Metal Complex - DNA Interactions

Metal Complex–DNA Interactions Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blac...

Author: Nick Hadjiliadis | Einar Sletten

135 downloads 1389 Views 7MB 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

Metal Complex–DNA Interactions

Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

Metal Complex–DNA Interactions Edited by NICK HADJILIADIS Department of Chemistry, University of Ioannina, Greece and EINAR SLETTEN Department of Chemistry, University of Bergen, Norway

A John Wiley & Sons, Ltd., Publication

This edition first published 2009 © 2009 Blackwell Publishing Ltd Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com. The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. 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 or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for every situation. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising herefrom. Library of Congress Cataloging-in-Publication Data Metal complexes : DNA interactions / edited by Nick Hadjiliadis, Einar Sletten. p. ; cm. Includes bibliographical references and index. ISBN 978-1-4051-7629-3 (cloth : alk. paper) 1. DNA-ligand interactions. 2. organometallic compounds. I. Hadjiliadis, Nick D. II. Sletten, Einar. III. Title: DNA interactions. [DNLM: 1. organometallic Compounds–chemistry. 2. Organometallic Compounds– pharmacology. 3. Antineoplastic Agents–therapeutic use. 4. DNA–chemistry. 5. Metals–therapeutic use. 6. Neoplasms–drug therapy. QV 290 M5831 2009] QP624.74.M48 2009 572.8′64–dc22 2008049834 A catalogue record for this book is available from the British Library Set in 10/12 Times by SNP Best-set Typesetter Ltd., Hong Kong Printed and bound in Great Britain by CPI Antony Rowe, Ltd, Chippenham, Wiltshire

To our families

Contents

Preface List of Contributors Acknowledgements A 1

2

Page xiii xv xix

BASIC STRUCTURAL AND KINETIC ASPECTS

1

Sequence-Selective Binding of Transition Metal Complexes to DNA Einar Sletten and Nils Åge Frøystein 1.1 Introduction 1.2 Ab initio Calculations and Photo-Cleavage Studies 1.3 NMR Spectroscopic Studies of Metal Binding to DNA Oligonucleotides 1.4 Summary of Theoretical and Experimental Evidence for Sequence–Selective Binding to DNA 1.5 Sequence-Specific Groove Binding Abbreviations References

3

Thermodynamic Models of Metal Ion–DNA Interactions Vasil Bregadze, Eteri Gelagutashvili and Ketevan Tsakadze 2.1 Introduction 2.2 Interactions of Metal Ions with DNA 2.3 Model and Mechanisms of Metal-Induced Formation of Point Defects 2.4 Conclusions Acknowledgements Abbreviations References

3 6 9 22 22 26 26 31 31 33 43 48 49 49 50

viii

3

4

B 5

6

Contents

Metal Ion Coordination in G-Quadruplexes Janez Plavec 3.1 Introduction 3.2 Cation Coordination and Stability of G-Quadruplexes 3.3 Structure of Sequences Consisting of Human Telomere Repeats 3.4 G-Quadruplexes Adopted by Promoter Regions 3.5 Bimolecular G-Quadruplexes of Analogues of the Oxytricha Telomere Repeat 3.6 Coordination of Cations within d[(TG4T)4] 3.7 Coordination of Cations within d[(G4T4G4)2] 3.8 Cation Movement within G-Quadruplexes 3.9 Summary Acknowledgements References Supramolecular Chemistry of Metal–Nucleobase Complexes P. J. Sanz Miguel, P. Amo-Ochoa, O. Castillo, A. Houlton and F. Zamora 4.1 Introduction 4.2 Discrete Architectures 4.3 Infinite Architectures 4.4 Overview of Metal Coordination Sites and Base-Base Hydrogen Bonding 4.5 M-DNA 4.6 Conclusion Acknowledgements References MEDICAL APPLICATIONS Platinum Drugs, Nucleotides and DNA: The Role of Interligand Interactions Giovanni Natile and Francesco Cannito 5.1 Introduction 5.2 Internucleotide Interactions 5.3 Guanine cis Amine Interactions 5.4 Solid-State Structures of Dynamic Nucleotides 5.5 Conformer Distribution in Cisplatin Adducts of G Derivatives 5.6 Retro Models Applied to Adducts with Tethered Guanine Bases 5.7 Conclusions and Perspectives Acknowledgements References Role of DNA Repair in Antitumour Effects of Platinum Drugs Viktor Brabec and Jana Kašpárková 6.1 Introduction 6.2 Human DNA Repair Systems

55 55 59 61 64 67 69 73 78 80 81 81 95

95 98 108 121 124 127 127 128 133 135 135 145 149 150 151 156 163 164 164 175 175 177

Contents

Specific Binding of Repair Proteins to DNA Modified by Antitumour Platinum Compounds 6.4 Repair of DNA Damage by Antitumour Platinum Compounds 6.5 Implications for Design of New Antitumour Platinum Compounds Acknowledgements References

ix

6.3

7

8

9

Telomeres and Telomerase: Potential Targets for Platinum Complexes Isabelle Ourliac-Garnier, Razan Charif and Sophie Bombard 7.1 Function of Telomeres 7.2 Structure of Telomeres 7.3 Telomerase 7.4 G-Quadruplex Structures and Small Molecules 7.5 Cisplatin 7.6 Interaction of Cisplatin and Related Platinum Complexes with G-Quadruplex Structures 7.7 Interaction of Cisplatin and Related Platinum Complexes with Telomeric DNA Duplexes 7.8 Interaction of Cisplatin and Related Platinum Complexes with Telomerase 7.9 Conclusion Acknowledgements Abbreviations References Towards Photodynamic Therapy of Cancer with Platinum Group Metal Polyazine Complexes David F. Zigler and Karen J. Brewer 8.1 Introduction 8.2 Study of Photophysics: Towards PDT 8.3 Photochemical Reactions of Metal Complexes with DNA 8.4 Designing Transition Metal Polyazines for DNA Photomodification 8.5 Cell Studies with Metal Complexes 8.6 Conclusions and Future Directions Acknowledgements Abbreviations References Platinated Oligonucleotides: Synthesis and Applications for the Control of Gene Expression Vicente Marchán and Anna Grandas 9.1 Therapeutic Applications of Synthetic Oligonucleotides 9.2 Platination as a Tool to Enhance Biological Effects

188 193 200 201 201

209 209 210 212 213 216 217 223 223 225 226 226 226

235 235 239 255 257 263 266 267 268 269

273 273 276

x

Contents

9.3 Synthesis of Platinated Oligonucleotides 9.4 Use of Platinated Oligonucleotides for Duplex Crosslinking 9.5 Use of Platinated Oligonucleotides for Triplex Crosslinking References

279 289 294 295

Rhodium– and Tin–DNA Interactions and Applications Kenneth D. Camm and Patrick C. McGowan 10.1 Introduction 10.2 Metal–DNA Interactions 10.3 Rhodium–DNA Interactions 10.4 Tin–DNA Interactions References

301

C

DNA RECOGNITION: NUCLEASES AND SENSORS

317

11

Groove-Binding Ruthenium(II) Complexes as Probes of DNA Recognition Jayden A. Smith, J. Grant Collins and F. Richard Keene 11.1 Introduction 11.2 Mononuclear Complexes 11.3 Dinuclear Complexes 11.4 Potential Biological Significance Abbreviations of Ligands References

10

12

13

DNA Recognition and Binding by Peptide–Metal Complex Conjugates Alexandra Myari and Nick Hadjiliadis 12.1 Introduction 12.2 Transition Metal Complex–Peptide Conjugates 12.3 Metallointercalator–Metallopeptide Conjugates 12.4 A Critical Survey and Future Perspectives 12.5 Conclusion Abbreviations References Artificial Restriction Agents: Hydrolytic Agents for DNA Cleavage Fabrizio Mancin and Paolo Tecilla 13.1 Introduction 13.2 DNA Hydrolysis 13.3 Free Ions and Mononuclear Complexes 13.4 Bimetallic Complexes 13.5 Conjugation with DNA-affine Subunits 13.6 Conjugation with Sequence-Selective Elements 13.7 The ARCUT System 13.8 A Critical Survey and New Perspectives

301 302 302 308 311

319 319 320 324 338 339 340

347 347 349 358 359 362 362 363

369 369 370 373 377 380 384 385 386

Contents

13.9 Conclusions Acknowledgements References 14

15

New Metallo-DNAzymes: Fundamental Studies of Metal–DNA Interactions and Metal Sensing Applications Zehui Cao and Yi Lu 14.1 Introduction 14.2 Metal Ions as Important Cofactors of DNAzymes 14.3 Selection of DNAzymes Using in vitro Evolution 14.4 Understanding Nucleic Acid Enzyme–Metal Ion Interactions 14.5 DNAzymes as Metal Ion Sensors 14.6 Summary Acknowledgements References Two-Metal-Ion-Dependent Catalysis in Nucleic Acid Enzymes Wei Yang 15.1 Chemistry of Nucleic Acid Synthesis, Cleavage and Strand Transfer 15.2 All DNA and RNA Polymerases Require Two Mg2+ Ions for Catalysis 15.3 Nucleases That Require Two Mg2+ Ions in the Active Site 15.4 Advantages of Two-Metal-Ion Catalysis: Specificity and Versatility 15.5 Concluding Remarks Acknowledgements References

D TOXICOLOGICAL ASPECTS 16

17

Structural Studies on the Mercury II-Mediated T-T Base-Pair Using NMR Spectroscopy Yoshiyuki Tanaka and Akira Ono 16.1 The History of the MercuryII–Mediated Thymine-Thymine Base Pair 16.2 Crystallographic Studies On HgII–Nucleobase Complexes 16.3 UV, UVCD and Vibrational (Ir/Raman) Spectral Studies 16.4 NMR Spectral Studies 16.5 Relationship Between 15n Chemical Shifts and Chemical Bonds 16.6 Applications of the T-HgII-T Base Pair 16.7 Biological Relevance of T-HgII-T Base Pairs 16.8 Concluding Remarks References Chromium-Induced DNA Damage and Repair Laura G. Little and Kent D. Sugden 17.1 Introduction 17.2 Nucleobase Oxidation

xi

389 389 390 395 395 396 397 399 403 410 411 411 415

415 417 420 428 430 430 430 437 439

439 440 441 444 452 454 455 457 457 463 463 468

xii

Contents

17.3 Sugar Oxidation 17.4 Chromium–DNA Binding 17.5 Cr(VI)-Induced DNA Crosslinks 17.6 Conclusions References 18

Arsenic-Induced Carcinogenicity: New Insights in Molecular Mechanism Andrea Hartwig and Tanja Schwerdtle 18.1 Introduction 18.2 Carcinogenicity and Cocarcinogenicity 18.3 Metabolism of Inorganic Arsenic 18.4 Modes of Action 18.5 Conclusions and Research Needs Acknowledgements Abbreviations References

Index

474 477 480 482 483

491 491 492 492 494 499 499 502 503 511

Preface

Metal ions and metal complexes have long been recognized as critically important components of nucleic acid chemistry both in regulation of gene expression and as promising therapeutic agents. Understanding how metal complexes interact with DNA has become an active research area at the interface between chemistry, molecular biology and medicine. In this volume, we attempt to bring together topics that span the breadth of this large area of research. The book is divided into four parts presenting recent developments in the field: Part A, Basic Structural and Kinetic Aspects, contains four chapters focusing on sequence-selective metal binding to DNA, thermodynamic models, metal coordination in G-quadruplexes and supramolecular networks of metal-nucleobases. Part B, Medical Applications, contains six chapters with the main focus on anticancer platinum drugs. Among the themes are the following: the role of interligand interactions, DNA repair in antitumour effects of platinum drugs, telomers and teleomerase as potential targets, photodynamic therapy, and synthesis and applications of platinated oligonucleotides for control of gene expression. A new and exciting field presented in Part C, DNA Recognition: Nucleases and Sensors, describes probes for DNA recognition, artificial restriction agents, metallo-DNAzymes for metal sensing applications and metal-ion-dependent catalysis in nucleic acid enzymes. Part D, Toxicological Aspects, deals with structural studies of mercury–DNA interactions, chromium-induced DNA damage and repair and the effect of arsenic-induced carcinogenicity. It has not been possible to cover all the topics that might find a place in a comprehensive coverage of metal–DNA interactions. The field is rapidly growing, spurred by significant improvements in experimental techniques. The applications of state-of-the art NMR spectroscopy, electrophoretic techniques, atomic force microscopy (AFM), etc. have dramatically expanded our knowledge on metal–DNA interactions during the last decade. The book presents pertinent up-to-date information for research scientists involved in, for example, the development of new and less toxic anticancer metallodrugs, the development of chemical nucleases and

xiv Preface

sensors, and the field of environmental and toxicological chemistry. Each chapter is written by experts on the various topics, and we hope this edition will prove to be a valuable contribution to the scientific community. Professor Nick Hadjiliadis (Ioannina, Greece)

Professor Einar Sletten (Bergen, Norway)

List of Contributors

Pilar Amo-Ochoa, Departamento de Química Inorgánica, Universidad Autónoma de Madrid, 28049 Madrid, Spain Sophie Bombard, Laboratoire de Chimie et Biochimie, Université Rene Descartes, 45 rue des Saints-Pères, 75270 Paris Cedex 06, France Viktor Brabec, Academy of Sciences of the Czech Republic, Institute of Biophysics, Kralovopolska 135, CZ-61265 Brno, Czech Republic Vasil Bregadze, E. Andronikashvili Institute of Physics, Georgian Academy of Sciences, 6 Tamarashvili Str, 0177 Tbilisi, Georgia Karen J. Brewer, Department of Chemistry, Virginia Tech, Blacksburg, VA 240610212, USA Kenneth D. Camm, School of Chemistry, University of Leeds, Leeds, LS2 9JT Francesco Cannito, Consorzio Interuniversitario di Ricerca in Chimica dei Metalli nei Sistemi Biologici (CIRCMSB), Bari, Via C. Ulpiani 27, I-70126 Bari, Italy Zehui Cao, Department of Chemistry, University of Illinois at Urbana-Champaign, A322 Chemical & Life Sciences Lab, 600 South Mathews Ave, Urbana IL 61801, USA Oscar Castillo, Departamento de Química Inorgánica, Facultad de Ciencia y Tecnología, Universidad del País Vasco, Apartado 644, E-48080 Bilbao, Spain

xvi

List of Contributors

Razan Charif, Laboratoire de Chimie et Biochimie, Université Paris Descartes, 45 rue des Saints-Pères, 75270 Paris Cedex 06, France J. Grant Collins, School of Physical, Environmental and Mathematical Sciences, University of New South Wales Australian Defence Force Academy, Canberra, ACT 2600, Australia Nils Åge Frøystein, Department of Chemistry, University of Bergen, Allégt. 41, 5007 Bergen, Norway Eteri Gelagutashvili, E. Andronikashvili Institute of Physics, Georgian Academy of Sciences, 6 Tamarashvili Str, 0177 Tbilisi, Georgia Anna Grandas, Departaments de Química Orgànica (Facultat de Química), IBUB Universitat de Barcelona, Martí i Franquès 1-11, 08028 Barcelona, Spain Nick Hadjiliadis, Department of Chemistry, University of Ioannina, Ioannina 45110, Greece Andrea Hartwig, Institut für Lebensmitteltechnologie und Lebensmittelchemie, Der Technische Universität Berlin, Gustav-Meyer – Allee 25, 13355 Berlin, Germany Andrew Houlton, Chemical Nanoscience Laboratory, School of Natural Sciences, Newcastle University, Newcastle upon Tyne, NE1 7RU, UK Jana Kašpárková, Academy of Sciences of the Czech Republic, Institute of Biophysics, Kralovopolska 135, CZ-61265 Brno, Czech Republic F. Richard Keene, School of Pharmacy & Molecular Sciences, James Cook University, Townsville, Queensland 4811, Australia Laura G. Little, Department of Chemistry, The University of Montana, Missoula, Montana 59812, USA Yi Lu, Department of Chemistry, University of Illinois at Urbana-Champaign, A322 Chemical & Life Sciences Lab, 600 South Mathews Ave, Urbana IL 61801, USA Fabrizio Mancin, Dipartimento di Scienze Chimiche, Università de Padova, Via Marzolo 1, I-35131, Padova, Italy Vicente Marchan, Departaments de Química Orgànica (Facultat de Química), IBUB Universitat de Barcelona, Martí i Franquès 1-11, 08028 Barcelona, Spain Patrick C. McGowan, School of Chemistry, University of Leeds, Leeds, LS2 9JT Pablo J. Sanz Miguel, Departamento de Química Inorgánica, Universidad Autónoma de Madrid, 28049 Madrid, Spain

List of Contributors xvii

Alexandra Myari, Department of Chemistry, University of Ioannina, Ioannina 45110, Greece Giovanni Natile, Dipartimento Farmaco-Chimico, Università degli Studi di Bari, Via E.Orabona 4, 70125 Bari, Italy Akira Ono, Department of Material & Life Chemistry, Faculty of Engineering, Kanagawa University, 3-27-1 Rokkakubashi, Kanagawa-ku, Yokohama, Kanagawa 221-8686, Japan Isabelle Ourliac-Garnier, Institut Curie, Centre de Recherche, 26 rue d’Ulm, 75248 Paris, Cedex 05, France Janez Plavec, Slovenian NMR Center, National Institute of Chemistry, Hajdrihova 19, SI-1000 Ljubljana, Slovenia Tanja Schwerdtle, Institut für Lebensmitteltechnologie und Lebensmittelchemie, Der Technische Universität Berlin, Gustav-Meyer – Allégt 25, 13355 Berlin, Germany Einar Sletten, Department of Chemistry, University of Bergen, Allgt. 41, 5007 Bergen, Norway Jayden A. Smith, School of Pharmacy & Molecular Sciences, James Cook University, Townsville, Queensland 4811, Australia Kent D. Sugden, Department of Chemistry, The University of Montana, Missoula, Montana 59812, USA Yoshiyuki Tanaka, Laboratory of Molecular Transformation, Graduate School of Pharmaceutical Sciences, Tohoku University, Aobayama, Aoba-ku, Sendai, Miyagi 980-8578, Japan Paolo Tecilla, Dipartimento di Scienze Chimiche, Università di Trieste, via Giorgeri, 1-34127 Trieste, Italy Ketevan J. Tsakadze, E. Andronikashvili Institute of Physics, Georgian Academy of Sciences, 6 Tamarashvili Str, 0177 Tbilisi, Georgia Wei Yang, National Institute of Diabetes and Digestive and Kidney Diseases, NIH, Bethesda, Maryland, MD 20892 USA Felix Zamora, Departamento de Química Inorgánica, Universidad Autónoma de Madrid, 28049 Madrid, Spain David F. Zigler, Department of Chemistry, Virginia Tech, Blacksburg, VA 240610212, USA

Acknowledgements

The editors would like to thank all authors for their important and high quality contributions to this book. Great thanks go to Jorunn Sletten for invaluable assistance in the editing process. Finally we thank Wiley for guidance through the production of this book.

Figure 3.2 (Plate 1) Variation in strand stoichiometry and topology of inter- and intramolecular G-quadruplex structures: (a) tetramolecular structure with all four parallel strands and all anti glycosidic torsion angles; (b) bimolecular and (c) monomolecular structures with antiparallel strands and alternating syn (in orange) and anti (in blue) orientations across glycosidic bonds

Figure 3.4 (Plate 2) Polymorphism of four repeats of human telomeric sequence: (a) intramolecular G-quadruplex formed by d[AG3(T2AG3)3] as determined by NMR spectroscopy in the presence of Na+ ions in solution;92 (b) the G-quadruplex fold with parallel GGG strands and double-chain reversal loops as determined by X-ray crystallography for the same sequence crystallized in the presence of K+ ions;104 (c) the topology of the four-repeat sequence in K+ ion containing solution (so called Hybrid-1 structure);204,205,207 (d) the fold of major form of the unmodified four-repeat sequence in K+ ion containing solution (so called Hybrid-2 structure).210 For clarity, only G-bases and their syn (in orange) and anti (in blue) orientations across glycosidic bonds are shown Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

Figure 3.5 (Plate 3) Representative structures of gene promoter regions: (a) G-quadruplex fold formed by guanine tracts of c-myc-2345 and c-myc-1245 sequences originating from the c-myc promoter, determined by NMR;231 (b) topology adopted by the c-myc-23456 sequence containing five of the six guanine tracts of the c-myc promoter region;232 (c) NMR structure of unprecedented intramolecular G-quadruplex formed by a G-rich sequence in the c-kit promoter in K+ solution233

Figure 3.6 (Plate 4) Topology of analogues of the Oxytricha nova telomere repeat sequence: (a) the structure of bimolecular d[(G4T4G4)2] quadruplex was one of the first high resolution structures of G-quadruplexes determined;96 (b) unique folding topology of d[(G3T4G4)2] with (3 + 1) strand orientations;244 (c) topology of d[(G4T4G3)2];243 (d) asymmetric structure of d[(G3T4G3)2]238

Figure 3.8 (Plate 5) Cation coordination within G-quadruplex structures adopted by d(TG4T) as revealed by X-ray crystallography: (a) Na+ ions (pink balls) within the 0.95 Å resolution structure (PDB ID 352D);157 (b) Na+ ions in the structure crystallized using lithium sulfate as the main precipitating agent (PDB ID 2O4F);246 (c) arrangement involving both Ca2+ (green balls) and Na+ (pink balls) cations crystallized in a mixed ion environment (PDB ID 2GW0);139 (d) Tl+ (red balls) and Na+ (pink ball) ions within the 2.2 Å resolution structure (PDB ID 1S45);247 (e) Tl+ (red balls) and Na+ (pink balls) ions within the 2.5 Å resolution structure (PDB ID 1S47);247 (f) Tl+ (red balls) and Na+ (pink balls) ions within the G-quadruplex exhibiting a T-quartet at the 3′-end of the strands (shown at the top of this structure) (PDB ID 1S47).247 Individual strands are shown in a wire frame model, while guanine bases are presented in stick representation to demonstrate the extent of out of plane bending. Cations are shown as coloured balls. Na+ with the smallest ionic radius can vary in coordination geometry from being within the plane of a G-quartet to being equidistant between two adjacent Gquartets. The larger ions are exclusively coordinated between two adjacent quartets

Figure 3.9 (Plate 6) Localization of cations within bimolecular G-quadruplexes: (a) Na+ ions (pink balls) within the d[(G4T4G4)2] quadruplex in crystals of the Oxytricha nova telomere end-binding protein with telomere DNA (PDB ID 1JB7);158 (b) K+ ions (grey balls) within d[(G4T4G4)]2 as revealed by X-ray crystal structure at 2.0 Å resolution (PDB ID 1JRN);99 (c) K+ ions (grey balls) within d[(G4T4G4)2] as revealed by X-ray crystal structure at 1.5 Å resolution (PDB ID 1JPQ);99 (d) five Tl+ ions (red balls) within d[(G4T4G4)2] as revealed by X-ray crystal structure at 1.55 Å resolution (PDB ID 2HBN);129 (e) ammonium ions within d[(G4T4G4)2] as established by NMR measurements in solution;162 (f) two ammonium ions within d[(G3T4G4)2] established by NMR studies in solution.176 Individual strands are shown in wire frame model, while guanine bases are presented in stick representation to demonstrate the extent of out of plane bending. Na+ ion can be localized within the plane of a G-quartet or between two adjacent G-quartets. The larger K+, Tl+ and 15NH4+ ions are exclusively coordinated between two adjacent quartets

platinum

hydrogen

nitrogen

carbonium

oxygen

phosphorus

Figure 5.9 (Plate 7) X-ray structures of two cis-A2Pt(5′-GMP)2 adducts showing how in the DHT conformer (top, based on data from Ref. 117) the 5′-phosphate is directed towards the cis amine, while in the ΛHT conformer (bottom, based on data from Ref. 118) the 5′phosphate is directed towards the cis guanine. For simplicity only the chelate ring atoms of the carrier ligands (Me4dach in the former case and Me4en in the latter case) are shown

HH1

∆HT1

HH2

platinum

nitrogen oxygen

carbonium phosphorus

Figure 5.13 (Plate 8) Minimum energy models (from NMR-restrained MMD calculations) of the anti/anti HH1 L (top) and anti/syn DHT1 L (centre) forms of (S,R,R,S)-bip-Pt(d(GpG)) and the anti/anti HH2 R (bottom) form of (R,S,S,R)-bip-Pt(d(GpG)) (the conformation of the 5′-G is given first, then that of 3′-G; only the chelate-ring atoms of the bip ligand are shown). Based on data from Refs 101 and 106

view from the major groove

view from the minor groove

platinum

nitrogen oxygen


Figure 5.15 (Plate 9) Central portion (four base-pairs) of oligonucleotides containing G/G intrastrand (top) and interstrand (bottom) crosslinks (based on data from Refs 114 and 143, respectively)

A H H N

T N

5' HO

-

N

O

O O P O O

O N

N

O O P OO-

O O P OO

N

O

H

O

N

N

H

N

N

3'

3'

O

H

HN

O

OO P O O

N

N

O

-

H

O

major groove

OH

5'

NH

minor groove

C

N

G

Figure 8.2 (Plate 10) Complimentary pairs of nucleotides that make up double helix DNA (Inset, A = adenine, T = thymine, G = guanine, C = cytosine, DNA = deoxyribonucleic acid). Atoms are carbon (grey), hydrogen (white), nitrogen (blue), oxygen (red) and phosphorous (pink)

N

N RuII N

N N

N

Cl

Cl

N

RhIII N

N N

N

N RuII N

N N

N

MC = [{(bpy)2Ru(dpp)}2RhCl2]Cl5 Cells treated with MC

illumination area

growth

after 48 hr growth period

calcein AM imaging of live cells

ethidium imaging of dead cells

Figure 8.20 (Plate 11) Micrographs of Vero cells pretreated with [{(bpy)2Ru(dpp)}2RhCl2]C l5, rinsed, and illuminated with 400–1000 nm light showing the high level of light-activated cell killing [bpy = 2,2′-bipyridine, dpp = 2,3-bis(2-pyridyl)pyrazine]. (From left to right) immediately after photolysis (light exposure within circle); after 48 hour growth period; live cell (green) visualized with calcein AM fluorescent dye; dead cell (red) visualized with ethidium homodimer-1 fluorescent dye84

Figure 14.2 (Plate 12)

Schematic representation of the in vitro selection process

Figure 14.5 (Plate 13) Single molecule FRET study of the 8-17 DNAzyme: (a) time traces of the FRET signals and changes upon injection of 100 mM Zn2+ at 21 s; (b) distribution of single DNAzyme molecules with different FRET efficiencies before and after injection of 20 mM Pb2+; (c) proposed reaction pathways of the DNAzyme in the presence of Zn2+/Mg2+ and Pb2+. (Reprinted with permission from Nat. Chem. Biol., 2007, 3, 763–768.)

Figure 14.7 (Plate 14) Design and performance of DNAzyme-based UO22+ sensor. Competing metals were tested from 10 µM to 1 mM, and UO22+ (the last three bars) was tested from 0 to 10 nM. Inset: Sequence, schematics and fluorescent detection of the DNAzyme-based UO22+ sensor. (Juewen Liu et al., A catalytic beacon sensor for uranium with parts-per-trillion sensitivity and millionfold selectivity, Proc. Natl. Acad. Sci. USA, 104, 2056–2061. Copyright 2007, National Academy of Sciences, USA.)

Figure 15.3 (Plate 15) Three types of catalytic centres found in DNA and RNA polymerases; (A) Phage T7 DNA polymerase represents the pol I-like active centre, which is found in the A, B and Y-family DNA polymerases, retroviral reverse transcriptases, phage RNA polymerases and viral RNA-dependent RNA polymerases. The catalytic residues (highlighted in blue and red) include two Asps and one Glu on two adjacent antiparallel strands (highlighted in cyan); (B) Human DNA pol b represents the pol b-like catalytic centre, which is found in the C and X-family DNA polymerases, poly-rA polymerases and CCA-adding enzymes. The catalytic residues include three conserved Asps on two adjacent parallel strands; (C) S. cerevisiae RNA pol II represents all bacterial and eukaryotic DNA-dependent RNA polymerases. The catalytic core consists of a b-barrel with three catalytically essential Asps on a loop (highlighted in cyan). All of these polymerases contain two absolutely conserved Asps, which together with the a-phosphate of (d)NTP jointly coordinate the two Mg2+ (shown as green spheres)

RNase H

MutH

Figure 15.4 (Plate 16) Two families of nucleases that have been shown to use two-metal-ion catalysis: (A) RNase H is a prototype of the superfamily, which includes nucleases, retroviral integrases, RuvC Holliday junction resolvase and dicer argonaute. The two catalytically essential Asps are located on two adjacent parallel strands (highlighted in cyan). Two additional carboxylates may be required for catalysis. Two Mg2+ are jointly coordinated by the carboxylates and the scissile phosphate; (B) MutH represents another superfamily, which includes type IIP restriction endonucleases, T7 endonuclease I (a Holliday junction resolvasae) and T4 exonuclease I. The catalytic residues, Asp (D), Glu (E) and Lys (K) are located on the diverging point of a b-hairpin (highlighted in cyan). The loop that contains the first Asp, which plays the central role in Mg2+ coordination, is often disordered in the MutH-family nucleases in the absence of a properly aligned substrate or divalent cations. A diagram of the metal ion coordination in each case is shown below the actual structure

Figure 15.6 (Plate 17) Coordination of the two metal ions changes during RNA cleavage by RNase H: (A) Enzyme–substrate complex; (B) Enzyme–product complex

Figure 15.7 (Plate 18) Coordination of the two metal ions in Group I intron. The two metal ions have similar coordination geometry (hence symmetric) and can reverse their roles as ‘A’ and ‘B’ metal ions in the successive steps of RNA splicing reaction. (Reproduced from Yang, W. et al., Making and breaking nucleic acids: two-Mg2+-ion catalysis and substrate specificity, Mol. Cell, 2006, 22, 5–13)

Part A Basic Structural and Kinetic Aspects


1 Sequence-Selective Binding of Transition Metal Complexes to DNA Einar Sletten and Nils Åge Frøystein

1.1 Introduction The biological significance of the interaction between metal ions and nucleic acids has become a rather well-established fact. One may mention the observed necessity for the presence of metal ions in many natural processes where nucleic acids play the dominant role. The effect of platinum-based chemotherapeutic drugs probably originates from their attack on DNA. Another aspect of metals in biological systems is the increased flux of metals in the environment during the last decades. An assessment of the toxic effect of an unnatural metal ion concentration must include information on the processes in which the metal can participate. In a comparison of metal carcinogenicity in humans based on several experimental factors, Cr and Ni turned out to be the most potent carcinogens.1 The nucleic acid monomers, guanine (G), adenine (A), thymine (T) and cytosine (C) have different metal ion affinities. The order of stability of 3d transition metal ion–nucleobase complexes are: G > A, C > T.2 At physiological pH the preferred binding sites on the nucleobases are: guanine N7, adenine N1 and/or N7, cytosine N3, thymine O4. For nucleotides the relationship between phosphate and base binding is dependent on the type of metal ion. Eichhorn and Shin3 studied the effect of various metal ions on the melting temperature of DNA (Figure 1.1). The authors suggest that magnesium ions increase Tm by binding to phosphate and Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

4

Sequence-Selective Binding to DNA

Mg Co Ni

75

Mn

Zn

Tm °C

65

55 Cd

45

Cu 35 0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

MOLES M+2/DNA (P)

Figure 1.1 Variation of Tm of solutions of DNA as a function of divalent metal ion concentration. (Reprinted with permission from J. Am. Chem. Soc., 1968, 90, 7327. Copyright 1968 American Chemical Society.)

stabilizing the double helix, whereas copper ions decrease Tm by binding to the bases and destabilizing the double helix. Based on the metal-induced variation in Tm they suggested that the relative metal affinity to the phosphate backbone of DNA follows the order Mg2+ > Co2+ > Ni2+ > Mn2+ > Zn2+ > Cd2+ > Cu2+. This implies that the binding of an individual metal ion may involve phosphate and base on the same molecule or form a linkage between two different nucleotides. An example of the latter situation is the mononucleotide–metal ion binding pattern observed for the Cu-(GMP) complex, where Cu2+ ions are bridging the GMP ligands through alternating N7–Cu–phosphate bonds (Figure 1.2).4 When nucleobases are incorporated into a duplex DNA matrix, the affinities towards metal ions are modified. It has been shown that several divalent metal ions, like Mn2+, Cu2+ and Pt2+ prefer GC-rich regions, while Hg2+, for example prefer ATrich regions.5,6 A more detailed picture indicates that metal binding to base residues is sequence-dependent, i.e. not all guanines in a particular sequence show identical affinity towards a specific type of metal ion.7–10 As a consequence, one may envisage designing metal complexes that can bind selectively to chosen sequences of DNA.

Introduction

5

Figure 1.2 The nucleobase–metal–phosphate linkage in the crystal structure of Cu(5′GMP) 4

Such complexes may be used as drugs that block specific gene expression associated with a certain disease. Metal ions can interact with nucleic acids in two distinct modes of binding: diffuse binding and site binding, both of which are important for the structure and function of nucleic acids.11a,b In the diffuse binding mode the metal and the nucleic acid retain their hydration layer and the interaction is through water molecules.11c This is a long-range Coulombic interaction, in which positive metal ions accumulate around the nucleic acid in a delocalized manner; for example, the counterion atmosphere that all nucleic acids possess is made up of diffusely bound positive ions. In the site-binding mode the metal is coordinated to specific ligands on the nucleic acid; the coordination can either be direct (termed inner-sphere) or through a water molecule (termed outer-sphere). In the outer-sphere binding mode only the innermost hydration layer of the metal is kept intact, and the metal and the nucleic acid ligand(s) to which the metal is coordinated share solvation shells. In inner-shell binding there is direct contact between the metal and the nucleic acid. Dehydration of the metal ion and the nucleic acid binding site therefore has to occur before an inner-shell bond is formed. The mechanism of inner-sphere binding is likely to be initiated by a diffuse binding mode, in which the metal and the nucleic acid are separated by no more than two layers of solvent molecules.11d This step is diffusion controlled. The next step is that the metal ion and the nucleic acid form an outer-sphere complex, separated only by one layer of solvent molecules. This step primarily depends on electrostatic attractions and hydrogen bonding between the metal and the nucleic acid.12 In the final step the metal and the nucleic acid come into direct contact (innersphere binding). Here the nucleophilicity of the coordination site plays a crucial role. In the last two steps steric effects are also important. Several attempts have been made to quantify the importance of accessibility and molecular electrostatic potential (MEP) at the site where the inner-sphere complex/covalent bond is formed.13,14 In these studies a reasonable correlation between these two important

6


factors exists, and has been used to predict which DNA site is the most reactive to metalation or methylation. In this chapter we present data on sequence-selective interactions between metal complexes and nucleic acids. In the outline we will distinguish between (i) site-selective inner-sphere metal coordination of nucleobases, and (ii) the selectivity of fully hydrated species located in the minor or major groove through hydrogen bonding and electrostatic interaction. In the former case a further distinction will be made between labile and nonlabile metals.

1.2 Ab initio Calculations and Photo-Cleavage Studies The highest occupied molecular orbital (HOMO) of DNA nucleobases plays a crucial role in metal coordination by interacting with the lowest unoccupied molecular orbital (LUMO) of metal ions.15 The calculations of HOMOs of macromolecules such as duplex DNA are extremely difficult. Consequently, there has been little focus on the role of the HOMO/LUMO in studies of DNA–metal ion interactions. Theoretical calculations of DNA bases have mostly focused on ionization potentials (IP) of monomeric nucleobases and the stability of the nucleobase pair in neutral and radical cation states.16 About ten years ago the Saito group published the first extensive studies on the variation of nucleobase IPs and localization of HOMOs as a function of base stacking, using high-level ab initio calculations.17 The IPs of four base monomers and 16 sets of nearest-neighbour stacked nucleobases in the B form were calculated. It was found that the GG/CC system has the lowest IP among ten possible stacked nucleobase pairs and that approximately 70% of the HOMO is localized on the 5′-G of 5′-GG-3′. These calculations indicate that the 5′-G of 5′-GG3′ is the most electron-donating site in B-DNA. The origin of IP lowering as a result of base stacking was further investigated by calculations of HOMO energy distribution as a function of the twist angle of a GG dimer.17 Within a normal range of twist angles for B-DNA (−25 ° to −45 °), IP values of GG are between 7.2 and 7.4 eV and the HOMO is predominantly localized on 5′-G. This implies that in B-form DNA the 5′-side G of the 5′-GG-3′ sequence is the most strongly interacting site with electrophiles. This principle may be very important in governing sequence-selective metal binding to DNA. According to the theoretical calculations, HOMOs of GG sites in duplex DNA should serve as the most reactive one-electron oxidation sites. In order to verify this assumption experimentally, Saito et al. performed laser flash photolysis of duplex DNA oligonucleotides with added photosensitizer (photocleaving aminoacid, PCA) and subsequent hydrolysis dephosphorylation with alkaline phosphatase (Scheme 1.1).18 The G3 : G4 ratio (84 : 16) determined by HPLC implies that the major degradation pathway of the hexamer involves the G3 site – the most readily oxidizable site according to theoretical calculations. It should be noted that photo-irradiation of the single-strand alone results in a nonselective cleavage at G3 and G4 in the ratio 1 : 1.

Ab initio Calculations

5´- T1T2G3G4T5A6

7

(T1 T2 + G3TA) + (TTG4 + TA) + PCA

3´-A-A-C-C-A-T

3´-A-A-C-C-A-T Scheme 1.1

Figure 1.3 Distribution of HOMOs (normalized to the largest value) in B-form duplex 5-mers obtained by ab initio calculations. The arrows indicate the distribution of the HOMOs normalized to the largest HOMO as 100 20

Further work by the Saito group involved ab initio calculations of HOMOs for a wide variety of double-stranded G-containing 5-mer sequences with B-form geometry using GUSSIAN 9x1at the HF/6-31G* level.19,20 For the quantum mechanical studies, all of the sugar backbones of the 5-mers were removed from the coordinate file and replaced by methyl groups. A few examples of the distributions of HOMOs in the duplex 5-mers are shown in Figure 1.3. The general trend for HOMO distribution is that the HOMO of stacked GG doublets is localized overwhelmingly on the 5′-G, regardless of the 3′- and 5′flanking residues (A, C or T). Bearing in mind that the model used is rather crude (all of the sugar backbone replaced by methyl groups), further discussion of more subtle sequence-specific differences is not warranted. Further experimental support for the theoretical results was obtained by studying the oxidation of oligodeoxynucleotides (ODN) with Co2+ ions and benzoyl peroxide using PAGE analysis of the reaction mixture after hot piperidine treatment. Sequence-dependent G-cleavage was observed for double-stranded ODN, whereas nonselective equal G cleavage was observed for single-stranded ODN. The relative rates of sequence oxidation were determined by densitometric assay of the ODN cleavage bands. Experimentally observed relative rates of G oxidation matches well with the calculated HOMOs of the G-containing sequences, implying that the Co2+ ion is coordinated more strongly to the G having the larger HOMO.20 Comparable theoretical calculations by Senthikumar et al.21 on stacked XGY triplets with B-form geometries, including sugar and phosphate groups, show that the site energy is strongly influenced by the type of nucleobase at the 3′ position.

8


When C or T is present at the 3′ position, the site energy at the guanine was found to be up to 0.44 eV higher than for A or G at this position. The influence of the base at the 5′ position was much smaller, the variation in site energy being less than 0.1 eV. The amount of charge on a certain G was calculated from the coefficients of the guanine fragment orbitals. It was concluded that the neighbouring base at the 3′position to a large extent determines the charge distribution and therefore the oxidative damage on a sequence of guanine bases.21 Further work by the Saito group on experimental mapping of G-rich hot spots included photo-induced cleavage of double-stranded 32P-end-labelled oligodeoxynucleotides (ODNs) 30-mers possessing two different G-containing sequences (5′TXGYT-3′) and a 5′-TTGGT-3′ step as a standard on the same strand.22 Under low photo-irradiation conditions, only the cleavage bands of 5′-Gs of the two GG steps and the middle step of the GGG triplet were observed by hot piperidine treatment. Quantitative densitometric assay was used to determine the relative amounts of cleavage products. The experimental results were compared to IP values calculated for 16 sets of base-paired G- and GG-containing 5-mers. A plot of the log of the relative reactivity (krel) toward photo-induced one-electron oxidation versus calculated IP is shown in Figure 1.4. A different explanation for the enhanced reactivity of the 5′-G of a GG step compared to the 3′-G is presented by the Schuster group.23 Theoretical calculations of base-paired quartets, d(5′-XGGX-3′)/d(5′-YCCY-3′), suggested that electronic factors may not be the primary determinant of the reaction selectivity for GG steps. Instead the authors propose, based on molecular dynamic (MD) simulations on BDNA oligomers, d(5′-GXXGGXXG-3′)/d(5′-GYYGGYYG-3′), where X = A,T,U and Y is the complementary base, that ‘there is an important steric contribution to the preference for reaction at the 5′-G in the GG doublets’. Photo-cleavage experiments carried out on the series of B-DNA oligomers showed that for GG steps in

Figure 1.4 A plot of the log of the relative reactivity (krel) toward photo-induced one-electron oxidation versus calculated ionization potential (IP). The relative reactivity was obtained from densitometric assay of G bands with TGG as a standard (krel = 1.0) under single-hit conditions. (Reprinted with permission from J. Am. Chem. Soc., 1998, 120, 12687. Copyright 1998 American Chemical Society.)

NMR Spectroscopic Studies

9

the context AGGA, the ratio of 5′ to 3′ reactivity was 1.8 ± 0.1, and for GG in the context TGGT, the ratio was 6.1 ± 0.3. The authors propose that the accessibility of H2O to the reaction site determined by the steric blocking by the methyl group plays the dominant role for the observed sequence-selectivity, rather than electronic effects.23

1.3 NMR Spectroscopic Studies of Metal Binding to DNA Oligonucleotides 1.3.1 NMR Methodology Most early nuclear magnetic resonance (NMR) studies on DNA involved complementary homopolymers and self-complementary, alternating copolymers, e.g. poly(dA)/poly/dT).24 The development of efficient and rapid methods of large-scale oligonucleotide syntheses has made it possible to design heteropolymeric sequences of high purity. Dodecamer (12 base pair) sequences adopting a normal B-DNA double-helical conformation, are assumed to complete a full turn of a right-handed helix. The structure of such a mini-helix is probably sufficiently close to that of real DNA to serve as a realistic model for determining preferred metal binding sites. The effects of adding paramagnetic metal ions to an aqueous solution of DNA fragments may be monitored by observing the decrease in spin-lattice (T1) and spinspin (T2) relaxation times (related to line-broadening) for protons close to the metal centres. Paramagnetic metal ions may be classified according to their electronic correlation times, i.e. as relaxation probes producing broad lines or as paramagnetic shift probes producing narrow lines. Divalent manganese is a typical relaxation probe with an estimated electronic relaxation time (ts = T1c = Tc) of 10−8–10−9 s, while nickel, which has a shorter ts in the range 10−10–10−12 s, is a typical chemical shift probe. Cobalt(II) in a low-spin coordination environment has an estimated ts between that of Mn2+ and Ni2+ in kinetically labile metal complexes. At low metal to nucleotide ratios paramagnetic shift effects of Ni2+ are difficult to detect. In this case geometric information about metal binding sites is most effectively obtained by measuring proton spin-lattice relaxation times (T1). Paramagnetic relaxation arises in NMR spectroscopy when an unpaired electron spin interacts with a nuclear spin. The large magnetogyric ratio of the electron compared to that of the proton makes the dipolar coupling to the electron spin a very effective means of relaxation for the nuclear spin.25,26 Scalar interactions between the electron and nuclear spins have similar effects. In the simplest possible case, a ligand molecule exchanges between a paramagnetic environment (e.g. bound to Mn(II), S = 5/2) and a ‘free’ state, when the ligand is present in solution in vast excess to the paramagnetic centre (e.g. 102–104). The effect of paramagnetic metal ions located at specific binding sites on DNA is observed as differential linebroadening of proton signals close to the binding site. Often, in 1D spectra of oligonucleotide molecules containing ten base pairs or more, key proton resonances may be severely overlapped, preventing an accurate assessment of the influence of

10


the added metal ions. In these cases, 2D NOESY experiments may be used to obtain sufficient resolution. For diamagnetic metal ions (no unpaired spin) the formation of a chemical bond is usually found to cause changes in the chemical shifts of proton resonances of hydrogen atoms in the proximity of the metal binding site. However, the coordination of Hg2+ ions to single nucleobases induces only rather insignificant 1H shift changes, as shown for thymidine and guanosine.27 This could be explained by a down-field chemical shift change, induced by metal binding, being cancelled by an up-field chemical shift caused by changes in ring current effects due to altered nucleobase stacking.28 The heteronuclei 13C, 15N and 31P may experience large shifteffects when a metal ion binds to nucleobase or phosphate groups. 1.3.2 Model Systems One of the first oligodeoxynucleotides studied by X-ray crystallography29 and subsequently by NMR spectroscopy30 was the famous Dickerson–Drew sequence, d(C1G2C3G4A5A6T7T8C9G10C11G12) (seq. I). The 1D and 2D NMR spectra used as bases for the three-dimensional NMR structure showed an unexpectedly strong line-broadening of the G4-H8 resonance. As a consequence, the G4-H8 to C3-1′H cross-peak was barely detectable in the 2D NOESY spectra, and the corresponding interproton distance was scored as >4 Å by using the usual r6 dependency to calculate the proton–proton distance from the intensity of the cross-peaks.31 Structure determination based on protom–proton distance constraints showed a distinct bend of the helix axis at the G4 residue.30 However, at a later stage, when EDTA was added to the solution the G4-H8 resonance was found to exhibit normal line width.7 Since trace amounts of paramagnetic impurities are often found to be present in nucleotide preparations it is recommended to always run samples through a CHELEX column to remove such impurities before NMR metal ion titration experiments are performed. 1.3.3 Oligionucleotide–Transition Metal Adducts Mn2+ Ions The first metal ion titration NMR experiment on a DNA oligonucleotide was carried out by Frøystein and Sletten on the Dickerson–Drew duplex d(C1G2C3G4 A5A6T7T8C9G10C11G12) (seq. I) which contains the Eco-RI recognition sequence GAATTC.7 Since most of the base proton signals are well resolved in the 1D spectrum, the effect of adding aliquots of a solution of MnCl2 was easily observed. A plot of metal-induced paramagnetic line-broadening of G-H8 showed a clear preference for G4 compared to G2, G10 and G12 (Figure 1.5). At a Mn2+: duplex ratio of 1 : 104 substantial line-broadening was already detected for the G4-H8 signal, while insignificant effects were observed for the other base proton resonances in the duplex. Of particular importance was the observation that the central adenine A-H2 protons showed no broadening effect, indicating that Mn2+ ions were not located in


11

Figure 1.5 Mn2+ induced line-broadening of G-H8 resonances of the Dickerson–Drew duplex [d(5′-C1G2C3G4A5A6T7T8C9G10C11G12)]2. G4 (open squares), G10 (filled squares). The circles represent the overlapping G2 and G12 resonances7

the minor groove (see discussion below). The amount of broadening was consistent with inner sphere coordination to G-N7. The binding to guanine by 3d transition metals was not surprising, considering the differences in thermodynamic stability of the corresponding complexes of the nucleoside and nucleotide monomers.32 However, the sequence-selective metal binding pattern manifested through the apparent preference for G4 rather than the residues G2, G10 or G12, was highly unexpected. Further NMR spectroscopic evidence for apparent sequence-selective metal binding to DNA is presented for the duplex d(G1C2C3G4A5T6A7T8C9G10G11C12) (seq. II) containing the Eco-RV recognition sequence GATATC.8 The 1D spectra with and without MnCl2, shown in Figure 1.6, exhibit differential G-H8 line-broadening in the order: G10 > G4 >> G1 ∼ G11. It is evident that the exposed terminal residues both in sequence II and the previous sequence I, although offering favourable accessibility, are not the preferred binding sites. One may also notice that the A5-H2 and A7-H2 protons, residing in the minor groove, are completely unaffected. A systematic search for a selectivity pattern was initiated by designing three self-complementary sequences: [d(TATGGTACCATA)]2 (III), [d(TATGGATCCATA)]2 (IV) and [d(TATGGCCATA)]2 (V), where the central triplet is GGX (X = A, T or C).10 Mn2+-induced line-broadening for G-H8 versus r = [Mn2+]/[duplex] is plotted in Figure 1.7a–c. The line-broadening of G-H8 for the 5′-G is practically the same for all three sequences. However, the linebroadening of 3′-G-H8 is seen to depend on the adjacent residue X following a

12


Figure 1.6 400 MHz 1H NMR spectra, the aromatic region, of the dodecamer [d(GCCGATATCGGC)]2: (a) the spectrum of Mn2+-free solution, and (b) with added MnCl2 at a Mn2+/duplex ratio of approximately 10−3. The H8 and H6 proton resonances of purine and pyrimidine residues are labelled according to their sequential assignment8

distinct sequence-selective pattern: GA > GT >> GC. The adenine A-H8 resonances, plotted as references, are not influenced by the paramagnetic ions. Elmroth and coworkers have shown that metalation of single-stranded oligonucleotides of the type d(TnGT16−n) reach a maximum in the central part of the oligomer indicating lack of sequence-selective influence on the reaction rate.33,34 This result is in accordance with NMR and photo-cleavage studies on metalation of single-stranded DNA oligomers.18 Based on Monte Carlo descriptions of the oligoelectrolyte properties of double-stranded DNA oligomers it has been postulated that outer-sphere Coulombic interactions cause cations to be localized preferentially in the interior rather than the terminal part of the DNA oligomers.35 For inner-sphere metalation other factors may dominate, as demonstrated for the palindromic hexanucleotide d(G1G2C3G4C5C6) (seq. VI).36 Addition of MnCl2 induces selective line-broadening on the terminal G1-H8 with no significant effect on G2 and G5, as shown in the plot of line-broadening versus added MnCl2. In contrast, the terminal G1-H8 in duplex II (see above) shows almost no line-broadening. Labiuk et al.37 have reported X-ray determinations of Co2+, Ni2+ and Zn2+ complexes of d(G1G2C3G4C5C6) (seq. VI), where the metal ions are coordinated only to the terminal guanine G1-N7 position, with no metal ions binding to nonterminal guanine positions. The authors concluded that in the regular B-DNA conformation the internal binding sites are not accessible to Co2+, Ni2+ and Zn2+, and that consequently these metal ions bind exclusively to the terminal region of double-helical B-DNA, irrespective of base sequence. This is in contrast to our studies, which show a clear sequence-selective binding pattern for 3d metal ions with no special


13

Figure 1.7 Line-broadening versus Mn2+/Duplex ratio of the H8 resonances of the guanosines 5′-G4 (*) and 3′-G5 (ⵧ) with the terminal adenine H8 protons (䊊) for comparison. (a) d(TATGGTACCATA)2; (b) d(TATGGATCCATA)2; (c) d(TATGGCCATA)2. (J. Vinje, J.A. Parkinson, P.J. Sadler, T. Brown, E. Sletten, Sequence-selective metalation of double-helical oligodeoxyribonucleotides with PtII, MnII and ZnII ions. Chem. Eur. J., 2003, 9, 1620–1630. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

preference for terminal base residues. Based on the NMR results from several DNA sequences the d(GGCGCC) duplex is expected to bind preferentially to the 5′-G in the 5′-GG step and that the 5′-G in the 5′-GC step should have negligible affinity, in agreement with the X-ray result.37 Crystallographic data of a metal–DNA complex containing an internal GG step would probably show binding at this site. Table 1.1 is a compilation of available line-broadening data including both controlled titration experiments and the effect of paramagnetic impurities (reversed by addition of EDTA). One may notice that only Gs in the context G-purine exhibit maximum broadening. For sequence no. 6, which contains no G-purine step, the 5′-G-H8 resonance in the context of the 5′-GT step exhibits maximum broadening. A close inspection of Table 1.1 reveals that the line-broadening of 5′-G-H8 follows a consistent pattern, where the influence on the 5′-G-H8 is dependent on the residue

14

Sequence-Selective Binding to DNA Table 1.1 Mn2+ paramagnetic induced broadening of G-H8 resonances. The residues which show maximum broadening are marked in bold and underlined. Restriction enzyme recognition sequences are marked in italic Sequence 1. 5′-CGCGAATTCGCG 2. 5′-GCCGATATCGGC 3. 5′-GCCGTATACGGC 4. 5′-GCCAGATCTGGC 5. 5′-GAATTTAAATTC 6. 5′-CGCGTATACGCG 7. 5′-GCCGTGCACGGC 8. 5′-GCCGTTAACGGC 9. 5′-GCCTGATCAGGC 10. 5′-CCAAGCTTGG 11. 5′-GCCGAATTCGGC 12. 5′-ATGGGTACCCAT 13. 5′-TATGGTACCATA 14. 5′-TATGGATCCAAT 15. 5′-TATGGCCATA 16. 5′-GGCGCC

Eco-RI Eco-RV AccI BglII

HpaI BclI

on the 3′-side according to the simple rule : 5′-GG ≥ 5′-GA > 5′-GT >> 5′-GC. Apparently, the residue on the 5′-side is less important. At higher metal ion concentrations most 1H NMR resonances undergo varying degrees of broadening. It must be stressed, however, that only relative broadening effects for each sequence are taken into account. For sequences 12, 13 and 14 the GGX (X = A, T, G) triplet is placed in an almost identical environment. The effects of varying the X-residue are clearly shown in Figure 1.7. In sequence no. 12 containing a GGG triplet, both linebroadening and T1 data show the following order: G3 = G4 > G5 consistent for, not only Mn2+, but also for Ni2+ and Co2+ (see Ni2+ ions below). Fe2+ Ions In early investigations of Fe2+-DNA binding using NMR, the T1 spin–spin relaxation time technique indicated that both the base and the phosphate groups interact with Fe2+ ions.38 Later on, Bertoncini et al.39, by using X-ray absorption near edge structure spectroscopy (XANES), demonstrated that while Fe3+ tends to associate with oxygen ligands, Fe2+ prefers to form inner-sphere coordination with nitrogen ligands on DNA. Linn and coworkers have made extensive studies on sequence-specific DNA cleavage by Fe2+-mediated Fenton reactions.40–42 Nicking of duplex DNA by the iron-mediated Fenton reaction occurs preferentially at a limited number of nucleotide sequences. Most notable are a purine nucleotide followed by three or more G residues, [RGGG], and purine nucleotides flanking a TG combination, [TTGR].42 The preferred reaction sites are probably a consequence of sequenceselective localization of the ferrous ions. Using 1H NMR to characterize Fe2+ binding within the duplex CGAGTTAGGGTAGC/GCTACCCTAACTCG it was shown that Fe2+ binds preferentially at the GGG sequence, most strongly towards its 5′-


15

end.42 These studies are especially interesting because of the presence of RGGG in a large majority of telomere repeats.40 Recent studies have implicated that telomere shortening during human aging may be accelerated by oxidative stress.43,44 Co2+ Ions The Co2+ ion in a low-spin environment has an estimated electronic relaxation time (ts) between those of Mn2+ and Ni2+. Selective line-broadening was observed when CoCl2 was added to the duplex, d(A1T2G3G4G5T6A7C8C9C10A11T12) VII.9 At r = Co2+/ phosphate = 1.8 × 10−3 Co2+ was found to bind preferentially at the GGG sequence with almost equal affinity based on selective line-broadening; for G3 and G4 (35 Hz) and significantly less (12 Hz) for G5. NOESY spectra show that the intensities of cross-peaks were reduced by paramagnetic effects from the nearby bound Co2+ ion. The differential H8 broadening of G3, G4 and G5 observed in the 1D spectra is reflected as quenched NOE effects for the intra- and inter-proton cross-peaks involving G-H8 … G-H1′ connectivity. As a caveat we should mention that minute paramagnetic impurities in DNA samples used for NMR structure determination seriously affect the validity of the structural analysis since calculations of proton– proton distances are based on the magnitude of cross-peak intensities. Ni2+ Ions The electronic relaxation time (ts) for Ni2+ in the range 10−10 to 10−12 s is much shorter than for Mn2+. Consequently, spin-spin T1 relaxation time measurements will give more reliable information on metal binding sites than line-broadening (T2) data. Figure 1.8a shows that the pattern for Ni2+ binding to the Dickerson–Drew sequence deduced from T1 data is almost identical to that observed for Mn2+ binding based on line-broadening data.9 The T1 results of NiCl2 titration of the d(A1T2G3G4G5T6A7C8C9C10A11T12) duplex, which contains a triple GGG sequence are plotted in Figure 1.8b. The affinities for G3 and G4 are identical within experimental errors, while the affinity of G5 in the context GT is, as expected, significantly lower. The T1 relaxation pattern follows exactly the same trend as was shown for the Co2+ line-broadening measurements.9 Abrescia et al.45 reported an X-ray structure determination of the Ni2+ /DNA complex of the sequence d(CGTATATACG)2. The Ni2+ ions were associated with N7 atoms of all the guanines and with none of the N7 atoms of adenines. Subsequently the same group published a low resolution X-ray structure determination of the duplex d(CGTGTACACG), where they found Ni2+ ions associated only with the terminal Gs and the phosphate backbone.46 This last result is not in accordance with the sequence-selectivity observed in solution. However, the discrepancy may be due to crystal packing effects, which lead to cross-linking of guanines by Ni2+ ions. Cu2+ Ions One of the first detailed UV spectroscopic studies on the influence of metal ions on DNA stability was published by Eichhorn and Shin.3 While some metal ions tend


16 a

b 4

G4 H8

10

G3 H8 G4 H8

3

6 A5 H8

4

G2 H8 G12 H8 G10 H8

2 0 0

5

10

15 20 r*104

25

30

T1−1

T1−1

8

G5 H8

2 1 A11 H2

0

0

5

10

15 20 r*104

25

30

Figure 1.8 Plot of T1−1 versus r = [Ni(II)]/[Phosphate] at 312 K for G-H8 and A-H8 resonances of selected residues along the sequences: (a) [d(C1G2C3G4A5A6T7T8C9G10C11G12)]2; (b) [d(A1 T2G3G4G5T6A7C8C9C10A11T12)]2. (Reprinted from Inorg. Chim. Acta, 273, 1–2, E. Moldrheim, B. Andersen, N. A. Froystein, E. Sletten, Interaction of manganese (II), cobalt (II) and nickel (II) with DNA oligomers studied by 1H NMR spectroscopy, 6. Copyright 1998, with permission from Elsevier.)

to stabilize the double-helical structure, Cu2+ salts were shown to induce a dramatic lowering of the melting temperature (Figure 1.1). Recent NMR studies of Cu2+ and 2+ Cu( en)2 binding to the Dickerson–Drew duplex show the same selectivity pattern as for the other transition metals.47 The increase in line width at an r = [Cu2+]/ [Phosphate] of 2.5 × 10−3 was approximately 10 Hz, a corresponding increase in line width for [Cu(en)2]2+ was obtained at r = 15 × 10−3. Oikawa and Kawanishi have investigated the telomere shortening induced by H2O2 plus Cu2+ and found predominant DNA damage at the 5′-site of 5′-GGG-3′ in a 48-base fragment of the telomere duplex.48 In contrast, when single-stranded DNA was used, the damage induced by oxidative stress occurred at every guanine. The difference in site specificity of DNA damage between double-stranded DNA and single-stranded DNA could be explained in terms of the lower ionization potentials of stacked guanine base pairs in double-stranded DNA, on the basis of theoretical calculations. Zn2+ Ions UV spectroscopic titration studies of ctDNA by ZnCl2 indicate that Zn2+ ions bind both to phosphate groups and base nitrogen atoms.3 Marzilli and coworkers carried out Zn2+ titration of duplex d(ATG3G4G5TACCCAT) (VII) containing a GGG triplet, and found the chemical shift variation for the G-H8 in the order G4 > G3 >> G5.49 Zn2+ titration experiments analogous to those using Mn2+ (see above) have been carried out for sequences III, IV and V.10 In contrast to paramagnetic metal systems, where very low metal/DNA ratios were employed, excess salt was added to the duplex solutions until an upper limit of Zn2+-induced chemical shift changes


17

Figure 1.9 Chemical shift versus Zn2+/duplex ratio plots for the H8 resonances of 5′-G4 (*) and 3′-G5 (䊐), with the terminal adenine H8 protons (o) for comparison: (A) d(TATGGTACCATA)2; (B) d(TATGGATCCATA)2; (C) d(TATGGCCATA)2. (J. Vinje, J. A. Parkinson, P.J. Sadler, T. Brown, E. Sletten, Sequence-selective metalation of double-helical oligodeoxyribonucleotides with PtII, MnII and ZnII ions. Chem. Eur. J., 2003, 9, 1620–1630. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

was reached. This upper level was observed at r = [Zn2+]/[duplex] ≈ 5, which corresponds to a [Zn2+]/[G] value of ≈1.2. Plots of chemical shifts versus r exhibit the same trend as for the Mn2+ titration data (Figure 1.9). In a recent multinuclear NMR study on Zn2+ binding to oligonucleotides the variation in 1H, 15N and 31P chemical shifts was monitored as a function of added ZnCl2.36 Measurements for three different sequences were carried out: d(GGCGCC) (III); d(GGTACCGGTACC)2 (VIII); d(GGTATATATACCGGTATA) (IX). The chemical shift pattern for sequence III followed closely the line-broadening pattern determined for Mn2+ (see above). For sequence VIII large chemical 15N shift effects were observed for G7-N7 at natural abundance, a clear proof of direct metal binding to this site. 31P spectra also show large chemical shifts for the central G7 residue indicating that at excess ZnCl2 concentration the Zn2+ ions have direct contact with the G7 phosphate group and/or that a change in the phosphodiester backbone conformation has occurred.36

18


Zn2+ titration of IX was carried out to check if the expected accumulation of negative electrostatic potential in the interior of a long sequence33,34 (compared to shorter ones) would lead to a higher Zn2+ affinity for the internal GG step, compared with that observed for VIII.36 The results clearly showed that there is no enhanced selectivity for the internal Zn2+ binding for the long sequence. [Pt(dien)]+ Ions Numerous publications have dealt with the binding pattern of cisplatin anticancer drugs to DNA (see Chapters 5–9 in this book). One of the first studies on this topic involved enzymatic digestion of platinated DNA using the well known anticancer drug cisplatin (cis-[PtCl2(NH3)2]).50 HPLC analysis of the cleavage products showed that cisplatin forms chelates spanning G and/or A residues with the following percentages : Pt-GG (65%), Pt-AG (25%), Pt-GA (0%). Apparently, the rule of sequence-selectivity proposed for labile metal ion complexes based on 1H NMR and photochemical cleavage studies does not apply to nonlabile platinum complexes. However, the mechanism of chelate formation does not necessarily follow the sequence-selectivity of the initial formation of monofunctional adducts. The fact that the 5′-monoadducts are formed more rapidly and chelate more slowly than the 3′monoadducts might reflect the inherently greater reactivity of the 5′-G compared to the 3′-G. The common view is that the monoaqua species of cisplatin make the initial attack on DNA bases, although recent results have lead Chottard and coworkers to suggest that cisplatin may undergo double hydrolysis before reacting with DNA.51 In most reported studies on the kinetics of monofunctional platination reactions of single- and double-stranded oligonucleotides, both NMR spectroscopic and chromatographic methods have been used to determine kinetic parameters. The reactions involving bifunctional platinum complexes have usually been carried out with single-stranded oligonucleotides. In these reactions an initial monofunctional adduct is formed, which subsequently ring-closes to form a bifunctional chelate. The proposed selectivity rule is based on the intrinsic binding properties of duplex DNA. Sadler and coworkers showed that in a reaction between 15N-labelled [PtCl(dien)]+ and a single-stranded 14-mer d(ATACATGGTACATA), little kinetic preference for platination of either 5′G or 3′G sites was observed, while the single-stranded 8mer d(ATACATGG) showed a distinct preference for 5′-G platination.52 Chottard and coworkers, using hairpin-forming oligomers as duplex models, concluded that the selectivity for monofunctional attack by Pt2+ on 5′and 3′ G-residues is dependent on the ligand in the trans position (e.g. Cl−, H2O, OH−, NH3).51 In a later study they proposed a model for sequence-selective binding of cisplatin to DNA duplexes, involving a combination of molecular potentials and N7 accessibility.53 In a study on selectivity of adduct formation between 15N -cisplatin and 14-mer duplexes containing central AGT and GAT residues, respectively, Hambley and coworkers concluded that ‘the purine base on the 3′side of the pair exerts substantially greater influence on the rate of binding at the 5′-base than does the 5′-base on the rate of binding at the 3′-base’.54 However, monofunctional binding of cisplatin to the TGAT


19

Scheme 1.2

sequence was found to be approximately an order of magnitude slower than binding to TAGT sequences, a result which does not agree with the proposed selectivity rule for labile metal ion complexes. Vinje et al.10 have investigated the monofunctional reaction kinetics of [PtCl(dien)]+ (dien = diethylenetriamine), with the oligodeoxyribonucleotides, III, IV and V, used for Mn2+ and Zn2+ titration (see above). The reaction mixtures were separated by HPLC and the chromatographic profiles showed a clear difference in the amount of 3′ versus 5′ monoplatinated species between the three duplexes (Figure 1.10). The reaction pathway of platination of seq. III is shown in Scheme 1.2. Reaction rates for Pt(dien) were determined based on 2D [1H, 15N] HSQC/ HMQC NMR spectroscopy using 15N-labelled Pt(dien). The time courses for reaction rates of each of the three duplexes are shown in Figure 1.11. Comparison of the three different duplexes containing the central sections GGX (X = A, T, C), respectively,10 shows that the selectivity for covalent platination matches that for adducts with labile metal ions Mn2+ and Zn2+. The reaction rate for platination is faster for the 5′-G than for the 3′-G. The reactivity of the 3′-G depends on the adjacent residue X in the following order: X = A > T >> C. For GGA the reaction rate is 1.2 times faster for 5′-G than for 3′-G, for GGT the rate of 5′-G platination is about eight times faster than that of 3′-G, and for GGC there is no significant adduct formation at the 3′-G (Table 1.2). When the reaction mixture of the platinated species was aged over several weeks, the relative amounts of 5′ and 3′ HPLC fractions changed. This was also confirmed by comparison of the NMR spectra of the aged solutions with those recorded in the initial kinetic experiments. It was suggested that Pt-N7 bond cleavage and isomerization had taken place, similar to that observed for platinated single-

20


Figure 1.10 Chromatograms of the reaction products from the reaction between Pt(dien) and DNA: (a) d(TATGGTACCATA)2; (b) d(TATGGATCCATA)2; (c) d(TATGGCCATA)2. The HPLC fractions were collected ca 1 day after the reactions were started. The assignments of the fractions are based on analysis of 2D NOESY spectra. (J. Vinje, J.A. Parkinson, P.J. Sadler, T. Brown, E. Sletten, Sequence-selective metalation of double-helical oligodeoxyribonucleotides with PtII, MnII and ZnII ions. Chem. Eur. J., 2003, 9, 1620–1630. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

stranded DNA.52 Leng and coworkers have suggested a catalytic property of the DNA double helix to explain the rearrangement that is observed when a stable trans-platinated single strand is annealed with its complementary strand.54 The occurrence of Pt–N bond cleavage may influence the results of kinetic analysis that are based on HPLC techniques, in which aliquots of the reaction mixture are collected at several time points and quenched with large amount of potassium chloride.


21

Figure 1.11 Experimental concentrations (NMR data) and theoretically-fitted curves for the reactions between [Pt(dien)]+ and DNA: (a) d(TATGGTACCATA)2; (b) d(TATGGATCCATA)2; (c) d(TATGGCCATA)2. Symbols: (ⵧ) Pt(dien); (o) Pt-G5′and (∆) Pt-G3′. (J. Vinje, J.A. Parkinson, P.J. Sadler, T. Brown, E. Sletten, Sequence-selective metalation of double-helical oligodeoxyribonucleotides with PtII, MnII and ZnII ions. Chem. Eur. J., 2003, 9, 1620–1630. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

Table 1.2 Rate constants (standard deviations in parentheses) for reactions of [PtCl(dien)]+ with DNA oligonucleotides. pH 5.6, 0.1 M NaClO410 Sequence

k5′ (M−1 s−1)

k3′ (M−1 s−1)

-GGTA-(a) -GGAT-(a) -GGCC-(b)

4.3 (6) 4.0 (6) 7.6 (6)

0.5 (1) 3.4 (5) (c)

Temperature: (a) 288 K; (b) 298 K; (c) Negligible 3′-G platination.

22


1.4 Summary of Theoretical and Experimental Evidence for Sequence–Selective Binding to DNA Paramagnetic and diamagnetic ions are shown to form both labile and nonlabile adducts with DNA duplexes, and in a series of G-containing duplexes a certain selectivity pattern emerges where the metal prefers the 5′-G in the following order: 5′-GG > 5′-GA > 5′-GT >> 5′-GC. The adjacent residue (Y) on the 5′-side (5′-YGG) is found to exert a negligible influence on the selectivity. For the monofunctional binding of [PtCl(dien)]+ to double helical DNA the variation in reaction rates follows qualitatively the same selectivity pattern as for the labile metal ions. Experimentally observed relative rates of G-oxidation are found to match well with the NMR results. Nonselective equal G cleavage is observed for single-stranded DNA, in line with the NMR experimental data. Ab initio molecular orbital calculations of stacked DNA bases with B-form geometry clearly indicated that the highest occupied molecular orbital (HOMO) localization on the 5′-G is highly sequencedependent. The degree of p–p interaction between the stacked bases influences the HOMO energies. However, it should be emphasized that for metal complexes the stereochemistry of the ligands also plays a significant role in determining the most favourable binding site.

1.5 Sequence-Specific Groove Binding Among the large number of contributions to the field of cation–groove interaction, probably a majority is based on X-ray crystallography. Sequence-specific nucleic acid conformation and dynamics are directly influenced by metal ions. DNA conformational heterogeneity has been explained by an electrostatic model where the local position and transient fluctuation of ions act through asymmetric neutralization of phosphate charges. Evidence from NMR spectroscopy, X-ray crystallography and molecular dynamics simulations has revealed that B-form duplexes interact in a sequence-specific manner with fully hydrated mono and divalent cations.55 1.5.1 Groove Geometry A noteworthy feature of B-DNA is the presence of two kinds of grooves, called the major groove (12 Å wide) and the minor groove (6 Å wide) (see figures in biochemistry text books). They arise because the glycosidic bonds of a base pair are not diametrically opposite each other. The minor groove contains the pyrimidine O2 and the purine N3 of the base pair, and the major groove is on the opposite side of the pair. In the minor groove, N3 and O2 can serve as hydrogen bond acceptors, and the amino group attached to guanine can be a hydrogen donor. In the major groove N7 is a potential hydrogen bond acceptor, as are O4 of thymine and O6 of guanine. The amino groups attached to adenine and cytosine, respectively, can serve as hydrogen donors.

Sequence-Specific Groove Binding 23

The width of the grooves is not constant, but is found to be sequencedependent. Alternative models have been presented to provide explanations for this sequence-dependent variation. NMR spectroscopic analyses of a series of DNA duplexes have shown that alternating (AT)n sequences are characterized by a rather wide minor groove.56,57 From a number of experimental and theoretical studies, it has been concluded that on average the minor groove of DNA A-tracts (several consecutive adenine residues) is significantly narrower than the minor groove of G-tracts.58–60 1.5.2 Monovalent Cations In the report of the first X-ray structure of a double-helical DNA duplex, [d(CGCGAATTCGCG)]2, published by Dickerson and Drew,58 it was proposed that a ‘spine of hydration’, composed of localized, geometrically arranged water molecules, in the minor groove, is an important structural component of A-tract DNA. However, this result is not fully conclusive since Na+ and H2O, being isoelectronic, are not easily distinguishable by X-ray diffraction. Egli and coworkers repeated the structural analyses of the same duplex using X-ray diffraction data to near-atomic resolution with crystals grown in the presence of Rb+ cacodylate.61 They found that a single Rb+ ion, with partial occupancy, was localized at the central ApT step at the bottom of the AATT minor groove (Figure 1.12). The ion replaces the water molecules that link the keto oxygen of thymines from opposite strands. The authors suggest that minor groove ion coordination appears to be an isolated event, highly sequence-dependent and unlikely to significantly affect the particular geometry of the A-tract in the Dickerson–Drew dodecamer. Further studies by Williams and coworkers on the same duplex using Tl+ to mimic K+ showed that none of the observed Tl+ sites surrounding the duplex were fully occupied.62 The

Figure 1.12 Coordination of Rb+ at the central ApT step in the Dickerson–Drew Duplex.61 (Reprinted from Chem. Biol., 9, 3, M. Egli, DNA-cation interactions: Quo vadis?, 10. Copyright 2002, with permission from Elsevier.)

24


most highly occupied sites (20–35%) were located within the G-tract major groove while the occupancy in the minor groove were estimated to be around 10%. The situation concerning penetration of Na+ ions into the spine of hydration is still an open question among X-ray crystallographers. The existence of just a single alkali metal ion coordination site in the Rb+-form crystal structure argues against a view that ions can invade the minor groove hydration spine along the entire length, or the existence of a mixed water-ion spine of hydration. In a recent high-resolution (1.1 Å) structure determination of the Dickerson–Drew duplex no experimental evidence for the presence of Na+ ions in the minor groove was found.63–65 A series of 23Na NMR quadrupolar relaxation studies have been carried out on B-DNA.66,67 A general conclusion from these studies is that monovalent counterion binding to DNA is loose and delocalized, without any dehydration or sequencespecific features. In an optimally designed test for sequence-specific Na+ binding in the minor groove Denisov and Halle have used a magnetic relaxation dispersion (MRD) technique, where the 23Na relaxation rate is measured over nearly two decades of resonance frequency.67 Comparison of Na+ MRD data from three dodecamers with different nucleotide sequences: CGCGAATTCGCG (abbreviated A2T2), CGAAAATTTTCG (A4T4) and CGCTCTAGAGCG (TA) showed that the most tightly bound Na+ ions reside in the minor groove. However, the occupancies are quite low corresponding to a binding constant KNa of 0.03 M−1 for TA and 0.1 M−1 for the other two dodecamers, and imply that Na+ binding in the minor groove is a rare event and is not likely to be detected by X-ray diffraction. These results are not necessarily inconsistent with higher occupancy at the cryogenic temperature (120–160 K) used in recent crystallographic studies. Even a modest binding enthalpy of 5 kJ mol−1 could increase the binding constant from 0.1 M−1 at 277 K to 1.7 M−1 at 120 K, which is sufficient to give 50% occupancy in a single binding site, as found for Rb+ at the ApT step in A2T2.61 The authors conclude that groove bound Na+ ions, with an occupancy of only a few percent at room temperature, are not likely to contribute importantly to the ensemble of DNA structures under physiological conditions.67

1.5.3 Divalent Cations At physiological concentrations the binding of divalent cations to DNA is both cation dependent and sequence dependent. From a sequence standpoint, specificity is contributed to by both the local molecular nucleophilicity (see above) and the hydrogen bond environment. These H-bond interactions reflect the greater hydration properties of divalent cations over monovalent. Generally, it is difficult to assess how contributions of base sequence or cation type influence groove specificity because of crystal packing effects. Chiu and Dickerson have examined a database of 28 cation-bound B-DNA structures spanning ten different crystal packing environments and showed that there is a correlation between experimental conditions and the number of observed cations.68 Hence, the locations of cations can be compared safely only between structures having similar crystallizing conditions, data collection methods and resolutions. Despite these difficulties, the authors find a

Sequence-Specific Groove Binding 25

Figure 1.13 Coordination of Mg2+ at the CpG step near one end of the Dickerson–Drew Duplex.63 (Reprinted from Chem. Biol., 9, 3, M. Egli, DNA-cation interactions: Quo vadis?, 10. Copyright 2002, with permission from Elsevier.)

strong correlation between divalent cation binding and base sequence for the 28 structures examined. For four high-resolution (0.99 Å) structures, the minor groove affinity for Mg2+ is GG > AG > AC and for Ca2+ is GG > AT ∼ AC, with cations positioned in the centre of the groove. For the major groove, the order for Mg2+ is GG > AG ∼ GT and for Ca2+ is GG > AG.68 Treshko et al.63 have redetermined the crystal structure of the Dickerson–Drew duplex at 1.1 Å resolution. Three ordered Mg2+ ions are present in the asymmetric unit, two hexahydrates and one pentahydrate complex. One Mg2+ is located in the major groove, close to the end of the duplex (Figure 1.13). The ion contacts the N7 and O6 edges of residue G2 and G22 from opposite strands via coordinated water molecules. None of the contacts between Mg2+ ions and DNA atoms on the floor of the major groove involve inner-sphere coordination.65 Hud and Feigon have studied the localization of Mn2+ in A-tract DNA by 1H NMR spectroscopy, using a series of self-complementary dodecamer oligonucleotides that contain the sequence motifs AnTn and TnAn, where n = 2, 3 or 4 flanked by 5′-CG or 5′-GC base pairs.69 At an Mn2+ to duplex ratio of 10−3 most of the aromatic base protons are severely broadened, however, in contrast to the Mn2+ NMR data referred to above, the adenine H2 resonances exhibit the largest broadening. The authors conclude that Mn2+ is localized in the minor groove with the position and degree of localization being highly sequence-dependent. In addition, G2-H8 in the sequence CGT is broadened by 7 Hz/µM MnCl2 while G1-H8 in the sequence GCT is broadened less than 2 Hz, a finding that is in agreement with the observation that the affinity of 5′-GT >> 5′-GC (see above).

