shaker set at 100 rpm. 9. Dip the slides briefly in two changes of distilled water (500 ml/ cm2). 10. Apply ECL+. A volume per slide of 1.5× that used for the primary and secondary antibodies is recommended, as the ECL spreads out less well over the membrane surface (see Note 8). The detection solution is applied to the membranes and left for 3 min at room temperature. 11. While the ECL is incubating, line a film cassette with blue roll and cut a transparency sheet to the required dimensions to overlay this. 12. Tip off the ECL solution. 13. Quickly transfer the slides to the film cassette, placing them on top of the blue roll to absorb any excess ECL (see Note 9). Cover with the pre-cut transparency sheet and take the film cassette to the darkroom. 14. A range of exposure times are recommended (e.g. 5  s, 15 s, 1 min, 5 min). 3.5. Image Analysis

1. Digitise the images using a flatbed scanner. Adjustments to the scan setting may be required to obtain useable images (see Note 10). Pictures in 8-bit greyscale, saved using the TIFF file format, are required for analysis. 2. Using the image analysis software, an appropriate analysis grid can be defined, saved, and stretched over each image. Correct assignation of each spot may require referral to a positive

420

S. Rinaldi et al.

Fig. 2. (a) Outline grid. Column and row headings represent the complex spotted at each location. The line of “x” from top left to bottom right represents a line of negative controls (methanol-spotted only). (b) Digitised, processed array overlain with outline grid to allow determination of positive spots and graphical presentation of result. (c) Analysis grid applied using image analysis software. Note that the spot outline radius for the large spot in the rightmost column, second row, has been increased to account for the increased signal generated at this location. (d) Output from image analysis software.

c­ ontrol standard, accurate measurement, or use of positivecontrol marker spots (see Note 11). Applying a labelled grid is also useful in this respect, and can also be used for pictorial representation of the processed grid (Fig. 2). 3. Adjustments to individual spot dimensions may be required to account for more intense spots. 4. Apply background correction (see Note 12). 5. Once these adjustments have been made, intensity values can be exported and analysed using a statistics package.

4. Notes 1. Different fixatives may be trialled if required. Enough UHU should be applied to prevent membrane detachment during the wash phase, but too much can result in an uneven membrane surface. This may require some trial and error to optimise. 2. We find it helpful to draw parallel lines on a sheet of card such that the slide edge can be lined up with one and the membrane edges with the next.

28  Combinatorial Glycoarray

421

3. Increased spot density risks losing the definition of individual spots during image analysis. 4. Programs can be checked before printing onto PVDF by replacing the glycolipids with a 1:10 (v:v) mixture of methylene blue and methanol and replacing the slides with a sheet of paper laid under the slide guide. 5. Failing to do this means that a volume of each sample is pulled straight through the machine and wasted. This menu also allows adjustment of the number of wash cycle between each sample. We would recommend increasing this from one (the default) to two, although this results in a longer time to complete each program. 6. It is most efficient to use a bath sonicator large enough to accept the sample rack direct from the ATS4, although it is also possible to sonicate the samples separately before loading into the rack if such a bath is not available. 7. We use arrays within 24 h of printing. Storage for longer periods in standard conditions results in signal loss. 8. Apply the ECL as quickly as possible following removal of the membranes from their distilled water rinse. If large numbers of slides are being processed simultaneously, performing this step in batches of six or so is recommended. It may be necessary to carefully spread the ECL out over the membrane using the side of a pipette tip while avoiding directly touching the membrane itself. 9. Ensuring that the slides are lined up accurately and parallel to each other at this stage facilitates spot assignation in the subsequent analysis. It may help to lay out a column of blank slides tight to the edge of the film cassette and line the processed slides up against these. 10. Increasing the brightness setting reduces background and improves spot definition. Care must be taken to process each image in the same way to allow comparison between different arrays. 11. We have used an anti-mouse IgG–HRP-linked secondary antibody, diluted 1:8,000 in 1:1 methanol:PBS, to mark the four corners of each array. 12. Background correction using spot edge average settings is useful to compensate for differences in background levels across the membrane. However, if the spot signals have run together and there are no clear margins around each spot, this will return falsely low values. In addition or as an alternative, negative-control spots can be defined.

