Synthetic Multivalent Molecules: Concepts and Biomedical Applications

SYNTHETIC MULTIVALENT MOLECULES SYNTHETIC MULTIVALENT MOLECULES Concepts and Biomedical Applications SEOK-KI CHOI A ...

Author: Seok-Ki Choi

40 downloads 1361 Views 6MB Size Report

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

DOWNLOAD PDF

SYNTHETIC MULTIVALENT MOLECULES

SYNTHETIC MULTIVALENT MOLECULES Concepts and Biomedical Applications

SEOK-KI CHOI

A JOHN WILEY & SONS, INC., PUBLICATION

Copyright 6 2004 by John Wiley & Sons, Inc. All rights reserved. Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, 978-646-8600, fax 978-750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best e¤orts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services please contact our Customer Care Department within the U.S. at 877-762-2974, outside the U.S. at 317-572-3993 or fax 317-5724002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print, however, may not be available in electronic format. Library of Congress Cataloging-in-Publication Data: Choi, Seok-ki, 1964– Synthetic multivalent molecules : concepts and biomedical applications / Seok-ki Choi. p. cm. Includes index. ISBN 0-471-56347-1 (cloth) 1. Multivalent molecules—Synthesis. 2. Multivalent molecules—Design. 3. Multivalent molecules—Physiological e¤ect. I. Title. RS431.M84 C48 2004 615 0 .31—dc22 2003020808 Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

CONTENTS

Preface

xix

Notes for Organization and Classification

xxi

Abbreviations 1

Introduction

xxiii 1

1.1 Nomenclature and Definitions, 1 1.1.1 Valency, 1 1.1.2 Linkers, 3 1.1.3 Sca¤olds, 4 1.1.4 Ligand Density, 7 1.1.5 Homo- and Heterovalent Molecules, 7 1.2 Mechanistic Aspects of Multivalent Interaction, 9 1.2.1 A‰nity Constant and Avidity, 9 1.2.2 Thermodynamics, 12 1.2.3 Kinetics, 18 1.2.4 Steric E¤ects, 19 1.3 Biological Roles of Multivalent Ligands, 20 2

Multivalent Molecules Applied to Viral Targets

23

2.1 Influenza Virus, 23 2.1.1 Hemagglutinin, 23 2.1.1.1 Divalent Sialic Acid, 23 2.1.1.2 Tetravalent Sialoside, 26 2.1.1.3 Dendrimers Presenting Sialosides, 26 v

vi

CONTENTS

2.1.1.4 2.1.1.5

2.2

2.3 2.4 2.5

Sialic Acid Displayed in Liposomes, 29 Polymerized Liposome Presenting Sialic Acid, 32 2.1.1.6 Sialic Acid in Langmuir–Blodget Monolayers, 35 2.1.1.7 Sialic Acid Presented on Biopolymer Surfaces, 36 2.1.1.8 Poly(acrylamide) Presenting Sialosides, 36 2.1.1.9 Poly(acrylic acid) Presenting Sialosides, 37 2.1.1.10 Poly(acrylamide) Presenting C-Sialosides, 37 2.1.1.11 Postmodification of Activated Polymers, 38 2.1.1.12 Modes of Action of Polymeric Sialosides, 39 2.1.1.13 OPTCOL Assay, 40 2.1.1.14 ter-Poly(acrylic acid) Presenting Sialosides, 41 2.1.1.15 Poly(glutamic acid) Presenting Lysogangliosides, 42 2.1.1.16 Poly(glutamic acid) Bearing Sialic Acid– Containing Trisaccharides, 42 2.1.1.17 Poly(acrylamide) Bearing Sialic Acid Linked at the C4 Position, 44 2.1.1.18 Poly(acrylamide) Tethering 9-OAcetylsialosides, 45 2.1.1.19 Neoglycoprotein Displaying Sialic Acids, 45 2.1.1.20 Natural b-Inhibitors, 47 2.1.2 Neuraminidase, 47 2.1.2.1 Poly(glutamic acid) Presenting NA Inhibitors, 48 Human Immunodeficiency Virus, 49 2.2.1 HIV-1 Protease, 49 2.2.1.1 Divalent Terminal Peptides, 50 2.2.2 HIV-1 Reverse Transcriptase, 51 2.2.2.1 Heterodimers Composed of NRTI and NNRTI, 51 2.2.3 Glycoprotein 120 (gp120) on Viral Surfaces, 53 2.2.3.1 Neoglycoprotein-Displaying CD4 Peptide, 55 2.2.3.2 Galactosyl Ceramide Immobilized on Viral Surfaces, 55 2.2.3.3 Multivalent Anions, 57 2.2.3.4 Bivalent Antagonists of CXCR4, 59 2.2.4 Surface Carbohydrates on HIV, 60 2.2.4.1 Oligomannose Sugars Present on gp120, 60 2.2.4.2 Modes of CVN Recognition, 62 Rotavirus, 63 Polyoma Virus, 64 Picorna Virus, 65

CONTENTS

vii

2.6 Respiratory Syncytial Virus, 66 2.7 Dengue Virus, 67 2.8 Nucleic Acids of Viruses, 68 2.8.1 RNA–Protein Interactions, 68 2.8.1.1 Neomycin B Linked to Acridine, 69 2.8.1.2 Neamine Linked to Pyrene–Carboxylic Acid, 69 2.8.2 RNA–Enzyme Interactions, 70 2.8.3 Binders to the Minor Groove of Viral DNA, 71 2.8.3.1 Hairpin-Shaped Polyamide Dimers, 71 2.8.3.2 H-Shaped Polyamide Dimers, 73 2.8.3.3 Naturally Occurring Dimers, 74 2.9 Synthetic Multivalent Vaccines, 77 2.9.1 Peptide-Based Anti-influenza Vaccines, 77 2.9.2 Gp41-Based Anti-HIV Vaccine, 78 2.9.3 Peptide-Based Anti-FMDV Vaccines, 78 3

Multivalent Molecules Applied to Bacterial Targets 3.1 Targets in Bacterial Cell Membranes, 79 3.1.1 d-Ala-d-Ala Peptide Precursors, 79 3.1.1.1 Mode of Action by Antibiotics of the Vancomycin Class, 79 3.1.1.2 Dimerization of Glycopeptide Antibiotics, 82 3.1.1.3 Di- and Trivalent Vancomycin, 84 3.1.1.4 Cyclic Dimers of Eremomycin, 96 3.1.1.5 Polyvalent Vancomycin Constructs, 97 3.1.1.6 Heterobivalent Vancomycin Constructs, 99 3.1.1.7 Multivalent Vancomycin Mimics, 103 3.1.1.8 Catalytic Vancomycin Mimics, 104 3.1.1.9 Presentation of Glycopeptide Molecules on Bacterial Surfaces, 105 3.1.2 Lipid A, 108 3.1.2.1 Cholate-Based Polymyxin Mimics, 108 3.1.2.2 ter-Cyclopentane–Based Polymyxin Mimics, 109 3.1.2.3 Natural Multivalent Molecules Bound to Lipid A, 110 3.1.3 Helicobacter pylori Adhesin Receptor, 112 3.1.3.1 Synthetic Multivalent Inhibitors of H. pylori, 112 3.1.4 Streptococcus suis Hemagglutinin, 117 3.1.5 Escherichia coli FimH Adhesin, 117 3.1.5.1 Physical Forces Involved in Bacterium–Cell Adhesion, 118

79

viii

CONTENTS

3.1.5.2 Polyvalent Inhibitors of E. coli Adhesin, 119 3.1.5.3 Presentation of Mannosides on Gold Nanoparticles, 119 3.1.6 E. coli Chemoreceptors, 120 3.1.6.1 Polyvalent Galactosides as an E. coli Chemoattractant, 121 3.1.6.2 Polyvalent Glucosides in the Chemotaxis of Gram-Positive Bacteria, 124 3.1.7 Nonbacterial Lectins, 124 3.1.8 Staphylococcus aureus E¿ux Pumps, 125 3.2 Bacterial Toxins, 126 3.2.1 Heat-Labile Enterotoxins, 126 3.2.1.1 Pentavalent Galactosides, 127 3.2.1.2 Multivalent Inhibitors Displayed on a Solid Support, 128 3.2.2 Cholera Toxin, 129 3.2.2.1 Galactoside-Presenting Pentavalent Inhibitors, 130 3.2.2.2 Mode of Action of Pentavalent Galactosides, 133 3.2.2.3 Dendrimer-Based Multivalent Lactosides, 133 3.2.2.4 Polymerized Liposomes Incorporating GM1 , 133 3.2.2.5 Liposomes Presenting Fluorescent GM1 , 136 3.2.3 Shigalike Toxin, 136 3.2.3.1 Bivalent Pk Trisaccharides, 136 3.2.3.2 Decavalent Pk Ligands, 137 3.2.3.3 Polyvalent Polymeric Pk Ligands, 139 3.2.3.4 Multivalent Presentation of P1 on a Solid Support, 140 3.2.4 Verotoxin, 141 3.2.5 Bacillus anthracis Anthrax Toxin, 141 3.2.5.1 Polyvalent Peptide Inhibitors, 142 3.3 Bacterial Enzymes, 143 3.3.1 Transpeptidases, 143 3.3.1.1 Divalent Substrates, 143 3.3.1.2 Multivalent b-Lactams, 143 3.3.2 E. coli Pyrophosphokinase, 146 3.3.3 Protein Tyrosine Phosphatase, 147 3.3.4 DNA Gyrase, 147 3.3.5 Dihydrofolate Reductase, 148 3.3.6 Sialidase (Neuraminidase), 150 3.3.7 NAD Synthetase, 151 3.4 Bacterial Nucleic Acids, 152 3.4.1 Ribosomal RNA, 152

CONTENTS

ix

3.4.1.1 Divalent Aminoglycosides, 152 3.4.1.2 Divalent Aminoglycoside Bound to a Dimerized A-Site, 153 3.4.1.3 Polyvalent Polymeric Amikacin Derivatives, 154 3.5 Multivalent Molecules as Synthetic Vaccines, 155 3.5.1 Synthetic Shigella Vaccines, 155 3.5.2 Synthetic Mycobacterium Vaccines, 155 3.5.3 Carbohydrate Vaccines against Leishmaniasis, 156 3.5.4 Capsular Polysialic Acid Presented on Liposomes, 157 3.5.5 Multivalent Carbohydrate Antigens from Streptococcus, 158 3.6 Fungal Cells, 160 3.6.1 Covalent Dimer of Amphotericin B, 160 4

Multivalent Molecules Applied to Cellular Targets 4.1 Carbohydrate-Recognition Receptors on Cell Surfaces, 162 4.1.1 Hepatic Mannose/N-Acetylgalactosamine-4-SO4 Receptors, 162 4.1.1.1 Glycoprotein-Based Multivalent Ligands, 162 4.1.2 Hepatic Asialoglycoprotein Receptors, 164 4.1.2.1 Triantennary Galactosides, 164 4.1.2.2 Lysine-Based Multivalent NAcetylgalactosides, 164 4.1.2.3 Trivalent Galactoside–Cholesterol Conjugates, 165 4.1.2.4 Trivalent Galactosides for Cell-Specific DNA Delivery, 168 4.1.2.5 Multivalent Binding and Intracellular Tra‰cking, 169 4.1.3 Mannose-Binding Proteins, 169 4.1.3.1 Bivalent Mannosides, 171 4.1.3.2 Mannose-Presenting Neoglycoproteins, 172 4.1.3.3 GlcNAc-Presenting Polymer, 172 4.1.3.4 Mannoside Mimics, 173 4.1.4 Mannose 6-Phosphate Receptor (Insulinlike Growth Factor II), 174 4.1.4.1 Neoglycoprotein-Based Mannose-6Phosphate, 175 4.1.4.2 Biotin/Avidin-Based Tetravalent Mannose6-Phospate, 176 4.1.5 Galectins, 177

162

x

CONTENTS

4.1.6 4.1.7

4.1.8 4.1.9

4.1.10

4.1.11 4.1.12

4.1.13

4.1.14

4.1.15

4.1.5.1 Glycodendrimer, 177 Sialoadhesin CD22, 179 Dendritic Cell–Specific ICAM-3–Grabbing Nonintegrin, 180 4.1.7.1 Mannoside Clusters on Peptide and Protein Sca¤olds, 180 4.1.7.2 Lewis Saccharide–Presenting Dendrimers, 181 Spermatogenesis, 181 Sperm–Egg Interactions, 182 4.1.9.1 Bindin, 182 4.1.9.2 b-1,4-Galactosyltransferase, 183 4.1.9.3 Fertilin b, 183 4.1.9.4 XL35 Lectin, 183 Fucose-Binding Receptor on Cancer Cells, 184 4.1.10.1 Fucose–Batracylin Conjugate Based on Neoglycoprotein, 184 4.1.10.2 Liposomes That Display GM3 –lac Ceramide, 185 Fibroblast Growth Factors, 186 Carbohydrate–Carbohydrate Interactions at Cell Surfaces, 186 4.1.12.1 Gg3 –GM3 Interaction, 186 4.1.12.2 Aggregation Factors of Marine Sponges, 187 4.1.12.3 Le x –Le x Interaction, 188 E-Selectin, 188 4.1.13.1 Divalent Sialyl Lewis x, 189 4.1.13.2 Trivalent Sialyl Lewis x Constructs, 189 4.1.13.3 Neoglycoproteins Displaying Sialyl Lewis x Ligands, 191 4.1.13.4 Liposomes Displaying Sialyl Lewis x, 192 4.1.13.5 Polyvalent Polymeric Sialyl Lewis x, 195 4.1.13.6 Sialyl Le x –Binding Multivalent Receptors, 198 L-Selectin, 198 4.1.14.1 Divalent Sulfated Lactoside, 200 4.1.14.2 Neoglycopolymers Presenting Sialyl Le x , 201 4.1.14.3 Poly(acrylamide) Sca¤olds That Present Sulfated Le x , 203 4.1.14.4 Poly(lactosamine) Sca¤olds That Present SLe x , 203 P-Selectin, 204 4.1.15.1 Sialyl Lewis x-RGD Conjugates, 205

CONTENTS

Copolymers Presenting Sialyl Le a Residues and Sulfated Tyrosines, 205 4.1.15.3 Neoglycopolymers Presenting Sulfated Galactosides, 206 4.1.15.4 Polymerized Liposomes Presenting Mimetics of Sialyl Lewis x, 206 4.1.16 Lectins, 209 4.1.16.1 Oligovalent Mannosides Binding to Con A, 209 4.1.16.2 Trivalent Mannosides on a Rigid Sca¤old, 209 4.1.16.3 Bifunctional Multivalent Glycosides Binding to Con A, 211 4.1.16.4 Heptavalent Mannosides Linked to bCyclodextrin, 211 4.1.16.5 Mannosides Displayed on a Benzene- or Peptide-Based Dendrimer, 211 4.1.16.6 Poly(oxanorbornene) Presenting Glucosides, 215 4.1.16.7 Poly(norbornene) Derivatives Bearing Two Types of Ligands, 217 4.1.16.8 Glycopolymer-Mediated Cell–Cell Adhesion, 217 4.1.16.9 Mannosides Displayed on Self-Assembled Monolayers, 218 4.1.16.10 Multivalent Galactosides Binding to Bauhinia purpurea Lectin, 218 4.1.16.11 Dendritic Sialosides Binding to Limax flavus Lectin, 219 4.1.16.12 Dendritic GlcNAc Binding to Wheat Germ Agglutinin, 220 4.1.16.13 Cyclodextrin-Based Multivalent Glucosides Binding to Pisum sativum Lectin, 222 4.1.16.14 Oligovalent Mannosides Binding to Dioclea grandiflora Lectin, 223 4.1.16.15 Polymeric Polyvalent Lactosides Binding to Psophocarpus tetragonolobus Lectin, 225 4.1.16.16 Resin-Displayed Multivalent Galactosides Binding to RCA Lectin, 225 4.1.16.17 Liposome-Based Multivalent Galactosides Binding to RCA Lectin, 226 4.1.16.18 Multivalent Display of Saccharides on Microtiter Well Surfaces, 227 4.2 Peptide and Hormone Recognition Receptors on Cell Surfaces, 228 4.1.15.2

xi

xii

CONTENTS

4.2.1

CD14 Glycoprotein, 228 4.2.1.1 Poly(acrylamide) Displaying Muramyl Dipeptides, 228 4.2.2 Major Histocompatibility Complex, 229 4.2.2.1 MHC Peptides Tethered to a Peptide Sca¤old, 229 4.2.3 Integrins, 231 4.2.3.1 Poly(norbornene) Presenting RGD and PHSRN Sequences, 231 4.2.3.2 GRGDS Displayed on Polystyrene Beads, 232 4.2.3.3 RGD Peptides Presented on SAMs, 233 4.2.3.4 Polymeric Nanoparticles Displaying av b3 Antagonist, 233 4.2.3.5 Divalent Antagonist of Integrin av b3 , 234 4.2.4 Immunoglobulin, 235 4.2.4.1 Pentavalent Peptides, 236 4.2.5 Estrogen Receptors, 236 4.3 Ligand-Mediated Receptor Dimerization, 236 4.3.1 Chemical Inducers of Dimerization, 236 4.3.1.1 Natural Heterodimerizers, 237 4.3.1.2 FK506 Dimer (FK1012), 238 4.3.1.3 FK506 Mimetics and CIDs, 239 4.3.1.4 Cyclosporin Dimer, 240 4.3.1.5 Heterodimeric CIDs Composed of FK506 and Cyclosporin, 241 4.3.1.6 Heterodimeric CIDs Composed of FK506 and SH2 Ligands, 241 4.3.1.7 Bivalent Ligands Binding to FKBP, 245 4.3.1.8 Rapamycin-Mediated Association of Inteins, 245 4.3.2 Cytokine Receptors, 246 4.3.2.1 Divalent EPO Mimetics, 247 4.3.2.2 Divalent Peptide Antagonists of IL-5, 247 4.3.2.3 Dimeric Mimetics of G-CSF, 247 4.3.2.4 Dimeric Mimetics of NT-3, 249 4.3.3 Vitamin D3 Receptors, 250 4.4 Enzymes, 250 4.4.1 Acetylcholinesterase, 250 4.4.1.1 Tetrahydroacridine Dimer, 250 4.4.1.2 Huperzine Dimer, 252 4.4.1.3 Surface Anchoring, 252 4.4.2 Matrix Metalloproteinases, 252 4.4.3 Tryptase, 253

CONTENTS

4.4.3.1

Bivalent Inhibitors Tethered to Cyclodextrin, 254 4.4.3.2 Bivalent Inhibitors Tethered to Diketopiperazine, 255 4.4.3.3 Benzamidine-Based Divalent Inhibitors, 255 4.4.4 Proteasomes, 256 4.4.4.1 Multivalent Nature of Proteasomes, 256 4.4.4.2 PEG-Linked Tripeptide Dimers, 257 4.4.5 Protein Kinase C, 258 4.4.5.1 Divalent Nature of Protein Kinase C, 258 4.4.5.2 Phorbol Dimers, 258 4.4.6 Syk Protein Tyrosine Kinase, 259 4.4.7 Protein Tyrosine Phosphatases, 259 4.4.7.1 Phosphophenol Dimers, 259 4.4.7.2 Divalent Tetronic Acids, 261 4.4.8 Human Leukocyte Elastase, 261 4.4.9 Thrombin, 261 4.4.10 Glucoamylase, 262 4.4.10.1 Acarbose Linked to b-Cyclodextrin, 262 4.4.11 Glycosyltransferase, 264 4.4.11.1 Divalent Inhibitors of a1,3-Galactosyltransferase, 264 4.4.11.2 Divalent Inhibitors of b1,4-Galactosyltransferase, 264 4.4.11.3 Divalent Inhibitors of a1,2-Fucosyltransferase, 266 4.4.11.4 Divalent Inhibitors of Sialyltransferase, 266 4.4.12 Riboflavin Synthase, 267 4.4.12.1 Lumazine Dimers, 267 4.4.13 Lactate Dehydrogenase, 267 4.4.13.1 Enzyme Oligomerization, 267 4.4.13.2 Cyclodextrin Dimers, 268 4.4.14 Zinc-Dependent Enzymes, 269 4.4.14.1 Histone Deacetylase, 269 4.4.14.2 Carbonic Anhydrase, 270 4.5 G-Protein-Coupled Receptors, 270 4.5.1 b-Adrenergic Receptors, 270 4.5.1.1 Divalent Ligands of b-AR, 271 4.5.1.2 Isoproterenol Presented on Polypeptides, 272 4.5.1.3 Practolol Dimers, 272 4.5.1.4 Propranolol Dimers, 272 4.5.1.5 Alprenolol Displayed on Dextran, 273 4.5.2 a-Adrenergic Receptors, 274

xiii

xiv

CONTENTS

4.5.2.1 Prazocin Dimers, 274 4.5.2.2 Yohimbine Dimers, 275 4.5.3 Opioid Receptors, 275 4.5.3.1 Oxymorphamine Dimers, 276 4.5.3.2 Bivalent Bridging, 276 4.5.4 Serotonin Receptors, 277 4.5.4.1 Carboxamidoindole Dimers, 278 4.5.4.2 Sumatriptan Dimer, 278 4.5.4.3 Divalent SERT Inhibitors, 278 4.5.4.4 Serotonin Displayed on Nanocrystals, 280 4.5.5 Muscarinic Acetylcholine Receptors, 280 4.5.5.1 Quinnuclidine Dimers, 280 4.5.6 Melanocortin-1 Receptor, 281 4.5.6.1 Divalent a-MSH Peptides, 281 4.6 Ion Channels, 282 4.6.1 Cyclic Nucleotide-Gated Channels, 282 4.6.1.1 Cyclic Nucleotides, 282 4.6.1.2 PEG-Linked cGMP Dimers, 282 4.6.2 IP3 -Gated Calcium Channels, 283 4.6.2.1 PEG-Linked IP3 Dimers, 283 4.6.3 Potassium Channels, 284 4.6.3.1 Bisquinolinium, 284 4.6.4 Nicotinic Acetylcholine Receptor, 285 4.6.4.1 Succinyldicholine, 285 4.7 Nucleic Acids, 286 4.7.1 DNA Intercalators, 286 4.7.1.1 Antraquinone Linked to an Aminosugar, 286 4.7.1.2 Daunorubicin Dimer, 287 4.7.1.3 Bisacridine, 288 4.7.1.4 Naphthalimide Dimers, 290 4.7.2 Minor Groove Binders, 291 4.7.2.1 Calicheamicin-Based Dimers, 291 4.7.2.2 Mitomycin Dimers, 292 4.7.2.3 Cyclopropa[c]pyrrolo [3,2-e]indole Dimers, 293 4.7.2.4 Pyrrolo[2,1-c][1,4]benzodiazepine Dimers, 294 4.7.2.5 PBD-Containing Heterodimers, 295 4.7.2.6 DNA-Binding Polyamides, 295 4.7.2.7 H-Shaped Polyamide Dimers, 296 4.7.2.8 Hairpin-Shaped Polyamide Dimers, 297 4.7.2.9 Hairpin-Shaped Extended Polyamides, 299 4.7.2.10 Ring-Shaped Polyamide Dimers, 299 4.7.2.11 Tetravalent Polyamides, 299

CONTENTS

xv

4.7.2.12

Polyamide–DNA Alkylator Conjugates, 299 4.7.2.13 Polyamides Linked to Camptothecin, 300 4.7.2.14 Polyamides Linked to Transcription Activators, 301 4.7.3 Divalent Recognition at Abasic Sites, 303 4.7.3.1 Adenine Tethered to Acridine, 304 4.7.4 Divalent Binding to Guanine–Guanine Pairs, 304 4.7.4.1 Naphthyridine Dimers, 304 4.7.5 Basic Leucine Zipper Proteins, 304 4.7.5.1 bZIP-Derived Peptide Dimers, 304 4.7.5.2 Divalent bZIP Moieties Linked to Chiral Templates, 305 4.7.5.3 Divalent bZIP Moieties Tethered to an Azobenzene Sca¤old, 305 4.8 Antibodies, 306 4.8.1 Anticancer Vaccines, 306 4.8.1.1 Tumor Cell Marker, 306 4.8.1.2 Globo-H Presented on Keyhole Limpet Hemocyanin, 307 4.8.1.3 GM3 Presented on Bovine Serum Albumin, 309 4.8.1.4 GM2 Presented on KLH, 310 4.8.1.5 TN Clusters Presented on KLH, 310 4.8.1.6 TF -Presenting Dendrimers, 312 4.8.1.7 Neoglycoproteins Presenting Multiple Tumor-Associated Antigens, 312 4.8.2 Surface Epitopes Not Related to Cancer, 313 4.8.3 Miscellaneous Epitopes, 314 4.9 Cell Surface Engineering, 315 4.9.1 Metabolic Incorporation of Unnatural Carbohydrate Antigens, 315 4.9.1.1 N-Butanoylsialic Acid, 316 4.9.1.2 N-Levulinoyl Neuraminic Acid, 316 4.9.1.3 Azidoacetyl Neuraminic Acid, 319 4.9.2 (Bio)chemical Modification of Cell Surface Antigens, 320 5

Chemical Synthesis of Multivalent Molecules 5.1 Selected Synthetic Methods for Multimerization, 321 5.1.1 Summary of Linker Chemistry, 321 5.1.2 Polymerization, 323 5.1.3 Self-Assembly, 325 5.1.4 Solid-Phase Synthesis, 325

321

xvi

CONTENTS

5.2 Combinatorial Chemistry, 325 5.2.1 Structure-Biased Combinatorial Synthesis, 325 5.2.1.1 Hetrodimeric Inhibitors of Picorna Virus, 325 5.2.1.2 Polyvalent Inhibitors of Influenza Virus, 327 5.2.1.3 Heterodimeric Inhibitors of NAD Synthetase, 328 5.2.1.4 Disulfide-Linked Heterodimers, 328 5.2.2 Target-Guided Combinatorial Synthesis, 330 5.2.2.1 FK-506 Binding Protein, 330 5.2.2.2 Caspase, 331 5.2.2.3 c-Src Kinase, 332 5.2.2.4 Acetylcholine Esterase, 333 5.2.2.5 Carbonic Anhydrase, 334 5.2.2.6 Kallikrein, 334 5.2.3 Dynamic Combinatorial Synthesis, 334 5.2.3.1 Disulfide Exchange, 336 5.2.3.2 Olefin Metathesis, 336 5.2.3.3 Transition Metal Complexes, 337 Appendix Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. Table 12. Table 13. Table 14.

339 Divalent Ligands Linked to Enzymes in Human Immunodeficiency Virus (HIV), 339 Multivalent Ligands Linked to gp120 Surface Receptors in Human Immunodeficiency Virus, 340 Multivalent Ligands Linked to Surface Receptors on Viruses, 341 Multivalent Receptors Targeting Multivalent Ligands in Bacterial Cell Walls, 343 Multivalent Ligands Targeting Bacterial Receptors in Membrane Surfaces, 344 Multivalent Ligands Linked to Bacterial Toxins, 345 Multivalent Molecules Targeting Bacterial Enzymes, 346 Multivalent Ligands Targeting Carbohydrate-Recognition Receptors on Cellular Surfaces, 347 Multivalent Ligands Targeting NoncarbohydrateRecognition Receptors on Cellular Surfaces, 350 Multivalent Ligands Linked to Selectins on Cell Surfaces, 352 Multivalent Ligands Linked to Lectins, 354 Multivalent Ligands Linked to Cellular Enzymes, 357 Multivalent Ligands Linked to G-Protein-Coupled Receptors (GPCRs), 359 Multivalent Ion Channel–Binding Molecules, 360

CONTENTS

xvii

Table 15. Homo- and Heterodivalent Chemical Inducers of Dimerization (CIDs), 361 Table 16. Multivalent Ligands Linked to Nucleic Acids, 362 Table 17. Synthetic Multivalent Antigens, 365 Table 18. Multivalent Ligands Displayed on Self-Assembled Monolayer (SAM) Made of Alkanethiolate on Gold, 367 Table 19. (Bio)chemical Modification of Cell Surface Antigens, 368 Table 20. Multivalent Targets, 369 References

371

Index

409

PREFACE

This book presents basic and advanced principles underlying the multivalent interactions that are prevalent in biological systems. To illustrate important or complex concepts, the book provides up-to-date examples of synthetic multivalent molecules, their design, and their biological benefits. Functional roles displayed by such molecules of both natural and synthetic origin are well documented in biology, where they exert unique and crucial activities at a level not readily achievable by monovalent molecules. The concept of multivalent design is now accepted as an e¤ective strategy—in particular, for designing ligands, inhibitors, and drugs that influence biological systems potently and selectively. Over the past 15 years, diverse diciplines have generated a growing interest in the biomedical application of multivalent design. The goal of this book is to share the findings in this exciting area of research by providing a systematic summary of experimentally tested case studies of multivalency. I believe that a single book serves best to collect such scattered research material in one place and to discuss it in a consistent and introductory format. The book focuses on practical examples of synthetic multivalent molecules reported broadly in the literature. It consists of five chapters. In Chapter 1 I introduce the multivalent molecule and its structural elements, describe the mechanistic basis that accounts for the benefits of multivalent interaction, and provide a short summary of biological functions displayed by multivalent molecules. In Chapters 2 to 4 I provide practical examples derived from biological targets in viral, bacterial, and mammalian cells, respectively. In each chapter I review in a similar format the design concept, synthesis, and biological activity of multivalent molecules: in particular, those of synthetic origin. xix

xx

PREFACE

Typically, the chapter format begins with a brief description of a target from a structural and functional viewpoint to provide a rationale for multivalent design, followed by the main discussion. In Chapter 5 I summarize various aspects of synthetic methods used in the synthesis of multivalent molecules, and I conclude with a summary of combinatorial approaches developed in the library design of multivalent molecules. The book ends with an appendix that presents tabular summaries of both examples treated in this book and untreated. The targets selected for discussion are comprehensive but, of course, do not purport to be a complete list. The book is written for a broad community of audiences, comprising educators, graduate students, and professional researchers in academia and the (bio)pharmaceutical industry, particularly those who perform interdisciplinary research in organic chemistry, chemical biology, biological chemistry, medicinal chemistry, pharmacology, and medicine. Specifically, it would be most valuable as a reference book for those scientists interested in finding new ideas and developments in areas of receptor–ligand interaction, carbohydrate-based medicines, enzyme inhibitors, toxin inhibitors, DNA(RNA)–drug association, antibiotics, antiviral agents, anti-inflammatory drugs, and anticancer therapeutics. It is my pleasure to express great gratitude to Professor Koji Nakanishi of Columbia University, who introduced me to the importance of multidisciplinary bioorganic studies and encouraged this publication. I am also indebted to Professor George M. Whitesides of Harvard University and sincerely appreciate his directing me to continue research in multivalency. In 1996, Professor Whitesides and several colleagues, including Professor John Gri‰n of Stanford University, Dr. Mathai Mammen of Harvard, and James Tananbaum of Sierra Ventures, founded Advanced Medicine, Inc. to develop opportunities in multivalent drug research and development. I joined Advanced Medicine in 1997 to help further develop ideas in multivalent drug design. Today, Advanced Medicine has evolved into a vibrant pharmaceutical company, Theravance, Inc., under the guidance of our Chairman, Roy Vagelos (former CEO of Merck Pharmaceuticals), and current CEO, Rick Winningham. I am indebted to my colleagues at Theravance for ongoing collaborations in medicinal chemistry. In particular, I am sincerely thankful to Dr. Ed Moran, Dr. Thomas Jenkins, and Dr. Mathai Mammen for their editorial assistance with this manuscript. I am also very grateful to Professor Fred Brewer at Albert Einstein College of Medicine and Dr. Obadiah Plante at Ancora Pharmaceuticals for their valuable comments during the review process. Publication of this book would not have been possible without the support and patience of Hyun-Joo, Gilbert, and other family members. Finally, I wish to thank the editors and sta¤ members at John Wiley & Sons, particularly Jonathan T. Rose, for their e¤orts to complete this project smoothly and in a timely manner. Palo Alto, California

Seok-Ki Choi

NOTES FOR ORGANIZATION AND CLASSIFICATION

Given diverse aspects of classification related to multivalent ligands and inhibitors, I wish to provide a road map to the structure and organization of this book. Its main contents consist of three chapters that are divided according to the origin of targets: viruses (2), bacteria (3), or mammalian cells (4). Within this division, materials are grouped further on the basis of target class and structural composition of multivalent molecules. Therefore, typical headings of each chapter start with an organism (e.g., influenza virus), proceed to a description of a target under a certain class (receptor: hemagglutinin), followed by a discussion of a group of multivalent molecules (e.g., divalent, trivalent, polyvalent, . . . , sialic acid). Under such a format, multivalent ligands from the same class can be reviewed collectively and comparatively as to their design concept, synthesis, and biological activity. A short outline of each chapter is given below. Chapter 2 covers multivalent examples from viral targets such as influenza virus and the human immunodeficiency virus (HIV). Several classes of multivalent inhibitors targeting viral surface proteins (influenza hemagglutinin), enzymes (influenza neuraminidase, HIV-protease), and nucleic acids (as intercalators and as binders at minor or minor groove) are presented. The chapter ends with a summary of synthetic multivalent antigens that are designed to mimic the surface of influenza, HIV, or foot-and-mouth disease virus. Chapter 3 focuses on bacterial cells, a pathogenic system that proves to be well validated for a multivalent approach. This chapter provides ample examples of multivalent inhibitors that display enhanced activity against a wide range of receptors or ligands expressed on a bacterial surface (d-ala-d-ala, lipid A, adhesins), toxins (cholera, anthrax), enzymes (transpeptidase), and nucleic acids (ribosomal RNA). xxi

xxii

NOTES FOR ORGANIZATION AND CLASSIFICATION

Chapter 4 treats multivalent molecules that inhibit or modulate mammalian cells. The list of tested macromolecular targets is extensive. Selected examples include surface receptors (hepatic asialoglycoprotein receptor, selectins), enzymes (acetylcholine esterase), G-protein-coupled receptors (opioid receptor), and ion channels (cyclicnucleotide-gated channel). Applications in nucleic acid are demonstrated by a daunorubicin dimer acting as a DNA intercalator, and a hairpin-shaped polyamide dimer acting as a minor groove binder. In addition, several types of synthetic vaccines are illustrated, including multivalent globoH presented on protein carrier as the one mimicking cancer cell surface. Discussion of such diverse cellular targets should help to better understand multivalent concepts and applications in drug development. Finally, the organization of the book aims for easy access to and concise comparison of multivalent molecules for a given receptor. However, readers might be interested in looking at the features of multivalent molecules organized from di¤erent aspects regardless of their target families, functions, or cellular origins. In a sense, it might be useful to compare multivalent design methods within a broadly defined boundary based on target location (membrane-bound versus soluble) and target valency (divalent, trivalent, . . . , multivalent). The nature of assembly of multivalent sites is considered to be important as well, such that targeting to a single-subunit multivalent receptor can be distinguished from a multisubunit receptor complex composed of either identical subunits (homo) or di¤erent subunits (hetero). To complement such interesting features, a table (Table 20) is given in the appendix as a cross-reference to present such information for major targets: for example, whether they are membrane-bound, soluble, homo-trivalent, or hetero-divalent. Readers are advised to utilize these additional aspects of classification when selecting materials.

ABBREVIATIONS

ACE AChE Ala AmB b-AR ASGP-R AT-III AZT BChE BSA bZIP CA-II Caspase-3 CD Ce¤ cGMP CID CNG Con A CPMV CRD CsA

a‰nity capillary electrophoresis acetylcholine esterase alanine amphotericin B b-adrenergic receptor asialoglycoprotein receptor antithrombin III azidothymidine butyrylcholinesterase bovine serum albumin basic leucine zipper protein carbonic anhydrase II cysteine aspartyl protease-3 cyclodextrin e¤ective local concentration cyclic guanosine monophosphate chemical inducer of dimerization cyclic nucleotide-gated (CNG) concanavalin A cowpea mosaic virus carbohydrate-recognition domain cyclosporin A xxiii

xxiv

ABBREVIATIONS

CTB CVN DC-SIGN DHFR DLS EC EGFR ELAM-1 ELISA EPO-R ERK FGF FKBP FMDV FRAP Gal Gal cer GalNAc GalTase G-CSF GGBP Glc Glc cer GlcNAc GlyCAM-1 gp120 GPCR GPI HA HAI HIV HLE HMPA HPPK HT IC50 ICAM-3 IgG

cholera toxin B subunit cyanovirin-N dendritic cell–specific ICAM-3 grabbing nonintegrin dihydrofolate reductase dynamic light scattering e¤ective concentration epidermal growth factor receptor endothelial leukocyte adhesion molecule-1 enzyme-linked immunosorbent assay erythropoietin receptor extracellular regulated kinase fibroblast growth factor FK506-binding protein foot-and-mouth disease virus FKBP-12-rapamycin-associated protein galactose galactosylceramide N-acetylgalactoside galactosyltransferase granulocyte-colony-stimulating factor glucose–galactose binding protein glucose glucosylceramide N-acetylglucosamine glycosylated cell adhesion molecule-1 glycoprotein 120 G-protein-coupled receptor glycosylphosphatidylinositol hemagglutinin hemagglutination inhibition human immunodeficiency virus human leukocyte elastase N-(2-hydroxypropyl)polymetharylamide 6-hydroxymethyl-7,8-dihydropterin pyrophosphokinase 5-hydroxytryptamine or serotonin concentration at 50% inhibition intracellular adhesion molecule-3 immunoglobulin G

ABBREVIATIONS

IgM IL-5 IP3 ITC JCP Ka Kd Ki K HAI i Km ko¤ kon KLH Lac cer L–B LDL Leb Lex LPS mAChR Man MBP MDR MHC MIC MMP a-MSH MTX NA nACh-R NAD NeuAc NNRTI NRTI NT-3 PA pA pAA

immunoglobulin M interleukin-5 inositol 1,4,5-triphosphate isothermal titration calorimetry jelly coat glycoprotein association constant dissociation constant inhibition constant hemagglutination inhibition constant Michaelis constant rate of dissociation rate of association keyhole limpet hemocyanin lactosylceramide Langmuir–Blodgett low-density lipoprotein Lewis b Lewis x lipopolysaccharide muscarinic acetylcholine receptor mannose mannose-binding protein multidrug resistance major histocompatibility complex receptor minimal inhibition concentration matrix metalloproteinase a-melanocyte-stimulating hormone methotrexate neuraminidase nicotinic acetylcholine receptor nicotine adenine dinucleotide N-acetylneuraminic acid (sialic acid) non-nucleoside-based reverse transcriptase inhibitor nucleoside-based reverse transcriptase inhibitor neurotrophin 3 protective antigen poly(acrylamide) poly(acrylic acid)

xxv

xxvi

ABBREVIATIONS

PAMAM PBP PEG PKC PNA pNAS PSA PSGL-1 PTPase RCA RGD RNase S ROMP RRE RSV RT SA SAM SIV SLea SLex SLT SNP SPR SSRI TAR VHR VP VRE VSE WGA ZP

poly(amide amine) penicillin-binding protein poly(ethylene glycol) protein kinase C peptide nucleic acid poly(N-acryloyloxysuccinimide) poly-a2,8-sialic acid P-selectin glycoprotein ligand-1 protein tyrosine phosphatase Ricinus communis agglutinin Arg-Gly-Asp ribonuclease S ring-opening metathesis polymerization rev-response element respiratory syncytial virus reverse transcriptase sialic acid self-assembled monolayer simian immunodeficiency virus sialyl Lewis a sialyl Lewis x Shigalike toxin single-nucleotide polymorphism surface plasma resonance selective serotonin reuptake inhibitor transactivator-response RNA element vaccinia VH1–related phosphatase viral protein vancomycin-resistant enterococci vancomycin-susceptible enterococci wheat germ agglutinin zona pellulcida

1 INTRODUCTION

1.1 1.1.1

NOMENCLATURE AND DEFINITIONS Valency

According to Mammen et al. [1], the valency of a molecule, or that of a biological entity such as a cell, virus, or bacterium, represents the number of separate structural units of the same or a similar type that are connected to the molecule or entity. Thus, if a molecule presents two tethered, identical copies of binding elements, such as a ligand, it is classified as a divalent molecule (ligand). Schematic examples are provided in Figure 1.1 to describe the concept of valency in the context of receptor and ligand interaction. For example, divalent binding occurs when a divalent ligand associates with a divalent receptor through the simultaneous interaction established between two receptor–ligand pairs. Similarly, multivalent or polyvalent interactions are defined as specific simultaneous associations of multiple ligands present on a molecular construct or biological surface that bind to multiple receptors presented on a complementary entity. The classification of multivalent molecules presented above is simply based on the structural aspects of molecules, such as the number of structurally identical ligands per multivalent construct; hence, the term structural valency is used to classify such molecules. However, not all ligands present on a multivalent molecule are involved in interactions with a multivalent receptor displaying

Synthetic Multivalent Molecules: Concepts and Biomedical Applications, By Seok-Ki Choi ISBN 0-471-56347-1 Copyright 6 2004 John Wiley & Sons, Inc.

1

2

INTRODUCTION

monovalent ligand

monovalent receptor

divalent ligand

divalent receptor

divalent association

n multivalent receptor

multivalent association

n multivalent ligand

trivalent ligand (structural valency)

trivalent receptor

monovalent association

n

divalent association (functional valency)

Figure 1.1 Definitions of structural and functional valence as illustrated in monovalent, divalent, and multivalent (polyvalent) interaction.

multiple sites, and a certain portion of the tethered ligands are functionally inactive (Figure 1.1). Thus, for a particular multivalent receptor–ligand association, the valency of a multivalent ligand is not necessarily equivalent to its structural valency. This property is described by Dam et al. [2] as functional valency. The concept of functional valency is of significant interest; however, it has been explored in only a limited number of studies. In this book the valency of a multivalent molecule normally refers to structural valency unless its functional valency is available otherwise. Multivalent interactions are now understood to be a ubiquitous strategy that has evolved in nature for a wide range of functions, including selective recognition of multivalent antigens by antibodies [3] (e.g., bivalent anti-DNP IgG [4], decavalent IgM, bivalent anticardiolipin antibody binding to b2 glycoprotein I lipoprotein [5]), neutrophil adhesion and rolling on the surface of an activated endothelial cell [6], and the tight adsorption of a virus particle or bacterium to a host cell surface [7–9] (Figure 1.2). These multivalent interactions are more potent and selective over the analogous monovalent interaction and are therefore only weakly inhibited by most monovalent ligands, especially when the binding cleft for the monovalent ligand is shallow. When designing inhibitor molecules to interfere with multivalent interactions, the most e¤ective strategy is to use multivalent molecules. Such multivalent molecules proved to be highly potent inhibitors: in particular, against surface– surface interaction as observed in virus–cell and cell–cell adhesion (Figure 1.2) [10,11].

NOMENCLATURE AND DEFINITIONS

3

IgG

antigenic surface virus (bacterium)

multivalent ligand

multivalent ligand

cell surface

cell multivalent ligand

cell surface

Figure 1.2 Representative examples of natural multivalent interactions and their synthetic multivalent inhibitors.

1.1.2

Linkers

The synthesis of multivalent molecules is performed by tethering multiple copies of a ligand or binding element with a linker. A linker provides not only a covalent connection but also appropriate spacing between tethered ligands such that multiple receptor–ligand pairs interact simultaneously without being forced to mismatch. A linker used for a multivalent molecule has to be stable chemically, biochemically, and enzymatically. In contrast, many linkers used in bioconjugate molecules, such as those designed into prodrugs or delivery tools [12], must be chemically or enzymatically labile in order to provide the release of monovalent drugs at a certain stage or site after delivery into a biological system. In addition, the use of linker in prodrugs is not intended for modulation of binding activity of prodrugs such that covalently attached ligands from prodrugs may not bind to its receptor in either monovalent or multivalent fashion until its ligands are released free. The linker must be at least a neutral contributor to biological activity, so as not to interfere with the intrinsic activity of the tethered ligand. In some cases, the linker may even contribute favorably provided that productive contacts are made between the linker and the target surface.

4

INTRODUCTION

In this book, the terms linkers and spacers are used interchangeably. A linker needs to be designed taking a number of factors into account, such as linker length between connected ligands, conformational property (i.e., flexible, rigid), and the nature of linker functional groups (e.g., amide, ether, amine). As long as the linker is long enough to allow bivalent occupation of tethered ligands at receptor sites, a rigid linker contributes more favorably to tight association than does a flexible one, as predicted from considering the conformational entropy of linkers, examined by Mammen et al. [13]. This is because a large number of linker configurations existing in the unbound state are going to a single configuration in the bound state, a thermodynamic feature associated with an entropic cost. Therefore, the larger the initial number of configurations, the greater this cost. A very important assumption, however, is that the rigid linker orients the ligands such that multivalent binding is possible. With this assumption, an unsaturated or aryl-incorporated linker, for example, experiences a lower conformational entropic cost upon association than does a saturated or thioether-containing flexible linker. Amide- and ether-based linkers provide an intermediate flexibility. However, the entropic e¤ect is often accompanied by an enthalpic contribution to the free energy of binding since the linker itself is able to participate in interaction with the receptor, or to contribute via a hydrophobic e¤ect by favorable transfer from an aqueous medium to a hydrophobic receptor domain. The experimental verification of such linker variation is discussed in later chapters. 1.1.3

Sca¤olds

Covalent linkage is a term generally associated with a sca¤olding or framework that serves as a molecular anchoring system where multiple chemically reactive sites are expressed as handles for ligand attachment. The valency and shape of a sca¤old exert a significant influence on the binding and functional ability of multivalent molecules that comprise such a sca¤old. Numerous classes of scaffold are commonly used in the design of multivalent molecules, as shown in Figure 1.3. Molecules of low valency (e.g., di- and trivalent molecules) are designed using a one-dimensional linear or branched chain. Examples include oligo(glycine)-spaced divalent sialoside as an influenza inhibitor [14], alkanespaced divalent sLe x as an E-selectin inhibitor [15], and Tris-linked trivalent lactoside as a ligand to the asialoglycoprotein receptor [16]. In addition to linear sca¤olding, multivalent molecules may also be built on a rigid and preorganized sca¤old that is well defined in structure, orientation, and conformation. Such a rigid system presents a known number of preassembled attachment sites that enables one to achieve the proper positioning between attached ligands. Rigid sca¤olds are found in diverse classes of molecular systems, ranging from small molecules [17] such as benzene and glucose, to macrocycles such as azacrown ether, to round molecules such as cyclodextrin and calix[n]arene. For example, azacrown ether proves to be an e¤ec-

NOMENCLATURE AND DEFINITIONS one-dimensional linear scaffold

5

rigid scaffold

scaffold linker ligand glucose benzene

branched scaffold

calixarene azacrown ether cyclodextrin macrocycle transition metal complex

flexible chain scaffold

polymer peptoid

multiantennary dendrimer (tree-like)

two- and three-dimensional surface scaffold

globular protein dendrimer latex bead nanocrystal (Au, Cd/Se)

micelle liposome self-assembled monolayer (Au) surface (silica, polystyrene, carbon nanotube)

Figure 1.3 Shape and size of various sca¤olds used in multivalent presentation.

tive framework for the activity of galactoside-presenting pentavalent ligands, which bind cholera toxin B pentamer very tightly [18]. Calix[n]arene consists of four (n ¼ 4), six (n ¼ 6), or eight (n ¼ 8) phenol units joined to form a vaselike structure (calix ¼ ‘‘vase’’ in Greek). Another well-known rigid system includes cyclodextrin, a cyclic oligosaccharide composed of six, seven, or eight a-d-glucose units, thus forming a-, b-, and g-cyclodextrin, respectively [19]. Sca¤olds such as cyclodextrin and calix[n]arene o¤er multiple repeats of a hydroxyl functional group located around the circumference of the narrow (primary) or the wide (secondary) face. These hydroxyl groups are readily derivatized for tethering ligands such as those based on carbohydrates [20–24]. The application of such rigid sca¤olds is represented by a heptavalent galactoside anchored on the primary face of b-cyclodextrin [25] and a tetravalent sialoside displayed on calix[4]arene. The surfaces of naturally occurring glycoproteins present multiple types of

6

INTRODUCTION

complex carbohydrates clustered in di- and triantennary configurations [26]. Such branched sca¤olds resemble a treelike structure that is classified as a dendrimer [27,28]. Compared to these natural glycodendrimerlike molecules, synthetic dendrimers o¤er templates that are more compact, with dense branching. Synthetic dendrimers allow the multivalent display of ligands in a medium range of valency (10- to 100-mer in a polymer) [29–32]. Depending on branching pattern and core structure, some dendrimers adopt a large spherical or pseudospherical structure, whereas others take on the shape of a half sphere. For example, a hexadecavalent sialoside that strongly inhibits the adhesion of influenza virus to red blood cells is designed on the template of an oligo(lysine)-derived half-spherical dendrimer [33]. The synthesis of multivalent molecules designed on a polymer backbone is readily achieved by polymerizing a ligand monomer or by modifying a preformed polymer by conjugating it with the ligand [34]. A polymer sca¤old provides a high-valency (>100-mer) system that is di‰cult to construct by using repeated connections of linear and branched sca¤olding. Polymeric polyvalent molecules are highly e¤ective in interfering with multivalent interactions, in particular those involving interactions of complementary micrometer-scale surfaces. This high level of interfering activity is attributed primarily to tight binding of one surface, but also partially to a steric e¤ect, a second mode of action observed in certain multivalent systems [35]. This steric occlusion involves blocking the complementary surfaces from approaching one another due to the presence of the interfering ligand, even if that ligand is incompletely bound to one surface. Typical examples of polymeric polyvalent ligands are based on a flexible framework that includes poly(acrylamide). For instance, the sialic acid present in amide side chains of poly(acrylamide) acts as a potent inhibitor of virus–cell adhesion [36]. Other polyvalent examples are based on conformationally rigid sca¤olds, such as polymers of unsaturated framework prepared by ring-opening metathesis polymerization. A specific example of the latter class includes polymeric mannoside, which provides multiple copies of mannoside as an inhibitor of lectin-mediated hemagglutination [37]. Unlike the natural peptide backbone, peptoid is made of synthetic a-amino peptide units in which the side chains are linked at the amide nitrogen rather than at the a-carbon. The resulting tertiary amide both removes a hydrogen bond critical to secondary structure and introduces a second stable rotomer of the amide bond. Both e¤ects are probably responsible for the unique properties that distinguish peptoids from a peptide sca¤old. An example peptoid-based multivalent molecule is illustrated by an N-substituted oligo(glycine) presenting multiple copies of lactoside or mannoside as side chains [38]. Several classes of sca¤old mimic a two-dimensional plane or three-dimensional spherical surface, as shown in Figure 1.3. First, globular proteins such as human albumin constitute a type of adaptable sca¤old because they contain reactive functional groups such as lysine, which are useful for amide coupling, and tyrosine, which is reactive to diazonium molecule for diazo coupling [39].

NOMENCLATURE AND DEFINITIONS

7

These proteins are soluble in water and lack intrinsic glycosylation, thus permitting neoglycoprotein preparation. Self-assembled monolayers (SAMs) on gold comprise a two-dimensional planar surface that enables the presentation of multivalent molecules. This system mimics a cell surface in some aspects, but it lacks lateral mobility as observed in a cell surface. Sugar-displaying SAMs have been designed to study multivalent sugar–lectin interactions [40,41]. Similar systems may be used to model the surface for bacterium–mammalian cell adhesion [42,43]. Liposomes and vesicles allow multivalent display on a spherical surface and are commonly used to simulate cells. In such a system, lateral motion is permitted and allows a multivalent display to ‘‘adapt’’ to a complementary surface, much as may occur in cellular systems. The utility of liposomes and vesicles is demonstrated by several examples, including sialyl ganglioside displayed on a liposome as a potent inhibitor of influenza virus [44], and d-Ala-d-Ala peptide ligand presented on micelles to mimic bacterial cell surfaces [45]. Multivalent display on a larger round surface is possible by using nanometer-sized spheres made of latex, silica, or gold. For instance, polystyrene beads are available for derivatization with 6-sulfo sLe x as a selectin ligand to produce multivalent nanoparticles that can e¤ectively inhibit selectin-mediated attachment of a human embryo to the uterine wall [46]. 1.1.4

Ligand Density

Two or more di¤erent types of molecules can be presented simultaneously, in multiple copies, on the surface of a liposome or on the side chains of a polymer. By varying, for instance, the proportions of bioactive ligand and inactive residues, it is possible to change the biological activity of such molecules. The properties of such molecules are described in terms of surface density or ligand density for liposomes and polymers, respectively. For polyvalent polymers, ligand density is defined as the number of attached ligands relative to the total number of side chains per polymer molecule. For liposomes bearing ligands, the surface density is defined as the average number of ligands relative to the total number of functional groups, including ligands, exposed on the surface of a liposome particle. Pertinent examples are illustrated using multivalent sialic acid, shown as either poly(acrylamide) [47] or liposome [48] in Figure 1.4. Ligand density constitutes a critical factor to be considered when designing multivalent molecules because it broadly a¤ects ligand distribution, interligand distance, and the shape and conformation of the polymer. All of these properties help to modulate the biological activity of multivalent molecules. 1.1.5

Homo- and Heterovalent Molecules

Multivalent interactions can refer to the simultaneous association occurring between multiple, identical pairs of receptor and ligand; such interactions are homovalent. In contrast, multivalent interactions can occur between more than

8

INTRODUCTION poly(SA-co-acrylamide)

O R

n O O O O O O O R NH R R R NH R

OHOH AcHN HO HO

(R = NH2) O

OH

OH

HO O CO2O

HN

HN O

HN O

HN O

HN O

O HO

OH

HN O

HO NHAc OH OH O

OHHO

CO2O HN

O

N H -O2C HO

O

O OH

HO AcHN OH

-O2C HO

OH

HN OH N

O

liposome

O

OH

O

OH

HO AcHN OH spacing

Side-chain density of sialic acid (χSA)

Surface density of sialic acid (χSA)

= [SA]/ [acrylamide]total = 2/8 = 0.25 or 25%

= [SA]/ [amide]total = 2/10 = 0.20 or 20%

Figure 1.4 Definition of ligand density in polymer- and liposome-based multivalent molecules.

one kind of receptor and ligand; such interactions are heterovalent. As with homovalent interactions, formation of multiple ligand–receptor pairs should contribute to the increased association strength of the entire complex. Figure 1.5 illustrates the functionally bivalent association between a heterobivalent ligand and a heterobivalent receptor. Structurally, a heterodimeric molecule closely resembles a bifunctional molecule, as the latter is also composed of two di¤erent ligands tethered through a linker. However, a bifunctional molecule is designed to bind in a monovalent manner to two distinct target receptors widely separated or located in a di¤erent compartment of a cell.

homobivalent association

heterobivalent association

bifunctional molecule


Figure 1.5 Representation of homovalent and heterovalent interaction.

MECHANISTIC ASPECTS OF MULTIVALENT INTERACTION

9

monovalent association Ka

mono

+ ∆G1mono = -RTln(Kamono) multivalent association Kamulti + n

n

n

∆GNmulti = -RTln(Kamulti) where N = (n + 2)

Figure 1.6 Comparison of thermodynamic parameters of association in monovalent and multivalent interaction.

1.2 1.2.1

MECHANISTIC ASPECTS OF MULTIVALENT INTERACTION A‰nity Constant and Avidity

Multivalent and monovalent interactions di¤er fundamentally in their respective definitions and calculations of association strength [1]. For a monovalent system composed of a receptor and a ligand, the a‰nity constant (Ka ) is the binding strength for a monovalent complex (Figure 1.6). This term is related to the free energy of association (DG mono ) by the Gibbs equation below. In a multivalent interaction between two entities presenting N tethered ligands and N tethered receptors, the association constant is defined as avidity (Kamulti ). This constant is a collective association constant that takes into consideration multiple interactions between two multivalent entities, and it is related to the free energy of binding calculated as for monovalent binding: DG mono ¼ RT ln(Kamono ) DGNmulti ¼ RT ln(Kamulti ) In estimating cooperativity in multivalent association, the free energy of multivalent binding (DGNmulti ) can be related to that of N monovalent associations: that is, N DG mono , given N independent receptor–ligand interactions. The ratio between the two indicates the degree of cooperativity, where the cooperativity coe‰cient (a) is defined as follows: a ¼ degree of cooperativity ¼

DGNmulti N DG mono

Depending on the magnitude of a, multivalent interaction is positively cooperative or synergistic (a > 1), noncooperative or additive (a ¼ 1), or negatively cooperative (a < 1). The term cooperativity is often used in biological systems

10

INTRODUCTION

as with the hemoglobin tetramer–oxygen interaction, which shows positive cooperativity [49]. However, the term is rarely used in multivalent systems, partly because few multivalent systems have been shown to demonstrate positive cooperativity. Moreover, multivalent interaction can be much tighter than monovalent binding, regardless of the size of the cooperativity constant. In practical terms, the contribution of a multivalent association is often expressed by the ratio (b) of multivalent avidity to monovalent a‰nity constant, introduced by Mammen et al. [1] (Figure 1.6): b¼

Kamulti Kamono

Thus, b represents the enhancement factor, a term that reflects the strength of a multivalent association relative to the monovalent association. The di¤erence between a and b is illustrated by a trivalent system based on a vancomycin receptor and a d-Ala-d-Ala ligand (Figure 1.7) [50]. Vancomycin is an antibiotic belonging to the glycopeptide class of receptors. Antibiotics in

Vancomycin (R)

Trivalent Receptor (R 3)

OHNH2 OH O HO H3C CH3O

H Vancomycin N

OH O O

HN

O HO H N

O NH

O

O H O N

N H N H

O

O

Cl H N

O

Cl

N O

H

OH O

O

CH3 NH

N H

N

H CONH2

CH3 CH3

O Vancomycin

OH OH

NH

HO Nα-Ac-Lys-D-Ala-D-Ala HN

O

Trivalent Ligand (L 3)

Nα-Ac-Lys-D-Ala-D-Ala

O

H N

O NH

O

N NH H

H3C

CH3 H N O

O OCH3

O Nα-Ac-Lys-D-Ala-D-Ala (L)

R

+

L

R•L

Kdmono ≈ 1.6 x 10-6 M; ∆G = -33 kJ/mol

R3

+

L3

R3•L3

Kdtri ≈ 4 x 10-17 M; ∆G = -94 kJ/mol

Figure 1.7 Interaction of a trivalent vancomycin with a trivalent d-Ala-d-Ala ligand.


11

this class target bacterial cell walls by binding to the d-Ala-Ala peptide precursor. In a monovalent system, vancomycin associates with d-Ala-d-Ala ligand with a dissociation constant (Kd ) value of 1:6 106 M (DG mono ¼ 33 kJ/mol). In a trivalent system composed of trivalent vancomycin and trivalent d-Ala-d-Ala designed by Rao et al. at Harvard [50], the receptor–ligand association is extremely tight, with a Kd value of 4 1017 M (DG tri ¼ 94 kJ/ mol). The avidity of the trivalent complex is much higher than the a‰nity of the monovalent interaction. Based on the value of the cooperativity constant [a ¼ 94 ð3 33Þ ¼ 0:95], this trivalent interaction is still negatively cooperative, although it is exceptionally tight. In terms of binding enhancement (b), the trivalent interaction leads to about a 4 10 10 -fold increase in binding strength relative to that of the corresponding monovalent interaction. In fact, the binding strength is 25 times higher than that of the avidin–biotin system—one of the strongest monovalent interactions in biological systems—thus clearly demonstrating the practical value of multivalency for designing a high-a‰nity system. Gargano et al. at Syracuse [51] proposed a simplified model that can be used in correlating the strength of multivalent association with that of monovalent association. This model makes it possible to estimate the multivalent enhancement factor (b) (Figure 1.8). The model is, however, designed under highly simplified conditions of multivalent systems that are not likely to fit real systems. The conditions include (1) equivalent binding sites on the multivalent receptor as a homobivalent system, (2) noncooperativity (a ¼ 1) such that first binding at one site is thermodynamically identical to second binding at neighboring site, (3) a flexible linker of optimal length to ensure bivalent association, and (4) no linker–receptor interaction, to avoid extra thermodynamic considerations. Because of such limiting conditions, this model should be used for

Kmono Eqn A

+ Kdi

Eqn B

+

K1 Eqn 1

Eqn 2

K2

K2 ≈ Kmono

+

K3 Eqn 3

K1 ≈ 4Kmono

+

+

K3 ≈ 10-2/2 M

Figure 1.8 Theoretical model used for deriving an equation that would make it possible to estimate di(multi)valent association constant.

12

INTRODUCTION

estimating the b term rather than predicting it accurately. It is illustrated by a simple divalent binding model composed of a dimeric receptor and a dimeric ligand, as described in Figure 1.8. This model suggests that the bivalent association constant (K di ) is expressed as three related association constants: K1 , K2 , and K3 . Each equilibrium constant is derived based on certain approximations, such as the number of possible permutations assumed in receptor–ligand site association in Eqn 1 in the figure (K1 ¼ 4K mono ), monovalent binding in Eqn 2 (K2 ¼ K mono ), and e¤ective local concentration (Ce¤ ) of bound divalent ligand in Eqn 3 (K3 A Ce¤ ¼ 102 /2 M, where two tethered ˚ apart [52]). An overall divalent association constant, which ligands are 30 A provides an estimate of the factor of divalent enhancement, is obtained as a product of the three equilibrium constants: K di ¼ K1 K2 K3 ¼ 2 102 (K mono ) 2 b¼

K di ¼ 2 102 (K mono ) K mono

This type of prediction is generalized to the multivalent association constant in a higher-order system as follows: K multi ¼ F (s 102 ) n1 (K mono ) n where n is the valency number, F is a statistical factor defined by the sys˚ )]. An application is illustrated by tem, and s ¼ 30/[interreceptor distance (A k a monovalent P trisaccharide ligand, which binds to pentavalent Shiga toxin (AB5 ) with a Ka value of 1 10 3 M 1 . If this monovalent ligand is converted to a polyvalent ligand that carries P k trisaccharide as side chains of poly(acrylamide), a binding constant(K penta ) can be estimated for the multivalent interaction with the toxin pentamer as follows: K penta (M 1 ) A 1(1 102 ) 51 (K mono ) 5 ¼ 108 (10 3 ) 5 ¼ 10 7 where F ¼ 1 and s ¼ 1. The predicted value is close to the experimental value of the binding constant, K penta A 1:2 10 7 M 1 , obtained from a cell-based assay [51]. This theoretical equation for calculating K multi indicates that the strength of the multivalent association increases exponentially as a function of valency. It also predicts that the intrinsic a‰nity of a monovalent system contributes significantly to the strength of multivalent association and that the intrinsic a‰nity is especially important in cases of low valency, such as for a bivalent or trivalent ligand. 1.2.2

Thermodynamics

The thermodynamic basis of multivalent interactions has been investigated by numerous groups of researchers [1,2]. Mammen et al. have provided a particu-


13

larly thorough description in their review article [1], and the following discussion borrows ideas from this review. To elucidate the thermodynamic basis of multivalent interactions, it is easiest to start by examining a simple, bivalent system. A bivalent association (Eqn B, Figure 1.8) can be compared with a monovalent association (Eqn A) under certain limiting conditions, as defined in Figure 1.8. The free energy of binding (DG) in a receptor–ligand association is made up of enthalpic (DH ) and entropic (DS ) components, in which the monovalent association is linked to the standard relationships between three thermodynamic parameters (DG mono , DH mono , and DS mono ), which are expressed as follows: DG mono ¼ DH mono T DS mono DS mono A DS mono (translational) þ DS mono (rotational) In the equations above, the total entropy change involved is derived from changes primarily in the translational and rotational entropies of the receptor and ligand upon association. The equation represents a qualitative summation. It assumes that the translational and rotational entropies make the greatest contribution and that other possible contributions, such as from the conformational, vibrational, and solvation entropies, are relatively insignificant (i.e., vibrational entropy of a water molecule is about 3000-fold smaller than its rotational or translational entropy) and too complicated to be treated here. For a bivalent association, the free energy of binding (DG di ) is expressed similarly, using standard equations that relate thermodynamic components in qualitative terms: DG di ¼ DH di T DS di DH di A 2DH mono DS di A DS mono (translational) þ DS mono (rotational) þ DS di (conformational‚ linker) where the association of a discrete ligand of a bivalent molecule with each of the bivalent sites is assumed to occur in an unstrained and independent manner. Under such binding conditions, the bivalent enthalpy change is equal to twice the enthalpy of monovalent binding. The change of entropy in the bivalent association is considered to consist of three components. The first two terms are equal to translational and rotational components from monovalent association. Such an estimation is based on certain limited considerations: (1) translational and rotational entropies of divalent and monovalent particles are equal because these entropies are only weakly dependent on mass and size; and (2) in both bivalent and monovalent associations, two particles associate and become one particle, resulting in the net loss of the free translation and rotation

14

INTRODUCTION

of a single particle. The third component results from a change in conformational entropy. This component takes into consideration the conformational entropy that is contributed by the change in entropy of the linker upon association. For a simple estimation of the thermodynamic advantages of a multivalent system, the free energies of association can be compared between two ideal systems. One is a multivalent system in which one multivalent receptor, with n sites, binds to one multivalent ligand, also with n sites, to become one perfectly matched multivalent complex. The second system is represented by n independent pairs of monovalent complexes composed of n copies of the monovalent ligand and n copies of the monovalent receptor. Thus, the free energy of association for each system is given collectively as follows: nDG mono ¼ nDH mono nT DS mono DG multi (n) ¼ DH multi (n) T DS multi (n) A nDH mono T DS mono T DS multi (n) (conformational‚ linker) Subtracting the two equations provides a free-energy di¤erence between the two systems: DDG ¼ DG multi (n) nDG mono ¼ (n 1)T DS mono T DS multi (n) Therefore, if the multivalent association is tighter than the monovalent association, it is due largely to the entropic contribution of monovalent binding (DS mono < 0). This comparison suggests that multivalent binding is favored because it su¤ers a relatively smaller entropy loss. In practical examples reported in the literature, the thermodynamic benefits of a multivalent system are estimated from a comparison with one pair of a monovalent system rather than with n pairs of a monovalent complex as a group. As an example, the free energy of bivalent association is compared with that of monovalent association, where the di¤erence between DG di and DG mono can be approximated by two terms, the enthalpy of monovalent binding and the conformational entropy. Such a relationship is extended similarly to a multivalent system: DG di ¼ 2DH mono T DS mono T DS di (conformational‚ linker) DG di DG mono ¼ DH mono T DS di (conformational‚ linker) DG multi (n) DG mono ¼ (n 1)DH mono T DS multi (conformational‚ linker) Based on this estimation, a negative value of the free-energy di¤erence (DDG ¼ DG multi DG mono ) represents multivalently enhanced binding. There-


15

fore, the enthalpy term contributes significantly and positively to the free energy of the multivalent association, whereas the conformational entropy causes a negative e¤ect because it is often associated with loss of conformational entropy of a linker upon receptor–ligand association. Such an enthalpy contribution has also been noted in ditopic binding of the cyclodextrin receptor dimer by Breslow et al. at Columbia [19]. In summarizing this section, thermodynamic aspect suggests that the enhancement factor (b ¼ eDDG/RT ) becomes greater in a system that involves a higher valency number (n), a tighter monovalent association (more negative DH mono ), and a rigid linker (a lower degree of conformational entropy loss), unless the linker interacts directly with the receptor. The thermodynamic basis of a tight multivalent association is illustrated by an experimental system (Figure 1.7) that is based on the association of a trivalent vancomycin molecule to a trivalent d-Ala-d-Ala molecule [50]. The trivalent binding results in a favorable value of DG tri (94 kJ/mol), which is dissected to enthalpic and entropic contributions (DH tri ¼ 167 kJ/mol; T DS tri ¼ 73 kJ/mol). It is a highly favored process compared with the association of monovalent vancomycin and d-Ala-d-Ala, which leads to a DG mono value of 33 kJ/mol, where DH mono is composed of DH mono (50.2 kJ/mol) and T DS mono (17.2 kJ/mol). Thus, in this system, DH tri has a value that is slightly more negative than three times the value of DH mono and thus is a major, driving factor for tight association. On the other hand, the absolute value of T DS tri is larger than that of T DS mono by 56 kJ/mol, which might be related to the loss in conformational entropy of trivalent complex, including the entropic loss contributed by linker. Clearly, favorable enthalpy makes a strong contribution to the very tight association of the trivalent system (K dtri ¼ 4 1011 M ). The thermodynamic basis for benefits of multivalent ligand–receptor association can be influenced by the system itself. For example, a multivalent ligand can interact with either a multivalent receptor presenting multiple connected sites (such as trivalent vancomycin and trivalent d-Ala-d-Ala) or unconnected multiple receptor sites. A series of studies performed by Dam et al. at Albert Einstein School of Medicine [2,53,54] provide important insights underlying the thermodynamics of multivalent binding of the latter class. Their studies are based on the association of concanavalin A dimer (Con A at pH 5.2) and divalent trimannoside cluster (Figure 1.9). The trimannoside dimer can form a soluble, one-dimensional cross-linked complex with Con A, with a divalent association constant (Kadi ) of 2:5 10 6 M 1 , which is about sixfold tighter than that of trimannoside monomer to Con A (Kamono ¼ 3:9 10 5 M 1 ). Such a slight enhancement in bivalent binding can be attributed to a combination of two separate, sequential association events where a trimannoside dimer cross-links to two lectins with a distinct binding a‰nity (first association: K a1 ¼ 1:6 10 7 M 1 ; second association: K a2 ¼ 8:8 10 5 M 1 ). Therefore, the first association is 18-fold tighter than the second one, which translates into

16

INTRODUCTION Binding parameters to concanavalin A (Con A)

Ka = 6.2 x 105 M-1 ∆G = -7.9 kcal/mol

HO HO HO

O HOO HO HO HO

First ligand association Ka1 = 1.6 x 107 M-1 ∆G1 = -9.8 kcal/mol

OH O

HO HO HO

O OH

Trimannoside monomer

OR

OH O O HO O

HO HO HO

OH O

O OH

OH O O O NH NH S

Second ligand association Ka2 = 8.8 x 105 M-1 ∆G2 = -8.1 kcal/mol

HO HO HO

OH O

NH O HOO

HO HO HO

CO2R

S

O OH

OH O

NH O

O Trimannoside dimer

Figure 1.9 Thermodynamic binding parameters in divalent trimannoside–lectin interaction.

a more negative value for the free energy of binding (DG1 ¼ 9:8 kcal/mol; DG2 ¼ 8:1 kcal/mol), indicating negative cooperativity generated on bivalent association. The two separate association events display similar enthalpies of binding (DH1 ¼ 12:5 kcal/mol; DH2 ¼ 12:3 kcal/mol), whereas the first association shows a more favorable entropy of binding (T DS1 ¼ 2:7 kcal/mol) than the second (T DS1 ¼ 4:2 kcal/mol), or binding of a trimannoside monomer to a lectin molecule (T DS ¼ 5:2 kcal/mol). This class of bivalent ligand– lectin association represents a multivalent system driven by entropy. Therefore, in the thermodynamic basis of binding, this system is clearly distinguished from a previous enthalpy-driven vancomycin/d-Ala-d-Ala system where a trivalent receptor presents three connected sites. Drug–DNA interactions represent another well-studied multivalent ligand– receptor system [55]. This type of system is exemplified by the association of hydroxyrubicin and daunorubicin to double-stranded DNA (Figure 1.10). Hydroxyrubicin is an anthracycline antibiotic composed of two parts. The first part is an adriamycinone moiety that intercalates into nucleotide bases. The second part is an attached sugar moiety that is able to bind at a DNA minor groove. Chaires et al. [56] reported that the free energy of binding (DG) of hydroxyrubicin to DNA is 7.0 kcal/mol, while that of adriamycinone is 5.7 kcal/mol. The higher a‰nity of hydroxyrubicin is attributable to the bivalent mode of interaction (intercalation plus minor groove recognition) compared to the monovalent mode of interaction (intercalation only) by adriamycinone.


17

base sugar

intercalator

phosphate

divalent association ∆Gdi minor groove binder

multivalent association ∆Gmulti


intercalator

∆Gmono

covalent dimer

I∆Gmulti I > I∆Gdi I > I∆Gmono I

O

OH

O

O

OH

OH

OH H3 CO

H3CO

Hydroxyrubicin (∆G = -7.0 kcal/mol )

OH

O

O

OH

O OH

OH

O OH O H3C O OH OH

O

O

H3CO

O OH O H3C O OH NH2

O OH O H3 C O OH NH

Daunorubicin HN OH O CH

3

OH

O

OH O

OCH3

OH HO H3CO

O

OH OH O

Adriamycinone (∆G = -5.5 kcal/mol )

Figure 1.10

OH O

Daunorubicin dimer

Mode of DNA binding displayed by daunorubicin and its dimer.

The involvement of multivalency in drug–DNA interactions is illustrated more dramatically by daunorubicin, a close analog of hydroxyrubicin. A daunorubicin dimer binds tightly to DNA with an association constant Ka of 2:7 10 11 M 1 (DG ¼ 15 kcal/mol), which is 16,000-fold greater than that of daunorubicin (DG ¼ 10 kcal/mol) [57]. This greatly enhanced a‰nity comes mostly from a large enthalpy contribution (DH ¼ 30 kcal/mol; T DS ¼ 15 kcal/mol) compared to daunorubicin (DH ¼ 11 kcal/mol; T DS ¼ 1 kcal/ mol). The favorable enthalpy displayed by the daunorubicin dimer, which is close to three times that of the daunorubicin monomer, serves as a major driving force enabling the otherwise unfavorable entropic cost arising from multi-

18

INTRODUCTION kon Kd = koff/kon

+ koff

+

kon

kondi

koff

koffdi

Figure 1.11 Comparison of association kinetics in monovalent and bivalent systems.

valent drug–DNA interaction to be overcome. This example illustrates the important role of enthalpy in the association of a multivalent ligand with a multivalent receptor. In summary the thermodynamic basis of multivalent association can be driven by enthalpy or entropy and depends on the nature of multivalent system as discussed with the three examples above, involving vancomycin, lectin, and DNA.

1.2.3

Kinetics

The thermodynamic aspects of multivalent receptor–ligand interactions are highlighted by enhanced binding a‰nity. These aspects are, however, inseparably related with the kinetic features of multivalent binding, which is characterized by a decrease in the rate of dissociation of the two interacting entities. The kinetic aspects of bivalent association are shown schematically in Figure 1.11, in which a bivalent ligand occupies two binding sites present on a divalent receptor. The dissociation of a bound dimer occurs in two stages, so that its di dissociation rate (ko¤ ) at the second site can be compared theoretically with the rate of dissociation (ko¤ ) of a bound monovalent ligand, as suggested by Kramer and Karpen [52]:

di ¼ ko¤

2ko¤ K d Ce¤ þ K d

where Ce¤ refers to the e¤ective local concentration of a ligand unit at the second site and Kd refers to the dissociation constant of the monovalent ligand. The e¤ective concentration of a bound divalent ligand is calculated from the volume of the molecular hemisphere, with a radius of r equal to the length of the linker:

Ce¤ ¼

1000 NA 23 pr 3


19

where NA is Avogadro’s number. The e¤ective concentration is closely associated with e¤ective molarity as a kinetic parameter commonly used in enzymatic or intramolecular reactions where e¤ective molarity contributes to rate acceleration relative to bimolecular reaction as investigated by Page and Jencks [58] and others. Thus, the rate of dissociation at the second site, di ko¤ , is defined as a linear function of the monovalent dissociation rate ko¤ multiplied by both a statistical factor of 2 (because dissociation can occur at either of two sites) and the probability K d =(Ce¤ þ K d ) that the second site is not occupied. Approximating the divalent dissociation rate is illustrated by the interaction of a bivalent cyclic guanosinemonophosphate (cGMP) tethered with a variable poly(ethylene glycol) spacer of molecular weight 2000 kDa. The spacer includes a cyclic nucleotide-gated channel which presents four cGMP sites distributed equally on four subunits comprising the channel [52]. With a linker length of ˚ , this bivalent ligand binds the tetrameric channel receptor with a K ddi val39 A ue of about 5 nM, which represents 600-fold tighter binding than that of the cGMP monomer (K d A 3 mM ). Assuming that the first site is occupied by a bivalent cGMP, the e¤ective concentration of cGMP at the second site is calculated to be 13.4 mM, which is much higher than the monovalent dissociation di constant (K d A 3 mM ). Under such circumstances, ko¤ at the second site is predicted to be about 1700-fold slower than ko¤ by monovalent cGMP. Thus, the decrease in dissociation rate is linked to the enhanced binding a‰nity of bivalent cGMP. Another good example of a slow dissociation rate by multivalent species involves a trivalent system composed of vancomycin and d-Ala-d-Ala (Figure 1.7). Compared to a monovalent system of vancomycin and d-Ala-d-Ala peptide (K d ¼ 1:6 10 6 M; kon ¼ 9:3 10 6 M 1 /s; ko¤ ¼ 14:88 s1 ), this trivalent system shows an exceptionally low dissociation constant Kd of 4 1017 M, tri and ko¤ is estimated to be 3 109 s1 [50]. Thus, for any trivalent receptorligand with a very tight association, a slower rate of dissociation can be inferred. 1.2.4

Steric E¤ects

As discussed, to design any multivalent ligand intended to target a cell surface receptor, the linker is an important factor to consider. In addition to its normal role as a spacer for bridging, it is implicated in the occurrence of undesired steric e¤ects caused by the size of receptor and the incompatible ligand spacing of multivalent ligands [59]. This unfavorable e¤ect is illustrated in Figure 1.12, where the interligand distance (LL ) is smaller than the distance between adjacent receptor sites (LR ). Under these circumstances, a certain fraction of unbound ligand units of a multivalent ligand are no longer available for receptor binding because these units are either obstructed directly by a bound receptor or they are excluded sterically by adjacent bound receptors.

20

INTRODUCTION optimal spacing: LR = LL LL

LR

cell surface receptor

short spacing: LR > LL LL

Figure 1.12 Definition of steric e¤ect in multivalent receptor–ligand complexation.

sterically hindered, and unavailable for receptor binding

Hlavacek et al. [59] proposed a theoretical model defining the steric e¤ect in a multivalent system that makes it possible to predict the relationship between the fraction of available ligand units and the fraction of bound receptor sites. It is illustrated using a linear multivalent ligand with a valency of 200 under arbitrary conditions where the number of ligand units covered by or in contact with bound receptor is 3 (LR > LL ; Figure 1.12). This model predicts that if the fraction of bound receptor sites by such a multivalent ligand is 5% (i.e., there are 10 bound ligands), about 75% of ligand units are still available for receptor binding. If the fraction of bound receptors increases to 10% (20 bound ligands), the fraction of ligand units available for receptor binding decreases to 50%. Therefore, a moderate increase in the fraction of bound receptor sites leads to a much greater degree of steric hindrance. In contrast, where linker spacing exceeds the optimal length between receptor sites, such that LL > LR , unbound ligand units are not covered or excluded sterically by bound receptors but instead are possibly available for receptor binding. However, the increase in linker length accounts for the decrease in e¤ective local concentration of the ligand, and thus leads to a weaker contribution to multivalent binding.

1.3

BIOLOGICAL ROLES OF MULTIVALENT LIGANDS

In biological systems, multivalent binding provides a broad range of benefits and unique roles that are not achievable with monovalent interactions. Here, we introduce selected features of multivalency benefits; detailed coverage is left for upcoming chapters. First, multivalent interaction is collectively much tighter than monovalent interaction. There are numerous examples of multivalent ligands that bind more strongly to their respective multivalent receptors,

BIOLOGICAL ROLES OF MULTIVALENT LIGANDS

21

as demonstrated in a diverse array of targets, including molecules as diverse as cell surface receptors, enzymes, G-protein-coupled receptors, ion channels, lectins, toxins, and nucleic acids found in cells and pathogens (viruses, bacteria). Notable examples include multivalent sialic acid molecules that bind to the hemagglutinin receptor on the surface of the influenza virus, where the multivalent enhancement factor ranges from 10 to 10 6 , depending on the degree of valency and type of sca¤old [60]. High-a‰nity multivalent ligands that target surface receptors show great potential as e¤ective inhibitors of surface–surface interactions, including cell–cell and cell–pathogen interactions, that occur in biological systems. Second, multivalent interactions can enhance receptor selectivity, as demonstrated by studies using multivalent ligands that bind to lectins [61], acetylcholine esterase [62], monoamine transporters [63], and serotonin receptors [64]. Third, multivalent interactions induce receptor clustering on the cell surface, as illustrated by bacterial galactoside-sensing receptors [65] and galactosyl transferase receptors on the surface of sperm [66,67]. Fourth, multivalent interactions provide a strategy for controlling signal transduction pathways within cells, as represented by the FK506 dimer, a prototype member of chemical inducers of dimerzation (CID) [68–70] and by the polyamide dimer as an artificial transcription factor [71]. Numerous excellent review articles deal with potential applications and characteristics of multivalent ligand–receptor interactions. The principles and biomedical applications of polyvalent interactions are discussed systematically by Mammen et al. [1]. The thermodynamic basis of multivalent binding is evaluated with respect to lectin–carbohydrate interaction by Brewer [54] and with respect to drug–DNA interaction by Chaires [55]. Classical examples of multivalent interactions, known as the cluster e¤ect, are described by Lee and Lee [26] in the context of carbohydrate–protein (C-type lectin and animal lectin) interaction, and multivalent selectin inhibitors are reviewed by Simanek et al. [72]. The mechanism of action by dimers of the FK506 class and its biomedical uses are reviewed by Crabtree and Schreiber [68]. Accounts by Roy [29] and Kiessling and Pohl [60] deal with chemical and biological aspects of multivalent neoglycoconjugates, and those by Yarema and Bertozzi [73] and Bertozzi and Kiessling [74] describe the novel roles played by multivalent carbohydrates in a chemical approach to glycobiology. Polyamide dimers are treated as highly selective DNA binders and their use in transcriptional control is described by Dervan and Bu¨rli [75] and Wemmer and Dervan [76]. The need for polyvalency when designing drugs to prevent pathogen–cell adhesion is described by Matrosovich [77]. Examples of multivalent drug candidates are documented in several therapeutic areas by Wright and Usher [11], while focused reviews include that of the use of multivalent glycan mimics as antiinfectives by Mulvey et al. [78], and rational design of a multivalent inhibitor of bacterial toxin by Fan and Merritt [79]. The role of noncovalent dimerzation of vancomycin in its antibiotic action is reviewed by Williams [80]. Cyclodextrin is a molecule that plays a dual role: Its hydrophobic cavity serves as a receptor,

22

INTRODUCTION

and multiple hydroxyl groups on its rims serve as handles for ligand attachment. Breslow et al. [19] reviewed the chelate e¤ect observed in ditopic binding of covalent cyclodextrin dimer. Fulton and Stoddart [21] discuss neoglycoconjugates designed on a multivalent sca¤old composed of cyclodextrin and calix[n]arene. Finally, the use of protein- and synthetic polymer-based polyvalent carbohydrates is reviewed by Bovin and Gabius [39].

2 MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS

2.1 2.1.1

INFLUENZA VIRUS Hemagglutinin

Influenza virus initiates mammalian cell infection by adhering to the cell surface (Figure 2.1) [81]. This viral adhesion is mediated by the binding of viral surface protein–hemagglutinin (HA) to sialic acid (i.e., N-acetylneuraminic acid), which is presented on cell surfaces [82,83]. The hemagglutinin is expressed as clusters of trimers on the surface of influenza viruses A and B, in ˚ apart from each other, as suggested by lowwhich the trimers are 105 to 109 A angle neutron-scattering experiments (Figure 2.1) [84,85]. The virus–cell association is very tight and involves multivalent interactions among multiple pairs of HA trimer and sialic acid ligand. In contrast, the association of monovalent sialoside to a HA trimer shows very weak binding (K d @ 2 mM ). 2.1.1.1 Divalent Sialic Acid. Glick and Knowles [86] tested the notion of multivalent recognition by designing sialic acid–based inhibitors that target influenza virus. Their multivalent design led to the synthesis of two classes of bivalent sialosides, 1 and 2 (Figure 2.2). In these two related molecular classes, two sialic acids are tethered to each end of a xylyl-based sca¤old either an oligo(glycine) or poly(ethylene glycol) linker of varying length. These dimers were biologically active with an ability to bind hemagglutinin target and to in-


23

24

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS

Influenza virus

Hemagglutinin (HA) trimer

Sialic acid

cell membrane

glycolipid

glycoprotein

HA trimer

105 Å

46 Å Sialic acid-binding site

Figure 2.1 Adhesion of influenza virus to mammalian cell via multivalent binding of viral hemagglutinin (HA) trimers to multiple sialic acids expressed on cell surfaces and the estimated distance between sialic acid–recognition sites located within a hemagglutinin trimer and neighboring HA trimers.

1 (n = 0 to 4)

HO OH

HO OH

CO2AcHN

O OH OH

OMe

CO2AcHN

α-O-Methyl Sialoside

O OH OH

O

O N H

O

H N O

H N n

NH n

O O

OH OH NHAc O CO2OH OH

Span ~27 to 57 Å

HO OH CO2AcHN

O OH OH

O

O N H

O

O

2 (n = 1 to 3) H N

H N n

O

O

O

O

O n

N H O

O CO2-

OH OH NHAc OH OH

Span ~49 to 65 Å

Figure 2.2 Bivalent sialosides linked with oligo(glycine) or poly(ethylene glycol) spacer.

INFLUENZA VIRUS

25

hibit influenza-cell adhesion. Specific a‰nity of the sialoside dimer to ligand recognition sites located within a hemagglutinin trimer was studied by using nuclear magnetic resonance (NMR) titration. These experiments were typically carried out in a virus-free system where binding was tested with a nonmembrane-bound, aqueous-soluble form of hemagglutinin (bromelain released form). For example, bivalent sialosides such as 1 (n ¼ 4) and 2 (n ¼ 1 to 3) showed a dissociation constant of 2.8 mM, which is identical to that of monovalent methyl sialoside. The similar binding a‰nities of the monovalent and bivalent species strongly suggest that the dimers bind monovalently. This result is somewhat surprising and unclear since some of the dimers have spacer ˚ —enough to span the distance of about 46 A ˚ required to lengths up to 65 A bridge to ligand sites within an HA trimer. Although dimeric sialosides (1 and 2) show virtually no increase in a‰nity to soluble HA trimer over the monovalent molecule, they are capable of inhibiting the adhesion of influenza virus (X-31) to chick red blood cells as measured in the hemagglutination inhibition (HAI) assay. In this assay, these same bivalent ligands show enhanced inhibitory activity, which is about 6- to 100-fold larger than that of methyl sialoside (K iHAI ¼ 2:5 mM ). With a specific, extended ˚ , two dimers, 1 (n ¼ 4) and 2 (n ¼ 2), display optilinker length of about 57 A mal binding with an inhibition constant K iHAI of 25 and 125 mM, respectively. In this assay, dimeric sialosides are likely to display certain mode(s) of action distinct from simple monovalent binding. One possibility is the involvement of bivalent binding, where sialic acid dimers tethered with a long spacer are able to bridge hemagglutinin trimers, which are clustered in multiple copies on the viral surface. Such binding should be able to discourage virus–cell adhesion, perhaps through tighter association. Natural sialic acids, as found in glycoproteins and glycolipids, are presented structurally as terminal sugars and in a multivalent format derived from complex carbohydrate clusters. Sabesan et al. [87,88] studied a series of bivalent sialoside-containing carbohydrate clusters as ligands that bind influenza virus in a bivalent mode (Figures 2.3 and 2.4). Each of the sialoside clusters (3 to 5) present two units of NeuAc-a2,6-Gal-b1,4-GlcNAc-b1 anchored on a galactoside sca¤old. They are arranged in specific orientations at positions 2,4-, 3,6-, and 4,6- on the galactoside. Depending on the relative orientation, the distance ˚ at positions 3,6 between two sialosides in each cluster varies from about 9 A ˚ and 4,6 to about 19 A at positions 2,4. These heptasaccharides serve as inhibitors of influenza virus A adsorption to resialylated human erythrocytes in the hemagglutination inhibition assay. For instance, dimers 4 and 5, which ˚ , show 10- and 8.4-fold increased posess a sialoside-to-sialoside distance of 9 A inhibitory activity, respectively, over that of the monomer control, methyl a˚, sialoside (K i ¼ 1:9 mM ). By contrast, 3, with an end-to-end distance of 19 A shows the same inhibitory activity as the monomer (K i ¼ 1:3 mM ). Thus, a ˚ supports optimal activity. At such a distance, a ligand separation of about 9 A divalent sialoside cluster is more likely to bind intramolecularly to the closest ligand recognition sites present on a hemagglutinin trimer rather than to bridge

26

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS HO OH AcHN HO

CO2OH O HO HO O

O HO O

HO

NHAc

HO HO

~19 Å (2,4-orientation)

AcHN O HO HO

-O2C O

O

HO OH AcHN OH HO

HO OO

-O2C O O HO AcHN OO HO HO HO HO HO


HO HO

NHAc OH O O O O OH HO AcHN O O HO OO

HO OH AcHN OH HO

OH

-O2C O

-O2C O

O

HO OH AcHN OH HO

O HO

OCH3

3

OCH3

4

OCH3

5

O

OH

O

HO HO

O O O

OH

HO HO


O

OH

OH HO O O NHAc O O O AcHN O OH

HO OO OH

O

-O2C

O

OH O O

O

OH

O HO OH OHHO

HO HO AcHN OH

Figure 2.3 Structure of three types of heptasaccharide clusters displaying di¤erently oriented bivalent sialosides.

two sites located on adjacent hemagglutinin trimers, which would require ˚. spanning a distance of more than 55 A 2.1.1.2 Tetravalent Sialoside. Tuzikov et al. [89] reported a tetravalent sialic acid designed on the sca¤old of an antennary structure, C[aCH2 NHCOCH2 NHa] 4 . This structure forms noncovalent oligomers as a consequence of selfassembly mediated by intermolecular hydrogen bonding. This tetrameric sialoside inhibits influenza virus–cell adhesion 1000-fold more e¤ectively than does monovalent sialoside. 2.1.1.3 Dendrimers Presenting Sialosides. Multivalent molecules designed with hyperbranched fractal structures, such as dendrimers, mimic natural mul-

INFLUENZA VIRUS OH O O

HO

OBz O O OBz

OH O O

O

O

OH OH

HO

i,ii

OCH3

OH

iv,v

OCH3

HO HO

O

OH

27

iii

OCH3 O

OTBDMS

NHAc OH O O O O OH HO AcHN O

OCH3 O

O HO HO

vi

OH HO HO

OH

NHAc OH O O O O OH HO AcHN O

HO OO

OCH3

vii

O

O

OH HO

HO OO

OH

HO HO

OH

HO HO

-O2C O O HO AcHN OO OH OH HO HO HO HO HO

-O2C

AcHN OH

NHAc OH O O O O OH HO AcHN O O

O

HO OO

HO OH HO

OH

O

OCH3 O

4 OH

reagents and conditions: i) n-Bu2SnO, benzene, Et4NBr, allyl bromide; ii) TBDMSCl, DMF; iii) PhCOCl, pyridine; Ir(COD)-[Ph2Me]2PF6, THF; 10% HgCl2, 90% aqueous acetone; iv) 2-deoxy-3,4,6-tri-O-acetyl-2-phthalimido-α,β-D-glucopyranosyl bromide, AgOTf, collidine, CH3NO2; v) NaOMe, MeOH; H2NNH2, MeOH; Ac2O, pyridine; NaOMe, MeOH; vi) UDPgalactose, bovine galactosyltransferase; vii) CMP-NeuAc, 2,6-sialyltransferase

Figure 2.4 Chemoenzymatic synthesis of divalent sialoside cluster.

tiantennary ligands present on glycoproteins. Roy et al. [90] reported the synthesis of dendritic a-thiosialosides as inhibitors of influenza A virus. These molecules consist of a-thiosialosides located on the tips of hyperbranched arms that emanate from an oligo(lysine)-based dendrimer (Figure 2.5). Examples of such sialic acid–presenting dendrimers include a simple divalent ligand (5; first-generation dendrimer) to a dense hexadecavalent dendrimer (8; fourthgeneration dendrimer). These dendrimers can block the adsorption of influenza virus A (strain X-31) to mammalian cells. Compared to sialic acid monomers, the dendrimers show enhanced inhibitory activity in virus-mediated hemagglutination of chick red blood cells: K iHAI ¼ 625 mM (5), 313 mM (6), 156 mM (7), and 19 mM (8). These activities are 5- to 158-fold greater than that displayed by methyl sialoside monomers (K iHAI ¼ 3 mM ). Moreover, higher valency leads to greater antiviral activity. Thus, these dendritic clustered sialosides belong to a class of e¤ective multivalent inhibitors of influenza virus.

28


AcHN HO

OH OH

AcHN HO

OH OH

HO HO O CO2O CO2OH S AcHN S OH OH OH O NH HO O AcHN OH O HO O OH CO2- NH HO AcHN O CO2- HN O S HO HO HN S O NH O HO O CO2NH NH O O O S O NH HN NH HN OH OH O O O O N AcHN HN H HN O HN HO O CO2NH O O HO S NH O NH O HN O O NH O N H HN O HO O OH OH O

5 (1st generation) 6 (2nd generation)

AcHN HO

HO

OH OH

O

S

CO2-

HO NHAc OH HO OH O S CO2-

O NH

O

HN O HO NHAc NH O OH HO OH HO HO NHAc O NH O HN OH O S CO2O O S CO - OH HO O 2 NHAc HO HN NH NH O OH HN O S HN O O O HN O CO2- OH O NH NH O HN HO HO HN O O O H NHAc S O O HN N N H N O CO2- OH N NH H O N O H H H OH N O N O H O H HN OH OH N S O HN N O NHAc O H CO2O O OH OH HN NH O (3rd generation) O HN S OH O OH -O2C NHAc HOHO

OH OH

HO O CO2S

7

AcHN HO

AcHN HO HO

OH OH

AcHN HO

OH OH AcHN S O AcHN OH O NH HO CO2O HO OH HO OH S O HO O CO2- NH HO AcHN O HN O HO O S CO NH CO 2 O 2 HN HO O NH S S O NH O O HO NHAc HO NHAc HN OH NH O O NH HO OH O OH HO HO O OH NH O O O OH HO NHAc HN HN AcHN O HO O NH N CO2- O S CO2OH HO S O CO2HN O O H HN O OH HO NHAc HO S CO O HN HO S NH O 2 O O NH O NH O OH HN O S HN O NH O NH O HN CO2- OH O HO OH NH HN HN O HN O NH O NH NH O CO2O O HN O AcHN H O O O O N HO HN S N HO HN N H NH H O NH H O N HO OH O HO HO N O O O H CO2- H O O HN O O NHAc S O N O H N NHH AcHN O HN N N N N N OH OH S CO2- OH O H O O NH H H HN HN O H OH O H HN O H N O O N HO -O2C H O O N N HO H O S O OH N S O NH OH O H AcHN O OH OH NHAc -O2C OH OH

O

CO2-

HO

O OH

HO HO

8 (4th generation)

Figure 2.5 Lysine-cored dendritic sialosides: first to fourth generations.

A representative synthesis of dendrimer 8 in the solid phase is shown in Figure 2.6. The majority of synthetic e¤orts lie in the construction of the dendritic l-lysine core anchored through b-alanyl spacer to poly[styrene-co-4(hydroxymethyl)phenoxymethyl]resin, also known as Wang resin. The sialoside derivative is tethered to each of the glycine ends of dendritic arms through an a-thio linkage, which, unlike the a-O-linkage, is strongly resistant to hydrolytic cleavage catalyzed by viral sialidase (neuraminidase).

INFLUENZA VIRUS OAc

AcO

AcO

AcHN

O OAcOAc

OAc

AcO

i

Cl

CO2Me AcHN

CO2Me

O OAcOAc

CO2Me AcHN

SAc

H2N

OAc

ii O OAcOAc

SH

NeuAc-SH

O Wang resin (polymer support)

29

O O

H N

Cl

O N H

O AcO

O Lys8-Lys4-Lys2-Lys

N H

16

O

O

iii-vi O

OAc CO2Me

O AcHN AcO OAc HO

H N

O

H N

S

O N H

O

H N

N H

16

O

O

O Lys8-Lys4-Lys2-Lys

O O

vii

OH CO2H

O AcHN HO OH

H N

S O

O N H

H N O

O

O Lys8-Lys4-Lys2-Lys 16

N H

OH

8

viii-x

reagents and conditions: i) AcSNa, Bu4NHSO4, EtOAc, Na2CO3, 25oC, 30 min (66%); ii) NaOMe, MeOH, -40oC, 5 min, then H+ resin; iii) N α,N ε-di-fmoc-L-Lys-OBt, DMF; iv) 20% piperidine; v) repeat steps–iii,iv– three more cycles; vi) ClCH2CO-Gly-Gly-OBt, DMF; vii) NeuAc-SH, 1% Et3N, DMF, 16 h, 25oC, viii) 95% aq. TFA, 1.5 h (66-99%); ix) NaOMe, MeOH, 1 h; x) 0.05 M NaOH, 2 h, 25oC, then H+ resin

Figure 2.6 Solid-phase synthesis of fourth-generation dendritic sialoside as an inhibitor of influenza virus.

A sialoside-presenting glycodendrimer with 16 sialoside residues (9) is prepared using a di¤erent strategy, as reported by Zanini and Roy [91]. This strategy is based on a solution-phase method that utilizes 3,3 0 -imino(propylamine) as a building block for the dendrimer core (Figure 2.7). Further, Reuter et al. [92] reported a unique class of polymer-based dendritic sialosides that are composed of dendrimer (9) linked in multiple copies to a polymer sca¤old. The resulting dendritic polymers are potent inhibitors of influenza hemagglutinin, with their activity dependent on the shape and valency of the dendritic polymers. For example, a comb-branched dendrigraft polymer shows enhancement in inhibitory activity up to 50,000-fold relative to monomeric sialoside. 2.1.1.4 Sialic Acid Displayed in Liposomes. Liposomes closely mimic natural cell membranes in shape, surface topology, and chemical functionality [3]. Liposomes o¤er several advantages when used as a multivalent sca¤old: (1) liposomes can present multiple copies of ligands on their surface; (2) as with a fluid bilayer, liposomes allow lateral movement of surface ligands, which aids in optimizing the strength of multivalent binding to the receptor; and (3) the lipid bilayer tends to limit ligand motion to a two-dimensional plane rather than a three-dimensional space, leading to a relatively minimal loss of translational and rotational entropy change upon multivalent association. The utility of liposomes as sca¤olds in multivalent design was demonstrated

30


iii

CbzHN O N H H N N O

N

O OH

O N H H N N O

N

O OtBu

N

CbzHN

CbzHN

O N OH

CbzHN

CbzHN

i

H2N

O N OtBu

H2N

CbzHN O N H H N N O N

CbzHN CbzHN

N

O OtBu

CbzHN H2N N

CbzHN

H2N

ii

H2N H2N

CbzHN CbzHN NHCbz NHCbz CbzHN NHCbz CbzHN N NHCbz CbzHN N NHCbz O N O N HN CbzHN N O O NH NHCbz NH N HN O O CbzHN NHCbz HN NH N N N H H CbzHN N N O NH N NHCbz N O NH HN OHN N O O O N N

ii, iv,v,vi,vii

O AcO OAc

NH HN O

CO2Me O SH AcHN AcO OAc

N BuO

NeuAc (Ac4)-SH

t

O

OH AcHN OH OHHO OH AcHN OH HO OH HO NHAc HO NHAc HO O CO2- O CO2HO HO O OH OH OH S HO OH AcHN OH S OH COHO 2O HO S O O OHOH O NH NH CO2- HO NHAc HO O CO2S CO2- S O AcHN HO O NH O OH S HO O OH O CO2NH N NH HO S NH O CO2-HOHO NHAc O S O HO OH NH N OH HN O NH N O AcHN O S CO - OH O HO N O 2 O CO2HO H NH HN O NH S NH N HO HO NHAc N HO OH O S O CO2- HO N O N AcHN OH N O O CO2- OH N H NH HO S O N HO NHO H HO HO N O O N N NHAc NH S O H N HO OH H H -O2C N O O CO2- OH N N N H O S N HN O OH AcHN OHN HN O N O OH OH O O OH N N S H OH N OH N H -O C O NHAc HO N O 2 HO -O2C H O S O HO HO AcHN OH OH AcHN HO

9 reagents and conditions: i) DIC, HOBt (42-82%); ii) H2, 10% Pd/C (97%); iii) 30% TFA (96%); iv) (ClCH2CO)2O (91%); v) NeuAc-SH, 1% Et3N, DMSO; vi) NaOMe, MeOH; vii) 0.05 M NaOH

Figure 2.7 Solution-phase synthesis of PAMAM-based dendritic sialoside.

i

31

INFLUENZA VIRUS 10

HO OH CO2AcHN

O OH OH

H N

O

O

O

H N

N O NH H O S

S

CH3

CH3

11 CO2-

O OH OH

O

O

O

NMe2

HO OH AcHN

H N

O

H N

O

O

H N O

O

O O S

CH3

N NH H

CH3

NMe2

HO OH

OH CO2OH O AcHN O O O OH OH HO HO AcHN

OH OH OH O OO O HO HO HO

O OH O HN O

p

OH

CH3

p < 8, q > 1

Ganglioside G2 q

CH3

Figure 2.8 Structure of sialic acid–presenting synthetic and natural lipids.

initially by Whitesides et al. [44], who designed liposome-based multivalent sialic acids as potent inhibitors of influenza virus. The liposomes are made of lipids that terminate in sialic acid residues and that contain either synthetic (10) or natural ganglioside G2 , as shown in Figures 2.8 and 2.9. When examined in the influenza virus–induced agglutination assay, such liposomes inhibit hemagglutination with extreme potency, as illustrated by those containing 10. These molecules show an HAI activity or K iHAI value of 20 nM when calculated on the basis of total sialic acid groups contained in this multivalent system. Such activity represents a 100-fold enhancement compared to the activity (K iHAI @ 200 mM ) of a monomer control (11) which is a close analog of 10 but is incapable of being incorporated into liposome. The lipid 10 is also more active than methyl sialoside monomer by a factor of 10 4 . This level of potent activity is observed from liposomes displaying sialic acid in the density range 2.5 to 18% mole fraction. This optimum range of sialic acid surface density correlates well with earlier measurements by Haywood [93] in which a liposome preparation containing 3% bovine ganglioside acted as a potent inhibitor of Sendai virus in the HAI assay. Thus, liposomes provide a very e¤ective multivalent system, particularly when designing inhibitors against the adsorption of influenza virus to mammalian cells. Liposomes play a valuable role as a model surface for studying the fusion of viruses and cell membranes and understanding the mechanism of viral invasion. To this end, Alford et al. [94] prepared liposomes presenting on their surfaces either ganglioside (GD1a ) or glycoporin, a membrane-bound protein

32

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS OH OH

AcHN HO HO O

CO2-

AcHN HO

O

OH OH

HO O O

CO2O OH

HO

10 NH

Ganglioside G2

O HO O OH

O

AcHN O

Me2N

O OH O OH HO

NH

O O S N H

O HO O OH O HN

O

HO NH

S

n

O O OH NH

n

Membrane

Figure 2.9 Membrane-bound form of sialic acid–presenting lipid.

commonly found on red blood cells. Glycoporin is heavily glycosylated with multiple sialic acid molecules. The glycoporin-bearing liposomes adhere to hemagglutinin-expressing cells through specific interactions involving hemagglutinin and sialic acid, with two or more copies of hemagglutinin trimer required at the fusion site [95]. 2.1.1.5 Polymerized Liposome Presenting Sialic Acid. Despite the fact that liposomes serve as an excellent sca¤old to mimic natural cell membranes, liposomes are potentially unstable and can disassemble reversibly to their lipid monomers. To overcome such an undesired property a multivalent sca¤olding strategy was developed by a group of researchers at the University of California–Berkeley led by Bednarski [96–98]. It is based on the use of polymerized liposomes. Covalently cross-linked lipid bilayers are constructed using synthetic lipid molecules containing a diacetylene functionality in the middle of a lipid chain while presenting a desired ligand, such as sialic acid, on their polar termini. A typical synthesis is illustrated by the two-step preparation of a polymerized liposome made of lipids 12 and 13: (1) sonication of an aqueous suspension of two sialic acid–derivatized lipids mixed at a certain molar ratio; (2) ultraviolet irradiation at 254 nm of the prepared liposome to cross-link lipids assembled in close proximity (Figure 2.10). The resulting polymerized liposome is stable, due to the nature of the poly(diacetylene) structure. It appears ellip-

INFLUENZA VIRUS OHOH CO2-O AcHN

O OH OH

12

O O

N H

N H

3

13

O HO

N H

+

i) sonication (liposome) ii) hν, 254 nm (cross-linking)

AcHN HO

OH OH

HO O CO2-

OHOH AcHN HO

O

HO

HN

O CO2HO O

CO2HN

HN

HO NHAc OH OH O

O

OHOH AcHN HO

CO2-O O

aqueous layer

N H

HO

33

HN O

HO HO NHAc O OH CO2- OH

polymerized liposome

O HO

N H

OHOH CO2-O AcHN

O OH OH

O O

N H

N H

3

O HO

N H O

HO

N H

OHOH CO2-O AcHN

O OH OH

poly(diacetylene)

O N H

O 3

N H

Figure 2.10 Structure of polymerized liposome presenting sialosides.

soidal in shape, is about 40 nm in length, and is about 15 nm in width. This sialyl-presenting liposome polymer is biologically active, demonstrated by its ability to adhere to influenza virus as assessed using the erythrocyte agglutination assay. Its activity depends largely on the surface density of sialic acid, which is defined as 12 o (12 þ 13). Liposomes presenting sialoside at a surface density between 5 and 10% are among the most e¤ective inhibitors of influenza virus, with K iHAI values of 0.57 and 0.33 mM, respectively. By contrast, poly-

34

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS OH OH

AcHN HO HO O HO OH AcHN HO O CO2HO O

CO2-

HO

O HN HN

HO NHAc OH OH O CO2-

O

HN

Blue color (λmax = 626 nm)

HN

polydiacetylene liposome

O

HO HO NHAc O OH CO2- OH

Influenza virus

virus

Red color (λmax = ~550 nm)

polydiacetylene liposome

Hemagglutinin Sialic acid

Figure 2.11 Colorimetric detection of influenza virus by use of polymerized liposome which contains poly(diacetylene) framework that makes it possible to elicit hypersensitive responses to molecular recognition events occurring on surfaces.

merized liposomes displaying a sialyl surface density outside the range 5 to 10% are virtually inactive, even at 10-fold higher concentration. The best activity by the polymerized liposome represents a 30,000-fold increase in potency over a-O-allyl sialoside as a comparable monomer. In one biological application, polymerized liposomes serve as a colorimetric sensor to detect pathogens. This optical application is made possible through the unique properties of the poly(diacetylene) molecules in the internal framework of the liposome (Figure 2.11) [96,97]. The conjugated backbone, with its alternating double and triple bonds, gives rise to intense absorption in the visible light range. The wavelength absorbed by the poly(diacetylene) backbone changes in response to molecular and environmental perturbations such as ligand–receptor binding on the surface, temperature, pH, solvent, and mechanical stress. For example, a polymerized liposome displaying sialic acid (12) at 5% density shows a deep blue color in the unperturbed state. When influenza virus adsorbs to the surface of the liposome through multivalent, specific recognition, the liposome solution changes immediately to a red color (Figure 2.11). The transition is visible and readily quantified by visible absorption spectroscopy. In summary, polymerized liposomes comprise an e¤ective multi-

INFLUENZA VIRUS

35

valent sca¤old for mimicking the cell surface, and they can serve as optical sensors that respond to multivalent binding. 2.1.1.6 Sialic Acid in Langmuir–Blodget Monolayers. Another application of functionalized poly(diacetylene) lipids is to use the Langmuir–Blodgett (LB) technique to form thin films on a solid support (Figure 2.12) [98]. Illustrated in Figure 2.12 is the preparation of an LB film monolayer that displays on its surface a poly(diacetylene) layer consisting of 5% sialic acid. Multivalent presentation of such biological ligands on a glass support in combination with poly(diacetylene) thin film enables this system to be used as an optical probe

OH OH

AcHN HO HO O

OH OH

AcHN HO HO O

CO2-

Multivalent ligands for recognition of pathogen

CO2-

O

O

NH

NH

O

O

O

O

HN

O

HN

OH

OH

OH

O

O

HN

O

HN

O HN

OH

OH

O

O

HN

O

HN

O

Polydiacetylene backbone as optical sensor

Monolayer support

glass support

Figure 2.12 Polymerized bilayer assembly on a glass support where the bottom monolayer of octadecylsiloxane is attached to a solid surface through siloxane linkage.

36


for detecting specific pathogens [99]. For example, a film with a blue color and a maximum absorption at 630 nm might shift to a red color with a maximum absorption of 540 nm upon exposure to influenza virus. This virus–biosurface recognition is reversible and specific, since it is based on the interaction of a hemaggultinin–sialic acid pair. Such a color shift can be impeded by specific inhibitors of viral HA activity, such as a-O-methyl sialoside. By extension, this application of the LB technique may enable the rapid screening of potential drug candidates. 2.1.1.7 Sialic Acid Presented on Biopolymer Surfaces. Displaying sialic acid in a multivalent manner is also possible on the surfaces of biomaterials. Takemoto et al. [100] demonstrated such an approach by covalently immobilizing fetuin—a glycoprotein that contains sialic acid—to the surface of a carboxymethylated dextran sensor chip. The ligand-presenting chip is coupled with surface plasma resonance (SPR) spectroscopy to study multivalent ligand– receptor recognition. This setup can be used to determine dissociation constants under flow conditions. For example, this approach was used to determine the dissociation constant between sialic acid–presenting biosurfaces and soluble hemagglutinin clusters (such as influenza virus A (X-31) rosettes, which contain 6 to 10 hemagglutinin trimers per rosette [101]). The interaction, which mimics the association of virus and cell, leads to an equilibrium Kd value of 1 107 M derived from on and o¤ rate constants (ka ¼ 2 10 3 M 1 /s; kd ¼ 2 104 s1 ). This multivalent binding is four orders of magnitude tighter than the monovalent interaction between an X-31 rosette and sialic acid and is attributable to a slower o¤-rate. 2.1.1.8 Poly(acrylamide) Presenting Sialosides. Mammen et al. [34], Spaltenstein and Whitesides [36], and Lees et al. [47] have demonstrated the practical use of synthetic polymers as multivalent sca¤olds to present bioactive ligands. The utility of such polymeric sca¤olding is illustrated by poly(acrylamide), which is prepared conveniently by free-radical polymerization of acrylamide monomers. For example, copolymerization of a mixture of monomers composed of unsubstituted acrylamide and sialic acid–containing acrylamide (14) yields poly(14-co-acrylamide), which presents multiple copies of sialosides as the side chains of poly(acrylamide) (Figure 2.13). This method is also convenient for preparing a series of polymeric, polyvalent molecules with a range of side-chain densities. Thus, copolymerization at various ratios of 14 to acrylamide monomer leads to the generation of a series of poly(14-co-acrylamide) displaying sialoside density in a range between 0 and 0.9 [side-chain density ¼ wsialic acid ¼ [14] o ([14] þ [acrylamide])]. These sialoside-containing copolymers are highly e¤ective in inhibiting the hemagglutination of red blood cells by influenza virus X-31. In particular, copolymers with wsialic acid of 0.2 to 0.5 are the most e¤ective inhibitors, showing a K iHAI value of about 2 107 M (calculated on a sialoside basis). The polyvalent polymer is more e¤ective than methyl a-sialoside monomer by a factor of about 10 4 to 10 5 . This example

INFLUENZA VIRUS OAc

AcO

HO

Cl AcHN

O OAc OAc

i,ii

HO 14

AcHN

CO2Me

AcHN

CO2O OH OH

O

O

O

O

NHCbz

iii,iv

CO2O OH OH

HN OH

O CO2O

v

H N

OH

NH2 O

O OH OH

CO2O OH OH

O

AcHN

AcHN

OH

OH

HO

HO

37

poly(14-co-acrylamide) (χSA = [14]/ [acrylamide] = 0.2)

O O

HN O

O O O O NH2 NH2 NH2 NH2 HO AcHN

NH2 O

O O

O O O HN NH2 NH2 NH2

OH

O

n O O NH2 NH2

O CO2-

O OH OH

O

reagents and conditions : i) HO(CH2)4O(CH2)3NHCbz, Ag-salicylate, benzene, 25oC, 3 days; ii) 1 M NaOH, 25oC, 12 h; iii) H2, 5% Pd/C, MeOH, 25oC, 6 h; iv) N-(acryloyloxy)succinimide, Et3N, H2O, 25oC, 12 h; v) CH2=CHCONH2, 4,4'-azobis(4-cyanopentanoic acid), hν (365 nm), 25oC, 5 h.

Figure 2.13

Synthesis of poly(acrylamide) presenting a-O-sialoside as a side chain.

demonstrates the e¤ectiveness and practicality of using multivalent polymers for designing influenza virus inhibitors. 2.1.1.9 Poly(acrylic acid) Presenting Sialosides. Matrosovich et al. [102] and Dikusar et al. [103] described a synthetic polymeric sialoside, poly(15-co-acrylic acid), whose polymer framework is based on poly(acrylic acid) (Figure 2.14). The polymer displays multiple copies of sialoside via amide side chains as well as a high density of negative charges along the polymer backbone. This polymeric sialoside e¤ectively inhibits hemagglutination, with a K iHAI value of 1.8 mM. 2.1.1.10 Poly(acrylamide) Presenting C-Sialosides. The a-O-sialoside–based inhibitors of influenza virus are designed by using a natural a-O-linkage. Accordingly, these molecules are chemically labile, particularly at acidic pH. They are also substrates of viral neuraminidase—a glycosidase that is present on the surface of the virus—and are susceptible to O-glycosidic cleavage, leading to the removal of sialoside ligands emanating from the polyvalent framework [104,105]. Paulsen and Matschulat [106], Nagy and Bednarski [107], and Lee and Lee [108] solved that issue by preparing C-glycosidic sialoside moieties

38


O X

O X

O X

O X

X

OH

HO

n O

O

+

O

X = p-nitrophenoxy

AcHN

i) 15

CO2O OH OH

O

O N H

15

NO2

ii) NaOH, H2O NH2

OH

HO AcHN

CO2O OH OH

O NH O

OH O

HN

OH O

O

Poly(15-co-acrylacid) HN O

O

O O O O OH OH OH OH

O O O HN OH OH OH O

NH -O2C

HO O HO OH AcHN OH

O

O

n O O OH OH

NH -O2C

O HO O HO OH AcHN OH

Figure 2.14 Synthesis of poly(acrylic acid) presenting sialoside as an amide side chain.

where the C-linkage is stable to chemical and enzymatic cleavage. The synthesis of such polymeric sialosides begins with the preparation of an N-acryloyloxy derivative (16) of a-C-sialic acid (NeuAc), followed by its copolymerization with acrylamide monomer (Figure 2.15). This method makes it possible to generate copolymers incorporating C-glycosidic NeuAc moieties as side chains with wNeuAc values of about 0.01 to 0.8 [109]. Biologically, the C-glycosidic copolymers are as e¤ective as O-glycosidic copolymers in inhibiting the agglutination of erythrocytes by influenza virus. For example, the optimal K iHAI values for O-glycosidic and C-glycosidic polymers are 0.2 and 0.4 mM, respectively, at side-chain densities ( wNeuAc ) of 0.3 to 0.4. The monomeric a-C-allyl sialic acid is a weak inhibitor of hemagglutination in the HAI assay, with a K iHAI value between 2.5 and 5 mM. 2.1.1.11 Postmodification of Activated Polymers. Making polyvalent sialoside polymers via copolymerization may introduce uncontrollable variables that influence the structure of resulting polymer molecules: (1) uneven distribution of sialoside chains along the polymer backbone, (2) polydispersity—heterogeneity in polymer size, and (3) tacticity—the orientation and stereochemistry of the side chains. Such variables are likely to be generated by di¤erent rates of polymerization among di¤erently substituted acrylamide monomers. Mammen et al. [34] developed a di¤erent approach to control the incorporation of such variables in polyvalent polymers. Their appraoch is based on a postmodification method, which relies on the use of activated polymers, such as poly(Nacryloyloxysuccinimide), that have reactive NHS esters as side chains (Figure

39

INFLUENZA VIRUS

AcO

OAc

AcHN

O OAc OAc

OH

HO

Cl

i,ii CO2Me

AcHN

OH

HO CO2H

iii,iv

AcHN

O OH OH

CO2O OH OH

O S

(and β-allyl isomer)

N H

16

OH OH AcHN HO

O

poly(16-co-acrylamide) χSA = [16] ÷ { [acrylamide] + [16]} = 0.2

CO2-

HO S v

NH2 O HN

HO HO AcHN

-O2C O OH OH

S

NH O

O O O O NH2 NH2 NH2 NH2 HO HO AcHN

NH2 O

O

O O O HN NH2 NH2 NH2 -O2C

O

n O O NH2 NH2

S

O OH OH

reagents and conditions: i) CH2=CHCH2SiMe3, BF3Et2O; ii) NaOH, 25oC, 12 h; iii) aminoethanethiol hydrochloride, 4,4'-azobis(4-cyanopentanoic acid), hν (365 nm); iv) N-(acryloyloxy)succinimide, Et3N, MeOH; v) CH2=CHCONH2, 4,4'-azobis(4-cyanopentanoic acid), hν (365 nm), H2O.