26


NMR spectroscopy70 and molecular dynamics (MD)71 studies have shown that AnTn and TnAn duplexes have unusual structures and dynamics. Anomalous broadening of the TpA adenine H2 resonance indicative of large amplitude base motion, has been observed for a series of nine unique four-nucleotide sequences.70 In AnTn sequences, the DNA assumes a unique structure characterized by a gradual and increasingly compressed minor groove, which reaches a minimum at the ApT step. In conclusion, this type of sequences does not exhibit a regular B-form conformation and thus minor groove cation binding may be unique to A-tract sequences. 1.5.4 Conclusion on Groove Binding The general trend is that ions bind in the minor groove of DNA A-tracts and in the major groove of DNA G-tracts. NMR spectroscopic studies indicate that sequencespecific binding of alkali and alkali earth ions appears to have only minor influence on the heterogeneity of DNA structures.67. The relative lack of hydration of alkali cations makes their interactions with base atoms largely electrostatic and relatively nonspecific. Divalent alkali earth cations tend to be fully hydrated and their interactions with duplex DNA are more sequence-specific through formation of hydrogen bonds to base atoms.55,67

Abbreviations MEP HOMO LUMO ODN MD EDTA XANES NOESY MRD

Molecular electrostatic potential Highest occupied molecular orbital Lowest occupied molecular orbital Oxidation oligodeoxynucleotides Molecular dynamics Ethylene diamine tetraacetic acid X-ray absorption near edge structure Nuclear overhauser effect spectroscopy Magnetic relaxation dispersion

References 1. Flessel, C.P.; Furst, A.; Radding, S.B.; A comparison of carcinogenic metals. In: Metal Ions in Biological Systems. Vol. 10, H. Sigel, Ed., Marcel Dekker, New York, 1980. 2. Frieden, E.; Alles, J.; Subtle interactions of cupric ions with nucleic acid components; J. Biol. Chem., 1958, 230, 797–804. 3. Eichhorn, G.L.; Shin, Y.A.; Interaction of metal ions with polynucleotides and related compounds. XII. The relative effect of various metal ions on DNA helicity; J. Am. Chem. Soc., 1968, 90, 7323–7328. 4. Sletten, E.; Lie, B.; Copper complex of guanosine-5′-monophosphate; Acta. Cryst., 1976, B32, 3301–3304.

References

27

5. Yamane, T.; Davidson, N.; On the complexing of deoxyribonucleic acid (DNA) by mercuric ion; J. Am. Chem. Soc., 1961, 83, 2599–2607. 6. Frøystein, N.Å.; Sletten, E.; Interaction of mercury(II) with the DNA dodecamer CGCGAATTCGCG. A 1H and 15N NMR study; J. Am. Chem. Soc., 1994, 116, 3240–3250. 7. Frøystein, N.Å.; Sletten, E.; The binding of manganese(II) and zinc(II) to the synthetic oligonucleotide [d(CGCGAATTCGCG)]2; Acta Chem. Scand., 1991, 45, 219–225. 8. Frøystein, N.Å.; Davis, J.T.; Reid, B.R.; Sletten, E.; Sequence-selective metal ion binding to DNA oligonucleotides; Acta Chem. Scand., 1993, 47, 649–657. 9. Moldrheim, E.; Andersen, B.; Frøystein, N.Å.; Sletten, E.; Interaction of manganese(II), cobalt(II) and nickel(II) with DNA oligomers studied by 1H NMR spectroscopy; Inorg. Chim. Acta., 1998, 273, 41–46. 10. Vinje, J.; Parkinson, J.A.; Sadler, P.J.; Brown, T.; Sletten, E.; Sequence-selective metalation of double-helical oligodeoxyribonucleotides with PtII, MnII and ZnII ions; Chem. Eur. J., 2003, 9, 1620–1630. 11. (a) Laing, L.G.; Gluic, T.C.; Draper, D.E.; Stabilization of RNA structure by Mg ions. Specific and nonspecific effects; J. Mol. Biol., 1994, 237, 577–587. (b) Misra, V.; Draper, D.E.; A thermodynamic framework for Mg2+ binding to RNA; PNAS, 2001, 98, 12456– 12461. (c) Manning, G.S.; The molecular theory of polyelectrolyte solutions with applications to the electrostatic properties of polynucleotides; Q. Rev. Biophys., 1978, 11, 179–246. (d) Strehlow, H.; Rapid Reactions in Solution, VCH, Weinheim, 1992. 12. Black, C.B.; Cowan, J.A.; Quantitative evaluation of electrostatic and hydrogen-bonding contributions to metal cofactor binding to nucleic acids; J. Am. Chem. Soc., 1994, 116, 1174–1178. 13. Kohn, K.W.; Hartley, J.A.; Mattes, W.B.; Mechanisms of DNA sequence selective alkylation of guanine-N7 positions by nitrogen mustards; Nucleic Acids Res., 1987, 15, 10531–10549. 14. Monjardet, V.; Elizondo-Riojas, M.-A.; Chottard, J.-C.; Kozelka, J.; A combined effect of molecular electrostatic potential and N7 accessibility explains sequence-dependent binding of cis-[Pt(NH3)2(H2O)2]2+ to DNA duplexes; Angew. Chem. Int. Ed., 2002, 41, 2998–3001. 15. Fukui, K.; Recognition of stereochemical paths by orbital interaction; Acc. Chem. Res., 1971, 4, 57–64. 16. Hutter, M.; Clark, T.; On the enhanced stability of guanine-cytosine base-pair radical cation; J. Am. Chem. Soc., 1996, 118, 7574–7577. 17. Sugiyama, H.; Saito, I.; Theoretical studies of GG-specific photocelavage of DNA via electron transfer: significant lowering of ionization potential and 5′-localization of HOMO of stacked GG bases in B-form DNA; J. Am. Chem. Soc., 1996, 118, 7063–7068. 18. Saito, I.; Takayama, M.; Sugiyama, H.; Nakatani, K; Tsuchida, I; Yamamoto, M.; Photoinduced DNA cleavage via electron transfer: demonstration that guanine residues located 5′ to guanine are the most electron-donating sites; J. Am. Chem. Soc., 1995, 117, 6406–6407. 19. Nakamura, T.; Nakatani, K.; Saito, I.; DNA HOMO as a new landmark for nucleic acid properties. Ab initio calculations and experimental mapping; Nucleic Acids Res., 1999, 42, 119–120. 20. Saito, I.; Nakamura, T.; Nakatani, K.; Mapping of highest occupied molecular orbitals of duplex by cobalt-mediated guanine oxidation; J. Am. Chem. Soc., 2000, 122, 3001–3006. 21. Senthikumar, K.; Grozema, F.C.; Guerra, C.; Bickelhaupt, F.M.; Siebbeles, L.D.A.; Mapping the sites for selective oxidation of guanines in DNA; J. Am. Chem. Soc.; 2003, 125, 13658–13659. 22. Saito, I.; Nakamura, T.; Nakatani, K.; Yoshioka, Y.; Yamaguchi, K.; Sugiyama, H.; Mapping of the hot spots for DNA damage by one-electron oxidation: efficacy of GG doublets and GGG triplets as a trap in long-range hole migration; J. Am. Chem. Soc., 1998, 120, 12686–12687.

28


23. Cleveland, C.L.; Barnett, R.N.; Bongiorno, A.; Joseph, J.; Liu, C.; Schuster, G.B.; Landman, U.; Steric effects on water accessibility control sequence-selectivity of radical cation reactions in DNA; J. Am. Chem. Soc., 2007, 129, 8408–8409. 24. Behling, R.W.; Kearns, D.R.; Proton two-dimensional nuclear Overhauser effect and relaxation studies of poly(dA).cntdot.poly(dT); Biochemistry, 1986, 25, 3335–3346. 25. Kowalewski, J.; Nordenskiöld, L.; Benetis, N.; Westlund, P-O.; Theory of nuclear spin relaxation in paramagnetic systems in solution; Prog. NMR Spectrosc., 1985, 17, 141–185. 26. Navon, G.; Valensin, G.; Nuclear relaxation times as a source of structural information. In: Metal Ions in Biological Systems, Vol. 21. H. Sigel, Ed. Marcel Dekker, New York, 1987. 27. Buncel, E.; Boone, C.; Joly, H.; Metal ion-biomolecule interactions. Part 13. NMR evidence for the formation of the mixed ligand thymidine-mercury-guanosine complex. A model for a putative Hg(Il) interstrand cross-linking structure of DNA; Inorg. Chim. Acta, 1986, 125, 167–172. 28. Young, P.R.; Nandi, U.S.; Kallebach, N.R.; Binding of mercury(II) to poly(dA-dT) studied by proton NMR; Biochemistry, 1982, 21, 62–66. 29. Drew, H.; Dickerson, R.E.; Structure of a B-DNA dodecamer. III. Geometry of hydration; J. Mol. Biol., 1981, 151, 535–556. 30. Nerdal, W.; Hare, D.R.; Reid, B.R.; Solution structure of the Eco-RI DNA sequence: refinement of NMR-derived distance geometry structures by NOESY spectrum backcalculation; Biochemistry, 1989, 28, 10008–10021. 31. Hare, D.R.; Reid, B.R.; Three-dimensional structure of a DNA hairpin in solution: twodimensional NMR studies and distance geometry calculations on d(CGCGTTTTCGCG); Biochemistry, 1986, 25, 5341–5350. 32. Eichhorn, G.L.; Complexes of nucleosides and nucleotides. In: Eichhorn, G. L., Ed. Inorganic Biochemistry, Elsevier, Amsterdam, 1973, Vol. 2, Chap. 33, 1192–1209. 33. Snygg, Å.S.; Ericson, A.; Elmroth, S.K.C.; Nonuniform rate of platination of guanine-N7 located in short DNA oligomers; Chem. Commun., 2001, 1190–1191. 34. Snygg, Å.S.; Brindell, M.; Stochel, G.; Elmroth, S.K.C.; A combination of access to preassociation sites and local accumulation tendency in the direct vicinity of G-N7 controls the rate of platination of single-stranded DNA; Dalton Trans., 2005, 1212–1227. 35. Olmsted, M.C.; Anderson, C.F.; Record, M.T.; Monte Carlo description of oligoelectrolyte properties of DNA oligomers: range of the end effect and the approach of molecular and thermodynamic properties to the polyelectrolyte limits; Proc. Natl. Acad. Sci., 1989, 86, 7766–7770. 36. Vinje, J.; Sletten, E.; Internal versus terminal metalation of double-helical oligodeoxyribonucleotides; Chem. Eur. J., 2006, 12, 676–688. 37. Labiuk, S.L.; Delbaere, L.T.J.; Lee, J.S.; Cobalt(II), nickel(II) and zinc(II) do not bind to intra-helical N(7) guanine positions in the B-form crystal structure of d(GGCGCC); J. Biol. Inorg. Chem., 2003, 8, 715–720. 38. Eisinger, J.; Schulman, R.G.; Szymanski, B.M.; Transition metal binding in DNA solutions; J. Chem. Phys., 1962, 36, 1721–1729. 39. Bertoncini, C.; Meneghini, R.; Cruz, D.Z.; Alves, M.C.M.; Tolentino, H.; Studies of Fe(II) and Fe(III)-DNA complexes by XANES spectroscopy. J. Synchtron Rad., 1999, 6, 417–418. 40. Henle, E.S.; Han, Z.; Tang, N.; Rai, P.; Luo, Y.; Linn, S.; Sequence-specific DNA cleavage by Fe2+-mediated Fenton reactions has possible biological implications; J. Biol. Chem., 1999, 274, 962–971. 41. Rai, P.; Cole, T.D.; Wemmer, D.E.; Linn, S.; Localization of Fe2+ At an RTGR sequence within a DNA duplex explains preferential cleavage by Fe2+ and H2O2; J. Mol. Biol., 2001, 312, 1089–1101. 42. Rai, P.; Wemmer, D.E.; Linn, S.; Preferential binding and structural distortion by Fe2+ at RGGG-containing DNA sequences correlates with enhanced oxidative cleavage at such sequence; Nucl. Acid. Res., 2005, 33, 497–510.

References

29

43. Saretzki, G.; von Zglinicki, T.; Replicative aging, telomeres and oxidative stress; Ann. N.Y. Acad. Sci., 2002, 959, 24–29. 44. Kawanishi, S.; Hiraku, Y.; Oikawa, S.; Mechanism of guanine-specific DNA damage by oxidative stress and its role in carcinogenesis and aging; Mutant. Res., 2001, 488, 65–76. 45. Abrescia, N.G.A.; Malinina, L.; Fernandez, L.G.; Huybh-Dinh, T.; Neidle, S.; Subirana, J.A.; Structure of the oligonucleotide d(CGTATAATACG) as a site-specific complex with nickel ions; Nucl. Acid. Res., 1999, 27, 1593–1599. 46. Abrescia, N.G.A.; Huybh-Dinh, T.; Subirana, J.A.; Nickel-guanine interactions in DNA: crystal structure of nickel-d[CGTGTACACG]2; J. Biol. Inorg. Chem., 2002, 7, 195–199. 47. Emwas, A.H.M.; Sletten, E.; NMR studies of interaction between Cu2+ and Cu(en)2 with the Dickerson–Drew duplex [d(CGCGAATTCGCG)]2. Manuscript (to be published). 48. Oikawa, S.; Kawanishi, S.; Site-specific DNA damage at GGG sequences by oxidative stress may accelerate telomere shortening; FEBS Lett., 1999, 453, 365–368. 49. Jia, X.; Zon, G.; Marzilli, L.G.; Multinuclear NMR investigation of zinc(2+) binding to a dodecamer oligodeoxyribonucleotide: insights from carbon-13 NMR spectroscopy; Inorg. Chem., 1991, 30, 228–239. 50. Fichtinger-Schepman, A.M.; Van der Veer, J.L.; den Hartog, J.H.J.; Reedijk, J.; Adducts of the antitumor drug cis-diamminedichloroplatinum(II) with DNA: formation, identification, and quantitation; Biochemistry, 1985, 24, 707–713. 51. Legendre, F.; Monjardet-Bas, V.; Kozelka, J.; Chottard, J.C.; A complete kinetic study of GG versus AG platination suggests that the doubly aquated derivatives of cisplatin are the actual DNA binding species; Chem. Eur. J., 2000, 6, 2002–2010. 52. Murdoch, P.D.; Guo, Z.J.; Parkinson, J.A.; Sadler, P.J.; Kinetics of formation and stability of {Pt(dien)}2+ complexes with octamer and 14-mer DNA oligonucleotides containing a GG sequence; J. Biol. Inorg. Chem., 1999, 4, 32–38. 53. Davis, M.S.; Berners-Price, S.J.; Hambley, T.W.; Rates of platination of AG and GA containing double-stranded oligonucleotides: insights into why cisplatin binds to GG and AG but not GA sequences in DNA; J. Am. Chem. Soc., 1998, 120, 11380–11390. 54. Dalbies, R.; Payet, D.; Leng, M.; DNA double helix promotes a linkage isomerisation reaction in trans-diamminedichloroplatinum(II)-modified DNA; Proc. Natl. Acad. Sci., 1994, 91, 8147–8151. 55. Hud, N.V.; Polak, M.; DNA-cation interactions: the major and minor grooves are flexible ionophores; Curr. Opin. Struct. Biol., 2001, 11, 293–301. 56. Chuprina, V.P.; Lipanov, A.A.; Fedoroff, O.Y.; Kim, S-G.; Kintanar, A.; Reid, B.R.; Sequence effects on local DNA topology; Proc. Natl. Acad. Sci., 1991, 88, 9087–9091. 57. Chuprina, V.P.; Sletten, E.; Fedoroff, O.Y.; Investigation of solution structure of d(GAATTTAATTC)2 by 1H NMR, molecular dynamics, mechanics, refinement by backcalculation of the NOESY spectrum and analysis of this structure using X-ray data; J. Biomol. Struct. Dyn., 1993, 10, 693–707. 58. Dickerson, R.F.; Drew, H.R.; Structure of a B-DNA dodecamer.II. Influence of base sequence on helix structure. J. Mol. Biol., 1981, 149, 761–786. 59. Burkhoff, A.M.; Tullius, T.D.; The unusual conformation adopted by the adenine tracts in kinoplast DNA; Cell, 1987, 48, 935–943. 60. Alexeev, D.G.; Lipanov, A.A.; Skuratovskii, I.Y.; Poly(dA)-poly(dT) is a B-type double helix with distinctively narrow minor groove; Nature, 1987, 325, 821–823. 61. Tereshko, V.; Minasov, G.; Egli, M.; A ‘hydrated’ spIne in a B-DNA minor groove; J. Am. Chem. Soc., 1999, 121, 3590–3595. 62. Howerton, S.B.; Sines, C.C.; VanDerveer, D.; Williams, L.D.; Locating monovalent cations in the groove of B-DNA; Biochemistry, 2001, 40, 10023–10031. 63. Tereshko, V.; Minasov, G.; Egli, M.; The Dickerson–Drew B-DNA dodecamer revisited at atomic resolution; J. Am. Chem. Soc., 1999, 121, 470–471. 64. Tereshko, V.; Minasov, G.; Egli, M.; Atomic-resolution crystal structure of B-DNA reveal specific influence of divalent metal ions on conformation and packing; J. Mol. Biol., 1999, 291, 83–99. 65. Egli, M.; DNA-cation interactions: Quo vadis? Chem. Biol., 2002, 9, 277–286.

30


66. Halle, B.; Denisov, V.P.; Water and monovalent ions in the minor groove of B-DNA oligonucleotides as seen by NMR; Biopolymers, 1998, 48, 210–233. 67. Denisov, V.P.; Halle, B.; Sequence-specific binding of counterions to B-DNA; Proc. Natl. Acad. Sci., 2000, 97, 629–633. 68. Chiu, T.K.; Dickerson, R.E.; 1 Å crystal structures of B-DNA reveal sequence-specific binding and groove-specific bending of DNA by magnesium and calcium; J. Mol. Biol., 2000, 301, 915–945. 69. Hud, N.V.; Feigon, J.; Characterization of divalent cation localized in the minor groove of AnTn and TnAn DNA sequence elements by 1H NMR spectroscopy and managnese(II); Biochemistry, 2002, 41, 9900–9910. 70. McAteer, K.; Ellis, P.D.; Kennedy, M.; The effects of sequence context on base dynamics at TpA steps in DNA studied by NMR; Nucleic Acids Res., 1995, 23, 3962–3966. 71. Leporc, S.; Mauffret, O.; Tevanian, G.; Lescot, E.; Monnot, M.; Fermandjian, S.; An NMR and molecular modelling analysis of d(CTACTGCTTTAG).d(CTAAAGCAGTAG) reveals that the particular behaviour of TpA steps is related to edge-to-edge contacts of their base-pairs in the major groove; Nucleic Acids Res., 1999, 27, 4759–4767.

2 Thermodynamic Models of Metal Ion–DNA Interactions Vasil Bregadze, Eteri Gelagutashvili and Ketevan Tsakadze

2.1 Introduction Recently the catalytic properties of DNA became of special interest to scientists. In particular, attention has been drawn to photo-processes1–7 and processes connected with charge8–15 and energy16–18 transfers. On one hand, transition- (G ↔ A, C ↔ T) and transversion- (A ↔ T, A ↔ C, G ↔ C, G ↔ T) type gene or point mutations, in rare cases, may lead to improvement of one or another characteristic of the organism. They are essential steps for artificial or natural selection in biology. On the other hand, more commonly, these mutations cause transformation of the cells, or their death19–20. This fact finds its application in medicine, specifically in chemotherapy and photo-chemotherapy. The mutations that lead to death of the cells are initiated by anticancer drugs, which often contain metals.4–7,21–25 The structure of the DNA double helix offers carte blanche for transition-type mutations via defects, such as wrong Watson–Crick pairs, where self-congruent cooperative keto–enol and amino–imine tautomeric transitions, named doubleproton transfer (DPT), take place.17 This is a property of cyclic structures in polar molecules. The nucleotide pairs in DNA naturally present such cyclic structures between complementary bases. The mechanism of DPT in a DNA duplex was first considered by Löwdin as early as in 1963.26,27 It is assumed that approximately one Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

32

Thermodynamic Models

− Table 2.1 Reaction rate constants of eaq with H3O+ and transition metal ions. (Values of Ke/aq 9 −1 −1 are given in 10 m s .)

Mn+

H+

Mn2+

Co2+

Ni2+

Ni(en)2+

Ni(en3)2+

Cu2+

Cu+

Zn2+

Ag+

Ke/aq

24

0.038

12

22

22.4

0.02

30

–

1.5

36

DPT takes place spontaneously in one out of 104 bp. Transition metal ions are known to interact with DNA (depending on anion charge, G-C composition of DNA, ionic strength etc.28), the effect of which is to increase the number of mispaired Watson– Crick base pairs. The interaction of H3O+ and transition metal ions with DNA depends on both the electronic structure of the ions and the structural, polyelectrolytic and dynamic characteristics of DNA in water solution. − Table 2.1 presents the reaction rate constants of hydrated electrons (eaq ) with 29,30 different types of ions. The reactivities of the ions presented in the table actually reflect the affinities of the hydrated electron with these ions in water solutions31 and, as will be shown in Section 2.3.3, they correlate well with the results of UV spectra of DNA–metal ion complexes. Native double helix DNA is a heavily charged polyanion (2e− charge per base pair), which in solution exists in the form of an ordered closely packed structure, as proved by the strong hyperchroism displayed upon melting of the double helix. Quantum chemical evaluations show an identical sequence of arrangement of electron-donor atoms for unscreened B-DNA and for that screened by Na+ ions:32–33 G-N7; G-N3; A-N3; T-O2; G-O6; A-N7 The NH2 group of adenine and guanine prevents interaction of metal ions with A-N7 and G-N3 due to steric crowding. In addition, G-N3, A-N3 and T-O2 are located in the minor groove, the width and the depth of which are 5.7 Å and 7.5 Å, respectively, in the B-form. For steric reasons all these atoms are unreachable for a hydrated transition metal ion, which has a diameter of approximately 7.5 Å. However, the omnipresent H3O+ ion can interact with electron-donor groups from both of the grooves. Interaction of ions with DNA also depends on ionic strength, GC composition, the order of the nucleotides in the helix and the conformational parameters of DNA (the angle of the coil twist, the propeller angle, the distance between adjacent phosphorus atoms of the same chain etc.). Nevertheless, one may say with a certain assurance that dynamic characteristics of DNA are the most significant factors governing metal ion–DNA interactions. First of all, this is important for modelling e.g. B→Z transitions, mechanisms of transcription, replication, carcinogenesis and mutagenesis. In this respect it is important to describe the characteristics of inner mobility in DNA, e.g. various small-amplitude oscillating movements of atoms, ′ inside the components of DNA (10−14–10−13 s); limited movements of about 0.1 Å phosphates, sugars and nucleobases around an equilibrium position; torsion and flexural oscillations of the double helix (10−10–10−8 s); large-amplitude movements of phosphates, sugars and bases occurring at the transition of DNA double helix from

Interactions of Metal Ions with DNA 33

one form to another (10−7–10−5 s).34 Thus for modelling of DNA structural changes stimulated by interactions with various transition metal ions it is necessary to estimate the correlation between the energy of interaction and the lifetime of the complexes formed.

2.2 Interactions of Metal Ions with DNA 2.2.1 Thermodynamic Adsorption Model of Complex Formation The catalytic characteristics of DNA are closely related to structural changes caused by interactions with different species, particularly with metal ions. Evidently, transition metal ions can locally influence the structure of DNA only when the lifetimes of the complexes are commensurable with the specific times of inner mobility in DNA (oscillation of small groups of atoms, untwisting of the double helix, opening of the individual base pairs, untwisting of helix binding of proteins and cell division). The movements last from 10−10 s to hundreds of seconds and longer. We have managed to find a correlation between dynamics and stability28 by applying Frenkel’s phenomenological thermodynamic approach,35–37 which was used as early as 1924 for the study of gas adsorption on the surface of solids. Frenkel introduced a new term ‘lifetime’ (t) for the adsorption state and connected it to the energy of interaction |∆E| between adsorbate and adsorbent surface by the following expression:

τ = τ 0a−s ⋅ exp ( ∆E kBT )

(2.1)

where τ 0a−s is the time of a single oscillation of the adsorbate on the surface, assumed to be 10−13–10−12 s and kB is the Boltzmann constant.36 The above model can, with good approximation, be applied to the study of interactions between small molecular species and biomacromolecules in solution. Using the relationship DG = −RT lnK, a simplified expression of (2.1) can be derived:

τ = τ0 ⋅ K

(2.2)

where t is the lifetime of a metal ion–macromolecule complex. Now we must clarify the value t0 for solutions. Frenkel’s τ 0a−s is the duration of the fluctuation excitation of adsorbing atoms or molecules interacting with a solid surface. It is supposed to be equal to the period of oscillation of the adsorbate relative to the adsorbent surface. In solutions, t0 describes the duration of relaxation of rotary and translation movements of the solvent molecules, ions, solvated ions or low-molecular-weight substances and lies between 10−11and 10−10 s. Thus, if we assume a pK in the range 4–6 for binding of DNA with two-fold positively charged metal ions of the first transition series, and t0 = 10−11 s, then the life span of these complexes is in the range 10−7–10−5 s.

34


The expression t = t0 · K allows us to correlate the stability constant of complex formation to the lifetime of the complex. Thus, the principal concept of molecular biophysics regarding biomolecules: structure – dynamics – function can be reformulated as: structure – stability – function. One may notice that such an approach highly simplifies and widens the time interval (from 10−10 s to 105–106 s and more) for investigating dynamic characteristics of macromolecules. A detailed description of the derivation of formulae (2.2) is given in reference28. 2.2.2 Stability Constants of DNA–M2+ Complexes Stability Constants Determined by Graphical Methods Based on Adsorption Isotherms Different methods suitable for calculation of stability constants have been developed, for the most part based on linearization procedures of the law of mass action.38,39 The improvement of computer technology led to the development of nonlinear curve fitting algorithms;40,41 among these the Scatchard and Klotz methods have been widely used to analyze experimental data.42,43 Let us describe, as an example, various models of Pb2+ binding to DNA using modified Scatchard plots and two types of the Klotz graphical representation. Figure 2.1A shows the Scatchard plot. The intersection of the curves with the abscissa gives the value n = n1 + n2 (n is the number of binding sites for Pb2+ per phosphate group of DNA at saturation). The intersection of the curves with the ordinate gives the value of the binding constant (K): Kn = k1n1 + k2n2.43 The parameters are calculated by the nonlinear least-squares fit to the equations: r m = 0.5  B ( r ) + B2 ( r ) + 4C ( r )  B ( r ) = k1n1 + k2 n2 − ( k1 + k2 ) r C ( r ) = k1k2 r ( n1 + n2 − r ) Figure 2.1B illustrates the Klotz method with intercepts a1, a2, a3, a4 for variables r/m and r, where two independent classes of binding sites are assumed. a1 = n1k1 + n2 k2 a = n + n 1 2  2 System:  2 2 2 a3 ( n1k1 + n2 k2 ) = a1 ( n1k2 + n2 k1 ) a4 = a2 2 k1k2 It is seen from Table 2.244 that the parameters determined by the Klotz and modified Scatchard analysis methods are almost identical. Consequently, it can be assumed that both models can be used to determine binding constants of Pb2+ to DNA. Based on the c2 test, the Scatchard method (c2 = 0.81) seems to give a better estimate than the Klotz method (c2 = 3.08).45

Interactions of Metal Ions with DNA 35 1.2

1.2

1.0

1.0

Pb(II)

A

0.6

0.6 0.4

0.4

0.2

0.2 0.0 0.0

B

0.8 r/m × 104

r/m × 104

0.8

a1

a4 a3

0.0 0.1

0.2

r

0.3

0.4

0.5

0.0

0.1

a2 0.2

r

0.3

0.4

0.5

Figure 2.1A: The Scatchard plot for binding of Pb(II) with DNA (2 mM NaNO3, 20 °C) as non-linear least square fit r m = 0.5 B (r ) + B2(r ) + 4C (r )  , where B(r) = k1n1 + k2n2 − (k1 + k2)r and C(r) = k1k2r(n1 + n2 − r). Figure 2.1B: The Klotz plot for binding of Pb(II) with DNA (2 mM NaNO3, 20 °C). Calculated values (k1, n1, k2, n2) were obtained from the Klotz equation system.

Table 2.2 Parameters of Pb2+ binding with DNA in 2 mM NaNO3, 20 °C Pb

Stability constant K × 104, M−1 Free Gibbs energy −∆G0, kcal mol−1 Number of binding sites at saturation n Microconstant k1 × 104, M−1 Number of binding sites n1 Microconstant k2 × 104, M−1 Number of binding sites n2 c2 distribution

Analysis by the Scatchard method

Analysis by the Klotz method

2.4 ± 0.5 5.96 0.48 ± 0.05 21.1 ± 0.7 0.043 ± 0.002 0.56 ± 0.01 0.44 ± 0.03 0.81

2.3 ± 0.2 5.94 0.50 ± 0.06 24.7 ± 1.2 0.034 ± 0.001 0.68 ± 0.01 0.47 ± 0.01 3.08

The stability constants (log K) for Mg2+, Cr3+, Mn2+, Co2+, Ni2+, Cu2+, Zn2+, Ag+, Cd and Pb2+ binding to DNA are in the range 4–6 (Table 2.3), and assuming a t0 value of 10−11 s, the lifetimes t fall in the range 10−7–10−5 s. For DNA this time corresponds to the movements of phosphates, sugars and bases with large amplitudes (see introduction). These movements are related to conformational changes and untwisting of the double helix. As metal ions of the first transition series form chelates with DNA, the stabilities of the complexes may be estimated from the DNA 2+

36


Table 2.3 Binding constants of metal ions with ctDNA (20 mM Na+). Metal ions

Cr3+

Co2+

Ni2+

Cu2+

Zn2+

Ag+

Cd2+

Pb2+

Log K

3.85

4.3

4.64

5.1

4.2

5.84

4.1

4.25

Ag+ Pb2+

1.0

Cu2+ Ni2+ Co2+ −Cr3+ Zn2+ Cd2+ H+

0.9

LogK

0.8 0.7 0.6

Mg2+ 0.5 ϕ

0.4 M-DNA 0.3 −3.0

−2.5

−2.0

−1.5

−1.0

−Log[Na(1)]

Figure 2.2 Efficacy of Na+ concentration on the stability constants (log K) for ctDNA complexes with Mg2+ 43 Co2+, Ni2+, Cu2+, Zn2+, Cd2+, Pb2+, Cr3+, Ag+ and ϕ, which is the electrostatic potential of a cylindrical polyion surface (r = 1 nm) obtained from the Poisson–Boltzman equation44 reduced to unity

dynamics. For example, the difference in pK between Co2+ and Cu2+ binding to ctDNA is only 0.8 (Table 2.3) compared to the much larger differences for binding of Co2+ and Cu2+ to bidentate ligands like NH2CH2CO2− and ethylendiamine (en) where the pK varies from 4.6 to 8.6 and from 6 to 11, respectively.46 Thus, when M2+ ions interact with DNA, the dynamic properties of DNA connected with big amplitude movements, e.g. unwinding the double helix and opening of base pairs, determine to a large extent the stability constants for complex formation. Effect of Ionic Strength Figure 2.2 is a graphic representation showing how pK varies with ionic strength for H+ and metal ions (Cr3+, Co2+, Ni2+, Cu2+, Zn2+, Ag+, Cd2+, Pb2+) binding to DNA. An NaCl salt solution was used for all metal ions except Ag+ and Pb2+ where NaNO3 was used. For Mg2+ the binding constants with poly(A-U) are given in reference.47


In addition, the dependence of the surface electrostatic potential (ϕ) of a cylindrical polyion (r = 1 nm) is plotted as a function of the ionic strength of the solution as obtained from the Poisson–Boltzmann equation.48 For Mg2+ the pK dependence and the surface potential are closely related, indicating a nearly exclusive binding of Mg2+ to the negatively charged phosphate groups of DNA. One may notice that the binding constants of Ag+ and Pb2+ are only weakly dependent on ionic strength. This is in agreement with the expected inner-sphere complexation of Ag+ ions involving nitrogen atoms of the nucleobases (soft binding). Divalent lead is one of the post-transition metal elements that exhibits the so-called ‘inert-pair effect’. This lone pair can cause a non-spherical charge distribution around the Pb2+ cation. ‘Stereochemically active lone pairs’ can take various forms in Pb2+ compounds.49 Recent data show that Pb2+ forms hemidirect coordination compounds both in complexes with water and in complexes with O and N atoms of ligands,50 which, in turn, are reactive centres in DNA for other metal ions. The stereochemistry of Pb2+ may be the reason of such a low-effect action of ionic strength. The log K dependencies for the transition metal ions Cr2+, Co2+, Ni2+, Cu2+ and 2+ Zn are in between the two extremes mentioned above. The slopes of their lines allow determination of the ‘softness’, or ‘transition degree’, of these ions assuming that the slope angle of the lines is proportional to this property. Effect of DNA G-C Content on Metal Complexation To quantify the effect of DNA heterogeneity upon metal ion binding, the values of pK for Ni2+ complexes with DNA of various origins (T7 phage, calf thymus, bovine spleen, salmon sperm, Spirulina platensis, E.coli) have been determined.51,52 In Figure 2.3

5.4 Ni(II)-DNA

E.coli T7

5.2 Spirulina platensis LogK

5.0

4.8 Bovine spleen

Calf thymus Salmon sperm Mice liver

4.6

40

42

44

46

48

50

52

(GC) %

Figure 2.3 Dependence of Log K versus G-C content

38

Thermodynamic Models 4.0 Ag+

M-DNA 20 mM Na+

3.5 Pb2+ 3.0 X m2 r

Cu2+ Co2+

2.5

2.0

Cd2+

Ni2+ y = −1.31 − 0.85x; R 0.92 SD 0.05; N7; p < 0.0001;

1.5 −6.0

−5.5

−5.0 pK

Zn2+ Cr3+ −4.5

−4.0

−3.5

Figure 2.4 Correlation between the covalence index Xm2r and − log K = pK for M–DNA complexes

pK for Ni2+ binding to different species is shown to increase with G-C content. The empirical dependence of pK versus G-C content has been derived: pK = 1.96 + 0.07 (% G-C). The observed trend to form stronger complexes with increasing G-C content may be due to the high probability of G-N7/Ni2+ complex formation. A linear correlation has been observed between log metal/poly d(GC), assessed at the midpoint of the structural transition curve, and the covalent index ( X m2 r ) 53,54 (Xm being the Pauling electronegativity and r the effective ionic radius55). The authors propose that the mid-phase mole ratio is inversely proportional to the stability constant for the metal ion–polydeoxynucleotide interaction. Such dependence between binding constants and X m2 r was also obtained in our case. The linear correlation of pK (for complexes of metal–ctDNA) and the covalent index X m2 r presented in Figure 2.4 suggests that the covalent contributions to the overall interaction energy are significant and demonstrates the dynamic nature of metal–DNA interactions (except in the case of Pb2+). These results are in good agreement with data in references 53 and 54. 2.2.3 UV Spectroscopic Study of Proton and Metal–DNA Interactions Table 2.4 shows spectral changes of UV absorption spectra of ctDNA and that of poly(dG-dC) · poly(dG-dC) caused by different ligands and macroscopic changes of the surrounding medium. Dn in the table is calculated from the expression Dn = Des/K, where Des = |De(nmax)| + |De(nmin)| is the total intensity of ultraviolet difference spectrum, K = |e(1)(nmax)| + |e(1)(nmin)|; e(1) is the first derivative of e(ν) function. Dn for d(pG) + d(pC) + H3O+ (pH = 2), H3O+, Cu+ and Ag+ were determined by absorption spectra of DNA complexes (more precisely this issue is observed in reference 17). Analyzing this acid bathochromic shift (the first line in the table) and

Interactions of Metal Ions with DNA 39 Table 2.4 Bathochromic shift (∆u) of UV absorption spectra of ctDNA and poly(dG-dC)·poly(dGdC) caused by different metal ions and H3O+ Ions

d(pG) + d(pC) + H3O+ H3O+ Mg2+ Mn2+ Co2+ Ni2+ Ni2+ + 40% ethanol Ni[(en)(H2O)24+ ] Ni[(en)(NO2−)(NCS−)]0 Ni2+ + EtBr Z-DNA poly(dGdC)·poly(dG-dC) 25% ethanol, 102 NaCl, (0.75 Ni2+/P) Cu2+ Cu+ Zn2+ Ag+

Bathochromic shift of the absorption band (∆u(cm−1)) ∼700 295 15 65 110 150 185 75 35 180 ∼700

190 ∼700 85 ∼500

Change of the Widening of the DPT formation absorption absorption band energy in G-C pair U intensity (%) (cm−1) and special characteristics (kcal mol−1) Insignificant 2.5↓

Insignificant

– 0.5 2.4 1.4 1.1 0.9

140 (1/2h)

0.8 0 1.3 0.2

1.5↓

7.4↓ 10↓

1000 (2/3h)

taking into account absorption of adenine and thymine, the amount of which totals 60% in ctDNA, we can roughly accept that the joint effect of the shift from protonation of guanine and cytosine is ∼700 cm−1 per 1 mol of G-C pair. Correlation Between Du and Ke/aq Interactions of aqua ions with the DNA double helix have a discrete nature due to steric requirements characterized as inner-sphere, outer-sphere or double outersphere interactions. The latter refers to interaction of hexa-aqua ions with the hydration layer of DNA. One may also take into account so-called ‘atmospheric’ interactions which involve mobile adsorption within the effective diameter of DNA depending on the ionic strength.56 UV spectroscopy is a sensitive method for detection and registration of the mentioned interactions. As for thermodynamic methods, they reflect the integral characteristics of these interactions. That is why stability constants (pK) are less dependent on G-C composition, ionic strength and the type of ions. The soft structures of transition metal–DNA interactions are also caused by the fact that each of the six water molecules surrounding the ions may exist in two positions (two nonbonding pairs of electrons in the water molecule are localized on ′ ).57 the sp3 hybrid orbitals of the oxygen and the distance between them is ~ 0.6 Å Thus, the water molecules located in a six-coordinated sphere of hydration can exist in 26 different states.

40

Thermodynamic Models 500

Ag+

400

*Cu+ H+

∆ν

300 Cu2+

Ni2+ + ethanol

200

100

Mn2+ Mg2+

0 0.0

Ni2+

Zn2+ Co2+

Ni(en)2+

Ni[(en)(NO2)(NCS−)]0 0.4

0.8

1.2

1.6

2.0

1/∆Ga

Figure 2.5 Correlation between ∆ν of ctDNA with different metal ions and 1/∆Ga. ∆Ga refers to the reaction of the hydrated electron with the ion under study and is related to the rate ∆Ga . The vertical factor for Cu+ is ×2 constant by the expression K e aq = 1011 exp − RT

(

)

More details follow from Figure 2.5, which exhibits the correlation between the shift of the absorption band (Dn) of DNA bound to the investigated ions and the reaction rate constants (Ke/aq) of hydrated electrons with the same ions, particularly with the 1/DGa values (Table 2.1). DGa is the hydrated electron activation energy in the expression: Ke/aq = 1011exp(−DGa/RT).29,30 At I = 0.01 M, a linear relationship between these values seems to exist only for Mn2+, Co2+, Ni2+ and Zn2+ while Mg2+, 0 Cu2+, H+, Ag+ and Cu+ deviate from the line. So do Ni(en)2+, Ni [( en)2( NO−2 )( NCS− )] 2+ and Ni (40% ethanol). From the UV, visible and near IR spectroscopic studies it may be concluded that Mn2+, Co2+, Ni+2 and Zn+2 form outer-sphere complexes with DNA. The small influence of these divalent ions on the DNA absorption spectrum and the correlation of Dn with 1/DG suggest that their interactions with the DNA nucleobases in solution take place through hydrogen bonding involving their hydrated coating. A 31P NMR spectroscopy study of the interaction between DNA and the paramagnetic ions Mn2+ and Co2+ showed that only 15 ± 5% of the total bound ions was directly coordinated to the phosphate groups (inner-sphere complexes). The remaining Mn2+ and Co2+ ions were bound either as outer-sphere complexes relative to the phosphate groups or elsewhere to the DNA, possibly to the bases.58 These results are in agreement with X-ray crystallographic data, which show that transition metals may coordinate to mononucleotides by direct nucleobase Nbinding and/or phosphate binding.59,60 The UV data indicate that Ag+, H+, Cu+ and, partially, Cu2+, interact directly with DNA without participation of water molecules. 0 Ions of Mg2+, Ni(en)2+ and Ni [( en)2( NO−2 )( NCS− )] , which stay below the line (Figure 2.5), have the lowest affinity to DNA nucleobases. In the case of direct binding of metal to nucleobases in DNA there is a certain probability of electron tunnelling transitions from the nitrogen bases to the metal


ion. As an example, for the DNA–Cu2+ complex a total electron transition of 0.2 quantum output is observed under UV irradiation.31 Thus, H+, Cu2+, Cu+ and Ag+ can form inner-sphere interactions with DNA while the barrier mechanism is typical for Mn2+, Co2+, Ni2+ and Zn2+. Hence, we may conclude that ions such as Mn2+, Co2+, Ni2+ and Zn2+, as well as most of the Cu2+ ions form outer-sphere complexes with DNA. Experiments with Ni2+ indicate that these ions prefer the GC/CG and CG/GC type base pairs causing a red shift of the DNA absorption spectrum. Influence of GC Composition The influence of the DNA nucleotide composition on the ultraviolet difference spectrum (UDS) of DNA–metal ion complexes has been investigated for Co2+, Ni2+, Cu2+ and Zn2+. Most attention has been paid to Ni2+ and its interactions with DNA of various origins: Clostridium perfringens (27% GC), mouse liver (C3HA line, 40% GC), calf thymus (40% GC), salmon sperm (44% GC), herring sperm (44% GC), E. coli (51% GC), Micrococcus luteus (72% GC). In addition, interaction of Ni+2 with the oligonucleotides poly(dG-dC).poly(dG-dC), poly(dG).poly(dC) and poly(dA-dT).poly(dA-dT) have been studied. These results are shown in Figure 2.6 where Des versus the GC percentage is plotted. The Des values are given for the relative concentration of 0.25 Ni+2 per nucleotide unit. Analysis of the data for the interaction of Ni+2 with DNA shows preferential binding to GC pairs. Ni2+-Induced B → Z Transition in poly(dG-dC) · poly(dG-dC) Polynucleotide Figure 2.7 presents the absorption spectra of poly(dG-dC)·poly(dG-dC) in the presence of 0.25% ethanol solution (10−2 M NaCl without (1) and with (2) Ni2+ ions at

600 Ni2+

Poly(dG-dC)Poly(dG-dC)

500 400 ∆εs

E.coli Micrococus luteus

Herring sperm Salmon sperm Calf thymus

300 200

Mice liver(C3HA line) Poly(dG)poly(dC) Cl.perfringers

100 0 −20

Poly(dA-dT)Poly(dA-dT) 0

20

40

60

80

100

GC% +2

Figure 2.6 Influence of Ni on the UV spectra of various DNAs and polynucleotides. Graphical illustration of the dependence of the mole extinction (∆es) on the percentage of GC pairs present

42


0.8

1 2

A

0.0 45 000

36 000 ∼ ν, cm−1

Figure 2.7 UV spectra of polynucleotide poly(G-C).poly(G-C) in 25% ethanol solution (10−2 M NaCl) without (1) and with (2) Ni2+ ions; concentration is 0.75 Ni/P

5 4 3 1000

2

∆ε

0 1 − 1000

6 7

−2000 45 000

40 000

35 000 ∼ ν, cm−1

Figure 2.8 UDS of polynucleotide poly(G-C).poly(G-C) in 25% ethanol solution at different concentrations of Ni2+: (1) 0.1 Ni/P, (2) 0.25 Ni/P, (3) 0.3 Ni/P, (4) 0.4 Ni/P, (5) 0.5 Ni/P, (6) 0.75 Ni/P, (7) 1 Ni/P

concentrations of 0.75 Ni2+ per nucleotide (P)). Ni2+ ions cause a significant red shift of 700 cm−1 and strong hypochromic effect (25%). Figure 2.8 shows the UDS of poly(dG-dC)·poly(dG-dC) in 0.25% ethanol solution with Ni2+/P ratios in the range 0.1 to 1.0. Figure 2.9 demonstrates a large difference between the UDS of Ni2+/ poly(dG-dC)·poly(dG-dC) recorded in water (1) and in 25% ethanol solution (2). In the first case a typical saturation line is shown while in the second case an S shape curve characteristic of conformational change is observed, e.g. B→A or B→Z transition. The CD spectra (Figure 2.10) recorded under the same conditions as for UDS show that at concentration Ni/P = 0.2 the polynucleotide is in the C conformation (commonly observed as a low humidity form of the lithium salt). At higher concentrations of Ni+2, the double helix is shown to undergo transition to the

Model and Mechanisms for Point Defects 43 3250

∆εs

2

1

0

1 Ni(II)/P

Figure 2.9 The dependence of UDS of polynucleotide poly(G-C).poly(G-C) on Ni2+ concentration in water (1) and in 25% ethanol solution (2)

7 6 4

1

εL − εR

2

3

5 4 −8 40 000

∼ ν, cm−1

32 000

Figure 2.10 CD spectra of polynucleotide poly(G-C).poly(G-C) in 25% ethanol solution at different concentrations of Ni2+: (1) without Ni2+, (2) 0.1 Ni/P, (3) 0.2 Ni/P, (4) 0.25 Ni/P, (5) 0.3 Ni/P, (6) 0.35 Ni/P, (7) 0.4 Ni/P

Z conformation. It was also demonstrated in our experiments that the Ni2+–poly(dGdC)·poly(dG-dC) complex forms the C conformation with wrong base pairs during the B→Z transition process.61–63 Thus, we have demonstrated that UV spectroscopy is a sensitive method for monitoring conformational transitions in double-helical DNA.

2.3 Model and Mechanisms of Metal-Induced Formation of Point Defects 2.3.1 Double Proton Transfer (DPT) in DNA: UV Spectroscopic Measurements The influence of pH on UV spectra of guanine and cytosine, as well as on keto–enol and amino-imino transformations in guanine and cytosine, has been known for a

44


long time.46,63,64 It should be noted that the transfers of protons between the donor and acceptor groups N1 and O6 in guanine and N4 and N3 in cytosine are definitely performed with the help of the hydrated water molecules; they form a kind of cyclic structure with the above groups in guanine and cytosine. Because of self-congruent transition of protons from N1 of guanine and N4 of cytosine to molecules of hydrated water, reorientation of water molecules takes place in both cases. Thus, we can state that the specific energy of H-bonds with two energy states and self-congruent transfer of protons in cyclic structures is a necessary and sufficient condition for keto–enol and amino–imine tautomeric transformations of polar organic molecules, particularly guanine, cytosine and adenine in solution. Water molecules act as a mediator in the process of forming the joint electron–proton complementary complex between the proton-acceptor/donor groups of the solute molecules and those of the solvent. As already mentioned, there is no need for water molecules as mediators in the DNA duplex because a pair of complementary bases always forms a cyclic structure. Double proton transfer (DPT) in DNA is demonstrated spectroscopically as bathochromic shifts, weak hypochromic effects and small widening of the absorption band (Table 2.4).14,17 We shall discuss the mechanism in detail, using GC pairs as an example for the following reasons: (i)

(ii) (iii) (iv) (v)

the effect of H+ interactions with guanine and cytosine,47,64 as well as interaction between transition metal ions and guanine are easily observed in UV-spectra; 65 spontaneous mutations of a genome occur more often at GC pairs than at AT pairs;66–68 guanine and cytosine are often populated with rare enol and imino forms;69–71 GC pairs are far less resistant to tunnelling transitions compared to AT pairs; 26,72,73 in the DNA duplex the site of preferable binding of H+ and metal ions is the endocyclic N7 of guanine located in the major groove.28,74

In general, the keto–enol and amino–imine tautomeric transformations in GC pairs are conditioned by the electric charges on G-N1 and C-N3 of endocyclic nitrogens, which play the principal role in H-bonding of the base pairs. Figures 2.11 and 2.12 present electronic configurations of atoms participating in the formation of Hbonds in GC and AT pairs. It is interesting to consider here the disturbance of the electronic structure of guanine in the pair of bases in a DNA duplex provoked by H+ and transition metal ions. When a positive ion interacts with a pyridine-type nitrogen (G-N7), a nonbonded pair of electrons located in a sp2 hybrid orbital can significantly overlap the p-electronic system of the indole ring of guanine inducing a decrease of electronic density of the ring, including the endocyclic N1 nitrogen atom. This will lead to a decrease in the depth of potential energy from the N1 side of guanine and an increase of possible proton tunnelling to N3 in cytosine, as illustrated in Figure 2.13. As early as in 1969, Scheiber and Daune47 pointed out quantitative similarities

Model and Mechanisms for Point Defects 45

Guanine

Watson-Crick base pair Cytosine

H

N2 pz

e

N1 sp

pz

sp2

O6 pz

sp2 N2

pz

O2

e

e H

H e

e

H wrong pair

sp2

N1 pz

2

sp2

O6

Watson-Crick base pair

sp N 3

Adenine py pz

sp2N

e

pz

sp2

sp2 N 1 1

e

4

sp2

H

sp2

sp2

O4

H e

e

2pz

H

N6

sp2 N

pz

sp N

pz 3

e

sp2

pz

spz

H

1

O4

N3 sp2

N4 pz

pz

H

wrong pair

O2

p pz sp N y 3

H

pz

e

2pz

sp2

pz

Thymine e

N6

sp2

sp2

pz

A

pz

B

Figure 2.11 The electron configuration of atoms of G-C (A) and A-T (B) pairs taking part in H-bonds before and after DPT

minor groove N2

G

N1

e N7 Mn+

O6

H

O2

H

N3 H

N4

e C

major groove

Figure 2.12 Double proton transfer (DPT) in a G-C pair of DNA

E 0

r

+

N

H

N

Figure 2.13 Hypothetical function of potential energy of N-H···N type H-bonds

46


between difference spectra of G-C + H+ and DNA + M2+ (M2+ = Cu2+, Ni2+ and Fe2+). Our investigations, carried out on metal ion–DNA complexes by thermodynamic and spectroscopic methods, in particular spectrophotometry in the UV, visible and near IR ranges, allowed us to propose a scheme for the structure and energetics of DNA interactions with metal ions.31,75 2.3.2 Depurination Interaction of H+ and Mn+ with N3 and N7 of guanine in DNA may induce DPT (1) and/or hydrolysis of the glycosidic linkage C1–N9 (2): C1+ C1+ + OH−

N9−

and

C1 OH

and

N9− + H+

(1)

N9H

(2)

The process of depurination may take place only in the unwound state of double-helical DNA. Crothers and collaborators76 studied the energetics of the opening of the central base pairs in different triplets of double-stranded RNA, which was often used as a model for the DNA double helix since crystallographic data for DNA was not known at the time. According to this study the energy of opening of a G-C pair at 25 °C is at a maximum and equals 7.5 kcal mol−1 in the sequence: G G G C C C corresponding to a minimum probability of 0.3 × 10−5 s, while the energy of opening an A-U pair is at a minimum (4 kcal mol−1) for the sequence: A A A U U U corresponding to a maximum probability of 120 × 10−5 s. By inserting these data in Equations (2.1) and (2.2), we obtain the times ∼3 × 10−6 s for G-C and ∼10−8 s for A-U. The probability of metal-induced DPT in DNA double helix is equal to the ratio of the wave shift caused by metal ion interactions with DNA G-C pairs to the wave shift after total protonation of guanine (pH 1) and cytosine (pH 2), which amounts to ∼700 cm−1:15 PDPT = ∆vMn+ ∆vG+C+H+ = ∆vMn + 700 cm −1 Considering the values of Dn from Table 2.4, we can assume that for the metal–DNA complexes studied by us these probabilities are in the range from 10−1

Model and Mechanisms for Point Defects 47

(for Mn2+) to 1 (for Cu+).15 Consequently, metal-induced depurination of the DNA chain in wrong Watson–Crick pairs depends mainly on the probability of base pair opening. In conclusion, interactions of H+ and Mn+ with N3 and N7 of purines will increase the probability of prototropic tautomerization of purines and, finally lead to the transition of N9 purines in pyrol-type nitrogens followed by consecutive depurination. Later the question of base pair opening was studied directly in DNA under different conditions using NMR and other methods.77–81 The results obtained by these authors are in agreement with the model proposed by Crothers.76 2.3.3 Formation of Point Defects For steric reasons, metal-ion-induced DPT may take place only at G-N7 in GC pairs, which can be approached by transition metal ions in DNA. DPT may also take place at G-N3 (GC pair) and T-O2 (AT pair) sites, but in this case only H+ ions can approach these sites. Thus, it may be concluded that transition metal ions cause DPT exclusively in GC pairs via G-N7, while H+ can do it at both GC and AT via G-N7, G-N3 and T-O2. DPT and depurination in vivo can lead to the gene mutations i.e. point mutations (transition and deletion, respectively). In the first case, the geometry of the base pairs that undergo DPT remains the same, or changes very slightly, and it may stay unnoticed in the processes of replication and correction. Nevertheless, the hydration of the correct Watson–Crick pairs differs from the wrong ones (transferred by DPT).14 We do not know the biological mechanisms of detection of these differences. In the view of solid-state physics, one of the major characteristics of these kinds of defect is their formation energy (U). In the case of DNA, this energy lies in the range 0 (for Cu+) to 2.4 kcal mol−1 (for Mg2+) (Table 2.4). The values of the energy (U) given in the table were obtained by the Boltzman distribution. In our case, PDPT = ∆vMn + 700 cm −1 represents the ratio of G-C pairs in keto and enol tautomeric forms: PDPT = nket/nen ≈ e−U/RT. Let us consider the probable mechanism of transition-type mutations in a G-C pair of DNA by an example involving the interaction of Cu2+ with G-N7. Schematically we can describe it as follows: the helix unwinding

Cu2+

G DPT G* C

C*

G*

mutation A

C

T

A*

A

C

T

* – rare tautomeric form of bases

48


Cytosine can stay in its imine form for quite a long period of time, in spite of the fact that this structure is energetically less advantageous compared to that of the amine form. The restoration process includes the time needed for the formation of a cyclic unidirectional structure of water molecules. Thus, when a wrong G-C pair opening takes place, the less hydrated cytosine needs to pass through the potential barrier in order to rebuild the surrounding water in a unidirectional cyclic structure. The time span depends on the number of water molecules participating in the cycle and in many cases exceeds the time needed for the closing of the helix. In the process of winding/unwinding of the double helix in the presence of transition metal ions, one may envisage the following alterations: (1) Mn+

G* C*

(2) M

n+

G* C

(3) Mn+

C*

G C*

(4) Mn+

G C

For the second and third examples, the defect that remains after closing of the disrupted pair is easily recognized by the repair system in vivo. In particular, one should notice that Cu2+ ions can easily be reduced to the Cu+ state (UV irradiation, g irradiation, ascorbic acid etc.)28,31 and the Cu+–DNA lifetime will increase by about 11 orders of magnitude, reaching 104 − 5 × 105 s.17 In this case, Cu2+, which is essential for life for many mammals, can be transferred into extremely toxic soft Cu+ ions. Depurination is the second process inducing point defects. Interaction of H+ with pyridine-type nitrogen atoms, and the preference of transition metal ions to bind to G-N7 may lead to the hydrolysis of the C-N3 glycosidic linkage, which is the weakest covalent bond in DNA with a torsion rotation barrier in the range 0.003–0.006 kcal mol−1 74,82 At the moment of base pair opening, the ions will initiate rupture of the bond and if it is not restored in vivo, severe damage of the double helix called deletion will take place. Even if the repair system restores the deleted base, this may not completely prevent damage, because interaction of metal ions with the DNA duplex may also lead to DPT (see above), and after depurination of guanine, cytosine adopts a rare tautomeric imine form. In this case, adenine instead of guanine will be inserted in the base pair.

2.4 Conclusions (i)

A thermodynamic model of metal ion interactions with DNA can be used to establish direct proportional correlation between dynamics and stability constants. This approach is based on the physical model of gas and vapour adsorption on the surface of solids. Thus, the principal concept of molecular biophysics regarding biomolecular structure–dynamics–function can be reformulated as structure–stability–function. One may notice that such an approach highly simplifies and widens the time intervals (from 10−10 s to 105–106 s and more) under investigation for dynamic characteristics of macromolecules.

Abbreviations

49

(ii) The dynamic properties of DNA connected with big amplitude movements like unwinding of the double helix and opening of base pairs determine the values of the stability constants of DNA complexes with transition metal ions, as well as the ranges of their variations. (iii) UV spectroscopic manifestation of double proton transfer (DPT) in GC pairs induced by H+ ions and transition metal ions, and the mechanisms of this phenomenon can be considered using a phenomenological quantum-mechanical approach to assess the effect of environment change on electron configuration of atoms in molecules participating in H-bonds. This consideration takes into account the specific energy of double-level H-bonds, and self-congruent transition of protons in cyclic structures. These structures form unified electronic– proton complementary molecular complexes between proton-donor and proton-acceptor groups. The latter is an indispensable and satisfactory condition for keto–enol and amino–imino tautomeric transformations in solutions of polar organic molecules. (iv) A comparison of the Scatchard plot and the Klorz method for determining stability constants, using, for example, the Pb2+/DNA interaction, shows that the former method gives the best results. (v) The use of UV spectroscopy has demonstrated that Ni2+ ions under relatively mild conditions (25% ethanol, 0.75 Ni2+/DNA(P), 10−2 M NaCl) induce B ↔ Z transitions in double-helical DNA. (vi) A mechanism for point defect formation induced by metal ions, leading to mutations of transition or deletion types in DNA, has been presented, using as an example the interaction of Cu2+ ions with DNA. It was shown that formation of metal-induced wrong Watson–Crick base pairs via DPT in water solutions requires energy within the range 0 (for Cu+) to 2.4 kcal mol−1 (for Mg2+).

Acknowledgements The authors express their gratitude to Prof. J. Monaselidze for useful discussions; Dr M. Kharatishvili for providing the CD spectra of DNA; Dr I. Khutsishvili, Dr E. Mchedlishvili and Dr Kh. Sologashvili for their active assistance in this work. The work was partly supported by the grant number 2.29.04 of the Georgian Academy of Sciences.

Abbreviations ctDNA Calf thymus DNA CD Circular dichroism DPT Double proton transfer e Molar extinction coefficient Des = |De(nmax)| + |De(nmin)|

50


EtBr En − eaq Ke/aq K IR M UDS

Ethidium bromide Ethylenediamine Hydrated electron − Rate constant of the reaction of eaq Stability constant Infrared Free hydrated metal ion Ultraviolet difference spectra

References 1. Ito, T.; Nikaido, G.; Nishimoto, S.; Effects of metal binding to mismatched base pairs on DNA-mediated charge transfer; J. Inorg. Biochem., 2007, 101, 1090–1093. 2. Loganathan, D.; Morrison, H.; Effect of ring methylation on the photophysical, photochemical and photobiological properties of cis-dichlorobis(1,10-phenanthroline)rhodium (III)chloride; Photochem. Photobiol., 2006, 82, 237–247. 3. Loganathan, D.; Morrison, H.; Photocisplatin reagents; Curr. Opin. Drug Disc. Devel., 2005, 8, 478–486. 4. Billadeau, M.A.; Morrison, H.; Photolytic covalent binding of metal complexes to DNA. In: Metal Ions in Biological Systems, Vol. 33, A. Sigel and H. Sigel, Eds., Marcel Dekker Inc., New York, Basel, 1996, Chap. 9, 269–292. 5. Loganathan, D.; Morrison, H.; Simultaneous photoconjugation of methylene blue and cis-Rh(phen)(2)Cl-2(+) to DNA via a synergistic effect; Photochem. Photobiol., 2000, 71, 369–381. 6. Menon, E.L.; Perera, R.; Navarro, M.; Kuhn, R.; Morrison, H.; Phototoxicity against tumor cells and sindbis virus by an octahedral rhodium bisbipyridyl complex and evidence for the genome as a target in viral photoinactivation; Inorg. Chem., 2004, 43, 5373–5381. 7. Boudvillain, M.; Pablies, R.; Leng, M.; Evidence for a catalytic activity of the DNA double helix n the reaction between DNA, platinium(II), and intercalators. In: Metal Ions in Biological Systems, Vol. 33, A. Sigel and H. Sigel, Eds., Marcel Dekker Inc., New York, Basel, 1996, Chap. 4, 87–102. 8. Gorodetsky, A.A.; Barton, J.K.; DNA-mediated electrochemistry of disulfides on graphite; J. Am. Chem. Soc., 2007, 129, 6074–6075. 9. Lu, W.; Vicic, D.A.; Barton, J.K.; Reductive and oxidative DNA damage by photoactive platinum(II) intercalators; Inorg. Chem., 2005, 44, 7970–7980. 10. Delaney, S.; Yoo, J.; Stemp, E.D.A.; Barton, J.K.; Charge equilibration between two distinct sites in double helical DNA; Proc. Nat. Acad. Sci. USA, 2004, 101, 10511–10516. 11. O’Neill, M.A.; Barton, J.K.; DNA-mediated charge transport chemistry and biology. In: Topics in Current Chemistry: Electron Transfer in DNA: I, ed. Schuster, G.B., SpringerVerlag, 2004, 67–115. 12. Delaney, S.; Barton, J.K.; Long-range DNA charge transport; J. Org. Chem., 2003, 68, 6475–6483. 13. Stemp, E.D.A.; Barton, J.K.; Electron transfer between metal complexes bound to DNA: Is DNA a wire? In: Metal Ions in Biological Systems, Vol. 33, A. Sigel and H. Sigel, Eds., Marcel Dekker Inc., New York, Basel, 1996, Chap. 11, 325–362. 14. Bregadze, V.; Khutsishvili, I.; Tsakadze, K.; Double proton transfer in DNA: The role of ions, ethanol and proteins; Bull. Georg. Acad. Sci., 2006, 173, 143–146. 15. Bregadze, V.; Tsakadze, K.; The analysis of wrong G-C pairs induced by the ions of transition metals and H3O+ in DNA, J. Biol. Phys. Chem., 2006, 6, 167–170. 16. Silvestri, A; Barone, G.; Ruisi, G.; Anselmo, D.; Riela, S.; Liveri, V.T.; The interaction of native DNA with Zn2+ and Cu2+ complexes of 5-triethyl ammonium methyl salicylidene orto-phenylendiimine; J. Inorg. Biochem., 2007, 101, 841–848.

References

51

17. Bregadze, V.G.; Khutsishvili, I.G.; Chkhaberidze, J.G.; Sologashvili, K.; DNA as a mediator for proton, electron and energy transfer induced by metal ions; Inorg. Chim. Acta, 2002, 339, 145–159. 18. Bregadze, V.G.; Chkhaberidze, J.G.; Khutsishvili, I.G.; Effect of metal ions on the fluorescence of dyes bound to DNA. In: Metal Ions in Biological Systems, Vol. 33, A. Sigel and H. Sigel, Eds., Marcel Dekker Inc., New York, Basel, 1996, Chap. 8, 253–266. 19. Chalote Auerbach, F.R.S.; Mutation Research: Problems, Results and Perspectives. Mir, Moscow, 1978. 20. Verne, G.; The Evolutionary Process: A Critical Review of Evolutionary Theory, Mir, Moscow, 1985. 21. Bloemink, M.J.; Reedijk, J.; Cisplatin and derived anticancer drugs: mechanism and current status of DNA binding. In: Metal Ions in Biological Systems, Vol. 32, A. Sigel and H. Sigel, Eds., Marcel Dekker Inc., New York, Basel, 1996, Chap. 19, 641–675. 22. Vinje, J.; Sletten, E.; NMR spectroscopy of anticancer platinum drugs; Anti-Cancer Agents Med. Chem., 2007, 7, 35–54. 23. Liu, Y.; Sivo, M.F.; Natile, G.; Sletten, E.; Antitumor trans-platinum complexes can form cross-links with adjacent purines; Angw. Chem. Int. Ed., 2001, 40, 1226–1228. 24. Liu, Y.; Vinje, J.; Pacifico, C.; Natile, G.; Sletten, E.; Formation of adenine-N3/guanine-N7 cross-link in the teaction of trans-oriented platinum substrates with dinucleotides; J. Am. Chem. Soc., 2002, 124, 12854–12862. 25. Karidi, K.; Garoufis, A.; Tsipis, A; Hadjiliadis, N.; Dendulk, H.; Reedijk, J.; Synthesis, characterization, in vitro antitumor activity, DNA binding properties and electronic structure (DFT) of the new complex cis-(Cl,Cl)[RuCl2(terpy)(NO)]Cl; J. Chem. Soc. Dalton, 2005, 1176. 26. Löwdin, P.O.; Proton tunneling in DNA and its biological implications; Rev. Mol. Phys., 1963, 35, 724–732. 27. Löwdin, P.O.; Quantum genetics and the aperiodic solid. Some aspects of the biological problems of heredity, mutations, aging and tumors in view of the quantum theory of the DNA molecule; Adv. Quant. Chem., 1966, 2, 213. 28. Bregadze, V.G.; Metal ion interactions with DNA: consideration on structure, stability, and effects from metal ion binding. In: Metal Ions in Biological Systems, Vol. 32, A. Sigel and H. Sigel, Eds., Marcel Dekker, New York, 1996, Chap. 12, 419–451. 29. Anbar M.; Neta P.; A compilation of specific bimolecular rate constants for the reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals with inorganic and organic compounds in aqueous solutions; Int. J. Appl. Radiat. Isotopes, 1967, 18, 493–523. 30. Hart, E.J.; Anbar, M.; The Hydrated Electron, Atomizdat, Moscow, 1973, 280. 31. Bregadze, V.G.; Nature of DNA interactions with cations: UV spectroscopic investigations and Marcus theory; Int. J. Quant. Chem., 1980, 17, 1213–1219. 32. Lavery, R.; Pullman, B.; The molecular electrostatic potential and steric accessibility of poly (dI.dC). Comparison with poly (dG.dC); Nucleic Acids Res., 1981, 9, 7041– 7052. 33. Lavery, R.; Cauchy, D.; De la luz, R.O. et al.; Molecular electrostatic potential of the BDNA helix. VII. Effect of screening by monovalent cations; Int. J. Quant. Chem., 1980, 7, 323–330. 34. Yakushevich, L.V.; DNA dynamics; Mol. Biol., 1989, 23, 652–662. 35. Frenkel, J.; Theorie der adsorption und verwandter erscheinungen; Z. Phys., 1924, 26, 117. 36. Frenkel, J.; Statistical Physics; M.-L. Acad. Sci. USRR, 1948, 760. 37. Slutsker, A.I.; Mikhailin, A.L.; Slutsker, I.A.; Microscopics of energy fluctuations of atoms in solids; Uspekhi Fizicheskikh Nauk, 1994, 164, 357–366. 38. Lineweaver, H.; Burk, D.; The determination of enzyme dissociation constants; J. Am. Chem. Soc., 1934, 56, 658–666. 39. Liliom K.; Orosz F.; Horvath L.; Ovadi J.; Quantitative evaluation of indirect ELISA. Effect of calmodulin antagonists on antibody binding to calmodulin; J. Immunol. Methods, 1991, 143, 119–125.

52


40. Glaser, R.W.; Program for simulation of ELISA experiments and affinity; J. Immunol. Methods, 1993, 160, 129–133. 41. Rovati, G.E.; Rodbard, D.; Munson, P.J.; DESIGH: computerized optimization of experimental design for estimating K and B(max) in ligand binding experiments. I. Homologous and heterologous binding to one or two classes of sites; Anal. Biochem., 1988, 174, 636–649. 42. Scatchard, G.; The attraction of proteins for small molecules and ions; Ann. N.Y. Acad. Sci., 1949, 51, 660–672. 43. Klotz I.; Hunston,D.L.; Properties of graphical representations of multiple classes of binding sites; Biochem., 1971, 10, 3065–3069. 44. Gelagutashvili, E.S.; Rcheulishvili, A.N.; Mosulishvili, L.M.; Constants of Pb(II) ion binding to DNA at different ionic strengths; Biophysics, 2001, 46, 957–959. 45. Taylor, J.R.; An Introduction to Error Analysis; Mir, Moscow, 1985, 272. 46. Angelici, R.J.; Stability of coordination compounds. In: Inorganic Biochemistry, Vol. 1, G.L. Eichhorn, Ed., Mir, Moscow, 1978, Chap. 2, 89–132. 47. Scheiber, J.P.; Daune, M.; Interactions des ions metalliques avec le DNA. IV. Fixation de l’ion cuivrique sur le DNA; Biopolymers, 1969, 8, 139–152. 48. Frank-Kamenetski, M.D.; Anshelevich, V.V.; Lulashin, A.B.; Polyelectronic model of DNA; Uspekhi Fizicheskikh Nauk, 1987, 151, 595–618. 49. Gillespie, R.J.; Hargittai, I.; The VSEPR Model of Molecular Geometry; Allyn and Bacon, Boston, MA, 1991. 50. Shimoni-Livny, L.; Glusker, J.P.; Bock, Ch.W.; Lone pair functionality in divalent lead compounds; Inorg. Chem., 1998, 37, 1853–1867. 51. Gelagutashvili, E.; Studies of interaction between Ni(II), Zn(II) and Cr(III) ions and DNA from Spirulina platensis; Proc. Georgian Acad. Sci., Chem. Ser., 2005, 31, 69–73. 52. Gelagutashvili, E.; Belokobilsky, A.; Sigua, K.; Tarkashvili, C.; Nature of binding of Ni(II) and Zn(II) ions with DNA; Proc. Georgian Acad. Sci., Chem. Ser., 1999, 25, 107–111. 53. Nieboer, E.; Richardson, D.H.S.; The replacement of the non descript term ‘heavy metal’ by a biologically and chemically significant classification of metal ions; Environ. Pollution (Series B), 1980, 1, 3–26. 54. Rossetto, F.E.; Nieboer, E.; The interaction of metal ions with synthetic DNA: Induction of conformational and structural transitions; J. Inorg. Biochem., 1994, 54, 167–186. 55. Pauling, L.; General Chemistry; Mir, Moscow, 1974, 839. 56. Vologodskii, A.V.; Cozzarelli, N.R.; Conformational and thermodynamic properties of supercoiled DNA; Ann. Rev. Biophys. Biomol. Struc., 1994, 23, 609–643. 57. Jukhnevich, G.V.; Infrared Spectroscopy of Water; Nauka, Moscow, 1973, 207. 58. Granot, J.; Kearns, D.R.; Interaction of DNA with divalent metal ions. III. Extent of metal binding: experiment and theory; Biopolymers, 1982, 21, 219–232. 59. Sletten, E.; Lie, B.; Copper complex of guanosine-5′-monophosphate; Acta. Cryst., 1976, B32, 3301–3304. 60. Aoki, K.; General conclusion from solid state studies of nucleotide-metal ion complexes. In: Metal Ions in Biological Systems, A. Sigel and H. Sigel, Eds., Vol. 32, Marcel Dekker, New York, 1997, Chap. 4, 479–501. 61. Pohl, F.M.; Jovin, T.M.; Salt-induced co-operative conformational change of a synthetic DNA: equilibrium and kinetic studies with poly (dG-dC); J. Mol. Biol., 1972, 67, 375–396. 62. Pohl, F.M.; Polymorphism of a synthetic DNA in solution; Nature, 1976, 260, 365–366. 63. Sitko, J.C; Mateescu, E.M.; Hansma, H.G.; Sequence-dependent DNA condensation and the electrostatic zipper; Biophys. J., 2003, 84, 419–431. 64. Cantor, C.R.; Schimmel, P.R.; Biophysical Chemistry, Part 3, Mir, Moscow, 1985. 65. Bregadze, V.G.; Khutsishvili, I.G.; Structure of DNA complexes with the transition metals: UV difference spectra; Proc. Acad. Sci. Georgia, Biol. Ser., 1994, 20 (1–6), 272–280. 66. Freese, E.B.; Molecular Genetics, Part 1, J.H. Taylor, Ed., Academic Press, 1963, 135. 67. Drake, J.W.; Spontaneous mutations accumulating in bacteriophage T4 in the complete absence of DNA replication; Proc. Nat. Acad. Sci. USA, 1966, 55, 738–743.

References

53

68. Danilov,V.I.; Quantum chemical Investigation of nucleic acid bases tautomer forms and problem of mutation; Biofizika, 1967, 12, 540–544. 69. Pullman, B; Pullman, A.; Le tautometrie des bases puriques et pyrimidiques et la theorie des mutations; Biochem. Biophys. Acta, 1962, 64, 403–405. 70. Pullman, B.; Pullman, A.; Electronic delocalization and biochemical evolution; Nature, 1963, 196, 1137. 71. Pullman, B.; Adenine complexes of thymine and uracil. In: Electronic Aspects of Biochemistry, B. Pullman, Ed., Academic Press, New York, 1964, 135. 72. Löwdin, P.O.; Effect of proton tunneling in DNA on genetic information and problems of mutations, aging and tumors; Biopolym. Symp., 1964, 1, 161–181. 73. Löwdin, P.O.; Some aspects of DNA replication; Incorporation errors and proton transfer. In: Electronic Aspects of Biochemistry, B. Pullman, Ed., Academic Press, 1964, 167. 74. Saenger, W.; Principles of Nucleic Acid Structure, Moscow, Mir; 1987, p. 584. 75. Bregadze, V.G.; Interpretation of ultra-violet difference spectrum of DNA complexes with metal ions of the first transition series; Biofizika, 1974, 19, 179–181. 76. Gralla, J.; Crothers, D.M.; Free energy of imperfect nucleic acid helixes, III. Small integral loops resulting from mismatches; J. Mol. Biol., 1973, 78, 301–319. 77. Dornberger, U.; Leijion, M.; Fritzsche, H.; High base pair opening rates in tracts of GC base pairs; J. Biol. Chem., 1999, 274, 6957–6962. 78. Bouvier, B.; Grubmuller, H.; A molecular dynamics study of slow base flipping in DNA using conformational flooding; Biophys. J., 2007, 93, 770–786. 79. Coman, D.; Russu, I.M.; A nuclear magnetic resonance investigation of the energetics of base pair opening pathways in DNA; Biophys. J., 2005, 89, 3285–3292. 80. Giudice, E.; Varnai, P.; Lavery, R.; Base pair opening within B-DNA: free energy pathways for GC and AT pairs from umbrella sampling simulations; Nucleic Acids Res., 2003, 31, 1434–1443. 81. Bhattacharya, P.K.; Cha, J.; Barton, J.K.; 1H NMR determination of base-pair lifetimes in oligonucleotides containing single base mismatches; Nucleic Acids Res., 2002, 30, 4740–4750. 82. Scott, R.A.; Scheraga H.A.; Method for calculating internal rotation barriers; J. Chem. Phys.; 1965, 42, 2209–2215.