422

S. Rinaldi et al.

References 1. Singer, S. J. and Nicolson, G. L. (1972) The fluid mosaic model of the structure of cell membranes. Science 175, 720–731. 2. Simons, K. and Ikonen, E. (1997) Functional rafts in cell membranes. Nature 387, 569–572. 3. Hakomori, S. I. (2002) Inaugural Article: The glycosynapse. Proc. Natl. Acad. Sci. USA 99, 225–232. 4. Todeschini, A. R., Dos Santos, J. N., Handa, K., and Hakomori, S. I. (2008) Ganglioside GM2/GM3 complex affixed on silica nanospheres strongly inhibits cell motility through CD82/cMet-mediated pathway. Proc. Natl. Acad. Sci. USA 105, 1925–1930. 5. Todeschini, A. R., Dos Santos, J. N., Handa, K., and Hakomori, S. I. (2008) Ganglioside GM2/GM3 complex affixed on silica nanospheres strongly inhibits cell motility through CD82/cMet-mediated pathway. Proc. Natl. Acad. Sci. USA 105, 1925–1930. 6. Karlsson, K. A. (2001) Pathogen-host proteincarbohydrate interactions as the basis of important infections. Adv. Exp. Med. Biol. 491, 431–443. 7. Miller-Podraza, H., Johansson, P., Angstrom, J., Larsson, T., Longard, M., and Karlsson, K. A. (2004) Studies on gangliosides with affinity for Helicobacter pylori: binding to natural and chemically modified structures. Glycobiology 14, 205–217. 8. Lalli, G., Bohnert, S., Deinhardt, K., Verastegui, C., and Schiavo, G. (2003) The journey of tetanus and botulinum neurotoxins in neurons. Trends Microbiol. 11, 431–437. 9. Avril, T., Wagner, E. R., Willison, H. J., and Crocker, P. R. (2006) Sialic acid-binding immunoglobulin-like lectin 7 mediates selective recognition of sialylated glycans expressed on Campylobacter jejuni lipooligosaccharides. Infect. Immun. 74, 4133–4141. 10. Crocker, P. R. and Redelinghuys, P. (2008) Siglecs as positive and negative regulators of the immune system. Biochem. Soc. Trans. 36, 1467–1471. 11. Schiavo, G. and van der Goot, F. G. (2001) The bacterial toxin toolkit. Nat. Rev. Mol. Cell Biol. 2, 530–537. 12. Eggens, I., Fenderson, B. A., Toyokuni, T., and Hakomori, S. (1989) A role of carbohydrate-carbohydrate interaction in the process of specific cell recognition during embryogenesis and organogenesis: a preliminary note. Biochem. Biophys. Res. Commun. 158, 913–920.

13. Fenderson, B. A., Zehavi, U., and Hakomori, S. (1984) A multivalent lacto-N-fucopentaose III-lysyllysine conjugate decompacts preimplantation mouse embryos, while the free oligosaccharide is ineffective. J. Exp. Med. 160, 1591–1596. 14. Regina, T. A. and Hakomori, S. I. (2008) Functional role of glycosphingolipids and gangliosides in control of cell adhesion, motility, and growth, through glycosynaptic microdomains. Biochim. Biophys. Acta 1780, 421–433. 15. Kaida, K., Morita, D., Kanzaki, M., Kamakura, K., Motoyoshi, K., Hirakawa, M., and Kusunoki, S. (2004) Ganglioside complexes as new target antigens in Guillain-Barre syndrome. Ann. Neurol. 56, 567–571. 16. Kaida, K., Kanzaki, M., Morita, D., Kamakura, K., Motoyoshi, K., Hirakawa, M., and Kusunoki, S. (2006) Anti-ganglioside complex antibodies in Miller Fisher syndrome. J. Neurol. Neurosurg. Psychiatry 77, 1043–1046. 17. Kaida, K., Morita, D., Kanzaki, M., Kamakura, K., Motoyoshi, K., Hirakawa, M., and Kusunoki, S. (2007) Anti-ganglioside complex antibodies associated with severe disability in GBS. J. Neuroimmunol. 182, 212–218. 18. Willison, H. J. (2005) Ganglioside complexes: new autoantibody targets in Guillain-Barre syndromes. Nat. Clin. Pract. Neurol. 1, 2–3. 19. Rinaldi, S., Brennan, K. M., Goodyear, C. S., O’Leary, C., Schiavo, G., Crocker, P. R., and Willison, H. J. (2009) Analysis of lectin binding to glycolipid complexes using combinatorial glycoarrays. Glycobiology 19, 789–796. 20. Rinaldi, S., Brennan, K. M., and Willison, H. J. (2010) Heteromeric glycolipid complexes as modulators of autoantibody and lectin binding. Prog. Lipid Res. 49, 87–95. 21. Lloyd, K. O., Gordon, C. M., Thampoe, I. J., and DiBenedetto, C. (1992) Cell surface accessibility of individual gangliosides in malignant melanoma cells to antibodies is influenced by the total ganglioside composition of the cells. Cancer Res. 52, 4948–4953. 22. Rinaldi, S. and Willison, H. J. (2008) Ganglioside antibodies and neuropathies. Curr. Opin. Neurol. 21, 540–546. 23. Greenshields, K. N., Halstead, S. K., Zitman, F. M., Rinaldi, S., Brennan, K. M., O’Leary, C., Chamberlain, L. H., Easton, A., Roxburgh, J., Pediani, J., Furukawa, K., Furukawa, K., Goodyear, C. S., Plomp, J. J., and Willison, H. J.