Figure 2.15 chain.

Synthesis of poly(acrylammide) presenting a-C-sialoside as an amide side

2.16). These activated polymers react with amines, and sequential addition of two di¤erent amines yields di¤erently functionalized polyvalent terpolymers. Sialoside-presenting polymers prepared by the postmodification method are potent inhibitors of influenza-caused hemagglutination of chick red blood cells. For example, poly(17-co-acrylamide) at a side-chain density of 35% shows the lowest K iHAI (0.3 nM) among the polyvalent sialosides of various side-chain densities. Based on an enzyme-linked immunosorbent assay (ELISA), this polymeric sialoside binds influenza virus very tightly, with a K iELISA value < 3 nM [110]. The inhibitory activity of poly(17-co-acrylamide) is influenced mostly by the density or mole fraction of the sialoside side chains, and potent activities are observed from polymers presenting densities in the range 20 to 80%. Other factors a¤ecting the activity of polyvalent molecules include molecular weight (Mp ) and polymer chain length (N, the number of monomer units per polymer molecule). The impact of such factors is linearly proportional to the activity if they lie in the range Mp 10 to 1000 kDa and N 100 to 40,000. Polymeric sialosides prepared by the postmodification method are significantly more e¤ective (up to 100-fold) than are their corresponding polymers obtained from copolymerization. 2.1.1.12 Modes of Action of Polymeric Sialosides. Poly(17-co-acrylamide) inhibits influenza virus adhesion through tight association of the polymer to the

40

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS O O

HO

i N O

O O

R

O R

O R

O R

NHS

n O

+

AcHN

R (R = NHS)

OH CO2O OH OH

S

NH2

17

poly(NAS)

N-Acryloyloxysuccinimide (NAS) HO OH AcHN HO

CO2O OH

poly(17-co-acrylamide) χSA = [17] ÷ {[17] + [acrylamide]} = 0.2

S NH O

ii,iii

n O O O O O O O NH NH NH2 NH2 NH2 NH2 NH 2 2

O O O O NH2 NH2 NH2 NH2 HO HO AcHN

-O2C O

S

OH OH reagents and conditions: i) azoisobutyronitrile (AIBN), benzene, 60oC, 24 h; ii) 17, Et3N, DMF, rt, 20 h, then 65oC, 6 h; iii) NH4OH, rt, 12 h.

Figure 2.16 Synthetic method that makes it possible to derive polyvalent copolymer from preactivated polymer.

surface of the virus and, as shown with monomeric inhibitors of influenza neuraminidase (NA), through steric stabilization of the virus (Figure 2.17). The ability of the polymer to inhibit hemagglutination is enhanced by a factor of 2 to 20 when assayed in the presence of Neu2en-NH2 , which is known as a potent and selective neuraminidase inhibitor [35]. For example, the K iHAI of poly(17-co-acrylamide) at wsialic acid of 100% is 200 nM in the absence of neuraminidase inhibitor, but decreases 20-fold, to 10 nM, in the presence of 0.5 mM Neu2en-NH2 . The neuraminidase inhibitor is a selective inhibitor of neuraminidase (K i ¼ 50 nM ) and only a poor inhibitor of hemagglutinin with K iHAI > 50 mM. Therefore, the neuraminidase inhibitor does not contribute directly to the inhibition of virus–cell adsorption. Instead, the enhanced activity results from competitive displacement of sialoside ligands bound at the neuraminidase site by the neuraminidase inhibitor. Through this displacement mechanism and the resulting steric stabilization of virus particles, the bound polymer backbone expands and becomes more e¤ective in preventing virus–cell interactions. 2.1.1.13 OPTCOL Assay. The HAI assay is of limited application when measuring exceptionally high activity, such as that of polyvalent polymeric inhibitors. The lowest measurable value of K iHAI in the assay is approximately equal to the minimal usable concentration of viral hemagglutinin, or the minimum concentration of hemagglutinin necessary to cause agglutination (esti-

INFLUENZA VIRUS

41

NA Influenza virus

HA NA inhibitor

Sialic acid

polymeric polyvalent inhibitor (tight association; steric blocking)

enhanced steric effect

NA inhibitor Inhibition of virus-cell adhesion Sialic acid

cell membrane

HO OH H O AcHN HO NH2

CO2H

Neu2en-NH2 Ki (NA) = 50 nM

Figure 2.17 Summary the modes of action displayed by polymeric polyvalent sialoside as an inhibitor of influenza virus.

mated to be 1 nM). Under such assumptions, all inhibitors with a K i < 1 nM therefore appear to be equally e¤ective and are indistinguishable in HAI assay. Other methods are necessary to determine potent inhibitory activity accurately and precisely. Mammen et al. at Harvard and the National Institute of Standards and Technology [111] developed a new bioassay based on an optical tweezer called OPTCOL (optically controlled collision). In this assay, the probability of adhesion of a single erythrocyte to a single virus-coated microsphere is evaluated in the presence of inhibitor. The technique makes it possible to distinguish the inhibition activity of the best polyvalent inhibitors prepared by the postmodification method. For example, poly(17-co-acrylamide) at an wSA value of 35% shows an inhibition constant K iOPTCOL value of 35 pM and a K iHAI value of 0.3 nM. 2.1.1.14 ter-Poly(acrylic acid) Presenting Sialosides. A primary mode of action of polyvalent sialosides can be attributed to specific multivalent interactions between silaoside and its receptor (hemagglutinin or neuraminidase) on the surface of influenza virus. Introducing a second, nonsialoside functionality to the side chains of the polymer provides an additional binding motif, perhaps through nonspecific binding of nonsialoside functional groups to the receptor. For example, hydrophobic groups, which are incorporated as side chains into the backbone of poly(acrylamide) or poly(acrylic acid) molecules displaying

42


O NH

HO OH

O O O O OH OH OH OH

n

poly(SA-co-acrylic acid) [χSA = 0.1]

CO2Na H 3C

O

S

O

HN

OH OH

O NH

HO OH HN O

O

O O O O OH OH OH NH

n

CO2H

CO2Na H3C

18

19 poly(SA-co-acrylic acid)(naphthylalanine) [χSA = 0.1; χNA = 0.12]

S

OH OH

Figure 2.18 Structure of poly(SA-co-acrylic acid) presenting side chains composed of sialoside and l-naphthylalanine.

sialoside groups, enhance the activity of the polymers in inhibiting virus-caused hemagglutination [112]. One such example is poly(SA-co-acrylic acid) (l-naphthylalanine), 19 (Figure 2.18). This terpolymer shows a potent HAI activity with a K iHAI value of 0.5 nM, an activity about 1000 times higher than that of the corresponding copolymer, poly(SA-co-acrylic acid), 18. As demonstrated by this example, the second functionality can dramatically improve HAI activity and should be considered when designing polymeric polyvalent ligands. 2.1.1.15 Poly(glutamic acid) Presenting Lyso-gangliosides. Kamitakahara et al. [113] conducted an important test of the hypothesis that incorporation of hydrophobic groups in a polymeric sialoside might contribute to enhanced binding as well as disruption of viral adhesion ability. The study is based on a unique polyvalent sialoside, which presents lyso-ganglioside (GM3 ) as amide side chains linked to a sca¤old of poly(glutamic acid) (degree of polymerization ¼ 540). In this polymer, the ganglioside serves as a source of sialic acid as well as providing long, hydrophobic alkyl chains (Figure 2.19). The polymeric, polyvalent ganglioside 20 shows potent inhibition activity with an IC50 value of 1:9 1012 M against influenza virus A in ELISA. It is about 103 -fold more potent than a corresponding monomer such as lyso-GM3 21 or GM3 , and about 105 -fold more potent than sialyllactose. The nanomolar activities of lysoGM3 and GM3 relative to sialyllactose can be attributed to their ability to form aggregates in aqueous solution, mimicking liposomes or micelles, both of which serve as a multivalent sca¤old. 2.1.1.16 Poly(glutamic acid) Bearing Sialic Acid–Containing Trisaccharides. Totani et al. [114] reported the chemoenzymatic synthesis of glycopolymers composed of poly(glutamic acid) bearing side chains of sialic acid–containing

43

INFLUENZA VIRUS 20 Poly(Glu)(Lyso-GM3) HO OH NaO2C H 3C

O

O

HN

OH OH O OH

OH OH

O

OH O O OH O

O HO

OH CH3 NH H N

Lysoganglioside GM3

O

HN O

O

OH OH O OH

OH OH

O HO

OH O O OH

m = 4; n = 270 O

O

OH C13H27

+

NH-Fmoc 5 eq

O

HN O

O

OH OH O OH

OH OH

O HO

OH O O OH O

C13H27 NH H N

NH-Fmoc

H 3C

HN O

O OH OH

O

OH OH O OH

O HO

OH O O OH O

OH

O

NH2

H N

H 3C C13H27

NH H N

N F B N F H3C

N F B N F H 3C

CH3

OH

O

HO OH NaO2C

H N O

NH2

HO OH NaO2C H3C

O

O

21 Lysoganglioside GM3

i

CH3

n

N

HO OH NaO2C H 3C

O m

O

NaO

Fluorescent tag (BODIPY-C3)

ONa N H

O

O

NH

O

H N

H 3C

N F B N F H3C

O

O SuO

ii

N F B N F H3C

CH3

O

OSu O

O

O N H

N

n O m O

O 20 iii

CH3

reagents and conditions: i) Et3N, MeOH, DMF, ~100%; ii) piperidine, DMF, 34%; iii) DMF, MeOH, then NaOH.

Figure 2.19 Synthesis of polyvalent lysoganglioside constructed on the backbone of poly(glutamic acid).

44


trisaccharide. They utilized a2,3- and a2,6-sialyltransferase in the synthesis in order to incorporate sialic acid in a regiospecific manner, thus producing molecules such as NeuAc-a2,6-Gal-b1,4-G1cNAc or NeuAc-a2,3-Gal-b1,4GlcNAc. The polyvalent glycopolymer displaying NeuAc-a2,6-Gal-b1,4GlcNAc shows significant inhibitory activity against infection by influenza viruses A and B in a cell protection assay. Such activity is enhanced upon increasing the molecular weight of the polymer and the density of sialic acid. 2.1.1.17 Poly(acrylamide) Bearing Sialic Acid Linked at the C4 Position. The most common site for introducing a linker to sialic acid is the C2 position; however, an alternative attachment site can be utilized including the 4-N substituent of sialic acid (22). Wu et al. [115] reported the synthesis of polymeric polyvalent sialic acid based on such linkage sites (Figure 2.20). The resulting polyvalent molecules weakly inhibit the influenza-caused agglutination of red blood cells, with a K iHAI value of around 106 M per sialic acid. This example shows that the attachment via C2 position may be preferred to 4-N position, although attachment through either position leads to polyvalent sialosides more potent than a monovalent sialoside.

poly(NAS)

O

O NHS

N

O O

O R

O R

O R

HO

n O

OH

CO2AcHN O S HO HN O

+

R

(R = NHS)

HN

NH4OH NH2

22 OH OH AcHN HO HN HN

O CO2O

S poly(22-co-acrylamide) (χSA = [22]/{[22] + [acrylamide]} = 0.25) NH O n O O O O O O NH2 NH2 NH2 NH NH2 NH2

O O O NH2 NH2 NH2

S

-O2C O

O

NH

NH

HO HO

OH NHAc

Figure 2.20 Synthesis of poly(acrylamide) presenting multiple copies of 4-N-linked sialoside.

INFLUENZA VIRUS AcO OAc

AcO OAc

AcO OAc

Cl H3C

O

HN O

CO2Me i

CO2Me

H 3C

OAc OAc

O

HN O

NO2

S

CO2Me

ii,iii H 3C

OAc OAc

CO2H H 3C

O

HN O

NH

S

O

H3C

O

vii

O H3C H 3C

O

O

OH OH

O O O O O NH NH2 NH2 NH2 NH2

O OH NaO2C HN

O

HN

n

O

S

OAc OAc

NH O

O O O O O NH NH2 NH2 NH2 NH2

HO OH NaO2C

vi

OH OH

HN O

HO OH iv,v

45

S

n

23 poly(SA-co-acrylylamide) [χSA = 0.1]

24 poly(9-O-acetyl-SA-co-acrylamide) [χSA = 0.1]

S

OH OH

reagents and conditions: i) 4-NO2C6H4SH, n-Bu4NHSO4, 1 M NaOH, CH2Cl2, 25oC; ii) NH4HCO2, 10% Pd/C, then SnCl2, EtOH, reflux; iii) CH2=CHCOCl, Et3N, CH2Cl2; iv) NaOMe; v) 0.1 M NaOH; vi) CH2=CHCONH2, (NH4)2S2O8; vii) MeC(OMe)3, TsOH

Figure 2.21 Synthesis of poly(acrylamide) presenting 9-O-acetylsialic acid as an amide side chain.

2.1.1.18 Poly(acrylamide) Tethering 9-O-Acetylsialosides. Unlike influenza viruses A and B, influenza virus C produces a hemagglutinin that recognizes 9O-acetyl-N-acetylneuraminic acid (9-O-acetylsialic acid) as a ligand determinant [116]. Roy et al. [117] reported the synthesis of a polyvalent sialoside (24) based on 9-O-acetylsialic acid as an inhibitor of influenza virus C (Figure 2.21). This polyvalent sialoside is designed on the backbone of poly(acrylamide) by tethering multiple copies of 9-O-acetylsialoside as side chains. In the HAI assay using murine erythrocytes, 24 inhibits influenza virus C at a concentration of 3.3 mM. Such activity is attributed to the specific binding of 24 to the influenza virus C as demonstrated in the microtiter plate assay. The binding ability of 24 is remarkable relative to an analogous polymer prepared from des-9-Oacetylsialic acid (23), which does not show any binding to the virus. This 9-Oacetylated glycopolymer (24) is found to be a poor substrate of 9-O-acetylsialic acid esterase, a hydrolytic enzyme located on the surface of influenza C. Thus, this type of polymeric sialoside represents an esterase-resistant polyvalent inhibitor. 2.1.1.19 Neoglycoprotein Displaying Sialic Acids. A variety of nonspecific molecules present in human and animal sera can inhibit the interaction of influenza virus and sialosides on cell surfaces [118,119]. Most natural inhibitors are glycoproteins that present multiple copies of sialosides on a protein surface.

46


Such natural inhibitors include bovine mucin (K i ¼ 20 mM ), fetuin (K i ¼ 8 mM ), human and bovine a2 M (K i ¼ 1 to 3 mM ), guinea pig a2 M (K i ¼ 50 nM ), and equine a2 M (K i ¼ 10 nM ) [34]. Hanaoka et al. [120] reported that a natural defense molecule known as g-inhibitor is a2 -macroglobulin, whose antiviral activity is attributed to the presence of multiple copies of sialosides displayed on the macromolecule. The proposed mechanism of action by such glycoproteins involves multivalent binding of its sialosides to influenza hemagglutinin. The term neoglycoproteins commonly refers to synthetic protein derivatives with multiple copies of carbohydrate ligands on the surface of a protein scaffold. They are prepared chemically or enzymatically by conjugating multiple ligands to a sugar-free protein, such as serum albumin (human or bovine), through the formation of an amide bond or via reductive amination. Modified BSA 26 is a typical neoglycoprotein (Figure 2.22). It displays up to 58 copies of sialyl-containing heptasaccharide 25 covalently attached to bovine serum albumin through amide linkages [88]. This neoglycoprotein inhibits the adhesion of influenza virus to human erythrocytes with an activity higher than that of divalent heptasaccharide 25, used as control. A second example of a synthetic glycoprotein is the neoglycoprotein 27, reported by Roy et al. [121] shown in Figure 2.23. It can display up to 23 copies

HO

OH Divalent heptasaccharide (25)

AcHN HO

CO2-

O HO

HO

Lysine

O NHAc O O

OHO O

HO HO

OH

HO OO

HO OH AcHN OH HO

OH

-O2C O

OH

Bovine serum albumin (BSA)

OH

HO

OH

AcHN HO

CO2-

O HO

EDC

HO

O NHAc O O

OHO O

HO HO

HO HO

OH

HO OO

HO OH HO

OH

O

H N

OH O O

O

OH AcHN O O

O

-O2C

AcHN OH

Figure 2.22

NH2 n = 58

H2N

+

O

OH AcHN O O

O

HO HO

OH O O

26 OH

Structure of sialic acid–presenting neoglycoprotein.

n

INFLUENZA VIRUS OAc

AcO AcHN

O OAc OAc

iv

O OAc OAc

OH

O AcHN OH OH

v

CO2H O

O H

HO

ii,iii

CO2Me

AcHN

CO2Me

HO

OAc

AcO

i

Cl

AcHN

O

HO AcHN

OH CO2H O OH OH

OH CO2H O OH OH

O

47

O

BSA H N n = 23

27 reagents and conditions: i) CH2=CHCH2OH, silver salicylate; ii) NaOMe, MeOH; iii) 0.1 M NaOH; iv) O3, then Me2S; v) bovine serum albumin, NaCNBH3, phosphate buffer, pH 7.0, 37oC.

Figure 2.23 Synthesis of neoglycoprotein conjugate via coupling of reductive amination.

of sialic acid per protein molecule, coupled through reductive amination to the e-amino residue of lysine. This multivalent sialylated protein e¤ectively inhibits the adsorption of influenza virus to chicken red blood cells [122]. 2.1.1.20 Natural b-Inhibitors. A di¤erent naturally occurring class of hemagglutinin inhibitors exhibits a unique mode of action relative to the a2 macroglobulins. Called b-inhibitors, they belong to a class of proteins found in normal bovine and mouse sera. Unlike the role played by a2 -macroglobulins (g-inhibitor) as a multivalent ligand, b-inhibitors do not target sialic acid– binding sites on hemagglutinin but instead, act as lectin receptors that can selectively recognize mannoside-containing ligands. For example, a viral hemagglutinin (X-31) monomer expresses on its surface a mannoside-containing oligosaccharide attached at Asn-165. The oligosaccharide is located proximally toward sialic acid-recognition sites on a neighboring hemagglutinin (Figure 2.24). Anders et al. [123] suggest that b-inhibitor binds to such mannoside residues on the tip of hemagglutinin molecules and sterically blocks access of the hemagglutinin site to sialic acid on the cell surface, thereby indirectly inhibiting virus–cell adhesion.

2.1.2

Neuraminidase

A second major protein located on the surface of influenza, neuraminidase (NA), exists as a tetrameric cluster with a surface density of 50 copies of tertamer per virion particle. Because of its critical role in viral pathogenicity, NA is the target of many drugs, including marketed drugs, such as zanamivir and oseltamavir, and other discovery-stage inhibitors [124–126]. The topological features of NA make it an attractive target for a multivalent approach.

48

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS inhibitor

HO OH AcHN O O OH AcHN HO

HO

HA1

HO

O O OH OH

OHO O HO

OH HO O O HO

sialic acid-site

OH O O

HO NHAc O O OH

NHAc OH OH

O OH O OH OH

O OH O Asn-165

sialic acid-site

Leu-226

HA2

HA3

O O HO

steric hindrance

OH OHO O HO OH OHO O OH HO O

O NHAc OH

HO NHAc HO

Figure 2.24 Anti-influenza activity by a b-inhibitor.

2.1.2.1 Poly(glutamic acid) Presenting NA Inhibitors. Honda et al. [127] reported the first class of polyvalent inhibitors that target influenza neuraminidase, designed on the basis of a derivative of zanamivir (Figure 2.25). A monomer ligand, 28—a 7-O-alkyl ether derivative—shows an inhibitory activity against sialidase, with an IC50 value of 22 nM, which is close to that of zanamivir (IC50 ¼ 11 to 29 nM ) [128]. Linking multiple copies of 28 to activated poly(glutamic acid) leads to the polymeric polyvalent molecules 29 and 30, in which the side-chain density of the inhibitor is controlled by varying the ratio of 28 to the reactive ester of activated poly(glutamic acid). In a solutionphase enzyme assay, 29 and 30 are active against sialidase, with IC50 values of 114 and 161 nM, respectively, although such activities are five- to sevenfold lower than that of a monomer control (28). However, in a plaque reduction assay that measures the extent to which influenza virus can replicate, the values of IC50 are 66 pM (29) and 72 pM (30). In contrast, unsubstituted poly(glutamic acid) is inactive. These activities represent 76- to 83-fold enhancement relative to 28. The mode of action by these polymers probably involves tight binding of the polymeric inhibitor to viral surfaces by binding multiple neuraminidase molecules simultaneously.

HUMAN IMMUNODEFICIENCY VIRUS OHO

OHOH H O

AcHN

CO2H

AcHN

OH HN H2N

O

NH

ONa

Oseltamivir

O

N H O

i

NH

O

O

H N

ii,iii m

OBt

O

NH2 O

H N O

(CH2)n O

OH HN H 2N

OBt

N H

activated poly(Glu)

O

AcHN

O O

poly(Glu) Mr = 50 - 70 kDa

OHO

OBt

H N

m ONa

CO2H

28 (n = 2,5)

O

H N

NH2

OH HN H 2N

O

H N

O

AcHN

CO2H

ONa

O O

O H 2N

Zanamivir

O

(CH2)n

H

49

NH

N H

CO2H

O

H N

N H O

m NH2

29 (n = 2; χNeu2en = 10%) 30 (n = 5; χNeu2en = 10%)

reagents and conditions: i) EDC, HOBt; ii) 28, pyridine; iii) aq. NH3; then dialysis against water.

Figure 2.25 Synthesis of polyvalent inhibitor that targets influenza neuraminidase.

2.2 2.2.1

HUMAN IMMUNODEFICIENCY VIRUS HIV-1 Protease

The enzymes and structural proteins of human immunodeficiency virus type 1 (HIV-1) play critical roles in producing mature and infectious virion particles [129]. These diverse proteins are first produced in the form of large precursor proteins encoded by the virus. They are subsequently generated after proteolytic cleavage of the precursor proteins by HIV-1 protease. This protease is active enzymatically only as a homodimer [130]. A principal driving force underlying the dimerization of HIV-1 protease is the favorable peptide–peptide association between the N- and C-terminal portions of the protease, which generates a four-stranded b-sheet (Figure 2.26). The homodimeric HIV-1 protease forms a stable complex with Kd values ranging from 39 pM to 440 nM [131,132]. The homodimeric complex contains a substrate-binding pocket that serves as a catalytic reaction site.

50

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS divalent peptide

N-terminal peptide

C-terminal peptide HIV-1 protease monomer (inactive)

HIV-1 protease dimer (active)

protease/divalent peptide complex (inactive)

Figure 2.26 HIV-1 protease dimerization and its inhibition by divalent interfacial peptide.

2.2.1.1 Divalent Terminal Peptides. One multivalent strategy for inhibiting HIV protease is to induce interference at the stage of self-dimerization by disrupting the dimeric peptide interface (Figure 2.26). Monomeric peptides with sequences corresponding to the N- and C-terminuses of HIV-1 protease have been reported that can block protease activity, perhaps by inhibiting enzyme dimerization [133–135]. This approach was elaborated by Babe et al. [136], who designed a contiguous, divalent peptide composed of either N- or Cterminal sequences of HIV-1 protease tethered with a tri(glycine) peptide ˚ long. Examples of such dimers linker. The linker was approximately 3.5 A include [HN-Pro-Gln-Ile-Thr-Leu]-(Gly)3 -[Cys-Thr-Leu-Asn-Phe-OH] and [H2 N-Met-Ser-Leu-Asn-Leu]-(Gly)3 -[Met-Ser-Leu-Asn-Leu-OH], both of which can inhibit HIV-1 protease. In a comparable approach, Zutshi et al. [137] reports cross-linked peptides from the N- and C-termini of the protease as potent inhibitors of HIV-1 protease (Figure 2.27). These divalent peptides are designed based on monomeric peptide sequences which show optimum inhibitory activity: N-terminus sequence ¼ H2 N-Ser-Thr-Leu-Asn-Phe-OH; C-terminus sequence ¼ HN-ProGln-Ile-Thr-Leu-Trp-OH [138]. Two identical (C-terminus) or di¤erent (N- and C-terminus) peptides are linked through aliphatic chains of C12 to C16 as an amide bond to a¤ord homodimers or heterodimers, respectively. The long, ˚ , which is equal to the disflexible alkyl chain is intended to span about 10 A tance separating the N-terminal ends in a dimeric form of the protease (Figure 2.26). In a standard enzyme assay, a series of heterodivalent peptides (31; n ¼ 10 to 16) show antiprotease activities that correlate with the length of the alkane spacer. The linker for optimal activity is observed at (CH2 )14 , with an IC50 value of 2.0 mM. Deleting one amino acid residue from the original sequence of C-terminus (32) or N-terminus (33) reduces the inhibitory activity about threefold, while removing two amino acids simultaneously (34) leads to an inactive peptide. A homodivalent molecule (35), which presents two Cterminal ends is also active, although it is less potent than the corresponding heterodivalent molecule (31; n ¼ 12).

HUMAN IMMUNODEFICIENCY VIRUS

O

N-Pro-Gln-Ile-Thr-Leu-Trp-OH

(CH2)n O

O

N-Pro-Gln-Ile-Thr-Leu-X-OH O

O

NH-Ser-Thr-Leu-Asn-Phe-OH

(CH2)12

(CH2)12


NH-Ser-Thr-Leu-Asn-Y-OH

O


32 (X = Trp; Y = bond) 33 (X = bond; Y = Phe) 34 (X = Y = bond)

31 (n = 10 to 16)

O

HN-Pro-Gln-Ile-Thr-Leu-Trp-OH +

51

N

O

O (CH2)10

O

35

O O

+

N

H2N-Ser-Thr-Leu-Asn-Phe-OH

O

O

O DIEA, DMSO 60oC, 24 h (~5% )

N-Pro-Gln-Ile-Thr-Leu-Trp-OH

(CH2)10 O

31


Figure 2.27 Structure of divalent interfacial peptides and their representative synthesis.

Homodivalent peptides tethered to a conformationally constrained, aromatic sca¤old inhibit the dimerization of HIV-1 protease [139]. Dimers of this class are shown in Figure 2.28, and their inhibitory activities depend largely on their peptide sequence. For example, in an enzyme assay, dimers 36 and 37 block protease activity with IC50 values of 4.2 and 2 mM, respectively. Such activity is attributed to their ability to form a complex with the enzyme at micromolar concentration, therefore preventing assembly of enzyme dimer. Removing one peptide from an active dimer inactivates the molecule, pointing to the importance of divalent association. 2.2.2

HIV-1 Reverse Transcriptase

An HIV-specific type 1 polymerase, HIV-1 reverse transcriptase (RT), catalyzes the synthesis of double-stranded DNA from the template of viral RNA [140]. Inhibitors of the enzyme include 20 ,30 -dideoxynucleoside analogs such as AZT and ddC (Figure 2.29), which are classified as nucleoside-based RT inhibitors (NRTIs). NRTIs competitively block the catalytic activity of RT. A second class of enzyme inhibitors comprises the nonnucleoside-based RT inhibitors (NNRTTs) such as HEPT. Unlike NRTIs, NNRTIs inhibit RT through a noncompetitive mechanism, occupying a hydrophobic pocket located proximal to the enzyme’s catalytic site and interfering sterically with reverse transcriptase activity. 2.2.2.1 Heterodimers Composed of NRTI and NNRTI. Pontikis et al. [141] prepared divalent inhibitors of RT by tethering one molecule of NRTI to an NNRTI molecule through a flexible linker. Such dimers are designed to occupy the distinct but proximal catalytic (NRTI) and NNRTI sites simultaneously. Examples of the dimers are shown in Figure 2.29 along with their representa-

52

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS O Peptide 1

N H

O Scaffold

Peptide 2

O

H N O

General structure of dimeric peptide

CH3 O

H3C H 3C H3CO O

CH3 O

H 3C H N

N H CH3 CH3

O H3CO H 3C

H N O CH3

O CH3

OH O

H3CO O

N H

O

H3C H N

N H H 3C

O

N H H3C

CH3 O

H N

O H3CO

O OH

H 3C H N O CH3

H3C

O N H H3C

OH

H N

O

O CH3

37 O OH

NH-Thr-Ile-Val-OMe

O i, ii

O

N CH3 O

O

N

N H

CH3

36

OH

CH3 O

O iii

N O

N

36

O OH O

NH-Thr-Ile-Val-OMe O

reagents and conditions: i) Ethyl 4-bromobutanoate, K2CO3, DMF; ii) KOH, MeOH; iii) H2N-peptide-OMe, isobutyl chloroformate

Figure 2.28 Synthesis of divalent peptides designed on conformationally constrained sca¤old.

tive synthesis (Figure 2.30). These heterodivalent molecules (39 to 41) can inhibit HIV-1 multiplication in infected cells at micromolar concentrations, although their activities, as reflected in e¤ective concentration (EC50 ) values, are up to about 100-fold lower than those of NRTI or an NNRTI monomer control, such as AZT, ddC, or 38. Such undesired activity can be attributed to several factors, including unfavorable physicochemical and metabolic properties of the dimers, such as high molecular weight, low cell permeability, and a low degree of phosphorylation of the nucleoside. Other factors may lower the a‰nity of the heterodimers to RT, such as inappropriate covalent attachment of, and nonoptimal spacer length between, the NRTI and NNRTI units. The impact of these factors on activity is illustrated by the dimer 41 which has

HUMAN IMMUNODEFICIENCY VIRUS NRTI

NNRTI

O

HO

O N O

CH3

HN

N

N

O

O N O

HO

O

O

NH2 CH3

HN

53

O

S

N

H3CO

O

HO

C2H5

HN

CH3

S CH3

O N3 38

HEPT

ddC

AZT

NRTI

NNRTI

Linker

O C2H5

HN O

O

HN

N (CH2)n H

C-5

O N HO O

CH3

O 39 (n = 4, 10) O

N3

C2H5

HN O N-3 H 3C HO

CH3

S

N

O

H N

O

N N

N (CH2)10 H O

H N

N

CH3

S CH3

O

40

O

O

O N3 N-4 HN (CH2)12 N HO

Figure 2.29

C2H5

HN

41

O

H N

N

CH3

S

O

CH3

O N O

Structure of divalent inhibitors of HIV-1 reverse transcriptase.

linkage at the N4 position rather than at the N3 or CS position. It is about fiveto 10-fold more active than 39 or 40. 2.2.3

Glycoprotein 120 (gp120) on Viral Surfaces

Infection of T-cells by HIV-1 virus is initiated by the interaction of gp120 on the surface of the virus with the CD4 protein on a T-cell membrane (Figure 2.31) [142]. Infection also relies on a T-cell coreceptor, either CCR5 or CXCR4, in the viral attachment stage. Such initial adhesion depends on the interaction of multiple receptor–ligand pairs and is a prerequisite to the adsorption of a second viral protein gp41 to the T-cell surface. This process initiates membrane fusion, which is a critical step in the cell entry mechanism (not shown). This initial virus–cell recognition is believed to involve very tight

54

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS O

O C2H5

HN O

CH3 i, ii

N H

C2 H 5

HN

S

N

O HO

CH3

CH3

S CH3

O

iv NH-(CH 2)12-NH2

O HN

N iii

O N HO O

O N HO O O C2H5

HN N

S CH3

O

N HO

O

H N

HN

CH3

N

O O

41

reagents and conditions: i) Br(CH2)4CO2Me, K2CO3, DMF, 60oC (19%); ii) 1 M NaOH, acetone (94%); iii) (2,4,6-triisopropylphenyl)sulfonyl chloride, Et3N, DMAP, CH2Cl2; then H2N(CH2)12NH2 (33%); iv) 2-chloro-4,6-dimethoxy-1,3,5-triazine (CDMT), N-Methylmorpholine (NMM), DMF (69%).

Figure 2.30 Synthesis of NRTI-NNRTI heterodimer.

HIV viral core viral RNA

gp120/41 trimer gp41 gp120 CD4

CD4

T cell membrane Co-receptor (CCR5 or CXCR4)

Figure 2.31

Specific adhesion of HIV to T-cells.

HUMAN IMMUNODEFICIENCY VIRUS O

O O

N

55

O

O

i S

S

N

+

N

HS ovalbumin

O

O

S

S

O

O

ii CD4 peptide

N H

S

S n (= 3-5)

CD4 (41-84)-ovalbumin CD4 (40-57)-ovalbumin CD4 (81-92)-ovalbumin

reagents and conditions: i) N-succinimidyl 3-(2-pyridyldithio)propionate (24 eq), phosphate buffer, 25oC, 1 h; then gel filtration chromatography; ii) CD4 peptide (5-20 eq), phosphate buffer, 25oC, 2h: CD4 peptide (41-84): GSFLTKGPSKLNDRADSRRSLWDQGNFPLIIKNLKIEDSDTYIC

Figure 2.32 Synthesis of CD4 peptide-based multivalent ligand presented on the surface of ovalbumin.

binding, based on the requisite multivalent interaction between an epitope on the C3 region of gp120 and a CD4 domain encompassing amino acid residues 40 to 57 [143,144]. 2.2.3.1 Neoglycoprotein-Displaying CD4 Peptide. Considering the specific nature of the HIV-cell interaction based on multiple pairing of the gp120–CD4 association, a multivalent molecule presenting ligands derived from the CD4 domain is predicted to bind to a viral surface presenting gpl20. Ghetie et al. [145] reported CD4-derived multivalent ligands prepared via the covalent tethering of three to five copies of CD4 peptide (sequence 40 to 57, 41 to 84, or 81 to 92) to an ovalbumin protein sca¤old (Figure 2.32). The CD4 peptide– ovalbumin conjugates can prevent recombinant CD4 protein from binding gp 120 with an avidity 10- to 20-fold greater than the a‰nity of monovalent peptide. For example, the inhibitory activity displayed by CD4 (41–84)– ovalbumin (IC50 ¼ 30 nM ) is about 20-fold more active than that of the CD4 (41–84) peptide, which translates to an increase of four- to sevenfold per peptide. The activity of the multivalent CD4 peptide depends on using the specific peptide sequence of the CD4 domain, supporting the supposition that residues 40 to 57 are required for CD4 to bind to gp120 [143,144]. 2.2.3.2 Galactosyl Ceramide Immobilized on Viral Surfaces. As discussed above, the CD4 surface receptor is a principal mediator of cellular attachment to HIV, although several lines of CD4-negative cells are nonetheless susceptible to infection by certain HIV strains. These conflicting observations suggest that more than one type of cellular receptor may play a role in viral invasion, thus providing an alternative mechanism for viral entry [146,147]. Galactosyl ceramide (Gal cer; Figure 2.33), expressed on the membranes of CD4-negative neural cell lines, binds to gp120 and probably provides an alternative route for virus attachment [148]. Harouse et al. [149] devised a binding assay based on high-performance thin-layer chromatography (HPTLC) using immobilized

56

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS Galactosyl ceramide OH HO

OH

OH OH

O

O

O

HO

O

OH

OH OH

HO

O

HN OH

O

O

OH OH

O

HN OH

O

HN OH

lipid bilayer

HO OH O HO OH

O N H

O O

O

S

O O

n

42 (n = 1, 2, 3)

O

Figure 2.33 Structure of galactosylceramide (Gal cer) presented as glycolipid in membrane and synthetic analogs of Gal cer.

glycolipids (Gal cer, GM1 , GD1a ) and measured the a‰nity of such multivalent glycolipids to gp120. The gp120 binds selectively to Gal cer (and galactosyl sulfatide), but not to glucosyl ceramide or other glycolipids. The value of the dissociation constant of Gal cer, 11.6 nM, is two- to threefold higher than that of gp120 to the CD4 receptor, which has a Kd value of 2 to 5 nM. Gal cer binding is confirmed further by studies using the anti-Gal cer antibody, which blocks the internalization and infection of HIV in CD4-negative cells [150]. The strength and specificity of the interaction implicate Gal cer as a specific cell surface ligand involved in viral entry to CD4-negative cells. In a di¤erent approach to identify the alternative mode of HIV infection via a non-CD4-dependent route, Conboy et al. [151] investigated the binding a‰nity of recombinant gp120 to naturally occurring glycosphingolipids (GSLs) displayed on the surface of planar lipid bilayers. Among the GSLs studied, galactosyl ceramide binds gp120 tightly; in contrast, lactosyl ceramide and glucosyl ceramide bind weakly. Synthetic analogs of galactosyl ceramide can recognize gp120, provided that they are presented on a planar membrane surface (Figure 2.33) [152]. These analogs consist of a galactosyl moiety that is tethered to the polar end of dioleoyl maleate through a spacer arm consisting of ˚ per ethylene mono-, di-, or triethylene glycol (supplying approximately 3 A


57

glycol). When Gal cer (42) is converted to a multivalent form by reconstitution at 5 mol% density in a planar membrane composed of 1,2-dioleoylphosphatidyl choline (DOPC), it binds to recombinant gp120 with a K a value in the range 2 to 5 10 6 for the spacer lengths tested. However, the amount of bound gp120 on the surface of the membrane is a¤ected by the spacer length as a consequence of di¤erent degrees of steric accessibility at the interface. For example, the Gal-presenting membrane that displays 42 (n ¼ 1) shows low surface coverage of gp120 at a level only twofold higher than that of nonspecific binding. Increasing the spacer length from n ¼ 1 to n ¼ 3 leads to a four- to sixfold increase in surface protein coverage compared to that of nonspecific binding. These data suggest a threshold spacer length that enables the galactosyl ligand to get into the carbohydrate-recognition region within gpl20. The binding to gp120 of Gal cer expressed in planar lipid bilayers is positively cooperative, with a cooperativity coe‰cient of about 1.5, and thus is much tighter than that of monomeric Gal cer [152]. 2.2.3.3 Multivalent Anions. The third variable region (V3) of gp120 has been implicated as a specific binding motif to the Gal cer ligand. This hypothesis is supported by several studies, including one in which monoclonal antibodies raised against the V3 region were able to block the binding of gp120 to multivalent Gal cer [153]. Thus, ligands that bind selectively to the V3 region should be e¤ective in blocking interactions between gp120 and Gal cer. For instance, suramin, a sulfonylated dimeric naphthalene derivative (Figure 2.34), can bind to gp120 (V3 ) as assessed by binding studies with tritium-labeled suramin. It inhibits the infection of colonic epithelial cells by HIV-1 at an IC50 value of 54 mg/mL [154]. NF036, a monovalent analog of suramin, is 100-fold less active in preventing viral infection, thus suggesting that the inhibitory activity of suramin derives from its divalent nature. A number of suraminlike polyanionic molecules can inhibit gpl20/CD4mediated adhesion of HIV to T-cells. Polyvalent anions 43 prepared from ‘‘telomerization’’—a type of chain reaction that produces a limited length of polymer (‘‘telomer’’)—of N-alkylated acrylamide monomers (Figure 2.35) [155] show defined molecular weights and present multiple negative charges in a ratio of 6 to 50 per polymer molecule. They block HIV-induced cytopathicity in CEM-4 cells. For example, 43 (n ¼ 23) shows an IC50 value of 1 mg/mL, an O

O

NaO3S HN

NaO3S

N NH SO3Na

HN

CH3 O

SO3Na

H 3C O

N H

SO3Na

NaO3S SO3Na

N H

Suramin

Figure 2.34

Cl NH SO3Na

O

NaO3S HN

NF036

Structure of suramin and its monomer NF036.

58

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS -O2C -O2C NH

O

O

N

Et "Telomerization" i) CH3(CH2)11SH (5 to 110 eq) AIBN, MeCN, reflux

Et C12H25

N

O

S

n O

N

Et

O

N

Et

43 (n = 1, 8, 15, 23)

ii) 2 M NaOH

HN

O CO2Me

CO2Me HN

O HN CO2-

CO2-

O CO2-

CO2-

Figure 2.35 Structure of polyvalent anionic surfactant prepared by telomerization of acryl amide monomer.

activity that is 20- to 40-fold higher than that of 43 (n ¼ 1, 8, or 15). This indicates that the inhibitory activities of the multivalent anions are influenced by the valency number, with higher valencies leading to greater activity. A di¤erent type of sca¤old, the polymerized micelle, enables the polyvalent display of anions (Figure 2.36). As illustrated by 45, polymerized micelles are prepared by g-polymerization of an aqueous micellar solution derived from undec-10-enoic acid (44) [156,157]. The advantage of this type of polymerization lies in the generation of polymers that have relatively small molecular weight (6 to 10 kDa) as a result of a low degree of polymerization. Such a beneficial result is possible since the number of monomers aggregated in the micelle is relatively small and the micelle stays intact during the progress of polymerization. The synthetic polyanion 45 can inhibit the replication of HIV in a cell culture assay with an IC50 value in the range 0.1 to 3.6 mg/mL. Multivalent anions are also designed on the basis of other sca¤olds, such as carbohydrate-derived oligomers (Figure 2.37). For example, molecules 46 and 47 are prepared by tethering multiple copies of monomer to multiple hydroxy groups present in cellobiose and g-cyclodextrin, respectively (Figure 2.38). This synthetic approach enables Leydet et al. and Schols et al. to prepare multivalent anions that present negative charges in the range 16 to 48 charges per molecule and are structurally better defined than polymer-based multivalent anions [158–160]. The multivalent anions 46 and 47 show strong inhibitory

59


-O2C

-O2C CO2-

-O2C CO2-

-O2C

-O2C 44

CO2-

-O2C CO2CO2CO2CO2-

-O2C

-O2C CO2-

-O2C CO2-

γ-Polymerization

-O2C CO2-

45

-O2C CO2CO2CO2CO2-

Figure 2.36 Polymerized micelle as a sca¤old for presentation of polyvalent anion.

activity against HIV-induced cytopathicity in MT-4 cells with IC50 values of 0.04 and 0.1 mM, respectively. Such activity is specific against HIV and relatively inactive against other viruses, including influenza, respiratory syncytial, and herpes simplex viruses. 2.2.3.4 Bivalent Antagonists of CXCR4. The chemokine receptors are another class of cellular surface proteins that play an essential role in HIV–cell adhesion. The class includes CCR5 and CXCR4, which are G-protein-coupled receptors (GPCRs) variably expressed on the surface of T-cells (Figure 2.31). Both are capable of interacting with HIV gp120 as coreceptors with CD4. Bicyclams, a family of bivalent azacrown molecules, are specific antagonists of

60

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS -O2C

S

S O

CO2-O2C

O O

O O

S

O

O O

O

CO2-O2C

S

S CO2-

O

S

O

S -O2C

CO2-

-O2C

cellobiose

-O2C

CO2- S

CO2CO2-

CO2CO2-

CO246

OR O

O

RO

OR

ORO

RO

O

NaO2C

CO2Na

O RO

R

O RO

γ-cyclodextrin

O O

S

O RO

OR

RO

OR

O RO

O OR

RO

OR OR O OR

RO ORO

O RO

RO O

O

S

CO2Na CO2Na

O

O OR

47

Figure 2.37 Multivalent anions derived from carbohydrate sca¤olds.

CXCR4. They include AMD-3100 and AMD-7049, both of which are composed of two identical cyclam units linked with a xylyl group (Figure 2.39). Bivalent molecules such as these protect cells from HIV infection [161], presumably by interfering with the interaction of gp120 and CXCR4. 2.2.4

Surface Carbohydrates on HIV

2.2.4.1 Oligomannose Sugars Present on gp120. The HIV surface presents a plethora of carbohydrates that are expressed in multiple copies. These sugars serve as viral surface ligands that are probably targeted by multivalent receptors. This notion is illustrated by cyanovirin-N (CVN), a mannose-binding natural protein of 11 kDa that is isolated from a cyanobacterium [162]. It is a potent inhibitor of all strains of both HIV and simian immunodeficiency virus (SIV) and is involved at the stage of gpl20-mediated virus–cell fusion [163]. The antiviral activity of CVN is attributed to its strong association to gp120 through its interaction with N-linked oligomannose sugars present on gp120 (Figure 2.40). Bewley et al. [164] suggested that each CVN molecule presents two distinct mannose-recognition sites that are located on opposite sides along the long axis of CVN. Each of the sites selectively recognizes a Man-al,2-linked

61

HUMAN IMMUNODEFICIENCY VIRUS OR O O OR OR RO RO O RO O OR O RO OR O O

RO

O RO

S

O S

O S

O

S CO2Na

CO2Na CO2Na

S

S

O

ii,iii CO2Na CO2Na

O

O RO

RO RO ORO

O

NaO2C

γ-Cyclodextrin (O-allyl): R = -CH2CH=CH2

O

OR O OR O OR O O OR

CO2Na

NaO2C

O O RO RO O OR RO O OR RO OR O

CO2Na

NaO2C

i

RO OR γ -CD OR O RO OR O O O OR RO RO O OR RO ORO OR O O RO

R = (CH2)3SCH(CO2Na)CH2CO2Na NaO2C

γ-Cyclodextrin: R = H

RO O

O

S

CO2Na CO2Na

47

O

O O

S

CO2Na CO2Na

O

CO2Na reagents and conditions: i) allyl iodide, NaH, DMF; ii) mercaptosuccinic acid, AIBN, THF, hν (254 nm); iii) NaOH, size exclusion chromatography (Sephadex gel G25)

Figure 2.38 Synthesis of multivalent anion tethered to cyclodextrin.

Man residue. The binding of each mannose-recognition site to the Man-a1,2Man disaccharide is not identical: The first binding leads to a high-a‰nity association, with a Ka value of 7:2 10 6 M 1 , while the second shows a lower Ka value of 6:8 10 5 M 1 [164]. Such sequential interaction indicates negative cooperativity. Nevertheless, with its two mannose-binding sites, CVN can bind gp120 with a nanomolar a‰nity constant, which probably results from association with multiple mannose ligands expressed on the gp120 trimer. While the structural details of bivalent binding remain to be established, two modes of

NH N

N HN

NHHN

NHHN AMD-3100

NH NH

N

N

HN HN

AMD-7049

Figure 2.39 Structure of AMD-3100 and AMD-7049 as a class of CXCR4 inhibitors.

62

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS Cyanovirin-N (CVN) high affinity site

low affinity site

intra-trimer inter-trimer

Man GlcNAc

gp120 trimer

enveloped virus (HIV)

Figure 2.40 Modes of binding of cyanovirin-N to oligomannose ligand present on HIV gp120.

bivalent association, between two gp120 trimers or within one gp120 trimer, have been suggested, as shown in Figure 2.40. 2.2.4.2 Modes of CVN Recognition. CVN shows unique structural and binding features distinct from those of most plant lectins, which exist as homodimers or homotetramers displaying a single carbohydrate recognition site per monomer [165]. Shenoy et al. [166] provided a thermodynamic basis for interaction between CVN and two di¤erent classes of al,2-linked oligovalent mannosides in water (Figure 2.41). First, a linear mannose ligand (48) binds to CVN in a 2:1 complex. In this model, the first binding is tighter than the second; the first dissociation constant, K d1 , of 3.48 mM is about 13-fold lower than the K d2 value from the second association. This di¤erence is not due to the di¤erence in enthalpic contribution because binding of 48 at either the first or second site leads to similar values of binding enthalpy. The di¤erent a‰nity at the two sites is due, rather, to the relatively unfavorable entropy change for the second binding event. Second, the binding of a trivalent antennary mannose ligand (49) to CVN leads to the formation of an overall 1:1 complex model in which one copy of branched 49 occupies two sites of the CVN molecule. Such binding generates two dissociation constants that di¤er approximately 10-fold: K d1 ¼ 0:27 mM; K d2 ¼ 2:70 mM. The overall binding a‰nity of 49 is about 10fold greater than that of 48 to CVN. The enhanced a‰nity of 49 can be attributed to a greater enthalpic contribution from the first binding event and to the gain in entropic contribution from the second. CVN binds the mannose-presenting gp120 with an equilibrium Ka value of 2:4 10 7 M 1 and an average stoichiometry of two molecules of CVN to one molecule of gp120. A similar binding feature is observed with respect to a‰nity

ROTAVIRUS HO OH O HO HO HO OO HO HO HO OO HO HO O

63

HO HO OH OH HO OH O O HO O HO HO HO HO HO HO HO HO OO OO OO HO HO HO HO HO HO HO HO OO OO O HO HO HO HO O OO HO

48

O

49

CVN

2:1 complex

CVN

1:1 complex CVN

CVN

n

Kd1 = 3.48 µM; Kd2 = 46.10 µM

Kd1 = 0.27 µM; Kd2 = 2.70 µM

∆H1 = -8.07 kcal/mol; ∆H2 = -8.76 kcal/mol T∆S1 = -0.51 kcal/mole; T∆S2 = -2.75 kcal/mol

∆H1 = -22.80 kcal/mol; ∆H2 = -0.38 kcal/mol T∆S1 = -13.71 kcal/mole; T∆S2 = 7.33 kcal/mol

Figure 2.41

Interaction between oligomannosides and cyanovirin-N.

and stoichiometry when CVN binds to a synthetic Man8 GlcNAc2 (51), a branched antennary mannose ligand with two CVN recognition sites (Figure 2.42): K a ¼ 5:4 10 1 M 1 . Another multivalent ligand, Man9 GlcNAc2 (50), binds CVN more tightly, with a Ka value of 1:3 10 8 M 1 and three CVN binding sites. The high specificity of CVN binding to Man8 GlcNAc2 and Man9 GlcNAc2 is associated with a multivalent binding feature: the presence of multiple copies of Man-a1,2-Man-a residues presented as terminal sugars on a dendritic sca¤old. By contrast, a lower-valency ligand such as Man6 GlcNAc2 (52), which displays only one terminal Man-a1,2-Man residue, shows a lower a‰nity to CVN.

2.3

ROTAVIRUS

Adhesion of the animal rotavirus to epithelial cells is mediated by the binding of a rotavirus receptor to a cell surface ligand comprised of sialyl-containing oligosaccharides expressed as various forms of glycoconjugates [167–169]. Like the influenza hemagglutinin–sialic acid interaction, such viral adsorption is based on tight binding driven by multivalency. Thus, monovalent sialoside serves as a selective ligand but binds rotavirus only weakly, with inhibition constant values larger than the millimolar concentration. Koketsu et al. [170] reported tighter binding of an inhibitor displayed on a liposome (Figure 2.43). This liposomal preparation is derived from 53, a phospholipid containing sialic

64


Man-α1,2-Man

HO HO HO HO HO HO

Man-α1,2-Man

Man-α1,2-Man

OH O OO

O HO O OH OO Man9GlcNAc2 (50) O HO HO O HOOHHO HOOH O HOO O OH OO O HO O O HO O O O O HO O HO HO OH NH O AcHN H HO HO OHHO NHAc HO HOOH

Man-α1,2-Man

HO HO HO HO HO HO

OH O OO O OH O

HOO HO O

HO HOOH HOO OO

O OH O HO O HO

OO HO O Man-α1,2-Man HO HOOHHO OHHO HO OH HO

HO HO HO

Man8GlcNAc2 (51)

O HO OO AcHN H

O O

NH NHAc

OH O O

HO O HO O HO HO OH HOO

OH O

OO HO O

Man-α1,2-Man HO HOOHHO OH HO

Man6GlcNAc2 (52)

O OH O HO O HO

O HO OO AcHN

O O

NH NHAc

Figure 2.42 Structure of N-linked multivalent oligomannosides (Man9 GlcNAc2 , Man8 GlcNAc2 , Man6 GlcNAc2 ).

acid on its end, and its mean diameter is estimated to be about 200 nm. It prevents cellular infection by human and simian rotaviruses with IC50 values of 4.35 and 16.1 mM, respectively, while those of a-methyl sialoside are 4.28 and 142 mM, respectively. Therefore, the inhibition of rotaviruses by a multivalent sialyllipid is about 103 - to 10 4 -fold more e¤ective than that of a monovalent ligand.

2.4

POLYOMA VIRUS

Polyoma virus is known to adhere to cell surfaces by recognizing a sialic acid ligand [81]. This interaction is inhibited competitively by a bivalent sialoside ˚ . This bivalent molecule linked with an oligo(glycine) spacer of roughly 57 A

PICORNA VIRUS O

O O

C15H31

O

O

OO P O

O

i

O + (H3C)3N

C7H14CH=CHC8H17

O

HO

OO P O

C7H14CH=CHC8H17

O

O Cl

AcHN

C15H31 O

OAc

AcO

65

O

O CO2Bn OAc OAc

AcO AcHN

ii

OAc

C15H31 O

CO2Bn O O OAc OAc

O

OO P O

O

C7H14CH=CHC8H17

O C15H31

O iii,iv

HO AcHN

OH

O CO2H O

OH OH

O

O

OO P O

O

C7H14CH=CHC8H17

53 reagents and conditions: i) 1,8-Octanediol, phospholipase D, H2O, EtOH; ii) AgOTf, CHCl3, iii) KOMe, MeOH; iv) H2, Pd/C

Figure 2.43 Synthesis of sialic acid–presenting phospholipid.

inhibits the agglutination of erythrocytes caused by polyoma virus with an activity about 500-fold greater than that displayed by a monomeric a-O-methyl sialoside [86].

2.5

PICORNA VIRUS

Picornavirus belongs to a family of nonenveloped RNA viruses encased in capsid proteins (VP1 to VP4) [171]. Other members of this virus family, such as poliovirus and coxsackievirus, include rhinovirus, hepatovirus, and enterovirus. A certain domain on the surface of this family of viruses contains broad depressions or canyons that play a major role as the adhesion of the virus to the cellular ligand [172,173]. The base of the canyon has an opening leading to a hydrophobic pocket made of the VP1 protein, which is occupied by a long, thin molecule, such as sphingosine, palmitate, and myristate, called a pocket factor (Figure 2.44). Tsang et al. [174] reported heteroaromatic ligands that resemble heterobivalent molecules and that are designed to bind a hydrophobic core of the VP1 capsid protein. Each of the molecules, such as 54 and 55, is composed of two heteroaromatic rings tethered by a short, nonflexible, 2-butene or pyridylmethyl linker such that each aryl domain occupies a subsite within the hydrophobic pocket [175]. Thus, 54 and 55 can e¤ectively inhibit picornavirus with minimal inhibition concentration (MIC) values of 3.2 and 0.26 mM, respectively, in a cell-based cytopathic assay. An analogous ligand, R-78206, is far more active, with an MIC value of 8 nM. It is also active against rhinovirus,

66

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS sphingosine (pocket factor)

Tethered

CH3 N S

S

N

N S

binding pocket of viral capsid

Cl

aromatic domain I aromatic domain II

CH3 N N

S

S

H3C N

N S

S

N

N

Cl

S

N S

55

54 O O

N

N

N

O R-78206

Figure 2.44 Hydrophobic pocket within picorna VP1 capsid protein and its complex with a pocket factor (hydrophobic linear alkane) or heterodimeric inhibitor.

with an IC50 value of 0.55 mM. Divalent linkage clearly enables the inhibitory activity of these molecules since untethered aromatic moieties comprising the inhibitors are inactive.

2.6

RESPIRATORY SYNCYTIAL VIRUS

Human respiratory syncytial virus (RSV) is a member of the paramyxovirus group of enveloped viruses [176]. Viral replication is inhibited by molecules that can interfere with cell-virus adsorption. Inhibitors of this type include heparinlike polysulfates and polysulfonates [177]. Viral replication is also inhibited by a di¤erent class of molecules known to interfere with virus–cell fusion mediated by a viral fusion protein. Examples of these are represented by 56 and 57, shown in Figure 2.45 [178–180]. These molecules are composed of a dianionic sca¤old domain and four benzenesulfonamides appended by eight arms that terminate in a primary carboxamide. They are active against human RSV with IC50 values of 0.05 mM (56) and 0.15 mM (57). Studies of both the structure–activity relationships and the biophysical measurements of these molecules attribute their activity to multiple structural and functional features. These include their ability to form multiple pairs of hydrogen bonds (due to the amide functionality), as well as electrostatic interactions (from anionic spacers) and hydrophobic interactions that facilitate the strong association with the

67

DENGUE VIRUS H2NOC(H2C) O N S H2NOC(H2C) O

NaO3S

NH

HN N

N N

NH

N H

N NH

HN O (CH2)CONH2 S N (CH2)CONH2 O

56

H2NOC(H2C)3 O N S H2NOC(H2C)3 O

NaO3S

NH

HN NH

N

N H

N

HN

NH

H2N

O (CH2)CONH2 S N O (CH2)CONH2

N N

SO3Na

H2NOC(H2C)3 O N S H2NOC(H2C)3 O

i

O (CH2)3CONH2 S N O (CH2)3CONH2

N

N

O (CH2)3CONH2 S N O (CH2)3CONH2

57

i O (CH2)CONH2 S NH O

O2N

O S Cl O

N N

SO3Na

H2NOC(H2C) O N S H2NOC(H2C) O

O 2N


HN

iv

i

O2N

N N

iii

O (CH2)CONH2 S N O (CH2)CONH2 NaO3S

N Cl


+

HN O (CH2)CONH2 S N O (CH2)CONH2

NH2

H 2N

v

SO3Na

reagents and conditions: i) Glycinamide, NaHCO3 (78%); ii) bromoacetamide, NaH, DMF (77%); iii) Fe, AcOH (76%); iv) 2,4,6-trichlorotriazine, buffer, pH 7, 0oC to 60oC (80%); v) DMSO, phosphate buffer, NaOH, microwave, 105oC, 1 h (10%)

Figure 2.45 Synthesis of fusion inhibitors of respiratory syncytial virus.

fusion protein. To bind the fusion protein as well as to show antiviral activity, molecules of this class must present at least three benzenesulfonamide appendages [181].

2.7

DENGUE VIRUS

The pathogenic dengue virus utilizes multiple copies of its envelope protein when attaching to the surfaces of target cells [182]. The viral surface protein binds with high a‰nity and specificity to a highly sulfated form of glycosaminoglycan, such as heparan sulfate, on the cell surface. This adhesion mechanism suggests the use of soluble, sulfated glycosaminoglycans and other polyanions as potential inhibitors against viral infection. Examples of poly-

56

68

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS OSO3O -O3SO O O OSO3O SO -O 3 -O3SO -O3SO OSO3O -O3SO OOSO3O OSO3O OSO3OSO3-O SO 3 -O3SO OSO3O O OSO3- OSO - O 3 OSO3OSO3O O O OSO3-O3SO Cyclodextrin sulfate (58)

-O3SO -

O3SO

OSO3OSO3OSO3NH O -O3SO -O SO O OSO3-3

OSO3-

OSO3OSO3O OSO3OSO3O OSO3OSO3Lactobionic acid sulfate (59) -O3SO NH

-O3SO O O O -O3SO RO RO O O O -O2C OSO3- R'HN O RO RO n -O2C OSO3- R'HN O R = H: Heparin R = SO3-; R' = SO3- (86%) & Ac (24%): Sulfated Heparin (60)

Figure 2.46 Structure of polyanions as an inhibitor of dengue virus.

anionic inhibitors (Figure 2.46) include suramin, sulfated b-cyclodextrin (58), sulfated lactobionic acid (59), and natural polyanionic polysaccharides such as heparin sulfate (60). Highly sulfated heparin decasaccharide e¤ectively blocks the binding of dengue virus envelope protein to the cell surface with a Kd value of 15 nM [183]. In a competition plate assay, O-sulfated heparin (60) shows a strong avidity to the viral envelope protein, with an IC50 value of 7.6 nM, which indicates a tighter association than that of densely sulfated heparin (760 nM) and suramin (700 nM) [184]. Therefore, inhibition of the dengue virus by polyanionic molecules is probably based on multivalent interaction, whose strength is controlled by three molecular parameters: the size of inhibitor, ˚ in length, a high density of anionic charge, which must be greater than 39 A and the structural flexibility of the inhibitor.

2.8 NUCLEIC ACIDS OF VIRUSES 2.8.1

RNA–Protein Interactions

Replication of HIV-1 is regulated by two types of RNA–protein interactions (Figure 2.47). The first interaction occurs between the transactivator protein

NUCLEIC ACIDS OF VIRUSES Rev recognition sequence ( C U U CU G GGCCAGA GAGC G CCGGUCU CUCG G A

GGCUGGU CCGACCG

A

UG

69

)

GG

AC G A

CGCAGCGUCA U

GCGUCGCAGU G A

A

TAR (HIV RNA) RRE domain II (HIV RNA)

Tat49-57

Ala-Ala-Ala-Arg-Lys-Lys-Arg-Arg-Gln-Arg-Arg-Arg-Ala-Ala-Ala-Cys

Rev34-50

Thr-Arg-Gln-Ala-Arg-Arg-Asn-Arg-Arg-Arg-Arg-Trp-Arg-Glu-Arg-Gln-Arg

Figure 2.47 Secondary structure of TAR and RRE fragments of HIV RNA and RNAbinding peptide sequences from Tat and Rev protein.

(Tat) and the transactivator-response RNA element (TAR) in RNA. A second type of regulatory interaction is formed when a Rev regulatory protein (Rev) binds to a Rev response element (RRE). RRE belongs to a short nucleotide domain located within the HIV-1 RNA. Serving as a recognition domain for regulatory Rev protein, RRE is responsible for controlling the active export of the HIV genomic RNA [185]. 2.8.1.1 Neomycin B Linked to Acridine. Known inhibitors of the HIV RNA site include the aminoglycoside class of molecules, such as neomycin. These competitive inhibitors block the association of Rev protein to the RRE domain [186]. Kirk et al. [187] demonstrated that a synthetic, divalent molecule composed of an aminoglycoside and a nucleic acid intercalator is a potent inhibitor of Rev–RRE association. Molecules of this type are represented by a heterodivalent conjugate (61) composed of neomycin B covalently linked to 9-aminoacridine through a short spacer (Figure 2.48). In a gel shift assay, neomycin–acridine conjugate 61 can displace a portion of the Rev protein (peptide Rev34 –50 ) from the Rev–RRE complex with an IC50 value of 0.65 mM and is more e¤ective than neomycin (IC50 ¼ 5:9 mM ). Its binding a‰nity to the RRE is estimated at 15 nM, which is comparable to the a‰nity of 12 nM displayed by the Rev peptide (Rev34 –50 ). This high a‰nity is greater than that of the monomer components such as 9-aminoacridine and neomycin B, which have inhibition constants of 24,000 and 800 nM, respectively. This divalent molecule binds to the same region of the RRE as that targeted by the Rev peptide and thus competitively inhibits the formation of the Rev–RRE complex. This example illustrates the design and enhanced a‰nity of a synthetic bivalent ligand to a viral RNA site. 2.8.1.2 Neamine Linked to Pyrene–Carboxylic Acid. A second example of intercalator-linked aminoglycosides was reported by Hamasaki and Ueno

70


HO HO HO

NH2 O NH2

H2N O O O H

NH2 i,ii

OH

i-Pr

OH NH2

O H2N O OH HO

NHBoc O HO HO NHBoc i-Pr BocHN O NHBoc O O O S O H OH O i-Pr OH O NHBoc BocHN O OH HO

Neomycin B RNA binding molecule intercalator

HO HO H N

iii,iv,v

S

NH2 O NH2 H 2N O NH2 O O H OH

61

N OH O NH2 H 2N O OH HO reagents and conditions: i) (Boc)2O, DMF, H2O, Et3N, 60oC, 2 h (72%); ii) 2,4,6-triisopropylbenzenesulfonyl chloride, pyridine, rt, 20 h (75%); iii) H2NCH2CH2SH, NaOEt, EtOH, rt, 4.5 h (80%); iv) 9-phenoxyacridine, phenol, 60-75oC, 1 h (84%); v) 4 M HCl, dioxane, HSCH2CH2SH, rt, 5 min (80%).

Figure 2.48 Synthesis of neomycin–acridine heterodimer.

[188], in which neamine is tethered to pyrene-l-carboxylic acid with an argininecontaining spacer (Figure 2.49). A tethered molecule (62) inhibits the formation of the HIV RRE–Rev complex with an IC50 value of 0.35 mM which indicates tighter binding than that of neamine (21 mM ) or 63 (0.46 mM ). It can also inhibit the formation of the TAR–Tat complex with an IC50 value of 0.028 mM, an activity much greater than that of neamine (28 mM ) or 63 (0.6 mM ). The potent inhibitory activity displayed by 62 is attributed to its strong a‰nity to TAR and RRE, with dissociation constants of 30 and 8.5 nM, respectively. 2.8.2

RNA–Enzyme Interactions

Some molecules not only bind to RNA in a sequence-specific manner but also interfere with RNA–protein interactions occurring at the same sequence. Futaki et al. [189] reported the synthesis and activity of a bivalent molecule that not only binds to specific RNA sequences but, most remarkably, enables an otherwise nonselective ribonuclease (RNase) to cleave at specific RNA sequences. The concept of this artificial RNase is shown in Figure 2.50. First, a natural RNase is assembled in situ as a consequence of the noncovalent association of two proenzyme components, S peptide and S protein, which are derived from two proteolytic fragments from bovine pancreatic RNase A. This complex behaves like a functionally active enzyme, but it lacks the RNA sequence selectivity. However, this nonselective enzyme can be converted to a

71

NUCLEIC ACIDS OF VIRUSES intercalator NHMtr

O

HN H 2N HO HO

NH2 O H2N O HO

NH2 NH2

NH O

HO HO

iv,iii

H2N O HO

OH

NH

O

O i,ii

NH2

HN NH

NH

NH2 NH2 OH

Neamine

HO HO

NH O H2N O HO

62 NH2 NH2 OH

minor groove binder iii NH2

HN H 2N

NH

O HO HO

NH O H2N O HO

63 NH2 NH2 OH

reagents and conditions: i) Fmoc-Arg(Mtr)OH, EDC, DMF, water, rt, 16 h; ii) Piperidine, DMF; iii) TMSBr, TFA, 0oC, 2 h; iv) PyOSu, DMF, water, rt, 20 h.

Figure 2.49 Synthesis of neamine-based aminoglycoside–intercalator conjugate.

selective enzyme by using a tailored S peptide. The design of such an S peptide is based on covalent linkage of an intact S-peptide component to a specific peptide which is capable of binding an RNA sequence selectively. An example is an S-peptide derivative (64), which is composed of a Rev peptide (which recognizes the RRE) attached to an S peptide (which recognizes S protein). The complex made of 64 and an S protein is an artificial RNase S, which cleaves a nucleotide strand of RNA at a loop site between G55 and C56, located close to the RRE domain (Figure 2.50). The single cleavage site is proximal to a major cleavage site (C56 and A57) displayed by natural RNase T1. This artificial RNase S functions even at 10 nM of 64, a concentration that is 10 times lower than that required for an unmodified S peptide. That di¤erence is attributed to the higher binding a‰nity of the modified RNase S to RNA. 2.8.3

Binders to the Minor Groove of Viral DNA

2.8.3.1 Hairpin-Shaped Polyamide Dimers. A hairpin-shaped polyamide comprises two units of polyamide that are tethered at each end with a short spacer such as g-aminobutyric acid. Each polyamide consists of multiple copies of pyrrole (Py) and imidazole (Im). The resulting composite molecule binds a minor groove of double-stranded DNA with high sequence selectivity [190– 193]. Such sequence-specific recognition has been explored extensively by Dickinson et al. at the California Institute of Technology [194] using polyamide

72

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS natural RNase S

+ S-peptide S-protein (inactive enzyme)

active but nonselective enzyme

RNase S (targeted active enzyme)

64

S-protein RNA S-peptide

GG 54 G 41 C GGUCUG CGCAGC CCGGAC GCGUCG A A G A A 79 U

Rev-peptide (RRE recognition)

a major cleavage site by natural RNase S

RRE-containing RNA (Rev peptide)

H2NCO-R-Q-R-E-R-W-R-R-R-R-N-R-R-A-Q-R-T

O N H

O S

(S-peptide) N K-E-T-A-A-K-F-E-R-Q-H-NIe-D-S-CONH2 H

64

Figure 2.50 Generation of active ribonuclease S (RNase S) via noncovalent association of S peptide and the assembly of sequence-selective RNase S by use of S peptide tethered covalently to RRE-recognition peptide (Rev peptide).

dimers such as 65 and 66 (Figure 2.51). Their observed sequence selectivity is rationalized according to empirical rules based on side-by-side heterocycle pairings, which state that a pyrrole opposite an imidazole (i.e., a Py/Im pair) targets a C–G base pair, whereas Im/Py targets a G–C pair and Py/Py binds both T–A and A–T pairs. The polyamide 65 binds very tightly, with a Kd value of 0.05 nM, to a specific DNA sequence that is immediately adjacent to the binding site for TATAbox binding protein (TBP), a transcription factor. The tight binding is attributed to eight pairs of hydrogen-bond interactions between the polyamide and the DNA sequence. When mismatched interactions occur between the polyamide and nucleotide sequence, as observed from 66 (which yields four pairs of matched interactions), the binding avidity becomes weaker, with a Kd value of 5.0 nM. The DNA sites targeted by 65 lie immediately next to a TATA-box sequence with a Kd range of 1 to 3 nM. In fact, its association with the polyamide influences the binding of TBP, as demonstrated by a gel mobility shift assay, in which 65 inhibits the binding of TBP at the TATA sequence. This in vitro functional activity is further translated in vivo such that 65 blocks HIV-1 transcription with an IC50 value of 60 nM. One micromolar concentration of

NUCLEIC ACIDS OF VIRUSES

73

ImPy-β-ImPy-γ-ImPy-β-ImPy-β-Dp (65)

CH3 O N

CH3 O N N H

N

(H3C)2N

H N

H N

H N O

O

N H

N H N

O

O N CH3

N CH3

H N

H N

N H

N H H N

O

O

N H

N

O

N CH3

CH3 O N

CH3 O N

NH

N H N

H N O N CH3

NH

H N

N

O N CH3

O

O

H N

H N

H N O

N

N H

N

N H

O

N CH3

CH3 O N

CH3 O N

(H3C)2N

H N

N

O N CH3

O N H

CH3 O N

CH3 O N

N CH3

O

ImIm-β-ImIm-γ-PyPy-β-PyPy-β-Dp (66)

Match

TATA box

5'-A G A T G C T G C A T A T A A G C A-3' Kd = 0.05 nM

65 3'-T C T A C G A C G T A T A T T

Mismatch

C G T-5'

TATA box

5'-A G A T G C T G C A T A T A A G C A-3' Kd = 5.0 nM

66 3'-T C

T A C G A C G T A T A T T

C G T-5'

Figure 2.51 Structure of dimeric polyamides (hairpin-shaped) and a schematic of an HIV-1 promoter region showing recognition sites by polyamides and transcription factor TBP (TATA box).

65 also induces an 80% reduction of HIV-1 replication. This example with polyamide demonstrates that multivalent recognition on DNA confers a highly specific binding a‰nity and associated functional activity. 2.8.3.2 H-Shaped Polyamide Dimers. An H-shaped polyamide dimer allows dsDNA and the polyamide to make multiple contacts in a 2:1 binding mode. This unique design is achieved by connecting the centers of two polyamide strands through a lateral bridge (Figure 2.52). Sharma et al. [195] reported such H-shaped polyamides designed with a flexible alkane linker (67, n ¼ 5; 7). These serve as minor groove binders, and consequently inhibit the action of

74


N N N H

N (CH2)n N

i,ii (90%)

iii,iv Cl3C

N O (CH2)n

H C O3 N Me2N

N H

O N H

N N

N (CH2)O n N

CCl3

NO2

H N N O CH3

O

(70%) Cl3C

N

N NMe2 ix,x

H2N

S

CH3 H N O

(38%) Me2N

N N (CH2)O n N

v-vii

N H N

N

O

(60%)

N O 2N

O2N

N

H C O3 N N H

H N

N

H N N O CH3

N

O

N

N H H3C

67 (n = 5, 7)

viii (62%)

NO2

N O (CH2)n O

N H

CCl3

S

NH2

N

reagents and conditions: i) K (metal); ii) Br(CH2)nBr; iii) CCl3COCl; iv) Et3N; v) Ac2O; vi) HNO3; vii) H2SO4; viii) N-Methyl-3-amino-5-[(3-dimethylaminopropyl-1-amino)carboxyl]pyrrole, DMF; ix) H2, Pd/C; x) 2-Amino-4-methyl-thiazole-5-carboxylic acid, DCC, HOBT.

Figure 2.52 Synthesis of H-shaped polyamide linked through a lateral bridge.

DNA gyrase in supercoiling DNA at micromolar concentrations. The level of activity by 67 is about 10-fold higher than that of natural distamycin, a member of natural polyamides. 2.8.3.3 Naturally Occurring Dimers. There are numerous examples in nature of molecules that exist as dimers or oligomers and exhibit unique biological activities not possible with the corresponding monomers (Figure 2.53). Such oligomeric molecules are represented by a family of over 70 forms of natural quinones that include dimers, trimers, cyclotrimers, and higher oligomers [196]. Conocurvone (69) is a deoxytrimeric quinone molecule that is probably derived from teretifolin B (68) [197,198]. This trimeric quinone strongly inhibits HIV-1 replication. At a concentration below the EC50 value of 20 nM, it averts completely the death of HIV-1-infected human lymphoblastoid cells by inhibiting replication of the virus. Such potent anti-HIV activity [81] is associated uniquely with trimeric presentation of quinone because the monomer 68 lacks such activity completely. Little is known regarding the mode of action of 69, but it may be related to its ability to form a helix, which is not possible with a monomer. The resulting helical molecule then binds the groove of DNA strand and interferes with the replication and transcription of virus-derived nucleic acid. Examples of dimeric quinones that interact with DNA possibly by such a mechanism include viomellein (70) [199] and michellamine (71), which belongs to the family of dimeric naphthylisoquinoline alkaloids that also includes jozimine C and korundamine [200]. Michellamine is highly active against HIV-1 and HIV-2 [201].

NMe2

NUCLEIC ACIDS OF VIRUSES

75

CH3 O O O

O OH H3C

O

H 3C

O

O O

O O

O O H 3C

CH3

CH3

H3C

CH3

CH3 CH3

CH3

CH3 69 (conocurvone)

68 (teretifolin B)

HO H3C O

CH3

OH O O

O OH OH O

H 3C

HN H3C H 3C

O 70 (viomellein)

OH H3CO HO CH3 CH3 NH

OH OCH3 HO 71 (michellamine B)

OH

CH3

Figure 2.53 Structure of naturally occurring oligomeric molecules that display antiHIV activity.

antigenic epitopes

membrane

pathogen (virus, bacterium)

antigenic epitopes

antigenic epitopes

Protein scaffold

Protein scaffold

synthetic multivalent vaccine

Figure 2.54 Surface of pathogen carrying multiple copies of antigenic epitopes, and synthetic multivalent molecules presenting multiple antigenic determinants.

76

MULTIVALENT MOLECULES APPLIED TO VIRAL TARGETS Dde-HN

O

O HN

NH-Dde

HN

NH2

i,ii,iii

iv O

O

O

O

O

NH-Fmoc

O O

NH v NH

HN NH

NH-Fmoc Solid support

M2

O

O

M2

HN

S1 S1 S2

NH

HN

O O

O

O

O

O NH-Fmoc

HN vi

O

O O

O O

HN

NH

O

HN NH

HN

H N

HN

H N

N H

HN RS

H N

NH

H N

HN HN

NH

S2

O

O O

72

O NH-Fmoc

HS

H N

N H

O O

NH NH

HN

M2

M2

O

O

HO

O

HN

HN

S1 HN H N H N HN vii,viii HN

O O

N H

O

NH

S2

H N

O

HN

NH O

O

NH

HN

NH

HN

O

NH

HN

NH

O O

S

O

O

H N

N H

O O

HN

O O

NH NH

NH2

NH2

M2

O O

O

S

H N

H N

73 O NH

OH

HO M2 = SLLTEVETPIRNEWGSRSNDSSDP S1 = SFERFEIFPKE S2 = HNTNGVTAASSHE

O HN

reagents and conditions: i) 20% piperidine, DMF; ii) standard F-Moc solid-phase peptide coupling with M2 peptide (to terminal Boc-protected serine); iii) 2% hydrazine, DMF; iv) F-Moc standard coupling of Fmoc-Gly-OH, and Dde-Lys(Fmoc)-OH; v) standard peptide addition of M2, S1, S2, and final coupling of Fmoc-Gly-OH, and Fmoc-Cys(Trt)-OH; vi) peptide cleavage from resin; vii) 2,2'-dithiodipyridine; isolation of a disulfide product, and its reaction with free thiol 72; viii) 20% piperidine, DMF.

Figure 2.55 Synthesis of an octavalent peptide construct carrying three types of antigenic determinants.

SYNTHETIC MULTIVALENT VACCINES

2.9

77

SYNTHETIC MULTIVALENT VACCINES

Synthetic vaccines include multivalent molecules that present multiple copies of antigenic determinants tethered to a sca¤old. They are designed to mimic the critical parts of antigenic sites and ectodomains found on the surface of pathogenic viruses and bacteria (Figure 2.54). With multiple determinants on its surface, a multivalent molecule is likely to cross-link antigen-specific immunoglobulin receptors on a B-cell and thus stimulate immune responses against specific antigens [3]. 2.9.1

Peptide-Based Anti-influenza Vaccines

Kragol et al. [202] reported a peptide-based multivalent vaccine designed to stimulate immune responses against influenza virus (Figure 2.55). This molecule (73) presents four copies of an ectodomain derived from the viral M2 protein, a membrane-bound ion channel, as a B-cell epitope, and two copies of S1 and S2 fragments as T-helper cell epitopes. Targeting B- and T-cells with three such distinct types of cross-linked epitopes makes the synthetic vaccine more likely to be a potent B-cell immunogen.

N

O

NH2

O

O

O N H2N

i) DMF, 3h

O

O

O

O

N

N O

O

n ii) Ac-CGSGVRGDFGSLAPRVARQL (FMDV peptide), H2O, 6h

water soluble carbon nanotube Ac-Cys-FMDV peptide S

O N

O

N O

Ac-Cys-FMDV peptide S

O

H N

O

O H N

N O

O

O

O

N n

Figure 2.56 Structure of foot-and-mouth VP1 peptide displayed on the surface of a single-walled carbon nanotube.

78

2.9.2


Gp41-Based Anti-HIV Vaccine

Wang and Singh [203] described the synthesis of multivalent peptides that present four copies of gp41 epitope incorporating the T20 domain, a sequence responsible for potent HIV inhibition. The design of this synthetic vaccine is based on the use of a carbohydrate-centered maleimide cluster, which is readily derivatized with a thiol-containing peptide. 2.9.3

Peptide-Based Anti-FMDV Vaccines

Synthetic vaccines targeting the foot-and-mouth disease virus (FMDV) are designed with a lipidated, lysine-cored dendrimer that displays four copies of a cyclic disulfide epitope derived from the virus [204]. This multivalent peptide is immunogenic and has been used for a vaccination trial. In another example, a unique framework based on single-walled carbon nanotubes (SWNTs) is used for displaying antigenic determinants. Pantarotto et al. [205] investigated the use of such material in the preparation of multivalent antigens derived from FMDV. This conjugate is composed of water-soluble amine-terminated SWNT (about 0.3 to 0.5 mmol/g) that is derivatized to present on its outer surface multiple copies of viral envelope peptide VPl of FMDV (Figure 2.56). The peptide on the surface retains its original conformation as well as its original activity and is readily recognized by anti-VP1 antibodies. This nanotube-based antigen induces a strong antibody response in mice.

3 MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS

3.1 3.1.1

TARGETS IN BACTERIAL CELL MEMBRANES d-Ala-d-Ala Peptide Precursors

3.1.1.1 Mode of Action by Antibiotics of the Vancomycin Class. Chlororemomycin, a member of the vancomycin group of antibiotics, exerts its antibacterial activity by binding a d-Ala-d-Ala terminus—a cell wall precursor presented on the surface of gram-positive bacteria [206]. Such binding interferes with the biosynthesis of bacterial cell walls (Figure 3.1), thus leading to the death of the bacterial cells. Mackay et al. at Cambridge [207] studied the association of chloroeremomycin in solution using a peptide ligand presenting d-Ala-d-Ala, which shows a strong association and a Ka value of 1.3 10 6 M 1 This tight binding stems primarily from the formation of multiple pairs of hydrogen bonds established between the antibiotic and the divalent ligand (Figure 3.2). In a peptide cell wall model of vancomycin-resistant bacteria where a terminal d-Ala residue is replaced by d-lactate (d-Lac), the association constant of chloroeremomycin with such a modified dipeptide terminus decreases about 1000-fold, to 920 M 1 [65,208]. This weak association is not only attributed to a reduction in the number of attractive hydrogen bonds but importantly, to the generation of a repulsive dipole–dipole interaction between the ester oxygen of


79

80

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS D-Ala D-Ala L-Lys D-Glu L-Ala HO O HO AcHN

N H

CH3OH O HO OO O O O HO AcHN AcHN

O O O

H3C

D-Ala D-Ala L-Lys D-Glu L-Ala

Transpeptidase-mediated transpeptidation

Inhibited by

HO HO HO AcHN

Vancomycin β-Lactam


OH O HO O NHAcHO AcHN O

O O O H3C

L-Ala D-Glu L-Lys D-Ala NH

CH3 O OH HO O O O O O HO AcHN AcHN

HN D-Ala L-Lys D-Glu L-Ala

N H

OH CH3 OH HO O O O O O O HO O NHAc NHAc O AcHN

L-Ala D-Glu L-Lys D-Ala D-Ala

Cross-linked peptidoglycan layer

CH3 O OH O O O O AcHN O O OP P O O-O

Bacterial inner membrane Immature peptidoglycan D-Ala D-Ala L-Lys D-Glu L-Ala HO HO HO AcHN

NH2


D-Ala D-Ala L-Lys D-Glu L-Ala CH3 O OH O O O O AcHN O O OP P O O-O

Lipid II intermediate

Transglycosylase-mediated transglycosylation Inhibited by Moenomycin Teicoplanin Chlorobiphenyl vancomycin

D-Ala D-Ala L-Lys D-Glu L-Ala HO HO HO AcHN


Figure 3.1 Biosynthetic pathway of a peptidoglycan layer in a bacterial cell wall and selected inhibitors that block it at di¤erent stages.

d-Lac and an amide oxygen (Figure 3.2) [209]. This model demonstrates that chloroeremomycin relies on the well-ordered receptor–ligand interaction with d-Ala-d-Ala as a primary mechanism of action. When studying the binding a‰nity of the vancomycin group of antibiotics, planar lipid bilayers or liposomes (see Section 1.1.3) that present multiple copies of precursor peptide ligands such as d-Ala-d-Ala or d-Ala-d-Lac on their surface are commonly used as models of bacterial cell membranes. For example, liposomes presenting multiple copies of d-Ala-d-Ala ligand on their surface, such as N a -docosanoyl-N e -acetyl-lysyl-d-alanyl-d-alanine, recognize chloroeremomycin with a Ka value of 2 10 7 M 1 as determined by surface plasmon resonance [210]. That association constant is 15-fold tighter than when performed with a free ligand in solution. Although the enhancement is

TARGETS IN BACTERIAL CELL MEMBRANES H2N HO OH O H3C CH3 HO O

(a)

O N H N H

O HO

Chloroeremomycin OH O O

H 2N HO O H3C CH3 O

O

O

Cl H N

O

Cl

N O

OH O

O N

H CONH2

H

N H

OH OH

N H

O

CH3 H N O

H 2N HO OH O H 3C CH3 HO O

O O HO

N H N H

O OCH3

OH H-bond

O O

H2N HO O H 3C CH3 O

O

O

Cl H N

O N O

H

Cl

dipole repulsion OH O

O

N H CONH2

N H

OH OH H N

H3C O

O H 3C N H

NHAc

CH3 NH CH3 CH3

O HO

H-bond

Nα,Nε-(Ac)2-Lys-(D)-Ala-(D)-Ala (VSE ligand)

NHAc

(b)

CH3

Ka = 1.3 x 106 M-1 O

H N

H3C

CH3 NH CH3

O HO

81

Ka = 920 M-1

O O O

OCH3

Nα,Nε-(Ac)2-Lys-(D)-Ala-(D)-Lac (VRE ligand)

Figure 3.2 Chloroeremomycin and its association to divalent ligand in solution: (a) tight association via multiple pairs of hydrogen bonds (dotted lines) with d-Ala-d-Ala ligand; (b) weak binding to d-Ala-d-Lac ligand as a result of a reduced number of hydrogen bonds and dipole–dipole repulsion (double-arrowed line).

somewhat modest, the observed binding a‰nity on the model surface is indeed close to a value observed with the surface of a bacterial cell (K a @10 8 M 1 ) [211]. This type of enhanced a‰nity from a vancomycin-sensitive cell wall model is also found to a greater extent in a model for a vancomycin-resistant cell wall. For example, the association constant of chloroeremomycin to a liposome presenting multiple d-Ala-d-Lac termini (N a -docosanoyl-glycyl-alanyld-g-glutamyl-N e -acetyl-lysyl-d-alanyl-d-alanine) is 2.5 10 5 M 1 , which is 250-fold higher than that obtained with free d-Ala-d-Lac ligand.

82

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS

Figure 3.3 Dimerization of glycopeptide antibiotic and its implication in its multivalent binding to d-Ala-d-Ala (or d-Lac) ligand present on cell membrane.

3.1.1.2 Dimerization of Glycopeptide Antibiotics. The examples described in the preceding section show that chloroeremomycin binds more tightly to a model membrane of a bacterial cell wall than it binds to free ligand. Williams [80] attributed this property to the greatly enhanced ability of the glycopeptide to form dimers on the cellular surface. For example, chloroeremomycin undergoes noncovalent dimerization, in the absence of any ligands, with a dimerization constant K dim value of 1.8 10 5 M 1 . In the presence of d-Ala-d-Ala ligand, K dim increases to 1.1 10 7 M 1 (Figure 3.3). This observation suggests that the enhanced a‰nity on a surface is probably attributed to enhanced chloroeremomycin dimerization, since receptor dimerization helps to reduce the loss of overall entropy generated upon receptor–ligand association. This example is considered as a cooperative system where ligand–receptor association enhances receptor dimerization as well as lowers the energy of divalent complexation. Glycopeptide antibiotics of the vancomycin class show varying degrees of self-dimerization with K dim ranging from about 500 to 10 8 M 1 (Figure 3.4) [212,213]. This ability to self-associate varies from a strong tendency, such as that of eremomycin, to the weak dimerization ability of vancomycin. As described earlier, antibiotic dimerization increases in the presence of bound ligand. For example, the dimerization constant of eremomycin increases about 100-fold, from K dim ¼ 3 10 6 M 1 in the absence of ligand to 3 10 8 M 1 upon incubation with N a, e -di-Ac-Lys-d-Ala-d-Ala [212,213]. With a K dim of 700 M 1 , vancomycin dimerizes relatively weakly in solution although it is highly active against gram-positive bacteria presenting d-Ala-dAla peptide precursors on their peptidoglycan layer. The drug shows potent

TARGETS IN BACTERIAL CELL MEMBRANES OHNH2 OH Glucose O HO OH H3C CH3O O Vancosamine O Cl O O HO O N H N H

O HO

Cl H N

O N O

OH O

O

N H

N H CONH2

H

83

Vancomycin Kdim (M-1) = 700 CH3 NH CH3 CH3

O OH OH HO HO OH Rhamnose HO Me O

HO OH Mannose OH OH O OH Arabinose O O OOH

O Glucose O Ristosamine HO H 3C

O O

O NH2 O O O

HO

N H N H

OHO

H N

N O

H

NH2

Ristocetin A Kdim (M-1) = 500

O OH

O HO HO OH HO

H 2N HO OH O H 3C CH3 HO O Epivancosamine H 2N HO O H 3C CH3 O O O

N H

CH3

O OH HO

HO OH O

N H

HO

O

HO

OH

O

N H N H

Glucose OH O O

O

O

X H N

O N O

O

H

X

O

N H CONH2

OH O N H

Eremomycin (X = H) Kdim (M-1) = 3 x106; Kdim (M-1) = 3 x108 (with ligand) CH3 NH CH3 CH3

Chloroeremomycin (X = Cl) Kdim (M-1) = 1.8 x105; Kdim (M-1) = 1.1 x107 (with ligand)

OH OH HO

Figure 3.4 Dimerization constant of glycopeptide antibiotics of the vancomycin class.

84


antibacterial activity against numerous strains of Staphylococcus aureus and vancomycin-susceptible enterococcus (VSE) with a minimal inhibition concetration (MIC) a 1 mM. However, the activity of vancomycin decreases about 700-fold against strains of vancomycin-resistant enterococcus (VRE), which present d-Ala-d-Lac on its surface, with an MIC of about 700 mM [214]. Such reduced activity of vancomycin is associated with its decreased binding a‰nity to the VRE ligand as reflected in the millimolar dissociation constant to Ac2 Lys-d-Ala-d-Lac (Kd = 1.8 mM), which is roughly 1600-fold weaker than when binding to Ac 2 -Lys-d-Ala-d-Ala (K d ¼ 1.1 mM ). The dimerization of glycopeptide antibiotics also occurs in the solid state. For example, vancomycin forms a dimeric complex with acetate or N-Ac-d-Ala [215–217]. The x-ray crystal structure reveals two di¤erent types of dimeric interaction, as shown in Figure 3.5. Ligand-independent association yields a back-to-back dimer, while ligand-mediated association yields a face-to-face dimer. This structural information provides a rationale for designing synthetic covalent dimers of vancomycin though the structure is based on solid state rather than solution phase. It is of particular relevance that vancomycin dimerization leads to increased a‰nity to its target, which can be translated to increased antimicrobial activity. 3.1.1.3 Di- and Trivalent Vancomycin. Gri‰n et al. at Stanford and Eli Lilly [218] reported that a covalent dimer derived from vancomycin enhanced antibacterial activity, possibly through a mechanism of divalent association, as shown in Figure 3.6. As summarized in Figure 3.7, the dimer is composed of two vancomycin carboxamides tethered with a hydrophobic linker at the carboxylic acid terminus of vancomcyin. The MIC of dimers 1, 2, and 3 against VRE in vitro are 32, 11, and 15 mM, respectively. Thus, dimer 2 shows an increase in antibacterial activity over the parent vancomycin by a factor of 60. The antibacterial activity of dimers 1, 2, and 3 against VSE strains is enhanced as well, but to a lower degree, with MIC values of 0.51, 0.50, and 0.29 mM, respectively. Such dimer activity represents only a two- to threefold enhancement relative to vancomycin. These vancomcyin dimers can bind a nonclustered form of the VRE model peptide ligand (Ac2 -Lys-d-Ala-d-Lac) in solution with dissociation constants of 180 mM (1), 440 mM (2), and 660 mM (3). Such dissociation strength indicates 3- to 10-fold tighter complex formation than that of the vancomycin monomer, which shows a Kd value of 1800 mM. The moderate decrease in Kd for dimer-ligand binding in solution does not appear to account fully for the 60-fold enhancement of the dimers’ in vitro antibacterial activity. The enhanced antibacterial activity may be accounted for by considering the additional contribution derived from the intrinsic ability of a covalent dimer to bind a cluster of d-Ala-d-Lac ligands present on a bacterial surface. The stability of the divalent vancomycin–ligand complex would be much greater than that achieved with monovalent ligand. The multivalent nature of vancomycin’s interaction with the d-Ala-d-Ala peptide ligand has prompted further biochemical studies. Using a di¤erent set

TARGETS IN BACTERIAL CELL MEMBRANES CH3 O

(a) -O

N H

O

OH

CH3 HO

OH O

CH3

H2NOC H N

H 3C

H N HN CH3 O HO

O

H

H N HO 2 H3C

O

OH3C O

OH OH

OH O O

O CH3O

N H N H

HO

OH

O

HO HO

O

OH O

O

O O HO

HO O

N H Cl

O

Cl

H N H N

O

N

Vancomycin (back-to-back dimer)

O

O

Cl H N

O

OH O

O

N H

N

H CONH2

H

NHOH 2

Cl

N O

CH3

OH OH

H N

H3C

HO

O H N HO 2 H3C

HO HO O CH3O

HO O O HO

N H N H

OH O O O

O

Cl H N

O H

OH O

O

N H

N

H CONH2

HO

-O2C -O2C O

N H NH

CH3 H N

NH O CH3

N-Ac-D-Glu-L-Dap-D-Ala-D-Ala

O OCH3

O

CH + 3 NH2 CH3 CH3

O OH OH O H 2N

CH3

Cl

N O

CH3 NH CH3

O

(b)

85

Vancomycin (face-to-face dimer)

O OH

OCH3

HO

CH3 CONH2 CH3 H O H H N N N H2N+ N CH3 O H O O HO Cl Cl

O

OH O H N H N

OH O

O OH

O O O HO

OH3C O

OH OH

CH3 NHOH 2

Figure 3.5 Structure of noncovalent vancomycin dimers formed in the presence of ligand: (a) back-to-back dimer with N-Ac-d-Ala; (b) face-to-face dimer with N-Ac-d-Glul-Dap-d-Ala-d-Ala.

86


D-Ala-D-Ala L-Lys D-Glu Immature peptidoglycan L-Ala CH3 O OH HO OH HO HO O O O O HO O OO O HO AcHN O AcHN AcHN AcHN O O OP P O H3C O O-O L-Ala D-Ala-D-Ala D-Glu L-Lys bacterial membrane CH3 O H ON N O OH O H CH3 OH HO O CH3 H2NOC Linker HN N H H O H 3C H H H O N N N N HN N CH3 O H O O O HO OH Cl Cl

O

O O O

HO

OH3C CH3 OH O HO H2N OH

HO HO HO AcHN


OH O O

HO O O NH

N H N H

O

O

Cl H N

O N O

Cl OH O

O N

H CONH2

H

N H

O OH OH HO

D-Ala-D-Ala L-Lys D-Glu L-Ala CH3 O OH HO O O O O O HO AcHN AcHN

O N H L-Lys D-Glu L-Ala

CH3 H N O

CH3 NH CH3 CH3

O OCH3

D-Ala-D-Ala


Figure 3.6 Proposed multivalent binding of a covalent vancomycin dimer to multiple d-Ala-d-Ala residues present on a bacterial cell surface.

of synthetic glycopeptide contructs, Rao and Whitesides [219] investigated the association of a divalent system derived from vancomycin and the d-Ala-d-Ala peptide ligand (Figure 3.8). A vancomycin dimer (4) presents two vancomycin units tethered at the C-terminus through a rigid linker, p-xylylenediamine, readily prepared by amide coupling. A Kd value of the divalent complex formed between 4 and 5 is about 1.1 nM when measured using a competitive binding assay based on a‰nity capillary electrophoresis (ACE). This bivalent interaction thus demonstrates about a 600-fold enhanced a‰nity relative to the corresponding interaction between vancomycin alone and diacetyl-l-Lys-dAla-d-Ala (Kd of 4.8 mM ). The free energy of association involved in such a divalent complex is 11.9 kcal/mol, while the monovalent complex generates 8 kcal/mol [220,221]. The di¤erence between the two G values—3.9 kcal/mol— might indicate a contribution from a second binding event driving the divalent complex formation. Indeed, the di¤erence in G values is equivalent to a high e¤ective molarity of about 20 mM for the second ligand unit in this particular association.

87

TARGETS IN BACTERIAL CELL MEMBRANES OHNH2 OH O HO O H3C CH3

OH O O

HO O O HO

N H N H

O

O

Cl H N

O

Cl

N O

H

OH O

O N

H CONH2

O OH OH

N H

CH3 NH

H2N-Linker-NH2 (0.5 eq) HBTU* (2.2 eq) DMSO, DMF

CH3 CH3

Vancomycin


HO

OH O O

OH HO

O O

CH3 H 3C

HO

OH

H N

HN CH3 O HO

H2NOC H N O

Cl

H N O

O

H N H N

O N H Cl

HN O

N H N H

O

Cl H N

O N O

H

O

Cl

O N

H CONH2

OH O N H

CH3 NH CH3 CH3

OH OH

O OH

HO

O 1: Linker = (CH2)6

O O HO

Linker

O NH

O


2: Linker = (CH2)2-S-S-(CH2)2 3: Linker = (CH2)3-S-S-(CH2)3

HBTU*: 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate

Figure 3.7 Synthesis of divalent vancomycin carboxamide (linked from C-terminus to C-terminus).

Using surface plasmon resonance (SPR) spectroscopy, Rao et al. [222] further studied the binding a‰nity of 4 to peptides presented on the surface of selfassembled monolayers (SAMs). The dimer 4 binds to the SAM, presenting 5% of d-Ala-d-Ala ( wL ¼ 0.05) with a calculated Kd value of about 0.5 nM. Such tight surface adhesion is ligand-specific, as demonstrated by the competitive displacement by free N-Ac-d-Ala-d-Ala, the addition of which reduces the amount of dimer bound on a surface. In comparison, vancomycin, with a Kd value of 1.1 mM, is less e¤ective in blocking binding of the vancomycin dimer to the SAM surface. The results from an SPR study indicate that 4 binds multivalent d-Ala-d-Ala displayed on the surface with an a‰nity >1000-fold enhanced relative to the parent molecule, vancomycin. As shown in Figure 3.9, Rao et al. [50] extended their investigation to include the association properties of a trivalent system derived from vancomycin and d-Ala-d-Ala. A tris-vancomycin carboxamide (6) is composed of three vancomycin units linked to the sca¤old of benzene–tricarboxylic acid through

88

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS OHNH2 OH O HO H3C CH3O

OH O O

OH HO CH3 H 3C

H N HN CH3 O HO

OH

HO O

O H2NOC H N O

Cl

H

O

N

N H Cl

O O

H O N

N O H

H N H N

N H N H

O

O

Cl H N

O N O

H

O N

H CONH2

O

O

Cl OH O N H

CH3 NH CH3 CH3

OH OH

OH HO

O O O

HO


4

O CH3 O -O O

N H

H HN N CH3 O

CH3 NH O O

O

HN

N NH H

H3C

CH3 H N O

O OCH3

O 5 [CH2CONεH(Nα-Ac)-L-Lys-D-Ala-D-Ala-CO2-]2

Figure 3.8 Divalent system of receptor–ligand pair derived from vancomycin and d-Ala-d-Ala.

an amide bond at the C-terminus of vancomycin (Figure 3.10). Similarly, a trivalent ligand (7) is made of three copies of d-Ala-d-Ala ligand tethered to benzenetricarboxylic acid. The trivalent vancomycin 6 makes a complex with the analogous trivalent ligand 7 at an extremely low dissociation constant of about 4 1017 M. This observed avidity of the trivalent vancomycin-d-Ala-dAla system is, in fact, 25 times greater than the a‰nity of the biotin–avidin system (K d A 1015 M ) [223], which belongs to one of the tightest naturally occurring systems. In contrast, 6 binds to a monovalent ligand, N a, e diacetyl-l-Lys-d-Ala-d-Ala, with a K d value of 2.7 106 M. This value is close to the dissociation constant of 1.6 106 M shown by the monovalent complex formed by vancomycin and N a, e -diacetyl-l-Lys-d-Ala-d-Ala. The thermodynamics of the multivalent interactions involved in the complex formation of 6 7 are introduced in Chapter 1, where the interactions are evaluated using isothermal titration calorimetry (ITC). This measurement makes it possible to quantify the contribution from an enthalpy change, DH . Since the enthalpy term is associated with the free-energy equation, DG ¼ DH T DS (DG calculated from a dissociation constant), it is also possible to calculate the

89

TARGETS IN BACTERIAL CELL MEMBRANES

H3C

CH3

H3C H N

H N

O HO

H N

CONH2

O

N H O

O

Cl O

HO OH

HN HO

OO

H NH N

O

HO OHO O

OH

Cl O

OH

O

OHNH2 OH O HO O CH3

O HN

CH3

H 2N HO

OH

H 3C

O O

CH3

O HN

O

Cl H N

HO O

H N

O

N H N H

H O N

O

NH

O O N O

H

Cl OH O

O

N H

N

H CONH2

CH3 CH3

O O

HO

CH3 NH

OH OH

NH

HO

6 O N H HN

HO HO

OH O-

O O

Cl

H

H

O

N

H2NOC

CH3 OH H3C O NH2 O OH OH

NH

O

N O

O

O O

CH3

O HN

O

H3C O

NH H N

OH H N H 3C

O CH3

Cl HO O

CHHN 3

HN

CH3

O

O

NH O

HN H C -O

3

N NH H

H 3C

CH3 H N O

O OCH3

O

CH3 7

HN

O

O

H N

O

O CH3

NH O

Figure 3.9 Trivalent system derived from vancomycin receptor and d-Ala-d-Ala ligand.

90

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS H2N

HN H 2N

H2N

i

O

ii,iii

H N

O NH2

NH-Boc

O

NH

NH2

NH2

Vancomycin HN

OH NH2 OH O HO H3C CH3O

OH O O

HN

O HO

iv H N

O O

NH

O H O N

N H N H

O

O

Cl H N

O N O

Cl

O

N H CONH2

H

O

N H

CH3 NH CH3 CH3

OH OH

Vancomycin NH

6

HO

D-Ala-D-Ala-Lys-L-(Nα-Ac)-HN α

OH O

t

O

v,iii

N -Ac-L-Lys-D-Ala-D-Ala-O- Bu D-Ala-D-Ala-Lys-L-(Nα-Ac)-HN

NH-(Na-Ac)-L-Lys-D-Ala-D-Ala O

O 7

reagents and conditions: i) (Boc)2O, TEA, rt; ii) 1,3,5-benzene tris(carbonyl chloride), CH2Cl2, 2 h; iii) TFA, CH2Cl2, 1 h; iv) vancomycin hydrochloride, HBTU, DIPEA, rt, overnight; v) 1,3,5-benzene tris(carbonyl chloride), DMF, rt

Figure 3.10 Synthesis of a trivalent vancomycin carboxamide and a trivalent ligand.

entropy of binding DS (Figure 3.11). For example, a high association constant of 6 7 yields a DG tri value of 94 kJ/mol, which is split into two components: an experimental value of DH tri (167 kJ/mol) and a calculated value of T DS tri (73 kJ/mol). Such binding parameters are compared with those generated from the association of monovalent vancomycin and d-Ala-d-Ala, which consist of DG mono (33 kJ/mol), DH mono (50.2 kJ/mol), and T DS mono (17.2 kJ/mol). This comparison indicates approximately an additive gain in enthalpy for the trivalent system relative to the monovalent system. On the other hand, the entropy change for a trivalent system represents a greater loss than the monovalent system by 56 kJ/mol. This entropic loss in the trivalent system is possibly related to conformational restriction of the linker during formation of the trivalent complex. Thus, the favorable enthalpy contribution is responsible for the very tight association of the trivalent system (Kdtri ¼ 4 1017 M ).


91

Monovalent receptor-ligand interaction

+ Ac2-L-Lys-D-Ala-D-Ala

Vancomycin

Kd = 1.6 x 10-6 M Vancomycin-D-Ala-D-Ala

∆GO = -33 kJ/mol, ∆HO = -50.2 kJ/mol, T∆SO = -17.2 kJ/mol

Trivalent receptor-ligand interaction

+ Kd = 4 x 10-17 M

Tris-vancomycin (6)

6•7 Tris-D-Ala-D-Ala (7)

∆GO = -94 kJ/mol, ∆HO = -167 kJ/mol, T∆SO = -73 kJ/mol

Figure 3.11 Thermodynamic parameters involved in monovalent and trivalent interaction of vancomycin and d-Ala-d-Ala ligand.

The structure of vancomycin contains at least three functionally distinct sites that can be derivatized to permit covalent dimerization. These functional sites include a carboxylic acid at the C-terminus, a secondary amino group at the Nterminus (N-Me-Leu), and a primary amino group located at a vancosamine residue. Tethering two of these sites with a linker in a combinatorial manner leads to six possible classes of orientational dimers, including a C-terminus to C-terminus linkage (head-to-head homodimers) as discussed already. Five additional types of linkage remain, including N-terminus to N-terminus (tail to tail), vancosamine to vancosamine (back to back), C-terminus to N-terminus, C-terminus to vancosamine, and N-terminus to vancosamine. Each linkage plays a unique role in presenting the binding pocket of vancomycin to the d-Ala-d-Ala ligand in a way that influences the binding a‰nity and antibacterial potency of the dimer. Staroske and Williams [224] reported vancomycin dimers produced by linking through an N-terminus to a C-terminus. These dimers show enhanced cooperativity in receptor–ligand interactions as well as better antibacterial activity (Figure 3.12). When they bind to multivalent ligands displayed on a bacterial cell surface, head-to-tail dimers such as 8 and 9 tend to undergo oligomerization more readily than do head-to-head dimers and generate highorder complexes, such as dimers of dimers. Synthesis of the N-terminus-to-Cterminus dimers requires several steps, including a key amide coupling between an Fmoc-protected linker acid and vancomycin. Such coupling requires regioselectivity at an N-Me-Leu residue of the antibiotic, which is controlled by use of certain coupling reagents. For example, when PyBOP [(benzotriazol-1-

92



OH NH2 OH O HO O CH3

OH

O O

O O O

O

Cl H N

O

Cl

i,ii HO

HO O N H N H

O HO

N O

H

OH

H3C

O

OH O

O

CH3 NH

N H

N H CONH2

O

CH3

O

Cl H N

O N O

H

O

CH3

O

HO

N H N H

O

Cl OH O

O

CH3 N (CH2)n-NH2 N H H O CONH2H3C CH3 N

OH OH

OH OH

n = 2, 5

HO

HO OH NH2 OH O HO H3C CH O

OH NH2 OH O HO O CH3

H3C

3

O O O HO O

iii O HO

N H N H

O O N O

O OH OH HO

Cl HO O

Cl H N

H

OH O O

OH

O

OH O

O CH3 N (CH2)n NH

N H H O C H 3 CONH2 CH3 N

N H N H

O

O

Cl H N

O N O

O

H

Cl

O N

H CONH2

OH O N H

CH3 NH CH3 CH3

OH OH HO 8 (n = 2) 9 (n = 5)

reagents and conditions: i) HO2C-(CH2)n-NHFmoc, PyBrOP (bromo-tris-pyrrolidino-phosphonium hexafluorophosphate), i-Pr2NEt, DMF (60%); ii) 10% piperidine, DMF; iii) vancomycin, HBTU (2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate), i-Pr2NEt, DMF, DMSO.

Figure 3.12 C- to N-terminus covalent (head-to-tail) dimers of vancomycin.

yloxy)tripyrrolidinophosphonium hexafluorophosphate) is used as a coupling agent, amide coupling occurs to a greater extent at the primary amine of the vancosamine sugar than at the secondary amine. In contrast, use of a close analog, PyBroP (bromotripyrrolidinophosphonium hexafluorophosphate], enables the acylation reaction to take place predominantly at the secondary amine from the N-terminus. Adamczyk et al. [225] and Stack and Thompson [226] at Abbott compared the self-association of back-to-back and head-to-head vancomycin dimers using surface plasmon resonance (SPR) technology (Figure 3.13). Each of the backto-back dimers, represented by 10 and 11, is composed of two vancomycin molecules tethered through a bisamide linkage at the vancosamine moiety. These dimers lack a basic nitrogen at the vancosamine moiety. Based on SPR binding studies, they adhere to a surface that presents an array of N e -Ac-Lysd-Ala-d-Ala ligands, and the association occurs in a bivalent manner. The surface adhesion does not, however, proceed to the higher-order self-association of dimers, which can build multiple layers of dimers at certain concentrations

93

TARGETS IN BACTERIAL CELL MEMBRANES O Vancomycin•HCl

+

N

O

O

O

O

O O

Et3N

N

n (n = 3, 6)

DMF O

back-to-back dimer O

OHO N H3C O H H 3C

O

OO

HO O O HO

N H N H

(CH2)n

OH OH

O

OH Cl O

Cl H N

O

NH OH OH O HO H3C CH3O

HO

N O

H

O

N H CONH2

OH O N H

O CH3 NH CH3 HO O CH

O

N H N H

3

O OH OH HO

OH O O

10 (n = 3) 11 (n = 6)

O

O

Cl H N

O N O

H

Cl

O N

H CONH2

OH O N H

CH3 NH CH3 CH3

OH OH HO

Figure 3.13 Vancosamine-to-vancosamine (back-to-back) dimers of vancomycin.

tested ( 16 mg/mL. The optimum activity occurs at a specific length of disulfide tether, with about 20 to 22 atoms in the linker (n ¼ 2; 3).

96

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS H3C OH N H3C O H O

OO

HO O O HO

N H N H

O

OH Cl O

Cl H N

O N O

H

O

n

OH OH [(PCy3)2Ru(CHPh)Cl2] C12H25NMe3Br OH O

O

N H CONH2

N H

H2O, CH2Cl2 (>95:5) (30-80%)

CH3 NH CH3 CH3

O OH OH HO H3C OH N H3C O H O

OO O HO O O HO

N H N H

Cl H N

N

O OH OH HO

H

n

OH OH n = 3, 4, 8

O N

H CONH2

OH O O

OH Cl O O

O

O

NH OH OH O HO H3C CH3O

O n

HO OH O N H

O CH3 NH CH3

O HO

N H N H

O

O

Cl H N

O N O

H

Cl

O N

H CONH2

OH O N H

O

CH3

CH3 NH CH3 CH3

OH OH HO

13 vancosamine-to-vancosamine dimer

Figure 3.16 Structure of olefin-linked (back-to-back) vancomycin dimers.

Therefore, proper positioning of two vancomycin receptor pockets is of importance for tight binding to d-Ala-d-Lac precursors expressed on the surface of resistant bacterial cells. In another series of vancosamine-to-vancosamine homodimers, the olefin metathesis method allows the introduction of an olefinic linkage in the middle of a linker (Figure 3.16). This olefinic bond is structurally rigid; it is locked in either a trans or cis configuration and is chemically stable relative to the disulfide bridge, which is readily cleavable by reducing agents. Vancomycin dimers of this class display improved activity against VRE (n ¼ 4: MIC ¼ 2 mg/mL), and the activity is dependent on the length of the olefinic linker (n ¼ 4 > 3 > 8). 3.1.1.4 Cyclic Dimers of Eremomycin. Eremomycin is another member of the glycopeptide antibiotic class. As reported by Sztaricskai et al. (Figure 3.17) [229], its cyclic dimer is linked with bisphenylisocyanate at the sites of N-methyl leucine and aminosugar. This dimer is, however, inactive against S. aureus, in


97

OH CH3 CH3 H 2N O O OH OH OO

OH CH3 CH3 H 2N O O

O O

O HO

N H N H

OH Cl O O

H N

N O

OH O

O

H CONH2

H

CH3 NH

N H

N

+ O

CH3

C

N

N

C

O

CH3

O OH OH HO

Eremomycin O

OH CH3 CH3 O O OH OH

H N

HN

OO

OH CH3 CH3 H 2N O O

O

NH

O

O

H N

N H N H

O HO

O

OH Cl O

N O

H

O

N H CONH2

OH CH3 CH3 O N H N

CH3

O OH HO CH3 CH3 HN O O OH OH OO

OH CH3 CH3 H 2N O O

O O

O HO

N H N H

N O

O OH OH HO

OH Cl O O

H N

H

O

N H CONH2

O

OH OH

HN

HN OH O N H

O N CH 3 CH3 CH3

Eremomycin cyclic dimer

Figure 3.17 Structure of eremomycin and its cyclic dimer.

contrast to a highly potent vancomycin monomer. Such a result possibly derives from an unfavorable conformation of cyclic dimers and, importantly, to the lack of basic amino groups at both the N-terminus and the aminosugar domain. 3.1.1.5 Polyvalent Vancomycin Constructs. Considering the high surface density of d-Ala-d-Ala peptide precursors in bacterial cell walls, it would be interesting to examine the antibacterial activity displayed by multivalent vancomy-

98

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS H3C OH N H3C O H O

OO

HO

i Vancomycin

reductive alkylation

O O HO

N H N H

OH Cl O

Cl H N

O N H

H

N

OH OH

O

O

O O

O H R OH O

O

N H CONH2

ii CH3 NH

N H

CH3

olefin metathesis

CH3

O OH OH HO

14

H3 C

OH N H3C O H O

OO O HO O O HO

N H N H

Cl H N O

O

H

O H

O N

H CONH2

OH O N H

NH HO O O OH OH CH3 H 3C

O

OH OH

OH Cl O O

N

O

N

O H

O H H O

CH3 NH

N

N

OO O H

H O

n

Ph HO

CH3 CH3

O

O

O Cl H N O N N O H H H N

HO

OH OH

Cl OH O

O

OH H3C O NH N CH3 H N H CH3 CONH2 O

OH OH

O

HO HO Vancomycin-HN 15 reagents and conditions: i) R-CH=O, NaCNBH3, i-Pr2NEt, MeOH-DMF (1:1), 70oC (80%); ii) Cl2Ru(PCy3)2=CHPh (10-16 mol%), MeOH, 36 h, rt (60%)

Figure 3.18 Synthesis of polyvalent vancomycin molecule via olefin-ring-opening metathesis polymerization.

cin molecules, particularly at high valency. Arimoto et al. [230] reported the synthesis and activity of a polyvalent vancomycin construct (15), presenting glycopeptide side chains on a polymer framework (Figure 3.18). The key synthetic approach to 15 is the olefin-ring-opening metathesis polymerization (ROMP) of 14, which contains a norbornene moiety-linked to the primary amino group of vancosamine. The reaction conditions for ROMP employ methanol as solvent and the Grubb’s ruthenium catalyst [231], which tolerates multiple types of normally reactive functional groups such as amine, acid, and alcohol present in vancomycin. Like vancomycin, the norbornene-derivatized monomer 14 is active against vancomycin-sensitive, gram-positive strains with MIC values below 0.5 mg/mL. However, the same molecule is inactive against vancomycin-resistant strains, with MIC values above 125 mg/mL. The polyvalent construct 15 is approximately 4- to 10-fold less active against vancomycin-sensitive species than the monomer 14, but it is more potent, by 8- to 60-


99

Heterodivalent vancomycin (b)

(a) Lipid II intermediate

(c)

Transglycosylase inhibitor


D-Ala D-Ala L-Lys D-Glu L-Ala HO HO HO AcHN


HO HO HO AcHN Vancomycin


D-Ala D-Ala L-Lys D-Glu L-Ala CH3 O OH O O O O AcHN O O OP P O O-O

Transglycosylase Bacterial inner membrane

Figure 3.19 Possible mechanisms of action displayed by covalent vancomycin conjugate as a bifunctional inhibitor of bacterial cell wall biosynthesis: (a) inhibition of transglycosylation; (b) simultaneous inhibition of transpeptidation and transglycosylation; (c) inhibition of transpeptidation.

fold, against VRE. It is unclear why the polymeric vancomycin becomes less active than vancomycin against vancomycin-sensitive strains. However, the observation raises several possibilities, indicating fundamental di¤erences in cell wall structure between VRE and VSE strains with regard to surface density and spatial distribution of peptide precursors, as well as the ability of giant polymers to penetrate bacterial targets located in the peptidoglycan layer. This specific example demonstrates the e¤ectiveness of applying the multivalency concept to the discovery of a novel class of antibiotics active against emergent vancomycin-resistant strains. 3.1.1.6 Heterobivalent Vancomycin Constructs. Heterobivalent molecules derived from glycopeptide antibiotics are intended to bind two distinct but proximal targets, including a d-Ala-d-Ala cell wall precursor. Such binding probably occurs in either monovalent or divalent manner, as depicted in Figure 3.19. Sun et al. at Princeton [232] reported the synthesis and activity of a vancomycin-based hybrid molecule that is designed to block two separate but spatially related steps responsible for cell wall biosynthesis, such as transpeptidation and transglycosylation. A heterodimer (18) is composed of vancomycin aglycone tethered covalently to 16, a truncated analog of moenomycin [233], which is a phospholipid-anchored inhibitor of transglycosylase (Figure 3.20). Like moenomycin, the disaccharide analog 16 shows antibacterial activity arising from its ability to inhibit the transglycosylation step [234,235]. With an MIC of 0.1 mg/mL, 18 is active against S. aureus and vancomycin-sensitive strains and is more potent than monovalent components such as 16 and 17 (which have similar MICs of 250 mg/mL), and the vancomycin aglycone (MIC of 2 mg/mL). The dimer 18 also shows anti-VRE activity, although it is fourfold less active

100

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Cl H2N HO HO HO OH HO

HO OH HN O

O HOO O AcHN

O OHO

O O HO

F3C HO HO HO O NH

CONH2 O

O OO

O P OH NHAc O O

HO2C

OH

Cl

O C25H41

NH

F3 C HO HO HO O NH

O HN O O O

O O OCONH2 HO P O C12H25

F 3C

NH O HN O O O

F3C

O H3CO

16

Moenomycin

OCONH2

O

OCH3

17 Cl F3C HO HO HO O NH

Cl

NH O HN O O O OCONH2

O AcO AcO AcO O NH

N3 O O O

F3C HO HO HO O NH

i,ii,iii

NH O HN O O O

OCONH2 SPh F 3C

O F3C

F 3C iv,v HO

OCONH2

O O

I HO

N H N H

O O

O

Cl H N

O N O

H

Cl

O N

H CONH2

OH O N H

O

CH3 NH CH3 CH3

OH OH HO

18

reagents and conditions: i) (CF3CO2)2Hg, BF3•ether. 2-chloroethanol, 83%; ii) PMe3, H2O, THF, EtOH, then 4-chloro-3-trifluoromethylphenyl isocyanate, CH2Cl2, DMF, 72%; iii) NaI, acetone, then NaOMe, MeOH, 86%; iv) O-allyl protected vancomycin algycone, Cs2CO3, DMF, 84%; v) PdCl2(PPh3)2, Bu3SnH, DMF, AcOH, 66%.