3 Metal Ion Coordination in G-Quadruplexes Janez Plavec

3.1 Introduction DNA is able to adopt a variety of secondary structures other than the canonical BDNA which may potentially add considerable informational capacity to DNA.1–4 The initial concept of DNA being exclusively the holder of linearly arranged genetic information is being revisited to add alternative DNA conformations as important informational elements. Guanine nucleosides and nucleotides, as well as DNA (and RNA) sequences that include tracts of contiguous guanine residues form fourstranded structures known as G-quadruplexes (or tetraplexes or G4 structures). Several reviews on G-quadruplexes have been published recently5–12 including the monograph edited by Neidle and Balasubramanian.1 G-quadruplexes have been proposed to play roles in DNA transactions such as replication, transcription and recombination.13–15 These structures are adopted by G-rich nucleic acid sequences which have been found to be abundant throughout the genomes of many organisms.16–21 Potential G-quadruplex-forming regions are ample in telomers, oncogene promoter sequences and other biologically relevant regions of the genome. The fact that G-quadruplexes play roles in crucial biological processes makes them targets for therapeutic intervention.22–46 G-quadruplex-forming oligonucleotide sequences, which have the ability to selectively act as inhibitors of signal transduction or transcription via binding to particular targets, are known as aptamers.47–54 In addition to Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

56

Metal Ion Coordination in G-Quadruplexes

their biological implications, the propensity for G-quadruplexes to form cationspecific supramolecular structures has opened the possibility for using guanine nucleoside and nucleotide analogues as well as G-rich DNA sequences for applications that range from environmental remediation to organic synthesis and nanotechnology.55–69 For example, molecular self-assembly of guanine-containing nucleosides has been shown to be useful for the construction of highly selective ionophores.70,71 G-rich sequences that fold into G-quadruplex structures can, of course, form Watson–Crick duplexes with their complementary C-rich strands. The formation of G-quadruplex or duplex structures within the cell will therefore depend on the relative stability of the G-quadruplex and the corresponding duplex.72–78 Furthermore, the C-rich strand might have the possibility of forming an intramolecular i-motif that likewise competes with duplex formation, which can also be influenced by cations or other changes in solution conditions (e.g. molecular crowding).79–83 The basic building block of a G-quadruplex is the G•G•G•G quartet, which is composed of four hydrogen-bonded guanine nucleotides in a horizontal planar arrangement (Figure 3.1). G-quartets are linked together by eight hydrogen bonds

Figure 3.1 The G-quartet: (a) four guanine residues that act both as acceptors and donors of H-bonds in Hoogsteen geometry form a cyclic G-quartet. R represents 2′-deoxyribose moiety; (b) side view of this planar building block of G-quadruplexes in cartoon presentation

Introduction

57

in a Hoogsteen pairing geometry. The presence of cations is required for G-quartet formation. Cations reduce repulsions amongst guanine carbonyl oxygen atoms and contribute to enhanced base-base stacking interactions.5–8,10,84–88 The coordination of cations by the closely spaced carbonyl oxygen atoms of a G-quartet was postulated long before the first high-resolution structure of a G-quadruplex was determined. It is noteworthy that G-quartets interact with dehydrated cations via inner sphere coordination.89 It is therefore not surprising that formation, stability and structural details of G-quadruplexes are dependent on cation species and cation concentration. 3.1.1 Stoichiometry DNA G-quadruplexes can be formed with different numbers of oligonucleotide strands (i.e. 1, 2 or 4, Figure 3.2). In a tetramolecular structure four strands associate to form a G-quadruplex. Such quadruplexes are usually comprised of four strands in a parallel orientation (Figure 3.2a). Parallel quadruplexes have all guanine residues in an anti conformation. Association of two oligonucleotides with two G-tracts, which can be different, leads to a bimolecular structure with a so-called fold-back topology (Figure 3.2b). As a result, some of the G residues adopt the syn conformation and strands adopt different alternations of parallel and antiparallel orientations. The topological model in Figure 3.2b shows a G-quadruplex with two parallel and two antiparallel strands. In this structure each strand has one parallel and one antiparallel neighbour. An example of such an arrangement has been established experimentally by the structure of d[(G4T4G4)2] (see below). A monomolecular Gquadruplex can be adopted by sequences consisting of four G-tracts, separated by

Figure 3.2 (Plate 1) Variation in strand stoichiometry and topology of inter- and intramolecular G-quadruplex structures: (a) tetramolecular structure with all four parallel strands and all anti glycosidic torsion angles; (b) bimolecular and (c) monomolecular structures with antiparallel strands and alternating syn (in orange) and anti (in blue) orientations across glycosidic bonds (See colour plate section)

58


three regions that can be any combination of residues, including G. An example of such an intramolecular structure with antiparallel strands and alternating syn and anti orientations across the glycosidic bonds is shown in Figure 3.2c. Quadruplexes are labelled as antiparallel when at least one of the four strands is antiparallel to the others. This type of topology is found in the majority of bimolecular and in many monomolecular G-quadruplex structures. Antiparallel folds exhibit both syn and anti guanine residues within individual G-quartets, which are arranged in a way that is in accordance with strand orientations. Recently, a formalism was proposed describing the interdependency of a set of structural descriptors as a geometric basis for folding of monomolecular Gquadruplex topologies.90 The formalism is a fundamental step towards the prediction of monomolecular G-quadruplex folding topologies from primary structures. In addition to establishing rules that will enable the prediction of folding topologies, the work of Webba da Silva90 suggested that there could be other folding topologies in addition to those that have been identified experimentally so far. 3.1.2 Loops G-quadruplex structures may also be classified according to the location of the loops that link the neighbouring G-tracts involved in G-quartets. The location and length of the loops in combination and in addition to the number of stacked G-quartets and the number and the pattern of the strand directionalities all contribute to a plurality of G-quadruplex structures. Edge-wise (or lateral) loops join adjacent Gstrands by spanning the edge of an outer G-quartet as shown schematically in Figure 3.3a. These loops are generally composed of two or more residues.91,92 This type of loop structure has been observed in two NMR solution-state structures of asymmetric G-quadruplexes of the d(TG4T2G4T) sequence93 and in the bimolecular Gquadruplex structure formed by the sequence d(G3CT4G3C).94 Two of the edge-wise loops can be located either on the same or on opposite sides of a G-quadruplex core, resulting in head-to-head or head-to-tail arrangements, respectively. Two distinct bimolecular G-quadruplexes were formed by d(G4T3G4).95 One of the structures exhibited a head-to-tail lateral loop dimer in which all adjacent strands were

Figure 3.3 Loop structures: (a) edge-wise loops connecting antiparallel strands on the same side of the G-quadruplex; (b) diagonal loops connecting opposite antiparallel strands; (c) external double-chain reversal loops connecting parallel strands

Cation Coordination and Stability

59

antiparallel. The other was a head-to-head hairpin quadruplex with one adjacent strand parallel and the other antiparallel.95 The second type of loop joins opposite antiparallel G-strands by spanning the diagonal of the outer G-quartet (Figure 3.3b). Such loops usually consist of three or more residues. Diagonal loops were observed, for example, in the structure formed by the Oxytricha nova telomeric sequence d(G4T4G4).96–100 Adjacent parallel strands can be connected by a doublechain reversal loop that links the 3′-end of a given G-tract with the 5′-end of the other (Figure 3.3c). This loop arrangement gives a G-quadruplex a propeller type topology,101 with loops that can consist of a single or several residues.102–104 A doublechain reversal loop has been found both in crystal and solution-state structures for G-quadruplexes adopted by human telomeric DNA sequences, and more recently in a number of nontelomeric quadruplexes (see below). Residues in all three loop orientations have been shown to contribute to the stability of G-quadruplex structures though base-pairing and stacking over the neighbouring G-quartets.95,105–114 It is noteworthy that loops can adopt diverse structures beyond the basic classification mentioned above, which can be targeted by small molecule ligands.

3.2 Cation Coordination and Stability of G-Quadruplexes Early NMR studies of alkali metal ion specificity suggested that 5′-GMP formed self-ordered structures consisting of stacked G-quartets in the presence of K+, Na+ and Rb+, but not Li+ or Cs+.115 The first experimental evidence for direct coordination of dehydrated cations between G-quartets was provided by the observation of broadening and shielding of the 23Na resonance in solutions of 5′-GMP.116,117 K+ and Na+ ions are the most extensively characterized cations with respect to their ability to stabilize G-quadruplex structures.118–122 A list of other cations that have been shown to promote G-quadruplex formation includes the monovalent cations Rb+, Cs+,123,124 NH4+ 125,126 and Tl+100,127–129 as well as the divalent cations Sr2+,124,130–133 Ba2+,134,135 and Pb2+.125,136–138 In general, low divalent cation concentrations initially stabilize G-quadruplexes,139 while increasing concentrations eventually become destabilizing. Mn2+ and europium cations that stabilize G-quadruplex have been localized in the grooves.140,141 Studies of 8-bromoguanosine gel melting-transition temperatures indicated the following relative G-quartet stabilizing propensities of divalent cations: Sr2+ >> Ba2+ > Ca2+ > Mg2+. Davis and coworkers have shown that, in the presence of Pb2+, lipophilic guanine analogues can associate to form quadruplexlike structures in organic solvents.142 These results affirmed earlier studies that Pb2+ induces a more stable and compact structure compared to K+, which is evident from the comparison of cation-O6 bond lengths, O6-O6 diagonal distances and interquartet separation.136 Studies by Hardin and coworkers have shown that Mg2+ and Ca2+ ions stabilize the tetramolecular parallel-stranded G-quadruplexes formed by the sequences d(CGCG3GCG) and d(TATG3ATA) more than monovalent alkali cations.143,144 A CD spectroscopic study by Sugimoto and coworkers has shown that the bimolecular G-quadruplex d[(G4T4G4)2] formed in the presence of NaCl is actually destabilized by only 1 mM concentrations of Mg2+, Ca2+, Mn2+, Co2+, or

60


Zn2+.76,145–147 Millimolar concentrations of Mg2+, Ca2+, Ba2+ and Sr2+ induce polymeric d(GGA)n repeat molecules to form quadruplexes.148 The studies of the melting temperature of gels formed by 8-bromoguanosine as a function of monovalent cation species showed that the relative ability to stabilize guanosine gels was K+ >> Rb+ > NH4+ > Na+ > Li+.149 The stabilization of G-quartet assemblies adopted by GMP follows the following order: K+ > NH4+ > Rb+ > Na+ > Cs+ > Li+.150 This general ordering has proven correct for G-quadruplex structures adopted by DNA oligonucleotides. Therefore, the melting temperature for a G-quadruplex is higher in the presence of K+ ions than in the presence of Na+ ions. The extent to which K+ increases the stability of a G-quadruplex over Na+ depends on the nucleotide sequence and can vary from as little as 2 °C to over 30 °C.107,124,151–155 The origins of these sequence-dependent differences are likely due to the K+ and Na+ forms of some G-quadruplexes being different in structure and/or in the number of coordinated cations. There is no simple explanation for the difference in melting temperatures for the Na+ and K+ forms for various G-quadruplexes. High-resolution X-ray and NMR structures have provided valuable insights into the effects of cation size and charge on G-quadruplex structure and on the location of ions within G-quadruplexes, all of which contribute to quadruplex stabilization.99,104,127,156–161 In addition, solution-state NMR studies have demonstrated that cations undergo dynamic exchange between coordination sites in the interior of a G-quadruplex and with bulk solution.162–165 The G-quadruplex formed by some G-rich sequences in the presence of K+ ions can be dramatically different in strand orientation and strand number from the Gquadruplex formed in the presence of Na+ ions. The size of a cation and its energy of (de)hydration both contribute to cation selectivity and the stability of a Gquadruplex. K+ is too large to be coordinated in the plane of the G-quartet, whereas a Na+ can fit in the centre of the G-quartet plane (see below). The resulting differences in K+ and Na+ coordination geometries and cation–guanine O6 distances contribute to a difference in the free energy provided by K+ versus Na+ coordination within a G-quadruplex. To a good approximation, hydration energies of monovalent ions are inversely proportional to their ionic radii.166 Thus, the free energy required to dehydrate K+ for coordination within a G-quadruplex is less than that required to dehydrate Na+. The net difference between the free energy of coordination within a G-quadruplex and the free energy of hydration ultimately determines cation selectivity by G-quadruplexes.6,153,167,168 A 1H NMR study by Hud and Feigon et al.168 on the bimolecular G-quadruplex adopted by d(G3T4G3) has shown that the preferred coordination of K+ over Na+ is actually driven by the greater energetic cost of Na+ versus K+ dehydration. The intrinsic free energy of Na+ coordination within d[(G3T4G3)2] is actually more favourable than K+ coordination.168 Subsequent computer simulations have provided additional support for the argument that cation dehydration is the dominant free energy that determines Na+ versus K+ selectivity by G-quartets.169–175 Recent quantitative analysis of the relative concentrations of the three dication forms of d[(G3T4G4)2] (i.e. 2 15NH4+, 15NH4+/K+ and 2 K+) at equilibrium, which are in slow exchange on the NMR timescale, showed that differences in standard Gibbs free energies are 5.7 kcal mol−1 between the di-15NH4+ and the 15NH4+/K+ forms, and 4.3 kcal mol−1 between the 15NH4+/K+ and the di-K+ forms.176 In comparison, the

Structure of Sequences Consisting of Human Telomere Repeats

61

conversion of d[(G3T4G3)2] from its Na+ form to K+ form is associated with a net free energy change (DDG0) of 1.7 kcal mol−1.168 3.2.1 Molecular Switches In addition to their central role in stabilizing G-quadruplexes, cations also play an important role in determining G-quadruplex structure. A systematic study of G-rich oligonucleotides containing runs of G residues of various lengths and with varying numbers of intervening residues, revealed a cation-dependent pattern for the Gquadruplex folds.48,75,105,107,109,110,113,122,177–179 The possibility of cation-induced changes in G-quadruplex structures has formed the basis of the hypothesis that DNA Gquadruplexes might function as molecular switches.180,181 The interconversion of Grich DNA sequences between antiparallel and parallel G-quadruplexes could be controlled by Na+/K+ concentrations.182–186 Thomas and coworkers used Raman spectroscopy to generate a phase diagram of antiparallel and parallel G-quadruplex structures, as controlled by concentrations of Na+ and K+ ions.180 At low cation concentrations d[(T4G4)4] folds into an antiparallel diagonal-loop G-quadruplex. Higher cation concentrations and K+ ions lead to the formation of parallel structures.180 An entirely different ion-dependent conformational switch has been observed for d(G2AG2AG). At a moderate ionic strength of 150 mM NaCl, this oligonucleotide forms a four-stranded quadruplex.103 The bimolecular quadruplex comprises a unique A•(G•G•G•G)•A hexad motif that is stabilized by sheared GA base pairs, in addition to the G-quartet.103 At low ionic strength d(G2AG2AG) forms a twostranded V-shaped ‘arrowhead’ motif.187 The cation binding sites associated with the two folds of this oligonucleotide suggest a rational for the cation-dependent switch. A cation-dependent loop polymorphism has been observed in the bimolecular G-quadruplex adopted by d(G3CT4G3C) that contains G•C•G•C quartets, along with G-quartets.94,188 The conformation of thymines in the edge-wise loops is different between Na+ and K+ forms. The loop conformation is affected by the presence of a K+ binding site in the T4 loop, whereas there is no such binding site for Na+ ions. The loop conformation, in turn, is sensitive to the conformation of the outer G-quartet, which is substantially different in the two cation forms. A K+ ion is postulated to penetrate the G•C•G•C quartet at the expense of hydrogen bonding between the G-C base pairs.94,188 Some of the aspects of cation coordination within G-quadruplexes and their consequences that are mentioned and introduced in a rather general and limited way in the section above are further discussed in greater detail with specific examples in the rest of this contribution.

3.3 Structure of Sequences Consisting of Human Telomere Repeats Human telomeric DNA ends are composed of tandem TTAGGG repeats. It has been known for over a decade that G-quadruplexes formed from human telomeric

62


repeat oligonucleotide sequences, e.g. d[(TTAGGG)4], exhibit different loop conformations depending upon whether Na+ or K+ ions are present.189 Studies on the structure and dynamics of the human telomeric G-quadruplex by single-molecule fluorescence resonance energy transfer (FRET) revealed two stable folded conformations in both Na+- and K+-containing buffers, with small differences between their enthalpies and entropies.74 Both folded conformations can be opened by the addition of a 21-mer complementary DNA oligonucleotide. The unfolding of both structures occurs at the same rate, which shows dependence on the monovalent cation present. Temperature-dependence studies in 100 mM KCl gave an apparent activation enthalpy and entropy of 6.4 kcal mol−1 and −52.3 cal mol−1 K−1, respectively, indicating that the unfolding is entropically driven and can occur easily. In contrast, in 100 mM NaCl the respective values are 14.9 kcal mol−1 and −23.0 cal mol−1 K−1, suggesting a more significant enthalpic barrier. These studies led to the conclusion that under near-physiological conditions, three structures coexist and can interconvert on a minute timescale.74,147,190–196 3.3.1 Single Repeat Sequences The oligonucleotide with a single repeat of the human telomere repeat sequence, d(TTAGGG), forms a tetramolecular parallel-stranded G-quadruplex at low K+ concentrations, which aggregates to form higher-order structures when the concentration of K+ is increased.197,198 3.3.2 Two-Repeat Sequences The X-ray structure of the two-repeat human telomeric sequence d(TAG3T2AG3T) formed in the presence of K+ ions exhibits a bimolecular structure with all strands in parallel orientation (PDB ID 1K8P).104 TTA residues form double-chain reversal (or propeller) loops. In addition, the end segments also participate in the formation of a T•A•T•A quartet, through pairing of the major groove edges of Watson–Crick A-T base pairs. K+ ions are coordinated between adjacent G-quartets.104 The same sequence has been shown to form both parallel and antiparallel G-quadruplex structures in the presence of K+ ions in solution.199 Both structures are bimolecular and comprise three stacked G-quartets. The two structures can coexist and interconvert in solution. They have different thermodynamic properties and different kinetics of folding and unfolding. d(TG4T2G4T), an analogous sequence with two G-for-A substitutions (originating from Tetrahymena), forms two asymmetric, bimolecular G-quadruplexes in solution that are involved in a similar equilibrium in the presence of Na+ ions.93 3.3.3 Four-Repeat Sequences The NMR structure of the monomolecular G-quadruplex formed by (almost) four repeats of the human telomere sequence, d[AGGG(TTAGGG)3], in the presence

Structure of Sequences Consisting of Human Telomere Repeats

63

Figure 3.4 (Plate 2) Polymorphism of four repeats of human telomeric sequence: (a) intramolecular G-quadruplex formed by d[AG3(T2AG3)3] as determined by NMR spectroscopy in the presence of Na+ ions in solution;92 (b) the G-quadruplex fold with parallel GGG strands and double-chain reversal loops as determined by X-ray crystallography for the same sequence crystallized in the presence of K+ ions;104 (c) the topology of the four-repeat sequence in K+ ion containing solution (so called Hybrid-1 structure);204,205,207 (d) the fold of major form of the unmodified four-repeat sequence in K+ ion containing solution (so called Hybrid-2 structure).210 For clarity, only G-bases and their syn (in orange) and anti (in blue) orientations across glycosidic bonds are shown (See colour plate section)

of Na+ ions, was reported several years ago.92 This monomolecular G-quadruplex is stabilized by the three stacked G-quartets with anti-syn-syn-anti conformations around individual G-quartets (Figure 3.4a). A core of the three G-quartets is held together by strands in alternating orientations. Two of the TTA loops connect the edges of the outer G-quartet on the same side of the structure. The third loop is a diagonal type loop and is positioned on the other side of the G-quadruplex core. Such loop arrangements restrict access to the outer G-quartets by potential ligands. The 5′- and 3′-ends of the oligonucleotide are found at the same end of the Gquadruplex (Figure 3.4a). The crystal structure of the K+ form shows a dramatically different fold with all four GGG segments being parallel (Figure 3.4b).104 All guanine residues are in the anti conformation. All three TTA loops are in a double-chain reversal conformation and connect the top of one GGG strand with the bottom of the other. The loop residues are positioned alongside the grooves rather than at the ends of the Gquartet stack and thus give the topology a propeller shape. The double-chain reversal

64


loops restrict access to the grooves, while the outer G-quartets represent large planar surface areas that were hoped to rationalize the selective binding of aromatic ligands. The 5′- and 3′-ends are at the opposite ends of the G-quadruplex (Figure 3.4b), which could support end-to-end stacking of successive G-quadruplexes in a larger DNA molecule. The crystal structure, in addition, revealed the position of K+ ions in the structure.104 In this 2.1 Å resolution crystal structure (PDB ID 1KF1) K+ ions are equidistant between the stacked G-quartets. The distances between K+ ions and each of the eight carbonyl oxygen atoms in the bipyramidal antiprismatic arrangement are 2.7 Å.104 Subsequently, several studies have demonstrated that an antiparallel Gquadruplex is formed by the human telomere repeat sequence in the presence of either Na+ or K+ ions in solution.74,192,200–202 In low salt conditions the human telomere DNA repeats tend to be either in disordered single-stranded states or parallel tetramolecular G-quadruplex structures.203 Determination of K+ ion structure in solution was elusive until recently when three groups reported the folding pattern of four repeats of human telomere sequence in the presence of K+ ions in solution.204–209 The main limitation for structure determinations was the existence of multiple conformers in equilibrium. The conformational equilibrium was driven into distinct predominant conformers that were amenable for structure determination by sequence modifications. Importantly, the modified oligonucleotide exhibited NMR spectral characteristics of the unmodified sequence. All these structures contain the (3 + 1) topology of the G-quadruplex core, in which three strands are oriented in one direction and the fourth is in the opposite direction (Figure 3.4c). The topology of this structure is different from the previously reported folds in the presence of Na+ ions in solution (Figure 3.4a) and K+ ions in the crystal (Figure 3.4b). This quadruplex contains three G-quartets with one exhibiting anti-syn-synsyn and two syn-anti-anti-anti conformations of guanine bases. One of the loops adopts a double-chain reversal structure, whereas the other two are edge-wise loops. In this fold, 5′- and 3′-ends of the sequence are located at opposite ends of the Gquadruplex core (Figure 3.4c). A related NMR study on a 26-nucleotide telomeric sequence d[(TTAGG)3TT] revealed that the wild-type human telomeric sequence adopts Hybrid 1 and Hybrid 2 conformations (Figures 3.4c and 3.4d) with a low energy barrier between them, which results in their coexistence in solution.210 Perusal of the topology of the Hybrid 2 structure shows the same (3+1) topology of the Gquadruplex core (Figure 3.4d). It consists of a single G-quartet with anti-syn-syn-syn and two with syn-anti-anti-anti conformations of guanine bases. The first two TTA loops closer to the 5′-end adopt an edge-wise structure, whereas the final loop adopts a double-chain reversal structure.

3.4 G-Quadruplexes Adopted by Promoter Regions Potential G-quadruplex-forming sequences are widespread in gene promoters. The analysis of bioinformatic data has shown that G-quadruplexes may be directly involved in gene regulation at the level of transcription. The promoter regions of

G-Quadruplexes Adopted by Promoter Regions 65

genes (1 kb upstream of the transcription start site) are significantly enriched in potential G-quadruplex motifs relative to the rest of the genome.21,211,212 Furthermore, the regions of the human genome that are both nuclease hypersensitive and within promoters show a remarkable 230-fold enrichment of G-quadruplex elements, compared to the rest of the genome.21 Sequences that have been studied for potential to form G-quadruplexes are c-myc,46,79,213–218 bcl2,219–222 VEGF,223 KRAS,224,225 RET,226 HIF-1a227 and c-kit.228–230 Structures adopted by these sequences, which are often composed of more than four G-tracts, are affected by their occurrence in the context of double-stranded DNA. The number of possible G-quadruplex structures is increased due to the fact that each of the G-tracts contains unequal number of guanines and that G-tracts are separated by different numbers of nucleotides. The c-myc gene and the respective protein are implicated in many cancers. The nuclease hypersensitivity element III1 controls up to 90% of total c-myc transcription. The guanine-rich sequence of this element contains six G-tracts, which result in multiple G-quadruplex folds that cannot be characterized individually by NMR. Early efforts in structure determination focused on the G-quadruplex that was formed by four of the six guanine tracts. Extensive search for suitable sequences afforded two sequences that exhibited sufficiently resolved NMR spectra, which corresponded to a single structure that enabled structure determination.231 One of the sequences named c-myc-2345 involved the second, third, fourth and fifth guanine tracts. The high-resolution structure of c-myc-2345 oligonucleotide in the presence of K+ ions has been determined by NMR (Figure 3.5a).231 This G-quadruplex represents an intramolecular propeller-type structure with all the strands in the parallel orientation. All guanines are in anti conformation. The structure consists of three G-quartets which are oriented in the same direction. All three loops are of doublechain reversal type. The structure resembles a parallel G-quadruplex with three double-chain reversal loops that was observed for the human telomere repeats in

Figure 3.5 (Plate 3) Representative structures of gene promoter regions: (a) G-quadruplex fold formed by guanine tracts of c-myc-2345 and c-myc-1245 sequences originating from the c-myc promoter, determined by NMR;231 (b) topology adopted by the c-myc-23456 sequence containing five of the six guanine tracts of the c-myc promoter region;232 (c) NMR structure of unprecedented intramolecular G-quadruplex formed by a G-rich sequence in the c-kit promoter in K+ solution233 (See colour plate section)

66


the crystal (see below and Figure 3.4b). An interesting feature of this G-quadruplex is the occurrence of two single nucleotide double-chain reversal loops, which were first found in the case of the c-myc promoter G-quadruplex.44,221,231 The loop in the middle of the sequence of c-myc-2345 contains two nucleotides (GA) while the other two loops consist of a single T. The topology of a related c-myc-1245 sequence is similar to the structure shown schematically in Figure 3.5a. The difference between the structures of c-myc-1245 and c-myc-2345 is in the loop length. In the former, loops closer to the 5′- and 3′-ends of the molecule contain a single residue (A or T), while the central loop contains six residues (T5A). Interestingly, the G-quadruplex adopted by c-myc-2345 is more stable by 15 °C. The difference can be attributed to different loop length and consequently distinct structures. An NMR structure determination of c-myc-2345 mutant (two guanines replaced by thymines) adopted the topology shown in Figure 3.5a.216 It is important to note that the three structures mentioned above improved the insight into the structural details of c-myc promoter by correcting some earlier low-resolution assessments.46 The recent study focused on the c-myc sequence containing five of the six guanine tracts.232 The sequence designated as c-myc-23456 consisted of the second, third, fourth, fifth and sixth guanine tracts of the 27-mer c-myc NHE III1 element. The topology of the G-quadruplex adopted in the presence of K+ is shown in Figure 3.5b. The structure consists of three G-quartets. Interestingly, guanines involved in the G-quartets originate from each of the five tracts (underlined here: 5′-dTGA GGG TGG GGA GGG TGG GGA AGG-3′). The four strands are in parallel orientations. All guanines are in the anti conformation except for the 3′-end guanine, which is a snap-back syn nucleotide. The loops are in a double-chain reversal conformation, except for the final one, which is a diagonal loop and contributes to the stability of the structure (Figure 3.5b). A new G-quadruplex fold was shown to be adopted by the c-kit promoter sequence positioned upstream of the transcription start site (Figure 3.5c).233 The ckit promoter encodes for a tyrosine kinase receptor and is therefore involved in a regulation of signal transduction. c-kit kinase is targeted in clinical treatments of several cancers and by small molecule inhibitors that promote G-quadruplex formation.228,229,233–235 The c-kit sequence that folded into a single structure in the presence of K+ ions consists of four GGG tracts. Its structure exhibits an unusual feature where an isolated guanine is involved in the formation of a G-quartet core, despite the presence of four G-tracts (guanines involved in the G-quartets are underlined here: 5′-dAGG GAG GGC GCT GGG AGG AGG G-3′).233 Strands are in parallel orientation and all guanines are in the anti conformation. The structure contains four loops. Two of them are single-nucleotide loops in a double-chain reversal conformation. The two-residue linker connects two edges of the outer G-quartet, while the longer five-residue loop allows the final two guanines to be inserted into the G-quartets. The latter loop contributes to the stability of the structure through Watson–Crick A-T base pairs. The 5′ and 3′-ends are on the opposite sides of the G-quadruplex (Figure 3.5c), which is potentially of relevance in the context of longer DNA as it allows for continuation of the sequence. The use of single molecule fluorescence resonance energy transfer enabled detection of G-quadruplexes in the context of extended DNA duplex.234

Bimolecular G-Quadruplexes 67

3.5 Bimolecular G-Quadruplexes of Analogues of the Oxytricha Telomere Repeat The NMR structure of the G-quadruplex formed by d(G4T4G4) in the presence of NaCl was one of the first G-quadruplex structures determined at high resolution.96,97 DNA oligonucleotide d(G4T4G4), which contains 1.5 repeats of the Oxytricha telomere folds into a well-behaved bimolecular G-quadruplex. The topology, shown in Figure 3.6a, exhibits a symmetrical bimolecular G-quadruplex with four G-quartets and thymine loops at opposite ends of the G-quartets. The glycosidic torsion angles are syn-syn-anti-anti around each G-quartet. Guanines are alternately in syn and anti conformations along each strand. Adjacent strands are alternately in parallel and antiparallel orientations. The two T4 loops span across diagonals of the outer G-quartets.97,236 NMR solution studies showed that d(G4T4G4) adopts the same

Figure 3.6 (Plate 4) Topology of analogues of the Oxytricha nova telomere repeat sequence: (a) the structure of bimolecular d[(G4T4G4)2] quadruplex was one of the first high resolution structures of G-quadruplexes determined;96 (b) unique folding topology of d[(G3T4G4)2] with (3 + 1) strand orientations;244 (c) topology of d[(G4T4G3)2];243 (d) asymmetric structure of d[(G3T4G3)2]238 (See colour plate section)

68


topology in the presence of K+ ions.98 A subsequent X-ray crystal structure of d[(G4T4G4)2] in the Na+ form, located in a cavity of the Oxytricha nova single-strand telomere-binding protein complex,158 revealed the same fold as the NMR structure. A recent X-ray study of the K+ form of d[(G4T4G4)2], which crystallizes in two crystal forms,99 also revealed identical folds to the NMR and the protein-associated structures. These results are likewise in accordance with the conformations deduced by Raman spectroscopy.237 The G-quadruplex formed by the related oligonucleotide d(G3T4G3) in the presence of K+ or Na+ ions has been determined independently by the groups of Feigon168,238 and Shafer239–241 by NMR in solution. d(G3T4G3) forms a bimolecular quadruplex (Figure 3.6d). Three G-quartets are formed from the GGG tracts. The conformation of guanine bases is syn-syn-anti-anti around each G-quartet. Adjacent strands are alternately in parallel and antiparallel orientations. Glycosidic conformations of the guanines are 5′-syn-syn-anti-(T4 loop)-syn-anti-anti in one strand and 5′-syn-anti-anti-(T4 loop)-syn-syn-anti in the other strand. The thymines in the loops span across the diagonal of the outer G-quartets. Unlike d[(G4T4G4)2], the bimolecular structure of d[(G3T4G3)2] is not symmetric, which results in separate signals for each monomer strand in NMR spectra. The unique feature of this topology is the stacking of the G-quartets, which includes both tail-to-tail and head-to-tail arrangements. The structures of d[(G4T4G4)2] (Figure 3.6a) and d[(G3T4G3)2] (Figure 3.6d) are closely related.97,98,185,240,242 However, the precise structure and dynamics of the T4 loops of both G-quadruplexes are cation dependent. A comparison of the Na+, K+ and 15NH4+ NMR structures of d[(G4T4G4)2] indicated that the coordination of Na+ ions within the planes of the outer G-quartets allows for the participation of the O2 carbonyl of the third loop thymine (T7) in the coordination of Na+ ions.98 In contrast, coordination of K+ or 15NH4+ ions between the planes of the G-quartets does not allow for coordination by the same loop thymine, which results in a different loop structure. When formed in the presence of K+ ions, there is also evidence of greater motion in the diagonal loops of the quadruplexes with respect to the Na+ structure.97,98,185,238,241 The sequence d(G4T4G3), which has one 3′ guanine missing from the sequence d(G4T4G4), has been shown to form a G-quadruplex with a significantly different structure and a greater cation-dependent polymorphism than d(G4T4G4).243 This sequence can be considered as an intermediate between d(G4T4G4) and d(G3T4G3). Our NMR study demonstrated that d(G4T4G3) forms a bimolecular G-quadruplex in the presence of Na+ ions. The topology of d[(G4T4G3)2], shown in Figure 3.6c, exhibits an asymmetric bimolecular fold-back structure consisting of three stacked G-quartets. The conformation of guanine bases is syn-syn-anti-anti around each Gquartet. The guanine bases of the two outer G-quartets exhibit a clockwise donor– acceptor hydrogen-bonding directionality, while those of the middle G-quartet exhibit anticlockwise directionality. The glycosidic torsion angle conformations of the guanine bases are 5′-syn–anti–syn–(anti-T4 loop)–anti–syn–anti in one strand and 5′-(syn)–anti–syn–anti–(T4 loop)–syn–anti–syn in the other strand. The two T4 loops both span diagonally across the outer G-quartets, but adopt different conformations. Each strand has neighbouring parallel and antiparallel strands. The two guanine residues not involved in G-quartet formation, G4 and G12 (i.e. the fourth

Coordination of Cations within d[(TG4T)4]

69

guanine base of one strand and the first guanine base of the other strand), adopt distinct conformations. G4 is stacked on top of an adjacent G-quartet and is part of the first loop. G12 is orientated away from the core of G-quartets and is stacked on the T7 base. The cation-dependent folding of the d[(G4T4G3)2] quadruplex structure is distinct from that observed for related sequences. While both d[(G4T4G4)2] and d[(G3T4G3)2] form bimolecular, diagonally looped G-quadruplex structures in the presence of Na+ and K+, we have observed this folding to be favoured for d[(G4T4G3)2] in the presence of Na+. The structure of d[(G4T4G3)2] exhibits a ‘slipped-loop’ element. The missing guanine residue at the 3′-end of d(G4T4G4), in terms of the primary structure, is compensated by slipping of the residues, which thus take its position within the 3D structure. Such a slipped strand apparently contributes to a thermodynamic stability of the structure. An NMR study on the folding of d(G3T4G4), a sequence with the 5′ terminal dG residue missing from the d(G4T4G4) sequence, established an unprecedented topology of a bimolecular G-quadruplex (Figure 3.6b).244 The structure consists of three G-quartets. The conformation of guanosine bases is anti-syn-syn-anti around one of the G-quartets, and syn-anti-anti-anti in the other two. Two guanine residues, G3 and G11, are involved in G-quartet formation. Three of the strands are in parallel, while one is in the antiparallel orientation. This type of topology resembles topological characteristics of (3 + 1) G-quadruplexes of the human telomere repeat sequences (see above). The outer G-quartets are spanned by diagonal as well as edgetype loops. Another unusual structural feature of d[(G3T4G4)2] is a leap between G19 and G20 over the middle G-quartet and chain reversal between G19 and G20 residues (Figure 3.6b). The examination of the influence of different monovalent cations on the folding of d(G3T4G4) showed that it forms a bimolecular G-quadruplex with the same general fold in the presence of K+, Na+ and ammonium ions.245

3.6 Coordination of Cations within d[(TG4T)4] Early studies on guanosine gels established a strong correlation between the melting temperature and the ionic radii of cations, which was an indication of site-specific ion binding by G-quartets.149 The strong interaction between cations and G-quartets originates from electrostatic interactions involving the guanine O6 oxygen atoms. On the other hand, electrostatic repulsions between cations within a G-quadruplex are substantial. Thus, the exact locations and coordination geometries of cations within a given G-quadruplex are the result of a balance between attractive interactions with carbonyl oxygen atoms and mutual cation repulsion. Cation coordination is not restricted to a particular geometry within a G-quadruplex (Figure 3.7). A series of stacked G-quartets produces a regular geometry, and potential cation coordination sites, with four O6 atoms within the plane of a G-quartet, or with eight O6 atoms between two stacked G-quartets. Ions such as K+ and 15NH4+ (ionic radii 1.33 Å and 1.43 Å, respectively) are too large to coordinate in the plane of a G-quartet, whereas Na+ (ionic radius 0.95 Å) is small enough to be coordinated within the plane of a G-quartet. Thus, as schematically illustrated in

70


Figure 3.7 Sodium ion localization within G-quadruplexes as revealed by X-ray crystal structures of d[(TG4T)4] with eight strands forming two sets of four G-quartets157

Figure 3.7, multiple cation coordination geometries are possible and observed (see below). The determination of G-quadruplex crystal structures of d[(TG4T)4] has provided a detailed view of cation coordination that is unmatched by other DNA structures.157 The X-ray crystal structure of d[(TG4T)4] was refined to a resolution of 0.95 Å (PDB ID 352D). This crystal structure confirmed that d(TG4T) forms a tetramolecular parallel-stranded G-quadruplex in the presence of Na+ ions (Figure 3.8a).157 The asymmetric unit of the d[(TG4T)4] crystal consists of four Gquadruplexes, with two quadruplexes being coaxially stacked in a 5′ to 5′ orientation, such that a continuous stack of eight G-quartets is formed.157,246 Seven Na+ ions are coordinated along the axial channel formed by the O6 atoms of eight G-quartets (Figure 3.8a). Close inspection of the crystal structure reveals that the central Na+ in this channel is positioned equidistant between the planes of two stacked G-quartets. The coordination geometry of this particular Na+ is bipyrimidal, with eight equidistant guanine O6 atoms. Na+ ions located away from this central position are displaced incrementally further from this perfect symmetrical bipyrimidal coordination. At the ends of this channel, terminal Na+ ions are in fact located in the planes of the end G-quartets, coordinated by the carbonyl groups of only four guanine residues (Figure 3.8a). Displacement of the Na+ ions from the central eight-coordinate geometry to the in-plane four-coordinate geometry is believed to arise from electrostatic repulsion between adjacent Na+ ions. The mutual repulsion of Na+ ions along the central axis of the G-quadruplex has also been hypothesized to be the origin of an out of plane bending observed for the terminal G-quartets of the eight stacked quartets.157

Coordination of Cations within d[(TG4T)4]

71

Figure 3.8 (Plate 5) Cation coordination within G-quadruplex structures adopted by d(TG4T) as revealed by X-ray crystallography: (a) Na+ ions (pink balls) within the 0.95 Å resolution structure (PDB ID 352D);157 (b) Na+ ions in the structure crystallized using lithium sulfate as the main precipitating agent (PDB ID 2O4F);246 (c) arrangement involving both Ca2+ (green balls) and Na+ (pink balls) cations crystallized in a mixed ion environment (PDB ID 2GW0);139 (d) Tl+ (red balls) and Na+ (pink ball) ions within the 2.2 Å resolution structure (PDB ID 1S45);247 (e) Tl+ (red balls) and Na+ (pink balls) ions within the 2.5 Å resolution structure (PDB ID 1S47);247 (f) Tl+ (red balls) and Na+ (pink balls) ions within the G-quadruplex exhibiting a T-quartet at the 3′-end of the strands (shown at the top of this structure) (PDB ID 1S47).247 Individual strands are shown in a wire frame model, while guanine bases are presented in stick representation to demonstrate the extent of out of plane bending. Cations are shown as coloured balls. Na+ with the smallest ionic radius can vary in coordination geometry from being within the plane of a G-quartet to being equidistant between two adjacent Gquartets. The larger ions are exclusively coordinated between two adjacent quartets (See colour plate section)

Analysis of details of the more recent crystal structure of d(TG4T) determined to 1.5 Å resolution (PDB ID 2O4F) shows that G-quartets within a bimolecular structure are not coplanar and are tilted with respect to the column axis towards the 5′-ends (Figure 3.8b).246 The crystallization of d(TG4T) from a solution containing both Li+ and Na+ ions showed only Na+ ions within the core of the quadruplex. A single bound Li+ ion has been observed at the surface of the bimolecular structure, but not within the G-quadruplex core.246 The high-resolution crystal structures of d(TG4T) also revealed the location of cations and water molecules located around the outside of the G-quadruplex.

72


Crystals of d(TG4T) were also formed from a solution that contained both NaCl and CaCl2.157 As in many other nucleic acid crystals divalent Ca2+ ions were found at crystal lattice contact points between nucleic acid assemblies (i.e. bridging between backbone phosphates of two G-quadruplexes). Single water molecules were found to provide a fifth oxygen atom for coordination of the outermost Na+ ions that are coordinated in the planes of the terminal G-quartets. Additionally, water molecules were found along the phosphate backbone, in the four grooves, and as part of the hydration shell of Ca2+ ions. There was no evidence that Ca2+ ions substituted for Na+ ions at any sites within the G-quadruplex. Crystallization from a solution containing MgCl2 resulted in no bound divalent cations.246 An earlier study established a single Mg2+ ion located at the surface of the quadruplex.247 In the recent crystallographic study139 on d(TG4T) a new crystal form for the tetramolecular quadruplex with four parallel strands has been found (PDB ID 2GW0). In these crystals, formed in a mixed Ca2+ and Na+ ion environment, the channel along the central cavity of the quadruplex dimer is populated by cations in an asymmetric way (Figure 3.8c). One half has three Na+ ions, whereas the other half has three Ca2+ ions. The ion channel is capped by water molecules. Na+ ions are irregularly positioned with respect to the G-quartet planes. The outermost Na+ is almost coplanar with a G-quartet, whereas the inner one is almost equidistant from two G-quartets. A Ca2+ ion sits at the dimer interface (itself in plane with four other Ca2+ ions each in a groove). Two Ca2+ ions are approximately midway between G-quartets.139 d(TG4T) has been found to produce two distinct crystal forms when crystallized from solutions containing both Na+ and Tl+ ions.247 The quality and resolution of data (i.e. 2.2 and 2.5 Å) combined with the very different electron densities of Tl+ and Na+ ions allowed these two cations to be distinguished in the G-quadruplex structure. One crystal form contained two G-quadruplexes in the asymmetric unit (PDB ID 1S45), whereas the other form contained three G-quadruplexes (PDB ID 1S47). In all cases, G-quadruplexes are stacked in a head-to-head fashion with a 5′ to 5′ orientation, as found in the Na+ crystal structure.157 The main difference lies in the position of metal ions within the G-quadruplex and the low occupancy of innerG-quadruplex cation coordination sites by Tl+ ions (i.e. occupancy levels between 0.15 and 0.70). The Tl+ ions coordinated within the G-quadruplex are positioned between two neighbouring G-quartet planes (Figures 3.8d–e).247 This coordination geometry would be expected for Tl+ ions, as the ionic radius of Tl+ (1.44 Å) is slightly larger than that of K+ and certainly too large to be coordinated within the plane of a G-quartet. Low Tl+ ion occupancy has been attributed to the higher concentration of Na+ ions in the crystallization solution.247 The tilt of G-quartets towards the 5′-end junction has been observed in the Tl+-containing structure with one dimer (PDB ID 1S45), and with one dimer and one monomer (PDB ID 1S47).247 It has been appreciated for some time that Sr2+ ions can promote G-quadruplex formation.130 The fact that Sr2+ ions can stabilize G-quadruplexes to a similar degree as K+ seemed somewhat enigmatic. The ionic radius of Sr2+ (1.13 Å) is in between that of Na+ and K+, however the energy from electrostatic repulsions between Sr2+ ions within a G-quadruplex would be four times that of monovalent cations, if Sr2+ ions were coordinated with a similar spacing along the central axis of a

Coordination of Cations within d[(G4T4G4)2]

73

G-quadruplex. The crystal structure of the RNA G-quadruplex formed by the sequence r(UG4U) in the presence of Sr2+ ions (PDB ID 1J8G), which was determined at ultra-high resolution (0.61 Å), seems to have provided an answer to this apparent enigma.160 Four strands of r(UG4U) form a parallel stranded quadruplex, similar to that described above for the analogous DNA sequence d(TG4T), but in the case of r(UG4U) the four strands within each quadruplex are symmetryrelated molecules. The tetramolecular G-quadruplexes of r(UG4U) stack with one another in opposite polarity (head-to-head or tail-to-tail) to form a pseudocontinuous column. Sr2+ ions are coordinated between every other pair of stacked G-quartets without any other cations coordinated at the intervening positions. The Sr2+ ions are associated with eight carbonyl oxygen atoms of adjacent G-quartets in a bipyramidal-antiprism geometry.160 Thus, it appears that the electrostatic repulsions between Sr2+ ions force these ions to leave vacant cation coordination sites within G-quadruplexes.

3.7 Coordination of Cations within d[(G4T4G4)2] 3.7.1 Crystal Structures The X-ray crystal structure at 1.86 Å resolution of the Oxytricha nova telomere end binding protein (OnTEBP) in complex with DNA and in the presence of NaCl has shown that d(G4T4G4) adopts a bimolecular G-quadruplex structure (PDB ID 1JB7)158 that is identical to the structure formed in the absence of the protein.99 Again, four guanine bases form four G-quartets, and four thymine residues form loops that span across the diagonal of the terminal G-quartets (Figure 3.6a). The structural features of the G-quadruplex in the protein–DNA complex are also very similar to those of the NMR structure,96,97 including the stacking of bases between G-quartets and the alternating syn-anti orientations across glycosidic bonds. The high-resolution X-ray diffraction data also reveals the positions of Na+ ions in the centre of the G-quadruplex.158 The two central Na+ ions are nearly coplanar within the central G-quartets and are in distorted, octahedral coordination environments (Figure 3.9a). The outer two Na+ ions are positioned above and below the planes of the outer G-quartets towards the T4 loops, and are coordinated by two O2 atoms of bases T5 and T7 in addition to the carbonyl oxygen atoms of the terminal G-quartet. Na+-thymine O2 coordination is consistent with the earlier proposal that changes observed by NMR in T nucleotide positions, when Na+ ions are exchanged for K+ or 15NH4+ ions, are driven by the loss of thymine O2 coordination.98 Electron density maps also revealed potential cation binding sites between the outside G-quartets and loops, which could not be unequivocally attributed to water molecules or Na+ ions.158 The bimolecular G-quadruplex formed by d(G4T4G4) in the presence of K+ ions has been crystallized in two different space groups and their structures have been refined to 2.0 and 1.49 Å resolution.99 One of these crystal structures (PDB ID 1JRN) has two G-quadruplexes per asymmetric unit, whereas the other crystal

74


Figure 3.9 (Plate 6) Localization of cations within bimolecular G-quadruplexes: (a) Na+ ions (pink balls) within the d[(G4T4G4)2] quadruplex in crystals of the Oxytricha nova telomere end-binding protein with telomere DNA (PDB ID 1JB7);158 (b) K+ ions (grey balls) within d[(G4T4G4)]2 as revealed by X-ray crystal structure at 2.0 Å resolution (PDB ID 1JRN);99 (c) K+ ions (grey balls) within d[(G4T4G4)2] as revealed by X-ray crystal structure at 1.5 Å resolution (PDB ID 1JPQ);99 (d) five Tl+ ions (red balls) within d[(G4T4G4)2] as revealed by X-ray crystal structure at 1.55 Å resolution (PDB ID 2HBN);129 (e) ammonium ions within d[(G4T4G4)2] as established by NMR measurements in solution;162 (f) two ammonium ions within d[(G3T4G4)2] established by NMR studies in solution.176 Individual strands are shown in wire frame model, while guanine bases are presented in stick representation to demonstrate the extent of out of plane bending. Na+ ion can be localized within the plane of a G-quartet or between two adjacent G-quartets. The larger K+, Tl+ and 15NH4+ ions are exclusively coordinated between two adjacent quartets (See colour plate section)

structure (PDB ID 1JPQ) has only one G-quadruplex per asymmetric unit. All three G-quadruplex structures exhibit the same bimolecular diagonal loop folding topology. This fold is identical to the structure determined for the same DNA sequence ten years earlier by NMR in the presence of both Na+ and K+ ions (Figure 3.6a).96–98 The central two G-quartets in each G-quadruplex are approximately coplanar. One guanine base in each terminal G-quartet is slightly more tilted than the others and stacks effectively with the adjacent 3′-thymine base (Figures 3.9b and 3.9c). The average rise between adjacent G-quartets is consistently 3.3 Å.99 A linear row of five equidistant K+ ions lies along the helical axis within the central core of all three G-


75

quadruplex crystal structures (Figure 3.9b). All K+ ions have full occupancy. The average K+–K+ distance is 3.38 Å. All three G-quadruplexes have three octahedrally coordinated K+ cations that are located equidistant between the planes of two adjacent G-quartets, with each K+ and eight O6 atoms forming a square antiprismatic arrangement. The outer K+ ions are located within the loops, where they also achieve octahedral coordination with the outer G-quartet O6 carbonyl oxygen atoms and the O2 atoms from the adjacent thymine bases, together with two oxygen atoms provided by water molecules. These potassium ions have only slightly increased mobilities (i.e. thermal factors) compared to the three central K+ ions.99 The crystal structure of the Tl+ form of the d[(G4T4G4)2] has been solved to 1.55 Å resolution (PDB ID 2HBN).129 This G-quadruplex contains five Tl+ ions (Figure 3.9d). The relative positions of these five cations are very similar to those found in the K+ crystal structure. The average distance between Tl+ ions is 3.6 Å. Three Tl+ ions are spread between adjacent G-quartet planes. They are coordinated by eight oxygens (one O6 from each of the surrounding guanines). Two Tl+ ions are coordinated in each of the two thymine loops (Figure 3.9d). They are coordinated by four guanine O6 atoms from the outer G-quartet plane and two thymine carbonyl groups. A water molecule was identified and located within the thymine loops and in close proximity to the loop-associated Tl+ ion.129

3.7.2 NMR Studies Ammonium ions have been shown to stabilize G-quartets to an extent that is similar to that observed for Na+.5,125 Hud and Feigon et al. used this fact to exploit 15NH4+ ion as a solution-state probe of monovalent cation coordination sites.162,248 These authors used 1H–1H and 1H–15N NMR correlation experiments to localize 15NH4+ ions within the bimolecular G-quadruplex d[(G4T4G4)2]. Dipole–dipole interactions (ROE) between the protons of the bound 15NH4+ ions and G-imino protons revealed that d[(G4T4G4)2] coordinates three 15NH4+ ions within its symmetric architecture (Figure 3.9e). In fact, only two distinct chemical shift environments were detected for bound 15NH4+ ions. One of them was assigned to the two symmetry-related outer sites, while the other was in the centre of the bimolecular G-quadruplex structure. The observed ROE cross-peak intensities also revealed that 15NH4+ ions in d[(G4T4G4)2] are positioned equidistant between stacked G-quartets (Figure 3.9e).162,176 NMR spectroscopy has also been used to monitor the competition between Na+ and 15NH4+ ions for coordination within G-quadruplexes at specific sites. For the G-quadruplex formed by d(G4T4G4) in the presence of 15NH4Cl, Na+ ions were shown to replace 15NH4+ ions at all three of the above mentioned outer- and innerbinding sites as the NaCl concentration is increased.163 Moreover, Na+ ions preferentially replace 15NH4+ ions coordinated at the inner site of d[(G4T4G4)2] quadruplexes.163 The 11-mer d(G3T4G4), a single base deletion of d(G4T4G4), has been shown by NMR spectroscopy to fold into an unusual asymmetric, bimolecular structure in the presence of 15NH4+, K+ and Na+ ions (Figure 3.6b).244 Multinuclear NMR studies

76


have demonstrated that the three G-quartets of d[(G3T4G4)2] create two cation coordination sites (Figure 3.9f).176 Analysis of NOE interactions between the protons of the bound 15NH4+ ions and G-imino protons revealed that ammonium ions are localized between the pairs of adjacent G-quartets within d[(G3T4G4)2] (Figure 3.9f). Subsequent titration of KCl into a solution of d[(G3T4G4)2] folded in the presence of 15NH4+ ions revealed a mixed mono-K+/mono-15NH4+ form that represents an intermediate in the conversion of the di-15NH4+ form into the di-K+ form.176 Similarly, 15NH4+ ions were found to replace Na+ ions inside this G-quadruplex. The preference for 15NH4+ over Na+ ions for the two internal coordination sites is considerably smaller than the preference for K+ over 15NH4+ ions. Interestingly, the two coordination sites within d[(G3T4G4)2] differ to such a degree that 15NH4+ ions bound to the site that is closer to the lateral-type loop are always replaced first during titration by KCl. That is, the second binding site is not occupied by a K+ ion until a K+ ion already resides at the first binding site.176 Recently, 23Na NMR demonstrated the presence of mixed cation species (Na+/K+, Na+,Rb+ and Na+/Sr2+) of 5′-GMP associates in solution.249 A noticeable progress in localization of cations has been made by Wu and coworkers who demonstrated that tightly bound metal ions can be detected directly by 23Na, 39K and 87Rb NMR in solution.249,250 The observation of quadrupolar nuclei with characteristic rapid relaxation that has traditionally116,117,251 led to prohibitively broad lines has been made possible with the appearance of NMR spectrometers with high magnetic fields where separation of signals for channel and free cations is achievable. 23Na NMR study has confirmed that three Na+ cations reside inside the bimolecular G-quadruplex d[(G4T4G4)2].252 There was no evidence of Na+ cations in the thymine loop. Solid-state NMR spectroscopy has also been used to localize cations within G-quadruplexes.150,252–254 Distinct 23Na chemical shifts for Gquadruplexes formed by self-association of 5′-GMP were interpreted based on ab initio calculations on a G-quadruplex model consisting of four stacked Gquartets and three channel cations. In this model each cation is sandwiched between two adjacent G-quartets which are separated by 3.4 Å and twisted by 45 °. The 23Na chemical shielding calculations with Na+ ion located within the G-quartet plane predicted 23Na chemical shift (d + 6 ppm), which does not agree with the experimental data (d − 17 ppm).249 It is noteworthy, however, that the localization of Na+ ions solely between the neighbouring G-quartet planes is not fully supported by the available crystal structures discussed above and shown in Figures 3.8 and 3.9. Tl+ has been shown to nicely substitute for K+ in the promotion of Gquadruplex formation.100,127,128 205Tl is a spin-½ nucleus with relatively high receptivity, in contrast to other monovalent cations. Its potential advantage could be a very large chemical shift range and thus high sensitivity to the chemical environment of a metal cation. As expected, the resonances of 205Tl+ ions associated with G-quartets were clearly resolved from the resonance of 205Tl+ ions in bulk solution. Strobel and coworkers have reported the NMR spectra for 205Tl+ ions in a solution containing the tetramolecular G-quadruplex d[(T2G4T2)4].127 The 205Tl NMR spectrum of this G-quadruplex shows a cluster of three peaks.127 The observation of three distinct 205 Tl peaks of approximately equal areas is consistent with a tetramer in which three Tl+ ions are localized between the layers of the four G-quartets. The ability of Tl+


77

ions to compete with Na+ ions for coordination within the G-quadruplex d[(G4T4G4)2] has been verified by solution-state 1H NMR.128 Differences in resonance peak line widths for the three Tl+ resonances suggests that ions move from one cation binding site to the other. The central peak was narrower, which was attributed to the slower exchange of the Tl+ ion at the middle coordination site. The outside Tl+ ions exchange with bulk solution more freely. The lifetime of Tl+ ions bound within the Gquadruplex was estimated to be at least 3 ms.127 One of the first NMR studies to determine the number of cations bound within a G-quadruplex in the solution state was performed using the sequence d(G3T4G3).168 In the presence of either NaCl or KCl, d(G3T4G3) forms a bimolecular fold-back structure (Figure 3.6d).238,240 1H NMR spectroscopy was used to follow the competition between Na+ and K+ ions for coordination sites within the G-quadruplex.168 Changes in 1H NMR spectra during these titration experiments indicated a gradual transition of d[(G3T4G3)2] from the G-quadruplex structure observed in the presence of only NaCl to the G-quadruplex structure observed in presence of only KCl. Although the Na+ and K+ forms of d[(G3T4G3)2] have the same molecular folds, there are relatively small structural differences between the two forms that are manifested by changes in 1H chemical shifts. No separate or additional 1H resonances were observed in samples of d[(G3T4G3)2] that contained various mixtures of NaCl and KCl, which demonstrated that the exchange rate of Na+ and K+ forms of this quadruplex at 25 °C is fast on the NMR timescale (ca. RNH2 > R2NH > R3N (R is an alkyl substituent),33 thus indicating that the steric hindrance and the hydrogen-bonding ability of the ligands can be important factors in determining the activity.34 The amines can act as hydrogen donors toward the O6 atom of a guanine or the phosphate groups of DNA.35 These interactions can be important from both a thermodynamic (stabilization of the Pt– DNA adduct) and a kinetic point of view (delivery of the Pt complex to the N7 of guanine). All these observations have resulted in a list of structural requirements for a platinum complex to be endowed with antitumour activity, the so-called structure–activity relationships (SAR):29 (1) Complexes should be neutral. (2) A cis geometry is required with general formula cis-[PtX2(amine)2] for platinum(II), and cis, trans, cis-[PtX2Y2(amine)2] for platinum(IV). (3) The X ligands (leaving groups) should be of intermediate strength (such as Cl−, SO2− 4 or carboxylate ligands). For platinum(IV) complexes the Y ligands should have trans orientation and can be Cl−, OH− or [OC(O)CnH2n+1]−. (4) The nonleaving amine ligands should contain at least one N-bound proton, necessary for hydrogen-bonding interactions with DNA (e.g. H-bonding to O6 of guanines or 5′ phosphate groups). All second and third generation platinum drugs (carboplatin, oxaliplatin) and those in the pipeline (satraplatin, picoplatin) satisfy the SAR given above. However it has been found that several series of platinum compounds which violate at least one of the above described structure–activity relationships are also antitumour active.36 For example, there are compounds with trans geometry37 or containing multiple metal centres that are antitumour active.38 Because of a different mode of interaction with DNA, such compounds are expected to exhibit a completely different spectrum of activity.

138

Platinum Drugs, Nucleotides and DNA

5.1.2 DNA Binding Because platinum belongs to the group of ‘soft’ metal ions, it reacts preferentially with N atoms rather than O atoms. In fact, the preferred binding site in DNA is N7 of purine nucleobases.39–45 It has been suggested that in vivo the positively charged cis-[Pt(NH3)2(OH2)2]2+ ion diffuses to the negatively charged DNA and then rapidly migrates along the helix to the preferred d(GpG) binding sites.46 A scheme with all potential binding sites is represented in Figure 5.2. In double-stranded DNA the only nitrogen atoms which are not involved in Watson–Crick hydrogen bonds and are available for coordination to Pt are N3 and N7 of purines. However N3 is sterically hindered, therefore only N7 is left (bold arrows in Figure 5.2). In single-stranded DNA N1 of adenine and N3 of cytosine are also capable of Pt binding (dotted arrows in Figure 5.2). Finally, all protonated nitrogen atoms of purines and pyrimidines have the lone pair of electrons delocalized on the aromatic p-system of the ring they belong or are attached to, therefore they become available for coordination to Pt only after deprotonation (empty arrows in Figure 5.2).

NH2

5 6 HO

4

3

cytosine (C)

N

2

1 N

O

O H

O H

H

H O

O

guanine (G) N 7 8 9 N

H

P

O

6 5 4

3 N

O H

O–

N 7 5 8 9 4 N

H O

H O

P O

NH2 NH 2

H

H

O

1 NH 2

O

P

adenine (A)

3 N O

5

H O

= preferential platination sites in double-strand DNA

N

H

H

= potential metal binding sites only after loss of a proton

1 2

O H

–

= additional metal binding sites in single-strand DNA

6

H

6

O

O

O–

H

H

OH

H

H

4 1 N

3 NH 2 O thymine (T)

H

Figure 5.2 Possible platinum binding sites on DNA

Introduction

139

N7 of both adenine and guanine can be platinated, however N7 of guanine shows a greater kinetic preference.35,42 This tendency results from the stronger basicity of that nitrogen and from possible simultaneous hydrogen-bond interactions between am(m)ine protons of cisplatin and O6 of guanine.34 In contrast, in the case of adenine, only repulsive interactions can be produced between the NH2 in position 6 of adenine and a platinum-bonded amine ligand.47 The reaction of cisplatin with DNA leads to the formation of at least six different types of Pt–DNA adducts, the most important being schematically depicted in Figure 5.3. The major adduct formed by cisplatin was found to be the 1,2intrastrand crosslink between adjacent guanine bases (d(GpG), 60–65%), followed by the 1,2-intrastrand crosslink between adjacent adenine and guanine bases (d(ApG), 20–25%); 1,3 and 1,4-intrastrand crosslinks between purines separated by one (d(GXG), X stands for a deoxyribonucleotide, 5–6%) or two intervening bases (d(GXXG)) can also occurr.48–51 It can be noticed how intrastrand d(GpG) and d(ApG) adducts together account for about 90% of the total. A small percentage of cisplatin was found to be involved in interstrand crosslinks between two guanines on opposite strands of the DNA, in monofunctional adducts in which the platinum

NH3 (b)

Pt NH3

NH3

Pt

(c) NH3

(d) NH3 NH3

(a)

Pt

NH3

NH3 Pt

PROTEIN

Figure 5.3 Main adducts formed in the interaction of cisplatin with DNA: (a) 1,2-intrastrand crosslink; (b) 1,3-intrastrand crosslink; (c) interstrand crosslink; (d) protein–DNA crosslink

140


moiety is coordinated to a single purine, and in DNA/protein crosslinks in which cisplatin coordinates a nucleobase and a protein molecule.49,52–56 The structure of platinated DNA is significantly distorted, resulting in a decrease in melting temperature,57 shortening,58 unwinding59 and local denaturation.60 Results obtained in numerous cell lines suggest that cisplatin-damaged DNA causes cell cycle perturbation and arrest in the G2-phase to allow for damage repair. In the case of inadequate repair, the cell eventually undergoes an abortive attempt at mitosis that results in cell death via an apoptotic mechanism.56,61–64 Transplatin is not able to form 1,2-intrastrand crosslinks because of the trans disposition of the two coordination sites; instead it may form 1,3-intrastrand crosslinks between two G residues separated by at least one base. The amount of monofunctional adducts and protein–DNA crosslinks is much higher for transplatin than for cisplatin. Another difference between the two isomers is that interstrand crosslinks formed by transplatin are between complementary G and C residues, whereas cisplatin forms only interstrand crosslinks between G residues. It was reported that, under physiological conditions, transplatin forms mainly monofunctional adducts, thus allowing the formation of protein–DNA crosslinks.64 In vivo, transplatin interstrand crosslinks might not form because of the slow rate (t1/2 ≈ 24 h) of the crosslinking reaction and the trapping of the monoadduct species by thiolcontaining amino acids.65,66 As stated earlier, DNA binding of platinum complexes is considered to be responsible for their cytotoxic effect. DNA containing platinum adducts is structurally distorted with respect to normal B-DNA, resulting in a loss of stability. The structural aspects of platinated DNA have been thoroughly investigated using NMR spectroscopy,67–71 X-ray crystallography72–74 and gel electrophoresis.75

5.1.3 Model Compounds cis-A2PtG2 complexes, in which A2 is a bidentate amine or two monodentate amines and G is a guanine derivative bound to platinum via N7, are the most simple models of the G/G crosslinks formed by cisplatin.76–83 However, even these most simple models can adopt conformations differing in the relative orientation of the two guanine bases. The G bases can be in a head-to-head arrangement (HH), in which both H8 atoms lie on the same side of the platinum coordination plane, or in a head-to-tail arrangement (HT), in which the H8 atoms lie on opposite sides of the coordination plane (LHT and DHT, Figure 5.4). In the solid state, the large majority of cis-A2PtG2 complexes adopt the conformation with the six-membered rings of the two guanines on opposite sides of the platinum coordination plane.84–89 In only a few cases (all complexes contained 9ethylguanine except one that contained 9-(2-methoxy-ethoxy)guanine) was the HH conformation found.85,90,91 In solution, cis-A2PtG2 complexes usually exhibit free rotation about the Pt–N7 bonds.92–95 In a pioneering study, Cramer demonstrated that guanosine rotation can be slowed sufficiently to detect atropisomers on the NMR timescale, if A2 is a sufficiently bulky chelate (e.g. N,N,N′,N′-tetramethylethylenediamine, Me4en).92 Two G

Introduction

N

N

N

N

N

141

N

Pt

Pt

Pt

∆HT

HH

ΛHT

Figure 5.4 DHT, HH, and ΛHT conformers. The G base is represented by an arrow, the tip indicating H8. The chiralities of the two HT conformers are defined according to the handedness of two straight lines: one perpendicular to the coordination plane and passing through the platinum atom and the second connecting the O6 atoms of the two Gs

H8 signals were detected. In an insightful analysis, it was shown that, when the cisA2Pt moiety has local C2 symmetry, the asymmetric sugar renders the two possible HT atropisomers magnetically nonequivalent, each with one H8 signal. Two H8 signals are also expected for the nonequivalent Gs of the HH atropisomer. The observation of only two of the four possible signals could be best explained by the presence of only the two HT atropisomers, consistent with most crystallographic results.84–89,96–99 The bulky A2 ligand(s) often lacked NH groups. Although N,N′dimethylethylenediamine (Me2en) complexes also formed HT atropisomers, the broad guanosine H8 signals revealed a rotation rate that was moderately fast on the NMR timescale.94 In conclusion, there is an intrinsic difficulty in investigating dynamic nucleos(t)ide complexes. The structure in the solid state may be very different from that in solution because of crystal-packing interactions. In solution, because of fast interconversion between possible conformers, only one set of signals, the average of those of individual conformers, is observed. 5.1.4 Retro Modelling The ‘dynamic motion problem’ led us to construct analogues of cisplatin with bulky ligands designed to reduce the dynamic motion by destabilizing the transition state for Pt-N7 rotation (Figure 5.5). An important feature of the design was to minimize steric effects in the ground state equilibrium species to allow conformers likely to be present in dynamic cis-A2PtG2 adducts to exist in the new adducts also. We introduced the term ‘retro modeling’76 to emphasize that the models we employed76–81,100–105 are more complicated than the relevant molecule, cisplatin. By reducing rotation rates by a billion fold,76,77 retro models have enabled us to understand the adducts of the highly fluxional cisplatin drug with DNA constituents.105 Retro model results76–78,80,81,100,101,103,104,106 have called into question two concepts that for many years were widely accepted from studies on dynamic platinum complexes: (i) untethered Gs adopt an HT conformation in preference to the HH conformation107 and (ii) single-stranded d(GpG) crosslinks favour the HH form, which undergoes slow rotation about the Pt-N7 bonds.108–112

142

Platinum Drugs, Nucleotides and DNA l Chelate-ring puckering

d Chelate-ring puckering

H

H

Me

Me N H

N Me

Pt

(S,R,R,S)-Me2dab-Pt

Me

Pt

H

(R,S,S,R)-Me2dab-Pt

H N

N

N

H

N

N

N

Pt

Pt

(S,R,R,S)-bip-Pt

(R,S,S,R)-bip-Pt

H

H

H

H H

N H

N Pt

N

N H

Pt

(S,R)-pipen-Pt

H

(R,S)-pipen-Pt

Me

Me

Me

Me N

Me

N Pt

N

N Me

Me

(R,R)-Me4dach-Pt Me

Pt

Me

(S,S)-Me4dach-Pt Me

Me

Me N

Me

N Pt

N

N Me

(R,R)-Me4dab-Pt

Me

Pt

Me

(S,S)-Me4dab-Pt

Figure 5.5 Me2dab, bip, pipen, Me4dach, and Me4dab ligands with in-plane bulk used to slow-down rotation about the Pt-N7 bonds in cis-A2PtG2 compounds

N,N′-dimethyl-2,3-diaminobutane (Me2dab) was the first carrier ligand used in such a detailed NMR retro model study.78,82,83 Me2dab differs from Me2en in having an additional methyl substituent on each carbon atom of the ethylene chain bridging the two nitrogens. These two methyl substituents not only increase the overall steric rigidity of the ligand, but also, by asymmetric induction, influence the configuration of the adjacent amine groups. Each N and each C in the ring is a centre of asymmetry. The Me2dab retro models provided the opportunity to define the solution structure of conformers by NMR methods, allowing us to identify the HH form of a cis-A2PtG2 adduct for the first time105 and also allowing us to use the known

Introduction

143

Me2dab configuration to define the absolute conformation of the HT forms in solution for the first time.78 The 2,2′-bipiperidine ligand (bip) is analogous to Me2dab, but is much more sterically rigid. This rigidity has two components: first, the chelate ring is part of a three-ring system, decreasing its fluxional character with respect to the Me2dab ligand; second, the CH groups projecting toward the G coordination sites are unable to rotate away from the G bases during G base rotation around the Pt-N7 bond. The bip ligand was able to decrease the dynamic motion roughly one billion times relative to (NH3)2 and ca. 100 times relative to Me2dab in cis-A2PtG2 complexes.76–78,101,104,106,113 Most of our work involved the C2-symmetrical isomers of Me2dab and bip, in which the chiral N, C, C and N centres in the chelate-ring had either S,R,R,S or R,S,S,R configurations. 5.1.5 Retro Models with Untethered Guanine Bases The use of Me2dab and bip ligands allowed the simultaneous observation by 1H NMR spectroscopy of all possible conformers (two HT and one HH conformer, Figure 5.6) that can be formed in cis-A2PtG2 complexes with a C2-symmetrical carrier ligand. Moreover, the conformers were present at equilibrium in different amounts depending upon the chiralities of the carrier ligands; thus these ligands were named chirality-controlling chelates (CCC). The relative amounts were the same for bip and Me2dab compounds with a given chirality and they differ only in dynamics.

Free 5'–GMP HHd

9.0

HTd

8.7

8.4

HHu

HTu

8.1

ppm

Figure 5.6 H8 region of the 1H NMR spectrum for (R,S,S,R)-Me2dab-Pt(5′-GMP)2 (Ref. 53 for related work)

144


The equivalent bases in each HT form had H8 chemical shifts intermediate between those of the nonequivalent bases in the HH form. Moreover the H8 signal of the major HT form appeared downfield from the H8 signal of the minor HT form.76–78,82,104 The H8 chemical shift is influenced by the positioning of the H8 atom with respect to the shielding cone of the cis G. The canting of the base in the direction that moves H8 away from the cis G will lead to less H8 shielding and hence to a downfield H8 signal relative to the average H8 signal (this relationship is called ‘6-in’ because the six-membered ring of G moves closer to the cis G). In contrast, a rotation of the base in the direction moving H8 toward the cis G will lead to greater H8 shielding,110 and hence to a more upfield H8 signal (this relationship is called ‘6-out’ because the six-membered ring of G moves further from the cis G, Figure 5.7). The G base of the HH form with O6 on the same side of the coordination plane as the N-H of the cis amine had an upfield H8 signal (HHu), indicative of an H8

S,R,R,S chirality of Me2dab or bip P

P

P O

N

N Pt N

N Pt N

P

P

O

O

O

Pt N H O

O P

ΛHT (major)

∆HT (minor)

HH

R,S,S,R chirality of Me2dab or bip P P

O

P

O

P N Pt N O

N Pt

N

O

N Pt

P

O

N O

P ΛHT (minor)

∆HT (major)

HH

Figure 5.7 HT and HH conformers in Me2dab or bip-Pt(5′-GMP)2 complexes having (S,R,R,S) or (R,S,S,R) chirality at the carrier ligand. N-CH2X groups (X = H and R for Me2dab and bip, respectively) are indicated by larger circles, N-H protons by smaller circles. Only the chelate ring atoms of the carrier ligands are shown. 5′-GMP is represented by an arrow, the tip indicating H8 and the tethered P the sugar-phosphate group

Internucleotide Interactions 145

pointing toward the cis G (‘6-out’ canting). The base with O6 on the same side of the coordination plane as the N-CH2X group of the cis amine (X = H or R for Me2dab and bip ligands, respectively) had a downfield H8 signal (HHd),76–78,82,104 indicative of an H8 pointing away from the cis G (no canting or ‘6-in’ canting). Differently from HH, in HT conformers the two bases are equivalent and the H8 of each G is on the same side of the coordination plane as the six-membered ring of the cis G. Under these circumstances the shielding effect of the cis G is expected to be significant even for no canting or moderate ‘6-in’ canting of the two nucleobases.

5.2 Internucleotide Interactions 5.2.1 Dipole–Dipole Interaction Between cis Guanines The major HT conformer found in Me2dab and bip-PtG2 adducts has G O6 located on the opposite side of the coordination plane from the cis amine N-H.78,104 The favoured form could not participate in a G O6/NH cis amine hydrogen bond, revealing that such an H-bonding might not be so important (Figure 5.7).100,104,114 The most plausible explanation for this behaviour is that stabilization of the HT conformers stems from dipole–dipole interactions between cis bases, and that this interaction is stronger for G bases having the six-membered ring of each guanine leaning towards the cis G (‘6-in’ canting). The ‘6-in’ canting, however, will bring the H8 of each guanine close to the cis amine and, if an N-Me substituent is on the same side of the platinum coordination plane as the H8, an unfavourable interaction between the two groups will limit the degree of ‘6-in’ canting and the stability of the conformer. This is the case for ‘minor’ HT conformers depicted in Figure 5.7. It is noteworthy that, in X-ray structures of HT conformers of cis-A2PtG2 complexes, much greater ‘6-in’ canting can be found when the H8 of each guanine is on the same side of the platinum coordination plane as an N-H of the cis amine115,116 than when it is on the side of an N-Me of the cis amine.117,118 A ‘6-in’ canting of ca. 20 ° (deviation from the guanine plane orthogonal to the coordination plane) will bring the electron-rich O6 atom of each guanine closer to the electron-deficient H8 atom of the cis base. Moreover, greater canting also reduces the dihedral angle between the planes of the two guanines (for a canting angle of 10 °, 20 °, 30 ° and 40 ° the dihedral angle is found to be 82 °, 74 °, 66 ° and 56 °, respectively) and possibly increases internucleotide stacking interactions. 5.2.2 Repulsion Between O6 Groups of cis Guanines In Me2dab and bip-PtG2 HH conformers, there is a large dispersion of H8 chemical shifts (∼1 ppm) that stems from a combination of electronic and steric effects. Electrostatic repulsion between electron-rich O6 atoms of cis Gs will tend to place the six-membered rings farther apart.119 The stereochemistry of the Me2dab and bip

146


ligands (for R,S,S,R or S,R,R,S configurations at the N, C, C and N asymmetric centres) places one N-H and one N-CH2X on each side of the platinum coordination plane. The guanine with the six-membered ring on the side of the coordination plane as the N-H of the cis amine can be canted ‘6-out’ because there is no clash with an amine substituent and O6 can act as an H-bond acceptor from the N-H. In contrast, the guanine with the six-membered ring on the side of the N-CH2X group of the cis amine will be much less canted and probably canted away from the cis amine (‘6-in’ canting). The former guanine will have a strongly shielded H8 signal, whereas the second guanine will have an unshielded H8 signal (HH conformers in Figure 5.7). Removal of the electron-rich O6 atom of guanine, such as in the case of adducts with benzimidazole (B), not only increases the yield of the HH conformer with respect to the HT conformers (the dipole–dipole interactions stabilizing the latter conformers are greatly reduced, if not absent), but also greatly reduces the dispersion of the B H2 chemical shifts even in the case of Me2dab or bip ligands (B H2 occupies the same position as H8 in G).119 This behaviour is sketched in Figure 5.8.