28  Combinatorial Glycoarray (2009) The neuropathic potential of anti-GM1 autoantibodies is regulated by the local glycolipid environment in mice. J. Clin. Invest 119, 595–610. 24. Varki, A. (1994) Selectin ligands. Proc. Natl. Acad. Sci. USA 91, 7390–7397.

423

25. Kanter, J. L., Narayana, S., Ho, P. P., Catz, I., Warren, K. G., Sobel, R. A., Steinman, L., and Robinson, W. H. (2006) Lipid microarrays identify key mediators of autoimmune brain inflammation. Nat. Med. 12, 138–143.

Index A Alkyne...................................3, 26, 58, 88, 156, 202, 204, 208, 306, 308, 311, 378, 380, 381, 384–385 Aminoglycosides....................................................... 303–317 Anthrose................................................................... 398, 399 Antibiotic resistance................................................. 303–317 Antibiotics................................................................ 303–317 Antibody...........................5–9, 106, 110, 121, 124, 130, 133, 138, 140, 141, 144, 145, 147, 149, 153, 167, 168, 170, 171, 175, 181, 233–235, 237, 238, 241–245, 247, 248, 285–301, 325, 331, 333–335, 348, 349, 352, 354, 357, 361–374, 395, 396, 398, 401, 403, 414, 416–419, 421 Array............................ 2, 5, 6, 8, 87–100, 145, 153, 156, 157, 161, 168, 189–191, 243–244, 246, 247, 269–284, 295, 296, 300, 314–316, 327, 330, 361–374, 395, 398–399, 402, 403, 414, 418, 421 Autoantibody............................................................ 297, 300 Azide...........................3, 22, 23, 26, 47, 58–62, 64, 66, 67, 88, 139, 141, 156, 185, 187, 198, 202, 203, 208, 209, 214, 288, 292, 306, 308, 311, 380, 381, 386, 402 Azide glycoside............................................................. 60–62

B

Chondroitin sulfate (CS)..........................222–225, 227–228, 321–324, 333, 340, 405 Click chemistry.......................................................... 26, 202 Combinatorial glycoarray......................................... 413–421 Covalent immobilization............. 2–4, 25, 171, 196, 394, 402 CS. See Chondroitin sulfate

D DDI. See DNA-directed immobilization Depolymerization of polysaccharides............... 338, 344–346 Deprotection.................................. 14, 17, 20–22, 24, 27, 60, 88, 158, 202, 236, 277, 278 Dermatan sulfate (DS)............................................. 323, 324 Designer microarrays................................................ 338, 350 Dextran..................................... 232, 238, 339, 344, 345, 355, 378, 379, 381–383, 385, 386 1,3-Dipolar cycloaddition....................26, 202, 208–209, 306 Dissociation constants.............................................. 107, 113 Divinyl sulfone (DVS).................................89, 90, 95–97, 99 DNA-directed immobilization (DDI)..................... 195–216 DPI. See Dual polarisation interferometry DS. See Dermatan sulfate Dual polarisation interferometry (DPI)................... 223–224 DVS. See Divinyl sulfone

Bacillus anthracis.......................................246, 394, 398, 399 Bacterial toxin................................................................... 413 Binding affinity................................................ 103–114, 147 Biochip............................................................. 247, 377, 378 Biointerface...................................................................... 185 Biosensor.................................................................. 183, 221 Biotin–streptavidin bridge................................................ 167 Biotinylated sialosides.............................................. 185–190 BSA-glycosides........................................................ 159–160

ELISA. See Enzyme-linked immunosorbent assay Enzymatic activity............................................................ 103 Enzymatic glycosylation............................109–110, 277, 279 Enzyme-linked immunosorbent assay (ELISA)........................... 91, 96, 137, 156, 167–170, 174, 180–181, 339, 362, 372, 377, 381, 388, 414 Epoxy-silane..................................................... 286, 289, 294