Figure 3.20 Structure of moenomycin analogs and a heterodimer composed of moenomycin disaccharide analog attached to vancomycin algycon.

than 16 (MIC ¼ 4 mg/mL), whose high anti-VRE activity is due to its inhibition of transglycosylation reaction and is not influenced by a vancomycin resistance mechanism. Nonetheless, the anti-VRE activity of 18 is far better than that of 17 or the vancomycin aglycone, tested either alone or in combination. This example illustrates the e¤ective application of covalent heterodimerization targeting receptors and enzymes located proximally on the surface of bacterial cells. A number of semisynthetic glycopeptides might be considered as functionally bivalent, including a vancomycin derivative displaying a hydrophobic substituent attached to the vancosamine sugar (Figure 3.21). Such examples include a chlorobiphenyl derivative (19) [236] that exhibits about an 80- to 160fold increase in antibacterial potency against both vancomycin-sensitive and vancomycin-resistant strains. Ge et al. [237] and Eggert et al. [238] attribute this enhanced activity to a bivalent mode of action by such vancomycin de-


101

H3C OH N H 3C O H O

OO

HO O O HO

N H N H

OH OH

O

OH Cl O

Cl H N

O N O

H

O N

H CONH2

Cl

OH O N H

CH3 NH CH3 CH3

O OH OH HO

19

H3C

OH N H 3C O H O

OO

HO O

H2N

H O N

N H N H

Cl

O

OH Cl O

Cl H N

O N O

H

O

OH OH

O

N H CONH2

O

Cl

OH O N H

CH3 NH CH3 CH3

OH OH HO

20

Figure 3.21 Semisynthetic vancomycin derivatives presenting hydrophobic aromatic substituent attached at a sugar moiety.

rivatives. Like the vancomycin–moenomycin heterodimer (18), the aglycone pocket of 19 binds the d-Ala-d-Ala peptide, while its hydrophobically modified disaccharide binds certain bacterial surface proteins involved in transglycosylation reaction. Such a mechanism is supported by a desleucyl analog derived from 19. This derivative lacks the ability to bind d-Ala-d-Ala but is still active against bacterial cells through its inhibition ability at the transglycosylation stage [237]. The anti-VRE activity displayed by this type of bivalent molecule is probably mediated primarily by the latter mechanism. Later, Eggert et al. [238] discovered a gene responsible for the susceptibility of Escherichia coli cells to 19. This gene encodes a protein, YfgL, that is identified as a lipoprotein located on the inner surface of the outer membrane of the bacterial cell wall, where it is normally involved in up-regulation of lytic transglycosylases (also known as autolysins). A second example of a heterobivalent glycopeptide acting at both the transglycosylation and transpeptidation stages was reported by Roy et al. at Merck

102


[239]. An inhibitor (20) is derived from vancomycin by attaching a bulky aromatic group to vancosamine sugar (Figure 3.21). It is active against both vancomycin-susceptible and vancomycin-resistant strains at an MIC value of about 0.25 to 2 mg/mL. As identified by a‰nity chromatography, the cellular targets of 20 in E. coli cells include a group of membrane proteins involved in cell wall biosynthesis. The group consists of five penicillin-binding proteins (PBP-1B, PBP-2, PBP-3, PBP-5, and PBP-6) and lytic murein transglycosylase A. This finding supports the fact that 20 interacts rather nonselectively with multiple bacterial proteins involved in peptidoglycan synthesis (PBP-1B, a dual transglycosylase–transpeptidase), cell wall remodeling (lytic transglycosylase A), cell growth (PBP-2), and division (PBP-3). However, such protein-binding ability itself is not enough to explain the potent and broad-spectrum antibacte-

Dansyl NH O NH2 HN AA3 AA2 O AA1 X

NH2 AA3 AA2 AA1

HN

NH2 Val Lys Phe

NH2 AA3 AA2 AA1

O -O N

HN

NH2 Val Lys Phe

HN

N

HN

NH

N

NH2 Val Lys Phe

(X = NH; O)

NH

N

N

N

21 O

22

NH

O

O R

OH O NH O

N H 2N

O

O

N H

H N O NH2

O N H

O

O

CH3 NH

N H

N H CONH2

CH3 CH3

Vancomycin mimic OH O NH O

N CH3 H3C HN CH3 O

H N

H2NOC H N O

H

H 2N O

N

N H

O

H N O

O

O

N H

O

O

H N O NH2

O N H

O

O

N H CONH2

N H

CH3 NH CH3 CH3

N HN

O

H N (CH2)nCH(OH)(CH2)n

N H

Mimic dimer (n = 1, 2, 3)

OH

Figure 3.22 Synthetic vancomycin mimics and their dimer that binds d-Ala-d-Ala (Lac) ligand.

103


rial activity by 20 because a hydrophobic substituent such as N-Ac-N 0 -(3,4dichlorobenzyloxyphenyl)putrescine is equally capable of binding PBP-1B, although it is 100-fold less active than 20. Therefore, 20 probably engages a heterobivalent mode of action in which it sequesters peptide precursors, such as d-Ala-d-Ala (and d-Ala-d-Lac), and interferes with transglycosylation assembly by binding to transglycosylase proteins. 3.1.1.7 Multivalent Vancomycin Mimics. Some synthetic molecules mimic vancomycin as a receptor in their ability to recognize d-Ala-d-Ala peptide residues. Such artificial peptide-binding receptors are less complex in structure and are easier to synthesize than natural glycopeptides of the vancomycin class. Monnee et al. [240] reported a library with 512 members composed of trimeric tripeptides. The library was prepared by a solid-phase method, where each peptide unit is linked to a triazacyclophane sca¤old such as 21, as shown in Figure 3.22. This library includes a selective receptor such as 22, which recognizes the d-Ala-d-Ala peptide ligand in solution, with a K a value of 2.66 10 4 M 1 . It is noteworthy that this receptor also recognizes the N-dansyl-Glyd-Ala-d-Lac—a vancomycin-resistant ligand—with a twofold lower binding constant of 1.01 10 4 M 1 . Ahrendt et al. [241] and Xu et al. [242] reported a novel class of vancomycin-mimicking receptors (Figure 3.22). They are designed on the basis of a monovalent mimic that is composed of a tripeptide attached to half of the vancomycin aglycon. As one of the active receptors identified among the 27,000-member library, the monomer mimic is slightly active in vitro against vancomycin-resistant enterococci, with an activity only fourfold greater than that of vancomycin. Dimers derived from this mimic tethered with a three- to seven-atom spacer show improved activity. For example, mimic dimers (n ¼ 1, 2, 3) show 16-, 8-, and 4-fold greater anti-VRE activity, respectively, than the monomer mimic.

HO O

H N

N H

N O

O

NH2

23

OH O

O N H

CH3

N CH3

NH2

N 24

O2N O N

O-

O O

O2N

O H H H N+

O H O N

Noncovalent complex

O

O-

HO O

O

+H3N

N O

O

- D-Lac N

H2O

O2N

H N

-24

O OH

O

Covalent intermediate

Figure 3.23 Proposed mode of action for catalytic vancomycin mimics that recognize d-Ala-d-Lac peptide as a substrate.

104


3.1.1.8 Catalytic Vancomycin Mimics. Recent advances in the design of small vancomycin mimics include receptors displaying catalytic activity to cleave the peptidoglycan precursor. Chiosis and Boneca [243] demonstrated that prolinebased di- and tripeptides display such activity (Figure 3.23). A tripeptide (23) e¤ectively cleaves the ester bond present in d-Ala-d-Lac under physiologically relevant conditions (23 o peptide ¼ @8; phosphate-bu¤ered solution, pH 7), in which 20% of the depsipeptide is cleaved within 24 hours. A dipeptide (24) is more e¤ective and cleaves 50% of the depsipeptide peptide within 24 hours. Selective cleavage of the d-Ala-d-Lac substrate results from complementary peptide binding in combination with a properly oriented nucleophile–electrophile assembly. The resulting serine-d-Ala ester intermediate is subsequently hydrolyzed to regenerate the original peptide. The catalytic activity of 24 shows a synergistic e¤ect when used in combination with vancomycin. For example,

Ph

HO

O O O AcHN

O

i,ii,iii

OBn

O

AcO AcO AcO AcHN

v,vi,vii

O AcO O O AcHN

HN O

HO HO HO AcHN

O

HO O O AcHN

HN O L-Ala γ-D-Glu L-Lys D-Ala D-Ala

O

O

viii,ix

O O P BnO OBn

O

xiii

O O P -O O O P -O O

HO HO HO AcHN

O AcO O O O AcHN BnO HN O

AcO AcO AcO ThPhHN O O

Br

AcO AcO AcO AcHN

O AcO O O AcHN

HN O L-Ala γ-D-Glu(OTMSE) L-Lys(NTEOC) D-Ala D-Ala(OTMSE)

CH3 OTMSE

7

Lipid II (25)

AcO HO O iv O AcHN BnO AcO HN O AcO AcO O ThPhHN CH3 OTMSE

O

HO O O AcHN


HO O O HO AcHN

O

HO O O AcHN


CH3 OTMSE

x,xi,xii

O O O P BnO OBn

HO O O HO AcHN

O

HO O O AcHN

O


3

reagents and conditions: i) Ala(OTMSE), DIPEA, HOBt, PyBOP, THF, CH2Cl2, rt, 2 h (81%); ii) TsOH, MeOH, 75oC, 30 min (89%); iii) pyridine, AcCl, -30oC, 15 min, then rt (80%); iv) AgOTf, 4 Å mol. sieves, CH2Cl2, -40oC, 3 h (60%); v) Amino-modified resin, 4 Å mol. sieves, butanol, 85oC, 24 h, then pyridine, Ac2O, rt (53%); vi) 10% Pd/C, H2, MeOH, 1 h (93%); vii) tetrazole, (BnO)2PN(i-Pr)2, CH2Cl2, -30oC to rt, 1 h, then m-CPBA, -40oC to rt, 1 h (55%); viii) TBAF, THF, rt, 45 min (87%); ix) H-γ-D-Glu(OTMSE)-Lys(TEOC)-D-Ala-D-Ala(OTMSE), DIPEA, HOBT, PyBOP, THF, CH2Cl2, rt, 2 h (61%); x) 10% Pd/C, H2, MeOH, 1 h (100%); xi) CDI, DMF, rt, 4 h, then undecaprenyl phosphate, rt, 48 h (39%); xii) TBAF, DMF, rt, 24 h, then 3% NaOMe, MeOH, 0oC, 1 h (35%); xiii) PBP1b, HEPES buffer (pH 7.5).

Figure 3.24 Synthesis of lipid II bacterial cell wall precursor (method A).

105


in the presence of 24 (50 mM ), vancomycin shows enhanced activity against vancomycin-resistant Enterococcus faecium by a factor of 8 over vancomycin alone. The synergistic e¤ect of 24 is attributed to its selective hydrolytic action against the d-Ala-d-Lac precursor present in the peptidoglycan layer of vancomycin-resistant cells. This example illustrates a potential benefit that can be achieved when the catalytic unit is linked covalently to vancomycin. Indeed, it was reported earlier by Shi and Gri‰n [243] that semisynthetic vancomycin molecules bearing an imidazole or amine at the C-terminus catalyze the hydrolysis of an activated carbamate attached to a Lys-d-Ala-d-Ala ligand. Such catalysis leads to a rate acceleration of several thousandfold compared to the uncatalyzed reaction. 3.1.1.9 Presentation of Glycopeptide Molecules on Bacterial Surfaces. Lipid II (25) is a transient species serving as an essential building block for cell wall biosynthesis (Figure 3.24). It is a substrate of the penicillin-binding proteins

O AcO O O O AcHN BnO HN O

AcO AcO AcO ThPhHN O

HN O

CH3 O

O

O AcO O O AcHN

AcO AcO AcO AcHN HN O

O

O HO

HN O

AcO AcO AcO AcHN

O i

HN O

CH3 NH

H N

H2NOC O

O AcO O O AcHN O

NH H N

O

O AcO AcO AcO AcHN HN

vi,vii O

O AcO O O AcHN O

O O O O P O P O O- O-

H2NOC O

O

NHCOCF3 NH H N

CO2Me

7 viii 3

H N

O O P HO OH

O

CH3 NH

O

H N O

CO2Me

SO2Ph

CH3 NH

NHCOCF3 H2NOC

O

O

O O O P BnO OBn

CH3 O

SO2Ph

O O P BnO OBn

O AcO O O AcHN

AcO AcO AcO AcHN

ii,iii

CH3 O

SO2Ph

iv,v

O AcO O O AcHN

AcO AcO AcO ThPhHN

i

Lipid II

NHCOCF3 NH H N

CO2Me

O reagents and conditions: i) H2, 10% Pd/C, MeOH, THF (94%); ii) dibenzyl-N,N-diethyl phosphoramidite, 1H-tetrazole, CH2Cl2, iii) 30% H2O2, THF, -78oC to rt (78%); iv) DBU, CH2Cl2; v) EDCI, NHS, DMF, then H2N-γ-D-Glu(CONH2)- Lys(NHCOCF3)-D-Ala-D-Ala(OMe), i-Pr2NEt (46%); vi) CDI, DMF, THF; vii) undecaprenyl monophosphate; viii) NaOH, H2O, 1,4-dioxane (24%)

Figure 3.25 Synthesis of lipid II bacterial cell wall precursor (method B).

106

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS D-Ala D-Ala L-Lys D-Glu L-Ala HO HO HO AcHN

Lipid II intermediate


+

D-Ala D-Ala L-Lys D-Glu L-Ala CH3

O

O HO AcHN

Bacterial inner membrane (Lactic acid bacterium)

HO HO HO AcHN


O

OOO POPO O- O

NH ON O

HO OH


Alexa 488 hydrazide

CH3 O OH O O O O n AcHN O O OP P O O-O

O D-Ala D-Ala L-Lys D-Glu L-Ala HO HO HO AcHN

OH O

reactive ketone epitope

O

NH

O Transglycosylase-mediated transglycosylation

26

O D-Ala D-Ala L-Lys D-Glu L-Ala

O N H

N NH-Alexa 488

NH


D-Ala D-Ala L-Lys D-Glu L-Ala CH3 O OH O O O O n AcHN O O OP P O O-O

Figure 3.26 Multivalent presentation on the surface of bacterial cell of nonnatural glycopeptide.

(PBPs)—a class of bifunctional enzymes that catalyze transglycosyalation (glycosyl transfer of lipid II)—as well as the transpeptidation (peptide cross-linking of lipid II) reaction. Due to the biologically unstable nature of the lipid II intermediate, it is extremely di‰cult to procure from natural sources. Schwartz et al. at Dupont [244] reported the total synthesis of lipid II (Figure 3.24). This synthetic lipid II serves as a substrate in polymerization reactions mediated by PBP-lb-catalyzed transglycosylation and transpeptidation. Shortly thereafter, VanNieuwenhze et al. at Lilly [245] reported a di¤erent method for the total


107

synthesis of lipid II. It is based on a convergent approach using a muramyl dipeptide intermediate (Figure 3.25). The surfaces of bacterial cell walls can be modified chemoenzymatically to present nonnatural ligands in a controlled manner. Sadamoto et al. [246] described such a method in a report in which a ketone-containing group is incorporated into a cell wall precursor (Figure 3.26). For example, a synthetic analog of a lipid I precursor (26) is expressed through the catalytic action of transglycosylase in multiple copies on the surface of gram-positive bacteria, such as lactic acid bacteria. The attached ketone functionality is then used as a handle for further derivatization of the modified peptidoglycan layers, as illustrated by the use of Alexa 488 hydrazide, a fluorescent probe that reacts selectively with ketone groups. The combination of such sequential reactions enables the display of multiple copies of artificial peptidoglycan residues on the outer surfaces of bacteria. These residues can then be visualized by fluorescence microscopy. This strategy applies equally for gram-negative bacteria when performed in the presence of EDTA, an agent added to increase passive permeability through the outer membrane. This technique opens a potential route for the modification of bacterial surfaces with diverse, nonnatural molecules that are otherwise not incorporated through normal biosynthetic pathways.

O

O

H N

N H

Polymyxin B2

NH3+

OH H N

O

NH3+ HN O HO

Lipid A

O

H N

N H

O

NH3+

NH3+ O -O O P HO HO O O

O NH

O O

H N O +H

3N

N H

O O

NH NH

NH3+

OH O

O O O O HO NH O P P O HN O O O O O O- OO HO O OH O O O O O

NH3+

Figure 3.27 Interaction of antibiotic polymyxin B2 and lipid A in gram-negative bacteria.

108


3.1.2

Lipid A

The outer membrane of a gram-negative bacterial cell wall provides a protective barrier against destructive action by various enzymes and chemicals generated by host cells. In particular, lipid A plays an essential role in maintaining cell wall integrity. Certain molecules associate tightly with lipid A and disrupt the organization of outer membranes, thus making bacterial cells more permeable and therefore susceptible to antibiotics. This class of molecules is represented by polymyxin B, a macrocyclic peptide possessing five positively charged ammonium cations derived from multiple diaminobutyric acid residues (Figure 3.27) [247]. The domain on polymyxin B that is involved directly in binding lipid A consists of a peptide-based macrocyclic sca¤old that presents on a single face, three of these ammonium cations. The key groups of lipid A involved in the interaction with the polymyxin include the sugar moieties and multiple phosphate anions. 3.1.2.1 Cholate-Based Polymyxin Mimics. Li et al. [248] reported polymyxin mimics that retain the minimal structural features displayed by polymyxin B that are required for biological activity (Figure 3.28). These mimics are designed using cholic acid as a sca¤old and present three amino groups oriented in a locked conformation similar to that predicted for polymyxin. Polymyxin mimics 27 and 28 inhibit the growth of E. coli with MIC values of 7.0 and 2.0 mg/mL, respectively. Such molecules can induce synergistic e¤ects when tested in combination with other antibiotics, such as erythromycin, which, with

CH3

CH3

CH3 i-iii

O

OH OH

CH3

CH3

OCH3

OTr iv-vi

O O O

OH

CH3

CH3

CH3 OTr

CH3 12

CH3

CH3

CH3 N

17

vii-x

O

7

3

O O

O O

O N3 N3

N3

CH3

R

NH

R

NH

R

NH

27: R = H xi 28: R = C(=NH)NH2

reagents and conditions. i) LiAlH4, THF (98%); ii) tritylchloride, Et3N, DMF (70%); iii) allylbromide, NaH, THF (96%); iv) 9-BBN, THF; H2O2, NaOH (80%); v) MsCl, CH2Cl2, Et3N (99%); vi) NaN3, DMSO (95%); vii) TsOH, MeOH (95%); viii) MsCl, CH2Cl2, Et3N (97%); ix) N-benzylamine (96%); x) LiAlH4, THF (99%); xi) H2NC(NH)SO3H, MeOH (89%).

Figure 3.28 Synthesis of cholate-based multivalent amines as polymyxin mimics.


109

an MIC of 70 mg/mL, is unable to penetrate through the outer membrane and is thus not very active against gram-negative bacteria. The antibacterial activity of erythromycin against E. coli is increased 70-fold, to 1 mg/mL, when measured in the presence of 27 or 28. A comparable synergism is also observed with novobiocin, which is nearly inactive against E. coli. When tested in combination with 28, the MIC value of novobiocin is reduced to 1 mg/mL. The enhanced susceptibility of gram-negative bacteria to those antibiotics that are inactive as single agents is attributed to the binding of multivalent mimics to lipid A. This association probably leads to disruption of the integrity of outer membrane, and thus increases its permeability to these antibiotics. 3.1.2.2 ter-Cyclopentane–Based Polymyxin Mimics. Hubbard et al. [249] reported a unique class of multivalent amines presented on a rigid sca¤old (Figure 3.29). The system is designed on a ter-cyclopentane trimer locked in a defined conformation with the required functionality for chemical derivatization. Such molecules present six copies of phenylalanine (29) or tryptophan (30) tethered to a ter-cyclopentane framework. They bind lipid A tightly under physiological conditions with K d values of 587 nM (29) and 592 nM (30). This binding strength is comparable to the K d value displayed by polymyxin [250]. Therefore, such molecules constitute a new class of synthetic polymyxin mimics, which are structurally unique but less complex than polymyxin. They are

O O i,ii

Ph

H

H

Ph

O

H

iii

O

H H

O

O

H

Ph

O

H

iv

O

O

O

O

v,vi

vii H

H

H 2N

O

O

O H 2N R

H

H

OH H OH

Ph O

O

O

NH2

O H

H

O

HO

O R

O

H

H

Ph O

O

HO

H

H

Ph

Ph

Ph

Ph O

H

viii,ix

O

H

Ph O

O

R

H H

O NH2 R

29: R = CH2Ph 30: R = N H

reagents and conditions: i) K3Fe(CN)6, K2OsO4, quinuclidine, K2CO3, t-BuOH, H2O (1:1); ii) NaIO4, THF, H2O (3:1), 0oC to rt (63%); iii) (i-PrO)2P(O)CH2C(O)CH2CH2Ph, t-BuOK, THF, 0oC to rt, 2h; then bis-aldehyde, -78oC to 4oC (90%); iv) AlMe3, 10 min; then AlCl3, CH2Cl2, 0oC, 10 min; then cyclopentadiene (10 eq), 4oC (87%), v) LiAlH4, THF, rt; vi) Bz-Cl, Et3N, CH2Cl2, rt; vii) O3, CH3OH, CH2Cl2 (1:1), -78oC; then NaBH4, 0oC to rt; viii) Boc-Phe or Boc-Trp, DCC, DMAP, CH2Cl2, DMF (7:3), rt, 18h; ix) TFA, Et3SiH, CH2Cl2, 12 h.

Figure 3.29 Synthesis of ter-cyclopentane-based multivalent amines as polymyxin mimics.

110

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Ph

Ph O H2N

O

O

O

O

O

H O

O

H

H

N H 2N H

porous silicon

NH2

O

O

O

H

O

n

O 10 eq

N H

N H

Si(O)3

NH2 N H

NH2

O

6% DMSO in water

30 (O)3Si

Si(O)3

O

O Ph

Ph O H 2N

O

O O O

HO

OH

O

O O

H N H2N H

O

O HN

NH

O

o

H

H

H

O NH2 N H

m

(m/o ~0.1)

N H

N H 31

Figure 3.30 Multivalent display of lipid A–recognition molecule through covalent attachment to silicon microcavity.

potentially active in neutralizing endotoxin, a liposaccharide, by associating tightly with lipid A. The structure of lipid A is highly conserved with that of several lipopolysaccharide (LPS) subtypes found from various gram-negative bacterial species. Because of its low structural variability, lipid A is often considered as a target of choice when designing a biological sensor to enable the selective detection of gram-negative pathogens. Chan et al. [251] reported an optical device called a silicon microcavity resonator, which utilizes the high a‰nity of 30 to lipid A. The silicon microcavity is derivatized with multiple copies of 30 at a surface density of 10% calculated on the basis of available attachment sites (Figure 3.30). When bound to lipid A, this functionalized resonator (31) shows a specific optical response seen as a red shift by 3 to 4 nm in the photoluminescence spectra of the biosensor. Indeed, the device responds selectively to lipid A–containing cell lysates prepared from gram-negative bacterial species, but not to lysates from gram-positive bacteria. This example illustrates the e‰cient application of multivalent organic receptors when coupled with a silicon microcavity resonator. 3.1.2.3 Natural Multivalent Molecules Bound to Lipid A. During the humoral immune response to bacterial pathogens, the outer membrane of the bacterial


111

cell wall presents common antigenic structures targeted by a complementary cascade system that kills bacterial cells by generating hydrophilic pores through cell membrane. The complement cascade involves the sequential and coordinated activation of 11 distinct complement proteins [252]. The classic complement pathway is initiated by the multivalent binding of IgG or IgM antibodies to antigens presented on outer membranes. This binding triggers the adsorption of C1 complement proteins C1q, C1r, or C1s to a complement-binding site exposed on the Fc region of the immunoglobulins. The C1q is a complex has a molecular weight of 410 kDa and is composed of six collagenlike helixes that are assembled together at one end (Figure 3.31) [253]. At another end, six free globular head groups function as a recognition site. In a study measuring the surface pressure relative to the area per molecule, Ringsdorf et al. [254] reported that the globular head of C1q binds multiple lipid A molecules presented on the surface. This type of recognition is specific to lipid A, as supported by the use of fluorescein-labeled anti-C1q antibody.

C1q

Lipid A

Lipid monolayer

Lipid A HO

NH3+ O HO

-O O P O O O

O O

OH O

O O O O NHHO O HN O P P O O O O O O- OO HO O OH O O O

NH3+

Figure 3.31 Complement component C1q and its multivalent binding to lipid monolayer composed of lipid A and DPPC.

112


3.1.3

Helicobacter pylori Adhesin Receptor

H. pylori is a human-specific gram-negative pathogen that causes gastritis and gastric ulcers [255]. This causative action occurs through bacterial colonization on epithelial cell surfaces, where pathogen–cell adhesion plays a critical role. Multiple pairs of bacterial receptor–cellular ligands are involved in such attachment [255]. First, BabA is 78-kDa adhesin located on the bacterial surface which selectively binds Lewis b (Fuc-a1,2-Gal-b1,3(Fuc-a1,4)GlcNAc-a1,3Gal-b1,4-Glc) glycoconjugates expressed on the cell surface. The surface of one H. pylori bacterium presents about 500 copies of BabA per Lewis b (Le b ) molecule, a density close to the number of fimbriae organelles on the surface of E. coli. Using multiple pairs of BabA–Le b molecules, the bacterium adheres tightly to the surface of epithelial cells with a high-a‰nity Ka value of 8 10 9 M 1 [256]. Other adhesin receptors on the surface of H. pylori include SabA, which recognizes sialyl-dimeric Lewis x glycosphingolipid; hemagglutinin– HpaA, which recognizes sialyl lactose (NeuAc-a2,3-(6)Gal-a1,4-Glc); a 65-kDa protein, which binds phosphatidyl ethanolamine; and Hsp 70, which recognizes neutrophil activating protein (NAP) and sulfated carbohydrates such as 30 -SO3 Gal-b1,R and 6 0 -SO3 -GlcNAc-bR. 3.1.3.1 Synthetic Multivalent Inhibitors of H. pylori. Multivalent interactions implicated in the adsorption of H. pylori to epithelial cells are inhibited by synthetic multivalent molecules. Boreń et al. [257] reported that a neoglycoconjugate composed of the Le b ligand in multiple copies on the surface of human serum albumin is an e¤ective inhibitor of such bacterial attachment (Figure 3.32). For example, a neoglycoconjugate (32) reduces the binding of H. pylori p466 to gastric mucosa by 93% at a concentration of 20 mg/mL, while a second glycoconjugate (33), displaying the H-1 oligosaccharide epitope (Figure 3.32), reduces it only by 48%. The inhibition of such pathogen–cell adsorption is achieved specifically by multivalent molecules presenting Le b ligand. Those presenting non-Le b ligands, such as H-2, Lea , Le x , and Le y , are all inactive at concentrations up to 100 mg/mL. The implication of BabA as a Le b -binding receptor is strongly supported by receptor activity-directed a‰nity tagging, developed by Boreń et al. (Figure 3.33). This photoa‰nity method, coupled with the use of a multivalent Le b glycoconjugate (34) that can be photo-cross-linked, makes it possible to isolate and purify BabA (75 kDa) from virulent strains of H. pylori. Localization of this adhesin protein on the bacterial surface is visualized by electron microscopy in H. pylori cells that were treated with biotinylated Le b glycoconjugate and then with an antibiotin antibody covalently linked to gold nanoparticles. It is often di‰cult to avoid lengthy processes in the chemical synthesis of complex branched oligosaccharides such as the Le b determinant. However, Randolph and Danishefsky [258] developed an e‰cient glycal methodology performed on solid support which enables the synthesis of a form of Le b that


113

Inhibition of cell-bacterium adhesion by Leb neoglycoconjugate

Adhesin (BabA)

H. pylori 32 or 33

Leb

Epithelial cells OH HO

Leb

human serum albumin

O 32

OH HO OH HO O HO O OAcHN HO O O HO HO O O O OH OH O OH HO

33

OH

N H n (=32)

human serum albumin

H-1

OH OH HO HO HO O OAcHN HO O O HO HO O O O OH OH O OH HO

O O N OH

O O N OH

N H n (=35)

OH

Figure 3.32 Multivalent Le b glycoconjugates constructed on human serum albumin protein and its proposed inhibition in the adhesion of H. pylori to epithelial cells.

supports bioconjugation (Figure 3.34). This approach makes it possible to generate Le b -O-allyl glycoside 35, whose terminal allyl group is oxidatively cleaved to an aldehyde functionality, which in turn is used as a handle for tethering the Le b molecule to human serum albumin through reductive amination [259]. The neoglycoconjugate 36 prepared in this way yields approximately 33 copies of Leb saccharide per albumin protein, conjugated to lysine residues. In a special environment where epithelial cells are subject to persistent infection and chronic inflammation, H. pylori exploits a second type of surface adhesion receptor, which is referred to as the sialic acid–binding adhesin (SabA). It selectively recognizes the sialyl-Le x ligand expressed as a form of glycosphingolipids in gastric epithelium [260]. The interaction between multiple pairs of SabA–sLe x thus becomes more significant during certain diseased states, such as chronic inflammation, because sLe x antigens are expressed at a higher level when stimulated by bacterial colonization. Such a secondary type of pathogen–cell adhesion is shown in Figure 3.35.

114

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS OH HO

Leb

albumin

O OH HO OH HO O HO O OAcHN HO O O HO O HO O O OH OH O OH HO

Sulfo-SBED

N H n

NH2

OH 34 multivalent Leb-cross-linker glycoconjugate

OH HO

O OH HO OH HO O HO O OAcHN HO O O HO O HO O O OH OH O OH HO

O O N OH

OH

O O N OH

N H n

BabA adhesin binding;

NH

then UV irradiation

O S S

disulfide bond O

N3 aryl azide photocross-linker

H HN N

NH

HN O

O

biotin

H N

S O

OH HO O OH HO OH HO O HO O OAcHN HO O O HO HO O O O OH OH O OH HO

O O N OH

N H n

NH O

S S

OH

O

N

H HN N

BabA adhesin

NH

HN O

H N

O

S O

photocross-linking BabA adhesin SH Reducing disulfide (DTT);

O

N

H HN N

then avidin affinity purification O

HN O

H N

NH S

O

Figure 3.33 Multifunctional Le b neoglycoconjugate and a receptor activity-directed a‰nity tagging method used for characterization of BabA as Le b -binding receptor.

115

TARGETS IN BACTERIAL CELL MEMBRANES Glycal 1 OTIPS

O

OSi(iPr)2- S O O O

O

HO HO i

BnO F O OBn OSi(iPr)2-S OTIPS O BnO OBn O O OHOO O ii OH

O

OBn O OBn OR'

O OR O OO O O

BnO

O

O OBn OBn

R = Si(iPr)2- S

; R' = TIPS

iii R = R' = H

Glycal 2 BnO OBn HO OTIPS O O OBn Bu SnO O 3 OTIPS HO O OTIPS OH I O O O OO O vi O NHSO2Ph

v

R = R' = TIPS

OTIPS O

O OBn BnO OBn

iv

R3O OR 3 O OR3 OR R1O OR R4O OR O O O OO O O R 4O NHR2 OR R1O O 1

OR O

O OR 3

R3O OR3 Leb-O-allyl glycoside (35)

OH OH

viii

O OH OH HO OH HO OH O O O OO O O HO NHAc OH HO O

OH O

ix

O

OH

O OH HO OH OH OH O OH HO OH O OO HO O

Leb neoglycoconjugate (36) OH

HO OH O O O NHAc OH HO

O

HSA

OH O OH

O

N n = 33 H

O OH HO OH

reagents and conditions: S = solid support; i) (a) 3,3-dimethyldioxirane, CH2Cl2, (b) glycal 1, ZnCl2, THF; ii) Sn(OTf)2, di-tert-butyl-pyridine, THF; iii) TBAF, AcOH, THF; iv) TIPSCl, imidazole, DMF; v) I(coll)2ClO4, PhSO2NH2, CH2Cl2; vi) glycal 2, 4 Å mol. sieves, THF; vii) (a) TBAF, AcOH, THF, (b) Na, NH3, (c) Ac2O, pyridine; viii) (a) 3,3-dimethyldioxirane, CH2Cl2, (b) allyl alcohol, ZnCl2, (c) NaOMe, MeOH; ix) (a) O3, (b) human serum albumin (HSA), NaCNBH3.

Figure 3.34 Synthesis of Le b -presenting neoglycoconjugate.

An additional surface receptor of H. pylori implicated in cell attachment includes HpaA, a 20-kDa hemagglutinin. It recognizes sialyllactose ligands on cell surfaces. Simon et al. [261] report that 3 0 -sialyllactoside inhibits the attachment of H. pylori to gastrointestinal cells. A multivalent molecule composed of 3 0 -sialyllactose displayed on serum albumin blocks such adhesion 100to 1000-fold more potently than does a monovalent 3 0 -sialyllactose.

116

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS BabA

SabA

H. pylori

A

sLe

Leb

x

chronic infection Epithelial cells (disease state)

Epithelial cells (normal state)

Figure 3.35 Simplified mode of adhesion of H. pylori to the surface of gastric epithelial cell in normal and disease states. OH

OH O

HO

OH

OH O

HO HO O

OH O O

HO

HO O

O N H

HO

OH O S

HO

S

OH

HO

37 OH

HO O HO

O OH

38

O HO

OH O

HO

OH

S O HO

OH O O

O N H

HO

H N

S

S

O

OH

O

O OH O HO

OH

HO

OH

O O HO

39

HO OH O

O

OH OH

OH OH

OH OH

OH

O

OH HO

HN

OH O

HO O HO HO

O OH O O

O

S

N H

H N

S

S S

O

40 O NH

OH O O HO

OH O OH O HO

OH OH

O HO HO HO

O

O

OH

O HO OH

HO

Figure 3.36 Structure of di- to tetravalent galabiosides as inhibitors of S. suis– mediated hemagglutination of erythrocytes.

TARGETS IN BACTERIAL CELL MEMBRANES OAc AcO

OAc

OAc O

AcO O AcO

AcO OAc O O

HO

OAc O O

AcO

Br

AcO OH

OAc O

AcO O

i

117

AcO

ii N3

OH O

HO O HO HO

OH O O

O NH2

+

HO

S

C

S

OH 3

iii,iv

40

O

reagents and conditions: i) NaN3, 15-crown-5, 75 oC, 24 h (97%); ii) NaOMe, MeOH; then H2, Pd/C (87%); iii) F5C6OH, DIC, DMF, 0oC to 22oC, 16 h; iv) HOBT, Et3N, DMF, 22oC, 16 h (55% over two steps).

Figure 3.37 Synthesis of tetravalent galabioside.

3.1.4

Streptococcus suis Hemagglutinin

Many bacterial species utilize cell oligosaccharide ligands displayed on the cell surface for bacterium-cell adhesion. Galabiose (Gal-a1,4-Gal) is one of the epitopes recognized by uropathogenic E. coli [262] and Streptococcus suis (gram-positive) [263]. In principle, this ligand is tailored to inhibit cell attachment by such bacteria. Hansen et al. [264] demonstrated that multivalent galabiosides are much more e¤ective than a monovalent galabioside in blocking the cell adhesion induced by S. suis hemagglutinin. Representative examples of bi- and tetravalent galabiosides are shown in Figures 3.36 and 3.37, which summarize the synthesis of a dendrimerlike tetravalent galabioside (40). In an assay that measures the inhibitory activity of agglutination of human erythrocytes mediated by S. suis bacteria, a divalent galabioside (39) a¤ords complete inhibition at 10 nM. This bivalent activity is enhanced 30-fold over a relevant monovalent galabioside (37). A second bivalent molecule (38), which is tethered through a shorter linker, has a lower activity reflected in an IC50 of 130 nM, representing only a twofold higher potency than that of 37. When the valency number is raised to 4, as in 40, the activity is enhanced 300-fold. This is the first example where bacterium–cell adhesion is inhibited completely by soluble saccharides of relatively low valency at nanomolar concentrations. 3.1.5

Escherichia coli FimH Adhesin

E. coli is a primary cause of urinary tract infections. Its pathogenicity correlates with its ability to adhere to the surface of host cells [265]. Pathogen–cell adsorption occurs through multiple specific interactions between the FimH adhesin located on the tips of the pili of E. coli type 1 and mannoside-containing oligosaccharide ligands present on the surface of bladder epithelial cells [266– 268]. Due to its multivalent nature, such surface–surface adhesion is strong

118


enough that microorganisms remain attached tightly to host cells under dynamic physiological conditions, including exposure to shear forces from flowing liquid. 3.1.5.1 Physical Forces Involved in Bacterium–Cell Adhesion. Liang et al. [269] investigated the adhesion forces involved in the interaction of uropathogenic E. coli with a model cellular surface (Figure 3.38). This study uses mannoside-terminated self-assembled monolayers (SAMs) as mimics of epithelial cells in combination with optical tweezers—also known as a gradient force

End-on attachment

Side-on attachment ~0.9 µm

Mannoside

~0.4 µm

mannoside-presenting model surface

HO HO HO

Mannoside

HO HO HO

OH O

OH O

S

S

HN

HN O

O

O

O

O

O

O

O

HO

HO

O

HO

HO

HO

HO

O

HO

HO

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

(CH2)9(CH2)9 (CH2)9(CH2)9(CH2)9(CH2)9(CH2)9 (CH2)9 (CH2)9(CH2)9

S

S

S

S

S

S

S

S

S

S

Self-assembled monolayers (SAM) on Au (gold)

Figure 3.38 Adhesion of E. coli to a mannoside-presenting SAM surface through two modes of attachment (side-on and end-on).


119

optical trap. Using this method it is possible to manipulate submicrometersized bioparticles freely. The E. coli bacterium can attach to the SAM surface that presents a-C-mannoside ligands with variable surface densities ( wman ¼ 101 , 103 , or 106 ). The surface density (wman ) is calculated on the basis of the background of a tri(ethylene glycol) group: wman ¼ mannoside o [mannoside þ (EG)3 ]. The optical tweezers are then used to detach bacteria adhering to the well-defined model surface. The physical force required to detach a bacterium adhered in a side-on orientation is measured in the range 3.5 to b18 pN from SAMs of mannoside density 10% ( wman ¼ 0.1). On the other hand, the force required to detach a bacterium adhering to a SAM surface with a lower density of mannoside at 105 % (A4 mannosides/mm 2 ) is smaller (3.8 to 7.8 pN ). Bacteria also attach to surfaces through an end-on orientation, where the estimated number of attachment points is one or two, even on a SAM with 10% mannoside density. The force needed to detach a bacterium in the end-on mode of attachment is much lower, around 1.7 pN. This force is interpreted as that required to break the monovalent interaction between a pilus and a mannoside molecule on a surface. It is apparent that uropathogenic E. coli bacteria rely on multiple specific interactions to attain tight binding to a model surface. Such multivalent binding as exhibited during side-on adhesion is much stronger than the monovalent binding that occurs during end-on adhesion. 3.1.5.2 Polyvalent Inhibitors of E. coli Adhesin. Li et al. [270] reported the synthesis and activity of poly(acrylamide)-based terpolymers presenting multiple copies of carbohydrate residues such as a-mannoside and lactoside as their side chains. This type of polyvalent mannoside inhibits the adhesion of E. coli to yeast cells. Such inhibitory activity can be attributed to the competitive binding of the polyvalent ligand to the surfaces of E. coli. Other synthetic examples include dendrimer-based sca¤olds used to display mannosides in an array. These may be synthesized using a one-pot reaction that includes monoand dimannose and an amine-terminated dendrimer, as reported by Hayes et al. [270]. 3.1.5.3 Presentation of Mannosides on Gold Nanoparticles. Using self-assembled thiolates on a gold surface, Lin et al. [271] reported the synthesis and biological characterization of multivalent mannosides presented on gold nanoparticles (Figure 3.39). The nanoparticle 41 (diameter ¼ 6 G 1 nm) is manipulated to present approximately 200 copies of covalently attached mannosides, with the synthesis shown in Figure 3.40. The bacterial receptor, adhesin FimH, interacts specfically with surface mannosides, causing the particles to adhere to E. coli type 1 pili. This adsorption is adhesin-specific, such that a mutant E. coli bacterium that lacks adhesin FimH on its pili cannot bind the sugar-coated nanoparticle. The adhesion is very tight, due to the multivalent nature of the binding, and the addition of monovalent mannose is insu‰cient to detach the particles from the pili, even at 20-fold higher mannose concentrations than that

120

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Wild-type E. coli

~6 nm

mannoside-presenting gold nanoparticle (41)

mutant

wild-type

E. coli mutated in type 1 pili

Figure 3.39 Selective adhesion of mannoside-presenting gold nanoparticles to the surface of E. coli type 1 pili.

presented on the surface of the nanoparticles. Detachment is observed when free mannose concentrations are much higher, such as 100-fold (10% detachment) and 2000-fold (roughly 90% detachment). This example illustrates that gold nanospheres presenting multivalent mannoside can serve as an e¤ective probe, capable of binding selectively and tightly to adhesin FimH on the surface of type 1 pili. 3.1.6

E. coli Chemoreceptors

Bacterial chemotaxis is a process whereby bacteria respond physically to specific chemicals [272,273]. It occurs through complex signal transduction events based on receptor-mediated molecular recognition. The E. coli chemotaxis is triggered by the binding of small molecules, such as sugars and amino acids to specific receptors located as clusters on the bacterial plasma membrane [274,275]. The small molecules act as chemoattractants or chemorepellents. For example, galactose binds to a soluble glucose–galactose binding protein HO HO HO

HO HO HO

OH O

AcO AcO AcO

i,ii OH

iii

AcO AcO AcO

Br

OH O

vi

O

OAc O

HO HO HO

2

iv,v

O

OH O O

S

OAc O

~200 mannoside per particle S

S

41 multivalent mannoside presented on gold nanoparticle

O O HO

OH OH OH

reagents and conditions: i) Ac2O, pyridine, DMAP, 90%; ii) HBr, AcOH, 80%; iii) 4-pentenyl alcohol, Hg(CN)2, 88%; iv) AcSH, AIBN, dioxane, 80%; v) cat. NaOMe, MeOH, 97%; vi) AuHCl4, NaBH4.

Figure 3.40 Synthesis of mannoside-presenting gold nanoparticle.

TARGETS IN BACTERIAL CELL MEMBRANES (a)

121

Galactose GGBP Out Membrane In

~90 Å (b)

(c)

Trg (dimeric chemoreceptor)

multivalent galactoside (short)

multivalent galactoside (long)

receptor clustering

Figure 3.41 E. coli chemotactic response via clustering of bacterial chemoreceptors (Trg) by polyvalent molecules: (a) monovalent galactose (Gal), which binds to GGBP and then to a dimeric form of Trg; (b) short polyvalent Gal, where its tethered ligands are unable to span the distance necessary to cluster dimeric Trg; (c) long polyvalent Gal, composed of su‰cient spacer length to make it possible to reorganize dimeric receptors into discrete clusters at a plasma membrane.

(GGBP) (Figure 3.41). Following formation of the galactose–GGBP complex, the complex is able to interact with a galactose-specific chemoreceptor (Trg), which initiates a signaling pathway that eventually guides bacteria to swim toward the galactose nutrient [276]. 3.1.6.1 Polyvalent Galactosides as an E. coli Chemoattractant. Gestwicki et al. at the University of Wisconsin [277] studied chemotactic responses in E. coli to polyvalent ligands such as 43, a neoglycopolymer bearing galactoside ligands at various valencies (Figure 3.42). These ligands cause chemotactic activity such as the tumbling behavior of bacteria, which can be measured as

122

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS HO O HN

O HN

O

Ph

i) [Cy3P]2Cl2Ru=CHPh C10H21N(CH3)3Br water, 1,2-dichloroethane

n-2 O

ii) ethyl vinyl ether

HO HO HO

OH OH OH

O

O

HN

O O

HO HO HO

OH

O

HN HO HO HO

O

OH

O

42 HO HO HO

O

43 (n = 10, 25, 50)

O

O

O

OH

O

OH HN

O

Ph 44 m O

O

HN

HN HO

HO HO HO

O

O

O

O

O

OH HO O HN

23

OH

O N H

O

N H

O

S F F BN N

O

HN O

OH OH OH

O O

O

O

OH OH OH

O

Ph

HO HO HO

O

O

HN HO HO HO

45 O

O

OH

Figure 3.42 Structure of polyvalent galactoside of varying valences that are synthesized through ROMP-mediated polymerization of galactose monomer.

tumbling frequency or mean angular velocity in a bacterial path. The polyvalent galactoside 43 (n ¼ 25) is about 100 times more active than monovalent 42 or galactose. This polymer induces low mean angular velocity even at very low concentrations (e.g., 1 mM per galactoside) and starts to trigger a significant response at 10 mM. Such a significant response can be displayed by a monovalent galactoside (42) but at a higher concentration, 1 mM. The enhanced activity of 43 is attributed to its ability to cluster the chemotactic Trg receptors through multivalent binding (Figure 3.41). This is an example of the important role of ligand valency in bacterial chemotaxis. To further evaluate the relationship between ligand valency and chemotactic activity, Gestwicki et al. [277] devised a capillary accumulation assay that


123

measures the number of bacteria accumulated as a function of the concentration gradient of attractants. The monovalent galactoside 42 and a short oligomer (43; n ¼ 10) show similar activities by eliciting a maximum chemotactic response at 1 mM. Higher-valency polymers (43; n ¼ 25, 50) generate more potent activity, with maximum responses observed at 0.25 mM (n ¼ 25) and 0.1 mM (n ¼ 50). A second generation of multivalent galactoside (44) containing a longer spacer between the attached ligand and polymer backbone is more active than 43 [278,279]. The degree of chemotactic activity is rationalized by considering the length of the glycopolymer molecule and its ability to induce receptor clustering. For ˚ , which is example, 43 (n ¼ 10) has an extended polymer length of roughly 50 A shorter than the spacing between Trg chemotactic receptors (approximately ˚ ) [280]. Therefore, it is unlikely that the polymer is able to cluster Trg re90 A ceptors (Figure 3.41b). Clustering of Trg receptors by polyvalent galactoside molecules is supported by fluorescence microscopic imaging performed using a galactose-bearing fluorescent polymer 45, in which the polyvalent ligand binds specifically to chemotactic receptors and causes receptor clustering on cell surfaces. The clustering of one class of chemoreceptors by a polyvalent ligand influences the response of a second class of chemoreceptors to a di¤erent chemoattractant. Kiessling et al. [277] reported one example using the polyvalent galactoside 43 (n ¼ 25) in E. coli, which not only presents galactose-sensing Trg receptors but also serine-recognition (Tsr) chemoreceptors. Both receptors respond to specific amino acids, such as serine and aspartic acid, as chemoattractants (Figure 3.43). Serine causes a maximum bacterial response at a (a) Serine

Galactose Out Membrane In

Tsr (dimeric)

(b)

Trg (dimeric)

multivalent galactoside

receptor clustering

Figure 3.43 Chemoreceptor clustering involved in potentiation of serine response.

124


concentration of about 100 mM. Pretreatment with 43 (10 mM per galactoside basis) lowers to 1 mM the concentration of serine required for a maximum response. This amplified response is attributed to clustering among the arrays of di¤erent bacterial chemoreceptors. This example illustrates the use of polyvalent ligands as a molecular probe in the study of receptor clustering and signal amplification. 3.1.6.2 Polyvalent Glucosides in the Chemotaxis of Gram-Positive Bacteria. Chemotactic responses are also observed in gram-positive bacteria. Gestwicki et al. [277] reported the correlation between ligand valency and glucose-induced chemotaxis of Bacillus subtilis. This bacterium, unlike E. coli, responds to glucose but not to galactose and galactoside-presenting polymers. In a capillary accumulation assay, a glucose-presenting polymer (47; n ¼ 10) elicited a maximum chemotactic response at 0.1 mM glucose concentration, which is 10-fold greater than 46 and fivefold greater than glucose (Figure 3.44). A highervalency polymer (47; n ¼ 25) at 10 mM generates a response that is 100-fold greater than that of 46. It is clear that ligand valency is strongly associated with chemotactic response. 3.1.7

Nonbacterial Lectins

Entamoeba histolytica is a causative parasite of amebiasis which colonizes in the large bowel. As an essential step toward infection, the colonization is ini-

HO O

HO O

HN O

Ph

HN

HO HO

n-2

OH O O

O

O

HN

OH 46

OH OH

O

HO HO

OH O

HN

O HO HO

OH

OH O

O

OH

47 (n = 10, 25)

BSA OH HO O HO OH n = 40 48

BSA OH HO O HO NHAc n = 39 49

Figure 3.44 Structure of Glu-bearing neoglycopolymer as a ligand of B. subtilis and GalNAc-presenting neoglycoproteins that bind to E. histolytica lectin.


125

tiated by parasitic adherence to the surface of host cells in the colon wall [281]. The pathogen–cell attachment involves specific interactions between lectin receptors on amebic surfaces and N-acetylgalactoside residues linked to epithelial glycoconjugates and colonic mucin. Lectins refer to a family of proteins that recognize specific carbohydrate moieties (for more details, see Section 4.1.16). The adhesion lectin is a 170-kDa heterodimer, whose adherence to target cells is inhibited selectively by N-acetylgalactosides only at 5 to 10 mM. Adler et al. [281] evaluated a series of multivalent glycoconjugates for their a‰nity to E. histolytica lectin. In a lectin-induced hemagglutination assay, natural glycoproteins such as fetuin and asialo-fetuin show inhibitory activity with a MIC value of 0.31 mM. Synthetic neoglycoconjugates that display multiple galactoside residues on protein surfaces show much greater activity (Figure 3.44). For example, the MIC value of 48 (Gal 40 BSA: 40 Gal residues linked to bovine serum albumin) and 49 (GalNAc39 BSA) are 0.15 mM and 5 nM, respectively. Compared to monovalent sugars such as galactoside, with an MIC of 0.71 mM, and N-acetylgalactosamine, with an MIC of 2.5 mM, the activity of 49 is enhanced about 140,000- and 500,000-fold, respectively. Radioligand binding studies with 49, using the displacement of [ 121 I]GalNAc39 BSA from E. histolytica membranes, provide a Kd value of 10 nM, indicating a 200,000-fold tighter binding for 49 over monovalent GalNAc. Therefore, a multivalent Ga1NAc presented on a serum protein shows much a higher avidity to E. histolytica lectin than do natural ligands. 3.1.8

Staphylococcus aureus E¿ux Pumps

QacR belongs to a member of multidrug-binding repressor proteins expressed in S. aureus and regulates the expression of multidrug resistance (MDR) e¿ux pumps [282]. QacA, another MDR pump, is involved in the transport of monoand bivalent cationic lipophilic antiseptics, such as quaternary ammonium ions, outside the cell. The structure of QacR comprises a DNA-binding domain and a drug-binding domain. It initiates the transcription of QacA gene when bound by structurally diverse lipophilic drugs, many of which also are substrates of QacA. Crystal structures of QacR–drug complexes reported by Schumacher et al. [283] suggest that drug binding induces the formation of an extended multidrug binding pocket which presents separate but linked drug-binding sites within a single protein. The complex with a divalent drug, dequalinium, is shown in Figure 3.45. Binding domain 1 creates an environment where a quinolinium group binds through p-interactions between Trp-61 and Tyr-93, and domain 2 interacts with a second quinolinium group sandwiched between Phe162 and Tyr-103. The decamethylene linker is stabilized through multiple van der Waals contacts with multiple hydrophobic side chains contributed from Leu-54, lle-99, Met-116, and Leu-119. Monovalent quaternary cations, including antimicrobial alkaloids, bind to either site 1 (berberine, rhodamine 6G) or site 2 (ethidium). Thus, divalent sites in the binding pocket appear to play a crucial role in the recognition of multiple drugs by bacterial cells. Such a

126

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Phe-162 Trp-61

N H

CH3 N+

Thr-89

OH

H3C

NH2 N +

H2 N Tyr-103 OH

HO

Tyr-93

Binding domain 1

Binding domain 2

Figure 3.45 Structure of dequalinium and its proposed divalent recognition by multidrug-binding repressor protein QacR from S. aureus.

divalent-recognition site is also observed in multidrug transporters, including bacterial QacA and QacB [284,285].

3.2 3.2.1

BACTERIAL TOXINS Heat-Labile Enterotoxins

LT is a heat-labile enterotoxin isolated from E. coli. It is a member of the AB5 family of bacterial toxins, which include Shiga toxin, Shigalike toxin, and pertussis toxin [286–289]. Structurally, LT is a heteromeric complex composed of one enzymatic A subunit as a lethal factor and five B subunits, which play an essential role in adsorption of the toxin to the target cell membrane. As shown in Figure 3.46, the five B-subunits are organized symmetrically, such that each

Galactoside binding site 1 Galactoside

45 Å

Pentavalent ligand

B5-ligand complex

B5 (LT)

Figure 3.46 Model for 1:1 complex between heat-labile enterotoxin (LT) B5 and a pentavalent galactoside.

BACTERIAL TOXINS

127

subunit contains an identical carbohydrate-binding site located in the same face. This pentameric face selectively recognizes galactoside-presenting sugar ligands such as natural ganglioside GM1 [290,291]. The enterotoxin initiates its intracellular invasion process by attachment to cell surfaces, a process attributed to specific, tight interactions between multiple galactoside-recognition sites on a B pentamer and multiple ligands derived from cell surfaces. 3.2.1.1 Pentavalent Galactosides. Fan et al. [292] reported high-a‰nity, multivalent inhibitors of LT-B5 designed through a modular approach. As in conventional rational drug design, this design relies heavily on known structural information of the target. Based on the distance and position of carbohydratebinding sites on B5 complex, a series of pentavalent ligands that are structurally complementary to the pentavalent receptor sites are designed (Figure 3.46). For example, a simple derivative of galactoside is tethered to an azacrown sca¤old through a flexible linker of varying lengths. Each pentavalent molecule presents ˚ , the estimated distance between tethered ligands separated by more than 45 A sugar-binding sites present on two nonadjacent B-subunits (Figures 3.47 and 3.48). As calculated in an extended conformation, the distance between non˚ (50) to 180 A ˚ (53). Howadjacent ligands is approximately in the range 60 A ever, the e¤ective length, calculated on the basis of a polymer model, is in the ˚ (50) to 45 A ˚ (53). In the latter model, the efsmaller approximate range 30 A fective length is proportional to the square root of the molecular weights of two tethered galactosides [52,293]. Such a linker distance suggests that some pentameric ligands bind the pentameric target in a multivalent way because of similarity in their molecular dimensions and shapes. In an ELISA assay [292], which evaluates the ability of various molecules to inhibit the binding of LT-B5 to ganglioside, pentavalent galactosides show greatly enhanced activity compared to that of a monovalent galactoside, as shown by the IC50 values of 242 mM (50), 16 mM (51), 6 mM (52), 0.56 mM (53), 5 mM (monomer 54), and 58 mM (galactose). This binding result provides a positive correlation between linker distance and a‰nity. For instance, ˚ (53), which the highest a‰nity is achieved at an e¤ective distance of 45 A matches the estimated distance required for a pentavalent association. This specific inhibitor shows a binding strength 105 -fold greater than that of monovalent galactose and 10 4 -fold (or 2000-fold on a valency-corrected basis) greater than that of a monomer control (54). However, 53 is still less active than is LT’s natural ligand, GM1 -OS, which shows an IC50 of 0.01 mM. In place of a plain galactoside, a more active ligand monomer such as GM1 -OS can be attached to a multivalent sca¤old to prepare a more potent pentavalent ligand. Indeed, a pentameric GM1 -OS equivalent to 53 leads to an inhibitory activity about 100 times higher per GM1 -OS ligand than that of the GM1 -OS monomer. By comparison, multivalent presentation of a weakly binding galactoside enhances the avidity of a multivalent ligand by four or five orders of magnitude, whereas a similar multivalent molecule presenting the more tightly binding GM1 -OS exhibits a less dramatic improvement.

128

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS HO HO OH O

HO

NH L = distance (at extended dimension)

O

HO OH

HO HO

Linker

O H N OH O

Linker

O

NH

O

OH OH OH

N N Linker

N N

N

Linker

Linker

O HN O

O OH HO OH

OH

HO HO OH O Linker =

N H

O N H

OH

HN

O

(CH2)2[CH2OCH2]3(CH2)2

O

O

N H n

N H

Azacrown O

50 (n = 1): L = ~60 Å (effective length = ~30 Å) 51 (n = 2): L = ~100 Å (effective length = ~35 Å) 52 (n = 3): L = ~140 Å (effective length = ~40 Å) 53 (n = 4): L = ~180 Å (effective length = ~45 Å)

HO HO HO

O H N OH

O

O

O

N H

N H

(CH2)2[CH2OCH2]3(CH2)2-NH2

54

Figure 3.47

Structure of pentavalent galactosides presented on 15-azacrown-5 sca¤old.

3.2.1.2 Multivalent Inhibitors Displayed on a Solid Support. Multivalent sugars immobilized on a solid support are e¤ective agents capable of neutralizing toxins. They act by tightly binding B5 units. Nilsson et al. [294] reported the synthesis of such glycopolymeric material presenting multiple sialyllactose or lactose immobilized on silylaminated Chromosorb P (Figure 3.49). In a cytotoxicity assay, sialyllactose-presenting Chromosorb polymers, represented by 57 and 58, show antibacterial activity by protecting cells from heat-labile enterotoxin of E. coli. Approximately 95% of toxic activity is blocked. These polymers also neutralize cholera toxin by blocking about 98% (57) and 95% (58) of toxin activity, while the nonderivatized Chromosorb P shows no block-

BACTERIAL TOXINS

129

R1 N

R1 N

R1 i,ii,iii

N CO(CH2)3NH2 N

N R1

R1

R2

R2 N

R2 N

N CO(CH2)3 N

O

O

N H

N H

N R2

R2

R3

R3 N

R3 N

N CO(CH2)3 N

N R3

R3

O MeO

iv

(CH2)2[CH2OCH2]3(CH2)2 -NH2 n

O

O

N H

N H

(CH2)2[CH2OCH2]3(CH2)2 NH n

50 (n = 1) 51 (n = 2) 52 (n = 3) 53 (n = 4)

O N H

(CH2)2[CH2OCH2]3(CH2)2-NHBoc

HO HO HO

HN

O

O H N OH

HO HO HO

O

O H N OH

55

O

O

O

O

N H

OMe

56

reagents and conditions: i) 55, MeOH, NaHCO3, H2O; ii) TFA; iii) repeat steps i and ii; iv) 56, NaHCO3, H2O

Figure 3.48 Synthesis of 15-azacrown-5-based pentavalent galactosides.

ing ability. This study illustrates the potential of sugar-containing solid supports as potential therapeutics in the treatment of bacterial toxin–mediated diarrheal disease. 3.2.2

Cholera Toxin

Cholera toxin, a member of the AB5 enterotoxin family, comprises one Asubunit—a 27-kDa enzyme that catalyzes ADP ribosylation of Gs-a protein— and five copies of an 11.6-kDa B-subunit—a lectin that recognizes galactoside ligand such as that found at the termini of GM1 molecules [295]. The distance between adjacent galactoside-recognition sites in the cholera toxin B-pentamer ˚ . Cholera toxin adheres tightly to cells through specific is estimated to be 31 A interactions between the five B-units and multiple GM1 gangliosides expressed on cell surfaces, with a Kd value ranging from 4.6 1010 to 5 108 M for

130

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS OH OH AcHN HO

CO2O HO HO O HO

Chromosorb P

O HO O OH

OH O N

H N

H N n (2.1 µmol incorporated/g)

O

OH O 57

-O2C HO HO

O

HO OH O HO O O OH

HO

O HO O OH Sialyllactose

H N n (2.4 µmol incorporated/g)

O

OH O

AcHN OH

CO2O HO HO O

H N

OH 58

OH OH AcHN HO

Chromosorb P OH O N

Chromosorb P OH OH

O OH

+

H2N

H N

H N O

i,ii

57

m

reagents and conditions: i) MeOH, AcOH, 60oC; ii) Ac2O

Figure 3.49 Multivalent sialyllactose immobilized on solid support as a bacterial toxin –binding agent.

the B-pentamer and GM1 ganglioside in solution [295–297]. The association of free GM1 gangliosides to the cholera toxin B-pentamer occurs in a cooperative manner as a result of enhanced intersubunit interaction upon sequential binding [298]. 3.2.2.1 Galactoside-Presenting Pentavalent Inhibitors. In a subsequent application of rational design, Zhang et al. [299] (see also Williams and Davies [300]) reported the synthesis of azacrown-cored multivalent ligands ending with branched galactosides. These multivalent molecules are designed to target two pairs of cholera toxin B-pentamer simultaneously (Figure 3.50). By virtue of a variable linker length, it is possible to optimize the binding of galactoside ligands, presented on the branched arms of a decavalent molecule, to a Bpentamer. These ligands are highly e¤ective inhibitors of cholera toxin B binding to ganglioside. When assessed by ELISA, the decavalent ligands show IC50 values of 191 mM (59; n ¼ 1), 5.3 mM (59; n ¼ 2), 1.4 mM (59; n ¼ 3), and 40 nM (59; n ¼ 4). The activities observed improve with increasing linker length in the range tested, and are 200- to about 1,000,000-fold greater than that of a monovalent galactoside, which yields an IC50 of 38.5 mM. In ELISA, 59 (n ¼ 4) makes a 1:1 complex with the cholera toxin B-pentamer with an apparent Kd value of about 40 nM. This value is supported by dynamic light scattering (DLS), which measures the hydrodynamic radius of the B-pentamer. The study indicates that a linker length of four iterations in 59 yields the

BACTERIAL TOXINS

131

OH OH OH HO OH OH OH HO

O O

HN HO HO

O

HO HN

OH NH

O

O

O O HN

O

O

N N

O Linker

NH O

O

O

NH HN

HN O HO HOHO

OH OH OH

O

O

H N

N H O

NH

O

OH

OH HO OH

O

OH

O

O NH

HN

O HO HOHO

O Linker =

O

O O

O

HO N H

N Linker

O

H N

O N Linker

O O

O

O

N

HN OH

NH

O

Linker

O Linker

HO

O

O

NH

OH OH OH

O

HN O

HO

HO HO

O NH

N H

O

OH

OH HO OH

O

O

HN (CH2)2[CH2OCH2]3(CH2)2 N nH

N H

O

OH

Azacrown O

59 (n = 1, 2, 3, 4)

Figure 3.50 Structure of branched galactosides presented on pentavalent arms of azacrown sca¤old.

greatest binding enhancement. At su‰ciently high concentrations of cholera toxin, a single molecule of 59 can associate with two cholera toxin B-pentamers by forming a sandwich complex (Figure 3.51). This arrangement is confirmed by crystal structure determination and DLS in solution [299]. Based on the DLS study, this sandwich complex, although less stable than 1:1 complex, appears considerably stable, with a Kd value estimated to be in the submicromolar range. Merritt et al. [301] reported the next generation of multivalent molecules composed of symmetric pentacyclen-based multivalent galactoside as an inhib-

132


+

B5 (cholera toxin)

decavalent ligand

sandwich complex

Figure 3.51 Formation of a 2:1 sandwich complex between cholera toxin B pentamer and a branched decavalent ligand.

itor of cholera toxin and E. coli heat-labile enterotoxin (Figure 3.52). A pentavalent inhibitor (62) is composed of five copies of m-nitrophenyl-a-d-galactoside (61), a monomer that has IC50 of 195 mM, reflecting a binding activity to cholera toxin B-receptor that is about 100 times higher than that of alkyl-a-dgalactoside. The pentamer 62 displays much greater activity, with an IC50 value

OH O HO HO OHO

OH O HO HO OHO

O

BocHN H 2N

O

O

O

OH

O

NO2 O

O MeO

OMe

NO2 OH O HO HO OHO

O

O

N H

i,ii

O

60 O

O

N H

OMe

O O

O

N H

iii NO2

iv O

61 HO

O OH HO HO

O OHO

Spacer

O N H

O

O

O

O

N H

HN

O HN Spacer

O

NO2

O

O N

O

O

N N

O O

H N Spacer

O

O

O

N H

O

HO

O OH

N H

NO2

N

N

O

O O

O N H

OH OH

O OH

N H

2

NO2

OH O HO HO OHO

NH2

O

N H 2

N Spacer H

NO2

N H

HO

O OH

OH OH

NO2 62

reagents and conditions: i) cyanuryl chloride; ii) TFA, CH2Cl2; iii) dimethyl squarate, NaHCO3, pH 8, H2O, rt , overnight; iv) squarate derivative of 15-azacrown-5, NaHCO3, pH 9, H2O, MeOH, rt, 36 h.

Figure 3.52 Synthesis of pentavalent m-nitrophenyl-a-d-Gal constructed on 15azacrown-5 core as an inhibitor of cholera toxin and E. coli heat-labile enterotoxin.

OH OH

BACTERIAL TOXINS

133

of 0.9 mM, which represents a 263-fold gain in inhibition activity compared to that of a single-finger ligand (60). Compared to the pentavalent galactoside 51 ˚ , 62 shows about an 18-fold encontaining an identical spacer length of 35 A hancement in activity. This example demonstrates that multivalent molecules derived from a high-a‰nity monomer can provide much better activity than a low-a‰nity monomer, although it depends on the system and is probably limited to relatively low valency molecules. 3.2.2.2 Mode of Action of Pentavalent Galactosides. With DLS analysis in solution along with crystal structure determination, Merritt et al. [301] demonstrated that the potent antitoxin inhibitory activity exhibited by 62 derives from a tight 1:1 association between a pentavalent ligand and a toxin pentamer. For example, the hydrodynamic radius (Rh ) of enterotoxin B-pentamer remains unchanged when bound by 62, even at high concentrations of the multivalent ligand, supporting the fact that toxin–ligand interaction leads to the formation of a 1:1 complex rather than multiple aggregation (>2:1) complexes. A crystal ˚ resolution provides further structure of the cholera toxin–62 complex at 1.46-A details about such interactions, by showing that each of five m-nitrophenyl-a-dgalactosides binds to each of five ligand sites present on the toxin with a fivefold axis (C5 ) of symmetry. 3.2.2.3 Dendrimer-Based Multivalent Lactosides. Although lactose binds weakly to the cholera toxin B (CTB)-subunit in the millimolar concentration range [302], the multivalent presentation of lactosides enhances the binding a‰nity to CTB. Vrasidas et al. [303] reported the synthesis of multivalent lactosides constructed on a dendrimer sca¤old (Figure 3.53). When evaluated for binding to CTB using a fluorescence quenching assay, the lactoside-displaying dendrimers show reduced K d values, such as 235 mM (63), 99 mM (64), and 33 mM (65). These values represent tighter binding by 38-, 46-, and 68-fold, respectively, than that of free lactose, which has a K d value of 18 mM. This binding result suggests a correlation between increasing ligand valency and increased binding a‰nity. The Hill coe‰cient (n) derived from three multivalent molecules binding to CTB pentamer lies in the range 1.4 to 1.5. An n value greater than 1 involves positive cooperation, possibly through a bridging mechanism. Thus, a dendritic molecule such as 65 is likely to bridge two gal˚ apart on the cholera toxin B-pentamer, actoside-recognition sites about 31 A provided that there is appropriate interligand spacing close to such a distance. 3.2.2.4 Polymerized Liposomes Incorporating GM1 . Taking advantage of the specific interaction between ganglioside and cholera toxin, Charych et al. [304] designed an artificial membrane-based sensor sensitive for detecting cholera toxin. The design of the biosensor incorporates GM1 (66) into a matrix of polymerized diacetylenic lipid (10,12-pentacosadiynoic acid) (Figure 3.54). As introduced in Chapter 2, polymerized liposomes presenting GM1 show a characteristic blue coloration, with a l max of about 630 nm that arises from the

134


OH OH

HO OH OH

HO HO HO

O OH OH HO O OH HO O

HO

HO

O

HO

O HO

NH

S

NH

S

O

O

O

NH

O

HN

O

O

NH

NH

O

NH S

NH

NH

NH S S NH

O

HO

O

HO

NH

NH S

O

O OH HO

O

O OH HO HO

O OH OH HO O OH HO O

HO

OH OH OH OH OH OH HO O HO O O OH OH O OH OH

O

HN O

O

O

OMe 63

64

O

OMe

OH OH OH OH OH OH OH OH OH OH HO O OH OH OH HO HO O HO O OH OH O HO O HO OH OH O O O HO OH OH OH HO O OH HO O OH OH HO OH OH OH HO O O OH HO O HO OH O HO O OH HO O HO NH HO HO O O OH S NH O O OH NH HO O NH HO S HO NH HO O S NH S OH NH NH S NH NH O HO S NH NH HO O O NH O S NH O O S NH O O NH O HN

O

HN

O

O

O HN

HN

O O O

O

O

O

NH

HN

O

O

O

65 MeO

O

Figure 3.53 Structure of lactose-presenting dendrimers.

BACTERIAL TOXINS OH GM1 (66) OH OH HO O O HO O O OH OH AcHN H OH O HO N O -O2C O O O OH O O HO H C 3 OH HO OH AcHN OH O

CH3 i) sonication (liposome);

CH3 +

ii) hν, 254 nm (cross-linking)

HO

OH OH OH HO O O HO O O OH OH AcHN OH O HO O -O2C O O O OH O HO n OH HO OH AcHN OH

135

GM1 Cholera toxin CO2-

Polymerized liposome (χ ~ 5%) blue color

OH OH OH HO O O HO O O OH OH AcHN OH O HO O -O2C O O O OH O HO n OH HO OH AcHN OH

CO2-

red color

Figure 3.54 Preparation and use of cross-linked poly(diacetylene) liposome presenting multiple copies of GM1 on its surface.

absorption of the conjugated poly(diacetylene) chromophore. When exposed to cholera toxin at 40 ppm, the liposome solution immediately undergoes color transition to red, with a l max value of about 550 nm. This red-to-blue color transition occurs because multiple toxin–GM1 recognition events that occur on the surface of the liposomes perturb the poly(diacetylene) framework. The lower limit of detection by such a method is about 0.1 nM of cholera toxin. The Ka value of cholera toxin to the GM1 ligand present on a poly(diacetylene) membrane is about 3 10 10 M 1 when studied by surface plasma resonance (SPR) spectroscopy [304]. This value is in good agreement with binding a‰nity of the toxin to GM1 on the natural surface of living cells. Such tight interaction is also supported by a separate SPR study reported by Kuziemko et

136


al. [305], where cholera toxin binds to GM1 displayed on the surface of biosensor membranes with a Kd value of 4.6 1012 M. The toxin also recognizes asialo-GM1 presented on the same biosurface but with 50-fold lower avidity, as reflected by the Kd value of 1.88 1010 M. 3.2.2.5 Liposomes Presenting Fluorescent GM1 . The sensitivity of bacterial detection can be improved to 0.05 nM of cholera toxin according to a technique developed by Song and Swanson [306]. This approach is based on an optical technique called fluorescence resonant energy transfer that is performed on biomimetic membranes. For example, cholera toxin is allowed to interact with fluorophore-labeled GM1 expressed on lipid bilayers in which the multivalent toxin–ligand interaction causes a dramatic change in fluorescence of the labeled GM1 . 3.2.3

Shigalike Toxin

Like other toxins in the AB5 family, Shigalike toxin (SLT) also comprises pentavalent B-subunits that serve as a sugar-binding lectin [307,308]. Each Bsubunit contains two nonidentical sugar-binding sites that recognize Gal-a1,4Glc-b1,4-Glc carbohydrate, or Pk trisaccharide, linked to glycolipid Gb3 or CD77 [309,310]. The distance between these two sites on a single unit is pro˚ , while the distance between two neighboring sites posed to be about 20 A ˚ [311]. Using SPR spectroscopy, located on two adjacent subunits is 18.5 A Nakjima et al. [312] studied the kinetics of Shiga toxin in the binding of isoforms 1 and 2 to cell-mimicking surfaces, including derivatized dextran surfaces that present multiple copies of Gb3 ceramide glycolipid. The interaction probably involves bivalent binding of the glycolipid with each Shiga toxin unit. This type of interaction is specific to Gb3 Cer surfaces since Shiga toxins 1 and 2 do not recognize GM3 -presenting surfaces and bind only weakly to GB4 Cer surfaces. 3.2.3.1 Bivalent Pk Trisaccharides. Lundquist et al. at Duke [313] reported the synthesis of bivalent ligands derived from C-linked Pk trisaccharide, a natural ligand recognized by pentavalent SLT-1B (Figure 3.55). For example, 68 and 69 are composed of two Pk ligands attached to a pentapeptide framework. The divalent ligand 68 inhibits SLT-1B in a competitive ELISA assay with an IC50 value of 20 mM, which represents about a 100-fold enhancement over the a‰nity of the monovalent Pk ligand (67). The thermodynamic parameters of binding of the divalent Pk ligand to SLT-1B were assessed by isothermal titration microcalorimetry. In these studies, the K a value of 68 is 11,000 M 1 . The binding stoichiometry (n), defined as the ratio of ligands to toxin units, is approximately 1. The association constant indicates about 20-fold stronger binding than that of the monomer 67, which has a K a value of 500 to 1000 M 1 and a binding stoichiometry of 1. A second ligand (69) composed of a hydrophilic spacer shows about 10-fold enhancement in binding, with a K a value of 68001

BACTERIAL TOXINS O H2N

CH3 O

H N

N H

O

CH3 O

H N

CH3 CH3

N H HN

O

O HO

NH2

HO

Pk trisaccharide

HO

O NH2

67

O HO

OH OH O HO O OH O HO O OHHO

H2N

CH3 O

H N O

OH O OH

N H

CH3 HO

O

NH2 NH

NH2

HO

O HO

68

O

OH HO O O

O HO

HO O OH O HO O OHHO

H2N OH O OH

H N

O N O H HO2C

HO H N

OH O OH HO

O NH2

N H HN

O

HO

O

NH2

NH2

NH O

OH

HO O OH

HO

OH OH O

OH

HO O OH

CH3

N H HN

O

OH HO O O

OH O HO OH

CH3 O

H N

137

69

O HO

OH HO O O

OH

HO O OH OH O HO OH

Figure 3.55 Monovalent and bivalent Pk trisaccharides displayed on a peptide sca¤old.

but with a distinct binding stoichiometry (n ¼ 2). It appears that these two dimers bind SLT-1B in di¤erent modes, depending on the nature of linker. The hydrophobic, bivalent ligand 68 binds a B-subunit with 1:1 stoichiometry (i.e., two Pk units bound per B-monomer), while a hydrophilic bivalent ligand (69) binds with a ratio of 2:1 (i.e., four Pk units bound per B-monomer). Based on the crystal structure of SLT-1B, Lundquist et al. suggested that 68 would bind within a subunit in which two Pk units of a bivalent ligand span two distinct ˚ apart, on a single B-subunit. Clearly, the sugar-recognition sites, roughly 20 A linker length is an important parameter that a¤ects both binding a‰nity and stoichiometry, possibly through little-studied e¤ects, such as contact interaction with a protein surface and dimer orientation. Another class of divalent ligands that bind to the Shigalike toxin has been reported by Kitov et al. [314]. A representative dimer from this class is composed of two Pk trisaccharides or galabiosides tethered with a bisamide linker at the O-2 0 atom. This dimer binds bivalently by occupying two sites present on a single B-subunit. The a‰nity of the dimer is enhanced 8- to 47-fold over that of the monomeric galabioside and is influenced by the length of the spacer. 3.2.3.2 Decavalent Pk Ligands. Kitov et al. at the University of Alberta [315] reported that a Pk-based multivalent carbohydrate ligand is a potent inhibitor of Shigalike toxins I and II (SLT-I and SLT-II). The structures of a divalent Pk

138


HO HO

OH O

HO HO

HO O OH O HO O OHHO

HO

HO O OH O HO O O HO

OH O OCH 3 OH

OH O

HO

OH O

O

HO O OH OH O O O OCH HO 3 O HO O O OH N H

N H

OH O OCH 3 OH

O

Pk trisaccharide methyl glycoside 70 HO HO

OH O

HO

OH O

HO

HO O OH O HO O O HO O

HO O OH OH O O O OCH HO 3 O HO O O OH N H O

O

N H

OH O OCH 3 OH

O

NH (CH2)7 HN O

R

O O

R

O

H N

S O O

R

O

O

N H

O

=R

R 71

Figure 3.56 Structure of multivalent Pk inhibitors of Shigalike toxin.

ligand (70) and a decavalent Pk ligand (71)—so-called ‘‘starfish’’ molecules— are shown in Figure 3.56. The dimer 70 can block the binding of a Pk-based glycoconjugate to SLT-I or SLT-II with an IC50 value of 55 mM and a K a value of about 105 M, which represent 40-fold higher activity than a monovalent Pk trisaccharide. However, 70 is far less active than the decavalent 71, which displays an IC50 value of 0.24 nM. This subnanomolar activity of the starfish molecule represents more than a 106 -fold increase in inhibitory activity over the Pk trisaccharide. It correlates well with the estimated dissociation constant ((K d @ 109 M ) of a B-subunit pentamer binding to the surface of a cell that expresses glycolipid Gb3 (Gal-a1,4-Gal-b1,4-Glc-b1-O-ceramide). In a cytotoxicity assay, the decavalent molecule protects cells from SLT I and II at micromolar concentrations. Kitov et al. determined the crystal structure of the decavalent Pk ligand 71 to the SLT-1B pentamer (Figure 3.57). The crystal structure shows that one molecule of 71 interacts with two pairs of B-subunit pentamers in the pseudotwofold symmetry of a 2:1 sandwich complex. The complex forms in a unique fashion such that the two ligands at the end of a single arm of 71 cannot bind adjacent sites on a single B-subunit. Rather, the two terminal ligands bind to di¤erent subunits in separate pentamers. Ultimately, five copies of the Pk li-

BACTERIAL TOXINS Starfish molecule (71)

139

Top view: bottom half of B5-ligand sandwich

Pk trisaccharide

Figure 3.57

Mode of binding of a ‘‘starfish’’ molecule to Shigalike toxin I.

gand interact with the second B-subunit pentamer, thus creating the sandwich complex illustrated in Figure 3.51. 3.2.3.3 Polyvalent Polymeric Pk Ligands. The use of a polymer framework makes it possible to increase the valency of Shiga toxin inhibitors. Gargano et al. at Syracuse [51] reported that such a polyvalent polymer is a potent inhibitor of the SLT-1B pentamer. Its design concept is shown in Figure 3.58. The polyvalent inhibitor is synthesized via the copolymerization reaction of an acrylamide, derived from Pk ligand 72, with unsubstituted acrylamide monomer. During the reaction, the ratios of the two acrylamides are varied to produce molecules with varying degrees of substitution. In a cell-based assay that measures the ability of the polymers to inhibit Shiga toxin 1, three variants of poly(72-co-acrylamide), with ligand densities (wPk ) of 0.1, 0.2, and 0.3, show IC50 values in the range 800 to 900 nM per Pk trisaccharide. These activities represent a 5000-fold enhancement in the inhibition of Shiga toxin 1 relative to a monovalent ligand such as methyl- or octyl-Pk trisaccharide, which have IC50 values greater than 5 mM. The potent activity of the polyvalent molecules is specific: Incorporation of lactose instead of the Pk saccharide in the polymer chain leads to inactive polymers. In a model of the AB5 toxins, Gargano et al. [51] proposed a theoretical equation useful for calculating the binding enhancement contributed from multivalency. The term binding enhancement (b) reflects the strength of a multivalent association relative to the monovalent association (for a detailed discussion, see Section 1.2). It is based on several limited conditions that define multivalent receptor–ligand interactions, such as equivalent binding sites, no cooperative interaction, a flexible linker with optimal length, and no linker– receptor interaction. The general equation for multivalent binding enhancement is expressed as F ½sK a ð102 Þ n1 where K a is the association constant of the monovalent ligand–receptor complex, n is the valency number (smaller than either the number of sites on a receptor or the number of tethered ligands on a multivalent molecule), F is a statistical factor defined by the system (b1), and s is a factor that adjusts the distance between receptor sites and is defined as 30

140

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Pk trisaccharide SLT-1 B subunit

Polyvalent Pk polymer

OH OH O

HO

HO O OH O HO O OHHO

OH O O OH

O

O +

N H

copolymerization

H2N acrylamide

Pk monomer 72 OH OH O

HO

HO O OH O HO O OHHO

OH O

O

OH

NH2 O HN OH OH O

HO

HO O OH O HO O OHHO

OH O OH

O

HN O

O O O O NH2 NH2 NH2 NH2 OH OH O HO HO O OH HO

O

NH2 O

O


O OHHO

OH O

O

n O O NH2 NH2

O

OH

poly(72-co-acrylamide) (χPk = [72]/{[72] + [acrylamide]} = 0.1, 0.2, 0.3)

Figure 3.58 Inhibition of Shiga toxin 1 by polyvalent Pk polymer and its synthesis.

divided by the distance between receptor sites in angstroms. For example, the binding enhancement contributed by a polyvalent Pk molecule binding to Shiga toxin 1 is calculated as follows: b A 1½1 10 3 ð102 Þ ð51Þ ¼ 10 4 per ligand, where K a ¼ 10 3 M 1 , s A 1, F ¼ 1, and n ¼ 5. Thus, the calculated value of binding enhancement is comparable to the value of (5 10 3 ) obtained experimentally. 3.2.3.4 Multivalent Presentation of P1 on a Solid Support. The application of a‰nity chromatography makes it possible to isolate and purify the B-subunit of

141

BACTERIAL TOXINS

HO

OH OH O

P1 antigen (73)

HO O OH O HO O OHHO

HO

agarose matrix OH H O O O (CH2)n NHAc

+

reductive amination H2N N H

NHNH2 m

OH OH O HO O OH O HO O OHHO

OH O O (CH2)n NHAc

H H N N

N N H H 74

AcHN (CH2)n O O HO

OHHO O O

OH

m

HO O OH OH O HO OH

Figure 3.59 Immobilization of P1 trisaccharide on agarose solid support used as an affinity matrix.

Shiga toxin. This technique is based on the tight and specific receptor–ligand interaction mediated by multivalency on the surface of a solid support. Pozsgay et al. [316] reported the preparation of an a‰nity matrix derived from the P1 trisaccharide ligand (73) (Figure 3.59). This a‰nity matrix (74) proves to be a useful tool that enables the isolation of subunit B on a preparative scale (100 mg). 3.2.4

Verotoxin

Another AB5 class of bacterial toxins is VT-1, produced by enterohemorrhagic E. coli. In the solution structure, its B pentamer is symmetric, with each Bsubunit presenting a site for binding globotriaosylceramide (Gb3 ) with a K d value of 1 10 3 M [317]. Due to this structural feature and low a‰nity displayed by a monomer, the toxin pentamer is targeted by multivalent Gb3 molecules or equivalents. For example, SAMs on a gold surface presenting multiple copies of a disaccharide ligand (Gal-a1,4-Gal, a truncated version of Pk trisaccahride) serve as a sensor to detect verotoxin (National Institute of Advanced Industrial Science and Technology, Japan). 3.2.5

Bacillus anthracis Anthrax Toxin

B. anthracis produces anthrax toxin, a causative agent responsible for severe life-threatening symptoms [318,319]. The adverse e¤ects of the toxin are initiated when bacterial protective antigen (PA) binds to a surface receptor on a mammalian cell. This surface binding initiates a series of events that includes proteolytic activation of PA to PA63 and self-association of the PA63 monomer into a heptameric complex (Figure 3.60). This complex is then utilized as a receptor capable of binding two bacterial enzymes, edema factor (EF, adenylate cyclase) and lethal factor (LF, a protease), with a high a‰nity (K d @1

142

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS self-association to heptamer PA 63

binding of EF/LF

receptor to PA63

Cytoplasm Endocytosis

polyvalent P1 (75) inhibition of EF/LF binding

Cytoplasm

75: poly(acrylamide)(P1) (χP1 = [P1]/ [acrylamide] = ~22/900 = 0.02)

poly(NAS)

O R

O R

O R

n O

O R

R

i,ii O O O O O O O O O O O O NH2 NH NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH2 NH NH2

n ~180 P1 peptide

(R = NHS) AcNH-HTSTYWWLDGAPK-CONH 2

AcNH-HTSTYWWLDGAPK-CONH 2

reagents and conditions: i) P1 peptide, Et3N, DMF, rt, 24 h; ii) NH4OH; then dialysis

Figure 3.60 Entry of anthrax toxin complex (PA 63 heptamer-EF/LF) into a cell via receptor-mediated endocytosis (top) and prevention of the cell entry by binding of polyvalent inhibitor to heptameric PA 63 (bottom).

nM ) to EF/LF. Formation of this high-a‰nity complex enables intracellular tra‰cking via endocytosis of the enzymes, which are lethal to macrophages and phagocytes. 3.2.5.1 Polyvalent Peptide Inhibitors. Mourez et al. at Harvard [320] demonstrated that a tailored multivalent molecule is e¤ective in preventing the harmful action of the toxin through interference at the stage of EF/LF binding to the heptamer complex (Figure 3.60). This multivalent molecule is designed on a sca¤old of poly(acrylamide), which presents through its side chains multiple

BACTERIAL ENZYMES

143

copies of P1 peptide, with about 22 peptides and 900 acrylamide units per molecule, thus yielding a ligand density of approximately 0.02 (Figure 3.60). The P1 peptide was discovered from a phage display library raised against PA63 heptamer. P1 binds to the heptamer with only low a‰nity, as reflected by the IC50 value of about 150 mM. A polyvalent inhibitor (75) inhibits the binding of a radiolabeled LF to PA63 on CHO cells with an IC50 value of 20 nM. This binding activity represents about a 7500-fold increase on a per peptide basis relative to a monomeric P1 peptide. In a cell culture assay that measures the cytotoxicity of PA63 , 75 blocks the toxicity with an IC50 value of 6 nM on a per peptide basis. Such activity is of significance since neither monomeric P1 nor the underivatized poly(acrylamide) control displays inhibitory activity.

3.3 3.3.1

BACTERIAL ENZYMES Transpeptidases

The bacterial cell wall is made up of multiple layers of polymeric cross-linked peptidoglycans in the inner and outer membranes of the bacterial cell [321,322]. The cross-linking of peptidoglycan precursors is catalyzed by transpeptidase, the enzyme that belongs to a family of penicillin-binding proteins and is the target inhibited by b-lactam antibiotics. During catalysis of the cross-linking reaction, this enzyme recognizes two separate Lys-d-Ala-d-Ala strands from two closely located, growing peptidoglycan strands, as depicted in Figure 3.61. One strand supplies a nucleophilic component and the second supplies an electrophilic component [323]. 3.3.1.1 Divalent Substrates. To understand precisely the mechanism of the transpeptidase-mediated cross-linking reaction, Lee et al. [324] designed a hybrid molecule (76) composed of cephalosporin and d-Ala-d-Ala. This hybrid molecule, which mimics those produced by transpeptidase-mediated cross-linking, is designed to bind bivalently to two nonidentical substrate-recognition sites present in the transpeptidase enzyme. An x-ray structure that is solved at ˚ resolution supports such a design principle, where, through bivalency, 76 1.2 A occupies an active site of a bifunctional d-Ala-d-Ala carboxypeptidase/transpeptidase. This structure serves as a snapshot of the enzyme bound with the cell wall components positioned to undergo the cross-linking reaction. 3.3.1.2 Multivalent b-Lactams. Divalent b-lactams are expected to target two separate penicillin-binding proteins (PBPs) and show higher a‰nity than the corresponding monomeric b-lactam. Rodriguez-Tebar et al. [325] reported improved activity from covalently linked penicillin dimers as shown in Figure 3.62. For instance, 77 is a dimer derivative of penicillin V, and 78 belongs to a divalent penicillin linked with a lengthy glycine–azobenzene spacer attached to the amino group at the C6 position. These dimers can cross-link two molecules

144

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS NH3+ CO2MurNAc

N H

Me H N O

O

H N

N H

CO2-

O

GluNAc

PBP acylation

Me H N Me

O

-(D)-Ala

CO2PBP enzyme

NH3+ site 1 for peptidoglycan strand 1

CO2MurNAc

N H

O

Me H N

N H

CO2-

O

GluNAc

O

H N

Me

O

O

Acyl enzyme intermediate

B

O

+H3N CO2-O H N

MurNAc

Me H N

N NH H

O

CO2Me

O N Me H

CO2-

site 2 for peptidoglycan strand 2

GluNAc

NH3+ CO2Cross-linking of peptidoglycan strands

MurNAc

N H

O

Me H N O

CO2-

GluNAc

N H

H N O

O

Me H N

O CO2-O

MurNAc

H N

N NH H

Me H N O

CO2Me

O N Me H

CO2-

GluNAc H N O

O

S

O

N

N NH H

CO2-O 76

Me H N O

CO2Me

O N H

CO2-

Figure 3.61 Transpeptidation that is catalyzed by a class of penicillin-binding proteins (PBP) and an example of its bivalent inhibitor.

BACTERIAL ENZYMES HO

O

O O O

N H

S

O

OH

O

O O

N

145

N N H

S OAc

OH 77 O HO

O

O O

N N H

S

O

H N O

N

H N

N H

O

N H N

H N

H 2N O

H

H2N

CO2K N S

O

OH

OAc

O

O

H N S

Cl

O

O

O

Cl

O O

N

i-iii

+ O

O

O KO2C

O OAc

CO2tBu

N H

H

O

Cl

S

N

O

AcO

S N

O

78

OAc CO2H

H N O

O

S

N

N H

O

OAc H

O

NH O

H

S OAc

N CO2K

H O 79

reagents and conditions: i) pyridine, CH2Cl2, 0oC to rt; ii) trifluoroacetic anhydride, CH2Cl2; iii) potassium α-caproate

Figure 3.62 Structure of penicillin dimers and cephalosporin trimer.

of penicillin-binding protein per penicillin dimer, such as PBP1b, 1c, or 3 isolated from E. coli. Kovaćs-Kulyassa et al. [326] reported the in vitro activity of trivalent blactam molecules composed of cephalopsorin linked to a trihydroxybenzne sca¤old (Figure 3.62). This class of trimeric b-lactams shows variable antibacterial activity and is sensitive to specific bacterial strains. For example, 79 is as active as cefuroxime against Streptococcus faecalis, with an MIC value of 3 mg/mL. However, it is 10-fold less active against Staphylococcus aureus than the cefuroxime monomer, which has an MIC value of 1.5 mg/mL. Polyvalent, polymeric b-lactam antibiotics have been designed on a copolymer sca¤old made of vinyl alcohol and vinyl amine or vinyl carboxhydrazide (Figure 3.63). Such polyvalent penicillins (80 and 81) present multiple copies of benzylpenicillin and phenoxymethylpenicillin as polymer side chains, as reported by Usakov et al. [327] and Panarin et al. [328]. Like the parent b-

146


HO HN

n

O

OH O

O O

N

n NH HN O

N H

S

O O

N N H

S

X

81 X = CH2: 57 mol% Penicillin G X = CH2O: 1.5 mol% Penicillin V

80 (2.5 mol% Penicillin G)

Figure 3.63 Structure of polyvalent polymeric b-lactams.

lactams, these macromolecules are water-soluble, stable, and active in antibacterial assays. 3.3.2

E. coli Pyrophosphokinase

Transfer of a pyrophosphate group from ATP to 6-hydroxymethyl-7,8-dihydropterin (HP) is catalyzed by HP pyrophosphokinase (HPPK), a microorganism-specific enzyme that is absent in humans. This reaction constitutes a first step in the biosynthesis of folate cofactors (Figure 3.64) [329]. The active site of this enzyme presents two proximal sites: one site for dihydropterin and a second site for ATP. These sites are arranged to bring two di¤erent substrates close together. Based on this structural knowledge, Shi et al. [330] designed bi-

6-Hydroxymethyl-7,8-dihydropterin (HP) ATP NH2 O N

HN H2N

N

OH

+

N H

O O O -O P O P O P O- O- O-

N O

O

N

N

Mg

N

H2 N AMP

HO OH

N

HN

2+

N

N H

O O O P O P OO- O-

HP pyrophosphate

HPPK (6-hydroxymethyl-7,8-dihydropterin pyrophosphokinase)

Monomer

Divalent inhibitor O

O N

HN H2 N

N

N

OH

N

HN H2 N

N

N

O O P O n O-

NH2 N O

N

N N

HO OH Petrin (82)

83 (n = 2, 3, 4)

Figure 3.64 Structure of bivalent inhibitors of 6-hydroxymethyl-7,8-dihydropterin pyrophosphokinase.

BACTERIAL ENZYMES

147

valent inhibitors (83) of HPPK, each of which is composed of a pterin moiety tethered to an adenosine moiety through varying numbers of phosphoryl groups. Binding and inhibition studies with E. coli HPPK show that 83 (n ¼ 4) has the highest a‰nity, and inhibition activity with a Kd value of 0.47 mM and an IC50 value of 0.44 mM. The activities of dimers tethered with shorter linker lengths are lower. Compared to unlinked monomers, a divalent inhibitor 83 (n ¼ 4) shows an a‰nity that is enhanced about 116 times over that of MgADP and is 76 times higher than that of 6-hydroxymethylpterin (82). Shi et al. ˚ resolution. determined the crystal structure of an 83: HPPK complex at 1.85-A The structure shows that the dimer indeed simultaneously occupies both dihydropterin and ATP-binding sites, thus supporting the notion of divalent recognition. 3.3.3

Protein Tyrosine Phosphatase

Certain members of the protein tyrosine phosphatase (PTPase) family are believed to be involved in various forms of diseases caused by Yersinia and Salmonella bacteria [331]. Since these bacteria do not possess phosphorylated tyrosines in their cellular system, they exploit PTPase to target the host. One class of PTPase inhibitors, the aryl-a-keto acids, bind to a PTPase catalytic site through interactions involving multiple contacts [332]. Chen and Seto [333] reported a multivalent approach in their design of aryl-a-keto acid–based PTPase inhibitors (Figure 3.65). For example, divalent inhibitors are constructed by linking two units of aryl-a-keto acid to a benzene sca¤old at three di¤erent orientations (ortho, meta, and para). In an enzyme assay against Yersinia PTPase, a dimer (85) in the meta orientation shows the best activity, with an IC50 value of 0.7 mM. This activity represents about a 214-fold increase relative to a monomer (84). The ortho and para orientational isomers of 85 are also active inhibitors of PTPase but are less e¤ective than the meta isomer by a factor of 12 and 6, respectively. The activity of a trimer (86) is enhanced as well. It is noteworthy that the orientation of divalent ligands substantially influences the molecule’s activity, perhaps by controlling interligand distance and the direction of ligand units. 3.3.4

DNA Gyrase

Psammaplin A (85) is a natural product isolated from Psammaplysilla sponge [334,335]. It is a symmetric dimer composed of two units of a bromotyrosine analog that are connected through a disulfide linkage. Its biological action includes antibacterial and antitumor activity, although its cellular target and mechanism of action have yet to be unveiled. Multiple enzyme targets, including DNA gyrase, topoisomerase II, farnesyl protein transferase, and leucine amino peptidase, are thought to be inhibited by this antibiotic [336,337]. Nicolaou et al. [338] reported a total syntheis of psammaplin A, starting from l-tyrosine (Figure 3.66).

148

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS O

O CO2H

O

HO2C

O

CO2H O

O

84 O

O

HO2C

O O

O

OH HO2C

85 (ortho, meta, para)

NH2

86

NH-Boc

i,ii

CO2H

CO2Me

iii

CO2Me O

HO

O

iv,v

CO2Me O

O

O

O

MeO2C O

NH-Boc

Boc-HN

CO2Me

HO

O

vi

O

HO2C

CO2H O

O 85 (meta)

reagents and conditions: i) SOCl2, MeOH (100%); ii) (Boc)2O, NaHCO3, H2O-dioxane (95%); iii) 1,3-bis(bromomethyl) benzene, Na, EtOH (49%); iv) TFA, CH2Cl2 (91%); v) HO2CCHO, CuSO4, AcOH, pyridine (19%), vi) NaOH, H2O (69%).

Figure 3.65 Synthesis of bivalent and trivalent aryl-a-ketocarboxylic acid.

The fluoroquinolone class of antibiotics, represented by ciprofloxacin, target bacterial type II topoisomerase enzymes, including DNA gyrase and topoisomerase IV [339]. Dimers derived from fluoroquinolone were reported by Kerns et al. [340,341] as potent antibacterial agents (Figure 3.67). The dimers are composed of either ciprofloxacin or norfloxacin linked with an unsaturated alkene or aryl spacer at a piperidine moiety. For instance, dimers 1 and 2 are more active than ciprofloxacin against S. aureus, including methicillin-sensitive and methicillin-resistant strains as well as glycopeptide-intermediate-resistant S. aureus strains. The activity of dimer 1 is much greater than that of dimer 2, and it is strongly influenced by the nature of the linker. This class of quinolone dimers underscores the validity of divalent design as a strategy for overcoming bacterial drug resistance. 3.3.5

Dihydrofolate Reductase

Dihydrofolate reductase (DHFR) is an NADPH-dependent enzyme that catalyzes the reduction of dihydrofolate to tetrahydrofolate [342]. Because of its unique existence in E. coli, DHFR serves as a drug target for treating bacterial infections. There are a variety of ligands to DHFR, including methotrexate

BACTERIAL ENZYMES OH

OH

Br

Br O

O S S

N H

HO N

N H

N

OH

Psammaplin A (85) OH

OH

OH i

O

OH

Br

Br

Br

ii

O OH

OH

O OH

O

N

NH2

NH2

iii

OH

O

CF3 OH

OH

Br iv

Br

v

vi

O

Psammaplin A

O OH

THPO N

THPO N O

O N

O

reagents and conditions: i) KBr, KBrO3; ii) TFAA, heat; iii) aq. TFA; iv) THPO-NH2; v) EDC, NHS; vi) cystamine, Et3N.

Figure 3.66 Syntheis of psammaplin A.

O

O F

HO

N

N R

NH

Ciprofloxacin (R = cyclopropyl) Norfloxacin (R = ethyl) O

O F

HO N R1

N

N N

N

R2 N OH

F O

O

Fluoroquinolone dimer-1: R1 = R2 = cyclopropyl Fluoroquinolone dimer-2: R1 = cyclopropyl; R2 = ethyl

Figure 3.67 Structure of symmetric fluoroquinolone dimers.

149

150

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS O NH2 N

N N

H 2N

CO2H OH

N H

O

N CH3

N

Methotrexate (MTX) O

O NH2 N

N H 2N

N

N

i

N CH3

N

H 2N

N

N

CO2H N H

N CH3

H N

N

CO2Bn OH

N H

N

N H2N

O NH2 N

NH2

OH

O

N CH3

N

O

O

O O

N H

CH3 N

H N HO2 C

ii,iii

N N

N

NH2 N

NH2 O

86 reagents and conditions: i) α-benzyl glutamate, PyBOP, DMSO; ii) 4,9-dioxa-1,12-diaminononane, PyBOP, NMP; iii) LiOH•H2O, MeOH.

Figure 3.68 Synthesis of methotrexate dimer.

(MTX), a folate analog that displays a high a‰nity (IC50 ¼ 0.6 nM). Although this enzyme exists as a monomer, it undergoes dimerization in the presence of a synthetic, divalent ligand. Kopytek et al. [342] reported such a molecule (86) based on MTX (Figure 3.68). The bis-MTX binds simultaneously to two DHFR enzymes, thus forming a ternary complex. According to gel filtration analysis, the binding is tight, with a K a value of 1.69 10 9 M 1 . The o¤-rate of the complex is about 100 times lower than that of MTX–DHFR complex. Thus, the divalent ligand ensures a longer half-life for the ternary complex. The enhanced strength of the ternary complex is attributed to the additional contribution from favorable protein–protein contacts. 3.3.6

Sialidase (Neuraminidase)

Bacterial sialidase or neuraminidase catalyzes the hydrolytic cleavage of Olinked sialic acid expressed as a form of terminal sugar in glycoproteins and glycolipids on cellular surfaces [343,344]. It belongs to a modular enzyme that presents two adjacent interacting domains located near an active site, one for catalytic action and another for the recognition of one or more lectin domains. Thobhani et al. [345] reported a multivalent mode of action displayed by bacterial sialidase in the hydrolysis of a sialic acid–presenting synthetic glycopolymer (Figure 3.69). Two sialidases from Clostridium perfringens and Vibrio clolerae can cleave O-silaloside with terminal galactosides presented in multi-

BACTERIAL ENZYMES

151

cleavage site HO

HO

AcHN

-O2C O O OH OH HO

OH

OH O O HO OH

O O NHAc

NHAc

Monovalent sialyllactoside

HO

HO CO2O

AcHN

O

OH OH HO

OH O O HO OH

OH O O NHAc

NH2 O

HO

NH

HO

AcHN

CO2O

O

OH OH HO

OH O O HO OH

HN O

O O O O NH2 NH2 NH2 NH2

NH2 O

O


O

n O O NH2 NH2

OH O O NHAc

HO HO

-O2C

AcHN

OH OH

O

OH

O O HO OH OH

O

OH O O NHAc

Polyvalent sialyllactoside poly(NeuAc-co-acrylamide); χSA = 0.18

Figure 3.69 Polyvalent 3 0 -sialyl N-acetyllactosamine presented as a side chain of poly(acrylamide).

ple copies on poly(acrylamide). The catalytic e‰ciencies (Vmax /Km ) of the two sialidases are, respectively, 46- and 32-fold greater when cleaving the poly(acrylamide) derivative than when cleaving the monovalent sialyllactoside. The enhanced enzymatic activity results from tighter binding of the polyvalent substrate, since the value of Km is lowered 55- to 142-fold compared to that of the monovalent substrate. This increased association is attributed to the occupation of the lectin domains by terminal galactosides generated by catalytic action of the sialidase [346]. Therefore, the partially hydrolyzed silalylglycopolymer binds in a multivalent manner using its generated galactosides (to the lectin site) as well as O-sialoside (substrate site). Thobhani et al. showed further that the polyvalent, galactoside-presenting polymer was 100-fold more e¤ective than the galactoside monomer for inhibiting the action of bacterial sialidase. 3.3.7

NAD Synthetase

Nicotinamide adenine dinucleotide (NAD) synthetase is a homodimeric enzyme responsible for the synthesis of NAD, which occurs via the catalytic conversion of deamido-NAD. The active site of this enzyme, as resolved by crystal structure, shows that two distinct but proximal sites are involved in the recognition of substrates such as ATP and nicotinic acid adenine dinucleotide (NaAD) [347,348]. On the basis of such structural information, Garcia et al.

152


[349] invoked rational design to synthesize dimeric inhibitors composed of N-methylnicotine tethered to an indole (or indolylaromatic group or structure) through (un)saturated spacers of varying length. The indole moiety serves as an adenine mimic of ATP. The resulting dimers show micromolar inhibitory activity against NAD synthetase from B. subtilis and are e¤ective inhibitors of gram-positive bacteria.

3.4 3.4.1

BACTERIAL NUCLEIC ACIDS Ribosomal RNA

3.4.1.1 Divalent Aminoglycosides. Aminoglycosides belong to a class of antibiotics displaying a broad spectrum of antibacterial activities. Their mode of action involves the inhibition of translation during bacterial protein synthesis, where they interact with A-site rRNA on the 16S domain of the ribosome [350]. For example, neamine interacts with the A-site by recognizing G–C base pairs selectively and forming a 2:1 complex (two drugs per rRNA) with a Kd value of 10 mM for each binding site (Figure 3.70). Therefore, connecting two identical copies of the drug with a spacer would be an e¤ective approach to enhancing the binding a‰nity to the targeted nucleic acid as proposed by Hendrix et al. and Sucheck et al. at Scripps [351]. This hypothesis was tested using a set of divalent neamine molecules with amine or amide linker functionalities and linker lengths that varied from 2 to 12 atoms. For example, a dimer (88) binds tightly to an immobilized rRNA sequence mimicking the A-site with a Kd value 1495

C CGCUG

AA

5'- G G C G U C A

1490

GUGG

G

CACC U

H 2N HO HO

C

O NHO2 HO

U

NH2 NH2 OH

Neamine covalent linking

N3 BnO BnO

H2N HO HO

O N3O O

N3

O

N3 OBn

i-iii

H2N HO HO

O NHO2 O HO

NH2

O NHO2 O

NH2 NH2 OH

NH2 OH

OH N

N H3C

CH3 12-atom spacer

87

88

reagents and conditions: i) N,N'-dimethyl-1,4-diaminobutane, EtOH, 95oC, 16 h; ii) P(CH3)3, THF, H2O; iii) 20% Pd(OH)2/C, H2, H2O, AcOH.

Figure 3.70 Recognition sites of neamine in an E. coli 16S ribosomal RNA A-site and synthesis of neamine dimer.

BACTERIAL NUCLEIC ACIDS

153

of 40 nM and in 1:1 stoichiometry. Such binding represents a 250-fold enhancement over neamine. This dimer also shows improved antibacterial activity, with an MIC value of 6.25 mM, which is about eightfold lower than that of neamine. 3.4.1.2 Divalent Aminoglycoside Bound to a Dimerized A-Site. Kirk and Tor [352] and Tok and Hu¤man [353] reported a series of homo- and heterodivalent aminoglycosides (89 to 91) designed by combining neomycin and tobramycin (Figure 3.71). Each monomeric aminoglycoside binds to a 16S E. coli ribosomal A-site with K d values of 0.44 mM for neomycin and 2.11 mM

16S A-site rRNA

CC GCUG 5'- G G C G U C

AA

16S A-site rRNA

GGCGUC A

CA

A

AA

C CGCUG

GU

GUGG

G

CACC U

C U

aminoglycoside covalent linking OH O

OH O

NH2 OH

HO H2N

89

NH2

NH2

O O NH 2

Neomycin

S OH OH NH2

S

O

S

H2 N O HO HO

Neomycin

HO

90

HO

HO H2N O O NH2

S

OH NH2 O O H 2N OH

NH2 OH S

O

O

NH2 NH2

O

S

Tobramycin

S

O

NH2

H 2N O O

NH2 OH

OH

Tobramycin

O

NH2 H 2N

OH O HO H2N 91

O O NH 2 NH2

OH O

H 2N O NH2 OH

S O HO O OH H2N

OH

NH2

NH2 OH S OH OH NH2

Neomycin

Tobramycin O

S S

O NH2 H 2N

S O HO O OH H 2N

H 2N O

O NH2 OH

OH

Figure 3.71 Structure of divalent aminoglycoside and its binding to a dimerized A-site in E. coli 16S ribosomal RNA.

154


for tobramycin. Divalent molecules can, in principle, bridge two recognition sites present on RNA structures. RNA models bearing such dimeric RNA binding sites are constructed artificially by grafting two A-sites into a dimerized RNA 16S A-site [353]. The K d value of a neomycin–neomycin homodimer (89) to a dimerized A-site (46-nucleotide construct) is lowered to 20 nM, thus representing about a 19-fold stronger association than that of monovalent neomycin. Such enhancement is attributed to divalent binding since the a‰nity of 89 to a monomeric 16S A-site of 27 nucleotides is not enhanced but is similar to monovalent recognition by neomycin. The binding a‰nities of other classes of dimers (90 and 91) to a dimerized A-site are not enhanced, but rather, resemble that of the monomer. This indicates that the binding activity of these divalent aminoglycosides is strongly influenced by the nature of the aminoglycosides. 3.4.1.3 Polyvalent Polymeric Amikacin Derivatives. Amikacin—a derivative of kanamycin with a 4-amino-2-hydroxybutanoyl group—belongs to a family of aminoglycoside antibiotics that act similarly to neomycin and kanamycin. Tanaka et al. [354] reported the synthesis of a polymeric, polyvalent amikacin. This aminoglycoside contains an amine moiety as its side chain, which provides a handle for derivatization (Figure 3.72). Thus, a 4-vinylbenzolyl monomer (92) is prepared and polymerized in the absence or presence of 9 equivalents of Kanamycin: R = H Amikacin: R = H2N

HO HO H2N

O OH

HO HO H 2N

OH

O HO

HO O

HO O HN

H 2N

O NH2

OH NH2

O

O

HO O

HO

HO O RHN

HO HO H2N

i

OH

O

OH

O HO

HO O HN H N

OH

OH NH2

ONH 2

HO O O NH2

OH NH2

ii or iii

O

OH

Amikacin

92

HO HO H 2N O

HO HO H2N

OH

O HO

HO O HN H N

HO O O NH2

OH NH2

O

O

HO

HO O HN H N

R

93 (Homopolymer)

HO O O NH2

OH NH2

O

OH

OH

n

OH

O

R

R

R

(R = CONH2) m

94 (Copolymer: [amikacin]/[acrylamide] = 0.1)

reagenst and conditions: i) 4-vinylbenzoic acid succinimidyl ester, MeCN, H2O, rt (51%); ii) 2,2'azobis- (2-amidinopropane) dihydrochloride (1 mol%), H2O, 60oC, 8 h (52%); iii) acrylamide (9 eq), 2,2'-azobis- (2-amidinopropane) dihydrochloride (3 mol%), H2O, 60oC, 4 h (70%).

Figure 3.72 Synthesis of polyvalent polymeric amikacin.

MULTIVALENT MOLECULES AS SYNTHETIC VACCINES

155

acrylamide to produce amikacin-presenting polymers such as homopolymer 93 and copolymer 94. In an assay that measures the ability of drugs to inhibit protein biosynthesis, a monovalent amikacin derivative (92) shows an IC50 value of 5 mM, which is 50-fold less active than amikacin itself. Some of this loss may stem from the conversion of the basic amine in amikacin to the neutral amide, as the protonated amine probably plays an essential role in amikacin’s binding. However, the polymeric amikacins 93 and 94 do not show reasonable activity. A nonoptimized spacer distance in these polymers is likely to fail to allow the amikacin side chains to bridge multiple rRNA sites. In a standard antibacterial growth assay, which assesses the ability of drugs to penetrate bacterial cell walls and to reach rRNA targets, both of the amikacin polymers are not active at 100 mg/mL, while the monovalent 92 shows only modest activity, with an MIC value of 40 to 60 mg/mL.

3.5 3.5.1

MULTIVALENT MOLECULES AS SYNTHETIC VACCINES Synthetic Shigella Vaccines

Carbohydrates expressed on the surface of pathogenic bacteria, such as capsular polysaccharides (CP) and O-specific polysaccharides (O-SP), constitute essential virulence factors. They serve as antigens for raising human antibodies in order to initiate immune reactions [355]. Several commercial CP-based vaccines, including Pneumovax 23 and Pnuimmune, target pneumococci and gram-positive bacteria [356]. Pozsgay [357] reported a chemical approach for preparing a synthetic vaccine against Shigella dysenteriae type 1. This bacterial strain presents on its surface an O-SP complex that consists of repeating tetrasaccharide units of composition Rha-Gal-G1cNAc-Rha. The synthetic vaccine encompasses multiple copies of the O-SP hexadecasaccharide fragment displayed on human serum albumin (Figure 3.73). Varying the ratio of saccharide to protein in the reaction leads to a series of multivalent glycoconjugates (95) with tetra to hexadeca valencies. 3.5.2

Synthetic Mycobacterium Vaccines

The microorganism Mycobacterium avium intracellularae (MAI) poses serious threats to immune-compromised persons. Its outer membrane surface presents species-specific oilgosaccharide haptens in the form of glycopeptide lipids [358]. Tri-, tetra-, and pentasaccharides are attached through O-glycosidic linkages to the threonine unit of a tetrapeptide (Figure 3.74). Representative examples of such glycopeptide antigens include Serovar 8 and 21, both of which contain immunodominant 4,6-pyruvylated b-d-Glc-Tal(6-deoxy)-Rha. Ziegler [358] reported the synthesis of a Serovar 21 analog, a trisaccharide neoantigen of MAI, and its glycoconjugate form, which is built on bovine serum albumin. This multivalent conjugate (96) presents 12 copies of the trisaccharide per albumin molecule.

156

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Hexadecasaccharide fragment of O-specific polysaccharide HO HO H C HO 3

O N

OH O

O

O

HO H 3C O OAcHN O HO

OH O HO HO HO O H 3C O HO O OAcHN O HO HO H 3C O O HO HO OH OH OH HO O HO

O OH

O

O 3

HO O H3C O O OAcHN O H3C HO O HO HO 3 OH O O HO HO HO O H C OH O 3 HO O OAcHN O HO HO H 3C O O HO HO OH OH HO HO

OMe OMe

(17 eq)

O HO

O

O

HN NH 2

i,ii

O O HN N H

iii H O

OH

O HO O H3C O O OAcHN O H C HO 3 O HO 3 HO OH O O HO HO HO O H3C OH O HO AcHN O O O HO HO H 3C O O 95 (n = 4 to 16) HO HO OH OH

O O HN N H

H N n

HSA

reagents and conditions: i) DMF, rt, 4 h; ii) AcOH, H2O, rt, 6 h; iii) human serum albumin, pH 7 borate-phosphate buffer, NaCNBH3, rt, 4 days.

Figure 3.73 Synthesis of neoglycoprotein displaying O-specific polysaccharide antigen as a synthetic vaccine against Shigella dysenteriae.

3.5.3

Carbohydrate Vaccines against Leishmaniasis

Leishmaniasis is a tropical disease caused by infection from a parasite species, Leishmania. On its cell surface, this parasite expresses multiple copies of unique lipophosphoglycans (LPGs) which are composed of a glycoylphosphatidylinositol (GPI) anchor, a repeating unit of phosphorylated disaccharide, and a cap oligosaccharide (Figure 3.75) [359]. Because of its abundance and the strong antigenicity of the phosphoglycan saccharide, LPG has attracted conceptual interest as a vaccine candidate. Synthetic vaccines based on natural LPGs are likely to be heterogeneous, due to the diversity of LPGs. Hewitt and Seeberger at MIT [360] developed a synthetic approach that enabled them to prepare a Leishmania carbohydrate vaccine. In this solution-phase method, the aminepresenting derivative of the cap tetrasaccharide is converted to a multivalent


Me O

H N

O MeO O H 3C MeO OH

Me O

N H

H3C

O

O

H N O Me

157

Ph O N H

Bacterial Cell

Alkyl

HO OHO HO H3C

Serovar 8: R = Me HO2C

Serovar 21: R = H

O H3C O RO

MeO2C O H3 C O BzO

OH

OH

O

NH-Cbz

O H3C

O O

H3C

O OBnO

i,ii,iii

BnO BzO O H3C O OAc

HO2C O H 3C O HO

OBz

NH2

O

HO OHO HO O H 3C O OH OH

trisaccharide unit of Serovar

O H3C

N H

O

HO2C O H 3C O HO

HO H 3C

BSA n (n = 12)

HO OHO

iv,v

H N

O O

OH

96

OH

reagents and conditions: i) cat. NaOMe, MeOH, rt, 24 h; ii) NaOH, H2O, rt, 48 h; iii) H2, 10% Pd/C, H2O, rt, 24 h; iv) bovine serum albumin (0.3 eq), glutardialdehyde solution (28%), pH 7.2 phosphate buffer, 4oC, 1 h; then NaBH4, 1 h; v) dialysis, then lyophilization.

Figure 3.74 Synthesis of neoglycoprotein displaying Serovar antigen as a synthetic vaccine against Mycobacterium avium.

form (97) by attaching multiple copies of the saccharide to keyhole limpet hemocyanin (KLH) as a protein sca¤old. This neoglycoprotein presents over 50 copies of the cap saccharide per hemocyanin molecule. 3.5.4

Capsular Polysialic Acid Presented on Liposomes

A polysaccharide, K1, found on the outer membrane of E. coli consists of 100 to 200 repeating units of a-2,8-linked sialic acid, with the terminal sialic acid connected to a lipid residue [361]. This type of multivalent carbohydrate cluster belongs to a class of acidic capsular polysaccharides [362]. To study the multivalent ligand–receptor recognition involving this type of polysialic acid, Frosch et al. [363] and Ringsdorf et al. [364] designed giant liposomes. The liposomes were composed of a lipid derivative of a2,8-linked polysialic acid incorporated into model membranes derived from phosphatidyl choline (Figure 3.76). The membrane display of the polysialic acid was verified by observing the binding

158


HO OH O HO HO HO OO HO OH HO HO HO OO O O HO HO HO O Cap tetrasaccharide

OHOPO32- OH OH O O O HO HO O

Phosphoglycan

HO OH O HO HO HO OO HO OH HO HO HO OO O O HO HO HO O

Glycosylphosphatidylinositol (GPI) anchor HO O O P O OH HOO HO OO O HO -O P O HO O OHO OH O O HOO HO HO O O HO O OH OH OH

n

i

NH2

Leishmania OH OH O

OH

HO

HO HO H2N O O HO

O-lipid

O OH


O S

N H

O

O Br O

Br

HN H N

N H NH

Br O

ii,iii

O Br


97 KLH O

O N H

S

N H n = 55

reagents and conditions : i) S-acetylmercaptoacetate pentafluorophenyl ester, DMF; ii) 0.2 M NH2OH, bromoacetyl derivative of keyhole limpet hemocyanin (KLH); iii) 2-aminoethanethiol.