5.2.3 Phosphate N1H cis G H-Bond Interaction, SSC Two relevant types of ionizable groups are present in cis-A2Pt(GMP)2 adducts. First, the monoanionic phosphates of the GMP ligands [ROP(OH)O2]− which have pKa values (5.6 to 6.2)84–86,112 lower than for free 5′-GMP (6.3)76,79,81-83 and become fully deprotonated by pH ∼7.5.77–80,84 Second, the N1H of each G which deprotonate with pKa values of 8.7 and ∼9.479,83,87,109 versus 9.5 for free 5′-GMP.79,80,84 For both the R,S,S,R and S,R,R,S chiralities of the Me2dab and bip-PtG2 complexes, the percentage of the DHT conformer in the case of 3′-GMP and the percentage of the LHT conformer in the case of 5′-GMP were higher near pH 7 than at lower pH.76,79–81,102,105 We have unravelled the reasons for this difference in the preferred conformer.104 The phosphate group position is different in 5′-GMP versus 3′-GMP. For 5′-GMP adducts, the 5′-phosphate group projects toward the cis nucleotide in the LHT form, and away from the cis nucleotide in the DHT form (assuming that the nucleotide maintains the preferred anti conformation, Figure 5.9). In contrast, for 3′-GMP adducts the 3′-phosphate is close to the cis nucleotide in the DHT form and away from it in the LHT form (this can be seen in a gedanken experiment in which the phosphates of Figure 5.9 are detached from the 5′ position and attached to the 3′ position of the ribose). Thus, the relative positions of phosphate groups to the cis G in a HT form with a given chirality are opposite for 3′-GMP versus 5′-GMP adducts. On these grounds the key factors stabilizing the DHT conformer for 3′-GMP derivatives and the LHT conformer for 5′-GMP adducts near pH 7 appear to be the phosphate/N1H cis G interactions. This represents a third type of internucleotide interaction. We call such interligand interaction ‘second-to-second sphere communication’ (SSC) because the interacting groups are at the periphery of the cis nucle-

Internucleotide Interactions 147

Figure 5.8 Different ratio of HH and HT conformers in adducts with G (top) and 5,6 dimethylbenzimidazole (bottom). The scheme does not intend to take into account the chirality of the HT conformer (D or Λ) (Ref. 119 for related work)

otides.81,104,105 This is not the case for the previous two types of internucleotide interactions which involve parts of ligands close to the metal and can be referred to as ‘first-to-first sphere communication’ (FFC). SSC interactions are optimal at pH 6–7, conditions in which the phosphate group is deprotonated while N1 is still protonated. Further support for the conclusion that in cis-A2Pt(5′-GMP)2 complexes the LHT conformer is favoured by G phosphate/N1H cis G hydrogen bonding (SSC) came also from our recently reported X-ray crystal structure of Me4en-Pt(5′GMP)2.118 Three unique features not present in all other cis-A2Pt(GMP)2 solid-state structures were found: a LHT conformation, very strong hydrogen-bond interactions between phosphate and N1H of cis nucleotides, and a very small dihedral angle between the planes of the two guanines, although both were nearly perpendicular to the coordination plane.

148


platinum

hydrogen

nitrogen

carbonium

oxygen

phosphorus

Figure 5.9 (Plate 7) X-ray structures of two cis-A2Pt(5′-GMP)2 adducts showing how in the DHT conformer (top, based on data from Ref. 117) the 5′-phosphate is directed towards the cis amine, while in the ΛHT conformer (bottom, based on data from Ref. 118) the 5′phosphate is directed towards the cis guanine. For simplicity only the chelate ring atoms of the carrier ligands (Me4dach in the former case and Me4en in the latter case) are shown (See colour plate section)

Guanine cis Amine Interactions 149

5.3 Guanine cis Amine Interactions 5.3.1 Steric Interaction between G and cis Amine As already pointed out in Section 5.2.1, in HT conformers having ‘6-in’ conformation, the relevant steric interactions involve the G H8 and the cis amine. Hence the major HT conformer found for Me2dab and bip-PtG2 adducts has the guanine H8 located on the same side of the coordination plane as the cis amine N-H and on the opposite side from the cis amine N-CH2X group (‘major’ HT conformers in Figure 5.7). In the HH conformer (Section 5.2.2) the guanines tend to have ‘6-out’ conformation and therefore the relevant interactions involve the guanine six-membered rings and the cis amine substituents. In order to have moderately stable HH conformers it is required that at least one guanine has the six-membered ring on the side of the coordination plane where the cis amine has a nonbulky substituent (such as an N-H). Steric clash between the six-membered ring of each guanine and the cis amine substituent positioned on the same side of the platinum coordination plane renders the HH conformer particularly unstable in adducts with tertiary amines (such as Me4dab and Me4dach reported in Figure 5.5)117,118,120 or N-donor heterocycles (such as phenanthroline and 2,9-dimethylphenanthroline).121 5.3.2 Guanine O6-NH cis Amine H-Bond Interactions At basic pH (∼10), the N1H deprotonation eliminates the SSC phosphate/N1H cis G interactions that stabilize the LHT conformer for 5′-GMP adducts and the DHT conformer for 3′-GMP adducts (Section 5.2.3). On the other hand, N1H deprotonation greatly increases the capacity of the G O6 to act as a hydrogen bond acceptor. It was observed that, upon N1H deprotonation at basic pH, the less abundant HT conformer of the moderately dynamic Me2dab-PtG2 model compounds becomes more favoured (Table 5.1). (Such an increase in the concentration of the minor HT conformer at basic pH is not observed for the R,S,S,R configuration of the diamine and 5′-GMP nucleotide for the reason discussed in the next section.)113 The ‘minor’ HTs have the G O6 of each guanine on the same side of the platinum coordination plane as the N-H of the cis amine (Figure 5.7); therefore, their increase in concentration at basic pH is likely to stem from G O6/NH Me2dab hydrogen bonding, now favoured as a result of an increased capacity of the G O6 to act as a hydrogen bond acceptor. 5.3.3 Phosphate/NH cis Amine H-Bond Interaction, FSC When G has a 5′-phosphate group, conformer distribution could also be influenced by phosphate/NH cis amine hydrogen-bond interactions. Such an interaction should favour the DHT conformer because for this conformer the 5′-phosphate protrudes toward the cis amine (assuming the nucleotides have the favoured anti conformation, Figure 5.7). As noted in the previous section, at basic pH the LHT conformer is unexpectedly disfavoured (Table 5.1) for (R,S,S,R)-Me2dab-Pt(5′-GMP)2 although

150


Table 5.1 Conformer percentages of Me2dab-PtG2 complexes at different pH values measured from the H8 signal intensities. Increase of pH from ∼3 to ∼7 highlights the effect of SSC (stabilization of DHT in the case of 3′-GMP adducts and stabilization of ΛHT in the case of 5′-GMP adducts), while further increase of pH from ∼7 to ∼10 highlights the effect of guanine O6/NH cis amine H-bond interaction (stabilization of DHT in the case of S,R,R,S configuration of the diamine and stabilization of ΛHT in the case of R,S,S,R configuration of the diamine). The latter stabilization (ΛHT favoured at high pH for R,S,S,R configuration of the diamine) is not observed in the case of 5′-GMP because of simultaneous stabilization of the DHT conformer stemming from 5′-phosphate/NH cis amine H-bond interaction as outlined in Section 5.3.3. (S,R,R,S) G 3′-GMP 5′-GMP

(R,S,S,R)

pH

% LHT

% DHT

% HH

pH

% LHT

% DHT

% HH

3.4a 7.2a 9.6a 3.2 7.0 10.4

80 60 37 92 98 20

13 35 60 2 1 72

7 5 3 6 1 8

3.3 6.9 9.3 3.1 7.3 10.0

16 4 86 7 12 10c

84 96 14 71 70 69

b b b

22 18 21

a

At 5 °C; all others were at room temperature. Not determined. The percentage of this conformer (ΛHT) does not increase at the expenses of the other HT conformer (DHT) when the pH is raised to 10. b c

the G O6 and the N-H of the cis amine are on the same side of the platinum coordination plane (Figure 5.7). We believe that the nonincrease of the LHT concentration at basic pH is due to a contemporary increase in stability of the DHT conformer with the consequence that, on moving from neutral to basic pH, the conformer concentrations remain practically unchanged. The stabilization at basic pH of the DHT conformer most probably stems from a phosphate/NH cis amine interaction (as depicted in Figure 5.7). Indeed, for the DHT conformer of (R,S,S,R)-Me2dabPt(5′-GMP)2 we match the situation in which the 5′-phosphate protruding towards the cis amine is on the same side of the platinum coordination plane as the N-H of the cis amine. In Me2dab-Pt(3′-GMP)2 adducts, conformers cannot be affected by phosphate/NH cis amine interaction for stereochemical reasons because the 3′phosphate is too remote from the cis amine N-H. Because the 5′-phosphate/NH cis amine hydrogen bonding involves one part of the ligand close to the metal (the NH) and one part far from the metal (the phosphate), we have called this interaction ‘first-to-second sphere communication’ (FSC). This is not the case for the previous two types of guanine/cis amine interactions which involve parts of ligands close to the metal and can be referred to as ‘first-to-first sphere communication’ (FFC).

5.4 Solid-State Structures of Dynamic Nucleotides In the solid state, numerous examples of DHT complexes of 5′-GMP and several related 5′-phosphate derivatives have been found, not only for Pt with various

Conformer Distribution in Cisplatin Adducts 151

carrier ligands, but also for other metals.85–93,96,110,113,115 The nucleotides in the structures have very similar relationships, with the purine bases having the same relative positions and the phosphate groups always hydrogen-bonded to the cis ligands in a similar manner. Only very recently has a LHT complex of 5′-GMP been reported,118 while no example of a cis bisnucleotide HH complex has yet been published. This contrast between the prevalence of FSC in the solid and its relative unimportance in solution supports the premise of our retro model studies that the structures in the solid state of dynamic nucleotide complexes may be very different from the solution structures. Interligand interactions, as they result from our retro model studies, can be grouped in two distinct categories: nucleotide/nucleotide and amine/nucleotide. Internucleotide interactions concern: (1) Electrostatic attraction in HT conformers between electron-rich guanine O6 and electron-deficient H8 of cis Gs, both on the same side of the platinum coordination plane. This interaction is responsible for these conformers generally having the six-membered ring of each guanine leaning towards the cis G so to bring the two oppositely charged moieties closer to each other (‘6-in’ conformation). (2) Electrostatic repulsion in the HH conformer between electron-rich O6s, both located on the same side of the platinum coordination plane. This interaction generally leads to ‘6-out’ canting of the guanines. (3) H-bond interaction between nucleotide phosphate and N1H of cis G (SSC). In the case of 5′-phosphate this interaction will favour LHT (since in LHT each anti 5′-GMP has the 5′-phosphate protruding towards the cis G), while in the case of 3′-phosphate will favour DHT (since in DHT each anti 3′- GMP has the 3′-phosphate protruding towards the cis G). Amine/nucleotide interactions can also be of different types: (1) Steric interaction between amine and cis G. In HT conformers (having preferentially the ‘6-in’ conformation) the relevant interaction involves the guanine H8 and the cis amine substituents, while in the HH conformer (having preferentially the ‘6-out’ conformation) the relevant interaction involves the sixmembered ring of the guanine and the cis amine substituents. (2) H-bond interaction between guanine O6 and N-H of the cis amine (when present). This interaction becomes significant after deprotonation of guanine N1H at basic pH. (3) H-bond interaction between nucleotide 5′-phosphate and N-H of the cis amine when present (FSC). For HT conformers, assuming an anti conformation of 5′GMP, only D chirality allows for such an interaction.

5.5 Conformer Distribution in Cisplatin Adducts of G Derivatives NMR methods do not allow detection of individual conformers in fluxional cisA2PtG2 complexes with two ammines or a primary diamine ligand. Because of this dynamic motion problem, we relied heavily on CD spectroscopy to assess the chirality of the major HT conformer.

152


5.5.1 CD Signature of Conformers The presence of enhanced CD signals for some cis-A2PtG2 complexes was first reported in 1980,122 and attempts were later made to interpret the results on a structural basis.123,124 However, a convincing interpretation of the results has depended on subsequent studies with less dynamic complexes.77,78,107 The CD spectra of (S,R,R,S)-Me2dab-PtG2 complexes all had the same shape, i.e., positive peaks at ∼285 and 230 nm and negative peaks at ∼250 and 210 nm. Because all the 2D NMR studies showed that the dominant atropisomer was LHT (Figure 5.7), this type of CD signal was designated as L. The CD spectra of enantiomeric (R,S,S,R)-Me2dab-PtG2 complexes had signs opposite to those of the corresponding peaks in the (S,R,R,S)-Me2dab-PtG2 CD spectra (negative peaks at ∼285 and 230 nm and positive peaks at ∼250 and 210 nm). Because the 2D NMR studies showed that the dominant atropisomer for (R,S,S,R)-Me2dab-PtG2 complexes had the DHT conformation (Figure 5.7), the latter type of CD signal was designated as D. Furthermore, the CD signal intensity roughly correlated with the percentage of the major HT form.78 Studies with the related bip complexes indicated that a mixture of 25% DHT, 50% HH, and 25% LHT had essentially no CD intensity.77 The CD intensity increased as the favoured HT form became dominant with time. Similarly, in a ‘pH jump’ experiment performed on (S,R)-pipen-Pt(5′-GMP)2 (Figure 5.5) in which we monitored the change of atropisomer distribution by using both NMR and CD spectroscopy,79 changes in H8 NMR signal intensities corresponded to changes in CD intensity. On the basis of the above experiments, we interpret the sign of the CD signal as a reflection of the D or L conformation of the major HT form. In particular, the pattern of a positive Cotton effect around 285 nm and a negative Cotton effect around 250 nm is characteristic of a LHT conformer; the opposite pattern is characteristic of a DHT conformer.76

5.5.2 Conformers in Cisplatin Adducts with 3¢-GMP: The Role of SSC The CD signals of cisplatin adducts of G derivatives with no phosphate groups are weak. No conformer is favoured because the absence of phosphate groups rules out both FSC and SSC interactions; moreover the two ammine ligands have no chiral property that can exert a stereochemical control through FFC interactions so favouring a particular conformation. This analysis applies not only to cisplatin, but also to analogous A2PtG2 complexes with achiral primary diamines because these also have weak CD signals. In contrast, at neutral pH, cis-(NH3)2Pt(3′-GMP)2, pn-Pt(3′-GMP)2 (pn = 1,3diaminopropane), and en-Pt(3′-GMP)2 all showed the D-type CD signal,105 indicating that the major atropisomer is DHT (Figure 5.10). The carrier ligand has a modulating effect on the CD intensity: weakest for en-Pt, strongest for cis-(NH3)2Pt. The 3′-phosphate group from each 3′-GMP can form a hydrogen bond with the N1H of the cis 3′-GMP in the DHT conformation (assuming the nucleotides keep the normally preferred anti conformation), whereas such hydrogen bonding is less

Conformer Distribution in Cisplatin Adducts 153 a b

∆c (M–1 cm–1)

5

c

a) (NH3)2Pt(3'–GMP)2 b) pnPt(3'–GMP)2 c) enPt(3'–GMP)2 pH = 7

0

–5

250

nm

a) (NH3)2Pt(5'–GMP)2

8

a b) pnPt(5'–GMP)2 b

4 ∆c (M–1 cm–1)

300

c) enPt(5'–GMP)2 pH = 7

c 0

–4

–8 250

300

nm

a) enPt(1–Me–5' GMP)2 b) pnPt(1–Me–5' GMP)2 pH = 7.8

a 6 4 ∆c (M–1 cm–1)

b 2 0 –2 –4 –6 250

300

nm

Figure 5.10 CD spectra of cis-A2PtG2 compounds when A2 = (NH3)2, pn, or en and G = 3′GMP, 5′-GMP, or 1-Me-5′-GMP (Ref. 105 for related work)

154


favourable in the LHT conformation. The preference for the DHT atropisomer for these three 3′-GMP complexes appears to be due to such a hydrogen bonding (SSC). Raising the pH from 3 to 7 increased the D type CD signal intensity, indicating that the DHT form is favoured upon 3′-phosphate group deprotonation, most likely due to formation of stronger phosphate/N1H cis G hydrogen bonds at neutral pH. Above pH 7 the D type CD signal starts to decrease, and the intensity is nearly zero at pH 9. At pH ∼9, where N1H deprotonation occurs,125 the N1 atom has lost its proton and no phosphate/N1H hydrogen bond can exist. Our NMR/CD studies with pipen, Me2dab and bip-Pt(3′-GMP)2 complexes support this interpretation.76,102

5.5.3 Conformers in Cisplatin Adducts with 5¢-GMP: Interplay between SSC and FSC At neutral pH, the CD signals of cis-(NH3)2Pt(5′-GMP)2, pn-Pt(5′-GMP)2 and enPt(5′-GMP)2 are of the L type, indicating that the major conformation is LHT (Figure 5.10).105 Intramolecular hydrogen bonding between the phosphate group of one 5′-GMP and the N1H of the cis 5′-GMP (SSC) is more favourable in the LHT than in the DHT conformer. The CD data point to substantial differences in signal intensity among 5′-GMP adducts of different carrier ligands [cis-(NH3)2, pn, en]. One obvious potential factor is the bite angle of the carrier ligand(s). The very low CD intensity of the en-Pt(5′GMP)2 complex suggests that the low bite angle of the en ligand reduces the strength of the SSC interaction. Raising the pH from ∼3 to 7 increased the L type CD signal intensity, indicating that the LHT form is more favoured upon 5′-phosphate group deprotonation, possibly because of stronger phosphate/N1H hydrogen bond at neutral pH. As the pH was raised further, the CD signals decreased in intensity and partially inverted at pH 10, suggesting a bias toward the DHT form after N1H deprotonation.105 The preference for the LHT atropisomer up to pH 9 is likely to stem from the 5′phosphate group of each 5′-GMP forming a hydrogen bond with the N1H of the cis 5′-GMP (SSC). In the LHT conformation the 5′-phosphate protrudes toward the cis nucleotide (assuming the nucleotides keep the preferred anti conformation). On the other hand, the bias towards the DHT form observed at pH 10 can be explained as follows. In the DHT conformation the 5′-phosphate is directed towards and can directly reach the cis amine and form an H-bond (FSC). It is worth nothing that the latter hydrogen bonding interaction is found in relevant X-ray structures. It is therefore most likely that the presence of 5′-phosphate/NH cis amine hydrogen bonding is responsible for the stabilization of the DHT form at high pH (such hydrogen bonding is masked at acidic and neutral pH by the dominating SSC interactions). 5.5.4 Conformers in Cisplatin Adducts with 1-Me-5′-GMP: The Role of FSC The CD signals of pn-Pt(1-Me-5′-GMP)2 and en-Pt(1-Me-5′-GMP)2 had signs opposite to those of 5′-GMP adducts (Figure 5.10).105 The signs indicate that the major atropisomer of the 1-Me-5′-GMP adducts is DHT.

Conformer Distribution in Cisplatin Adducts 155

Because no phosphate/N1H hydrogen bond is possible for 1-Me-5′-GMP derivatives (SSC), the only important interaction influencing the HT atropisomer population is phosphate/NH cis amine H-bonding (FSC). The DHT conformer has the 5′-phosphate protruding toward the cis amine and able to form a phosphate/NH cis amine H-bond. This conclusion is also supported by the observation that in the absence of a cis G (which rules out any SSC interaction), such as in the complex Me3dien-Pt(5′GMP) (Me3dien = N,N′,N′′-trimethyldiethylenetriamine, with all three methyl groups on the same side of the platinum coordination plane), H-bond formation between the phosphate and the N-H of the cis amine becomes the dominant interaction.126,127 5.5.5 Conformers in Oxaliplatin Adducts with Guanines Lacking a Phosphate Group: The Role of FFC In an early investigation of the reaction of enantiomeric dach platinum complexes with guanine derivatives,123 Pasini and coworkers observed that the CD spectrum of (R,R)-dach-PtG2 (G = 9-methylguanine) was characterized by a positive Cotton effect centred at 230 and 280 nm and a negative Cotton effect centred at 260 nm; the corresponding Cotton effects for the (S,S)-dach-PtG2 enantiomer had the opposite sign (Figure 5.11). The Cotton effect was assigned to coupling between p–p* electronic transitions centred on the guanine bases and was taken as a clear indication of transmission of chirality from the dach ligand to the coordinated cis guanines mediated by the amine protons.123 The details of the mode of transmission of chirality given at that time were not convincing and have remained obscure until very recently.128 cis-A2PtG2 complexes, with G = a guanine derivative lacking a phosphate substituent, can have only FFC interactions. However, whereas each N in retro-model carrier ligands had substituents of very different bulk on the two sides of the

(R,R)DACH-Pt(9–EtG)2 (S,S)DACH-Pt(9–EtG)2

∆ε/ (mol–1 cm–1)

2 1 0

200

250

300

λ / nm

350

Figure 5.11 CD spectra of (R,R)-dach-Pt(9-EtG)2 (broken line) and (S,S)-dach-Pt(9-EtG)2 (solid line) at pH 7 and room temperature

156


platinum coordination plane, the dach ligand has an N-H on each side of the plane. Nevertheless, a slight difference in the stereochemistry of the two protons (one N-H has ‘quasi axial’ and the other ‘quasi equatorial’ character) is sufficient to induce a significant change in the relative stabilities of the dach-PtG2 DHT and LHT conformers. At acidic and neutral pH the stability of HT conformers is governed by G to G dipole–dipole interaction, which is greater for the six-membered ring of each guanine leaning toward the cis G. Such a ‘6-in’ canting of the two guanines can be hampered by the steric interaction between the H8 of each guanine and the substituent on the cis amine that is on the same side of the platinum coordination plane. Such a repulsion is greater for a ‘quasi equatorial’ N-H than for a ‘quasi axial’ N-H. Therefore the DHT conformer is stabilized in the (S,S)-dach-PtG2 complex, while the LHT conformer is stabilized in the (R,R)-dach-PtG2 species (this can be seen in Figure 5.7 if the larger circles now represent ‘quasi equatorial’ protons and the smaller circles ‘quasi axial’ protons). Indeed the CD spectrum is typical of a DHT conformer in the former case and of a LHT conformer in the latter case.128 At basic pH, deprotonation of the guanine N1-H renders the O6 a much better H-bond acceptor; therefore, the stability of the HT conformers is governed by the guanine O6/NH of the cis amine H-bond interaction. Such a guanine O6/NH cis amine interaction is more favourable for a ‘quasi axial’ than for a ‘quasi equatorial’ N-H. As a consequence, at basic pH the CD spectrum of (S,S)-dach-PtG2 has the signature of a LHT conformer, while the CD spectrum of (R,R)-dach-PtG2 has the signature of a DHT conformer (this can also be seen in Figure 5.7, assuming the larger circles to represent ‘quasi equatorial’ protons and the smaller circles ‘quasi axial’ protons).128 5.5.6 Quantifying the Atropisomer Distribution in Adducts of Cisplatin or Analogous A2Pt(5¢-GMP)2 Complexes with Achiral Primary Diamines CD spectroscopy is an empirical method at the current state of theoretical understanding of CD spectral transitions in these types of complexes, therefore it cannot be used for a quantitative estimate of different conformers, for which we have to rely upon NMR data. Very recently, by lowering the temperature down to 238 K, it has been possible to observe different conformers of (cis-1,4-dach)Pt(5′-GMP)2 (1,4-dach = 1,4diaminocyclohexane for which the N–Pt–N bite angle is ≥97 °). The percentages of the HH, LHT and DHT conformers were estimated from intensities of 1H NMR signals to be 33, 57 and 16%, respectively. This is the first case in which the abundance of the three possible rotamers has been determined in A2Pt(5′-GMP)2 adducts with primary diamines.129

5.6 Retro Models Applied to Adducts with Tethered Guanine Bases The cis-(NH3)2Pt(GpG) and cis-(NH3)2Pt(d(GpG)) complexes, the simplest Glinked models of the major cisplatin–DNA adduct, have been characterized by a

Retro Models Applied to Adducts 157

number of techniques, including CD spectroscopy,109,130 1H NMR spectroscopy108,130 and molecular modeling.110 On the basis of the observation that both nucleotides of each complex were anti (Figure 5.12),130 the crosslinks were initially assigned structures in the HH1 conformer class. Studies of retro models have shown that there are three other classes, namely, HH2, DHT1, and LHT2 (Figure 5.13). Members of adjacent classes are separated by ca. 180 ° rotation of one G about its Pt-N7 bond. Variants in nonHH1 conformer classes have been found for bip-Pt(d(GpG)) and bip-Pt(GpG) adducts.101,103,106 Moreover, the bases can have right-handed (R) or left-handed (L) canting (handedness of canting is defined by the following two straight lines: one connecting the platinum-coordinated N7 atoms of the two guanines and the second connecting the N1 and C8 atoms of a given guanine). The base canting is determined by the chirality of the amine: (S,R,R,S)-bip induces left-handed canting (L, upper row of Figure 5.7) while (R,S,S,R)-bip induces right-handed canting (R, lower row of Figure 7) of the guanine bases. (S,R,R,S)-bip-

O N7

N

O H3' H2'

H5'' H4'

NH2

N

H1' O

H2''

O

O

(a) N7

P O

H

H8

OH

H5'

N

H5'

H

N

H8

O

N

O H3' H2'

H5'' H4'

NH2

N

H1' H2''

HO

H H2N

O N

OH

N

H H H

N

O H H HO

H H H

(b)

N

O

H NH2

N

N

OH H H

N

N

O H

H

H

H

H

H

HO

(c)

Figure 5.12 Structure of d(GpG) (a) and schematic anti (b) and syn (c) guanine glycosidic torsion angle alignments. The H8 and sugar H1′ protons are closer to each other (2.5 Å) in the syn alignment compared to a longer distance (3.7 Å) in the anti alignment, a feature that can be readily monitored by NMR

158


HH1

∆HT1

HH2

platinum

carbonium

nitrogen oxygen

phosphorus

Figure 5.13 (Plate 8) Minimum energy models (from NMR-restrained MMD calculations) of the anti/anti HH1 L (top) and anti/syn DHT1 L (centre) forms of (S,R,R,S)-bip-Pt(d(GpG)) and the anti/anti HH2 R (bottom) form of (R,S,S,R)-bip-Pt(d(GpG)) (the conformation of the 5′-G is given first, then that of 3′-G; only the chelate-ring atoms of the bip ligand are shown). Based on data from Refs 101 and 106 (See colour plate section)

Pt(d(GpG)) and (S,R,R,S)-bip-Pt(GpG) both favoured two variants, anti/anti HH1 L (with N and S puckers for the 5′ and 3′-G residues, respectively) and anti/syn DHT1 L (both sugars having mainly N puckers, top and centre molecular models in Figure 5.13).103,106 When these complexes were kept at pH 10 for several days, the DHT1 L variant became favoured over the HH1 L variant, demonstrating that DHT1 L is more favourable when N1 is deprotonated.103,131 For (R,S,S,R)-bip-Pt(d(GpG))


at physiological pH, the two major conformers were anti/anti HH1 R and anti/anti HH2 R (bottom model in Figure 5.13). For both conformers the sugar pucker was N and S for the 5′ and 3′-G, respectively.101 When the sample was kept for 1 day at pH 10, where N1 was not protonated, the LHT2 R conformer became ∼30% abundant.103 In contrast to the derivative with d(GpG), for (R,S,S,R)-bip-Pt(GpG) only the HH1 R variant was observed at any pH, and only a small population of LHT2 R was observed at pH 10.103 5.6.1 Interligand Interactions Stabilizing Different Conformers in Adducts with Tethered Guanines Investigations of adducts of retro models with d(GpG) and GpG dinucleotides, not only revealed the dynamic behaviour of these intrastrand crosslink models, with the possibility for each guanine to undergo ca. 180 ° rotation about the Pt-N7 bond, but also highlighted the role of ‘weak’ interactions, not unlike those discovered in the investigations with untethered cis guanines. In particular, in the HH conformer, a key interaction appeared to be the repulsion between the electron-rich O6 atoms of the two guanines. Because of such repulsions, the six-membered ring of each guanine is forced to move further away from the cis guanine and closer to the cis amine (‘6-out’ relationship). If the cis amine has an N-H on the same side of the platinum coordination plane as the O6 of the guanine, an H-bond could possibly be formed.101 Between HT conformers, the DHT1 L appears to be a major component in the case of adducts with the (S,R,R,S)-bip ligand, which in theory can provide two N-Hs for H-bond formation with the O6 of the two guanines (this can be seen in Figure 5.7, in which, however, the two guanines are not tethered).106 The LHT2 R conformer can also form two guanine O6/NH cis amine H-bonds in adducts of (R,S,S,R)bip (this can also be seen in Figure 5.7, although here the two guanines are not tethered); however, at acidic pH the LHT2 R conformer is a minor form compared to the HH1 R and HH2 R (the latter only for the d(GpG) adduct) conformers, thus indicating that the LHT2 R conformer is intrinsically less stable. The LHT2 R conformer concentration increases at high pH, and this could be attributed to O6 becoming a better hydrogen-bond acceptor when N1 is deprotonated.103 5.6.2 Multiple Conformers in Cisplatin Adducts with Tethered Guanine Bases The cis-(NH3)2Pt(GpG) complex is probably very dynamic, undergoing rapid rotation about the Pt-N7 bonds. The H8 signals of cis-(NH3)2Pt(GpG) are relatively sharp (∼10 and 4 Hz for 5′-G H8 and 3′-G H8, respectively) at 21 °C but broader (∼20 and 5 Hz, respectively) at 5 °C.103 Therefore, exchange between conformer classes or between variants within a conformer class must be moderately fast at 21 °C and slower at 5 °C. Comparison to spectral data for bip-Pt(GpG) indicates that cis-(NH3)2Pt(GpG) exists as a mixture of variants from both the DHT1 and HH1 conformer classes;103 thus, the most likely explanation for the broadening is exchange between conformer classes. Analysis of the NMR shifts and CD signal

160


shapes indicates that at low pH cis-(NH3)2Pt(GpG) exists mostly as the HH1 R conformer, as suggested previously in the literature;123 however, other conformers account for perhaps one-third of the complex. At high pH, the conformer distribution changes. Although it appears to be reasonable that the HH1 L variant is now a more abundant form, it is not the exclusive form. Other forms likely to be present include the DHT1 form.103 As found for the GpG adduct, cis-(NH3)2Pt(d(GpG)) also exists as a dynamic mixture of various forms of comparable populations, in contrast to the adducts with longer oligonucleotides, which exist mainly as the HH1 L variant.106 In the solid state, cis-(NH3)2Pt(d(pGpG)) crystallizes with an equal population of HH1 R and HH1 L variants.132 We believe that a mixture of these two variants also exists in solution. In addition, the computed anti/syn DHT1 model of cis-(NH3)2Pt(d(GpG)) is also structurally very similar to that computed for the analogous bip adduct. Furthermore, we computed that this conformer has one of the lowest energies of the cis-(NH3)2Pt(d(GpG)) models and we have found this form in many adducts containing a wide variety of carrier ligands; therefore, it is most likely that the DHT1 form is also part of the dynamic mixture of cis-(NH3)2Pt(d(GpG)) conformers. 5.6.3 Wrapping of the Positively Charged Metallic Core by Flanking Residues in Cisplatin Adducts with Single-Stranded Oligonucleotides With the exception of very simple species such as cis-(NH3)2Pt(d(GpG)), an unusual but characteristic feature of larger oligonucleotides crosslinked at N7 by platinum(II) is a large difference (∼1 ppm) in the H8 signals, with the 3′-G H8 signal downfield in single-stranded adducts. The guanine orientation dictated by the sugar–phosphate backbone is responsible for the shift difference.108 The same remarkable difference in H8 chemical shifts of the HH conformer was found in Me2dab and bip-PtG2 complexes with untethered guanines, suggesting that interligand interactions in the coordination sphere may dictate guanine orientations comparable to those induced by the sugar–phosphate backbone of the GpG moiety within larger oligonucleotides. On the basis of the unusual 31P NMR signals observed in the LPt– d(TCTCGGTCTC) adducts (L = cis-(NH3)2 or en), it was concluded that this G/G crosslinked single strand is a distorted coil with the 5′ residues ‘wrapped around’ the Pt binding site.133 The 5′-region distortion is probably caused both by electrostatic interactions between the negatively charged nucleotide sequences flanking the crosslink and the cationic [cis-(NH3)2Pt]2+ or [en-Pt]2+ moieties and by hydrogenbond formation between the phosphate and the Pt-NH groups; the resulting local distortion at phosphodiester groups explains the rather unusual 31P NMR shifts.134 Such ‘wrapping’ of the cationic [cis-(NH3)2Pt]2+ moiety by flanking residues results in more stable coils and in local ‘melted’ regions of longer DNA, thereby decreasing the Tm of DNA.134,135 Treatment of the self-complementary dodecamer d(A1T2G3G4G5T6A7C8C9C10 A11T12) (G3) under conditions in which G3 is in the duplex (DS) form with cisplatin or en-PtCl2 gives single-stranded (SS) products: G3/G4 adducts in a coil form and G4/G5 adducts in a hairpin form stabilized by a favourable A7-G4 stacking interaction at the loop–stem junction (Figure 5.14).67,136–138


Figure 5.14 Schematic representation of the reaction of G3 (d(A1T2G3G4G5T6A7C8C9C10A11 T12)) with: (A) platinum complexes with secondary nitrogen donors having steric bulk (small loops close to N) or (B) platinum compounds without steric bulk on the nitrogen atoms (Ref. 139 for related work)

The ‘wrapped’ SS coil form becomes less stable in bip-Pt–G3 and Me2dab-Pt– G3 adducts because the hydrophobic and bulky bip and Me2dab carrier ligands prevent the flanking regions of the negative DNA strand from approaching the cationic Pt centre as closely as these regions can approach the Pt in adducts with less hindered bis-ammine or primary diamine ligands. These effects involving the bip and Me2dab ligands while destabilizing the ‘wrapped’ form, favour the duplex form; therefore, the bip-Pt–G3 and Me2dab-Pt-G3 adducts with the crosslink in a HH1 conformation favour the duplex form (Figure 5.14).139 This is particularly true for (R,S,S,R) configuration of the diamine which induces the characteristic R canting of platinated guanines in duplexes. In the case of (S,R,R,S) configuration of the diamine, the duplex appears to be in equilibrium with a small amount of a single-stranded (most likely coil) form. The facts that the (S,R,R,S) chirality favours L canting and that this canting is characteristic of coils suggest a reasonable explanation for this result. An important outcome of this investigation is that the conformation and annealing propensity of SS oligonucleotides with G/G intrastrand crosslinks can be modulated by the stereochemistry of the platinum carrier ligands. 5.6.4 Flexibility in Cisplatin Adducts with Double-Stranded Oligonucleotides A few X-ray structures of double-stranded (DS) oligonucleotides containing a cisA2Pt moiety (A2 = two ammines or a diamine) crosslinking two G residues of the same strand72,73,140,141 or of opposite strands74 have been reported. In the intrastrand

162


view from the major groove

view from the minor groove

platinum

nitrogen oxygen


Figure 5.15 (Plate 9) Central portion (four base-pairs) of oligonucleotides containing G/G intrastrand (top) and interstrand (bottom) crosslinks (based on data from Refs 114 and 143, respectively) (See colour plate section)

crosslinks (intra-CL) the cis-A2Pt moiety is located in the major groove, and the six-membered rings of the two guanines are on the same side of the platinum coordination plane (head-to-head, HH, conformation, Figure 5.15). In contrast, when platinum crosslinked G residues of opposite strands (inter-CL), the cis-A2Pt moiety

Conclusions and Perspectives

163

is located in the minor groove, and the six-membered rings of the two guanines are on opposite sides of the platinum coordination plane (head-to-tail, HT, conformation, Figure 5.15). Solution NMR data have been interpreted by analogy with the results of the X-ray investigations.46,70,114,142,143 A detailed description of these adducts appears in some excellent reviews.144–146 However, even in the case of DS oligonucleotides, which are, per se, somewhat rigidly structured, some dispute exists as to the fine structure of more dynamic regions, and the contributions of the sugar–phosphate backbone and of interligand interactions in determining the overall stereochemistry remain rather uncertain. For instance, no two NMR studies on duplexes have led to the same proposed structure46,68,70,107,147 and the diverse proposed solution structures generally lack the features found in the X-ray structure of a 16-mer bound to rat HMG1.148 These variations in structure may be caused by spectral complications arising from dynamic motion centred at the crosslink. For instance, the dihedral angle between crosslinked guanines has been found to span all values between 25 ° and 75 °. However, in a relatively recent study of a 9-mer duplex, an NMR-based structural model with many of the elements of the X-ray structure of a 16-mer bound to rat HMG1 was reported,114 and this 9-mer has many spectral features common to duplexes with a pyrimidine–G–G–pyrimidine sequence. If, as appears to be the case, the structure of crosslinked double-stranded oligonucleotides is far from being rigid, it is likely that the conformation is deeply influenced by ‘weak’ interligand interactions, which are difficult to detect, but which can result in a different biological activity. This appears to be the case for oxaliplatin, [Pt(oxalato)(dach)] (dach = 1,2-diaminocyclohexane), for which only the R,R enantiomer ((R,R)-dach carrier ligand) has been approved for clinical use.141,144 We also investigated the mutagenic activity of several platinum compounds with chiral diamines and found that the S,S enantiomer was always a far greater mutagenic agent than the R,R form, particularly in the case of dach and dab adducts (dab = 2,3diaminobutane).149 In an insightful analysis the distortion of double-stranded oligonucleotides modified at a single site by dach-Pt or dab-Pt residues of different chiralities (either R,R or S,S) was monitored by chemical probes. It was found that different diamine enantiomers induced different types of distortion and, as a consequence of these structural differences, the affinity of the platinated oligonucleotides for proteins (e.g. HMG proteins) was different, as well as their processing by repair systems.150,151

5.7 Conclusions and Perspectives In general, adducts of cisplatin (and cisplatin analogues) with nucleotides and SS and DS oligonucleotides are highly dynamic systems that are difficult to investigate with conventional techniques. In most cases intermolecular interactions will affect the solid-state conformation; solution data are generally interpreted on the basis of a single model even when the system is likely to be a mixture of rapidly interconverting forms. Retro models have proven to be particularly useful in unravelling the

164


existence of different conformations, and in the most favourable cases have also allowed the full characterization of their stereochemistry. Moreover, several interligand interactions, some of which had remained unrecognized for a long time, have been revealed and a hierarchy among them has been established. Our analysis, based on solution results with adducts containing specially designed carrier ligands, has allowed us to rationalize computational, solution-state, and solid-state results on cis-(NH3)2Pt crosslink adducts in a more consistent manner than previous interpretations based on the cis-(NH3)2Pt adducts themselves. This success verifies the value of the retro modelling approach. In addition, by using retro models we found dramatic differences in the relative stability of different forms of oligonucleotides (coil, hairpin and duplex) dependent on the carrier ligand. We believe that in the future the use of retro models for the stabilization of single conformers and the characterization of their stereochemistry will play an important role in the interpretation of experimental data pertaining to dynamic systems and in the validation of results coming from theoretical investigations.

Acknowledgements G. Natile gratefully acknowledges the collaboration with Prof. Dr L. G. Marzilli (Louisiana State University) which, over a period of nearly two decades, has led to the major discoveries presented in this article. Thanks are also due to the many students and postdoctoral researchers who have contributed and whose names are listed in the references as coauthors of previous papers. Support and sponsorship from European COST Actions D39/0004/06 ‘Structure, recognition, and processing of DNA damage by antitumour metal-based compounds’, the Italian Ministry for University and Research (MUR, Rome), the University of Bari and the Interuniversity Consortium for Research in the Chemistry of Metal Ions in Biological Systems (C.I.R.C.M.S.B.) are also acknowledged.

References 1. Drobnik, J.; Antitumour activity of platinum complexes; Cancer Chemother. Pharmacol., 1983, 10(3), 145–149. 2. Pil, P.; Lippard, S.; Cisplatin and related drugs. In Encyclopedia of Cancer; Bertino, J.R., Ed., Academic Press, San Diego, 1997, 392–410. 3. Bierbach, U.; Qu, Y.; Hambley, T.W.; Peroutka, J.; Nguyen, H.L.; Doedee, M.; Farrell, N.; Synthesis, structure, biological activity, and DNA binding of platinum(II) complexes of the type trans- PtCl2(NH3)L (L = planar nitrogen base). Effect of L and cis/trans isomerism on sequence specificity and unwinding properties observed in globally platinated DNA; Inorg. Chem., 1999, 38, 3535–3542. 4. Di Blasi, P.; Bernareggi, A.; Beggiolin, G.; Piazzoni, L.; Menta, E.; Formento, M.; Cytotoxicity, cellular uptake and DNA binding of the novel trinuclear platinun complex BBR 3464 in sensitive and cisplatin resistant murine leukemia cells; Anticancer Res., 1998, 18, 3113–3117.

References

165

5. Scanlon, K.J.; Kashani-Sabet, M.; Miyachi, H.; C. Sowers, L.; Rossi, J.; Molecular basis of cisplatin resistance in human carcinomas: Model systems and patients; Anticancer Res., 1989, 9, 1301–1312. 6. Perez, R.P.; Hamilton, T.C.; Ozols, R.F.; Young, R.C.; Mechanisms and modulation of resistance to chemotherapy in ovarian cancer; Cancer, 1993, 71, 1571–1580. 7. Safaei, R.; Holzer, A.K.; Katano, K.; Samimi, G.; Howell, S.B.; The role of copper transporters in the development of resistance to Pt drugs; J. Inorg. Chem., 2004, 98, 1607–1613. 8. Samimi, G.; Safaei, R.; Katano, K.; Holzer, A.K.; Rochdi, M.; Tomioka, M.; Goodman, M.; Howell, S.B.; Increased expression of the copper efflux transporter ATP7A mediates resistance to cisplatin, carboplatin and oxaliplatin in ovarian cancer cells; Clin. Cancer Res., 2004, 10, 4661–4669. 9. Holford, J.; Beale, P.J.; Boxall, F.E.; Sharp, S.Y.; Kelland, L.R.; Mechanisms of drug resistance to the platinum complex ZD0473 in ovarian cancer cell lines; Eur. J. Cancer, 2000, 36, 1984–1990. 10. Yang, P.; Ebbert, J.O.; Sun, Z.; Weinshilboum, R.M.; Role of the glutathione metabolic pathway in lung cancer treatment and prognosis: a review; J. Clin. Oncol., 2006, 24, 1761–1769. 11. Ferry, K.V.; Hamilton, T.C.; Johnson, S.V.; Increased nucleotide excision repair in cisplatin-resistant ovarian cancer cells: role of ERCC1-XPF; Biochem. Pharmacol., 2000, 60, 1305–1313. 12. Chang, Y.J.; Kim, M.H.; Kim, H.B.; Lee, D.Y.; Kim, S.H.; Kim, H.Y.; You, H.J.; Small interfering RNA-induced suppression of ERCC1 enhances sensitivity of human cancer cells to cisplatin; Biochem. Biophys. Res. Commun., 2005, 327, 225–233. 13. Reed, E.; ERCC1 and clinical resistance to platinum-based therapy; Clin. Cancer Res., 2005, 11, 6100–6102. 14. Fink, D.; Nebel, S.; Aebi, S.; Zheng, H.; Cenni, B.; Nehmé, A.; Christen, R.D.; Howell, S.B.; The role of DNA mismatch repair in platinum drug resistance; Cancer Res., 1996, 56, 4881–4886. 15. Bassett, E.; Vaisman, A.; Tropea, K.A.; McCall, C.M.; Masutani, C.; Hanaoka, F.; Chaney, G.; Frameshifts and deletions during in vitro translesion synthesis past Pt-DNA adducts by DNA polymerases β and η; DNA Repair, 2002, 1, 1003–1016. 16. Albertella, M.R.; Green, C.M.; Lehmann, A.R.; O’Connor, M.J.; A role for polymerase η in the cellular tolerance to cisplatin-induced damage; Cancer Res., 2005, 65, 9799–9806. 17. White, S.C.; Lorigan, P.; Margison, G.P.; Margison, J.M.; Martin, F.; Thatcher, N.; Anderson, H.; Ranson, M.; Phase II study of SPI-77 (sterically stabilised liposomal cisplatin) in advanced nonsmall cell lung cancer; Br. J. Cancer, 2006, 92, 822–828. 18. Dragovich, T.; Mendelson, D.; Kurtin, S.; Richardson, K.; Von Hoff, D.; Hoos, A.; A phase 2 trial of the liposomal DACH platinum L-NDDP in patients with therapy-refractory advanced colorectal cancer; Cancer Chemoter. Pharmacol., 2006, 58, 759–764. 19. Rademaker-Lakhai, J.M., Terret, C.; Howell, S.B.; Baud, C.M.; De Boer, R.F.; Pluim, D.; Beijnen, J.H.; Schellens, J.H.; Droz, J.P.; A phase I and pharmacological study of the platinum polymer AP5280 given as an intravenous infusion once every 3 weeks in patients with solid tumours; Clin. Cancer Res., 2004, 10, 3386–3395. 20. Rice, J.R.; Gerberich, J.L.; Nowotnik, D.P.; Howell, S.B.; Preclinical efficacy and pharmacokinetics of AP5346, a novel diaminocyclohexane-platinum tumourtargeting drug delivery system; Clin. Cancer Res., 2006, 12, 2248–2254. 21. Campone, M.; Rademaker-Lakhai, J.M., Bennouna, J.; Howell, S.B.; Nowotnik, D.P.; Beijnen, J.H.; Schellens, J.H.; Phase I and pharmacokinetic trial of AP5346, a DACHplatinum-polymer conjugate, administered weekly for three out of every 4 weeks to advanced solid tumour patients; Chemoter. Pharmacol., 2007, 60, 523–533. 22. Armstrong, D.K.; Bundy, B.; Wenzel, L.; Huang, H.Q.; Baergen, R.; Lele, S.; Copeland, L.J.; Walker, J.L.; Burger, R.A.; Gynecologic Oncology Group; Intraperitoneal cisplatin and paclitaxel in ovarian cancer; N. Engl. J. Med., 2006, 354, 34–43.

166


23. Morgan, A.S.; Sanderson, P.E.; Borch, R.F.; Tew, K.D.; Niitsu, Y.; Takayama, T.; Von Hoff, D.D.; Izbicka, E.; Mangold, G.; Paul, C.; Broberg, U.; Mannervik, B.; Henner, W.D.; Kauvar, L.M.; Tumour efficacy and bone-marrow sparing properties of TER286, a cytotoxin activated by glutathione S-transferase; Cancer Res., 1998, 58, 2568–2575. 24. Townsend, D.M.; Shen, H.; Staros, A.L.; Gate, L.; Tew, K.D.; Efficacy of a glutathione Stransferase p-activated prodrug in platinum-resistant ovarian cancer cells; Mol. Cancer Ther., 2002, 1, 1089–1095. 25. Kavanagh, J.J.; Gershenson, D.M.; Choi, H.; Lewis, L.; Patel, K.; Brown, G.L.; Garcia, A.; Spriggs, D.R.; Multi-institutional phase 2 study of TLK286 (TELCYTA, a glutathione S-transferase P1–1 activated glutathione analog prodrug) in patients with platinum and paclitaxel refractory or resistant ovarian cancer; Int. J. Gynecol. Cancer, 2005, 15, 593–600. 26. Plumb, J.A.; Strathdee, G.; Sludden, J.; Kaye, S.B.; Brown, R.; Reversal of drug resistance in human tumour xenografts by 2′-deoxy-5-azacytidine-induced demethylation of the hMLH1 gene promoter; Cancer Res., 2000, 60, 6039–6044. 27. Kelland, L., The resurgence of platinum-based cancer chemotherapy; Nature Rev. Cancer, 2007, 7, 573–584. 28. Lippert, B. Trans-diammineplatinum(II): What makes it different from cis-DDP? Coordination chemistry of a neglected relative of cisplatin and its interaction with nucleic acids. In: Metal Ions in Biological Systems, Vol. 33, A. Sigel and H. Sigel, Eds., Marcel Dekker, New York, Basel, Hong Kong, 1996, 105–141. 29. Connors, T.A.; Cleare, M.J.; Harrap, K.R.; Structure–activity relationships of the antitumour platinum coordination complexes; Cancer Treatment Rep., 1979, 63(9–10), 1499–1502. 30. Reedijk, J.; Metal-ligand exchange kinetics in platinum and ruthenium complexes; Platinum Metals Rev., 2008, 52, 2–11. 31. Harrap, K.R., Preclinical studies identifying carboplatin as a viable cisplatin alternative; Cancer Treat. Rev., 1985, 12, 21–33. 32. Calvert, A.; Newell, H.D.R.; Gumbrell, L.A.; Oreilly, S.; Burneli, M.; Boxall, F.E.; Siddik, Z.H.; Judson, I.R.; Gore, M.E.; Wiltshaw, E., Carboplatin dosage: prospective evaluation of a simple formula based on renal function; J. Clin. Oncol., 1989, 7, 1748–1756. 33. Bloemink, M.J.; Reedijk, J.; Cisplatin and derived anticancer drugs: Mechanism and current status of DNA binding. In: Metal Ions in Biological Systems, Vol. 32, A. Sigel and H. Sigel, Eds., Marcel Dekker, New York, Basel, Hong Kong, 1996, 641–685. 34. Reedijk, J.; Clues for platinum antitumour chemistry: Kinetically controlled metal binding to DNA; Proc. Natl. Acad Sci. USA, 2003, 100, 3611–3613. 35. Reedijk, J.; The relevance of hydrogen bonding in the mechanism of action of platinum antitumour compounds; Inorg. Chim. Acta, 1992, 198–200, 873–881. 36. Farrell, N.; Nonclassical platinum antitumour agents: perspectives for design and development of new drugs complementary to cisplatin; Cancer Invest., 1993, 11, 578–589. 37. Natile, G.; Coluccia, M.; Current status of trans-platinum compounds in cancer therapy; Coord. Chem. Rev., 2001, 216, 383–410. 38. Farrell, N., Polynuclear platinum drugs. In: Metal Ions in Biological Systems, Vol. 41, A. Sigel and H. Sigel, Eds., Marcel Dekker, New York, 2004, 252–296. 39. Ishida, S.; Lee, J.; Thiele, D.J.; Herskowitz, I.; Uptake of the anticancer drug cisplatin mediated by the copper transporter Ctr1 in yeast and mammals; Proc. Natl. Acad. Sci. USA, 2002, 99, 14298–14302. 40. Lin, T.X.; Okuda, J.; Holzer, A.; Howell, S.B.; The copper transporter CTR1 regulates cisplatin uptake in Saccharomyces cerevisiae; Mol. Pharmacol., 2002, 62, 1154–1159. 41. Goren, M.P.; Wright, R.K.; Horowitz, M.E.; Cumulative renal tubular damage associated with cisplatin nephrotoxicity; Cancer Chemoth. Pharm., 1986, 18, 69–73. 42. Blommaert, F.A.; van Dijk-knijnenburg, H.C.M.; Dijt, F.J.; den Engelse, L.; Baan, R.A.; Berends, F.; Fichtinger-Schepman, A.M.; Formation of DNA adducts by the anticancer drug carboplatin: different nucleotide sequence preferences in vitro and in cells; Biochemistry, 1995, 34, 8474–8480.

References

167

43. Rahn, R.O.; Chromatographic analysis of the adducts formed in DNA complexed with cis-diamminedichloroplatinum(II); J. Inorg. Biochem., 1984, 21, 311–321. 44. Andrews, P.A.; Howell, S.B.; Cellular pharmacology of cisplatin: perspectives on mechanisms of acquired resistance; Cancer Cell – Mon. Rev., 1990, 2, 35–43. 45. Mansy, S.; Thomson, A.J.; Rosenberg, B.; Binding of cis- and trans-dichlorodiamminepla tinum(II) to nucleosides. I. Location of the binding sites; J. Am. Chem. Soc., 1973, 95, 1633–1640. 46. Dunham, S.U.; Dunham, S.U.; Turner, C.J.; Lippard, S.J.; Solution structure of a DNA duplex containing a nitroxide spin-labeled platinum d(GpG) intrastrand crosslink refined with NMR-derived long-range electron-proton distance restraints; J. Am. Chem. Soc., 1998, 120, 5395–5406. 47. Arpalahti, J., Platinum(II)-nucleobase interactions: A kinetic approach. In: Metal Ions in Biological Systems, Vol. 32, A. Sigel and H. Sigel, Eds., Marcel Dekker, New York, Basel, Hong Kong; 1996, 379–395. 48. Fichtinger-Schepman, A.M.; Lohman, P.H.; Reedijk, J.; Detection and quantification of adducts formed upon interaction of diamminedichloroplatinum(II) with DNA, by anion-exchange chromatography after enzymatic degradation; Nucleic Acids Res., 1982, 10, 5345–5356. 49. Fichtinger-Schepman, A.M.; van der Veer, J.L.; den Hartog, J.H.; Lohman, P.H.; Reedijk, J.; Adducts of the antitumour drug cis-diamminedichloroplatinum(II) with DNA: formation, identification, and quantitation; Biochemistry, 1985, 24, 707–713. 50. Eastman, A.; Reevaluation of interaction of cis-dichloro(ethylenediamine)platinum(II) with DNA; Biochemistry, 1986, 25, 3912–3915. 51. Eastman, A.; The formation, isolation and characterization of DNA adducts produced by anticancer platinum complexes; Pharmacol. Ther., 1987, 34, 155–166. 52. Eastman, A.; Interaction cross links and sequence specificity in the reaction of cis-dichloro(ethylenediamine)platinum(II) with DNA; Biochemistry, 1985, 24, 5027– 5032. 53. Plooy, A.C.M.; Fichtinger-Schepman, A.M.J.; Schutte, H.H.; van Dijk, M.; Lohman, P.H.M.; The quantitative detection of various Pt-DNA-adducts in chinese hamster ovary cells treated with cisplatin: Application of immunochemical techniques; Carcinogenesis, 1985, 6, 561–566. 54. Fichtinger-Schepman, A.M.J.; van Oosterom, A.T.; Lohman, P.H.M. Berends, F.; Cisdiamminedichloroplatinum(II)-induced DNA adducts in peripheral leukocytes from seven cancer patients: quantitative immunochemical detection of the adduct induction and removal after a single dose of cis-diamminedichloroplatinum(II); Cancer Res., 1987, 47, 3000–3004. 55. Reed, E.; Ostchega, Y.; Steinberg, S.M.; Yuspa, S.H.; Young, R.C.; Ozols, R.F.; Poinier, M.C.; An evaluation of platinum-DNA adduct levels relative to known prognostic variables in a cohort of ovarian cancer patients; Cancer Res., 1990, 50, 2256–2260. 56. Hambley, T.W.; Platinum binding to DNA: structural controls and consequences; J. Chem. Soc. – Dalton Trans., 2001, 2711–2718. 57. Hermann, D.; Houssier, C.; Guschlbauer, W.; Dissociation of double-stranded poly(I)·poly(C) by cis-diammine-dichloro-platinum(II); Biochim. Biophys. Acta, 1979, 564, 456–472. 58. Cohen, G.L.; Bauer, W.R.; Barton, J.K.; Lippard, S.J.; Binding of cis- and trans-dichloro diammineplatinum(II) to DNA: evidence for unwinding and shortening of the double helix; Science, 1979, 203, 1014–1016. 59. Macquet, J.P.; Butour, J.L.; A circular dichroism study of DNA platinum complexes. Differentiation between monofunctional, cis-bidentate and trans-bidentate platinum fixation on a series of DNAs; Eur. J. Biochem., 1978, 83, 375–385. 60. Scovell, W.M.; Capponi, V.J.; Cis-diamminedichloroplatinum(II) modified DNA stimulates far greater levels of S1 nuclease sensitive regions than does the modification produced by the trans-isomer; Biochem. Biophys. Res. Commun., 1982, 107, 1138– 1143.

168


61. Sorenson, C.M.; Eastman, A.; Mechanism of cis-diamminedichloroplatinum induced cytotoxicity: role of G2 arrest and DNA double strand breaks; Cancer Res., 1988, 48, 4484–4488. 62. Simonian, P.L.; Grillot, D.A.M. and Nunez, G.; Bcl-2 and Bcl-xL can differentially block chemotherapy-induced cell death; Blood, 1997, 90, 1208–1216. 63. Eastman, A.; The mechanism of action of apoptosis by cisplatin: From adducts to apoptosis; Cisplatin, Chemistry and Biochemistry of a Leading Anticancer Drug, B. Lippert, Ed., Wiley-VCH, 1999, 111–134. 64. Brabec, V.; Leng, M.; DNA interstrand crosslinks of trans-diamminedichloroplatinum(II) are preferentially formed between guanine and complementary cytosine residues; Proc. Nat. Acad. Sci. USA, 1993, 90, 5345–5349. 65. Boudvillain, M.; Dalbies, R.; Aussourd, C.; Leng, M.; Intrastrand crosslinks are not formed in the reaction between transplatin and native DNA: relation with the clinical inefficiency of transplatin; Nucleic Acids Res., 1995, 23, 2381–2388. 66. Bemal-Méndez, E.; Boudvillain, M.; Gonz, F.; Leng, M.; Chemical versatility of transplatin monofunctional adducts within multiple site-specifically platinated DNA; Biochemistry, 1997, 36, 7281–7287. 67. Iwamoto, M.; Mukundan, S.; Marzilli, L.G.; DNA adduct formation by platinum anticancer drugs. Insight into an unusual GpG intrastrand crosslink in a hairpin-like DNA oligonucleotide using NMR and distance geometry methods; J. Am. Chem. Soc., 1994, 116, 6238–6244. 68. Yang, D.Z.; van Boom, S.S.G.E.; Reedijk, J.; Van Boom, J.H.; Wang, A.H.J.; Structure and isomerization of an intrastrand cisplatin-crosslinked octamer DNA duplex by NMR analysis; Biochemistry, 1995, 34, 12912–12920. 69. Fouchet, M.H.; Guittet, E.; Cognet, J.A.H.; Kozelka, J.; Gauthier, C.; LeBret, M.; Zimmermann, K.; Chottard, J.C.; Structure of a nonanucleotide duplex crosslinked by cisplatin at an ApG sequence; J. Biol. Inorg. Chem., 1997, 1, 83–92. 70. Gelasco, A.; Lippard, S. J.; NMR solution structure of a DNA dodecamer duplex containing a cis-diammineplatinum(II) d(GpG) intrastrand crosslink, the major adduct of the anticancer drug cisplatin; Biochemistry, 1998, 37, 9230–9239. 71. Teuben, J.M.; Bauer, C.; Wang, A.H.J.; Reedijk, J.; Solution structure of a DNA duplex containing a cis-diammineplatinum(II) 1,3-d(GTG) intrastrand crosslink, a major adduct in cells treated with the anticancer drug carboplatin; Biochemistry, 1999, 38, 12305–12312. 72. Takahara, P.M.; Rosenzweig, A.C.; Frederick, C.A.; Lippard, S.J.; Crystal structure of double-stranded DNA containing the major adduct of the anticancer drug cisplatin; Nature, 1995, 377, 649–652. 73. Takahara, P.M.; Frederick, C.A.; Lippard, S.J.; Crystal structure of the anticancer drug cisplatin bound to duplex DNA; J. Am. Chem. Soc., 1996, 118, 12309–12321. 74. Coste, F.; Malinge, J.M.; Serre, L.; Shepard, W.; Roth, M.; Leng, M.; Zelwer, C.; Crystal structure of a double-stranded DNA containing a cisplatin interstrand crosslink at 1.63 Å resolution: hydration at the platinated site; Nucl. Acids Res., 1999, 27, 1837–1846. 75. Bellon, S.F.; Coleman, J.H.; Lippard, S.J.; DNA unwinding produced by site-specific intrastrand crosslinks of the antitumour drug cis-diamminedichloroplatinum(II); Biochemistry, 1991, 30, 8026–8035. 76. Ano, S.O.; Intini, F.P.; Natile, G.; Marzilli, L.G.; Retro models of Pt anticancer drug DNA adducts: chirality-controlling chelate ligand restriction of guanine dynamic motion in (2,2′-bipiperidine)PtG2 complexes (G = guanine derivative); Inorg. Chem., 1999, 38, 2989–2999. 77. Ano, S.O.; Intini, F.P.; Natile, G.; Marzilli, L.G.; Viewing early stages of guanine nucleotide attack on Pt(II) complexes designed with in-plane bulk to trap initial adducts. Relevance to cis-type Pt(II) anticancer drugs; J. Am. Chem. Soc., 1997, 119, 8570–8571. 78. Marzilli, L.G.; Intini, F.P.; Kiser, D.; Wong, H.C.; Ano, S.O.; Marzilli, P.A.; Natile, G.; Chirality-Controlling Chelate (CCC) ligands in analogues of platinum anticancer agents. Influence of N9 substituents of guanine derivatives (G) on the distribution of chiral

References

79.

80.

81. 82. 83.

84. 85.

86. 87. 88. 89. 90.

91.

92.

93.

169

conformers of (CCC)PtG2 with CCC = N,N′-dimethyl-2,3-diaminobutane; Inorg. Chem., 1998, 37, 6898–6905. Wong, H.C.; Coogan, R.; Intini, F.P.; Natile, G.; Marzilli, L.G.; New steric hindrance approach employing the hybrid ligand 2-aminomethylpiperidine for diminishing dynamic motion problems of platinum anticancer drug adducts containing guanine derivatives; Inorg. Chem., 1999, 38, 777–787. Sullivan, S.T.; Ciccarese, A.; Fanizzi, F.P.; Marzilli, L.G.; Cisplatin-DNA crosslink models with an unusual type of chirality-neutral chelate amine carrier ligand, N,N′dimethylpiperazine (Me2ppz): Me2ppzPt(guanosine monophosphate)2 adducts that exhibit novel properties; Inorg. Chem., 2000, 39, 836–842. Sullivan, S.T.; Ciccarese, A.; Fanizzi, F.P.; Marzilli, L.G.; Cisplatin–DNA crosslink retro models with a chirality-neutral carrier ligand: evidence for the importance of ‘SecondSphere Communication’; Inorg. Chem., 2001, 40, 455–462. Xu, Y.; Natile, G.; Intini, F.P.; Marzilli, L.G.; Stereochemically controlled ligands influence atropisomerization of Pt(II) nucleotide complexes. Evidence for Head-to-Head and stable L Head-to-Tail atropisomers; J. Am. Chem. Soc., 1990, 112, 8177–8179. Kiser, D.; Intini, F.P.; Xu, Y.H.; Natile, G.; Marzilli, L.G.; Atropisomerization of cisbis(5′GMP)-platinum(II)-diamine complexes with nonC2-symmetrical asymmetric diamine ligands containing NH groups directed to one side of the coordination plane; Inorg. Chem., 1994, 33, 4149–4158. Lippert, B.; Platinum nucleobase chemistry; Prog. Inorg. Chem., 1989, 37, 1–97. Schollhorn, H.; Raudaschl-Sieber, G.; Muller, G.; Thewalt, U.; Lippert, B.; DNAintrastand guanine, guanine crosslinking by cisplatin: comparison of three model compounds with Head-Head orientation of the nucleobases; J. Am. Chem. Soc., 1985, 107, 5932–5937. Grabner, S.; Plavec, J.; Bukovec, N.; Leo, D.D.; Cini, R.; Natile, G.; Synthesis and structural characterization of platinum(II) acyclovir complexes; J. Chem. Soc. Dalton Trans., 1998, 1447–1451. Sinur, A.; Grabner, S.; A platinum(II) diammine complex: cis-Pt(C8H11N5O3)2(NH3)2 Cl2·2H2O; Acta Crystallogr., 1995, C 51, 1769–1772. Sherman, S.E.; Gibson, D.; Wang, A.H.J.; Lippard, S.J.; X-ray structure of the major adduct of the anticancer drug cisplatin with DNA: cis- Pt(NH3)2(d(pGpG)); Science, 1985, 230, 412–417. An abstract of an incompletely refined structure containing a LHT atropisomer – Bau, 2− R.; Gellert, R.W.; Structure of Pt (en ) ( 5′-GMP )2 anion; Biochimie, 1978, 60, 1040. Lippert, B.; Raudaschl, G.; Lock, C.J.L.; Pilon, P.; ‘Real’ model compounds for intrastrand crosslinking of two guanine bases by cisplatin: crystal structures of cis-diamminebis(9ethylguanine-N7)platinum(II) dichloride trihydrate, Pt(NH3)2(C7H9N5O)2 Cl2·3H2O, and cis-diamminebis(9-ethylguanine-N7)platinum(II) sesquichloride hemibicarbonate sesquihydrate Pt(NH3)2(C7H9N5O)2 Cl1.5(HCO3)0.5.1.5H2O; Inorg. Chim. Acta, 1984, 93, 43–50. Cini, R.; Grabner, S.; Bukovec, N.; Cerasino, L.; Natile, G.; Synthesis and structural characterization of a new form of bis(acyclovir)(ethylenediamine)platinum(II) – Correlation between the puckering of the carrier ligand and the canting of the nucleobases; Eur. J. Inorg. Chem., 2000, 1601–1607. (a) Cramer, R.E.; Dahlstrom, P.L.; Seu, M.J.T.; Norton, T.; Kashiwagi, M.; Crystal and molecular structure of cis- Pt(NH3)2(Guo)2 Cl3/2(ClO4)1/2·7H2O and anticancer activity of cis- Pt(NH3)2(Puo)2 Cl2 complexes; Inorg. Chem., 1980, 19, 148–154. (b) Cramer, R.E.; Dahlstrom, P.L.; Evidence for diastereomers of the platinum(N,N,N′,N′-tetrame2+ thyl-ethylenediamine) (guanosine)2 cation; J. Am. Chem. Soc., 1979, 101, 3679–3681. (a) Marcelis, A.T.M.; Korte, H.J.; Krebs, B.; Reedijk, J.; Crystal structure and dynamic proton NMR of (1,2-bis(pyridin-2-yl)ethane-N,N′)bis(9-methylhypoxanthine-N7) platinum(II) dinitrate; Inorg. Chem., 1982, 21, 4059–4063. (b) Marcelis, A.T.M.; van der Veer, J.L.; Zwetsloot, J.C.M.; Reedijk, J.; Rotation and conformation of purine ligands in cis-bis(6-oxopurine)platinum compounds; Inorg. Chim. Acta, 1983, 78, 195–203.

170


94. Cramer, R.E.; Dahlstrom, P.L.; Diastereotopic exchange in the square-planar platinum(II) complex, (N,N′-dimethylethylenediamine)bis(guanosine)platinum(II); Inorg. Chem., 1985, 24, 3420–3424. 95. (a) Dijt, F.J.; Canters, G.W.; den Hartog, J.H.J.; Marcelis, A.T.M.; Reedijk, J.; Reaction products of cis-diammineplatinum(II) compounds with 5′-guanosine monophosphate characterized by high-frequency proton NMR; J. Am. Chem. Soc., 1984, 106, 3644–3647. (b) Miller, S.K.; Marzilli, L.G.; Interaction of platinum antitumour agents with guanine nucleosides and nucleotides. Platinum-195 and proton NMR spectroscopic characterization of compound III; Inorg. Chem., 1985, 24, 2421–2425. 96. Zn: Miller, S.K.; van der Veer, D.G.; Marzilli, L.G.; Models for the interaction of zinc(2+) ion with DNA. The synthesis and x-ray structural characterization of two octahedral zinc complexes with monomethyl phosphate esters of 6-oxopurine 5′-monophosphate nucleotides; J. Am. Chem. Soc., 1985, 107, 1048–1055. 97. Cu: Chiang, C.C.; Sorrell, T.; Kistenmacher, T.J.; Marzilli, L.G.; Preparation and crystal and molecular structure of sodium bis(inosine 5′-monophosphate)(diethylenetriamine) copper(II) decahydrate. Possible implications for intrastrand crosslinking of polynucleotides by copper(II) and platinum(II) complexes; J. Am. Chem. Soc., 1978, 100, 5102–5110. 98. Poojary, M.D.; Manohar, H.; The cis-binding of inosine 5′-monophosphate and guanosine 5′-monophosphate to cobalt(III) in 1:1 metal-nucleotide complexes. X-ray crystal structures of ternary complexes with ethylenediamine; J. Chem. Soc., Chem. Commun., 1982, 533–534. 99. Cd: Goodgame, D.M.L.; Jeeva, I.; Reynolds, C.D.; Skapski, A.C.; Multiplicity of cadmium binding sites in nucleotides. X-ray evidence for the involvement of O2′ and O3′ as well as phosphate and N7 in inosine 5′-monophosphate; Nucleic Acids Res., 1975, 2, 1375–1379. 100. Sullivan, S.T.; Ciccarese, A.; Fanizzi, F.P.; Marzilli, L.G.; A rare example of three abundant conformers in one retro model of the cisplatin-DNA d(GpG) intrastrand cross link. Unambiguous evidence that guanine O6 to carrier amine ligand hydrogen bonding is not important. Possible effect of the Lippard base pair step adjacent to the lesion on carrier ligand hydrogen bonding in DNA adducts; J. Am. Chem. Soc., 2001, 123, 9345–9355. 101. Ano, S.O.; Intini, F.P.; Natile, G.; Marzilli, L.G.; A novel Head-to-Head conformer of d(GpG) crosslinked by Pt: new light on the conformation of such crosslinks formed by Pt anticancer drugs; J. Am. Chem. Soc., 1998, 120, 12017–12022. 102. Wong, H.C.; Intini, F.P.; Natile, G.; Marzilli, L.G.; Diminishing dynamic motion problems of platinum anticancer drug adducts of guanine derivatives with the hybrid ligand approach: evidence for cis interligand interactions especially between 3′-GMPs; Inorg. Chem., 1999, 38, 1006–1014. 103. Williams, K.M.; Cerasino, L.; Natile, G.; Marzilli, L.G.; Cis-Pt(NH3)2(GpG) properties interpreted through comparison with Retro-Model GpG adducts having carrier ligands designed to slow dynamic motion and control crosslink handedness; J. Am. Chem. Soc., 2000, 122, 8021–8030. 104. Saad, J.S.; Scarcia, T.; Shinozuka, K.; Natile, G.; Marzilli, L.G.; Modification of SecondSphere Communication, leading to an unusually high abundance of the Head-to-Head conformer of cisplatin crosslink retro models; Inorg. Chem., 2002, 41, 546–557. 105. Wong, H.C.; Shinozuka, K.; Natile, G.; Marzilli, L.G.; Second-sphere ‘communication’ between two cis-bound guanine nucleotides. Factors influencing conformations of dynamic adducts of cis-type platinum anticancer drugs with guanine nucleotides as deduced by circular dichroism spectroscopy; Inorg. Chim. Acta, 2000, 297, 36–46. 106. Marzilli, L.G.; Ano, S.O.; Intini, F.P.; Natile, G.; New concepts relevant to cisplatin anticancer activity from unique spectral features providing evidence that adjacent guanines in d(GpG), intrastrand-crosslinked at N7 by a cis-platinum(II) moiety, can adopt a Head-to-Tail arrangement; J. Am. Chem. Soc., 1999, 121, 9133–9142. 107. Ano, S.O.; Kuklenyik, Z. Marzilli L.G.; Structure and dynamics of Pt anticancer drug adducts from nucleotides to oligonucleotides as revealed by NMR methods. In: Cispla-

References

108.

109. 110. 111. 112. 113.

114. 115. 116. 117.

118. 119. 120. 121.

122. 123.

171

tin: Chemistry and Biochemistry of a Leading Anticancer Drug, B. Lippert, Ed.,WileyVCH, Weinheim; 1999, 247–291. den Hartog, J.H.J.; Altona, C.; Chottard, J.C.; Girault, J.P.; Lallemand, J.Y.; de Leeuw, F.A.A.M.; Marcelis, A.T.M.; Reedijk, J.; Conformational analysis of the adduct cisPt(NH3)2{d(GpG)}+ in aqueous solution. A high field (500–300 MHz) nuclear magnetic resonance investigation; Nucleic Acids Res., 1982, 10, 4715–4730. Girault, J.P.; Chottard, G.; Lallemand, J.Y.; Chottard, J.C.; Interaction of cisPt(NH3)2(H2O)2 (NO3)2 with ribose deoxyribose diguanosine phosphates; Biochemistry, 1982, 21, 1352–1356. Kozelka, J.; Fouchet, M.H.; Chottard, J.C.; H8 chemical shifts in oligonucleotides 2+ crosslinked at a GpG sequence by cis-Pt ( NH 3 )2 : a clue to the adduct structure; Eur. J. Biochem., 1992, 205, 895–906. Neumann, J.M.; Tran-Dinh, S.; Girault, J.P.; Chottard, J.C.; Huynh-Dinh, T.; DNA fragment conformations. A 1H-NMR conformational analysis of the d(G-G)-chelated platinum-oligonucleotide d(A-T-G-G)cisPt; Eur. J. Biochem., 1984, 141, 465–472. Berners-Price, S.J.; Ranford, J.D.; Sadler, P.J.; 1H, 15N NMR investigations of Pt-NH hydrogen bonding in d(GpG), d(pGpG), and d(TpGpG)-N7,N7 adducts of Pt(en)2+ in aqueous solution; Inorg. Chem., 1994, 33, 5842–5846. Saad, J.S.; Scarcia, T.; Natile, G.; Marzilli, L.G.; Factors influencing conformer equilibria in retro models of cisplatin–DNA adducts as revealed by moderately dynamic (N,N′dimethyl-2,3-diaminobutane)PtG2 retro models (G = a guanine derivative); Inorg. Chem., 2002, 41, 4923–4935. Marzilli, L.G.; Saad, J.S.; Kuklenyik, Z.; Keating, K.A.; Xu, Y.; Relationship of solution and protein-bound structures of DNA duplexes with the major intrastrand crosslink lesions formed on cisplatin binding to DNA; J. Am. Chem. Soc., 2001, 123, 2764–2770. Marzilli, L.G.; Chalilpoyil, P.; Chiang, C.C.; Kistenmacher, T.J.; Platinum(II) antitumour agents. A new class of intrastrand crosslinking models exhibiting significant intracomplex base-base interactions; J. Am. Chem. Soc., 1980, 102, 2480–2482. Barnham, K.J.; Bauer, C.J.; Djuran, M.I.; Mazid, M.A.; Rau, T.; Sadler, P.J.; Outer-sphere macrochelation in Pd(en)(5′-GMP-N7)2·9H2O and Pt(en)(5′-GMP-N7)2·9H2O: X-ray crystallography and NMR spectroscopy in solution; Inorg. Chem., 1995, 34, 2826–2832. Benedetti, M.; Tamasi, G.; Cini, R.; Natile, G.; X-ray structure and circular dichroism of pure rotamers of bis guanosine-5′-monophosphate(-1) (N,N,N′,N′-tetramethylcyclohexyl1,2-diamine)platinum(II) complexes that have R,R and S,S configurations at the asymmetric diamine; Chem. Eur. J., 2003, 9, 6122–6132. Benedetti, M.; Tamasi, G.; Cini, R.; Marzilli, L.G.; Natile, G.; The first pure LHT rotamer of a complex with a cis-metal(nucleotide)2 unit: a cis-Pt(amine)2(nucleotide)2 LHT rotamer with unique molecular structural features; Chem. Eur. J., 2007, 13, 3131–3142. Trani, M.; Cannito, F.; Natile, G.; Marzilli, P.A.; Marzilli, L.G.; Slowly exchanging cisPtA2B2 atropisomers with A2 = chiral diamine ligands and B = 5,6-dimethylbenzimidazole; Eur. J. Inorg. Chem., 2005, 2826–2835. Benedetti, M.; Saad, J.S.; Marzilli, L.G.; Natile, G.; Chiral discrimination in the formation reaction and at equilibrium for N,N,N′,N′-tetramethyl-1,2-diaminocyclohexane–PtG2 complexes; J. Chem. Soc. Dalton Trans., 2003, 872–879. Margiotta, N.; Papadia, P.; Fanizzi, F.P.; Natile, G.; Mono- and bis-guanosine adducts of platinum complexes with carrier ligands having in-plane steric bulk: the case of 1,10phenanthroline and 2,9-dimethyl-1,10-phenanthroline; Eur. J. Inorg. Chem., 2003, 1136–1144. Marzilli, L.G.; Chalilpoyil, P.; Enhanced CD spectra in cis-bis(nucleotide)platinum(II) complexes. Comparison with the CD enhancement induced in DNA by the antitumour agent, cisPt; J. Am. Chem. Soc., 1980, 102, 873–875. Gullotti, M.; Pacchioni, G.; Pasini, A.; Ugo, R.; Complexes of platinum(II) with chiral diamines and guanosine. Stereochemical investigation related to the mechanism of the antitumour activity of cis-bis(amine)platinum(II) type complexes; Inorg. Chem., 1982, 21, 2006–2014.