C

F

Cancer biomarker............................................................. 155 Carbohydrate arrays...................................244–246, 269, 398 Carbohydrate chemical synthesis.................................. 13–27 Carbohydrate immobilization........................5, 378, 397, 404 Chemoenzymatic synthesis.................................. 34, 47, 188 Chemoselectivity.............................................................. 103

Fluorescence............................5, 97, 112, 139, 191, 192, 214, 233, 245, 341, 380, 403, 408 Fluorescence detection..................................................... 139 Fluorous based carbohydrate microarray.................. 149–153 Florous-coated glass slide................................................. 151 Fluorous-tagged carbohydrate.................................. 150–152

E

Yann Chevolot (ed.), Carbohydrate Microarrays: Methods and Protocols, Methods in Molecular Biology, vol. 808, DOI 10.1007/978-1-61779-373-8, © Springer Science+Business Media, LLC 2012

425

Carbohydrate Microarrays: Methods and Protocols 426  Index

  

G

I

GAG. See Glycosaminoglycans Ganglioside........................33, 41, 42, 49, 120, 357, 414, 416 Ganglioside complexes....................................... 42, 414, 416 GBP. See Glycan-binding protein b1,3-and b1,6-Glucan oligosaccharides................... 351, 355 b-Glucan polysaccharides................................................. 337 Glycan-binding motif....................................................... 264 Glycan-binding protein (GBP).........138, 168, 251, 405–409 Glycans array...................................... 97, 138, 145, 156, 157, 168, 260–263, 269, 296, 395, 398–399, 402 derivatization...................................................... 137–147 microarray....................................... 1–9, 47, 51, 104–109, 137–147, 251–266, 286, 287, 298 Glycoarray................................. 1–9, 195–216, 357, 413–421 Glycobiology.............................................................. 81, 301 Glycoconjugate..................................... 5, 32, 35, 57–67, 103, 167–182, 196–198, 201, 213, 214, 216, 241, 242, 247, 338, 402, 404 Glyco-epitope........................................................... 246, 409 Glycolipid................................9, 32, 103, 120, 123, 128, 132, 241, 246, 290–292, 298, 413–418, 421 Glycomics..............................................2, 242, 252, 286, 401 Glycomimetics...........................................202, 209, 215, 216 Glyco-peptide arrays.........................270, 271, 276–279, 284 Glycopolymers...................................169, 171–177, 179–181 Glycoprotein............................32, 94, 99, 103, 117, 120, 138, 139, 141, 146, 155, 156, 241, 246, 286, 291, 399 Glyco-SAM......................................................... 88, 91–100 Glycosaminoglycan-binding protein................................ 321 Glycosaminoglycan (GAG) microarray.................... 323, 333 Glycosaminoglycans (GAG)................................7, 8, 70, 80, 118–120, 132, 321–335, 357 Glycosylation.............7, 14–16, 18–20, 23, 27, 33, 35, 37–41, 43, 44, 62, 63, 109–110, 120, 156, 277, 279, 285 Glycosyltransferases............. 61, 104, 105, 107, 110, 269–284 Gold.................................................... 98, 189, 232, 269-284 Growth factor...............7, 8, 80, 223, 231–235, 237, 323, 333

IC50. ...................................................... 7, 104, 107, 111-113 Immobilization chemistry..................................................99 Immobilization techniques....................................... 4, 25–27 Immune deficiency........................................................... 361 Immunoassay..................................... 245, 364, 403, 407, 408 Influenza virus..........9, 50, 120, 134, 251, 252, 257–259, 266

H HA. See Hyaluronic acid Heparan sulfate (HS)..........................80, 232–234, 323, 406 Heparin (Hp)........................4, 7, 8, 70, 71, 79–83, 222–226, 228, 232–234, 236, 323, 324 High-throughput analysis.......................5, 69, 104, 221, 377 HILIC.............................................................. 289, 293, 299 HS. See Heparan sulfate Hyaluronic acid (HA).............................................. 323, 324 Hydrazide.................................................. 108, 225, 401-410 Hydrazide-linked carbohydrates....................................... 105 Hydrazide-conjugated carbohydrates............................... 104 Hydrazide surface..................................................... 227, 406

L Lactoferrin (Lf )........................................................ 223, 228 Lectin................................1, 6, 57, 87, 91, 94, 96–97, 99, 104, 121, 124, 130, 137–147, 150, 153, 161–162, 164, 168, 184, 190, 191, 214, 216, 223, 224, 227, 247, 286, 290, 295, 300, 337, 338, 354, 378, 380–382, 384, 388, 389, 403, 407, 408, 410, 414, 416–419 FITC-ConA............................................................... 150 inhibitors.................................................................... 155 Luminex................................................................... 361–374