Figure 3.75 Synthesis of multivalent Cap tetrasaccharide presented on keyhole limpet hemocyanin as a carbohydrate vaccine against leishmaniasis.

of a fluorescently labeled anti-K1 antibody to the membranes. The antibody– liposome interaction leads characteristically to nonhomogeneous distribution, or patching, of fluorescence images, which is caused by lateral di¤usion of the antigen–antibody complex in the membrane and by clustering of multiple K1 antigens by the bivalent antibody. 3.5.5

Multivalent Carbohydrate Antigens from Streptococcus

The cell walls of members of the b-homolytic Streptococcus group A contain a well-defined polysaccharide composed of multiple repeats of rhamnopyranoside with N-acetylglucosamine (GlcNAc) attached as a branched sugar. Targeting such carbohydrate antigens is of importance in the application of immuno-


AcHN

H OH

-O2C

HO HO

O OH

159

OH

O AcHN

CO2O OH OH

O OP nO O

O O

n = 100 - 200

O

(CH2)14CH3

liposome

(CH2)14CH3

phosphatidylcholine DMPC

O

-O2C

HO HO

O

AcHN

OH

OH

H OH O AcHN

O OH OH

HO -O2C H HO OH O O n CO2n AcHN AcHN O O OH OH HO O -O P O HO O -O P O O

CO2-

Liposome

Capsular polysaccharide mimetic of outer bacterial membrane

Figure 3.76 Multivalent capsular polysaccharide K1 displayed in a model membrane.

HO HO OH HO O O OH HO NHAc O O O HO HO O HO O O AcHN O HO HO HO OH

ii,iii

O

i

HO HO OH HO O O OH HO NHAc O O O HO HO O HO O O AcHN O HO HO HO OH HO HO OH HO O O OH HO NHAc O O O HO HO O HO O O AcHN HO HO

HO HO OH HO O O OH HO NHAc O O O HO HO O HO O O AcHN O HO HO HO OH

O

S

O

O

N H

N H 16

O

O

N H

N H 9

O

S

NH2

Bovine serum albumin

98

O

S

Ovalbumin

99

reagents and conditions: i) cysteamine, UV; ii) diethyl aquarate; iii) bovine serum albumin.

Figure 3.77 Structure of neoglycoconjugate presenting either penta- or hexasaccharide corresponding to Streptococcus group A polysaccharide.

160

MULTIVALENT MOLECULES APPLIED TO BACTERIAL TARGETS Ion channel

lipid bilayer

divalent AmB

AmB OH Me HO

O Me

OH OH

OH

OH

Me

OH OH O

HO

i,ii

CO2H

Me O

O Amphotericin B (AmB)

O Me

OH OH

Me

O 100

NH2

OH Me iii

HO

O Me

OH OH

CO2H

Me

OH OH O

CO2H

O Me

OH

OH

Me HO

Me

NH

OH

OH

O

HO H2 N

OH OH

Me

OH OH O

CO2H

Me

Me O Divalent AmB

Me

OH OH O

Me

OH

HO

OH

Me

O

O

Me OH

HO HN

O

O N H

R

O

OH

HO N H

Me

NH

101: R = (CH2)6 102: R = (CH2)2C(=O)O(CH2)2OC(=O)(CH2)2 reagents and conditions: i) FmocNH(CH2)5CHO, NaCNBH3; ii) piperidine; iii) bis-succinimidyl ester of 1,8-octanedioic acid (or ethylene glycol diacid), Et3N.

Figure 3.78 Proposed mode of action by amphotericin B (AmB) and synthesis of AmB dimer.

stimulation agents and in immunodiagnosis [365]. Auzanneau and Pinto [366] (see also Hu et al. [367]) reported the synthesis of neoglycoproteins 98 and 99. These molecules present multiple copies of hexa- and pentasaccharides whose sequences are identical with Streptococcus group A polysaccharides (Figure 3.77). 3.6 3.6.1

FUNGAL CELLS Covalent Dimer of Amphotericin B

Fungal infections are treated by amphotericin B (AmB), which is a member of the class of polyene antibiotics. The activity of AmB is attributed to its ability

FUNGAL CELLS

161

to undergo self-assembly into AmB octamer-ergosterol, an oligomeric complex, thus forming ion-permeable channels across lipid bilayers (Figure 3.78) [368]. Tethering two neighboring monomeric AmBs through a spacer provides a possible route for preorganizing dimeric constructs and stabilizing the functional complex. Matsumori et al. [369] tested this design strategy by constructing AmB dimers. The model for this complex suggests that the optimal spacer between AmB units would extend outside the fungal membrane. Therefore, a chemical handle was installed on the glycoside domain of AmB to allow dimerization with a linker. Reductive amination of an N-Fmoc-protected aminoaldehyde with AmB a¤ords such a derivative (100), after deprotection of the Fmoc group. To generate dimeric AmB, the primary amine of 100 is coupled with an activated bisester with a variable linker length (101 and 102). In a hemolytic assay using human red blood cells, both of the dimers are highly active, with an e¤ective dose (ED50 ) value of 0.3 mM, which is fivefold higher than that of AmB. Despite such in vitro activity, the two dimers are less active in inhibiting the growth of Candida albicans, where their activities are about 26 (101) to 13 (102) times lower than that by free AmB, which has an ED50 value of 0.3 mM.

4 MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS

4.1 CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES 4.1.1 Hepatic Mannose/N-Acetylgalactosamine-4-SO4 Receptors Mannose (Man)/N-acetylgalactosamine-4-SO4 (GalNAc-4-SO4 ) receptors are expressed on the surface of hepatic endothelial cells and macrophages [370]. These receptors recognize physiological glycoprotein molecules, such as the hormones lutropin and thyrotropin, that bear carbohydrate moieties terminating in Man and GalNAc-4-SO4 . Depending on the site of expression, this receptor exists as either a dimer, as in hepatic cells, or as a monomer, as in macrophages. The hepatic Man/GalNAc-4-SO4 receptor has the important role of regulating physiological levels of lutropin. It does so by selective binding of the GalNAc-4-SO4 moieties presented by lutropin, thus mediating the clearance of lutropin from the circulation (Figure 4.1). 4.1.1.1 Glycoprotein-Based Multivalent Ligands. Lutropin hormone interacts multivalently with a dimeric Man/GalNAc-4-SO4 receptor. The hormone– receptor binding is tight. Depending on the number of GalNAc-4-SO4 moieties involved in the association, the K d value varies from 8.3 nM, with three to four GalNAc-4-SO4 moieties, to 267 nM, with two GalNAc-4-SO4 moieties [371].


162

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES OH OH O O O HO HO AcHN AcHN HO HO HO

163

-O3SO

O O

Lutropin (LH) n n~4

GalNAc-4-SO4 Man receptor

GalNAc-4-SO4 receptor LH

membrane

Man-specific Man/GalNAc-4-SO4 receptor (Monomeric)

GalNAc-4-SO4-specific Man/GalNAc-4-SO4 receptor (Dimeric)

Macrophage

Hepatic endothelial cell

Figure 4.1 Man-specific (macrophage) and GalNAc-4-SO4 -specific (hepatic endothelial) form of Man/GalNAc-4-SO4 receptor. The dimeric receptor interacts selectively with lutropin, which presents multiple copies of terminal GalNAc-4-SO4 moieties.

Lutropin binds much more tightly to the receptor than does a monovalent GalNAc-4-SO4 ligand (inhibition constant Ki ¼ 16.8 mM ). Roseman and Baenziger [372] further elucidated the role of multivalent binding of Man/ GaINAc-4-SO4 receptors in hepatocytes. Although Man/GalNAc-4-SO4 receptors can recognize the GalNAc-4-SO4 ligand selectively in a multivalent environment, the receptors do not display high ligand-binding specificity during monovalent interaction. For example, this receptor shows similar binding activity toward two trisaccharides, with Ki values of 25.8 mM (1) and 16.2 mM (2) (Figure 4.2). These molecules di¤er only with respect to location of the sulfate in position 3 or 4. However, multivalent presentation of the ligands on a sca¤old of serum albumin substantially increases the binding a‰nity displayed by the resulting molecules, with Ki values of 13 nM for 3 and 170 nM for 4. This study shows that ligand specificity is achieved through multivalent association. Moreover, the rate of clearance from the blood circulation di¤ers between the two molecules such that 3 is cleared about 12-fold faster than 4. This 12-fold di¤erence in clearance rate is comparable to the about a 13-fold di¤erence in a‰nities.

164 -O3SO

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH O

HO AcHN

HO

OH O O HO AcHN HO HO HO

-

OH O

O3SO AcHN

O O O

OH O

HO AcHN

OH O

O HO AcHN HO HO HO

O O

OCO2Me

2 (GalNAc-3-SO4)

HO -O

O

3SO

OH O

AcHN

O O

O

OCO2Me

1 (GalNAc-4-SO4)

-O3SO


BSA


O O O

O ~7

3

4

BSA O ~7

Figure 4.2 Structure of GalNAc-SO4 -presenting ligands.

4.1.2

Hepatic Asialoglycoprotein Receptors

Mammalian liver cells express on their surfaces multiple copies of the asialoglycoprotein receptor (ASGP-R), which recognizes b-d-galactoside (Gal) or 2-acetamido-2-deoxy-b-d-galactoside (GalNAc) displayed on soluble glycoproteins [373]. This hepatic receptor plays a role in the uptake of such glycoproteins and in regulating their cellular level. For instance, it is involved in the rapid clearance of desialylated glycoproteins from the blood circulation, provided that prior to desialylation, such asialoglycoproteins present on their surface multiple galactoside residues attached to sialic acid. 4.1.2.1 Triantennary Galactosides. A series of trivalent b-d-galactosides were designed in the form of antennary clusters [374]. Although structurally homologous, these ligands di¤er in the distance between neighboring galactosides (Figure 4.3). Such trivalent galactosides bind asialoglycoprotein receptors located in parenchymal liver cells, in competition with [ 125 I]asialoorosomucoid ligand. Relative to the weakly binding monomeric GalNAc, which has a Ki value of 870 mM, the binding a‰nities of the trivalent ligands are enhanced up to 4000-fold in manner, which depends on the spacer length, with Ki values of >390 mM (5), 19 mM (6), 11 mM (7), and 0.2 mM (8). Thus, if presented at optimum length, vicinal galactosides within a cluster are bound more tightly by asialoglycoprotein receptors than are those with shorter spacing. 4.1.2.2 Lysine-Based Multivalent N-Acetylgalactosides. Valentijin et al. [375] developed a solid-phase methodology for preparing lysine-based di- and tri-

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES OH OH O O OH

HO

OH OH O O HO OH

OH OH O O OH

HO

OH OH O O HO OH

O C N 3 H

O O

O

C N 3 H

O O

O

O

C N 3 H

O O

O

O

O

HO

O

C N 3 H

H N

OMe O

O

O

O

O

O

O

O

H N

OMe

H N

OMe

H N

OMe

165

5 (4 Å)

6 (9 Å)

7 (13 Å)

8 (20 Å)

OH OH O

HO

9Å

O O

OH OH O O HO OH

O O

O

O C

O

N H

H N

OMe O

O

O 9Å O OHO OH HO HO

6 (9 Å)

Figure 4.3 Structure of glycoclusters composed of triantennary galactoside.

antennary galactosides (Figure 4.4). These multivalent ligands are composed of Gal or GalNAc ligands displayed on a sca¤old derived from a lysine peptide. They bind asialoglycoproteins on the surface of parenchymal liver cells more tightly than monovalent ligands, with Ki values of 95,000 nM (9), 30 nM (10), 650 nM (11), and 4 nM (12). The results indicate that GalNAc-presenting molecules are more active than the corresponding Gal-presenting molecules. Trimers show higher a‰nity than do dimers constructed on comparable scaffolds. The highest a‰nity exhibited by 12 indicates >200,000-fold enhancement over monovalent GalNAc. 4.1.2.3 Trivalent Galactoside–Cholesterol Conjugates. Multivalent antennary galactosides that bind selectively to liver cells play unique roles of biological

166

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH OH O O R

HO OH OH O O R

OH OH O O R

H N

HO

O

OH OH O O HO R

O

O

NH O 9: R = OH 10: R = NHAc

BzO BzO

NH

O

OH OH O O HO R

OH

O

H N

HO

N NH H

OH O

O 11: R = OH 12: R = NHAc

OBz O O

BzO BzO

OBz O O R

R O

i,ii

NH

O

i,iii,iv

NH

Alloc-HN

O

O O

O

Fmoc-HN O R = OBz; R = NHAc

N H

Fmoc-HN

solid support

BzO BzO

O O

OBz O O R

BzO BzO

OBz O O R BzO BzO

O

H N

11, 12 O

O

O

OBz O O R

v

NH

N NH H

O O

O

reagents and conditions: i) piperidine, DMA; ii) Fmoc-Lys(Alloc)-OH, BOP, DIEA; iii) Pd(PPh)3, AcOH, NMM, CHCl3; iv) carboxylic acid derivative of β-D-galactoside, BOP, DIEA; v) 1 M NaOH.

Figure 4.4 Solid-phase synthesis of lysine-based cluster galactoside.

significance. Biessen et al. [376] reported the functional activity of a triantennary galactoside conjugate (13) which presents a cholesterol derivative on one end of the conjugate (Figure 4.5). This bifunctional hybrid is designed on the assumption that its cholesterol moiety would associate with cholesterol-recognizing receptors such as lipoproteins, while its multivalent sugar ligand guides

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES

167

LDL

13 LDL

ASGP-R

Gal

membrane (liver cell)

intracellular uptake

OH OH O O OH

O

HO

O

O

O

O

O

H N

C N 3 H

OH O

O

BOP DMAP

O H2 N

OH OH O O HO OH

O

O O

O

ligand for ASGP-R

O

O

O

C N 3 H 13

H N

H N O

O O

O LDL-binding domain

Figure 4.5 Simplified model for targeted delivery of low-density lipoprotein into hepatic cells by a cholesterol-presenting multivalent galactoside.

the hybrid to hepatic cells through selective interaction with hepatic asialoglycoprotein receptors [377]. Like nonhybrid galactoside clusters, the conjugate 13 shows high a‰nity to asiologlycoprotein receptors, with a Ki value of 200 nM. Due to its amphiphilic nature, this ligand forms micelles in aqueous solution with an average radius of about 5 nm, at a surface density of about 20 molecules per micelle. It is also incorporated into lipoproteins, such as the lowdensity lipoprotein (LDL) present in human serum. Indeed, 13 is associated with LDL at a ratio of about 210 molecules per LDL. Upon intravenous injection into rats, this hybrid ligand causes a decrease in the level of cholesterol present in serum. Such cholesterol-lowering activity is attributed to the targeted

168

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS

delivery of LDL to hepatic cells, which results from the multivalent, tight interaction of the LDL–glycolipid complex with asialoglycoprotein on liver cells (Figure 4.5). 4.1.2.4 Trivalent Galactosides for Cell-Specific DNA Delivery. The specific a‰nity of multivalent galactoside molecules to hepatocytes is utilized further for the selective delivery of DNA to mammalian cells. In a study reported by Ren et al. [378], mono-, di-, and trivalent galactosides are covalently conjugated with a molecular construct comprising seven primary amines that serve to form a complex with DNA (Figure 4.6). In a gene transfection assay mea-

NH2 O TBDMSO

O

9

AcO AcO AcO AcO

O

i N

TBDMSO

N

9

OAc O O

O

N H

ii

N

NH2

O O

OAc OAc O O

NH

O O

OAc

iii-vi

O N

N H

N H

OTBDMS

O

9

NH2

NH2 OH OH O O HO OH OH OH O O HO OH

O O O O

N

N H

N

NH2

N

N

N

NH2

O

O N H

H N

O

NH

N

O

R=

H 2N

9

ASGP-R recognition

14

NH2

N

H 2N DNA association domain

OH OH O O HO OH OH OH O O HO OH

OH OH O O HO OH O O

O HN

NH

N

O O

O

N

O

H N

O O

N H

9

O

NH-R

15

reagents and conditions: i) N(CH2CH2NH2)3, DMAP, MeCN, 88oC; ii) acetylated galactosyl derivative, Et3N, THF, 69oC; iii) Bu4NF, THF, 82oC; iv) CDI, DMAP, CH2Cl2, 87oC; v) DAB-dendrimer (NH2)8, THF, reflux; vi) NaOMe, rt.

Figure 4.6 Synthesis of multivalent galactoside–polyamine conjugate.


169

suring DNA transfer to hepatocarcinoma cells, these galactoside–amine conjugates increase the level of DNA transferred into cells, as assessed by subsequent expression of the transfected genes. For example, the levels of gene expression observed in the presence of 14 and 15 are 300- to 500-fold higher than that displayed by a monovalent galactoside–polyamine conjugate. Such a large enhancement is obtained specifically in hepatocytes and is inhibited competitively by the addition of galactoside-presenting asialofetuin. The enhanced gene transfer is explained by asialoglycoprotein-mediated endocytosis to hepatic cells [379]. Another use of cell-specific targeting is based on peptide nucleic acid (PNA) conjugates that present multivalent lactoside residues [380]. This macromolecule e‰ciently enters liver cells through asialoglycoprotein–lactoside recognition. It also blocks telomerase activity, provided that the PNA provides an antisense telomerase sequence. A similar example is a cholate-based, trivalent galactoside linked covalently to an antisense oligonucleotide [381]. Similarly, the goal of designing this molecule is to enable the carrier-mediated delivery of a therapeutic nucleic acid. 4.1.2.5 Multivalent Binding and Intracellular Tra‰cking. Polyvalent galactosides presented on the side chains of an HMPA [N-(2-hydroxypropyl)methacrylamide] copolymer can attach to the surface of human HepG2 cells (Figure 4.7) [382]. The polymers containing lactoside (16, with galactoside densities of 20 and 30%) and trivalent galactoside (17, with galactoside densities of 20 and 30%) bind more strongly than does galactoside-free HMPA, a polymer control. Furthermore, its binding is selectively inhibited by a large excess of free lactoside. There is a correlation between binding capability and the density of galactoside on HMPA such that polymers with higher galactoside densities show higher adsorption rates, as observed from 16 and 17. After receptor– ligand-mediated multivalent binding to hepatic cells, 16 and 17 are internalized and transported into lysosomal compartments of the cell. The extent of the resulting receptor-mediated endocytosis is determined specifically by galactosidepresenting polymers and is dependent on the nature of the polyvalent ligand (17 is greater than 16). The extent of receptor-mediated endocytosis is also related to the carbohydrate density: Molecules with 30%, 20%, and 10% carbohydrate density experience successively lower rates of endocytosis. In a separate study, a radiolabeled multiantennary GalNAcLe x enabled the discovery that the multivalent ligand predominantly targets hepatic cells [383]. Such ASGP-mediated specific cell recognition is also achieved by using polymers presenting a nongalactoside ligand, such as a modified glucose at the C6 position [384]. 4.1.3

Mannose-Binding Proteins

Mannose-binding proteins (MBPs) are calcium-dependent lectins that bind to mannose and N-acetylglucosamine as specific ligands [385,386]. These man-

170

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS 17 (χgal = 10, 20, or 30%)

16 (χgal = 10, 20, or 30%)

O

NH O OH

NH O

OH

HN

Fluorescein-HN

NH O

NH

OH O HO

HO O OH

O

NH O OH

OH

S

HO

NH O

O

O

NH O OH

HN

Fluorescein-HN

OH

NH O

S

HO HO

OH HO HO HO

NH O

NH

OH

OH O O OH O O OH

O NH O O

OH

HO HO OH

fluorescence-labeled HMPA copolymer

ASGP-R

Gal

membrane (liver cell)

internalization

Figure 4.7 Structure of HMPA copolymer–based polyvalent galactoside and its role in intracellular uptake.

nose-binding lectins are expressed as two closely related forms and are classified according to the site of expression, such as serum (MBP-A) and liver (MBP-C). An important role played by MBPs includes triggering the immune response by binding to mannose ligands present at or above a threshold density on the surface of pathogenic bacteria and fungi. This binding subsequently activates the complement pathway, which is targeted to kill foreign organisms. Although their function relies on their ability to recognize mannose with high specificity, MBPs show only weak a‰nity to monovalent sugar ligands, with Kd values in the millimolar range [387]. However, with multiple carbohydrate-recognition domains (CRDs) displayed per protein subunit on the membrane surface,


171

MBPs bind very tightly to target cells, such as bacterial and fungal cells, that present dense and repetitive arrays of Man-type ligands. 4.1.3.1 Bivalent Mannosides. Quesenberry et al. at Johns Hopkins [388] proposed that MBPs display two distinct modes of binding, as shown in Figure 4.8. A subunit in MBP-A, from serum, contains a CRD with dual a‰nity to

GlcNAc or Man site Multivalent antennary glycosides Man site Man GlcNAc

MBP-A (rat serum)

MBP-C (rat liver)

Multivalent glycoprotein

MBP-A

MBP-C

OH

18 (GlcNAc)2

O

H N

AcHN

N H

O O

N H

H N

AcHN

H N

AcHN

O

O

OH OH OH

O

OH OH OH

O

O

O

O

O

OH

19 (Man)2

O

H N

AcHN

N H

O O

N H

H N

OH

O H N

O

O OH

O

OH OH OH OH OH OH

OH

20 (GlcNAc)(Man)

O

H N

AcHN

N H

O O

N H

H N

AcHN O

O

O

O

OH OH OH

O H N O

OH

OH OH OH

Figure 4.8 Model for interaction of mannose-binding protein with its multivalent ligands, which include divalent glycosides presenting either GlcNAc or mannoside.

172


Man or GlcNAc, while MBP-C, from liver, presents two CRDs, with one site selective to Man and another to Man or GlcNAc. This di¤erence in ligand specificity and the valency of CRDs distinguishes the monovalent nature of the MBP-A subunit from the bivalent nature of the MBP-C subunit, and this is reflected in the binding a‰nities ofmannoside- and GlcNAc-based divalent glycosides (Figure 4.8). For example, MBP-A is inhibited by the galactoside ligands 18, 19, and 20, with IC50 values of 0.5, 0.56, and 0.44 mM, respectively, which suggests no substantial benefit from divalent glycosides relative to monovalent sugars such as Man (1.5 mM) and G1cNAc (0.9 mM). However. the binding profile to MBP-C is notably improved, with IC50 values of 0.35 mM (18), 0.05 mM (19), and 0.14 mM (20). Thus, bivalent glycosides 19 and 20 exhibit a 22- and 8-fold increase in their a‰nity to MBP-C, respectively, over monomer. 4.1.3.2 Mannose-Presenting Neoglycoproteins. Multivalent mannosides presented on the surface of a glycoprotein such as bovine serum albumin have increased a‰nity to MBPs. For example, a tetravalent mannoside, BSA(Man)4 , shows IC50 values of 0.2 mM to MBP-A and 0.7 mM to MBP-C. These values represent about 10 4 - and 103 -fold enhancements over free mannoside [389]. When the valency of the glycoprotein increases to 20, as with BSA(Man)20 , its a‰nities to MBPs are much higher, with IC50 values of 2 nM (MBP-A) and 5 nM (MBP-C). This example suggests a mechanism that enables the human body to distinguish membrane surfaces derived from the self from those of pathogens. ‘‘Self ’’ membrane surfaces probably contain less dense mannose clusters, and MBP clusters are likely to bind only weakly to such surfaces. In contrast, MBP clusters will probably bind tightly to the dense mannose arrays present on pathogenic surfaces, thus facilitating the discovery and demise of pathogens by the immune system. 4.1.3.3 GlcNAc-Presenting Polymer. Multivalent presentation of mannoside not only increases avidity to MBP, but also modulates selectivity to MBPs derived from various tissues. Fan et al. [390] reported the synthesis of the neoglycopolymer 21, in which poly(acrylamide) is used to present multiple copies of synthetic Man9 GlcNAc as side chains. The multivalent molecule is prepared by copolymerizing acrylamide and a Man9 GlcNAc-terminated-acrylamide derivative in aqueous solution (Figure 4.9). This synthetic glycopolymer inhibits the binding of MBP-CRD coated on a solid surface to BSA(Man)30 ligand, with IC50 values of 3.5 mM for MBP-A and 74.5 nM for MBP-C. The a‰nity of 21 to MBP-C is enhanced about 180 times relative to that of a nonpolymeric ligand control, Man9 GlcNAc, and is also higher than that of natural soybean agglutinin, which has an IC50 value of 13.2 mM. The a‰nity of 21 to MBP-A is relatively low compared to MBP-C, yet it is still of great interest because there are few high-a‰nity ligands to MBP-A, and neither Man9 GlcNAc nor soybean agglutinin shows significant activity.

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES HO OH O HO HO HO OO HO HO HOO

173

O H 2N

O

OH OO O HO Man9GlcNAc2 HO O HOOH OH HOOH O HOO OH OO O HO O HO OO O O O HO O HO HO OH OH O O H AcHN HO NHAc HOOH OHOH HO

HO OH O HO HO HO OO HO HO

ammonium persulfate TEMED, H2O O N H

21 [Man9GlcNAc2]/[CONH2] = 1/44

O HOO OH OO n O HO HO O HOOH OH O O O O O O HOOH NH NH NH NH NH 2 2 NH 2 2 2 O HOO OH OO O HO HO O OO O O O HO O HO HO OH OH O O H AcHN HO NHAc HOOH OHOH HO

Figure 4.9 Structure of neoglycopolymer presenting Man9 GlcNAc2 as a side chain.

4.1.3.4 Mannoside Mimics. Grandjean et al. [391] and Angyalosi et al. [392] designed multivalent mannoside analogs using d-quinic and shikimic acid as ligands. These carbocyclic acids are likely to serve as mannoside mimics because their structures possess a vicinal diol at the C4 and C5 positions (Figure 4.10), as displayed by a mannoside ligand bound at the calcium site of MBP. The acids are incorporated into fluorescence-labeled multivalent molecules constructed on a lysine–peptide sca¤old with a valency of 2 to 8. Glycomimetics such as octavalent 22 and 23 undergo internalization as a result of MBP-mediated uptake in human dendritic cells, although the level of internalization is about twofold (22) and fourfold (23) lower than that of an equivalent octavalent mannoside. Such internalization is achieved only if mannoside or its mimic is presented as the ligand; a similar multivalent construct made of galactonoyl ligand is not internalized. Tetravalent mimics (24 and 25) show a twofold higher level of mannose-receptor-specific internalization than that of a tetravalent mannoside. Thus, multivalent constructs derived from quinic and shikimic acids are recognized by human mannose receptors. In a separate study, Schuster et al. [393] reported the solid-phase synthesis of mannose mimics derived from shikimic acid as ligands to a C-type lectin and a mannose-binding protein.

174

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH 5 OH O OR HO HO 4

5 O HO HO OH OH 4 OH

D-Mannose

5 HO HO 4 OH

(D)-Quinic acid

CO2H

Shikimic acid

R-HN

O

H N

R-HN O

S

OH N H

CO2H O

O CO2H O 22: R =

HO HO

OH

OH

NH

NH

HN

O S

HN

O

O

O

CO2H NH2

O NH

O

NH O

N H

O

HN

O

NH-R

HO HO

N H

H N

O 23: R =

S

OH

R-HN R-HN

NH-R

O

O

NH

S

N H

O HN

HN

HO2C

O H N

HO2C

S

O

O

R-HN O

S

N H

24: R =

CO2H OH

OH O

25: R =

HO HO

OH

CO2H O

O HO HO

NH-R

O

H N

R-HN

NH-R

N H

NH

NH O HN O

OH

NH-R

HN

H N

S O

O

N H

O S N H

H N O

CO2H

O

NH2 O

NH-R

Figure 4.10 Structure of multivalent mannoside mimics.

4.1.4

Mannose 6-Phosphate Receptor (Insulinlike Growth Factor II)

Mannose-6-phosphate/insulinlike growth factor II receptor (Man-6-P/IGF-R) contains two Man-6-P-binding domains located in the extracytoplasmic region of the receptor [394]. The receptor recognizes Man-6-P as a specific ligand, although its a‰nity to various Man-6-P ligands varies with ligand valency. For instance, it binds to monovalent Man-6-P with a relatively low a‰nity, re-


175

flected by a Kd value of 6 to 8 mM. However, the receptor binds with high a‰nity to glycoproteins presenting divalent Man-6-P groups, as reflected by a Kd value of 1 to 2 nM [395]. 4.1.4.1 Neoglycoprotein-Based Mannose-6-Phosphate. Byrd and MacDonald [396] and Byrd et al. [397] studied the mechanisms responsible for the higha‰nity association of Man-6-P receptors with multivalent ligands, with particular emphasis on intrareceptor bridging and interreceptor dimerization (Figure 4.11). This receptor displays two Man-6-P sites and binds to a pentavalent ligand present on bovine serum albumin, (Man-6-P)5 , with a K d value of 0.54 nM and a stoichiometry of 1:1 (Figure 4.11a). In comparison, nonnatural Man-6-P receptors that contain only one Man-6-P recognition site, such as truncated receptor or mutated receptors that lack a second Man-6-P site, are also able to bind the pentavalent ligand (Man-6-P)5 in the nanomolar a‰nity range but with a 2:1 stoichiometry, resulting in the receptor dimerization (Figure 4.11b). Therefore, it is likely that the interaction between Man-6-P/IGF-R and multivalent ligands involves multivalency, leading to a high-a‰nity association as well as receptor oligomerization on cell surfaces.

(a) Man-6-P site Man-6-P/IGF-II

Man-6-P BSA(Man-6-P)5

membrane Dimerization of receptor (1:1 stoichiometry) (b) mutated Man-6-P/IGF-II

truncated Man-6-P/IGF-II

Dimerization of receptor (2:1 stoichiometry)

Figure 4.11 Hypothetical model for interaction of Man-6-P/IGF-R with multivalent ligand presenting Man-6-P: (a) dimerization of natural receptors via divalent bridging; (b) dimerization of mutated or truncated receptors that presents one Man-6-P binding domain per receptor.

176


4.1.4.2 Biotin/Avidin-Based Tetravalent Mannose-6-Phospate. Biotin and avidin display a noncovalent but very tight interaction, with a K d value of 1015 M. Such a strong association provides a system that makes it possible to design a novel type of multivalent ligand. Rothenberg et al. [398] reported the preparation of a pentamannose phosphate that is linked to biotin through a 2,6-diaminopyridine spacer (Figure 4.12). This mannose phosphate–biotin conjugate (26) is converted to a noncovalent form of multivalent construct when its terminal biotin unit associates to tetravalent avidin. The resulting tetravalent complex binds to a mannose-6-phosphate receptor immobilized on the surface, whereas a monovalent mannose phosphate shows only poor a‰nity.

O

O HN HO

NH

H2 N

N

NH2

HN

H2 N

S

N

i

O

H N

NH +

S

OPO32OH O O HO

HO HO

O

PO32OH O OH 4

pentamannose phosphate

O 2-

iii

HO HO

OPO3 OH O O HO

2-

OPO3 OH O O HO 3

HN

2-

PO3 OH H OH N

N

NH iii

H N

S O

26 fluorescent mannose-phosphate biotin conjugate

HO HO

HO HO

OPO32OH O O HO

OPO32OH O O HO OPO32OH O O HO 3

OPO32OH O O HO 3

PO32OH H OH N Avidin

2-

PO3 OH H OH N

biotin (Kd = 10-15 M to avidin)

N HO H HO 2O3P

N HO H HO 2O3P

OH O O HO 2O3PO

OH O 2-

O HO O3PO

OH O3 O HO 2O3PO

OH O3 O HO 2O3PO

OH OH

OH OH

reagents and conditions: i) EDC, NHS; ii)Schiff base formation, and reductive amination; iii) avidin

Figure 4.12 Tetravalent presentation of mannose phosphate through biotin–avidin interaction.

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES CD45

177

CD45 Galectin

Galactoside

CD3

CD3

Figure 4.13 Cross-linked complexes formed among galactin-1, CD45, and CD3.

4.1.5

Galectins

Galectins, a class of mammalian lectins, consist structurally of homodimers that recognize N-linked complex carbohydrates terminating in Gal residues such as N-acetyllactosamine [399]. The CRDs of galectins are presented bivalently, where each site is located in the opposite face. Galectins are involved in cell–cell adhesion, cell migration, and growth regulation through a mechanism based on multivalent interaction [400–402]. For example, galectin-1 isolated from calf spleen can induce apoptosis in susceptible T-cells [403]. Its proposed mechanism of action includes a cross-linking of specific T-cell receptors on cell surfaces, followed by the segregation of T-cell receptors (Figure 4.13) [404]. Evidence for this mechanism is further based on galectin-induced spatial segregation of the CD45–CD3 complex on human T-cell surfaces caused by the formation of an extended lectin–carbohydrate lattice [405]. 4.1.5.1 Glycodendrimer. Andre´ et al. [406] reported galectin inhibitors that are designed to target CRDs of galectins. The opposite orientation of the two galectin CRDs makes it di‰cult to design multivalent ligands capable of binding simultaneously. Disubstituted benzoic acid can be used to attach two, four, and eight copies of lactoside moieties, thus creating multiple generations of wedgelike glycodendrimers (Figure 4.14). These glycodendrimeric molecules are potent inhibitors against homodimeric galectin-1. For example, a tetravalent, second-generation glycodendrimer (27) shows inhibitory activity, with an IC50 value of 27 mM. The activity of the second-generation glycodendrimer is enhanced 1667-fold per lactoside relative to free lactose. In addition, 28 blocks hemagglutination mediated by galectin-3 at micromolar concentrations in an assay that is commonly used as a cell adhesion model. The hemagglutination blocking activity is 144 times greater per lactoside than that of a monovalent lactoside. A separate study by Andre´ et al. [407] showed that the binding activity of a lactose-presenting, polyamidoamine starburst dendrimer to

178

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH OH OH

OH OH OH HO

HO

O O

OH OH

OH HO

OH HO

O

HO

O

HO

O

NH

NH S

HO

O HO OH

O HO OH HO

O

HO

O

OH

HO

NH

S NH

S

S

NH

NH

O O

OH OH

OH

O NH NH

O

O

O

O HN

HN

O

O

O

O 27 O

MeO

OH OH HO OH OH OH OH OH OH HO O OH OH OH O HO HO HO OH OH O O HO O HO OH OH O O O HO OH OH OH HO O OH HO O OH OH HO OH OH OH HO O O OH HO O HO OH O HO O OH HO O HO NH HO HO O O OH S HO NH O O OH NH O O NH HO S HO NH OH S NH S OH NH NH S NH NH O HO S NH NH HO O O NH O S NH O O S NH O O NH HO

O HN

O

HN

O

O

O HN

HN

O O O

O

O

O

NH

HN

O

O

O

28 MeO

O

Figure 4.14 Structure of wedgelike glycodendrimers as inhibitors of mammalian galectins.


179

galectin-1 is influenced by the method of dendrimer generation, and is therefore influenced by the molecule’s shape and ligand density. 4.1.6

Sialoadhesin CD22

CD22 belongs to a class of cell adhesin glycoproteins that are expressed uniquely on the surface of B-cells and B-cell-derived tumor cells [408,409]. It recognizes glycans terminated with sialic acid residues. As one of two isoforms of CD22 (a and b), CD22b binds to endogenous ligands presenting a NeuAc(a2,6)Lac trisaccharide unit (29) as a binding motif (Figure 4.15). Van Rossenberg et al. [410] and Sliedregt et al. [411] found that some disaccharides, and even monovalent sialic acids, show comparable or higher a‰nity than that of a natural trisaccharide in an assay that measures their ability to bind to CD22b. For example, the N-nitrobenzoyl derivative of NeuAc(a2,6)glucosamine (30)

OHOH 29

AcHN

CO2H OH O O OH O OH HO OH HO O

OHOH CO2H

OH O OMe OH

AcHN

O OH

OH

O HO HO

O

OMe 30

NH O

O 2N

OHOH CO2H AcHN

O OH

OH

O HO HO

O

O

NH O

O O2N

HN

OHOH AcHN O

O

OH

O OH HO HO

O

O

CO2H AcHN

NO2

O

O

O

NH

N H

O HN

H N

O

O

OH OH HO

O

O

O n

N H

O CO2H

OH NHAc OH OH

31 (n = 0, 1, 2)

O

NO2

O2N O HN O

O OHOH

CbzHN

CO2H AcHN

O

OH

O OH HO HO

O O

O

NH

O

N H

O

O2N

OHOH NO2

O O

O

32

CO2H

HN NH

O

O

OH OH OH HO O NHAc O

O

O O N H

O

HN O

O

OH OH OH HO O NHAc O CO2H

OHOH

Figure 4.15 Structure of Neu5Ac(a2,6)Lac-displaying multivalent ligands to CD22.

180


exhibits an IC50 value of 26 mM, an activity about ninefold greater than that of 29, which has an IC50 value of 244 mM. Tethering this glucosamine-derived disaccharide unit to a Tris- or glutamate-based sca¤olding creates a multivalent molecule. The resulting di-, tri-, and tetravalent ligands (31; n ¼ 0; 1; 2) inhibit CD22b with about 10-fold higher activity than that of 30. A second type of trivalent ligand constructed on a Tris sca¤old (32) binds potently to CD22b with about 30-fold higher a‰nity than 30, and about 270-fold higher activity than that of the natural binding unit, 29. This example demonstrates the utility of multivalency in designing ligands that target dimeric CD22 receptors. 4.1.7

Dendritic Cell–Specific ICAM-3–Grabbing Nonintegrin

The dendritic cell–specific ICAM-3-grabbing nonintegrin (DC-SIGN) is a Ctype (calcium-dependent) lectin present on the surface of dendritic cells. This molecule mediates intercellular communication between dendritic cells and Tcells by binding to the intercellular adhesion molecule-3 (ICAM-3) molecules present on T-cells. It also binds to the gp120 envelope glycoprotein on the surface of human immunodeficiency virus (HIV). The crystal structure of oligosaccharides bound to the DC-SIGN CRD show that this receptor recognizes mannoside-containing high densities of oligosaccharides displayed on protein surfaces of ICAM-3 and gp120 [412]. The DC-SIGN receptor is believed to form tetramers on cell surfaces [413]. 4.1.7.1 Mannoside Clusters on Peptide and Protein Sca¤olds. Feinberg et al. [412] studied the mechanism of carbohydrate recognition by DC-SIGN receptors using ligands based on mannosides (Figure 4.16). Free mannose shows weak inhibition activity, with a Ki value of 13.1 mM, while Man9 GlcNAc2 (33), a high-a‰nity mannoside cluster, shows enhanced activity, with a Ki value of 0.016 mM. The significant di¤erence in a‰nity between the two ligands is HO OH O HO HO HO O O HO HO

HOO HO HO OH

O O

O HOO

OH O O HOO

OH OHOH

O O O O

O HO

OH HO O O O HO AcHN

HO O OR O HO AcHN

HO O HO HO OH

BSA

O 30

OH OHOH

OH OHOH

Man9GlcNAc2 (33)

Man30BSA (34)

OH OHOH

Figure 4.16 Structure of a triantennary mannoside cluster and a mannoside-presenting neoglycoprotein.


181

attributed to the manner of presentation of multiple mannosides. When the binding is tested for Man30 BSA (34), DC-SIGN shows a much higher a‰nity, with a K d value of 169 nM. Multivalent ligands probably enhance the binding a‰nity because multiple CRDs from an oligomeric cluster of DC-SIGN receptors are involved in the binding. 4.1.7.2 Lewis Saccharide–Presenting Dendrimers. Frison et al. [414] reported oligolysine-based dendrimers displaying four or five copies of Lewis saccharides, such as Le a , Le b , and Le x . These glycodendrimers show a higher a‰nity to DC-SIGN receptors than do mannoside-presenting dendrimers of the same type. Upon binding, the Le glycodendrimers undergo multivalency-mediated endocytosis into cells expressing DC-SIGN receptors. A separate study by Van Die et al. [415] suggested that human DC-SIGN binds to Le x determinants displayed multivalently on a neoglycoconjugate by using a region within the CRD that di¤ers from the site for high-mannose ligands. 4.1.8

Spermatogenesis

Male germ cells interact closely with Sertoli cells in testes through direct cell– cell adhesion. This is a first step that leads to multiple stages that are essential to germ cell maturation. This maturation process, spermatogenesis, enables germ cells to undergo di¤erentiation into mature spermatozoa [416]. This type of intercellular communication is based on a specific interaction between a lectin receptor present on Sertoli cells and certain N-glycans displayed on germ cells (Figure 4.17) [417]. Akama et al. [418] reported that carbohydrate ligands involved in such interactions comprise triantennary carbohydrates terminated with multiple copies of GlcNAc residues. Their involvement is supported by an observation that the adhesion of mouse germ cells to Sertoli cells is blocked by an N-glycan oligosaccharide which presents three copies of terminal GlcNAc and a fucose residue attached to the core of the glycan chain. Thus, it is likely

GlcNAc Mannose Fucose lectin

Sertoli cell

Germ cell tri-antennary N-glycan

Figure 4.17 Proposed model describing intercellular communication between Sertoli cell and germ cell mediated by lectin–carbohydrate interaction.

182


that this type of rare carbohydrate cluster is involved in mouse spermatogenesis. 4.1.9

Sperm–Egg Interactions

4.1.9.1 Bindin. During fertilization, molecules on the surface of egg cells play a direct role in species-specific interactions with sperm [419]. In the sea urchin, the receptor on sperm is bindin, a cell surface component of acrosome granules (Figure 4.18). The complementary ligands on egg surfaces, known as O-linked glycans, are attached to cell surface glycoproteins [420]. Kitazume-Kawaguchi et al. [421] reported that the basic structure of the glycan ligand is a polysialic acid that terminates in a sulfated sialic acid, 9-(SO4 )Neu5Gc-a2,5-Neu5Gc. When sperm bind to the outer coat of the eggs, formation of the acrosomal process ensues. The sperm adheres to the plasma surface of the egg by means of the sperm bindin receptor binding to 9-(SO4 )-Neu5Gc-terminated polysialic acid molecules on the egg surface. This association is specific, as supported by identification of a free oligosaccharide composed of a sulfated sialic acid (such as 35), which can bind to sperm that have undergone the acrosomal reaction and can prevent the fertilization of eggs [422,423]. Beads that display on the surface multiple copies of a neoglycoprotein containing sulfated sialic acid can elicit sperm binding, which leads to the formation of stable aggregates of sperm and beads.

acrosomal process

actin polymerization; cell adhesion

Bindin receptor

Sperm acrosomal vesicle nucleus

carbohydrate (35)

jelly coat/vitelline layer Egg O OHOH -O S OH HO OH OHOH OO H H O CO2H O CO2- N O H CO2O CO2- N O N N O HO O OR O HO O HO HO HO O HO HO O HO HO 35

Figure 4.18 Sperm–egg interaction in a sea urchin and an oligosaccharide ligand on the egg surface involved in the process.

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES (a)

183

(b) acrosomal process

GalTase GlcNAc

cell adhesion integrin α6β1

Sperm

acrosomal vesicle

nucleus

fertilin β

zona pellucida and perivitelline layer

Egg

Figure 4.19 Mammalian gamete recognition: (a) adhesion of sperm to the outer surface (zona pellucida) of egg through GalTase-GlcNAc binding; (b) binding of acrosomereacted sperm at the side face of its head to the plasma surface of the egg through fertilin b–integrin a6 b1 interaction.

4.1.9.2 b-1,4-Galactosyltransferase. In the mouse, sperm–egg interaction initiates when multiple copies of b-1,4-galactosyltransferase (GalTase) present on the surface of the sperm heads bind to oligosaccharide ligands displayed in the egg coat, or zona pellulcida (ZP) [66,67]. The GalTase selectively recognizes GlcNAc presented on ZP3 glycoprotein (Figure 4.19). This surface-to-surface interaction involves multivalency-mediated clustering of GalTase, followed by intracellular signaling events, leading to acrosomal reactions in the sperm. The sperm-ZP3 adhesion is blocked by the addition of multivalent molecules presenting GlcNAc, including natural ZP3 or other glycoproteins. 4.1.9.3 Fertilin b. Expressed as an integral membrane protein on the extracellular surface of sperm, fertilin b is involved in sperm–egg adhesion at a stage that requires its binding to a6 b1 intergrin, which is present on the egg plasma membrane (Figure 4.19) [424]. Since the fertilin b-integrin interaction takes place at the cellular interface, it can be modulated by multivalent molecules that present ligands binding to either fertilin b or integrin. This notion is tested by Gupta and Sampson [425] and Gupta et al. [426], who designed a POPCbased liposome that presents a fertilin b peptide sequence at a density of 1 mol% on its surface. The resulting liposome inhibits sperm–egg binding with an IC50 value of 5.2 mM and is 100-fold more active than a monomeric fertilin b peptide. Moreover, this multivalent mimic of fertilin b blocks sperm–egg association completely. This degree of blocking cannot be achieved with monomeric peptides. 4.1.9.4 XL35 Lectin. An animal lectin present on Xenopus laevis oocytes, XL35 binds galactoside moieties displayed in the jelly coat glycoprotein (JCP)

184

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH OH O

NH2

HO

HO O HO HO

O O OH

+

O O O O OR OR OR O O N R O

Et3N, DMF; then NH4OH

36

OH OH O HO HO O HO HO

O O O O NH NH2 NH2 NH2

37 (χsuagr = 8, 17 mol%)

O O OH

Figure 4.20 Structure of multivalent galactosylglucoside presented on the side chain of poly(acrylamide).

that surrounds the frog egg [427]. The interaction of XL35 with JCP plays an essential role in blocking polyspermy, the adhesion of multiple sperm cells to a single egg. XL35 is an oligomeric glycoprotein that presents collectively 9 to 12 carbohydrate-binding units. In an e¤ort to understand the mechanism of XL35 binding to jelly coat mucin, Arranz-Alex et al. [427] studied poly(acrylamide) presenting multiple side chains of melibiose (Gal-a1,6-Glu) as mimics for natural JCP (Figure 4.20). One such example is a glycopolymer (37), which was modified to contain 8 or 17 mol% melibiose (36) as side chains of the polymer. The derivitized glycopolymers block the binding of XL35 to JCP with IC50 values of 16 mM (8 mol%) or 7 mM (17 mol%). These activities, which are close to the activity of natural JCP (Ki ¼ 4.3 mM ), represent a marked enhancement over the binding of the monovalent ligand 36, which shows an IC50 value of 14 mM. In a SPR study, 37 (17 mol%) binds to immobilized XL35 with a K d value of 1.67 mM, while a monovalent melibiose binds weakly, with a K d value of 1.57 mM. Thus, the XL35 lectin helps protect the egg by binding tightly to multiple sugar ligands present on JCP in frog oocytes. 4.1.10

Fucose-Binding Receptor on Cancer Cells

4.1.10.1 Fucose–Batracylin Conjugate Based on Neoglycoprotein. Certain cancer cells present lectin-binding surface receptors that recognize particular carbohydrate epitopes. For example, the colon cancer cell line SW480 contains fucose binding receptors [428]. Lerchen et al. [429] developed multivalent molecules as a molecular tool for delivering anticancer drugs to specific cancer targets (Figure 4.21). The molecular drug delivery system is based on a neoglycoprotein composed of multiple copies of batracylin, BAY H204, an anti-

185

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES O O H2N

OHO

OHOR

N

lysine

O

H N

H N

i,ii

N

S

N

BSA

Batracylin

NH2

N m = 15-30

iii

38

O HO OR 39: R = OH (fucose); n = 9 40: R = OCH2CO2H; n = 22

O OH N N

tyrosine

n HO

O

H N

H N S

N

m

N

reagents and conditions: i) CSCl2, ii) bovine serum albumin (BSA), iii) NaNO2, HCl (diazotization).

Figure 4.21 Synthesis of neoglycoconjugates presenting multiple copies of anticancer drug (batracylin) along with fucose as cancer cell–targeting ligand.

cancer drug. In addition, the molecule contains multiple copies of a fucose derivative; these serve as ligands that target the surface receptors of cancer cells. A drug-carrying glycoconjugate (39) undergoes intracellular uptake via endocytosis into SW480 tumor cells, although only weakly. However, it is taken up e‰ciently by hepatocytes that present fucose-binding receptors. A second glycoconjugate (40), carrying modified fucose residues as the ligand, shows selective and improved cellular uptake by SW480 cells but little uptake by control cells. This uptake inhibits cell growth and induces cytotoxicity against SW480, while a carbohydrate-free drug conjugate (38) lacks any activity against SW480, presumably due to an inability to permeate the cell. 4.1.10.2 Liposomes That Display GM3 –Lac Ceramide. Certain tumor cells express elevated levels of specific types of glycosphingolipid (GSL) on their surface. These cell adhesion molecules contribute to metastasis and the invasion of tumor cells into other cells and tissues [430]. For example, melanoma cells express a high number of the gangliosphingolipid (GM3 ), molecules on their surface, and the surface density of GM3 is associated directly with the metastasic competency of the melanoma cells. This type of cancer-specific expression provides an opportunity to develop multivalent GSL molecules as antimetastasis agents in cancer therapy. Otsuji et al. [431] reported the construction of liposomes that present multiple copies of GM3 –Lac ceramide or GM3 –Gg3 ceramide. The multivalent ligands block the adhesion of B16 melanoma cells to endothelial cells in vitro and furthermore, inhibit melanoma metastasis in vivo.

186

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OSO3O

O HO -O3SHN

OSO3O -O C 2 O O HO HO O HO AcHN OH -O2C O

Heparin

OSO3O O -O3SO OH -O3SHN O n

O

~40 Å Heparin FGF receptor

membrane Fibroblast growth factor (FGF)

Dimerization of FGF receptor

Figure 4.22 Dimerization of fibroblast growth factor (FGF) receptors mediated by binding of heparin-linked FGF dimer.

4.1.11

Fibroblast Growth Factors

Fibroblast growth factors (FGFs) comprise a family of multifunctional proteins that mediate cellular functions by binding to membrane-bound FGF receptors [432]. Activation of the protein kinase activity of an FGF receptor requires oligomerization of the receptor, which occurs when the extracellular domains bind to a dimeric form of FGF [433]. The FGF dimer is formed by the creation of a complex with heparinlike multivalent molecules. DiGabriele et al. [434] reported the crystal structure of a heparin-linked, biologically active form of FGF dimer (Figure 4.22). This heparin decasaccharide ligand binds to two FGF molecules through five or six contacts made by monosaccharide units with the protein. 4.1.12

Carbohydrate–Carbohydrate Interactions on Cell Surfaces

4.1.12.1 Gg3 –GM3 Interaction. Oligosaccharides present on cell membranes are in certain cases involved in carbohydrate–carbohydrate interactions that mediate cellular recognition events. Among various studies investigating carbohydrate–carbohydrate interactions, Hakomori’s works describe the role of Le x –Le x in embryogenesis, the Gg3 (gangliotriaosyl ceramide)–GM3 interaction between lymphoma and melanoma cells, and the Gg3 –GM3 interaction in signal transduction [435–438]. Using SPR spectroscopy, Matsuura et al. [439] provided quantitative estimates of the binding of polyvalent glycosides to


187

n

OH HO OH HO O O HO O O HO OH AcHN

OH O

NH O RHN

O RHN

O

OH 41 (R = Gg3)

Gg3

n

OH O HO

HO HO

O

OH O

NH O RHN

O RHN

O

OH

OH Lac

42 (R = Lac)

n

HO HO

OH O HO

O

OH O

NH O RHN

O RHN

O

OH

OH Cel

43 (R = Cel)

Figure 4.23 Structure of polyvalent glycocopolymer presenting either Gg3, Lac, or Cel as side chain.

glycolipid monolayers. The polyvalent molecules are based on polystyrene, which presents multiple copies of Gg3 , Lac, or Cel on its side chains (Figure 4.23). For example, 41 binds to GM3 monolayers with a Ka value of 2.5 10 6 M 1 , whereas a monovalent Gg3 does not adhere to the monolayer. Two other glycopolymers (42 and 43) bind with a 30- to 60-fold lower a‰nity, suggesting carbohydrate specificity. Similarly, 41 shows ligand selectivity when bound to glycolipid monolayers, with relative a‰nities of GM3 > LacCer > GlcCer. This type of homophilic carbohydrate interaction is also reported to play an important role in rainbow trout fertilization, where (KDN)GM3 on sperm heads binds to Gg3 -like epitopes present on the surface of the egg [440]. 4.1.12.2 Aggregation Factors of Marine Sponges. Cell–cell adhesion in marine sponges occurs via two events. First, in a calcium-independent process, extracellular proteoglycan-like molecules known as aggregation factors (AFs) bind on the cell surface. Second, in a calcium-dependent process, AFs selfassociate. These cell–cell adhesion events are responsible for cell aggregation and sorting [441]. A sulfated disaccharide (44) isolated from the sponge M. prolifera is identified as a surface epitope acting as an AF [442]. To study selfinteractions of the sponge, Haseley et al. [443] developed a model surface com-

188

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS BSA OH O HO HO NaO3SO

O OHO

S

O

OH

NHAc

O HO HO NaO3SO

O OHO

H N

H N

O

O

n

OH

n = 7.8

NHAc 45

44

2+ 2+ Ca2+ Ca Ca

45

Ca2+

2+ 2+ Ca2+ Ca Ca Ca2+

Ca2+ CMC sensor

Figure 4.24 Structure of neoglycoprotein presenting multiple copies of sulfated disaccharide epitope and a schematic for calcium-induced self association.

posed of a neoglycoprotein (45) attached to the surface of a carboxylmethylated dextran sensor used for SPR measurement (Figure 4.24). In the absence of Ca 2þ , 45 undergoes no observable self-interactions. However, upon addition of Ca 2þ , 45 undergoes self-aggregation to form multiple layers. Such aggregation is specific to Ca 2þ ; other divalent cations, such as Mg 2þ and Mn 2þ , do not induce such self-association. The binding strength of 45 in self-aggregates is estimated to have a K a value of about 10 5 M 1 (ka ¼ 100 M 1 /s; kd ¼ 0:001 s1 ). 4.1.12.3 Le x –Le x Interaction. Homophilic carbohydrate interactions between Le x antigens [444] are implicated in molecular compaction. De la Fuente et al. [445] reported a model surface that presents Le x carbohydrates through an alkanethiolate linkage on gold nanoparticles. Nanoparticles at a mean diameter of about 20 nm become aggregated in an aqueous solution containing Ca 2þ but remain dispersed in the absence of such divalent ions. This calcium-dependent aggregation is attributed to specific self-recognition between Le x –Le x molecules mediated by Ca 2þ . 4.1.13

E-Selectin

The recruitment of white blood cells to injured tissues involves specific interactions between the glycoprotein E-selectin [endothelial leukocyte adhesion molecule-1 (ELAM-1)] present on the surface of activated endothelial cells and a cell surface carbohydrate ligand displayed on neutrophils (Figure 4.25) [446– 448]. The basic structural motif recognized by E-selectin includes a branched


189

Leukocyte

rolling

Sialyl Lewis x E-selectin

Endothelial cell

Figure 4.25 E-selectin-mediated adhesion of leukocyte to endothelial cell.

tetrasaccharide terminated with a sulfated trisaccharide, such as 3 0 -O-SO3 Lewis x, or with a sialic acid, such as sialyl Lewis x (SLe x ) or sialyl Lewis a (SLe a ). The a‰nity of the monovalent ligands, including SLe x , to E-selectin is relatively weak, with an IC50 value of 1 to 2 mM [449,450]. However, the cell– leukocyte adhesion mediated by E-selectin is very tight, suggesting the involvement of multiple specific interactions at the cell interfaces [451,452]. 4.1.13.1 Divalent Sialyl Lewis x. To identify E-selectin antagonists that are more active than sialyl Lewis x, DeFrees et al. at Scripps [453] designed a series of bivalent sialyl Lewis x ligands composed of two copies of sialyl Lewis x linked to a galactoside template (Figure 4.26). These ligands thus represent positional isomers which di¤er only in the spacing between adjacent sialyl Lewis x. These dimers block the adhesion of HL-60 cells to immobilized Eselectin more potently than does sialyl Lewis x. For example, the IC50 value of 46 (3,6-orientation) is 0.2 mM, which indicates sixfold more activity than that of sialyl Lewis x alone. Other orientational dimers (47 and 48) are three times more potent than sialyl Lewis x. Thus, the point of attachment on a galactoside sca¤old influences significantly the a‰nities of the resulting dimers in the order 3,6-linked > 4,6-linked > 2,6-linked > monomer. The hierarchy of activities probably stems from di¤erences in distance and orientation between the two presented ligands. In another report, Wittmann et al. [454] studied dimers of sialyl Lewis x linked through oligo(ethylene glycol) spacers of varying lengths (Figure 4.27). Dimers (49) show inhibitory activity against immobilized E-selectin, with IC50 values of 70 mM (n ¼ 3) and 81 mM (n ¼ 6). Such activities represent approximately a five- to sixfold enhancement over the sialyl Le x monomer. ¨ hrlein [455] re4.1.13.2 Trivalent Sialyl Lewis x Constructs. Baisch and O x ported di- and trivalent sialyl Le molecules attached to an oligopeptide

190

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH OH -O2C HO OH O OH AcHN O NHAc O O O OH OH O 6 OH OH O O O OEt OH OH HO AcHN O O OH OH 3 O OH HO O O O HO -O2C OH O HO OH O AcHN 46 (3,6-linked) OH OH

OH OH OH OH O OH HO -O2C HO NHAc O O O O O AcHN O OR OH OH OH HO

HO OH

Sialyl Lex

OH OH OH OH CO2OHOH OH OH OHOH HO O O -O2C OH OH O OH O OH AcHN O OH NHAc O O O OH OH OO OH HO O NHAc HO O O OH HO O OH O OH 6 HO O HO NHAc O OH O(CH2)5CO2Me -O2C OH 6 O O HO O OH O HO O(CH2)5CO2Me O O 4 O O AcHN OH AcHN O OH HO O OH OH OH 2 OH OH O O HO HO -O2C O O HO O OH O OH AcHN 47 (4,6-linked) OH OH 48 (2,6-linked) AcHN HO

O

OHOH HO HO HO HO HO

NHAc OH O O O OEt NHAcO OH O

HO i

HO OH

HO

O HO O OH HO OH HO O O OH

NHAc OH O O O OEt NHAc O OH O

ii

OH

OHOH -O2C OH OH AcHN O NHAc O O HO OH OH OH O O O OH HO O OEt AcHN O OH O OH HO HO O O HO -O2C OH O HO OH O AcHN

OH

iii 46

OH

reagents and conditions: i) UDP-glucose, uridine 5'-diphosphogalactose 4'-epimerase, β-1,4-galactosyl transferase; ii) CMP-N-acetylneuraminic acid, α-2,3-sialyltransferase, iii) GDP-fucose, fucosyltransferase.

Figure 4.26 Chemoenzymatic synthesis of galactoside-anchored bivalent sialyl Le x .

191

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES OH OH OH O OH OH HO -O2C HO NHAc O O O O O AcHN O O OH OH OH HO

O

H N

N

NH2 +

O

O

O O

O

O

O

O n

O

N

O

n = 2 to 6

Et3N, DMF

OH OH OH OH O OH HO HO -O2C NHAc O O O O O AcHN O O OH OH OH HO OH OH O OH OH HO -O2C HO NHAc O O O O O AcHN O O OH OH OH HO

H N

NH

O

O O

OH

H N O

O N H

n

O

49 (n = 3, 6) x

Figure 4.27 Synthesis of divalent sialyl Le molecules of varying linker length.

framework. These constructs closely mimic the clusters of sialyl Le x moieties presented on natural glycoproteins (Figure 4.28). The synthesis combines both chemical and enzymatic methods. A trivalent sialyl Le x ligand (51; n ¼ 4) shows a‰nity to E-selectin with an IC50 value of 0.14 mM per sialyl Le x , which is fourfold more active than a monovalent sialyl Le x . A second trivalent ligand (51; n ¼ 5) is slightly better than monomer control, with an IC50 value of 0.39 mM. Sprengard et al. [456] reported a trivalent molecule that presents three sialyl Lewis x moieties attached to an oligopeptide sca¤old. The synthesis is based on a cyclic heptapeptide framework prepared by a solid-phase method and includes coupling of a partially protected sialyl Lewis x to the sca¤old in the solution phase (Figure 4.29). This trivalent ligand (52) blocks the binding of E-selectin to sialyl Lewis x–presenting tumor cells, such as cell line HL-60, with an IC50 value of 0.35 to 0.6 mM, thus representing two- to threefold higher activity than the SLe x monomer. 4.1.13.3 Neoglycoproteins Displaying Sialyl Lewis x Ligands. Bovine serum albumin (BSA) is used as a sca¤olding to present SLe x in multiple copies as a ligand for E-selectin, as reported by Welply et al. at Monsanto [457]. For example, a BSA conjugate presenting 16 copies of SLe x per protein shows inhibitory activity against cell adhesion mediated by E-selectin, with an IC50 value of 1 mM. This activity is about three orders of magnitude greater than that of the monomer. The structure of this neoglycoconjugate can be altered to attach biotin. The resulting multivalent ligand provides a way to stain the surface of

192

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS HO HO HO OH AcHN O O

HO HO HO

O +

HO HO HO

O

H N

H N NHAc

O

CO2tBu

NHAc

HN i

HO HO HO

O

NHAc

O (CH2)n NH

NH-CO-(CH2)n-NH2 AcHN

50

O

CO2tBu

O

NH NHAc

H N

O

HO O H HO N HO t CO2 Bu AcHN O

ii,iii

NH NHAc

O

N H HN

NH-CO-(CH2)n

O

O O

HO HO HO

n = 4, 5

O iv,v,vi

(CH2)n NH AcHN

O

O O

HO HO HO

NH NHAc

OHOH AcHN HO

HO

CO2- OH OH O O HO O OH O

O

AcHN OH

OH

OH COHO 2HO O O O O AcHN O OH OH OH HO

O

H N NHAc

HO OH NH-CO-(CH2)n

O

OH

O

H N O

N H HN

tBu

CO2

O

51 (n = 4, 5)

(CH2)n NH

OHOH AcHN OH

OH CO2-OH HO O O O O AcHN O OH OH OH HO

O

O

O O

NH NHAc

OH OHOH reagents and conditions: i) DMF, DCC, HOBt; ii) TFA, 0oC; iii) 50, DCC, HOBt, DMF; iv) UDP-Gal, galactosyl transferase, pH 7.5; v) CMP-sialic acid, α(2,3)-sialyl transferase, pH 6.5; vi) GDP-fucose, fucosyl transferase, pH 6.5.

Figure 4.28 Chemoenzymatic synthesis of trivalent sialyl Le x constructed on a peptide sca¤old.

endothelial cells expressing E-selectin. Such cells are colored in the presence of bound biotin-presenting multivalent SLe x and Texas red-linked avidin. 4.1.13.4 Liposomes Displaying Sialyl Lewis x. Liposomes serve as an e¤ective sca¤old for the multivalent presentation of selectin ligands. DeFrees et al. [458] and Allen et al. [459] reported a phospholipid molecule terminated with sialyl Le x on its distal end. As shown in Figure 4.30, the sialyl Le x moiety is tethered via poly(ethylene glycol) (PEG), which is long and flexible. The incorporation


193

O HO HO HO AcHN

O

O OH OH

O

HN O OBn BnO O O O O OH

HN O

Me O

NH2 NHAc

+

O

NH

O

NH

HO

OBn BnO OBn

O O

i-iii

NH

Me NH O

O

OH

HN O

HO HO

OH HO2C OH HO O O O O AcHN O OH OH OH

H O N NHAc O HN OH Me HN OH O OH HO OH O NH HO HO2C OH O NH O HO HO OH O O O O O O AcHN NH N HO O OH OH OH NHAcH HO HO Me NH O HO O AcHN NHAc OH OO O O HN N O O OHOH H H CO OH OH O 2 OH HO OH O

52 reagents and conditions: i) HATU, HOAt, DIPEA; ii) H2, Pd/C, MeOH, dioxane, AcOH; iii) NaOH, H2O, MeOH, pH 10.6.

Figure 4.29 Synthesis of a trivalent sialyl Le x built on cyclic oligopeptide.

of a PEG spacer is intended to minimize potential nonspecific protein–ligand associations. Liposomes averaging 100 nm in diameter are prepared by forced filtration of a mixture of 54, cholesterol, and phospholipid through membrane filters. The liposomes that present 0.5 to 5 mol% of 54 block the binding of HL60 cells to immobilized E-selectin with an IC50 value of 0.2 mM. This activity represents a 5000-fold more potent inhibition than that of the native sialyl Le x oligosaccharide, as well as about a 750-fold greater inhibition than that of a nonlipidated derivative of sialyl Le x (53). This enhancement in inhibitory activity is attributed to multivalent interactions between E-selectin and the multivalent presentation of sialyl Le x . Stahn et al. [460] reported a separate study that shows assay-dependent inhibition of E-selectin activity by liposome-based SLex ligands. For example, in an assay that examines the ability of cells to bind to anti-SLe x antibody in solution, liposomes displaying SLe x ligands inhibit antibody binding more e¤ectively than monovalent SLe x by seven orders of magnitude. The same liposomes inhibit the binding of cells to immobilized E-selectin by five orders of magnitude. In addition to the functional groups required for E-selectin recognition, the conformation of SLe x carbohydrates bound to E-selectin is considered an important factor when designing SLe x mimics [461]. Wong et al. [462] designed a fucosyl-containing glycopeptide (55) as a SLe x mimic (Figure 4.31). It shows a

194

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OHOH NaO2C OH OH OH HO OH AcHN O O O O O OH O OEt OH O O OH OH NH O OH H3 C OHOH

PEG Liposome

SLex

OAc OAc MeO2C OAcOAc AcHN O O OAc OAc O O HO OAc

OAc O

OAcOAc

i,ii,iii

O

O

NH2

OAc

~100 nm

OAc OAc MeO2CAcO AcHN O OAc OAc O

iv,v

OH O

OAcOAc O

O

OAc

OEt

O O O O OH NH O OH H3 C O OHOH

OH OH O OH O

OH O

O O

OH O

OH NH O OH H3 C O OHOH

H N

OEt + HS

O n

O

O Br

N H

vi

OAc O

O O

NH2

OH

OHOH NaO2C HO AcHN O OH OH O

O

OAc NH O OH H3 C O OHOH

OEt 53

OHOH NaO2C HO AcHN O O OH OH

OAc

H N

O

O

O P O O-

O O C17H35 C17H35

O O

n = 42-48

OHOH O

O

OH O N H

OEt

54 H N

S O

O

O n

H N O

O

O P O O-

O O C17H35 C17H35

O O

reagents and conditions: i) p-Nitrobenzoyl chloride, NaHCO3, CH2Cl2; ii) tri-O-benzyl-α/β-L-fucopyranosyl fluoride, AgClO4, SnCl2, TMU, CH2Cl2; iii) H2, Perlman's catalyst, MeOH, H2O; iv) NaOMe, MeOH, H2O; v) bromoacetic anhydride, EtOH, H2O, NaHCO3; vi) Na2HPO4, EDTA, pH 7.2.

Figure 4.30 Liposome presenting multiple copies of sialyl Le x ligand on its surface tethered through PEG linker.

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES E-selectin recognition units HO OH O OH HO HO AcHN

-O2C HO OH O

O

OH OH

NHAc

O O OH HO O

-O

O OR

2C

OH OH H3 C O OH OH O CH3 OH O NH CO2Et N H O 55

Sialyl Lewis x

BnO2C

iii

HO2C

iv

HO2C

OBn OBn H3C O OBn OBn O CH3 OH O OH NH N O H O OH OH H3C O OH OH O CH3 OH H O N NH N O H O OH OH H3C O OH OH O CH3 OH H O N NH N O H O 56

195

H2N

O i,ii

O

O

N3 3 HO2C

OH OH H3C O OH OH O CH3 OH H O N NH N O H O

O

NH2 3

H N 3

O

H N 3

O O HO

n reagents and conditions: i) EDC, HOBt, CH2Cl2; ii) H2, Pd(OH)2/C, EtOH, H2O, dioxane, AcOH; iii) di-yne-acid NHS ester, Et3, DMF; iv) di-yne acid, HEPES buffer, sonication; then hν, 0oC.

Figure 4.31 Synthesis of sialyl Le x mimetic and its conversion to polymerized liposome.

binding a‰nity to E-selectin with an IC50 value of 0.5 mM, an activity equipotent to sialyl Le x [462]. When 55 is derivatized as a distal group at the end of di-yne-incorporated lipid, the resulting lipid can be converted to polymerized liposome (56), which displays 55 on its surface. This polymerized liposome shows enhanced inhibitory activity to E-selectin with an IC50 of 30 mM. Its activity represents about a 16-fold enhancement over 55. 4.1.13.5 Polyvalent Polymeric Sialyl Lewis x. A linear polymer can be used as a sca¤old to present multiple copies of sialyl Le x on its side chains, as shown in Figure 4.32. Thoma et al. [463] reported such a synthetic glycopolymer con-

196

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS Cl NH2

H N

O N H

NH2

NH2

O

H N O

N H

60

O

NH2

NH

O

O

H N

O

O

H N

i N H

HN

HN

N H

O

O

O

60

NH

O

Cl

Cl

Cl

HO OH O OH HO AcHN

OH

-O2C HO OH O

OH OH

ii,iii

O

O O OH HO O

O HO

S OH

S

N H

H N

57 SLex-thiol (20 mol%)

OH

O

NHAc O O

NH

HN

O

H N

O

O N H

O

S HN

N H

O

O

OH

O

60

NH S

OH OH

reagents and conditions: i) chloroacetic anhydride, 2,6-dimethylpyridine, DMF, 0oC; ii) sialyl Lewis x thiol (20 mol%), DBU, DMF; iii) excess thioglycerol, DBU, DMF

Figure 4.32 Synthesis of polyvalent poly(lysine) presenting sialyl Le x as a side chain.

sisting of commercially available poly(lysine) containing on average 240 lysine monomers per polymer molecule. Its amine-terminated side chains are derivatized with chloroacetyl groups, which then react with sialyl Le x -terminated alkanethiol groups for covalent attachment. Varying the ratio of sialyl Le x thiol to the pendant chloroacetyl group on the polymer chain controls the sidechain density of sialyl Le x . Such a method leads to synthesis of 57, which contains about 20 mol% of sialyl Le x relative to the total number of side chains composed of glycerol and sialyl Le x . However, this glycopolymer shows little inhibitory activity to E-selectin at 0.1 mM. The E-selectin activity improves on the use of a ligand that binds more tightly than sialyl Le x [464]. For example, sialyl Le x mimic 58 (Figure 4.33) shows a high a‰nity to E-selectin, with an IC50 value of 36 mM. This value represents a 30-fold enhancement over sialyl Le x [465]. When this ligand is converted to a polyvalent polymeric ligand (59, 20 mol %), the polymer shows an IC50 value of 2 mM and an 18-fold higher a‰nity to E-selectin than that of the monomer control (58). The size of such a glycopolymer influences the Eselectin activity. Thus, an analog of 59 (20 mol%; n ¼ 1200), which is composed of a greater number of monomer units, shows an IC50 value of 0.2 mM.


197

HO OH O OH -O2C HO OH O

O O O OH HO

O

58 HO OH O OH

HO OH O OH

-O2C HO OH

-O2C HO OH O

O

O O

OH

i

O

O

O NH2

ii,iii

O O

O NH

OH

SH

O HO OH O OH -O2C HO OH O

O

O O

O NH

OH

OH S

O

H N

59 SLex mimetic (5 to 100 mol%)

O S

HO OH

NH

HN

O

H N

O

O N H

O

S HN

N H

O

O

OH

O

n = 60 or 300

NH S

OH OH

reagents and conditions: i) thiobutyrolactone, Et3N, MeOH; ii) chloroacetylated poly(lysine), DBU, DMF; iii) thioglycerol, DBU, DMF.

Figure 4.33 side chain.

Synthesis of polyvalent poly(lysine) presenting mimics of sialyl Le x as a

The density of sialyl Le x mimetics also influences the activity such that the density of 59 responsible for the highest activity (IC50 ¼ 0.05 mM ) is observed in the range 35 to 50 mol% (n ¼ 1200). This optimized activity represents roughly a 50,000-fold improvement relative to sialyl Le x . Glycopolymers containing 58 at a density outside the optimum range show lower activity. Sialyl Le a also serves as a ligand that binds to E-selectin. Weitz-Schmidt et al. [466] reported the synthesis of a polyvalent ligand that displays sialyl Le a carbohydrates (Figure 4.34). It is prepared by reacting a derivative of sialyl Le a with the activated ester of poly(acrylic acid), which leads to a poly(acrylamide)based ligand (60). A glycopolymer containing a 20 mol% density of sialyl Le a binds tightly to immobilized E-selectin with an IC50 value of 2.4 mM. The E-

198

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS HO OH O OH OH -O2C HO OH HO O O O O O O AcHN OR OH OH OH AcHN OH

Sialyl Lewis a

HO OH O OH -O2C HO OH HO HO O O O O O AcHN OH OH OH AcHN OH

O

O

O OR

n O OR

+

O O

i) DMF, rt NH2

ii) 2-ethanolamine

R = 4-nitrophenyl NO2 HO OH O OH OH COHO HO HO 2OH O O O O O AcHN OH OH OH AcHN

O O

n NH O HN

O HN

OH

OH

O

60 (χSLea = 2 - 20 mol%)

Figure 4.34 Structure of poly(acrylamide)-based polyvalent sialyl Le a .

selectin a‰nity is a¤ected by the SLe a density in the polymer chains such that a lower density of sialyl Le a , such as 2 or 5 mol%, leads to inactive ligands. 4.1.13.6 Sialyl Le x –Binding Multivalent Receptors. In designing a nonnatural receptor that mimics E-selectin, a multivalent strategy may find success. Using phage display, Kwon et al. [467] that identified a 12-amino acid peptide (AHWIPRYSSPAT) that binds weakly to sialyl Le x -conjugated bovine serum albumin, with a K d value of >1 mM. When this peptide is tethered to a lysinebased sca¤olding and displayed in multimeric form, its a‰nity to BSA-sialyl Le x conjugate increases 14-fold, with a resulting K d value of 70 mM for a dimer, [AHWIPRYSSPAT]2 K. The a‰nity increases 10,000-fold, with a Kd value of 0.1 mM for a tetramer, [[AHWIPRYSSPAT ]2 K]2 K. This example illustrates that the valency of the carbohydrate-binding receptor plays a major role in enhancing the binding a‰nity to its cognate multivalent ligand. 4.1.14

L-Selectin

L-selectin, one of numerous carbohydrate-binding proteins located on the surface of leukocytes, is involved in the binding and rolling of leukocytes on endothelial cells [468] (Figure 4.35). Physiological molecules that bind to Lselectin include mucinlike glycoproteins, such as the glycosylated cell adhesion molecule-1 (GlyCAM-1), PSGL-1, and CD34 [469,470]. These glycoproteins


199

Leukocyte

rolling

L-selectin

Sialyl Lewis x E-selectin

Endothelial cell

HO OH OH AcHN HO

HO

NHAc OH OH

O

CO2O OH O HO OH OH OH HO O OH O O OH OH OH O OHO O OH HO O OO NHAc HO HO O NHAc O OH OH O O OH AcHN HO O HO O HO CO2-

OH CO2OH OH OH O O OO OH HO O HO OO HO OH O NHAc HO HO NHAc O O O O HO OH HO HO OH OH O O NHAc O O AcHN O OH O O NHAc O O HO OH OH HO O O OR -O2C O O HO HO O OH HO AcHN O O O O 61 O O OH AcHN OH OH HO O HOHO OH O OH O O HO NHAc OH O OH O O O HO -O2C 62 O O HO HO AcHN O O HO OH O OH AcHN OH O OH OH O HO O HO

HO HO

NHAc O OR

O OH O

-O2C O HO HO

OH

OH NHAc

Figure 4.35 Adhesion and rolling of leukocyte on an endothelial cell mediated by l-selectin and its ligand on a cellular interface, and the structure of branched multivalent sialyl Le x .

200


present on their surface multiple copies of anionic saccharide epitopes, including sulfated (sialyl) Le x and Le a . These sialyl carbohydrates serve as ligands to L-selectin. This multivalent display is probably the basis for the high-avidity interaction with L-selectin on leukocytes. For example, a divalent cluster of sialyl Le x (61) blocks selectin-mediated lymphocyte–endothelium adhesion with an IC50 value of 150 nM (Figure 4.35). This activity is considerably more than that of a monovalent sialyl Le x , which has an IC50 value greater than 10 mM [471]. When the valency of sialyl Le x is increased to 4, as in a branched sialyl Le x (62), the molecule blocks selectin-mediated adhesion much more effectively, with an IC50 of 5 nM [472]. As a naturally occurring divalent ligand for L-selectin, 63 is composed of two sialyl Le x saccharides linked in a linear manner (Figure 4.36). The chemical synthesis of this ligand was reported by Nicolaou et al. [473]. 4.1.14.1 Divalent Sulfated Lactoside. The a‰nities of multivalent mimics of natural L-selectin ligands can readily be modulated, first, by controlling distance and orientation between saccharide units, and second, by varying the nature of the sca¤old. Kiessling et al. [474] reported the synthesis of a class of multivalent ligands that binds L-selectin (65). These molecules display 3 0 sulfoallyllactoside (64), a simple analog of sialyl Le x , through asparagines Lex

Lex

HO OH O OH HO

OH

OHOH

-O2C O

AcHN

NHAc OHOH O O

O O OH HO O

O

OH OH

HO OH O OH NHAc O(CH2)6CO2But

O O OH HO

O

O

63

BnO OBn O OBn BnO MeO2C AcO OBn BnO O O O O O AcHN OBn OBn OAc

BnO OBn O OBn NPhth O F

AcHN

O OBn OBn

O

O OAc

O O

i

OHOBn HO

O OH

OBn

BnO OBn O OBn BnO BnO MeO2C AcO OBn

+

O O

NHAc t O O(CH2)6CO2Bu OBn

BnO OBn O OBn NPhth OHOBn O O O

OBn

OH

ii-v O O

NHAc t O O(CH2)6CO2Bu OBn

reagents and conditions: i) AgOTf, HfCp2Cl2, CH2Cl2; ii) LiI, pyridine, 120oC; iii) hydrazine hydrate, MeOH, benzene; iv) Ac2O, MeOH, CH2Cl2; v) H2, Pd/C, MeOH, rt, 2d.

Figure 4.36 Synthesis of sialylated divalent Lewis x.

63

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES HO OH O O HO HO OH

OH O O OH

HO OH O O NaO3SO HO OH

HO OH O

i,ii

OH O O OH

O NH2 +

AcHN

N H O

CH3 H N n O

HO

64

OH O O OH

O HO OH

NaO3SO

O

iii,iv,v

CH3

N H O

201

CONH2

vi

HO n = 1, 3

O AcHN HO OH O

NaO3SO

OH

O HO

OH HN O O OH

N H O

CH3 H N n O HN

O N H O

CH3 CONH2 HO O O HO

OH O

HO O

OSO3Na

HO OH

65 (n = 1, 3) reagents and conditions: i) Bu2SnO, MeOH; ii) SO3•NMe3, THF; iii) O3, MeOH, -77oC, then PPh3; iv) BnNH2, HOAc, BH3•pyr, MeOH, H2O; v) Pd(OH)2, H2; vi) HBTU, HOBt, DMF, rt.

Figure 4.37 Synthesis of sulfated bivalent lactoside as l-selectin ligand.

linkages on the framework of an alanine peptide (Figure 4.37). By varying the size of the peptide sca¤old of 65, the distance between the sulfated ligand units can be adjusted. 4.1.14.2 Neoglycopolymers Presenting Sialyl Lex . The glycoprotein, GlyCAM-1, is a heavily O-glycosylated mucin. GlyCAM-1 displays multiple copies of certain saccharide epitopes, such as 6- or 6 0 -sulfated sialyl Le x , that are recognized by L-selectin. Sanders et al. [475] reported the synthesis of multivalent mimics of natural GlyCAM-1. The mimics are prepared by polymerizing sulfated Le x monomers in aqueous solution via ROMP. The reaction is catalyzed by a ruthenium metal complex (Figure 4.38). The ROMP reaction generates neoglycopolymers such as 66 and 67, which present 15 copies of sulfated Le x ligand per polymer chain. This synthetic method is versatile, since it works well in water, and with substrates containing unprotected polar and charged functionalities. It also enables the synthesis of neoglycopolymers of various size by varying the ratio of initiator to monomer. These glycopolymers actively inhibit the binding of L-selectin to heparin, with IC50 values of 39 mM (580 mM per Le x ; 66) and 35 mM (520 mM per Le x ; 67). They are more active than sialyl Le x monomers, which show an IC50 value of 3 mM. Glycopolymer 66, which displays 3 0 ,6-disulfoLe x , potently blocks L-selectin-mediated cell rolling when evaluated under conditions of physiological shear flow, with an IC50 value of 15 mM per sialyl Le x . It is more potent than monovalent sialyl Le x , which re-

202

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS HO HO AcHN

OHOH

-O2C O

OH OH

O

O

OSO3O OR

OO

OH O OH OHOH

H 3C

NHAc

GlyCAM-1 epitope

Ph O

HO OH -O

O O O OH H3C O OH OHOH

3SO

OSO3HN O O OH

PCy3 Ph Cl Ru Cl PCy3

n = 15

HO OH -O

O O O OH H 3C O OH OHOH

3SO

DTAB, DCE pH 6, 60oC

OSO3HN O O

O

OH 66

Ph HO OSO3 -O

O O OH O H3C O OH OHOH

3SO

O OH HN O O OH

PCy3 Ph Cl Ru Cl PCy3 DTAB, DCE pH 6, 60oC

HO

OSO3-

O O O OH H3C O OH OHOH

-O SO 3

n = 15 OH HN O O OH

O

67 x

Figure 4.38 Synthesis of multivalent disulfo Le neoglycopolymer as l-selectin ligand.

mains weakly active, with an IC50 value of 2.5 mM. In contrast, the second polymeric ligand 67, presenting 3 0 ,6 0 -disulfoLe x , shows no inhibition at all against cell rolling. This example illustrates that under shear flow, the avidity of multivalent, sulfated Le x to L-selectin is a¤ected not only by multivalent ligand presentation but also by the display pattern of sulfate anions along a polymer framework. The neoglycopolymer 66 shows not only tight binding to an L-selectin array but also functional activity of biological significance. When it is incubated with human neutrophils, it induces shedding of the extracellular portion of the Lselectin protein from the cell surface [476,477]. As a result, the neoglycopolymer decreases the density of L-selectin on cell surfaces, which is expected to decrease the leukocytes’ adhesion ability. A neoglycopolymer that contains a fluorescent probe allows the visualization of polyvalent ligands interacting with L-selectin on cell surfaces. In Figure 4.39 is shown as an example of the synthesis of a fluorescence-capped neoglycopolymer (68), which possesses multiple copies of 3,6-disulfogalactoside on its side chains [478]. Upon interacting with Jurkat cells, this neoglycopolymer binds to localized sites on cell surfaces. The resulting patches are visualized by adding a fluorescent, anti-L-selectin antibody. Cells lacking L-selectin do not bind 68. The binding pattern of 68 to


203

O HO OH O O HO OH

N3

PCy3 Ph Cl Ru Cl PCy3

HO OSO3Na O O NaO3SO OH

i,ii,iii

Ph HO OSO3Na HN O O NaO3SO OH

v

NH2

O

RuLm n O

n HO OSO3Na HN O O NaO3SO OH

O

O

O

O

MeO

TMS

O

TMS

vi,vii O

O HO2C O

Ph n HO OSO3Na HN O O NaO3SO OH

O

end-capping

O Ph

HO OSO3Na HN O O NaO3SO OH

iv

O

O

O

S N H

N H

N H

OH Fluorescent probe

O 68

reagents and conditions: i) Bu2SnO, MeOH, heat; ii) SO3•Pyr, pyridine; iii) H2, Pd(OH)2/C; iv) (±)-norbornene 5-carboxylic acid NHS ester, N-methylmorpholine, DMF; v) dodecyltrimethyl ammonium bromide (DTAB), bis-tris buffer, pH 5.9, dichloroethane, 40oC; vi) 50 mM NaOH, 60oC; vii) 5-((5-aminopentyl)thioureidyl) fluorescein, EDCI, NHS, H2O.

Figure 4.39 Synthesis of end-capped neoglycopolymer carrying multiple sulfated galactosides and a fluorescent probe attached at its distal end.

L-selectin on leukocyte surfaces is similar to that observed upon adhesion of multivalent sialyl Le x to activated endothelial cells that express E-selectin [479]. 4.1.14.3 Poly(acrylamide) Sca¤olds That Present Sulfated Le x . As a polymer sca¤old, poly(acrylamide) allows the polyvalent display of 3 0 -sulfo-Le x for Lselectin binding. Sasaki et al. [480] reported examples such as a terpolymer that presents both a 3 0 -sulfated Le x mimic (side chain density ¼ 7%) and 6-sulfo-GlcNAc (6%) as side chains. This polymer blocks the binding of Lselectin to SLe x with an IC50 value of 3 mM per glycoside. It is more active than the free ligands, which remain inactive at 50 mM. 4.1.14.4 Poly(lactosamine) Sca¤olds That Present SLe x . A branched poly(lactosamine) is used as a sca¤old by Renkonen et al. [481] in their design of a nanomolar inhibitor of lymphocyte adhesion. This glycopolymer-based ligand, which is readily synthesized with chemoenzymatic methods, comprises four

204


P-selectin

+++

+++

Tyr-OSO3-

Tyr-OH

SLex

SLex

PSGL-1 (peptide 9 to 18) Kd = 10-3 M

Kd = 10-8 M

HO

OH CO2-

AcHN HO

O

OH OH OH OH O OH

O

HO

O HO

HO

O - OH CO2OH

HO

AcHN

OO HO

O OH OH

O

NHAc OH O

O OH

O

O NHAc O

Gln-Ala-Glu-Tyr-Glu-Tyr-Leu-Asp-Tyr-Asp-Phe-Leu-Pro-Glu N H OSO3OSO3- OSO3-

H N Glu-Pro-Pro O

P-selectin epitope of PSGL-1

Figure 4.40 Interaction of P-selectin and PSGL-1 (P-selectin glycoprotein ligand-1).

copies of SLe x attached to the backbone of poly(LacNAc) in a b-1,3 0 or b-1,6 0 orientation. 4.1.15

P-Selectin

As a natural ligand to P-selectin present on human neutrophils, P-selectin glycoprotein-1 (PSGL-1) plays a primary role in leukocyte recruitment during inflammation [482,483]. It is an O-glycosylated, mucinlike molecule carrying multiple copies of sialyl Le x residues, which serve as a minimal recognition motif to P-selectin (Figure 4.40). Optimal binding by PSGL-1 occurs in cooperation with several O-sulfated tyrosines that are located proximally to the sialyl Le x , which is attached to threonine. These sulfated tyrosines contribute to the enhanced a‰nity of PSGL-1 to P-selectin by enabling bivalent recognition between the ligand and P-selectin. Indeed, the sulfated glycopeptide binds to Pselectin with a Kd value of about 108 M—approximately 105 more tightly than the unsulfated glycopeptide ligand, which has a Kd value of about 103 M [484]. Koeller et al. [482] reported the synthesis of a PSGL-1 binding determinant by combining chemical solution and solid-phase methods and an enzymatic glycosyltransferase reaction. The method provides a complex sialyl Le x attached heptapeptide (69) (Figure 4.41).

205


AcHN OAc AcO AcO O O O AcO AcO AcHN O

Rink amide resin H2 CH3

N-Glu(But)-HN i-iv

OH O O HO AcHN O

NH-Glu(But)-HN O

O AcHN HO AcO O HO

v,vi CH3

Ac-Tyr-Asp(But)-Phe-Leu-Pro-Glu(But-HN )

OH

FmocHN

AcHN OAc AcO AcO O O O AcO AcO AcHN O

vii,viii,ix CH3 NH-Glu-NH2

Ac-Tyr-Asp-Phe-Leu-Pro-Glu-HN O

HO OH OH O OH -O2C HO OH HO NHAc OH O O O O O O O O AcHN HO OH HO OH OH AcHN O 69

CH3 NH-Glu-NH2

Ac-Tyr-Asp-Phe-Leu-Pro-Glu-HN O

reagents and conditions: i) HBTU, HOBt, NMM, DMF; ii) morpholine, DMF; iii) Fmoc-AA-OH, HBTU, HOBt, NMM, DMF; iv) repeat ii and iii, then Ac2O, Pyr; v) TFA, H2O, EDT; vi) NaOH, MeOH; vii) β1,4-galactosyl transferase; viii) α2,3-sialyl transferase; ix) α1,3-fucosyltransferase.

Figure 4.41 Synthesis of sialyl Le x -attached heptapeptide as P-selectin-recognition glycopeptide epitope.

4.1.15.1 Sialyl Lewis x-RGD Conjugates. A heterobivalent glycoconjugate comprised of a sialyl Lewis x moiety attached to a tripeptide, Arg-Gly-Asp (RGD), provides two recognition elements. One element recognizes selectin and the other recognizes the integrin receptor. Depending on the spacer length and interligand orientation, the RGD motif can bind to an accessory site located close to the SLe x binding site in P-selectin. Two typical RGD-sialyl Lewis x ligands (70 and 71) are designed based on such a notion (Figure 4.42). With an IC50 value of 26 mM, the ligand 70, tethered with short alanine spacer, actively inhibits P-selectin-mediated cell adhesion more potently than sialyl Le x [485]. The second ligand, 71 with a longer spacer, binds to immobilized Pselectin with a Ka value of 6.6 10 7 M 1 [486]. This a‰nity represents about a 100- to 1000-fold enhancement over either sialyl Le x (K a ¼ 8.8 10 5 M 1 ) or RGDS (K a ¼ 9.7 10 4 M 1 ) alone. In the presence of both P-selectin and intergrin b1, 71 interacts with the two di¤erent receptors simultaneously, thus forming a ternary complex. 4.1.15.2 Copolymers Presenting Sialyl Le a Residues and Sulfated Tyrosines. Synthetic polymers possessing multiple copies of sialyl Le a and O-sulfated tyrosine in their side chains closely mimic the natural PSGL-1 ligand. Pochechueva et al. [487] reported the synthesis of a molecule in which poly(acrylamide) presents sialyl Le a along with sulfated tyrosine as a secondary recognition motif of P-selectin (Figure 4.43). One such glycopolymer (72) is a potent inhibitor of P-selectin, with an IC50 value of 0.25 mM per carbohydrate. Such an inhibition is 480 times greater than that of 73, which does not carry

206

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OHOH OH O OH -O2C HO OH HO NHAc CH3 O O O NH O O O O N AcHN OH HO H OH OH O O

O

H N O

NH3+

N H

O-

NH

70 (sLex-Ala-Asp-Gly-Arg)

OHOH OH O OH -O2C OHOH HO NHAc O O O O O O O AcHN OH HO OH OH

H2N

O +H

3N

N H

H N O

+

H N

NH2

O

H N

N H

O

H2N

+

N H O

O

NH

OH

O

OO

O-

NH2

71 (sLex-spacer-Lys-Gly-Arg-Gly-Asp-Ser)

Figure 4.42 RGD-linked sialyl Le x glycoconjugates.

sulfated tyrosine residues. With an IC50 value of 1.3 mM, poly(acrylic acid)based polyvalent sialyl Le a (74) is far better than poly(acrylamide)-based ligand 73. The enhanced a‰nity from the anionic polymer is attributed to the involvement of carboxylate side chains in multivalent binding to P-selectin sites, in a role similar to that of sulfated tyrosine. 4.1.15.3 Neoglycopolymers Presenting Sulfated Galactosides. Although sLe a is a P-selectin ligand, its complex structure makes large-scale preparation di‰cult. In contrast, mimics of P-selectin ligand are relatively simple in structure, but bind weakly. However, these can be manipulated into multivalent molecules for tighter binding. Manning et al. [488] reported the synthesis of a polyvalent neoglycopolymer that presents sulfated galactosides as side chains (Figure 4.44). For example, 75, which presents the 3,6-disulfogalactoside ligand, inhibits cell adhesion to immobilized P-selectin with considerable activity, as reflected in an IC50 value of 167 mM per galactoside. In blocking P-selectin-cell adhesion, it is about 20 times more e¤ective than sialyl Le x , which has an IC50 value of 3.4 mM. However, a neoglycopolymer (76) carrying monosulfated galactosides at the 3-position shows only weak inhibitory activity, with an IC50 value of about 1 mM. This example illustrates that high-a‰nity binding to Pselection requires multiple 6-sulfogalactosides, a structural feature consistent with that of the GlyCAM-1 ligand, which displays multiple copies of sulfated sialyl Le x at the 6-position of the galactoside determinant [489]. 4.1.15.4 Polymerized Liposomes Presenting Mimetics of Sialyl Lewis x. The liposome serves as an excellent tool for the multivalent presentation of Pselectin ligands on the outer surface. Spevak et al. [490] (see also Bruehl et al.

207

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES P-selectin

P-selectin

+++

+++

polymer

SLex

Tyr-OSO3-

HO OH O OH OH CO2-OH OH HO O O O O O O AcHN O OH OH OH AcHN

i-iii

OH

+

n O RO

O RO

O

O

NH2

Sialyl Lewis a-O(CH2)3NH2 R = 4-nitrophenyl NO2

HO OH O OH OH CO2-OH OH HO HO O O O O O AcHN OH OH OH AcHN

O

n O

NH O HN

O HN

O

HO2C OH

72 (χSLea = 15 mol%, χsulfo-tyr = 20 mol%) 73 (χSLea = 20 mol%, χsulfo-tyr = 0 mol%)

OSO3HO OH O OH CO2-OH OH HO = 20 mol%) O O O O O AcHN OH OH OH AcHN HO HO

74 (χSLea

O O

n NH O -O

O

-O

O

reagents and conditions: i) Et3N, DMF; ii) tyrosine O-sulfate; iii) 2-ethanolamine, rt

Figure 4.43 Bivalent interaction of P-selectin and synthetic polyvalent sLe a .

Ph

Ph n = 17

HO OSO3Na HN O O NaO3SO OH

75

O

n = 17 HO OH HN O O NaO3SO OH

O

76

Figure 4.44 Sulfated neoglycopolymer as P-selectin ligand.

208

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS P-selectin recognition units HO OH O OH

HO OH O OH HO

HO

-O2C HO OH O

AcHN

O

OH OH

-O2C HO OH

NHAc

O O OH HO O

O

O OR

Sialyl Lewis x

O O OH HO O

OH O OR

O

O O OH HO O

OH O OR

i) di-yne acid, sonication OH Ac O N S

ii) hν (254 nm), 0oC H N O O

HO OH O OH -O2C HO OH O

O HO O OH HO Lactoside

Sialyl Lewis x mimic (77)

HO OH O OH -O2C HO OH

HO OH HO

O O OH HO O

HO O

OH O

N Ac

S

N H O

78 (5 mol%)

HO

n

polymerized liposome

Figure 4.45 Polymerized glycoliposome presenting sialyl Le x -mimics.

[491]) reported the creation of a polymerized liposome presenting 5 mol% of a sialyl Lewis x mimetic, such as 77 and lactoside (Figure 4.45). A glycoliposome (78) potently inhibits P-selectin adhesion to cells, with an IC50 value of 2 nM, and its level of activity decreases by either increasing or decreasing the surface density of the carbohydrate mimetic. Another glycoliposome, 79, presents lactoside on its surface and is also active, with an IC50 value of 15 nM. However, its activity is sevenfold weaker than that of 78. Overall, these polymerized glycoliposomes are far more active than monovalent 77 or polymerized liposomes displaying no carbohydrate, which have IC50 values greater than 5 mM.


4.1.16

209

Lectins

Lectins are a family of proteins that recognize specific carbohydrate moieties attached to glycoproteins and glycolipids [328]. Being ubiquitous proteins found widely in plant, animal, and bacterial cells, lectins are involved in diverse cellular events, such as cellular recognition, cell–cell adhesion, bacterial colonization, and signal transduction [20,492]. An example of a well-characterized plant lectin includes concanavalin A (Con A), a mannose-binding protein with a molecular weight of 29 kDa. Con A monomers assemble into various highorder aggregates whose dominant species are controlled by pH [493]. For example, Con A exists predominantly as a tetramer at pH 7, with four mannose binding sites per tetramer. These binding sites are separated from each other by ˚ [494]. Similar oligomeric structures are also observed in soybean 65 to 70 A lectin isolated from Glycine max (SBA), which recognizes galactoside ligands and cross-links multiantennary oligosaccharides that terminate in galactoside residues [495]. Isolectins are composed of many di¤erent subunits. For example, Gri¤onia simplicifolia-1 (GS-1) is a tetrameric lectin comprising lectin monomers A and B. The A subunit binds specifically to a-d-GalNAc and the B subunit binds to a-d-Gal [496]. Through various combinations of the two subunits, the isolectins exist as a mixture of five di¤erent tetramers (A4 , A3 B, A2 B2 , AB3 , B4 ), each of which should display unique a‰nity to multivalent carbohydrates present on cell surfaces. 4.1.16.1 Oligovalent Mannosides Binding to Con A. To study the multivalency e¤ects from Con A, Dimick et al. and Corbell et al. [497] synthesized dendritic mannosides whose sugar ligands are displayed through a rigid aromatic sca¤old (Figure 4.46). A divalent ligand (80) inhibits Con A–mediated agglutination of red blood cells with an IC50 value of 380 mM, an activity only twofold greater than that of methyl a-mannoside. Increasing the ligand valency enhances the inhibitory activity, as illustrated by a tetravalent molecule (81) whose activity is 31 times higher than that of methyl a-mannoside. Such enhanced activity is not attributed to a favorable free energy of binding of 81 to Con A, since 81 shows a free energy of binding identical with that of methyl a-mannoside. Rather, the improved binding is driven by a large favorable entropy, with unfavorable enthalpy—conditions that are considered typical of aggregation. Therefore, as observed in an agglutination inhibition assay, the enhanced activity by a given tetravalent ligand is not necessarily related to increased a‰nity but to its ability to inhibit the formation of cross-linked lectins. 4.1.16.2 Trivalent Mannosides on a Rigid Sca¤old. A conformationally rigid sca¤old makes it possible to fix the distance between tethered ligands. A trivalent mannoside, 83 (Figure 4.46), is designed in such a sca¤old. The three mannose residues are appended to a well-defined macrocyclic lactone which ˚ of separation between adjacent mannosides in the extended conprovides 35 A

210


HO HO HO

HO HO

OH O

HO HO HO

OH OH O HO

NH

HO HO HO

HN

O

OH O

O

O

O

HO OH HO O HO

HO HO

O

O O

O

HN NH

O

NH

HN

O

N3 80

O

O

OH O O

N3 O 82

H N

OH OH

HN

O

HO HO HO

O

NH O

O

HO

81

O

O

O

OH OH

O OH

O O

HO HO HO

OH O O

O O

NH O

O

O H O

H

H

O

O

O O H

H O

83 HN

NH

O H O 35 Å O O O O

OH OH OH O OH

Figure 4.46 Structure of dendritic mannosides and a trivalent mannoside prepared on a macrocyclic sca¤old.

˚ sepaformation [498]. This interligand distance is smaller than the 65 to 70 A ration required for bridging mannose recognition sites within Con A tetramer. Hence, this trivalent molecule is likely to interact with multiple Con A tetramers rather than with a single tetramer. In an SPR study designed to measure the binding of Con A to a mannose-presenting glycolipid surface, a monovalent ligand (82) shows inhibitory activity, with a Ki value of 99 mM, a value close to that of methyl a-mannoside, which shows a Ki value of 65 mM. In contrast, the trivalent ligand 83 shows di¤erent modes of binding. At concentrations greater than 50 mM, it inhibits lectin binding; however, at concentrations less than 10 mM, it promotes lectin binding. Burke et al. [498] attributed this dual activity to the formation of soluble clusters between 83 and multiple Con A tetramers, a mechanism that allows some unoccupied saccharide sites of the lectin–


211

ligand complex to interact with the surface. According to other assays, 83 interacts with Con A at least 10 times more potently than does 82. 4.1.16.3 Bifunctional Multivalent Glycosides Binding to Con A. Fijimoto et al. [499] designed an octavalent glucoside, 84, built on a calix[4]arene scaffold to utilize multivalent interaction in host–guest complex formation. This calixarene-based glycoside acts as a bifunctional molecule: It presents a binding pocket for a guest molecule, such as fluorescent eosin Y, within a calixarene ring system, and it binds multiple glucoside ligands, which enable tight binding to Con A (Figure 4.47). For example, 84 binds to Con A immobilized on Sepharose gel with a K a value of 1 10 6 M 1 , a‰nity about 200-fold higher than that of methyl a-glucopyranoside. Using its hydrophobic cavity, 84 also makes a complex with eosin Y, with a K a value of 7.5 10 5 M 1 . When measured in the presence of eosin Y, 84 binds to Con A on gel with a slightly improved a‰nity, as demonstrated by the K a value of 2 10 6 M 1 . Such improved association is probably linked to the formation of a ternary complex, as shown in Figure 4.47. 4.1.16.4 Heptavalent Mannosides Linked to b-Cyclodextrin. In a separate report, Ortega-Caballero et al. [500] studied the inclusion ability and Con A binding of a dendritic heptavalent mannoside (85) designed with a b-cyclodextrin core (Figure 4.47). Using its cyclodextrin hydrophobic pocket, this multivalent ligand has the ability to form a complex with guest molecules, such as 2-naphthalenesulfonate, with a Ka value of 2.53 10 3 M 1 . In addition, it can inhibit Con A binding to mannan with an IC50 value of 67 nM in an enzyme-linked lectin assay (ELLA). The ability of this multivalent ligand to inhibit Con A binding is 17 times greater than that of methyl a-mannoside, a monovalent ligand. This and other examples [501–503] point to a potential application of bifunctional multivalent molecules as tools for the selective delivery of probes and drugs to clustered targets present on biological surfaces. 4.1.16.5 Mannosides Displayed on a Benzene- or Peptide-Based Dendrimer. Ashton et al. at UCLA [504] developed convergent synthetic routes that allow the synthesis of glycodendrimers based on a 1,3,5-trisubstituted benzenoid core that serves as a sca¤old for the display of 9 to 36 copies of mannoside ligands (Figure 4.48). An example of the dendritic mannosides is the 18-mer 86. These dendrimers are active in inhibiting the binding of Con A to yeast mannan in ELLA. Optimal potency is observed from 9- and 18-mer forms of 86, which are four times more active than methyl a-mannoside or a dendritic wedge. A second example of a mannoside-presenting dendrimer includes the 16-mer 87 (Figure 4.49), as reported by Page´ et al. [505] (see also Woller and Cloninger [506]). This dendritic ligand displays 16 mannosides appended to multiple branches of an oligolysine peptide. This 16-mer is identified as a potent inhibitor of Con A–mannan binding, with an IC50 value of 1.6 mM, representing a 578-fold (66-fold per mannoside) enhancement in a‰nity relative to methyl amannoside.

212

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS Con A tetramer Eosin Y

84 Glucoside

Host-guest complex Ka = 7.5 x 105 M-1

HO HO

84 R = (CH2)10CH3

HO HO

HO HO

OH O

OH O

O OH HO

OH O

O OH HO

HO HO

OH OH HO

O NH

O

OH OH H N HO O OH O

O OH HO

85

HO O

R

R

R

R

O

O

O O

HO

O

O OH OH HO N O HHO O HO HO

O

O

OH OH

OH OH

O OH

S

O OH HO

O HO

O HO

OH O OH O OH S O OH

HO HO HO

OH OH

OH OH OH

OH O

O OH OH

OH OH

O OH OH HO N O HHO O HO HO O OH OH HO NH O HO O HO HO

O

OH OHHN

S OH O S

O

O

HOHO HO

HO HO HO

O OH OH HO NHHO O O HO HO

OH OH H N HO O

O OH HO

O S O HO

HO OH

S O

HO

OOH

OH

HO O HO OH O O

S O HO

OH OH OH

OH OHOH

Figure 4.47 Formation of ternary complex from Con A, bifunctional glucoside, and a guest molecule; and the structure of dendritic glycoside built on calix[4]arene or bcyclodextrin sca¤old.

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES AcO AcO AcO AcO AcO AcO

OAc O O

OAc O

AcO AcO OO AcO AcO O

HO O O

NH

N H

O

O N

AcOO AcO AcO O AcO AcO O O AcO AcOOAcAcO O

i,ii

O

NH2

+ HN O HN

O

OH O

NH

O

HO

O HO HO OH

AcO AcOOAc HO HO HO HO OH HO HO HO OH OH OH O O O O O OH O

O

HN

HOOH OH OOH O O OHHO OH OH O O OH O

HO HO HO

HN OH O

O

N HN

HO HO O O HO HO O HO O HO O HO OH

O

N

O O

NH

N H

O

OH O O OH HO OH O OH HO OH OH

O HN

86

O

OH OH O OH O OH

O

NH O

H N O

O

H N

O

NH

O O

HO O HO HO HO

O

NH

N

O

NH O

HO O

HO HOOH

O HO O

HO HO OH

O O HO

O O HO

OH HO OH

OO HO

OH OH OH

OH OHOH

reagents and conditions: i) DCC, HOBt, CH2Cl2, DMF; ii) NaOMe, MeOH.

Figure 4.48 Synthesis of mannopyranoside-containing dendrimer.

213

214

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS HO HO OH OH

HO HO HO OH

HO HO

O

O

OH OH

HOOHOH HN

O O

HN

S

O NH

O

N H

O

NH HN

O

O

HN

HN HO

HO HOOH

O O

S

O

O

NH

O

O

H N

HN NH

S

O

O

O NH HO

O

HN

H N

O N O H

O

N H

O

O

S

HO

O NH

NH

NH

O HO OH HO OH

O HO

OH OH OH OH

H N

S

HN O NH O S O HN

O

O

O

O

HO OH OH

H N

S

O O

O

S

O

OH

N H

HN

HN O

NH

HN

O

O

N NHH NH

O

HOOH OH

OH

O

O

O HO HO HOOH

O

NH

O

O

NH

O

NH

HN S

H N

H NH N O

S

NH

NH

N NHH

O O

O O

O

HN

S

O

NH

O

O

H O N N H O

S

N H

O

NH

OH O

NH

O

O

NH

NH H N

HN

O

O

N H

HOOH OH OH

O

S

O HN

HN O NH

O

O

NH

NH O S O

O

HO HO HO

HN O

O

OH

OH

O

S

HN O S

HO HO HO

O

O HO

OH OH OH

OH HO OH 87

Figure 4.49 Structure of lysine-based dendrimer presenting mannoside.

OH OH OH


215

Con A tetramer

~65 Å

glycoside ligand Neoglycopolymer 89

carbohydrate recognition domain

TESO O TESO TESO TESO

+ OH

H O

O

O

H O

i.ii

H O

HO OH HO

O

O HO

H O

H O

n OH

O O

OH

OH O

OH

88

O

HO OH HO

OH OH

OH O

iii

H O

O O

HO

O

OH

89 O

O H O HO OH HO

OHX O

H O

n OH

O O X HO

90 (x = OCH2)

OH O

OH

H O HO OH HO

H O

n OH

O O X

X

HO O

OH O

OH

OH 91 (x = CH2) 92 (x = OCH2)

reagents and conditions: i) N-methyl 2-chloropyridinium iodide, DMAP, Pr3N, CH2Cl2; ii) HF, pyridine; iii) RuCl3, H2O, 55oC.

Figure 4.50 Multivalent neoglycopolymer binding to Con A tetramer and structure of neoglycopolymer presenting carbohydrate ligand to Con A.

4.1.16.6 Poly(oxanorbornene) Presenting Glucosides. Mortell et al. at the University of Wisconsin [507] reported the synthesis of polymeric, polyvalent glucosides derived from a 7-oxanorbornene monomer. Theirs syntheses are based on the polymerization of a 7-oxanorbornene glucoside catalyzed by a ruthenium metal complex in aqueous solution (Figure 4.50). A neoglycopolymer (89) presents multiple glucosides attached to a polymer skeleton in the aconfiguration preferred for Con A–binding. The C-glycosidic linkage o¤ers resistance against the chemical and biological hydrolysis that occurs with natural O-linkages. This neoglycopolymer strongly inhibits Con A–mediated ag-

216


glutination of erythrocytes, with an IC50 value of 25 mM, a 2000-fold activity enhancement relative to methyl a-d-glucoside or 88. However, a second neoglycopolymer (90), whose glycoside ligand is attached through an O-linkage with three atoms of distance, shows only weak activity, with an IC50 value of 2 mM (Figure 4.50) [508]. The inhibitory activities of these polyvalent neoglycopolymers are controlled by a combination of several molecular parameters, such as the intrinsic a‰nity of the monovalent ligand, the nature of the polymer–ligand linkage, and the length of the polymer–ligand spacer. For instance, the equivalent polymer possessing fucoside as the side chain is inactive. Additionally, the neoglycopolymer 91 presenting a-mannoside—a better ligand to Con A than a-glucoside—is highly active in the C-linked form, with an IC50 value of 0.5 mM, while the corresponding O-linked mannoside-presenting polymer (92) has an IC50 value of only 13 mM. Provided that tethered epitopes are able to span targeted carbohydrate recognition sites in the Con A tetramer, the size and ligand density of a glycopolymer comprise the major parameters that govern multivalent interactions. Kanai et al. [509], Strong and Kiessling [510], and Mann et al. [511] studied the e¤ect of the length of mannoside-presenting glycopolymers on the ability of Con A to induce hemagglutination (Figure 4.51). Glycopolymers (94) with various mannoside valencies (wmann ¼ 100%; valency ¼ n ¼ 10 to 143) are prepared under living ROMP conditions. This method allows the generation of

Ph H O

H O

Ph

PCy3 Cl Ru Cl PCyPh

N

H O

H O

CH n 3

H O N

ii

i

OH O

H O

N

3

O

n

O

OH

OH OH 93

O

OH O

OH O

OH

OH

OH OH

OH OH

94 (n = 10, 25, 52, 143)

95 (n = 10, 25, 52, 143)

Ph H O

H O

H O

N

H O N

+ O

OH O

OH OH

OH

i

O HO

PCy3 Cl Ru Cl PCyPh 3

OH O

HO

n H H O O

H O

H O

m

N

N O

OH O OHHO O HO OH OH

OH

O

OH

OH

96 reagents and conditions: i) H2O, ClCH2CH2Cl, DTAB, rt; ii) TsNHNH2, H2O, 100oC.

Figure 4.51 Synthesis of mannoside-presenting neoglycopolymers of varying length and density.


217

monodisperse polymers of defined length by varying the ratio of monomer to ruthenium catalyst. The inhibitory activity of 94 in Con A–induced hemagglutination increases exponentiallyin relation to the polymer length and the mannoside valency. For example, as the polymer size increases, the potency of 94 relative to monomer 93 is enhanced 100-fold (n ¼ 10), 300-fold (n ¼ 25), 2000-fold (n ¼ 52), and 2500-fold n ¼ 143). A neoglycopolymer (95) whose polymer backbone is made of a saturated chain, and thus is more flexible than the unsaturated backbone in 94, shows a similar trend in the relationship between activity and length. Such an exponential correlation is linked with multivalent binding, where the variation of length of the polymer makes it possible to achieve improved potency. 4.1.16.7 Poly(norbornene) Derivatives Bearing Two Types of Ligands. In a continuing study, Cairo et al. [512] reported a neoglycopolymer (96) prepared by copolymerization of a mixture of two di¤erent glycomonomers (Figure 4.51). This glycopolymer presents both mannoside and galactoside ligands as side chains. The role of the galactoside unit is to control spatial separation between neighboring mannoside ligands. This copolymer promotes the formation of Con A clusters through multivalent associations, as supported by data from a quantitative precipitation assay. The number of Con A tetramers bound per polymer molecule increases with a higher density of mannoside epitopes. For example, 96 (n ¼ 143, m ¼ 0)—a neoglycopolymer whose side chains are composed entirely of mannoside—binds 16 Con A tetramers. However, 96 (n þ m ¼ 116) has a mannoside density of 10% and binds three Con A tetramers. In terms of cluster stoichiometry calculated on the basis of mannoside available per Con A, a glycopolymer (96) with a lower epitope density binds more e‰ciently, by using fewer epitopes, than the equivalent glycopolymer with higher density. This concept is illustrated by calculating the number of mannosides bound per Con A tetramer: The molecule with 100% mannoside density binds nine mannosides per Con A, while the molecule with 10% mannoside density binds only two mannosides per Con A. The clusters of Con A bound by the mannoside-containing glycopolymers can be visualized by transmission electron microscopy (TEM) using biotin-labeled Con A and streptavidin-presenting gold nanoparticles, where the glycopolymer promotes the aggregation of gold particles in dimers and trimers [513]. 4.1.16.8 Glycopolymer-Mediated Cell–Cell Adhesion. Through its multiple sugar-binding units, Con A promotes the aggregation of mammalian cells that display multiple copies of diverse glycans terminating in mannoside (Figure 4.52). Such cell aggregation is also possible in the presence of lectin clusters induced by a neoglycopolymer. Gestwicki et al. [514] demonstrated that a mannoside-presenting glycopolymer, 97, induced Con A clustering with a stoichiometry of four to six Con tetramers per polymer. Thus, upon incubation with Con A and Jurkat cells, this glycopolymer mediates cell–cell aggregation more e¤ectively, with a 60% enhancement over that of Con A alone. The gly-

218


cell Neoglycopolymer 97

Glycan Con A tetramer

cell

Ph HO HO HO

OH O O

H O HN

ROMP HO HO HO

H O

OH O

n 97 (n = 142)

HN O

Figure 4.52 Lectin-mediated cell–cell aggregation mediated by neoglycopolymer.

copolymer itself induces such cell aggregation poorly, through a mechanism that involves the mannose-binding receptor. 4.1.16.9 Mannosides Displayed on Self-Assembled Monolayers. Presentation of carbohydrate arrays on self-assembled monolayers (SAMs) on a gold surface serves as a convenient system for mimicking natural cell surfaces that display diverse arrays of complex carbohydrates. Houseman and Mrksich [515] reported a versatile chemical strategy based on specific chemical reactions occurring on the surfaces of SAMs. This approach utilizes a mixed SAM that presents multiple terminal hydroquinones on the background of a SAM containing penta(ethylene glycol). This biologically inert surface inhibits nonspecific protein adsorption (Figure 4.53). The next steps include the oxidation of hydroquinone to quinone and reaction with a cyclopentadiene–mannoside conjugate via a Diels–Alder reaction, which leads to a mannoside-presenting SAM on gold. Mannose ligands on the surface of this chip bind Con A with an IC50 value of 55 mM when assessed by SPR. This strategy may be used to introduce other carbohydrates to the SAMs and to perform enzymatic modification of the exposed carbohydrates. 4.1.16.10 Multivalent Galactosides Binding to Bauhinia purpurea Lectin. TentaGel is a commercial resin made of ethylene glycol–grafted polystyrene. The gel is commonly used as a solid support in the solid-phase synthesis of oligosaccharides. Liang et al. [516] reported that this resin is useful for synthesizing a multivalent carbohydrate library and for on-bead screening where sugars are displayed on the surface of resin in multiple copies. One such example is shown in Figure 4.54, where disaccharide Gal-b1,3-GalNAc-b-SPh residues are synthesized on the surface of TentaGel beads. The disaccharide-displaying resin

219

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES HO HO HO

OH

H N

O O

O O O

HO

O

O OH

HO

HO O

O O

O

O O

S

O

O

O

O O

O O

S

S

O

i oxidation

S

S

O

O O

O O

S

O

O O ii Diels-Alder reaction

S

S

O

O

O O

O

O O

OH O

O

O O

S

HO O

O

O O

HO

O

O

O O

O

O O

OH O

O

O O

S

HO

HO

OH O

O

O

O

O O

O O

O O

O O

O O

S

S

98 reagents and conditions: i) excess 1,4-benzoquinone, H 2O; ii) mannoside-cyclopentadiene conjugate, H2O.

Figure 4.53 Preparation of carbohydrate chip based on self-assembled monolayer.