172


124. Pasini, A.; De Giacomo, L.; Aspects of the stereochemistry of bis-guanosine-5′monophosphate diamine platinum complexes; Inorg. Chim. Acta, 1996, 248, 225–230. 125. Song, B.; Oswald, G.; Bastian, M.; Sigel, H.; Lippert, B.; Extent of the acidification by N7-coordinated cis-diammine-platinum(II) on the acidic sites of guanine derivatives; Metal-Based Drugs, 1996, 3, 131–141. 126. Carlone, M.; Fanizzi, F.P.; Intini, F.P.; Margiotta, N.; Marzilli, L.G.; G. Natile; Influence of carrier ligand NH hydrogen bonding to the O6 and phosphate group of guanine nucleotides in platinum complexes with a single guanine ligand; Inorg. Chem., 2000, 39, 634–641. 127. Carlone, M.; Marzilli, L.G.; Natile, G.; Platinum complexes with NH groups on the carrier ligand and with only one guanine or hypoxanthine derivative. Informative models for assessing relative nucleobase and nucleotide hydrogen-bond interactions with amine ligands in solution; Inorg. Chem., 2004, 43, 584–592. 128. Benedetti, M.; Marzilli, L.G.; Natile, G.; Rotamer stability in cis- Pt(diA)G2 complexes (diA = diamine derivative and G = guanine derivative) mediated by carrier-ligand amine stereochemistry as revealed by circular dichroism spectroscopy; Chem. Eur. J., 2005, 11, 5302–5310. 129. Ranaldo, R.; Margiotta, N.; Intini, F.P.; Pacifico, C.; Natile, G.; Conformer distribution in (cis-1-4-DACH)bis(guanosine-5′-phosphate)platinum(II) adducts: a reliable model for DNA adducts of antitumoural cisplatin; Inorg. Chem., 2008, 47, 2820–2830. 130. Chottard, J.C.; Girault, J.P.; Chottard, G.; Lallemand, J.Y.; Mansuy, D.; Interaction of cisdiaquodiammineplatinum dinitrate with ribose dinucleoside monophosphates; J. Am. Chem. Soc., 1980, 102, 5565–5572. 131. Lebwohl, D.; Canetta, R.; Clinical development of platinum complexes in cancer therapy: an historical perspective and an update; Eur. J. Cancer, 1998, 34, 1522–1534. 132. Sherman, S.E.; Gibson, D.; Wang, A.H.J.; Lippard, S.J.; Crystal and molecular structure of cis-Pt(NH3)2 d(pGpG), the principal adduct formed by cis-diamminedichloroplatinu m(II) with DNA; J. Am. Chem. Soc., 1988, 110, 7368–7381. 133. Kline, T.P.; Marzilli, L.G.; Live, D.; Zon, G.; Investigations of platinum amine induced distortions in single- and double-stranded oligodeoxyribonucleotides; J. Am. Chem. Soc., 1989, 111, 7057–7068. 134. Pilch, D.S.; Dunham, S.U.; Jamieson, E.R.; Lippard, S.J.; Breslauer, K.J.; DNA sequence context modulates the impact of a cisplatin 1,2-d(GpG) intrastrand crosslink on the conformational and thermodynamic properties of duplex DNA; J. Mol. Biol., 2000, 296, 803–812. 135. Poklar, N.; Pilch, D.S.; Lippard, S.J.; Redding, E.A.; Dunham, S.U.; Breslauer, K.J.; Influence of cisplatin intrastrand crosslinking on the conformation, thermal stability and energetics of a 20-mer DNA duplex; Proc. Natl. Acad. Sci. USA, 1996, 93, 7606–7611. 136. Kline, T.P.; Marzilli, L.G.; Live, D.; Zon, G.; NMR studies of an oligonucleotide with an unusual structure induced by platinum anticancer drugs; Biochem. Pharmacol., 1990, 40, 97–113. 137. Villanueva, J.M.; Jia, X.; Yohannes, P.G.; Doetsch, P.W.; Marzilli, L.G.; Cisplatin (cis2+ Pt(NH3)2Cl2) and cis-Pt ( NH 3 )2 ( H 2O)2 intrastrand crosslinking reactions at the telomere GGGT DNA sequence embedded in a duplex, a hairpin, and a bulged duplex: Use of Mg2+ and Zn2+ to convert a hairpin to a bulged duplex; Inorg. Chem., 1999, 38, 6069–6080. 138. Yohannes, P.G.; Zon, G.; Doetsch, P.W.; Marzilli, L.G.; DNA hairpin formation in adducts with platinum anticancer drugs: gel electrophoresis provides new information and a caveat; J. Am. Chem. Soc., 1993, 115, 5105–5110. 139. Beljanski, V.; Villanueva, J.M.; Doetsch, P.W.; Natile, G.; Marzilli, L.G.; Marked dependence on carrier-ligand bulk but not on carrier-ligand chirality of the duplex versus single-strand forms of a DNA oligonucleotide with a series of G-Pt(II)-G intrastrand crosslinks modeling cisplatin-DNA adducts; J. Am. Chem. Soc., 2005, 127, 15833–15842.

References

173

140. Spingler, B.; Whittington, D.A.; Lippard, S.J.; 1,2-d(GpG) intrastrand crosslink formed by oxaliplatin with duplex DNA: a crystallographic study; J. Inorg. Biochem., 2001, 86, 440–440. 141. Spingler, B.; Whittington, D.A.; Lippard, S.J.; 2.4 Å crystal structure of an oxaliplatin 1,2-d(GpG) intrastrand crosslink in a DNA dodecamer duplex; Inorg. Chem., 2001, 40, 5596–5602. 142. Huang, H.; Zhu, L.; Reid, B.R.; Drobny, G.P.; Hopkins, P.B.; Solution structure of a cisplatin-induced DNA interstrand crosslink; Science, 1995, 270, 1842–1845. 143. Paquet, F.; Perez, C.; Leng, M.; Lancelot, G.; Malinge, J.-M.; NMR solution structure of a DNA decamer containing an interstrand crosslink of the antitumour drug cis-diammi nedichloroplatinum(II); J. Biomol. Struct. Dyn., 1996, 14, 67–77. 144. Hambley, T.W.; The influence of structure on the activity and toxicity of Pt anti-cancer drugs; Coord. Chem. Rev., 1997, 166, 181–223. 145. Reedijk, J.; Improved understanding in platinium antitumour chemistry; J. Chem. Soc. Chem. Commun., 1996, 801–806. 146. Jamieson E. R. Lippard, S J.; Structure, recognition and processing of cisplatin-DNA adducts; Chem. Rev., 1999, 99, 2467–2498. 147. Herman, F.; Kozelka, J.; Stoven, V.; Guittet, E.; Girault, J.P.; Huynh-Dinh, T.; Igolen, J.; Lallemand, J.Y.; Chottard, J.C.; A d(GpG)-platinated decanucleotide duplex is kinked. An extended NMR and molecular mechanics study; Eur. J. Biochem., 1990, 194, 119–133. 148. Ohndorf, U.M.; Rould, M.A.; He, Q.; Pabo, C.O.; Lippard, S.J.; Basis for recognition of cisplatin-modified DNA by high-mobility-group proteins; Nature, 1999, 399, 708–712. 149. Coluccia, M.; Correale, M.; Giordano, D.; Mariggiò, M.A.; Moscelli, S.; Fanizzi, F.P.; Natile, G.; Maresca, L.; Mutagenic activity of some platinum complexes with monodentate and bidentate amines; Inorg. Chim. Acta, 1986, 123, 225–229. 150. Malina, J.; Kasparkova, J.; Natile, G.; Brabec, V.; Recognition of major DNA adducts of enantiomeric cisplatin analogs by HMG box proteins and nucleotide excision repair of these adducts; Chem. Biol., 2002, 9, 629–638. 151. Kasparkova, J.; Vojtiskova, M.; Natile, G.; Brabec, V.; Unique properties of DNA interstrand crosslinks of antitumour oxaliplatin and the effect of chirality of the carrier ligand; Chem. Eur. J., 2008, 14, 1330–1341.

6 Role of DNA Repair in Antitumour Effects of Platinum Drugs Viktor Brabec and Jana Kašpárková

6.1 Introduction Understanding the molecular and biochemical mechanisms associated with the biological effects of the existing platinum agents provides important insights for designing new and more efficient platinum-based drugs. There is a large body of experimental evidence suggesting that the success of platinum complexes in killing tumour cells results from their ability to damage DNA by forming various types of covalent adducts.1–4 Hence the research into DNA interactions with platinum antitumour drugs has predominated.5,6 However, cells have developed efficient tools to correct any chemical changes in DNA so that the genetic information is retained uncorrupted. Agents that damage DNA involve: (i) ionizing and ultraviolet radiation; (ii) highly reactive oxygen radicals produced during normal cellular respiration, as well as by other biochemical pathways; (iii) chemicals in the environment, such as various hydrocarbons, including some found in cigarette smoke, some plant, fungal and microbial products, e.g. the aflatoxins produced in mouldy peanuts and not least (iv) chemicals used in chemotherapy, especially in chemotherapy of cancer (see Table 6.1). Major types of DNA damage are: (i) all four of the bases in DNA (A, T, C, G) which can be covalently modified at various positions. One of the most frequent is the loss of an amino group (‘deamination’) resulting, for example, in a C being Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

176

DNA Repair in Antitumour Effects of Platinum Drugs

Table 6.1 Examples of chemotherapeutic agents that damage DNA Drug Bleomycin Busulfan Chlorambucil Cyclophosphamide Melphalan Mitomycin Carboplain Cisplatin Oxaliplatin Dactinomycin (actinomycin D) Daunorubicin (daunomycin) Doxorubicin Epirubicin Idarubicin Irinotecan Mitoxantrone Nogalamycin Proflavin Berenil Hoechst 33258 Lexitropsins (netropsin, distamycin and analogues)

Mechanism of action Cuts DNA strands between GC or GT Alkylating agent forming inter- and/or intrastrand crosslinks

Form intrastrand and intersand cross-links Inhibit DNA and RNA synthesis by intercalating between base pairs of the DNA/RNA strand, thus preventing the replication of rapidly growing cancer cells; they also create iron-mediated oxygen free radicals that damage the DNA and cell membranes; are also capable of inhibiting topoisomerase II enzymes, preventing DNA from relieving torsional stress during replication Inhibit DNA and RNA synthesis by binding noncovalently to the minor groove of DNA

converted to a U; (ii) mismatches of the normal bases due to a failure of proofreading during DNA replication; (iii) breaks in the backbone which can involve both strands (double-strand break) or be limited to only one strand (single-strand break) (ionizing radiation is a frequent cause, but some chemicals produce breaks as well); (iv) crosslinks – covalent linkages can be formed between bases on the same DNA strand (‘intrastrand’) or on the opposite strands (‘interstrand’), between two DNA double helices or between DNA and proteins. Several chemotherapeutic drugs used against cancers including antitumour platinum drugs crosslink DNA. Damage to the DNA molecules is identified by the cell, which uses a collection of mechanisms to correct this damage. This process, called DNA repair, is defined as a biological process during which alterations in the chemistry of DNA (DNA damage) are removed and the integrity of the genome restored. The broader field is now more appropriately referred to as the field of biological responses to DNA damage, and in this context even the term ‘DNA damage’ has been broadened to include phenomena such as the arrest of DNA synthesis in the absence of defined DNA damage.7 The adducts formed by platinum antitumour drugs cause structural damage to DNA and can alter or eliminate the cell’s ability to replicate and transcribe the gene that the affected DNA encodes. These lesions also induce potentially harmful mutations in the cell’s genome, which affect the survival of its daughter cells after it undergoes mitosis. After DNA is modified by platinum drugs, cellular repair systems recognize this damage. DNA repair is not perfect, however.This means that over

Human DNA Repair Systems 177

time the repair process will fail and a cell will enter one of the following states: (i) apoptosis or cellular suicide and necrosis; these are processes by which the cell destroys itself; (ii) senescence, a process whereby the cell becomes irreversibly dormant; (iii) uncontrolled cellular division; this can result in tumour formation and cancer. Thus, DNA repair has a significant role in modulating cytotoxicity of platinum drugs. The interruption of pathways that are involved in repairing platinated DNA represents a powerful approach to cancer treatment. It is therefore not surprising that repair of damage to DNA induced by platinum antitumour drugs and its downstream effects in tumour cells is, at present, a very active area of molecular pharmacology. The purpose of this review is to highlight recent discoveries related to repair of DNA damage induced by anticancer platinum drugs. The topics discussed do not involve mechanisms underlying repair of DNA damage by platinum compounds in bacterial or yeast systems.

6.2 Human DNA Repair Systems All macromolecules in a living cell, i.e., nucleic acids, proteins, lipids and carbohydrates can be damaged when the cells are exposed to deleterious agents, including antitumour metal-based drugs. A major defence against damage to cells is DNA repair. Although other biomacromolecules may be damaged and subject to repair, the significance and exceptionality of DNA repair in cellular processes is accentuated by the fact that only failure to repair DNA damage has fatal consequences for the cells. If other cellular molecules are damaged, they can be degraded and newly synthesized via transcription and translation. Thus, for instance, damage to cellular proteins is usually not corrected since, due to the large number of different chemical modifications, it seems to be chemically and evolutionally impossible to develop the required set of enzymes to correct damage to proteins. Therefore, very often protein degradation is the only way to remove such damaged proteins from the protein pool. Ribozyme mediated strategies exist that can be used to amend damaged genetic instructions at the RNA level.8–10 An obvious limitation of RNA repair is, however, the fact that repair of RNA only results in transient production of corrected genetic instructions, as the damaged DNA template is not amended. DNA repair, which is present in all organisms examined, including bacteria, yeast, drosophila, fish, amphibians, rodents and humans, is involved in processes that minimize cell death, mutations, replication errors, persistence of DNA damage and genomic instability. All organisms must keep their DNA intact and free of lesions to obtain faithful transmission of genetic information. There are several different DNA repair pathways in mammalian cells, as described in the sections 6.2.1–6.2.4. 6.2.1 Direct Reversal of Damage This type of DNA repair is relatively rare and involves chemical reversal of damage. It does not involve breakage of the phosphodiester backbone and is specific to the

178


type of damage. Only a few types of DNA damage are repaired in this way, particularly pyrimidine dimers resulting from exposure to ultraviolet (UV) light and alkylated guanine residues that have been modified by the addition of methyl or ethyl groups at the O6 position of the purine ring. Enzymes that catalyse direct repair reactions are widespread in both prokaryotes and eukaryotes, including humans. This type of DNA repair has not been observed for DNA damaged by antitumour metal-based drugs. 6.2.2 Excision Repair Although direct repair is an efficient way of dealing with particular types of DNA damage, excision repair is a more general means of repairing a wide variety of chemical alterations to DNA, including those induced by platinum antitumour drugs. In excision repair, the damaged DNA is recognized and removed, either as free bases or as nucleotides. The resulting gap is then filled in by synthesis of a new DNA strand, using the undamaged complementary strand as a template. Three types of excision repair – base-excision repair (BER), nucleotide-excision repair (NER) and mismatch repair (MMR) – enable cells to cope with a variety of different kinds of DNA damage. Base Excision Repair Common small lesions, especially methylated or oxidized bases, or uracil in DNA resulting from spontaneous deamination of cytosine are corrected by this type of DNA repair. An altered base is specifically recognized by one of many specialized glycosylases. The recognition process may be sensitive to enhanced DNA flexibility or to decreased base-pair stability as well as to steric effects like protrusions in the grooves. The glycosylase removes the offending base, giving an abasic site or AP site (apurinic/apyrimidinic site). The common feature of the glycosylases is that they tend to flip out the base and do chemistry on the extrahelical base. Excision of a damaged base and subsequent processing of the resulting single-strand break lead to an intermediate, containing a 3′-hydroxyl, suitable for priming strand resynthesis by a polymerase (Table 6.2). The majority of BER proceeds through the so-called ‘short-patch’ subpathway in which a single nucleotide is removed and replaced. In this pathway (Figure 6.1), DNA resynthesis is carried out by polymerase b (Table 6.2), a member of the X family of polymerases (this DNA polymerase has the unique ability to repair single-stranded DNA gaps smaller than six nucleotides; it fills single nucleotide gaps in DNA produced by the BER pathway of mammalian cells).11 DNA ligase I or III completes the repair by sealing DNA ends.12 Nucleotide Excision Repair In NER a small region of the strand surrounding the damage is removed from DNA as an oligonucleotide. The small gap left in the DNA helix is filled by the sequential action of DNA polymerase and DNA ligase. NER is generally thought of as a repair system for bulky adducts, that recognizes a wide range of substrates, such as damage caused by UV irradiation and chemical agents, including platinum anticancer drugs.

Human DNA Repair Systems 179 Table 6.2 The survey of currently known human DNA polymerases and their function.a89,154 Polymerase name Polymerase a (synonymes are DNA primase, RNA polymerase) Polymerase b Polymerase g Polymerase d Polymerase e Polymerases h, i, k, and Rev1 Polymerase z Polymerase q, l, j, s, and m Others, but their nomenclature is ambiguous

Polymerase proposed function Acts as a primase (synthesizing a RNA primer), and then as a DNA polymerase elongating that primer with DNA nucleotides Is implicated in repairing DNA (BER; single-strand break repair) Replicates mitochondrial DNA Is the main polymerase on the lagging strand in eukaryotes, it is highly processive and has 3′ → 5′ exonuclease activity Is the primary leading strand DNA polymerase in eukaryotes, and is also highly processive and has 3′ → 5′ exonuclease activity Y-family DNA polymerases, TLS B-family DNA polymerase, TLS Not well characterized

a None of the eukaryotic polymerases can remove primers (5′ → 3′ exonuclease activity); that function is carried out by other enzymes. Only the polymerases that deal with elongation (g, d and e) have proofreading ability (3′ → 5′ exonuclease).

NER is a complex process (Figure 6.2) in which basically the following steps can be distinguished: (i) recognition of a DNA lesion; (ii) separation of the double helix at the DNA lesion site; (iii) single strand incision at both sides of the lesion; (iv) excision of the lesion-containing single stranded DNA fragment; (v) DNA repair synthesis to replace the gap and (vi) ligation of the remaining single stranded nick. In step (i) two modes of recognition can be distinguished (Figure 6.2): repair of lesions over the entire genome, referred to as global genome NER (GG-NER), and repair of transcription-blocking lesions present in transcribed DNA strands, hence called transcription-coupled NER (TC-NER). GG-NER removes damage in both transcribed and untranscribed DNA strands in active and inactive genes throughout the genome. This pathway employs several damage-recognition proteins that scan the genome and recognize distortions in double-helical DNA. Unlike GGNER, TC-NER is initiated, not by specific recognition of the DNA damage site by proteins in the NER pathway, but rather by stalling RNA polymerase II. Human GG-NER has been characterized at the biochemical level in considerable detail.13 Six repair factors, XPA, RPA, XPC-RAD23B, TFIIH, XPG and XPFERCC1, are necessary and sufficient to remove damage from DNA. A current model for human NER is as follows: XPA, RPA and XPC-RAD23B locate the damage site; whether the primary damage recognition factor is RPA (or XPA/RPA complex) or XPC-RAD23B is unclear, inconsistent findings in regard to the order of the recognition assembly have been reported. Another NER protein thought to cooperate in DNA damage recognition is XPE protein14 although several studies

180


Figure 6.1 Base excision repair mechanism in mammalian cells

have raised some doubts about the role of XPE in excision repair.14–16 The recognition assembly recruits the TFIIH transcription/repair factor, which contains six polypeptides including helicases XPB and XPD that unwind the DNA around the damage site; the XPG and XPF-ERCC1 subunits are responsible for the 3′ and 5′ dual incisions, respectively. The dual-incision leads to the removal of a singlestranded fragment of DNA with a single-strand gap of 25∼30 nucleotides. The resulting gap in DNA is filled by DNA polymerase d or e (Table 6.2) by copying the undamaged strand. Proliferating cell nuclear antigen (PCNA) assists the DNA polymerase in the reaction, and replication protein A (RPA) protects the other DNA strand from degradation during NER. Finally, DNA ligase I seals the nicks to finish NER.17,18 Mismatch Repair MMR provides several genetic stabilization functions; it corrects errors of DNA replication, ensures the fidelity of genetic recombination and participates in the earliest steps of checkpoint and apoptotic responses to several classes of DNA damage. Hence, a primary function of the MMR pathway is to correct persistent DNA replication errors and avoid the accumulation of deleterious mutations. MMR


Figure 6.2 Nucleotide excision repair mechanism in mammalian cells

repairs base–base mismatches and insertion–deletion loops generated during DNA replication and recombination.19 The MMR system is also of biomedical interest because MMR-deficient tumour cells are resistant to certain cytotoxic chemotherapeutic drugs, including cisplatin.20,21 MMR has evolved to correct errors of DNA polymerases that escape their 3′ → 5′ exonuclease proofreading activity. These errors, mismatches, can be identified in DNA because they fail to form Watson–Crick base pairs. However, because

182


neither nucleotide is damaged or modified, it is not obvious which strand carries the correct genetic information and which carries the error; thus, mismatch repair cannot be accomplished by a mechanism such as BER or NER, which simply excise the damaged base, or a short DNA fragment containing the damage, respectively (see above). Unlike BER and NER, postreplicative MMR has to be targeted exclusively to the newly synthesized strand, which carries, by definition, the erroneous genetic information.21 However, as the parent and daughter strands cannot be distinguished at the site of the mispair, this has to happen elsewhere. The most important features of the current models of the molecular mechanism of MMR are the following:22 The MMR process (Figure 6.3) is initiated by the binding of the hMutSα (hMSH2/ hMSH6 heterodimer), which undergoes an ATP-driven conformational change and recruits the hMLHl/hPMS2 heterodimer (also in an ATP-dependent manner). This ternary complex can translocate in either direction along the DNA. Importantly, the binding of a mismatch or modified DNA base(s) by the hMSH2/hMSH6 heterodimer, or by the hMSH2/hMSH6/hMLHl/hPMS2 complex, does not bring about DNA incision at this site, nor does it lead to damage excision. When it encounters a strand discontinuity that may be bound by a proliferating cell nuclear antigen, loading of an exonuclease initiates degradation of the nicked strand towards the mismatch. Notably, the exonucleolytic degradation of DNA takes place only if a pre-existing strand discontinuity is present in the vicinity. If the exonuclease dissociates before it reaches the mismatch, the single-stranded gap is stabilized by RPA. Loading of the second hMutSα/hMLHl/hPMS2 complex at the mismatch stimulates a second round of exonucleolytic degradation. This process is repeated until the mismatch is removed. Also importantly, the modified DNA bases are not removed during the DNA degradation step unless they are present in the strand containing the discontinuity. The RPA-stabilized single-stranded gap is filled in by the replicative polymerase and the remaining nick can be sealed by DNA ligase.23,24 6.2.3 Repair of Double-Strand Breaks Double-strand breaks, in which the phosphodiester backbones of both strands in the double helix are interrupted, are particularly hazardous to the cell because they can lead to genome rearrangements. Two mechanisms exist to repair double-strand breaks: nonhomologous end joining (NHEJ) and homologous recombination repair (HHR) (also known as template-assisted repair). Nonhomologous End Joining DNA NHEJ is a predominant pathway of DNA double-strand break repair in mammalian cells, and defects in the pathway cause radiosensitivity at the cellular and whole-organism levels. NHEJ is referred to as ‘nonhomologous’ because the break ends are directly ligated without the need for a homologous template, in contrast to homologous recombination (see below), which requires a homologous sequence to guide repair. NHEJ typically utilizes the short homologous DNA sequences in the single-stranded overhangs that are often present at the ends of double-strand breaks and are used to promote restorative repair. When these overhangs are com-


Figure 6.3 Mismatch repair mechanism in mammalian cells

patible, NHEJ almost always repairs the break accurately, with no sequence loss.25 This mechanism does not have quality control on the basis of sequence homology and, hence, is not error-free, and can lead to deletions and inversions. Imprecise repair leading to loss of nucleotides can also occur, but is much less common. Interestingly, cisplatin inhibits repair of double-strand breaks by NHEJ.26–28

184


Figure 6.4 Nonhomologous end-joining mechanism in mammalian cells

The major NHEJ pathway (Figure 6.4) relies on a set of core proteins. The two DNA ends of the double-strand break are recognized and bound by the ring-shaped heterodimer Ku70/Ku80,which recruits the DNA-dependent protein kinase catalytic subunit (DNA-PKcs).29 The assembled DNA-dependent protein kinase (DNAPK) holoenzyme then exhibits serine-threonine protein kinase and DNA end-bridging activities.30 Among other functions, the kinase activity regulates DNA end access to processing enzymes like the DNA-PKcs-associated Artemis nuclease.31,32 Finally, the XRCC4/DNA ligase IV complex is responsible for the ligation step.33 Another core NHEJ factor is Cernunnos-XLF, a factor with a predicted structural similarity to XRCC4 that has been identified as an XRCC4-interacting protein.34 DNA polymerases l and m (Table 6.2) fill gaps during NHEJ.35 Homologous Recombination Repair HRR acts on double-strand breaks generated at broken replication forks or occurring within replicated DNA. Recombinational repair requires the presence of an identical or nearly identical sequence to be used as a template for repair of the break. The enzymatic machinery responsible for this repair process is nearly identical to the machinery responsible for chromosomal crossover during meiosis. This pathway allows a damaged chromosome to be repaired using a sister chromatid or a homologous chromosome as a template.


The HRR process (Figure 6.5) requires the assembly of multienzymatic complexes.36 These complexes include the Rad51 family members such as Rad51, Rad54, Rad51B, Rad51C, Rad51D, Xrcc2 and Xrcc3.37 To initiate the repair of doublestrand breaks in DNA, RAD51 needs to bind to the single-stranded DNA that is produced by nucleolytic resection at the break site. The protein shows a weak specificity for binding single-stranded DNA compared with double-stranded DNA, but this specificity is enhanced by interactions with another key recombination protein, which is known as RAD52. Invasion of a resected end of the double-strand break into duplex DNA takes place in the RAD51 filament and requires the binding of a high energy nucleotide cofactor such as ATP.

6.2.4 DNA Damage Bypass Even when DNA repair and cell cycle checkpoint control are fully functional, some DNA lesions often persist through replication of the genome. Factors that contribute to the persistence of DNA damage include: (i) high levels of damage; (ii) poorly repaired lesions; (iii) inefficiently repaired genomic regions and (iv) DNA damage incurred during the S phase of the cell cycle. Since many lesions that persist despite DNA repair and cell cycle checkpoints hamper or counteract the replication apparatus, cells have evolved a damage tolerance system to allow complete replication in the presence of DNA damage. There are two reasons why it is important for the cell to be able to move replication forks past unrepaired damage. First, long-term blockage of replication forks leads to cell death. Second, replication of damaged DNA provides a sister chromatid that can be used as template for subsequent repair by homologous recombination. DNA damage bypass or translesion synthesis (TLS) is a DNA damage tolerance process that allows the DNA replication machinery to replicate past DNA lesions (Figure 6.6). It involves exchanging regular DNA polymerases by specialized translesion polymerases (Table 6.2), often with larger active sites that can facilitate the insertion of bases opposite damaged nucleotides. The polymerase switching is thought to be mediated by, among other factors, the post-translational modification of the replication processivity factor – PCNA. Many of the TLS enzymes (Table 6.2) belong to the recently described ‘Y-family’ of DNA polymerases. By possessing a spacious preformed active site, these enzymes can physically accommodate a variety of DNA lesions and facilitate their bypass. Flexible DNA-binding domains and a variable binding pocket for the replicating base pair further allow these TLS polymerases to select specific lesions to bypass and favour distinct nonWatson–Crick base pairs. Consequently, TLS polymerases tend to exhibit much lower fidelity (higher propensity to insert wrong bases) than the cell’s regular polymerases when copying normal DNA, which results in a dramatic increase in mutagenesis. Occasionally this can be beneficial, but it often speeds the onset of cancer in humans. However, many TLS polymerases (Table 6.2) are extremely efficient at inserting correct bases opposite specific types of damage. For example, polymerase h mediates error-free bypass of lesions induced by UV irradiation, whereas polymerase z introduces mutations at these sites. From a cellular perspective, risking the introduction of point mutations during TLS may be preferable to resorting to more drastic

186


Figure 6.5 Homologous recombination repair mechanism in mammalian cells


Figure 6.6 Mechanism of translesion synthesis in mammalian cells

mechanisms of DNA repair, which may cause gross chromosomal aberrations or cell death. Cells use both transcriptional and posttranslational regulation to keep these low-fidelity polymerases under strict control and limit their access to a replication fork.38 Lesion bypass directly utilizes the damaged template. Conceptually, lesion bypass can be divided into two steps: (i) nucleotide incorporation opposite the lesion (i.e., translesion synthesis), followed by (ii) extension of DNA synthesis. After a short stretch of extension, normal DNA synthesis by the replication apparatus can then resume. Significantly, lesion bypass can be either error-free, whereby the correct

188


nucleotide opposite the damage is predominantly incorporated, or error-prone, whereby an incorrect nucleotide is frequently incorporated opposite the damage. Consequently, error-free lesion bypass is a mutation-avoiding mechanism, and errorprone lesion bypass is a mutation-generating mechanism. Several proteins, including novel Y-family polymerases that have been recently identified in a multitude of organisms, are involved in DNA damage bypass pathways. Humans have four such (Y-family) polymerases (Table 6.2) – polymerase k, polymerase i, polymerase h, and Rev1 – each with a unique DNA damage bypass and fidelity profile. Polymerase h, for example, is unique in its ability to replicate through UV-induced cyclobutane pyrimidine dimers, while polymerase k is inefficient at replicating through a T-T dimer, but can readily extend from mispaired termini. Polymerase i is perhaps the most unusual with varied efficiencies and fidelities opposite different template bases. Notably, polymerases involved in TLS make up an appreciable fraction of all known eukaryotic DNA polymerases (Table 6.2). Some of the TLS polymerases are capable of replicating accurately past certain types of DNA damage. Rev1 does not have a Greek name since it’s not considered a true DNA polymerase; it can insert only one nucleotide at a time, usually in a template-independent fashion. Polymerases i and k are related to polymerase h but have somewhat different bypass specificities and are not usually error-free. Polymerase q is especially competent at bypassing AP sites, where it inserts an A residue. Since most AP sites are consequences of purine loss, and A is a purine base, repair by polymerase q is frequently error-free.

6.3 Specific Binding of Repair Proteins to DNA Modified by Antitumour Platinum Compounds A number of proteins that are components of various cellular DNA repair systems bind to DNA modified by antitumour platinum drugs specifically (Table 6.3). When these proteins are absent, repair of DNA damage by platinum complexes is diminished; sensitivity of cells to the platinum drug is enhanced, which is consistent with the hypothosis that these proteins modulate biological effects of these metallodrugs. Some of these proteins initiate specific cell signalling pathways. Recognition of DNA adducts of antitumour platinum compounds by DNA repair systems has mostly been examined in the case of DNA adducts of cisplatin and its analogs cis-diamminecyclobutanedicarboxylatoplatinum(II)] (carboplatin) and [(1R,2R-diamminocyclohexane)oxalatoplatinum(II)] (oxaliplatin). Cisplatin binds to DNA, forming mainly intrastrand crosslinks between adjacent purine residues (∼90%) (1,2-GG or 1,2-AG intrastrand crosslinks).39,40 Other minor adducts are 1,3-GXG intrastrand crosslinks (X = A, C, T), interstrand crosslinks (∼6% in linear DNA) preferentially between guanine residues in the 5′GC/5′-GC sequence41 and monofunctional lesions. The adducts formed by cisplatin in DNA affect its secondary structure.42 The formation of major 1,2-crosslinks of cisplatin locally unwinds DNA duplex and bends it towards the major groove, expos-

Specific Binding of Repair Proteins to DNA

189

Table 6.3 Examples of human proteins involved in DNA repair that were shown to bind specifically to DNA modified by antitumour platinum compounds Protein 3-methyladenine DNA glycosylase (AAG) XPA RPA XPC-HR23B RNA polymerase II PARP-1 hMutSα (heterodimer of hMSH2 and hMSH6) Ku80 SSRP1

Proposed function

Reference

BER NER NER, HRR NER TC-NER TC-NER MMR NHEJ Transcription elongation

97 51 51,53 60 64–67 69 70,71 76,78 95

ing a wide, shallow minor groove surface to which several classes of proteins bind. The interstrand crosslink also induces several irregularities in DNA.43 The crosslinked guanine residues are not paired with hydrogen bonds to the complementary cytosines, which are located outside the duplex. The cis-diammineplatinum(II) bridge resides in the minor groove and the double helix is locally reversed to a lefthanded, Z-DNA-like form, markedly locally unwound and bent towards the minor groove. The 1,3-GXG intrastrand crosslink formed by cisplatin locally unwinds and bends the helix axis towards the major groove, but in contrast to 1,2-intrastrand adducts, DNA is locally denatured and flexible at the site of the 1,3-adduct.44 The monofunctional adducts of cisplatin distort DNA in a sequence-dependent manner.45,46 These distortions disturb stacking interactions in double-helical DNA, slightly unwind the duplex, but no intrinsic bending is induced by these adducts. Direct analogues of cisplatin, such as carboplatin and oxaliplatin, produce adducts similar to those produced by the parent drug, though different in their relative rates of formation.47–49 6.3.1 Base Excision Repair System The human 3-methyladenine DNA glycosylase (AAG) is a BER enzyme that removes a number of damaged bases from DNA, including adducts formed by some chemotherapeutic agents. AAG readily recognizes cisplatin adducts, such as 1,2-GG, 1,2-AG and 1,3-GTG intrastrand crosslinks. 1,N-6-ethenoadenine, which is repaired efficiently by AAG, is recognized considerably more weakly than cisplatin adducts.96 6.3.2 Nucleotide Excision Repair System The NER proteins that exhibit enhanced affinity to cisplatin-modified DNA are mainly those involved in the first step in repair pathways, i.e. in damage recognition. The repair proteins that probably have attracted the most attention are those that are absent in patients suffering from the NER deficiency characteristic of the disease

190


Xeroderma pigmentosum (XP). The minimal factors necessary for removal of damaged nucleotides also include XPA and RPA which are among the major damage-recognition proteins involved in the early stages of NER. XPA (32 kDa) and RPA (70, 34 and 14 kDa subunits) are able to bind damaged DNA independently, although RPA interaction stimulates XPA binding to damaged DNA. XPA belongs among proteins that bind preferentially to bent or kinked DNA duplexes.50 Interestingly, XPA protein binds with considerable selectivity to the rigid double-stranded kink induced by the 1,2-GG intrastrand crosslink of cisplatin.51 In contrast, XPA fails to recognize the site of strong duplex destabilization generated by the same crosslink, but formed by the bifunctional dinuclear platinum complex [{trans-PtCl(NH3)2}2H2N(CH2)4NH2]Cl251 that fails to impose a rigid kink, but, instead, increases DNA flexibility in a nondirectional manner.52 RPA, on the other hand, recognizes the helical destabilization caused by the dinuclear complex more effectively than the cisplatin-induced kink. RPA binds with a relatively low affinity to double-stranded DNA containing the 1,2-GG intrastrand crosslink.53 The preference of RPA for the more helix destabilizing crosslink54 is consistent with previous studies suggesting that RPA is recruited to damaged double-stranded DNA through interactions with single-stranded sites.53,55 It has been suggested51 that XPA in conjunction with RPA constitutes a regulatory factor that monitors DNA bending and unwinding. In addition, it has been demonstrated56 that the particular manner in which XPA and RPA proteins align themselves with respect to the 1,3-GNG intrastrand crosslink involves their binding in close proximity to the adduct, with XPA in contact with both the platinated and unplatinated strands of DNA, and RPA binding preferentially to the unplatinated strand. It has also been shown57 that a direct interaction between RPA and XPA proteins facilitates the assembly of a preincision complex during the NER processing of DNA containing a 1,2-GG intrastrand crosslink. A ternary complex of RPA and XPA both bound to DNA containing the 1,2-GG intrastrand crosslink displays modest specificity for platinated versus unplatinated DNA. In addition, the RPA–XPA complex exhibits a greater affinity for binding cisplatin-damaged duplex as compared to the RPA or XPA proteins alone. It has also been demonstrated57 that the role of XPA is to stabilize the doublestranded DNA structure by inhibiting the strand separation activity of RPA. XPA is also believed to participate in DNA damage recognition of cisplatin adducts, whereas it has no affinity for DNA adducts of clinically ineffective transplatin.58 The effects of the lesions induced by single, site-specific 1,2-GG or 1,3-GTG intrastrand adducts of cisplatin on the energetics of DNA have been examined as well.59 The results have confirmed that DNA bending is the specific determinant responsible for high-affinity interactions of XPA with damaged DNA. An additional important factor is a change of thermodynamic stability of DNA induced by the damage. These results confirm that RPA preferentially binds to DNA distorted in such a way that hydrogen bonds between complementary bases are interrupted. RPA also binds to nondenatured distortions in double-helical DNA, but the affinity of RPA to these distortions is insensitive to alterations of thermodynamic stability of damaged DNA. XPC-HR23B displays an enhanced binding affinity for 1,3-GTG intrastrand crosslinks compared to unmodified DNA.60 In addition, XPC-hHR23B associates

Specific Binding of Repair Proteins to DNA

191

with the 1,2-GG and 1,3-GXG intrastrand crosslinks with similar rates. The more rapid rate of dissociation of XPC-HR23B from the 1,2-GG compared to the 1,3GXG intrastrand crosslink is likely a result of the lack of localized melting of the duplex DNA containing the 1,2-GG intrastrand adduct. RPA is one of the proteins that belongs to the initial damage-sensing factors initiating NER (see above). The relative binding affinities of RPA to the different cisplatin adducts correlate with the repair of the adducts observed in an in vitro NER assay;61 RPA has higher affinity for the 1,3-GTG intrastrand than for the 1,2GG intrastrand crosslink of cisplatin, and the 1,3-GTG adduct is more efficiently repaired by the human exonuclease than the 1,2-GG adduct.53 Thus, binding of RPA to damaged DNA is a sensitive indicator to predict whether mammalian NER could be effective in the removal of damaged nucleotides. BBR3464 is the novel phase II antitumour polynuclear platinum drug ([(trans-PtCl(NH3)2)2(m-trans-Pt-(NH3)2(NH2 (CH2)6NH2)2)]4+), which forms on DNA so-called long-range intra- and interstrand crosslinks. The effect of 1,4-GG intrastrand and interstrand crosslinks of BBR3464 on RPA binding was examined62 (the platinated bases in 1,4-crosslinks are separated by two intervening base pairs). Interestingly, RPA did not bind to DNA containing interstrand crosslinks of BBR3464. Thus, these results predict very low efficiency of the mammalian NER pathway to remove the interstrand crosslinks of BBR3464, subsequently proved by in vitro experiments.62,63 TC-NER is initiated by stalling RNA polymerase II. Therefore, cisplatin DNA adduct processing by RNA polymerase II has been investigated.64–67 It was found that RNA polymerase II stalls in front of a cisplatin intrastrand crosslink because it does not pass a translocation barrier that impairs delivery of the bulky crosslink into the active site. In vitro assays indicate that a considerable fraction of stalled RNA polymerase II proteins remains strongly associated with DNA damaged by cisplatin following RNA polymerase II arrest.66 RNA polymerase II stalled in front of a cisplatin adduct (1,3-GTG intrastrand crosslink) serves as a decoy to sequentially recruit TFIIH, XPA, RPA, XPG and XPF repair factors in an ATP-dependent manner. This RNA polymerase II/repair complex allows the ATP-dependent removal of the cisplatin lesion.65 In conclusion, the anticancer drug cisplatin forms DNA intrastrand crosslinks that stall RNA polymerase II, and in this way cisplatin triggers TC-NER.67 In addition, recent observations demonstrate a role for poly-(ADPribose)polymerase-1 (PARP-1) in TC-NER68 and PARP-1 binding to 1,2-GG intrastrand crosslinks in nuclear cell extracts.69 Thus, the latter observations indicate that poly(ADP-ribosylation) activity may be crucial for cellular NER of platinated DNA.

6.3.3 Mismatch Repair System Similarly to the case of NER, the enhanced affinity to DNA damaged by cisplatin is exhibited mainly by the MMR proteins involved in damage recognition, i.e. by one of two heterodimers of MutS homologues, hMutSa (heterodimer of hMSH2 and hMSH6) or hMutLa (hMLH1 and hPMS2 heterodimer). The purified hMSH2 protein binds to DNA containing a single 1,2-GG intrastrand crosslink.70 Moreover, hMSH2

192


displays selective affinity for DNA modified by cisplatin, but not for DNA modified by transplatin70 and preferentially recognizes cisplatin-modified DNA over oxaliplatinmodified DNA.71 The hMutSα heterodimer recognizes the 1,2-GG intrastrand crosslink of the cisplatin adduct, but it has no affinity for the 1,3-GTG intrastrand crosslink of transplatin.72 The binding affinity of hMutSα for the 1,2-GG intrastrand crosslink of cisplatin is enhanced in duplexes containing a thymine incorporated opposite the 5′ or the 3′ guanine of the cross-link.73,74 The data suggest that hMutSα may interact with cellular platinum-DNA lesions (especially with intrastrand crosslinks superimposed on a mismatch73) and influence the DNA repair and signalling pathway, although a detailed mechanism of these processes is not known. 6.3.4 Binding of Proteins Involved in the Nonhomologous End-Joining System DNA-PK is a core protein of the NHEJ pathway and binds to DNA modified by cisplatin. Its K80 subunit, which is responsible for the binding,75 also has a high affinity to the major adduct of cisplatin, the 1,2-GG intrastrand crosslink, which is only twofold lower than its affinity to DNA ends.76 DNA adducts of cisplatin do not alter the ability of Ku to bind DNA ends, but do impair the translocation of Ku proteins along DNA,77 resulting in reduced affinity of DNA-PKc to the Ku-DNA complex, and reduced kinase activity.78 6.3.5 Homologous Recombination Repair System The Rad51-guided HRR system plays an important role in the recognition and repair of DNA interstrand crosslinks. It is not unexpected that cells deficient in the HRR pathway become hypersensitive to cisplatin as well as to other interstrand crosslinking agents. The repair of interstrand crosslinks has been shown to involve components of the NER, HRR and TLS pathways.79 However, the manner by which HRR proteins specifically bind platinum adducts is unclear. 6.3.6 Processing of DNA Modified by Antitumour Platinum Compounds by Translesion Synthesis DNA Polymerases Several TLS DNA polymerases (Table 6.2) bypass 1,2-GG intrastrand crosslinks of cisplatin.80–85 DNA polymerases that bypass cisplatin adducts in vitro include DNA polymerases b, m and h, whereas polymerases a, i, k and l are unable to perform TLS past platinum adducts.80,82,83,86 Each DNA polymerase displays a distinct specificity in its lesion-bypass properties, including bypass ability, fidelity and extension ability. For example, DNA polymerase h bypasses platinum adducts most efficiently in error-free TLS,81,85 whereas polymerase m is the most error-prone enzyme. Moreover, two DNA polymerases often work together to complete TLS.87 Although polymerase z is unable to bypass certain DNA lesions, including those by platinum agents, the enzyme has the ability to extend TLS once nucleotides are inserted opposite DNA adducts by other polymerases.88,89 Little is known about the TLS past platinum interstrand crosslinks, although a single DNA polymerase is not likely to

Repair of DNA Damage by Antitumour Platinum Compounds 193

be able to bypass this lesion. It has been suggested, however, that TLS may occur during the repair of interstrand crosslinks.79,90 Interestingly, polymerase h bypasses 1,2-GG crosslinks of oxaliplatin more readily than 1,2-GG crosslinks of cisplatin.91 This suggests that the conformations of cisplatin- and oxaliplatin-GG adducts are different and that these conformational differences have an influence on the ability of distantly related DNA polymerases to perform translesion synthesis. 6.3.7 Binding of Other DNA Repair Proteins The high-mobility-group (HMG) SSRP1 protein (structure-specific recognition protein 1) is a member of a conserved chromatin-remodelling complex (FACT/ DUF/CP) implicated in DNA replication, basal and regulated transcription, and also in DNA repair.92,93 SSRP1 is the first HMG-domain protein found to bind selectively to DNA adducts of cisplatin.94,95 The role of this protein in repair of DNA adducts of platinum compounds is, however, unknown.

6.4 Repair of DNA Damage by Antitumour Platinum Compounds 6.4.1 Base excision Repair The presence of cisplatin adducts in the reactions inhibits the excision of 1,N-6ethenoadenine by BER DNA glycosylase AAG. AAG readily recognizes cisplatin intrastrand adducts while the 1,N-6-ethenoadenine adduct, which is repaired efficiently by AAG, is recognized considerably more weakly. Despite the affinity of AAG for cisplatin adducts, AAG is unable to remove any of these adducts from DNA. It has been suggested that cisplatin adducts could titrate AAG away from its natural substrates, resulting in higher mutagenesis and/or cell death because of the persistence of AAG substrates in DNA.96 Another intriguing possibility is that AAG binds to cisplatin adducts and, because it cannot remove them, it conveys the adducts to the NER pathway through the interaction of AAG with the hHR23 proteins.97 Thus the interaction between AAG and DNA adducts of cisplatin can increase the efficiency of DNA repair and, in the case of overexpression of this protein, could lead to cisplatin resistance.97,98 Hence, cisplatin may exhibit a synergistic effect in potentiating the toxic effects of agents that are substrates for AAG. 6.4.2 Nucleoide Excision Repair A major pathway used by human cells for the removal of platinum adducts from DNA is NER. Typical experiments demonstrating NER of DNA adducts of platinum complexes were performed using linear DNA fragments of ∼150 base pairs containing a platinum adduct in the centre of one strand. To allow for determination of excision activity in human or rodent cell extracts, the modified sequences con-

194


tained a 32P-labeled phosphate at the phosphodiester bond near the lesion on its 5′ side. Following incubation with cell extract or a reconstituted system containing highly purified NER factors, damage-containing oligonucleotides are released as radiolabelled DNA fragments that are visualized after gel electrophoresis and autoradiography. Incubation in mammalian cell extracts or a reconstituted system containing highly purified NER factors demonstrated that both major 1,2- and minor 1,3-intrastrand crosslinks of cisplatin are efficiently repaired, the 1,3-crosslink being repaired much more efficiently.61 Both crosslinks bend and unwind DNA to a similar extent, but in contrast to 1,2-intrastrand crosslinks, the 1,3-intrastrand adduct of cisplatin locally denatures and induces a flexibility at the site of the 1,3-adduct.44 Hence, the differential NER can be attributed to the different DNA structures induced by the 1,2- and 1,3-intrastrand cisplatin adducts. Denatured base pairs and locally enhanced flexibility induced in DNA by platinum complexes may be factors enhancing recognition of platinum adducts by NER systems. On the other hand, in the extracts from human HeLa cells, the 1,2-AG intrastrand crosslink is excised more efficiently than the 1,2-GG intrastrand adduct61 although these 1,2-intrastrand crosslinks distort the DNA in a similar way (DNA unwinding and bending are similar6,99,100 and hydrogen bonds in the distorted base pairs are preserved). NER of 1,2-, but not of 1,3-intrastrand crosslinks, is blocked upon addition of an HMG-domain protein (HMG = high mobility group).101 Several mechanisms have been considered for how HMG-domain proteins might modulate the sensitivity of cells to cisplatin.61 One prominent hypothesis is that HMG-domain proteins block cisplatin–DNA adducts from the damage recognition needed for repair. Hence, in the cells where this shielding mechanism is responsible for their sensitivity to cisplatin, 1,2- and not 1,3-intrastrand crosslinks of this metallodrug are most likely candidates for genotoxic lesions relevant to the antitumour effects of cisplatin. An in vitro NER of a site-specific cisplatin interstrand crosslink has also been studied using mammalian cell-free extracts containing HMG-domain proteins at levels insufficient to block NER of the 1,2-intrastrand adducts.61 Repair of the interstrand crosslink formed by cisplatin was not detected. Similarly, in cell strains derived from patients with Fanconi’s anaemia, NER of cisplatin-interstrand crosslinks is not observed, although NER can readily occur in these cells.102,103 Fanconi’s anaemia cells have been described as being extremely sensitive to interstrand crosslinking agents, and it was suggested that this high sensitivity to cisplatin could be explained by the incapability of these cells to repair cisplatin interstrand crosslinks.104 On the other hand, repair of these lesions has been detected with the aid of a repair synthesis assay, which measures the amount of new DNA synthesized after damage removal in whole-cell extracts.105 In this way, however, the repair could also result from a mechanism different from that of NER. DNA interstrand crosslinks pose a special challenge to repair enzymes because they involve both strands of DNA and therefore cannot be repaired using the information in the complementary strand for resynthesis. Quite recently the processing of stalled forks caused by DNA interstrand crosslinks has been proposed to be an important step in initiating mammalian interstrand crosslink repair. It has been demonstrated106 that the XPFERCC1 complex makes an incision 5′ to a psoralen interstrand crosslink on Y-shaped DNA and that the XPF-ERCC1 complex generates an interstrand crosslink-specific incision on the 3′-side of this lesion. The interstrand crosslink-specific 3′ incision,


along with the 5′ incision, on the crosslinked Y-shaped DNA results in the separation of the two crosslinked strands (the unhooking of the interstrand crosslink) and the induction of a double-strand break near the crosslinked site. These results implicate the XPF-ERCC1 complex in initiating interstrand crosslink repair by unhooking the interstrand crosslink, which simultaneously induces a double-strand break at a stalled fork. That this mechanism asserts itself also in repair of interstrand crosslinks of cisplatin and other platinum complexes awaits further experimental confirmation. The differential recognition of the 1,2-GG intrastrand crosslink of cisplatin and dinuclear platinum complex [{trans-PtCl(NH3)2}2H2N(CH2)4NH2]Cl2 with divergent effects on DNA conformation (see above) by XPA and RPA prompted testing of NER activity in response to these two platinum lesions.51 Incubation of linear DNA fragments containing the platinum crosslink in HeLa cell extract demonstrated that 1,2-GG intrastrand crosslinks of both platinum complexes were able to elicit oligonucleotide excision. These results are consistent with the view, and support the hypothesis that XPA-RPA constitutes a universal sensor of nucleotide damage that operates through indirect readout of DNA conformations. XPA-RPA is guided by the synergism between two different DNA binding domains: XPA recognizes abnormal rigid bending of the deoxyribose-phosphate backbone, whereas RPA recognizes base pair disruption. These two nucleic acid interaction modules are used to monitor integrity of the Watson–Crick helix by a double-check probing mechanism that confers an extremely wide recognition capacity for structural distortions. Cisplatin also forms DNA–protein cross-links107 The DNA–protein cross-links formed by cisplatin inhibit removal of these ternary lesions from DNA by NER systems more effectively than plain DNA intrastrand or monofunctional adducts. Thus, the bulky DNA–protein crosslinks formed by cisplatin represent a more distinct and persisting structural motif recognized by the components of downstream cellular systems processing DNA damage considerably different from that of the plain DNA adducts of this metallodrug. In addition, in cell-free media, cisplatin forms DNA–protein crosslinks considerably more effectively than the clinically ineffective transplatin. This implies that there is a positive correlation between the efficiency of mononuclear bifunctional platinum complexes to form DNA–protein crosslinks and their antitumour effects. Thus, these results are consistent with previous findings demonstrating that DNA–protein crosslinks persist longer in cells exposed to cisplatin than in those exposed to the chemotherapeutically inactive trans analogue,108–110 which suggests relevance of such lesions to the antitumour effects of cisplatin. The failure to remove bulky DNA–protein crosslinks has been suggested to be due to the steric hindrance caused by the size of the crosslinked protein,which may interfere sterically with the assembly of the mammalian NER system. Thus, the results demonstrating failure of the mammalian NER system to remove the DNA–Pt–protein crosslinks107 reinforce the view that DNA–protein crosslinks formed by platinum drugs may be among the critical lesions relevant to their antitumour effects. In eukaryotic cells, DNA is packed and the smallest structural unit of eukaryotic DNA packaging, is the nucleosome. The nucleosome is made up of about 146 base pairs of DNA and four pairs of proteins called histones. Histones H2A, H2B, H3 and H4 form the core of the nucleosome, around which the DNA is wrapped,

196


while histone H1 sits on the base of the nucleosome at the junction between nucleosome DNA and linker DNA, extending along the DNA into the linker region.111 Nucleosomes modulate many cellular processes so that it is of interest to know how NER of DNA adducts of platinum compounds is affected when these adducts are present in nucleosomes. Nucleosomes were prepared containing either a site-specific 1,2-GG or 1,3 GTG intrastrand crosslink of cisplatin.112 It was found that the nucleosome significantly inhibited NER, the inhibition being more pronounced in the case of the 1,3intrastrand crosslink. Interestingly, excision from native nucleosomal DNA was higher than the level observed with recombinant material. This result reveals that posttranslational modification of histones can modulate NER from damaged chromatin. Despite its success, cisplatin has several disadvantages: it is active against a limited range of human tumours, can cause severe side effects and its administration may result in acquired resistance. The drawbacks, coupled with the toxicity of cisplatin and its analogues, have been the impetus for the development of platinum drugs with improved pharmacological properties and a broader range of antitumour activity. Antitumour activity of platinum drugs is multifactorial in nature and also includes contributions from differential DNA repair mechanisms. Therefore it is of interest to know whether NER and other repair pathways discriminate between cisplatin and novel platinum drugs that exhibit altered biological effects. A study of this kind was focused on NER of DNA adducts formed by a third generation antitumour platinum drugs oxaliplatin [(1R,2R-diamminocyclohexane) oxalato-platinum(II)] and satraplatin [bis-acetatoamminedichlorocyclohexylamine platinum(IV), JM216].113 Interestingly, the types of DNA lesions generated by the three platinum drugs, cisplatin, oxaliplatin and satraplatin, are repaired in vitro with similar kinetics by the mammalian NER pathway. This result is consistent with reports that major DNA adducts of cisplatin, oxaliplatin and satraplatin cause similar distortions in DNA.114,115 Besides oxaliplatin ([Pt(R,R-DACH)]2+, DACH = 1,2-diaminocyclohexane), another enantiomeric form of this complex exists, [Pt(S,S-DACH)]2+. Whereas conformational alterations induced in DNA by the major adduct of [Pt(R,R-DACH)]2+ are not markedly different from those induced by the adduct of cisplatin, conformational alterations induced by the major adduct formed by [Pt(S,S-DACH)]2+ in the TGGT sequence are different.115 The major differences resulting from the modification by the two enantiomers consist in sequence-dependent thermodynamic destabilization and conformational distortions. Interestingly, the 1,2-GG intrastrand crosslinks of [Pt(R,R-DACH)]2+ bind to HMG-domain proteins with a similar affinity as the same crosslink of cisplatin.115 In contrast, the crosslink formed by [Pt(S,S-DACH)]2+ in the TGGT sequence binds to the HMG-domain protein with a noticeably lower affinity. Similar results have also been obtained with DNA modified by [Pt(DAB)]2+ (DAB = 2,3-diaminobutane) enantiomers (which are closely related to [Pt(DACH)]2+ enantiomers).116–118 Intrastrand crosslinks between neighbouring guanine residues of both enantiomers of [Pt(DAB)]2+ were efficiently removed from DNA by NER in an in vitro


assay using mammalian or rodent cell-free extracts.117 Consistent with the different affinity of the intrastrand crosslinks of both enantiomers to HMG-domain proteins, addition of HMG-domain protein only blocked NER of the crosslinks of [Pt(R,RDAB)]2+, whereas NER of the crosslinks formed by [Pt(S,S-DAB)]2+ in the TGGT sequence remained unaffected. Thus, these results suggest that very fine structural modifications, such as those promoted by enantiomeric ligands in bifunctional platinum(II) compounds, can modulate the ‘downstream effects’, such as specific protein recognition by DNA processing enzymes and other cellular components. More specifically, as a consequence of enantiomorphism of carrier amine ligands of cisplatin analogues, their major DNA crosslinks not only exhibit different conformational features, but are also processed differently by cellular components, including NER. It may be also suggested that these differences are associated with different antitumour and other biological activities of the two enantiomers. The original empirical structure–activity relationships considered the trans isomer of cisplatin and other transplatin analogues to be inactive.119 However, several new analogues of transplatin, which exhibit a different spectrum of cytostatic activity including activity in tumour cells resistant to cisplatin have been identified. Examples are analogues containing heterocyclic amine ligands, aliphatic ligands or iminoether groups.3,120,121 The molecular mechanism of antitumour activity of transplatinum compounds has been extensively investigated. Unlike cisplatin or transplatin, modification of DNA by antitumour trans compounds leads to monofunctional and bifunctional intra- and interstrand crosslinks in different proportions. In addition, structure of intrastrand and interstrand crosslinks of antitumour trans complexes and conformational distortions induced by these adducts in DNA are mostly fundamentally different from that of the crosslinks of cisplatin and parent transplatin. Thus, it is not surprising that cellular processing of DNA adducts of antitumour transplatin analogues, including NER, is also different. The interstrand crosslinks of these trans-platinum compounds are not removed from DNA in the same way as the corresponding adducts of cisplatin. In contrast to cisplatin, intrastrand crosslinks of antitumour trans-platinum analogues are removed from DNA by NER with markedly lower efficiency than the major adducts of cisplatin.122,123 Hence, a general property of DNA adducts of these new antitumour trans-platinum compounds is that they can persist for a considerably long time in cells, which is a prerequisite for the toxicity of these lesions in tumors sensitive to these transplatin analogs, without being protected from the recognition and attack of the NER components. An intriguing example of how NER modulates the antitumour effects of transplatinum compounds is the case of trans-[PtCl2(E-iminoether)2]. The replacement of both NH3 groups in transplatin by an iminoether ligand results in a marked enhancement of its cytotoxicity. It is more cytotoxic than its cis congener and exhibits significant antitumour activity, including activity in cisplatin-resistant tumour cells.124 The new trans compound (trans-[PtCl2(E-iminoether)2]) forms mainly stable monofunctional adducts at guanine residues when reacted with DNA.125 These adducts locally distort DNA, are not recognized by HMG-domain proteins and are readily removed from DNA by the NER system. The monofunctional adducts of trans-[PtCl2(E-iminoether)2] readily crosslink proteins, which markedly enhances

198


the efficiency of this adduct to terminate DNA polymerization in vitro and to inhibit removal of this adduct from DNA by the NER system.126 Hence, DNA–protein ternary crosslinks produced by trans-[PtCl2(E-iminoether)2] persist considerably longer than the noncrosslinked monofunctional adducts, which potentiates toxicity toward tumour cells. Thus, trans-[PtCl2(E-iminoether)2] belongs to a class of platinum antitumour drugs in which activation of the trans geometry is associated with an increased efficiency to form DNA–protein ternary crosslinks. Polynuclear platinum compounds comprise a unique class of anticancer platinum agents with distinct chemical and biological properties different from mononuclear platinum drugs.127 The lead compound is BBR3464, a trinuclear, bifunctional DNA binding agent. Quite surprisingly, in comparison with cisplatin, enhanced cellular uptake of the charged polynuclear platinum compounds has been demonstrated,128,129 but the enhanced uptake is not sufficient to explain the increased toxicity of BBR3464 in tumour cells.130 The interactions of antitumour polynuclear platinum compounds with target DNA (for review see, for instance, ref. 127) are distinct from the mononuclear-based cisplatin family and, indeed, unlike those of any DNA-damaging agent in clinical use. The ability of BBR3464 to induce long-range delocalized intraand interstrand crosslinks,131 which are not produced by conventional mononuclear platinum compounds, suggests that BBR3464 may escape, at least in part, the classical mechanism of cisplatin resistance related to DNA damage recognition and repair. The crosslinks of BBR3464 distort the DNA conformation. In contrast to distortions induced by the major crosslinks of cisplatin, the crosslinks of BBR3464 do not extensively unwind and rigidly bend DNA. Hence they are not substrates for damaged-DNA binding proteins, such as HMG-domain proteins.62,63 On the other hand, while intrastrand adducts of BBR3464 are readily removed from DNA by the NER systems, the interstrand crosslinks are not. It has been suggested62,63 that interstrand, and not intrastrand, crosslinks of BBR3464 could persist for a sufficiently long time in cells to potentiate toxicity of this drug.

6.4.3 Mismatch Repair Recent observations support the view that another cellular repair mechanism, such as MMR, can affect the antitumour efficiency of cisplatin,132,133 and that dysfunction of this type of DNA repair may result in the resistance of tumour cells to cisplatin, or in drug tolerance.134 The function of the MMR is to scan newly synthesized DNA and remove mismatches that result from nucleotide incorporation errors made by the DNA polymerases. To explain cisplatin tolerance, it is assumed that replication bypass of DNA adducts of cisplatin leads to mutations. During MMR, the strand to be corrected is nicked, a short fragment containing the mismatch is excised, and a new DNA fragment is synthesized. The MMR system always replaces the incorrect sequence in the daughter strand, which would leave the cisplatin adduct unrepaired. This activity initiates a futile cycle. During DNA synthesis to replace the excised short fragment, the DNA polymerases would again incorporate mutations followed by attempts to remove them. The repeated nicks in DNA formed at each ineffective cycle of repair


could trigger a cell death response. Thus, MMR recognition of cisplatin adducts on DNA may trigger a programmed cell death pathway rendering MMR-proficient cells more sensitive to DNA modification by cisplatin than MMR-deficient cells. In other words, loss of MMR would then increase the cell’s ability to tolerate lesions formed on DNA by cisplatin. The MMR is likely to act mainly in the postreplicative phase, when the highest concentration of mispairs is expected in the newly synthesized DNA. This would be in contrast to NER, which probably acts independently of the cell cycle. Thus, rapidly proliferating tumour cells would be more prone to futile repair attempts capable of triggering cell death. Although the level of resistance to cisplatin that accompanies loss of MMR is relatively small, loss of MMR has been shown to be sufficient to account for the failure of treatment with cisplatin in several model systems.135–137 Nevertheless, a direct connection between the MMR pathway and cytotoxicity of cisplatin has not been established. 6.4.4 Nonhomologous End-Joining Radiation exposure can produce many types of DNA damage, including doublestrand breaks. In a local multiply damaged site, where an adduct of cisplatin occurs close to a single-strand break, the damage may be initially repaired by NER. However, during the dual incision step, a double-strand break, which is the most lethal type of a cell damage, is produced. Alternatively, if an adduct of cisplatin is close to a double-strand break, it could physically block cisplatin removal and repair by the NER pathway, or the cisplatin adduct could hinder the double-strand break repair pathways by restricting the access of the protein complexes involved. Another possibility is that the NER and proteins involved in repair of double-strand breaks may compete for access to the local multiply damaged site. Investigations into the efficiency of NER have shown that NER is less efficient if cisplatin adducts are near to a double-strand break due to binding of the heterodimer Ku70/80 at the DNA ends.138 In addition, a decreased amount of NER proteins binds to the lesion if a double-strand break is present.139,140 Double-strand break NHEJ, which requires DNA-PK, is also inhibited by cisplatin-damaged DNA in cell extracts.26,27 This process may in part explain the radiosensitizing effect of cisplatin administered during concurrent chemoradiation. 6.4.5 Homologous Recombination Repair Several studies have shown that HRR plays a role in the repair of DNA damage by cisplatin in mammalian cells.21,141,142 HRR of unrepaired cisplatin adducts at stalled replication forks appears to be a mechanism of cellular resistance to cisplatin in NER-deficient and MMR-deficient cells. In addition to its function in NER and its ability to uncouple cisplatin interstrand crosslinks, the ERCC1/XPF structurespecific nuclease has a role in the repair of cisplatin adducts related to recombination processing.143 Homologous recombination might be initiated prior to excision of interstrand crosslinks of cisplatin. Because of its probable role in HRR of DNA adducts of cisplatin, ERCC1 is an attractive therapeutic target, especially in MMRdeficient cancers, where increased recombinational bypass of cisplatin lesions

200


enhances resistance to this drug. It has been demonstrated144 that downregulation of ERCC1 significantly sensitizes the MMR-deficient prostate cancer cell line to cisplatin, where inhibition of the NER pathway had no effect. In addition, a recent study21 suggests that homologous recombination status might represent a novel prognostic marker and possibly also a therapeutic target.

6.5 Implications for Design of Antitumour Platinum Compounds The occurrence of inherent and acquired resistance to platinum is a major problem that undermines efforts to effectively treat cancer. This limitation in cancer treatment can be overcome by elucidating the mechanisms underlying resistance to platinum drugs and then developing ways to treat resistant tumours effectively or prevent its occurrence.145 DNA repair is a major resistance mechanism to platinumbased chemotherapy. Evidence for increased repair of platinum–DNA damage in resistant cancer cells has been demonstrated by a variety of cellular assays. In human cells, major intrastrand crosslinks of cisplatin are removed from DNA mainly by the NER system.61 Importantly, the repair of major intrastrand crosslinks of cisplatin may be blocked by high-mobility-group (HMG) domain proteins, and perhaps by other proteins that bind to damaged DNA,3,97,146,147 causing drug resistance to be lower. Thus, the ability of cancer cells to recognize damage induced in DNA by platinum antitumour drugs and initiate DNA repair is an important mechanism for therapeutic resistance and has a negative impact upon therapeutic efficacy. Pharmacological inhibition of DNA repair, therefore, has the potential to enhance efficiency of antitumour chemotherapy. It implies that design of new antitumour platinum drugs should focus on those new platinum complexes that form major adducts with DNA that are difficult to remove by DNA repair systems. This strategy has been employed by designing and synthesizing antitumour dinuclear platinum complexes [(cis-{Pt-(NH3)2})2(m-OH)(m-pz)]2+ (pz = pyrazolate)148 or dinuclear azole-bridged platinum compounds.149–152 These new platinum complexes exhibit higher toxic effects in some tumour cell lines and form 1,2-GG instrastrand crosslinks, which bend DNA markedly less than cisplatin. It has been suggested that these minor structural changes are not recognized by the DNA repair proteins and in this way escape repair. This assumption must, however, await further experimental confirmation. It cannot be excluded that the bulkiness of the adduct in addition to its propensity to induce only very weak conformational distortions is an important factor involved in the initiation of DNA repair. Moreover, the use of inhibitors of DNA repair or DNA damage signalling pathways also appears to provide an opportunity to enhance or alter the antitumour effects of platinum agents and to target the genetic differences that exist between normal and tumour tissue as well.153 On the other hand, use of DNA repair inhibitors or their combination with such anticancer therapies is also likely to enhance the risk of mutagenic DNA damage and hence the risk of secondary cancer. Thus, concepts for antitumour strategies based on DNA repair inhibitors must await further studies.

References

201

Acknowledgements This work was supported by the Ministry of Education of the Czech Republic (MSMT LC06030, 6198959216, ME08017, OC08003), the Academy of Sciences of the Czech Republic (Grants 1QS500040581, KAN200200651, AV0Z50040507 and AV0Z50040702), the Grant Agency of the Academy or Sciences of the Czech Republic (IAA400040803) and the Grant Agency of the Ministry of Health of the Czech Republic (NR8562-4/2005). J.K. is the international research scholar of the Howard Hughes Medical Institute. The authors also acknowledge that their participation in the EU COST Action D39 enabled them to exchange regularly the most recent ideas in the field of anticancer metallodrugs with several European colleagues.

References 1. Johnson, N.P.; Butour, J.-L.; Villani, G.; Wimmer, F.L.; Defais, M.; Pierson, V.; Brabec, V.; Metal antitumor compounds: The mechanism of action of platinum complexes; Prog. Clin. Biochem. Med., 1989, 10, 1–24. 2. Brabec, V.; Kasparkova, J.; DNA interactions of platinum anticancer drugs. Recent advances and mechanisms of action. In Metal Compounds in Cancer Chemotherapy, Perez-Martin, J.-M., Fuertes, M.A. and Alonso, C., Eds., Research Signpost, Trivandrum, Kerala (India), 2005, 187–218. 3. Brabec, V.; DNA modifications by antitumor platinum and ruthenium compounds: their recognition and repair; Prog. Nucleic Acid Res. Mol. Biol., 2002, 71, 1–68. 4. Siddik, Z.H.; Cisplatin: mode of cytotoxic action and molecular basis of resistance; Oncogene, 2003, 22, 7265–7279. 5. Brabec, V.; Kleinwächter, V.; Butour, J.L.; Johnson, N.P.; Biophysical studies of the modification of DNA by antitumour platinum coordination complexes; Biophys. Chem., 1990, 35, 129–141. 6. Jamieson, E.R.; Lippard, S.J.; Structure, recognition, and processing of cisplatin–DNA adducts; Chem. Rev., 1999, 99, 2467–2498. 7. Friedberg, E.C.; A brief history of the DNA repair field; Cell Res., 2008, 18, 3–7. 8. Watanabe, T.; Sullenger, B.A.; RNA repair: a novel approach to gene therapy; Adv. Drug Deliv. Rev., 2000, 44, 109–118. 9. Bellacosa, A.; Moss, E.G.; RNA repair: Damage control; Curr. Biol., 2003, 13, R482–R484. 10. Wang, L.K.; Lima, C.D.; Shuman, S.; Structure and mechanism of T4 polynucleotide kinase: an RNA repair enzyme; EMBO J., 2002, 21, 3873–3880. 11. Pelletier, H.; Sawaya, M.R.; Wolfle, W.; Wilson, S.H.; Kraut, J.; Crystal structures of human DNA polymerase b complexed with DNA: Implications for catalytic mechanism, processivity, and fidelity; Biochemistry, 1996, 35, 12742–12761. 12. Sharma, R.A.; Dianov, G.L.; Targeting base excision repair to improve cancer therapies; Mol. Aspects Med., 2007, 28, 345–374. 13. Hutsell, S.Q.; Sancar, A.; Nucleotide excision repair, oxidative damage, DNA sequence polymorphisms, and cancer treatment; Clin. Cancer Res., 2005, 11, 1355–1357. 14. Kulaksiz, G.; Reardon, J.T.; Sancar, A.; Xeroderma pigmentosum complementation group E protein (XPE/DDB2): Purification of various complexes of XPE and analyses of their damaged DNA binding and putative DNA repair properties; Mol. Cell. Biol., 2005, 25, 9784–9792. 15. Keeney, S.; Eker, A.P.M.; Brody, T.; Vermeulen, W.; Bootsma, D.; Hoeijmakers, J.H.J.; Linn, S.; Correction of the DNA repair defect in xeroderma pigmentosum group E by

202

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

DNA Repair in Antitumour Effects of Platinum Drugs injection of a DNA damage-binding protein. Proc. Natl. Acad. Sci. USA; 1994, 91, 4053–4056. Itoh, T.; Linn, S.; Ono, T.; Yamaizumi, M.; Reinvestigation of the classification of five cell strains of xeroderma pigmentosum group E with reclassification of three of them; J. Investig. Dermatol., 2000, 114, 1022–1029. Reardon, J.T.; Sancar, A.; Recognition and repair of the cyclobutane thymine dimer, a major cause of skin cancers, by the human excision nuclease; Genes Dev., 2003, 17, 2539–2551. Reardon, J.T.; Sancar, A.; Nucleotide excision repair; Prog. Nucleic Acid Res. Mol. Biol., 2005, 79, 183–235. Jiricny, J.; The multifaceted mismatch-repair system; Nature Rev. Mol. Cell Biol., 2006, 7, 335–346. Iyer, R.R.; Pluciennik, A.; Burdett, V.; Modrich, P.L.; DNA mismatch repair: Functions and mechanisms; Chem. Rev., 2006, 106, 302–323. Pani, E.; Stojic, L.; El-Shemerly, M.; Jiricny, J.; Ferrari, S.; Mismatch repair status and the response of human cells to cisplatin; Cell Cycle, 2007, 6, 1796–1802. Stojic, L.; Brun, R.; Jiricny, J.; Mismatch repair and DNA damage signalling; DNA Repair, 2004, 3, 1091–1101. Dzantiev, L.; Constantin, N.; Genschel, J.; Iyer, R.R.; Burgers, P.M.; Modrich, P.; Defined human system that supports bidirectional mismatch-provoked excision; Mol. Cell, 2004, 15, 31–41. Genschel, J.; Modrich, P.; Mechanism of 5′-directed excision in human mismatch repair; Mol. Cell, 2003, 12, 1077–1086. Budman, J.; Chu, G.; Processing of DNA for nonhomologous end-joining by cell-free extract; EMBO J., 2005, 23, 849–860. Boeckman, H.J.; Trego, K.S.; Turchi, J.J.; Cisplatin sensitizes cancer cells to ionizing radiation via inhibition of nonhomologous end joining; Mol. Cancer Res., 2005, 3, 277–285. Diggle, C.P.; Bentley, J.; Knowles, M.A.; Kiltie, A.E.; Inhibition of double-strand break non-homologous end-joining by cisplatin adducts in human cell extracts; Nucleic Acids Res., 2005, 33, 2531–2539. Pastwa, E.; Malinowski, M.; Non-homologous DNA end joining in anticancer therapy; Curr. Cancer Drug Targets, 2007, 7, 243–250. Downs, J.A.; Jackson, S.P.; A means to a DNA end: The many roles of Ku; Nature Rev. Mol. Cell Biol., 2004, 5, 367–378. Collis, S.J.; DeWeese, T.L.; Jeggo, P.A.; Parker, A.R.; The life and death of DNA-PK; Oncogene, 2005, 24, 949–961. Drouet, J.; Frit, P.; Delteil, C.; de Villartay, J.P.; Salles, B.; Calsou, P.; Interplay between Ku, artemis, and the DNA-dependent protein kinase catalytic subunit at DNA ends; J. Biol. Chem., 2006, 281, 27784–27793. Goodarzi, A.A.; Yu, Y.P.; Riballo, E.; Douglas, P.; Walker, S.A.; Ye, R.Q.; Harer, C.; Marchetti, C.; Morrice, N.; Jeggo, P.A.; Lees-Miller, S.P.; DNA-PK autophosphorylation facilitates Artemis endonuclease activity; EMBO J., 2006, 25, 3880–3889. Grawunder, U.; Zimmer, D.; Fugmann, S.; Schwarz, K.; Lieber, M.R.; DNA ligase IV is essential for V(D)J recombination and DNA double-strand break repair in human precursor lymphocytes; Mol. Cell, 1998, 2, 477–484. Ahnesorg, P.; Smith, P.; Jackson, S.P.; XLF interacts with the XRCC4-DNA ligase IV complex to promote DNA nonhomologous end-joining; Cell, 2006, 124, 301–313. Daley, J.M.; Vander Laan, R.L.; Suresh, A.; Wilson, T.E.; DNA joint dependence of Pol X family polymerase action in nonhomologous end joining; J. Biol. Chem., 2005, 280, 29030–29037. Xu, Z.-Y., Loignon, M.; Han, F.-Y., Panasci, L.; Aloyz, R.; Xrcc3 induces cisplatin resistance by stimulation of Rad51-related recombinational repair, S-phase checkpoint activation, and reduced apoptosis; J. Pharmacol. Exp. Ther., 2005, 314, 495–505. West, S.C.; Molecular views of recombination proteins and their control; Nature Rev. Mol. Cell Biol., 2003, 4, 435–445. Yang, W.; Woodgate, R.; What a difference a decade makes: Insights into translesion DNA synthesis; Proc. Natl. Acad. Sci. USA, 2007, 104, 15591–15598.