M MALDI-ToF MS................................. 6, 160, 197, 211, 216, 269–284, 294, 298, 301, 379 Maleimide-linked glycans................................................ 105 Maleimide-conjugated carbohydrates....................... 104, 107 Mannose–6-phosphate (Man–6-P).......................... 137–147 Man–6-P receptor (MPR)........................138, 141, 145, 147 Mass spectrometry (MS)..............................5–7, 14, 90, 118, 126, 132, 158, 160, 163, 189, 205, 206, 269–284, 286, 287, 294, 298, 299, 301, 323, 338, 340, 341, 345–350, 355, 379 Microarray................................ 1–9, 14, 47, 69–85, 103–114, 117–134, 137–147, 149–153, 157, 183–193, 196, 231–239, 241–248, 251–266, 285–301, 303–317, 321–335, 337–357, 377–389, 394, 401–410 Microwave-assisted reaction............................................. 401 MPR. See Man–6-P receptor MS. See Mass spectrometry

N Neoglycolipid....................................117–134, 337–357, 380 Neoglycoprotein....................................................... 155–164 N-glycans...................118, 138, 144, 290, 292, 293, 298, 356 Nitrocellulose................................. 4, 25, 117–134, 242, 244, 338, 340–341, 347, 351, 356, 380 Non-covalent immobilization of carbohydrates....................5

O Oligosaccharide...................... 2, 13, 32, 69–85, 87, 117–134, 149, 156, 195, 221–229, 241, 299, 323, 337–357, 378, 394, 402, 413 Oxime............................................ 4, 119-121, 131, 132, 356 Oligosaccharide building block.............................. 15, 16, 27



Carbohydrate Microarrays: Methods and Protocols 427 Index     

P Phosphorylation..................................31, 137–147, 235, 304 Photo-coupling......................................................... 397, 402 Phthalimide...................................................... 394–397, 402 Physical adsorption................................... 171–173, 177–180 Pneumococcal polysaccharide................................... 361–374 Polyacrylamide......................................................... 167–182 Pyrrolylated oligosaccharide............ 71, 74, 76, 77, 80, 81, 84 Polysaccharide...........................4, 8, 25, 32, 33, 80, 118, 120, 156, 231–239, 241, 246, 247, 323, 324, 338, 339, 342–346, 350, 353–355, 361–374, 378–381, 384–387, 389, 395, 402–404, 408, 409 Polysaccharide microarrays............................... 231–239, 384 Polystyrene.........169, 171–174, 178, 289, 339, 368, 377–389 Protecting group............................... 6, 14, 15, 17, 18, 20–23, 27, 35–45, 213, 278, 283 PVDF membrane......................................414, 415, 417, 418 P-type lectin............................................................. 137–147

Self-assembled monolayer (SAM)....................6, 25, 87–100, 184, 270, 271, 274, 276–278, 283 Shotgun microarray.......................................................... 285 Sialic acid......................................... 9, 31–51, 120, 132, 186, 188, 189, 223, 228, 251, 252, 257–259, 264, 298, 299, 357 Sialoside............................................35–40, 43–51, 183–193 Sialylation..............................................35–50, 106, 109, 189 Sialyltransferase....................... 35, 45–49, 106, 109, 186, 188 SPOT synthesis.........................................270, 271, 276–279 SPR. See Surface plasmon resonance SPRi. See Surface plasmon resonance imaging Starburst dendrimer monolayer........................................ 401 Surface immobilization........... 13–27, 88, 195, 196, 247, 311 Surface plasmon resonance (SPR)....................1, 6, 7, 69–85, 91, 97, 98, 100, 104, 183–193, 200, 212, 270, 323 Surface plasmon resonance imaging (SPRi)................ 69–85, 88, 89, 91, 94, 97–100, 183–193

R

T

Reductive amination.... 4, 5, 26, 117, 119, 121, 131, 156, 158, 159, 162–163, 298 RNA 7, 303–305, 309–311, 315–317

Thiolated glycoside........................................... 88, 92–93, 99 Thiols..................................3, 6, 22–27, 88, 92, 93, 95, 98, 99, 104, 105, 107–108, 114, 156, 224, 225, 228, 229, 402 Triazole..................................................................... 387, 389

S Saccharides..................................4, 6, 16, 17, 26, 70, 80, 119, 124, 125, 132, 133, 195, 221, 222, 246, 338, 344, 354–356, 381, 393–399, 402, 413

V Virus fluorescent labelling........................................ 252–254