(99) binds Arachis hypogaea lectin, which recognizes Gal-b1,3-GalNAc as a specific ligand and induces aggregation as a result of multivalent lectin–ligand recognition. The sugar-coated resin also shows a‰nity to Bauhinia purpurea lectin, which recognizes Gal-b1,3-GalNAc-b-SPh. Such recognition is selective to Gal-b1,3-GalNAc-b-SPh present on the bead and thus does not occur with beads that contain altered disaccharides such as Gal-b1,3-GlcNAc-b-SPh and molecules with altered stereochemistry, such as b to a, at the anomeric GalNAc linkage. The avidity between the B. purpurea lectin and Gal-b1,3-GalNAc-bSPh presented on beads is influenced by the surface density of the disaccharide. In a related study, Horan et al. [517] addressed the e¤ect of surface density by using SAMs on gold (100) with SPR analysis. The a‰nity of lectin to sugardisplaying SAMs varies with epitope density, and tighter binding is observed at a low density (wsugar ¼ 0.1) rather than a high density (wsugar ¼ 0.6). This result may point to steric congestion, which is likely to occur at higher surface density. 4.1.16.11 Dendritic Sialosides Binding to Limax flavus Lectin. The L. flavus lectin binds to the human a1 -acid glycoprotein (orosomucoid), which presents sialic acid–terminated carbohydrates as the epitope required for such interac-

220


Ph

Ph

O O O

AcO

S

N3

OH

O

PivO

Ph OPiv O

TentaGel

OPiv

S O

HO

OH O OH

HO O

OH

HO O

n

OH O

S NHAc

H N

O

n

O

HO

HO

iv-vi

H N

N3

OH O

iii

H N O

HO O

S O

O O O

O N3

H2 N

O

PivO

HO

i,ii

+

O O

99

O

OH O

n

Au

S NHAc

O

H N

O

N H

O

O

O 6

S

100

reagents and conditions: i) HOBt, HBTU, DIEA, NMP, rt, 12 h; ii) H2NNH2, DMF, rt, 9 h; iii) Gal-(OPiv)4-1-SOPh, (CF3SO2)2O, DTBMP, CH2Cl2, -60oC to 0oC; iv) HSCH2CO2H, rt; v) 20% TFA in CH2Cl2, rt; vi) LiOH, 80% MeOH, THF, rt.

Figure 4.54 Synthesis of multivalent Gal-b1,3-GalNAc-b-SPh presented on either TentaGel resin or a self-assembled monolayer on gold.

tion. Zanini and Roy at the University of Ottawa [518] reported the synthesis of a glycodendrimeric construct that displays multiple copies of a-thiosialic acid tethered to a dendrimer sca¤olding made of a 3,3 0 -iminobis(propylamine) core (Figure 4.55). These glycodendrimers include tetra- to dodecavalent sialosides. A dodecamer (101) is the most e¤ective molecule in inhibiting the binding of L. flavus lectin to orosomucoid, with an IC50 value of 8.2 nM (98.6 nM per sialoside residue). It is 182-fold more potent than monovalent sialic acid, NeuAc-a-N3 . A second type of sialodendrimer is the spherical dendrimer designed on the Starburst PAMAM sca¤old, as exemplified by 102 (Figure 4.56). With an IC50 value of 1.13 nM (36.2 nM per sialoside), this 32-mer shows better inhibitory activity than the 12-mer (101) [519]. Dendrimers analagous to 102, with valencies lower than 32, are less active by one order of magnitude: 16-mer > 8-mer > 4-mer. Other glycodendrimers include starburst dendrimers that present lactosides or galactosides. Their ability to inhibit the activity of plant lectins is a¤ected by the size or generation of the dendrimer and the binding-site orientation of the lectin [407,519]. 4.1.16.12 Dendritic GlcNAc Binding to Wheat Germ Agglutinin. Zanini and Roy, and Baek and Roy [519], reported the synthesis of oligolysine-based dendrimers that display multiple GlcNAc residues (103 and 104). These dendrimers are prepared by a solid-phase synthetic method, using Wang resin (Figure 4.57). The GlcNAc dendrimers show a‰nity to wheat germ agglutinin

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES CbzHN H N

N CbzHN

NH2 O

O H N

+

OH

i-v

N

H 2N

NH2

N

CbzHN

N

O

CbzHN AcHN OH HO OH OH OH AcHN HO O HO CO2HO O CO2 S S O HO OH AcHN CO2HO O HO S HO

OH

CO2-

O H N

HO NHAc OH HO OH O

O NH NH N O NH

NH

O S AcHN O OH OH

N

N O

NH

CO2- HO NHAc HO OH O NH OH O S CO2NH S

O

HN

O

N

N H

O

O N

N

N H

NH

HN N -O

HO HO AcHN

2C

O

O

N

HN

S O HN O O OH -O2C S OH O HO OH HO OH NHAc

O

O

HN

N H

N

O

N HN

N H N H

S

OH NHAc

OH CO2- OH

O S

-O

HO O

2C

OH OH NHAc HO OH O

HN O O S

S O2C O OH O OH OH NHAc HO OH HO OH HO NHAc

-O

2C

101

reagents and conditions: i) DIC, HOBt, DIPEA; ii) H2, 10% Pd/C, MeOH; iii) (ClCH2CO)2O; iv) methyl 1-α-SH-sialic acid (OAc)4, Et3N, DMSO; v) NaOMe, MeOH, rt; then 0.05 M NaOH, rt.

Figure 4.55 Synthesis of tethered a-thiosialic acid–presenting dendrimer.

221

222


AcHN OH AcHN OH HO OH OH HO OH HO O AcHN OHHO O HO CO2CO2HO S AcHN OHHO O CO2- S HO - S NH HO OH HO O CO2 NH S AcHN S NH S NH NH HO O CO2 NH S NH HO OHHO HN HN S NH S O O AcHN HN O NH CO2HO O O HN N S NH HO S NH HO OH O N NH HN NH S AcHN CO2O NH O HN O HO NH N S HO O NH O NH S N HO OH HN H H CO2O N N AcHN H O O HN O HO S N N N HO S O N H O N N H NH O O HO OH

CO2-

O S AcHN OH OH

H H N N S H H N N

O N HO N H O

N

O

H N O

N

N

NHAc OH OHHO NHAc OH O OH HO NHAc HO CO2- O OH S -HO OH S CO2 O HO NHAc S CO2HO OH O OH S NH NH S CO NH S HO 2 NHAc NH S NH HO OH HN O NH S OH S NH OHN CO2- HO O HN NH HO NHAc N OHN S O OH NH S N O NH - OH CO2 HN O NH S O NH HO HO N O H NHAc N H N N S O N H O OH S O CO2- OH H NH N N N HN HO HO H N O H O N OH S O NHAc S HN N N -O C N N N OH 2 H O OH O H H HO

HO

HN N

O N

NH

O HN

N

H N

O

H OH S N S OH O N N O NHAc -O C H 2 H NH N S HO OH HN O HN O OH S N HN OH O NH O HN O O2C NHAc S HN HN NH N O HO OH HN S OH N HN S O NH O OH O2C O NH HN NHAc O O S S HN NH NH HO OH OH HN S HN O2 C O OH HN HN HN S S NHAc S HO OH HN -O C HN OH O 2 S OH S O2C O OHHO NHAc S HO O2C O2C OH O OH HO O OH HO NHAc OH HO HO OH HO NHAcHO NHAc

NH O

O O NH

NH

HN HO -O2C S HO O H N N H N O S N AcHN O N H HN O OH OH S NH -O C HN O 2 HN O HO N S HO HN O S NHO N AcHN OH OH -O C HN NH O 2 O HN S HO NH S HN O NH HO HN S HN AcHN OHOHO2C S HN S HN HO O HN S -O C HO OH S 2 AcHN OH O -O C S HO 2 OH S HO O O 2C OH HO OH AcHN O OH HO OHHO NHAcHO OH NHAc

N

N O H

N

102

Figure 4.56 acid.

Structure of 32-valent PAMAM-based dendrimer presenting a-thiosialic

(WGA), which recognizes the GlcNAc ligand. The dendrimers inhibit WGA binding to GlcNAc-containing porcine mucin, with IC50 values of 0.51 mM (103) and 0.088 mM (104). Such activity by GlcNAc dendrimers represents at least a 25- and a 170-fold enhancement, respectively, relative to monovalent allyl a-d-GlcNAc. 4.1.16.13 A Cyclodextrin-Based Multivalent Glucosides Binding to Pisum sativum Lectin. Garcıá-Lo´pez et al. [520] reported the synthesis of b-cyclodextrin–based heptavalent glycosides, which are prepared by per-glucosylating the primary hydroxymethyl groups of b-cyclodextrin (Figure 4.58). The biological activity of these cyclodextrin-based glycosides is demonstrated by

223


O H O O S N N N NHAc H O H O HO H O O HO N S HO N N NHAc H O H HO HO HO HO HO HO

HO HO HO

O S NHAc

O S NHAc

HO HO HO

O S NHAc

O S NHAc

O N H

O

NH

HN O

O

NH

O

HN

O O

HN O

O

N O NH H

O N H

O OH

H N

NH

O

N NH H

104

Figure 4.57 drimer.

O

O N H

O

O

103

OH

H N

H N O

O

N NH H O

O

O

O

HN

O

O

H N

H N

O N H

N H

N H

O

H N

O

O H O O S N N N NHAc H O H O HO H O O S HO N HO N N NHAc H O H HO HO HO

O

H N

H N

N H

O O

H N O

H N

N H

O

AcHN S O

AcHN S O

H N

N H

OH OH OH

OH OH OH AcHN S O

O O H N

O

OH OH OH AcHN OH S O OH OH

Structure of tetra- and octavalent GlcNAc presented on oligolysine den-

inhibiting P. sativum lectin (pea lectin), which recognizes d-glucoside and dmannoside as ligands. For example, 105 shows an IC50 value of 90 mM, an activity about 105 times greater than that of methyl b-d-glucoside, or 15 times greater per glucoside. The second molecule (106), composed of glucoside tethered to cyclodextrin, with a shorter spacer, shows poor activity, with an IC50 value of 870 mM. Synthesis of a heptavalent sialoside displayed on the rim of bcyclodextrin is also reported [521]. 4.1.16.14 Oligovalent Mannosides Binding to Dioclea grandiflora Lectin (DGL). A lectin from the D. grandiflora seed recognizes a trimannoside epitope presented through an asparagine linkage to a glycoprotein. Brewer et al. at Albert Einstein College of Medicine studied the thermodynamic basis of DGL binding to oligovalent trimannoside by using microcalorimetry [2,41]. The divalent (108) and tetravalent (109) molecules present multiple trimannosides attached to an aromatic core (Figure 4.59). Because of the relatively short distance between neighboring trimannosides, individual epitopes are unlikely to bind simultaneously to multiple CRD sites present on a lectin oligomer. A reference monomer (107) binds DGL in 1:1 stoichiometry, with a K a value of 39 10 4 M 1 . Increasing the ligand valency to two (108) and four (109) leads

224

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS HOOHOH

HO OH HO HO

HO O

O S

S O

HO HO HO

O

H O N

S

NH

O O

NH

O OHHO O OH

O HO

OH

O HO

O

HO

O

HO

OH O 105 HO OH O HO O OH O OH OHHO HN O O HN O O S HN O

O

S

N O H

O OH

S O

S

OH HO HO OH

O

O

OH OH OH

OH OH OH

OH

OH OHOH HOOHOH

HO OH HO HO

HO

O

O S

O OH HO HO HO

O

O S

OH

O

S O OHHO

O HO

O HO

HO

OH O 106 HO O OH O HO O OH HO O OH OH S O O S O O O HO OH S OH HO O OH

S

OH OH OH OH O

OH OH OH OH

OH OHOH

Figure 4.58 Structure of dendritic glucoside based on b-cyclodextrin core.

to 15- and 166-fold increases in binding a‰nity, respectively, with respective K a values of 6 10 6 M 1 and 6.5 10 7 M 1 . Based on thermodynamic parameters, the enhanced binding results mostly from favorable binding enthalpy, such that the DH value observed per multivalent molecule is very close to the sum of DH values contributed from individual epitopes from a multivalent ligand. Unlike binding enthalpy, binding entropy does not contribute favorably to the multivalent association. Therefore, it is likely that the multivalent ligand binds multiple independent CRD sites from separate lectin molecules.

CARBOHYDRATE-RECOGNITION RECEPTORS ON CELL SURFACES HO HO HO HO HO HO

HO HO HO

OH O O HOO

HO HO HO

OH O

HO HO HO

OMe

O OH

HO HO HO

107 HO HO HO

HO HO HO HO HO HO

HO HO HO

OH O O HO O

OH O O

O OH OH O O HO O

O OH OH O

NH S

O OH

CO2H

S

OH O

NH NH O

O

O OH

108 OH O O

O

O NH HN

NH

S

S HN

NH

HN

NH O

O HO

S

S OH O

O

HN CONH- (CH2)6 -HNOC

O HO O

O NH

OH O O HOO

225

O

O

O

109 O HO

OH O OOH O

OH OH OH

OH OH O HO OH OH O OH OH OH OOH O OH OH O HO OH

Figure 4.59 Structure of multivalent construct presenting trimannoside.

4.1.16.15 Polymeric Polyvalent Lactosides Binding to Psophocarpus tetragonolobus Lectin. A biotin end-labeled multivalent glycopolymer serves as a tool in the microfabrication and engineering of biosurfaces, especially when designing a device for detecting lectins and antibodies. Sun et al. [522] reported the synthesis of a biotin-terminated poly(acrylamide) with lactoside side chains (110, Figure 4.60). This glycopolymer is selectively immobilized to the portion of the surface that is microfabricated with streptavidin, and the resulting glycopolymer surface recognizes in turn the galactoside-binding lectin of P. tetragonolobus. The lectin adsorption occurs selectively on the region microfabricated with glycopolymer. 4.1.16.16 Resin-Displayed Multivalent Galactosides Binding to RCA Lectin. Micrometer-sized beads that present specific carbohydrates on their surfaces are used in flow cytometry to screen sugar receptors. Vetter et al. [523] reported this method, which is based on synthetic sugar-coated TentaGel beads approximately 10 mm in diameter, which present multiple copies of lactoside (111) (Figure 4.61). The glycoconjugated beads bind the fluorescence-labeled Ricinus communis agglutinin (RCA120), a lectin that presents on its B-subunit three galactoside-binding sites [524]. Fluorescence-activated particle sorting makes it possible to detect the complex of fluorescence-labeled RCA120 with the glyco-

226

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS O HN

O NH

NH2

H N

S

HN

NH

i

N2+

H N

S

O

ii

BF4-

O

O HN

NH

•OCN

H N

S

+

OH HO HO O HO O HO HO

O

O O OH

O

iii

N H

radical initiator O HN 110 (χLac = 10/70 per polymer)

NH H N

S

O O O O O O O NH2 NH2 NH NH2 NH2 NH2 NH

O

HO HO HO

HO O OOH OH

O

O

OH HO HO HO

HO O OOH

O

n

O OH

OH

reagents and conditions: i) NaNO2, HBF4, H2O, THF; ii) NaOCN; iii) acrylamide, H2O, THF, 50oC.

Figure 4.60 Synthesis of polyvalent polymeric lactoside terminated with biotin.

conjugated beads and to isolate the bound beads from the unbound. The lectin–sugar recognition is ligand selective, such that beads presenting nonlactosides, such as sialyllactose and LacNAc, do not associate with the RCA lectin. 4.1.16.17 Liposome-Based Multivalent Galactosides Binding to RCA Lectin. Liposomes are commonly used to present multiple copies of a sugar ligand on a cell-like surface. However, the nature of ligand presentation, either as a cluster or as a simple sugar monomer, influences the biological properties displayed by


iii

O OH OH

i,ii



H O N OH O

H O N OH O

O

O O

N O

O N H n

TentaGel

111 reagents and conditions: i) aq. NH4HCO3, 5 d; ii) disuccinimidyl suberate, HOBt, DMSO; iii) H2N-TentaGel, DMF.

Figure 4.61 Synthesis of lactoside-presenting TentaGel resin.


O

O

O O

CH2

O

HO

C18H37

+

CH3

H m

m = 3, 12

O O O

CH3

CH3

CH3

N C18H37

CH3 CN

CN N

CN

C18H37 N C18H37

N

i

O

O N

OO

O O

OO

O

O

O

O OH

O OH

O OH

O

OH HO HO HO HO HO OHHOHO HO OH OHOH

227

CH3

CH3

CH3

112 (degree of polymerization = n = 6, 15)

Galactoside

ii liposomal membrane

reagents and conditions: i) THF, 70oC; ii) DMPC, sonication.

Figure 4.62 Preparation of liposome presenting oligogalactoside cluster.

liposomes. Liposomes that present sugar clusters more closely mimic a natural cell surface because carbohydrates present on cell surfaces are expressed as clusters. Kitano et al. [525] studied such multivalent liposomes with respect to their ability to bind to the RCA120 lectin. For example, sonication of 1,2dimyristoylphosphatidylcholine (DMPC) lipid with an amphiphile terminated with a block of multiple galactosides (112) induces the formation of liposomes displaying multiple sugar clusters (Figure 4.62). Upon addition of RCA, the solution of liposomes composed of 112 becomes turbid, due to RCA-mediated liposome–liposome cross-linking. The turbidity change increases in exponential fashion upon an increase in galactoside-terminated lipid density on the liposome surface. It also depends on the degree of polymerization of 112. At a low molar ratio of 112 to DMPC, the liposomes show a more rapid turbidity change, with higher degrees of galactoside clusters, on the order of m ¼ 12 > 3. At a high ratio, the liposomes show a gradual decrease in turbidity change. Thus, steric congestion occurs on liposome surfaces that are overloaded with galactosides, which blocks the binding of lectin and cross-linking. 4.1.16.18 Multivalent Display of Saccharides on Microtiter Well Surfaces. Diverse sets of specific carbohydrates can be displayed on the surface of 96-well microtiter plates to enable screening for corresponding carbohydrate receptors. Bryan et al. [526] developed one such method, based on hydrophobic interaction, that allows noncovalent anchorage of simple carbohydrates to microtiter

228

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH HO O HO HO

i O

OH HO O HO HO

ii

O

HO HO HO HO OH HO OHHO OH HO OH HO O HO HO

O HO

O HO

O

O

O

O

iii

O

(CH2)8

(CH2)8

(CH2)8

113

surface of microtiter plate reagents and conditions: i) CH2=CH(CH2)10CH3, Cl2Ru(PCy3)2=CHPh; ii) H2, 10% Pd/C; iii) loading on microtiter plate, washing.

Figure 4.63 Multivalent display of saccharides on a microtiter plate.

surfaces. For example, a galactosyl lipid (113), which is synthesized via olefin metathesis, is adsorbed to microtiter surfaces and displayed in a multivalent fashion (Figure 4.63). This glycosurface recognizes the ricin B chain in ELISA. Fazio et al. explored this display strategy further to optimize the presentation of diverse oligosaccharide arrays on microtiter plates [527]. A second example is based on a copper(I)-catalyzed 1,3-dipolar cycloaddition occurring at the interface of a sugar-containing alkyl azide in solution, and alkyne lipid adsorbed on the surfaces of microtiter wells (Figure 4.64). For example, an azide tethered to a trisaccharide (114) reacts with an -alkyne lipid presented on the surface. This reaction leads to regioselective formation of a 1,4-triazole, and as a result, covalent attachment of the trisaccharide sugar to the well surface. The linked carbohydrate is converted further to a sialyl Lewis x residue, through enzymatic fucosylation. This approach makes it possible to display 11 di¤erent sugars, ranging from mono- to tetrasaccharides on microtiter plates.

4.2 PEPTIDE AND HORMONE RECOGNITION RECEPTORS ON CELL SURFACES 4.2.1

CD14 Glycoprotein

4.2.1.1 Poly(acrylamide) Displaying Muramyl Dipeptides. CD14 is a surface receptor present on monocytes and macrophages that induces the production of proinflammatory mediators such as TNF-a and IL-1 [528]. It is anchored to the cell membrane through glycosylphosphatidylinositol (GPI). Exposure of CD14 to bacterial cell wall components initiates the proinflammatory response,

229

PEPTIDE AND HORMONE RECOGNITION RECEPTORS Sialyl Lewis x AcHN HO

OH OH

HO O

CO2O

HO 114

AcHN HO

HO O

OH OH O

AcHN HO

HO

- + N N N

O

NHAc O

1,3-dipolar cycloaddition N

i

O

OH OH OH OH O O HO OH OO HO

O OH HO

O N N

O

NHAc O

enzymatic fucosylation N

ii

N N

O

O NH

CO2O

OH OH O

NHAc

OH OH

HO O

CO2O

O OH HO O

OH OH

NH

NH

microtiter well reagents and conditions: i) CuI (5 eq) , DIPEA (5 eq), MeOH; ii) GDT-fucose, fucosyltransferase, 37 oC, 1 h.

Figure 4.64 1,3-Dipolar cycloaddition-based multivalent display of sialyl Le x on the surface of a microtiter plate well.

which invokes liposaccharides (LPSs), peptidoglycan, and lipoteichoic acid. Siriwardena et al. [528] elucidated the mechanism involved in such CD14 binding and the structural features of the CD14 ligands by using neoglycopeptide polymers composed of multiple cross-linked muramyl dipeptides. These polymers mimic the bacterial peptidoglycan layer. The synthetic polymer is made of a poly(acrylamide) backbone that presents pendant muramyl dipeptides on the side chains at varying densities (Figure 4.65). When allowed to interact with monocytes, the polymer (115) induces the production of TNF-a in a dose-dependent manner, while the monovalent muramyl dipeptide or underivatized poly(acrylamide) fails to induce TNF-a production. TNF-a production is inhibited by an anti-CD14 monoclonal antibody. 4.2.2

Major Histocompatibility Complex

4.2.2.1 MHC Peptides Tethered to a Peptide Sca¤old. MHC proteins are expressed as heterodimeric glycoproteins on the surface of antigen-presenting cells (APCs). An integral part of the immune system, the MHC complex displays peptides derived from foreign antigens in an array of MHC–peptide

230

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS NH2

HO O

OH O O NHAc O H O N N H O

S

n O O OR OR O

O +

NH2

R

O

O N

Et3N, DMF; then NH4OH

CO2H

n O O O NH2 NH NH2

HO O

OH O O NHAc O H O N N H O

S 115 (χpeptide = 4, 8%) NH2 CO2H

Figure 4.65 Synthesis of petidoglycan-derived neoglycopolymer.

complexes on the APC surface [529]. The MHC–peptide complexes are recognized by antigen-specific T-cells in a process that initiates T-cell activation. APC-T-cell recognition occurs through multivalent interactions between an array of T-cell receptors and an array of antigenic peptides presented in the context of the MHC (Figure 4.66) [530]. In a study involving oligomerization of MHC proteins by biomimetics, Cochran and Stern [531] provide insights into the molecular basis of T-cell receptor binding and activation by multivalent MHC–peptide complexes. Multiple copies of the MHC HLA-DR1 protein were attached to a peptide framework, providing dimers (116), trimers, and Cell

MHC Antigenic peptide

T-cell receptor

T Cell

multivalent MHC

T Cell

Figure 4.66 Multivalent association occurring between arrays of MHC and T-cell receptors and that for artificial multivalent MHC.

231

PEPTIDE AND HORMONE RECOGNITION RECEPTORS 45 Å MHC peptide (HLA- DR1) S O N

HO

O

O

O

S O

O O

O

NH

N

NH

CO2H N

HN S

H O H O H O H O N N N N N N N NH2 H O H H O OH O OH O CO2 116 (divalent)

50 Å

S

S O

HO

O

O CO2 H N

HN S

O

N

S

S O

O O

O NH

O

O

N O

O

NH

N

O

O NH

N

NH

-

H O H O H O H N N N N N N N H O H O H O O OH OH CO2-

HO OH H H O H O H H O H O N N N N N N N N N N H O H O H O H O O O OH OH CO2CO2-

O

H O N NH2

117 (tetravalent)

Figure 4.67 Structure of multivalent MHC peptides tethered to a peptide backbone.

tetramers (117) (Figure 4.67). These multivalent MHC proteins bind to the surface of T-cells in a competitive binding assay, with IC50 values of 1.8 mM (116), 0.31 mM (trimer), and 91 nM (117). Thus, the binding a‰nity is valencydependent, as illustrated by about a 41-fold enhancement by the tetrameric MHC–peptide over the monovalent MHC–peptide. The multivalent MHC– peptide molecules are also functionally active since they can induce T-cell activation under conditions not possible with monomeric MHC–peptide complex. 4.2.3

Integrins

4.2.3.1 Poly(norbornene) Presenting RGD and PHSRN Sequences. Extracellular matrix proteins such as fibronectin can attach tightly to cell surfaces. The cell–matrix adhesion is attributed to the specific interaction of integrin receptors on the cell surface with a matrix component, the RGD peptide ligand [532]. This attachment is tight because multiple RGD domains bind simultaneously with multiple integrin receptors. Multivalent synthetic molecules that present the RGD peptide mimic natural RGD-containing proteins by their

232


O

O

H N

N H

N H

O

O

O

H N

OH

N H O

O

n O

i,ii

O

O

HN

HN

H N

O N H

O

HN NH O S O

H N

N H O

O

OH O

OH

HN HN

OH

O

NH2

O 118 poly(norbornene)(GRGDS) H2N

NH H N

HN O HO

O

H N O

O

N H HO

N H N

119 poly(norbornene)(GRGDS, PHSRN)

HN O

N O O

NH2 m

n O

n/m ~ 1

H N

HN O

N H

H N O

OH

O N H O

OH O

OH

HN HN

O

NH2

reagents and conditions: i) Cl2Ru(PCy3)(N,N-di-Mes-2,3-dihydroimidazole)=CHPh, CH2Cl2, MeOH, 55oC; ii) TFA, TIS, H2O.

Figure 4.68 Synthesis of poly(norbornene) presenting RGD-containing peptide.

ability to interfere with cell–matrix protein adhesion [533,534]. Maynard et al. at Caltech [534] reported the synthesis of polyvalent RGD, or RGD and PHSRN peptides, appended on a poly(norbornene) framework prepared by ROMP (Figure 4.68). (The PHSRN sequence is included in the study since it ˚ apart coexists with RGD in fibronectin, where both sequences are 30 to 40 A and are located on the same face of the matrix protein [535].) The resulting polymers composed of RGD peptides, 118 and 119, block the attachment of human fibroblast cells to fibronectin-coated surfaces, with IC50 values of 0.18 mM (118) and 0.04 mM (119). Thus, they are 7- to 33-fold more potent than the monovalent GRGDS peptide. In addition, the polymer 119, displaying both RGD and PHSRN as coligands, shows greater adhesion activity than that of homomeric 118. 4.2.3.2 GRGDS Displayed on Polystyrene Beads. Synthetic multivalent RGD ligands play an important role in understanding the functions mediated by integrins in cytoskeletal organization and signal transduction. To this end,

233

PEPTIDE AND HORMONE RECOGNITION RECEPTORS

Miyamoto et al. [536] coated polystyrene beads with a GRGDSPC peptide or fibronectin. The beads bind integrin receptors and cause their aggregation, while soluble, monovalent GRGDS simply occupies the receptor, without causing aggregation. The aggregation induced by bound polyvalent beads further triggers intracellular signal transduction, which requires tyrosine phosphorylation and the clustering of several cytoskeletal proteins. 4.2.3.3 RGD Peptides Presented on SAMs. Synthetic surfaces that present multivalent RGD peptides serve as useful models for studying integrin-RGD– mediated cell adhesion [537]. In particular, SAMs on gold are highly convenient for displaying biofunctional groups at varying densities, thereby mimicking a cell surface [538]. Yousaf et al. [539] explored such SAM-based applications by performing selected chemical reactions using reactants attached to SAM surfaces. In one such example, RGD-presenting SAM surfaces were prepared using Diels–Alder cycloaddition (Figure 4.69). Fibroblast cells adhere selectively to a region microfabricated with RGD ligands on a SAM surface. 4.2.3.4 Polymeric Nanoparticles Displaying av b 3 Antagonist. Endothelial cells increase the expression levels of some surface molecules during vascular remodeling, and angiogenesis—a process also required for tumor growth. Among the surface molecules whose expression levels increase, integrin a v b3 H2 N

+ NH2

HN -

O H2 N HO

O2C H N

O N H

O

O

H N O

H N

N H

O

O O

Diels-Alder reaction O

O

S

S

O

O

O O

O

HO O

O

O O

O

i

HO

HO

O

O O

HO

O

O O

O

O

O

OH

O O

O

O O

O O

O O

O O

O O

O O

O O

O O

O O

S

S

S

S

S

OH

S

reagents and conditions: i) RGD-cyclopentadiene conjugate, H 2O.

Figure 4.69 peptide.

Biomimetic surface made of self-assembled monolayer that presents RGD

234

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS multivalent Integrin antagonist HN HN N

HN N HN

HN N HN

O

O

O

O NH NH O NH CO2H CO2H HN CO2H HN O SO HN O O SO OS HN HN O HN O O HN NH N O O N N O O NH NH O O O NMe3+ O + O O Me3N O -O P O O NH O O O NH O O O O O O O O

gene-binding domain Integrin antagonist

+

+ + +

+ +

+ nanoparticle +

+ +

+ +

+

+

~40-50 nm m n

Figure 4.70 Nanoparticle presenting multiple integrin antagonist prepared by polymerization of self-assembled layers of lipid.

[540] binds to a matrix protein and also assists with internalization of bound pathogens. Hood et al. [541] described a bifunctional construct that presents multiple ligands of a v b3 along with a domain that allows conjugation with a gene (Figure 4.70). The construct is designed to bind genes and deliver them to cells that express avb3 . A bifunctional conjugate of a selective a v b3 antagonist, with an IC50 value of 0.04 mM and a copy of the luciferase gene, was attached to nanoparticles of diameter 40 to 50 nm. The resulting nanoparticles deliver the reporter gene selectively to avb3 -bearing cells. In an in vivo tumor model, the adminstration of nanoparticles conjugated with Raf(), a mutant gene that blocks angiogenesis upon expression, induces tumor regression rapidly. This example suggests the therapeutic potential of a multivalent system in its application to targeted gene delivery. 4.2.3.5 Divalent Antagonist of Integrin av b 3 . Silletti et al. [542] invented a divalent antagonist that inhibits the binding of integrin avb3 to matrixmetallo-

PEPTIDE AND HORMONE RECOGNITION RECEPTORS O F 3C

HO2C O

NH

235

O

H HN N

NH H N

O

O

CO2H HN

O

CF3

O O

120

Figure 4.71 Structure of a divalent inhibitor of integrin av b3 binding to matrixmetalloproteinase 2.

proteinase 2 (MMP-2) (Figure 4.71). The C2 symmetric homodimer 120 is composed of two units of a Lys–Gly–benzamide derivative. The homodimer binds directly to integrin avb3 and interferes with RGD-independent interaction between integrin and MMP-2. This dimer also displays antiangiogenic and antitumor activity. 4.2.4

Immunoglobulin

The immune response is mediated by multivalent interactions between divalent immunoglobulins (Igs) and their antigens. Igs can bind tightly and selectively to antigenic determinants present in multiple copies on the surfaces of bacteria and viruses [3]. IgM antibodies are composed of pentameric subunits that bind an antigenic surface with high avidity, whereas the IgM monomer binds to a single antigenic determinant with low a‰nity (Figure 4.72). Each IgG molecule binds two antigens. The binding of IgG to antigens on a cell surface leads to IgG

IgM antigenic determinant

pathogen

C1 complex

C1q

IgG

pathogenic surface

Figure 4.72 Multivalent interaction implicated in immune responses such as IgGantigen, IgM-antigen, and specific recognition of bound IgG by C1 complex.

236


the formation of an IgG–antigen cluster and subsequent recruitment of the pentameric complement factor, C1q, to multiple bound IgG molecules. The formation of the multivalent, specific IgG–C1q complex triggers a complement pathway that leads to the destruction of antigen-presenting pathogens and cells. C1q binds more tightly to IgG clusters than to individual IgG molecules. For instance, the a‰nities of C1q to the IgG monomer, dimer, and tetramer increases with increasing valency, with association constants of 100 mM, 1 mM, and 3 nM, respectively [543]. 4.2.4.1 Pentavalent Peptides. High-avidity ligands to immunoglobulins can be designed by converting a weakly binding peptide ligand to a multivalent ligand. Terskikh et al. [544] illustrated such an approach by identifying a pentavalent peptide called a peptabody which binds tightly to amIgM antibody derived from B-cell lymphoma. This peptabody molecule is a type of selfassembled oligomeric complex composed of five peptides, each of which is designed to present a peptide domain called S peptide that recognizes IgM and a pentamerization domain fused to the ligand domain. The pentavalent molecule shows high-avidity binding to IgM molecules coated on the surface, with an IC50 value of 1 nM, which represents a 2 10 5 -fold enhancement over the monovalent S peptide. 4.2.5

Estrogen Receptors

Estrogen receptors are involved in the regulation of gene expression in cell nuclei. These receptors are activated via the binding of steroid hormones, such as estradiol and estrone [545]. Estrogen receptors dimerize where some of their bivalent ligands are able to interfere with the dimerization [546,547]. Rabouin et al. [548] described the activities of symmetric dimers composed of 17b-estradiol linked with either oligo(ethylene glycol) or an alkane spacer at position C17 . In cells expressing the estrogen receptor, dimers with a di- or tri(ethylene glycol) spacer display inhibitory action as potent as that of a monomer unit (IC50 ¼ 71 mM ), although they are twofold less active than is estradiol itself. The inhibition activity is influenced significantly by tether distance, such that dimers linked with a larger or smaller spacer are inactive. The results above, although unexpected, indicate the importance of additional designing factors. For example, the oligo(ethylene glycol) linker might be undesirable due to its hydrophilicity (hydrated) and flexibility, which can cause significant loss of entropy upon binding to hydrophobic receptor domains. 4.3 4.3.1

LIGAND-MEDIATED RECEPTOR DIMERIZATION Chemical Inducers of Dimerization

Immunosuppressants are molecules that decrease the activity of the immune system and are typically used to prevent the rejection of transplanted organs.

LIGAND-MEDIATED RECEPTOR DIMERIZATION

H 3C O N

H3CO HO

H

H O

OCH3 OCH3

O OH H O O H3C OH O

CH3

CH3 CH3

237

CH3 HO CH3 H3C CH3 Me O Me O Et Me N O N N N N H 3C Me O H H O O H3C CH3 NMe CH CH3 3 MeN OH3CH O H O N N N N H3C H CH3 O Me O H3C CH3 CH3 CH3

FK506

Cyclosporin CH3

H3C

H 3C

O

O H OH OMe O

O N H O

O

O

CH3

OH

H

OMe OH

CH3 CH3 CH3 CH3 Rapamycin

Figure 4.73 Structure of naturally occurring immunosuppressants including FK506, rapamycin, and cyclosporin A.

The natural product class of immunosuppressive drugs are represented by FK506, cyclosporin A (CsA), and rapamycin (Figure 4.73). Such molecules permeate cell membranes and are small compared to proteins. These immunosuppressants are considered to be bivalent in structural and functional aspects since they are composed of two orthogonal protein–binding surfaces, which mediate protein-protein dimerization [68,69,549]. Their primary cellular targets, called immunophilins, include FKBP (FK506-binding protein) and cyclophilin. 4.3.1.1 Natural Heterodimerizers. FK506 is a macrocyclic lactone isolated from soil microorganisms. In the cytoplasm, FK506 binds to FKBP-12 (FK506-binding protein 12) with a subnanomolar a‰nity and a Kd value of 0.4 nM (Figure 4.74) [550]. The formation of this binary complex generates a new composite surface comprised of structural elements from FK506 and FKBP. This binary complex subsequently associates with its next target, calcineurin, which is a cellular enzyme with phosphatase activity, resulting in formation of a ternary complex. Once bound, calcineurin is unable to perform its catalytic action, which is linked to the cellular signaling process necessary for nuclear transcription. Rapamycin behaves like FK506 by binding to FKBP. However, a resulting binary complex inhibits the protein kinase FRAP (FKBP-12rapamycin-associated protein) rather than calcineurin. Cyclosporin exerts its

238

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS FK506

cell membrane

Calcineurin

FKBP-12

FKBP/FK506

FKBP/FK506/Calcineurin

Cyclosporin A (CsA) cell membrane

Calcineurin

Cyclophilin

Cyclophilin/CsA

Cyclophilin/CsA/Calcineurin

Rapamycin

cell membrane

FRAP

FKBP-12

FKBP/Rapamycin

FKBP/Rapamycin/FRAP

Figure 4.74 Protein–protein association mediated by FK506, cyclosporin A, and rapamycin.

immunosuppressive action through association with cyclophilin and calcineurin. In summary, the immunosuppressants selected function as natural heterodimerizers: They bring two di¤erent proteins into close proximity and inhibit the function of bound proteins, o¤ering a way to control intracellular signaling processes. 4.3.1.2 FK506 Dimer (FK1012). FK1012 is a synthetic dimer of FK506 composed of two monomers that are linked through a C21 allyl group (Figure 4.75) [551]. This dimer does not retain immunosuppressive activity, since the position of the linker is directed at the calcineurin-binding domain of FK506 and thus does not allow the dimer to bind calcineurin. Instead, it binds two molecules of FKBP with high a‰nity (0.2 nM) and thus induces the homodimerization of FKBP. The members of this class of molecules are defined broadly as chemical inducers of dimerization (CIDs). With such specific ability,

LIGAND-MEDIATED RECEPTOR DIMERIZATION H 3C O N

H3CO HO

H

H O

O OH H O O H3C OH O

CH3 CH3

239

OCH3 OCH3 CH3 O H N

O

N H

O

O CH3 H3C H3CO

FK1012

H3CO

OH

CH3 CH3

O

OH O O H O H HO O N

H

OCH3

O CH3

Figure 4.75 Structure of FK1012 as a synthetic homodimer of FK506.

FK1012 has served as a valuable tool in understanding and controlling intracellular signal transduction. Schreiber at Harvard and Crabtree at Stanford, as reported by Pruschy et al. [551], explored various roles of FK1012 using engineered cells that contain diverse conditional alleles that encode, for example, FKBP-fused chimeric receptors (Figure 4.76). In T-lymphocytes, FK1012 dimerizes myristoylated FKBP-12 and FKBP chimeric FKBP-Sos receptors (guanine nucleotide exchange factors). The divalent association brings FKBP12 and the chimeric Sos receptor into close proximity near the cell membrane [552]. This type of membrane recruitment of Sos triggers Ras activation. In a similar example, FK1012 activates Src-like tyrosine kinase as a result of protein heterodimerization and membrane localization, which leads to the initiation of caspase-mediated programmed cell suicide, or apoptosis [553,554]. FK1012 also permeates the nucleus, where it induces protein heterodimerization by bringing a transcriptional activation domain close to a promoter domain bound on DNA [555]. 4.3.1.3 FK506 Mimetics and CIDs. Many chemical inducers of dimerization (CIDs) are derived from nonnatural molecules. These molecules are designed to be less complex in structure and are modified more easily than are naturally derived CIDs. These properties make nonnatural CIDs suitable candidates for structure-based optimization of their pharmacological properties. Amara et al. [556] reported 122 (AP1510) and 123 as examples of nonnatural CIDs (Figure 4.77). Comprising two synthetic monomers tethered with a short linker, the molecules simultaneous bind to two FKBPs, such as FK1012. In cells encoding a myristoylated FKBP–Fas fusion receptor, the dimers induce apoptosis in a dose-dependent manner. For example, 122 is active at nanomolar concentration and is three times more potent than FK1012. Both 122 and 123 are also active in controlled expression of specific genes in a conditional system where DNA transcription is triggered by heterodimerization of a FKBP-fused DNAbinding domain with a FKBP-fused transcriptional domain bound on DNA strands. For instance, the dimer 122 is slightly more potent than either FK1012

240

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS FK1012 membrane localization of Src kinase cell membrane

GTP

Ras

GDP Myristoylated FKBP-12

FKBP-12

Sos Src-family kinase

membrane localization of Sos

nuclear membrane

FKBP-12

Fas domain

Caspase-8 Fas-mediated programmed cell death

Activation domain/ FKBP DNA binding domain

plasmid DNA activating transcription

Figure 4.76 Exploration of FK1012-mediated signal transduction with designed conditional alleles as illustrated in membrane localization of Sos and Src kinase, Fasmediated assembly of caspase, and activation of DNA transcription.

or 123, which contains a longer spacer. Addition of a monovalent competitor, such as FK506 or 121, inhibits transcriptional activation, with an IC50 value of 6 nM (FK506). These data suggest that the mode of action by 122 is linked to specific binding to FKBP-fused proteins. 4.3.1.4 Cyclosporin Dimer. Cyclosporin A (CsA) is a cell-permeable macrocyclic peptide with high a‰nity to cyclophilin (CyP), at a Kd value of 5 nM. CsA presents two distinct surfaces that bind to two orthogonal targets, CyP and calcineurin. CsA can be converted into a homodimer that presents two identical protein-binding surfaces (Figure 4.78). For example, a cyclosporin dimer (124) that is composed of two CsAs tethered with a xylyl spacer can bind two CyP proteins simultaneously [557]. However, since its original binding surface to calcineurin is sterically blocked, due to the presence of linker domain, this dimer cannot recognize calcineurin and thus is ine¤ective as an immunosuppressant. In a study reported by Belshaw et al. [557], 124 is found to trigger apoptosis, even at nanomolar concentrations, in engineered T-cells en-

241

LIGAND-MEDIATED RECEPTOR DIMERIZATION H3CO

Ch3O H3CO

OtBu

O

+

OH

N O

O

H3CO

i,ii

O

OH

O

O

OH

N O

O

O O

O 121

H3CO

OCH3

H3CO iii,iv

H N

O O

N O

O

O

Ln

OCH3

H N

O O

O

O

N OO

O

122: Ln = CH2CH2 123: Ln = (CH2CH2O)2CH2CH2

reagents and conditions: i) DCC, DMAP, CH2Cl2; ii) TFA, CH2Cl2; iii) disuccinimidyl carbonate, pyridine, CH3CN, iv) H2N-Ln-NH2, Et3N, CH3CN.

Figure 4.77 Structure of synthetic analogs of FK1012.

coding chimeric receptors composed of CyP fused to Fas receptor. Apoptosis is caused by CID-mediated receptor aggregation and localization of multiple Cyp–Fas fusion receptors at the inner surfaces of cell membranes. 4.3.1.5 Heterodimeric CIDs Composed of FK506 and Cyclosporin. A composite surface made of two protein-binding surfaces, one from FK506 and another from CsA, serves as a heterodimeric CID that can recruit two orthogonal targets into close proximity. Indeed, a heterodimer (125) induces heterodimerization of FKBP and CyP (Figure 4.79) [558]. Such dimerization activates a Fas signaling pathway in cells encoding membrane-anchored FKBP as well as soluble CyP-fused Fas receptors. The signaling pathway is triggered by dimer-mediated oligomerization of Fas receptors at membrane surfaces. In a separate example, this heterodimerization induces transcriptional activation within a nucleus through recruitment of an FKBP-fused enhancer domain to a CyP-fused DNA promoter. 4.3.1.6 Heterodimeric CIDs Composed of FK506 and SH2 Ligands. CID constructs can also mediate protein–protein contacts among proteins that do not belong to the immunophilin family. Briesewitz et al. [559] illustrated the general application of this strategy by designing heterodivalent ligands that target Src-family tyrosine kinases as well as FKBP. The Src-family kinase proteins contain an SH2 domain composed of a relatively small peptide module of about 100 residues. This domain binds specific peptides containing phosphotyrosine, a residue found widely in cell-signaling proteins. Heterodimers 126 and 127 are composed of FK506, or a synthetic ligand to FKBP, tethered

242


CH3 CH3 HO H 3C CH3 Me O Me O Et Me N N O N N N H 3C Me O H H O O H 3C CH3 NMe CH CH3 3 MeN OH3CH O H O N N N N H 3C H O CH3 O HMe CH3 3C CH3 CH3

CH3 HO CH3 Me O N

i-iv

OH H 3C Et Me Me O O N N H3C N N Me O H H O O H 3C CH3 NMe CH CH3 3 MeN OH3C O H O H N N N N H3C H O CH3 O HMe 3C CH3 CH3 CH3

Cyclosporin H3C

CH3

H3C N H

v,vi O

CH3 HO CH3 Me O N

O

N H3C H Et Me Me O O N N N N H3C Me O H H O O H3C CH3 NMe CH CH3 3 MeN OH3CH O H O N N N N H3C H O CH3 O HMe 3C CH3 CH3 CH3

CH3 Me O N

CH3 H CH3 N N O H O O CH3 NMe CH3 CH3 NMe H3C CH3 O O H H O Me N N CH3 N O N N Me Et O Me O Me H H3C CH3 N O OH O

O

CH3

124

reagents and conditions: i) Ac2O; ii) NBS, AIBN; iii) NEt4OAc; iv) NaOMe; v) 2 eq. CO(Im)2 vi) 1,4-diamino-p-xylene.

Figure 4.78 Synthesis of cylcosporin homodimer (CsA)2 .

through a short linker to a phosphotyrosine SH2 ligand, Tyr(OPO3 )-Glu-GluIle (Figure 4.80). These heterodimers bind to FKBP, although at a lowered a‰nity relative to the respective monovalent FKBP ligands. For example, 126 associates with FKBP-52 with a Kd value of 150 nM, an a‰nity threefold lower than that of FK506 to FKBP-52. In addition, 126 can bind to the Fyn kinase SH2 domain to form a binary complex with a Kd value of 340 nM. This binary complex subsequently forms a ternary complex with FKBP-52, where the a‰nity of 126 to FKBP-52 is enhanced, possibly due to favorable protein–protein interactions mediated by such heterovalent ligands. The heterodimer 127 binds to FKBP-12 with a Kd value of 60 nM, which represents a threefold lower a‰nity than that of nontethered, synthetic ligand to FKBP (SLF). This heterodimer also forms a complex with the Fyn SH2 domain, with a Kd value of 180 nM. This dimer forms a ternary complex between the SH2 domain and FKBP-12, but with sixfold reduced a‰nity to the SH2 domain and a Kd value of 1000 nM. The weakened association illustrates

243

LIGAND-MEDIATED RECEPTOR DIMERIZATION CH3 HO CH3 Me O N

O

Br H3C Et Me Me O O N N H3C N N Me O H H O O H3C CH3 NMe CH CH3 3 MeN OH3CH O H O N N N N H3C H CH3 O HMe O 3C CH3 CH3 H3C

N

O

O

OH

CH3 CH3

O

O

H

O CH3

OH O O H O H HO O N

H 3C H3CO

OCH3

O CH3

H3CO

iv,v

i,ii,iii

O

CH3

O

HO CH3 Me O N

OH H 3C Et Me Me O O N N H 3C N N Me O H H O O H 3C CH3 NMe CH CH3 3 MeN OH3CH O H O N N N N H 3C H O CH3 O HMe 3C CH3 CH3 H3C

H 2N

N H

OH

CH3 CH3

O

H

O CH3

OH O O H O H OH O N

H 3C H3CO

OCH3

O CH3

H3CO

vi CH3

O

O

HO CH3 Me O N

HN H 3C Et Me Me O O N N H 3C N N Me O H H O O H 3C CH3 NMe CH CH3 3 MeN OH3CH O H O N N N N H 3C H O CH3 O HMe 3C CH3 CH3 H3C

N H

O CH3 H 3C H3CO H3CO

OH

CH3 CH3

O

OH O O H O H OH O N

H

O CH3

125

reagents and conditions: i) ethyl hydrogen malonate, isopropyl magnesium bromide, THF; ii) Et4NOAc•4H2O, DMF, 90oC; iii) NaOMe, MeOH; then H2O; iv) H2N(CH2)6NHBoc, DCM; v) HF, MeCN; vi) azaBop, lutidine, DCM.

Figure 4.79

Synthesis of FK-CsA heterodimer.

OCH3

244

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS H 3C O N

FK506

OCH3

O OH H O O H3C OH O

H O

H

H3CO

OCH3

CH3 ligand to Fyn SH2 domain CO2H N

O

CH3 CH3

HO

O

O (a) FKBP

O

H N

N H

N H

O

Fyn kinase (Src-family)

OPO3-

O-

126

O-

127

O

CO2-

synthetic ligand to FKBP CO2MeO MeO

H N

O O

O

O

O

O

O

H N

N H

N H

O

H

N

OPO3-

O

O

CO2-

synthetic ligand to FKBP

O OMe

(b) FKBP

MeO

unknown cell target

O

O

O

O

H N

O

MeO

128

N O

O

H

HO

N

O MeO

OMe OMe

N CO2Et CO2Et variable target domain

SLF-MTx dimer synthetic ligand to FKBP (SLF)

Methotrexate (Mtx)

MeO (c) FKBP

O

MeO λcI

H N

DHFR αNTD

O

RNA polymerase RNA binding RNA binding

O N

O H

O

O CO2H

H N 8

O

NH2

N H N Me

N N

N N

O

Figure 4.80 Structure of heterodimers composed of two orthogonal ligands capable of binding to two distinct cellular targets, including FKBP and (a) Fyn SH2 domain or (b) unknown receptor; (c) dimerzation of FKBP and DHFR hybrid proteins on RNA mediated by a dimeric ligand composed of methotrexate linked to FK506 mimic.

NH2


245

a negative contribution generated upon formation of a ternary complex. Based on nuclear magnetic studies of [15 N]FKBP-52, Rosen et al. [560] attributed such reduced a‰nity to an attractive interaction between Tyr(OPO3 )-Glu-GluIle and FKBP-52 residues, which causes a conformational change of the phosphotyrosine ligand, thus perturbing its targeted interaction with the SH2 domain. A third example of heterodivalent CIDs is based on a synthetic library, with members presenting the FKBP ligand linked to a tetrahydrooxazepine, 128 (Figure 4.80). Koide et al. [561] designed these ligands to recruit both FKBP and unknown cell targets that are recognized by the tetrahydrooxazepine moiety. Members of this class of molecules can permeate cell membranes, bind FKBP, and inhibit dimerization of FKBP-fused receptors mediated by homodimeric FKBP-binding ligands. Finally, the application of FK506-based heterodimers has been expanded to a system involving the transcriptional activation of genes. Altho¤ and Cornish [562] reported such dimers, composed of a ligand that mimics FK506 (SLF) linked to methotrexate (Mtx) through a 10-atom spacer (Figure 4.80). Designed to bind FKBP12 and dihydrofolate reductase (DHFR), the SLF-Mtx heterodimer brings these two proteins into close contact. It also dimerizes two hybrid proteins, composed of FKBP12-lcI and DHFR-aNTD (where aNTD is the Nterminal domain of the a-subunit of RNA polymerase). Both the cI and aNTD proteins bind to RNA directly or indirectly via RNA polymerase. The protein dimerization induced by the SLF-Mtx ligands activates the transcription of specific genes in the E. coli system. 4.3.1.7 Bivalent Ligands Binding to FKBP. Synthetic divalent ligands that target FKBP are designed to bind two intramolecular sites within FKBP rather than targeting two FKBPs. Such dimers show enhanced a‰nity but cannot induce receptor dimerization. For example, Shuker et al. at Abbott Laboratories [563] reported a ‘‘screening by NMR’’ method that makes it possible to discover multiple small ligand molecules that bind weakly to multiple, distinct sites on FKBP. Linking two selected ligands leads to a hetrodimer that binds two proximal sites within a single FKBP receptor and generates a single ligand with a high a‰nity to FKBP. This example illustrates that weakly binding monomeric ligands be tethered together to form a single ligand capable of binding tightly, through divalency. 4.3.1.8 Rapamycin-Mediated Association of Inteins. The concept of CIDmediated protein heterodimerization is integrated with protein trans-splicing, a reaction that is turned on only when the N- and C-terminal halves of a protein splicing element (termed intein) are brought to intimate proximity [564]. Mootz and Muir [565] reported the conditional trans-splicing of a protein. The method encompasses two complementary fusion proteins. The first fusion protein encodes FKBP fused to N-terminal intein linked to protein A. The second encodes

246


FKBP

FRB

InteinN

InteinC

Protein A

Protein splicing Rapamycin

Protein B

Hetero-dimerization Protein A-B

Figure 4.81 Rapamycin-mediated protein dimerization and splicing.

FKBP-rapamycin-binding protein (FRB) fused to a C-terminal intein linked to protein B (Figure 4.81). These two proteins do not interact in the absence of rapamycin, but they do interact in the presence of rapamycin because rapamycin binds to two protein surfaces derived from the two fusion proteins. The association of two inteins catalyzes the generation of protein-splicing products. Such catalysis in a trans-splicing reaction involves rapamycin-mediated protein dimerization, which can be inhibited by adding FK506 or ascomycin, two molecules that bind tightly to FKBP. 4.3.2

Cytokine Receptors

Cytokines and interleukins are endogenous secreted peptides and proteins. They interact with white blood cells by binding to cytokine receptors present on cellular surfaces [3,566]. Upon binding, the cytokine receptors dimerize, which triggers a series of intracellular signaling events mediated by a pathway involving Janus kinase (JAK) and the signal transducers and activators of transcription (STAT) (Figure 4.82). This pathway is known simply as the JAK/ STAT pathway. A member of the cytokine receptor superfamily, the erythropoietin receptor (EPO-R), controls the proliferation and di¤erentiation of red blood cell precursors [566]. Agonists to EPO-R include the erythropoietin (EPO) protein and

cytokine receptor dimerization

cell membrane

STAT recruitment

cytokine receptor STAT

Figure 4.82 Cytokine-induced dimerization of cytokine receptor leading to STAT recruitment.


247

EPO-mimetic peptides (EMP). The red blood cell precursors are activated when EPO binds to the extracellular portion of EPO-R and homodimerizes the receptor. EPO-R dimerization subsequently turns on a downstream signaling pathway through the recruitment and dimerization of STAT [567]. The EPO-R dimerization is achieved by the binding of several divalent species, including endogenous EPO, peptide mimetics and their dimers, bivalent antibodies, and chemical cross-linking reagents [568,569]. 4.3.2.1 Divalent EPO Mimetics. As mentioned, a dimeric ligand, composed of small molecules and/or peptides that mimic EPO, can activate EPO-R through receptor dimerization (Figure 4.83). Examples of such ligands include non-peptide-based 129 and 130. Goldberg et al. [570] reported the divalent 129, which actively stimulates cell proliferation, although it shows a lower intrinsic level of activity than that of natural EPO. A second example is the multivalent ligand 130, which also increases proliferation activity [571]. A peptide-based dimer (131) consists of two EPO-mimetic peptides linked through a long, flexible poly(ethylene glycol) spacer [572]. This dimer binds tightly to EPO-R, with an IC50 value of 2.5 nM. Its a‰nity for EPO-R is about 1000-fold enhanced over that of a monomeric EMP-PEG3400 . It serves as an agonist of cell proliferation, with an ED50 value of 0.1 nM and a potency that is far greater than that of the monovalent EMP-PEG3400 or known EMP1 peptides, with an ED50 value of 0.1 mM. 4.3.2.2 Divalent Peptide Antagonists of IL-5. Interleukin-5 (IL-5), a cytokine molecule derived from T cells, is involved in controlling erythrocyte production by binding the cells of eosinophil and basophil lineage. Its target is the interleukin-5 receptor (IL-5R), a transmembrane, heterodimeric complex composed of an extracellular, agonist-binding a-chain and a cytoplasmic, signal-transducing b-chain. England et al. [573] reported a peptide dimer as an antagonist of IL-5R whose primary sequence is not associated with IL-5 (Figure 4.83). This peptide (132) displays two identical copies of a 19-amino acid peptide cross-linked through a disulfide bridge. It shows a high a‰nity to the a-chain of IL-5R, with an IC50 value of 0.4 nM, which is close to that of the native ligand. Based on gel filtration studies, this dimeric peptide binds simultaneously to two a-chains of IL-5R. Such divalent binding also makes it possible to dimerize a chimeric receptor that consists of an IL-5 a-subunit fused to the intracellular domain of the epidermal growth factor receptors (EGFRs). 4.3.2.3 Dimeric Mimetics of G-CSF. Granulocyte-colony-stimulating factor (G-CSF) is a protein hormone that controls the production and activation of cells of granulocytic lineage. Its mode of action involves dimerization of the GCSF receptor, which triggers the JAK/STAT-signaling process through the cytoplasmic domain. The signalling process can be activated by a nonpeptide mimic of G-CSF. As an example, Tian et al. [574] reported a benzimidazolecontaining small molecule (133) (Figure 4.83). A dimer with twofold rotational

248

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS H N

O

H N

MeO2C NHCbz

O

O

N

N O

O

O

NHCbz N H

CO2Me

N H

129 O RO O N H

RO

O O

H N

N O

O N H

RO

O

N H

O

N

N H

O

O

N

O N H

N

N

H N

H N

O

O

O N H

O

O

OR

H N

N

NH

N H

O

O

H N

OR

H N

HN

N H HN

OR

H N O OR

NH H N

130

O

O O NH2

HN O

CH3

R

Cl N H O

EMP

N H

erythropoietin mimetic peptide (EMP)

O O

O n

O

PEG 3400

S

N H

GGLYACHMGPMTWVCQPLRG-CO2H

131

S

TGGGDGYVCVEWARCPTCK S S

N

HN

HN

TGGGDGYVCVEWARCPTCK S

H N

N

N

N

N

NH N H

NH

N

S

132 133

Figure 4.83 Structure of multivalent ligands to erythropoietin receptor, IL-5 receptor, and G-CSF receptor.

249


symmetry, this ligand binds to the extracellular domain of the G-CSF receptor and triggers receptor activation through ligand-induced receptor dimerization. 4.3.2.4 Dimeric Mimetics of NT-3. Neurotrophins, such as neurotrophin 3 (NT-3), and nerve growth factor (NGF) induce cell signal transduction by binding to membrane-bound target receptors. These receptors belong to the family of tyrosine kinase receptors and include TrkC, the NT-3 receptor, and TrkA, the NGF receptor. Some divalent ligands have been designed to dimerize the neurotrophin receptors. Zhang et al. and Pattarawarapan et al. [575] presented such a dimeric ligand (135) composed of two b-turn cyclic peptides (134) as a functional mimic of NT-3 (Figure 4.84). This dimeric ligand selectively stimulates cells that contain TrkC. Its level of activity is much higher than that displayed by a monovalent ligand, such as NT-3 or 134. In a second example, two identical peptidomimetics of the Trk ligand are tethered to a triazine sca¤old. With a fluorescent tag attached, the mimetic is screened directly for binding to Trk-expressing cells by fluorescene-activated cell sorting (FACS). OH O HN O2N

HN O HN NH

134 β-turn mimetic of NT-3

O NH2 O

O

HO

OH O HN O2 N

HN O HN NH

H N

O HN

O

N H

O

O HN

NH O

O

H N

O

NH O

N H

HN NH O NH

O

H2N O H2NOC

N NH HO

NH

O

135

HO

OH

O

NO2

O

O

O HO

OH

HN

NH

NH2

N O HHN

O CONH2

O

O N H

HN O

O

N

N

N N

H N

NH

N

O

O O

HN

O

HO2C

fluorescein tag

OH

Figure 4.84 Structure of bivalent b-turn peptidomimetics as ligand to neurotrophin receptor.

250

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS H3C OH H

HO

H3C

CH3

OH

1α,25-dihydroxyvitamin D3 (calcitriol)

n

HO H

HO

n =1, 3

OH

HO

OH H

OH

136

Figure 4.85 Structure of vitamin D3 (calcitriol) dimer.

4.3.3

Vitamin D3 Receptors

Vitamin D3 is a precursor of calcitriol (1a,25-dihydroxyvitamin D3 ), a hormone involved in a broad range of biological process, from the promotion of cellular di¤erentiation to the inhibition of tumor cell proliferation (Figure 4.85). Calcitrol targets vitamin D3 receptors (VDRs) in the nucleus, where they regulate the activation of gene transcription. Their mode of action includes homodimerization or heterodimerization of VDR with the retinoid X receptor. Therefore, a dimeric form of calcitriol ligand is expected to bind and dimerize VDR. Such a ligand should be important for understanding the role of dimeric VDR and its interaction with DNA. Sestelo et al. [576] reported the synthesis of such dimeric ligands. Calcitriol dimers comprising two calcitriols linked with six- or 10-carbon atom spacers at the C11 position, such as 136, are synthesized through a convergent synthetic strategy.

4.4 4.4.1

ENZYMES Acetylcholinesterase

Acetylcholinesterase (AChE) is an enzyme that catalyzes the hydrolysis of a neurotransmitter, acetylcholine (ACh), providing a route by which the level of Ach is regulated in the neuronal synapse [577]. Control of Ach and other neurotransmitters at the cellular level is critical, given that their decrease is often associated with certain neurodegenerative diseases, including Alzheimer’s disease. One potential way to deal with such disease is to limit the catalytic action of AChE through enzyme inhibitors [578]. 4.4.1.1 Tetrahydroacridine Dimer. A proposed substrate-binding pocket of AChE is composed of two sites. One site, in the catalytic domain, has a high a‰nity to substrate. The second site, with low a‰nity, is located near the entrance to the catalytic pocket (Figure 4.86) [579]. Pang et al. [580], Carlier et al. and Han et al. [581], and Hu et al. [582] reported inhibitors of AChE that are

ENZYMES

251

covalently linked dimer Peripheral site (low affinity)

Catalytic site (high affinity) acetylcholine esterase (AchE)

NH2

HN

(CH2)n

NH

i,ii N

N

THA

N 137 (n = 7, 10)

Cl

HN

(CH2)7

NH

iii Cl

N

Cl

N

N

Cl

138

Cl

HN

(CH2)7

NH

iv,v N

N

N 139

reagents and conditions: i) KOH, DMSO, rt; ii) Br(CH2)nBr (0.5 eq); iii) 1,7-diaminoheptane, PhOH, NaI, 180oC; iv) 1,7-diaminoheptane (3 eq), 1-pentanol, 160 oC; v) 4-chloroquinoline, 1-pentanol, 160oC.

Figure 4.86 Catalytic pocket of acetylcholinesterase presenting a high-a‰nity substrate site and a low-a‰nity site near the entrance to a second substrate and divalent inhibitors of AChE composed of tetrahydroacridine.

designed on the basis of a structure-based divalent approach. This approach leads to a dimeric inhibitor comprising two tethered inhibitors expected to bind simultaneously to two proximal sites, thereby possibly achieving enhanced affinity and selectivity. For example, 9-amino-1,2,3,4-tetrahydroacridine (THA), a drug used for treatment of Alzheimer’s disease, is dimerized through an alkane spacer at the 9-position of THA. An inhibitor, 137 (n ¼ 7), shows potent inhibitory activity against rat AChE, with an IC50 value of 0.4 nM, which indicates 1475-fold enhanced activity over monovalent THA. The dimerization promotes binding a‰nity at AChE compared to butyrylcholinesterase (BChE). For example, 137 (n ¼ 7) has decreased activity at BChE (IC50 ¼ 390 nM ), which is ninefold less active than the THA monomer (IC50 ¼ 44 nM ). Dimer potency and selectivity are influenced by the length of spacer as well as the structural features of the monomeric inhibitor. A dimer with a longer

252

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS Me NH

NH Me

NH2

O

HN O

H2 N

(-)-Huperzine A

NH

O

O HN

140

(CH2)n NH

141 (n = 10, 12, 13)

Figure 4.87 Structure of bivalent AChE inhibitors based on huperzine A.

spacer (137; (n ¼ 10) is about sevenfold less active to AChE than a shorter dimer (n ¼ 7). Addition of a chlorine atom on a THA ring, as exemplified by 138, increases the activity about threefold relative to 137 (n ¼ 7). In contrast, replacement of a THA unit with quinoline, as exemplified by 139, reduces inhibition activity about sevenfold. 4.4.1.2 Huperzine Dimer. Wong et al. [583] proposed a mode of bivalent binding at AChE by a tetrahydroquinolone dimer on the basis of evidence obtained from x-ray structural analysis. The structural determination was performed with dimers derived from Huperzine A, an alkaloid with strong inhibitory activity against AChE (Figure 4.87). For example, the dimer, 141 (n ¼ 10) is active against AChE isolated from Torpedo californica, with a Ki value of 0.8 mM. Thus, 141 is >200 times more active than Huperzine A. A second dimer with a longer linker length (141, n ¼ 12) is about fivefold less active than the construct with a linker length of 10. The crystal structures of the dimers bound to AChE show that one ligand unit occupies a substrate site at the bottom of the enzyme active site, while the second unit binds in a peripheral site located at the top entrance, as shown in Figure 4.86 [583]. 4.4.1.3 Surface Anchoring. In addition to the bivalent nature of enzyme catalytic pockets, certain AChEs, such as those present in vertebrate synapses, are anchored to basal lamina. This surface anchoring involves multivalent interactions established between multiple heparin sulfates displayed on the basal surface and the collagen tail subunit (ColQ) of AChE, where ColQ presents two nonidentical heparin-binding sites with di¤erent a‰nity [584]. In this way, AChE is anchored and positioned at a neuromuscular junction. 4.4.2

Matrix Metalloproteinases

Stromelysin, a zinc-dependent endoproteinase involved in matrix degradation and tissue remodeling, belongs to the family of matrix metalloproteinases (MMPs). Abnormal regulation of the processes controlled by the MMPs is associated with arthritis and tumor metastasis. Inhibitors of this class of enzymes include peptides as well as natural products such as pycnidione and tetracyclines. Using an NMR screening tool, Hajduk et al. [585] found that stromelysin contains a variety of sites, around a catalytic site, which recognize small,

ENZYMES Low-affinity site (Kd = 17 mM)

High-affinity site (Kd = 0.02 mM) OH

253

covalent link O

H N

H N

OH O Zn

OH O Zn

NC

NC

Stromelysin

H N

O n

H N

O

OH

n

O

OH

O

NC 142 (n = 1, 2)

NC

143 (n = 1, 2)

Figure 4.88 Design of high-a‰nity ligand of stromelysin based on covalent linkage of two low-a‰nity ligands and examples of such a divalent inhibitor.

distinct molecules, although at low a‰nity. One such site binds acetohydroxamic acid with a Kd value of 17 mM. This molecule is found as a structural element chelated to a zinc ion at the active site of MMP (Figure 4.88). A second binding element includes a group of biphenyl derivatives that target a large hydrophobic domain present at the catalytic site. The biphenyl derivatives bind with a higher a‰nity, as reflected in the Kd value of 0.02 mM [586]. Covalent linkage of the two molecular elements creates a high-a‰nity ligand to the enzyme. For example, 142 (n ¼ 1) and 143 (n ¼ 1) show Ka values of 25 and 15 nM, respectively. The a‰nities of such dimers are influenced by linker length to such a great extent that a variation of only one atom results in about a 100-fold di¤erence in a‰nity. 4.4.3

Tryptase

Human b-tryptase is a trypsinlike serine protease present in mast cells. It catalyzes the hydrolysis of peptide bonds at the C-terminus of argine and lysine residues [587]. Its uncontrolled action is associated with the pathogenesis of allergic and inflammatory disorders in asthma. This enzyme exists as a tetramer consisting of four identical subunits [588], each of which contains an active site directed toward a central pore (Figure 4.89). At the bottom of each subunit lies an S1 catalytic subsite presenting negatively charged Asp-189. The shortest distance between two S1 pockets from neighboring subunits is approximately ˚ , and the longest distance is about 45 A ˚ from two subunits located in a 33 A diagonal relationship. Due to its well-characterized active-site topology and subunit assembly, b-tryptase presents a target uniquely suited to multivalent ligand design.

254

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS 33 Å B

A

Asp189

Divalent bridging

45 Å

45 Å

Monovalent binding

active site C

D

β-Tryptase tetramer

OH O O HO OH O OH HO O O O HO N 13 Å OH HO H HO O OH O O HO OH HO O OH OH OH OH O O HO O OH O

NH2

O

O S

N H

OMe O

144

H2N HO N S

O

O HO 145

146: R = bond 147: R = -HNCH2CO148: R = -HN(CH2)2CO-

O

O S

OH H2N O O HO OH HO O OH O O O HO N R S OH NH H O HO O OH O O HO OH HO O OH OH OH OH O O HO O OH O

NH2

R

O HO

Figure 4.89 Active-site topology of b-tryptase tetramer and its interaction with divalent inhibitor designed on a cyclodextrin sca¤old.

4.4.3.1 Bivalent Inhibitors Tethered to Cyclodextrin. Schaschke et al. [589] described bivalent inhibitors of b-tryptase (146 to 148), which are composed of two weak ligands (144) tethered to a b-cyclodextrin sca¤old (Figure 4.89). This ˚ sca¤old is favored because of its rigidity and fixed distance of roughly 13 A between two opposing side arms that ensure the optimal fitting of binding units within the multimeric enzyme complex. Dimers 147 and 148 show strong inhibitory activity against human b-tryptase, with Ki values of 0.6 and 2.5 nM, respectively. Such activities are much higher than that of a monomer (144; Ki ¼ 17 mM ) or a shortly linked dimer (146; Ki ¼ 30 mM ). Examples of such bivalent ligands illustrate the importance of linker distance. The dimer 147 is

ENZYMES

255

BocHN

MeO O

N H

H N

HO

O NH2 n

+

O

O

O

i,ii

O OH

N H

H 2N

O MeO O

N H

H N n

H N O

O

N H

O

O N H

149 (n = 1, 2, 3)

H n N O

O OMe NH2

reagents and conditions: i) i-Pr2NEt, EDC, HOBt, DMF; ii) 95% aq. TFA, 0oC to rt.

Figure 4.90

Synthesis of symmetric diketopiperazine-based inhibitors of b-tryptase.

likely to bind tightly, through optimal bivalent positioning of its two binding units, while a shorter dimer (146) might bind monovalently, due to its insu‰cient linker length. 4.4.3.2 Bivalent Inhibitors Tethered to Diketopiperazine. Schaschke et al. [590] reported bivalent inhibitors (149) designed on a sca¤old of a diketopiperazine (Figure 4.90). With trans-geometry and a rigid ring system, the diketopiperazine serves as a good template for displaying two linked ligands appropriately positioned toward proximal subsites. The dimers 149 show potent inhibitory activity against b-tryptase, with Ki values of 18 nM (n ¼ 1), 10 nM (n ¼ 2), and 350 nM (n ¼ 3). These activities represent 39- to 1370-fold enhancements relative to a relevant monomer control, such as an aminomethylsubstituted phenylalanine methyl ester, which has a Ki value of 13.7 mM. 4.4.3.3 Benzamidine-Based Divalent Inhibitors. Serine proteases of the trypsin family are inhibited by a class of molecules presenting an amidine moiety, an isosteric equivalent of the guanidine side chain in arginine [591]. However, because of the similarity of active sites among various enzymes within the trypsin class, monovalent amidines are nonselective. For example, benzamidine shows inhibitory activity against tryptase, bovine trypsin, and human plasmin with a Ki values of 22, 33, and >100 mM, respectively. Its selectivity to tryptase can, however, be enhanced by use of bivalent strategy in a system where tryptase exists as tetramer, a unique feature that distinguishes tryptase from monomeric proteases. Burgess et al. [592] reported the synthesis of divalent inhibitors designed to bind to two adjacent subsites within a tryptase tetramer (Figure 4.91). Dimer 150 is composed of two benzamidine inhibitors tethered with a flexible ˚ between two carbonyl groups in the alkane spacer that provides about 22 A fully extended conformation. Such a spacer distance is su‰cient for spanning

256

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS NC

i,ii

NH2

H2N

OH

O

O

O

O N H

iii NC

O

O

N H

iv CN

~22Å O

O N H

H2N NH

O

O 150

N H

NH2 NH

reagents and conditions: i) 1,5-pentanediol, PPh3, diisopropyl azodicarboxylate, THF, rt; ii) LiAlH4, THF, reflux; iii) 4-cyanobenzoyl chloride, K2CO3, H2O, dioxane; iv) NH4Cl, AlMe3, benzene, 145oC, 17-24 h.

Figure 4.91 Synthesis of bivalent benzamidine-based inhibitor of b-tryptase tethered with flexible linker.

two sites. This dimer is not only an extremely potent inhibitor of lung tryptase, with a Ki value below 0.01 nM, but is also highly selective to the tryptase compared to trypsin and plasmin, by factors of 100,000 and 450,000, respectively. As an example of how linker length influences dimer activity, the addition or subtraction of two carbons in the central (CH2 )5 chain in 150 leads to a 20- to 100-fold decrease in inhibitory activity. 4.4.4

Proteasomes

4.4.4.1 Multivalent Nature of Proteasomes. An enzyme responsible for degradation of misfolded peptides in the cytosol and the nucleus, the proteasome belongs to a class of multicatalytic proteases that display low substrate specificity and a nonselective cleavage pattern [593]. In mammals, it generates antigenic peptide fragments that are later presented on cell surfaces through MHC class I receptors, which are implicated in pathological inflammatory processes. The multimeric nature of this enzyme complex presents a unique macromolecular topology suited for simultaneous binding of multiple inhibitors. For example, the 20S proteasome from Saccharomyces cerevisiae consists overall of 28 subunits, a1–a7, b1–b7, a10 –a7 0 , b10 –b7 0 , which are arranged as a stack of four rings of heptameric complexes (Figure 4.92) [594]. Of these subunits, only the b1, b2, and b5 subunits from each b-ring serve as active proteases responsible for the proteolytic activities of the proteasome. The range of distances (28 ˚ ) between catalytic sites of the active b-subunits is as indicated in Figure to 64 A 4.92, suggesting that this enzyme complex is a valid target for a multivalent approach.

ENZYMES

257

C2 symmetry 1

1'

28Å

2 2'

43Å 64Å

49Å

5

5'

HO2C

O

H N

N H

O

O

H N

N H

O

O

nO

O

H N

N H

O

O

H N

N H

O

CO2H

151 (n = 19-25) HN

HN

NH2 NH

HO2C

O

H N

N H

O

NH

N H

O

NH H2 N

O

H N

NH H2 N

O

nO

O

H N O

N H

H N O

NH2 NH

O N H

CO2H

152 (n = 19-25)

NH

Figure 4.92 Heptameic ring arrangement of central b-subunits of 20S yeast proteasome and examples of peptide-based dimeric inhibitors.

4.4.4.2 PEG-Linked Tripeptide Dimers. Loidl et al. [595,596] described tripeptide-based divalent inhibitors (151 and 152) which are tethered through poly(ethylene glycol) (Figure 4.92). This linker is flexible and long enough to span two active b-subunits in the intra- and interring mode, which requires a ˚ . The hydrophobic peptide dimer 151 inhibits maximal distance of about 65 A the chymotrypsin-like activity of proteasome with an IC50 value of 17 nM, which is about 124-fold more active than the free tripeptide itself or a monovalent tripeptide-PEG. This dimer does not e¤ectively inhibit trypsinlike activity. In contrast, the arginine-containing peptide dimer 152 is a potent inhibitor of the trypsinlike activity of the proteasome, with an IC50 value of 71 nM. This activity is about 115-fold enhanced over that of either a monomeric peptide or peptide–PEG conjugate. The enhanced activity is attributed to divalent tight binding to adjacent b-subunits in which two of the six b-subunit sites are occupied by its tripeptide moieties backed by the x-ray crystal structure [595].

258

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS Unoccupied CRD

Bound DAG (or phorbol ester)

Bound divalent ligand membrane

cytoplasm

Protein kinase C (active)

Protein kinase C (resting) O

OH

OH

H

H OH

i

H

OH OHO

OH

H

O

OAc

OAc

H

OHO

O

(CH2)n O

OH

ii,iii

OH OHO

OAc

OAc

H

Phorbol

H

OH

OH

O OH

H OH

153 (n = 6, 10, 20) O

O

O

O OAc

OAc O

H

O OH

OHO

H

O O P O O-

NH2 CO2H

154 OH

reagents and conditions: i) Ac2O, Et3N, CH2Cl2, THF; ii) HO2C(CH2)nCO2H, Et3N, 2,4,6-trichlorobenzoyl chloride; then, phorbol ester, DMAP, toluene; iii) HClO4, MeOH.

Figure 4.93 Activation of protein kinase C presenting two adjacent cysteine-rich domains (CRD) by divalent phorbol ester.

4.4.5

Protein Kinase C

4.4.5.1 Divalent Nature of Protein Kinase C. Protein kinase C (PKC) belongs to the family of serine–threonine kinases that are involved in a diverse cellular regulation processes. This enzyme consists of multiple isozymes grouped by their activators and cofactors. For example, PKC isozymes a, b, g are activated by endogenous PKC activators such as phosphatidyl-l-serine and sn-1,2diacylglycerol (DAG), which bind to cysteine-rich domains located in regulatory sites (Figure 4.93). These isozymes are also stimulated by the high-a‰nity binding of a phorbol ester to cysteine-rich domains, except in the case of the PKC-g isozyme, which requires simultaneous occupation of two phorbol ester molecules in cysteine-rich domains [597]. Due to such an arrangement of cysteine-rich domains, dimers derived from phorbol esters can target two cysteinerich domains simultaneously in PKC-g and provide a mechanism that makes the dimer more selective to this isozyme. 4.4.5.2 Phorbol Dimers. Wender et al. at Stanford [598] reported phorbol dimers (153) tethered with a hydrophobic spacer of varying length (Figure

ENZYMES

259

4.93). The dimer 153 (n ¼ 10) shows potent binding ability, with a Ki value of 0.1 nM. It is thus 13-fold more active than a monovalent phorbol ester, phorbol-12,13-dibutyrate. Varying the linker length reduces the a‰nity, with Ki values of 4.5 nM (n ¼ 6) and 2.7 nM (n ¼ 20). These phorbol dimers promote PKC activity, as observed from the natural phorbol ester. In a parallel approach, Sodeoka et al. [599] reported the synthesis of a heterodimer (154) composed of a phorbol ester tethered to phosphatidyl-l-serine. It binds tightly to rat PKC isozymes but with about 10-fold weaker a‰nity than a monovalent phorbol (phorbol 12-myristate, 13-acetate). This heterodimer is a partial agonist of PKC. 4.4.6

Syk Protein Tyrosine Kinase

Activation of mast cells results in the release of inflammatory mediators and is regulated by the binding of the cytoplasmic domain of membrane-bound immunoglobulin E (FceRI receptor) to the syk protein tyrosine kinase [600,601]. This kinase is a multidomain protein composed of two tandemly located Src homology 2 (SH2) domains and a tyrosine kinase. Each of the SH2 domains recognizes a phosphotyrosine residue displayed in an immunoreceptor tyrosinebased activation motif (ITAM) of the g-chain of immunoglobulin E (IgE). A natural ligand to an SH2 domain includes a consensus sequence of pTyr-X-XLeu, which is expressed in duplicate copies within the ITAM domain. Thus, interaction of syk kinase and mast cell IgE is based on divalent binding. This concept is illustrated by the binding of a divalent epitope peptide ( pTyr-ThrGly-Leu-Asn-Thr-Arg-Ser-Gln-Glu-Thr-pTyr-Glu-Thr-Leu), which, with a Kd value of 37 nM, shows about a 5000-fold higher a‰nity than the monovalent peptide ( pTyr-Glu-Thr-Leu) [602–604]. A main role for the intervening peptide sequence between two phosphotyrosine ligands is to place the two ligands in proper spacing and orientation for simultaneous interaction with tandem SH2 domains. In keeping with this idea, Dekker et al. [605] designed synthetic dimers composed of two phosphotyrosine tetrapeptides which are tethered with a flexible hexa(ethylene glycol) spacer or rigid propynylphenyl spacer. A dimer containing the rigid spacer shows a binding a‰nity to syk tandem SH2 domains as tight as that of natural divalent phosphotyrosine peptide, with a Kd value of 29 nM, while an oligo(ethylene glycol)-linked dimer shows about a 10-fold lower a‰nity. 4.4.7

Protein Tyrosine Phosphatases

4.4.7.1 Phosphophenol Dimers. Whereas protein tyrosine kinases catalyze the phosphorylation of tyrosine residues, protein tyrosine phosphatases (PTPases) catalyze the opposite action, removal of the phosphoryl group from phosphotyrosine residues in a protein substrate. An x-ray crystal structure of PTPase 1B shows the existence of two phosphotyrosine-binding sites located proximally, one as a high-a‰nity catalytic site and another as a noncatalytic low-a‰nity

260

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS +H

2-O

3N

CO2OPO32-

2-O PO 3

3PO

155 bis(p-phosphophenyl)methane

Phosphotyrosine

CO2H O diaryloxamic acid (ligand for catalytic site)

Diaryloxamic acid-derived dimer OH

N O

O

O

HO N H

AcHN

site 2 ligand O

linker

OH

HO

O

O O

CH3

O

156 (RK-682)

OH

HO

O

O

O

O

HO

157 (tail to tail)

OH

CH3 O

O

O

O

OH

HO

O

O O

O

O O

O OH

158 (head to head)

CH3

O O

O

Figure 4.94 Phosphotyrosine-mimicking divalent substrate of protein tyrosine phosphatase 1B and examples of divalent tetronic acid as inhibitor of ERK phosphatase.

site [606]. For example, O-phosphorylated tyrosine binds to this enzyme with a K m value of 4.9 mM at the active sites. It also binds to lower-a‰nity sites on the same enzyme (Figure 4.94). However, bis( p-phosphophenyl)methane (155), a symmetric dimer derived from phosphophenol, binds to PTPase 1B equally well at both sites, and with enhanced a‰nity, as reflected by the K m value of 16 mM. The added binding contribution of this symmetric binder enhances the a‰nity 300-fold. The use of such dual sites at PTPase 1B in drug design was further elaborated by Puius et al., Szczepankiewicz et al., and Liu et al. at Abbott Laboratories [606]. Their approach is based on a linker-fragment strategy using an NMR-based screening method. With this method, they have identified a heterodimer that acts as a selective inhibitor of PTPase 1B, with a Ki of 22 nM (Figure 4.94). The heterodimer is composed of diaryloxamic acid as a ligand for a primary catalytic site tethered with a pentane chain to naphthylene car-

ENZYMES

261

boxylic acid as a site 2 binder. This dimeric ligand is more active than either diaryloxamic acid or the diaryloxamic acid linker, by factors of about 1800 and 50, respectively. 4.4.7.2 Divalent Tetronic Acids. Dual-specificity PTPases catalyze the hydrolysis of monophosphoesters, such as phosphotyrosine, phosphothreonine, and phosphoserine, that are present in protein substrates [607]. Vaccinia VH1related phosphatase (VHR) belongs to this phosphatase family and serves as a regulator of extracellular regulated kinases (ERKs) [608]. VHR is inhibited by 156 (RK-682) at an IC50 value of 11.6 mM (Figure 4.94). Such inhibition occurs in a competitive manner in which the acidic 3-acetyltetronic acid of 156 interacts as a phosphate mimetic with an active site loop. The VHR enzyme appears to present two recognition sites for 156, as suggested from binding kinetics reported by Usui et al. [609]. These two unique proximally located sites are composed of an active site and a noncatalytic site. Such proposed active-site structures make it possible to design divalent tetronic acids using tail-to-tail (157) or head-to-head (158) linkage. These dimers inhibit VHR in a 1:1 binding stoichiometry, at IC50 values of 12 mM (157) and 1.83 mM (158). Thus, ligand dimerization such as in 158 leads to about a sixfold enhancement in activity compared to 156 as a monomer. 4.4.8

Human Leukocyte Elastase

Human leukocyte elastase (HLE), a serine protease localized in polynuclear leukocytes, is involved in the degradation of phagocytized proteins [610]. It is composed of highly basic amino acids, including 18 arginine residues on its surface, although these are partially neutralized by six neighboring acidic residues. Four arginines are adjacent to the entrance of an active site over a length ˚ . The positive charge of such a protein surface enables a spanning about 40 A multiply charged anionic substrate to adhere to the surface through a charge– charge interaction. Steric interference can be invoked to preclude access of the substrate to the active site and thus to inhibit binding at the catalytic pocket entrance. Regan et al. [611] described linear oligomers composed of aromatic anions, such as 159 and 160. These molecules are expected to generate multiple electrostatic contacts with HLE surfaces (Figure 4.95). Oligomers (159) inhibit this enzyme noncompetitively, with Ki values of 8300 nM (n ¼ 2), 130 nM (n ¼ 3), and 120 nM (n ¼ 4). A second multivalent molecule (160), which contains a longer space between adjacent anions, shows higher activity, with a Ki value of 20 nM. The activity is thus influenced by the distance and valency of displayed anions. 4.4.9

Thrombin

Blood coagulates as a consequence of a complex cascade process mediated by multiple interacting factors, such as thrombin and factor Xa [612]. These fac-

262


CO2H O

CO2H O

O OH

OH

n

O

159 (n = 2, 3, 4) O

CO2H

CO2H

CO2H

CO2H

O

OMe

O

OMe

O

OMe

OMe

O

OMe

O

OMe

O

Me

160

Figure 4.95 Structure of anionic diphenylmethane oligomers as inhibitor of human leukocyte elastase.

tors represent two serine proteases involved in downstream events of the process. Anti-thrombin III (AT-III) blocks coagulation by forming a complex with these proteases. The inactivation rate is accelerated by several orders of magnitude when AT-III is bound to negatively charged natural polysaccharides, such as heparin and heparan sulfate, which are composed of 20 to 100 repeats of sulfated carbohydrate units [613]. According to a model proposed by Van Boeckel and Petitou [614], a catalytic contribution from heparin molecules stems from its ability to form a ternary complex with AT-III and thrombin (Figure 4.96). Such complex formation is mediated by three structural features of heparin: (1) a pentasaccharide domain that binds tightly to arginine- and lysine-rich regions of AT-III, (2) a long spacer domain, and (3) a second binding motif to thrombin. This mode of heparin binding suggests that synthetic coagulation inhibitors can be designed by tethering two distinct binding groups through a long, flexible linker. For example, the heterobivalent molecule 161 is composed of a pentasaccharide domain for AT-111 binding and a trisaccharide region for thrombin recognition separated by a 53-atom spacer [615,616]. In contrast, the homobivalent molecule 162 possesses two identical pentasaccharides that bind to both AT-III and thrombin. These synthetic bivalent molecules are as active as heparin in AT-III-mediated inhibition of thrombin and are far better than a monovalent pentasaccharide molecule, which is virtually inactive. 4.4.10

Glucoamylase

4.4.10.1 Acarbose Linked to b-Cyclodextrin. Hydrolytic cleavage of b-dglucose from starch and related polysaccharides is catalyzed collectively by

ENZYMES

263

Multivalent sulfated saccharide (heparin)

- -

- -

-

Antithrombin III (AT-III)

-

Inactivation of thrombin

Thrombin

AT-III-binding domain OSO3OSO3OSO3O OMe O OMe

161 O

O

O O OOSO3-O2C OMe O OSO3OSO3-

CO2O O OMe

H N

OSO3-

H N

S

3

O O3SO

OSO3 OSO3O OSO3-

O OSO3-

O

O

CO2O O OMe H N

OO O OSO3-O2C OMe O OSO3OSO3-

-

O3SO

O OSO3-

OMe OSO3-

O O

3SO

O OSO3-

3

S S

O

3

OSO3N H

O3SO

O OSO3-

O OSO3OSO3-

OSO3-

O O

-

O OSO3-

OSO3-

162 O

O

5

Thrombin-binding domain

OSO3-

3

-O

O OSO3-

-

O

N H

O -

OSO3-

OMe OSO3O

O

O

3

O OSO3-

O

O OSO3-

7

O OSO3-

O

CO2O O OMe

OO O OSO3-O2C OMe O OSO3OSO3-

O OSO3-

OMe OSO3-

OSO3-

O

Figure 4.96 Antithrombin III–mediated inactivation of thrombin enhanced by multivalent sulfated carbohydrates such as heparin and examples of synthetic mimics.

multiple enzymes. Among these, glucoamylase is a glucohydrolase that acts at nonreducing ends [617,618]. Glucoamylase from Aspergillus niger presents two distinct receptor domains that perform substrate recognition: a catalytic site and a starch-binding domain. The two sites are separated spatially by an Oglycosylated, flexible peptide spacer. The optimal performance of this hydrolytic enzyme depends on the cooperative interaction of two binding domains with substrate. Payre et al. [619] designed a bivalent ligand (163), which consists of acarbose as a potent active-site inhibitor of glucoamylase (with a Kd value of 1012 M ) and a b-cyclodextrin moiety, which mimics the amylose substrate with its ability to bind to the starch-binding domain (Figure 4.97). To allow the binding motifs to span the two relevant domains within glucoamylase, these two binding motifs are tethered through a flexible spacer that varies ˚ , as estimated in the extended conformation. Divalent in length from 14 to 73 A

264

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS HO

H 3C

OH O

O

OH O

OH

HO HO HO N OH OHO OH OHO OH OH H Acarbose

HO

H3C

O

OH O

OH O

HO HO HO N OH OHO OH OHO OH OH H

163 (n = 2, 3; m =1, 2, 4)

S

O

S

OH O O HO OH HO O OH O O HO OH OH HO HO O OH β-Cyclodextrin O O OH OH HO O OH HO OH HO O O O HO n m

O

O HO

Figure 4.97 Structure of a bivalent ligand to glucoamylase.

substrate 163 shows Kd values indicating an a‰nity lowered by four orders of magnitude relative to that of acarbose: 5 108 M (m ¼ 0); 1 108 M (m ¼ 4, n ¼ 3) [620]. The lowered association strength is attributed to an unfavorable entropic cost originating from steric hindrance between b-cyclodextrin and the starch-binding domain since this bivalent binding does not cause any loss in binding enthalpy. 4.4.11

Glycosyltransferase

4.4.11.1 Divalent Inhibitors of a1,3-Galactosyltransferase. a1,3-Galactosyltransferase (a1,3-GalTase) catalyzes the transfer of a galactosyl moiety in the a1,3 configuration to galactoside terminated in oligosaccharide. The active site of this enzyme comprises two proximal domains designed to recognize both uridine diphosphate (UDP) and galactoside as substrates (Figure 4.98). The catalysis involves the formation of a glycosyl–enzyme complex, which is followed by transfer of a galactoside while retaining the configuration at its anomeric center. This model of enzyme catalysis supports using a multivalent approach in the design of a1,3-GalTase inhibitors. Kim et al. [621] tested this concept by designing a synthetic divalent molecule (164) composed of an Niminosugar–based galactoside mimic and a uridine moiety, which are linked with a hydrophilic spacer as a replacement for pyrophosphate. This bivalent molecule is a potent and selective inhibitor of a1,3-GalTase, with a Ki value of 4.4 mM, whereas it is inactive against b1,4-galactosyltransferase. Free monomers are less active against a1,3-GalTase, with Ki values of 55 mM (165) and 5 mM (166). 4.4.11.2 Divalent Inhibitors of b1,4-Galactosyltransferase. Unlike a1,3GalTase, b1,4-GalTase catalyzes the transfer of a galactoside to an acceptor

265

ENZYMES UDP-galactoside O

O

O-

O

OH OH O HO O O HO O P O P O O- O-

OH OH O HO HO

NH O

N O

-UDP

O OH OH O

HO

OR

HO OH OH

α1,3-GalTase

Gal transfer at α-configuration bivalent inhibitor O O

OOH OH O HO N spacer HO

NH O

N O

OH OH

O OH OH O HO N HO

NH

OH S OH

O

N O

O OH OH O HO N HO

OH

H2N

165

164

S OH

OH

OH OH

NH

OH

OH

O

N O

OH OH 166

Figure 4.98 Mode of binding of a1,3-galatosyltransferase with bivalent inhibitor.

molecule such as GlcNAc while retaining the b1,4-configuration (Figure 4.99). In addition, this catalytic reaction is based on a di¤erent mechanism of transfer reaction, in which an acceptor sugar reacts with UDP-galactoside in an SN 2like displacement mode, which inverts the molecular configuration. Hashimoto et al. [622] constructed a bivalent molecule (167) composed of two substrates tethered through a short spacer. It is a competitive inhibitor of b1,4-GalTase with Ki values of 1.35 mM against GlcNAc and 3.3 mM against UDP-Gal. This

SN-2-like transition state AcHN OR O HO HO O OH OH HO O NH HO O O N O HO O P O P O O O- OOH OH β1,4-GalTase

AcHN OH OH HO OMe O O HO O HO O O NH O O N O O O P O P O O- OOH OH 167

Figure 4.99 SN -2-like transfer model for b1,4-galactosyltransferase.

266

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS SN2-like transition state OH HO

OH

OH

168

O OPh GDP-fucose OH

O O H 3C P P O O O O OH HO -O OOH

O

HO

O N N

NH N

NH2

OH 169

OH OH

HO

α1,2-Fucosyltransferase

NeuAc-CMP

NH2

O -O P O N O O O AcHN H O CO2H OH HO OH HO OH

α-2,3 or α-2,6

OH HO O HO O HO OH LacNAc

O OPh O O O P P O O O- OOH

O N O

N

NH N

NH2

OH OH

O OPh O O P OO-

NH2

N

HO

OH

N O -O P O N O O O AcHN H O CO2H OH HO OH HO S OH (CH2)n S O HO O HO OH Divalent inhibitor of sialyltransferase (n = 1 to 5) HO

NHAc O OMe OH

OH

NHAc O OMe NHAc

Figure 4.100 SN 2-like transition state for glycosylation catalyzed by glycosyltransferase such as a1,2-fucosyltransferase and sialyltransferase, and examples of their bivalent inhibitor.

competitive inhibition against donor and acceptor substrates also implies bivalent occupation at the two substrate sites. 4.4.11.3 Divalent Inhibitors of a1,2-Fucosyltransferase. Another related example of glycosylation is observed from a1,2-fucosyltransferase, which catalyzes the transfer of an a-fucosyl moiety to the 2-OH of b-d-galactoside [623]. This enzyme recognizes guanosine 5 0 -diphosphofucose (GDP-fucose) as its substrate. Like b1,4-GalTase, fucosylation by a1,2-fucosyltransferase occurs through an SN 2-like displacement mechanism with inversion of the molecular configuration at the anomeric carbon of the fucosyl residue (Figure 4.100). On the basis of this proposed mechanism, Palcic et al. [623] designed a bivalent inhibitor (168) which is composed of a b-d-galactoside acceptor linked to guanosine diphosphate via a short linker. This dimer inhibits the membrane-bound form of a1,2-fucosyltransferase as a competitive inhibitor, with Ki values of 2.3 mM against phenyl galactoside and 16 mM against GDP-Fuc. This increased inhibitory activity represents about a 50-fold enhancement over a monomeric molecule (169) against free substrates. 4.4.11.4 Divalent Inhibitors of Sialyltransferase. The bivalent substrate approach is also applicable for designing inhibitors of sialyltransferase. Invok-

ENZYMES

267

ing a bivalent strategy, Hinou et al. [624] designed molecules composed of cytidine monophosphatesialic acid (CMP-NeuAc) linked to N-acetyllactosamine (LacNAc) through a one- to five-atom spacer (Figure 4.100). That design reflects the configuration of two substrates, CMP-NeuAc and LacNAc, bound at the active-site domains of sialyltransferase, which catalyzes the transfer of a donor sialyl group to an acceptor galactoside terminus with an a2,3 or a2,6 linkage [624]. The dimeric molecules potently inhibit sialyltransferase activity by occupying both the donor and acceptor sites. In particular, dimers with a spacer length of 4 units are about 40- to 80-fold more active at acceptor sites than is LacNAc.

4.4.12

Riboflavin Synthase

4.4.12.1 Lumazine Dimers. The biosynthesis of riboflavin occurs through dismutation between two identical 6,7-dimethyl-8-(d-ribityl)lumazines catalyzed by riboflavin synthase. This reaction involves the transfer of a four-carbon unit from a lumazine molecule bound at the donor site of the enzyme to a second molecule bound at the acceptor site, thus forming riboflavin and pyrimidinedione (Figure 4.101). The presence of two distinct substrate-binding sites within the enzyme supports the application of a bivalent approach in the rational design of inhibitors. Cushman et al. [625] designed bivalent lumazine molecules (170) composed of two identical lumazine molecules tethered at the N3 position with a C3 –C5 alkane chain spacer. The short linker length is chosen on the basis of mechanistic considerations of the dismutation reaction, which requires close proximity of the two bound lumazine molecules, possibly via a stacked, opposing arrangement [625]. This bivalent lumazine molecule (170; n ¼ 4) shows inhibitory activity against riboflavin synthase, with a Ki value of 37 mM. Its potency is about twofold greater than that of the monomer 6,7-dimethyl-8(d-ribityl)lumazine. The inhibitory activity of the bivalent lumazine is dependent on linker length, as illustrated by the reduced activity of the other two dimers.

4.4.13

Lactate Dehydrogenase

4.4.13.1 Enzyme Oligomerization. Many enzymes require oligomerization for optimal catalytic activity, and some of them are active only in the aggregated state [626]. Representative examples of such enzymes include l-lactate dehydrogenase, fumarase, and sorbitol dehydrogenase, which are aggregated as tetramers, as well as citrate synthase, d-galactose dehydrogenase, and adenosine deaminse, which form dimers. This protein multimerization is often driven by hydrophobic adhesion occurring between hydrophobic patches on the enzyme surface, as reviewed by Zutshi et al. [626] (Figure 4.102). Maletic et al. and Breslow et al. at Columbia [627] demonstrated that molecules capable of interfering with this hydrophobic interaction serve as potential inhibitors of

268


N

H 3C

2

H 3C

HO

OH riboflavin synthase

OH N O

H3C

NH

N

N

H3C

NH

H2 N O

riboflavin

N

CH3

N

CH3 Nu

H 2C

N

H 3C

N

OH HO

HO

O

HO

N

+

NH

OH OH H N O

acceptor site OH

donor site

N H HO

OH

O

6,7-dimethyl-8-(D-ribityl)lumazine

HN

HO

OH N O

N

O

O

OH

OH

OH HO

O

OH N O

NH2 CH3

N

HN O

NH O

OH

OH

OH

N H HO

HO

HO

N

CH3 H

OH N O NH

N O

OH

HO riboflavin synthase OH HO

H 3C

N

OH OH N O

170 (n = 3, 4, 5) H 3C

N

N O

O (CH2)n

N O

N HO

HO

N

CH3

N

CH3

OH

HO

Figure 4.101 Mechanism of riboflavin synthase–catalyzed conversion of two lumazine molecules to riboflavin.

multimeric enzymes. They include dimeric molecules derived from b-cyclodextrin (b-CD), which contains a large hydrophobic pocket that selectively binds hydrophobic and aromatic side chains from polypeptides. For example, a monovalent b-CD binds l-Phe-d-Pro, with a Kd value of 180 M 1 . When this monomer is converted to a bivalent b-CD, the dimer binds with higher a‰nity to aromatic peptides. Further increases in valency continue to enhance the affinity to a multivalent, hydrophobic peptide, as illustrated by a cyclodextrin trimer. The trimer enables tritopic binding, which is tighter than the ditopic binding supplied by the cyclodextrin dimer [628]. 4.4.13.2 Cyclodextrin Dimers. Leung et al. [628] investigated the specific peptide binding ability of bivalent cyclodextrin constructs in an e¤ort to find bCD dimers targeting hydrophobic patches present in enzyme proteins. Dimers

ENZYMES

269

dimerization bivalent β-CD inactive enzyme OH

O HO

O

OH

O OHHO O OH

O

OH

inactive enzyme

active enzyme

HO

O HO

OH

O

O N

O

OH

O OHHO O OH O

OH

O HO

O X X HO OH hydrophobic OH O O OH HO cavity O OH HO O OH O O OH O O OH OH HO O OH OH HO OH O O OH OHHO HO O OH O O O OH O HO 171 (X = O, NH) HO

Figure 4.102 Disruption of protein–protein dimerization by cyclodextrin dimer.

(171) are composed of two identical b-CDs tethered together with a rigid pyridyl group. These bivalent b-CDs block dimerization of lactate dehydrogenase and citrate synthase in a concentration-dependent manner. The ability of 171 to interfere with enzyme dimerization is inferred to result from its ability to bind the protein–protein interface. For example, dimer 171 (X ¼ O) blocks the catalytic activity of two enzymes, with IC50 values of 140 mM for citrate synthase and 30 mM for lactate dehydrogenase. Structurally, lactate dehydrogenase contains a number of hydrophobic residues, such as Tyr and Trp, exposed on its dimerization surface. It is noteworthy that the ability of inhibitory molecules to bind to the dimer interface is linked directly to their ability to inhibit enzyme activity. In addition to inhibiting protein oligomerization, cyclodextrin dimers reportedly supply a diverse range of important biological applications, such as mimetics of hydrolytic enzyme catalysis, inducers of protein folding, and carriers of porphyrin-based photosensitizers in photodynamic tumor therapy [19,629]. 4.4.14

Zinc-Dependent Enzymes

4.4.14.1 Histone Deacetylase. Multivalency accounts for the improved activity displayed by a certain class of inhibitors that target zinc-dependent enzymes. The basic design components of these enzyme inhibitors include an active-site binder linked to a component that binds to a peripheral site (Figure 4.103). For instance, Finnin et al. and Richon et al. [630] reported the x-ray structure of suberoylanilide hydroxamic acid (SAHA) bound at histone deacetylase (HDAC), an enzyme that catalyzes the removal of an N-acetyl moiety

270

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS zinc-coordination domain O

H N

O N H

O

N H

OH

trichostatin A

O H2N S O O

CH3 CH3

N

SAHA

zinc-coordination domain

O

OH

N H

O

O

variable secondary group

H N

NH2 O

Figure 4.103 Structure of bivalent inhibitors of zinc-dependent enzymes, histone deacetylase, and carbonic anhydrase.

derived from an e-amino group of a lysine in the histone protein. The SAHA inhibitor binds into a tube-shaped pocket of the enzyme active site. Its hydroxamic acid unit coordinates the zinc metal at the bottom of the pocket while its aliphatic spacer chains and anilide capping group make multiple contacts within the hydrophobic tube and pocket entrance, respectively. Trichostatin A, another HDAC inhibitor, prevents cell growth and as such is used as an antitumor agent. It binds the inside of the HDAC enzyme pocket in a similar manner. 4.4.14.2 Carbonic Anhydrase. A second example of bivalent inhibitors of zinc enzymes is reported by Boriack et al. [631] in their design of nanomolar inhibitors targeting human carbonic anhydrase II (CAII), which catalyzes the generation of bicarbonate in aqueous humor (Figure 4.103). Members of this series of inhibitors are composed of benzenesulfonamide to coordinate zinc in the enzyme active site. The benzenesulfonamide moiety is tethered by a tri(ethylene glycol) spacer to an amino acid, such as phenylalanine or leucine, as secondary binding groups. The x-ray structure of the enzyme–inhibitor complex indicates multiple interactions, including sulfonamide–zinc coordination, association of the ethylene glycol linker with a hydrophobic patch in the active site and binding of a pendant amino acid to the enzyme surface.

4.5 4.5.1

G-PROTEIN-COUPLED RECEPTORS b-Adrenergic Receptors

The G-protein-coupled receptors (GPCRs) includes the b-adrenergic receptor (bAR) family, which helps regulate the pulmonary and circulatory systems [3].

G-PROTEIN-COUPLED RECEPTORS

271

This receptor family is composed of at least three subtypes, including b1 -AR, b2 -AR, and b3 -AR. Members of this class of receptors are activated by endogenous catecholamines, such as epinephrine, and synthetic ligands that include nonselective isoproterenol and albuterol, which is marketed as a drug for the treatment of asthma. The stimulatory action by b2 -AR induces dilation of bronchial smooth muscles and vasodilation of peripheral blood vessels, whereas stimulation by b1 -AR leads to an increase in pulmonary blood pressure. Thus, targeting a specific subtype of b-AR would be desired for eliciting or inhibiting specific physiological actions. 4.5.1.1 Divalent Ligands of b-AR. Two main groups of bivalent ligands bind to GPCRs. Members of the interreceptor group can bridge two receptors while members of the intrareceptor group target one receptor by occupying simultaneously a primary ligand site and a proximal nonligand site (Figure 4.104). It is possible that bivalent ligands would enhance potency and receptor selectivity

bivalent binding within a receptor

bivalent bridging between two receptors

bivalent ligand

membrane

receptor

HO

OH H N

CH3 CH3

HO

OH H N

HO

N H

HO

Isoproterenol

HO

OH

OH H N

HO

OH OH

172 (Hexoprenaline)

H 3C H3C

CH3 CH3 CH3

Albuterol (salbutamol)

OH

H N

OH

CH3

OH OH H N

OH

HO OH

CH3 H 3C

N H

CH3 CH3 CH3

Albuterol dimer OH N N OH (NH2)

173 (χisoproterenol = 2.5%)

H N

HN

O N H m O

OH

H N O HN

O N nH

O

O

OH

Figure 4.104 Bivalent binding to membrane-bound receptor and structure of isoproterenol, albuterol, and their multivalent ligand to b-adrenergic receptor.

272


via bivalent binding to the same or homologous sites. For example, hexoprenaline (172) is a symmetric bivalent ligand of b-AR. It is composed of two catecholamines tethered through a relatively short hexane linker. This bivalent agonist shows tissue selectivity with a greater e¤ect on peripheral vascular smooth muscle than on heart muscle, and stimulation by this agonist leads to a longer duration of action in hemodynamic e¤ects than that of isoproterenol [632]. The increased selectivity and activity duration probably derive from hexoprenaline’s ability to bind bivalently to b-AR. As a second example, a homodimer derived from albuterol was synthesized and its structure reported by Haddad et al. at Boehringer [632]. This dimer is generated as a side product during synthesis or upon long-term storage of albuterol. Its pharmacological activity needs to be evaluated. 4.5.1.2 Isoproterenol Presented on Polypeptides. Isoproterenol can be converted to a multivalent form displayed on a polymer support, as reported by Verlander et al. [633]. One such example is 173, which presents multiple copies of isoproterenol. The isopreterenol moieties are attached through diazo linkage to a polypeptide sca¤old composed of phenylalanine and gluatamine derivatives. Like isoproterenol, this polymeric polyvalent isoproterenol actively elicits positive chronotropic responses in heart tissue, suggesting that the ligands tethered to the polymer sca¤old can still interact with b-ARs present on cell surfaces. However, e¤ective dose levels are one to two orders of magnitude lower than that displayed by free isoproterenol. 4.5.1.3 Practolol Dimers. Practolol, a member of the aryloxypropanolamine class of b-AR ligands, is a selective b1 -AR antagonist (Figure 4.105). Kizuka and Hanson [634] synthesized dimers (174) composed of two practolol pharmacophores linked with an alkane spacer of short-to-medium length. The dimers bind with di¤erent a‰nities to di¤erent tissues, possibly through binding to b1 -AR expressed in varying levels, depending on tissues. For example, the dimer 174 (n ¼ 8) shows tighter binding to heart tissue, with a K d value of 0.1 mM, than to lung tissue, with a K d value of 1.4 mM. These dimer a‰nities are enhanced 62- and 26-fold in respective tissues over those of monovalent practolol. Shortening the linker length from eight to two atoms decreases the binding a‰nity in tissue. 4.5.1.4 Propranolol Dimers. Propranolol is an example of b2 -selective adrenergic antagonists. Kierstead et al. [635] reported the synthesis of divalent ligands derived from propranolol, which include 175 and 176 (Figure 4.106). These dimers, which vary in linker length and composition, show activity similar to that of a b2 -AR blocker, with an EC50 value of about 0.04 nM. This activity is enhanced about 14-fold over that of monovalent propranolol. In addition, they bind more tightly to b2 -AR than to b1 -AR, with a selectivity ratio of about 15,000 (175) and 1400 (176). These dimers are more selective than free propranolol, which shows only a 23-fold selectivity ratio. These results thus


273

CH3 H 3C

N H

O OH

Practolol

HN O

CH3 H3C

NH

CH3

O

O

i,ii

H3C

+

N

O OH

HN NH2

O

CH3

CH3 H 3C

N H

O

O

N OH H

OH

CH3

iii,iv (CH2)n

HN O

NH O

174 (n = 2, 4, 6, 8) reagents and conditions: i) MeOH, reflux; ii) HCl, EtOH, reflux; iii) bis(acid chloride), Et3N, CH2Cl2; iv) H2, 5% Pd/C, MeOH, HOAc, rt.

Figure 4.105 Synthesis of bivalent practolol analogs as b-adrenergic antagonist.

support the belief that bivalent antagonists enhance selectivity as well as potency. 4.5.1.5 Alprenolol Displayed on Dextran. Alprenolol is also a potent antagonist of b-ARs (Figure 4.107). Pitha et al. [636] invented an alprenolol-derived, multivalent, macromolecular construct that interacts with surface receptors.

CH3 H3C

N H

O OH

Propranolol

CH3

CH3 H3C

N H

O OH

O

N H

O

N OH H

CH3

175

CH3 H 3C

O

OH

CH3 O

O OH

O

O

O

O

O

N OH H

CH3

176

Figure 4.106

Synthesis of bivalent aryloxypropanolamine as b-adrenergic antagonist.

274

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH NH(CH3)2

O

Alprenolol

OH O

NH(CH3)2 OH S

Dextran

O

O OH

O n

177 (14-atom spacer)

OH

OH NH(CH3)2

O

NH(CH3)2

O

O S

N H NHAc

178 (9-atom spacer)

OH O

S n

O n

179 (5-atom spacer)

Figure 4.107 Structure of polyvalent alprenolol as b-adrenergic antagonist.

For example, ligands 177 to 179 present multiple copies of alprenolol ligands tethered to a dextran matrix through a hydrophilic linker of variable length. Despite its large size, these macromolecules bind to b-ARs expressed on erythrocyte membranes. Higher a‰nity is observed from multivalent ligands containing relatively long spacers, such as 177, with a Kd value of 28 nM; lower a‰nities are displayed by those ligands designed with a shorter spacer, such as 178 (1.4 mM ) and 179 (20 mM ). However, even the best a‰nity is about 11fold lower relative to that of the alprenolol monomer. Taken together, these data imply that receptor sites are rather constricted, and that the ligand, when tethered to a macromolecular sca¤old, experiences only limited access to the receptor sites. 4.5.2

a-Adrenergic Receptors

4.5.2.1 Prazocin Dimers. Inhibition of a1 -adrenergic receptors (a1 -ARs) has been considered for the treatment of benign prostatic hyperplasia [637]. A diverse collection of structurally unrelated molecules interact with this receptor. Among these, prazocinin is a potent a1 -AR antagonist with moderate selectivity. In particular, high-a‰nity selective ligands that are able to distinguish the three a1 -AR subtypes, a1A , a1B , a1D , have contributed greatly to elucidating the relationship between subtype selectivity and ligand structure. Bolognesi et al. [638] reported bivalent ligands based on prazocin in an e¤ort to identify selective ligands (Figure 4.108). Dimers 180 and 181 are composed of two identical prazocin analogs linked with a polyamine linker. The 22-atom linker is attached at a di¤erent site in the two dimers. One dimer (180) shows high antag-


275

NH2 H3CO H3CO

N N

Prazocin

N N

O O

NH2 H3CO H3CO

N

22-atom spacer N

N

H N (CH2)6

N O

3

3 N (CH2)6 H

N (CH2)6 H

O

N H

N N

N N

180

OCH3 OCH3

NH2 NH2 H3CO H3CO

O

N N

4 N (CH2)6 H

N N O

Figure 4.108

N (CH2)6 H

N (CH2)6 H

N

H N

N

N 4

181

N

OCH3 OCH3

NH2

Divalent antagonist of a-adrenergic receptor.

onistic activity to a1D -AR, with a Kb (a constant indicative of antagonistic activity) value of 109:3 , although it is slightly less potent than prazocin, whose Kb value is 109:71 . But the approximately 100-fold selectivity of 180 for a1D AR over a1A -AR is better than the approximately 10-fold selectivity provided by prazocin. The second dimer 181, linked at a di¤erent attachment site, is less active and less selective than prazocin. 4.5.2.2 Yohimbine Dimers. Ligand dimerization proves an e¤ective strategy for identifying subtype-selective ligands of the a2 -AR. Yohimbine is a potent a2-AR antagonist, with a Ki value of 0.42 nM and about a fivefold selectivity for a2a -AR over the a2b -AR. Coupling of yohimbinic acid with a;o-diaminoalkane through an amide linkage generates a series of yohimbine dimers (182; n ¼ 2 to 12) (Figure 4.109) [639]. A dimer with a (CH2 )6 linker shows greater receptor subtype selectivity. This molecule preferentially binds a2a -AR over a2b -AR by a factor of 123. However, its a‰nity to a2a -AR, with a Ki of 1.35 nM, is slightly lower than that of yohimbine. 4.5.3

Opioid Receptors

The opioid receptors comprise a family of GPCRs belonging to the rhodopsin superfamily. The opioid receptors are activated by endogenous ligands that include morphine, opioid peptides, and other synthetic molecules [640]. This family of receptors is further divided into three subtypes: d-, k-, and mreceptors. The opioid receptors undergo dimerization on cell surfaces. For example, receptors present on the surface of neuroblastoma cells form receptor clusters upon ligand binding [641,642]. The distance between receptors in the

276


HN H O

N

H

MeO

H

HO Yohimbine

HN

HN H O HO

H

N H

HO Yohimbinic acid

+

H2N

(CH2)n

DCC

H

HOBt

N

NH2

OH H N (CH2)n H

O

H

H O N H HO

H

N H

NH 182 (n = 2 to 12)

Figure 4.109 Synthesis of bivalent yohimbines.

dimers varies, depending on the receptor subtype. In this context, a bivalent ligand would be expected to show high selectivity to a particular subtype of opioid receptor as a function of bridging distance. 4.5.3.1 Oxymorphamine Dimers. Portoghese [643], Portoghese et al. [644], and Erez et al. [645] extensively investigated the divalent strategy by designing a series of ligands selective for specific opioid receptors (Figure 4.110). For example, bivalent ligands (183) are composed of two biologically active ()oxymorphamine enantiomers tethered with an oligo(glycine) spacer. These dimers show agonistic activity in m-receptors. In particular, 183 with n ¼ 2 is the most active among this series of dimers. It is 36-fold more potent than free morphine and about 19-fold more active than a monomer control (185). The binding data suggest that the length of spacer at n ¼ 2 is close to an optimal distance, which enables this bivalent ligand to bridge two discrete recognition sites, possibly located on two neighboring m-receptors. The proposed mechanism of bivalent bridging is further supported by a mesomeric dimer (184; n ¼ 2) composed of an active isomer () and an inactive isomer (þ) tethered together with an oligo(glycine) spacer. The binding activity of this heterodimer is only 7.6-fold higher than that of morphine. While the heterodimer’s activity is lower than that of the homodimer described above (183; n ¼ 2), these two dimers are identical, except in the optical nature of the ligands. 4.5.3.2 Bivalent Bridging. Bivalent bridging requires the dimer spacer to be long enough to span the gap between two opioid recognition sites displayed by dimerized receptors (Figure 4.110). As shown in the figure, that distance is es˚ if the receptor dimer is formed at the transmembrane timated as about 32 A ˚ if it is formed by interlocking, condomain (TM) 4,5 interface, or about 27 A tact dimerization at the TM5,6 interface [643]. Given that the spacer length in

G-PROTEIN-COUPLED RECEPTORS (-)-oxymorphone

(-)-oxymorphone

N CH3

O O

H3 C N

183 (n = 1, 2, 3, 4)

OH

HO

H N

N H

HO

O n

N H O n

N H

O

OH

H3C N

N CH3 184 (n = 2)

OH O O

O

(+)-oxymorphone

(-)-oxymorphone

HO

277

H N

N H

HO

O n O

H N

N H

O

OH

O n

(-)-oxymorphone N CH3 OH

185 O

O

HO

H N

N H

NHAc O

~32 Å

~27 Å 11

6 7

1 2

5 3

4

3

4 5

2 7

6

TM4,5-interfaced opioid receptor dimer

2 1

7

3 4

6

5

5

6

4

3 1

2

7

ligand site TM5,6-interfaced opioid receptor dimer

Figure 4.110 Structure of bivalent oxymorphamine ligands and a schematic for twodimensional arrangement of seven TM subunits from dimerized opioid receptors at either a TM4,5-interface or a TM5,6-interface.

˚ , it is more likely that bivalent binding by 185 occurs through 183 is about 22 A interaction with receptor dimers formed at the TM5,6 interface. 4.5.4

Serotonin Receptors

Serotonin [5-hydroxytryptamine (5-HT)] receptors play a diverse array of physiological roles and are implicated in the treatment of migraines and irritable bowel syndrome [637]. They are classified into seven receptor subtypes (5HT1 to 5-HT7 ) which are expressed in various tissues and exert specific physiological actions. Ligands to the 5-HT receptors include 5-HT as an endogenous agonist and a group of synthetic indole derivatives.

278

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS H N

O i

O O

O

(CH2)n

H N O

H N

O iii

H 2N

N H

(CH2)n

(CH2)n

H N O

O

O

O

H N

ii

H N

186 (n = 2 to 12)

O NH2

N H

reagents and conditions: i) α,ω-diaminoalkane, NaBH(OAc)3, CH2Cl2, rt; ii) 50% aq. H2SO4, THF, rt; iii) 5-carboxamidoindole, EtOH, pyrrolidine, reflux, 3 days.

Figure 4.111 ceptor.