References

203

39. Fichtinger-Schepman, A.M.J.; Van der Veer, J.L.; Den Hartog, J.H.J.; Lohman, P.H.M.; Reedijk, J.; Adducts of the antitumor drug cis-diamminedichloroplatinum(II) with DNA: Formation, identification, and quantitation; Biochemistry, 1985, 24, 707–713. 40. Eastman, A.; The formation, isolation and characterization of DNA adducts produced by anticancer platinum complexes; Pharmacol. Ther., 1987, 34, 155–166. 41. Lemaire, M.A.; Schwartz, A.; Rahmouni, A.R.; Leng, M.; Interstrand cross-links are preferentially formed at the d(GC) sites in the reaction between cis-diamminedichloro platinum(II) and DNA; Proc. Natl. Acad. Sci. USA, 1991, 88, 1982–1985. 42. Gelasco, A.; Lippard, S.J.; Anticancer activity of cisplatin and related complexes. In Metallopharmaceuticals I. DNA Interactions, Clarke, M.J. and Sadler, P.J.; Eds., Springer, Berlin, 1999, 1–43. 43. Brabec, V.; Chemistry and structural biology of 1,2-interstrand adducts of cisplatin. In Platinum-based Drugs in Cancer Therapy, Kelland, L.R. and Farrell, N.P.; Eds., Humana Press Inc., Totowa, NJ, 2000, 37–61. 44. Teuben, J.M.; Bauer, C.; Wang, A.H.J.; Reedijk, J.; Solution structure of a DNA duplex containing a cis-diammineplatinum(II) 1,3-d(GTG) intrastrand cross-link, a major adduct in cells treated with the anticancer drug carboplatin; Biochemistry, 1999, 38, 12305–12312. 45. Brabec, V.; Reedijk, J.; Leng, M.; Sequence-dependent distortions induced in DNA by monofunctional platinum(II) binding; Biochemistry, 1992, 31, 12397–12402. 46. Brabec, V.; Boudny, V.; Balcarova, Z.; Monofunctional adducts of platinum(II) produce in DNA a sequence-dependent local denaturation; Biochemistry, 1994, 33, 1316–1322. 47. Blommaert, F.A.; van Dijk-Knijnenburg, H.C.M.; Dijt, F.J.; Denengelse, L.; Baan, R.A.; Berends, F.; Fichtinger-Schepman, A.M.J.; Formation of DNA adducts by the anticancer drug carboplatin: Different nucleotide sequence preferences in vitro and in cells; Biochemistry, 1995, 34, 8474–8480. 48. Woynarowski, J.M.; Chapman, W.G.; Napier, C.; Herzig, M.C.S.; Juniewicz, P.; Sequenceand region-specificity of oxaliplatin adducts in naked and cellular DNA; Mol. Pharmacol., 1998, 54, 770–777. 49. Woynarowski, J.M.; Faivre, S.; Herzig, M.C.S.; Arnett, B.; Chapman, W.G.; Trevino, A.V.; Raymond, E.; Chaney, S.G.; Vaisman, A.; Varchenko, M.; Juniewicz, P.E.; Oxaliplatininduced damage of cellular DNA; Mol. Pharmacol., 2000, 58, 920–927. 50. Zlatanova, J.; Van Holde, K.; Binding to four-way junction DNA: A common property of architectural proteins?; FASEB J., 1998, 12, 421–431. 51. Missura, M.; Buterin, T.; Hindges, R.; Hubscher, U.; Kasparkova, J.; Brabec, V.; Naegeli, H.; Double-check probing of DNA bending and unwinding by XPA-RPA: an architectural function in DNA repair; EMBO J.; 2001, 20, 3554–3564. 52. Kasparkova, J.; Mellish, K.J.; Qu, Y.; Brabec, V.; Farrell, N.; Site-specific d(GpG) intrastrand cross-links formed by dinuclear platinum complexes. Bending and NMR studies; Biochemistry, 1996, 35, 16705–16713. 53. Patrick, S.M.; Turchi, J.J.; Replication protein A (RPA) binding to duplex cisplatindamaged DNA is mediated through the generation of single-stranded DNA; J. Biol. Chem., 1999, 274, 14972–14978. 54. Hofr, C.; Farrell, N.; Brabec, V.; Thermodynamic properties of duplex DNA containing a site-specific d(GpG) intrastrand crosslink formed by an antitumor dinuclear platinum complex; Nucleic Acids Res., 2001, 29, 2034–2040. 55. Lao, Y.; Gomes, X.V.; Ren, Y.J.; Taylor, J.S.; Wold, M.S.; Replication protein A interactions with DNA. III. Molecular basis of recognition of damaged DNA; Biochemistry, 2000, 39, 850–859. 56. Hermanson-Miller, I.L.; Turchi, J.J.; Strand-specific binding of RPA and XPA to damaged duplex DNA; Biochemistry, 2002, 41, 2402–2408. 57. Patrick, S.M.; Turchi, J.J.; Xeroderma pigmentosum complementation group A protein (XPA) modulates RPA-DNA interactions via enhanced complex stability and inhibition of strand separation activity; J. Biol. Chem., 2002, 277, 16096–16101. 58. Payne, A.; Chu, G.; Xeroderma pigmentosum group e binding factor recognizes a broad spectrum of DNA damage; Mutation Res., 1994, 310, 89–102.

204


59. Brabec, V.; Stehlikova, K.; Malina, J.; Vojtiskova, M.; Kasparkova, J.; Thermodynamic properties of damaged DNA and its recognition by xeroderma pigmentosum group A protein and replication protein A; Arch. Biochem. Biophys., 2006, 446, 1–10. 60. Hey, T.; Lipps, G.; Sugasawa, K.; Iwai, S.; Hanaoka, F.; Krauss, G.; The XPC-HR23B complex displays high affinity and specificity for damaged DNA in a true-equilibrium fluorescence assay; Biochemistry, 2002, 41, 6583–6587. 61. Zamble, D.B.; Mu, D.; Reardon, J.T.; Sancar, A.; Lippard, S.J.; Repair of cisplatin-DNA adducts by the mammalian excision nuclease; Biochemistry, 1996, 35, 10004–10013. 62. Kasparkova, J.; Zehnulova, J.; Farrell, N.; Brabec, V.; DNA interstrand cross-links of the novel antitumor trinuclear platinum complex BBR3464. Conformation, recognition by high mobility group domain proteins, and nucleotide excision repair; J. Biol. Chem., 2002, 277, 48076–48086. 63. Zehnulova, J.; Kasparkova, J.; Farrell, N.; Brabec, V.; Conformation, recognition by high mobility group domain proteins, and nucleotide excision repair of DNA intrastrand cross-links of novel antitumor trinuclear platinum complex BBR3464; J. Biol. Chem., 2001, 276, 22191–22199. 64. Tornaletti, S.; Patrick, S.M.; Turchi, J.J.; Hanawalt, P.C.; Behavior of T7 RNA polymerase and mammalian RNA polymerase II at site-specific cisplatin adducts in the template DNA; J. Biol. Chem., 2003, 278, 35791–35797. 65. Laine, J.P.; Egly, J.M.; Initiation of DNA repair mediated by a stalled RNA polymerase IIO; EMBO J., 2006, 25, 387–397. 66. Jung, Y.; Lippard, S.J.; RNA polymerase II blockage by cisplatin-damaged DNA – Stability and polyubiquitylation of stalled polymerase; J. Biol. Chem., 2006, 281, 1361–1370. 67. Damsma, G.E.; Alt, A.; Brueckner, F.; Carell, T.; Cramer, P.; Mechanism of transcriptional stalling at cisplatin-damaged DNA; Nature Struct. Mol. Biol., 2007, 14, 1127–1133. 68. Flohr, C.; Burkle, A.; Radicella, J.P.; Epe, B.; Poly(ADP-ribosyl)ation accelerates DNA repair in a pathway dependent on Cockayne syndrome B protein; Nucleic Acids Res., 2003, 31, 5332–5337. 69. Zhang, C.X.; Chang, P.V.; Lippard, S.J.; Identification of nuclear proteins that interact with platinum-modified DNA by photoaffinity labeling; J. Am. Chem. Soc., 2004, 126, 6536–6537. 70. Mello, J.A.; Acharya, S.; Fishel, R.; Essigmann, J.M.; The mismatch-repair protein hMSH2 binds selectively to DNA adducts of the anticancer drug cisplatin; Chem. Biol., 1996, 3, 579–589. 71. Vaisman, A.; Varchenko, M.; Umar, A.; Kunkel, T.A.; Risinger, J.I.; Barrett, J.C.; Hamilton, T.C.; Chaney, S.G.; The role of hMLH1, hMSH3, and hMSH6 defects in cisplatin and oxaliplatin resistance: Correlation with replicative bypass of platinum-DNA adducts; Cancer Res., 1998, 58, 3579–3585. 72. Duckett, D.R.; Drummond, J.T.; Murchie, A.I.H.; Reardon, J.T.; Sancar, A.; Lilley, D.M.; Modrich, P.; Human MutSa recognizes damaged DNA base pairs containing O6methylguanine, O4-methylthymine, or the cisplatin-d(GpG) adduct; Proc. Natl. Acad. Sci. USA, 1996, 93, 6443–6447. 73. Mu, D.; Tursun, M.; Duckett, D.R.; Drummond, J.T.; Modrich, P.; Sancar, A.; Recognition and repair of compound DNA lesions (base damage and mismatch) by human mismatch repair and excision repair systems; Mol. Cell. Biol., 1997, 17, 760–769. 74. Yamada, M.; O’Regan, E.; Brown, R.; Karran, P.; Selective recognition of a cisplatinDNA adduct by human mismatch repair proteins; Nucleic Acids Res., 1997, 25, 491–495. 75. Turchi, J.J.; Henkels, K.; Human Ku autoantigen binds cisplatin-damaged DNA but fails to stimulate human DNA-activated protein kinase; J. Biol. Chem., 1996, 271, 13861–13867. 76. Turchi, J.J.; Patrick, S.M.; Henkels, K.M.; Mechanism of DNA-dependent protein kinase inhibition by cis-diamminedichloroplatinum(II)-damaged DNA; Biochemistry, 1997, 36, 7586–7593.

References

205

77. Kysela, B.; Doherty, A.J.; Chovanec, M.; Stiff, T.; Ameer-Beg, S.M.; Vojnovic, B.; Girard, P.-M.; Jeggo, P.A.; Ku stimulation of DNA ligase IV-dependent ligation requires inward movement along the DNA molecule; J. Biol. Chem., 2003, 278, 22466– 22474. 78. Turchi, J.J.; Henkels, K.M.; Zhou, Y.; Cisplatin-DNA adducts inhibit translocation of the Ku subunits of DNA-PK; Nucleic Acids Res., 2000, 28, 4634–4641. 79. Noll, D.M.; Mason, T.M.; Miller, P.S.; Formation and repair of interstrand cross-links in DNA; Chem. Rev., 2006, 106, 277–301. 80. Vaisman, A.; Lim, S.E.; Patrick, S.M.; Copeland, W.C.; Hinkle, D.C.; Turchi, J.J.; Chaney, S.G.; Effect of DNA polymerases and high mobility group protein 1 on the carrier ligand specificity for translesion synthesis past platinum-DNA adducts; Biochemistry, 1999, 38, 11026–11039. 81. Bassett, E.; Vaisman, A.; Havener, J.M.; Masutani, C.; Hanaoka, F.; Chaney, S.G.; Efficiency of extension of mismatched primer termini across from cisplatin and oxaliplatin adducts by human DNA polymerases beta and eta in vitro; Biochemistry, 2003, 42, 14197–14206. 82. Vaisman, A.; Chaney, S.G.; The efficiency and fidelity of translesion synthesis past cisplatin and oxaliplatin GpG adducts by human DNA polymerase beta; J. Biol. Chem., 2000, 275, 13017–13025. 83. Vaisman, A.; Masutani, C.; Hanaoka, F.; Chaney, S.G.; Efficient translesion replication past oxaliplatin and cisplatin GpG adducts by human DNA polymerase h; Biochemistry, 2000, 39, 4575–4580. 84. Havener, J.M.; McElhinny, S.A.N.; Bassett, E.; Gauger, M.; Ramsden, D.A.; Chaney, S.G.; Translesion synthesis past platinum DNA adducts by human DNA polymerase mu; Biochemistry, 2003, 42, 1777–1788. 85. Bassett, E.; King, N.M.; Bryant, M.F.; Hector, S.; Pendyala, L.; Chaney, S.G.; CordeiroStone, M.; The role of DNA polymerase eta in translesion synthesis past platinum-DNA adducts in human fibroblasts; Cancer Res., 2004, 64, 6469–6475. 86. Daube, S.S.; Tomer, G.; Livneh, Z.; Translesion replication by DNA polymerase delta depends on processivity accessory proteins and differs in specificity from DNA polymerase beta; Biochemistry, 2000, 39, 348–355. 87. Prakash, S.; Prakash, L.; Translesion DNA synthesis in eukaryotes: A one- or twopolymerase affair; Gene Dev., 2002, 16, 1872–1883. 88. Wu, F.; Lin, X.J.; Okuda, T.; Howell, S.B.; DNA polymerase zeta regulates cisplatin cytotoxicity, mutagenicity, and the rate of development of cisplatin resistance; Cancer Res., 2004, 64, 8029–8035. 89. Prakash, S.; Johnson, R.E.; Prakash, L.; Eukaryotic translesion synthesis DNA polymerases: Specificity of structure and function; Annu. Rev. Biochem., 2005, 74, 317–353. 90. Dronkert, M.L.G.; Kanaar, R.; Repair of DNA interstrand cross-links; Mutation Res., 2001, 486, 217–247. 91. Chaney, S.G.; Campbell, S.L.; Bassett, E.; Wu, Y.; Recognition and processing of cisplatinand oxaliplatin-DNA adducts; Crit. Rev. Oncol. Hematol., 2005, 53, 3–11. 92. Cao, S.; Bendall, H.; Hicks, G.G.; Nashabi, A.; Sakano, H.; Shinkai, Y.; Gariglio, M.; Oltz, E.M.; Ruley, H.E.; The high-mobility-group box protein SSRP1/T160 is essential for cell viability in day 3.5 mouse embryos; Mol. Cell. Biol., 2003, 23, 5301–5307. 93. Krohn, N.M.; Stemmer, C.; Fojan, P.; Grimm, R.; Grasser, K.D.; Protein kinase CK2 phosphorylates the high mobility group domain protein SSRP1, inducing the recognition of UV-damaged DNA; J. Biol. Chem., 2003, 278, 12710–12715. 94. Toney, J.H.; Donahue, B.A.; Kellett, P.J.; Bruhn, S.L.; Essigmann, J.M.; Lippard, S.J.; Isolation of cDNAs encoding a human protein that binds selectively to DNA modified by the anticancer drug cis-diamminedichloroplatinum(II); Proc. Nat. Acad. Sci. USA, 1989, 89, 8328–8332. 95. Yarnell, A.T.; Oh, S.; Reinberg, D.; Lippard, S.J.; Interaction of FACT, SSRP1, and the high mobility group (HMG) domain of SSRP1 with DNA damaged by the anticancer drug cisplatin; J. Biol. Chem., 2001, 276, 25736–25741.

206


96. Kartalou, M.; Samson, L.D.; Essigmann, J.M.; Cisplatin adducts inhibit 1,N-6-ethenoadenine repair by interacting with the human 3-methyladenine DNA glycosylase; Biochemistry, 2000, 39, 8032–8038. 97. Kartalou, M.; Essigmann, J.M.; Recognition of cisplatin adducts by cellular proteins; Mutation Res., 2001, 478, 1–21. 98. Wozniak, K.; Blasiak, J.; Recognition and repair of DNA-cisplatin adducts; Acta Biochim. Pol., 2002, 49, 583–596. 99. Bellon, S.F.; Coleman, J.H.; Lippard, S.J.; DNA unwinding produced by site-specific intrastrand crosslinks of the antitumor drug cis-diamminedichloroplatinum(II); Biochemistry, 1991, 30, 8026–8035. 100. Bellon, S.F.; Lippard, S.J.; Bending studies of DNA site-specifically modified by cisplatin, trans-diamminedichloroplatinum(II) and cis-Pt(NH3)2(N3-cytosine)Cl+; Biophys. Chem., 1990, 35, 179–188. 101. Huang, J.C.; Zamble, D.B.; Reardon, J.T.; Lippard, S.J.; Sancar, A.; HMG-domain proteins specifically inhibit the repair of the major DNA adduct of the anticancer drug cisplatin by human excision nuclease; Proc. Natl. Acad. Sci. USA, 1994, 91, 10394–10398. 102. Plooy, A.C.M.; Fichtinger-Schepman, A.M.J.; Schutte, H.H.; van Dijk, M.; Lohman, P. H.M.; The quantitative detection of various Pt-DNA-adducts in chinese hamster ovary cells treated with cisplatin; application of immunochemical techniques; Carcinogenesis, 1985, 6, 561–566. 103. Dijt, F.J.; Fichtinger-Schepman, A.M.J.; Berends, F.; Reedijk, J.; Formation and repair of cisplatin-induced adducts to DNA in cultured normal and repair-deficient human fibroblasts; Cancer Res., 1988, 48, 6058–6062. 104. Fujiwara, Y.; Tatsumi, M.; Sasaki, M.S.; Cross-link repair in human cells and its possible defect in Fanconi’s anemia cells; J. Mol. Biol., 1977, 113, 635–649. 105. Calsou, P.; Frit, P.; Salles, B.; Repair synthesis by human cell extracts in cisplatin-damaged DNA is prefentially determined by minor adducts; Nucleic Acids Res., 1992, 20, 6363–6368. 106. Fisher, L.A.; Bessho, M.; Bessho, T.; Processing of a psoralen DNA interstrand cross-link by XPF-ERCC1 complex in vitro; J. Biol. Chem., 2008, 283, 1275–1281. 107. Chvalova, K.; Brabec, V.; Kasparkova, J.; Mechanism of the formation of DNA-protein cross-links by antitumor cisplatin; Nucleic Acids Res., 2007, 35, 1812–1821. 108. Olinski, R.; Wedrychovski, A.; Schmidt, W.N.; Briggs, R.C.; Hnilica, L.S.; In vivo DNAprotein cross-linking by cis- and trans-diamminedichloroplatinum(II); Cancer Res., 1987, 47, 201–205. 109. Plooy, A.C.M.; Van Dijk, M.; Lohman, P.H.M.; Induction and repair of DNA cross-links in Chinese hamster ovary cells treated with various platinum coordination compounds in relation to platinum binding to DNA, cytotoxicity, mutagenicity, and antitumor activity; Cancer Res., 1984, 44, 2043–2051. 110. Olinski, R.; Briggs, R.C.; DNA-protein cross-linking in L1210 cells and resistant to cisdiamminedichloroplatinum(II); Mol. Biol. Rep., 1991, 15, 81–86. 111. Richmond, T.J.; Davey, C.A.; The structure of DNA in the nucleosome core; Nature, 2003, 423, 145–150. 112. Wang, D.; Hara, R.; Singh, G.; Sancar, A.; Lippard, S.J.; Nucleotide excision repair from site-specifically platinum-modified nucleosomes; Biochemistry, 2003, 42, 6747–6753. 113. Reardon, J.T.; Vaisman, A.; Chaney, S.G.; Sancar, A.; Efficient nucleotide excision repair of cisplatin, oxaliplatin, and bis-aceto-ammine-dichloro-cyclohexylamine-platinum(IV) (JM216) platinum intrastrand DNA diadducts; Cancer Res., 1999, 59, 3968–3971. 114. Hartwig, J.F.; Lippard, S.J.; DNA binding properties of cis-Pt(NH3)(C6H11NH2)Cl2, a metabolite of an orally active platinum anticancer drug; J. Am. Chem. Soc., 1992, 114, 5646–5654. 115. Malina, J.; Novakova, O.; Vojtiskova, M.; Natile, G.; Brabec, V.; Conformation of DNA GG intrastrand cross-link of antitumor oxaliplatin and its enantiomeric analog; Biophys. J., 2007, 93, 3950–3962. 116. Malina, J.; Hofr, C.; Maresca, L.; Natile, G.; Brabec, V.; DNA interactions of antitumor cisplatin analogs containing enantiomeric amine ligands; Biophys. J., 2000, 78, 2008–2021.

References

207

117. Malina, J.; Kasparkova, J.; Natile, G.; Brabec, V.; Recognition of major DNA adducts of enantiomeric cisplatin analogs by HMG box proteins and nucleotide excision repair of these adducts; Chem. Biol., 2002, 9, 629–638. 118. Delalande, O.; Malina, J.; Brabec, V.; Kozelka, J.; Chiral differentiation of DNA adducts formed by enantiomeric analogues of antitumor cisplatin is sequence-dependent; Biophys. J., 2005, 88, 4159–4169. 119. Reedijk, J.; Improved understanding in platinum antitumour chemistry; Chem. Commun., 1996, 801–806. 120. Farrell, N.; Current status of structure-activity relationships of platinum anticancer drugs: Activation of the trans geometry. In Metal Ions in Biological Systems, Vol. 32, Sigel, A. and Sigel, H.; Eds., Marcel Dekker, Inc., New York, Basel, Hong Kong, 1996, 603–639. 121. Coluccia, M.; Natile, G.; Trans-platinum complexes in cancer therapy; Anti-Cancer Agents Med. Chem., 2007, 7, 111–123. 122. Kasparkova, J.; Novakova, O.; Marini, V.; Najajreh, Y.; Gibson, D.; Perez, J.-M.; Brabec, V.; Activation of trans geometry in bifunctional mononuclear platinum complexes by a piperidine ligand: Mechanistic studies on antitumor action; J. Biol. Chem., 2003, 278, 47516–47525. 123. Marini, V.; Christofis, P.; Novakova, O.; Kasparkova, J.; Farrell, N.; Brabec, V.; Conformation, protein recognition and repair of DNA interstrand and intrastrand cross-links of antitumor trans-[PtCl2(NH3)(thiazole)]; Nucleic Acids Res., 2005, 33, 5819–5828. 124. Coluccia, M.; Nassii, F.; Loseto, F.; Boccarelli, A.; Marigió, M.A.; Giordano, D.; Intini, F.P.; Caputo, P.; Natile, G.; A trans-platinum complex showing higher antitumor activity than the cis congeners; J. Med. Chem., 1993, 36, 510–512. 125. Brabec, V.; Vrana, O.; Novakova, O.; Kleinwachter, V.; Intini, F.P.; Coluccia, M.; Natile, G.; DNA adducts of antitumor trans-[PtCl2(E-imino ether)2]; Nucleic Acids Res., 1996, 24, 336–341. 126. Novakova, O.; Kasparkova, J.; Malina, J.; Natile, G.; Brabec, V.; DNA-protein crosslinking by trans-[PtCl2(E-iminoether)2]. A concept for activation of the trans geometry in platinum antitumor complexes; Nucleic Acids Res., 2003, 31, 6450–6460. 127. Farrell, N.; Polynuclear platinum drugs. In Metal Ions in Biological Systems, Vol. 42, Sigel, A. and Sigel, H.; Eds., Marcel Dekker, Inc., New York, Basel, 2004, 251–296. 128. Perego, P.; Gatti, L.; Caserini, C.; Supino, R.; Colangelo, D.; Leone, R.; Spinelli, S.; Farrell, N.; Zunino, F.; The cellular basis of the efficacy of the trinuclear platinum complex BBR 3464 against cisplatin-resistant cells; J. Inorg. Biochem., 1999, 77, 59–64. 129. Roberts, J.D.; Beggiolin, G.; Manzotti, C.; Piazzoni, L.; Farrell, N.; Comparison of cytotoxicity and cellular accumulation of polynuclear platinum complexes in L1210 murine leukemia cell lines; J. Inorg. Biochem., 1999, 77, 47–50. 130. Farrell, N.; Polynuclear charged platinum compounds as a new class of anticancer agents: Toward a new paradigm. In Platinum-based Drugs in Cancer Therapy, Kelland, L.R. and Farrell, N.P.; Eds., Humana Press Inc, Totowa, NJ, 2000, 321–338. 131. Brabec, V.; Kasparkova, J.; Vrana, O.; Novakova, O.; Cox, J.W.; Qu, Y.; Farrell, N.; DNA modifications by a novel bifunctional trinuclear platinum Phase I anticancer agent; Biochemistry, 1999, 38, 6781–6790. 132. Fink, D.; Nebel, S.; Aebi, S.; Zheng, H.; Cenni, B.; Nehme, A.; Christen, R.D.; Howell, S.B.; The role of DNA mismatch repair in platinum drug resistance; Cancer Res., 1996, 56, 4881–4886. 133. Aebi, S.; Fink, D.; Gordon, R.; Kim, H.K.; Zheng, H.; Fink, J.L.; Howell, S.B.; Resistance to cytotoxic drugs in DNA mismatch repair-deficient cells; Clin. Cancer Res., 1997, 3, 1763–1767. 134. Johnson, S.W.; Ferry, K.V.; Hamilton, T.C.; Recent insights into platinum drug resistance in cancer; Drug Resist. Updates, 1998, 1, 243–254. 135. Drummond, J.T.; Anthoney, A.; Brown, R.; Modrich, P.; Cisplatin and adriamycin resistance are associated with MutL alpha and mismatch repair deficiency in an ovarian tumor cell line; J. Biol. Chem., 1996, 271, 19645–19648.

208


136. Fink, D.; Nebel, S.; Aebi, S.; Nehme, A.; Howell, S.B.; Loss of DNA mismatch repair due to knockout of MSH2 or PMS2 results in resistance to cisplatin and carboplatin; Int. J. Oncol., 1997, 11, 539–542. 137. Ferry, K.V.; Fink, D.; Johnson, S.W.; Nebel, S.; Hamilton, T.C.; Howell, S.B.; Decreased cisplatin damage-dependent DNA synthesis in cellular extracts of mismatch repair deficient cells; Biochem. Pharmacol., 1999, 57, 861–867. 138. Calsou, P.; Frit, P.; Salles, B.; Double strand breaks in DNA inhibit nucleotide excision repair in vitro; J. Biol. Chem., 1996, 271, 27601–27607. 139. Frit, P.; Calsou, P.; Chen, D.J.; Salles, B.; Ku70/Ku80 protein complex inhibits the binding of nucleotide excision repair proteins on linear DNA in vitro; J. Mol. Biol., 1998, 284, 963–973. 140. Frit, P.; Canitrot, Y.; Muller, C.; Foray, N.; Calsou, P.; Marangoni, E.; Bourhis, J.; Salles, B.; Cross-resistance to ionizing radiation in a murine leukemic cell line resistant to cis-dic hlorodiammineplatinum(II): Role of Ku autoantigen; Mol. Pharmacol., 1999, 56, 141–146. 141. Bhattacharyya, A.; Ear, U.S.; Koller, B.H.; Weichselbaum, R.R.; Bishop, D.K.; The breast cancer susceptibility gene BRCA1 is required for subnuclear assembly of Rad51 and survival following treatment with the DNA cross-linking agent cisplatin; J. Biol. Chem., 2000, 275, 23899–23903. 142. Caldecott, K.; Jeggo, P.; Cross-sensitivity of gamma-ray-sensitive hamster mutants to cross-linking agents; Mutation Res., 1991, 255, 111–121. 143. De Silva, I.U.; McHugh, P.J.; Clingen, P.H.; Hartley, J.A.; Defects in interstrand cross-link uncoupling do not account for the extreme sensitivity of ERCC1 and XPF cells to cisplatin; Nucleic Acids Res., 2002, 30, 3848–3856. 144. Cummings, M.; Higginbottom, K.; McGurk, C.J.; Wong, O.G.-W., Koberle, B.; Oliver, R.T.D.; Masters, J.R.; XPA versus ERCC1 as chemosensitising agents to cisplatin and mitomycin C in prostate cancer cells: Role of ERCC1 in homologous recombination repair; Biochem. Pharmacol., 2006, 72, 166–175. 145. Selvakumaran, M.; Pisarcik, D.A.; Bao, R.; Yeung, A.T.; Hamilton, T.C.; Enhanced cisplatin cytotoxicity by disturbing the nucleotide excision repair pathway in ovarian cancer cell lines; Cancer Res., 2003, 63, 1311–1316. 146. Brabec, V.; Kasparkova, J.; Modifications of DNA by platinum complexes: Relation to resistance of tumors to platinum antitumor drugs; Drug Resist. Updates, 2005, 8, 131–146. 147. Jung, Y.; Lippard, S.J.; Direct cellular responses to platinum-induced DNA damage; Chem. Rev., 2007, 107, 1387–1407. 148. Teletchéa, S.; Komeda, S.; Teuben, J.-M., Elizondo-Riojas, M.A.; Reedijk, J.; Kozelka, J.; A pyrazolato-bridged dinuclear platinum(II) complex induces only minor distortions upon DNA-binding; Chem. Eur. J., 2006, 12, 3741–3753. 149. Komeda, S.; Lutz, M.; Spek, A.L.; Yamanaka, Y.; Sato, T.; Chikuma, M.; Reedijk, J.; A novel isomerization on interaction of antitumor-active azole-bridged dinuclear platinum(II) complexes with 9-ethylguanine. Platinum(II) atom migration from N2 to N3 on 1,2,3-triazole; J. Am. Chem. Soc., 2002, 124, 4738–4746. 150. Komeda, S.; Kalayda, G.V.; Lutz, M.; Spek, A.L.; Yamanaka, Y.; Sato, T.; Chikuma, M.; Reedijk, J.; New isomeric azine-bridged dinuclear platinum(II) complexes circumvent cross-resistance to cisplatin; J. Med. Chem., 2003, 46, 1210–1219. 151. Komeda, S.; Bombard, S.; Perrier, S.; Reedijk, J.; Kozelka, J.F.; Kinetic study of azolebridged dinuclear platinum(II) complexes reacting with a hairpin-stabilized doublestranded oligonucleotide; J. Inorg. Biochem., 2003, 96, 357–366. 152. Spiegel, K.; Magistrato, A.; Carloni, P.; Reedijk, J.; Klein, M.L.; Azole-bridged diplatinum anticancer compounds. Modulating DNA flexibility to escape repair mechanism and avoid cross resistance; J. Phys. Chem. B, 2007, 111, 11873–11876. 153. Madhusudan, S.; Hickson, I.D.; DNA repair inhibition: a selective tumour targeting strategy; Trends Mol. Med., 2005, 11, 503–511. 154. Hubscher, U.; Maga, G.; Spadari, S.; Eukaryotic DNA polymerases; Annu. Rev. Biochem., 2002, 71, 133–163.

7 Telomeres and Telomerase: Potential Targets for Platinum Complexes Isabelle Ourliac-Garnier, Razan Charif and Sophie Bombard

7.1 Function of Telomeres Telomeres are nucleoprotein structures that protect the very end of linear chromosomes. Telomeric DNA consists of tandemly repeated G-rich sequences of DNA, which differ from one species to another, being (TTAGGG)n in humans. These sequences of about 4–10 kbp protrude at the 3′ extremity by a single strand of the G-rich sequence of about 50 to 200 bases. Firstly, telomeres preserve the stability and the structural integrity of chromosomes;1–4 indeed, telomere dysfunction leads to chromosome instability and abnormalities, senescence or apoptosis, and represents potentially the most widespread cause of genome instability in cancer.5 So, concretely, telomeres avoid end-to-end fusion of chromosomes by preventing the chromosomes ends from being recognised as DNA double-strand breaks (DSB) and consequently from being repaired.6 Secondly, telomeres preserve the genome integrity by protecting chromosomes from a loss of genetic information due to the ‘end replication problem’: a progressive erosion of the 5′ extremity of each chromosome that happens at each round of replication.7–10 Indeed, DNA replication involves RNA primers, named Okazaki fragments, which are necessary for DNA synthesis from the 5′ to the 3′ extremities. The last replication fork is supposed to be subtelomeric, consequently, telomeric DNA is replicated unidirectionally and the G-rich strand is synthesized discontinuously, whereas the C-rich one is replicated continuously (Figure 7.1). Then, conventional DNA polymerases replace Okazaki Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

210

Telomeres and Telomerase 3'

G C

5'

S-phase

G C

3' 5'

Fork progression

exonuclease 3'

3' 5'

5'

Figure 7.1 End replication of telomeres and processing. The G-rich strand is replicated discontinuously, whereas the C-rich one is replicated continuously. The RNA Okazaki fragments (in dark grey) at the 5′ end of the daughter DNA (dotted lines) is not replaced by conventional DNA polymerases, the parental G-rich strand is consequently not completely replicated, leading to the G-overhang formation. Moreover, the 5′ end of the C-rich strand of telomeres (daughter or parental strand) can also be digested by a 5′ to 3′ nuclease which contributes to the formation of the G-overhang

fragments by DNA, except at the 5′ end of the daughter DNA, which becomes shorter. Consequently, the parental G-rich strand is not completely duplicated leading to the formation of a G-rich overhang. Moreover, the telomere shortening mechanism is also due to a particular processing of the C-rich strand.11,12 Particularly, it has been proposed that Apollo, a protein involved in DNA repair and colocalised with telomeric proteins, contributes to the resection that occurs at the 5′ C-rich strand thanks to its 5′-exonuclease activity.13 Consequently, in normal cells, telomeres erode progressively at each round of replication and can shorten until a length limit (Hayflick limit) that induces a signal in the cell, leading to senescence.14 Telomeres are therefore considered as a mitotic clock that allows the cell to divide a limited number of times before entering senescence.15,16

7.2 Structure of Telomeres Telomeres have a well-organised structure comprising telomeric DNA and a package of specific proteins that are called the telosome17 or the shelterin18 and cap the chromosome extremity. The shelterin is composed of six proteins: TRF1, TRF2, POT1, TIN2, TPP1 and Rap1. Three of these are specifically and directly bound to

Structure of Telomeres TPP1

AT

POT1 TPP1

ERCC1/ XPC

MRE 11 Rad 50 Nbs1

Ku ATM kinase

POT1 POT1

TRF2

TIN2

Ku

TPP1

WRN

N2

ATM kinase

MRE 11 Rad 50 Nbs1

PI

N

X1

AT Mk

ina se

Ku

TA 1 e NK t2

TR F1

BL M

PI N X1

PARP2 WRN

Ku

F1

ATM kinase

WRN

TI

TRF1

M

TANK1 et2

TIN2

ERCC1/ XPC

TRF2

hRAP1

e

TR

hRAP1 PINX1

BL

s na ki

TIN2 BLM

hRAP1 TIN2 PARP2

PARP2

M

NK 2 TA et 1

Ku

TIN2 TPP1

211

TRF2

ATM kinase

Ku

ERCC1/ XPC MRE 11 Rad 50 Nbs1

Figure 7.2 The shelterin complex and associated proteins. T-loop formation has been proposed as a structure capping the end of telomeres. It consists of the invasion of the 3′ Goverhang into the duplex telomeric DNA. Associated telomeric proteins are involved in this complex: TRF1, TRF2 and POT1 bind directly to telomeric DNA, whereas TPP1, TIN2 and RAP1 interact with them. Other proteins that are not specific for telomeres have also been associated with this complex. They include Apollo, which appears to protect chromosome ends from being processed as DNA damage, and DNA damage proteins: PINX1, Ku 70/80, TANK1/2, ATM-kinase, PARP2, ERCC1/XPC, MRE11, Rad 50, Rbs1

telomeric DNA: TRF1 and TRF2 bind to the double-stranded structure through their Myb-type DNA binding domain19–21 and POT1 binds to the G-rich 3′ overhang through two oligonucleotide/oligosaccharide binding domains.22 The three other proteins allow the connection between the double-stranded telomeric DNA and its 3′ extremity: TIN2 binds to TRF1 and TRF2; TPP1 binds to TIN2 and POT1; Rap1 binds to TRF2.23 Two telomeric proteins, TRF2 and POT1 are directly involved in the maintenance of the structure of telomeres.24 In particular, as shown by an electron microscopy study, TRF2 is thought to promote the formation of the t-loop, which consists of an invasion of the G-rich overhang in the double-stranded telomeric DNA,25 and is supposed to protect the chromosome extremity from degradation26 (Figure 7.2). Moreover, it has been found that expression of a dominant-negative mutant allele of TRF2 induces: (i) recognition of telomeres as DSB, (ii) activation of the ATM/p53 DNA damage response pathway, (iii) end-to-end fusions of chromosomes and (iv) accelerated senescence entry.27–29 TRF2 is overexpressed in some cancer cell lines,30–32 and it has been shown that the inhibition of TRF2 can reduce tumourigenicity of melanoma cancer cell lines.33 This suggests that TRF2 is strongly involved in telomere capping in cancer cells and could be a good target for anticancer therapy. POT1 has been shown to act either as an activator or as a repressor of the enzyme responsible of telomere elongation, telomerase (see below), via its interaction with TRF1 and TPP1.34–38 Dysfunctional or uncapped telomeres, created through destabilization of the shelterin by TRF2 or POT1 inhibition, leads to the DNA damage response,3,35,39–42 loss of the 3′-G-overhang and aberrant homologous recombination.43,44 In conclusion, the protective function of telomeres depends essentially on the stability and integrity of the shelterin, which could be perturbed by modification of the amount of telomeric proteins bound to DNA.

212

Telomeres and Telomerase

7.3 Telomerase Cells escape limits of proliferation by maintaining a constant length of their telomeres. This is mainly due to telomerase, a telomere-specific reverse transcriptase that was first identified in 1985.45 Telomerase is activated and expressed in 85% of cancer cells, in germ cells, some stem cells46,47 and at a very low rate in some epithelial and intestinal cells.48 It functions by adding telomeric repeats GGTTAG at the end of the 3′-G-overhang of telomeres, using an RNA template complementary to the telomeric sequence.49 This mechanism prevents replicative senescence, thus conferring on cells unlimited proliferation.50 Indeed, it has been shown that when telomerase was introduced into normal cells, it resulted in an extension of their life span.51 Therefore, telomerase has been proposed as one of the hallmarks of cancer cells52 and as a cancer marker or prognosis factor.53 Telomerase consists of two main components: a protein component reverse transcriptase, hTERT, (127 kD) that contains conserved catalytic reverse transcriptase motifs,54 and a large RNA of 451 nucleotides, hTR, (153kD). It contains a template complementary to telomeric sequences, and other critical conserved regions essential for the assemblage of hTERT with hTR or for telomerase activity55,56 (Figure 7.3). Recently, a third partner, dyskerin, a pseudouridine synthase belonging to the H/ACA box ribonucleoproteins, has been identified as an essential component of the telomerase complex57,58 that probably binds to the H/ACA motif of hTR. Moreover the telomerase trimeric complex has been proposed to be associated in cells with about 30 other regulatory proteins involved in biogenesis, trafficking and recruitment to telomeres.57,59 hTR

hTR

CAAUCCCAAUC TTAGGGTTAGG

CR4–5

hTERT

Pseudoknot (CR2–3) 3' template

5'

H/ACA (CR6–8)

CR7

Figure 7.3 Telomerase is composed of two main components: a protein component that contains the catalytic reverse transcriptase motif (hTERT) and a large RNA of 451 nucleotides (hTR) containing the template region able to elongate the 3′ extremity of telomeres. Other proteins, that are not necessary for the in vitro telomerase activity, have been associated with telomerase in vivo, such as dyskerin. In addition to the template, many structural domains of hTR are conserved among the species (in grey boxes). They are involved in the assemblage of hTR with hTERT, in telomerase activity or in hTR processing

G-Quadruplex Structures

213

Since telomerase activity is elevated in the vast majority of tumours, telomerase participates in the immortality of cells, thanks to telomere elongation. Since telomeres are critically shortened in tumours versus normal tissues, it is thus conceivable to inhibit telomerase as a strategy for cancer therapy. Inhibition of telomerase has been performed using oligonucleotides targeting hTR or the mRNA of hTERT, or nucleoside analogues and nonnucleoside inhibitors targeting the catalytic centre of hTERT. They have been shown to induce progressive telomere shortening and senescence or apoptosis of treated cancer cells.60–63

7.4 G-Quadruplex Structures and Small Molecules It has been recognised that the 3′-G-overhang of the telomere, (TTAGGG)n, is able to fold, in vitro, into G-quadruplex structures. The structures consist of G-quartets that are stacked in a planar arrangement of four guanines associated by Hoogsteentype hydrogen bonds. Four consecutive repeats of guanines in the telomeric sequence are required for its folding in an intramolecular G-quadruplex structure. The monovalent cations Na+ or K+ are required to stabilise these structures by interacting with the eight carbonyl oxygens of two adjacent G-quartets64,65 (Figure 7.4A). Although there is no direct proof of G-quadruplexes in vivo, recent biological evidence is in favour of their existence in the ciliate Stylonychia66 and in humans.67 Moreover, some proteins have been shown to promote their formation68 and to induce their unfolding.69 G-quadruplexes can also be formed outside the telomere region, because intramolecular G-quadruplex-forming sequence motifs are prevalent in the genome,70,71 and particularly enriched in gene promoters.72 Moreover, it has been shown that gene expression depending on these promoters could be modulated by G-quadruplex binders.73 Crystallographic and NMR studies of human telomeric quadruplex structures have revealed a high degree of polymorphism in the structures, depending on the folding of the DNA backbone, the orientation of the strands (parallel or antiparallel) and the conformation of the TTA loops (reverse, lateral or transversal). Four main G-quadruplex structures have been described for telomeric DNA: parallel,74 antiparallel75 and two mixed-hybrid structures76,77 (Figures 7.4B–E). Each structure has been determined in the presence of a particular monovalent cation (Na+ or K+), but it is now clear that these structures coexist in solution, whatever the cation.78–81 Since the substrate of telomerase is the linear extremity of the 3′-G-overhang of telomeric DNA, it has been shown that its folding in G-quadruplex structures may impede its recognition by telomerase and may interfere with telomere elongation.82 Therefore, the stabilization of quadruplex DNA structures by small molecules has been proposed as a new strategy in order to inhibit telomerase and interfere with telomere maintenance in tumour cells.53 Many molecules have been synthesised in order to target G-quadruplexes and then tested on telomerase activity and cell proliferation. The paradigm for telomerase inhibitors is that senescence of cancer cells will occur after many rounds of replication because of progressive telomere shortening. Some of the G-quadruplex binders have been shown to follow this rule.83 However, the test used


214

A

T12 3'

T11 5'

A13 G10

G8

5'

G22

G14

A1

A1

A7

G14

G2

G2

G21

G9

G20

G9

A13 A19 G15

T6 G21

G3

G15

G3 G8

T17

G10

G20

G16

T5 G4

G16 T5

T6

G4

A19

G22 3'

T18

T17

B

C T12

T18 T17

A13

T11 5'

G8

G10

5'

G14

G20

G9

A19 G15

G3

G4

G22

A7 T5

T6 D

G2

G9

G3

G10

G20 G15

T18

T17 G16

G16

T6

G21 G8

A19

A1 A7

A1 G2

T18

G21 G14

T5 G4

T11 G22

A13 3'

3'

T12

E

Figure 7.4 Schematic structures of the G-quadruplexes formed by the human telomeric sequence: (A) the G-quartet consists of four guanines linked through Hoogsteen-type hydrogen bonds; (B) d[AGGG(TTAGGG)3] antiparallel, Na+ solution;75 (C) d[AGGG(TTAGGG)3] parallel, K+ (crystal); (D) and (E) d[TAGGG(TTAGGG)3])77 and d[TAGGG(TTAGGG)3TT],76 K+ solution, mixed-hybrids. Syn and anti guanines are coloured white and grey, respectively

G-Quadruplex Structures

215

Figure 7.5 Chemical formulae of G-quadruplex ligands that inhibit cell proliferation in a way independent of telomerase elongation

for appreciation of in vitro telomerase inhibition, TRAP (telomeric repeat amplification protocol) has been revaluated. The results indicate that the inhibitory effect of many G-quadruplex binders on telomerase elongation has been overestimated because these molecules were shown to be good inhibitors of the PCR step used for the amplification protocol.84 Moreover, some of these G-quadruplex binders (telomestatin, RHPS4, BRACO 19) (Figure 7.5) have been shown to inhibit cell proliferation in a way independent of telomerase elongation.85–90 It has been suggested that they stabilize the multiple G-quadruplex structures that can be formed transiently along telomeric DNA,91,92 thus altering the recognition of the telomeric sequences by proteins from the shelterin (TRF2 and POT1)85–89 or that they can inhibit telomere replication.93 The resulting increase in genomic instability has been found to depend on the type of cancerous cells.33,90 This selective toxicity for cancer cells can be explained by the fact that telomeres from normal and cancer cells exhibit differences in structure, stability and accessibility.94 There is therefore a growing interest in considering G-quadruplexes as potential targets for new antitumour drugs.64,65,91–93,95,96 The past few years have seen considerable development in design and synthesis of novel and efficient G-quadruplex binders.97 At present, the most efficient one is telomestatin.85–88,98–100 In general, an effective G-quadruplex binder is planar and, in most of cases, positively charged. On the basis of the very few structural data of G-quadruplex binders available nowadays, it has been proposed that the interaction occurs via π-stacking and electrostatic interactions.101–106 Design of new binders that are easily synthesized is still a challenge. A promising recent approach relies on the use of metallo-organic complexes, based on the assumption that the central metal could be positioned over the cation channel of the quadruplex, thereby optimizing the stacking interactions with the

216


Figure 7.6 Chemical formulae of selected metallo-organic G-quadruplex ligands

external G-quartet.107 Therefore, transition metals are considered very appealing, thanks to their multiple and controllable geometries, the possible modifications of their coordination environment and their easy synthetic access. Moreover, it has been highlighted that the metal centre is a key parameter governing the recognition of G-quadruplex using a family of terpyridine metallo-organic complexes.108 The pertinence of this approach has been demonstrated by the high selectivity of saphen (Ni(II)),109 terpyridine (Cu(II) and Pt(II)),108 porphyrins (Mn(III),110 Cu(II)111,112 and Ni(II)113, 114) and phthalocyanine derivatives (Zn(II) and Ni(II)115,116) for Gquadruplexes (Figure 7.6).

7.5 Cisplatin The antitumour effect of the drug cisplatin, cis-[Pt(II)Cl2(NH3)2], was discovered in 1965.117 The cellular target of cisplatin, which triggers its anticancer activity, is

Interactions of Platinum Complexes with G-Quadruplex Structures 217

assumed to be DNA. Cisplatin becomes activated intracellularly by the aquation of the two chloride leaving groups, leading to cis-[Pt(II)(NH3)2(H2O)2]2+, and subsequently covalently binds to DNA, mainly forming intrastrand crosslinks on DNA between two adjacent guanines d(GpG) (65%) or adjacent adenine and guanine, d(ApG) (25%), by coordinating to N7 of purine bases.118–120 The antitumour effect of cisplatin is attributed to the formation of the major GG adduct because the tumour response correlates with the level of GG adducts. The recognition and processing by DNA damage-response proteins120 induce signal transduction pathways that lead to apoptosis or necrosis. However, although cellular responses to platinum-induced damage have been widely studied119, 121 all the biological targets are not fully revealed. Since human telomeres consist of long tandem repeats of TTAGGG sequences, they are potential targets for platinum complexes. Given that telomeric nucleosomes are intrinsically more mobile under physiological conditions than nucleosomes formed on an average DNA sequence,94 one can envisage that platination of telomeres could be kinetically favoured over average DNA. It is thus conceivable that cisplatin may exert part of its anticancer activity via the formation of DNA adducts at telomeres, altering the telomere structure and thus impeding its access to telomerase, and/or lead to a loss of specific telomere binding proteins. It is also possible that cisplatin could influence directly the telomerase activity by binding to its RNA core, hTR. This review will present the advances in this field.

7.6 Interaction of Cisplatin and Related Platinum Complexes with G-Quadruplex Structures Since metallo-organic complexes appeared recently as promising new G-quadruplex binders,97 square planar Pt(II) complexes with extended aromatic ligands have been prepared. Indeed, they provide a p-surface that is compatible with the G-quartet motif. Although some of them have been extensively studied as duplex DNA binders, they have been less studied for their recognition of G-quadruplex DNA. Only a few examples of platinum–organic G-quadruplex binders are known to date. They belong to two different classes: (i) Pt(II) complexes without any labile ligand, unable to undergo platination reaction on the G-quadruplexes and therefore considered as reversible binders; (ii) mono-functional Pt(II) complexes containing one labile ligand (Cl−) potentially able to covalently link G-quadruplexes and considered as potential irreversible binders in addition to their stabilizing effect. (i) Bipyridine and phenanthroimidazole ethylenediamine platinum (II) complexes (Figure 7.7A) have been shown to interact with a 12-mer intermolecular quadruplex-forming sequence (T4G4T4)4 with two platinum complexes per quadruplex.122 Binding constants are in the range 106–107 M−1 increasing with the p-surface of the complexes. They induce significant stabilization of the intermolecular G-quadruplex structure and increased preference of two orders of magnitude for quadruplex over duplex. This binding mode is consistent with an end-stacking platinum complex located at each side of the quadruplex structure. Molecular modelling studies of the Pt(II) complexes with the quadruplex

218


Figure 7.7 Chemical formulae of selected platinum–organic G-quadruplex ligands known to stabilize specifically G-quadruplexes over duplex DNA. Platinum complexes (A) are unable to platinate G-quadruplexes once bound to it, whereas platinum complexes (B) are potentially able to platinate G-quadruplexes since they are monofunctional. However, this possibility has not been yet investigated

structure of (T4G4T4)4 showed p-stacking interaction of the aromatic system with the guanines of both external G-quartets. Surprisingly, the Pt(II) is offcentre due to the presence of three hydrogen bonds between the ethylenediamine ligand and the DNA phosphate backbone.122 (ii) A mono-substituted phenanthroline has been synthesized and used as a tridendate N,N,N-ligand to coordinate to Pt(II)123 (Figure 7.7B). This design was based on molecular modelling suggesting that the Pt(II) ion could be positioned over the cation channel at the centre of an accessible G-quartet, optimizing the stacking interactions. In the same context, three Pt(II)–terpyridine complexes have been synthesized (Figure 7.7B).108 The monosubstituted phenanthroline– Pt(II) complex induces a high degree of stabilization for quadruplexes (increase in the melting temperature over 20 °C) and is selective for quadruplex structure over duplex by a factor of 40. A comparable increase in melting temperature (11–16 °C) was observed with the Pt(II)–terpyridine complexes. Moreover, the metal-free ligand has considerably lower effect on the stabilization than Pt(II) complexes, suggesting that the presence of a Pt(II) centre plays an important role in quadruplex stabilization. Since both platinum complexes are monofunctional, (one Cl− labile ligand), they should be able to link covalently to purines of G-quadruplexes. However, this possibility has not yet been investigated by the authors. Studies from our group pointed out that G-quadruplex structures, against all expectations, were potential targets for Pt(II) coordination.78,124–127). We first analysed the possible platination of two quadruplex structures of the Tetrahymena

Interactions of Platinum Complexes with G-Quadruplex Structures 219 5'

T14

G3

G15

G4

T14

5'

T13 G3

G12

T13 G15

G12

3' G24

G21

G16 G5

G16

T20

G10

G6

G17

G22

G17

G23

G5 T19

G22

G21

G9 T7 G18

T8

T19

G11

G4

G11

G23

G10 G6 G9 T7

G24

G18 T8

3'

T20 A

B

Figure 7.8 Schematic structures of the G-quadruplex formed by the telomeric sequence of Tetrahymena. The antiparallel structure (A) has been solved by chemical probes128 and the mixed-hybrid one (B) has been solved by NMR130

sequence, (T2G4)4, which have been structurally determined by Williamson et al.128, Henderson et al.129 and Wang and Patel.130 These structures contain four guanines in their loops in addition of the three G-quartets (Figure 7.8). The four guanines of a G-quartet have their N7 atom involved in hydrogen bonding. Therefore, within a G-quadruplex structure, we postulated that only the four guanines not included in the G-quartets should react with Pt(II) complexes. The two quadruplex structures exhibit various loop and strand orientations (Figure 7.8), in which different guanines, not included in the G-quartets, are exposed: G3, G9, G15, G21 in the antiparallel structure (Figure 7.8A) and G6, G9, G15, G24 in the hybride structure (Figure 7.8B). Using the monofunctional triammine complex [Pt(II)(NH3)3(H2O)]2+, we have shown that only the four guanines G3, G9 G15 and G21, (but not G6 and G24) are platinated, suggesting the existence of the antiparallel structure as the major one in solution. However, the four guanines were not equally accessible for platination. Molecular dynamics simulations were used in order to interpret these results and allowed us to conclude that the relative accessibility of the N7 atom of the free guanines depends on their stacking with the neighbouring bases and on their binding with the central cation, positioned within the loops.124 Therefore [Pt(II)(NH3)3(H2O)]2+ could be considered as a useful tool to check the accessibility of the nucleophilic sites of purines. Furthermore, the difunctional Pt(II) complexes, cis- and trans[Pt(II)(NH3)2(H2O)2]2+, were shown to crosslink both ends of the quadruplex structure of (T2G4)4, since G3–G15 and G9–G21 crosslinks were formed. Molecular dynamics simulations indicated that these crosslinks occurred without perturbing the overall structure and the quartet stacking.124

220


It was therefore challenging to determine if such Pt(II) complexes were able to cross link the quadruplex structures of the human telomeric sequence, AG3(T2AG3)3. Indeed, they contain only four adenines in the loops and stem that are accessible to Pt(II) complexes (Figure 7.3). Moreover it was of pharmacological interest to establish if such a crosslink might exist and have similar cellular effects to those of compounds stabilizing the G-quadruplex structures.96,131 The results of the in vitro platination of the human G-quadruplex structures by the monofunctional platinum complex [Pt(II)(NH3)3(H2O)]2+ was surprising because, in addition to the four adenines, some guanines were platinated, whereas their N7 atom was expected to be involved in hydrogen bonds into the G-quartets. Thanks to unconstrained molecular dynamics simulations run on the antiparallel and on the parallel structures (Figure 7.9), it was possible to rationalize the platination of these specific guanines. Indeed, the simulations clearly showed that both structures were flexible and that the G-quartet located at the transversal loop side of the antiparallel structure could be disrupted, allowing the N7 atoms of G2, G10, G14 and G22 to be platinated. Moreover, two guanines located in the G-quartet of the 5′ extremity of the parallel structure were also shown to slip out of the G-quartet, rendering their N7 atoms (G8 and G20) accessible to platination. Since the adenines of the loops and some guanines of the G-quartets are potential platination sites, crosslinks of human telomeric G-quadruplex structures by difunctional Pt(II) complexes may be envisioned. Actually, we showed that the antiparallel structure was crosslinked by the difunctional Pt(II) complexes, cis- and trans-[Pt(II)(NH3)2(H2O)2]2+, between an adenine and a guanine, (A1–G10, A13–G22), as well as the two adenines located at both ends of the structure (A1–A13, A7–A19).125 These results indicate that purines of the parallel structure (A–A or A–G) are not close enough to be crosslinked by the mononuclear difunctional platinum

G2

G20

G10 G14

G-quartets

G8

G22

Anti-parallel structure Parallel structure of the human telomeric DNA sequence AG3(T2AG3)3

Figure 7.9 Conformation of the antiparallel and the parallel structures AG3(T2AG3)3 from molecular dynamics simulations. The top G-quartet of the anti-parallel structure can be completely disrupted whereas the top G-quartet of the parallel structure can lose two guanines. The slipping out of these guanines render their N7 atom accessible to platinum complexes. (Adapted from Biochemistry, 2005, 44, 10620–10634.)

Interactions of Platinum Complexes with G-Quadruplex Structures 221

complexes. In order to try to crosslink the parallel structure, we used the dinuclear difunctional platinum complexes, [{trans-PtCl(NH3)2}2H2N(CH2)nNH2]Cl2 (n = 2 or 6), in which two platinum coordination units are linked by a flexible diamine linker that allows long range crosslinks on duplex DNA.132 We found that the parallel structure was mainly crosslinked between two guanines at the 5′-end G-quartet (G8–G10) and (G2–G20), respectively, and that the antiparallel structure was crosslinked between two guanines at the 5′-end G-quartet (G2–G10), (G10–G22) and (G2–G14).78 All these results highlight that human G-quadruplex structures may serve as favourable platination targets, since some of their guanines can reversibly leave the G-quartet plane, thus rendering their N7 atom accessible to platinum complexes. Moreover, we have showed that platination of human telomeric G-quadruplexes by cisplatin was kinetically favoured in vitro by a factor of about two over a duplex containing a unique GG platination site.126 Based on these results we have designed new platinum complexes combining a G-quadruplex binder133 with Pt(II) in order to target more specifically Gquadruplex structures. Firstly, a platinum–quinacridine complex (Pt-MPQ) has been specifically designed to bind covalently in order to stabilize G-quadruplex structures.127 The platinum moiety can reach the opposite external G-quartet and bind to an accessible purine (Figure 7.10). Actually, this novel platinum hybrid complex is shown to trap preferentially the antiparallel quadruplex structure by linking to the guanines of the 5′-end G-quartet and also enhance quadruplex versus duplexDNA selectivity. More importantly, Pt-MPQ displays cytotoxic activities analogous to that of cisplatin on various cell lines. Another monofunctional hybrid platinum complex, a Pt(II)-acridine derivative, Pt-ACRAMTU, (Figure 7.10) has also been synthesized. Initially, it was known to form adenine adducts in duplex DNA. Later, it has been shown to bind preferentially to the adenines of the TTA loops of the G-quadruplex of human telomeric sequences and to react significantly faster with G-quadruplex than with duplex DNA.134 It was proposed that the acridine was stacked on one of the external G-quartets, which will favour the platination of an adenine of the neighbouring TTA loops. Interestingly, the two hybrid platinum complexes (Pt-MPQ and Pt-ACRAMTU) show major differences in their binding mechanisms to Gquadruplexes. It is conceivable that the length of the linker between the platinum moiety and the G-quadruplex binder governs the site of platination. However, the reversible stacking reaction is quicker than the irreversible platination reaction in both cases, suggesting that it is the binding reaction that will govern the platination site. For Pt-ACRAMTU, the binding of the acridine and the platination reaction will occur at the same side of the G-quadruplex. In this case, the guanines of the G-quartet are protected from platination due to the p-stacking interaction with the acridine. Thus, only N7 and N1 atoms of adenines in the loops are free and susceptible to interact with Pt(II). For Pt-MPQ, the quinacridine moiety binds at one side of the G-quadruplex and platination can occur at the opposite external G-quartet, thanks to the linker. Since the guanines of the 5′-end G-quartet of the antiparallel structure have been platinated by Pt-MPQ, it suggests that the quinacridine moiety binds at the opposite 3′-end G-quartet. Moreover, it should be noted that no adenine

222

Telomeres and Telomerase T12 3'

Pt 5'

A13

G10

G22

A1 G14

G2

G21

G9 G3

G15

G8 G4

G20 A19

G16

T18

T5 T12 5'

G10

A13 Pt

3' G22

A1 G14

G2

G21

G9 G15

G3 G8 G4

G20 G16

T5

T6

T17

A19 T18

Figure 7.10 Chemical formulae of two platinum–organic G-quadruplex binders that have been shown to stabilize selectively and to bind covalently to G-quadruplexes. The proposed binding mode on the antiparallel human telomeric G-quadruplex is indicated. For Pt-MPQ, the binding and platination sites are located at both sides of the G-quadruplex,127 whereas, for Pt-ACRAMTU,134 both sites are located at the same side of the G-quadruplex. Platination sites are represented by circles

was found to be platinated, indicating that the guanines, even when they are thought to be involved in the G-quartet, react more rapidly than adenines that are about ten times less nucleophilic.135 It is worth noticing that binding of the quinacridine moiety on the 5′-end G-quartet cannot be excluded, but it would prevent the platination of the G-quadruplex, since no platinable guanines are available on the opposite 3′-end.78 Since the four adenines and some guanines of the flexible G-quartets of the human G-quadruplex structure are efficiently platinated, the design of new platinum–hybrid compounds using different G-quadruplex binding motifs, represents an exciting challenge. Preliminary studies with these conjugates have revealed interesting biological responses worth further investigation.

Interactions of Platinum Complexes with Telomerase 223

7.7 Interaction of Cisplatin and Related Platinum Complexes with Telomeric DNA Duplexes Due to the presence of triple runs of guanines, a telomeric DNA duplex is a potential target for cisplatin. Firstly, it has been shown that the platinum content of a plasmid containing a 800bp (TTAGGG)n sequence is higher than in the remaining plasmid sequence, in proportion to the greater number of guanines.136 Secondly, it has been shown that longer telomeres are associated with resistance to cisplatin in melanoma cells.137 It is thus conceivable that the telomeric DNA–cisplatin interaction could be relevant for drug sensitivity/resistance status depending on the lengths of the telomere. However, it is worth noticing that this hypothesis has been invalidated in other cell lines.138 Thirdly, modified guanines are critical for association of telomere binding proteins since it has been shown that the presence of single or multiple 8-oxo-G lesions on telomeric sequences interfere with the in vitro binding of TRF1 and TRF2.139 All these results enable us to hypothesize that the modification of DNA by cisplatin could occur at telomeres and thus disturb the recognition of telomeric DNA by TRF1 and TRF2. Based on this hypothesis, we analysed the in vitro binding of TRF1 and TRF2 on the telomeric sequences (T2AG3)4/(C3TA2)4 platinated at a defined GGG site by cisplatin. We showed that the presence of cisplatin decreases the affinity of TRF2 more dramatically than that of TRF1 for these sequences.140 These results suggest that if there is any platination of telomeres by cisplatin in cancerous cells, it could have some impact on the binding of TRF2 and consequently could affect the integrity of the telomeres.

7.8 Interaction of Cisplatin and Related Platinum Complexes with Telomerase The RNA component of telomerase, hTR, is necessary for the reverse transcriptase activity of the enzyme. The template and the conserved structured regions (loops and pseudo-knot) of hTR are accessible binding sites,141 consequently hTR is a potential target for antitelomerase drugs. Antisense oligonucleotides, peptide nucleic acids (PNAs) directed against the template142 and small molecules directed against the tertiary structures143,144 have been shown to inhibit efficiently telomerase activity. The most promising analogue is GRN163L, a lipid conjugate to thiophosphoramidate145 that is used in breast cancer therapy.146 Since electrophilic agents could interact with the accessible nucleobases of hTR, the effect of cisplatin and some Pt(II) derivatives has been evaluated in vitro on telomerase activity using the TRAP assay.147 Particularly, novel platinum complexes bearing aromatic amines as bulky carrier groups have been prepared for this purpose.148,149 Inhibition of telomerase, dependent on Pt(II) complex interactions, has been observed, indicating platination of hTR; however, we can not exclude that the effect may be due to platination of essential nucleophilic amino acids of the proteic unit, hTERT. It is know that TRAP assay occurs with cells extracts during 30 minutes of incubation, followed by PCR amplification of the telomeric repeats.150 However, this incubation timescale is

224


insufficient to allow total hydrolysis of the platinum complexes;151 indeed, the half life of hydrolysis of cisplatin and carboplatin is about 2 and 286 h, respectively. Therefore, the in vitro assay of telomerase activity is not fully informative, since a part of the platinum complexes remains inactive. Many workers have also investigated the cellular effects of cisplatin on telomere lengths, and telomerase residual activity and expression. However, the three parameters have never been taken into account at the same time. Until now, the effect of cisplatin on telomeres has not been well clarified. Concerning the effect of cisplatin on telomere length, the first study showed that telomere loss in Hela-treated cells (20 kbp in telomere length) was dependent on cisplatin dose. Low concentration of cisplatin (0.5 mM) induces telomere shortening after 24 h of treatment, whereas higher concentrations that blocked DNA replication in the S phase did not have any effect on telomere length.152 Cisplatin was also shown to induce the shortening of telomeres in hepatoma cells after 24 to 72 h treatment (0.8 to 50 mM).153 Since the rapid loss in telomere length was observed after only one to three rounds of replications, it probably did not result from telomerase inhibition, which induces a progressive shortening of telomeres.83 It can thus be proposed that platination of telomeres may inhibit telomere replication and/or could uncap the telomere, as already suggested for rapid loss of telomeres from quadruplex-binder-treated cells.93 A gradual shortening of telomeres was observed when yeast NER (nucleotide excision repair) mutant cells were treated by doses of cisplatin that did not affect cell viability.154 This suggests that the NER pathway, which is involved in repair of cisplatin adducts,119,155,156 may play a critical role in the repair and maintenance of telomeres. However, in another recent study it was found that cisplatin treatment (0.5–500 mM) induced no loss of telomeres after 2 to 72 h in three cell lines, independent of their initial telomere length (4–80 kbp) (neuroblastoma, Hela, acute lymphoblastic T cell).138 The results from these different studies were not conclusive and gave only an indication of the possibility that cisplatin could interfere with telomeres. It could depend on the susceptibility of the various cell lines used in the studies toward cisplatin. Moreover, the high cytotoxic activity of cisplatin makes it difficult to discriminate between overall cellular effects and the specific effect on telomerase and/or telomeres. Telomerase activity of cisplatin-treated cells has also been investigated. Telomerase expression (hTERT, mRNA or hTR quantification) has been controlled in parallel, in some cases, in order to find out if the decrease in telomerase activity was due to its downregulation or to its inactivation. Since the evaluation of this activity has been performed with TRAP assay150 after cell extraction, telomerase inhibition can reflect only irreversible inactivation of telomerase. This inactivation, which occurs upon cell treatment, could be attributed to covalent binding of cisplatin to telomerase (hTR or hTERT), but there is no proof of this event at this time. The results are still the subject of controversy. Upon cisplatin treatment, a decrease in telomerase activity has been found in human testicular cancer cells,157 associated with a downregulation of hTR. A decrease in telomerase activity has also been observed in hepatoma cells,153 breast cancer cells in culture and in xenograph tumours.158 In contrast, no decrease of telomerase activity has been detected after

Conclusion

225

treatment by cisplatin of haematopoietic cancer cells when used at concentrations resulting in cytostatic growth inhibition159 in human ovarian adenocarcinoma cell lines,160 neuroblastoma, Hela and acute lymphoblastic T cells138 and of nasopharyngeal cancer cells.161 No significant downexpression of hTERT has been observed in MCF-7 cells treated with cisplatin.162 Since telomerase plays a role in cellular resistance to apoptosis, which is the primary mode of cell death induced by several drugs, combining telomerase inhibition with chemotherapeutic treatment may prove more efficient than each approach on its own. Many studies have been done in order to investigate the relationship between telomerase, telomere dysfunction and vulnerability to drug-induced apoptosis, particularly by cisplatin. Inhibition of telomerase activity or expression increased the susceptibility of glioblastoma cells,163 and of drug-resistant and drugsensitive promyelocytic leukemia and breast cancer cells164 to cisplatin-induced apoptosis. Results were controversial for melanoma cells, since neither an increase in sensibility for cisplatin was detected when telomerase was inhibited30 nor a precise relationship was found between telomerase activity and cellular sensitivity to cisplatin.137,165 Moreover, telomerase activity of cisplatin-treated cells has also been shown to depend on the period of treatment of prostate cell lines.166 Other studies revealed that the high level of hTERT expression and/or telomerase activity are related to chemosensitivity to cisplatin in ovarian carcinoma,167 and in oesophageal cancer cell lines,168 suggesting that telomerase could be a marker of prognosis for cisplatin treatment. A study of cisplatin treatment of telomere dysfunctional and telomere functional mice cells suggested that telomere dysfunction does not govern chemosensitivity to cisplatin.169 Based on all available data one may conclude that telomerase may play a role in tumour cell proliferation by modulating the cellular response to chemotherapy, particularly for drugs that affect DNA integrity.

7.9 Conclusion In vitro experiments have shown that telomerase and telomeres are good targets for cisplatin and some of its derivatives. The various results from cellular studies have been explained by differential suppression of hTERT expression and/or inhibition of telomerase activity that could lead to telomere shortening. However, this inhibition seems to depend on the cell lines, the doses and the experimental conditions. An unambiguous conclusion on the effects of cisplatin treatments may be obtained by comparing different cell lines under strict identical experimental conditions. Final conclusions must therefore await further studies. There is one way that has not yet been investigated extensively, but that appears to be of importance due to the increasing role of telomeres in cell viability; it concerns the possible platination of telomeric DNA. This event could lead to the wellknown cellular responses to cisplatin treatment leading to apoptosis or senescence.119 This hypothesis is supported by the fact that telomeric DNA is more accessible than genomic DNA94 and that telomere dysfunction induced by noncovalent quadruplex DNA binders is favoured in cancer cells over normal cells.33 Platination of telomeric

226


DNA could lock the 3′- G-overhang in a G-quadruplex structure that is no longer recognized by telomerase; this could in turn inhibit replication of telomeres or uncap telomeres by impeding the binding of telomeric proteins. The latter approach is at present under investigation in our laboratory. Targeting telomere mechanisms could be of growing interest for future successful therapeutic anticancer strategies, since telomeres rather than telomerase may be the universal target for cancer therapy.

Acknowledgements We wish to thank Dr M.P. Teulade-Fichou, Dr D. Monchaud, H. Bertrand and Dr E. Segal-Bendirdjian for helpful discussions, and the Association pour la Recherche contre le Cancer (grants 3482 and 4835) for financial help.

Abbreviations TRF1 TRF2 POT1 TIN2 TPP1 Rap1 hTR hTERT 3′-G-overhang Pt-MPQ

Telomere repeat factor 1 Telomere repeat factor 2 Protection of telomere 1 TRF1 interacting nuclear protein 2 Tripetidylpeptidase 1 Repressor/activator protein 1 Human telomerase RNA component. Human telomerase reverse transcritase Single strand of the 3′ extremity of telomere that consists of G-rich repeated sequences Platinum–quinacridine complex

References 1. Blackburn, E.H.; Telomere states and cell fates; Nature, 2000, 408(6808), 53–56. 2. Karlseder, J.; Smogorzewska, A.; de Lange, T.; Senescence induced by altered telomere state, not telomere loss; Science, 2002, 295(5564), 2446–2449. 3. Takai, H.; Smogorzewska, A.; de Lange, T.; DNA damage foci at dysfunctional telomeres; Curr. Biol., 2003, 13(17), 1549–1556. 4. de Lange, T.; Telomere-related genome instability in cancer; Cold Spring Harbour Symposium; Quant. Biol., 2005, 70, 197–204. 5. Londono-Vallejo, J.A.; Telomere instability and cancer; Biochimie, 2008, 90(1), 73–82. 6. de Lange, T.; Protection of mammalian telomeres; Oncogene, 2002, 21(4), 532–540. 7. Watson, J.; Origin of concatemeric T7 DNA; Nat. New Biol., 1972, 239, 197–201. 8. Olovnikov, A.M.; A theory of marginotomy. The incomplete copying of template margin in enzymic synthesis of polynucleotides and biological significance of the phenomenon; J. Theor. Biol., 1973, 41(1), 181–190.

References

227

9. Gilson, E.; Geli, V.; How telomeres are replicated; Nat. Rev. Mol. Cell Biol., 2007, 8(10), 825–838. 10. Verdun, R.E.; Karlseder, J.; Replication and protection of telomeres; Nature, 2007, 447(7147), 924–931. 11. Cimino-Reale, G.; Pascale, E.; Alvino, E.; Starace, G.; D’Ambrosio, E.; Long telomeric C-rich 5′-tails in human replicating cells; J. Biol. Chem., 2003, 278(4), 2136–2140. 12. Sfeir, A.J.; Chai, W.; Shay, J.W.; Wright, W.E.; Telomere-end processing the terminal nucleotides of human chromosomes; Mol. Cell., 2005, 18(1), 131–138. 13. Lenain, C.; Bauwens, S.; Amiard, S.; Brunori, M.; Giraud-Panis, M.J.; Gilson, E.; The Apollo 5′ exonuclease functions together with TRF2 to protect telomeres from DNA repair; Curr. Biol., 2006, 16(13), 1303–1310. 14. Shay, J.W.; Wright, W.E.; Hayflick, his limit, and cellular ageing; Nat. Rev. Mol. Cell Biol., 2000, 1(1), 72–76. 15. Wright, W.E.; Shay, J.W.; Cellular senescence as a tumor-protection mechanism: the essential role of counting; Curr. Opin. Genet. Dev., 2001, 11(1), 98–103. 16. Shay, J.W.; Wright, W.E.; Hallmarks of telomeres in ageing research; J. Pathol., 2007, 211(2), 114–123. 17. Liu, D.; O’Connor, M.S.; Qin, J.; Songyang, Z.; Telosome, a mammalian telomereassociated complex formed by multiple telomeric proteins; J. Biol. Chem., 2004, 279(49), 51338–51342. 18. de Lange, T.; Shelterin: the protein complex that shapes and safeguards human telomeres; Genes Dev., 2005, 19(18), 2100–2110. 19. Chong, L.; van Steensel, B.; Broccoli, D.; Erdjument Bromage, H.; Hanish, J.; Tempst, P. et al.; A human telomeric protein; Science, 1995, 270(5242), 1663–1667. 20. Bilaud, T.; Brun, C.; Ancelin, K.; Koering, C.E.; Laroche, T.; Gilson, E.; Telomeric localization of TRF2, a novel human telobox protein; Nat. Genet., 1997, 17(2), 236–239. 21. Broccoli, D.; Smogorzewska, A.; Chong, L.; de Lange, T.; Human telomeres contain two distinct Myb-related proteins, TRF1 and TRF2; Nat. Genet., 1997, 17(2), 231– 235. 22. Baumann, P.; Cech, T.R.; Pot1, the putative telomere end-binding protein in fission yeast and humans; Science, 2001, 292(5519), 1171–1175. 23. O’Connor, M.S.; Safari, A.; Xin, H.; Liu, D.; Songyang, Z.; A critical role for TPP1 and TIN2 interaction in high-order telomeric complex assembly; Proc. Natl. Acad. Sci. USA, 2006, 103(32), 11874–11879. 24. Smogorzewska, A.; de Lange, T.; Regulation of telomerase by telomeric proteins; Annu. Rev. Biochem., 2004, 73, 177–208. 25. Griffith, J.D.; Comeau, L.; Rosenfield, S.; Stansel, R.M.; Bianchi, A.; Moss, H. et al.; Mammalian telomeres end in a large duplex loop [see comments]; Cell, 1999, 97(4), 503–514. 26. de Lange, T.; T-loops and the origin of telomeres; Nat. Rev. Mol. Cell. Biol., 2004, 5(4), 323–329. 27. van Steensel, B.; Smogorzewska, A.; de Lange, T.; TRF2 protects human telomeres from end-to-end fusions; Cell, 1998, 92(3), 401–413. 28. Denchi, E.L.; de Lange, T.; Protection of telomeres through independent control of ATM and ATR by TRF2 and POT1; Nature, 2007, 448(7157), 1068–1071. 29. Celli, G.B.; Denchi, E.L.; de Lange, T.; Ku70 stimulates fusion of dysfunctional telomeres yet protects chromosome ends from homologous recombination; Nat. Cell Biol., 2006, 8(8), 885–890. 30. Biroccio, A.; Gabellini, C.; Amodei, S.; Benassi, B.; Del Bufalo, D.; Elli, R. et al., Telomere dysfunction increases cisplatin and ecteinascidin-743 sensitivity of melanoma cells; Mol. Pharmacol., 2003, 63(3), 632–638. 31. Nakanishi, K.; Kawai, T.; Kumaki, F.; Hiroi, S.; Mukai, M.; Ikeda, E. et al.; Expression of mRNAs for telomeric repeat binding factor (TRF)-1 and TRF2 in atypical adenomatous hyperplasia and adenocarcinoma of the lung; Clin. Cancer Res., 2003, 9(3), 1105–1111.