Synthesis of bivalent indoles as high-a‰nity ligand to an 5-HT1D; 1A re-

4.5.4.1 Carboxamidoindole Dimers. LeBoulluec et al. [646] described a bivalent approach that enables the identification of selective, high-a‰nity ligands for 5-HT receptors (Figure 4.111). For instance, bivalent ligands (186) are composed of two 5-carboxamidoindoles linked with a range of alkane chains (n ¼ 2 to 12) at the 3-position of indole. These indole dimers display subnanomolar a‰nity to 5-HT1D and 5-HT1A subtypes but show very poor a‰nity to a 5-HT1E subtype, with IC50 values ranging from 0.1 to >1 mM. An optimal selectivity to 5-HT1D receptors is observed with a linker of (CH2 )7 , where the IC50 values upon binding to the 5-HT1D and 5-HT1A subtypes are 0.05 and 1.83 nM, respectively. Increasing the linker length (n > 10) leads to high a‰nity but generates nonselective ligands to both subtypes. 4.5.4.2 Sumatriptan Dimer. Sumatriptan is a 5-HT1B, 1D -selective agonist that is used as an e¤ective drug for the treatment of acute migraine attacks. Perez et al. [647] reported a sumatriptan dimer tethered with a p-xylyl spacer at the sulfonamide nitrogen (Figure 4.112). This dimer (187) shows a better a‰nity profile than does a sumatriptan monomer. It binds to a 5-HT1B receptor with a Ki value of 0.64 nM, representing about a 30-fold enhanced a‰nity, and to a 5-HT1D receptor, with a Ki value of 0.89 nM and about a 10-fold increased a‰nity relative to the monomer. 4.5.4.3 Divalent SERT Inhibitors. Monoamine transporters are membranebound receptors that reuptake and control the synaptic level of endogenous neutrotransmitters such as dopamine, serotonin, and norepinephrine. The individual transporters derive their name from their binding selectivity to various monoamines, including the dopamine transporter (DAT), the serotonin transporter (SERT), and the norepinephrine transporter (NET). Interfering with serotonin transport by selective serotonin-reuptake inhibitors (SSRIs) such as fluoxetine constitutes an e¤ective way to treat depression.


H3C

CH3 N

CH3 NH + O O

Cl

Cl

NaH

S

DMF, rt

N H Sumatriptan

H 3C

CH3 N

CH3 N S O O

N H

279

CH3 N CH3

CH3 N S O O

N H

187

Figure 4.112 receptor.

Synthesis of bivalent sumatriptan as high-a‰nity ligand to an 5-HT1B; 1D

Tamiz et al. [648,649] prepared a series of bivalent inhibitors composed of piperidine-based low-a‰nity SERT inhibitors (188), which are linked with an alkane chain at the amide functionality of the monomer (Figure 4.113). Dimers (189; n ¼ 2 to 9) show inhibitory activities against all three monoamine transporters that are more potent than that of monomer, 188. In particular, 189 (n ¼ 5) is a potent and selective inhibitor of SERT. This molecule has a Ki value of 1.2 nM and selectivity ratios of 1633 (DAT/SERT) and 328 (NET/ SERT). In contrast, a divalent ligand (190, an optical isomer of 189) is about

Cl

O N H

188 N Me Cl

N Me

Cl

Cl

O

CH3

Cl

O N H

(CH2)n

O

N H

N Me

N Me

O N H

N H N Me

189 (n = 2 to 9)

Me

(CH2)n

190 (n = 5)

H O N

O

N

N H H

N

Me

H Cl

Cl

aryltropane dimer

Figure 4.113 Structure of bivalent inhibitors of serotonin transporter.

280

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS ~30 Å H2 N OH

H2 N

i,ii

O N H

n

N H

Serotonin (5-HT)

CdSe nanocrystal

191 (n = ~200)

H2 N O

O

O

CdSe/ZnS nanocrystal

S m

N H 192 (m = ~160)

reagents and conditions: i) Me4NOH, methanol, 30 min; ii) Cd-Se nanocrystals, 24 h.

Figure 4.114 Structure of multivalent serotonin tethered to CdSe nanocrystal.

10-fold less active with SERT than is 189 (n ¼ 5) with selectivity ratios lowered by fivefold (SERT/DAT) and 23-fold (SERT/NET) relative to those of 189. The results suggest the presence of dual ligand-binding sites at SERT. Bridging two proximal sites by 189 would be responsible for the greater than 8000-fold increase in potency and selectivity compared to monomer 188. Replacement of the N-methyl–containing ring structure and varying the attachment position can lead to altered binding selectivity, as illustrated by an aryltropane dimer linked with an octane spacer, which shows 130-fold higher selectivity toward DAT [650]. 4.5.4.4 Serotonin Displayed on Nanocrystals. Rosenthal et al. [651] reported nanocrystals used as a type of sca¤old for presenting multiple serotonin ligands on a surface (Figure 4.114). Multivalent ligands 191 and 192 are composed of multiple copies of serotonin derivatives attached to the surface of CdSe or ˚ . These multivalent molCdSe/ZnS nanocrystals with a diameter of about 30 A ecules inhibit serotonin reuptake by human SERT with EC50 values of 99 mM (191) and 115 mM (192). Such activities are, however, one to two orders of magnitude lower than that displayed by free serotonin. Due to the inherent fluorescence activity of the CdSe/ZnS nanocrystals, 192 can be used as a fluorescent probe when targeting cell surface receptors. For example, 192 labels SERT-expressing cells in a selective manner; the labeling is inhibited by SERT antagonists, such as paroxetine. 4.5.5

Muscarinic Acetylcholine Receptors

4.5.5.1 Quinnuclidine Dimers. Acetylcholine, a natural neurotransmitter, mediates a majority of physiological actions regulated by the muscarinic acetylcholine receptor (mAChR). This family of receptors contains five subtypes


281

S N N O 193 N

S N N O

1,4-orientation

1,3-orientation N

N 194

O

N N S

S N N O

O

N S N

N

N 195

Figure 4.115 Structure of bivalent agonists of muscarinic acetylcholine receptor linked at two di¤erent orientations.

with highly homologous sequences, particularly in the ligand-binding domain. This extensive sequence similarity makes it di‰cult to design a subtype-selective ligand. Christopoulos et al. [652] explored a bivalent approach in the design of ligands that selectively and potently bind mAChR (Figure 4.115). The dimeric ligands are designed on the basis of a monovalent ligand (193) that is a quinnuclidine-based, nonselective agonist of mAChR, with submicromolar binding a‰nity. Dimer examples include 194 and 195 as ligands merged through their phenyl groups in di¤erent orientations. Thus, each dimer presents two identical quinnuclidine units presented in the 1,4- and 1,3-orientations. The dimer 194 is a potent agonist to both the M1 (pKi ¼ 9.89) and M2 (pKi ¼ 10.10) subtypes of human mAChR expressed in CHO cells, with >100-fold selectivity over other subtypes. Moreover, this bivalent molecule is over 100-fold more active than the monomer 193 in binding the two subtypes. In contrast, the orientational isomer 195 is only 10-fold more active than 193 in all subtypes, with virtually no selectivity. 4.5.6

Melanocortin-1 Receptor

4.5.6.1 Divalent a-MSH Peptides. The melanocortin-1 receptor (MC-1 R) is a member of the seven-transmembrane domain GPCR family. Activated by peptide ligands such as melanocortin peptides and melanocyte-stimulating hormone (MSH), MC-1 R is involved primarily in pigmentation and steroidogenesis [653]. Carruthers and Lerner [654] reported divalent ligands derived from natural peptides that enhance the a‰nity to MC-1R and induce receptor clustering (Figure 4.116). This dimeric approach is based on a prior example of a godanotropin-releasing hormone receptor. In this example, a dimeric form of the ligand shows enhanced stimulatory activity and causes microaggregation of the receptor [655]. A divalent form of the a-MSH peptide (197) is composed of two identical copies of a pharmacophore sequence, MEHFRWG, tethered with a long peptide spacer linked to the C-terminus of the ligand. A terminal Cys-b-Ala serves as a dimerization point enabling the formation of a

282

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS 196

SYSMEHFRWGKPV-GGG-εAhx-(KKKKK)4-βAla

agonist domain

spacer

SYSMEHFRWGKPV-GGG-εAhx-(KKKKK)4-C-βAla S S

197

SYSMEHFRWGKPV-GGG-εAhx-(KKKKK)4-C-βAla

Figure 4.116 Structure of bivalent peptide ligand at melanocortin-1 receptor.

disulfide bond. The long, hydrophilic linker is intended to span two GPCR re˚ apart, assuming its conformational flexibility perceptor sites more than 40 A mits proper bridging at the two sites [656]. This dimeric peptide stimulates melanocortin receptors in the melanocyte dispersion assay with an EC50 value of 51 nM, which reflects about a sevenfold increase in activity over a monomer control (196). Another example of a peptide-based multivalent ligand that targets a membrane-bound receptor is enkephalin, a neurotensin peptide presented on the surface of liposomes, which shows significantly greater receptor a‰nity than that of its monomer [657].

4.6 4.6.1

ION CHANNELS Cyclic Nucleotide-Gated Channels

4.6.1.1 Cyclic Nucleotides. Cyclic nucleotides are involved in mediating physiological actions of allosteric e¤ector proteins. For example, cyclicnucleotide-gated (CNG) channels, found in vertebrate photoreceptors, are activated by cyclic guanosine monophosphate (cGMP); two protein kinases, PKG and PKA, are activated by cGMP and cyclic adenosine monophosphate (cAMP), respectively [658]. Each of these proteins displays four ligand sites per functional protein complex. For example, the CNG channel, a complex composed of four identical subunits embedded in the membrane, contains one cGMP recognition site per subunit (Figure 4.117) [659]. The presence of such multiple cGMP sites enables a bivalent cGMP molecule to achieve a high a‰nity association as a result of simultaneous binding at two neighboring cGMP sites. 4.6.1.2 PEG-Linked cGMP Dimers. Kramer and Karpen [660] designed a series of bivalent cGMP molecules tethered together with a varied spacer length. These dimers are composed of a sulfhydryl derivative of cGMP linked covalently with a bis(vinylsulfone)-terminated poly(ethylene glycol) spacer of variable size (Figure 4.117). For example, a dimer (197) presents two tethered ˚ (n ¼ 6), 39 A ˚ cGMPs separated by average spacer lengths of about 15 A ˚ (n ¼ 45), or 120 A (n ¼ 454). These dimers are agonists of CNG channels from the rat olfactory system, with K d values of 0.85 mM (n ¼ 6), 12 nM (n ¼ 45),

ION CHANNELS cGMP binding site

too-short linker

bivalent cGMP (length ~30-39 Å)

too-long linker

cytosol

out

283

tetrameric CNG channel O N

HN H2N

SH N

N O

-O P O

O S O

+

O OH

O

O S O

O n

pH 7.5, 37oC

MW = 282 (n = 6), 2000 (n = 45) MW = 20,000 (n = 454)

>2 eq

O

O N

HN H 2N

S N

N O

O S O

O n

O

NH

S N

N

NH2

O O

O -O P O

N

O S O

OH

197 (n = 6, 45, 454)

-O P O

O

OH

Figure 4.117 Tetrameric subunit arrangement of a CNG channel and its interaction with poly(ethylene glycol)-linked cGMP dimers.

and 250 nM (n ¼ 454). Thus, an optimum activity is observed with a linker at ˚ , a distance that may match the spacing between two n ¼ 45, or roughly 39 A nearby cGMP sites. Such dimers show greater a‰nities than free cGMP (Kd ¼ 31 mM ) and cGMP–poly(ethylene glycol) monomer (Kd ¼ 1:4 mM ). Unlike monovalent agonists, whose Hill coe‰cient is about 2, the dimer 197 shows a Hill coe‰cient of about 1, suggesting a bivalent mode of binding. 4.6.2

IP3 -Gated Calcium Channels

4.6.2.1 PEG-Linked IP3 Dimers. The inositol 1,4,5-triphosphate (IP3 ) receptor is a ligand-gated calcium channel that is activated by IP3 . It mediates the release of Ca 2þ from intracellular stores [661]. The mammalian IP3 receptor consists of tetrameric subunits, either homomeric or heteromeric, in which each subunit presents an IP3 -recognition site located close to the pore-forming region of a subunit. Riley et al. [662] and Riley and Potter [663] investigated the distance between IP3 -binding sites on IP3 receptors using divalent IP3 probes. Their dimers are composed of two IP3 ligands tethered through a poly(ethylene ˚ (Figure 4.118). These PEGglycol) spacer that varies in length from 10 to 80 A linked dimers bind to type 1 IP3 receptors in the rat membrane more tightly than does free, monomeric IP3 , which shows a Kd value of 11 nM, or a PEG-

284


HO O -O PO 3 -O 3PO

O OPO3-

HO O OH

HO O -

O- 3PO O3PO

H N

O

O

N HHO O

n

O

O OPO3-

-O PO 3 -O 3PO

n

H N

OPO3OH

H N

O O

O HO

-O PO 3 -O 3PO

199 (n = 5, 32, 75, 179)

O

O

O

OPO3OH

O

198 (n = 2, 11)

OH

-O PO -3 O3PO

O

H N

OPO3OH

CH3 6

200

Figure 4.118 Structure of bivalent inositol 1,4,5-triphosphate.

linked monomer 200, which shows a Kd value of 23 nM. The highest a‰nities to the tetrameric IP3 receptor are observed from dimers with shorter linkers, ˚ spacer) and 199 (n ¼ 6; *15 A ˚ spacer), which such as in 198 (n ¼ 2; *10 A show Kd values of 1.8 and 3.1 nM, respectively. These a‰nities represent binding that is 7 to 13 times tighter than that of 200. In contrast, IP3 dimers bind to monomeric IP3 receptors with an a‰nity similar to that displayed by IP3 monomers. The binding studies suggest that the four IP3 -binding sites on ˚. the tetrameric IP3 receptor are separated by less than 15 A 4.6.3

Potassium Channels

4.6.3.1 Bisquinolinium. The potassium channels belong to a family of membrane-bound proteins that mediates selective transport of potassium ions across membranes [664]. This channel protein exists uniquely as a tetramer, comprising four identical or similar pore-forming subunits (Figure 4.119) [665]. Because of sequence similarities between channel subunits, the potassium channel is likely to present multiple binding sites to regulatory molecules. For example, dequalinium (201), a bisquinolinium linked with a decane chain, is a relatively potent blocker of a low-conductance Ca 2þ -activated potassium channel (SK Ca ), with an IC50 value of 0.49 mM [666–668]. A cyclophane analog (202) also blocks the potassium channel, with activities influenced significantly by linker length. This trend is seen in the various IC50 values of 0.26 mM (n ¼ 10) and 2 nM (n ¼ 5). The removal of linker abolishes most of the activity exhibited by bisquinolinium blocker. Another type of dimer is illustrated by 203 (GW 0430), a neuromuscular blocker in clinical development for muscle relaxation [669,670]. It is a potent but ultrashort-acting agent because of the hydrolytic instability of the chlorofumarate linker.

ION CHANNELS

2

1

K-channel tetramer (top view)

Pore

4

CH3 H2 N

N

285

3

+

N

(CH2)n

HN

H3 C

+

NH

NH2 N

Me

Me (CH2)n

N

+

+

201 (Dequalinium)

202 (n = 10, 5) MeO N +

MeO MeO

Me

Cl O

OMe

O O

O

MeO

N Me + MeO

OMe

OMe

OMe OMe

203

Figure 4.119 Potassium channel composed of four identical pore-forming subunits and examples of its channel blockers.

4.6.4

Nicotinic Acetylcholine Receptor

4.6.4.1 Succinyldicholine. The nicotinic acetylcholine receptor (nAChR), part of the well-studied family of ligand-gated ion channels, is a membrane-bound glycoprotein comprised of five subunits (e.g., a2 bgd in muscle type) [671]. The five subunits are arranged pseudosymmetrically around an axis perpendicular to the plane of the membrane surface, forming a central pore that functions as a gated channel (Figure 4.120). This channel opens when the receptor is activated, as a result of the binding of agonists, such as acetylcholine, an endogenous quaternary ammonium molecule. Each nAChR pentamer presents two acetylcholine-binding sites present on two a-subunits [672]. However, acetylcholine molecule binds with unequal a‰nity to the two sites. The high-a‰nity site is located at the interface between subunits a and d, while the low-a‰nity site is at the a and g interface. Acetylcholine binds at the two sites with a Hill coe‰cient of 1.4, suggesting cooperative interactions, and an equilibrium dissociation constant of 1.3 mM [673]. Succinyldicholine is a dimer composed of two cholines linked through a short succinyl bis ester. With a Kd value of 0.35 mM, it is an nAChR agonist with a higher a‰nity than that of acetylcholine. Thus, succinyl dicholine binds about four times more tightly than acetylcholine, although insu‰cient linker length probably precludes bivalent bridging at the two defined acetylcholine sites.

286

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS cross-sectional view of nAChR

~46 Å

Ach site Pore

membrane

Ach site

cytosol top view of nAchR

Pore (d ~ 20-25 Å)

+ Me3N

O O

CH3

+ Me3N

O O

O O

Acetylcholine (Ach)

NMe3 +

Succinyldicholine

Figure 4.120 Muscle-type nicotinic acetylcholine receptor composed of a pentameric subunit complex (a2 bgd): for clarity, the cross-sectional view shows only three subunits.

4.7 4.7.1

NUCLEIC ACIDS DNA Intercalators

There are many naturally occurring antitumor antibiotics that bind to DNA, as represented by actinomycin D (dactinomycin) and the anthraquinone class of molecules, which include anthracyclines, such as aclacinomycin, pluramycins, and aureolic acids, such as olivomycin A (Figure 4.121) [674]. These molecules bind DNA with high a‰nities and sequence selectivity. Their strong DNA binding activity is attributed to the molecules’ bivalent mode of interaction, based on their thermodynamics of binding (discussed briefly in Chapter 1). Specifically, an aromatic moiety supplies a DNA intercalating role, and its tethered carbohydrate residue binds within the DNA minor groove. This type of bivalent interaction also occurs when the antitumor antibiotic, chartreusin, binds to DNA [675]. 4.7.1.1 Anthraquinone Linked to an Aminosugar. On the basis of a bivalent binding model such as that described above, Toshima et al. [676,677] designed artificial DNA-binding molecules that are composed of an anthraquinone moiety linked to a 2,6-dideoxyaminosugar, both of which are common structural components found in many natural DNA-binding molecules (Figure 4.122). These synthetic molecules bind selectively to a TGC sequence, with potency influenced by the nature of the aminosugar moiety. For example, 204, presenting a d-aminosugar, binds stronger than 205, which contains a corre-

NUCLEIC ACIDS O

287

CO2Me Me OH

Intercalation OMeO

OH O

Me 6Å

O

O NMe2

Minor groove binding Me

O

O OH Me O

Me OH O O Me2N Me AcO Me O

OMe Me

Me

O

HO

O

OAc Me

O

O

O

NMe OH 2

i-Pr O

Pluramycin A

Me

N

O Me O O O O

O HN

N

NH2

Me

O

O

Me

Me

Me

Olivomycin A

N

O NH O

OH

Me

Intercalator

N

Me

HO O Me Me O O HO

Me

Me

NH O O H O Me N O O O

OMeOH

O H OH OH O Me O O Me OHOO

O

N

Aclacinomycin

H

O

Me

Me

Me

O

O

N N

Me

Me Minor groove binder

Me

Me

Actinomycin D (Dactinomycin)

Figure 4.121 Bivalent interaction of anthraquinone antibiotic with DNA.

sponding l-sugar, indicating that the sugar moiety contributes to the binding. The binding activity is also translates into cytotoxicity, with IC50 values of 9 mM (204) and 58 mM (205). Two more examples of artificial hybrids include 206 and 207, which contain a 2-phenylquinoline aromatic moiety as the intercalator as well as the photoreactive group. These dimers not only bind to DNA but also cleave DNA strands through a photomediated radical mechanism that leads to cytotoxicity. Their cytotoxicity increases when evaluated under ultraviolet irradiation. 4.7.1.2 Daunorubicin Dimer. A member of the anthracycline class of antibiotics, daunorubicin is functionally a bivalent molecule. Its anthraquinone unit intercalates into DNA oligonucleotides with sequence selectivity. Its dau-

288

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS O Me HO Me2N

Me HO Me2N

O Me2N HO O Me

O OH O

OH O

204

205

O O

O

O

Me HO Me2N

O O

O

O

N N 206

207

Figure 4.122 Structure of artificial DNA binding molecules as an aromatic– carbohydrate conjugate.

nosamine sugar moiety binds to the DNA minor groove [678]. The ability to interact bivalently with DNA confers on daunorubicin a strong binding a‰nity, with a Ka value of 1.6 10 7 M 1 . Based on the established structure of the anthracycline–DNA complex, the drug binds to DNA in a 2:1 stoichiometry, whereby each drug molecule intercalates at either end of an oligonucleotide, while its aminosugar moiety occupies the minor groove in an orientation facing toward one another (Figure 4.123). The distance between the two aminosugars ˚ . Chaires et al. reported a synthetic anthracycline is estimated to be less than 7 A dimer (208) composed of two daunorubucin moieties [679]. This dimer contains a covalent linkage at the aminosugar, tethered through a p-xylyl spacer such that the dimer mimics the spatial arrangement of two free daunorubicin molecules in the bound state. It binds DNA more tightly than does the natural monomer. The Ka value of 208 to DNA is 2.7 10 11 M 1 , which represents a 16,000-fold enhancement over daunorubicin. A significant portion of this huge enhancement in a‰nity comes from a large enthalpic contribution, where DH ¼ 30 kcal/mol; T DS ¼ 15 kcal/mol, as compared to daunorubicin, where DH ¼ 11 kcal/mol; T DS ¼ 1 kcal/mol). These di¤erences in DH and T DS indicate that bivalent binding is driven by enthalpy rather than entropy in this particular system. Additional antibiotics, derived from naturally occurring DNA binders, have been designed in a dimeric form. The dimer 209 represents such a molecule. It is composed of two covalently linked diazobenzofluorene glycosides whose monomeric structure resembles kinamycin [680]. This dimer demonstrates potent cytotoxic activity against cancer cell lines, at IC50 values of 0.01 to 98 ng/mL. The molecular mechanism probably involves dsDNA cleavage. 4.7.1.3 Bisacridine. A hairpin loop is a common structural element found in RNA as well as in gene transcription regions involved in DNA replication [681]. A number of anticancer molecules, such as ethidium bromide, actino-

NUCLEIC ACIDS O

OH

289

O OH

CH3O

O OH O H3C O OH NH

4 > 7 > 5 > 8. 4.7.2.8 Hairpin-Shaped Polyamide Dimers. Swalley et al. [715], Trauger et al. [716], de Clairac et al. [717], and Baird and Dervan [718] reported a novel series of synthetic polyamides, enabling sequence-selective recognition of doublehelical DNA within the minor groove. Like natural polyamides, these synthetic polyamides are composed of N-methylimidazole and N-methylpyrrole connected by amide bonds in multiple repeats and diverse combinations, which modulates their a‰nity and sequence selectivity to DNA. According to the empirical base-pairing rule suggested by White et al. [719], Im/Py, which consists of an imidazole ring on one polyamide ligand complemented by a pyrrole ring on a second polyamide ligand, targets G C base pairs selectively, while Py/Py targets T A pairs e¤ectively (Figure 4.131).

Me N O

Me N O N H

N 5' Im

A

C

3'

T

N Me

A

A

A Py H N

H2 N

NH2

T

Py

O

N H Py

Py G

T

Me N O

tethering to hairpin dimer

T

Py H N

N Me

O

Me N O

O

Me N O N H

N

+ NHMe2

N Me Me N O N H

NH O

H N

HN O

O

H N

N Me

O

H N N Me

O

NH N Me

225 (ImPyPy-γ-PyPyPy-β-Dp)

Figure 4.131 Base-pairing rule in polyamide–DNA interaction and structure of hairpin-shaped dimeric polyamide.

298


A heterodimer (225) is designed on the basis of a side-by-side antiparallel binding mode. This hairpin-shaped dimer is prepared by tethering two complementary polyamides at both ends with an g-aminobutyric acid spacer as the turn motif [720]. This dimer binds tightly to dsDNA containing a target sequence of 5 0 -TGTTA ACAAT-5 0 with a K a value of 2.9 10 8 M 1 . The high a‰nity is attributed to multivalent association. This association is so sensitive to DNA sequences that its binding a‰nity decreases about 60-fold when bound to a single mismatched sequence such as 5 0 -TGTCA ACAGT-5 0 . A second dimer (226), which is composed of six rings (ImImPy-y-PyPyPy-R-Dp), closely resembles 225, as the second Im ring replaces the second Py ring in 225 (Figure 4.132). This dimer shows altered sequence specificity and preferentially targets a sequence of 5 0 -TGGTT-3 0 , with a K a value of 1.0 10 8 M 1 [721].

Me N O

Me N O N H

N

+ NHMe2

Me N O N H

N

Me N O

Me N O NH

HN

N H

N

Me N O N H

N

O H N

O

O

N Me

H N N Me

O

NH N Me

O

H N

HN O

N Me

5'- T

G Im

β

+

3'- A

Py Py C C

5'- T

Py A

A -5'

3'- A

Me N O

Me N O N H

N

N H

N

H N

N H

O

O

H N

O

N Me

N

O

N Me

N Me

N Me

O

G

T

T -3'

Im

Py

β

Py Py C C

Py

5'- T

NH

H N

N Me

G Im

+

A

A -5'

G

Me N O

O H N

O

Im Py

Me N O + NHMe2

T -3'

T

G

Me2HN +

H N

227 (PyPyPy-γ-ImImPy-β-Dp)

226 (ImImPy-γ-PyPyPy-β-Dp)

HN

NH

O

H N

HN

O N H

N

+

β

3'- A

NH

G Im

C

C

Im Py

Py

Im

Im G

Py Py C C

G

A -3'

T -5'

N Me

O

228 (ImImPyPy-γ-ImImPyPy-β-Dp)

Me N O

Me N O + NHMe2

N

N H

Me N O

Me N O N H

N H

Me N O N H

NH O

H N

HN O

O

H N N Me

H N

O

N Me

O

H N N Me

O

H N N Me

O

N

NH

5'- T

β + 3'- A

G

T

T

A

C

Im

Py

Py

Py

Py

Py C

Py A

Py Py Im T -5' A T G

A -3'

N Me

229 (ImPyPyPyPy-γ-ImPyPyPyPy-β-Dp)

Figure 4.132 Structure of hairpin-shaped dimeric polyamides and their preferred sites of interaction with dsDNA.

NUCLEIC ACIDS

299

Hairpin-shaped dimeric polyamides are prepared by attaching two polyamides. The carboxy terminus of one polyamide is linked to the amino terminus of the second (225, 226), or vice versa (227) [722–724]. For example, the linker location in dimer 227 utilizes an amino-to-carboxy terminus, as opposed to that of 226. According to the base-pairing rule, these two dimers target an identical sequence (5 0 -TGGTT ACCAA-5 0 ), although the association constant of 226 is about sixfold greater than that of 227. 4.7.2.9 Hairpin-Shaped Extended Polyamides. The number of consecutive heterocycle units participating in a DNA interaction a¤ects the binding a‰nity of polyamide dimers. Heterodimers 228 and 229 present an array of eight and 10 heterocycle rings composed of Im and Py, respectively. Relative to dimers composed of six rings, these extended dimers show enhanced a‰nity to their preferred DNA sequences by two orders of magnitude [725–727]. For example, the K a value of 228 to dsDNA presenting a sequence of 5 0 -TGGCCA-3 0 is 9.7 10 9 M 1 and that of 229 to a sequence of 5 0 -TGTTACA-3 0 is 1.0 10 10 M 1 [728,729]. 4.7.2.10 Ring-Shaped Polyamide Dimers. Cyclic dimerization at both ends of two dimeric polyamides, such as 231, a¤ords a macrocyclic dimer (232) (Figure 4.133) [730]. This dimer is conformationally rigid and adopts a single orientational configuration relative to 231 while maintaining DNA-binding ring units ordered in the right place. It binds tightly to the DNA minor groove at a target sequence of 5 0 -AGTACT-3 0 . The K a value of this interaction is 7.6 10 10 M 1 , an a‰nity enhanced about eightfold over that of a hairpin-shaped dimer (231). It is enhanced about 3600-fold over that of a monomer (230). 4.7.2.11 Tetravalent Polyamides. Covalent dimerization of a hairpin-shaped dimer (226) in a head-to-head orientation yields a tetravalent molecule (233) [731]. Unlike hairpin-shaped dimers, which target five-base-pair sites in the minor groove of DNA, this tetramer targets 10 to 12 base pairs with high a‰nity and selectivity. For instance, 233 (n ¼ 3) binds to an 11-base-pair sequence of 5 0 -TGGCATTACCA-3 0 , with a K a value of 7 10 9 M 1 , which represents a 70-fold increased a‰nity over 226. 4.7.2.12 Polyamide–DNA Alkylator Conjugates. Many DNA alkylating agents used as drugs show poor binding selectivity to the dsDNA minor groove, a property that causes undesired e¤ects, such as severe toxicity. Attaching sequence-selective alkylating agents to polyamide sca¤olds is expected to yield sequence-selective drugs (Figure 4.134). For example, Baraldi et al. [732] reported a heterodimer (234) that carries a PBD unit derived from DC-81 as a G C-targeting DNA alkylator tethered to a distamycin-derived polyamide. This dimer makes a more stable complex with dsDNA than does free distamycin or DC-81 alone. It inhibits tumor cell growth at submicromolar concentrations, which represents three to 25 times greater activity than that

300


Figure 4.133 Structure of diverse multivalent polyamides, including cyclic dimer and head-to-head linked tetramer.

displayed by either DC-81 or distamycin. Such enhanced activity is attributed to tighter binding. Three other heterodimers (235 to 237) are designed by a similar concept to present both cyclopropylpyrroloindole (CPI) and polyamide. The dimer 235 selectively binds and alkylates A T-rich sequences in the minor groove of DNA. The dimer strongly inhibits tumor cell growth at nanomolar concentrations, reflecting an activity that is over than 20-fold more potent than free CPI units [733]. A second dimer (236) binds in a hairpin conformation to dsDNA and preferentially alkylates multiple adenine bases at the sequences 5 0 AAAAA-3 0 and 5 0 -TCAG-3 0 [734]. The CPI–polyamide 237 makes a complex by binding to a specific dsDNA sequence in 2:1 stoichiometry and alkylates both adenines and guanines located proximally on opposite strands [735]. 4.7.2.13 Polyamides Linked to Camptothecin. Covalent coupling of a polyamide with a secondary binding motif makes it possible to direct the attached group to a site at (or near) a coding domain within the DNA minor groove. Such molecules include a heterodimer (238) composed of three modules: a poly-

301

NUCLEIC ACIDS Distamycin

HN

Me N O

Me N O

Me N O N H

N H

+ NH2

N H

HN

O

N H

O N

O

N N

H 234

PBD

N

MeO

O

HN

N H

+ NH3

N H

N

O

235

Me N

O

N H

N

N CO2Me

OH O

N H

O

O

NH

237

236

N O

+ NH2

Me Me N O

N H

NH2 N H

NH CPI N

O Me

N H

OH

N

Cl Me N O

Me N O

Me

O

Me N O

Me N O

Me N O

NH2

Me

Cl cross-link

5'- C

3'- G

T

G

+

Im Py

CPI A C

A -3'

5'- C

T

G

A

Im

Py

CPI T -5'

3'- G

A

C

C

G

A -3'

CPI

Py

Im

T

G

C

T -5'

cross-link

Figure 4.134 Structure of polyamide linked to DNA alkylator.

amide linked with a seven- to 18-atom flexible linker to a camptothecin moiety that recognizes and stabilizes a topoisomerase I-DNA cleavage complex (Figure 4.135) [736]. This heterodimer shows a high a‰nity to a sequence of 5 0 -AGTATT-3 0 , with a K a value of 1 to 3 10 8 M 1 . In the presence of topoisomerase I, the heterodimer induces DNA cleavage predominantly at sites three to four base pairs apart from a polyamide target sequence. 4.7.2.14 Polyamides Linked to Transcription Activators. A novel application of polyamide-based functional molecules is illustrated in the design of an artificial gene transcription factor [737]. In principle, this type of molecule mimics the functional aspects shown by natural activator proteins, which recognize a specific DNA sequence and recruit the transcriptional machinery to nearby promoters to initiate transcription. A multivalent polyamide (239) represents such a class of artificial transcription activators (Figure 4.136) [738]. It consists of a hairpin-shaped eight-ring polyamide that enables selective sequence targeting, an oligo(proline) spacer that forms a rigid helical structure, and a 16amino acid peptide, VP2, that functions as an activating region. The distance ˚ (Pro6 ) to about 45 A ˚ supplied by the proline spacer is varied from about 18 A (Pro15 ), long enough to optimize transcription activity. The binding a‰nity of

302

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS Me N O

Me N O + NHMe2

N H

N

Me N O N H

Me N O N H

NH O

H N

HN

H N

O

O

Camptothecin O

O H N

O

N Me

5'- T

+ 3'- A

N

O

N Me

A

β T

H N

N

O

N Me

G

T

A

T

Py

Py

Py

Py Py

Py

Py

T

A

A

O

N H

O N N O

N Me

Im

C

R

N NH H

238 R = (CH2)5 R = (CH2)11 R = (CH2CH2O)4(CH2)4

T

C

G

C

OHO CH 3

A -3'

A

Camptothecin

A

C

G

T

G

T -5'

Topo I-induced cleavage site

Figure 4.135

Mode of action of polyamide–camptothecin conjugate.

Activation domain

RNA pol II

Spacer

+

β

Im

Py

Py

Py

Py

Py

Py

Py

5'-

DNA RNA pol II

DNA-binding domain

T

G

T

T

A

Im

Py

Py

Py

Py

Py Py

Py

A

T

+

β

3'-

A C

HS O O NH Me N O

Me N O + NHMe2

N

N H

N N H

N

N H

O

A

T -3' gene transcription A - 5'

H N DFDLDMLGDFDLDMLG n O VP2 (activation domain)

Me N O

O N H

NH O

H N

HN O

O

H N N Me

O

H N N Me

O

239 (n = 6, 9, 12, 15)

H N N Me

O

NH N Me

Figure 4.136 Polyamide-based artificial transcription factor and its activation of gene transcription.

NUCLEIC ACIDS

303

239 to dsDNA containing a target sequence of 5 0 -TGTTAT-3 0 varies within only a factor of 2 over these linker distances, with a K d value of 3.3 nM (n ¼ 12). In transcription reactions measured in vitro, these dimers enhance the activation of gene transcription 10- to 37-fold relative to the level observed in the absence of 239. Maximum transcription activity is observed from dimers ˚ ). This distance represents optimal spacing bewith the linker of n ¼ 12 (36 A tween DNA and the activation region required for e‰cient transcription. 4.7.3

Divalent Recognition at Abasic Sites

One of the most frequent DNA lesions, an abasic (apurinic, apyrimidinic, or collectively, AP) site is an abnormal site present on DNA that lacks a covalently attached heterocyclic base. The base is removed as a result of hydrolysis of an N-glycosidic bond (Figure 4.137) [739]. It is caused spontaneously by chemical activities, from cancer drugs and alkylating agents, physical assaults, C

G

O

O

O

O

-O O P O O

O O P Abasic site O OHO O

T

O

-O O P O O

O

O

O O

-O O P O O

O P O O-

C

G

O

Purine-intercalator O heterodimer

-O O P O O

O P O O-

O

O

O O

-O O P O O

A

O P O O-

T

O

O

-O O P O O

O P O O-

O

O

O NH2

NH2

N

N

Adenine

N

N

N

N

N

H2 N

N

+ H2 N + + NH 2

Cl

N NH2 N

Cl NH

Acridine

O P O O-

N

NH2 + NH2

NH N

OMe

240

OMe

241

Figure 4.137 Basic site in dsDNA and its interaction with purine–acridine heterodimer.

304


such as irradiation, and by enzymatic activities, such as from DNA N-glycosylases. AP sites can remain unrepaired by DNA repair enzymes, leading to gene mutation and eventually, cell death. Hence, the AP sites are the main antitumor targets of alkylating drugs. In this context, molecules that recognize AP sites can interfere with repair processes and enhance of biological activity of anticancer drugs. 4.7.3.1 Adenine Tethered to Acridine. Berthet et al. [740], Belmont et al. [741], and Alarcon et al. [742] reported heterodimers 240 and 241, which are composed of an adenine nucleic base and an acridine intercalator, tethered with a positively charged 10-atom spacer. Such spacers also enhance the association with DNA through electrostatic attraction to the negatively charged phosphate ions in DNA. The adenine unit selectively recognizes the thymine base located at an abasic site, while the acridine intercalates into DNA. In a thermal denaturation study, dimer 240 stabilizes a duplex DNA containing an abasic site (5 0 -GCGTGTGTGCG CGCACXCACGC-5 0 , where X represents an abasic site), as judged by an increase in the Tm value of dsDNA relative to that measured in the absence of dimer. This e¤ect is attributed to the formation of 1:1 complex between 240 and the oligonucleotide with an abasic site. A heterodimer (241) that is linked with a strongly basic guanidine-fused spacer (pK a > 11) binds more strongly to an abasic site than does 240. The tighter interaction probably stems from the attraction between a cationic guanidinium group with anionic phosphate.

4.7.4

Divalent Binding to Guanine–Guanine Pairs

4.7.4.1 Naphthyridine Dimers. A single nucleotide di¤erence between two otherwise identical DNA sequences constitutes a single-nucleotide polymorphism (SNP) [743]. SNPs occur throughout the human genome at a rate of one in every 500 to 1000 base pairs. High-a‰nity ligands that bind to SNPs could enable their rapid detection and mapping. Nakatani et al. [743] reported divalent ligands that selectively target a typical SNP, the G G mismatch (Figure 4.138). A dimer (242) is composed of two units of 2-amino-1,8-naphthyridine, a heterocycle that recognizes a guanine base by forming a complementary array of hydrogen bonds. It binds to dsDNA with G G mismatched pairs in 1:1 stoichiometry, with a K a value of 1:9 107 M1 . It binds selectively to G G sites with an a‰nity that is 360-fold greater than its a‰nity for G A or G T mismatched pairs. 4.7.5

Basic Leucine Zipper Proteins

4.7.5.1 bZIP-Derived Peptide Dimers. Peptides derived from basic leucine zipper proteins (bZIPs) can be dimerized through formation of a disulfide bond between the two identical peptides, which bind dsDNA with sequence selectiv-

NUCLEIC ACIDS

305

242 5'

5' G

A

G

C

T

C

G

G

A T

A

G

T

C

G G

C

A

G

T

3'

3' G-G mismatch

DNA NH N

O

N

Me

N H

NH Me

N

N

H N

Me O

242

N N

N H

G N H

N

N

O

O NH

H N

N H O

N

N

H N G

H N

N O

N

Me

H

N DNA

Figure 4.138 Recognition of G-G mismatch by divalent naphthyridine.

ity [744]. For example, two identical helical peptides derived from the GCN4 basic region can be linked together. The resulting homodimer binds dsDNA, with each of the two helical peptides directly contacting the major groove on opposite sides (Figure 4.139) [745]. This dimer recognizes a sequence called CRE domain (5 0 -ATGACGTCAT-3 0 ) more strongly and selectively than does the peptide monomer. 4.7.5.2 Divalent bZIP Moieties Linked to Chiral Templates. Synthetic bZIP peptide dimers are designed by tethering bZIP peptides to sca¤olds that may include a metal complex, b-cyclodextrin–adamatane system, or a biphenyl group [746–749]. A specific example is 243, which presents two GCN4 peptides attached with N-to-N polarity on the C2 -chiral template used as a dimerization module [750]. This dimer displays sequence selectivity di¤erent from that of native dimers, which have C-to-C polarity. Thus, sequence selectivity is dependent on peptide orientation. 4.7.5.3 Divalent bZIP Moieities Tethered to an Azobenzene Sca¤old. Varying the distance between two linked peptides modulates the binding a‰nity of a peptide dimer to DNA. A dimer (244) was designed on an azobenzene sca¤old that can be used to adjust the interpeptide distance through the application of light [751]. In a thermally stable trans isomer, the two linked peptides are sep˚ . When irradiated at 430 nm, the trans conarated by a distance of 12 to 13 A formation is isomerized to a cis form, reducing the interpeptide distance to 8 to ˚ . The cis isomer of 244 binds across a 20-base-pair duplex DNA fragment 9A containing a sequence termed CREB. Its a‰nity is enhanced 60 to 70 times over that of the trans form.

306

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS dimeric peptide basic α helix (bZIP)

sequence-specific DNA binding

O

H N

N H

O O

S-peptide O

peptide =

O

H N

N H

O

S-peptide

CPAALKRARNTEAARRSRARLKLQ-CONH2

243 (N-to-N polarity)

12-13 Å N N

peptide

peptide UV trans

cis

N N

8-9 Å

Vis peptide

peptide 244 (C-to-C polarity)

peptide = CH2OC(=O)CH2S-CGGQLKRARSRRAAETNRARKLAAPD-NHAc

Figure 4.139 Sequence-specific DNA binding by dimeric bZIP a-helix peptide.

4.8 4.8.1

ANTIBODIES Anticancer Vaccines

4.8.1.1 Tumor Cell Marker. Glycosphingolipids (GSLs) comprise a class of glycosylated ceramides expressed on the surface of mammalian cells. The GSLs are grouped in three series, globo-, ganglio-, and lacto-, based on the structure of attached complex carbohydrates. Normally, these molecules play essential roles in mediating cell–cell adhesion, communication, and signal transduction processes. Studies performed by Hakomori and Murakami [752] and Mora et al. [753] show that certain types of GSLs are found in specific tumor cells at abnormally high levels. For example, globo-H-based GSL is highly expressed in breast and ovarian cancer cells, while GM3 , of theganglio series, and Le y – Le x , of the lacto series, are found abundantly on the surfaces of melanoma and lung cancer cells, respectively (Figure 4.140) [754]. Little is known about the cellular mechanisms responsible for the enhanced levels of GSL, although it probably results from abnormal carbohybrate expression and metabolism. A higher surface density of specific GSLs on tumor cells contributes to the strong cell–cell adhesion required for tumor cell metastasis and the active signal

ANTIBODIES

307

MBr1 antigen (breast and ovarian cancer)

OH OH OH HO OHO O HO O O O HO O NHAc HO O H 3C O OH HO OH HO Globo H OH

O

HO O HO

OH

O

O

NHCOC17H35 C13H27

OH

OH

OHOH CO2-

AcHN HO HO HO AcHN

O

HO O HO OH HO -O2C O O O O OH

OH OH

HO O O

OH

NHAc

Disialosylgalactosyl globoside (renal cell carcinoma)

O HO O HO HO

O

HO O HO

O

AcHN

OH CO2-OH HO O O O HO OH OH OH O

O OH

O

NHCOC17H35 C13H27

OH

OH

HO OH

O

OH

NHCOC17H35 C13H27

GM3 (melanoma)

OH

H3C OH Ley-Lex O (lung, colorectal cancer) OH OH OH OH OH HO OH HO OHOH O O HO O O O O O O HO HO O O OO OH NHAc O NHAc OH H 3C O OH H3C O OH OH OH OH OH

O OH

O

NHCOC17H35 C13H27 OH

Figure 4.140 Structure of representative examples of tumor-associated glycosphingolipid.

transduction required for rapid cell growth. Specific types of GSL molecules may serve as cell surface markers associated with certain types of cancer. Due to the high level of expression on the surface of tumor cells, GSLs induce an antibody response through the immunological defense mechanism [755]. Stimulation of the immune system occurs primarily through interaction with exposed hydrophilic carbohydrate moieties of the GSLs, which function as tumor-associated antigens. Thus, one potentially e¤ective cancer therapy has been built on vaccines that precisely mimic the carbohydrate-displaying surface of tumor cells. Such vaccines include a class of multivalent molecules that carry multiple copies of tumor-associated antigens displayed on a carrier protein or sca¤old [756]. However, the development of carbohydrate-based vaccines has been slow, due partially to the di‰culty of synthesizing such complex carbohydrate antigens. Recently, however, the advancement of synthetic methodologies has enabled the large-scale preparation of specific carbohydrate antigens [757–761]. 4.8.1.2 Globo-H Presented on Keyhole Limpet Hemocyanin. A hexasaccharide, globo-H, constitutes one of several immunogenic molecules present in

308

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS 245 OH

OH

O HO O

HO HO

O

OH Globo H allyl glycoside O HO O O 244 NHAc HO O HO O O HO O O HO HO OH OH

O O OH OH OH

H3C

OH OH OH HO OHO O HO O O O HO O NHAc HO O H 3C O OH HO OH HO OH

O ii

O

O

OH OH OH HO OHO O HO O O O HO O NHAc HO O H3C O OH HO OH HO HO

O

N H

OH

OH

N

i

246 HO O O HO

iii

N

H N

O O

247 HO O O HO

O O

O

OH

OH

O

H2N-HN

N H

N

H N

O O

KLH

S

H N

n Å 720

NH OH

OH HO HO

OHO O O

H3C

O OH OH OH

OH O HO O NHAc

248 (carrier protein = KLH; n = 152) 249 (carrier protein = BSA; n = 17)

O

OH O HO HO

O OH

HO O HO

O OH

O

N Hn

carrier protein

reagents and conditions: i) O3, Me2S; ii) NaCNBH3; iii) thiolated KLH; then dialysis.

Figure 4.141 Synthesis of multivalent globo-H tethered to carrier protein.

high density on the surface of prostate, breast, and ovarian cancer cells. Park et al. at Columbia [762] reported the total synthesis of the globo-H or Hakomori MBrl antigen, which is believed to be a human breast tumor–specific antigen. This antigen (244) is prepared as a globo-H allylglycoside construct by applying glycal assembly methodology (Figure 4.141). An allyl group located at the terminus of the hexasaccharide construct is introduced as a handle that is used during conjugation to the protein sca¤olding. Ozonolysis of 244 generates an aldehyde functionality, which is then reductively coupled with the amino group of the bifunctional spacer, 4-(4-N-maleimidomethyl)cyclohexane-l-carboxhydrazide (245). The maleimido group of 246 becomes cross-linked to thiolderivatized keyhole limpet hemocyanin (KLH), a highly immunogenic carrier protein. This process results in the preparation of 247, a multivalent globo-H presented on the KLH protein with 720 globo-H moieties per protein [763]. Park et al. demonstrated that the carbohydrate vaccine 247 could induce a high titer of specific IgM antibodies against globo-H in human trials. Two similar examples of globo-H conjugates (248 and 249) were prepared from 244, with some di¤erences in the design of multivalent presentation, such as the

ANTIBODIES

309

valency number, the length and nature of the spacer, and the type of carrier protein [764,765]. When used as a vaccine in conjunction with an immunologic adjuvant in the mouse, 248 induces the production of globo-H-specific IgM and IgG antibodies. The antibody titer is greater than that elicited by 249, which presents a lower number of globo-H antigens on bovine serum albumin (BSA). These antibodies react with globo-H-positive MCF-7 breast cancer cells and e¤ectively induce complement-mediated cytotoxicity against such tumor cells. 4.8.1.3 GM3 Presented on Bovine Serum Albumin. Overexpression of a ganglioside sphingolipid, GM3 , is a broad biochemical marker linked to several di¤erent types of cancers, such as skin, colon, breast, and lung cancers [766]. Raising antibodies against GM3 presented in a multivalent format may find potential application in anticancer therapy. For instance, a murine, GM3 specific monoclonal antibody (mAb) inhibits the growth of melanoma cells [767]. Lee et al. [768] and Delgado et al. [769] reported GM3 -based, synthetic, multivalent constructs (250 and 251) which are prepared using a chemoenzymatic approach. These multivalent molecules display 17 or 10 copies of the GM3 epitope, respectively, per BSA carrier molecule Figure 4.142). These conjugates can enable the isolation of single-chain antibodies isolated against breast and melanoma cells. This technique identified a number of antibodies that bind to 250 with Kd values ranging from ibodies are identified with K d values for 250 of 4.2 101 to 2.1 105 M. The GM3 linear trisaccharide is viewed as a tumor-associated antigen, but some immunization experiments suggest that instead, 252 is the actual antigen.

GM3 epitope HO HO

-O2C HO O AcHN O OH OH

OH

HO O O HO

OH

O

O

O O

OH

N

N H

H N

S

BSA

NH n = 17

O

250 HO HO AcHN

-O2C HO O O OH OH

OH O

HO O HO

OH

O

Ac N

OH

N Me

O

H N

N

S

O

H N NH

O

BSA n = 10

251 HO HO

OH O

O HO O HO AcHN

O

HO O HO

O OH

O

NHCOC17H35 C13H27 OH

X

OH OH

252 (X = O; GM3 lactone) 253 (X = CH2)

OH OH O

HO AcHN OH O OH -O2C HO O O O AcHN O HO OH OH OH

O

O

NHCOC17H35 C13H27

OH

GM2 glycosphingolipid

Figure 4.142 Structure of ganglioside-based synthetic vaccine.

OH

310


This lactonic form of GM3 bears a spirocyclic conformation and does not carry a negatively charged carboxylate ion. As a potential hapten for immunization, Tietze and Keim [770] reported the synthesis of 253, a hydrolysis-stable GM3 lactone that mimics sphingolipid molecules. 4.8.1.4 GM2 Presented on KLH. The GM2 sphingolipid is also overexpressed on the surfaces of melanoma, sarcoma, and renal human cancer cells [771]. Vaccination has been attempted with GM2 -KLH conjugates that present natural GM2 extracted from mammalian tissues. These conjugates induce the production of antibodies with limited activity and specificity, possibly reflecting the heterogeneity of natural ceramides. The use of synthetic GM2 prepared in pure form would help to solve such an issue. Castro-Palomino et al. [772] and Stauch et al. [773] reported a synthetic method enabling the e‰cient synthesis of GM2 in a form useful for immunotherapy. 4.8.1.5 TN Clusters Presented on KLH. A family of natural glycoproteins, mucins are expressed on the surface of epithelial cells [774]. These glycoproteins carry multiple copies of carbohydrate clusters attached to serine and threonine residues in O-linked form, where the carbohydrates involved in attachment commonly include N-acetylgalactosamine. The mucins isolated from normal cells are composed of large carbohydrate clusters, whereas those from malignant cells present shorter carbohydrate chains. Examples include TN monosaccharide, TF (Thomsen–Friedenreich) disaccharide, and STN (sialyl TN ), all of which are commonly found on colon and prostate cancer cells (Figure 4.143) [775].

OH HO O HO AcHN O O

OH OH HO OHO O O HO OH AcHN O O

H (or Me) H N

N H

N H

O

O

TF (Thomsen-Friedenreich)

TN

HO

OH

AcHN HO

HO OH AcHN O HO HO

H (or Me) H N

CO2-

OH O O HO AcHN O O N H STN

H (or Me) H N

CO2O HO O OH HO OHO O O HO OH AcHN O O N H

O

H (or Me) H N O

2,6-ST

Figure 4.143 Structure of tumor-associated O-linked antigens.

ANTIBODIES

HO HO OH HO O HO OH HO O OH OH HO HO O O O O HO AcHN OH AcHN O Me O Me H H O N N AcHN N H O O Me O NHAc HO O O O OH OH 254

OH OH

AcHN HO

OH

HO OHHO O CO2CO2- HO O HO O HO OH O HO O HO OHO O O O HO AcHN OH AcHN O Me O H O H H O N N N OH N SH AcHN N H O H O O O Me O NHAc HO OH OH O O OH O OH HO OH HO O 255 OH O OH -O2C NHAc AcHN HO

O

HO OH

311

OH

OH

HO HO O O HO HO AcHN O AcHN O H O H N N AcHN N H O O O NHAc O

OH OH OH

O

O

N H N

O O

N O

O SH O

KLH

256

OH OH HO HO O O HO HO AcHN O AcHN O H O H N N AcHN N H O O O NHAc OH O OH OH

O H N

N S O

O

O KLH N H n = 317

257

Figure 4.144 Structure of trivalent cluster of TN , TF , and 2,3-ST-O-linked antigens.

When conjugated to a carrier protein, these small antigens increase antibody titers against STN [776]. Kuduk et al. [777] and Schwarz et al. [778] reported e‰cient routes leading to the synthesis of trivalent clusters of mucin-related Oantigens (254 to 256), which present three copies of the TF , STN , or TN antigen on a peptide backbone (Figure 4.144). The trivalent TN cluster 256 is covalently linked to the KLH carrier protein using MBS (m-maleimidobenzoic acid N-hydroxysuccinimide ester), a¤ording 257, which displays about 317 copies of the sugar cluster per protein sca¤old. Immunization of mice with 257 induces high IgM and moderate IgG titers against TN . The anti-TN antibodies recognize the surface of TN -positive cells, inducing complement activation and cell lysis.

312

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS TF epitope

258

OH OH HO OHO O O HO OH AcHN O

S

H N

H N

TF epitope

259

OH OH HO OHO O O HO OH AcHN O OH OH HO OHO O O HO OH AcHN O OH OH HO OHO O O HO OH AcHN O OH OH HO HO O O O HO OH AcHN O

BSA n

O

O S

NH H N

S

NH O

S O

O

NH O

O

O

NH

S

OMe

NH O N H

S

S

NH O

Figure 4.145 Structure of multivalent TF -antigen presented in neoglycoconjugate and glycodendrimer.

4.8.1.6 TF -Presenting Dendrimers. A neoglycoprotein conjugate (258) presents multiple copies of TF carbohydrate epitopes in a form that is not built on a peptide. It induces TF -specific monoclonal antibodies that are used to detect breast adenocarcinomas (Figure 4.145) [779]. Roy et al. [780] reported using the TF carbohydrate epitope in designing multivalent glycodendrimers based on an N,N 0 -bis(acrylamido)acetic acid core or an l-lysine dendrimer. One such example is 259, a tetravalent TF dendrimer. This dendrimer potently inhibits the binding of a TF -specific monoclonal IgG3 antibody to a multivalent array of TF carbohydrates, with an IC50 value of 18 nM. This activity represents a 128-fold enhancement relative to that of the TF monomer. 4.8.1.7 Neoglycoproteins Presenting Multiple Tumor-Associated Antigens. The presentation of multiple types of antigens on a single molecular construct might elicit multiple, distinct antibodies that could target a range of tumorassociated antigens (Figure 4.146). Borman [781], Allen et al. [782], and Williams et al. [783] developed this strategy in connection with cancer immunotherapy. Representative examples include KLH-based synthetic vaccines (260 and 261), in which a set of three cancer cell antigens, such as globo-H, TN , and Ley -containing glycopeptides, are displayed simultaneously, either in a partially nonnative structure (260) or with the native mucin architecture (261) [784]. The synthetic construct, 260, is more immunogenic than the mucin-based construct, 261. Antibodies induced in response to 260 recognize three antigen

313

ANTIBODIES OH OH OH HO OHO O HO O O O HO O NHAc HO O H3C O OH HO OH HO OH Globo-H

O

HO O HO

O

HO OH O HO AcHN O

O

OH

OH

H H3C OH N AcHN O OH O OH OH OH HO OH HO O O O O HO O O O Ley O NHAc OH H3C O OH OH OH

α-TN

O

O

H N

N H

O

H N

O

N

S O

N H n

O

260

HO OH HO OH HO OH O O O α-TN HO O HO HO AcHN AcHN O O H O H H3C OH N N O AcHN N OH H O OH O OH OH HO HO O HO AcHN O O O O HO O O O O O NHAc OH Ley HO OH H3C O OH 261 OH OH TF

O O

H N

N

S O

O

N H n

OHOH AcHN O OH OH

T-cell epitope of tetanus toxin H N

H2N-Y-S-Y-F-P-S-V O

O O

O

O

N H

CO2-

OH O O HO AcHN O H O N N H O

Sialyl-TN (cancer cell antigen) H N

S-A-P-D-T-R-P-A-P-CO2H

O

262

Figure 4.146 Structure of multiantigenic carbohydrate-based vaccine.

components and react strongly with specific tumor cells expressing each of the three tumor-associated antigens. In a separate approach, Keil et al. [785] reported a divalent vaccine construct (262) that contains a sialylTN mucin peptide tethered to a peptide epitope derived from tetanus toxin with a flexible PEG spacer. This heterodivalent vaccine induces a cytotoxic T cell response. 4.8.2

Surface Epitopes Not Related to Cancer

Following the process of xenotransplantation, in which an animal organ is transplanted into a human, acute rejection occurs due to the interaction of antibodies in human serum with epitopes on the surface of the nonhuman cells [786]. For example, xenoactive antigens on porcine endothelial cells include a-

314

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS OH HO O OH OH HO HO HO O O O HO OH

OH OH HO O O O HO OH NHAc O

O OH

endothelial cell surface

α-Gal epitope

O BSA

N H

H N

S

O

L-Sialic acid (L-NeuAc) OH OH CO2H O NHAc O HO HO 12

263

Figure 4.147 Structure of a-Gal epitope on animal endothelial cell and a multivalent glycoprotein presenting an unnatural optical isomer of sialic acid, L-NeuAc.

galactosyl epitopes. These are commonly expressed in the form of Gal-a1,3-Gal residues (Figure 4.147). The interaction of this epitope with human anti-a-Gal antibodies needs to be minimized to reduce rejection following xenotransplantation. Prior e¤orts at blocking such interactions include either decreasing the level of anti-Gal antibodies in serum through a‰nity column chromatography, or inhibiting such immune reactions using a-Gal oligosaccharide antagonists [787]. Liu et al. [788] created a recombinant mucin–immunoglobulin conjugate, which displays 140 copies of O-linked Gal-a1,3-Gal epitopes per conjugate. This multivalent galactoside is highly e¤ective in absorbing antipig antibodies. The technical aspects of synthesizing an a-Gal inhibitor were addressed by Fang et al. [789] in a report describing a chemoenzymatic approach that makes it possible to prepare a-Gal-containing pentasaccharide such as Gala1,3Galb1,4-GlcNAcb1,3-Galb1,4-Glcb-R. Kozlov et al. [790] reported a multivalent construct that presents unnatural, optical isomeric sugars. Such molecules are prepared in combination with phage display selection to identify their mirror-image receptors. A neoglycoconjugate (263) is composed of the carrier protein BSA possessing multiple copies of l-sialic acid (l-NeuAc), a sugar that is the mirror image of natural sialic acid (d-NeuAc). This multivalent molecule binds anti-NeuAc(d) antibodies five to 30 times less tightly than does an equivalent construct composed of d-NeuAc. 4.8.3

Miscellaneous Epitopes

Although a variety of multivalent immunoconjugates have been reported, we conclude this section by discussing a unique class of multivalent conjugates which provide important applications in (bio)chemical reactions. By displaying transition-state epitopes on their carrier surface, these conjugates are designed to induce a novel class of antibodies called catalytic antibodies [791–793]. These antibodies, as reported by Schultz and Lerner [792,793], are capable not only of

CELL SURFACE ENGINEERING + Me NH O

+ Me NH O OMe O O

+ Me NH O OMe

O -O

enzyme or catalytic Ab

cocaine

OMe

O OH breakdown of cocaine

KLH

O

H N

Spacer

H N

KLH n

+ NH O

Spacer O

n

OMe

O O P -O O

O P -O O

264 (cocaine hydrolysis) Spacer = (CH2)5, (CH2)9

O

O HO

N H

N+ CH3

N H

HO

+

OH

tetrahedral TS

+ Me NH O

315

265 (cocaine hydrolysis) Spacer = (CH2)3, (CH2)5CONH(CH2)2 KLH

O

N n H

266 (glycoside hydrolysis)

H3C

Si

CH3

ON+

H N

O

KLH n

N H

267 (phosphate ester hydrolysis)

Figure 4.148 Antibody-catalyzed hydrolysis of cocaine and synthetic multivalent conjugates presenting a transition-state mimetic in representative hydrolysis.

recognizing their epitopes but also catalyzing reactions such as those of enzymes. For example, 264 and 265 are designed to represent epitopes relevant to cocaine hydrolysis (Figure 4.148) [794,795]. A broad range of chemical reactions [796–827] are catalyzed by catalytic antibodies raised by such multivalent epitopes, ranging from peptide-bond cleavage to pericyclic reactions that include glycoside hydrolysis (266) [796,797] and phosphoester hydrolysis (267) [800–802].

4.9 4.9.1

CELL SURFACE ENGINEERING Metabolic Incorporation of Unnatural Carbohydrate Antigens

N-Acetyl neuraminic acid is involved in diverse cellular functions, including cell–cell recognition and antigenicity. It is expressed abundantly at 106 to 109 copies per cell in glycoconjugate form [828]. A series of cellular events is involved in its biosynthesis and cellular expression: (1) enzyme-mediated metabolic conversion of N-acetylmannosamine to N-acetyl neuraminic acid, which occurs in the cytoplasm; (2) subsequent glycoconjugation at the Golgi complex; and (3) tra‰cking to cell surfaces (Figure 4.149). Based on the sialic acid biosynthetic pathway, Kayser et al. [829] reported that nonphysiological sialic acids, such as N-propanoyl neuraminic acid, could be expressed on cell surfaces

316

MULTIVALENT MOLECULES APPLIED TO CELLULAR TARGETS N-Alkanoylmannosamine

N-Alkanoylneuraminic acid

O

HO OH CO2H H3Cm(CH2) N O OH OH O O

(CH2)mCH3 HO HN O OH HO (m = 0, 1, 2) HO

cell surface n

cytoplasm

metabolic conversion to unnatural sialic acid

H3Cm(CH2) O

OHOH CO2H N O OH OH OH

H3Cm(CH2) O

OHOH CO2H N O OH OH O

incorporation to glycoconjugate Golgi complex

Figure 4.149 Strategy for expression of unnatural carbohydrates on a cell surface via metabolic biosynthesis.

by feeding cells N-propanoylmannosamine instead of N-acetylmannosamine. This metabolic strategy provides a convenient way to modify the acyl group present in N-acetyl neuraminic acid in vivo, thus facilitating the modulation of biological functions carried out by natural sialic acids. 4.9.1.1 N-Butanoylsialic Acid. Poly-a2,8-sialic acid (PSA) is a linear homopolymer composed of about 100 or fewer sialic acids linked iteratively in a2,8linkage (Figure 4.150). It is found mainly on the surface of neural cells as a part of neural cell adhesion molecules (NCAMs) and also on tumor cells, such as

O HO HO HO HO O HO HO HO

HN CH3 O OH

AcHN

OH

N-butanoylmannosamine

CO2O OH OH

HN CH3 O OH

O OH CO2AcHN

O OH OH HO

N-acetylmannosamine polysialic acid (PSA)

AcHN

O

m = 100 CO2-

O OH OH

O n

neuron or tumor cell

Figure 4.150 Metabolic biosynthesis of poly(a2,8-sialic acid) expressed on neural and cancer cells.

CELL SURFACE ENGINEERING

317

neuroblastomas and small cell lung carcinomas, as a cell surface marker. In a natural system, the biosynthesis of PSA relies on the conversion of Nacetylmannosamine to CMP-sialic acid. Interfering with such a pathway is expected to alter the expression of PSA. Mahal et al. [830] and Borman [831] at Berkeley demonstrated that incubation with N-butanoylmannosamine, an acyl analog of N-acetylmannosamine, led to inhibition of the expression of PSA on neural cells. The inhibitory activity is ascribed to the generation of Nbutanoylsialic acid through a metabolic mechanism and the termination of the PSA biosynthetic chain. 4.9.1.2 N-Levulinoyl Neuraminic Acid. Mahal and Bertozzi [832] applied the metabolic incorporation strategy to engineering cell surfaces with nonnatural antigens, including various carbohydrates, nonphysiological functional groups, and synthetic conjugates. For example, the expression of nonnatural Nlevulinoyl neuraminic acid (SiaLev) is induced on cell surfaces at a density of 1.8 10 6 copies per cell (268) when cells are fed with a respective metabolic precursor, N-levulinoylmannosamine (Figure 4.151) [833]. This SiaLev conGal-LevNeuAc O HO HN O OH O HO HO

OH

CH3

HO HO

R1-ONH2

N-levulinoylmannosamine

O O OH N H 3C

HO OH CO2H N O OH OH O n O 269

metabolic biosynthesis

Lac-LevNeuAc OH O O

OH OHOH CO2H N O OH OH O n O

O H3C

HO HO

O O HO OH

O HN

OH

N

H3C

cell

HO OH CO2H N O OH OH O n O

R2-CONHNH2

268

270

O HN

NH

R4-ONH2 S R3-CONHNH2 O

OH

OHO N

O OHO

Biotin-LevNeuAc H N O H 3C

N


S N N

O OH

N H

N H

O N H3C

OHOH CO2H N O OH OH O n O

271

HO MRI contrast reagent (DPTA-labeled)

272

Figure 4.151 Structure of N-levulinoyl neuraminic acid (SiaLev) presented on a cell surface and its derivatization based on ketone-selective reaction.

318


tains ketone functionality in its N-acyl moiety and is able to react selectively with aminooxy- and hydrazide-terminated functional groups. Thus, the surface displaying SiaLev undergoes chemoselective ligation to an oxime or acylhydrazone moiety and enables the attachment of additional molecular entities at a SiaLev site. This type of reaction leads to the display of various novel epitopes with defined surface densities, as illustrated by nonnatural oligosaccharides (269 and 270) [834], biotin-labeled molecules (271) [835], and DTPA-labeled molecules (272) [836]. Artificial cell surfaces presenting nonnatural sialosides, such as 268, show neoantigenic properties that upon treatment with anti-SiaLev antibody are subject to complement-mediated cell lysis. For example, cellular surfaces derivatized with biotin (271) show selective binding to a disulfide-linked avidinricin A bifunctional conjugate. Such ricin-bound cells are killed by the toxic activity of the conjugate, which is initiated after cellular internalization of the conjugate. Through another metabolic strategy, delivery of magnetic resonance contrast reagents to cell surfaces is e¤ectively achieved by the selective introduction of DTPA on cellular surfaces. In another application, Lee et al. [837] described a strategy useful for viral gene transfer to cells that are otherwise rarely infected under normal conditions. For instance, NIH-3T3 cells that are metabolically induced to present SiaLev on their surfaces react with biotinhydrazide and are then coupled with an avidin-linked antiadenovirus antibody. This modified cell surface, which presents a multivalent array of adenovirus antibody, shows enhanced binding to adenovirus and leads to an increase in the level of viral gene transfer. O N3

HO HN O OH HO HO

N3

metabolic N-azidoacetylmannosamine biosynthesis

OHOH CO2H cell N O OH OH O n O 273

Ph Ph P

RO

O

H N O

HO OH CO2H OMe N O N OH OH O n Ph O Ph P O aza-ylide

OMe

Staudinger H reaction RO N

RO

HO OH CO2H Ph N O N OH OH O n Ph P O O OMe H N O

O Ph O Ph P O

O H 2O

HN

NH H N

-MeOH S

O

O

O

N H

H N


O

O

274 R

Figure 4.152 Expression of azidosialic acid on a cell surface and its conversion to biotin derivative via a Staudinger reaction with biotinylated phosphine.

CELL SURFACE ENGINEERING HO OH O O HO O OMe HO O OH OH OH OH

H N

H O-GDP O OH HO OH

OHO

O

HO O O RO HO n AcHN

+

319

cell

GDP-fucose

Blood group B trisaccharide

Lewis α1,3/4fucosyltransferase

HO OH O O OHO HO O OMe HO O OH OH OH OH

HO RO O

H N O

O O n NHAc

cell

O OH HO OH

275

Figure 4.153 Enzymatic transfer of trisaccharide epitope to a cell surface for multivalent presentation of artificial glycoconjugate.

4.9.1.3 Azidoacetyl Neuraminic Acid. The Staudinger reaction that occurs between an azide and a phosphine provides a tool for the selective modification of cell surface antigens. A research group led by Bertozzi [838–840] produced a Staudinger reaction on the surface of Jurkat cells expressing azidoacetyl neuraminic acid (273) (Figure 4.152). The nonnatural neuraminic acid is generated by the cellular machinery upon the addition N-azidoacetylmannosamine. Upon treatment with a water-soluble biotin-linked phosphine, it reacts with the phosphine and is converted to a biotin-tethered neuraminic acid (274). The mechanism for this type of a Staudinger-based ligation reaction includes for-

O

HO H N

NN+ N

H N O

O

CO2H

+ 60

O

117 eq. N H

cowpea mosaic virus HO

O

O

O

CO2H i [3+2] cycloaddition

H N

H N O

N N N

O

N H

O

60

276 reagents and conditions: i) CuSO4 (1 mM), copper wire, tris(triazolyl)amine (2 mM), phosphate buffer, pH 8, 4oC, 16 h (94%).

Figure 4.154 Multivalent display of fluorescein tag on the surface of cowpea mosaic virus.

320


mation of an azaylide intermediate and its subsequent conversion to an amide through hydrolysis. This example demonstrates the biological utility of physiologically orthogonal chemistry in connection with cell surface engineering. 4.9.2

(Bio)chemical Modification of Cell Surface Antigens

Multivalent presentation of ligands on cell surfaces is also possible by performing chemical and biochemical reactions occurring on natural surfaces. Enzymatic synthesis provides a good example, as described in a report by Srivastava et al. [841] that describes the decoration of cell surfaces with nonnatural sugars (Figure 4.153). This approach relies on fucosyltransferase activity, where a preassembled blood group B trisaccharide, linked to fucose GDP, is transferred regioselectively to a hydroxyl group at the C3 position of GlcNAc present in a glycoconjugate, a¤ording 275. The transfer reaction enables selective decoration of cell surfaces by presenting multiple copies of the blood group B trisaccharide. The surfaces of virus particles carry certain functional groups that can be chemically modified by reactive reagents such as thiols, maleimides, and succinimidyl esters. Based on this chemistry, Gillitzer et al. [842] reported the multivalent presentation of synthetic ligands on the surface of a viral protein cage. Huisgen’s [3 þ 2] cycloaddition, which occurs between an azide and acetylene, is used to derivatize the surface of a virus with functional molecules. Wang et al. [843,844] reported such a chemical method enabling the multivalent display of fluorescein molecules on the outer surface of cowpea mosaic virus (CPMV). This RNA virus is composed of 60 identical of protein units assembled around the genomic RNA (Figure 4.154). The method begins by derivatizing the protein comprising the virus with azide-presenting amide groups on its surface, resulting in 60 azides per virion. The surface azides then undergo a copper(I)catalyzed cycloaddition reaction with free alkyne terminated in fluorescein, generating a nonnatural virus conjugated with 60 copies of fluorescein tag per virion. This type of CPMV surface modification includes derivatization with cysteine side chains, which can be used in thiol-specific reactions, such as disulfide exchange and chemosorption to metallic gold [845].

5 CHEMICAL SYNTHESIS OF MULTIVALENT MOLECULES

5.1 5.1.1

SELECTED SYNTHETIC METHODS FOR MULTIMERIZATION Summary of Linker Chemistry

Linker chemistry serves as a tool to interconnect two or more biologically active fragments or pharmacophores or to attach them to a certain sca¤old. As illustrated by numerous examples in previous chapters, linker chemistry comprises a critical component in the chemical design of multivalent molecules. In this section we review briefly a list of linker chemistries used in multivalent synthesis. A carboxylate ester linkage is not used often because of its hydrolytic instability. It is nonetheless important when such functionality has to be installed to maintain the biological activity of the desired multivalent molecules. Representative examples include the synthesis of a phorbol ester dimer [846]. An amide linkage is more stable than an ester linkage and is thus used more generally in synthesizing multivalent molecules [214,572,790,848–858]. Synthetic strategies to construct amide linkages consist of conventional coupling methods as well as novel methods, such as a modified Staudinger reaction [838,839]. An ether linkage results from the preparation of divalent molecules by invoking processes such as Williamson ether synthesis [635,861], Mitsunobu coupling [847], other methods employing catalysts such as Ag2 CO3 [859] and ZnC12 [862,863], and using an activated glycosyl donor [788,860,864–866]. Thioether linkages can be prepared in a similar manner, via the Williamson method [867– Synthetic Multivalent Molecules: Concepts and Biomedical Applications, By Seok-Ki Choi ISBN 0-471-56347-1 Copyright 6 2004 John Wiley & Sons, Inc.

321

322

CHEMICAL SYNTHESIS OF MULTIVALENT MOLECULES

869], Michael-type 1,4-addition of a thiol to an activated vinyl-containing group [870], and photocatalyzed addition of an alkyl thiol [871,872]. Amine linkages may be prepared by N-alkylation of amines [351,634], N-reductive alkylation [638,873–876], and nucleophilic displacement of aromatics [666] or methyl squarate [292,443]. Sulfonamide linkages result from the N-alkylation reaction of a sulfonamide [647]. Urea [2,879–881] and thiourea linkages may be formed from reaction of amines with isocyanate [877] or thioisocyanate [878], respectively. Additional examples of covalent linkages are based on functionalities such as carbamate [299,551,662,663], oxime [882–884], and azo groups [885]. Certain well-known (i.e., ‘‘named’’) reactions are excellently suited for multivalent synthesis. As a representative of multicomponent condensation reactions, the Ugi reaction occurs in a mixture of four functionally distinct components—acid, aldehyde, amine, and isocyanide—that react together to generate a diamide [886–888]. Wittig olefination is used less frequently in the synthesis of multivalent molecules because of its more rigorous reaction conditions, but still constitutes a valuable approach [889,890]. A Diels–Alder cycloaddition occurring between a diene and a dienophile accounts for the formation of a six-membered ring [41]. For example, this type of cycloaddition strategy is used for the regiospecific synthesis of a dimeric taxoid [891], which targets the b-tubulin component of microtubules. The dimeric taxoid presents multiple taxol binding sites, as determined by photoa‰nity cross-linking [892– 894]. In a dipolar 1,3-cycloaddition, alkynes undergo a cyclization reaction with a dipolar species such as an alkyl azide (R-N3 ) and a nitrile N-oxide (R-CNO), resulting in the formation of a triazole [895] and an isoxazole [896], respectively. The cycloaddition of azide with an unactivated simple alkyne leading to a triazole is also possible by use of an organic-soluble copper catalyst, as illustrated in the synthesis of multivalent neoglycoconjugates [897]. Olefin metathesis ligation is a well-known catalytic process with numerous practical applications in (bio)polymers. It occurs between olefinic reactants presenting terminal alkene moiety, and as a result generates an olefin-ligated dimer [898]. Descotes et al. [899] demonstrated the application of this type of ligation by introducing olefin metathesis in the synthesis of dimeric molecules such as a glucoside dimer. The synthesis is catalyzed by a chloroaryloxide tungsten complex. However, this reaction requires the use of an organic solvent and a protected sugar. Using a potentially more flexible system, a group led by Kiessling [900,901] reported the practical application of olefin metathesis in the synthesis of multivalent carbohydrates by employing Grubb’s catalyst under conditions that permit the use of unprotected carbohydrates in aqueous solvent. This strategy was later applied to the synthesis of diverse multivalent molecules, such as a vancomycin-derived olefinic dimer [228], a divalent sialoside [902], and an olefin-linked 1a,25-dihydroxyvitamin D3 dimer [903]. In addition, alkene metathesis is readily applied to solid-phase synthesis, as demonstrated by Olenyuk et al. [714] in their preparation of H-shaped polyamide dimers tethered with an aliphatic linker of varying lengths.

SELECTED SYNTHETIC METHODS FOR MULTIMERIZATION

323

The disulfide linkage readily forms from the reaction of twothiol functionality, under mild oxidative conditions or through thiol–disulfide exchange [904]. This class of linkage makes it possible to generate a library of dimers derived from, for example, vancomycin [228] or psammaplin A [338], and to prepare melamine-based dendrimers bearing multiple copies of disulfide-linked octapeptides and captopril [904]. 5.1.2

Polymerization

Among various approaches that enable the synthesis of multivalent ligands, polymerization is highly e¤ective for rapidly creating multiple units of ligands in long linear chains. Salpenstein and Whitesides demonstrated the use of a polymer as a multivalent sca¤old by designing polyvalent sialoside based on a synthetic polymer. The multivalent product was identified as a potent inhibitor of influenza virus adhesion. Free-radical copolymerization of acrylamide with a derivative that contains sialic acid generates a polymeric, polyvalent sialoside (1) (Figure 5.1) [36]. Varying the ratio between the two acrylamides enables controlling the density of sialic acid on the polymer chain. Such direct polymerization commonly results in a mixture of homologous polymers of variable composition and length, as influenced by the reaction conditions and the nature of the monomers. Thus, such polymers tend to display polydispersity, with nonuniform size distribution, and polytacticity, or variation in side-chain arrangement. A postmodification approach can be used to reduce the incidence of these problems. As illustrated in Figure 5.1, a postmodification strategy begins with the preparation of the preformed polymer containing an activated group, such as poly(N-acryloyloxysuccinimide) (2) [34]. The preformed polymers are more homogeneous because of homopolymerization, and they may fractionate on the basis of molecular size by gel permeation chromatography. Sequential reaction of the NHS ester present on the polymer chain with primary amines derived from sialoside, benzylamine, and then ammonia leads to a terpolymer (3) which displays each of the respective groups as side chains. This derivatization method provides a convenient way to control the density of ligands, simply by varying the ratio of amine reactants to NHS ester. The copolymerization technique or the postmodification approach are being used with increasing frequency to prepare molecules such as a polyvalent lactoside [905] and a polyvalent muramyl dipeptide [528], both of which are linked through amide side chains on poly(acrylamide). A subclass of olefin metathesis, the ring-opening metathesis polymerization (ROMP), o¤ers an e‰cient strategy for polymerizing substrates containing internal olefinic bonds. This method was introduced for multivalent synthesis by a group led by Kiessling [507], as illustrated in Figure 5.2. For example, 7oxonobornene, presenting two tethered unprotected glucosides, is converted to a polymeric, polyvalent glucoside (4) through ROMP catalyzed by a ruthenium catalyst in aqueous solution. Copolymerization of vinyl alcohol and a vinyl amide monomer is also used for polyvalent presentation, as reported in the synthesis of polyvalent penicillin G and V [327,328] and in the preparation of a

324

CHEMICAL SYNTHESIS OF MULTIVALENT MOLECULES Copolymerization OH

HO AcHN

CO2O OH OH

H N

O

O

+

i

H 2N O

O NH2 O O

HN

OH

HO AcHN

NH2 O

CO2O OH OH

n O O NH2 NH2

HO

O

O

O

O O O O O HN NH2 NH2 NH2 NH2 NH2

HO

AcHN

-O2C O O OH OH

O

1

Postmodification NHS O

O O O

ii

N

O N-acryloyloxysuccinimide

O

O N

O

O

O R

R

O R

O R

n O

O R

R

O 2 (R = NHS)

N-hydroxysuccinimide (NHS)

HO AcHN

NH2

OH

NH2 O

CO2O OH OH

S HO

iii,iv

HN

OH

AcHN

O

n O O O O O O O O O NH NH NH2 NH NH2 NH2 NH NH2 NH2 2

CO2O OH OH

S

S -O2C

3

O

HO

OH HO AcHN OH reagents and conditions: i) 4,4'-azobis(4-cyanopentanoic acid), hν (365 nm), 25oC, 5 h; ii) azoisobutyronitrile, benzene, 60oC, 24 h; iii) Et3N, DMF, rt; iv) benzylamine, then NH4OH.

Figure 5.1 Synthesis of polyvalent sialoside by either copolymerization of acrylamide monomers or postmodification of preformed activated polymer.

O O O HO HO HO

OH

n

O O

O HO

O

ROMP O

O HO HO HO

RuCl3 H2O, 55oC

HO HO

OH

HO

O

O O

O HO

4

O

HO HO HO

Figure 5.2 Synthesis of polyvalent polymeric glycoside by ring-opening metathesis polymerization.

COMBINATORIAL CHEMISTRY

325

polyanioic adhesion inhibitor of the human immunodeficiency virus (HIV) [155]. 5.1.3

Self-Assembly

A prototypical model of self-assembly, the liposome provides a water-exposed outer surface on which ligand molecules are displayed in multivalent fashion, thereby mimicking the native cell surface. An early example of multivalent inhibitors of influenza virus was designed on a liposome sca¤old, resulting in a strong inhibitor of virus-cell adhesion [44]. This system has been used widely for the preparation of multivalent ligands and inhibitors binding to a broad range of targets. Certain liposomes can be polymerized within their hydrophobic interior, thus converting them to irreversible and stable liposomes [98,99]. Known examples of such polymerized liposomes include those assembled from diacetylene-incorporated lipids presenting sialic acid, lactoside, glucoside, biotin, and amino acids on their polar terminus. Self-assembled monolayers (SAMs) on gold are prepared by the chemosorption of alkanethiolates. This convenient system allows for multivalent expression of a diverse class of biological ligands on the SAM surface [906]. For example, a SAM-based biosurface has served as a critical tool for studying ligand–receptor interactions, such as carbohydrate–lectin [908] and carbohydrate–toxin [41], and as a model surface for cell–pathogen attraction [42]. The technique has even been applied to the display of nucleic acids [907]. 5.1.4

Solid-Phase Synthesis

In a change from conventional solution-phase synthesis, solid-phase synthesis builds molecules while they are covalently attached to the surface of a solid support, to provide synthetic and processing convenience. It remains a practical way for synthesizing an array of complex natural biomolecules, such as peptides [909], nucleic acids [910], and sugars [911,912], as well as synthetic molecules such as vancomycin [913] and polyamide dimers [714]. The solid-phase display of ligand molecules is considered multivalent by definition, and as such is often used directly for examining receptor–ligand interactions, as illustrated by the study of lectin–sugar binding [914].

5.2 5.2.1

COMBINATORIAL CHEMISTRY Structure-Biased Combinatorial Synthesis

5.2.1.1 Hetrodimeric Inhibitors of Picorna Virus. A causative agent of the common cold, picorna virus is a nonenveloped RNA virus with an icosahedron shape [915]. Virus–host cell adhesion is mediated by a viral capsid protein, VP1, located in a ‘‘hydrophobic canyon’’ on the outer surface of the virus

326

CHEMICAL SYNTHESIS OF MULTIVALENT MOLECULES aromatic domain II

pore

O N

aromatic hydrophobic domain I N O spacer O H 3C

binding pocket of viral capsid (VP1) of picorna virus Hetrocycle core

(Cl)H

Linker

N

Br

Br

SH S N

Br Br

SH N SH N

K2CO3 DMF, rt

+

Br

Br

X X = CH, N

Me

thioether-based dimers (75-member library)

S

Linker

S Heterocycle

Heterocycle

example:

N SH N Me 2 eq

N Br

Br 1 eq

S S

N

S

Me N N

5

Figure 5.3 Hydrophobic pocket of VP1 capsid protein in complex with heterobivalent inhibitor, and construction of a divalent library.

(Figure 5.3). Blocking of this adhesion site constitutes a valid approach for inhibiting the virus. According to the x-ray structure of the picorna virion capsid protein, the floor of the canyon appears to present two neighboring domains that can be occupied by an inhibitor composed of two aromatic groups tethered with a hydrophobic chain. From a structural standpoint, an e¤ective inhibitor is likely to be heterodimeric. There are numerous types of capsid inhibitors that comprise two heteroaryl groups connected by a four- to six-atom linker. As reported by Tsang et al. [174], a group of collaborators led by Whitesides and Hogle at Harvard [174] developed a combinatorial strategy enabling the e¤ective synthesis and biological identification of divalent picorna inhibitors. As summarized in Figure 5.3, the library is generated by the formation of a thioether linkage between a heteroarylthiol, with five heterocycles, and a dihalide (as source of linker) coupled in a 2:1 molar ratio. The entity of such a heteroaryl head and spacer is chosen on the basis of the structure of known inhibitors. The resulting library contains a mixture of 25 homodimers and 50 heterodimers, based on theoretical calculation. Without isolating individual compounds, the library is evaluated as a mixture in an antiviral assay. The assay begins by incubating the entire library with picorna virus and relies on mass spectrometric analysis for identifying molecules bound to the virus. This struc-


327

ture-based approach enabled the rapid discovery of a number of tight-binding inhibitors, such as heterodimer 5. 5.2.1.2 Polyvalent Inhibitors of Influenza Virus. Influenza virus initiates cellular infection by adsorption to cell surfaces [81]. The virus-cell adhesion is mediated through multivalent interactions between hemagglutinin—the viral receptor—and sialic acid ligands present on the cell surface. Inhibition of this interaction constitutes a rational approach for interfering with the first step in viral invasion. Among a variety of influenza inhibitors that target hemagglutinin, polyvalent polymeric sialoside is the most e¤ective in blocking the viral adhesion in vitro [34]. The potent blocking activity is attributed primarily to tight binding of the polymer to a viral surface and is achieved through interaction between multiple receptor–ligand pairs (Figure 5.4). In addition, a steric

polyvalent inhibitor

sialoside NA inhibitor

hemagglutinin (HA)

neuraminidase (NA)

membrane influenza virus

O

O

O

i

O

O

HO

O O

O

+

O

AcHN

OH CO2O OH OH

S

R-NH2 NH2

ii

poly(acrylic anhydride)

poly(acrylic acid) (NeuAc; R)

HO AcHN

OH

HO

O HN

CO2O OH OH

O

n O O O O O O O O OH NH OH OHHN OH NH OH R R

S HO HO AcHN

-O2C

6

S

O OH OH

(R-NH2 = amino acids, primary amines, aminosugars) reagents and conditions: i) azoisobutyronitrile (AIBN), benzene, 60oC; ii) water, sonication, 30 min.

Figure 5.4 Inhibition of influenza virus by polyvalent molecule and combinatorial generation of terpolymeric sialoside presenting secondary binder.

328


mechanism, mediated by the hydrated layer of a polymer chain positioned between the two surfaces, further impedes the virus–cell attachment (see Section 2.1.1.12) [35]. Hemagglutinin, which constitutes about 80 to 90% of total viral proteins, serves as a primary drug target on the surface of the influenza virus. However, less abundant secondary proteins are still accessible, such as neuraminidase and M2 channel proteins. This heterogeneous array of surface macromolecules lends itself to the use of multivalent inhibitors to target hemagglutinin and secondary proteins simultaneously. An example reported by Choi et al. [112] illustrates this concept (Figure 5.4). The figure shows an array of polyvalent molecules that are designed to present sialoside as a primary ligand and the R group as a secondary binder presented on the framework of poly(acrylic acid). This combinatorial library of terpolymers is synthesized by coreacting a library of amines (RNH2 ) and a modified sialic acid–containing amine with poly(acrylic anhydride) in aqueous solution. In a virus–cell adhesion inhibition assay, a set of secondary R groups are identified that enhance the antiviral activity of the terpolymers. For example, 6 poly(acrylic acid) (NeuAc; l-3-(2 0 naphthyl)alanine), a terpolymer presenting sialoside at approximately 5 mol% and l-3-(2 0 -naphthyl)alanine at approximately 6 mol%, is about 104 times more potent than a corresponding copolymeric poly(acrylic acid) (NeuAc). The Ki value of the terpolymer is approximately 0.5 nM. 5.2.1.3 Heterodimeric Inhibitors of NAD Synthetase. The structure-biased synthesis of dimeric molecules is applicable for enzyme targets. For example, Garcia et al. [349] reported a combinatorial library of divalent inhibitors that target nicotine adenine dinucleotide (NAD) synthetase (Figure 5.5). The structure of a representative inhibitor (7) is composed of three modules: indole (subunit A), which binds to an adenine site; pyridinium ion (subunit B), which recognizes a nicotine site; and an alkane tether that links the two units. Systematic variation of the three domains of the heterodimer makes it possible to discover divalent inhibitors more active than the parental units. 5.2.1.4 Disulfide-Linked Heterodimers. Certain natural products exist as dimers, as illustrated by marine-derived psammaplin A (Figure 5.6) [916,917]. It is structurally a disulfide-linked homodimer presenting two oxime units, and functionally a potent antibacterial agent against methicillin-resistant Staphylococcus aureus (MRSA). A team led by Nicolaou at Scripps [336] generated a series of disulfide-linked homo- and heterodimeric analogs of the antibacterial agent in an e¤ort to optimize the activity and study the mode of action of psammaplin. The design of the library is based on the variation of two domains, subunit A and subunit B, while maintaining the alkyl disulfide linker, as found in natural products. Heterodimers are generated through a disulfide exchange reaction that occurs between two di¤erent homodimers. This combinatorial synthesis led to the identification of a diverse class of divalent molecules more active than psammaplin A.

329

COMBINATORIAL CHEMISTRY NAD synthetase 7 O2N

O N+

N H

adenine site

OH

nicotine site

structure-biased library design

ii) linker variation

i) variation at subunit A

O2N

O N+

N H

iii) variation at subunit B

OH

heterodimeric inhibitor

Figure 5.5 Combinatorial approach for a divalent inhibitor of NAD synthetase.

OH

OH psammaplin A

Br

Br

O

HO

N

O N H

S S

N H

N

OH

subsite 2

subsite 1 bacterial target

F

F O

OH

OH

Br

Br HO

HO

N

O N H

N

N H

i) variation at subunit A

S S

N H

N

S S

N

OH

OMe

OH OH

N H

homodimer

O

O

F

F

Cl O

ii) variation at subunit B HO

N

N H

S S

S N O HO

heterodimer

Figure 5.6 Combinatorial strategy for the generation of disulfide-linked psammaplin A analogs.

330


5.2.2

Target-Guided Combinatorial Synthesis

5.2.2.1 FK-506 Binding Protein. In a conventional combinatorial approach, dimers are designed according to available structural and mechanistic information regarding a particular receptor–ligand interaction. Normally, these dimers are then generated in the absence of targets. A target-guided approach depends less on such knowledge because with this technique the target proteins assist directly in the identification of the monomeric building blocks. This is because the technique screens and selects for building blocks that are able to occupy multiple proximal sites present in the proteins. For example, Shuker et al. [563] reported a rapid target-guided method that makes it possible to identify high-a‰nity ligands to the FK506-binding protein (FKBP) (Figure 5.7). It is based on structure–activity relationship studies, with nuclear magnetic resonance used as a screening tool (SAR by NMR). The strategy begins by identifying a group of structurally homologous ligands that recognize a specific site in FKBP, such as 8 (Kd ¼ 2 mM ) (step 1). This method is applied repetitively to identify a second group of ligands, which recognize a nearby second site, such as the weakly binding 9 (step 2). Once the two discrete ligands are discovered as monomeric building blocks, a tether is introduced to connect the two ligands covalently, thus leading to a heterodivalent ligand (step 3). Variation of the linker further optimizes the binding activity of dimers. A heterodimer (10) generated in such a way binds FKBP very tightly, with a Kd value of

step 1: screen for, and optimize first ligand.

step 2: screen for, and optimize second ligand.

8 (Kd = 2 µM)

9 (Kd = 0.1 mM) O

N O

MeO

OMe OO

N O

OMe OMe

MeO

OMe HO OO

HN

OH

OMe OMe

FK506 binding protein (FKBP)

step 3: tether ligands, and optimize length. 10 (n = 1 to 4) O N O

MeO

O OO

OMe OMe

n O HN

OH Kd = 19 nM (n = 1) Kd = 32 nM (n = 2) Kd = 49 nM (n = 3) Kd = 228 nM (n = 4)

Figure 5.7 Discovery of heterodivalent ligand for FK506-binding protein by NMR screening.

331

COMBINATORIAL CHEMISTRY X

PG S

X

PG S

SH

tethered ligand screening of first library

PG S

S

deprotection

+ enzyme or receptor

R

X

PG S

X

PG S

S S

R SH

S

+

R

R

O Caspase

S HO2C

H N

S O

X = leaving group PG = protecting group

S

S S

dynamic screening of second library

S S

SS

S

Identification of divalent elements

synthesis of stable heterodimer

S S

O O S N H

O

CO2H H OH

HO2C

O O S N H

H N

CO2H OH

O divalent inhibitor

Figure 5.8 Discovery of heterodivalent inhibitor of caspase-3 by a tether approach.

19 nM, which represents more than a 100-fold enhancement in binding strength relative to the nontethered ligands. 5.2.2.2 Caspase. Another target-guided approach for discovering divalent molecules that target two proximal sites on a protein was reported by Erlanson et al. at Sunesis [918,919]. The method invokes covalent tethering based on the chemoselective reactivity of a cysteine thiol located at the active site of cysteine aspartyl protease-3 (caspase-3), an enzyme involved in programmed cell death (Figure 5.8). A library of binding elements for the first site is captured by the chemoselective reaction of a thiol with halide-derivatized monomer ligands after coincubation with the enzyme. This reaction occurs selectively even though the binding elements have weak a‰nity to the site as a free ligand. This first ligand is then modified to display a thiol, which then undergoes a disulfide exchange reaction with a second array of disulfide monomers, resulting in a dynamic equilibrium that favors the trapping of a more tightly binding second element. The identification of two binding elements to discrete enzyme sites, coupled with the knowledge of intersite distance, makes it possible to design a heterodimer linked with a more stable linker rather than disulfide. Using this strategy, the authors discovered high-a‰nity inhibitors of caspase-3, one ex-

332


O N R

O N

O N

O N

NR

NR

O N O N

c-Src kinase

N

X

R

R

R

O N R

X = O, NH, S

1. Synthesis of random monomer library, and screening

R'

O N

OH OH

Et

2. Identification of bound monomer elements

O N R1

(CH2)n

O N

n = 2 to 6

3. Heterodimer synthesis by tethering, and screening R2

O N

c-Src kinase

N

(CH2)2 11

O N

OH Et

OH

Figure 5.9 Target-guided combinatorial synthesis of heterodivalent ligands of c-Src kinase.

ample being a salicylsulfonamide linked to an aspartyl aldehyde via a tolulylamide spacer (Kd ¼ 0:2 mM ). 5.2.2.3 c-Src Kinase. The utility of target-guided dimer synthesis is demonstrated further in a study reported by Maly et al. [920] and Borman [921] at Berkeley in the discovery of c-Src kinase inhibitors (Figure 5.9). This approach begins with the identification of potential binding elements to c-Src kinase from a randomly generated library. For example, such a library is composed of 305 discrete members of oxime-derived compounds where the oxime functionality is introduced as a chemical handle. Each of these monomers is screened for its ability to bind to c-Src kinase, which leads to the discovery of a set of 37 binding elements, including those that interact weakly. These monovalent building blocks are then combinatorially cross-linked using a flexible alkane chain through an oxime linkage, thus generating a new dimer library composed of homo- and heterodivalent ligands. Screening these against the multiple kinases

333


leads to the discovery of a number of active inhibitors. For example, a dimer made of carbazole and catechol units (11) is highly active, with an IC50 value of 64 nM against c-Src kinase. The inhibitor is selective to c-Src kinase over Fyn kinase, which shows an IC50 value of 5 mM, and Lyn kinase, which shows an IC50 value of 13 mM. 5.2.2.4 Acetylcholine Esterase. A di¤erent approach to generating targetguided combinatorial dimers is illustrated by the discovery of inhibitors of acetylcholinesterase. This enzyme catalyzes the hydrolysis of the cationic neurotransmitter, acetylcholine, into acetate and choline. Structurally, it presents two distinct sites, a catalytic site and a peripheral site, located within roughly ˚ of each other [579]. Rostovtsev et al. [922], Demko and Sharpless [923], 12 A and Borman [924] at Scripps reported the application of a target template method aimed at the discovery of acetylcholinesterase inhibitors. Their strategy is based on combinatorial in situ click chemistry, utilizing a set of azides and a set of terminal acetylenes as monomeric building blocks (Figure 5.10). These monomers are selected as potential enzyme-binding elements. In an enzymefree reaction, a mixture of azides and acetylenes undergo random crosscoupling reactions according to Huisgen’s [3 þ 2] cycloaddition rule, which leads to the formation of diverse triazoles produced in a mixture of syn and

monomer building blocks H 2N

NH2

all combination (98 members) N+

H 2N

(CH2)l

N3

(CH2)n

N+

NH2

[3+2] cycloaddition

R1 R2

N N NH +

R1

N N NH

+ H N

R2 N3

(CH2)m

(CH2)o

H N

syn

anti

N

N

(l = 2 to 6; m = 1 to 3; n = 6, 7; o = 2 to 6)

acetylcholinesterase templated "click chemistry" a single product isolated NH2

H2N

N+

NH2 N N N

H N

[3+2] cycloaddition N

syn H2N

N+

N N N

H N

12

Figure 5.10 Huisgen [3 þ 2]cycloaddition–based combinatorial synthesis of acetylcholinesterase inhibitor.

N

334

H 2N

O S O

CHEMICAL SYNTHESIS OF MULTIVALENT MOLECULES NH2

S O

+

H 2N

O S O

NH2

Cl

13 (50%) H 2N S

O S O

O

SH +

CA II H2N

Cl CO2H

O S O

S CO2H 14 (92%)

CO2H 14 (50%)

Figure 5.11 Target-templated alkylation of thiol with alkyl chloride, generating carbonic anhydrase inhibitors.

anti regioisomers at roughly a 1:1 ratio. Assuming all possible combinations of the reactants, 98 compounds result. When an equivalent reaction is performed in the presence of acetylcholinesterase as template, only a single product (12) is generated as a syn isomer out of 98 possible combinatorial dimers. This heterodimer binds acetylcholinesterase very tightly and is the most potent inhibitor ever identified. 5.2.2.5 Carbonic Anhydrase. The concept of templated synthesis is also described by Nguyen and Huc [925] in the design of carbonic anhydrase inhibitors. Their strategy utilizes an enzyme template to form thioether products from two types of reactive monomeric building blocks. The reactants are composed of a thiol-containing sulfonamide as the primary inhibitor and an alkyl chloride as the potential secondary binding element (Figure 5.11). When a mixture of one sulfonamide-thiol and two types of alkyl chloride proceeds to competitive alkylation, in the absence of carbonic anhydrase II (CA II), the reaction produces a mixture of two thioether products (13 and 14) in equal proportion. When performed in the presence of the enzyme, the reaction forms almost exclusively a single product, (14; 92%). With a Ki value of 84 nM, this heterodimer shows about a ninefold higher a‰nity to the enzyme than does 13. The high-a‰nity ligand is probably produced from more tightly bound reactants, which are brought into close proximity within the enzyme under kinetically controlled conditions. 5.2.2.6 Kallikrein. Another example, described by Mosbach et al. [926], utilizes a template-directed approach in the generation of enzyme inhibitors by using a molecularly imprinted synthetic polymer instead of a biological template. Use of this method makes it possible to produce a micromolar inhibitor of the kallikrein proteinase in a fourfold larger amount than in the absence of the template. 5.2.3

Dynamic Combinatorial Synthesis

The traditional combinatorial approach of dimer synthesis is usually performed in the absence of a biological target and provides a collection of randomly


335

28-member library of dimers traditional combinatorial synthesis irreversible tethering

random irreversible reaction etc

+

monomer building blocks

dynamic combinatorial synthesis +

+

etc.

+

+

etc.

reversible tethering divalent target

+

+ accumulation of high-affinity dimers

Figure 5.12 Features that distinguish dynamic combinatorial synthesis from traditional combinatorial approach.

generated diverse molecules in a relatively short time. In contrast, dynamic combinatorial synthesis is carried out in the presence of a target and relies on a reversible tethering reaction that drives the reaction equilibrium to favor the formation of ligands that bind more tightly (Figure 5.12) [927–930]. Thus, dynamic synthesis enables a weighted generation of high-a‰nity dimers through interaction with the target. It also facilitates the isolation of a reduced number of selected soluble dimers. The dynamic method is, however, limited by the small number of available chemical reactions that are both reversible at chemical equilibrium and mild enough not to damage a biological target under the conditions of the reaction. However, examples of such reactions are known and

336


Sugar

H N

O

(CH2)5 O

disulfide exchange S

2

pH 7.4

10-member library Man-Man, Man-Gal, Man-Ara, Man-Xyl Gal-Gal, Gal-Xyl, Gal-Ara, Xyl-Xyl, Xyl-Ara, Ara-Ara

Sugar = mannoside, galactoside, xyloside, arabinoside Con A

Man-Man

HO HO HO

OH O

O

H N

(CH2)5 O

S S

H N

(CH2)5

15 (major)

O

O

OH OOH OH OH

+ HO HO HO

OH O

O

H N

(CH2)5 O

S S

H N

(CH2)5

O Sugar

O

Man-Gal, Man-Ara, Man-Xyl

Figure 5.13 Disulfide exchange–based dynamic combinatorial generation of divalent saccharides binding to concanavalin A.

include the exchange reactions occurring between disulfides [931], oximes [932], imines [933,934], and acylhydrazones [935] and also those involving metalion-mediated chelate formation [936] olefin metathesis [937]. 5.2.3.1 Disulfide Exchange. Disulfide interchange is suited for dynamic combinatorial synthesis because of the rapid and reversible equilibrium that occurs between disulfides andthiols under mild conditions (pH > 7) in aqueous solution. Ramstro¨m and Lehn [938] demonstrated the utility of such reactions in the synthesis of divalent saccharides that bind concanavalin A (see Section 4.1.16), a tetravalent lectin that recognizes a d-mannoside ligand (Figure 5.13). For example, a mixture of four disulfide-linked carbohydrate homodimers undergoes a disulfide scrambling reaction in the absence of the lectin, generating all 10 possible combinations of homo- and heterodimers in comparable ratios. When an exchange reaction occurs in the presence of concanavalin A, a dmannoside homodimer (15) is produced in higher proportions than from a control reaction with no lectin. Such an enhanced yield of the dimer is attributed to its particular a‰nity to concanavalin A, which is higher than that of the three other d-mannoside-containing heterodimers. 5.2.3.2 Olefin Metathesis. Olefin metathesis serves as another established route leading to the reversible dimerization of ligands in the presence of biological targets. Nicolaou et al. [228] reported experiments based on such a concept in the target-accelerated combinatorial synthesis (TACS) of dimers derived from vancomycin. This approach consists of an olefin metathesis-based (or disulfide) synthesis of vancomycin dimers, where the formation of dimer products in a library is monitored in the presence or absence of d-Ala-d-Ala


337

ligand, a bacterial peptidoglycan precursor known to promote noncovalent dimerization of vancomycin. The presence of the ligand selectively increases the rate of formation of those dimers that tend to be more potent in antibacterial activity. In olefin metathesis approach, the dimers generated represent a mixture of conformational isomers because their linker framework can take both trans and cis conformation. These active dimers need to be further characterized to find which isomer is more active. In disulfide approach, dimers generated are based on disulfide linkage, which is uniform but relatively unstable. In summary, either metathesis or disulfide approach enables to discover active dimers rapidly as well as to a¤ord information regarding optimal linker framework, on the basis of which second generation of dimers can be synthesized. 5.2.3.3 Transition Metal Complexes. Transition metal–based coordination chemistry also provides a combinatorial method for generating a dynamic mixture of stereoisomers whose ratio can be controlled by use of biological targets. Sakai et al. [939] demonstrated the utility of such an approach by using a GalNAc-derived monovalent ligand and its multivalent species formed through a transition coordination complex (Figure 5.14). In the target-free system, a reaction between three molecules of GalNAc-derived bipyridyl ligand and Fe(II) generates trivalent GalNAc presented as a coordination complex in a statistically predictable mixture of four stereoisomers, including 16. When performed in the presence of the lectin, Vicia villosa B4, the reaction produces a GalNAc-recognizing multivalent protein, 16, as a single major isomer (85%).

four steroisomers OH

OH Fe (II)

O HO AcHN O

N N

Fe

Fe (II) lectin (Vicia villosa B4) OH

Fe

+ Fe

Fe

OH O

HO AcHN O

N N

N Fe

O O NHAc HO HO

N

16 (15%) N

N

OH

16 (85%)

OH O NHAc OH O OH HO

OH

O HO AcHN O

=

Figure 5.14 Dynamic synthesis of transition metal complex–based trivalent GalNAc in the presence of lectin Vicia villosa B4.

338


Such high product selectivity is attributed to interactions between the lectin and the trivalent ligand during product generation, where the assembly reaction is expected preferentially to form trivalent ligands with higher a‰nity. This major isomer (16) shows the highest lectin a‰nity by factors of 3 to 18 among the four isomers.