228


32. Gu, J.; Spitz, M.R.; Zhao, H.; Lin, J.; Grossman, H.B.; Dinney, C.P. et al.; Roles of tumor suppressor and telomere maintenance genes in cancer and aging – an epidemiological study; Carcinogenesis, 2005, 26(10), 1741–1747. 33. Biroccio, A.; Rizzo, A.; Elli, R.; Koering, C.E.; Belleville, A.; Benassi, B. et al.; TRF2 inhibition triggers apoptosis and reduces tumourigenicity of human melanoma cells; Eur. J. Cancer, 2006, 42(12), 1881–1888. 34. Colgin, L.M.; Baran, K.; Baumann, P.; Cech, T.R.; Reddel, R.R.; Human POT1 facilitates telomere elongation by telomerase; Curr. Biol., 2003, 13(11), 942–946. 35. Hockemeyer, D.; Palm, W.; Else, T.; Daniels, J.P.; Takai, K.K.; Ye, J.Z. et al.; Telomere protection by mammalian Pot1 requires interaction with Tpp1; Nat. Struct. Mol. Biol., 2007, 14(8), 754–761. 36. Etheridge, K.T.; Compton, S.A.; Barrientos, K.S.; Ozgur, S.; Griffith, J.D.; Counter, C.M.; Tethering telomeric double- and single-stranded DNA binding proteins inhibits telomere elongation; J. Biol. Chem., 2008, 283(11), 6935–6941. 37. Wang, F.; Podell, E.R.; Zaug, A.J.; Yang, Y.; Baciu, P.; Cech, T.R. et al.; The POT1-TPP1 telomere complex is a telomerase processivity factor; Nature, 2007, 445(7127), 506–510. 38. Xin, H.; Liu, D.; Wan, M.; Safari, A.; Kim, H.; Sun, W. et al.; TPP1 is a homologue of ciliate TEBP-beta and interacts with POT1 to recruit telomerase; Nature, 2007, 445(7127), 559–562. 39. Blasco, M.A.; Telomeres and human disease: ageing, cancer and beyond; Nat. Rev. Genet., 2005, 6(8), 611–622. 40. Hockemeyer, D.; Sfeir, A.J.; Shay, J.W.; Wright, W.E.; de Lange, T.; POT1 protects telomeres from a transient DNA damage response and determines how human chromosomes end; EMBO J., 2005, 24(14), 2667–2678. 41. Celli, G.B.; de Lange, T.; DNA processing is not required for ATM-mediated telomere damage response after TRF2 deletion; Nat. Cell Biol., 2005, 7(7), 712–718. 42. Zhu, X.D.; Niedernhofer, L.; Kuster, B.; Mann, M.; Hoeijmakers, J.H.; de Lange, T.; ERCC1/XPF removes the 3′ overhang from uncapped telomeres and represses formation of telomeric DNA-containing double minute chromosomes; Mol. Cell, 2003, 12(6), 1489–1498. 43. Wu, L.; Multani, A.S.; He, H.; Cosme-Blanco, W.; Deng, Y.; Deng, J.M. et al.; Pot1 deficiency initiates DNA damage checkpoint activation and aberrant homologous recombination at telomeres; Cell, 2006, 126(1), 49–62. 44. He, H.; Multani, A.S.; Cosme-Blanco, W.; Tahara, H.; Ma, J.; Pathak, S. et al.; POT1b protects telomeres from end-to-end chromosomal fusions and aberrant homologous recombination; EMBO J., 2006, 25(21), 5180–5190. 45. Greider, C.W.; Blackburn, E.H.; Identification of a specific telomere terminal transferase activity in Tetrahymena extracts; Cell, 1985, 43(2 Pt 1), 405–413. 46. Greider, C.W.; Telomerase activity, cell proliferation, and cancer; Proc. Natl. Acad. Sci. USA, 1998, 95(1), 90–92. 47. Broccoli, D.; Young, J.W.; de Lange, T.; Telomerase activity in normal and malignant hematopoietic cells; Proc. Natl. Acad. Sci. USA, 1995, 92(20), 9082–9086. 48. Masutomi, K.; Yu, E.Y.; Khurts, S.; Ben-Porath, I.; Currier, J.L.; Metz, G.B. et al.; Telomerase maintains telomere structure in normal human cells; Cell, 2003, 114(2), 241–253. 49. Morin, G.B., The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG repeats; Cell, 1989, 59(3), 521–529. 50. Stewart, S.A.; Weinberg, R.A.; Telomeres: cancer to human aging; Annu. Rev. Cell Dev. Biol., 2006, 22, 531–557. 51. Bodnar, A.G.; Ouellette, M.; Frolkis, M.; Holt, S.E.; Chiu, C.P.; Morin, G.B. et al.; Extension of life-span by introduction of telomerase into normal human cells [see comments]; Science, 1998, 279(5349), 349–352. 52. Hanahan, D.; Weinberg, R.A.; The hallmarks of cancer; Cell, 2000, 100(1), 57–70. 53. Mergny, J.L.; Riou, J.F.; Mailliet, P.; Teulade-Fichou, M.P.; Gilson, E.; Natural and pharmacological regulation of telomerase; Nucleic Acids Res., 2002, 30(4), 839–865.

References

229

54. Lingner, J.; Hughes, T.R.; Shevchenko, A.; Mann, M.; Lundblad, V.; Cech, T.R.; Reverse transcriptase motifs in the catalytic subunit of telomerase [see comments]; Science, 1997, 276(5312), 561–567. 55. Chen, J.L.; Blasco, M.A.; Greider, C.W.; Secondary structure of vertebrate telomerase RNA; Cell, 2000, 100(5), 503–514. 56. Chen, J.L.; Greider, C.W.; Determinants in mammalian telomerase RNA that mediate enzyme processivity and cross-species incompatibility; EMBO J., 2003, 22(2), 304– 314. 57. Cohen, S.B.; Graham, M.E.; Lovrecz, G.O.; Bache, N.; Robinson, P.J.; Reddel, R.R.; Protein composition of catalytically active human telomerase from immortal cells; Science, 2007, 315(5820), 1850–1853. 58. Mergny, J.L.; Lacroix, L.; Riou, J.F.; Telomerase: a dimeric menage a trois; Chembiochem., 2007, 8(13), 1483–1485. 59. Autexier, C.; Lue, N.F.; The structure and function of telomerase reverse transcriptase; Annu. Rev. Biochem., 2006, 75, 493–517. 60. Shay, J.W.; Wright, W.E.; Telomerase therapeutics for cancer: challenges and new directions; Nat. Rev. Drug Discov., 2006, 5(7), 577–584. 61. Olaussen, K.A.; Dubrana, K.; Domont, J.; Spano, J.P.; Sabatier, L., Soria, J.C.; Telomeres and telomerase as targets for anticancer drug development; Crit. Rev. Oncol. Hematol., 2006, 57(3), 191–214. 62. Pendino, F.; Tarkanyi, I.; Dudognon, C.; Hillion, J.; Lanotte, M.; Aradi, J. et al.; Telomeres and telomerase: Pharmacological targets for new anticancer strategies?; Curr. Cancer Drug Targets, 2006, 6(2), 147–180. 63. Kelland, L.; Targeting the limitless replicative potential of cancer: the telomerase/telomere pathway; Clin. Cancer Res., 2007, 13(17), 4960–4963. 64. Davis, J.T.; G-quartets 40 years later: from 5′-GMP to molecular biology and supramolecular chemistry; Angew. Chem. Int. Ed. Engl., 2004, 43(6), 668–698. 65. Burge, S.; Parkinson, G.N.; Hazel, P.; Todd, A.K.; Neidle, S.; Quadruplex DNA: sequence, topology and structure; Nucleic Acids Res., 2006, 34(19), 5402–5415. 66. Schaffitzel, C.; Berger, I.; Postberg, J.; Hanes, J.; Lipps, H.J.; Pluckthun, A.; In vitro generated antibodies specific for telomeric guanine-quadruplex DNA react with Stylonychia lemnae macronuclei; Proc. Natl. Acad. Sci. USA, 2001, 98(15), 8572–8577. 67. Granotier, C.; Pennarun, G.; Riou, L.; Hoffschir, F.; Gauthier, L.R.; De Cian, A. et al.; Preferential binding of a G-quadruplex ligand to human chromosome ends; Nucleic Acids Res., 2005, 33(13), 4182–4190. 68. Paeschke, K.; Simonsson, T.; Postberg, J.; Rhodes, D.; Lipps, H.J.; Telomere end-binding proteins control the formation of G-quadruplex DNA structures in vivo; Nat. Struct. Mol. Biol., 2005, 12(10), 847–854. 69. Opresko, P.L.; Mason, P.A.; Podell, E.R.; Lei, M.; Hickson, I.D.; Cech, T.R. et al.; POT1 stimulates RecQ helicases WRN and BLM to unwind telomeric DNA substrates; J. Biol. Chem., 2005, 280(37), 32069–32080. 70. Huppert, J.L.; Balasubramanian, S.; Prevalence of quadruplexes in the human genome; Nucleic Acids Res., 2005, 33(9), 2908–2916. 71. Todd, A.K.; Johnston, M.; Neidle, S.; Highly prevalent putative quadruplex sequence motifs in human DNA; Nucleic Acids Res., 2005, 33(9), 2901–2907. 72. Huppert, J.L.; Balasubramanian, S.; G-quadruplexes in promoters throughout the human genome; Nucleic Acids Res., 2007, 35(2), 406–413. 73. Grand, C.L.; Han, H.; Munoz, R.M.; Weitman, S.; Von Hoff, D.D.; Hurley, L.H. et al.; The cationic porphyrin TMPyP4 down-regulates c-MYC and human telomerase reverse transcriptase expression and inhibits tumor growth in vivo; Mol. Cancer Ther., 2002, 1(8), 565–573. 74. Parkinson, G.N.; Lee, M.P.; Neidle, S.; Crystal structure of parallel quadruplexes from human telomeric DNA; Nature, 2002, 417(6891), 876–880. 75. Wang, Y.; Patel, D.J.; Solution structure of the human telomeric repeat d[AG3(T2AG3)3] G-tetraplex; Structure, 1993, 1(4), 263–282.

230


76. Phan, A.T.; Kuryavyi, V.; Luu, K.N.; Patel, D.J.; Structure of two intramolecular Gquadruplexes formed by natural human telomere sequences in K+ solution; Nucleic Acids Res., 2007, 35(19), 6517–6525. 77. Luu, K.N.; Phan, A.T.; Kuryavyi, V.; Lacroix, L.; Patel, D.J.; Structure of human telomere in K+ solution: an intramolecular (3 + 1) G-quadruplex; J. Am. Chem. Soc., 2006, 128, 9963–9970. 78. Ourliac-Garnier, I.; Elizondo-Riojas, M.A.; Redon, S.; Farrell, N.P.; Bombard, S.; Crosslinks of quadruplex structures from human telomeric DNA by dinuclear platinum complexes show the flexibility of both structures; Biochemistry, 2005, 44(31), 10620–10634. 79. He, Y.J.; Neumann, R.D.; Panyutin, I.G.; Intramolecular quadruplex conformation of human telomeric DNA assessed with I-125-radioprobing; Nucleic Acids Res., 2004, 32(18), 5359–5367. 80. Ying, L.; Green, J.J.; Li, H.; Klenerman, D.; Balasubramanian, S.; Studies on the structure and dynamics of the human telomeric G quadruplex by single-molecule fluorescence resonance energy transfer; Proc. Natl. Acad. Sci. USA, 2003, 100(25), 14629–14634. 81. Ren, J.; Qu, X.; Trent, J.O.; Chaires, J.B.; Tiny telomere DNA; Nucleic Acids Res., 2002, 30(11), 2307–2315. 82. Zahler, A.M.; Williamson, J.R.; Cech, T.R.; Prescott, D.M.; Inhibition of telomerase by G-quartet DNA structures; Nature, 1991, 350(6320), 718–720. 83. Riou, J.F.; Guittat, L.; Mailliet, P.; Laoui, A.; Renou, E.; Petitgenet, O. et al.; Cell senescence and telomere shortening induced by a new series of specific G-quadruplex DNA ligands; Proc. Natl. Acad. Sci. SA, 2002, 99(5), 2672–2677. 84. De Cian, A.; Cristofari, G.; Reichenbach, P.; De Lemos, E.; Monchaud, D.; TeuladeFichou, M.P. et al.; Reevaluation of telomerase inhibition by quadruplex ligands and their mechanisms of action; Proc. Natl. Acad. Sci. USA, 2007, 104(44), 17347–17352. 85. Gomez, D.; Paterski, R.; Lemarteleur, T.; Shin-Ya, K.; Mergny, J.L.; Riou, J.F.; Interaction of telomestatin with the telomeric single-strand overhang; J. Biol. Chem., 2004, 279(40), 41487–41494. 86. Gomez, D.; O’Donohue, M.F.; Wenner, T.; Douarre, C.; Macadre, J.; Koebel, P. et al.; The G-quadruplex ligand telomestatin inhibits POT1 binding to telomeric sequences in vitro and induces GFP-POT1 dissociation from telomeres in human cells; Cancer Res., 2006, 66(14), 6908–6912. 87. Gomez, D.; Wenner, T.; Brassart, B.; Douarre, C.; O’Donohue, M.F.; El Khoury, V. et al.; Telomestatin induced telomere uncapping is modulated by POT1 through G-overhang extension in HT1080 human tumor cells; J. Biol. Chem., 2006, 38721–38729. 88. Tahara, H.; Shin-Ya, K.; Seimiya, H.; Yamada, H.; Tsuruo, T.; Ide, T.; G-quadruplex stabilization by telomestatin induces TRF2 protein dissociation from telomeres and anaphase bridge formation accompanied by loss of the 3′ telomeric overhang in cancer cells; Oncogene, 2006, 25(13), 1955–1966. 89. Salvati, E.; Leonetti, C.; Rizzo, A.; Scarsella, M.; Mottolese, M.; Galati, R. et al.; Telomere damage induced by the G-quadruplex ligand RHPS4 has an antitumor effect; J. Clin. Invest., 2007, 117(11), 3236–3247. 90. Burger, A.M.; Dai, F.; Schultes, C.M.; Reszka, A.P.; Moore, M.J.; Double, J.A. et al.; The G-quadruplex-interactive molecule BRACO-19 inhibits tumor growth, consistent with telomere targeting and interference with telomerase function; Cancer Res., 2005, 65(4), 1489–1496. 91. Oganesian, L.; Bryan, T.M.; Physiological relevance of telomeric G-quadruplex formation: a potential drug target; Bioessays, 2007, 29(2), 155–165. 92. Maizels, N.; Dynamic roles for G4 DNA in the biology of eukaryotic cells; Nat. Struct. Mol. Biol., 2006, 13(12), 1055–1059. 93. De Cian, A.; Lacroix, L.; Douarre, C.; Temime-Smaali, N.; Trentesaux, C.; Riou, J.F. et al.; Targeting telomeres and telomerase; Biochimie, 2008, 90(1), 131–155. 94. Pisano, S.; Marchioni, E.; Galati, A.; Mechelli, R.; Savino, M.; Cacchione, S.; Telomeric nucleosomes are intrinsically mobile; J. Mol. Biol., 2007, 369(5), 1153–1162.

References

231

95. Phan, A.T.; Kuryavyi, V.; Patel, D.J.; DNA architecture: from G to Z; Curr. Opin. Struct. Biol., 2006, 16(3), 288–298. 96. Neidle, S.; Parkinson, G.; Telomere maintenance as a target for anticancer drug discovery; Nat. Rev. Drug Discov., 2002, 1(5), 383–393. 97. Monchaud, D.; Teulade-Fichou, M.P.; A hitchhiker’s guide to G-quadruplex ligands; Org. Biomol. Chem., 2008, 6(4), 627–636. 98. Shin-ya, K.; Wierzba, K.; Matsuo, K.; Ohtani, T.; Yamada, Y.; Furihata, K. et al.; Telomestatin, a novel telomerase inhibitor from Streptomyces anulatus; J. Am. Chem. Soc., 2001, 123(6), 1262–1263. 99. Kim, M.Y.; Vankayalapati, H.; Shin-Ya, K.; Wierzba, K.; Hurley, L.H.; Telomestatin, a potent telomerase inhibitor that interacts quite specifically with the human telomeric intramolecular G-quadruplex; J. Am. Chem. Soc., 2002, 124(10), 2098–2099. 100. Tauchi, T.; Shin-ya, K.; Sashida, G.; Sumi, M.; Okabe, S.; Ohyashiki, J.H. et al.; Telomerase inhibition with a novel G-quadruplex-interactive agent, telomestatin: in vitro and in vivo studies in acute leukemia; Oncogene, 2006, 25(42), 5719–5725. 101. Gavathiotis, E.; Heald, R.A.; Stevens, M.F.; Searle, M.S.; Recognition and stabilization of quadruplex DNA by a potent new telomerase inhibitor: NMR studies of the 2:1 complex of a pentacyclic methylacridinium cation with d(TTAGGGT); Angew. Chem. Int. Ed. Engl., 2001, 40(24), 4749–4751. 102. Gavathiotis, E.; Heald, R.A.; Stevens, M.F.; Searle, M.S.; Drug recognition and stabilisation of the parallel-stranded DNA quadruplex d(TTAGGGT)4 containing the human telomeric repeat; J. Mol. Biol., 2003, 334(1), 25–36. 103. Fedoroff, O.Y.; Salazar, M.; Han, H.; Chemeris, V.V.; Kerwin, S.M.; Hurley, L.H.; NMRBased model of a telomerase-inhibiting compound bound to G-quadruplex DNA; Biochemistry, 1998, 37(36), 12367–12374. 104. Haider, S.M.; Parkinson, G.N.; Neidle, S.; Structure of a G-quadruplex-ligand complex; J. Mol. Biol.; 2003, 326(1), 117–125. 105. Clark, G.R.; Pytel, P.D.; Squire, C.J.; Neidle, S.; Structure of the first parallel DNA quadruplex-drug complex; J. Am. Chem. Soc., 2003, 125(14), 4066–4067. 106. Cocco, M.J.; Hanakahi, L.A.; Huber, M.D.; Maizels, N.; Specific interactions of distamycin with G-quadruplex DNA; Nucleic Acids Res., 2003, 31(11), 2944–2951. 107. Reed, J.E.; Arnal, A.A.; Neidle, S.; Vilar, R.; Stabilization of G-quadruplex DNA and inhibition of telomerase activity by square-planar nickel(II) complexes; J. Am. Chem. Soc., 2006, 128(18), 5992–5993. 108. Bertrand, H.; Monchaud, D.; De Cian, A.; Guillot, R.; Mergny, J.L.; Teulade-Fichou, M.P.; The importance of metal geometry in the recognition of G-quadruplexDNA by metal-terpyridine complexes; Org. Biomol. Chem., 2007, 5(16), 2555– 2559. 109. Reed, J.E.; Arnal, A.A.; Neidle, S.; Vilar, R.; Stabilization of G-quadruplex DNA and inhibition of telomerase activity by square-planar nickel(II) complexes; J. Am. Chem. Soc., 2006, 128(18), 5992–5993. 110. Dixon, I.M.; Lopez, F.; Tejera, A.M.; Esteve, J.P.; Blasco, M.A.; Pratviel, G. et al.; A Gquadruplex ligand with 10000-fold selectivity over duplex DNA; J. Am. Chem. Soc., 2007, 129(6), 1502–1503. 111. Evans, S.E.; Mendez, M.A.; Turner, K.B.; Keating, L.R.; Grimes, R.T.; Melchoir, S. et al.; End-stacking of copper cationic porphyrins on parallel-stranded guanine quadruplexes; J. Biol. Inorg. Chem., 2007, 12(8), 1235–1249. 112. Keating, L.R.; Szalai, V.A.; Parallel-stranded guanine quadruplex interactions with a copper cationic porphyrin; Biochemistry, 2004, 43(50), 15891–15900. 113. Maraval, A.; Franco, S.; Vialas, C.; Pratviel, G.; Blasco, M.A.; Meunier, B.; Porphyrinaminoquinoline conjugates as telomerase inhibitors; Org. Biomol. Chem., 2003 1(6), 921–927. 114. Dixon, I.M.; Lopez, F.; Esteve, J.P.; Tejera, A.M.; Blasco, M.A.; Pratviel, G. et al.; Porphyrin derivatives for telomere binding and telomerase inhibition; Chembiochem., 2005, 6(1), 123–132.

232


115. Ren, L.; Zhang, A.; Huang, J.; Wang, P.; Weng, X.; Zhang, L. et al.; Quaternary ammonium zinc phthalocyanine: inhibiting telomerase by stabilizing G-quadruplexes and inducing G-quadruplex structure transition and formation; Chembiochem., 2007, 8(7), 775–780. 116. Zhang, L.; Huang, J.; Ren, L.; Bai, M.; Wu, L.; Zhai, B. et al.; Synthesis and evaluation of cationic phthalocyanine derivatives as potential inhibitors of telomerase; Bioorg. Med. Chem., 2008, 16(1), 303–312. 117. Rosenberg, B.; Van Camp, L.; Krigas, T.; Inhibition of cell division in Escherichia coli by electrolysis products from a platinum electrode; Nature, 1965, 205(4972), 698–699. 118. Wang, D.; Lippard, S.J.; Cellular processing of platinum anticancer drugs; Nat. Rev. Drug Discov., 2005, 4(4), 307–320. 119. Jung, Y.; Lippard, S.J.; Direct cellular responses to platinum-induced DNA damage; Chem. Rev., 2007, 107(5), 1387–1407. 120. Jamieson, E.R.; Lippard, S.J.; Structure, recognition, and processing of cisplatin-DNA adducts; Chem. Rev., 1999, 99, 2467–2498. 121. Kelland, L.; The resurgence of platinum-based cancer chemotherapy; Nat. Rev. Cancer, 2007, 7(8), 573–584. 122. Kieltyka, R.; Fakhoury, J.; Moitessier, N.; Sleiman, H.F.; Platinum phenanthroimidazole complexes as G-quadruplex DNA selective binders; Chemistry, 2007, 14(4), 1145–1154. 123. Reed, J.E.; Neidle, S.; Vilar, R.; Stabilisation of human telomeric quadruplex DNA and inhibition of telomerase by a platinum-phenanthroline complex; Chem. Commun. (Camb.), 2007, (42), 4366–4368. 124. Redon, S.; Bombard, S.; Elizondo-Riojas, M.A.; Chottard, J.C.; Platination of the (T2G4)4 telomeric sequence: a structural and cross-linking study; Biochemistry, 2001, 40(29), 8463–8470. 125. Redon, S.; Bombard, S.; Elizondo-Riojas, M.A.; Chottard, J.C.; Platinum cross-linking of adenines and guanines on the quadruplex structures of the AG3(T2AG3)3 and (T2AG3)4 human telomere sequences in Na+ and K+ solutions; Nucleic Acids Res., 2003, 31, 1605–1613. 126. Ourliac Garnier, I.; Bombard, S.; GG sequence of DNA and the human telomeric sequence react with cis-diammine-diaquaplatinum at comparable rates; J. Inorg. Biochem., 2007, 101(3), 514–524. 127. Bertrand, H.; Bombard, S.; Monchaud, D.; Teulade-Fichou, M.P.; A platinumquinacridine hybrid as G-quadruplex ligand; J. Inorg. Biol. Chem., 2007, 12, 1003–1014. 128. Williamson, J.R.; Raghuraman, M.K.; Cech, T.R.; Monovalent cation-induced structure of telomeric DNA: the G-quartet model; Cell, 1989, 59(5), 871–880. 129. Henderson, E.R.; Moore, M.; Malcolm, B.A., Telomere G-strand structure and function analyzed by chemical protection, base analogue substitution, and utilization by telomerase in vitro; Biochemistry, 1990, 29(3), 732–737. 130. Wang, Y.; Patel, D.J.; Solution structure of the Tetrahymena telomeric repeat d(T2G4)4 G-tetraplex; Structure, 1994, 2(12), 1141–1156. 131. Cuesta, J.; Read, M.A.; Neidle, S.; The design of G-quadruplex ligands as telomerase inhibitors; Mini Rev. Med. Chem., 2003, 3(1), 11–21. 132. Kasparkova, J.; Farrell, N.; Brabec, V.; Sequence specificity, conformation, and recognition by HMG1 protein major DNA interstrand cross-links of antitumor dinuclear platinum complexes; J. Biol. Chem., 2000, 275(21), 15789–15798. 133. Hounsou, C.; Guittat, L.; Monchaud, D.; Jourdan, M.; Saettel, N.; Mergny, J.L. et al.; Gquadruplex recognition by quinacridines: a SAR, NMR, and biological study; ChemMedChem., 2007, 2(5), 655–666. 134. Rao, L.; Bierbach, U.; Kinetically favored platination of adenine in the G-rich human telomeric repeat; J. Am. Chem. Soc., 2007, 129(51), 15764–15765. 135. Martin, R.B.; Platinum complexes: Hydrolysis and binding to N(7) and N(1) of purines. In: Cisplatin: Chemistry and Biochemistry of a Leading Anticancer Drug, Lippert, B., Ed., Wiley-VCH, Zürick, 1999, 183–206. 136. Burstyn, J.N.; Heiger-Bernays, W.J.; Cohen, S.M.; Lippard, S.J.; Formation of cis-diammi nedichloroplatinum(II) 1,2-intrastrand cross-links on DNA is flanking-sequence independent; Nucleic Acids Res., 2000, 28(21), 4237–4243.

References

233

137. Villa, R.; Folini, M.; Perego, P.; Supino, R.; Setti, E.; Daidone, M.G. et al.; Telomerase activity and telomere length in human ovarian cancer and melanoma cell lines: correlation with sensitivity to DNA damaging agents; Int. J. Oncol., 2000, 16(5), 995–1002. 138. Jeyapalan, J.C.; Saretzki, G.; Leake, A.; Tilby, M.J.; von Zglinicki, T.; Tumour-cell apoptosis after cisplatin treatment is not telomere dependent; Int. J. Cancer, 2006, 118(11), 2727–2734. 139. Opresko, P.L.; Fan, J.; Danzy, S.; Wilson, D.M., 3rd; Bohr, V.A.; Oxidative damage in telomeric DNA disrupts recognition by TRF1 and TRF2; Nucleic Acids Res., 2005, 33(4), 1230–1239. 140. Ourliac-Garnier, I.; Poulet, A.; Amiard, S.; Gilson, E.; Giraud-Panis, M.; Bombard, S.; Platination of telomeric DNA by cisplatin disrupts recognition by TRF2 and TRF1; manuscript in preparation 2008. 141. Chen, J.L.; Blasco, M.A.; Greider, C.W.; Secondary structure of vertebrate telomerase RNA; Cell, 2000, 100(5), 503–514. 142. Norton, J.C.; Piatyszek, M.A.; Wright, W.E.; Shay, J.W.; Corey, D.R.; Inhibition of human telomerase activity by peptide nucleic acids; Nat. Biotechnol., 1996, 14(5), 615–619. 143. Keppler, B.R.; Jarstfer, M.B.; Inhibition of telomerase activity by preventing proper assemblage; Biochemistry, 2004, 43(2), 334–343. 144. Keppler, B.R.; Jarstfer, M.B.; A high-throughput assay for a human telomerase proteinhuman telomerase RNA interaction; Anal. Biochem., 2006, 353(1), 75–82. 145. Herbert, B.S.; Gellert, G.C.; Hochreiter, A.; Pongracz, K.; Wright, W.E; Zielinska, D. et al.; Lipid modification of GRN163, an N3′–>P5′ thio-phosphoramidate oligonucleotide, enhances the potency of telomerase inhibition; Oncogene, 2005, 24(33), 5262–5268. 146. Gellert, G.C.; Dikmen, Z.G.; Wright, W.E.; Gryaznov, S.; Shay, J.W.; Effects of a novel telomerase inhibitor, GRN163L, in human breast cancer; Breast Cancer Res. Treat., 2006, 96(1), 73–81. 147. Colangelo, D.; Osella, D.; Telomerase inhibition and cancer: might platinum based drugs have a future as anti-telomerase pharmacological approach?; Curr. Med. Chem., 2005, 12(26), 3091–3102. 148. Colangelo, D.; Ghiglia, A.L.; Viano, I.; Cavigiolio, G.; Osella, D.; Cis-[Pt(Cl)2(pyridine)(5SO3H-isoquinoline)] complex, a selective inhibitor of telomerase enzyme; Biometals, 2003, 16(4), 553–560. 149. Furuta, M.; Nozawa, K.; Takemura, M.; Izuta, S.; Murate, T.; Tsuchiya, M. et al.; A novel platinum compound inhibits telomerase activity in vitro and reduces telomere length in a human hepatoma cell line; Int. J. Cancer, 2003, 104(6), 709–715. 150. Kim, N.W.; Piatyszek, M.A.; Prowse, K.R.; Harley, C.B.; West, M.D.; Ho, P.L. et al.; Specific association of human telomerase activity with immortal cells and cancer [see comments]; Science, 1994, 266(5193), 2011–2015. 151. Knox, R.J.; Friedlos, F.; Lydall, D.A.; Roberts, J.J.; Mechanism of cytotoxicity of anticancer platinum drugs: evidence that cis-diamminedichloroplatinum(II) and cis-diammine(1,1-cyclobutanedicarboxylato)platinum(II) differ only in the kinetics of their interaction with DNA; Cancer Res., 1986, 46(4 Pt 2), 1972–1979. 152. Ishibashi, T.; Lippard, S.J.; Telomere loss in cells treated with cisplatin; Proc. Natl. Acad. Sci. USA, 1998, 95(8), 4219–4223. 153. Zhang, R.G.; Zhang, R.P.; Wang, X.W.; Xie, H.; Effects of cisplatin on telomerase activity and telomere length in BEL-7404 human hepatoma cells; Cell Res., 2002, 12(1), 55–62. 154. Ishii, K.; Yang, W.L.; Cvijic, M.E.; Kikuchi, Y.; Nagata, I.; Chin, K.V.; Telomere shortening by cisplatin in yeast nucleotide excision repair mutant; Exp. Cell Res., 2000, 225, 95–101. 155. Brabec, V.; Kasparkova, J.; Modifications of DNA by platinum complexes. Relation to resistance of tumors to platinum antitumor drugs; Drug Resist. Updat., 2005, 8(3), 131–146. 156. Sedletska, Y.; Giraud-Panis, M.J.; Malinge, J.M.; Cisplatin is a DNA-damaging antitumour compound triggering multifactorial biochemical responses in cancer cells: importance of apoptotic pathways; Curr, Med. Chem. Anticancer Agents, 2005, 5(3), 251–265.

234


157. Burger, A.M.; Double, J.A.; Newell, D.R.; Inhibition of telomerase activity by cisplatin in human testicular cancer cells; Eur. J. Cancer, 1997, 33(4), 638–644. 158. Raymond, E.; Sun, D.; Izbicka, E.; Mangold, G.; Silvas, E.; Windle, B. et al.; A human breast cancer model for the study of telomerase inhibitors based on a new biotinylatedprimer extension assay; Br. J. Cancer, 1999, 80(9), 1332–1341. 159. Akiyama, M.; Horiguchi-Yamada, J.; Saito, S.; Hoshi, Y.; Yamada, O.; Mizoguchi, H. et al.; Cytostatic concentrations of anticancer agents do not affect telomerase activity of leukaemic cells in vitro; Eur. J. Cancer, 1999, 35(2), 309–315. 160. Akeshima, R.; Kigawa, J.; Takahashi, M.; Oishi, T.; Kanamori, Y.; Itamochi, H. et al.; Telomerase activity and p53-dependent apoptosis in ovarian cancer cells; Br. J. Cancer, 2001, 84(11), 1551–1555. 161. Ku, W.C.; Cheng, A.J.; Wang, T.C.; Inhibition of telomerase activity by PKC inhibitors in human nasopharyngeal cancer cells in culture; Biochem. Biophys. Res. Commun., 1997, 241(3), 730–736. 162. Spiropoulou, T.; Ferekidou, L.; Angelopoulou, K.; Stathopoulou, A.; Talieri, M.; Lianidou, E.S.; Effect of antineoplastic agents on the expression of human telomerase reverse transcriptase beta plus transcript in MCF-7 cells; Clin. Biochem., 2004, 37(4), 299–304. 163. Kondo, Y.; Kondo, S.; Tanaka, Y.; Haqqi, T.; Barna, B.P.; Cowell, J.K.; Inhibition of telomerase increases the susceptibility of human malignant glioblastoma cells to cisplatininduced apoptosis; Oncogene, 1998, 16(17), 2243–2248. 164. Ward, R.J.; Autexier, C.; Pharmacological telomerase inhibition can sensitize drugresistant and drug-sensitive cells to chemotherapeutic treatment; Mol. Pharmacol., 2005, 68(3), 779–786. 165. Folini, M.; De Marco, C.; Orlandi, L.; Daidone, M.G.; Zaffaroni, N.; Attenuation of telomerase activity does not increase sensitivity of human melanoma cells to anticancer agents; Eur. J. Cancer, 2000, 36(16), 2137–2145. 166. Chen, Z.; Koeneman, K.S.; Corey, D.R.; Consequences of telomerase inhibition and combination treatments for the proliferation of cancer cells; Cancer Res. 2003, 63(18), 5917–5925. 167. Buttitta, F.; Pellegrini, C.; Marchetti, A.; Gadducci, A.; Cosio, S.; Felicioni, L. et al.; Human telomerase reverse transcriptase mRNA expression assessed by real-time reverse transcription polymerase chain reaction predicts chemosensitivity in patients with ovarian carcinoma; J. Clin. Oncol., 2003, 21(7), 1320–1325. 168. Asai, A.; Kiyozuka, Y.; Yoshida, R.; Fujii, T.; Hioki, K.; Tsubura, A.; Telomere length, telomerase activity and telomerase RNA expression in human esophageal cancer cells: correlation with cell proliferation, differentiation and chemosensitivity to anticancer drugs; Anticancer Res., 1998, 18(3A), 1465–1472. 169. Lee, K.H.; Rudolph, K.L.; Ju, Y.J.; Greenberg, R.A.; Cannizzaro, L.; Chin, L. et al.; Telomere dysfunction alters the chemotherapeutic profile of transformed cells; Proc. Natl. Acad. Sci. USA, 2001, 98(6), 3381–3386.

8 Towards Photodynamic Therapy of Cancer with Platinum Group Metal Polyazine Complexes David F. Zigler and Karen J. Brewer

8.1 Introduction This chapter focuses on the current state of the art of platinum group polyazine chromophores, their photochemistry with DNA and phototoxicity toward intact cells. It is intended to provide the basic principles and concepts for a researcher entering the field of inorganic photodynamic therapy (PDT). Early on the reader will be introduced to the basic principles of electronic excited states and PDT. In keeping with the theme of this book, special focus will be given to Ru(II), Os(II) and Rh(III) polyazine photochemistry with DNA. Comprehensive reviews have been published on the photochemistry of monometallic polyazine complexes with DNA,1,2 nonplatinum pharmaceuticals,3,4 photochemical action of metallointercalators and metalloinsertors of DNA,5 copper complex photochemistry with DNA,6 photoactive metal complexes in medicine7 and bioinorganic photochemistry.8 Many reviews have been published assessing the mechanism of PDT action at the tissue, cellular and molecular levels.9–11 This chapter will serve as an addendum to all of these reviews and include discussion of multimetallic complexes with highlights of the most recent work in this rapidly expanding field. Multimetallic and supramolecular systems are particularly promising as PDT agents due to the ease of tuning properties and reactivity by structural modification. The chapter will be punctuated by cell studies with inorganic polyazine chromophores, offering a glimpse into the Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

236

Towards Photodynamic Therapy of Cancer

future of transition metal polyazine complexes as PDT agents. Polyazine metal complexes have properties that make them a promising class of future PDT agents, which will be highlighted herein. Photodynamic therapy is a modern treatment that uses light and a chromophoric molecule (PDT agent) to locally photosensitize tissues and to selectively kill targeted cells. PDT has potential applications in the treatment of localized tumours and intravascular diseases. Following administration of the PDT agent, the lesion of interest is irradiated with visible or near-infrared (NIR) light. Complicated photochemical reactions ensue, inducing apoptosis and/or necrosis, hopefully resulting in ablation of the lesion. PDT holds particular promise in the treatment of cancer as it tends to have fewer systemic toxic effects commonly associated with traditional radiation and chemotherapies. 8.1.1 Cancer Cancer is a disease caused by uncontrolled growth of abnormal cells, impacting the functioning of bodily tissues and organs. A mass of cancer cells, also known as a tumour or lesion, has many stages of growth. Initial mass formation is typified by rapid growth concomitant with poor vascularization within the lesion. Continued rapid growth results in the centre of the mass becoming hypoxic and undergoing necrosis. Abnormal cells may break off of the primary tumour and migrate throughout the body in a process called metastasis. Current cancer treatments are relatively crude, having severe side effects. These include surgery (invasive, risk of infection, complications), chemotherapy (poor cytospecificity/ selectivity, systemic toxicity effects) and radiation therapy (radiation sickness, increased risk of secondary tumours). Ideal treatments of cancer could be used in the ablation of lesions without the side effects which greatly impact the quality of life of the patient. 8.1.2 A Brief History of PDT and Current Clinical Treatments Photodynamic therapy has been expanding as a treatment of a wide range of localized maladies. Photodynamic action was first described in 1900 by Oscar Raab.12 In 1905 von Tappeiner and Jesionek reported the first successful clinical use of PDT to treat basal cell carcinoma with a topical application of eosin red and light.13 PDT studies were largely abandoned until the late 1970s, when animal trials using hematoporphyrin derivatives gave promising results in the treatment of animal cancers. Approval of the hematoporphyrin derivative, Photofin®, for use in PDT did not occur in the US until 1995 (Figure 8.1).14 Within the US, 5-aminolevulinic acid (Levulan®) is the only other approved chemical for use in PDT, gaining approval in 2000. Levulan® acts as a prodrug, enhancing porphyrin anabolism. The European Union, Norway and Iceland have approved Foscan®, a chlorin (chlorophyll derivative), for the treatment of head and neck cancers. At the time of writing there were three chromophores in current clinical trials: metoxifin lutetium (Lutex), tin ethyl

Introduction Me

R

Me Na O

Me Me

O H

N H N

237

N H N

N

N SnII N N

Me

O O

O NaO2C

n

hematoporphyrin derivative n = 1-4, Photofrin®

tin ethyl etiopurpurin Puryltintm OH

NH N

N

N

LuIII

N

HN

N

OH

tetraphenylchlorin Foscan® O

O

N

O

O

O

OMe

N

O

O

O

OMe

metoxifin lutetium Lutex

NH3+Cl-

OH

5-aminolevulinic acid Levulan® Figure 8.1 Chemical structures of compounds in clinical use and clinical trial for PDT in the United States15

etiopurpurin (Puryltin™) , and mono-L-aspartyl chlorine, with Lutex and Puryltin™ being the only inorganic centred chromophores in clinical trial.15 8.1.3 Characteristics of PDT Agents Ideal PDT agents absorb light strongly in the red or near-infrared (NIR) region of the spectrum, accumulate within the target tissue, efficiently perform the intended

238


action, have low systemic toxicity and are eliminated rapidly from the body. The red to NIR region of the spectrum (650–850 nm) is the light that has the greatest depth of penetration in human tissue and is known as the ‘therapeutic window’. Potent light absorbers with moderately long-lived and reactive electronic excited states are of particular interest. Chromophores with long-lived electronic excited states have a high probability of approach to a substrate during the excited state lifetime. Ironically, excited-state energy and lifetime often are inversely related, with longer wavelength light absorbers having short-lived excited states.16 Some current PDT agents show some cytospecificity, accumulating in tumour cells and enhancing the photosensitizer’s PDT function.14 Uptake of the PDT agent is often complicated and is impacted by the size of the chromophore, its charge and its lipophilicity. Chromophore size, charge and lipophilicity are important characteristics to consider in the design of PDT agents for elimination from the body. PDT agents that have hydrophobic architectures are retained for long periods within the body, leading to prolonged patient photosensitivity.15 Inorganic Complexes as PDT Agents Current PDT treatments employ organic chromophores as photosensitizers, but inorganic chromophores show great promise as PDT agents.1,8 The first approved PDT agent was a hematoporphyrin derivative, Photofrin® (Figure 8.1). Other photosensitizers have been approved, all of which are based on organic constructs.14 PDT has not gained widespread clinical use, owing in part to the newness of it as a therapy. In addition, the stigma of heavy-metal-containing compounds has led to limited use of these promising species as PDT agents. Two metal-containing photosensitizers are in clinical trials, metoxifin lutetium and tin ethyl etiopurpurin. Transition-metal-centred complexes with porphyrin, phthalocyanine, naphthalocyanine or polypyridine-type ligands have interesting ES properties, making them good targets for future therapies. This chapter will focus on inorganic systems with excited state properties and bioactivity that makes them of interest as future PDT agents.

8.1.4 DNA and PDT Action Deoxyribonucleic acid (DNA) is a natural supramolecule with complex secondary and tertiary structure whose integrity is central to the functioning of a cell (Figure 8.2). The importance of DNA to cellular function has directed research focus on transition metal complexes as potential PDT agents based on their ES reactions with DNA.2 Several structures of DNA have been reported with B-form DNA being the most prominent. A single turn of the B-form double helix is ca. 12 nm in length and averages 10.4 p-stacked base pairs.17 Two hydrophobic grooves run along the surface of the DNA molecule. The major groove is broad and shallow, and the minor groove is narrow and deep, both lined by the sugar phosphate backbone, giving DNA its polyanionic character. These structural features can be used in the rational design of DNA photomodification agents.

Study of Photophysics: Toward PDT 239

A H H N

T N

5' HO

N

O

O O P O O

3'

HN N O

O O P OO-

N N

O

N

O O P OO

N

H O

OO P O O

3'

N

N

O

-

O

H H

N

O

O

N

N H

NH

major groove

OH

5' C

minor groove

G

Figure 8.2 (Plate 10) Complimentary pairs of nucleotides that make up double helix DNA (Inset, A = adenine, T = thymine, G = guanine, C = cytosine, DNA = deoxyribonucleic acid). Atoms are carbon (grey), hydrogen (white), nitrogen (blue), oxygen (red) and phosphorous (pink) (See colour plate section)

8.2 Study of Photophysics: Toward PDT The discussion of transition metal polyazine photophysics requires explanation of general photophysical processes. Absorption of a photon by a chromophore causes an electronic transition and populates the chromophore’s electronic excited state. There are several internal pathways by which the excited chromophore might relax. The decay processes of electronically excited transition metal polyazine complexes have been the subject of study for decades, and continue to be a strong focus in research. Thorough studies of transition metal polyazine chromophores have led to many applications of the complexes in photochemistry. 8.2.1 Electronic Excitation Absorption of a photon of light by a chromophore in its electronic ground state (GS) is coincident with an electronic transition, populating an electronic excited state (ES) of the chromophore. The electronic transition promotes an electron from a low energy occupied orbital to an unoccupied orbital of higher energy (Figure 8.3). The lowest energy transition promotes an electron from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO). The orbitals involved in the transition must be electronically coupled to allow electronic transition. The probability of an electronic transition is governed by selection rules. The symmetry selection rule for intensity of an electronic transition is expressed in Equation (8.1):

240

Towards Photodynamic Therapy of Cancer high energy electron LUMO hν

E

low energy hole HOMO 1

GS

1

ES

Figure 8.3 The change in electronic configuration following excitation by light of a singlet ground state chromophore (1GS) to a singlet electronic excited state (1ES) (HOMO = highest occupied molecular orbital, LUMO = lowest unoccupied molecular orbital) ES GS f ∝ ∫ ψ e1 µ ψ e1 dv

(8.1)

where f, the oscillator strength, is proportional to the integral of the product of the GS and ES wavefunctions (y)and m, the electric dipole moment operator. The spin selection rule states that for an electronic transition to be allowed, the spin multiplicity of the ground state and excited state must be the same. Electronic transitions which are allowed by donor–acceptor orbital electronic coupling, symmetry and electron spin are the most intense, with extinction coefficients, e, of ≥103 M−1 cm−1. Transitions which are allowed by spin, but formally forbidden by symmetry have e between 100 and 102 M−1 cm−1. Truly spin-forbidden transitions are not observed. Relaxation of the selection rules can give rise to observation of transitions that are formally forbidden. In the presence of a heavy atom, spin–orbit coupling relaxes the spin-selection rule, giving rise to absorptions that are formally spin forbidden displaying e as high as 103 M−1 cm−1. Electronic transitions often are accompanied by vibronic excitation. The conversion between electronic energy surfaces, moving from the 1GS to the 1ES, is dictated by the Franck–Condon principle (Figure 8.4). Electronic transitions following absorption of a photon, hn, are rapid, populating the 1ES in a nuclear configuration consistent with the 1GS. The 1ES is populated in a hot vibronic state (nn, where n is an integer >0). Vibronic relaxation to the lowest vibronic states is rapid with a rate constant kvib > 1012 s−1. Depopulation of the 1ES to the 1GS occurs at varying rates, with rate constants of the decay processes greatly depending on the nature of the 1 ES. 8.2.2 Unimolecular Electronic Excited State Decay The quantum efficiency of ES processes is an important description of the relative kinetics of ES decay. Decay from the ES by a specific pathway, x, is quantified by the quantum yield, Fx. The quantum yield is the probability of a molecule in its excited state decaying by a specific pathway. Quantum yield is defined as the ratio of the observed first-order rate constant of the process of interest, kx, divided by the sum of the rate constants of all pathways depopulating that state, Σk, Equation 8.2:


kvib > 1012s–1

ν6 ν5

ν7

1

ES

ν4 ν3

ν2 ν1 ν0

E

hν

kr = 103–1010s–1

hν' ν6

ν7

1

GS

ν5

ν4 ν3 ν2 ν1 ν0

reGS

reES

r Figure 8.4 Morse potential energy surface diagram for electronic (1GS, 1ES) and vibrational (νn) states of a chromophore. Included are typical first-order rate constants. kvib = vibronic relaxation rate constant, kr = rate constant of emission of light with average energy hν′, r = internuclear distance, reGS = GS equilibrium internuclear distance, reES = ES equilibrium internuclear distance.

Φ x = Φ ES

kx ∑k

(8.2)

where FES is the quantum efficiency of population of the reactive ES. An important descriptor of the electronic excited state is its inherent lifetime, t0, simply defined as (Σk)−1. Conversion between electronic states of systems is typically represented with Jablonski or state diagrams. These diagrams are simplified versions of the Morse potential energy surfaces, as shown in Figure 8.4, where each electronic state is represented as a line instead of a surface. Decay processes of electronic excited states, ES, may occur by multiple pathways (Figure 8.5). Radiative processes are typically represented as straight arrows ( ) and nonradiative processes are presented as wavy arrows ( ). Internal conversion, kic, is nonradiative relaxation without change in electron spin, while intersystem crossing, kisc, is accompanied by a change in spin. A molecular orbital depiction of the processes is shown in Figure 8.6. Fluorescence, kf, is defined as emission of a photon and decay to the ground state without change in electronic spin. Radiative decay with a change in spin state is defined as phosphorescence, kp. It should be noted that several examples in the literature incorrectly define fluorescence or phosphorescence by the lifetime of the

242


1

ES

k isc 3

kf

E hν

ES

kp k ic k nr

1

GS

Figure 8.5 Jablonski diagram illustrating relative energies of states. Each bar ( ) represents the potential energy surface of an electronic state with lines above the state meant to reflect excited vibronic states. ES are excited states and GS is the ground state. Arrows represent electronic excited state decay via fluorescence (kf), internal conversion (kic), intersystem crossing (kisc), phosophorescence (kp) and nonradiative (knr) pathways

kisc

hν

E 1

GS

1

ES

3

ES

Figure 8.6 Orbital representation of intersystem crossing (kisc) from singlet excited state (1ES) to triplet excited state (3ES)

emitting ES. Nonradiative relaxation of an ES to the ground state is sometimes expressed by knr. 8.2.3 Electronic Excited States and Unimolecular Decay of Transition Metal Polyazine Transition metal polyazine complexes are potent light absorbers in the ultraviolet (UV) and visible region of the spectrum (Figure 8.7.)18 All polyazine complexes absorb UV light strongly due to intense internal ligand (IL) transitions of the aromatic ligands (Figure 8.8). IL transitions are called ligand centered (LC) in some literature. Ruthenium(II) and osmium(II) polyazine complexes absorb strongly in the visible region and have low-lying ES that are typically emissive, long-lived and reactive.19 Mononuclear Rh(III) polyazine complexes absorb strongly in the UV-A region due to IL transitions. The ES of rhodium(III) and dirhodium(II) polyazine complexes have reactivity that is strongly coupled to the type of polyazine ligands


np ns

t1u

a1g,t1u

σ*M

a1g

eg

σ*M

t1g,t2g, t1u,t2u π*

LF

(n-1)d

t2g,eg

t2g

IL

t1u,t2u,t1g,t2g a1g,t1u,eg

Metal Orbitals

L

t1u,t2u,t1g,t2g

π*

MLCT

πM

t1u,t2u,t1g,t2g

πL a1g,t1u,eg

σL

Molecular Orbitals

π

σ

Ligand Orbitals

Figure 8.7 Molecular orbital diagram of an octahedral metal complex, depicting common electronic transitions associated with transition metal (M) complexes with polyazine ligands (L) (IL = internal ligand, MLCT = metal to ligand charge transfer, LF = ligand field)

employed. Combining chromophores into mixed-metal supramolecules gives complexes with photophysical properties that are unique compared to the mononuclear analogues.20 Tuning the ES properties of the transition metal polyazine complexes allows for tuning of their photochemical properties. Table 8.1 summarizes the photophysical properties of transition metal polyazine complexes used in the study of photochemical degradation of DNA or light activated cell studies. Ruthenium(II) and Osmium(II) Polyazine Complexes Ru(II) and Os(II) complexes are interesting as photosensitizers as they possess reactive ES that are easily tuned by modification of their ligand coordination sphere.19 Mononuclear Ru(II) polyazine complexes typically have an absorption band in the 400–500 nm region due to the Ru(dp) → TL(p*) charge transfer transition (MLCT), where TL = terminal polyazine ligand (Figure 8.7). The electronic absorption spectrum of [Ru(bpy)3]2+ , bpy = 2,2′-bipyridine, is presented in Figure 8.9. The initially populated 1MLCT state undergoes intersystem crossing with unit efficiency (Fisc ≈ 1) to give the 3MLCT state (Figure 8.10).21 For tris(chelate) polyazine complexes of Ru(II), the 3MLCT state is long-lived at room temperature (t0 = 10– 1000 ns) and displays strong phosphorescence (Fp ≈ 0.10).19 The ES populated following metal-centred d → d excitation is the ligand field state (LF). Though higher in energy, the nonemissive 3LF state is thermally accessible at room temperature from the 3MLCT state. Thermal population of the 3LF state competes with phosphorescence or nonradiative decay pathways of the 3MLCT state as a decay pathway. Modification of the coordination environment of Ru(II) polyazine systems modulates the electronic properties of the chromophore. Bis(chelate) polyazine complexes of Ru(II) with a cis-Ru(TL)2 moiety tend to have absorption spectral

244


Terminal Ligands N

N N

N

N

N

N

N

N

N

N

NH

N

N

N

NH

N

N

dppz

phi

R

N

N

NH

N

N

R

tpy

Ph2phen, R = H dsdp, R = SO3-

N

N

N

N

N

N

N

N

N

N

chrysi

NH

phzi

N

N

N

N

PHEHAT

ddz

NH

NH

N

R

phen, R = H Me4phen, R = Me

N

HAT

R

N

R

N

TAP

bpy R

N

DAP

Bridging Ligands

N

N

N

N

N

N

bpm

N

N

N

N

N

N

N

N

N

dpp

N

dpq

dpb

Figure 8.8 Polyazine ligands commonly used in the synthesis of transition metal complexes for the photomodification of DNA

features similar to the tris(chelate) systems. Emission quantum efficiencies from the 3 MLCT states of these complexes, however, are often orders of magnitude smaller. For example, the Fp of [Ru(bpy)3]2+ at room temperature is 0.06 versus Fp = 0.002 reported for [(bpy)2Ru(NH3)2]2+ at 157 K.19,22 This reduction in emission for [(bpy)2Ru(NH3)2]2+ is due to the more rapid interstate conversion between the

– 610 596 614 743 645

636 – – – 610 – – 692 690 698 – – – – –

345, 490 452 462 – 478g 466 437 452 552g 588g 682g 460 – – 480 474 486 277c 358 355 355 375

cis-[(bpy)2Ru(NH3)2](BF4)2 [Ru(bpy)3]Cl2 [Ru(phen)3]Cl2 [Ru(Ph2phen)3]Cl2 [Os(bpy)3](PF6)2 [(phen)Ru(HAT)2]Cl2

[Ru(TAP)3]Cl2

[(dppz)Ru(HAT)2]Cl2 [(bpy)2Ru(DAP)](PF6)2 [Ru(DAP)3](PF6)2 [Os(DAP)3](PF6)2 [(bpy)2Ru(ddz)]Cl2 [(phen)2Ru(ddz)]Cl2 Na4[Ru(dsdp)3] [(bpy)2Ru(dpp)]Cl2 [(phen)2Ru(dpp)]Cl2 [(Ph2phen)2Ru(dpp)]Cl2 [Rh(phen)3]Cl3 [(phen)2Rh(phi)]Cl3

cis-[(phen)2RhCl2]Cl cis-[(Me4phen)2RhCl2]Cl cis-[(5,6-Me2phen)2RhCl2]Cl

602

em (RT ) λmax (nm)

Complexb

abs λmax (nm)

723 704 710

– 698 727 763 – – – – – – 567 – – – –

1.090 – – – 0.180h 0.210h 3.600 – – – 4,100 –

–

0.052f 0.385 0.421 0.909 0.060 0.835

741f – – – 710 – –

t (RT) (µs)

em λmax (77 K )c (nm)

MLCT

3

LF LF 3 LF

3

3

MLCT, MLCT 3 MLCT 3 MLCT 3 MLCT 3 MLCT 3 MLCT 3 MLCT 3 MLCT 3 MLCT 3 IL 3 IL

3

3

3

LF MLCT 3 MLCT 3 MLCT 3 MLCT 3 MLCT

3

Lowest Energy ESd PB-M Type II-PC Type I & II-PC – – PB-L,Type I-PO Type II-PC PB-L,Type I-PO Type II-PC Type I-PO Type II-PC – – Type II-PC Type II-PC Type II-PC Type II-PC Type II-PC Type II-PC Type I-PC PB-M, Type I-PC PB-M PB-M PB-M

DNA Photochemistrye

30 66 66

72 77 77 77 76,81 76 76 75 75 75 26,73 58

50,67

22 2,56 2,56,74 2,56 24 72

Ref.

Table 8.1 Photophysical properties of some mononuclear transition metal polyazine complexes used in DNA photomodification schemesa


– – – – – – – – –

432 434 432 542 525 540 540 615 525

cis-[Rh2(m-O2CMe)2(dppz) (h1-O2CMe)(MeOH)](O2CMe) cis-[Rh2(m-O2CMe)2(dppz)2](O2CMe)2 cis-[Rh2(m-O2CMe)2 (dppz)(bpy)](O2CMe)2 [{(bpy)2Ru(dpp)}2 Ru(dpp)PtCl2](PF6)6 [{(bpy)2Ru(dpp)}2RhCl2](PF6)5 [{(bpy)2Os(dpp)}2RhCl2](PF6)5 [{(tpy)RuCl(dpp)}2RhCl2](PF6)3 [{(bpy)2Ru(bpm)}2RhCl2](PF6)5 [{(bpy)2Ru(dpp)}2IrCl2](PF6)5 – – – – –

–

– –

– – – – –

–

– –

–

–

554, 710i –

t (RT) (µs)

em λmax (77 K )c (nm)

3

MMCT MMCT 3 MMCT 3 MLCT 3 MLCT

3

MLCT

3

3

MMLCT MMLCT

3

MMLCT

3

IL, 3LF

3

Lowest Energy ESd

PC PC PC – –

Type II-PC

Type I & II-PC Type I & II-PC

Type I-PC, PB-M Type I & II-PC

DNA Photochemistrye

43 44 44 43 43

78

33 83

33

65

Ref.

b

Absorption maxima and room temperature emission maxima, reported in aqueous solutions, unless otherwise noted. bpy = 2,2′-bipyridine; phen = 1,10-phenanthroline; Ph2phen = 4,7-diphenyl-1,10-phenanthroline; dppz = dipyrido[3,2-a:2′,3′-c]phenazine; HAT = 1,4,5,8,9,12-hexaazatriphenylene; TAP = 1,4,5,8-tetraazaphenanthrene; DAP = 1,12-diazapyrylene; ddz = dibenzo[h,j]dipyrido[3,2-a:2′,3′-c]phenazine; dsdp = 4,7-di(3-sulphurylphenyl)-1,10-phenanthroline; dpp = 2,3-bis(2-pyridyl)pyrazine; phi = 9,10-phenanthrenequinone diimine; Me4phen = 3,4,7,8-tertramethyl-1,10-phenanthroline; 5,6-Me2phen = 5,6-dimethyl-1,10-phenanthroline, tpy = 2,2′:6′,2″-terpyridine; bpm = 2,2′-bipyrimidine c In 4:1 EtOH/MeOH d LF = ligand field, MLCT = metal-to-ligand charge transfer, IL = internal ligand, MMLCT = metal-metal to ligand charge transfer, MMCT = metal to metal charge transfer 5 PB-L = photobinding to ligand, PB-M = photobinding to metal, Type I = direct ES & DNA reaction following electron transfer, Type II = photosensitized oxygen mediated reactions, PC = photocleavage, PO = photooxidation d 157 K. g In room temperature acetonitrile h Multiexponential, reported as the average of three components i Emission is observed from both 3LC and 3LF states in 4 : 1 EtOH/MeOH glass

a

–

em (RT ) λmax (nm)

380

abs λmax (nm)

cis-[(phen)(dppz)RhCl2]Cl

Complexb

Table 8.1 Continued

246 Towards Photodynamic Therapy of Cancer

Study of Photophysics: Toward PDT 247 8

bpy(π→π*)

ε x10-4 (M-1cm-1)

7

N

6

N

5

N RuII N

N N

4 3 2

Ru(dπ)→bpy(π*) CT

1 0 200

300

400

500

600

Wavelength (nm)

Figure 8.9 Electronic absorption spectrum of tris(2,2′-bipyridine)ruthenium(II), [Ru(bpy)3]2+, in CH3CN at RT

1

IL

∆E (x103 cm-1)

30

kic

25 20

1

kisc

MLCT

3

LF

3

MLCT knr hν' kp hν

15 10

knr'

5 0

1

GS

Figure 8.10 Jablonski-type diagram of [Ru(bpy)3]2+. bpy = 2,2′-bipyridine, 1GS = singlet electronic ground state, 1MLCT = singlet metal to ligand charge transfer excited state, 3MLCT = triplet MLCT excited state, 1IL = singlet internal ligand excited state, 3LF = triplet ligand field excited state. Relative energies are adapted from Juris, Balzani, Barigletti, Campagna and Belser.19

3

MLCT state and lower-lying 3LF state (Figure 8.11). Excitation directly into higher LF states has been linked to photoaquation of complexes with the cis-RuII(TL)2 moiety, where complexes such as [(bpy)2RuCl2] photochemically generate [(bpy)2RuCl(OH2)]+ and [(bpy)2Ru(OH2)2]2+.23 Tris(chelate) polyazine complexes of Os(II) generally absorb and emit visible light, but at longer wavelengths and with decreased emission efficiency than the Ru(II)-centred analogs.24,25 A shorter 3MLCT state lifetime (and smaller Fp) is a result of several contributing factors. The energy gap law states that as the difference in energy of two electronic states decreases, the vibronic coupling of the states increases, enhancing knr and leading to more efficient ES deactivation.24 Another factor to consider is the large spin–orbit coupling of Os(II) versus Ru(II). A larger spin–orbit coupling constant relaxes the spin selection rule, increasing the intensity of spin forbidden transitions. This gives rise to a low energy ‘tail’ of the visible region

248

Towards Photodynamic Therapy of Cancer 1

∆E (x103 cm-1)

30 25 20

1 1

IL LF

kisc'

3

kisc

MLCT

3

3

MLCT knr hν' k

15

nr'

hν

10

kp

LF2 LF1

knr''

krxn

krxn'

5 1

0

GS

Figure 8.11 Jablonski-type diagram of [(bpy)2Ru(NH3)2]2+. bpy = 2,2′-bipyridine, 1GS = singlet electronic ground state, 1MLCT = singlet metal to ligand charge transfer excited state, 3 MLCT = triplet MLCT excited state, 1IL = singlet internal ligand excited state, 3LF = triplet ligand field excited state. Relative energies are adapted from Singh and Turro22

1

IL

∆E (x103 cm-1)

30 kic

25 20

1

MLCT

15

3

LF

kisc 3

hν' hν

10

hν'' kp

MLCT

knr

5 0

1

GS

Figure 8.12 Jablonski-type diagram of [Os(bpy)3]2+. bpy = 2,2′-bipyridine, 1GS = singlet electronic ground state, 1MLCT = singlet metal to ligand charge transfer excited state, 3MLCT = triplet MLCT excited state, 1IL = singlet internal ligand excited state. Relative energies are adapted from Kober, Caspar, Lumpkin and Meyer24

electronic absorption spectra of Os(II) polyazine complexes. The low energy tail is due to direct population of the 3MLCT state (Figure 8.12).24 The larger spin–orbit coupling constant also enhances decay of the 3MLCT state to the 1GS. Unlike the Ru(II) counterparts, Os(II) polyazine 3LF states are not thermally accessible from the 3MLCT state. Therefore, excitation of the Os(II) chromophore with visible light generally does not result in ligand substitution. Chromophores based on Ru(II) and Os(II) polyazine complexes have found great utility in the study of DNA photomodification. Rhodium Polyazine Complexes Rhodium polyazine complexes have distinct, reactive electronic excited states that make them attractive as PDT agents. As with other transition metal polyazine complexes discussed in this chapter, the ligand environment around the central rhodium has major impact on the ES properties. Mononuclear tris(chelate) Rh(III) polyazine complexes are potent light absorbers in the UV region of the spectrum due to strong

Study of Photophysics: Toward PDT 249 1

35

IL

kisc

∆E (x103 cm-1)

30

3

IL

25 hν

20

kp

15

knr

10 5 1

0

GS

Figure 8.13 Jablonski-type diagram of [Rh(phen)3]3+. phen = 1,10-phenathroline, 1GS = singlet electronic ground state, 1IL = singlet internal ligand excited state. Relative energies are adapted from Crosby and Elfring26

1

kisc'

IL LF

30

∆E (x103 cm-1)

1

25

3

IL

kisc

20 15

hν

0

hν'

knr

5 1

LF2

3

LF1

kic kp

10

3

krxn

GS

Figure 8.14 Jablonski-type diagram of [(phen)2RhCl2] +. phen = 1,10-phenathroline, 1GS = singlet electronic ground state, 1IL = singlet internal ligand excited state, 3LF = triplet ligand field excited state. Relative energies are adapted from Demas and Crosby27

IL transitions. These Rh(III) complexes lack the visible light absorption bands as their MLCT transitions occur at much higher energy than analogous Ru(II) and Os(II) systems (Figure 8.13).26 Coincidently, the complexes emit from a 3IL state. Bis(chelate) polyazine complexes of Rh(III) with a cis-RhIIIX2 moiety, where X is a halide, also absorb only in the UV and are emissive in low temperature glasses.27 A weak absorption band observed between 350–400 nm arises from LF transitions for the [(TL)2RhX2]+ systems. The emission from these complexes, when excited into the IL or LF bands, is broad and significantly Stokes shifted relative the lowest absorption band (∼14 000 cm−1).28 The initially populated 1LF state intersystem crosses to an upper 3LF2 state (Figure 8.14). The 3LF2 state decays by internal conversion to the emissive 3LF1 state. At room temperature, photoaquation (krxn) competes with nonradiative decay to give the mono- and disubstituted solvato complexes, [[(TL)2RhX(S)]2+ and [(TL)2Rh(S)2]3+ (S = solvent).23,29 Morrison and coworkers have exploited this decay pathway to photobind to DNA, mimicking the

250


∆E (x103 cm-1)

30 25 20

1

IL

kisc

MMLCT

kisc'

3

MM-MM

kisc"

3

1 1

15

3

MM-MM

10

hν knr

5 0

IL MMLCT krxn

1

krxn'

GS

Figure 8.15 Jablonski-type diagram of cis-[Rh2(m-O2CCH3)2(dppz)2]2+. dppz = dipyrido[3,2a:2′,3′-c]phenazine, 1GS = singlet electronic ground state, 1IL = singlet internal ligand excited state, 1MMLCT = singlet metal–metal to ligand charge transfer state, 1MM-MM = Rh-Rh(dp*dΣ*) state. Relative energies are adapted from Angeles-Boza, Bradley, Fu, Wicke, Bacsa, Dunbar, Turro33,82

DNA thermal binding properties of the well known complex cis-[Pt(NH3)2Cl2] (cisplatin).30,31 Dirhodium(II) complexes possess unique ES compared to the mononuclear Rh(III) complexes due to the strong electronic coupling of the two metal centres. Complexes of the form [Rh2(m-O2CCH3)4] absorb weakly in the near UV region owing to a low lying Rh-Rh(dp*) → Rh-Rh(ds*) (MM-MM) transitions (Figure 8.15).32 Inclusion of ligands with stabilized p*-acceptor orbitals, e.g. dibenzo[h,j]dipyrido[3,2-a:2′,3′-c]phenazine (dppz), gives complexes with transitions at ∼430 nm attributed to IL transitions and Rh-Rh(dp*) → dppz(p*) charge transfer transitions (MMLCT).33 The 3MMLCT state is thought to be a long-lived state with an oxidized dirhodium core. These complexes display DNA photooxidation reactions, consistent with reaction of the 3MMLCT state. Excited States of Polyazine Bridged Ru–Rh Complexes Polyazine bridged, mixed-metal supramolecules have been developed that incorporate multiple metal centres and have more complicated photophysics than the mononuclear analogues.34 The mixed-metal complexes often retain some ES properties of the subunits used to assemble the supramolecule. Metal centres with strong electronic coupling, e.g. di-Rh(II) complexes above, have electronic transitions that involve orbitals delocalized over the metal centres. At the other extreme, metal centres that are covalently linked, but without strong electronic coupling, behave like the mononuclear analogues. Ru(II)–Rh(III) dyads linked via an alkyl tether have electronic properties consistent with each subunit, indicating weak electronic interaction between coupled chromophores.35–39 The polyazine bridged complexes have intercomponent interactions, giving rise to states that are unique to the supramolecular complex without full delocalization of orbitals. Mixed-metal supramolecules that have weakly to moderately coupled metal centres exhibit unique ES with interesting ES dynamics, giving rise to useful pho-


1

∆E (x103 cm-1)

30

IL kic

25 20 1MLCT 15

kisc 3

hν

10

MLCT kp

ket knr

5 0

3

krxn

MMCT

knr'

1

GS

Figure 8.16 Jablonski-type diagram of [{(bpy)2Ru(dpp)}2RhCl2]5+. bpy = 2,2′-bipyridine, dpp = 2,3-bis(2-pyridyl)pyrazine, 1GS = singlet electronic ground state, 1MLCT = singlet metal to ligand charge transfer, 3MLCT = triplet MLCT, 1IL = singlet internal ligand excited state. Relative energies adapted from Molnar, Jensen, Vogler, Jones, Laverman, Bridgewater, Richter and Brewer.41

tophysical properties. Metals bridged by 2,3-bis(2-pyridyl)pyrazine (dpp) electronically couple through the aromatic pyrazine ring.40 Mononuclear Ru(II) complexes with dpp have a strong, visible MLCT transition and a LUMO localized on the abs pyrazine ring of the dpp ligand (e.g. [(bpy)2Ru(dpp)]2+, λmax = 480 nm in water). Bridging the complex to another metal centre stabilizes the dpp(p*) LUMO, shifting abs the MLCT transition to lower energy (e.g. [(bpy)2Ru(dpp)Ru(bpy)2]4+, λmax = 526 nm ). II III Trimetallic complexes with Ru (dpp) subunits bridged to a cis-Rh Cl2 centre such as [{(bpy)2Ru(dpp)}2RhCl2]5+ have interesting properties (Figure 8.16).41 The HOMO of these trimetallic complexes is localized on the Ru(dp) orbitals like other Ru(II) polyazine complexes, but the LUMO is Rh(ds *) in nature. Direct Ru(dp) → Rh(ds*) (metal-to-metal) charge transfer (MMCT) excitation is forbidden by symmetry and the relatively weak HOMO–LUMO electronic coupling. Irradiation with visible light, however, populates the 1MLCT state. Intersystem crossing generates the 3 MLCT state, which is higher in energy than the 3MMCT state.42 Coupling of the dpp(p*) and Rh(ds *) acceptor orbitals facilitates intramolecular electron transfer to give the 3MMCT state. These trimetallic complexes have been shown to photocleave DNA.43,44 Substitution of the bridging ligand dpp for molecular bridges with lower energy acceptor orbitals decreases the photochemical reactivity, implicating the 3MMCT state.44 8.2.4 Bimolecular Excited State Interactions Ground state light absorbers (LA) and their electronic excited states (*LA) are interesting for study, but it is the interaction of *LA with substrate molecules that interests PDT researchers. Bimolecular interactions resulting in quenching of *LA following excitation (Equation 8.3) and competing with unimolecular radiative decay (Equation 8.4) and nonradiative decay (Equation 8.5) can be classified into basic groups: bimolecular deactivation, knr ′ (Equation 8.6); excited state electron transfer, ket, to (Equation 8.7) or from (Equation 8.8) *LA; excited state energy

252


transfer, ken (Equation 8.9); and photochemical reaction, krxn (Equation 8.10); where Q = quencher, ED = electron donor, EA = electron acceptor. Ia LA + hν  → *LA

(8.3)

kr *LA  → LA + hν ′

(8.4)

knr *LA  → LA

(8.5)

knr′ *LA + Q  → LA + Q

(8.6)

ket *LA + ED  → LA − + ED+

(8.7)

ket′ *LA + EA  → LA + + EA −

(8.8)

ken *LA + Q  → LA + *Q

(8.9)

krxn *LA + reactant  → photoproducts

(8.10)

Current research efforts toward PDT agents study the photochemistry of an excited chromophore with DNA. Transition metal polyazine complexes have a rich history of photochemistry with DNA, focusing strong research efforts. These efforts have given rise to recent discoveries of previously unknown photochemistry. The field is expanding rapidly to accommodate increasing interest in DNA photochemical modification and inorganic-based PDT. In addition, information about fundamental photophysical properties is being uncovered. Excited State Electron Transfer Theory Reductive and oxidative quenching of *LA is driven by the relative redox potentials of the ED or EA and *LA. The *LA reduction E(*LAn/LAn−1) and *LA oxidation E (*LAn/LAn+1) potentials are estimated with the following equations: E(*LA n LA n−1 ) ≈ E(LA n LA n−1 ) − ∆E 0−0

(8.11)

E ( *LA n LA n+1 ) ≈ E ( LA n LA n+1 ) + ∆E 0−0

(8.12)

where E(LAn/LAn−1) and E(LAn/LAn+1) are the respective first reduction and oxidation potentials in V of the GS chromophore, and ∆E0−0 is the difference of energy between lowest vibronic states of the GS and ES in eV (see Figure 8.4). The ∆E0−0 em is commonly estimated from the λmax with a Stokes shift applied to account for the shift of the excited state energy surface relative to the ground state. The rate of ES quenching is governed by the driving force of electron transfer as described by Marcus.45 For a bimolecular reaction under pseudo-first order conditions:

[ ED]o or [ EA ]o >> [ ES]


the observed rate constant of electron transfer (kobs,et) is defined as: kobs,et = vNκ exp ( −∆G‡ RT )

(8.13)

where vN is the nuclear frequency factor and k is the electron transfer coefficient. The free energy of activation, ∆G‡, is expressed as:

∆G‡ =

λ ∆G0    1 + 4 λ 

2

(8.14)

where l is the reorganizational energy and DG0 is the change in standard free energy of redox reactants and products. Experimentally ket can be determined by Stern– Volmer kinetic analysis as described below. Type I Photooxidation Reactions in PDT Reductive quenching of an excited photosensitizer by DNA constitutes a Type I photooxidation reaction, as defined by Foote.46 For example, consider the mechanism of decay of the 3MLCT state of [(phen)Ru(HAT)2]2+ in the presence of DNA, phen = 1,10-phenanthroline, HAT = 1,4,5,8,9,12-hexaazatriphenylene. 2+ a → 1 * [( phen) Ru ( HAT )2 ] [( phen) Ru ( HAT )2 ]2 + + hv I 1

3

kisc 2+  → 3 * [( phen) Ru ( HAT )2 ]

(8.16)

knr 2+  → [( phen) Ru ( HAT )2 ]

(8.17)

kp 2+  → [( phen) Ru ( HAT )2 ] + hv ′

(8.18)

2+

* [( phen) Ru ( HAT )2 ] 3

2+

* [( phen) Ru ( HAT )2 ] 2+

* [( phen) Ru ( HAT )2 ]

(8.15)

2+ 2+ q * [( phen) Ru ( HAT )2 ] + DNA k → [( phen) Ru ( HAT )2 ] + DNA

(8.19)

2+ + et * [( phen) Ru ( HAT )2 ] + DNA k → [( phen) Ru ( HAT )2 ] + DNA +

(8.20)

Kinetically the quantum yield of the electron transfer reaction (Equation 8.20) is expressed as:

Φ et = Φ isc

ket [ DNA ] kp + knr + ( ket + kq ) [ DNA ]

(8.21)

where Fet is the quantum yield of electron transfer, Fisc is the efficiency of population of the 3MLCT state, kp and knr are the respective unimolecular phosphorescence and nonradiative decay rate constants, ket is the rate constant of electron transfer, kq is rate constant of bimolecular deactivation and [DNA] is the effective concentration of the DNA quencher in molarity of base pairs. Efficiency and rates of excited state reactions can be determined by probing the quantum yield of phosphorescence Fp:

Φ po = 1 + τ o (ket + kq ) [ DNA ] Φp

(8.22)

254


with Fp being the quantum yield of phosphorescence in the presence of quencher and Φ p0 is the phosphorescence yield in the absence of quencher. Stern–Volmer analysis consists of plotting Φ p0 Φ p as a function of [DNA] times the ES lifetime t0 in the absence of quencher, providing a linear relationship with a slope equal to kSV, where kSV = ket + kq. In the absence of measurable ES emission, ket may be determined by observing the changing in ES lifetime as measured by transient absorption spectroscopy because:

Φ p0 Φ p = τ τ 0

(8.23)

where t is the excited state lifetime in the presence of quencher. Transient spectroscopy can differentiate electron transfer quenching and other decay pathways probing the formation of the electron transfer product. Measuring the rate of formation of the product can also allow for the determination of ket. Excited State Energy Transfer Theory Two mechanisms of excited state energy transfer (Equation 8.9) have been described and involve either ES dipole resonance coupling or electron exchange between the *LA donor and quencher, Q (Equation 8.9). These two mechanisms have been described individually by Förster47 and Dexter.48 Förster described *LA quenching by excited state energy transfer through space by a Q.47 Resonant coupling of the chromophore and quencher excited state dipoles allow for radiationless energy transfer from the *LA donor to the acceptor Q. Förster energy transfer efficiency is governed by the spectral overlap of *LA emission and Q absorption, interchromophore distance and relative dipole orientation. Förster excited state energy transfer is the dominant pathway when *LA → LA and Q → *Q transitions are spin allowed. Förster excited state energy transfer is the governing principle of fluorescence resonance energy transfer (FRET) techniques used in confocal microscopy. Another mechanism of excited state energy transfer, described by Dexter, involves electron exchange between the *LA and Q.48 Electron exchange requires some electronic coupling of the *LA donor and the Q acceptor orbitals. Dexter transfer is not bound by the spin selection rule and is thought to be the dominant mechanism of *LA energy transfer when there is poor spectral overlap of *LA emission and Q absorption, e.g. photosensitization of 3O2 → 1O2. Study of excited state energy transfer quenching dynamics can be performed using Stern–Volmer kinetic analysis. Details of this analysis are outlined above for excited state electron transfer. In some instances, observation of the rise-time and/or decay of the *Q state can also be observed and aids in understanding this bimolecular interaction for a specific system. Type II Photooxidation Reactions in PDT Type II photooxidation reactions of biomolecules, as defined by Foote,46 involves photosensitization of molecular oxygen (3O2), generating singlet oxygen (1O2) which reacts with a biologically important substrate molecule. Photosensitization of 1O2

Photochemical Reactions of Metal Complexes with DNA

255

from 3O2 requires excited state energy transfer by the Dexter mechanism. The 3 MLCT state of [Ru(phen)3]2+, has been shown to produce 1O2 from 3O2 (Equation 8.28).49 a → [Ru ( phen)3 ]2 + + hv I 1

* [ Ru ( phen)3 ]

2+

3

3

3

* [ Ru ( phen)3 ]

1

* [ Ru ( phen)3 ]

2+

(8.24)

kisc 2+  → 3 * [ Ru ( phen)3 ]

2+

knr  →

(8.25)

[Ru ( phen)3 ]2 +

(8.26)

2+ 2+ p *[ Ru ( phen )3 ] k → [ Ru ( phen )3 ] + hv′

2+ ken * [ Ru ( phen)3 ] + 3 O2  →

[Ru ( phen)3 ]2 + +

(8.27) 1

O2

(8.28)

knr′ → 3 O2 O2 

(8.29)

kp′ O 2  → 3 O2 + hv ′′

(8.30)

krxn DNA + 1 O2  → DNA photoproduct

(8.31)

1

1

Using the steady-state approximation on the [1O2], the quantum efficiency of oxygen mediated photomodification of DNA (Frxn) can be described by: ken [ 3 O2 ] krxn [ DNA ]    Φ rxn = Φ isc   kp + knr + ken [ 3 O2 ]   knr′ + kp′ + krxn [ DNA ] 

(8.31)

where Fisc is the quantum yield of population of the 3MLCT state, ken is the rate constant for ES energy transfer to oxygen, 3O2, dissolved in solution and krxn is the rate constant for the reaction of 1O2 with DNA, where [DNA] is expressed as concentration of DNA base pairs. The rate constant ken is determined experimentally by measuring emission from the 3MLCT state of [Ru(phen)3]2+ in the absence of DNA, at different concentrations of 3O2. The rate constant krxn can be elucidated from DNA photoproduct analysis or by observing 1O2 emission decay.