APPENDIX

TABLE 1. Divalent Ligands Linked to Enzymes in Human Immunodeficiency Virus (HIV) Enzyme HIV-1 protease

Multivalent Ligand

Notes

Section

Divalent protease peptide linked with glycine spacer Divalent protease peptide linked with 10- to 16atom spacer Divalent tripeptide linked to pyridine-based sca¤old

Inhibition of protease dimerization Inhibition of protease dimerization (IC50 ¼ 2 mM ) Inhibition of protease dimerization (IC50 ¼ 2 mM ) Potent inhibition of HIV-1 proteinase (Bioorg. Med. Chem. Lett. 1993, 3, 503) No synergy observed

2.2.1.1

Divalent penicillin G derivative linked with 3- to 4-atom spacer HIV-1 reverse transcriptase

Dimer composed of NRT and NNRT inhibitor

2.2.1.1

2.2.1.1

2.2.2


339

340

APPENDIX

TABLE 2. Multivalent Ligands Linked to gp120 Surface Receptors in Human Immunodeficiency Virus Multivalent Ligand Pentavalent CD4 peptide (aa 41–84) linked to ovalbumin Galactosyl ceramide (Gal cer) on surface Multivalent Gal cer in liposome Multivalent Gal cer in lipid bilayer Divalent anti-Gal cer antibody Divalent suramin (binding to V3 region) Polyvalent o-acrylate (telomerized) Polymeric undec-11enoate (micelle) Multivalent mercaptoalkanoate Multivalent thiomalate of cyclodextrin Bivalent cyclam (CXCR4 antagonist) Cyanovirin-N (CVN)

Notes

Section

Binding to gp120; not inhibited by antibodies in HIVþ sera

2.2.3.1

Binding to gp120 (Kd ¼ 11.6 nM )

2.2.3.2

Binding to gp120 (Ka ¼ 1.6 10 9 M 1 )

2.2.3.2

Binding to gp120 (Ka ¼ 2.5 10 6 M 1 ) (cooperativity constant ¼ @1.5) Inhibition of gp120 binding to CD4 cell

2.2.3.2 2.2.3.2

Inhibition of gp120–Gal cer interaction

2.2.3.3

Inhibition of HIV infection (IC50 ¼ 0.1 mM )

2.2.3.3

Inhibition of HIV infection (IC50 ¼ 0.1 mg/mL) Inhibition of HIV infection (IC50 ¼ 40 nM )

2.2.3.3

Inhibition of gp120–CD4 interaction

2.2.3.3

Inhibition of gp120 binding to CXCR4

2.2.3.4

Bivalent binding to Man9 GlcNAc2 on gp120 (K a ¼ 2.4 10 7 M 1 ; 2 CVN per gp120)

2.2.4

2.2.3.3

APPENDIX

341

TABLE 3. Multivalent Ligands Linked to Surface Receptors on Viruses Receptor Hemagglutinin (HA) (influenza virus A)

Multivalent Ligand Divalent sialic acid (SA) ˚ linked with 27–57 A spacer (PEG, Gly)

Divalent SA linked with oligo(glycine) spacer Divalent SA linked to galactoside HA (influenza)

HA (influenza A)

HA (influenza)

HA (Sendai virus) HA (influenza A)

Tetravalent SA tethered to glycine-terminated dendrimer Multivalent SA on dendrimer Multivalent SA on dendrimer Multivalent SA on liposome Multivalent SA (glycophorin) presented on liposome Polyvalent SA in polymerized liposome Polyvalent SA in polydiacetylene on LB monolayer Multivalent SA on liposome Multivalent SA (fetuin) immobilized on dextran surface Polyvalent SA presented on poly(acrylamide) Polyvalent SA (Cglycoside) presented on poly(acrylamide) Polyvalent SA presented on poly(acrylamide) Polyvalent SA presented on poly(acrylamide)

Notes

Section

Inhibition of hemagglutination (HAI) activity: 500-fold enhanced over monomer 100-Fold higher a‰nity than monovalent SA Inhibition of virus–cell adhesion (@10-fold enhanced) Inhibition of virus–cell adhesion (1000-fold enhanced) HAI activity

2.1.1.1

Synthetic inhibitor of influenza HAI activity (KiHAI ¼ 20 nM ) Multiple HA–SA interactions required at fusion site Optical sensor for viral recognition Optical censor for viral recognition Inhibition of virus–cell adhesion Binding kinetics on surface (HA trimer–SA) HAI activity: @10 3 to 10 6 -fold enhanced over SA monomer HAI activity: Ki ¼ @10 nM Adhesion dynamics between mesoscale particles (OPTCOL) Synergy (2- to 20-fold) by neuraminidase inhibitor

2.1.1.1

2.1.1.1

2.1.1.2

2.1.1.3 2.1.1.3 2.1.1.4 2.1.1.4

2.1.1.5 2.1.1.6

2.1.1.4 2.1.1.7

2.1.1.8

2.1.1.10

2.1.1.13

2.1.1.12

(Continued)

342

APPENDIX

TABLE 3. (Continued) Receptor

Multivalent Ligand Polyvalent SA (C4 tethered) presented on poly(acrylamide) Polyvalent SA presented on poly(acrylic acid) Polyvalent (lysoganglioside GM3 ) on poly(Glu)

HA (influenza A, B) HA (influenza C)

HA (influenza A)

HA (influenza)

Neuraminidase (NA) (influenza A) Dengue virus Picorna virus (capsid protein) Polyoma virus

Respiratory syncytial virus (RSV) Rota virus Hepatitis B virus

Polyvalent NeuAc-a2,6Gal-b1,4-GlcNAc on poly(Glu) Polyvalent 9-O-acetylsialic acid presented on poly(acrylamide) Multivalent SA on g-inhibitor (a2 -macroglobulin) Multivalent SA linked to bovine serum albumin (BSA) Multivalent SA linked to BSA or tetanus toxoid

Polyvalent 4-guanidinoNeuAc presented on poly(Glu) Heparin sulfate (polyanion) Divalent benzothiazole linked with 6- to 7-atom spacer Divalent SA linked to oligo(lysine) Multivalent RFI-641 tethered to biphenylbis(sulfonate) Multivalent SA linked to phospholipid Multivalent viral surface protein on polystyrene bead

Notes

Section

HAI activity at mM

2.1.1.17

Combinatorial library of polyvalent polymers Specific SA–HA and nonspecific hydrophobic interaction Inhibition of viral infection

2.1.1.14

2.1.1.15

2.1.1.16

HAI activity; resistance to 9-Oacetylesterase Natural defense molecule against virus

2.1.1.18

Inhibition of virus– human erythrocyte adhesion Synthesis of neoglycoprotein (Can. J. Chem. 1990, 68, 2045) Enhanced antiviral activity over monomer Binding to envelope protein Heterobivalent library

2.1.1.19

HAI activity: 500-fold enhanced over monomer Inhibiting viral adsorption and fusion (IC50 ¼ 50 nM ) 10 4 -Fold increase in inhibitory activity Adhesion of bead to hepatoblastoma (EMBO J. 2001, 20, 4443)

2.1.1.19

2.1.2.1

2.7 2.5

2.4

2.6

2.3

TABLE 4. Multivalent Receptors Targeting Multivalent Ligands in Bacterial Cell Walls Surface Ligand d-Ala-d-Lac

d-Ala-d-Ala

Multivalent Receptor Divalent vancomycin (head-to-head) linked with hexane spacer Divalent vancomycin (head-to-head) linked with p-xylylene Divalent vancomycin (head-to-tail) linked with 4- to 7-atom spacer Divalent vancomycin (back-to-back) linked with 5- to 8-atom spacer Divalent vancomycin (back-to-back): disulfide and olefin spacer Divalent vancomycin (multiple combination of orientation) Divalent eremomycin (cyclic dimer) Trivalent vancomycin (head-to-head) linked to benzene sca¤old Polyvalent vancomycin (back-to-back) linked with 14-atom spacer Vancomycin merged with moenomycin Trivalent peptide receptor mimic displayed on triazacyclophane Divalent vancomycin mimic linked with 3- to 7-atom spacer Synthetic receptor

Lipid A, gramnegative

Lipopolysaccharide Ergosterol Candida albicans

Polymyxin B mimic: trivalent amine linked to cholyl sca¤old Polymyxin B mimic: tetravalent amine tethered to ter-cyclopentane C1q (complement system) Divalent amphotericin B (AmB) linked with bisamide

Notes Increased activity against VRE (@20fold enhancement) Divalent receptor– ligand binding (Kd ¼ 1.1 nM; @600-fold increase) Synthesis and characterization Bivalent binding to ligand on surface (SPR study) Enhanced activity against VRE van A

Section 3.1.1.3

3.1.1.3

3.1.1.3

3.1.1.3

3.1.1.3

Enhanced activity against VRE van A

3.1.1.3

Lack of antibacterial activity Trivalent receptor– ligand association (Kd ¼ 4 1017 M ) Enhanced antibacterial activity against VRE

3.1.1.4

Designed to bind two distinct targets Ka (d-Ala-d-Ala) ¼ 26,600 M 1 ; Ka (d-Ala-d-Lac) ¼ 10,100 M 1 Inhibition of VRE: 60-fold more active than vancomycin Catalytic hydrolysis of d-Ala-d-Lac Inhibition of bacterial growth

3.1.1.6

Inhibition of bacterial growth

3.1.2.2

Adsorption of C1q to lipid A surface >10-Fold weaker antibiotic activity than AmB monomer

3.1.2.3

3.1.1.3

3.1.1.5

3.1.1.7

3.1.1.7

3.1.1.8 3.1.2.1

3.6.1

343

344

APPENDIX

TABLE 5. Multivalent Ligands Targeting Bacterial Receptors in Membrane Surfaces Receptor E. coli receptor

E. coli fimbriae

Chemoreceptor (E. coli)

Chemoreceptor (B. subtilis) BabA adhesin (H. pylori)

HpaA adhesin (H. pylori) H. pylori receptor Multidrug e¿ux pump (E. coli QacA) Repressor protein QacR (S. aureus) Strep. suis adhesin

Multivalent Ligand Polyvalent Man/Gal on poly(acrylamide) Man on self-assembled monolayer Man presented on gold nanoparticle Trivalent a-Manpresenting glycodendrimer Polyvalent Gal on neoglycopolymer Polyvalent Gal on neoglycopolymer Polyvalent Gal on neoglycopolymer Polyvalent Glu linked to neoglycopolymer Multivalent Le b glycoconjugate on serum albumin Multivalent fucosylated Le b neoglycoprotein Multivalent 3 0 sialylLac presented on neoglycoprotein Multivalent SA presented on neoglycoprotein Divalent cyanine

Notes Inhibition of yeast agglutination to E. coli Adhesion forces: monovalent (@1.7 pN ), polyvalent (b 18 pN ) Selective binding to type 1 pili

Section 3.1.5.2 3.1.5.1

3.1.5.3

Inhibition of fimbriae-mediated bacterium–cell adhesion

3.1.5

Eliciting bacterial chemotaxis; clustering chemoreceptors Increased chemotactic activity; valence and tether length e¤ect Clustering Gal-sensing receptor; higher sensitivity to serine R Increased chemotactic activity; valence dependence Inhibition of bacterial adhesion to human gastric epithelium

3.1.6.1

K a (to bacterium) ¼ 8 10 9 M 1 ; @500 Le b bound per bacterium 1000-Fold increase in inhibition of bacterial adhesion

3.1.6.1

3.1.6.1 3.1.6.2 3.1.3

3.1.3

3.1.3

Inhibition of H. pylori–induced hemagglutination

3.1.3

Km ¼ 13.50 mM

3.1.8

Divalent quinolinium and dequalinium

Presenting two proximal drugrecognition sites

3.1.8

Di- and tetravalent galabioside

1000-Fold increased inhibition of bacterial hemagglutination

3.1.4

APPENDIX

345

TABLE 6. Multivalent Ligands Linked to Bacterial Toxins Toxin

Multivalent Ligand

Anthrax toxin (PA63 )

Multivalent dodecapeptide in poly(acrylamide)

Cholera toxin (CT)

Multivalent ganglioside (GM) on polymerized liposome Fluorophore-labeled GM1 presented on lipid bilayer Multivalent gangliosidepresenting membrane

Cholera toxin (CT); E. coli enterotoxin Heat-labile enterotoxin; cholera toxin E. coli heatlabile enterotoxin (LT) B5 Shiga toxin

Shiga toxin isoforms (ShTX-1 and 2) Shigalike toxin (SLT)-1B

Shigalike toxin I (SLT-I) Tetanus toxin (TeNT) Verotoxin (VT)

Ganglioside tethered to PAMAM dendrimer Multivalent sialylLac on polymer support

Notes Inhibition of lethal factor binding to PA63 (10 5 -fold enhanced) Colorimetric sensor for toxin binding Fluorescent detection of CT (< 0:05 nM ) Binding a‰nity to CT by SPR: GM1 > GM2 > GM3 > asialoGM1 Inhibition of toxin adhesion to GM1-coated surface E¤ective neutralization of bacterial toxin

Section 3.2.5.1

3.2.2.4

3.2.2.5 3.2.2.4

3.2.2.3

3.2.1.2

Pentavalent Gal tethered to azacrown sca¤old

Inhibition of LT-B5 : 10 5 fold enhanced over Gal monomer

3.2.1.1

Multivalent P1 trisaccharide presented on agarose gel Polyvalent P k trisaccharide on poly(acrylamide) Multivalent glycolipid Gb3 Cer in lipid layer

A‰nity purification of B subunit >5000-Fold enhanced a‰nity over monomer Kinetics of multivalent binding of ShTX-2 and ShTX-1

3.2.3.4

Divalent P k trisaccharide tethered to pentapeptide

Inhibition of SLT-1B: Ka (M 1 ); 20-fold enhanced over monomer Bivalent occupation of sites 1 and 2 on B subunit 10 7 -Fold enhanced activity; pentavalent complex (x-ray) Presence of two distinct GM sites (J. Biol. Chem. 2001, 276, 32274) Positive cooperativity in binding

3.2.3

Divalent P k trisaccharide linked with bisamide ˚ long) (@11 A Decavalent P k trisaccharide on glucose-cored dendrimer Multivalent ganglioside (GT1b) Multivalent globotriaose

3.2.3.3 3.2.3

3.2.3

3.2.3.2

3.2.4

346

APPENDIX

TABLE 7. Multivalent Molecules Targeting Bacterial Enzymes Enzyme

Multivalent Molecule

Notes

Methicillinresistant Staph. aureus (MRSA) Mycobacterium tuberculosis

Bromotyrosine linked with disulfide spacer (psammaplin A) Divalent flavone linked with hexane spacer

Antibacterial activity against MRSA

Dihydrofolate reductase Nicotine adenine dinucleotide synthetase Protein tyrosine phosphatase (Yersinia) Pyrophosphokinase (E. coli)

Methotrexate dimer

Trivalent 4-methoxyphenylglyoxylic acid linked to benzene Dihydropterin linked to adenosine

Penicillin-binding protein (PBP)

Cephalosporin linked to Lys-d-Ala-d-Ala

Sialidase

Topoisomerase IV

Transglycosylase

Nicotine linked to indole

Bivalent b-lactam (penicillin V) linked with oligo(glycine) spacer Trivalent cephalosporin linked to trihydroxybenzene Polyvalent b-lactam linked to polymer SA-presenting poly(acrylamide) Divalent fluoroquinolone linked with rigid aryl spacer Moenmomycin linked to coumarin; presented on liposome 3,4-Dichlorobenzyloxybenzyl linked to vancomycin

Inhibiting bacterial growth (Bioorg. Med. Chem. Lett. 2001, 11, 2101) Enhanced divalent association Inhibition of NAD synthetase

Section 3.3.4

3.3.5 3.3.7

18- to 214-Fold increased inhibition

3.3.3

>100-Fold higher affinity than ATP or pterin Cross-link to active site groove of transpeptidase Higher a‰nity to PBP; cross-link two PBP molecules Similar or lower activity than lactam monomer Retained antibacterial activity Enhanced catalytic e‰ciency for polyvalent substrate Improved activity over monomer

3.3.2

3.3.1.1

3.3.1.2

3.3.1.2

3.3.1.2 3.3.6

3.3.4

Interaction with artificial membrane

3.1.1.6

Hydrophobic domain binding to PBP

3.1.1.6

347

APPENDIX

TABLE 8. Multivalent Ligands Targeting Carbohydrate-Recognition Receptors on Cellular Surfaces Receptor

Multivalent Ligand

B16 melanoma cell

Multivalent ganglioside (GM3 )

Cancer cell

Man-6-phosphonate displayed on liposome

Lectin on cancer cell

Fucosylated biantennary N-glycan linked to BSA Muramyl dipeptide linked to poly(acrylamide) Divalent NeuAc-a2,6Gal linked to asparagine Trivalent NeuAc-a2,6Lac linked to Trisbased sca¤old Trivalent NeuAc-a2,6Glc Polyvalent NeuAc-a2,6Gal linked to polymer (ROMP) Multivalent sulfated glycoside (heparin)

CD14

CD22 (B lymphocyte) CD22

ColQ (AChE)

C-type lectin (DC-SIGN) DC-SIGN

Dodecapeptide (bacteriophage display) Fibroblast growth factor (FGF) Fucose-binding receptor (colon cancer) Galectin-1

Nonavalent mannoside (Man9 GlcNAc2 ) Pentavalent Le a, b, x linked to lysine-based dendrimer Multivalent SA and KDO presented on solid support Divalent heparin decasaccharide Multivalent Fuc and batracylin linked to BSA Divalent Man-presenting glycoprotein (CD45CD3)

Notes Inhibition of lung metastasis (Chem. Biol. 1997, 4, 97) Adhesion of vesicle to cancer cell (Chem. Commun. 2001, 85) Specific a‰nity to cancer cell (J. Med. Chem. 2002, 45, 478) Induced production of proinflammatory mediator Chemoenzymatic synthesis

Section

4.1.6

4.1.6

300-Fold higher a‰nity over monomer

4.1.6

58-Fold enhanced a‰nity over monomer Binding to B cell surface

4.1.6

Presence of two nonidentical heparinrecognition sites 130-Fold tighter binding to DC-SIGN Higher a‰nity to DCSIGN than MBP

4.4.1.3

Nanomolar a‰nity (ChemBiochem 2001, 2, 741) Formation of active dimer between FGF and heparin Lectin-mediated drug delivery to cancer cell Clustering of T-cell surface receptor

4.1.6

4.1.7.1 4.1.7.2

4.1.11

4.1.10.1

4.1.5

(Continued)

348

APPENDIX

TABLE 8. (Continued) Receptor Galectin-1, 7, and 3 (mammalian) Asialoglycoprotein receptor (ASGP-R) Hepatic ASGP-R

Multivalent Ligand Di- to octavalent Lac linked to branched benzoic acid Trivalent Gal

Trivalent Gal linked to ˚ cholesterol (20 A spacer) Multivalent Lac linked to peptide nucleic acid (PNA) Trivalent Gal tethered with carbamate linkage Trivalent Gal linked to antisense RNA

Hepatic lectin

Trivalent GalNAc linked to lysine-based sca¤old Triantennary GalNAc Le x Trivalent peptide

Hepatocyte (rabbit)

Trivalent Gal linked with 3-atom spacer

Hepatocyte

Multivalent GlcNAc and Gal linked to poly(acrylamide) Multivalent carbohydrate

HepG2 (hepatocarcinoma) Lewis x glycosphingolipid

Polyvalent Gal attached to HPMA copolymer Multivalent Lewis x linked to liposome

Mannose-binding protein (MBP)

Trivalent Man linked to serine-based peptide Trivalent Man linked to Tyr-Asp-based peptide

Notes

Section

150-Fold higher a‰nity per lactoside basis

4.1.5

High a‰nity to hepatic ASGP-R (spacer– a‰nity correlation) High a‰nity to ASGP-R (Ki ¼ 200 nM )

4.1.2.1

4.1.2.3

Liver cell–specific uptake of PNA

4.1.2.4

Specific gene delivery to HepG2

4.1.2.4

Guided delivery of antisense RNA to ASGPR expressing cell High a‰nity to ASGP-R: Ki ¼ 4 nM Biodistribution and clearance by liver tissue Conformational analysis (Acc. Chem. Res. 1995, 28, 321) 10 3 -Fold increased a‰nity (J. Biol. Chem. 1982, 257, 939) Synthetic glycopolymer (Methods Enzymol. 1994, 242, 221) Adhesion of hepatocyte to immobilized sugar (J. Biol. Chem. 1979, 254, 10830) Internalization by human HepG2 Homophilic carbohydrate–carbohydrate interaction Binding to MBP

4.1.2.4

20 to 30-Fold enhanced a‰nity to liver MBP (not to serum MBP)

4.1.3

4.1.2.2 4.1.2.5

4.1.2.5 4.1.12

4.1.3

APPENDIX

349


Multivalent Ligand Hexavalent Man linked to lysine-based dendrimer Tetra- and octavalent Man linked to dendrimer Multivalent Man linked to BSA Polyvalent Man linked to poly(acrylamide)

Mannose receptor

Mannose/GalNAc-4-SO4 receptor (liver) Mannose/GalNAc-4-SO4 receptor (macrophage) Mannose/GalNAc-4-SO4 receptor (liver) Mannose-6phosphate receptor (MPR) Mannose-6phosphate/ IGF2R Natural killer cell receptor (sNKR-P1A) Sperm bindin receptor XL35 lectin

Sulfated GlcNAcb1,3-Fuc

Di- to octavalent Man bioisotere (quinic acid, shikimic acid) Tetravalent GalNAc-4SO4 -bearing lutropin (hormone) Multivalent Manpresenting glycoprotein

Notes

Section

MBP-guided endocytosis in human dendritic cell

4.1.3

Thiourea-bridged glycoside cluster

4.1.3

10 4 - to 10 6 -Fold enhanced a‰nity to MBP (serum, liver) @2500-Fold enhanced a‰nity to liver MBP over monomer Selective recognition by Man receptor and intracellular uptake Regulation of hormone clearance

4.1.3.2

4.1.3.3

4.1.3.4

4.1.1.1

Selective recognition of multivalent ligand by receptor

4.1.1.1

Heptavalent GalNAc-4SO4 linked to BSA

2000-Fold enhanced a‰nity to receptor

4.1.1.1

Tetravalent pentaMan phosphate tethered to biotin (to avidin) Multivalent Man-6phosphate linked to BSA Multivalent GlcNAc linked to dendrimer

Tighter binding to MPR than monomer

4.1.4.2

Receptor dimerization by multivalent Man-6-P

4.1.4.1

9-(OSO32 )Neu5Gcdisplaying neoglycoprotein Polyvalent melibiose linked to poly (acrylamide) Multiple sulfated GlcNAc-b1,3-Fuc linked to BSA

Formation of glycocluster-receptor lattice (FEBS Lett. 1999, 426, 243) Inhibition of sperm–egg fertilization

4.1.9

Inhibition of XL35 binding to jelly coat protein

4.1.9.4

Carbohydrate selfrecognition in cell–cell adhesion

4.1.12

(Continued)

350

APPENDIX

TABLE 8. (Continued) Receptor Tumor cell

Multivalent Ligand Divalent LacNAc

Polyvalent lactoside linked to poly(lysine)

Tumor cell (gangliosphingolipid)

Multivalent GM3 -Lac ceramide in liposome

Notes Inhibition of tumor cell adhesion to laminin (Carbohydr. Res. 1996, 296, 229) Inhibition of tumor cell colonization (Carbohydr. Res. 1993, 245, 175) Blocking melanoma metastasis

Section

4.1.10.2

TABLE 9. Multivalent Ligands Targeting Noncarbohydrate-Recognition Receptors on Cellular Surfaces Receptor Class II major histocompatibility (MHC)

Endothelial cell

Ephrin-A2

Erythropoietin (EPO) receptor Estrogen receptor

Folate receptor

Multivalent Ligand Multivalent epitope HA306–318 linked with 12–36-peptide spacer Di- to octavalent MHCII peptide ligand Multivalent cyclo(RGDfK) lipopeptide on liposome Multivalent EphA3-Fc cluster Divalent EPO linked with PEG (3400) Divalent 17b-estradiol linked with PEG spacer Multivalent folate tethered to polyetherbased dendrimer Multivalent folate tethered to PEG grafted liposome

Notes

Section

Trigger immune response; valence and spacer dependence

4.2.2

Enhanced avidity to Tcell receptor; receptor clustering Selective adhesion to endothelium (Chem. Eur. J. 2001, 7, 1095) Shedding of EphA2 from cell surface (Science 2000, 289, 1360) 10 3 -Fold increased activity Binding to estrogen receptor-expressing cell

4.2.2

12 Folates per dendrimer (Bioconj. Chem. 1999, 10, 1115) Binding to surface folate receptor (Bioconj. Chem. 1999, 10, 289)

4.3.2 4.2.5

APPENDIX

351


Multivalent Ligand

Notes

Hepatocyte

Divalent bile acid linked to oligonucleotide

Hydroxyapatite crystals in bone

Multivalent RGDpresenting lipid (selfassembled)

Integrin receptor

Multivalent GRGDS– peptide linked to poly(norbornene) Multivalent RGD– peptide presented on polystyrene bead av b3 -selective antagonist presented on nanoparticle Polyvalent cyclic (RGDfK) linked to poly(acrylamide) Sperm fertilin b-derived peptide displayed in liposome Divalent IL-5 antagonist with disulfide linkage

Increased biliary excretion (Bioorg. Med. Chem. Lett. 2001, 11, 945) pH-Induced self-assembly of peptide–amphiphile (Science 2001, 294, 1684) 10-Fold greater inhibition in cell–fibronectin adhesion Clustering integrin

av b3 Integrin receptor av b5 Integrin receptor a6 b1 Integrin receptor (egg plasma) Interleukin-5 receptor (human IL5Ra) Neurotrophin receptor (TrkC) NIH3H3I cell

Peroxisome proliferatoractivated receptor P-Glycoprotein

Phytoalexin

Divalent b-turn peptidomimetic linked to arylamide spacer Multivalent GRGDS linked to liposome Trimeric phenoxyisobutyric acid Divalent ()-stipiamide linked with poly(ethylene glycol) Trivalent proline-based glycopeptide

Section

4.2.3.1

4.2.3.2

Selective gene delivery to av b3 integrin-bearing cell Adhesion to osteoblast cell (Science 1995, 267, 883) Inhibition of sperm–egg binding (100-fold enhanced potency) Induction of receptor dimerization

4.2.3.4

Designed for receptor dimerization

4.3.2.4

Binding to cell membrane (Chem. Commun. 1999, 1687) Structure-biased library design (Chem. Biol. 1997, 4, 909) Binding to P-glycoprotein (Tetrahedron Lett. 2001, 42, 3819) Phytoalexin elicitor activity (Bioorg. Med. Chem. 1996, 4, 1873)

4.1.9.3

4.3.2.2

352

APPENDIX

TABLE 10. Multivalent Ligands Linked to Selectins on Cell Surfaces Selectin Selectin (E, L, P)

E-Selectin

Multivalent Ligand

Notes

Section

Di- and trivalent sialyl Lewis x (sLe x ) Polyvalent sLe x presented on polymerized liposome Multivalent sialyl Lewis a linked to poly(aspartic acid) Bivalent sLe x anchored to galactose

Enhanced a‰nity to selectins High-a‰nity binding to selectins Multivalent ligand (Bioorg. Med. Chem. Lett. 1997, 7, 1705) Higher activity of dimer than sLe x (J. Am. Chem. Soc. 1995, 117, 66) Inhibitor of E-selectinmediated cell adhesion Inhibition of E-selectinmediated cell adhesion (@750-fold increase) Use in cell-free E-selectin assay 50,000-Fold enhanced activity (J. Am. Chem. Soc. 2001, 123, 10113) Inhibition of E-selectinmediated neutrophil–cell adhesion Inhibition of E-selectin: @17-fold higher than sLe x 2- to 3-Fold more active than sLe x (Angew. Chem. Int. Ed. 1996, 35, 321) 4-Fold enhanced a‰nity (Angew. Chem. Int. Ed. 1996, 35, 1812) Binding to E-selectin on surface (10 5 -fold enhanced a‰nity) Inhibition of lymphocyte binding to endothelium Promotion of L-selectin shedding from cell surface Inhibition of lymphocyte adhesion to endothelium

4.1.13.1

Divalent sLe x Multivalent sLe x -PEGDSPE on liposome Multivalent sialyl Lewis (x, a) linked to poly(lysine) Polyvalent sLe x presented on poly(acrylamide) Multivalent sLe x tethered x to BSA (BSAsLe16 ) Multivalent fucodipeptide linked to polymerized liposome Trivalent sLe x linked to cyclic heptapeptide Trivalent sLe x linked to peptide sca¤old Multivalent sLe x displayed on liposome L-Selectin L-Selectin (neutrophil) L-Selectin

Tetravalent sLe x linked to branched oligosaccharide Multivalent 3 0 ,6-disulfated Le x linked to neoglycopolymer Divalent sLe x anchored to galactoside

4.1.13.4

4.1.13.1 4.1.13.4

4.1.13.5

4.1.13.3

4.1.13.4

4.1.13.4

4.1.14 4.1.14.2

4.1.14

APPENDIX

353

TABLE 10. (Continued) Selectin L-Selectin (B cell) L-Selectin

P-Selectin

P-Selectin (neutrophil) P-Selectin

P-Selectin; integrin

Multivalent Ligand

Notes

Section

Disulfo-a-d-Gal linked to fluorescent neoglycopolymer Divalent 3 0 -sulfolactoside linked to Asn-based glycopeptide Polyvalent 3 0 ,6-disulfo Le x (Glc) presented on neoglycopolymer Multivalent sLe x presented on poly(lactosamine)

Binding to cell surface (fluorescence microscopy)

4.1.14.2

Selectin-targeted glycopeptide

4.1.14.1

Inhibition of L-selectinmediated leukocyte rolling Nanomolar inhibitor of L-selectin-dependent lymphocyte adhesion Micromolar inhibition of L-selectin

4.1.14.2

Poly(acrylamide) displaying sulfoLe x mimic/sulfoGlcNAc Polyvalent sulfatide mimetic Multivalent sulfated galactoside displayed on polymer P-selectin glycoprotein ligand on lipid bilayer Polyvalent sLe x presented on polymerized acidic liposome Multivalent sLe x analog linked to polymerized liposome SLe x linked to RGDA SLe x linked to RGD peptide with lysine

Potent inhibitor of Pselectin Blocking of HL60 cell binding to immobilized selectin-Ig Ka ¼ 1.4 10 7 M 1 ; binding force: @110 pN Nanomolar inhibition of selectin adhesion Potent selectin inhibition: IC50 ¼ 2 nM vs. 5 mM (sLe x ) IC50 ¼ 26 mM Higher a‰nity to P-selectin than sLe x ; stable complex

4.1.14.4

4.1.14.3

4.1.15 4.1.15.3

4.1.15 4.1.15.3

4.1.15.3

4.1.15.1 4.1.15.1

354

APPENDIX

TABLE 11. Multivalent Ligands Linked to Lectins Lectin

Multivalent Ligand

Notes

Abrus lectin

Multivalent Gal expressed on erythrocyte

Bauhinia purpurea lectin

Multivalent Gal-b1,3GalNAc linked to tentagel bead Multivalent Gluderivatized polymer

K d ¼ 8 10 6 M 1 @ 4 10 8 M 1 (J. Biol. Chem. 1976, 251, 3977) Carbohydrate-derivatized bead as a model cell surface Potent inhibitor of Con A– induced agglutination (J. Am. Chem. Soc. 1994, 116, 12053) 300-Fold enhanced activity over monomer (Curr. Opinion Struc. Biol. 1996, 6, 692) A‰nity influenced by anomer, valence, and linkage type Enhanced activity by HAI and ITC (Tetrahedron Asymmetry 2000, 11, 95) Visualization of ligand– receptor complex using TEM Dynamic combinatorial library Transporter of eosin Y to saccharide-recognition lectin Correlation of agglutination inhibition activity and valence Selective inhibition of Con A

Con A

Hexadecavalent a-Man linked to PAMAM dendrimer Heptavalent Man linked to cyclodextrin Multivalent Man linked to dendrimer Multivalent Man linked to ROMP-derived polymer Divalent Man linked with disulfide spacer Multivalent Glu linked with cyclophane Multivalent Man linked to neoglycopolymer Divalent Man and Glu linked with bicyclic diester Trivalent Man linked to benzene-triacid 3- to 36-Valent a-Man linked to benzenecored dendrimer Multivalent Glu linked to polymerized liposome

30-Fold increased a‰nity in HAI; no enhancement in DG 4-Fold enhanced a‰nity

Reversible interaction with Con A (Angew. Chem. Int. Ed. 1981, 20, 91)

Section

4.1.16.10

4.1.16.4

4.1.16.6

5.2.3.1 4.1.16.3

4.1.16.7

4.1.16

4.1.16.1

4.1.16.5

355

APPENDIX

TABLE 11. (Continued) Lectin

Multivalent Ligand

Notes

Multivalent Man linked to lipid bilayer

Binding to Con A (SPR) (J. Am. Chem. Soc. 1998, 120, 10575) Synergistic formation of soluble Con A clusters (J. Am. Chem. Soc. 2000, 122, 4518) Oligonucleotide in blocktype to graft-type (Bioconj. Chem. 2001, 12, 776)

Trivalent Man linked to rigid macrocycle

Dioclea grandiflora lectin Entamoeba histolytica lectin Galanthus nivalis agglutinin (GNA) Glycine max lectin (tetrameric) Lectin

Limax flavus lectin (LFA) Peanut lectin (PNA)

Man linked to poly(methacrylamide)oligonucleotide Divalent Man-cluster

Multivalent GalNAc linked to bovine serum albumin Multivalent Man linked to branched mannopentaose Di- to multivalent Gal

Tetra- to octavalent Man linked through thiourea Multivalent Glu tethered to Tris-based dendrimer Trivalent Gal tethered to benzenetrimethanol Multivalent Man linked to b-cyclodextrinbased dendrimer Multivalent Man linked to sixth-generation PAMAM dendrimer Multivalent sialic acid presented on dendrimer Multivalent Gal linked to b-cyclodextrin

Enhanced a‰nity; thermodynamics (J. Biol. Chem. 2000, 275, 14223) 140,000-Fold enhanced activity over monomer

Section

3.1.7

Formation of cross-linked complex (Structure 1996, 4, 1339) Formation of cross-linked lectin lattices (Biochem. 1988, 27, 1034) Synthesis of cluster glycosides (Angew. Chem. Int. Ed. 1996, 35, 1953) Synthesis of glycodendrimer (Chem. Eur. J. 1996, 2, 1115) Synthesis by glycosidation of glycal epoxide (J. Org. Chem. 1997, 62, 1543) Synthesis and binding a‰nity (Chem. Commun. 2000, 1489) Synthesis and lectin binding activity Up to 182-fold enhanced activity over monomer

4.1.16.11

4.1.16.11

Soluble cross-linked clusters with PNA (J. Am. Chem. Soc. 1994, 116, 12053) (Continued)

356

APPENDIX

TABLE 11. (Continued) Lectin Ricinus communis agglutinin (RCA120)

Multivalent Ligand

Notes

Section

Multivalent Gal linked to liposome and polymer

Specific recognition of RCA; size and density dependence

4.1.16.17

Multivalent Lac displayed on tentagel solid support

Lectin–sugar interaction (flow cytometry) (Bioconjugate Chem. 1995, 6, 319) Association of RCA–gold nanoparticle (J. Am. Chem. Soc. 2001, 123, 8226) Formation of cross-linked lattice

RCA

Multivalent Lac linked to gold nanoparticle

Soybean agglutinin (SBA) Sponge geodia cydonium lectin

Divalent Gal

Vicia villosa agglutinin

Wheat germ agglutinin (WGA)

Multivalent Galb1,4GlcNAc presented on asialofetuin 8- to 16-Valent aGalNAc linked to calixarene dendrimer Trivalent GalNAcbipyridyl linked to iron (II) metal complex Multivalent GlcNAc linked to peptidebased dendrimer Multivalent SA linked to peptide-based dendrimer

Inhibition of lectin– glycoprotein binding (FEBS Lett. 1998, 427, 134) HAI activity: 250-fold increase (Angew. Chem. Int. Ed. 1999, 38, 369) Lectin-mediated dynamic combinatorial synthesis

20-Fold increased inhibition activity over monomer Inhibition of lectin (Synthetic Oligosaccharides, ACS Symp. Ser., 1994, 560, 104)

4.1.16

5.2.3.3

4.1.16.12

TABLE 12. Multivalent Ligands Linked to Cellular Enzymes Enzyme

Multivalent Ligand

Acetylcholine esterase (AChE)

Divalent 9-amino-THA linked with 7- to 10atom spacer

AChE; BuChE

Divalent 9-amino-THA linked with 5- to 8-atom spacer Divalent huperzine A linked with 10- to 13atom alkane spacer Divalent pyridinium linked with acylhydrazone spacer Divalent heparin pentasaccharide linked with 50-atom spacer Sulfated pentasaccharide linked to trisaccharid through PEG spacer Divalent benzenesulfonamide linked to thioether Divalent benzenesulfonamide linked to hydrophobics Multivalent phenoxyacetic acid linked to diphenylmethane Multivalent Gal- and GlcNAc-ZP3 glycoprotein (egg coat)

AChE

Antithrombin III (AT-III) AT-III; thrombin

Carbonic anhydrase II (CAII)

Elastase (human leukocyte; HLE) Galactosyltransferase (sperm b-1,4GalTase) b-1,3-GalTase Glycosyltransferase Glucoamylase

Divalent inhibitor of Gal mimic linked to uridine Substrate-based divalent inhibitor b-Cyclodextrin linked to acarbose through PEG spacer

Lymphocytic leukemia

Divalent steroid fused to pyrazine

O-MethylguanineDNA methyltransferase

(4-Bromothenyl)guanine linked to Glu

Notes

Section

Divalent binding to catalytic and peripheral site; high a‰nity >600-Fold enhanced a‰nity; 100-fold higher selectivity Binding to two domains in catalytic pocket of enzyme Deconvolution of dynamic combinatorial library AT-III-mediated inhibition of thrombin

4.4.1

Twofold higher antithrombin activity than heparin Combinatorial synthesis Dynamic combinatorial synthesis

4.4.9

Competitive and specific enzyme inhibition (Ki ¼ 20 nM ) GalTase-ZP3 mediated sperm–egg adhesion leading to acrosomal reaction Potent, selective inhibition Enhanced inhibitory activity Change of enzyme conformation upon bivalent ligand occupation Potent antitumor activity (Angew. Chem. Int. Ed. 1996, 35, 611) Selective uptake and activity in tumor cell (J. Med. Chem. 2001, 44, 4050)

4.4.8

4.4.1.1

4.4.1.2

5.2.2.4

4.4.9

4.4.14.2 5.2.2.5

4.4.11.2

4.4.11.1 4.4.11 4.4.10

(Continued)

357

358

APPENDIX

TABLE 12. (Continued) Enzyme

Multivalent Ligand

Phospholipase A2 (PLA2 )

Multivalent phosphatidylcholine linked to polymerized liposome

Proteasome (Yeast 20S)

Divalent Arg-Val-Arg linked with PEG spacer

Protein kinase C (PKC)

Bivalent alkane-linked phorbol Bivalent bisindolylmaleimide

Divalent quinone (torreyanic acid)

PKC, PKA

Divalent staurosporine

Protein tyrosine phosphatase 1B

Bis-(4-phosphophenyl)methane

Protein-tyrosine phosphatase (VHR) Riboflavin synthase

Divalent tetronic acid (RK-682) linked through ester Bivalent lumazine

c-Src protein kinase

Divalent aromatics linked with carbazole oxime

Stromelysin (matrix metalloproteinase)

Acetohydroxamic acid linked to 3-cyanomethylbiphenyl Divalent phosphotyrosinecontaining tetrapeptide

Syk protein tyrosine kinase b-Tryptase (human)

Tryptase (human lung)

Divalent 3aminomethylbenzenesulfonamide at cyclodextrin Divalent benzamidine linked with bisarylether spacer

Notes PLA2 catalysis at interface; color sensor (Angew. Chem. Int. Ed. 1999, 38, 655) >100-Fold enhanced activity; bivalent binding at catalytic site 13-Fold enhanced activity Competitive inhibition: Ki ¼ 14 nM (J. Med. Chem. 1992, 35, 994) 5- to 10-Fold enhanced activity (J. Org. Chem. 1996, 61, 3232) Potent inhibitor of protein kinase (J. Biol. Chem. 1994, 269, 9568) Divalent binding to catalytic and peripheral site 10- to 66-Fold higher inhibition activity than monomer Increase in inhibition activity Combinatorial, target-oriented ligand assembly Potent nonpeptide inhibitor (IC50 < 25 nM ) 5000-Fold higher a‰nity than monomer 10 4 -Fold higher activity; two dimers bound per enzyme tetramer Subnanomolar inhibitory activity; bridging two active sites

Section

4.4.4

4.4.5

4.4.7.1

4.4.7.2

4.4.12 5.2.2.3

4.4.2

4.4.6

4.4.3.1

4.4.3.3

359

APPENDIX

TABLE 13. Multivalent Ligands Linked to G-Protein-Coupled Receptors (GPCRs) GPCR

Multivalent Ligand

Notes

Adenosine receptor

Divalent adenosine analog linked at adenine

a1 -Adrenergic receptor (a1 -AR) a2 -AR

Divalent prazosin linked to benextramine with disulfide Divalent yohimbine linked with 2- to 12atom alkane spacer Multivalent isoproterenol linked through azobenzene to peptide Divalent albuterol linked with hexamethylene (hexaprenalin) Divalent aryloxy-propanolamine linked with 3- to 20-atom spacer Divalent practolol linked with 8- to 14-atom spacer Multivalent alprenolol tethered to dextran

Selective, coactivation of A1 and A3 -R (J. Biol. Chem. 2000, 275, 30272) Subtype and functional selectivity to a1D

b-Adrenergic receptor (b-AR)

Melanocortin-1

Monoamine transporter Muscarinic acetylcholine receptor (mAChR) Opioid receptor

Divalent a-MSH peptide linked to receptor disulfide spacer Divalent ligand linked with alkane spacer through ester Divalent NNC 11–1314 linked with phenyl group Divalent naltrexamine tethered with 8-atom spacer Divalent naltrexamine tethered with oligo(glycine) Divalent naltrexone linked with pyrrole spacer

Section

4.5.2.1

123-Fold enhanced a‰nity to a2a -AR over a2b -AR

4.5.2.2

Positive chronotropic activity

4.5.1.2

Enhanced agonistic activity

4.5.1.1

Selectivity to b1 and b2 -AR influenced by linkage sites

4.5.1.4

160-Fold higher a‰nity, depending on spacer length

4.5.1.3

10- to 8000-Fold lower potency, depending on spacer length 7-Fold increase in activity (EC50 ¼ 52 nM )

4.5.1.5

8000-Fold enhanced potency for serotonin reuptake transporter Increased activity and selectivity

4.5.6

4.5.4.3

4.5.5

Potency and selectivity influenced by spacer length

4.5.3

Correlation of a‰nity and spacer length at m-receptor

4.5.3

Kappa-selective antagonist (J. Med. Chem. 2000, 43, 1573) (Continued)

360

APPENDIX

TABLE 13. (Continued) GPCR

Serotonin receptor (5-HT receptor)

Multivalent Ligand

Notes

Divalent b-naltrexamine linked with PEG spacer Divalent enkephalin

Bridging proximal recognition sites (J. Med. Chem. 1982, 25, 847) Enhanced a‰nity and selectivity to d-receptor (Nature 1982, 297, 333) Increased potency and selectivity to 5-HT receptor

Divalent sumatriptan linked with xylyl group Divalent carboxamidoindole linked with 2- to 12-atom spacer Divalent serotonin linked with piperazine spacer Serotonin presented on Cd/Se/Zn nanocrystal

Increased potency and selectivity to 5-HT receptor

Potent full agonist of 5HT1B=1D (J. Med. Chem. 1996, 39, 4920) Fluorescent probe targeting 5HT receptor-expressing cell

Section

4.5.4.2

4.5.4.1

4.5.4.4

TABLE 14. Multivalent Ion Channel–Binding Molecules Ion channel Ca 2þ -activated Kþ channel (SK ca )

Cyclic nucleotide– gated (CNG) channel IP3 –gated calcium channel Nicotinic acetylcholine receptor (nAch-R)

Multivalent Molecule

Notes

Section

Divalent quinolinium linked with decamethylene spacer Divalent quinolinium linked with aromaticcyclophane spacer Divalent quinolinium linked with aliphaticcyclophane spacer Divalent cGMP linked with PEG spacer

Potent, selective blocker of SK ca

4.6.3

Potent channel blocker (IC50 ¼ 3 nM )

4.6.3

Potent channel blocker (IC50 ¼ 2 nM )

4.6.3

Activation of CNG channel: 1000-fold increase in potency 7- to 13-Fold enhanced activity Presence of two agonist sites per receptor complex Binding kinetics (Anal. Chem. 1999, 71, 3157)

4.6.1

Divalent inositol triphosphate (IP3 ) Divalent choline (succinylcholine) a-BgTx linked to biotin (then to streptavidingold)

4.6.2 4.6.4

361

APPENDIX

TABLE 15. Homo- and Heterodivalent Chemical Inducers of Dimerization (CIDs) Receptor Bacterial DNA gyrase (GyrB) Cyclophilin-Fas fusion FK-506 binding protein (FKBP)

FKBP-12

FKBP-12; Fyn SH2 domain FKBP-12; dihydrofolate reductase (DHFR) FKBP-caspase fusion protein FKBP-CD3 z fusion FKBP-Fas fusion protein

Divalent Ligand Coumermycin dimer Cyclosporin A dimer: (CsA)2 Divalent FK-506: FK-1012 FK506 analog: heterodimer of lowa‰nity monomers FK-1012 congeners

AP-1867 linked to tetrahydrooxazepine FK-1012 analog

FK-506 tethered to pTYEEI (Fyn SH2 ligand) FK-506 analog linked to methotrexate (DHFR ligand) FK-1012 Divalent FK-506 (linker variation) FK-1012

AP-1510 (dimer)

FKBP-Gal4; cyclophilinGFP FKBP-Gal4; cyclophilinVP16 FKBP12/c-Kit fusion protein; FRAP/Fit-3

FK-506 linked to cyclosporin A (FK-CsA) FK-CsA

Notes

Section

Raf-1 activation by dimerization (Nature 1996, 383, 178) Control of programmed cell death Transcriptional on–o¤ switch

4.3.1.4 4.3.1.2

Tight binding: Kd ¼ 19 nM (Science 1996, 274, 1531) Structure–activity relationship (Bioorg. Med. Chem. 1998, 6, 1309) Library of AP-1867 linked to THOX (heterodimerization) Structure of ternary complex (x-ray) (Bioorg. Med. Chem. Lett. 1998, 8, 1) Control of receptor–ligand association by protein interaction Dimerization of two hybrid proteins Intracellular cross-linking of caspase causing apoptosis Oligomerization of surface receptor (Science 1993, 262, 1019) Control of cellular apoptosis by dimerization of Fas protein Induction of apoptosis by dimerization of Fas protein Recruitment of Fas fusion to plasma membrane to apoptosis Inducible nuclear localization of cytosolic protein

4.3.1.6

4.3.1.6

4.1.6

4.3.1.2

4.3.1.2

4.3.1.3 4.3.1.5

4.3.1.5

FK-CsA

Inducible transcription activation

4.3.1.5

FK-1012, AP-1510, rapamycin

Homo- and heterodimerization of c-Kit and Fit-3

4.3.1.3

(Continued)

362

APPENDIX


Divalent Ligand

FKBP-Sos fusion protein

FK-1012

FKBPthrombopoietin Granulocytecolony stimulating factor (G-CSF) Src-like tyrosine kinase Vitamin D receptor

AP-1903 (dimer) SB247464 (merged dimer)

FK-1012 Divalent vitamin D3 and calcitriol

Notes Membrane localization of Sos (Proc. Natl. Acad. Sci. USA 1995, 92, 9810) Dimerization of growth receptor (Blood 2000, 95, 430) Receptor dimerization at extracellular domain

Dimerization and recruitment of Src kinase at membrane Synthesis of dimeric ligand for receptor dimerization

Section

4.3.2.3

4.3.1.6 4.3.3

TABLE 16. Multivalent Ligands Linked to Nucleic Acids Nucleic Acid DNA (HIV)

DNA

Multivalent Ligand Trivalent teretifolion B (naphtoquinone) Divalent juglone (5hydroxynaphtoquinone) Ancistrobrevine B; michellamine B Divalent mitomycin C linked at C-1 with aryl linker Divalent bZIP linked to biphenyl Noncovalent bZIP dimer formed by cyclodextrin dimer Divalent bZIP linked to 9,10-dihydrophenathrene

Multivalent ellipticine linked to HPMA copolymer Divalent daunorubicin linked with p-xylyl spacer

Notes

Section

Selective anti-HIV activity: inhibition of viral replication Anti-HIV activity

2.8.3.3

Anticytopathic activity

2.8.3.3

Interstrand cross-linking of DNA; enhanced cytotoxicity Sequence-specific DNA recognition Sequence-specific DNA recognition (Biochem. 1999, 38, 1626) Sequence-specific recognition of palindromic DNA (J. Am. Chem. Soc. 1997, 119, 3649) Enhanced cytotoxic activity (Bioconj. Chem. 2001, 12, 711) DNA binding by bisintercalation (K a ¼ 2:7 10 11 M 1 )

4.2.2

2.8.3.3

4.7.5.2

4.7.1.2

APPENDIX

363

TABLE 16. (Continued) Nucleic Acid

Multivalent Ligand Daunorubicin linked to paclitaxel

3-Aminoglycoside linked to 2-quinoline (intercalator) 3-Aminoglycoside linked to 9,10-anthraquinone Divalent diazobenzofluorene glycoside (lomaiviticin A, B) Naphthyridine dimer

DNA major groove

Adenine linked to acridine linked with basic spacer Divalent basic region (BR) of bZIP linked to azobenzene Acridine dimer linked with spermine Divalent naphthalimide linked with positively charged spacer Divalent 9-methylphenazine linked with cationic spacer

DNA minor groove (HIV-1)

Netropsin tethered to distamycin with 5- to 7atom spacer

DNA minor groove

Divalent polyamide py/im (hairpin shape) Divalent calicheamicin oligosaccharide Divalent py/im polyamide (hairpin shape) Divalent polyamide (H shape) Divalent polyamide (ring shape)

Notes Multiple arrays of DNA– microtubule (Bioorg. Med. Chem. Lett. 2000, 10, 261) Photocatalyzed cleavage of supercoiled DNA Sequence-selective binding via intercalation Cleavage of dsDNA; anticancer activity Selective recognition of guanine–guanine mismatch Selective binding to abasic site–containing DNA Light-controlled divalent binding of BR peptides (cis conformation) 10 4 -Fold higher a‰nity Bis-intercalate to DNA with linker bound in major groove High GC-selectivity; 20fold enhanced a‰nity (J. Med. Chem. 2001, 44, 1407) 10-Fold enhanced inhibition of DNA gyrase (J. Org. Chem. 2000, 65, 1102) Inhibition of DNA binding to transcription factors 10 3 -Fold higher DNA affinity; enhanced sequence selectivity Recognition of 6 to 9 base pairs with high a‰nity Enhanced a‰nity Enhanced a‰nity

Section

4.7.1.1 4.7.1.1 4.7.1.2

4.7.4

4.7.3 4.7.5.3

4.7.1.3 4.7.1.4

2.8.3.1 4.7.2

4.7.2.8 4.7.2.7 4.7.2.10 (Continued)

364

APPENDIX

TABLE 16. (Continued) Nucleic Acid

Multivalent Ligand Divalent polyamide linked with 3-aminopropionic acid Duocarmycin linked to CC-1065 with aliphatic/ aromatic spacer Bivalent bizelesin linked with indolylurea CC-1065 linked to poly(pyrrole) Polyamide linked to camptothecin Polyamide linked to VP2 peptide CC-1065 linked to DC-81 linked with 4-atom spacer DC-81 linked to pyrrolobenzodiazepine DC-81 linked to polypyrrole

RNA (RRE; HIV-1)

Chartreusin: intercalator linked to minor groove binder Neomycin B linked to acridine through short spacer Neamine linked to pyrene through arginine spacer Rev peptide linked to S peptide

rRNA (E. coli 16S) rRNA A-site (E. coli 16S) rRNA (E. coli S30)

Divalent neamine linked with butane spacer Divalent aminoglycoside Polyvalent amikacin tethered to poly(vinylbenzene)

Notes Recognition of 16-base pair with subnanomolar a‰nity DNA cross-linking molecule (J. Am. Chem. Soc. 1989, 111, 6428) DNA cross-linking molecule Sequence-specific alkylation (J. Med. Chem. 2001, 44, 2536) Induction of DNA cleavage at 3 or 4 base pairs from polyamide site Enhanced activation of gene transcription DNA interstrand A-G cross-linking (J. Am. Chem. Soc. 2001, 123, 4865) DNA interstrand G-crosslinking; 440-fold more potent than melphalan Increased antiproliferative activity over DC-81 or polypyrrole Enthalpy-driven divalent binding of antitumor drug to DNA Nanomolar inhibition of Rev response element of RNA 10 3 -Fold increased inhibition of TAR and Rev response element Sequence-selective cleavage of RNA strand by artificial RNase S Kd ¼ 40 nM (250-fold enhancement) 19-Fold increased a‰nity to dimeric A-site Decreased antibacterial activity (IC50 > 0:1 mM )

Section 4.7.2.9

4.7.2.3

4.7.2.13

4.7.2.14

4.7.2.4

4.7.2.12

4.7.1

2.8.1.1

2.8.1.2

2.8.2

3.4.1.1 3.4.1.2 3.4.1.3

APPENDIX

365

TABLE 17. Synthetic Multivalent Antigens Antibody Raised

Multivalent Antigen

Notes

Section

Antibody against breast cancer antigen

Globo H linked to keyhole limpet hemocyanin (KLH) Multivalent Globo H linked to KLH protein

Induced strong humoral immune response in patient Strong induction of IgM and IgG; reactive with breast cancer cell Synthetic vaccine for prostate cancer (Proc. Natl. Acad. Sci. USA 1999, 96, 5710) Multiantigenic synthetic vaccine

4.8.1.2

Antibody (IgM) against Globo H

Multivalent Globo H linked to KLH

Antibody against multiple antigens

Multivalent Globo H, aTN , and Le y linked to KLH Multivalent TF, Le y , and TN tumor antigens

Anti-Gal antibody (pig) Anti-c-myc mAb

Anti-NeuAc antibody Anti-SiaLev antibody Catalytic antibody (cocaine hydrolysis) Catalytic antibody (glycoside hydrolysis) Catalytic antibody (phosphoester hydrolysis) Catalytic antibody (aldolase activity) Gp-120 (HIV-1)

Mucin-fused immunoglobulin (140 copies of Gal per Ig) Multivalent c-myc epitope peptide on polymerized liposome Multivalent NeuAc linked to BSA Multivalent NlevulinoylNeuAc linked to KLH Multivalent TS analog of benzoyl ester hydrolysis on carrier protein Multivalent TS analog of ketal hydrolysis on carrier protein Multivalent TS analog of phosphate linked to protein Multivalent TS analog of aldol reaction on carrier protein Multivalent nonamannan present on glycoprotein

Synthesis of molecular construct containing multiple sugar antigens E¤ective binding to antigal antibody Specific antigen–antibody interaction (J. Am. Chem. Soc. 2001, 123, 417) Synthesis of neoglycoprotein Complement-mediated cell lysis (Chem. Biol. 2001, 8, 265) Specific hydrolysis of cocaine: Mab: k cat =Km ¼ 2:3 10 4 Mab: k cat =k uncat ¼ 870; glycosidase activity

4.8.1.2

4.8.1.7

4.8.1.5

4.8.2

4.8.2

4.8.3

4.8.3

Ab: k cat =k uncat A 364; Km ¼ 3:57 mM

4.8.3

Ab: catalytic activity in selective retroaldol and b-elimination Synthetic vaccine for highmannose sugar (Angew. Chem. Int. Ed. 1996, 35, 197)

4.8.3

(Continued)

366

APPENDIX

TABLE 17. (Continued) Antibody Raised E. coli

Foot-and-mouth virus HIV-1

Influenza virus

Leishamania parasite (leishmaniasis) Monoclonal Ab JAA-F11 (IgG3) Mycobacterium avium intracellulare (MAI) S. aureus

Shigella dysenteriae T-cell

Tumor cell

Multivalent Antigen

Notes

Multivalent K1-capsular polysialic acid in liposome VP1 peptide presented on single-walled carbon nanotube (SWNT) Tetravalent gp41 epitope (T20) linked to maleimide cluster Octavalent ectodomain of M2 peptide on oligo(Gly-Lysine) Multiple tetrasaccharide presented on KLH

Adsorption of anti-K1 Ab to liposome

3.5.4

Immunogenic

2.9.3

Immunogenic peptide vaccine

2.9.2

Peptide vaccine presenting three di¤erent antigens

2.9.1

Preparation of synthetic vaccine

3.5.3

Hexavalent breast cancer T-antigen linked to dendrimer Multivalent MAI-specific trisaccharide presented on BSA Multivalent cell wall polysaccharide linked to protein Multivalent O-specific polysaccharide on albumin Divalent GalNAc linked to T-cell epitope Tumor associated MUC1 antigen linked to tetanus toxin epitope Multivalent Gal-b1,3GalNAc linked to dendrimer Multivalent GM3 -lactone presented on KLH Multivalent GM3 -lactam Trivalent quinic acid (Gal-a1,3-Gal mimic) Multivalent GM2 ceramide linked to KLH

128-Fold increased inhibition activity (J. Am. Chem. Soc. 2001, 123, 1809) Preparation of synthetic vaccine: @12 saccharide antigens per albumin Preparation of multivalent antigen

Section

3.5.2

3.5.5

Preparation of synthetic vaccine: @11 copies of saccharide per albumin Binding to MHC; stimulating T-cells (Bioconj. Chem. 2001, 12, 325) Induction of cytotoxic T-cells (Angew. Chem. Int. Ed. 2001, 40, 366) Synthesis of T-antigenbearing glycodendrimer

3.5.1

Synthetic vaccine to GM3 expressing tumor Synthetic vaccine to GM3 lactam-expressing tumor Mimic of tumor-associated antigen (Tetrahedron 2001, 57, 3281) Synthetic vaccine for GM2 -expressing cancer cell

4.8.1.3 4.8.1.3

4.8.1.4

APPENDIX

367

TABLE 18. Multivalent Ligands Displayed on Self-Assembled Monolayer (SAM) Made of Alkanethiolate on Gold Receptor

Multivalent Ligand

Anti-fluorescein antibody

Fluorescein-presenting SAM

Bauhinia purpurea lectin BS-II lectin Bandeiraea simplicifolia Cell (fibroblast)

Gal-b1,3-GalNActerminated SAM GlcNAc-terminated SAM

Cell surface presenting Lewis x or lactose Chloroeremomycin DNA

Fibronectin (3T3 fibroblast cell) GalTase

Benzophenonecontaining RGD peptide photocrosslinked to SAM Lewis x or lactosepresenting SAM

d-Ala-d-Ala(Lac)presenting SAM Complementary oligonucleotide presented on SAM RGD-cyclopentadiene attached to quinone SAM via Diels–Alder GlcNAc-terminated SAM

Mammalian cell

Various functional group-presenting SAM

Nicotinic acetylcholine receptor Strepavidin

Bungarotoxin (BgTx)– biotin to streptavidincoated gold Biotin-terminated SAM

Uropathogenic E. coli

Mannoside-presenting SAM

Vancomycin (dimer)

d-Ala-d-Alapresenting SAM

Notes Binding force of single antigen–antibody pair (Proc. Natl. Acad. Sci. USA 1998, 95, 7402) Correlation of sugar selectivity and density K d @ 80 nM (multivalent binding) (Angew. Chem. Int. Ed. 1999, 38, 782) Photocontrolled micropatterning of peptide (Chem. Biol. 1997, 4, 731)

Section

4.1.16.10

Multivalent model for homophilic cell–cell adhesion

4.1.12.3

E¤ect of surface ligand density to antibiotic binding Measuring force to stretch and unbind dsDNA (Chem. Biol. 1997, 4, 519) Control of cell adhesion via redox switch

3.1.1.1

GalTase-catalyzed glycosylation on SAM (Angew. Chem. Int. Ed. 1999, 38, 782) Micropatterning of cells on SAM surface (J. Am. Chem. Soc. 1993, 115, 5877) Binding kinetics and parameters (Anal. Chem. 1999, 71, 3157) A‰nity-ligand density correlation (Langmuir 1991, 7, 1837) Measuring force involved in bacterium-surface adhesion 10 3 -Fold enhanced a‰nity to vancomycin dimer over monomer

4.2.3.3

3.1.5.1

3.1.1.3

368

APPENDIX

TABLE 19. (Bio)chemical Modification of Cell Surface Antigens Cell Surface Antigen Poly-a2,8-sialic acid (Npropanoyl) on neuron and HeLa cell Sialic acid(N-propanoyl) Sialyl-azidoAc on Jurkat cell SiaLev (ketone) on Jurkat cell; biotin– hydrazide SiaLev (ketone) on HUVEC cell; biotin– hydrazide de SiaLev (ketone) on Jurkat cell SiaLev (ketone) on cell

Fucose-containing synthetic glycoconjugate on cell Fluorescein displayed on surface of cowpea mosaic virus (CPMV) Ketone-presenting lipid I precursor incorporated on bacterium surface

Notes Expression of unnatural poly-a2,8-sialic acid (SA) on neural cell adhesion molecule (Science 2001, 294, 380) Expression of nonphysiological SA Derivatization with biotin-linked phosphine through Staudinger reaction Selective avidin binding to cell surface; selective delivery of cytotoxic avidin–ricin A conjugate Bind avidin–antiadenovirus conjugate; adhesion of adenoviral vector and gene transfer Derivatization with aminooxy-terminated DTPA for Eu (II) chelation; diagnosis of tumor by MRI Derivatization with aminooxy-terminated Gal and hydrazide-terminated Lac Recognition by anti-SiaLev specific antibody; altered immunoreactivity (Chem. Biol. 2001, 8, 265) Fucosyltransferase-mediated expression of nonnatural carbohydrates

Section

4.9.1.1 4.9.1.3 4.9.1.2

4.9.1.2

4.9.1.2

4.9.1.2

4.9.2

[3 þ 2]-Cycloaddition-mediated modification of protein surface on virus

4.9.2

Use of ketone group for derivatization with hydrazides

4.1.1.9

APPENDIX

369

TABLE 20. Multivalent Targets Target HIV-1 reverse transcriptase E. coli pyrophosphokinase Protein tyrosine phosphatase Bacterial sialidase NAD synthetase Protein kinase C Syk protein tyrosine kinase Protein tyrosine phosphatase Glucoamylase Glycosyltransferase Riboflavin synthase Carbonic anhydrase Histone deacetylase NAD synthetase Caspase cSrc kinase Cyanovirin-N (CVN) Galectin FK-506 binding protein Acetylcholinesterase Transpeptidase S. aureus e¿ux pump Nicotinic ACh-R Man/GalNAc-4-SO4 receptor Mannose 6-P receptor P-Selectin Mannose-binding protein Immunoglobulin Ricinus communis agglutinin Matrix metalloproteinase Human leukocyte elastase Thrombin Nucleic acid (DNA) Nucleic acid (RNA) FKBP Cyclophilin Dihydrofolate reductase Fibroblast growth factor Cytokine receptor EPO receptor IL-5 receptor G-CSF recptor Neurotrophin receptor E. coli chemoreceptor Vitamin D receptor

Type

Valency of System

Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Membrane, soluble Membrane Membrane Membrane Membrane Membrane Membrane Membrane Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Soluble Membrane Membrane Membrane Membrane Membrane Membrane Membrane Nucleus

Divalent sites Divalent substrate sites Divalent substrate sites Divalent sites Divalent substrate sites Divalent sites Divalent sites Divalent sites Divalent substrate sites Divalent substrate sites Divalent substrate sites Divalent sites Divalent sites Divalent substrate sites Divalent sites Divalent sites Divalent sites Divalent sites Divalent sites Divalent sites Divalent substrate sites Divalent sites Divalent agonist sites Divalent sites Divalent sites Divalent sites Di- or trivalent sites Di- to decavalent sites Trivalent sites Oligovalent sites Oligovalent sites Oliogovalent sites Oliogovalent sites Oliogovalent sites Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced dimer Ligand-induced oligomer Ligand-induced dimer

Section 2.2.2 3.3.2 3.3.3 3.3.6 3.3.7 4.4.5 4.4.6 4.4.7 4.4.10 4.4.11 4.4.12 4.4.14 4.4.14 5.2.1 5.2.2 5.2.2 2.2.4 4.1.5 5.2.2 4.4.1 3.3.1 3.1.8 4.6.4 4.1.1 4.1.4 4.1.15 4.1.3 4.2.4 4.1.16 4.4.2 4.4.8 4.4.9 2.8;4.7 2.8;3.4 4.3.1 4.3.1 3.3.5 4.1.11 4.3.2 4.3.2 4.3.2 4.3.2 4.3.2 3.1.6 4.3.3 (Continued)

370

APPENDIX

TABLE 20. (Continued) Target HIV-1 protease Estrogen receptor Influenza hemagglutinin Tryptase Lactate dehydrogenase Influenza neuraminidase CNG channel IP3 -gated calcium channel Potassium channel Concanavalin A Grionia simplicifolia-1 Heat-labile enterotoxin Cholera toxin Shigalike toxin Verotoxin B. anthracis anthraxtoxin DC-SIGN XL35 lectin Proteasome Picorna virus capsid HIV-1 glycoprotein 120 d-Ala-d-Ala peptide Lipid A H. pylori adhesin S. suis hemagglutinin E. coli FimH adhesin E. histolytica lectin Asialoglycoprotein receptor Sialoadhesin CD22 Bindin b-1,4-Galactosyltransferase Fertilinb Fucose-binding receptor E-Selectin L-Selectin CD14 glycoprotein MHC receptor Integrin b-Adrenergic receptor a-Adrenergic receptor Opioid receptor Serotonin receptor Mucarinic ACh-R Melanocortin-1 receptor

Type

Valency of System

Section

Soluble Nuclear membrane Membrane Soluble Soluble Membrane Membrane Membrane Membrane Soluble Soluble Soluble Soluble Soluble Soluble Soluble Membrane Soluble Soluble Surface Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane Membrane

Dimer Dimer; multivalent display Trimer; multivalent display Tetramer Tetramer; multivalent display Tetramer; multivalent display Tetramer; multivalent display Tetramer; multivalent display Tetramer; multivalent display Tetramer, dimer Tetramer (isolectin) Pentamer Pentamer Pentamer Pentamer Heptamer Oliogomer (tetramer) Oligomer Oligomer 60-mer Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display Multivalent display

2.2.1 4.2.5 2.1.1 4.4.3 4.4.13 2.1.2 4.6.1 4.6.2 4.6.3 4.1.16 4.1.16 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 4.1.7 4.1.9 4.4.4 5.2.1 2.2.3 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.1.7 4.1.2 4.1.6 4.1.9 4.1.9 4.1.9 4.1.10 4.1.13 4.1.14 4.2.1 4.2.2 4.2.3 4.5.1 4.5.2 4.5.3 4.5.4 4.5.5 4.5.6

REFERENCES

[1] Mammen, M., Choi, S.-K., Whitesides, G. M. Angew. Chem. Int. Ed. 1998, 37, 2755. [2] Dam, T. K., Roy, R., Das, S. K., Oscarson, S., Brewer, C. F. J. Biol. Chem. 2000, 275, 14223. [3] Alberts, B., Bray, D., Lewis, J., Ra¤, M., Roberts, K., Watson, J. D. Molecular Biology of the Cell, 2nd ed., Garland Publishing, New York, 1989. [4] Yang, T., Baryshnikova, O. K., Mao, H., Holden, M. A., Cremer, P. S. J. Am. Chem. Soc. 2003, 125, 4779. [5] Willems, G. M., Janssen, M. P., Pelsers, M. A. L., Comfurius, P., Galli, M., Zwaal, R. F., Bevers, E. M. Biochemistry 1996, 43, 13833. [6] Rosen, S. D., Bertozzi, C. R. Curr. Biol. 1996, 6, 261. [7] Baranowski, E., Ruiz-Jarabo, C. M., Domingo, E. Science 2001, 292, 1102. [8] Rossmann, M. G. Protein Sci. 1994, 3, 1712. [9] Karlsson, K.-A. Curr. Opin. Struct. Biol. 1995, 5, 622; Isberg, R. R. Science 1991, 252, 934. [10] Borman, S. Chem. Eng. News 2000, Oct. 9, 48. [11] Wright, D., Usher, L. Curr. Org. Chem. 2001, 5, 1107. [12] Romań, J. S., Gallardo, A., Levenfeld, B. Adv. Mater. 1995, 7, 203. [13] Mammen, M., Shakhnovich, E. I., Whitesides, G. M. J. Org. Chem. 1998, 63, 3168. [14] Glick, G. D., Toogood, P. L., Wiley, D. C., Skehel, J. J., Knowles, J. R. J. Biol. Chem. 1991, 266, 23660.


371

372

REFERENCES

[15] DeFrees, S. A., Kosch, W., Way, W., Paulson, J. C., Sabesan, S., Halcomb, R. L., Huang, D.-H., Ichikawa, Y., Wong, C.-H. J. Am. Chem. Soc. 1995, 117, 66. [16] Connolly, D. T., Townsend, R. R., Kawaguchi, K., Bell, W. R., Lee, Y. C. J. Biol. Chem. 1982, 257, 929. [17] Roy, R., Park, W. K. C., Wu, Q., Wang, S.-N. Tetrahedron Lett. 1995, 36, 4377. [18] Zhang, Z., Merritt, E. A., Ahn, M., Roach, C., Hou, Z., Verlinde, C. L. M. J., Hol, W. G. J., Fan, E. J. Am. Chem. Soc. 2002, 124, 12991. [19] Breslow, R., Belvedere, S., Gershell, L., Leung, D. Pure Appl. Chem. 2000, 72, 333. [20] Wenz, G. Angew. Chem. Int. Ed. 1994, 33, 803. [21] Fulton, D. A., Stoddart, J. F. Bioconjug. Chem. 2001, 12, 655. [22] Dondoni, A., Kleban, M., Hu, X., Marra, A., Banks, H. D. J. Org. Chem. 2002, 67, 4722. [23] Casnati, A., Sansone, F., Ungaro, R. Acc. Chem. Res. 2003, 36, 246; Marra, A., Scherrmann, M.-C., Dondoni, A., Casnati, A., Minari, P., Ungaro, R. Angew. Chem. Int. Ed. 1994, 33, 2479. [24] Hayashida, O., Kato, M., Akagi, K., Aoyama, Y. J. Am. Chem. Soc. 1999, 121, 11597. [25] Yasuda, N., Aoki, N., Abe, H., Hattori, K. Chem. Lett. 2000, 706; Meunier, S. J., Roy, R. Tetrahedron Lett. 1996, 37, 5469. [26] Lee, Y. C., Lee, R. T. Acc. Chem. Res. 1995, 28, 321. [27] Naylor, A. M., Goddard, W. A., III. J. Am. Chem. Soc. 1989, 111, 2339. [28] Tomalia, D. A., Berry, V., Hall, M., Hedstrand, D. M. Macromolecules 1987, 20, 1167. [29] Roy, R. Curr. Opin. Struct. Biol. 1996, 6, 692. [30] Turnbull, W. B., Stoddart, J. F. J. Biotechnol. 2002, 90, 231. [31] Bezouska, K. J. Biotechnol. 2002, 90, 269. [32] von der Lieth, C. W., Frank, M., Lindhorst, T. K. J. Biotechnol. 2002, 90, 311. [33] Roy, R., Zanini, D., Meunier, S. J., Romanowska, A. J. Chem. Soc. Chem. Commun. 1993, 1869. [34] Mammen, M., Dahmann, G., Whitesides, G. M. J. Med. Chem. 1995, 38, 4179. [35] Choi, S. K., Mammen, M., Whitesides, G. M. Chem. Biol. 1996, 3, 97. [36] Spaltenstein, A., Whitesides, G. M. J. Am. Chem. Soc. 1991, 113, 686. [37] Mortell, K. H., Gingras, M., Kiessling, L. L. J. Am. Chem. Soc. 1994, 116, 12053. [38] Roy, R., Saha, U. K. Chem. Commun. 1996, 210; Yuasaa, H., Kamataa, Y., Kuronob, S., Hashimotoa, H. Bioorg. Med. Chem. Lett. 1998, 8, 2139. [39] Bovin, N. V., Gabius, H.-J. Chem. Soc. Rev. 1995, 413. [40] Horan, N., Yan, L., Isobe, H., Whitesides, G. M., Kahne, D. Proc. Natl. Acad. Sci. USA 1999, 96, 11782. [41] Houseman, B. T., Mrksich, M. Chem. Biol. 2002, 9, 443. [42] Liang, M. N., Smith, S. P., Metallo, S. J., Choi, I. S., Prentis, M., Whitesides, G. M. Proc. Natl. Acad. Sci. USA 2000, 97, 13092. [43] Lin, C.-C., Yeh, Y.-C., Yang, C.-Y., Chen, C.-L., Chen, G. F., Chen, C.-C., Wu, Y.-C. J. Am. Chem. Soc. 2002, 124, 3508.

REFERENCES

373

[44] Kingery-Wood, J. E., Williams, K. W., Sigal, G. B., Whitesides, G. M. J. Am. Chem. Soc. 1992, 114, 7303. [45] Westwell, M. S., Bardsley, B., Try, A. C., Dancer, R. T., Williams, D. H. J. Chem. Soc. Chem. Commun. 1996, 589. [46] Genbacev, O. D., Prakobphol, A., Foulk, R. A., Krtolica, A. R., Singer, M. S., Yang, Z.-Q., Kiessling, L. L., Rosen, S. D., Fisher, S. J. Science 2003, 299, 405. [47] Lees, W. J., Spaltenstein, A., Kingery-Wood, J. E., Whitesides, G. M. J. Med. Chem. 1994, 37, 3419. [48] Spevak, W., Nagy, J. O., Charych, D. H., Schaefer, M. E., Gilbert, J. H., Bednarski, M. D. J. Am. Chem. Soc. 1993, 115, 1146. [49] Del Mar Hernandez, M., Jose, M. V. Anal. Biochem. 2003, 313, 226. [50] Rao, J., Lahiri, J., Isaacs, L., Weis, R. M., Whitesides, G. M. Science 1998, 280, 708; Rao, J., Lahiri, J., Weis, R. M., Whitesides, G. M. J. Am. Chem. Soc. 2000, 122, 2698. [51] Gargano, J. M., Ngo, T., Kim, J. Y., Acheson, D. W. K., Lees, W. J. J. Am. Chem. Soc. 2001, 123, 12909. [52] Kramer, R. H., Karpen, J. W. Nature 1998, 395, 710. [53] Dam, T. K., Roy, R., Page, D., Brewer, C. F. Biochemistry 2002, 41, 1359; Dam, T. K., Roy, R., Page, D., Brewer, C. F. Biochemistry 2002, 41, 1351. [54] For review, see: Dam, T. K., Brewer, C. F. Chem. Rev. 2002, 102, 387. [55] Chaires, J. B. Biopolymers 1998, 44, 201. [56] Chaires, J. B., Satyanarayana, S., Suh, D., Fokt, I., Przewloka, T., Priebe, W. Biochemistry 1996, 35, 2047. [57] Chaires, J. B., Leng, F., Przewloka, T., Fokt, I., Ling, Y.-H., Perez-Soler, R., Priebe, W. J. Med. Chem. 1997, 40, 261. [58] Page, M. I., Jencks, W. P. Proc. Natl. Acad. Sci. USA 1971, 68, 1678. [59] Hlavacek, W. S., Posner, R. G., Perelson, A. S. Biophys. J. 1999, 76, 3031. [60] Kiessling, L. L., Pohl, N. L. Chem. Biol. 1996, 3, 71. [61] Lee, R. T., Lee, Y. C. Glycoconj. J. 2000, 17, 543. [62] Pang, Y.-P., Quiram, P., Jelacic, T., Hong, F., Brimijoin, S. J. Biol. Chem. 1996, 271, 23646. [63] Tamiz, A. P., Zhang, J., Zhang, M., Wang, C. Z., Johnson, K. M., Kozikowski, A. P. J. Am. Chem. Soc. 2000, 122, 5393. [64] LeBoulluec, K. L., Mattson, R. J., Mahle, C. D., McGovern, R. T., Nowak, H. P., Gentile, A. J. Bioorg. Med. Chem. Lett. 1995, 5, 123. [65] Gestwicki, J. E., Kiessling, L. L. Nature 2002, 415, 81. [66] Miller, D. J., Macek, M. B., Shur, B. D. Science 1992, 357, 589. [67] Gong, X., Dubois, D., Miller, D. J., Shur, B. D. Science 1995, 269, 1718. [68] Crabtree, G. R., Schreiber, S. L. Trends Biochem. Sci. 1996, Nov. [69] Clemons, P. A. Curr. Opin. Chem. Biol. 1999, 3, 112. [70] Cardenas, M. E., Sanfridson, A., Shane, N., Heitman, J. Trends Biotechnol. 1998, 16, 427. [71] Arora, P. S., Ansari, A. Z., Best, T. P., Ptashne, M., Dervan, P. B. J. Am. Chem. Soc. 2002, 124, 13067.

374

REFERENCES

[72] Simanek, E. E., McGarvey, G. J., Jablonowski, J. A., Wong, C. H. Chem. Rev. 1998, 98, 833. [73] Yarema, K. J., Bertozzi, C. R. Curr. Opin. Chem. Biol. 1998, 2, 49. [74] Bertozzi, C. R., Kiessling, L. L. Science 2001, 291, 2357. [75] Dervan, P. B., Bu¨rli, R. W. Curr. Opin. Chem. Biol. 1999, 3, 688. [76] Wemmer, D. E., Dervan, P. B. Curr. Opin. Struct. Biol. 1997, 7, 355. [77] Matrosovich, M. N. FEBS Lett. 1989, 252, 1. [78] Mulvey, G., Kitov, P. I., Marcato, P., Bundle, D. R., Armstrong, G. D. Biochimie 2001, 83, 841. [79] Fan, E., Merritt, E. A. Curr. Drug Targets Infect. Disord. 2002, 2, 161. [80] Williams, D. H. Nat. Prod. Rep. 1996, 469. [81] Wiley, D. C., Skehel, J. J. Annu. Rev. Biochem. 1987, 56, 365; Harrison, S. C. Curr. Opin. Struct. Biol. 1995, 5, 157. [82] Wilson, I. A., Skehel, J. J., Wiley, D. C. Nature 1981, 289, 366. [83] Weis, W., Brown, J. H., Cusack, S., Paulson, J. C., Skehel, J. J., Wiley, D. C. Nature 1988, 333, 426. [84] Mellema, J. E., Andree, P. J., Krygsman, P. C. J., Kroon, C., Ruigrok, R. W. H., Cuasack, S., Miller, A., Zulauf, M. J. Mol. Biol. 1981, 151, 329. [85] Cusack, S., Ruigrok, R. W. H., Krygsman, P. C. J., Mellema, J. E. J. Mol. Biol. 1985, 186, 565. [86] Glick, G., Knowles, J. R. J. Am. Chem. Soc. 1991, 113, 4701. [87] Sabesan, S., Duus, J. Ø., Domaille, P., Kelm, S., Paulson, J. C. J. Am. Chem. Soc. 1991, 113, 5865. [88] Sabesan, S., Duus, J. Ø., Neira, S., Domaille, P., Kelm, S., Paulson, J. C., Bock, K. J. Am. Chem. Soc. 1992, 114, 8363. [89] Tuzikov, A. B., Chinarev, A. A., Gambaryan, A. S., Oleinikov, V. A., Klinov, D. V., Matsko, N. B., Kadykov, V. A., Ermishov, M. A., Demin, I. V., Demin, V. V., Rye, P. D., Bovin, N. V. Chembiochem 2003, 4, 147. [90] Roy, R., Meunier, S. J., Romanowska, A. In ACS Symposium Series 560, Kovać, P., Ed., American Chemical Society, Washington, DC, 1994, pp. 104–119. [91] Zanini, D., Roy, R. J. Org. Chem. 1996, 61, 7348. [92] Reuter, J. D., Myc, A., Hayes, M. M., Gan, Z., Roy, R., Qin, D., Yin, R., Piehler, L. T., Esfand, R., Tomalia, D. A., Baker, J. D. Bioconjug. Chem. 1999, 10, 271. [93] Haywood, A. M. J. Mol. Biol. 1974, 83, 427. [94] Alford, D., Ellens, H., Bentz, J. Biochemistry 1994, 33, 1977. [95] Ellens, H., Bentz, J., Mason, D., Zhang, F., White, J. Biochemistry 1990, 29, 9697. [96] Reichert, A., Nagy, J. O., Spevak, W., Charych, D. J. Am. Chem. Soc. 1995, 117, 829. [97] Charych, D., Cheng, Q., Reichert, A., Kuziemko, G., Stroh, M., Nagy, J. O., Spevak, W., Stevens, R. C. Chem. Biol. 1996, 3, 113. [98] Charych, D. H., Nagy, J. O., Spevak, W., Bednarski, M. D. Science 1993, 261, 585.

REFERENCES

375

[99] Spevak, W., Nagy, J. O., Charych, D. H. Adv. Mater. 1995, 7, 85. [100] Takemoto, D. K., Skehel, J. J., Wiley, D. C. Virology 1996, 217, 452. [101] Skehel, J. J., Bayley, P. M., Brown, E. B., Martin, S. R., Waterfield, M. D., White, J. M., Wilson, I. A., Wiley, D. C. Proc. Natl. Acad. Sci. USA 1982, 79, 968. [102] Matrosovich, M. N., Mochalova, L. U., Marinina, V. P., Byramova, N. E., Bovin, N. V. FEBS Lett. 1990, 272, 209. [103] For synthesis of preactivated polymer, see: Dikusar, M. A., Kubrakova, I. V., Chinarev, A. A., Bovin, N. V. Bioorg. Khim. 2001, 27, 457. [104] Burmeister, W. P., Ruigrok, R. W., Cusack, S. EMBO J. 1992, 11, 49. [105] Air, G. M., Laver, W. G. Proteins 1989, 6, 341. [106] Paulsen, H., Matschulat, P. Liebigs Ann. Chem. 1991, 487. [107] Nagy, J. O., Bednarski, M. D. Tetrahedron Lett. 1991, 32, 3953. [108] Lee, R. T., Lee, Y. C. Carbohydr. Res. 1974, 37, 193. [109] Sparks, M. A., Williams, K. W., Whitesides, G. M. J. Med. Chem. 1993, 36, 778. [110] Sigal, G. B., Mammen, M., Dahmann, G., Whitesides, G. M. J. Am. Chem. Soc. 1995, 118, 3789. [111] Mammen, M., Helmerson, K., Kishore, R., Choi, S.-K., Phillips, W. D., Whitesides, G. M. Chem. Biol. 1996, 3, 757. [112] Choi, S.-K., Mammen, M., Whitesides, G. M. J. Am. Chem. Soc. 1997, 119, 4103. [113] Kamitakahara, H., Suzuki, T., Nishigori, N., Kanie, O., Wong, C.-H. Angew. Chem. Int. Ed. 1998, 37, 1524. [114] Totani, K., Kubota, T., Kuroda, T., Murata, T., Hidari, K. I.-P. J., Suzuki, T., Suzuki, Y., Kobayashi, K., Ashida, H., Yamamoto, K., Usui, T. Glycobiology 2003, 13, 315. [115] Wu, W.-Y., Jin, B., Krippner, G. Y., Watson, K. G. Bioorg. Med. Chem. Lett. 2000, 10, 341. [116] Herrler, G., Rott, R., Klenk, H.-D., Mu¨ller, H.-P., Shukla, A., Schauer, R. EMBO 1985, 4, 1503. [117] Roy, R., Andersson, F. O., Harms, G., Kelm, S., Schauer, R. Angew. Chem. Int. Ed. 1992, 31, 1478. [118] Krizanova, O., Rathova, V. Curr. Top. Microbiol. Immunol. 1969, 47, 125. [119] Gottschaik, A., Belyavin, G., Biddle, F. In Glycoproteins: Their Compositions, Structure and Function, Gottschaik, A., Ed., Elsevier, New York, 1972, pp. 1082– 1096. [120] Hanaoka, K., Pritchett, T. J., Takasaki, S., Kochibe, N., Sabesan, S., Paulson, J. C., Kobata, A. J. Biol. Chem. 1989, 264, 9842; Pritchett, T. J., Paulson, J. C. J. Biol. Chem. 1989, 264, 9850. [121] Roy, R., Laferriere, C. A. Can. J. Chem. 1990, 68, 2045. [122] Roy, R., Gamian, A., Chomik, M. In Proceedings of the Japanese–German Symposium on Sialic Acids, Schauer, R., Yamakawa, T., Eds., Verlag Wissenschaft Bildung, Kiel, Germany, 1988, pp. 264–265. [123] Anders, E. M., Hartley, C. A., Jackson, D. C. Proc. Natl. Acad. Sci. USA 1990, 87, 4485.

376

REFERENCES

[124] Meanwell, N. A., Krystal, M. Drug Dev. Today 1996, 1, 316; Kim, C. U., Lew, W., Williams, M. A., Liu, H., Zhang, L., Swaminathan, S., Bischofberger, N., Chen, M. S., Mendel, D. B., Tai, C. Y., Laver, G., Stevens, R. C. J. Am. Chem. Soc. 1997, 119, 681; Hanessian, S., Bayrakdarian, M., Luo, X. J. Am. Chem. Soc. 2002, 124, 4716. [125] Guo, C.-T., Sun, X.-L., Kanie, O., Shortridge, K. F., Suzuki, T., Miyamoto, D., Hidari, K. I.-P., Wong, C.-H., Suzuki, Y. Glycobiology 2002, 12, 183. [126] Sabesan, S., Neira, S., Wasserman, Z. Carbohydr. Res. 1995, 267, 239. [127] Honda, T., Yoshida, S., Arai, M., Masuda, T., Yamashita, M. Bioorg. Med. Chem. Lett. 2002, 12, 1929. [128] Honda, T., Masuda, T., Yoshida, S., Arai, M., Kaneko, S., Yamashita, M. Bioorg. Med. Chem. Lett. 2002, 12, 1925. [129] Swanstrom, R., Erona, J. Pharmacol. Ther. 2000, 86, 145. [130] Wlodawer, A., Miller, A., Jaskolski, M., Sathyanarayana, B., Baldwin, E., Weber, I., Selk, L., Schneider, J., Kent, S. Science 1989, 245, 616. [131] Darke, P. L., Jordan, S. P., Hall, D. L., Zugay, J. A., Shafer, J. A., Kuo, L. C. Biochemistry 1994, 33, 98. [132] Grant, S. K., Deckman, I. C., Culp, J. S., Minnich, M. D., Brooks, I. S., Hensley, P., Debouck, C., Meek, T. D. Biochemistry 1992, 31, 9491. [133] Schramm, H. J., Nakashima, H., Schramm, W., Wakayama, H., Yamamoto, N. Biochem. Biophys. Res. Commun. 1991, 179, 847. [134] Zhang, Z., Poorman, R., Maggiora, L., Heinrickson, R., Kezdy, F. J. Biol. Chem. 1991, 266, 15591. [135] Schramm, H., Billich, A., Jaeger, E., Rucknagel, K., Arnold, G., Schramm, W. Biochem. Biophys. Res. Commun. 1993, 194, 595. [136] Babe, L. M., Rose, J., Craik, C. S. Protein Sci. 1992, 1, 1244. [137] Zutshi, R., Franciskovich, J., Shultz, M., Schweitzer, B., Bishop, P., Wilson, M., Chmielewski, J. J. Am. Chem. Soc. 1997, 119, 4841. [138] Franciskovich, J., Houseman, K., Mueller, R., Chmielewski, J. Bioorg. Med. Chem. Lett. 1993, 3, 765. [139] Bouras, A., Boggetto, N., Benatalah, Z., de Rosny, E., Sicsic, S., ReboudRavaux, M. J. Med. Chem. 1999, 42, 957. [140] Meneńdez-Arias, L. Trends Pharmacol. Sci. 2002, 23, 381; Jacobo-Molina, A., Arnold, E. Biochemistry 1991, 30, 6351. [141] Pontikis, R., Dolle´, V., Guillaumel, J., Dechaux, E., Note, R., Nguyen, C. H., Legraverend, M., Bisagni, E., Aubertin, A.-M., Grierson, D. S., Monneret, C. J. Med. Chem. 2000, 43, 1927. [142] Baranowski, E., Ruiz-Jarabo, C. M., Domingo, E. Science 2001, 292, 1102. [143] Lasky, L. A., Nakamura, G., Smith, D. H., Fennie, C., Shimasaki, C., Patzer, E., Berman, P., Gregory, T., Capon, D. Cell 1987, 50, 975. [144] Arthos, J., Deen, K. C., Chaikin, M. A., Fornwald, J. A., Sathe, G., Sattentau, Q. J., Clapham, P. R., Weiss, R. A., McDougal, J. S., Pietropaolo, C., Axel, R., Truneh, A., Maddon, P. J., Sweet, R. W. Cell 1989, 57, 469. [145] Ghetie, V., Slaughter, C., Wheeler, H. T., Uhr, J. W., Vitetta, E. S. Proc. Natl. Acad. Sci. USA 1991, 88, 5690.

REFERENCES

377

[146] Chiodi, F., Fuerstenberg, S., Gidlung, M., Asjo, B., Fenyo, M. J. Virol. 1987, 61, 1244. [147] Li, X. L., Moudgil, T., Vinters, H. V., Ho, D. D. J. Virol. 1990, 64, 1383. [148] Fantani, J., Cook, D. G., Nathanson, N., Spitalik, S. L., Gonzalez-Scarano, F. Proc. Natl. Acad. Sci. USA 1993, 90, 2700. [149] Harouse, J. M., Bhat, S., Spitalnik, S. L., Laughlin, M., Stefano, K., Silberberg, D. H., Gonzalez-Scarano, F. Science 1991, 253, 320. [150] Fantini, J., Cook, D. G., Nathanson, N., Spitalnik, S. L., Gonzalez-Scarano, F. Proc. Natl. Acad. Sci. USA 1993, 90, 2700. [151] Conboy, J. C., McReynolds, K. D., Gervay-Hague, J. J., Saavedra, S. S. Angew. Chem. Int. Ed. 2000, 39, 2882. [152] Gu, Y., LaBell, R., O’Brien, D. F., Saavedra, S. S. Angew. Chem. Int. Ed. 2001, 40, 2320. [153] Cook, D. G., Fantini, J., Spitalnik, S. L., Gonzalez-Scarano, F. Virology 1994, 201, 206. [154] Yahi, N., Sabatier, J.-M., Nickel, P., Mabrouk, K., Gonzalez-Scarano, F., Fantini, J. J. Biol. Chem. 1994, 269, 24349. [155] Leydet, A., Barragan, V., Boyer, B., Monte´ro, J. L., Roque, J. P., Witvrouw, M., Este, J., Snoeck, R., Andrei, G., De Clercq, E. J. Med. Chem. 1997, 40, 342. [156] Leydet, A., Barthelemy, Ph., Boyer, B., Lamaty, G., Roque, J. P., Bousseau, A., Evers, M., Henin, Y., Snoeck, R., Andrei, G., Ikeda, S., Reymen, D., De Clercq, E. J. Med. Chem. 1995, 38, 2433. [157] Leydet, A., El Hachemi, H., Boyer, B., Lamaty, G., Roque, J. P., Snoeck, R., Andrei, G., Ikeda, S., Reymen, D., Witvrouw, M., De Clercq, E. J. Med. Chem. 1996, 39, 1626. [158] Leydet, A., Jeantet-Segonds, C., Bouchitte´, C., Moullet, C., Boyer, B., Roque, J. P., Witvrouw, M., Este, J., Snoeck, R., Andrei, G., De Clercq, E. J. Med. Chem. 1997, 40, 350. [159] Leydet, A., Moullet, C., Roque, J. P., Witvrouw, M., Pannecouque, C., Andrei, G., Snoeck, R., Neyts, J., Schols, D., De Clercq, E. J. Med. Chem. 1998, 41, 4927. [160] Schols, D., De Clercq, E., Witvrouw, M., Nakashima, H., Snoeck, R., Pauwels, R., Van Schepdael, A., Claes, P. Antiviral Chem. Chemother. 1991, 2, 45. [161] Borman, S. Chem. Eng. News 2001, May 21, 63. [162] Esser, M. T., Mori, T., Mondor, I., Sattentau, Q. J., Dey, B., Berger, E. A., Boyd, M. R., Lifson, J. D. J. Virol. 1999, 73, 4360. [163] Boyd, M. R., Gutafson, K. R., McMahon, J. B., Shoemaker, R. H., O’Keefe, B. R., Mori, T., Gulakowski, R. J., Wu, L., Rivera, M. I., Laurencot, C. M., Currens, M. J., Cardellina, J. H., Buckhei, R. W., Nara, P. L., Pannell, L. K., Sowder, R. C., Henderson, L. E. Antimicrob. Agents Chemother. 1997, 41, 1521. [164] Bewley, C. A., Otero-Quintero, S. J. Am. Chem. Soc. 2001, 123, 3892. [165] Bourne, Y., Bolgiano, B., Liao, D., Strecker, G., Cantau, P., Herzberg, O., Feizi, T., Cambillau, C. Nat. Struct. Biol. 1994, 1, 863. [166] Shenoy, S. R., Barrientos, L. G., Ratner, D. M., O’Keefe, B. R., Seeberger, P. H., Gronenborn, A. M., Boyd, M. R. Chem. Biol. 2002, 9, 1109. [167] Karlsson, K.-A. Curr. Opin. Struct. Biol. 1995, 5, 121.

378

REFERENCES

[168] Yolken, R. H., Willoughby, R. E., Wee, S. B., Misku¤, R., Vonderfecht, S. J. Clin. Invest. 1987, 79, 148. [169] Willoughby, R. E. Glycobiology 1993, 3, 437. [170] Koketsu, M., Nitoda, T., Sugino, H., Juneja, L. R., Kim, M., Yamamoto, T., Abe, N., Kajimoto, T., Wong, C.-H. J. Med. Chem. 1997, 40, 3332. [171] Rueckeert, R. R. In Virology, 3rd ed., Fields, B. N., Knipe, D. M., Howley, P. M., Eds., Lippincott-Raven, Philadelphia, PA, 1996, pp. 609–654. [172] Rossmann, M. G. Protein Sci. 1994, 3, 1712. [173] Rossmann, M. G., Bella, J., Kolatkar, P. R., He, Y., Wimmer, E., Kuhn, R. J., Baker, T. S. Virology 2000, 269, 239. [174] Tsang, S. K., Cheh, J., Isaacs, L., Joseph-McCarthy, D., Choi, S.-K., Pevear, D. C., Whitesides, G. M., Hogle, J. M. Chem. Biol. 2001, 8, 33. [175] Joseph-McCarthy, D., Tsang, S. K., Fillman, D. J., Hogle, J. M., Karplus, M. J. Am. Chem. Soc. 2001, 123, 12758. [176] Heilman, C. A. J. Infect. Dis. 1990, 161, 402. [177] De Clercq, E. Int. J. Antimicrob. Agents 1996, 7, 193. [178] Nikitenko, A. A., Raifeld, Y. E., Wang, T. Z. Bioorg. Med. Chem. Lett. 2001, 11, 1041. [179] Ding, W.-D., Mitsner, B., Krishnamurthy, G., Aulabaugh, A., Hess, C. D., Cutler, M., Feld, B., Gazumyan, A., Raifeld, Y., Nikitenko, A., Lang, S. A., Gluzman, Y., O’Hara, B., Ellestad, G. A. J. Med. Chem. 1998, 41, 2671. [180] For a review, see: Gazumyan, A., Mitsner, B., Ellestad, G. A. Curr. Pharm. Des. 2000, 6, 525. [181] Razinkov, V., Gazumyan, A., Nikitenko, A., Ellestad, G., Krishnamurthy, G. Chem. Biol. 2001, 8, 645. [182] Wimmer, E. In Cellular Receptors for Animal Viruses, Wimmer, E., Ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1994, pp. 1–13. [183] Chen, Y., Maguire, T., Hileman, R. E., Fromm, J. R., Esko, J. D., Linhardt, R. J., Marks, R. M., Nat. Med. 1997, 3, 866. [184] Marks, R. M., Lu, H., Sundaresan, R., Toida, T., Suzuki, A., Imanari, T., Hernaíz, M. J., Linhardt, R. J. J. Med. Chem. 2001, 44, 2178. [185] Frankel, A. D., Young, J. A. Annu. Rev. Biochem. 1998, 67, 1. [186] Zapp, M. L., Stern, S., Green, M. R. Cell 1993, 74, 969. [187] Kirk, S. R., Luedtke, N. W., Tor, Y. J. Am. Chem. Soc. 2000, 122, 980. [188] Hamasaki, K., Ueno, A. Bioorg. Med. Chem. Lett. 2001, 11, 591. [189] Futaki, S., Araki, M., Kiwada, T., Nakase, I., Sugiura, Y. Bioorg. Med. Chem. Lett. 2001, 11, 1165. [190] Wade, W. S., Mrksich, M., Dervan, P. B. J. Am. Chem. Soc. 1992, 114, 8784. [191] Mrksich, M., Wade, W. S., Dwyer, T. J., Geierstanger, B. H., Wemmer, D. E., Dervan, P. B. Proc. Natl. Acad. Sci. USA 1992, 89, 7586. [192] White, S., Szewczyk, J. W., Turner, J. M., Baird, E. E., Dervan, P. B. Nature 1998, 391, 468. [193] Trauger, J. W., Baird, E. E., Dervan, P. B. Angew. Chem. Int. Ed. 1998, 37, 1421.

REFERENCES

379

[194] Dickinson, L. A., Gulizia, R. J., Trauger, J. W., Baird, E. E., Mosier, D. E., Gottesfeld, J. M., Dervan, P. B. Proc. Natl. Acad. Sci. USA 1998, 95, 12890. [195] Sharma, S. K., Tandon, M., Lown, J. W. J. Org. Chem. 2000, 65, 1102. [196] Thomson, R. H. Naturally Occurring Quinones, Vol. 2, Academic Press, London, 1971. [197] Laatsch, H. Angew. Chem. Int. Ed. 1994, 33, 422. [198] Decosterd, L. A., Parsons, I. C., Gustafson, K. R., Cardellina, J. H., II, McMahon, J. B., Cragg, G. M., Murata, Y., Pannell, L. K., Steiner, J. R., Clardy, J., Boyd, M. R. J. Am. Chem. Soc. 1993, 115, 6673. [199] Laatsch, H. Ann. Chem. 1987, 297. [200] Bringmann, G., Zagst, R., Scha¨¤er, M., Hallock, Y. F., Cardellina, J. H., II, Boyd, M. R., Angew. Chem. Int. Ed. 1993, 32, 1190; Hallock, Y. F., Cardellina, J. H., Scha¨¤er, M., Bringmann, G., François, G., Boyd, M. R. Bioorg. Med. Chem. Lett. 1998, 8, 1729; Bringmann, G., Holenz, J., Weirich, R., Ru¨benacker, M., Funke, C. Tetrahedron 1998, 54, 497. [201] Manfredi, K. P., Blunt, J. W., Cardellina, J. H., II, McMahon, J. B., Pannell, L. L., Cragg, G. M., Boyd, M. R. J. Med. Chem. 1991, 34, 3402. [202] Kragol, G., Otvos, L., Jr., Feng, J., Gerhard, W., Wade, J. D. Bioorg. Med. Chem. Lett. 2001, 11, 1417. [203] Wang, L. X., Singh, S. Bioorg. Med. Chem. 2003, 11, 159. [204] De Oliveira, E., Villen, J., Giralt, E., Andreu, D. Bioconjug. Chem. 2003, 14, 144. [205] Pantarotto, D., Partidos, C. D., Gra¤, R., Hoebeke, J., Briand, J.-P., Prato, M., Bianco, A. J. Am. Chem. Soc. 2003, 125, 6160. [206] Perkins, H. R. Biochem. J. 1969, 111, 195; Nieto, M., Perkins, H. R. Biochem. J. 1971, 123, 780. [207] Mackay, J. P., Gerhard, U., Beauregard, D. A., Maplestone, R. A., Williams, D. H. J. Am. Chem. Soc. 1994, 116, 4573. [208] Entress, R. M. H., Dancer, R. J., O’Brien, D. P., Try, A. C., Cooper, M. A., Williams, D. H. Chem. Biol. 1998, 5, 329. [209] McComas, C. C., Crowley, B. M., Boger, D. L. J. Am. Chem. Soc. 2003, 125, 9314; Wright, G. D., Walsh, C. T. Acc. Chem. Res. 1992, 25, 468; Walsh, C. T., Fisher, S. L., Park, I.-S., Prahalad, M., Wu, Z. Chem. Biol. 1996, 3, 21. [210] Cooper, M. A., Williams, D. H., Cho, Y. R. Chem. Commun. 1997, 1625. [211] Beauregard, D. A., Maguire, A. J., Williams, D. H., Reynolds, P. E. Antimicrob. Agents Chemother. 1997, 41, 2418. [212] Mackay, J. P., Gerhard, U., Beauregard, D. A., Westwell, M. S., Searle, M. S., Williams, D. H. J. Am. Chem. Soc. 1994, 116, 4581. [213] Gerhard, U., Mackay, J. P., Maplestone, R. A., Willaims, D. H. J. Am. Chem. Soc. 1993, 115, 232. [214] Sundram, U. N., Gri‰n, J. H., Nicas, T. I. J. Am. Chem. Soc. 1996, 118, 13107. [215] Loll, P. J., Miller, R., Weeks, C. M., Axelsen, P. H. Chem. Biol. 1998, 5, 293. [216] Schafer, M., Schneider, T. R., Sheldrick, G. M. Structure, 1996, 4, 1509. [217] Loll, P. H., Bevivino, A. E., Korty, A. E., Axelsen, P. H. J. Am. Chem. Soc. 1997, 119, 1516.

380 [218] [219] [220] [221] [222] [223] [224] [225] [226] [227]

[228] [229] [230] [231] [232] [233] [234]

[235] [236] [237] [238] [239] [240] [241] [242]

REFERENCES

Sundram, U. N., Gri‰n, J. H. J. Org. Chem. 1995, 60, 1102. Rao, J., Whitesides, G. M. J. Am. Chem. Soc. 1997, 119, 10286. Nieto, M., Perkins, H. R. Biochem. J. 1971, 123, 773. Nieto, M., Perkins, H. R. Biochem. J. 1971, 123, 789. Rao, J., Yan, L., Xu, B., Whitesides, G. J. Am. Chem. Soc. 1999, 121, 2629. Green, N. M. Biochem. J. 1963, 89, 585. Staroske, T., Williams, D. H. Tetrahedron Lett. 1998, 39, 4917. Adamczyk, M., Moore, J. A., Rege, S. D., Yu, Z. Bioorg. Med. Chem. Lett. 1999, 9, 2437. Stack, D. R., Thompson, R. G. EP0801075 A1, 1997. Gri‰n, J. H., Linsell, M. S., Nodwell, M. B., Chen, Q., Pace, J. L., Quast, K. L., Krause, K. M., Farrington, L., Wu, T. X., Higgins, D. L., Jenkins, T. E., Christensen, B. G., Judice, J. K. J. Am. Chem. Soc. 2003, 125, 6517. Nicolaou, K. C., Hughes, R., Cho, S. Y., Winssinger, N., Labischinski, H., Endermann, R. Chem. Eur. J. 2001, 7, 3824. Sztaricskai, F., Batta, G., Dinya, Z., Miroshnikova, O. V., Preobrazhenskaya, M. ´ ., Boda, Z. J. Antibiot. 2001, 54, 314. N., Herna´di, F., Koncz, A Arimoto, H., Nishimura, K., Kinumi, T., Hayakawa, I., Uemura, D. Chem. Commun. 1999, 1361. Schwab, P., France, M. B., Ziller, J. W., Grubbs, R. H. Angew. Chem. Int. Ed. 1995, 34, 2039. Sun, B., Chen, Z., Eggert, U. S., Shaw, S. J., LaTour, J. V., Kahne, D. J. Am. Chem. Soc. 2001, 123, 12722. Van Heijenoort, Y., Derrien, M., van Heijenoort, J. FEBS Lett. 1978, 89, 141. Sofia, M. J., Allanson, N., Hatzenbuhler, N. T., Jain, R., Kakarla, R., Kogan, N., Liang, R., Liu, D. S., Silva, D. J., Wang, H. M., Gange, D., Andersen, J., Chen, A., Chi, F., Dulina, R., Huang, B. W., Kamau, M., Wang, C. W., Baizman, E., Branstrom, A., Bristol, N., Goldman, R., Han, K. H., Longley, C., Midha, S., Axelrod, H. R., J. Med. Chem. 1999, 42, 3193. Baizman, E. R., Branstrom, A. A., Longley, C. B., Allanson, N., Sofia, M. J., Gange, D., Goldman, R. C. Microbiology 2000, 146, 3129. Nagarajan, R. J. Antibiot. 1993, 46, 118. Ge, M., Chen, Z., Onishi, H. R., Kohler, J., Silver, L. J., Kerns, R., Fukuzawa, S., Thompson, C., Kahne, D. Science 1999, 284, 507. Eggert, U. S., Ruiz, N., Falcone, B. V., Branstrom, A. A., Goldman, R. C., Silhavy, T. J., Kahne, D. Science 2001, 294, 361. Roy, R. S., Yang, P., Kodali, S., Xiong, Y., Kim, R. M., Gri‰n, P. R., Onishi, R., Kohler, J., Silver, L. L., Chapman, K. Chem. Biol. 2001, 139, 1. Monnee, M. C. F., Brouwer, A. J., Verbeek, L. M., van Wageningen, A. M. A., Liskamp, R. M. J. Bioorg. Med. Chem. Lett. 2001, 11, 1521. Ahrendt, K. A., Olsen, J. A., Wakao, M., Trias, J., Ellman, J. A. Bioorg. Med. Chem. Lett. 2003, 13, 1683. Xu, R., Greiveldinger, G., Marenus, L. E., Cooper, A., Ellman, J. A. J. Am. Chem. Soc. 1999, 37, 1563.

REFERENCES

381

[243] Chiosis, G., Boneca, I. G. Science 2001, 293, 1484; Shi, Z., Gri‰n, J. H. J. Am. Chem. Soc. 1993, 115, 6482. [244] Schwartz, B., Markwalder, J. A., Wang, Y. J. Am. Chem. Soc. 2001, 123, 11638. [245] VanNieuwenhze, M., Mauldin, S. C., Zia-Ebrahimi, M., Winger, B. E., Hornback, W. J., Saha, S. L., Aikins, J. A., Blaszczak, L. C. J. Am. Chem. Soc. 2002, 124, 3656. [246] Sadamoto, R., Niikura, K., Sears, P. S., Liu, H., Wong, C.-H., Suksomcheep, A., Tomita, F., Monde, K., Nishimura, S.-I. J. Am. Chem. Soc. 2002, 124, 9018. [247] Vaara, M. Antimicrob. Agents Chemother. 1993, 37, 354. [248] Li, C., Peters, A. S., Meredith, E. L., Allman, G. W., Savage, P. B. J. Am. Chem. Soc. 1998, 120, 2961. [249] Hubbard, R. D., Horner, S. R., Miller, B. J. J. Am. Chem. Soc. 2001, 123, 5810. [250] David, S. A., Bechtel, B., Annaiah, C., Mathan, V. I., Balaram, P. Biochim. Biophys. Acta 1994, 1212, 167. [251] Chan, S., Horner, S. R., Fauchet, P. M., Miller, B. L. J. Am. Chem. Soc. 2001, 123, 11797. [252] For review, see: Ritchie, G. E., Mo¤att, B. E., Sim, R. B., Morgan, B. P., Dwek, R. A., Rudd, P. M. Chem. Rev. 2002, 102, 305. [253] Reid, K. B. M., Porter, R. Biochem. J. 1976, 155, 19; Knobel, H. R., Villiger, W., Isliker, H. Eur. J. Immunol. 1975, 5, 78. [254] Ringsdorf, H., Schlarb, B., Venzmer, J. Angew. Chem. Int. Ed. 1988, 27, 113. [255] Lingwood, C. A. Curr. Opin. Chem. Biol. 1998, 2, 695. [256] Ilver, D., Arnqvist, A., Ogren, J., Frick, I.-M., Kersulyte, D., Incecik, E. T., Covacci, A., Engstrand, L., Boreń, T. Science 1998, 279, 373. [257] Boreń, T., Falk, P., Roth, K. A., Larson, G., Normark, S. Science 1993, 262, 1892. [258] Randolph, J., Danishefsky, S. J. Angew. Chem. Int. Ed. 1994, 33, 1470. [259] Bernstein, M. A., Hall, D. L. Carbohydr. Res. 1980, 78, C1. ˚ ng[260] Mahdavi, J., Sondeń, B., Hurtig, M., Olfat, F. O., Forsberg, L., Roche, N., A stro¨m, J., Larsson, T., Teneberg, S., Karlsson, K.-A., Altraja, S., Wadstro¨m, T., Kersulyte, D., Berg, D. E., Dubois, A., Petersson, C., Magnusson, K.-E., Norberg, T., Lindh, F., Lundskog, B. B., Arnqvist, A., Hammarstro¨m, L., Boreń, T. Science 2002, 297, 573. [261] Simon, P. M., Goode, P. L., Mobasseri, A., Zopf, D. Infect. Immun. 1997, 65, 750. [262] Kihlberg, J., Hultgren, S. J., Normark, S., Magnusson, G. J. Am. Chem. Soc. 1989, 111, 6364. [263] Haataja, S., Tikkanen, K., Nilsson, U., Magnusson, G., Karlsson, K.-A., Finne, J. J. Biol. Chem. 1994, 269, 27466. [264] Hansen, H., Haataja, S., Finne, J., Magnusson, G. J. Am. Chem. Soc. 1997, 119, 6974. [265] Mulvey, M. A., Lopez-Boado, Y. S., Wilson, C. L., Roth, R., Parks, W. C., Heuser, J., Hultgren, S. J. Science 1998, 282, 1494. [266] Abraham, S. N., Sun, D. X., Dale, J. B., Beachey, E. H. Nature 1988, 336, 682.

382

REFERENCES

[267] Kuehn, M. J., Heuser, J., Normark, S., Hultgren, S. J. Nature 1992, 356, 252. [268] Sokurenko, E. V., Schembri, M. A., Trintchina, E., Kjaergaard, K., Hasty, D. L., Klemm, P. Mol. Microbiol. 2001, 41, 675. [269] Liang, M. N., Smith, S. P., Metallo, S. J., Choi, I. S., Prentis, M., Whitesides, G. M. Proc. Natl. Acad. Sci. USA 2000, 97, 13092. [270] Li, J., Zacharek, S., Chen, X., Wang, J., Zhang, W., Janczuk, A., Wang, P. G. Bioorg. Med. Chem. Lett. 1999, 7, 1549; Hayes, W., Osborn, H. M. I., Osborne, S. D., Rastall, R. A., Romagnoli, B. Tetrahedron Lett. 2002, 43, 7683. [271] Lin, C.-C., Yeh, Y.-C., Yang, C.-Y., Chen, C.-L., Chen, G.-F., Chen, C.-C., Wu, Y.-C. J. Am. Chem. Soc. 2002, 124, 3508. [272] Parkinson, J. S. Cell 1993, 73, 857. [273] Alon, U., Surette, M. G., Barkai, N., Leibler, S. Nature 1998, 393, 18. [274] Grebe, T. W., Stock, J. Curr. Biol. 1998, 8, R154; Adler, J., Hazelbauer, G. L., Dahl, M. M. J. Bacteriol. 1973, 115, 824. [275] Kim, S.-H., Wang, W., Kim, K. K. Proc. Natl. Acad. Sci. USA 2002, 99, 11611. [276] Hazelbauer, G. L., Adler, J. Nat. New Biol. 1971, 230, 101. [277] Gestwicki, J. E., Strong, L. E., Kiessling, L. L. Chem. Biol. 2000, 7, 583; Gestwicki, J. E., Kiessling, L. L. Nature 2002, 415, 81. [278] Gestwicki, J. E., Strong, L. E., Borchardt, S. L., Cairo, C. W., Schnoes, A. M., Kiessling, L. L. Bioorg. Med. Chem. 2001, 9, 2387. [279] Lamanna, A. C., Gestwicki, J. E., Strong, L. E., Borchardt, S. L., Owen, R. M., Kiessling, L. L. J. Bacteriol. 2002, 184, 4981. [280] Barnakov, A. N., Downing, K. H., Hazelbauer, G. L. J. Struct. Biol. 1994, 74, 77. [281] Adler, P., Wood, S. J., Lee, Y. C., Lee, R. T., Petri, W. A., Jr., Schnaar, R. L. J. Biol. Chem. 1995, 270, 5164; McCoy, J. J., Mann, B. J., Petri, W. A. J. Infect. Immun. 1994, 62, 3045. [282] Mayer, S., Boos, M. S., Beyer, A., Fluit, A. C., Schmitz, F. J. J. Antimicrob. Chemother. 2001, 47, 869. [283] Schumacher, M. A., Miller, M. C., Grkovic, S., Brown, M. H., Skurray, R. A., Brennan, R. G. Science 2001, 294, 2158. [284] Mitchell, B. A., Paulsen, I. T., Brown, M. H., Skurray, R. A. J. Biol. Chem. 1999, 274, 3541. [285] Paulsen, I. T., Brown, M. H., Littlejohn, T. G., Mitchell, B. A., Skurray, R. A. J. Biol. Chem. 1996, 93, 3630. [286] Merritt, E. A., Hol, W. G. J. Curr. Opin. Struct. Biol. 1995, 5, 165. [287] Schengrund, C. L. Biochem. Pharmacol. 2003, 65, 699. [288] Fan, E., Merritt, E. A., Verlinde, C. L. M. J., Hol, W. G. L. Curr. Opin. Struct. Biol. 2000, 10, 680. [289] Thompson, J. P., Schengrund, C.-L. Biochem. Pharmacol. 1998, 56, 591. [290] Sixma, T. K., Pronk, S. E., Kalk, K. H., Van Zanten, B. A., Berghuis, A. M., Hol, W. G. J. Nature 1992, 355, 561. [291] Merritt, E. A., Sarfaty, S., Feil, I. K., Hol, W. G. J. Structure 1997, 5, 1485. [292] Fan, E., Zhang, Z., Minke, W. E., Hou, Z., Verlinde, C. L. M., Hol, W. G. J. J. Am. Chem. Soc. 2000, 122, 2663.

REFERENCES

383

[293] Knoll, D., Hermans, J. J. Biol. Chem. 1983, 258, 5710. [294] Nilsson, U. J., Heerze, L. D., Liu, Y.-C., Armstrong, G. D., Palcic, M. M., Hindsgaul, O. Bioconjug. Chem. 1997, 8, 466. [295] Cuatrecasas, P. Biochemistry 1973, 12, 3547. [296] Masserini, M., Freire, E., Palestini, P., Calappi, E., Tettamanti, G. Biochemistry 1992, 31, 2422. [297] Lauer, S., Goldstein, B., Nolan, R. L., Nolan, J. P. Biochemistry 2002, 41, 1742. [298] Goins, B., Freire, E. Biochemistry 1988, 27, 2046; Scho¨n, A., Freire, E. Biochemistry 1989, 28, 5019. [299] Zhang, Z., Merritt, E. A., Ahn, M., Roach, C., Hou, Z., Verlinde, C. L. M. J., Hol, W. G. J., Fan, E. J. Am. Chem. Soc. 2002, 124, 12991. [300] For review, see: Williams, S. J., Davies, G. J. Trends Biotechnol. 2001, 19, 356. [301] Merritt, E. A., Zhang, Z., Pickens, J. C., Ahn, M., Hol, W. G., Fan, E. J. Am. Chem. Soc. 2002, 124, 8818. [302] Mertz, J. A., McCann, J. A., Picking, W. D. Biochem. Biophys. Res. Commun. 1996, 226, 140. [303] Vrasidas, I., de Mol, N. J., Liskamp, R. M. J., Pieters, R. J. Eur J. Org. Chem. 2001, 4685. [304] Charych, D., Cheng, Q., Reichert, A., Kuziemko, G., Stroh, M., Nagy, J. O., Spevak, W., Stevens, R. C. Chem. Biol. 1996, 3, 113. [305] Kuziemko, G. M., Stroh, M., Stevens, R. C. Biochemistry 1996, 35, 6375. [306] Song, X., Swanson, B. I. Anal. Chem. 1999, 71, 2097. [307] See verotoxin as a close example: Shimizu, H., Field, R. A., Homans, S. W., Donohue-Rolfe, A. Biochemistry 1998, 37, 11078. [308] Nyholm, P.-G., Magnusson, G., Zheng, Z., Norel, R., Binnington-Boyd, B., Lingwood, C. A. Chem. Biol. 1996, 3, 263. [309] Imai, Y., Matsuura, Y., Ono, Y., Ishikawa, T., Ito, Y. J. Biochem. 2001, 130, 665. [310] Kitova, E. N., Kitov, P. I., Bundle, D. R., Klassen, J. S. Glycobiology 2001, 11, 605. [311] Ling, H., Boodhoo, A., Hazes, B., Cumings, M. D., Armstrong, G. D., Brunton, J. L., Read, R. J. Biochemistry 1998, 37, 1777. [312] Nakjima, H., Kiyokawa, N., Katagiri, Y. U., Taguchi, T., Suzuki, T., Sekino, T., Mimori, K., Ebata, T., Saito, M., Nakao, H., Takeda, T., Fujimoto, J. J. Biol. Chem. 2001, 276, 42915. [313] Lundquist, J. J., Debenham, S. D., Toone, E. J. J. Org. Chem. 2000, 65, 8245. [314] Kitov, P. I., Shimizu, H., Homans, S. W., Bundle, D. R. J. Am. Chem. Soc. 2003, 125, 3284. [315] Kitov, P. I., Sadowska, J. M., Mulvey, G., Armstrong, G. D., Ling, H., Pannus, N. S., Read, R. J., Bundle, D. R. Nature 2000, 403, 669. [316] Pozsgay, V., Trinh, L., Shiloach, J., Robbins, J. B., Donohue-Rolfe, A., Calderwood, S. B. Bioconjug. Chem. 1996, 7, 45. [317] Richardson, J. M., Evans, P. D., Homans, S. W., Donohue-Rolfe, A. Nat. Struct. Biol. 1997, 4, 190. [318] Sellmann, B. R., Mourez, M., Collier, R. J. Science 2001, 292, 695.

384

REFERENCES

[319] Leppla, S. H. In Bacterial Toxins and Virulence Factors in Diseases: Handbook of Natural Toxins, Vol. 8, Moss, J., Iglewski, B., Vaughan, M., Tu, A., Eds., Marcel Dekker, New York, 1995, pp. 543–572. [320] Mourez, M., Kane, R. S., Metallo, S., Deschatelets, P., Sellman, B. R., Whitesides, G. M., Collier, R. J. Nat. Biotechnol. 2001, 19, 958. [321] Walsh, C. Nature 2000, 406, 775. [322] Labiscinski, H., Maidohof, H. In Bactaerial Cell Wall, Ghuysen, J.-M., Hakenbeck, J., Eds., Elsevier, Amsterdam, 1994, pp. 23–39. [323] Kelly, J. A., Knox, J. R., Zhao, H., Fre´re, J. M., Ghuysen, J.-M. J. Mol. Biol. 1989, 154, 281. [324] Lee, W., McDonough, M. A., Kotra, L. P., Li, Z.-H., Silvaggi, N. R., Takeda, Y., Kelly, J. A., Mobashery, S. Proc. Natl. Acad. Sci. USA 2001, 98, 1427. [325] Rodriguez-Tebar, A., Aran, V., Vazquez, D. Eur. J. Biochem. 1984, 139, 287. ´ ., Herczegh, P., Sztaricskai, F., Szabo´, P. J. Antibiot. 2000, [326] Kovaćs-Kulyassa, A 53, 1207. [327] Usakov, S. N., Trukmanova, L. B., Drozdova, E. U., Monkelova, T. M. Dokl. Akad. Nauk. SSSR 1961, 141, 1117. [328] Panarin, E. F., Solovskii, M. W., Ekzenphylarov, O. N. Khim. Farm. Zh. 1975, 5, 24. [329] Talarico, T. L., Ray, P. H., Dev, I. K., Merrill, B. M., Dallas, W. S. J. Bacteriol. 1991, 173, 7029. [330] Shi, G., Blaszczyk, J., Ji, X., Yan, H. J. Med. Chem. 2001, 44, 1364. [331] Bliska, J. B., Guan, K. L., Dixon, J. E., Falkow, S. Proc. Natl. Acad. Sci. USA 1991, 88, 1187. [332] Chen, Y. T., Onaran, M. B., Doss, C. J., Seto, C. T. Bioorg. Med. Chem. Lett. 2001, 11, 1935. [333] Chen, Y. T., Seto, C. T. J. Med. Chem. 2002, 45, 3946. [334] Quinõa´, E., Crews, P. Tetrahedron Lett. 1987, 52, 3584. [335] Arabshahi, L., Schmitz, F. J. J. Org. Chem. 1987, 52, 3584. [336] Nicolaou, K. C., Hughes, R., Pfe¤erkorn, J. A., Barluenga, S. Chem. Eur. J. 2001, 7, 4296. [337] Kim, D., Lee, I. S., Jung, J. H., Yang, S.-I. Arch. Pharm. Res. 1999, 22, 25. [338] Nicolaou, K. C., Hughes, R., Pfe¤erkorn, J. A., Barluenga, S., Roecker, A. J. Chem. Eur. J. 2001, 7, 4280. [339] Drlica, K. Curr. Opin. Microbiol. 1999, 2, 504. [340] Kerns, R. J., Rubak, M. J., Kaatz, G. W., Vaka, F., Cha, R., Grucz, R. G., Diwadkar, V. U., Ward, T. D. Bioorg. Med. Chem. Lett. 2003, 13, 1745. [341] Kerns, R. J., Rybak, M. J., Kaatz, G. W., Vaka, F., Cha, R., Grucz, R. G., Diwadkar, V. U. Bioorg. Med. Chem. Lett. 2003, 13, 2109. [342] Kopytek, S. J., Standaert, R. F., Dyer, J. C. D., Hu, J. C. Chem. Biol. 2000, 7, 313. [343] Taylor, G. Curr. Opin. Struct. Biol. 1996, 6, 830. [344] Crennell, S., Garman, E., Laver, G., Vumir, E., Taylor, G. Structure 1994, 2, 535.

REFERENCES

385

[345] Thobhani, S., Ember, B., Siriwardena, A., Boons, G.-J. J. Am. Chem. Soc. 2003, 125, 7154. [346] Gaskell, A., Crennell, S., Taylor, G. Structure 1995, 3, 1197. [347] Rizzi, M., Nessi, C., Mattevi, A., Coda, A., Bolognesi, M., Galizzi, A. EMBO J. 1996, 15, 5125. [348] Hara, N., Yamada, K., Terashima, M., Osago, H., Shimoyama, M., Tsuchiya, M. J. Biol. Chem. 2003, 278, 10914. [349] Garcia, J. G., Brouillette, C. G., Pierson, M. C., DeLucas, L. J., Brouillette, W. J. Book of Abstracts, 218th ACS National Meeting, New Orleans, LA, Aug. 22–26, 1999, MEDI-297; Garcia, J. G., DeLucas, L. J., Brouillette, W. J. Book of Abstracts, 216th ACS National Meeting, Boston, Aug. 23–27, 1998, MEDI-248. [350] Formy, D., Recht, M. I., Blanchard, S. C., Puglisi, J. D. Science 1996, 274, 1367. [351] Hendrix, M., Alper, P. B., Priestley, E. S., Wong, C.-H. Angew. Chem. Int. Ed. Engl. 1997, 36, 95; Sucheck, S. J., Wong, A. L., Koeller, K. M., Boehr, D. D., Draker, K.-A., Sears, P., Wright, G. D., Wong, C.-H. J. Am. Chem. Soc. 2000, 122, 5230. [352] Kirk, S. R., Tor, Y. Bioorg. Med. Chem. Lett. 1999, 7, 1979. [353] Tok, J. B.-H., Hu¤man, G. R. Bioorg. Med. Chem. Lett. 2000, 10, 1593. [354] Tanaka, H., Nishida, Y., Furuta, Y., Kobayashi, K. Bioorg. Med. Chem. Lett. 2002, 12, 1723. [355] Sood, R. K., Fattom, A., Pavliak, V., Naso, R. B. Drug Discov. Today 1996, 1, 381. [356] Maclead, C. M., Hodges, R. G., Heidelberger, M., Bernhard, W. G. J. Exp. Med. 1945, 82, 445. [357] Pozsgay, V. Angew. Chem. Int. Ed. 1998, 37, 138; Pozsgay, V. J. Org. Chem. 1998, 63, 5983. [358] Gurjar, M. K., Viswanadham, G. Tetrahedron Lett. 1991, 32, 6191; Ziegler, T. Angew. Chem. Int. Ed. 1992, 31, 1358. [359] Turco, S. J., Descoteaux, A. Annu. Rev. Microbiol. 1992, 46, 65. [360] Hewitt, M., Seeberger, P. H. J. Org. Chem. 2001, 66, 4233. [361] McGuire, E. J., Binkley, S. B. Biochemistry 1964, 3, 247. [362] Timmis, K. N., Boulnois, G. J., Bitter-Suermann, D., Cabello, F. C. Curr. Top. Microbiol. Immunol. 1985, 118, 197. [363] Frosch, M., Go¨rgen, I., Boulnois, G. J., Timmis, K. N., Bitter-Suermann, D. Proc. Natl. Acad. Sci. USA 1985, 82, 1194. [364] Ringsdorf, H., Schlarb, B., Venzmer, J. Angew. Chem. Int. Ed. Engl. 1988, 27, 113. [365] Paoletti, L. C., Pinel, J., Johnson, K. D., Reinap, B., Ross, R. A., Kasper, D. L. J. Infect. Dis. 1999, 180, 892. [366] Auzanneau, F.-I., Pinto, B. M. Bioorg. Med. Chem. 1996, 4, 2003. [367] For a review that deals with protein-based multivalent vaccine, see: Hu, M. C., Walls, M. A., Stroop, S. D., Reddish, M. A., Beall, B., Dale, J. B. Infect. Immun. 2002, 70, 2171.

386

REFERENCES

[368] Bolard, J. Biochim. Biophys. Acta 1986, 864, 257. [369] Matsumori, N., Yamaji, N., Matsuoka, S., Oishi, T., Murata, M. J. Am. Chem. Soc. 2002, 124, 4180. [370] Baenziger, J. U., Kumar, S., Brodbeck, R. M., Smith, P. L., Beranek, M. C. Proc. Natl. Acad. Sci. USA 1992, 89, 334. [371] Roseman, D. S., Baenziger, J. U. Proc. Natl. Acad. Sci. USA 2000, 97, 9949. [372] Roseman, D. S., Baenziger, J. U. J. Biol. Chem. 2001, 276, 17052. [373] Van Den Hamer, C. J., Morell, A. G., Scheinberg, I. H., Hickman, J., Ashwell, G. J. Biol. Chem. 1970, 245, 4397; Bider, M. D., Wahlberg, J. M., Kammerer, R. A., Spiess, M. J. Biol. Chem. 1996, 271, 31996. [374] Biessen, E. A., Beuting, D. M., Roelen, H. C. P. F., van de Marel, G. A., van Boom, J. H., van Berkel, T. J. C. J. Med. Chem. 1995, 38, 1538. [375] Valentijin, A. R. P. M., van de Marel, G. A., Sliedregt, L. A. J. M., Berkel, T. J. C., Biessen, E. A. L., van Boom, J. H. Tetrahedron 1997, 53, 759. [376] Biessen, E. A. L., Broxterman, H., van Boom, J. H., van Berkel, T. J. C. J. Med. Chem. 1995, 38, 1846. [377] Kempen, H. J. M., Hoes, C., van Boom, J. H., Spanjer, H. H., Langendoen, A., van Berkel, T. J. C. J. Med. Chem. 1984, 27, 1306. [378] Ren, T., Zhang, G., Liu, D. Bioorg. Med. Chem. 2001, 9, 2969. [379] For a review of lectin-mediated drug delivery, see: Yamzaki, N., Kojima, S., Bovin, N. V., Andre´, S., Gabius, S., Gabius, H.-J. Adv. Drug Deliv. Rev. 2000, 43, 225. [380] Zhang, X., Simmons, C. G., Corey, D. R. Bioorg. Med. Chem. Lett. 2001, 11, 1269. [381] Maier, M. A., Yannopoulos, C. G., Mohamed, N., Roland, A., Fritz, H., Mohan, V., Just, G., Manoharan, M. Bioconjug. Chem. 2003, 14, 18. [382] David, A., Kopeckova´, P., Rubinstein, A., Kopecek, J. Bioconjug. Chem. 2001, 12, 890. [383] Yang, Y., Thomas, V. H., Man, S., Rice, K. G. Glycobiology 2000, 10, 1341. [384] Kim, S.-H., Goto, M., Akaike, T. J. Biol. Chem. 2001, 276, 35312. [385] Weis, W. I., Taylor, M. E., Drickamer, K. Immunol. Rev. 1998, 163, 19; Weis, W., Drickamer, K. Structure 1994, 2, 1227. [386] Weis, W. I., Drickmer, K., Hendrickson, W. A. Nature 1992, 360, 127. [387] Ng, K. K.-S., Kolatkar, A. R., Park-Snyder, S., Feinberg, H., Clark, D. A., Drickamer, K., Weis, W. I. J. Biol. Chem. 2002, 277, 16088. [388] Quesenberry, M. S., Lee, R. T., Lee, Y. C. Biochemistry 1997, 36, 2724. [389] Wang, L.-X., Tang, M., Suzuki, T., Kitajima, K., Inoue, Y., Inoue, S., Fan, J.-Q., Lee, Y. C. J. Am. Chem. Soc. 1997, 119, 11137. [390] Fan, J.-Q., Quesenberry, M. S., Takegawa, K., Iwahara, S., Kondo, A., Kato, I., Lee, Y. C. J. Biol. Chem. 1995, 270, 17730. [391] Grandjean, C., Angyalosi, G., Loing, E., Adriaenssens, E., Melnyk, O., Pancre´, V., Auriault, C., Gras-Masse, H. ChemBiochem 2001, 2, 747. [392] Angyalosi, G., Grandjean, C., Lamirand, M., Auriault, C., Gras-Masse, H., Melnyk, O. Bioorg. Med. Chem. Lett. 2002, 12, 2723.

REFERENCES

387

[393] Schuster, M. C., Mann, D. A., Buchholz, T. J., Johnson, K. M., Thomas, W. D., Kiessling, L. L. Org. Lett. 2003, 5, 1407. [394] Rini, J. M., Lobsanov, Y. D. Curr. Opin. Struct. Biol. 1999, 9, 578. [395] Tong, P. Y., Gregory, W., Kornfeld, S. J. Biol. Chem. 1989, 264, 7962. [396] Byrd, J. C., MacDonald, R. G. J. Biol. Chem. 2000, 275, 18638. [397] Byrd, J. C., Park, J. H. Y., Scha¤er, B. S., Barmroudi, F., MacDonald, R. G. J. Biol. Chem. 2000, 275, 18647. [398] Rothenberg, B. E., Hayes, B. K., Toomre, D., Manzi, A. E., Varki, A. Proc. Natl. Acad. Sci. USA 1993, 90, 11939. [399] Di Virgilio, S., Glushka, J., Moremen, K., Pierce, M. Glycobiology 1999, 9, 353. [400] Rabinovich, G. A., Baum, L. G., Tinari, N., Paganelli, R., Natoli, C., Liu, F.-T., Iacobelli, S. Trends Immunol. 2002, 23, 313. [401] Sacchettini, J. C., Baum, L. G., Brewer, C. F. Biochemistry 2001, 40, 3009. [402] Brewer, C. F. Biochem. Biophys. Acta 2002, 1572, 255. [403] Perillo, N. L., Pace, K. E., Seilhamer, J. J., Baum, L. G. Nature 1995, 378, 736. [404] Pace, K. E., Lee, C., Stewart, P. L., Baum, L. G. J. Immunol. 1999, 163, 3801. [405] Brewer, C. F., Miceli, M. C., Baum, L. G. Curr. Opin. Struct. Biol. 2002, 12, 616. [406] Andre´, S., Pieters, R. J., Vrasidas, I., Kaltner, H. K., Kuwabara, I., Liu, F.-T., Liskamp, R. M. J., Gabius, H.-J. ChemBiochem 2001, 2, 822. [407] Andre´, S., Ortega, P. J., Perez, M. A., Roy, R., Gabius, H. J. Glycobiology 1999, 9, 1253. [408] Stamenkovic, I., Seed, B. Nature 1990, 345, 74. [409] van Rossenberg, S. M. W., Sliedregt, L. A. M., Autar, R., Piperi, C., Van der Merwe, A. P., van Berkel, T. J. C., Kuiper, J., Biessen, E. A. L. J. Biol. Chem. 2001, 276, 12967. [410] van Rossenberg, S. M. W., Sliedregt, L. A. J. M., Autar, R., Piperi, C., Van der Merwe, A. P., van Berkel, T. J. C., Kuiper, J., Biessen, E. A. L. J. Biol. Chem. 2001, 276, 12967. [411] Sliedregt, L. A. J. M., van Rossenberg, S. M. W., Autar, R., Valentijn, A. R. P. M., vander Marel, G. A., van Boom, J. H., Piperi, C., van der Merwe, P. A., Kuiper, J., van Berkel, T. J. C., Biessen, E. A. L. Bioorg. Med. Chem. 2001, 9, 85. [412] Feinberg, H., Mitchell, D. A., Drickamer, K., Weis, W. I. Science 2001, 294, 2163. [413] Curtis, B. M., Scharnowske, S., Watson, A. J. Proc. Natl. Acad. Sci. USA 1992, 89, 8356. [414] Frison, N., Taylor, M. E., Soilleux, E., Bousser, M.-T., Mayer, R., Monsigny, M., Drickamer, K., Roche, A.-C. J. Biol. Chem. 2003, 278, 23922. [415] Van Die, I., van Vliet, S. J., Nyame, A. K., Cummings, R. D., Bank, C. M. C., Applemelk, B., Geijtenbeek, T. B. H., van Kooyk, Y. Glycobiology 2003, 13, 471. [416] Kerr, J. B. Microsc. Res. Tech. 1995, 32, 364. [417] Muramatsu, T. Science 2002, 295, 53. [418] Akama, T. O., Nakagawa, H., Sugihara, K., Narisawa, S., Ohyama, C., Nishimura, S.-I., O’Brien, D. A., Moremen, K. W., Millań, J. L., Fukuda, M. N. Science 2002, 295, 124.

388

REFERENCES

[419] For review, see: Mengerink, K. J., Vacquier, V. D. Glycobiology 2001, 11, 37. [420] Foltz, K. R., Lennarz, W. J. J. Cell. Biol. 1990, 111, 2951. [421] Kitazume-Kawaguchi, S., Inoue, S., Inoue, Y., Lennarz, W. J. Proc. Natl. Acad. Sci. USA 1997, 94, 3650. [422] Dhume, S. T., Lennarz, W. J. Glycobiology 1995, 5, 11. [423] Dhume, S. T., Stears, R. L., Lennarz, W. J. Glycobiology 1996, 6, 59. [424] Primako¤, P., Myles, D. G. Science 2002, 296, 2183. [425] Gupta, S., Sampson, N. S. Org. Lett. 2001, 3, 3333. [426] Gupta, S., Li, H., Nicole, N. S., Sampson, S. Bioorg. Med. Chem. Lett. 2000, 8, 723. [427] Arranz-Alex, E., Tracy, A. S., Siriwardena, A., Pierce, J. M., Boons, G.-J. J. Am. Chem. Soc. 2002, 124, 13035. [428] Gabius, H. J., Engelhardt, R., Hellmann, T., Midoux, P., Monsigny, G. A., Nagel, G. A., Vehmeyer, K. Anticancer Res. 1987, 7, 109. [429] Lerchen, H.-G., Baumgarten, J., Piel, N., Kolb-Bachofen, V. Angew. Chem. Int. Ed. 1999, 38, 3680. [430] Rosenfelder, G., Young, W. W., Hakomori, S. Cancer Res. 1977, 37, 1333. [431] Otsuji, E., Park, Y. S., Tashiro, K., Kojima, N., Toyokuni, T., Hakomori, S. Int. J. Oncol. 1995, 6, 319. [432] Jaye, M., Schlessinger, J., Dionne, C. A. Biochim. Biophys. Acta 1992, 1135, 185. [433] Herr, A., Ornitz, D. M., Sasisekharan, R., Venkataraman, G., Waksman, G. J. Biol. Chem. 1997, 272, 16382. [434] DiGabriele, A. D., Lax, I., Chen, D. I., Svahn, C. M., Jaye, M., Schlessinger, J., Hendrickson, W. A. Nature 1998, 393, 812. [435] Eggens, I., Fenderson. B., Toyokuni, T., Dean, B., Stroud, M., Hakomori, S. J. Biol. Chem. 1989, 264, 9476. [436] Kojima, N., Hakomori, S. J. Biol. Chem. 1991, 266, 17552. [437] Iwabuchi, K., Yamamura, S., Prinetti, A., Handa, K., Hakomori, S. J. Biol. Chem. 1998, 273, 9130. [438] Kojima, N., Fenderson, B. A., Stroud, M. R., Goldberg, R. I., Habermann, R., Toyokuni, T., Hakomori, S. Glycoconj. J. 1994, 11, 238. [439] Matsuura, K., Kitakouji, H., Sawada, N., Ishida, H., Kiso, M., Kitajima, K., Kobayashi, K. J. Am. Chem. Soc. 2000, 122, 7406. [440] Yu, S., Kojima, N., Hakomori, S.-I., Kudo, S., Inoue, S., Inoue, Y. Proc. Natl. Acad. Sci. USA 2002, 99, 2854. [441] Popescu, O., Misevic, G. N. Nature 1997, 386, 231. [442] Spillmann, D., Thomas-Oates, J. E., van Kujik, J. A., Vliegenthart, J. F. G., Misevic, G., Burger, M. M. Finne, J. J. Biol. Chem. 1995, 270, 5089. [443] Haseley, S. R., Vermeer, H. J., Kamerling, J. P., Vliegenthart, J. F. G. Proc. Natl. Acad. Sci. USA 2001, 98, 9419. [444] Boubelik, M., Floryk, D., Bohata, J., Draberova, L., Macak, J., Smid, F., Draber, P. Glycobiology 1998, 8, 139. [445] de la Fuente, J. M., Barrientos, A. G., Rojas, T. C., Rojo, J., Canada, J., Fernańdez, A., Penade´s, S. Angew. Chem. Int. Ed. 2001, 40, 2258.

REFERENCES

389

[446] Paulson, J. C. In The Receptors, Conn, M., Ed., Academic Press, New York, 1985, Vol. 2, pp. 131–219. [447] Springer, T. A., Lasky, L. A. Nature 1991, 349, 196. [448] Lawrence, M. B. Curr. Opin. Chem. Biol. 1999, 3, 659. [449] Rosen, S. D., Bertozzi, C. R. Curr. Opin. Cell Biol. 1994, 6, 663. [450] DeFrees, S. A., Gaeta, F. C. A., Lin, Y.-C., Ichikawa, Y., Wong, C.-H. J. Am. Chem. Soc. 1993, 115, 7549. [451] Feizi, T. Curr. Opin. Struct. Biol. 1993, 3, 701. [452] Sears, P., Wong, C.-H. Proc. Natl. Acad. Sci. USA 1996, 93, 12086. [453] DeFrees, S. A., Kosch, W., Way, W., Paulson, J. C., Sabesan, S., Halcomb, R. L., Huang, D.-H., Ichikawa, Y., Wong, C.-H. J. Am. Chem. Soc. 1995, 117, 66. [454] Wittmann, V., Takayama, S., Gong, K. W., Weitz-Schmidt, G., Wong, C.-H. J. Org. Chem. 1998, 63, 5137. ¨ hrlein, R. Angew. Chem. Int. Ed. 1996, 35, 1812. [455] Baisch, G., O [456] Sprengard, U., Schudok, M., Schmidt, W., Kretzschmar, G., Kunz, H. Angew. Chem. Int. Ed. 1996, 35, 321. [457] Welply, J. K., Abbas, S. Z., Scuddler, P., Keene, J. L., Broschat, K., Casnocha, S., Gorka, C., Steininger, C., Howard, S. C., Schmuke, J. J. Glycobiology 1994, 4, 259. [458] DeFrees, S. A., Phillips, L., Guo, L., Zalipsky, S. J. Am. Chem. Soc. 1996, 118, 6101. [459] Allen, T. M., Brandeis, E., Hansen, C. B., Kao, G. Y., Zalipsky, S. Biochem. Biophys. Acta 1995, 1237, 99. [460] Stahn, R., Schafer, H., Kernchen, F., Schreiber, J. Glycobiology 1998, 8, 311. [461] Sche¿er, K., Ernst, B., Katopodis, A., Magnani, J. L., Wang, W. T., Weisemann, R., Peters, T. Angew. Chem. Int. Ed. Engl. 1995, 34, 1841. [462] Lin, C.-C., Kimura, T. J., Wu, S.-H., Weitz-Schmidt, G., Wong, C.-H. Bioorg. Med. Chem. Lett. 1996, 6, 2755. ¨ hrlein, R., Ernst, B., Schwarzenbach, F., Duthaler, [463] Thoma, G., Magnani, J. L., O R. O. J. Am. Chem. Soc. 1997, 119, 7414. [464] Thoma, G., Duthaler, R. O., Magnani, J. L., Patton, J. T. J. Am. Chem. Soc. 2001, 123, 10113. [465] Thoma, G., Kinzy, W., Bruns, C., Patton, J. T., Magnani, J. L., Ba¨nteli, R. J. Med. Chem. 1999, 42, 4909. [466] Weitz-Schmidt, G., Stokmaier, D., Scheel, G., Nifant’ev, N., Tuzikov, A. B., Bovin, N. V. Anal. Biochem. 1996, 238, 184. [467] Kwon, M., Jeong, S., Lee, K. H., Park, Y.-K., Yu, J. J. Am. Chem. Soc. 2002, 124, 13996. [468] Kansas, G. S. Blood 1996, 88, 3259. [469] Baumheuter, S., Singer, M. S., Henzel, W., Hemmerich, S., Renz, M., Rosen, S. D., Lasky, L. A. Science 1993, 262, 436. [470] Lawrence, M. B. Curr. Opin. Chem. Biol. 1999, 3, 659. [471] Maaheimo, H., Renkonen, R., Turunen, J. P., Penttila, L., Renkonen, O. Eur. J. Biochem. 1995, 234, 616. [472] Toppila, S., Lauronen, J., Mattila, P., Turunen, J. P., Penttila¨, L., Paavonen, T., Renkonen, O., Renkonen, R. Eur. J. Immunol. 1997, 27, 1360.

390

REFERENCES

[473] Nicolaou, K. C., Hummel, C. W., Iwabuchi, Y. J. Am. Chem. Soc. 1992, 114, 3126. [474] Bru¨ning, T., Kiessling, L. L. Tetrahedron Lett. 1996, 37, 2907. [475] Sanders, W. J., Gordon, E. J., Dwir, O., Beck, P. J., Alon, R., Kiessling, L. L. J. Biol. Chem. 1999, 274, 5271. [476] Gordon, E. J., Sanders, W. J., Kiessling, L. L. Nature 1998, 392, 30. [477] Gordon, E. J., Strong, L. E., Kiessling, L. L. Bioorg. Med. Chem. Lett. 1998, 6, 1293. [478] Gordon, E. J., Gestwicki, J. E., Strong, L. E., Kiesling, L. L. Chem. Biol. 2000, 7, 9. [479] Welply, J. K., Abbas, S. Z., Scudder, P., Keene, J. L., Broschat, K., Casnocha, S., Gorka, C., Steininger, C., Howard, S. C., Schumuke, J. J. Glycobiology 1994, 4, 259. [480] Sasaki, K., Nishida, Y., Tsurumi, T., Uzawa, H., Kondo, H., Kobayashi, K. Angew. Chem. Int. Ed. 2002, 41, 4463. [481] Renkonen, O., Toppila, S., Penttila, L., Salminen, H., Helin, J., Maaheimo, H., Costello, C. E., Turunen, J. P., Renkonen, R. Glycobiology 1997, 7, 453. [482] Koeller, K. M., Smith, M. E. B., Wong, C.-H. J. Am. Chem. Soc. 2000, 122, 742. [483] Rosen, S. D., Bertozzi, C. R. Curr. Biol. 1996, 6, 261. [484] Leppanen, A., Mehta, P., Ouyang, Y.-B., Ju, T., Helin, J., Moore, K. L., van Die, I., Canfield, W. M., McEver, R. P., Cummings, R. D. J. Biol. Chem. 1999, 274, 24838. [485] Sprengard, U., Kretzschmar, G., Bartnik, E., Hu¨ls, C., Kunz, H. Angew. Chem. Int. Ed. 1995, 34, 990. [486] Matsuda, M., Nishimura, S.-I., Nakajima, F., Nishimura, T. J. Med. Chem. 2001, 44, 715. [487] Pochechueva, T. V., Galanina, O. E., Bird, M., Nifantiev, N. E., Bovin, N. V. Chem. Biol. 2002, 9, 757. [488] Manning, D. D., Hu, X., Beck, P., Kiessling, L. L. J. Am. Chem. Soc. 1997, 119, 3161. [489] Diacovo, T. G., Puri, K. D., Warnock, R. A., Springer, T. A., Vonandrian, U. H. Science 1996, 273, 252. [490] Spevak, W., Foxall, C., Charych, D. H., Dasgupta, F., Nagy, J. O. J. Med. Chem. 1996, 39, 1018. [491] For related article, see: Bruehl, R. E., Dasgupta, F., Katsumoto, T. R., Tan, J. H., Bertozzi, C. R., Spevak, W., Ahn, D. J., Rosen, S. D., Nagy, J. O. Biochem. 2001, 40, 5964. [492] Karlsson, K.-A. Curr. Opin. Struct. Biol. 1995, 5, 622. [493] Goldstein, I. J., Poretz, R. D. In The Lectins: Properties, Functions, and Applications in Biology and Medicine, Liener, I. E., Sharon, N., Goldstein, I. J., Eds., Academic Press, New York, 1986. [494] Mann, D. A., Kanai, M., Maly, D. J., Kiessling, L. L. J. Am. Chem. Soc. 1998, 120, 10575. [495] Bhattacharyya, L., Haraldsson, M., Brewer, C. F. Biochemistry 1988, 27, 1034.

REFERENCES

391

[496] Knibbs, R. N., Takagaki, M., Blake, D. A., Goldstein, I. J. Biochemistry 1998, 37, 16952. [497] Dimick, S. M., Powell, S. C., McMahon, S. A., Moothoo, D. N., Naismith, J. H., Toone, E. J. J. Am. Chem. Soc. 1999, 121, 10286; Corbell, J. B., Lundquist, J. J., Toone, E. J. Tetrahedron: Asymmetry 2000, 11, 95. [498] Burke, S. D., Zhao, Q., Schuster, M. C., Kiessling, L. L. J. Am. Chem. Soc. 2000, 122, 4518. [499] Fijimoto, T., Shimizu, C., Hayashida, O., Aoyama, Y. J. Am. Chem. Soc. 1998, 120, 601. [500] Ortega-Caballero, F., Gimeńez-Martıńez, J. J., Garcıá-Fuentes, L., OrtizSalmeroń, E., Santoyo-Gonza´lez, F., Vargas-Berenguel, A. J. Org. Chem. 2001, 66, 7786. [501] For other bifunctional dendritic glycosides (mannoside and galactoside) presented on b-cyclodextrin, see: Baussanne, I., Benito, J. M., Mellet, C. O., Fernańdez, J. M. G., Defaye, J. ChemBiochem 2001, 2, 777. [502] Vargas-Berenguel, A., Ortega-Caballero, F., Santoyo-Gonza´lez, F., GarcıáLope´z, J., Gimeńzez-Martıńez, J. J., Garcıá-Fuentes, L., Ortiz-Salmeroń, E. Chem. Eur. J. 2002, 8, 812. [503] Mellet, C., Defaye, J., Fernańdez, J. M. G. Chem. Eur. J. 2002, 8, 1982. [504] Ashton, P. R., Hounsell, E. F., Jayaraman, N., Nilsen, T. M., Spencer, N., Stoddart, J. F., Young, M. J. Org. Chem. 1998, 63, 3429. [505] Page´, D., Zanini, D., Roy, R. Bioorg. Med. Chem. 1996, 4, 1949. [506] For mannoside-functionalized dendrimer based on PAMAM, see: Woller, E. K., Cloninger, M. J. Org. Lett. 2002, 4, 7. [507] Mortell, K. H., Gingras, M., Kiessling, L. L. J. Am. Chem. Soc. 1994, 116, 12053. [508] Mortell, K. H., Weatherman, R. V., Kiessling, L. L. J. Am. Chem. Soc. 1996, 118, 2297. [509] Kanai, M., Mortell, K. H., Kiessling, L. L. J. Am. Chem. Soc. 1997, 119, 9931. [510] Strong, L. E., Kiessling, L. L. J. Am. Chem. Soc. 1999, 121, 6193. [511] Mann, D. A., Kanai, M., Maly, D. J., Kiessling, L. L. J. Am. Chem. Soc. 1998, 120, 10575. [512] Cairo, C. W., Gestwicki, J. E., Kanai, M., Kiessling, L. L. J. Am. Chem. Soc. 2002, 124, 1615. [513] Gestwicki, J. E., Strong, L. E., Kiessling, L. L. Angew. Chem. Int. Ed. 2000, 39, 4567. [514] Gestwicki, J. E., Strong, L. E., Cairo, C. W., Boehm, F. J., Kiessling, L. L. Chem. Biol. 2002, 9, 163. [515] Houseman, B. T., Mrksich, M. Chem. Biol. 2002, 9, 443. [516] Liang, R., Loebach, J., Horan, N., Ge, M., Thompson, C., Yan, L., Kahne, D. Proc. Natl. Acad. Sci. USA 1997, 94, 10554. [517] Horan, N., Yan, L., Isobe, H., Whitesides, G. M., Kahne, D. Proc. Natl. Acad. Sci. USA 1999, 96, 11782. [518] Zanini, D., Roy, R. J. Am. Chem. Soc. 1997, 119, 2088. [519] Zanini, D., Roy, R. J. Org. Chem. 1998, 63, 3486; Zanini, D., Roy, R. Bioconjug. Chem. 1997, 8, 187; Baek, M.-G., Roy, R. Bioorg Med. Chem. 2002, 10, 11.

392

REFERENCES

[520] Garcıá-Lo´pez, J. J., He´rnandez-Mateo, F., Isac-Garcıá, J., Kim, J. M., Roy, R., Santoyo-Gonza´lez, F., Vargas-Berengue, A. J. Org. Chem. 1999, 64, 522. [521] Roy, R., Hernańdez-Mateo, F., Santoyo-Gonza´lez, F. J. Org. Chem. 2000, 65, 8743. [522] Sun, X.-L., Faucher, K. M., Houston, M., Grande, D., Chaikof, E. L. J. Am. Chem. Soc. 2002, 124, 7258. [523] Vetter, D., Tate, E. M., Gallop, M. A. Bioconjug. Chem. 1995, 6, 319. [524] Frankel, A. E., Burbage, C., Fu, T., Tagge, E., Chandler, J., Willingham, M. C. Biochem. 1996, 35, 14749. [525] Kitano, H., Sohda, K., Kosaka, A. Bioconjug. Chem. 1995, 6, 131; Kitano, H., Ohno, K. Langmuir 1994, 10, 4131. [526] Bryan, M. C., Plettenburg, O., Sears, P., Rabuka, D., Wacowich-Sgarbi, S., Wong, C.-H. Chem. Biol. 2002, 9, 713. [527] Fazio, F., Bryan, M. C., Blixt, O., Paulson, J. C., Wong, C.-H. J. Am. Chem. Soc. 2002, 124, 14397. [528] Siriwardena, A., Jørgensen, M. R., Wolfert, M. A., Vandenplas, M. L., Moore, J. N., Boons, G.-J. J. Am. Chem. Soc. 2001, 123, 8145. [529] Krishna, S., Benaroch, P., Pillai, S. Nature 1992, 357, 164. [530] Cochran, J. R., Aivazian, D., Cameron, T. O., Stern, L. J. Trends Biochem. Sci. 2001, 26, 304. [531] Cochran, J. R., Stern, L. J. Chem. Biol. 2000, 7, 683. [532] Miyamoto, S., Akiyama, S. K., Yamada, K. M. Science 1995, 267, 883. [533] Komazawa, H., Saiki, I., Igarashi, Y., Azuma, I., Kojima, M., Orikasa, A., Ono, M., Itoh, I. J. Bioact. Compat. Polym. 1993, 8, 258. [534] Maynard, H., Okada, S. Y., Grubbs, R. H. J. Am. Chem. Soc. 2001, 123, 1275. [535] Leahy, D. J., Aukhil, I., Erickson, H. P. Cell 1996, 84, 155. [536] Miyamoto, S., Akiyama, S. K., Yamada, K. M. Science 1995, 267, 883. [537] Marchi-Artzner, V., Lorz, B., Hellerer, U., Kantlehner, M., Kesseler, H., Sackmann, E. Chem. Eur. J. 2001, 7, 1095. [538] Singhvi, R., Kumar, A., Lopez, G. P., Stephanopoulas, G. N., Wang, D. I. C., Whitesides, G. M., Ingber, D. E. Science 1994, 264, 696. [539] Yousaf, M. N., Houseman, B. T., Mrksich, M. Angew. Chem. Int. Ed. 2001, 40, 1093. [540] Xiong, J.-P., Stehle, T., Diefenbach, B., Zhang, R., Dunker, R., Scott, D. L., Joachimiak, A., Goodman, S. L., Arnaout, M. A. Science 2001, 294, 339. [541] Hood, J. D., Bednarski, M., Frausto, R., Guccione, S., Reisfeld, R. A., Xiang, R., Cheresh, D. A. Science 2002, 296, 2404. [542] Silletti, S., Kessler, T., Goldberg, J., Boger, D. L., Cheresh, D. A. Proc. Natl. Acad. Sci. USA 2001, 98, 119. [543] Male, D., Champion, B., Cooke, A. In Advanced Immunology, Male, D., Champion, B., Cooke, A., Eds., Gower Medical, London, 1987, pp. 2.1–2.13. [544] Terskikh, A. V., Le Doussal, J.-M., Crameri, R., Fisch, I., Mach, J.-M., Kajava, A. V. Proc. Natl. Acad. Sci. USA 1997, 94, 1663. [545] Stryer, L. Biochemistry, 2nd ed., W.H. Freeman, San Francisco, 1981.

REFERENCES

393

[546] Bergmann, K. E., Wooge, C. H., Carlson, K. E., Katzenellenbogen, B. S., Katzenellenbogen, J. A. Steroid Biochem. Mol. Biol. 1994, 49, 139. [547] Groleau, S., Nault, J., Lepage, M., Couture, M., Dallaire, N., Be´rube´, G., C.-Gaudreault, R. Bioorg. Chem. 1999, 27, 383. [548] Rabouin, D., Perron, V., N’Zemba, B., C.-Gaudreault, R., Be´rube´, G. Bioorg. Med. Chem. Lett. 2003, 13, 557. [549] Lin, H., Cornish, V. W. Angew. Chem. Int. Ed. 2001, 40, 871. [550] Siekierka, J. J., Hung, S. H., Poe, M., Lin, C. S., Sigal, N. H. Nature 1989, 341, 755. [551] Pruschy, M. N., Spencer, D. M., Kapoor, P. M., Miyaki, H., Crabtree, G. R., Schreiber, S. L. Chem. Biol. 1994, 1, 163. [552] Holsinger, L. J., Spencer, D. M., Austin, D. J., Schreiber, S. L., Crabtree, G. R. Proc. Natl. Acad. Sci. USA 1995, 92, 9810. [553] Spencer, D. M., Graef, I., Austin, D. J., Schreiber, S. L., Crabtree, G. R. Proc. Natl. Acad. Sci. USA 1995, 92, 9805. [554] MacCorkle, R. A., Freeman, K. W., Spencer, D. M. Proc. Natl. Acad. Sci. USA 1998, 95, 3655. [555] Ho, S. N., Biggar, S. R., Spencer, D. M., Schreiber, S. L., Crabtree, G. R. Nature 1996, 382, 822. [556] Amara, J. F., Clackson, T., Rivera, V., Guo, T., Keenan, T., Natesan, S., Pollock, R., Yang, W., Courage, N. L., Holt, D. A., Gilman, M. Proc. Natl. Acad. Sci. USA 1997, 94, 10618. [557] Belshaw, P., Spencer, D. M., Crabtree, G. R., Schreiber, S. L. Chem. Biol. 1996, 3, 731. [558] Belshaw, P. J., Ho, S. N., Crabtree, G. R., Schreiber, S. L. Proc. Natl. Acad. Sci. USA 1996, 93, 4604. [559] Briesewitz, R., Ray, G. T., Wandless, T. J., Crabtree, G. R. Proc. Natl. Acad. Sci. USA 1999, 96, 1953. [560] Rosen, M. K., Amos, C. D., Wandless, T. A. J. Am. Chem. Soc. 2000, 122, 11979. [561] Koide, K., Finkelstein, J. M., Ball, Z., Verdine, G. L. J. Am. Chem. Soc. 2001, 123, 398. [562] Altho¤, E. A., Cornish, V. W. Angew. Chem. Int. Ed. 2002, 41, 2327. [563] Shuker, S. B., Hajduk, P. J., Meadows, R. P., Fesik, S. W. Science 1996, 274, 1531. [564] Mills, K. V., Lew, B. M., Jiang, S., Paulus, H. Proc. Natl. Acad. Sci. USA 1998, 95, 3543. [565] Mootz, H., Muir, T. W. J. Am. Chem. Soc. 2002, 124, 9044. [566] Syed, R. S., Reid, S. W., Li, C., Cheetham, J. C., Aoki, K. H., Liu, B., Zhan, H., Osslund, T. D., Chirino, A. J., Zhang, J., Finer-Moore, J., Elliott, S., Sitney, K., Katz, B. A., Matthews, D. J., Wendoloski, J. J., Egrie, J., Stroud, R. M. Nature 1998, 395, 511. [567] Bazan, F. J. Proc. Natl. Acad. Sci. USA 1990, 87, 6934. [568] Wrighton, N. C., Farrell, F. X., Chang, R., Kashyap, A. K., Barbone, F. P., Mulcahy, L. S., Johnson, D. L., Barrett, R. W., Jolli¤e, L. K., Dower, W. J., Science 1996, 273, 458.

394

REFERENCES

[569] Cwirla, S. E., Balasubramanian, P., Du‰n, D. J., Wagstrom, C. R., Gates, C. M., Singer, S. C., Davis, A. M., Tansik, R. L., Mattheakis, L. C., Boytos, C. M. Science 1997, 276, 1696. [570] Goldberg, J., Jin, Q., Ambroise, Y., Satoh, S., Desharnais, J., Capps, K., Boger, D. L. J. Am. Chem. Soc. 2002, 124, 544. [571] Qureshi, S. A., Kim, R. M., Konteatis, Z., Biazzo, D. E., Motamedi, H., Rodrigues, R., Boice, J. A., Calaycay, J. R., Bednarek, M. A., Gri‰n, P., Gao, Y. D., Chapman, K., Mark, D. F. Proc. Natl. Acad. Sci. USA 1999, 96, 12156. [572] Johnson, D. L., Farrell, F. X., Barbone, F. P., McMahon, F. J., Tullai, J., Kroon, D., Freedy, J., Zivin, R. A., Mulcahy, L. S., Jolli¤ee, L. K. Chem. Biol. 1997, 4, 939. [573] England, B. P., Balasubramanian, P., Uings, L., Bethell, S., Chen, M.-J., Schatz, P., Yin, Q., Chen, Y.-F., Whitehorn, E. A., Tsavaler, A., Martens, C. L., Barrett, R. W., McKinnon, M. Proc. Natl. Acad. Sci. USA 2000, 97, 6862. [574] Tian, S.-S., Lamb, P., King, A. G., Miller, S. G., Kessler, L., Luengo, J. I., Averill, L., Johnson, R. K., Gleason, J. G., Pelus, L. M., Dillon, S. B., Rosen, J. Science 1998, 281, 257. [575] Zhang, A. J., Khare, S., Gokulan, K., Linthicum, D. S., Burgess, K. Bioorg. Med. Chem. Lett. 2001, 11, 207; Pattarawarapan, M., Reyes, S., Xia, Z., Zaccaro, M. C., Saragovi, H. U., Burgess, K. J. Med. Chem. 2003, 46, 3565. [576] Sestelo, J. P., Mourinõ, A., Sarandeses, L. A. J. Org. Chem. 2000, 65, 8290; Sestelo, J. P., Mourinõ, A., Sarandeses, L. A. Org. Lett. 1999, 1, 1005. [577] Gentry, M. K., Doctor, B. P. In Cholineesterases: Structure, Function, Mechanism, Genetics, and Cell Biology, Massoulie, J., Bacou, F., Barnard, E. A., Chatonnet, A., Doctor, B. P., Eds., American Chemical Society, Washington, DC, 1991, pp. 394–398. [578] Brufani, M., Filocamo, L., Lappa, S., Maggi, A. Drugs Future 1997, 22, 397. [579] Harel, M., Schalk, I., Ehret-Sabatier, L., Bouet, F., Goeldner, M., Hirth, C., Axelsen, P. H., Silman, I., Sussman, J. L. Proc. Natl. Acad. Sci. USA 1993, 90, 9031. [580] Pang, Y.-P., Quiram, P., Jelacic, T., Hong, F., Brimijoin, S. J. Biol. Chem. 1996, 271, 23646. [581] Carlier, P. R., Chow, E. S.-H., Han, Y., Liu, J., Yazal, J. E., Pang, Y.-P. J. Med. Chem. 1999, 42, 4225; Han, Y. F., Li, C. P.-L., Chow, E., Wang, H., Pang, Y.-P., Carlier, P. R. Bioorg. Med. Chem. 1997, 7, 2569. [582] Hu, M.-K., Wu, L.-J., Hsiao, G., Yen, M.-H. J. Med. Chem. 2002, 45, 2277. [583] Wong, D. M., Greenblat, H. M., Dvir, H., Carlier, P. R., Han, Y.-F., Pang, Y.-P., Silman, I., Sussman, J. L. J. Am. Chem. Soc. 2003, 125, 363. [584] Deprez, P., Inestrosa, N. C., Krejci, E. J. Biol. Chem. 2003, 278, 23233. [585] Hajduk, P. J., Sheppard, G., Nettesheim, D. G., Olejniczak, E. T., Shuker, S. B., Meadows, R. P., Steinman, D. H., Carrera, G. M., Jr., Marcotte, P. A., Severin, J., Walter, K., Smith, H., Gubbins, E., Simmer, R., Holzman, T. F., Morgan, D. W., Davidsen, S. K., Summers, J. B., Fesik, S. W. J. Am. Chem. Soc. 1997, 119, 5818. [586] Gooley, P. R., O’Connell, J. F., Marcy, A. I., Cuca, G. C., Salowe, S. P., Bush, B. L., Hermes, J. D., Esser, C. K., Hagmann, W. K., Springer, J. P., Johnson, B. A. Nat. Struct. Biol. 1995, 1, 111.

REFERENCES

395

[587] Johnson, D. A. In Handbook of Proteolytic Enzymes, Barrett, A. J., Rawlings, N. D., Woessner, J. F., Eds., Academic Press, San Diego, CA, 1998. [588] Pereira, P. J. B., Bergner, A., Macedo-Ribeiro, S., Huber, R., Matschiner, G., Friz, C. P. Nature 1998, 392, 306. [589] Schascheke, N., Matschiner, G., Zettl, F., Marquardt, U., Bergner, A., Bode, W., Sommerho¤, C. P., Moroder, L. Chem. Biol. 2001, 8, 313. [590] Schaschke, N., Domimik, A., Matschiner, G., Sommerho¤, C. P. Bioorg. Med. Chem. Lett. 2002, 12, 985. [591] Baker, B. R., Erickson, E. H. J. Med. Chem. 1967, 10, 1123. [592] Burgess, L. E., Newhouse, B. J., Ibrahim, P., Rizzi, J., Kashem, M. A., Hartman, A., Brandhuber, B. J., Wright, C. D., Thomson, D. S., Vigers, G. P. A., Koch, K. Proc. Natl. Acad. Sci. USA 1999, 96, 8348. [593] Goldberg, A. L., Stein, R., Adams, J. Chem. Biol. 1995, 2, 503. [594] Groll, M., Ditzel, L., Lo¨we, J., Stock, D., Bochtler, M., Bartunik, H. D., Huber, R. Nature 1997, 386, 463. [595] Loidl, G., Groll, M., Musiol, H.-J., Huber, R., Moroder, L. Proc. Natl. Acad. Sci. USA 1999, 96, 5418. [596] Loidl, G., Musiol, H.-J., Groll, M., Huber, R., Moroder, L. J. Pept. Sci. 2000, 6, 36. [597] Quest, A. F., Bell, R. M., J. Biol. Chem. 1991, 266, 18330. [598] Wender, P. A., Koehler, M. F. T., Wright, D. L., Irie, K. Synthesis 1999, 1401. [599] Sodeoka, M., Arai, M. A., Adachi, K., Uotsu, K., Shibasaki, M. J. Am. Chem. Soc. 1998, 120, 457. [600] Turner, M., Schweigho¤er, E., Colucci, F., Di Santo, J. P., Tybulewicz, V. L. Immunol. Today 2000, 21, 148. [601] Sada, K., Takano, T., Yanagi, S., Yamamura, H. J. Biochem. (Tokyo) 2001, 130, 177. [602] Chen, T., Repetto, B., Chizzonite, R., Pullar, C., Burghardt, C., Charm, E., Zhao, Z., Carroll, R., Nunes, P., Basu, M., Danho, W., Visnick, M., Kochan, J., Waugh, D., Gilfillan, A. M. J. Biol. Chem. 1996, 271, 25308. [603] Grucza, R. A., Bradshaw, J. M., Mitaxov, V., Waksman, G. Biochemistry 2000, 39, 10072. [604] Ottinger, E. A., Botfield, M. C., Shoelson, S. E. J. Biol. Chem. 1998, 273, 729. [605] Dekker, F. J., de Mol., N. J., Fischer, M. J. E., Liskamp, R. M. J. Bioorg. Med. Chem. Lett. 2003, 13, 1241. [606] Puius, Y. A., Zhao, Y., Sullivan, M., Lawrence, D. S., Almo, S. C., Zhang, Z.-Y. Proc. Natl. Acad. Sci. USA 1997, 94, 13420; Szczepakiewicz, B. G., Liu, G., Hadjuk, P. J., Abad-Zapatero, C., Pei, Z., Xin, Z., Lubben, T., Trevillyan, J. M., Stashko, M. A., Ballaron, S. J., Liang, H., Huang, F., Hutchins, C. W., Fesik, S. W., Jirousek, M. R. J. Am. Chem. Soc. 2003, 125, 4087; Liu, G., Xin, Z., Liang, H., Abad-Zapatero, C., Hajduk, P. J., Janowick, D. A., Szczpankiewicz, B. G., Pei, Z., Hutchins, C. W., Ballaron, S. J., Stashko, M. A., Lubben, T. H., Berg, C. E., Rondinone, C. M., Trevillyan, J. M., Jirousek, M. R. J. Med. Chem. 2003, 46, 3437. [607] Fauman, E. B., Saper, M. A. Trends Biochem. Sci. 1996, 21, 413.

396

REFERENCES

[608] Yuvaniyama, J., Denu, J. M., Dixon, J. E., Saper, M. A. Science 1996, 272, 1328. [609] Usui, T., Kojima, S., Kidokoro, S.-I., Ueda, K., Osada, H., Sodeoka, M. Chem. Biol. 2001, 153, 1. [610] Starkey, P. M., Barrett, A. J. Biochem. J. 1976, 155, 265. [611] Regan, J., McGarry, D., Bruno, J., Green, D., Newman, J., Hsu, C.-Y., Kline, J., Barton, J., Travis, J., Choi, Y. M., Volz, F., Pauls, H., Harrison, R., Zilberstein, A., Ben-Sasson, S. A., Chang, M. J. Med. Chem. 1997, 40, 3408. [612] For review, see: Linhardt, R. J. J. Med. Chem. 2003, 46, 1. [613] Driguez, P. A., Lederman, I., Strassel, J.-M., Herbert, J.-M., Petitou, M. J. Org. Chem. 1999, 64, 9512. [614] Van Boeckel, C. A. A., Petitou, M. Angew. Chem. Int. Ed. 1993, 32, 1671. [615] Westerduin, P., Basten, J. E. M., Broekhoven, M. A., de Kimpe, V., Kuijpers, W. H. A., van Boeckel, C. A. A. Angew. Chem. Int. Ed. 1996, 35, 331. [616] van Boeckel, C. A. A. Pure Appl. Chem. 1997, 69, 389. [617] Aleshin, A., Golubev, A., Firsov, L. M., Honzatko, R. B. J. Biol. Chem. 1992, 267, 19291. [618] Aleshin, A., Firsov, L. M., Honzatko, R. B. J. Biol. Chem. 1994, 269, 15631. [619] Payre, N., Cottaz, S., Boisset, C., Borsali, R., Svensson, B., Henrissat, B., Driguez, H. Angew. Chem. Int. Ed. 1999, 38, 974. [620] Sigurskjold, B. W., Christensen, T., Payre, N., Cottaz, S., Driguez, H., Svensson, B. Biochemistry 1998, 37, 10446. [621] Kim, Y. J., Ichikawa, M., Ichikawa, Y. J. Am. Chem. Soc. 1999, 121, 5829. [622] Hashimoto, H., Endo, T., Kajihara, Y. J. Org. Chem. 1997, 62, 1914. [623] Palcic, M. M., Heerze, L. D., Srivastava, O. P., Hindsgaul, O. J. Biol. Chem. 1989, 264, 17174. [624] Bruner, M., Horenstein, B. A. Biochemistry 2000, 39, 2261; Takashima, S., Tsuji, S., Tsujimoto, M. J. Biol. Chem. 2002, 277, 45719; Hinou, H., Sun, X.-L., Ito, Y. J. Org. Chem. 2003, 68, 5602. [625] Cushman, M., Mavandadi, F., Yang, D., Kugelbrey, K., Kis, K., Bacher, A. J. Org. Chem. 1999, 64, 4635; Schott, K., Kellermann, J., Lottspeich, F., Bacher, A. J. Biol. Chem. 1990, 265, 4204. [626] Daugherty, D. L., Gellman, S. H. J. Am. Chem. Soc. 1999, 121, 4325; Pallitto, M. M., Ghanta, J., Heinzelman, P., Kiessling, L. L., Murphy, R. M. Biochemistry 1999, 38, 3570; Zutshi, R., Brickner, M., Chmielewski, J. Curr. Opin. Chem. Biol. 1998, 2, 62. [627] Maletic, M., Wennermers, H., McDonad, D. O., Breslow, R., Still, W. C. Angew. Chem. Int. Ed. 1996, 35, 1490; Breslow, R., Yang, Z., Ching, R., Trojandt, G., Odobel, F. J. Am. Chem. Soc. 1998, 120, 3536; Breslow, R., Zhang, B. J. Am. Chem. Soc. 1992, 114, 5882. [628] Leung, D. K., Atkins, J. H., Breslow, R. Tetrahedron Lett. 2001, 42, 6255; Leung, D. K., Yang, Z., Breslow, R. Proc. Natl. Acad. Sci. USA 2000, 97, 5050. [629] Yan, J., Breslow, R. Tetrahedron Lett. 2000, 41, 2059; Wilson, D., Perlson, L., Breslow, R. Bioorg. Med. Chem. 2003, 11, 2649; Ruebner, A., Yang, Z., Leung, D., Breslow, R. Proc. Natl. Acad. Sci. USA 1999, 96, 14692.

REFERENCES

397

[630] Finnin, M. S., Donigian, J. B., Cohen, A., Richon, V. M., Rifkind, R. A., Marks, P. A., Breslow, R., Pavletich, N. P. Nature 1999, 401, 188; Richon, V. M., Webb, Y., Merger, R., Sheppard, T., Jursic, B., Ngo, L., Civoli, F., Breslow, R., Rifkind, R. A., Marks, P. A. Proc. Natl. Acad. Sci. USA 1996, 93, 5705. [631] Ann Boriack, P., Christianson, D. W., Kingery-Wood, J., Whitesides, G. M. J. Med. Chem. 1995, 38, 2286. [632] Turnheim, K., Kraupp, O. Eur. J. Pharmacol. 1971, 15, 231; Haddad, N., Xu, Y., Baron, J. A., Yee, N. K. Tetrahedron Lett. 2002, 43, 1135. [633] Verlander, M. S., Venter, J. C., Goodman, M., Kaplan, N. O., Saks, B. Proc. Natl. Acad. Sci. USA 1976, 73, 1009. [634] Kizuka, H., Hanson, R. N. J. Med. Chem. 1987, 30, 722. [635] Kierstead, R. W., Faraone, A., Mennona, F., Mullin, J., Guthrie, R. W., Crowley, H., Simko, B., Blaber, L. C. J. Med. Chem. 1983, 26, 1561. [636] Pitha, J., Zjawiony, J., Lefkowitz, R. J., Caron, M. G. Proc. Natl. Acad. Sci. USA 1980, 77, 2219. [637] Hardman, J. G., Limbird, L. E., Eds., Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 9th ed., McGraw-Hill, New York, 1996. [638] Bolognesi, M. L., Marucci, G., Angeli, P., Buccioni, M., Minarini, A., Rosini, M., Tumiatti, V., Melchiorre, C. J. Med. Chem. 2001, 44, 362. [639] Zheng, W., Lei, L., Lalchandani, S., Sun, G., Feller, D. R., Miller, D. D. Bioorg. Med. Chem. Lett. 2000, 10, 627. [640] Lord, J. A., Waterfield, A. A., Hughes, J., Kosterlitz, H. W. Nature 1977, 267, 495. [641] Milligan, G., Rees, S. Annu. Rep. Med. Chem. 2000, 35, 271. [642] Hazum, E., Chang, K.-J., Cuatrecasas, P. Science 1979, 206, 1077. [643] Portoghese, P. S. J. Med. Chem. 2001, 44, 2259. [644] Portoghese, P. S., Larson, D. L., Yim, C. B., Sayre, L. M., Ronsisvalle, G., Lipkowski, A. W. J. Med. Chem. 1985, 28, 1140. [645] Erez, M., Takemori, A. E., Portoghese, P. S. J. Med. Chem. 1982, 25, 847. [646] LeBoulluec, K. L., Mattson, R. J., Mahle, C. D., McGovern, R. T., Nowak, H. P., Gentile, A. J. Bioorg. Med. Chem. Lett. 1995, 5, 123. [647] Perez, M., Pauwels, P. J., Fourrier, C., Chopin, P., Valentin, J.-P., John, G. W., Marien, M., Halazay, S. Bioorg. Med. Chem. Lett. 1998, 8, 675. [648] Tamiz, A. P., Zhang, J., Zhang, M., Wang, C. Z., Johnson, K. M., Kozikowski, A. P. J. Am. Chem. Soc. 2000, 122, 5393. [649] Tamiz, A. P., Bandyopadhyay, B. C., Zhang, J., Flippen-Anderson, J. L., Zhang, M., Wang, C. Z., Johnson, K. M., Tella, S., Kozikowski, A. P. J. Med. Chem. 2001, 44, 1615. [650] Fandrick, K., Feng, X., Janowsky, A., Johnson, R., Cashman, J. R. Bioorg. Med. Chem. Lett. 2003, 13, 2151. [651] Rosenthal, S. J., Tomlinson, I., Adkins, E. M., Schroeter, S., Adams, S., Swafford, L., McBride, J., Wang, Y., DeFelice, L. J., Blakely, R. D. J. Am. Chem. Soc. 2002, 124, 4586. [652] Christopoulos, A., Grant, M. K. O., Ayoubzadeh, N., Kim, O. N., Sauerberg, P., Jeppesen, L., El-Fakahany, E. J. Pharmacol. Exp. Ther. 2001, 298, 1260.

398

REFERENCES

[653] Yang, Y.-K., Dickinson, C., Haskell-Luevano, C., Gantz, I. J. Biol. Chem. 1997, 272, 23000. [654] Carruthers, M. D., Lerner, M. R. Chem. Biol. 1996, 3, 537. [655] Conn, P. M., Rogers, D. C., Stewart, J. M., Niedel, J., She‰eld, T. Nature 1982, 296, 653. [656] Schertier, G. F. X., Villar, C., Henderson, R. Nature 1993, 362, 770. [657] Zhao, J., Kimura, S., Imanishi, Y. J. Pept. Sci. 1996, 2, 245. [658] Zagotta, W. N., Siegelbaum, S. A. Annu. Rev. Neurosci. 1996, 19, 235. [659] Ruiz, M., Karpen, J. W. Nature 1997, 389, 389. [660] Kramer, R. H., Karpen, J. W. Nature 1998, 395, 710. [661] Joseph, S. K., Lin, C., Pierson, S., Thomas, A. P., Maranto, A. R. J. Biol. Chem. 1995, 270, 23310. [662] Riley, A. M., Morris, S. A., Nerous, E. P., Correa, V., Potter, B. V. L., Taylor, C. W. J. Biol. Chem. 2002, 277, 40290. [663] Riley, A. M., Potter, B. V. L. Chem. Commun. 2000, 983. [664] Kaczorowski, G. J., Garcia, M. L. Curr. Opin. Chem. Biol. 1999, 3, 448. [665] Doyle, D. A., Cabral, J. M., Pfuetzner, R. A., Kuo, A., Gulbis, J. M., Cohen, S. L., Chait, B. T., MacKinnon, R. Science 1998, 280, 69. [666] Chen, J.-Q., Galanakis, D., Ganellin, C. R., Dunn, P. M., Jenkinson, D. H. J. Med. Chem. 2000, 43, 3478. [667] Rosa, J. C., Galanakis, D., Piergentili, A., Bhandari, K., Ganellin, C. R., Dunn, P. M., Jenkinson, D. H. J. Med. Chem. 2000, 43, 420. [668] Rosa, J. C., Galanakis, D., Ganellin, C. R., Dunn, P. M., Jenkinson, D. H. J. Med. Chem. 1998, 41, 2. [669] Samano, V., Ray, J. A., Thompson, J. B., Mook, R. A., Jr., Jung, D. K., Koble, C. S., Martin, M. T., Bigham, E. C., Regitz, C. S., Feldman, P. L., Boros, E. C. Org. Lett. 1999, 1, 1993. [670] Boros, E. E., Samano, V., Roy, J. A., Thompson, J. B., Jung, D. K., Kaldor, I., Kolbe, C. S., Martin, M. T., Styles, V. L., Mook, R. A., Feldman, P. L., Savarese, J. J., Belmont, M. R., Bigham, E. C., Boswell, G. E., Hashim, M. A., Patel, S. S., Wisowaty, J. C., Bowers, G. D., Mosleley, C. L., Walsh, J. S., Reese, M. J., Rutkowske, R. D., Sefler, A. M., Spitzer, T. D. J. Med. Chem. 2003, 46, 2502. [671] Arias, H. R. Biochim. Biophys. Acta 1998, 1376, 173. [672] Martinez, K. L., Corringer, P. J., Edelstein, S. J., Changeux, J. P., Me´rola, F. Biochemistry 2000, 39, 6979. [673] Sine, S. M., Quiram, P., Papanikolaou, F., Kreienkamp, H.-J., Taylor, P. J. Biol. Chem. 1994, 269, 8808. [674] Uhlmann, E., Peyman, A. Chem. Rev. 1990, 90, 543; Krugh, T. R. Curr. Opin. Struct. Biol. 1994, 4, 351. [675] Barcelo´, F., Capo´, D., Portugal, J. Nucleic Acids Res. 2002, 30, 4567. [676] Toshima, K., Ouchi, H., Okazaki, Y., Kano, T., Moriguchi, M., Asai, A., Matsumura, S. Angew. Chem. Int. Ed. 1997, 36, 2748. [677] Toshima, K., Takano, R., Maeda, Y., Suzuki, M., Asai, A., Matsumura, S. Angew. Chem. Int. Ed. 1999, 38, 3733.

REFERENCES

399

[678] Chaires, J. B. In Advances in DNA Sequence Specific Agents, Hurley, L. H., Chaires, J. B., Eds., JAI Press, Greenwich, CT, 1996, Vol. 2, pp. 141–167. [679] Chaires, J. B., Leng, F., Przewloka, T., Fokt, I., Ling, Y.-H., Perez-Soler, R., Priebe, W. J. Med. Chem. 1997, 40, 261. [680] He, H., Ding, W.-D., Bernan, V. S., Richardson, A. D., Ireland, C. M., Greenstein, M., Ellestad, G. A., Carter, G. T. J. Am. Chem. Soc. 2001, 123, 5362. [681] McMurray, C. T., Wilson, W. D., Douglass, J. O. Proc. Natl. Acad. Sci. USA 1991, 88, 666. [682] Wang, A. H. J., Ughetto, G., Quigley, G. J., Hakoshima, T., van der Marel, G. A., van Boom, J. H., Rich, A. Science 1984, 225, 1115; Lown, J. W., Hanstock, C. C. J. Am. Chem. Soc. 1982, 104, 3213; Dervan, P. B., Becker, M. M. J. Am. Chem. Soc. 1968, 100, 1968; Slama-Schwok, A., Teulade-Fichou, M.-P., Vigneron, J.-P., Taillandier, E., Lehn, J.-M. J. Am. Chem. Soc. 1995, 117, 6822. [683] Lokey, R. S., Kwok, Y., Guelev, V., Pursell, C. J., Hurley, L. H., Iverson, B. L. J. Am. Chem. Soc. 1997, 119, 7202. [684] Guelev, V., Sorey, S., Ho¤man, D. W., Iverson, B. L. J. Am. Chem. Soc. 2002, 124, 2864. [685] Gamage, S. A., Spicer, J. A., Finlay, G. J., Stewart, A. J., Charlton, P., Baguley, B. C., Denny, W. A. J. Med. Chem. 2001, 44, 1407. [686] Lee, D. M., Dunne, T. S., Siegel, M. M., Chang, C. C., Morton, G. O., Borders, D. B. J. Am. Chem. Soc. 1987, 104, 3464. [687] Gomez-Paloma, L., Smith, J. A., Chazin, W. J., Nicolaou, K. C. J. Am. Chem. Soc. 1994, 116, 3697. [688] Ho, S. N., Boyer, S. H., Schreiber, S. L., Danishefsky, S. J., Crabtree, G. R. Proc. Natl. Acad. Sci. USA 1994, 91, 9203. [689] Nicolaou, K. C., Smith, B. M., Ajito, K., Komatsu, H., Gomez-Paloma, L., Tor, Y. J. Am. Chem. Soc. 1996, 118, 2303. [690] Bifulco, G., Galeone, A., Gomez-Paloma, L., Nicolaou, K. C., Chazin, W. J. J. Am. Chem. Soc. 1996, 118, 8817. [691] Bifulco, G., Galeone, A., Nicolaou, K. C., Chazin, W. J., Gomez-Paloma, L. J. Am. Chem. Soc. 1998, 120, 7183. [692] Liu, C., Smith, B. M., Ajito, K., Komatsu, H., Gomez-Paloma, L., Li, T., Theodorakis, E., Nicolaou, K. C., Vogt, P. K. Proc. Natl. Sci. Acad. USA 1996, 93, 940. [693] Iyer, V. N., Szybalski, W. Science 1964, 145, 55. [694] Tomasz, M., Lipman, R., Chowdary, D., Pawlak, J., Verdine, G. L., Nakanishi, K. Science 1987, 235, 1204. [695] Na, Y., Li, V.-S., Nakanishi, Y., Bastow, K. F., Kohn, H. J. Med. Chem. 2001, 44, 3453. [696] Boger, D. L., Johnson, D. S. Angew. Chem. Int. Ed. 1996, 35, 1438. [697] Seaman, F. C., Hurley, L. H. J. Am. Chem. Soc. 1996, 118, 10052. [698] Seaman, F., Chu, J., Hurley, L. J. Am. Chem. Soc. 1996, 118, 5383. [699] Seaman, F. C., Hurley, L. Biochemistry 1993, 32, 12577. [700] Mitchell, M. A., Johnson, P. D., Williams, M. D., Aristo¤, P. A. J. Am. Chem. Soc. 1989, 111, 6428.

400

REFERENCES

[701] Mitchell, M. A., Kelly, R. C., Wienienski, N. A., Hatzenbuhler, N. T., Williams, M. G., Petzold, G. L., Slightom, J. L., Siemieniak, D. R. J. Am. Chem. Soc. 1991, 113, 8994. [702] Sun, D., Hurley, L. H. J. Am. Chem. Soc. 1993, 115, 5925. [703] Fukuda, Y., Seto, S., Furuta, H., Ebisu, H., Oomori, Y., Terashima, S. J. Med. Chem. 2001, 44, 1396. [704] Thompson, A. S., Sun, D., Hurley, L. H. J. Am. Chem. Soc. 1995, 117, 2371. [705] Thurston, D. E. In Molecular Aspects of Anticancer Drug–DNA Interactives, Macmillan, London, 1993, pp. 54–88. [706] Gregson, S. J., Howard, P. W., Hartley, J. A., Brooks, N. A., Adams, L. J., Jenkins, T. C., Kelland, L. R., Thurston, D. E. J. Med. Chem. 2001, 44, 737. [707] Kamal, A., Ramesh, G., Laxman, N., Ramulu, P., Srinivas, O., Neelima, K., Kondapi, A. K., Sreenu, V. B., Nagarajarma, H. A. J. Med. Chem. 2002, 45, 4679. [708] Zhou, Q., Duan, W., Simmons, D., Shayo, Y., Raymond, M. A., Dorr, R. T., Hurley, L. H. J. Am. Chem. Soc. 2001, 123, 4865. [709] Dervan, P. Science 1986, 232, 464. [710] Mrksich, M., Dervan, P. B. J. Am. Chem. Soc. 1993, 115, 2572. [711] Wade, W. S., Mrksich, M., Dervan, P. B. J. Am. Chem. Soc. 1992, 114, 8783. [712] Dwyer, T. J., Geirstanger, B. H., Mrksich, M., Dervan, P. B. J. Am. Chem. Soc. 1993, 115, 9900. [713] Mrksich, M., Dervan, P. B. J. Am. Chem. Soc. 1994, 116, 3663. [714] Olenyuk, B., Jitianu, C., Dervan, P. B. J. Am. Chem. Soc. 2003, 125, 4741. [715] Swalley, S. E., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1999, 121, 1113. [716] Trauger, J. W., Baird, E. E., Dervan, P. B. Angew. Chem. Int. Ed. 1998, 37, 1421. [717] de Clairac, R. P. L., Geierstanger, B. H., Mrksich, M., Dervan, P. B., Wemmer, D. E. J. Am. Chem. Soc. 1997, 119, 7909. [718] Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1996, 118, 6141. [719] White, S., Baird, E. E., Dervan, P. B. Chem. Biol. 1997, 4, 569. [720] Parks, M., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1996, 118, 6147. [721] Parks, M. F., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1996, 118, 6153. [722] Hawkins, C. A., de Clairac, R. P., Dominey, R., Baird, E. E., White, S., Dervan, P. B., Wemmer, D. E. J. Am. Chem. Soc. 2000, 122, 5235. [723] White, S., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1997, 119, 8756. [724] Woods, C. R., Ishii, T., Boger, D. L. J. Am. Chem. Soc. 2002, 124, 10676. [725] Urbach, A. R., Szewczyk, J. W., White, S., Turner, J. M., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1999, 121, 11621. [726] Floreancig, P. E., Swalley, S. E., Trauger, J. W., Dervan, P. B. J. Am. Chem. Soc. 2000, 122, 6342. [727] Trauger, J. W., Baird, E. E., Dervan, P. B. Chem. Biol. 1996, 3, 369. [728] Swalley, S. E., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1997, 119, 6953. [729] White, S., Turner, J. M., Szewczyk, J. W., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1999, 121, 260.

REFERENCES

401

[730] Herman, D. M., Turner, J. M., Baird, E. E., Dervan, P. B. J. Am. Chem. Soc. 1999, 121, 1121. [731] Weyermann, P., Dervan, P. B. J. Am. Chem. Soc. 2002, 124, 6872. [732] Baraldi, P. G., Balboni, G., Cacciari, B., Guiotto, A., Manfredini, S., Romagnoli, R., Spalluto, G., Thurston, D. E., Howard, P. W., Bianchi, N., Rutigliano, C., Mischiati, C., Gambari, R. J. Med. Chem. 1999, 42, 5131. [733] Baraldi, P. G., Balboni, G., Pavani, M. G., Spalluto, G., Tabrizi, M. A., De Clercq, E., Balzarini, J., Bando, T., Sugiyama, H., Romagnoli, R. J. Med. Chem. 2001, 44, 2536. [734] Toth, J. L., Price, C. A., Madsen, E. C., Handl, H. L., Hudson, S. J., Hubbard, R. B., III, Bowen, J. P., Kiakos, K., Hartley, J. A., Lee, M. Bioorg. Med. Chem. Lett. 2002, 12, 2245. [735] Tao, Z.-F., Saito, I., Sugiyama, H. J. Am. Chem. Soc. 2000, 122, 1602. [736] Wang, C., Dervan, P. B. J. Am. Chem. Soc. 2001, 123, 8657. [737] Mapp, A. K., Ansari, A. Z., Ptashne, M., Dervan, P. B. Proc. Natl. Acad. Sci. USA 2000, 97, 3930. [738] Arora, P. S., Ansari, A. Z., Best, T. P., Ptashne, M., Dervan, P. B. J. Am. Chem. Soc. 2002, 124, 13067. [739] Scha¨rer, O. D., Deng, L., Verdine, G. L. Curr. Opin. Chem. Biol. 1997, 1, 526. [740] Berthet, N., Constant, J.-F., Demeunynck, M., Michon, P., Lhomme, J. J. Med. Chem. 1997, 40, 3346. [741] Belmont, P., Jourdan, M., Demeunynck, M., Constant, J.-F., Garcia, J., Lhomme, J. J. Med. Chem. 1999, 42, 5153. [742] Alarcon, K., Demeunynck, M., Lhomme, J., Carrez, D., Croisy, A. Bioorg. Med. Chem. Lett. 2001, 11, 1855. [743] Nakatani, K., Sando, S., Kumasawa, H., Kikuchi, J., Saito, I. J. Am. Chem. Soc. 2001, 123, 12650. [744] Talanian, R. V., McKnight, C. J., Kim, P. S. Science 1990, 249, 769. [745] Cuenoud, B., Schepartz, A. Science 1993, 259, 510. [746] Cuenoud, B., Schepartz, A. Proc. Natl. Acad. Sci. USA 1993, 90, 1154. [747] Morii, T., Tanaka, T., Sato, S.-I., Hagihara, M., Aizawa, Y., Makino, K. J. Am. Chem. Soc. 2002, 124, 180. [748] Ueno, M., Murakami, M., Makino, K., Morii, T. J. Am. Chem. Soc. 1993, 115, 12575. [749] Morii, T., Shimomura, M., Morimoto, M., Saito, I. J. Am. Chem. Soc. 1993, 115, 1150. [750] Morii, T., Saimei, Y., Okagami, M., Makino, K., Sugiura, Y. J. Am. Chem. Soc. 1997, 119, 3649. [751] Caamanõ, A. M., Va´zquez, M. E., Martıńez-Costas, J., Castedo, L., Mascarenãs, J. L. Angew. Chem. Int. Ed. 2000, 39, 3104. [752] Hakomori, S., Murakami, W. T. Proc. Natl. Acad. Sci. USA 1968, 59, 254. [753] Mora, P. T., Brady, R. O., Bradley, R. M., McFarland, V. W. Proc. Natl. Acad. Sci. USA 1969, 63, 1290. [754] Hakomori, S.-I., Zhang, Y. Chem. Biol. 1997, 4, 97.

402

REFERENCES

[755] Ito, M., Suzuki, E., Naiki, M., Sendo, F., Arai, S. Int. J. Cancer 1984, 34, 689. [756] Cunto-Amesty, G., Luo, P., Monzavi-Karbassi, B., Lees, A., Kieber-Emmons, T. Vaccine 2001, 19, 2361. [757] Nicolaou, K. C., Caulfield, T., Kataoka, H., Kumazawa, T. J. Am. Chem. Soc. 1988, 110, 7910. [758] Bilodeau, M. T., Park, T. K., Hu, S., Randolph, J. T., Danishefsky, S. J., Livingston, P. O., Zhang, S. J. Am. Chem. Soc. 1995, 117, 7840. [759] Lassaletta, J. M., Carlsson, K., Garegg, P. J., Schmidt, R. R. J. Org. Chem. 1996, 61, 6973. [760] Ito, Y., Numata, M., Sugimoto, M., Ogawa, T. J. Am. Chem. Soc. 1989, 111, 8508. [761] Kameyama, A., Ishida, H., Kiso, M., Hasegawa, A. Carbohydr. Chem. 1991, 10, 549. [762] Park, T. K., Kim, I. J., Hu, S., Bilodeau, M. T., Randolph, J. T., Kwon, O., Danishefsky, S. J. J. Am. Chem. Soc. 1996, 118, 11488. [763] Slovin, S. F., Ragupathis, G., Adluri, S., Ungers, G., Terry, K., Kim, S., Spassova, M., Bornmann, W. G., Fazzari, M., Dantis, L., Olkiewicz, K., Lloyd, K. O., Livingston, P. O., Danishefsky, S. J., Scher, H. I. Proc. Natl. Acad. Sci. USA 1999, 96, 5710. [764] Ragupathi, G., Park, T. K., Zhang, S., Kim, I. J., Graber, L., Adluri, S., Lloyd, K. O., Danishefsky, S. J., Livingston, P. O. Angew. Chem. Int. Ed. 1997, 36, 125. [765] Wang, Z.-G., Williams, L. J., Zhang, X.-F., Zatorski, A., Kudryashov, V., Ragupathi, G., Spassova, M., Bornmann, W., Slovin, S. F., Scher, H. I., Livingston, P. O., Lloyd, K. O., Danishefsky, S. J. Proc. Natl. Acad. Sci. USA 2000, 97, 2719. [766] Hoon, D. S., Wang, Y., Sze, L., Kanda, H., Watanabe, T., Morrison, S. L., Morton, D. L., Irie, R. F. Cancer Res. 1993, 53, 5244. [767] Dohi, T., Nores, G., Hakomori, S. Cancer Res. 1988, 48, 5680. [768] Lee, K. J., Mao, S., Sun, C., Gao, C., Blixt, O., Arrues, S., Hom, L. G., Kaufmann, G. F., Ho¤man, T. Z., Coyle, A. R., Paulson, J., Felding-Habermann, B., Janda, K. J. Am. Chem. Soc. 2002, 124, 12439. [769] Delgado, M., Lee, K. J., Altobell, L., III, Spanka, C., Wentworth, P., Jr., Janda, K. J. Am. Chem. Soc. 2002, 124, 4946. [770] Tietze, L. F., Keim, H. Angew. Chem. Int. Ed. 1997, 36, 1615. [771] Hamilton, W. B., Helling, F., Lloyd, K. O., Livingston, P. O. Int. J. Cancer 1993, 53, 566. [772] Castro-Palomino, J. C., Ritter, G., Fortunato, S. R., Reinhardt, S., Old, L. J., Schmidt, R. R. Angew. Chem. Int. Ed. 1997, 36, 1998. [773] Stauch, T., Greilich, U., Schmidt, R. R. Liebigs Ann. 1995, 2101. [774] Carlstedt, I., Davies, J. R. Biochem. Soc. Trans. 1997, 25, 214. [775] Springer, G. F. Science 1984, 224, 1198. [776] Jennings, H. J., Ashton, F. E., Gamian, A., Michon, F. In Towards Better Carbohydrate Vaccines, Bell, R., Torrigiani, G., Eds., Wiley, London, 1987, pp. 307– 327. [777] Kuduk, S. D., Schwarz, J. B., Chen, X.-T., Glunz, P. W., Sames, D., Ragupathi, G., Livingston, P. O., Danishefsky, S. J. J. Am. Chem. Soc. 1998, 120, 12474.

REFERENCES

403

[778] Schwarz, J. B., Kuduk, S. D., Chen, X.-T., Sames, D., Glunz, P. W., Danishefsky, S. J. J. Am. Chem. Soc. 1999, 121, 2662. [779] Rittenhouse-Diakun, K., Xia, Z., Pickardt, D., Morey, P. S., Baek, M. G., Roy, R. Hybridoma 1998, 17, 165. [780] Roy, R., Baek, M.-G., Rittenhouse-Olson, K. J. Am. Chem. Soc. 2001, 123, 1809; Baek, M.-G., Roy, R. Bioorg. Med. Chem. 2001, 9, 3005. [781] Borman, S. Chem. Eng. News 2001, Mar. 19, 40. [782] Allen, J. R., Harris, C. R., Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 1890. [783] Williams, L. J., Harris, C. R., Glunz, P. W., Danishefsky, S. J. Tetrahedron Lett. 2000, 41, 9505. [784] Ragupathi, G., Coltart, D. M., Williams, L. J., Koide, F., Kagan, E., Allen, J., Harris, C., Glunz, P. W., Livingston, P. O., Danishefsky, S. J. Proc. Natl. Acad. Sci. USA 2002, 99, 13699. [785] Keil, S., Claus, C., Dippold, W., Kunz, H. Angew. Chem. Int. Ed. 2001, 40, 366. [786] Rother, R. P., Squinto, S. P. Cell 1996, 36, 185; Cooper, D. K. C., Koren, E., Oriol, R. Immunol. Rev. 1994, 141, 31. [787] Vetere, A., Donati, I., Campa, C., Semeraro, S., Gamini, A., Paoletti, S. Glycobiology 2002, 12, 283. [788] Liu, J., Weintraub, A., Holgersson, J. Xenotransplantation 2003, 10, 149. [789] Fang, J., Li, J., Zhang, Y., Wang, J., Guo, Z., Zhang, W., Yu, L., Brew, K., Wang, P. G. J. Am. Chem. Soc. 1998, 120, 6635. [790] Kozlov, I. A., Mao, S., Xu, Y., Huang, X., Lee, L., Sears, P. S., Gao, C., Coyle, A. R., Janda, K., Wong, C.-H. ChemBiochem 2001, 2, 741. [791] Lerner, R. A., Benkovic, S. J., Schultz, P. G. Science 1991, 252, 659. [792] Pollack, S. J., Jacobs, J. W., Schultz, P. G. Science 1986, 234, 1570. [793] Tramontano, A., Janda, K. D., Lerner, R. A. Science 1986, 234, 1566. [794] Yang, G., Chun, J., Arakawa-Uramoto, H., Wang, X., Gawinowicz, M. A., Zhao, K., Landry, D. W. J. Am. Chem. Soc. 1996, 118, 5881. [795] Matsushita, M., Ho¤man, T. Z., Ashley, J. A., Zhou, B., Wirsching, P., Janda, K. D. Bioorg. Med. Chem. Lett. 2001, 11, 87. [796] Shabat, D., Sinha, S. C., Reymond, J.-L., Keinan, E. Angew. Chem. Int. Ed. 1996, 35, 2628. [797] Yu, J., Hsieh, L. C., Kochersperger, L., Yonkovich, S., Stephans, J. C., Gallop, M. A., Schultz, P. G. Angew. Chem. Int. Ed. 1994, 33, 339. [798] Reymond, J.-L., Angew. Chem. Int. Ed. 1995 , 34, 2285. [799] Reymond, J.-L., Chen, Y. Tetrahedron Lett. 1995, 36, 2575. [800] Rosenblum, J. S., Lo, L.-C., Li, T., Janda, K. D., Lerner, R. A. Angew. Chem. Int. Ed. 1995, 34, 2275. [801] Li, T., Hilton, S., Janda, K. D., J. Am. Chem. Soc. 1995, 117, 3308. [802] Li, T., Janda, K. D., Ashley, J. A., Lerner, R. A. Science 1994, 264, 1289. [803] Jacobs, J. W., Schultz, P. G., Sugasawara, R., Powell, M. J. Am. Chem. Soc. 1987, 109, 2174. [804] Tramontano, A., Janda, K. D., Lerner, R. A. Proc. Natl. Acad. Sci. USA 1986, 83, 6736.

404 [805] [806] [807] [808] [809] [810] [811] [812] [813] [814] [815] [816] [817] [818] [819] [820] [821] [822] [823] [824] [825] [826] [827] [828] [829] [830] [831] [832] [833]

REFERENCES

Pollack, S. J., Nakayama, G. R., Schultz, P. G. Science 1988, 242, 1038. Baldwin, E., Schultz, P. G. Science 1989, 245, 1104. Janda, K. D., Benkovic, S. J., Lerner, R. A. Science 1989, 244, 437. Janda, K. D., Weinhouse, M. I., Schloeder, D. M., Lerner, R. A., Benkovic, S. J. J. Am. Chem. Soc. 1990, 112, 1274. Pollack, S. J., Hsiun, P., Schultz, P. G. J. Am. Chem. Soc. 1989, 111, 5961. Napper, A. D., Benkovic, S. J., Tramontano, A., Lerner, R. A. Science 1987, 238, 1041. Wirsching, P., Ashley, J. A., Benkovic, S. J., Lerner, R. A. Science 1991, 252, 680. Janda, K. D., Schloeder, D., Benkovic, S. J., Lerner, R. A. Science 1988, 241, 1188. Iverson, B. L., Lerner, R. A. Science 1990, 243, 1184. Benkovic, S. J., Napper, A. D., Lerner, R. A. Proc. Natl. Acad. Sci. USA 1988, 85, 5355. Janda, K. D., Lerner, R. A., Tramontano, A. J. Am. Chem. Soc. 1988, 110, 4835. Hilvert, D., Nared, K. D. J. Am. Chem. Soc. 1988, 110, 5593. Hilvert, D., Carpenter, S. H., Nared, K. D., Auditor, M.-T. M. Proc. Natl. Acad. Sci. USA 1988, 85, 4953. Jackson, D. Y., Jacobs, J. W., Sugasawa, R., Reich, S. H., Bartlett, P. A., Schultz, P. G. J. Am. Chem. Soc. 1988, 110, 4841. Hilvert, D. H., Hill, K. W., Nared, K. D., Auditor, M.-T. M. J. Am. Chem. Soc. 1989, 111, 9261. Braisted, A., Schultz, P. G. J. Am. Chem. Soc. 1990, 112, 7430. Balan, A., Doctor, B. P., Green, B. S., Torten, M., Zi¤er, H. Chem. Commun. 1988, 106. Cochran, A. G., Sugasawara, R., Schultz, P. G. J. Am. Chem. Soc. 1988, 110, 7888. Shokat, K. M., Leumann, C. J., Sugasawara, R., Schultz, P. G. Nature 1989, 338, 269. Jackson, D. Y., Schultz, P. G. J. Am. Chem. Soc. 1991, 113, 2319. Keinan, E., Sinha, S. C., Sinha-Bagchi, A., Benory, E., Ghozi, M. C., Eshhar, Z., Greene, B. S. Pure Appl. Chem. 1990, 62, 2013. Cochran, A. G., Schultz, P. G. J. Am. Chem. Soc. 1990, 112, 9414. Shokat, K. M., Leumann, C. J., Sugasawara, R., Schultz, P. G. Angew. Chem. Int. Ed. 1988, 27, 1172. Jeanloz, R. W., Codington, J. F. In Biological Roles of Sialic Acid, Rosenberg, A., Schengrund, C. L., Eds., Plenum Press, New York, 1976, pp. 201–238. Kayser, H., Zietler, R., Kannicht, C., Grunow, D., Nuck, R., Reutter, W. J. Biol. Chem. 1992, 267, 16934. Mahal, L. K., Charter, N. W., Angata, K., Fukuda, M., Koshland, D. E., Jr., Bertozzi, C. R. Science 2001, 294, 380. Borman, S. Chem. Eng. News 2001, Oct. 22, 56. Mahal, L. K., Bertozzi, C. R. Chem. Biol. 1997, 4, 415. Lemieux, G. A., Bertozzi, C. R. Chem. Biol. 2001, 8, 265.

REFERENCES

405

[834] Yarema, K. J., Mahal, L. K., Bruehl, R. E., Rodriguez, E. C., Bertozzi, C. R. J. Biol. Chem. 1998, 273, 31168. [835] Mahal, L. K., Yarema, K. J., Bertozzi, C. R. Science 1997, 276, 1125. [836] Lemieux, G. A., Yarema, K. J., Jacobs, C. L., Bertozzi, C. R. J. Am. Chem. Soc. 1999, 121, 4278. [837] Lee, J. H., Baker, T. J., Mahal, L. K., Zabner, J., Bertozzi, C. R., Wiemer, D. F., Welsh, M. J. J. Biol. Chem. 1999, 274, 21878. [838] Saxon, E., Bertozzi, C. Science 2000, 287, 2007. [839] Saxon, E., Luchansky, S. J., Hang, H. C., Yu, C., Lee, S. C., Bertozzi, C. R. J. Am. Chem. Soc. 2002, 124, 14893. [840] Rouhi, M. Chem. Eng. News 2000, Mar. 27, 26. [841] Srivastava, G., Kaur, K. J., Hindsgaul, O., Palcic, M. M. J. Biol. Chem. 1992, 267, 22356. [842] Gillitzer, E., Willits, D., Young, M., Douglas, T. Chem. Commun. 2002, 20, 2390. [843] Wang, Q., Chan, T. R., Hilgraf, R., Fokin, V. V., Sharpless, K. B., Finn, M. G. J. Am. Chem. Soc. 2003, 125, 3192. [844] Wang, Q., Kaltgrad, E., Lin, T., Johnson, J. E., Finn, M. G. Chem. Biol. 2002, 9, 805. [845] Wang, Q., Lin, T., Johnson, J. E., Finn, M. G. Chem. Biol. 2002, 9, 813. [846] Wender, P. A., Koehler, M. F. T., Wright, D. L., Irie, K. Synthesis 1999, 1401. [847] Larock, R. C. Comprehensive Organic Transformations, VCH, New York, 1989. [848] Dimick, S. M., Powell, S. C., McMahon, S. A., Moothoo, D. N., Naismith, J. H., Toone, E. J. J. Am. Chem. Soc. 1999, 121, 10286. [849] Ashton, P. R., Boyd, S. E., Brown, C. L., Jayaraman, N., Nepogodiev, S. A., Stoddart, J. F. Chem. Eur. J. 1996, 2, 1115. [850] Tao, Z.-F., Saito, I., Sugiyama, H. J. Am. Chem. Soc. 2000, 122, 1602. [851] Ashton, P. R., Hounsell, E. F., Jayaraman, N., Nilsen, T. M., Spencer, N., Stoddart, J. F., Young, M. J. Org. Chem. 1998, 63, 3429. [852] Ashton, P. R., Boyd, S. E., Brown, C. L., Jayaraman, N., Stoddart, J. F. Angew. Chem. Int. Ed. 1997, 36, 732. [853] Woods, C. R., Ishii, T., Boger, D. L. J. Am. Chem. Soc. 2002, 124, 10676. [854] Sprengard, U., Schudok, M., Schmidt, W., Kretzschmar, G., Kunz, H. Angew. Chem. Int. Ed. 1996, 35, 321. [855] Sprengard, U., Kretzschmar, G., Bartnik, E., Hu¨ls, C., Kunz, H. Angew. Chem. Int. Ed. 1995, 34, 990. [856] Patel, A., Lindhorst, T. K. J. Org. Chem. 2001, 66, 2674. [857] Kopytek, S. J., Standaert, R. F., Dyer, J. C. D., Hu, J. C. Chem. Biol. 2000, 7, 313. [858] Ben, R. N., Eniade, A. A., Hauer, L. Org. Lett. 1999, 1, 1759. [859] Bielfeldt, T., Peters, S., Meldal, M., Bock, K., Paulsen, H. Angew. Chem. Int. Ed. 1992, 31, 857. [860] Kuduk, S. D., Schwarz, J. B., Chen, X.-T., Glunz, P. W., Sames, D., Ragupathi, G., Livingston, P. O., Danishefsky, S. J. J. Am. Chem. Soc. 1998, 120, 12474.

406

REFERENCES

[861] Baraldi, P. G., Balboni, G., Cacciari, B., Guiotto, A., Manfredini, S., Romagnoli, R., Spalluto, G., Thurston, D. E., Howard, P. W., Bianchi, N., Rutigliano, C., Mischiati, C., Gambari, R. J. Med. Chem. 1999, 42, 5131. [862] Patch, R. J., Chen, H., Pandit, C. R. J. Org. Chem. 1997, 62, 1543. [863] Marra, A., Dondoni, A., Sansone, F. J. Org. Chem. 1996, 61, 5155. [864] Grice, P., Ley, S. V., Pietruszka, J., Priepke, H. W. Angew. Chem. Int. Ed. 1996, 35, 197. [865] Castro-Palomino, J., Ritter, G., Fortuanato, S. R., Reinhardt, S., Old, L. J., Schmidt, R. R. Angew. Chem. Int. Ed. 1998, 36, 1998. [866] Merritt, J. R., Naisang, E., Fraser-Reid, B. J. Org. Chem. 1994, 59, 4443. [867] Roy, R., Hernańdez-Mateo, F., Santoyo-Gonza´lez, F. J. Org. Chem. 2000, 65, 8743. [868] Payre, N., Cottaz, S., Boissert, C., Borsali, R., Svesson, B., Henrissat, B., Driguez, H. Angew. Chem. Int. Ed. 1999, 38, 974. [869] Knapp, S., Malolanarasimhan, K. Org. Lett. 1999, 1, 611. [870] Cochran, J. R., Stern, L. J. Chem. Biol. 2000, 7, 683. [871] Fulton, D. A., Stoddart, J. F. J. Org. Chem. 2001, 66, 8309. [872] Fulton, D. A., Stoddart, J. F. Org. Lett. 2000, 2, 1113. [873] Roy, R., Laferrie´re, C. A. Can. J. Chem. 1990, 68, 2045. [874] Danishefsky, S. J., Behar, V., Randolph, J. T., Lloyd, K. O. J. Am. Chem. Soc. 1995, 117, 5701. [875] Pozsgay, V. Angew. Chem. Int. Ed. 1998, 37, 138. [876] Koshida, S., Suda, Y., Arano, A., Sobel, M., Kusumoto, S. Tetrahedron Lett. 2001, 42, 1293. [877] Zhang, A. J., Khare, S., Gokulan, K., Linthicum, D. S., Burgess, K. Bioorg. Med. Chem. Lett. 2001, 11, 207. [878] Lindhorst, T. K., Kieburg, C. Angew. Chem. Int. Ed. 1996, 35, 1953. [879] Woller, E. K., Cloninger, M. J. Org. Lett. 2002, 4, 7. [880] Unverzagt, C., Andre´, S., Seifert, J., Kojima, S., Fink, C., Srikrishna, G., Freeze, H., Kayser, K., Gabius, H.-J. J. Med. Chem. 2002, 45, 478. [881] Lerchen, H.-G., Baumgarten, J., Piel, N., Kolb-Bachofen, V. Angew. Chem. Int. Ed. 1999, 38, 3681. [882] Rodriguez, E. C., Winans, K. A., King, D. S., Bertozzi, C. R. J. Am. Chem. Soc. 1997, 119, 9905. [883] Rodriguez, E. C., Marcaurelle, L. A., Bertozzi, C. R. J. Org. Chem. 1998, 63, 7134. [884] Liu, H., Brock, A., Wong, C.-H., Schultz, P. G. J. Am. Chem. Soc. 2003, 125, 1702. [885] Verlander, M. S., Venter, J. C., Goodman, M., Kaplan, N. O., Saks, B. Proc. Natl. Acad. Sci. USA 1976, 73, 1009. [886] Ugi, I. Angew. Chem. Int. Ed. 1982, 21, 810. [887] Bradley, H., Fitzpatrick, G., Glass, W. K., Kunz, H., Murphy, P. V. Org. Lett. 2001, 3, 2629. [888] Lockho¤, O. Angew. Chem. Int. Ed. 1998, 37, 3436.

REFERENCES

407

[889] Dondoni, A., Kleban, M., Marra, A., Banks, H. D. J. Org. Chem. 2002, 67, 4722. [890] Allen, J. R., Harris, C. R., Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 1890. [891] Cheng, Q., Oritani, T., Horiguchi, T., Yamada, T., Hassner, A. Tetrahedron 2000, 56, 1181. [892] Rao, S., He, L., Chakravarty, S., Ojima, I., Orr, G. A., Horwitz, S. B. J. Biol. Chem. 1999, 274, 37990. [893] Rao, S., Krauss, N. E., Heerding, J. M., Swindell, C. S., Ringel, I., Orr, G. A., Horwitz, S. B. J. Biol. Chem. 1994, 269, 3132. [894] Rao, S., Orr, G. A., Chaudhary, A. G., Kingston, D. G. I., Horwitz, S. B. J. Biol. Chem. 1995, 270, 20235. [895] Fazio, F., Bryan, M. C., Blixt, O., Paulson, J. C., Wong, C.-H. J. Am. Chem. Soc. 2002, 124, 14397. [896] Calvo-Flores, F. G., Isac-Garcıá, J., Hernańdez-Mateo, F., Pe´rez-Balderas, F., Calvo-Asıń, J. A., Sanche´z-Vaquero, E., Santoyo-Gonza´lez, F. Org. Lett. 2000, 2, 2499. [897] Pe´rez-Balderas, F., Ortega-Munõz, M., Morales-Sanfrutos, J., Hernańdez-Mateo, F., Calvo-Flores, F. G., Calvo-Asıń, J. A., Isac-Garcıá, J., Santoyo-Gonza´lez, F. Org. Lett. 2003, 5, 1951. [898] Ivin, K. J. Olefin Metathesis, Academic Press, London, 1983. [899] Descotes, G., Ramza, J., Basset, J.-M., Pagano, S. Tetrahedron Lett. 1994, 35, 7379. [900] Mortell, K. H., Gingras, M., Kiessling, L. L. J. Am. Chem. Soc. 1994, 116, 12053. [901] Owen, R. M., Gestwicki, J. E., Young, T., Kiessling, L. L. Org. Lett. 2002, 4, 2293. [902] Gan, Z., Roy, R. Tetrahedron 2000, 56, 1423. [903] Sestelo, J. P., Mourino, A., Sarandeses, L. A. Org. Lett. 1999, 1, 1005. [904] Umali, A. P., Simanek, E. E. Org. Lett. 2003, 5, 1245. [905] Aravind, S., Park, W. K. C., Brochu, S., Roy, R. Tetrahedron Lett. 1994, 35, 7739. [906] Mrksich, M., Whitesides, G. M. ACS Symp. 1997, 680, 361; Mrksich, M. Chem. Soc. Rev. 2000, 29, 267. [907] Noy, A., Vezenov, D. V., Kayyem, J. F., Meade, T. J., Lieber, C. M. Chem. Biol. 1997, 4, 519. [908] Smith, E. A., Thomas, W. D., Kiessling, L. L., Corn, R. M. J. Am. Chem. Soc. 2003, 125, 6140. [909] Merrifield, B. J. Am. Chem. Soc. 1963, 85, 2149. [910] Gravert, D. J., Janda, K. D. Curr. Opin. Chem. Biol. 1997, 1, 107. [911] Danishefsky, S. J., McClure, K. F., Randolf, J. T., Ruggeri, R. B. Science 1993, 260, 1307. [912] Plante, O. J., Palmacci, E. R., Seeberger, P. H. Science 2001, 291, 1523. [913] Nocolaou, K. C., Winssinger, N., Hughes, R., Smethurst, C., Cho, S. Y. Angew. Chem. Int. Ed. 2002, 39, 1084. [914] Debenham, S. D., Snyder, P. W., Toone, E. J. J. Org. Chem. 2003, 68, 5805. [915] Richman, D. D. Trends Microbiol. 1994, 2, 401.

408

REFERENCES

[916] Arabshahi, L., Schmitz, F. J. J. Org. Chem. 1987, 52, 3584. [917] Quinoa, E., Crews, P. Tetrahedron Lett. 1987, 28, 3229. [918] Erlanson, D. A., Lam, J. W., Wiesmann, C., Luong, T. N., Simmons, R. L., DeLano, W. L., Choong, I. C., Burdett, M. T., Flanagan, W. M., Lee, D., Gordon, E. M., O’Brien, T. Nat. Biotechnol. 2003, 21, 308. [919] Erlanson, D. A., Braisted, A. C., Raphael, D. R., Randal, M., Stroud, R. M., Gordon, E. M., Wells, J. A. Proc. Natl. Acad. Sci. USA 2000, 97, 9367. [920] Maly, D. J., Choong, I. C., Ellman, J. A. Proc. Natl. Acad. Sci. USA 2000, 97, 2419. [921] Borman, S. Chem. Eng. News 2000, May 15, 53. [922] Rostovtsev, V. V., Green, L. G., Fokin, V. V., Sharpless, K. B. Angew. Chem. Int. Ed. 2002, 41, 2596. [923] Demko, Z. P., Sharpless, K. B. Angew. Chem. Int. Ed. 2002, 41, 2110. [924] Borman, S. Chem. Eng. News 2002, Feb. 11, 29. [925] Nguyen, R., Huc, I. Angew. Chem. Int. Ed. 2001, 40, 1774. [926] Mosbach, L., Yu, Y., Andersch, J., Ye, L. J. Am. Chem. Soc. 2001, 123, 12420. [927] Otto, S., Furlan, R. L. E., Sanders, J. K. M. Drug Disc. Today 2002, 7, 117. [928] Lehn, J.-M., Eliseev, A. V. Science 2001, 291, 2331. [929] Lehn, J.-M. Chem. Eur. J. 1999, 37, 2828. [930] Cousins, C. R. L., Poulsen, S.-A., Sanders, J. K. M. Curr. Opin. Chem. Biol. 2000, 4, 270. [931] Otto, S., Furlan, R. L. E., Sanders, J. K. M. J. Am. Chem. Soc. 2000, 122, 12063. [932] Nazarpack-Kandlousy, N., Zweigenbaum, J., Henia, J., Elisev, A. V. J. Comb. Chem. 1999, 1, 199. [933] Huc, I., Lehn, J.-M. Proc. Natl. Acad. Sci. USA 1997, 94, 2106. [934] Goodwin, J. T., Lynn, D. G. J. Am. Chem. Soc. 1992, 114, 9197. [935] Bunyapaiboonsri, T., Ramstro¨m, O., Lohmann, S., Lehn, J.-M., Peng, L., Goeldner, M. ChemBiochem. 2001, 2, 438. [936] Goral, V., Nelen, M. I., Eliseev, A. V., Lehn, J.-M. Proc. Natl. Acad. Sci. USA 2001, 98, 1347. [937] Grubbs, R. H., Chang, S. Tetrahedron 1998, 54, 4413. [938] Ramstro¨m, O., Lehn, J.-M. ChemBiochem 2000, 1, 41. [939] Sakai, S., Shigemasa, Y., Sasaki, T. Tetrahedron Lett. 1997, 38, 8145.

INDEX

Abasic site, 303 Acarbose, 262 Acetohydroxamic acid, 253 Acetylcholine esterase, 21, 250, 333 receptor, muscarinic, 280 receptor, nicotinic, 285 N-Acetylgalactosamine Entamoeba histolytica lectin, 124 Gri¤onia simplicifolia lectin, 209 receptor, 162 Vicia villosa B4 lectin, 337 N-Acetylglucosamine cell wall, streptococcus, 158 GalTase, 183 mannose-binding protein, 169 spermatogenesis, 181 N-Acetylneuraminic acid, 23, 314 anti-NeuAc antibody, 314 hemagglutinin, 23 metabolic biosynthesis, 315 See also Sialic acid a1 -Acid glycoprotein, 219. See also Orosomucoid Aclacinomycin, 286 Acridine, 69 9-amino-, 69

tetrahydro-, 250 Acrosome, 182 Actinomycin D, 286 Adenine, 304 Adenylate cyclase, 141 Adhesin Bab A, 112 E. coli Fim H, 117 H. pyroli Hpa A, 115 H. pyroli SabA, 112 sialoadhesin, 179 Adrenergic receptor a-, 274 b-, 270 Adriamycinone, 16 A‰nity capillary electrophoresis, 86 chromatography, 140 constant, 9 Aggregation factor, 187 d-Ala-d-ala cell wall precursor, 79 divalent, 86 heterodivalent, 143 monovalent, 19 presented on liposome, 7 trivalent, 15


409

410

INDEX

d-Ala-d-lac cell wall, vancomycin-resistant, 81, 96 monovalent, 84 presented on liposome, 81 Alprenolol, 273 Amikacin, 154 Aminoglycoside, 69 divalent, 152 neamine, 70, 152 neomycin, 69 2-Amino-1,8-naphthyridine, 304 Amphotericin B, 160 Anthracycline, 16 Anthraquinone, 286 Anthrax toxin, 141 Antibody anticancer, 306 anti-a-Gal, 313 catalytic, 314 IgG, 2 IgM, 111, 235, 308 Antigen(s) immunoglobulin, 235 tumor-associated, 306 xenoactive, 314 Antithrombin, 262 AP1510, 239 Apoptosis, mediated by caspase, 239 Fas, 239 galectin, 177 Arg-gly-asp, 205, 231 Aryl-a-ketocarboxylic acid, 147 Aryloxypropanolamine, 272 Asialoglycoprotein receptor, 4, 164 Autolysin, 101 Avidity, 9 Avidin biotin interaction, 11, 88 tetravalent system, 176 Azacrown CXCR4 antagonist, 59 galactoside-presenting, 127 sca¤old, 4 See also Sca¤old Azo linkage, 272 AZT, 51 B cell, 77, 236 BabA, 112. See also Adhesin Basic leucine zipper protein, 304 Batracylin, 84 Benzamidine, 253 Bicyclam, 59

Bifunctional ligand binding to av b3 , 234 binding to Con A, 211 definition, 8 Bindin, 182 Binding enthalpy of, 13 entropy of, 13 Biotin avidin interaction, 11, 88 presented on cell surface, 317 tethered to polymer, 226 tetravalent construct, 176 See also Avidin Bis(methidium)spermine, 289 Bovine serum albumin, presentation of GalNAc, 125, 162 GlcNAc, 188 GM3 , 309 mannose, 172 Mycobacterium vaccine, 155 sca¤old, 6 sialic acid, 47, 191 See also Sca¤old Butyrylcholineesterase, 250 Calcineurin, 238 Calcitriol, 250 Calicheamicin, 291 Calix[n]arene bifunctional, 211 sca¤old, 5 See also Sca¤old Camptothecin, 300 Cap saccharide, 157 Capsid protein, 65, 325 Carbohydrate-recognition domain Dioclea grandiflora, 223 galectin, 177 mannose-binding protein, 169 Carbonic anhydrase, 270, 334 5-Carboxamidoindole, 278 Carboxymethylated dextran neoglycoprotein attached to, 188 sialic acid presenting, 36 See also Sca¤old Carboxypeptidase, 143 Carzinophilin A, 290 Caspase-3, 239, 331 CC 1065, 293 CCR5, 53, 59 CD4, 33, 55 CD14, 228 Cefuroxime, 143

INDEX Cell adhesion molecule endothelial leukocyte, 188 glycosylated, 198 intercellular, 180 neural, 316 Cellobiose, 60 Cell surface antigen, 320 engineering, 315 Cell wall, 79. See also Peptidoglycan Cephalosporin, 143 Ceramide galactosyl, 55 glucosyl, 56 lactosyl, 185 Chartreusin, 286 Chemoreceptor, 120 Chemotaxis, 120 Cholic acid, 108. See also Sca¤old Chromosorb P, 128 Ciprofloxacin, 147. See also Fluroquinolone Citrate synthase, 267 Click chemistry, 333 Cluster e¤ect, 21 Combinatorial synthesis, 325 dynamic, 334 structure-biased, 325 target-accelerated, 336 target-guided, 330 Complement component, 110, 236 Complex binary, 237 ternary, 237 Concanavalin A, 15, 209, 211. See also Lectin Conocurvone, 74 Cooperativity choleratoxin B binding, 130 definition, 9 trivalent vancomycin, 10, 90 Copolymerization acrylamide, 36, 139 acrylic acid, 37 norborene, 217 vinyl amide, 323 See also Polymerization CXCR4, 53, 59 Cyanovirin-N, 60, 62 Cyclic GMP, 19, 282 Cyclic nucleotide–gated channel, 19, 282. See also Ion channel Cyclodextrin dimer, 269 divalent tryptase inhibitor, 254

glucoside presented in, 211 mannoside presented in, 211 multivalent anion, 57, 68 sca¤old, 4 substrate of glucoamylase, 262 See also Sca¤old Cyclophilin, 240. See also Immunophilin Cyclopropylpyrroloindole, 293 Cyclosporin A, 240 dimer, 240 Cytokine dimerization, induced by, 246 receptor, 246 Daunorubicin, 17, 287 DC-81, 294 DC-SIGN, 180 Dendrimer galabioside-presenting, 117 lactoside-presenting, 133, 177 Lewis saccharide–presenting, 181 mannoside-presenting, 211 sca¤old, 6 sialoside-presenting, 27, 219 See also Sca¤old Dequalinium, 125, 284 Diaryloxamic acid, 260 Diazobenzofluorene, 288 Diazo coupling, 6, 272 Diels–Alder reaction, 218, 233 Dihydrofolate reductase, 148, 244 Dihydropterin pyrophosphokinase, 146 Dimerization chemical inducer of, 21, 236 cytokine induced, 246 enzyme, 267 of glycopeptide antibiotics, 82 of HIV protease, 49 of mannose-6-phosphate receptor, 174 of opioid recepor, 275 See also Oligomerization Dipolar 1,3-cycloaddition, 229 Distamycin, 295 Disulfide aminoglycoside dimer, 152 bZIP peptide dimer, 306 dynamic combinatorial synthesis, 334 linkage, 323 a-MSH peptide, 281 psammapline A, 147, 329 vancomycin dimer, 95 See also Linker DNA gyrase, 147 Duocarmycin, 293

411

412

INDEX

Dynamic light scattering, 130 Dynemicin A, 291 E. coli adhesion, 117 dihydrofolate reductase, isolated from, 148 penicillin binding protein, isolated from, 141 susceptibility of, 101 synergism in, 108 uropathogenic, 117 Ectodomain, 77 Edema factor, 141. See also Anthrax toxin E¤ective molarity, 19, 86 E¿ux pump, 125 multidrug resistance, 125 Endocytosis, mediated by anthrax toxin, 141 glycodendrimer, 177 neoglycopolymer, 167 neoglycoprotein, 184 Endothelial cell av b3 -expressing, 234 leukocyte adhesion to, 2, 188, 203 Enediyne antibiotic, 291 Enhancement factor definition, 10 in polyvalent system, 139 in trivalent system, 11 Enterococci vancomycin-resistant, 84, 95 vancomycin-susceptible, 84, 93 Enzyme-linked immunosorbent assay, 39, 127 Enzyme-linked lectin assay, 211 Epidermal growth factor receptor, 247 Epithelial cell bacterial adhesion, 112 infected by HIV-1, 57 virus adhesion, 63 Eremomycin, 81, 96 chloro-, 79, 81 Erythrocyte adhesion to influenza A/B, 23, 36 adhesion to influenza C, 45 adhesion to polyoma virus, 64 Con A–mediated agglutination, 215 hematopoiesis, 246 Erythromycin, 108 Erythropoietin, 247 Espramicin, 291 Estradiol, 236 Estrogen receptor, 236 Ethidium, 125, 290 Extracellular matrix, 231

Fas receptor, 239 Fertilin b, 183 Fetuin, 46 asialo-, 125 Fibroblast growth factor, 186 Fibronectin, 231 Fimbriae, 112 FK506 binding protein, 238, 330 dimer, 21, 238 mimic, 239 Fluorescent resonance energy transfer, 136 Fluoroquinolone, 148 ciprofloxacin, 149 norfloxacin, 149 Folate cofactor, 148 Fucose-binding receptor, 184 Fucosyltransferase, 266. See also Glycosyltransferase Functional group, synthesis of amide, 321 amine, 322 carbamate, 322 ester, 321 ether, 321 oxime, 322 sulfonamide, 322 thioether, 321 urea, 322 Galabioside, 117 Galactose-sensing receptor, 21 Galactoside asialoglycoprotein, 164 cholera toxin, 129 galactosyl transferase, 264 glucose–galactose binding protein, 120 heat-labile enterotoxin, 127 lectin, 218 sca¤old, 26 Galactosyltransferase, 183, 264. See also Glycosyltransferase Galectin, 177 Gal-a1,3-Gal, 314 Ganglioside, types and interaction G2 , 31 Gb3 , 136 GD1a , 31 Gg3 –GM3 interaction, 186 globo-H, 307 GM1 , 127, 133 GM2 , 310 GM3 , 42, 185, 306 Lea , 112, 181

INDEX Leb , 112, 181 Ley –Lex , 306 sialyl-, 112 Glucoamylase, 262 Glucose–galactose binding protein, 121 Glucoside, ligand of Con A, 215 glucose–galactose binding protein, 121 Glycoconjugate, 22, 184 Glycodendrimer, 177, 219, 312 Glycolipid Gal cer–terminated, 55 Gb3 cer–terminated, 136 sialic acid–terminated, 31 Glycoporin, 31. See also Ganglioside, GD1a Glycoprotein gp41, 78 gp120, 53, 60 jelly coat, 182 Glycosphingolipid, 56, 112, 185, 306. See also Ganglioside Glycosylphosphatidylinositol, 156, 228 Glycosyltransferase, 264 Godanotropin-releasing hormone receptor, 281 Gold nanoparticle, presenting Lex carbohydrate, 188 mannoside, 119 streptavidin, 217 See also Sca¤old G-protein-coupled receptor, 59, 270 Granulocyte-colony-stimulating factor, 247 Grubb’s catalyst, 98, 322 Guanine–guanine mismatch, 304 Hakomori MBr1 , 307. See also Ganglioside Hemagglutination, induced by Lectin, 125, 177 Virus, 23 Hemagglutinin H. pylori, 112 influenza, 21, 23 S. suis, 117 Hemoglobin, 10 Heparan sulfate, 66, 263 Heparin fibroblast growth factor, 186 inhibition of dengue virus, 67 selectin binding, 201 surface anchoring, 252 thrombin, 262 Heterovalency, 7. See also Homovalency Hexoprenaline, 271 Histone deacetylase, 269 Homovalency, 7. See also Heterovalency

H. pylori, 112 Huisgen’s [3 þ 2] cycloaddition, 320 Human serum albumin, presentation of H-1, 113 Leb , 113 O-specific polysaccharide, 155 sca¤old, 6 sialic acid, 45 See also Sca¤old Huperzine A, 252 Hydrodynamic radius, 130 Hydroxyrubicin, 16 5-Hydroxytryptamine, 277 presented on nanocrystal, 280 receptor, 277 See also Serotonin Immunoglobulin, 2, 111, 235, 306. See also Antibody Immunophilin, 238 Immunosuppressant, 238 Influenza virus, 4, 21, 327 b inhibitor, 47 g inhibitor, 46 Inositol-1,4,5-triphosphate, 283 Insulinlike growth factor, 174 Integrin, 205, 233 Intercalator, 286 Interleukin, 247 Ion channel, 282 Isoproterenol, 272 Isothermal titration calorimetry, 88 Kallikrein proteinase, 334 Kanamycin, 154 Kedarcin, 291 Keyhole limpet hemocyanin, 157, 307, 310 Kinase extracellular regulated, 261 Fyn, 242, 333 JAK tyrosine, 246 Lyn, 333 (protein) tyrosine, 258 Src-like tyrosine, 239 Syk protein tyrosine, 259 b-Lactam, 143. See also Penicillin Lactate dehydrogenase, 267 Lactobionic acid, 68 Lactoside, 133, 170, 177, 200, 226 Langmuir–Blodgett, 35 Lectin Arachis hypogaea, 219 Bauhinia purpurea, 218

413

414

INDEX

Lectin (Continued ) Con A, 209 Dioclea grandiflora, 223 Entamoeba histolytica, 124 Glycine max, 209 Gri¤onia simplicifolia, 209 isolectin, 209 Limax flavus, 219 P. tetragonolobus, 225 Piscum sativum, 222 Ricinus communis agglutinin, 225 wheat germ agglutinin, 220 XL35, 183 Leishmaniasis, 156 Leukocyte elastase, 261 Lewis x, 181, 201 Lex –Lex interaction, 188 See also Ganglioside Ligand density definition, 7 of polyvalent polymer, 36, 217 Linker alkane, 93, 256 binding contribution, 14 definition, 3 di-sulfide, 95 peptide, 137 steric e¤ect, 19 Lipid A, 108 Lipid II, 105 Lipophosphoglycan, 156 Lipopolysaccharide, 110. See also Lipid A Liposome, presentation of d-ala-d-ala, 80 Cap polysialic acid, 157 galactoside, 226 GM1 , 136 GM3 –Lac ceramide, 185 peptide, fertilin , 183 sca¤old, 29, 325 sialic acid, 29, 32 sLex , 192, 206 See also Sca¤old Low-density lipoprotein, 167 Lumazine, 267 Lutropin, 162 a2 -Macroglobulin, 46. See also Influenza virus, g-inhibitor Macrophage, 142, 162 Major groove, 290 Major histocompatibility complex, 229, 253 Mannosamine N-acetyl, 315

N-propanoyl, 315 Mannose bacterial adhesion, 117 binding protein, 169 binding to Con A, 15, 209 DC-SIGN receptor, 180 Dioclea grandiflora lectin, 223 Mannose/N-acetylgalactosamine-4-SO4 receptor, 162 Mannose 6-phosphate receptor, 174 Matrix metalloproteinase, 252 Melanocortin-1 receptor, 281 a-Melanocyte stimulating hormone, 281 Melibiose, 184 Metabolic biosynthesis, 315 Metalloproteinase, 234 Metastasis, 185, 306 Methicilin, 148 Methotrexate, 148, 245 Michael 1,4-addition, 322 Michellamine, 74 Microtiter well, 227 Minor groove binder, 16, 71, 291 Mitomycin, 292 Mitsunobu coupling, 321 Moenomycin, 99 Monoamine transporter, 278 dopamine, 278 norepinephrine, 278 serotonin, 278 M2 protein, 77 Mucin, 46, 310 Multimerization, 321 Multivalency, 1 Muramyl dipeptide, 228 NAD synthetase, 151, 328 Naphthalimide, 290 Naphthyridine, 304 Neamine, 69, 152. See also Aminoglycoside Neocarzinostatin, 291 Neoglycoconjugate, 185. See also Neoglycopolymer; Neoglycoprotein Neoglycopolymer, presentation of N-acetylglucosamine, 172 galactoside, 121 glucoside, 215 mannoside, 217 sLea , 200 sLex , 201 Neoglycoprotein, presentation of N-acetylgalactosamine, 125 CD4 peptide, 55 fucoside, 184

INDEX galactoside, 125 H-1, 112 Leb , 112 mannoside, 172, 180 O-specific polysaccharide, 155 sialic acid, 45, 327 sLex , 191 TF , 312 Neomycin B, 69, 153 Netropsin, 295 Neuraminic acid N-acetyl, 23 N-azidoacetyl, 319 N-butanoyl, 316 N-levulinoyl, 316 N-propanoyl, 316 See also Sialic acid Neuraminidase, 37, 47, 150. See also Sialidase Neurotrophin, 249 Neutrophil, 2, 188 Nicotinic acetylcholine receptor, 285. See also Ion channel Novobiocin, 109 Nuclear receptor, 236, 250 Nucleic acid, 68, 152, 286 Olefin metathesis, preparation of glucoside dimer, 322 vancomycin dimer, 96, 322, 336 Oligoglycine, dimerzation of oxymorphone, 276 sialic acid, 23, 64 See also Linker Oligolysine, presentation of N-acetylglucosamine, 220 Lewis saccharide, 179 mannoside, 214 sialic acid, 27 See also Sca¤old Oligomannose, 60 Oligomerization of amphotericin B, 160 of DC-SIGN receptor, 180 of enzyme, 267 of vancomycin dimer, 82 See also Dimerization Oligoproline, 301 Olivomycin A, 286 OPTCOL, 40 Opioid receptor, 275 Optical tweezers, 119 Orientation of aryl-a-keto acid dimer, 147 binding site, 223

415

lactoside dimer, 200 quinnuclidine dimer, 280 sialic acid dimer, 190 tetronic acid, 261 vancomycin dimer, 93 Orosomucoid, 219 asialo-, 164 Oseltamivir, 47 Oxymorphamine, 276 PA63 , 141 Penicillin, 143 penicillin-binding protein, 102, 105, 143 Pentamannose phosphate, 176 Peptabody, 236 Peptide nucleic acid, 169 Peptidoglycan, 80, 143, 229. See also Cell wall Peptoid, 6 Phage display anti-GM3 antibody, 309 anti-NeuAc antibody, 314 P1 peptide, 141 sLex receptor, 198 Phorbol, 258 Phosphophenol, 259 Photoa‰nity labeling, 112, 322 Pk trisaccharide, 12, 136 Pluramycin, 287 Pocket factor, 65 Poly(acrylamide), presentation of lactoside, 225 Lex , 203 mannoside, 119, 172 muramyl peptide, 228 Pk trisaccharide, 140 sialic acid, 6, 36, 44, 323 sLea , 195, 203 See also Sca¤old Poly(acrylic acid), presentation of sialic acid, 37, 41, 327 sLea , 205 See also Sca¤old Poly(N-acryloyloxysuccinimide), preparation of poly(acrylamide) preparation and use, 40, 323 P1 peptide–presenting, 141 sialic acid–presenting, 40 Polyamide, 295 conjugate, 299 H-shaped, 73, 296 hairpin-shaped, 71, 297, 299 ring-shaped, 299 tetravalent, 299

416

INDEX

Polyamidoamine, presentation of mannoside, 213 sialic acid, 29, 220 Poly(diacetylene), 195, 208 Langmuir–Blodgett film, 35 liposome, 33, 133, 195 Polydispersity, 38, 323 Poly(ethyleneglycol), dimerization of cGMP, 19, 282 EPO-mimetic peptide, 247 inositol triphosphate, 283 sialic acid, 23 sLex , 191 tryptase inhibitor, 257 See also Linker Poly(glutamic acid), 42, 48 Poly[N-(2-hydroxypropyl)methacrylamide], 169 Polylactosamine, 203 Polylysine, 191 Polymerization, 323 acrylamide, 37, 45, 139, 323 acrylic anhydride, 327 N-acryloyloxysuccinimide, 323 Langmuir–Blodgett film, 35 liposome, 32, 133, 195 norbornene, 217 7-oxanorbornene, 215 ring-opening metathesis, 98, 323 styrene, 154 See also Copolymerization Polymyxin B, 108 Poly(norbornene), 217, 231 poly(7-oxanorbornene), 215 Polytacticity, 38 Polystyrene, 187, 232 Polyvalency, 1 Postmodification, 38, 323 Potassium channel, 284. See also Ion channel Practolol, 272 Prazocin, 274 Prodrug, 3 Proinflammatory mediator, 228 Propranolol, 272 Protease, 49 serine, 253, 261 Proteasome, 256 Protective antigen, 141 Protein trans-splicing, 245 Protein tyrosine phosphatase, 147, 259 Psammaplin A, 147, 329 Pyrophosphokinase, 146 Pyrrolo[2,1-c][1,4]benzodiazepine, 294

QacR, 125 Quinic acid, 173 Quinnuclidine, 280 Quinoxaline, 289 Rapamycin, 237, 245 Rate constant, 18 Receptor clustering, 21, 123 Reductive amination, 47, 161 Rev regulatory protein, 69 Rev response element, 69 Reverse transcriptase, 51 non-nucleoside-based inhibitor, 51 nucleotide-based inhibitor, 51 Rhamnopyranoside, 158 Riboflavin synthase, 267 Ribonuclease S, 70 Ribosomal RNA, 152. See also Nucleic acid Ricinus communis agglutinin, 225. See also Lectin Ring-opening metathesis polymerization, 323 polyvalent glucoside, 215 polyvalent RGD peptide, 231 vancomycin dimer, 96 See also Polymerization SabA, 112. See also Adhesin Sca¤old albumin, 55, 159 azacrown, 4, 61, 128 benzene, 51, 88, 145, 148 cholic acid, 108 dendrimer, 5, 28, 133, 212, 221 diazobenzene, 145, 305 galactoside, 26, 189 nanocrystal, 280 poly(ethyleneglycol), 19, 165, 191, 257 polymer, 36, 58, 139, 207, 336 self-assembled monolayer, 4, 87, 218, 233 Selectin E-selectin, 188 L-selectin, 198 P-selectin, 204 P-selectin glycoprotein ligand, 204 Self-assembled monolayer, presentation of d-ala-d-ala, 87 Gal-GalNAc, 219 mananoside, 118, 218 RGD peptide, 233 sca¤old, 4, 7, 325 See also Sca¤old

INDEX Self-assembly, 161, 325 Sensor colorimetric, 34 optical, 110 Serine-recognition chemoreceptor, 123 Serotonin receptor, 277 reuptake inhibitor, 278 transporter, 278 See also 5-Hydroxytryptamine Sertoli cell, 181 SH2 domain, 241 Shikimic acid, 173 Sialic acid antibody, 314 capsular polysialic acid, 157 influenza, 21, 23, 327 Limax flavus lectin, 219 poly(a2,8-sialic acid), 157, 316 See also N-acetylneuraminic acid Sialidase, 47, 150 9-O-acetylsialic acid esterase, 45 Sialyllactose, 115, 130, 151 Sialyl Lewis a, 188, 205 Sialyl Lewis x E-selectin, 188 H. pylori SabA, 112 L-selectin, 198 P-selectin, 204 6-sulfo, 201 Sialyltransferase, 266 Silicon microcavity resonator, 110 Single nucleotide polymorphism, 304 Single-walled carbon nanotube, 77 Solid-phase synthesis, 28, 103, 325 Spacer, 3. See also Linker O-specific polysaccharide, 155 S-peptide, 71 Spermatogenesis, 181 Sperm–egg interaction, 182 Squarate, 322 Staudinger reaction, 318, 321 Steric e¤ect, 19, 261 Streptococus suis, 117 Stromelysin, 252 Succinyldicholine, 285 Sumatriptan, 278 Suramin, 68 Surface plasma resonance, 36, 87, 92, 135, 186, 218 Synergism in E. coli, 104, 108 in influenza virus, 40

TAR–Tat complex, 70 Taxol, 322 T cell receptor, 177, 230 Telomerization, 58 Templated syntheis, 334 TentaGel, 218. See also Polymerization Teretifolin B, 74 Tetronic acid, 261 TF disaccharide, 310, 312 Thrombin, 261 Thyrotropin, 162 TNF-a, 229 TN monosaccharide, 310 Tobramycin, 153 Tomaymycin, 294 Topoisomerase, 148, 290 Toxin, bacterial anthrax, 141 cholera, 129 endotoxin, 110 heat-labile, 126 Shiga, 12, 136 verotoxin, 141 Transactivator protein, 68 transactivator-response RNA element, 68 Transcription factor, 72 Transglycosylase, 80, 99 Transition metal–based chemistry, 337 Transpeptidase, 80, 99, 143 transpeptidation, 80, 99 Triazole, 229, 333 Trichostatin A, 270 Triostin A, 290 Trypsin, 255 Tryptase, 253 Tyrosine kinase receptor, 249 Ugi reaction, 322 Vaccinia VH1–related phosphatase, 261 Vaccine, 77 Mycobacterium, 153 Shigella, 155 Valency, 1 Vancomycin, 79 catalytic, 104 dimer, 84, 91, 336 dimerization, 82 heterovalent, 99 mimic, 103 polymer, 97 trimer, 10, 15, 19, 84

417

418

INDEX

Viomellein, 74 Virus cowpea mosaic, 320 dengue, 67 foot-and-mouth disease, 78 hepatovirus, 65 human immunodeficiency, 49, 60 picorna, 65, 325 polyoma, 64 respiratory syncytial, 59, 66 rhinovirus, 65 rotavirus, 63 sendai, 31 simian immunodeficiency, 60

Vitamin D3 , 250 Wang resin, 28, 220 Wheat germ agglutinin, 220. See also Lectin Williamson ether synthesis, 321 Wittig olefination, 322 Xenotransplantation, 313 Yohimbine, 275 Zanamivir, 47 Zinc-dependent enzyme, 269 Zona pellulcida, 18

Synthetic Multivalent Molecules: Concepts and Biomedical Applications

Synthetic instruments: concepts and applications

Synthetic Instruments Concepts and Applications

Multivalent functions

Biomedical Data and Applications

Biomedical Applications

Stereochemistry of Radical Reactions: Concepts, Guidelines, and Synthetic Applications

Biomedical Applications

Cellular and Biomolecular Recognition: Synthetic and non-Biological Molecules

Calculus: Concepts and Applications

Proteomics: Biomedical and Pharmaceutical Applications

Calculus: Concepts and applications

Biology: Concepts and Applications

Biomedical Applications of Nanotechnology

Biomedical Applications Polymer Blends

Polyphosphazenes for Biomedical Applications

Chromium-VI Reagents: Synthetic Applications

Basic concepts of synthetic differential geometry

Internal Flow: Concepts and Applications

Pharmaceutical Biotechnology. Concepts and Applications

Environmental biotechnology: concepts and applications

Regression Analysis: Concepts and Applications

Understanding Hypertext Concepts and Applications

Pharmaceutical Biotechnology. Concepts and Applications

IP Multicasting: Concepts and Applications

Clinical Pharmacokinetics: Concepts and Applications

Quantum mechanics, concepts and applications

Mouse Genetics: Concepts and Applications

Pharmaceutical biotechnology : concepts and applications

Quantum Mechanics: Concepts and Applications

Electropolymerization: Concepts, Materials and Applications

Synthetic Multivalent Molecules: Concepts and Biomedical Applications

Synthetic instruments: concepts and applications

Synthetic Instruments Concepts and Applications

Multivalent functions

Biomedical Data and Applications

Biomedical Applications

Stereochemistry of Radical Reactions: Concepts, Guidelines, and Synthetic Applications

Biomedical Applications

Cellular and Biomolecular Recognition: Synthetic and non-Biological Molecules

Calculus: Concepts and Applications

Proteomics: Biomedical and Pharmaceutical Applications

Calculus: Concepts and applications

Biology: Concepts and Applications

Biomedical Applications of Nanotechnology

Biomedical Applications Polymer Blends

Polyphosphazenes for Biomedical Applications

Chromium-VI Reagents: Synthetic Applications

Basic concepts of synthetic differential geometry

Internal Flow: Concepts and Applications

Pharmaceutical Biotechnology. Concepts and Applications

Environmental biotechnology: concepts and applications

Regression Analysis: Concepts and Applications

Understanding Hypertext Concepts and Applications

Pharmaceutical Biotechnology. Concepts and Applications

IP Multicasting: Concepts and Applications

Clinical Pharmacokinetics: Concepts and Applications

Quantum mechanics, concepts and applications

Mouse Genetics: Concepts and Applications

Pharmaceutical biotechnology : concepts and applications

Quantum Mechanics: Concepts and Applications

Electropolymerization: Concepts, Materials and Applications

Recommend Documents