8.3 Photochemical Reactions of Metal Complexes with DNA DNA plays important roles in the functioning of cells because it is central to protein expression (transcription) and cell replication. Damage to DNA manifests itself in many forms, some of which are detrimental to the integrity of the data stored on the DNA. Damage to DNA disrupts cell function by causing carcinogenesis, mutagenesis, apoptosis and cell death. There are several reported mechanisms of excited state photosensitizer reactions inducing DNA damage.2,5,50–53 Careful design of DNA photochemical experiments can yield surprising amounts of information about the mechanism of DNA/metal complex interaction as well as information about the

256


chromophore’s excited state dynamics. Characteristic changes to DNA are observed for photobinding of metal complexes, photooxidation of nucleobases and photocleavage of the sugar phosphate backbone. Knowledge of the decomposition pathways of DNA photolyzed with a metal complex helps to understand a possible mechanism of PDT action. 8.3.1 DNA Targets and Analysis Study of the photochemical reactions of metal complexes with DNA requires careful choice of the type of DNA used and method of analysis. Helpful descriptions of DNA structure and analysis can be found in the book by Calladine.17 Mono- and oligonucleotides are used as targets when chemical analysis techniques such as mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy are used to study the photoproducts. Mononucleotides are available commercially and have found utility in experiments to understand binding of metal complexes to DNA and ES quenching studies. Oligonucleotides (oligos) are short strands of DNA (15–100 base pairs) that have known sequence and structural properties. Oligos are better models for nuclear DNA because they can be obtained as double stranded DNA. High-resolution poly(acrylamide) gel electrophoresis (HR-PAGE) is instrumental in understanding sequence specificity of photochemical reactions with oligonuclear DNA leading to irreversible changes. Circular plasmid DNA extracted from bacteria is obtained with a supercoiled tertiary structure that makes this type of DNA an excellent target for photomodification schemes.54,55 Changes to the topographical nature of supercoiled circular plasmid DNA are indicative of several decomposition pathways that can be analysed using agarose gel electrophoresis (Figure 8.17). Finally, genomic DNA is commercially available as a freeze dried extract from various tissue sources and is high molecular weight, with > 106 base pairs per molecule. Genomic DNA is used to understand DNA–metal interactions that require large volumes and/or high DNA concentrations for analysis using selective precipi-

λ Cir Lin 23.1 kbp9.4 kbp6.6 kbp4.4 kbp-

Form II

2.3 kbp2.0 kbp-

Form III Form I

Figure 8.17 Photograph of a typical 0.8% agarose gel illustrating the electrophoretic migration (moving from top to bottom) of three different forms of plasmid DNA. The lanes from left to right are as follows: l is a lambda DNA/HindIII digest molecular weight marker (kbp = 1,000 base-pairs), Cir is circular plasmid DNA containing native supercoiled (Form I) and open-circular (Form II) plasmid DNA, and Lin is linear plasmid DNA (Form III). Band assignments are based on Vinograd and Lebowitz55

Designing Transition Metal Polyazines for Photomodification

257

tation, changes to spectroscopy and viscosity measurements. Each form of DNA has specific advantages and disadvantages toward use in photochemical experiments and multiple types are required to get a more complete understanding of the metal complex–DNA photochemical mechanism.

8.4 Designing Transition Metal Polyazines for DNA Photomodification Photomodification of DNA by transition metal polyazine complexes is attractive because of the structural diversity of this class of complexes.18,19 Two major design considerations for chromophores applicable as DNA photomodification agents are: the impact of structural changes on the interaction of the chromophore with DNA, and the impact of chromophore structure on the efficiency and mechanism of their DNA photochemistry. Recent reports present photochemical mechanisms that help elaborate previously described photochemistry (Table 8.1) and have prompted cell studies to explore application in PDT. 8.4.1 Ground State Interactions of Transition Metal Polyazine Complexes and DNA The mechanism of DNA modification by photosensitizers is often multistep with the first step involving some form of metal complex–DNA association.2,50,51 Smaller, compact polyazine transition metal complexes tend to have an electrostatic interaction with DNA.56 Lipophilic substituents on the polyazine ligands give complexes that bind in the hydrophobic minor or major groove of the DNA double helix.57 Ligands with extended, planar aromatic substituents favour intercalation between DNA bases.5,51 More recently, metal complexes have been shown to insert between strands of the double helix, a behaviour that is previously not described.51 Binding of transition metal polyazine complexes to DNA is enantioselective, a reflection of the chiral nature of DNA and the often chiral nature of polyazine chelated metal complexes. Barton has pioneered exploiting the different binding modes as shapeselective probes of DNA and RNA structure.5,51,58,59 The strength of the DNA–metal complex interaction has a direct impact on the photosensitizer function as a DNA modification tool. Electrostatic interaction of cationic transition metal polyazine complexes with DNA can be described by the polyelectrolyte approximation of counterion condensation theory.60 DNA is polyanionic with a single negative charge per nucleotide in aqueous solution. Usually Na+ acts as the counterion to the DNA negative charge. Transition metal polyazine complexes studied for DNA photomodification can have large negative to large positive charges. Exchange efficiency of Na+ at the DNA surface by an incoming complex ion is a factor of the charge of the complex, the concentration of free metal complex and Na+ concentration (ionic strength). Complexes with large positive charges that are structurally compact have stronger electrostatic interactions with DNA. The prototypical transition metal polyazine complex [Ru(bpy)3]2+ exhibits ionic binding to DNA.56

258


Groove binding of transition metal polyazine complexes to DNA is controlled by complex lipophilicity and molecular shape. Also known as surface binding, groove binding is association of a species with the hydrophobic pocket formed between the sugar-phosphate backbones of each strand. The complex [Ru(Me4phen)3]2+ is a wellknown groove binder.57 The methyl substituents on the phen ligand contribute to the complex’s overall lipophilicity, but provide steric bulk that prevents more intimate interactions with the DNA. Extended, planar polyazine ligands that can intercalate into the base stack of DNA have been shown to enhance metal complex–DNA electronic interactions prior to photolysis.5,51 Two ligands often included in a metal complex’s architecture that promote intercalation are dppz and 9,10-phenanthrenequinone diimine (phi). Complexes with bulky planar ligands have been reported to bind by an alternative mechanism. The complex [(phen)2Rh(chrysi)]3+ reported by Barton and coworkers inserts its a-chrysenequinone diimine (chrysi) ligand into the bases stack accompanied by displacement of a base pair. Photocleavage of DNA by metalloinsertor complexes was shown to be an effective method of identifying base mismatches in the DNA.61 It is important to note that the mode of DNA–metal complex interaction prior to photolysis is not always clear. The reader is referred to a review by Kirsch DeMesmeaker, Lecompte and Kelly for a historical account of ‘The controversial case of [Ru(phen)3]2+.2 Hexacoordinate transition metal complexes with two or three bidentate ligands tend to have strongly enantioselective binding interactions with DNA.59 DNA in its most common B form is a right-handed double helix (Figure 8.2), interacting more strongly with ∆ enantiomers of tris(chelate) metal complexes. Barton has shown that the complex ∆-[Ru(Ph2phen)3]2+ binds strongly in the major groove of the doublehelix, intercalating the base pairs. The other enantiomer Λ-[Ru(Ph2phen)3]2+ does not bind as strongly, owing to the steric repulsion of the bulky ancillary ligands with the sugar-phosphate backbone of DNA provided by the stereochemistry around the metal centre.

8.4.2 Mechanisms of DNA Photochemical Degradation Photochemical degradation of DNA has been pursued in recent years toward applications in PDT, and DNAase and RNAase biomimetics. The mechanisms of DNA degradation under photolytic or oxidative stress have been thoroughly reviewed, so minimal explanation of these pathways is given below.52,53,62 Photoinitiated binding of metal complexes has been studied, coined ‘photo-cisplatin’ agents. Photobinding through an ancillary ligand also leads to DNA covalent modification. Oxidation of the DNA bases by ES electron transfer reaction and hydrogen atoms abstraction are important to the photochemical function of some transition metal polyazine complexes. Photocleavage of DNA by the transition metal polyazine ES in the presence or absence of dissolved 3O2 reflects separate mechanisms of action. Many of the transition metal polyazine complexes studied for their photochemical interaction with DNA exhibit multiple ES reactions with DNA.


259

Design Aspects: Toward Photobinding DNA Metal complexes can be designed to bind DNA after excitation. The variety of ES and their photochemistry provide a range of mechanisms for DNA photobinding. Photobinding DNA through a metal centre. Ruthenium(II) complexes of the form cis-[(TL)2RuX2]n+, where X = halide (n = 0) and X = H2O, NH3 or aliphatic amine (n = 2), have been investigated for their DNA thermal binding properties and more recently as DNA photobinding agents.3 The complexes cis-[(bpy)2RuCl2] and cis-[(bpy)2Ru(OH2)2]2+ readily covalently bind to DNA under physiological conditions. Photobinding to DNA of cis-[(bpy)2RuCl2] and cis-[(bpy)2Ru(OH2)2]2+ shows little enhancement over thermal binding pathways.63 Singh and Turro reported the DNA photobinding properties of cis-[(bpy)2Ru(NH3)2]2+, but have shown that the complex did not appreciably thermally bind DNA.22 The mononuclear complex absorbs strongly at 290 nm and 243 nm, consistent with IL transitions. abs MLCT transitions in the visible region occur at λmax = 490 nm . The complex ex undergoes photoaquation when irradiated with near-UV light (λmax = 350 nm or 400 nm, Faq = 0.024 ± 2 or 0.018 ± 2). Excitation with light > 450 nm, however, did not lead to observable chemical reaction. The authors indicate the photoaquation takes place from a 3LF state that is thermally inaccessible from the 3MLCT state. A photoproduct of 400 nm excitation of cis-[(bpy)2Ru(NH3)2]2+, cis-[(bpy)2Ru(OH2)2]2+, has been demonstrated to undergo ligand substitution with mononucleotides and was shown to bind readily to double stranded (ds) and single stranded (ss) DNA. The authors predict that the interacting electronic excited states can be tuned by modifying the complex structure to favour photoaquation following visible light excitation. Octahedral polyazine complexes of Rh(III) with a cis-dihalide moiety are also known to undergo ligand substitution upon photolysis with near-UV light.23,29 Morrison and coworkers have studied the DNA photobinding properties of cis[(TL)2RhCl2]+.30,64–66 They demonstrated that [(phen)2RhCl2]+, when irradiated with 308 nm light, photobound to ss- or dsDNA with Fb ≈ 0.006 and 0.001 respectively.30 Dark controls confirm that electronic excitation of the metal complex is required for [(phen)2RhCl2]+ to bind DNA. The authors have also proposed a photooxidation step is necessary to initiate the sequence giving the photosubstitution products.64 Previous assignment of the photolabile state of [(phen)2RhCl2]+ was to the lowest 3 LF state.29 Exchanging phen for methyl substituted 4,7,7,8-tetramethyl-1,10-phenanthroline (Me4phen) or 5,6-dimethyl-1,10-phenanthroline (5,6-Me2phen) had varying impact on the photochemistry.66 The lowest absorption band of [(5,6Me2phen)2RhCl2]+ is red shifted compared to both [(phen)2RhCl2]+ and [(Me4phen)2RhCl2]+. Enhanced photoaquation for [(Me4phen)2RhCl2]+ (Faq = 0.63) compared to [(phen)2RhCl2]+ (Faq = 0.03) when photolysed with 347 nm light was seen. A similar complex, [(phen)(dppz)RhCl2]+ was shown to both photobind and photocleave DNA when irradiated with 311 nm light. The planar dppz ligand is known to facilitate intercalative binding of the metal complex to DNA. The Faq of [(phen)(dppz)RhCl2]+, when irradiated with 355 nm light is reported to be 0.068, reduced compared to [(phen)2RhCl2]+ (Faq = 0.087). Competitive decay to a low lying n → p* state localized on the dppz ligand is possible in the dppz-containing

260

Towards Photodynamic Therapy of Cancer H2N

O HN N N

N

N N

N Ru N

N

HN N N

N

N

N R

O

N

N

N

[(TAP)2Ru(TAP-NH-dGMP)]

N N N

N

N

N Ru N

N N

N N

R

N N

N

N N

N

[(HAT)(phen)Ru(HAT-O-dGMP)]

Figure 8.18 Photoadducts of guanisine monophosphate (dGMP) and different Ru(II) polyazine complexes (TAP = 1,4,5,8-tetraazaphenanthrene, HAT = 1,4,5,8,9,12-hexaazatriphenylene, phen = 1,10-phenanthroline, R = 1′-deoxyribose-5′-phosphate)50,72

complex. The authors noted photocleavage by [(phen)(dppz)RhCl2]+ may occur through a low lying 3IL state. Photobinding DNA through a coordinated ligand. Another mechanism has been implicated in photobinding of the metal complex to DNA through covalent attachment to an ancillary ligand. Moucheron, Kirsch-De Mesmaeker and Kelly have reviewed an extensive series of ruthenium(II) complexes containing polyazaaromatic ligands that are reductively quenched by guanine.50 Excitation of [Ru(TAP)3]2+ , TAP = 1,4,5,8-tetraazaphenanthrene, with visible light gives rise to weak emission from the 3MLCT state. Reductive quenching was observed for dGMP (ket = 2.2 × 109 M−1 s−1), calf thymus (CT) DNA and poly(dG-dC). DNA photoproduct analysis by gel electrophoresis indicated retarded migration of the DNA relative to the nonphotolysed sample. This result is indicative of increased size of the DNA and/or decrease in the overall charge. Change in size and charge would be expected in the event of metal-complex binding to DNA. The metal complex–DNA photoproduct was digested into nucleotides and studied by 1H-NMR spectroscopy and MS. The authors determined the major product was metal complex covalently bound through the 2-carbon of a TAP ligand at the C3-amine of guanine (Figure 8.18). The authors concluded that the complex must bind to the minor groove of DNA to allow for formation of the observed photoadduct. More recently, photoadduct formation during photolysis of DNA with [(phen)Ru(HAT)2]2+ was reported.67 The analogous 2-carbon on the ancillary ligand is again bound to guanine, but at the 5C-oxo site, forming the ether linked photoadduct. The large planar HAT ligands intercalate the DNA base stack in the hydrophobic major groove, lending to preferential formation of the ether linked photoadduct. Design Aspects: Toward Photocleavage of DNA Photoinitiated strand scission of DNA can occur due to excited state electron transfer, hydrogen atom abstraction from a sugar and by reaction of DNA with 1O2.2,52,53,62 The details of these DNA cleavage mechanisms have been outlined above.


261

The ability of an excited state chromophore to oxidize DNA is a reflection of the redox properties of DNA bases and the excited chromophore. Guanine has been identified as the DNA nitrogenous base with the strongest reducing power [E(G/ G+) ≈ 0.92 V versus Ag/AgCl].68–71 When incorporated in dsDNA, oxidation occurs preferentially at the 5′-G of a multi-G stack, suggested to result from a destabilization of the 5′-G donor orbitals, shifting the oxidation potential (E0) of 5′-GGG-3′ to ca. 0.42 V (versus Ag/AgCl).62 Radical cations of the oxidized bases decompose by multiple pathways, including cleavage of the sugar phosphate backbone and base-labile nucleotide oxidation products.52 The chromophore in its ES or an unstable redox state may also be involved in cleavage schemes. Hydrogen atom abstraction from a sugar of DNA by *[(phen)2Rh(phi)]3+ has been proposed to initiate a reaction cascade ending with single-strand scission.58 Early studies have indicated that the 3MLCT states of [Ru(bpy)3]2+ under anoxic and hypotonic conditions are quenched coincidentally with single-strand cleavage of circular plasmid DNA, Frxn = 1.2 × 10−6 when [DNA]/ [Ru(bpy)3]2+ = 9. Recently the photochemistry of the complex [(TAP)2Ru(dppz)]2+ with DNA was reported, showing mixed photochemical pathways.72 The authors noted a mechanism by which the complex photobinds to DNA through a TAP ligand as described above for [Ru(TAP)3]2+. Interestingly, the dppz complex has a strong emission from a long-lived 3MLCT state in room temperature and in aqueous solution (t0 = 1.09 µs). For other Ru(II)-polyazine complexes with the dppz ligand, emission in water is notably weak due to low lying n → p* 3IL state(s) that quenches the emission from the 3MLCT state, possibly through hydrogen bonding of the water to the lone pairs on the nitrogens on the dppz ligand. Intercalation into DNA, however, protects the nitrogens on dppz from water, decreasing quenching of the 3 MLCT state and greatly enhancing emission. Ru(II)-complexes with dppz are often called ‘DNA light switches’.73 The emission of [(TAP)2Ru(dppz)]2+ is efficiently ‘switched off’ upon binding to calf-thymus DNA (t = 60 ns), a very unusual result. Oxygen mediated photocleavage of DNA. Photocleavage of DNA represents a major research effort in the study of DNA photomodifications with application to the development of new PDT agents. Several mechanisms of DNA photocleavage by platinum group metal polyazine complexes have been proposed that require 3O2. DNA photocleavage was reported for [Ru(TL)3]2+, TL = bpy, phen, in aerated solution .74 Mongelli, Brewer and coworkers reported a series of mononuclear Ru(II) complexes containing a potential molecular bridge, [(TL)2Ru(dpp)]2+, where TL = bpy, phen or Ph2phen.75 Enhancement of DNA photocleavage was noted for complexes with more lipophilic terminal ligands, [(Ph2phen)2Ru(dpp)]2+ being the most active, consistent with earlier reports by Barton.56 Hergueta-Bravo, Orellana and coworkers investigated the 3O2 dependence of DNA photocleavage by a series of Ru(II) complexes incorporating the intercalating ligand dibenzo[h,j]dipyrido[3,2a:2′,3′-c]phenazine (ddz).76 The 3MLCT of [(bpy)2Ru(ddz)]2+ state is a less efficient photosensitizer of 1O2 generation relative to [Ru(bpy)3]2+. Enhanced DNA photocleavage by [(bpy)2Ru(ddz)]2+, however, was noted. The increase in DNA photocleavage by [(bpy)2Ru(ddz)]2+ was attributed to strong intercalative binding to the DNA prior to photolysis. A similar observation was made for the complex [(bpy)2Ru(DAP)]+, DAP = 1,12-diazaperylene.77

262


The coupling of Ru polyazine chromophores to a cis-PtIICl2 site provides a forum to attach a DNA photocleavage unit to a DNA covalent binding unit. This attachment provides a means to deliver the Ru polyazine chromophore to the DNA target. An example of such a complex, [{(bpy)2Ru(dpp)}2Ru(dpp)PtCl2]6+, binds to DNA through the Pt(II) centre and photocleaves DNA via an oxygen-mediated pathway.78 Oxygen independent photocleavage of DNA. Swavey, Holder and Brewer examined the DNA photocleavage ability of polyazine bridged mixed-metal supramolecules. The supramolecules were constructed with two Ru(II) or two Os(II) light absorber subunits connected to a central cis-RhIIICl2 moiety.42,44 A wide array of supramolecules constructed with a dpp molecular bridge produced nicks in supercoiled circular plasmid DNA with >450 nm light photolysis. The gel showing the photocleavage of DNA by a member of this series, [{(bpy)2Ru(dpp)}2RhCl2]5+, is shown in Figure 8.19. The mechanism of DNA photocleavage is unclear, but this complex functions as a DNA photocleavage agent in the presence or absence of dissolved 3O2. DNA photocleavage is proposed to result from intramolecular ES electron transfer quenching of the 3MLCT state by the cis-RhCl2 center. The resultant metal to metal charge transfer (3MMCT) state has a formally oxidized Ru centre and a formally reduced Rh centre, both of which are capable of mediating DNA cleavage in the absence of 3O2. Importantly, nicks in the DNA backbone caused by most types of photocleavage often are irreparable by the repair enzyme DNA ligase.53

no MC MC MC +hν

N

N II

N

Cl

Ru N N

N

Cl III

N

Rh N N

N

N

N RuII N

N N

N

MC = [{(bpy)2Ru(dpp)}2RhCl2]Cl5

Form II Form I λ

pUC18 DNA

Array of 5W LEDs

Figure 8.19 Photograph of electrophoresis gel showing the efficient photocleavage of pUC18, supercoiled circular plasmid DNA (Form I), with the metal complex [{(bpy)2Ru (dpp)}2RhCl2]Cl5, generating open circular pUC18 DNA (Form II) [bpy = 2,2′-bipyridine, dpp = 2,3-bis(2-pyridyl)pyrazine]. DNA was photolysed with 460 nm centred light from a 5W LED (inset image). From left to right the lanes are: lambda DNA/ HindIII digest molecular weight marker (λ), pUC18 DNA without treatment with metal complex (no MC), pUC18 DNA incubated with metal complex in the dark at 5:1 base pairs to metal complex (MC), and pUC18 DNA photolysed in the presence of metal complex at 5:1 base pairs to metal complex (MC+hν)

Cell Studies with Metal Complexes 263

8.5 Cell Studies with Metal Complexes Cellular uptake and photoinitiated cell death is indicated for several metal polyazinebased light absorbers. Given the short timeframe of this research effort, this is a very promising result, with no strong understanding of means to design systems that will transport into the nucleus. Several groups indicate uptake of the metal complex in the cell to be a function of the ancillary ligands lipophilicity.1,66,79 Metal complexes representing different constructs outlined in preceding sections are promising as PDT agents. Surprisingly few cell studies with Ru, Os or Rh polyazine chromophores have been performed, given the breadth of knowledge of the photochemistry of these complexes with DNA. Cell studies of some Ru, Os and Rh polyazine complexes have revealed their excellent phototoxicity and minimal dark toxicity.33,65,66,79–84 Table 8.2 is a synopsis of the cell studies with transition metal polyazine complexes. Table 8.2 Transition metal polyazine complexes used in cell studies with applications in PDT Complexa [Ru(bpy)3]Cl2 and [Ru(phen)3]Cl2 [(bpy)2Ru(ddz)]Cl2 [(TL)2Ru(dppz)]Cl2TL = bpy, phen, mcmb, Ph2phen

cis-[(TL)2RhCl2]Cl TL = phen, 5,6-Me2phen, Me4phen, Ph2phen cis-[(phen)(dppz)RhCl2]Cl

[(phen)(chrysi)Rh(dmb(arg)8-DYE)](O2CCF3)11 cis-[Rh2(µ-O2CMe)2(dppz) (bpy)](O2CMe)2 [{(bpy)2M(dpp)}2RhCl2]Cl5, M = Ru or Os

Description of Interaction

Ref.

• Cellular uptake following photolysis • Phototoxic at concentrations >10−4 M • Selective uptake and luminescence enhancement by nonviable cells • Cell photokilling not investigated • Cellular uptake as a function of lipophilicity, size and overall charge • Faster uptake for complexes with lipophilic ligands (e.g. TL = Ph2phen) • Cell photokilling not investigated • Uptake investigated as function of lipophilicity • Uptake observed only for TL = Me4phen • Phototoxic to tumour cell lines at 55 µM and irradiation with >400 nm light • Penetrates cellular membranes and viral coats • Phototoxic to tumour cells at 40 µM and irradiation with 311 nm light • Deactivates viral DNA within intact virus • Metal complex with octa(arginine) tail and fluorescent dye • Rapid uptake observed at 5 µM • Cell photokilling not investigated • Dark toxicity reported LC50 = 208 µM • Phototoxic LC50 = 44 µM with 30 min, 400– 700 nm irradiation • No dark toxicity observed (tested up to 120 µM) • Phototoxic at 5 µM, with 4 min , 400–1000 nm irradiation

80 81

79

66

65

82

33,83

84

a bpy = 2,2′-bipyridine; phen = 1,10-phenanthroline; mcmb = 4-carboxy-4′-methyl-2,2′-bipyridine, Ph2phen = 4,7diphenyl-1,10-phenanthroline; dppz = dipyrido[3,2-a:2′,3′-c]phenazine; 5,6-Me2phen = 5,6-dimethyl-1,10-phenanthroline; Me4phen = 3,4,7,8-tertramethyl-1,10-phenanthroline; chrysi = a-chrysenequinone diimine; ddz = dibenzo [h,j]dipyrido[3,2-a:2′,3′-c]phenazine; dpp = 2,3-bis(2-pyridyl)pyrazine.

264


8.5.1 Cellular Uptake and Phototoxicity of Transition Metal Polyazine Complexes Recent studies of the interaction of platinum group polyazine complexes with cells indicate that uptake of metal complexes by cells is greater with a compromised cellular envelope. Dobrucki investigated [Ru(bpy)3]Cl2 and [Ru(phen)]Cl2 as confocal microscopy imaging dyes of J144 mouse macrophages and normal human fibroblasts (NFB).80 The metal complexes were internalized by the cells with endocytosis. Neither metal complex penetrated intact plasma membranes under dark conditions. Visible light photolysis of cells with internalized metal complex resulted in redistribution of the observed luminescence within the cell. Dobrucki suggests that Type II photooxidation of the cell membrane compromises its integrity, allowing passage of the metal complexes. Photoinduced cell death was noted for concentrations of [Ru(phen)3]2+ greater than 0.2 mM and irradiation with 490 nm light. JimenezHernandez, Orellana, et al. studied [(bpy)2Ru(ddz)]2+ as a selective stain of nonviable rat hepatocytes.81 A commonly used technique to image nonviable cells with fluorescence microscopy uses intercalation of the ethidium homodimer ion into DNA, which greatly enhances ethidium emission. This provides red fluorescent imaging of the nonviable cells. The planar aromatic ddz ligand of [(bpy)2Ru(ddz)]2+ readily intercalated DNA, also exhibiting a potent DNA light switch effect. Upon passing through the leaky membrane of dead cells, [(bpy)2Ru(ddz)]2+ is rapidly bound by nuclear DNA to give an emissive label of nonviable cells. A metal complex bound to an octa(arginine) tether [(arg)8] exhibited enhanced cellular uptake and association with nuclear DNA.82 Brunner and Barton studied the cellular uptake of [(phen)(chrysi)Rh{dmb-(arg)8-DYE}](O2CCF3)11, where DYE is a fluorescein or thiazole orange fluorescent tag. Uptake of the fluorescein dyelabelled metal complex was studied by confocal microscopy of metal complex bound HeLa cells. Confocal microscopy images of cells incubated with 5 µM showed that the dye was rapidly incorporated into the nucleus. The authors suggest that the design of the studied complexes could be adapted to build and understand the cellular uptake mechanisms of other metal complexes. Mononuclear ruthenium(II) and rhodium(III) polyazine complexes exhibit cellular uptake mechanisms related to the lipophilicity of the metal complex. Puckett and Barton studied the cellular uptake of a series of [(TL)2Ru(dppz)]Cl2 complexes, with TL = bpy, phen, Ph2phen, and 4-carboxy-4′-methyl-2,2′-bipyrindine (mcmb).79 Uptake was observed using confocal microscopy with direct excitation of the Ruex centred chromophore λmax = 488 nm . The complex [(Ph2phen)Ru(dppz)]Cl2 displayed the largest mean luminescence intensity inside the cell. The authors related uptake to the octanol/water partition coefficient of the metal complexes, suggesting a passive transport mechanism. Minimal emission was observed within the nucleus, suggesting that transport across the nuclear envelop is retarded. Loganathan and Morrison reported a similar enhanced cellular uptake with increasing lipophilicity for cis-[(TL)2RhCl2]Cl as a function of TL methylation.66 The authors report that [(Me4phen)2RhCl2]Cl, at 55 µM and with >400 nm light irradiation, caused a 20% reduction in cell growth relative to dark controls. The authors were unable to assess the mechanism of cell killing, but noted [(Me4phen)2RhCl2]Cl to be an efficient photobinder of DNA. Another DNA photobinding complex, [(phen)(dppz)RhCl2]Cl, was reported to be phototoxic at 40 µM with 311 nm irradiation.65 This complex was

Cell Studies with Metal Complexes 265

also shown to deactivate the DNA of a virus, indicating transport through intact viral envelope. 8.5.2 Polymetallic Complex Phototoxicity Studies Four recent examples of multimetallic polyazine complexes, in two separate studies, were reported to have a dramatic increase in toxicity upon photolysis. Angeles-Boza, Dunbar, Turro and coworkers studied the dirhodium(II) complexes, cis-[Rh2(m-O2 CMe)2(dppz)(TL)](O2CMe)2, where TL = MeOH and MeCOOH; dppz; or bpy.33,83 There was little enhancement of toxicity toward human skin fibroblasts upon photolysis in the presence of cis-[Rh2(m-O2CMe)2(dppz)(MeOH)(h1- O2CMe)](O2CMe). The dark LC50 values for cis-[Rh2(m-O2CMe)2(dppz)2](O2CMe)2 and cis-[Rh2(m-O2 CMe)2(dppz)(bpy)](O2CMe)2 were 140 µM and 210 µM, respectively. This is slightly less toxic than cisplatin (LC50 = 131 µM) and much less toxic than hematoporphyrin under the same conditions (LC50 = 20 µM). Each complex displayed comparable phototoxicity (LC50 ≈ 40 µM) with >395 nm light irradiation for 15 min. The complex cis-[Rh2(m-O2CMe)2(dppz)(bpy)](O2CMe)2 had the greatest difference between light and dark toxicity. Holder, Storrie, Brewer et al. investigated the phototoxicity of mixed-metal supramolecules of the form [{(bpy)2M(dpp)}2RhCl2]Cl5, where M = Ru or Os.84 No appreciable dark toxicity was observed in Vero cells for any concentration of either metal complex, up to 120 µM. Excess metal complex was washed from the cells prior to photolysis. Irradiation with > 460 nm for 5 min resulted in dramatic decrease in cell viability (Figure 8.20). As the concentration of [{(bpy)2Ru(dpp)}2RhCl2]Cl5 in N

N RuII N

N N

N

Cl

Cl

N

RhIII N

N N

N

N RuII N

N N

N

MC = [{(bpy)2Ru(dpp)}2RhCl2]Cl5 Cells treated with MC

illumination area

growth

after 48 hr growth period

calcein AM imaging of live cells

ethidium imaging of dead cells

Figure 8.20 (Plate 11) Micrographs of Vero cells pretreated with [{(bpy)2Ru(dpp)}2RhCl2]C l5, rinsed, and illuminated with 400–1000 nm light showing the high level of light-activated cell killing [bpy = 2,2′-bipyridine, dpp = 2,3-bis(2-pyridyl)pyrazine]. (From left to right) immediately after photolysis (light exposure within circle); after 48 hour growth period; live cell (green) visualized with calcein AM fluorescent dye; dead cell (red) visualized with ethidium homodimer-1 fluorescent dye84 (See colour plate section)

266


media used to pre-treat the cells was increased from 5 to 120 µM, the number of viable cells following photolysis approached zero. The Os complex, [{(bpy)2Os(dpp)}2 RhCl2]Cl5, exhibited decreased phototoxicity relative to the Ru complex, with the ratio of cells present before and after photolysis decreasing toward unity as a function metal complex concentration. The difference in phototoxicity is quite surprising given the similar DNA photoreactivity. This may be due to the lifetime of the reactive state of each complex, with the Os systems providing shorter ES lifetimes. The cells studies of these metal complexes show great promise. Neither study of multimetallic complexes could comment on the mechanism of cell killing. Both report the photosensitizers of interest to be efficient DNA photocleavage agents.33,43,44,83

8.6 Conclusions and Future Directions Platinum group metal polyazine complexes have pronounced DNA photochemistry and show unusual promise as photodynamic therapy (PDT) agents. Recent discoveries contribute to the understanding of the electronic excited state interactions of these systems with DNA. Studies of photoactivated interactions of platinum group complexes with biomolecules also provide insight into the basic photophysics of these metal complexes. DNA can be a good probe of the excited state dynamics of metal complexes. Polyazine complexes of transition metal centres, especially ruthenium and osmium, exhibit strong light absorbing properties in the UV and in the visible region, with interesting photoreactivity. Excitation populates reactive electronic excited states with properties very different than the ground state molecules. Ruthenium(II) polyazine complexes are noted visible light absorbers with typically long-lived Ru(dp) → dpp(p*) charge transfer (3MLCT) states known to photosensitize 1O2. Osmium(II) polyazine complexes have similar light absorbing properties to ruthenium(II), but with electronic transitions shifted to lower energy. The phosphorescence of 3MLCT states provides a convenient handle to reactivity of these 3MLCT states. Rhodium(III) polyazine complexes, though poor visible light absorbers, have interesting reactive internal ligand (3IL) and ligand field (3LF) electronic excited states that photocleave and photobind DNA. The studies to date provide significant motivation for continued study of Rh(III) complexes. Polyazine complexes incorporating dirhodium(II) centres possess electronic excited states arising from the delocalized metal–metal(dp*) highest occupied molecular orbital, giving them potent photochemical reactivity. Polyazine bridged complexes of ruthenium(II) and rhodium(III) are both visible light absorbers and possess the reactive Rh(III) centre. Systems can be designed to have a lowest lying Ru(dp) → Rh(dσ*) charge transfer (3MMCT) state that is reactive with, among other substrates, DNA. Structural modifications of these assemblies provide a unique forum to modulate properties. The forum of photomodification of DNA by metal polyazine complexes has been developing over the last three decades. Researchers in the field have discovered multiple modes of photochemical action, including metal complex mediated

Acknowledgements

267

photooxidation of DNA bases, photobinding metal polyazine chromophores to DNA and photocleavage of the DNA backbone by the excited inorganic chromophore. These studies show both the complexity of the interactions of metal complexes with DNA in the excited state and the potential promise of these complexes in PDT. The excited states of metal polyazine complexes are easily tuned by modification of the ligand environment, as is tuning of ground state interactions of these systems with DNA. Metal complexes that associate with DNA to a greater extent prior to photolysis tend to have enhanced efficiency of DNA photomodification. Given the typically low concentration of DNA, metal complex–DNA interactions are key to photoreaction efficiency. Systems incorporating extended planar, aromatic, polyazine ligands or systems containing lipophilic polyazine ligands generally interact strongly with DNA. The studies of metal complexes and their light activated modification of DNA show the promise of metal complexes in this forum. The studies summarized herein illustrate examples of a variety of photoreactions with DNA. Many systems react with DNA by yet to be understood mechanisms. Surprisingly few studies of platinum group metal polyazine complexes with cells have been undertaken and it is clear that much is to be learned in this forum. The utility of metal complexes as chromophores in photoinitiated cell killing schemes remains clear. The photochemical mechanisms of platinum group metal polyazine complexes leading to cell death have yet to be explored. The more detailed studies of DNA photomodification noted for these systems suggest this as a possible mechanism of cell photokilling. Cellular uptake mechanisms of metal polyazine complexes are currently being investigated to understand the relationship between ligand lipophilicity and uptake. While Barton and coworkers show that rational design strategies can be used to develop platinum group metal polyazine complexes which readily enter cells, many surprises are being uncovered. Studies of multinuclear polyazine complexes illustrate the utility of this class of photosensitizer in future photodynamic therapy schemes. The low dark toxicity and high light activated cell killing makes these supramolecular systems particularly promising to develop new, potent PDT agents. Continuing studies within the field of platinum group metal polyazine complexes as PDT agents, will ultimately lead to their clinical use in the treatment of cancer.

Acknowledgements Special thanks to Prof. Brenda S.J. Winkel for her frequent questions, discussions and collaborations on studies of mixed-metal supramolecules and their photochemistry with DNA. The authors would especially like to thank the National Science Foundation for their generous support of this work (CHE-0408445). The authors thank Theralase, Inc. for continued collaboration to investigate the PDT action of supramolecular complexes. The authors would also like to thank Dr Shamindri M. Arachchige, Dr Krishnan Rangan, Avijita Jain, Samantha Hopkins and Anne-Marie C. Overstreet for help with this chapter.

268


Abbreviations 5,6-Me2phen bpm bpy DAP ddz dpp dppz dsdp EA ED en ES et f GS HAT HOMO ic IL isc kx LA *LA LF LUMO Me4phen MLCT MMCT MMLCT nr p PDT Ph2phen phen phi q Q rxn TAP tpy Fx

5,6-dimethyl-1,10-phenanthroline 2,2′-bipyrimidine 2,2′-bipyridine 1,12-diazapyrylene dibenzo[h,j]dipyrido[3,2-a:2′,3′-c]phenazine 2,3-bis(2-pyridyl)pyrazine dipyrido[3,2-a:2′,3′-c]phenazine 4,7-di(3-sulfurylphenyl)-1,10-phenanthroline electron acceptor electron donor excited state energy transfer excited state light absorber excited state electron transfer fluorescence ground state light absorber 1,4,5,8,9,12-hexaazatriphenylene highest occupied molecular orbital internal conversion internal ligand intersystem crossing rate constant of process ‘x’ ground state light absorber excited state light absorber ligand field lowest unoccupied molecular orbital 3,4,7,8-tetramethyl-1,10-phenanthroline metal to ligand charge transfer metal to metal charge transfer metal–metal to ligand charge transfer nonradiative decay phosphorescence photodynamic therapy 4,7-diphenyl-1,10-phenanthroline 1,10-phenanthroline 9,10-phenanthrenequinone diimine bimolecular deactivation quencher photochemical reaction 1,4,5,8-tetraazaphenanthrene 2,2′:6′,2″-terpyridine quantum yield of process ‘x’

References

269

References 1. Ali, H.; van Lier, J.E.; Metal complexes as photo- and radiosensitizers; Chem. Rev., 1999, 99, 2379–2450. 2. Kirsch-De Mesmaeker, A.; Lecomte, J.-P.; Kelly, J.M.; Photoreactions of metal complexes with DNA, especially those involving a primary photo-electron transfer; Top. Curr. Chem.; 1996, 177, 25–76. 3. Clarke, M.J.; Non-platinum chemotherapeutic metallopharmaceuticals; Chem. Rev., 1999, 99, 2511–2533. 4. Clarke, M.J.; Ruthenium metallopharmaceuticals; Coord. Chem. Rev., 2003, 236, 209–233. 5. Erkkila, K.E.; Odom, D.T.; Barton, J.K.; Recognition and reaction of metallointercalators with DNA; Chem. Rev., 1999, 99, 2777–2795. 6. McMillin, D.R.; McNett, K.M.; Photoprocesses of copper complexes that bind to DNA; Chem. Rev., 1998, 98, 1201–1219. 7. Stochel, G.; Wanat, A.; Kulis, E.; Stasicka, Z.; Light and metal complexes in medicine; Coord. Chem. Rev., 1998, 171, 203–220. 8. Szacilowski, K.; Macyk, W.; Drzewiecka-Matuszek, A.; Brindell, M.; Stochel, G.; Bioinorganic photochemistry: frontiers and mechanisms; Chem. Rev., 2005, 105, 2647– 2694. 9. Castano, A.P.; Demidova, T.N.; Hamblin, M.R.; Mechanisms in photodynamic therapy: part one – photosensitizers, photochemistry and cellular localization; Photodiagn. Photodyn. Ther., 2004, 1, 279–293. 10. Castano, A.P.; Demidova, T.N.; Hamblin, M.R.; Mechanisms in photodynamic therapy: part two – cellular signaling, cell metabolism, and modes of cell death; Photodiagn. Photodyn. Ther., 2005, 2, 1–23. 11. Castano, A.P.; Demidova, T.N.; Hamblin, M.R.; Mechanisms in photodynamic therapy: part three – photosensitizer pharmicokinetics, biodistribution, tumor localization and modes of tumor destruction; Photodiagn. Photodyn. Ther., 2005, 2, 91–106. 12. Allison, R.R.; Mota, H.C.; Sibata, C.H.; Clinical Pd/Pdt in North America: An historical review; Photodiagn. Photodyn. Ther., 2004, 1, 263–277. 13. Kessel, D.; Photodynamic therapy: from the beginning; Photodiagn. Photodyn. Ther., 2004, 1, 3–7. 14. Allison, R.R.; Downie, G.H.; Cuenca, R.; Hu, X.-H.; Childs, C.J.H.; Sibata, C.H.; Photosensitizers in clinical PDT; Photodiagn. Photodyn. Ther., 2004, 1, 27–42. 15. Triesscheijn, M.; Baas, P.; Schellens, J.H.M.; Stewart, F.A.; Photodynamic therapy in oncology; Oncologist, 2006, 11, 1034–1044. 16. Englman, R.; Jortner, J.; The energy gap law for radiationless transitions in large molecules; Mol. Phys., 1970, 18, 145–164. 17. Calladine, C.R.; Understanding DNA: The Molecule and How It Works, 3rd Edn, Elsevier Academic Press, San Diego, CA, 2004. 18. Kalyanasundaram, K.; Photochemistry of Polypyridine and Porphyrin Complexes, Academic Press, San Diego, CA, 1992. 19. Juris, A.; Balzani, V.; Barigletti, F.; Campagna, S.; Belser, P.; Von Zelewsky, A.; Ru(II) polypyridine complexes photophysics, photochemistry, electrochemistry, and chemiluminescence; Coord. Chem. Rev., 1988, 84, 85–277. 20. Balzani, V.; Moggi, L.; Scandola, F.; In Supramolecular Photochemistry; Balzani, V., Ed., D. Reidel Publishing Company, 1987, 1–28. 21. Demas, J.N.; Taylor, D.G.; Intersystem crossing yields in ruthenium(II) and osmium(II) photosensitizers; Inorg. Chem., 1979, 18, 3177–3179. 22. Singh, T.N.; Turro, C.; Photointiated DNA binding by Cis-[Ru(Bpy)2(Nh3)2]2+; Inorg. Chem., 2004, 43, 7260–7262. 23. Ford, P.C.; Wink, D.; Dibenedetto, J.; In Progress in Inorganic Chemistry, Vol. 30, Lippard, S.J., Ed., John Wiley & Sons, Inc., New York, 1983,213–271.

270


24. Kober, E.M.; Caspar, J.V.; Lumpkin, R.S.; Meyer, T.J.; Application of the energy gap law to excited-state decay of osmium(II)-polypyridine complexes: calculation of relative nonradiative decay rates from emission spectral profiles; J. Phys. Chem., 1986, 90, 3722–3734. 25. Caspar, J.V.; Kober, E.M.; Sullivan, B.P.; Meyer, T.J.; Application of the energy gap law to the decay of charge-transfer excited states; J. Am. Chem. Soc., 1982, 104, 630–632. 26. Crosby, G.A.; Elfring, W.H., Jr.; Excited states of mixed ligand chelates of ruthenium(II) and rhodium(III); J. Phys. Chem., 1976, 80, 2206–2211. 27. Demas, J.N.; Crosby, G.A.; Quantum efficiencies of transition-metal complexes.I. d-d Luminescence; J. Am. Chem. Soc., 1970, 92, 7262–7270. 28. Crosby, G.A.; Spectroscopic investigations of excited states of transition-metal complexes; Acc. Chem. Res., 1975, 8, 231–238. 29. Muir, M.M.; Huang, W.-L.; Photoaquation of some complexes of rhodium(III); Inorg. Chem., 1973, 12, 1831–1835. 30. Mahnken, R.E.; Billadeau, M.A.; Nikonowicz, E.P.; Morrison, H.; Toward the development of photo cis-platinum reagents. Reaction of cis-dichlorobis(1,10-phenanthroline)rh odium(III) with calf thymus DNA, nucleotides, and nucleosides; J. Am. Chem. Soc., 1992, 114, 9253–9265. 31. Sherman, S.E.; Lippard, S.J.; Structural aspects of platinum anticancer drug interactions with DNA; Chem. Rev., 1987, 87, 1153–1181. 32. Bradley, P.M.; Fu, P.K.-L.; Turro, C.; Excited state properties of Rh2(O2CCH3)4: solution photochemistry and photoinitiated DNA cleavage; Comments Inorg. Chem., 2001, 22, 393–426. 33. Angeles-Boza, A.M.; Bradley, P.M.; Fu, P.K.-L.; Wicke, S.E.; Bacsa, J.; Dunbar, K.R.; Turro, C.; DNA binding and photocleavage in vitro by new dirhodium(II) dppz complexes: correlation to cytotoxicity and photocyotoxicity; Inorg. Chem., 2004, 43, 8510–8519. 34. De Cola, L.; Belser, P.; Photoinduced energy and electron transfer processes in rigidly bridged dinuclear Ru/Os complexes; Coord. Chem. Rev., 1998, 177, 301–346. 35. Furue, M.; Hirata, M.; Kinoshita, S.; Kushida, T.; Kamachi, M.; Intramolecular electrontransfer of covalently-linked polypyridine ruthenium(II)/rhodium(III) binuclear complexes in the excited state. observation of the Marcus inverted region; Chem. Lett., 1990, 2065–2069. 36. Indelli, M.T.; Bignozzi, C.A.; Harriman, A.; Schoonover, J.R.; Scandola, F.; Four intercomponent processes in a Ru(II)-Rh(III) polypyridine dyad: electron transfer from excited donor, electron transfer to excited acceptor, charge recombination, and electronic energy transfer; J. Am. Chem. Soc., 1994, 116, 3768–3779. 37. Kalyanasundaram, K.; Gratzel, M.; Nazeeruddin, M.K.; Excited-state interactions in ligand-based chromophore-quencher complexes containing rhodium(III) and ruthenium(II) polypyridyl units; J. Phys. Chem., 1992, 96, 5865–5872. 38. Kalyanasundaram, K.; Nazeeruddin, M.K.; Photophysics and photoredox reactions of ligand-bridged binuclear polypyridyl complexes of ruthenium(II) and their monomeric analogues; Inorg. Chem., 1990, 29, 1888–1897. 39. Kleverlaan, C.J.; Indelli, M.T.; Bignozzi, C.A.; Pavanin, L.; Scandola, F.; Hasselman, G.M.; Meyer, G.J.; Stepwise charge separation in heterotriads. binuclear Ru(II)-Rh(III) complexes on nanocrystalline titanium dioxide; J. Am. Chem. Soc., 2000, 122, 2840–2849. 40. Fuchs, Y.; Lofters, S.; Dieter, T.; Shi, W.; Morgan, R.; Strekas, T.C.; Gafney, H.D.; Baker, A.D.; Spectroscopic and electrochemical properties of dimeric ruthenium(II) diimine complexes and determination of their excited state redox properties; J. Am. Chem. Soc.; 1987, 109, 2691–2697. 41. Molnar, S.M.; Jensen, G.E.; Volger, L.M.; Jones, S.W.; Laverman, L.; Bridgewater, J.S.; Richter, M.M.; Brewer, K.J.; Photochemical properties of mixed-metal supramolecular complexes; J. Photochem. Photobiol. A, 1994, 80, 315–322. 42. Swavey, S.; Brewer, K.J.; Synthesis and study of Ru,Rh,Ru triads: modulation of orbital energies in a supramolecular architecture; Inorg. Chem.; 2002, 41, 4044–4050. 43. Swavey, S.; Brewer, K.J.; Visible light induced photocleavage of DNA by a mixed-metal supramolecular complex: [{(bpy)2Ru(dpp)}2RhCl2]5+; Inorg. Chem., 2002, 41, 6196–6198.

References

271

44. Holder, A.A.; Swavey, S.; Brewer, K.J.; Design aspects for the development of mixedmetal supramolecular complexes capable of visible light induced photocleavage of DNA; Inorg. Chem., 2004, 43, 303–308. 45. Marcus, R.A.; Chemical and electrochemical electron transfer theory; Annu. Rev. Phys. Chem.; 1964, 15, 155–196. 46. Foote, C.S. Definition of Type I and Type II Photosensitized Oxidation; Photochem. Photobiol. 1991, 54, 659. 47. Forster, T.; Transfer mechanisms of electronic excitation; Discuss. Faraday Soc., 1959, 27, 7–17. 48. Dexter, D.L.; A theory of sensitized luminescence in solids; J. Chem. Phys., 1953, 21, 836–850. 49. Demas, J.N.; Diemente, D.; Harris, E.W.; Oxygen quenching and charge-transfer excited states of ruthenium(II) complexes. Evidence for singlet oxygen production; J. Am. Chem. Soc., 1973, 95, 6864–6865. 50. Moucheron, C.; Kirsch-De Mesmaeker, A.; Kelly, J.M.; Photoreactions of ruthenium(II) and osmium(II) complexes with deoxyribonucleic acid (DNA); J. Photochem. Photobiol. B, 1997, 40, 91–106. 51. Zeglis, B.M.; Pierre, V.C.; Barton, J.K.; Metallo-intercalators and metallo-insertors; J. Chem. Soc. Chem. Commun., 2007, 4565–4579. 52. Armitage, B.; Photocleavage of nucleic acids; Chem. Rev., 1998, 98, 1171–1200. 53. Pogozeleski, W.K.; Tullius, T.D.; Oxidative strand scission of nucleic acids: Routes initated by hydrogen abstraction from the sugar moiety; Chem. Rev., 1998, 98, 1098–1107. 54. Vinograd, J.; Lebowitz, J.; Physical and topological properties of circular DNA; J. Gen. Physiol., 1966, 49, 103–125. 55. van der Maarel, J.R.C.; Zakharova, S.S.; Jesse, W.; Backendorf, C.; Egelhaaf, S.U.; Lapp, A.; Supercoiled DNA; Plectonemic structure and liquid crystal formation; J. Phys. Condens. Matter, 2003, 15, S183–S189. 56. Kumar, C.V.; Barton, J.K.; Turro, N.J.; Photophysics of ruthenium complexes bound to double helical DNA; J. Am. Chem. Soc. 1985, 107, 5518–5523. 57. Mei, H.Y.; Barton, J.K.; Chiral probe for A-form helices of DNA and RNA: tris(tetrame thylphenanthroline)ruthenium(II); J. Am. Chem. Soc., 1986, 108, 7414–7416. 58. Sitlani, A.; Long, E.C.; Pyle, A.M.; Barton, J.K.; DNA photocleavage by phenanthrene diimine complexes of rhodium(III): Shape selective recognition and reaction; J. Am. Chem. Soc., 1992, 114, 2303–2312. 59. Barton, J.K.; Metals and DNA: Molecular left-handed compliments; Science, 1986, 233, 727–734. 60. Manning, G.S.; Ray, J.; Counterion condensation revisted; J. Biomol. Struct. Dyn., 1998, 16, 461–476. 61. Junicke, H.; Hart, J.; Kisko, J.; Glebov, O.; Kirsch, I.R.; Barton, J.K.; Bioinorganic chemistry special feature: A rhodium complex for high-affinity DNA base-pair mismatch recognition; Proc. Natl. Acad. Sci. USA, 2003, 100, 3737–3742. 62. Burrows, C.J.; Muller, J.G.; Oxidative nucleobase modifications leading to strand scission; Chem. Rev., 1998, 98, 1109–1151. 63. Poklar, N.; Pilch, D.S.; Lippard, S.J.; Redding, E.A.; Dunham, S.U.; Breslauer, K.; Influence of cisplatin intrastrand crosslinking on the conformation, thermal stability, and energetics of a 20-mer DNA duplex; Proc. Natl. Acad. Sci. USA, 1996, 93, 7606–7611. 64. Morrison, H.; Harmon, H.; ‘Hot spots’ associated with the photoinduced binding of cisdichloro bis(1,10-phenanthroline)rhodium(III) chloride to HIV-1 and C-Raf DNA; Photochem. Photobiol., 2000, 72, 731–738. 65. Menon, E.L.; Perera, R.; Navarro, M.; Kuhn, R.J.; Morrison, H.; Phototoxicity against tumor cells and sindbis virus by an octahedral rhodium bisbipyridyl complex and evidence for the genome as a target in viral photoinactivation; Inorg. Chem., 2004, 43, 5373–5381. 66. Loganathan, D.; Morrison, H.; Effect of ring methylation on the photophysical, photochemical and photobiological properties of cis-dichlorobis(1,10-phenanthrolin)rhodium (III)chloride; Photochem. Photobiol., 2006, 82, 237–247.

272


67. Blasiu, R.; Nierengarten, H.; Luhmer, M.; Constant, J.-F.; Defrancq, E.; Dumy, P.; van Dorsselaer, A.; Moucheron, C.; Kirsch-De Mesmaeker, A.; Photoreaction of [Ru(HAT)2phen]2+ with guanosine-5′-monophosphate and DNA: Formation of new types of photoadducts; Chem. Euro.J., 2005, 11, 1507–1517. 68. Kawanishi, S.; Oikawa, S.; Murata, M.; Tsukitome, H.; Saito, I.; Site-specific oxidation at GG and GGG sequences in double-stranded DNA by benzoyl peroxide as a tumor promoter; Biochemistry, 1999, 38, 16733–16739. 69. Prat, F.; Houk, K.N.; Foote, C.S.; Effect of guanine stacking on the oxidation of 8-oxoguanine in B-DNA; J. Am. Chem. Soc., 1998, 120, 845–846. 70. Saito, I.; Takayama, M.; Sugiyama, H.; Nakatani, K.; Photoinduced DNA cleavage via electron transfer: Demonstration that guanine residues located 5′ to guanine are the most electron-donating sites; J. Am. Chem. Soc., 1995, 117, 6406–6407. 71. Sugiyama, H.; Saito, I.; Theoretical studies of GG-specific photocleavage of DNA via electron transfer: Significant lowering of ionization potential and 5′-localization of homo of stacked GG bases in B-form DNA; J. Am. Chem. Soc., 1996, 118, 7063–7068. 72. Ortmans, I.; Elias, B.; Kelly, J.M.; Moucheron, C.; Kirsch-De Mesmaeker, A.; [Ru(TAP)2(dppz)]2+: A DNA intercalating complex, which luminesces strongly in water and undergoes photo-induced proton-coupled electron transfer with guanosine-5′monophosphate; Dalton Trans., 2004, 668–676. 73. Friedman, A.E.; Chambron, J.-C.; Sauvage, J.-P.; Turro, N.J.; Barton, J.K.; Molecular ‘light switch’ for DNA; Ru(bpy)2(dppz)2+; J. Am. Chem. Soc., 1990, 112, 4960–4692. 74. Fleisher, M.B.; Waterman, K.C.; Turro, N.J.; Barton, J.K.; Light-induced cleavage of DNA by metal complexes; Inorg. Chem., 1986, 25, 3549–3551. 75. Mongelli, M.T.; Heinecke, J.; Mayfield, S.; Okyere, B.; Winkel, B.S.J.; Brewer, K.J.; Variation of DNA photocleavage efficiency for {(TL)2Ru(dpp)]Cl2 complexes where TL = 2,2′bipyridine, 1,10-phenanthroline, or 4,7-diphenyl-1,10-phenanthroline; J. Inorg. Biochem., 2006, 100, 1983–1987. 76. Hergueta-Bravo, A.; Jimenez-Hernandez, M.E.; Montero, F.; Oliveros, E.; Orellana, G.; Singlet-oxygen mediated DNA photocleavage with Ru(II) polypyridyl complexes; J. Phys. Chem. B, 2002, 106, 4010–4017. 77. Chouai, A.; Wicke, S.E.; Turro, C.; Bacsa, J.; Dunbar, K.R.; Wang, D.; Thummel, R.P.; Ruthenium(II) complexes of 1,12-diazaperylene and their interactions with DNA; Inorg. Chem., 2005, 44, 5996–6003. 78. Miao, R.; Mongelli, M.T.; Zigler, D.F.; Winkel, B.S.J.; Brewer, K.J.; A multifunctional tetrametallic Ru-Pt supramolecular complex exhibiting both DNA binding and photocleavage; Inorg. Chem., 2006, 45, 10413–10415. 79. Puckett, C.A.; Barton, J.K.; Methods to explore cellular uptake of ruthenium complexes; J. Am. Chem. Soc., 2007, 129, 46–47. 80. Dobrucki, J.W.; Interaction of oxygen-sensitive luminescent probes Ru(phen)32+ and Ru(bipy)32+ with animal and plant cells in vitro. Mechanism of phototoxicity and conditions for non-invasive oxygen measurements; J. Photochem. Photobiol. B, 2001, 65, 136–144. 81. Jimenez-Hernandez, M.E.; Orellana, G.; Montero, F.; Portoles, M.T.; A ruthenium probe for cell viability measurement using flow cytometry, confocal microscopy and timeresolved luminescence; Photochem. Photobiol., 2000, 72, 28–34. 82. Brunner, J.; Barton, J.K.; Targeting DNA mismatches with rhodium intercalators functionalized with a cell-penetrating peptide; Biochemistry, 2006, 45, 12295–12302. 83. Angeles-Boza, A.M.; Bradley, P.M.; Fu, P.K.-L.; Shatruk, M.; Hilfiger, M.G.; Dunbar, K. R.; Turro, C.; Phototoxicity of a new Rh2(II,II) complex: Increase in cytotoxicity upon irradiation similar to that of Pdt agent hematoporphyrin; Inorg. Chem., 2005, 44, 7262–7264. 84. Holder, A.A.; Zigler, D.F.; Tarrago-Trani, M.T.; Storrie, B.; Brewer, K.J.; Photobiological impact of [{(bpy)2Ru(dpp)}2RhCl2]Cl5 and [{(bpy)2Os(dpp)}2rhcl2]Cl5 [bpy = 2,2′-bipyridine; dpp = 2,3-bis(2-pyridyl)pyrazine] on Vero cells; Inorg. Chem., 2007, 46, 4760–4762.

9 Platinated Oligonucleotides: Synthesis and Applications for the Control of Gene Expression Vicente Marchán and Anna Grandas

9.1 Therapeutic Applications of Synthetic Oligonucleotides The genetic information necessary for making a living organism is encoded in DNA (deoxyribonucleic acid), a very large biopolymer associated with many copies of several proteins in chromosomes. The individual units of DNA that contain the instructions for producing proteins are called genes. However, the flow of information from DNA to proteins is mediated by a different nucleic acid, mRNA (messenger ribonucleic acid), which differs from DNA in that deoxyribose is replaced by ribose and thymine by uracil. According to the central tenet of molecular biology, the information encoded by DNA in genes is transcribed into information in RNA (although in some cases, such as in retroviruses, information flows from RNA to DNA), which in turn is translated into the functional protein. The availability of molecules that specifically target either protein or DNA or RNA sequences is essential for the treatment of diseases of genetic, viral or bacterial origin and malignant tumours. In this respect, the extraordinary capacity for recognition of nucleic acids through hydrogen bonding between complementary bases may allow the use of nucleotide-based molecules to control the expression of a particular gene. In 1978 Zamecnik and Stephenson reported for the first time the use of small nucleic acid fragments, called antisense oligonucleotides (AS-O), as therapeutic Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

274

Platinated Oligonucleotides

transcription

action may be blocked by complex formation with synthetic oligonucleotides (aptamers)

translation

dsDNA

mRNA

protein

blocked by antigene oligonucleotides.

blocked by antisense oligonucleotides, ribozymes, siRNAs.

Triplex formation:

Duplex formation:

antisense oligonucleotides

ribozymes

Figure 9.1 Use of synthetic oligonucleotides for the control of gene expression

agents that could inhibit gene expression.1 Antisense technology is based on the specific hybridization of synthetic single-stranded oligonucleotides (some 15–20 bases in length) to a complementary mRNA sequence through Watson–Crick pairing, which prevents the expression of the target gene2,3 (Figure 9.1). Unlike conventional drugs, which bind proteins and modulate their functions, antisense oligonucleotides act at the mRNA level, inhibiting protein biosynthesis through two major mechanisms. When a highly stable heteroduplex is formed between AS-O and mRNA, the ribosomal machinery cannot process the information carried by that mRNA, resulting in the interruption of protein biosynthesis. However, the most potent and widely demonstrated mechanism involves the cleavage of mRNA by RNase H, a ubiquitous cellular nuclease that degrades the RNA strand of hybrid RNA–DNA duplexes. After the improvements in technology over the last 20 years, one antisense drug (VitraveneTM) has been approved for patients with AIDS-related retinitis,4 and several other molecules are currently in advanced clinical trials. The main problems associated with the use of synthetic oligonucleotides are finding accessible sites on the highly structured mRNA target, their low cellular stability, the difficulty of crossing cell membrane barriers, and the formation of structures that may affect binding affinity and specificity. To overcome these problems, chemists have synthesized oligonucleotide analogues modified almost everywhere in the chain (nucleobases and sugar-phosphate backbone), resulting in molecules with enhanced cellular stability (‘first generation’ of AS-O, e.g. phosphorothioate oligodeoxynucleotides), and in lower toxicity and higher affinity for the target mRNA (‘second generation’ of ASO, e.g. 2′-O-modified oligoribonucleotides). Recently, chemical modifications have provided new DNA and RNA analogues (‘third-generation’ of AS-O), with improved target affinity, biostability and pharmacokinetics, which are expected to generate new antisense-based drugs in the near future.5 As well as the antisense approach, two additional anti-mRNA strategies (Figure 9.1) have been developed, namely ribozymes and RNA interference (RNAi), both

Therapeutic Applications of Synthetic Oligonucleotides

275

of which also aim to inhibit the expression of a particular gene.6 Synthetic ribozymes mimic the mode of action of naturally occurring ribozymes,7 which bind to mRNA and cleave a precise phosphodiester bond. One of the most commonly studied is the hammerhead ribozyme, which contains two binding arms for substrate recognition and a catalytic domain for cleavage.8 Like AS-O, ribozymes have been chemically modified to increase their stability in cells and their delivery to specific tissues without compromising activity, and some of them are currently in clinical trials for evaluation as therapeutic agents.9 The recent discovery of RNA interference (RNAi)10 has triggered the use of technologies based on this natural gene-silencing mechanism in biological research and their evaluation for the treatment of many diseases.11,12 RNAi is mediated by small interfering RNAs (siRNAs), which are double-stranded oligoribonucleotides typically 21–25 nucleotides in length, and two nucleotide 3′-OH overhangs that are incorporated into an active ribonucleoprotein complex, the RNA-induced silencing complex (RISC).13 Unwinding of the two chains of the siRNA duplex is followed by hybridization of the antisense strand with mRNA, which, depending on the degree of complementarity, either guides RNA cleavage by the activated RISC complex or blocks translation. Synthetic siRNAs seem to be more active than AS-O and could very possibly be developed as a new therapeutic strategy. However, nonspecific effects and toxicity may be associated with high activity, and, although siRNAs are fairly stable in cells, chemical modifications that increase cellular stability and facilitate delivery may be required to ensure their efficiency in vivo for future therapeutic applications.14 Another emerging technology is that based on targeting microRNAs (miRNAs), which are single-stranded nonprotein coding RNAs (21–23 bases in length) that regulate gene expression in plants and animals.15,16 Unlike siRNA-mediated RNAi that inhibits translation of a single gene, an miRNA may target hundreds of different mRNAs. The evidence of miRNA participation in a wide range of cellular functions in mammals (e.g. cell growth and apoptosis, heart and brain development and insulin secretion), together with their involvement in diseases such as neurodegenerative disorders and cancer, makes them attractive new druggable targets.17 Recently, chemically modified anti-miRNA oligonucleotides have proved valuable tools to specifically silence miRNAs in vivo, enabling their functions to be unravelled in a new, promising therapeutic strategy.18,19 Small synthetic oligonucleotides can also be used to target specific sequences of the double helix of DNA (antigene strategy, Figure 9.1).20 These molecules, called triplex-forming oligonucleotides (TFOs),21 recognize and bind polypurine/polypyrimidine tracts in DNA duplexes through non Watson–Crick hydrogen-bonding interactions (Hoogsteen bonding) forming a supramolecular structure that contains three strands. TFOs have been used to selectively inhibit or regulate gene expression at the transcriptional level, and to correct genetic defects arising from point mutations.22 Nevertheless, problems such as the low stability of triple helices, the need to improve their delivery and stability in the intracellular environment and their access to the highly packed nuclear DNA still have to be suitably addressed. In addition to interacting with each other through specific hydrogen bonding, structured nucleic acids (DNA and RNA) also bind to a wide range of target

276


molecules, from organic dyes to proteins. Some oligonucleotides (25–40 bases in length), called aptamers, show a highly organized tertiary structure that enables them to bind specifically to their target (generally a protein) with high affinity (Figure 9.1). Aptamers, which can be selected from combinatorial nucleic acid libraries through an iterative in vitro process (SELEX, systematic evolution of ligands by exponential enrichment),23,24 can, in principle, target virtually any protein. Among the many aptamers reported, only MacugenTM, a polyethylene glycol-engrafted oligonucleotide that specifically binds the vascular endothelial grown factor-165 isoform, has been approved for treating age-related macular degeneration.25 Although many aspects such as delivery must be improved to develop new efficient aptamer-based drugs, aptamer technology is emerging as a powerful and versatile tool in various fields of biotechnology, target validation, bioanalysis, diagnostics and imaging.26,27

9.2 Platination as a Tool to Enhance Biological Effects As stated above, one of the requirements for making drugs out of synthetic oligonucleotides is that they efficiently recognize and interact with their targets. Only sufficiently stable complexes can prevent the ribosome machinery from translating the mRNA code into a polypeptide sequence, stop replication or transcription, or block the function of a protein.28 Chemical modification of the chain modulates the affinity of synthetic oligonucleotides for their targets. Substitution of sulfur for one of the oxygens of the phosphodiester bridge (phosphorothioate or S-oligonucleotides),29 which is the modification introduced in the only antisense oligonucleotide approved for therapeutic use,4 slightly decreases the stability of the modified-oligonucleotide–mRNA duplex.30 Conversely, the stability of the duplex can be improved, for instance, by the introduction of a methylene bridge that links the 4′-C and 2′-O and freezes the North conformation of the ribose ring (locked nucleic acids or LNAs),31 or by replacement of the charged sugar-phosphate backbone with noncharged ones, as in peptide nucleic acids (PNAs)32 or morpholino oligonucleotides.33 The stability of the final complex, either duplex or triplex, can also be improved by forming a crosslink between the oligonucleotide and the target. Duplexes can be stabilized by covalently linking their ends. This covalent union has been chemically established by a short oligonucleotide sequence, as in naturally occurring hairpin duplexes, or by nonoligonucleotide linkages (Figure 9.2) such as disulfide bridges,34 alkyl chains,35,36 aromatic moieties such as stilbenes37 or phenanthrenes,38 and metal complexes.39,40 Nevertheless, if oligonucleotides are to be used as therapeutic agents, and recognize and link to their targets, only the synthetic chain can be modified. In that case, a group attached or appended to the synthetic oligonucleotide is expected to give rise to the desired covalent union upon hybridization. Examples of this approach are the use of 2-arylthioethyl- or quinone methide-modified oligonucleotides, as described by Sasaki41 and Rokita,42 respectively (Figures 9.3A and B). In both cases, the target is alkylated upon duplex formation, yielding the interchain crosslink.

Platination to Enhance Biological Effects 277 O

S

N

A N

N

O

O N

B

O

S

N

O

O

(CH2)n

N

N

O

HN

NH

(CH2)3

(CH2)3

NH N

O

O

N

N

O C

NH

O

(CH2)n O

N

N

D O

NH (CH2)n

HN

O

(CH2)n

Ph

N

E

N NH (H2C)3

M

2+

N

= oligonucleotide chain

HN (CH2)3

Figure 9.2 Alternatives to the crosslinking of two oligonucleotide chains by chemical derivatization: disulfide bridges (A), alkyl chains (B), stilbenes (C), phenantrenes (D) and metal complexes (E)

Another alternative that has been pursued is that a metal atom, especially platinum, could be made to establish the crosslink between the synthetic oligonucleotide and the target. In the context of antisense, antigene or anti-microRNA strategies, the platinum(II)-mediated attachment of small synthetic oligonucleotides to their complementary targets (DNA or RNA) is expected to produce a more persistent blockage than that obtained by simple hydrogen-bonding pairing. In this chapter we review the attempts to achieve high therapeutic efficacy as a result of the formation of platinum crosslinks between oligonucleotide chains at the duplex or triplex level. Over the past 15 years, a great deal of effort has been devoted to the development of methodologies to synthesize platinated oligonucleotides, but the scope of applications of such molecules is limited because of the difficulties associated with their preparation (see Sections 9.3 and 9.4).

278


Figure 9.3 Crosslinking promoted by annealing. The reactive species is either a vinyl group (A) or a quinone-methide moiety (B)

Although experiments leading to platinated oligonucleotides have often been followed by crosslinking studies with no clear-cut separation between the two aspects of the study, in this chapter the description of these studies has been split into three sections. Section 9.3 covers most of the methods developed to prepare platinated oligonucleotides, either using preplatinated monomers that are incorporated into oligonucleotide chains, or platinating fully assembled oligonucleotides (partially protected or unprotected, unmodified or containing modified nucleobases). Then, the use of platinated oligonucleotides in studies whose final goal is the formation of crosslinked duplexes, most often to achieve improved antisense effects, is described in Section 9.4. Finally, Section 9.5 reviews the results of experiments aiming at the formation of complexes between platinated oligonucleotides and dsDNA, namely the formation of triple helices (antigene approach). Although the purpose of that study was not to establish a crosslink, it is worth mentioning that the effect of platinating the sense strand of an siRNA duplex was also evaluated in a recent publication.43

Synthesis of Platinated Oligonucleotides

279

9.3 Synthesis of Platinated Oligonucleotides 9.3.1 Platination of Unmodified Oligonucleotides In general, platination of unprotected oligonucleotides in neutral solutions leads to complex mixtures because of the presence of multiple coordination sites at the heterocyclic nucleobases (N7 in guanine, N7 and N1 in adenine, and N3 in cytosine). Although N7 of guanine is the preferred binding site,44 the formation of many different platinated products certainly complicates the isolation of the desired adduct, making its preparation impossible in some cases. An alternative to decreasing the complexity of the crude is to reduce the pH of the reaction medium, in order to keep cytosines and adenines protonated (N3 and N1, respectively) and thus unreactive.45 However, these conditions cause depurination.46,47 Another widely used approach is to reduce the number of guanines to one or two residues, but this alternative prevents the oligonucleotide that is to be platinated from targeting any sequence.48 Given the constraints derived from the structure of platinum(II) complexes together with their preference for guanines, it is possible to control the evolution of the reaction to yield a predominant adduct by choosing the number and position of guanines in the chain as well as the conditions of the reaction. For example, single-stranded oligonucleotides containing two neighbouring guanines react with cisplatin to afford a major 1,2-intrastrand crosslink with platinum bonding both guanines (Figure 9.4A). In contrast, the linear geometry of transplatin prevents formation of these chelates, but it can yield 1,3-intrastrand crosslinks at GNG and ANG sequences (where N is any nucleotide residue other than G) (Figure 9.4B). The geometry of coordination also explains that closure of monofunctional adducts to form interstrand crosslinks is more frequent with transplatin than with cisplatin, especially between the N7 of guanine and the N3 of its base-paired cytosine in a DNA duplex.49 This is because the distortion in the double helix is smaller than with cisplatin-mediated interstrand crosslinks.50,51 Although the preference of transplatin for GNG triplets allows the preparation of platinated oligonucleotides for use in crosslinking experiments (see Section 9.4.2),52 the restriction in sequence imposed by the fact that the oligonucleotide cannot contain other guanines prevents the targeting of any nucleic acid. Additional difficulties associated with this alternative are practical aspects of purification and characterization of GNG-platinated oligonucleotides, and the low stability of the final adduct in some cases. In this respect, it has been reported that the chelates formed by reaction of transplatin with guanines at CGNG sequences rearrange into 1,4-intrastrand crosslinks by platinum migration from the 5′-guanine to cytosine (Figure 9.4C).45,53 The approaches in which transplatin or a binuclear complex reacts with an oligonucleotide to yield a reactive monofunctional adduct are also problematic (see Section 9.4.2).45–48 One difficulty is the low stability of platinated oligonucleotides with one labile bond, since they may react with other nucleobases present in the oligonucleotide (‘suicide’ reaction). In addition, there are many cellular

280


A

Cl

GG

Pt

Cl

B

Cl H3N

GNG

NH3 NH3

Pt

GG

Here,

H3N

=

Pt

NH3

NH3 Cl

H3N

GNG

Here,

=

CGNG

Here,

=

Pt

N = A, C, T

NH3

C CGNG N = A, C, T

D

Cl

G

H3N

Pt

NH3 PG

Pt

NH3

Cl/H2O

PG H3N Pt

H3N

NH3

G

deprotection

H3N Pt

NH3

G

PG = protecting group = oligonucleotide

Figure 9.4 Reactions of cisplatin with GG-containing oligonucleotides (A) and of transplatin with GNG-containing chains (N = A, C, T) (B). trans-Pt(NH3)2 1,3-chelates isomerize to 1,4chelates when the GNG sequence is flanked by a 5′ C (C). The reaction of guanine-containing oligonucleotides with suitably protected platinum complexes affords platinated oligonucleotides with one labile bond (D)

nucleophiles that may render the complex unreactive before it reaches its target. One solution envisages transplatin being replaced by a trans-(NH3)2 platinum complex in which the metal is linked to one labile substituent and a protecting group selectively removable upon interaction with the target. However, this ideal protecting group has not yet been discovered. So far, only thymine derivatives (Pt-N3 linkage) have been used as protecting groups in the preparation of platinated oligonucleotides with one labile bond. Their use has not been extended, however, because they are labile to acidic conditions that may induce depurination (Figure 9.4D).46,54 Over the years, chemists have applied various methods to clarify the structure of platinated oligonucleotides. The presence of platinum had been assessed by atomic absorption or induced coupled plasma mass spectrometry, but these techniques were replaced by modern mass spectrometric techniques that allow the ionization of oligonucleotides, such as ESI or MALDI (ESI = electrospray ionization, MALDI = matrix-assisted laser desorption ionization). The formation of a coordination compound was inferred from the evidence that reaction of the platinated oligonucleotide with CN− or thiourea had afforded the nonmodified chain. NMR


281

provides both proof of the presence of platinum (when the use of 195Pt enriched complexes allows 195Pt spectra to be recorded) and, based on the comparison between the chemical shifts of platinated and nonplatinated oligonucleotides, information on the coordination point. This item of information, however, has also been obtained from gel electrophoresis analysis after chemical reactions that modify nonplatinated nucleobases and cleave the chain, and, more recently, from mass spectrometric analysis after digestion with exonucleases. Both NMR and X-ray crystallography provide structural information at the atomic level. The first structure was solved more than 10 years ago, and showed the distortion caused in a duplex by the formation of a 1,2-intrachain cis-[Pt(NH3)2] adduct at a GG doublet.55 Highresolution pictures have also been obtained for duplexes with one strand coordinated to other platinum complexes,56,57 as well as for cis-58 and trans-[Pt(NH3)2] crosslinked duplexes.51 9.3.2 Platination of Unprotected Oligonucleotides with an Appending Chelating Group The McLaughlin group explored the alternative of platinating oligonucleotides incorporating an ethylenediamine group.59,60 They introduced the phosphoramidite derivative of the bis-Fmoc-protected chelator (Fmoc = 9-fluorenylmethoxycarbonyl) at the 5′-end of an immobilized oligothymidine. After ammonia treatment and purification of the oligonucleotide derivative, reaction with K2PtCl4 afforded the expected dichloroplatinum(II) derivative. However, even though formation of a chelate favours coordination with the diamine, the high reactivity of K2PtCl4 precludes the inclusion of nucleobases other than thymine and thus limits the scope of this approach (Figure 9.5).

Fmoc HN

NiPr2 N P OCNE Fmoc HO Tn

H2N

NH

Tn

K2PtCl4

NH

H2N Cl

Pt

Tn

Cl

= solid matrix Fmoc = 9-fluorenylmethoxycarbonyl CNE = 2-cyanoethyl

Figure 9.5 Post-assembly platination of an oligonucleotide bearing an appending chelator

9.3.3 Solid-Phase Synthesis Using Platinated Building Blocks Biopolymers, such as peptides and oligonucleotides, which are made up of the repetition of a basic unit (amino acid or nucleoside, respectively), are commonly assembled through solid-phase technologies. This process can be automated because the

282


incorporation of each synthon always involves the same series of steps, namely coupling of a suitably protected building block and removal of a temporary protecting group to liberate the functional group to which the next unit has to be attached. The elongation of oligonucleotide chains in automatic synthesizers utilizes the phosphite triester approach,61 which is the most effective methodology, even though it involves the largest number of steps. Nevertheless, the H-phosphonate method62,63 has also provided satisfactory results for the preparation of some oligonucleotide analogues. It can be used in automatic synthesizers, provided that the standard synthesis cycles are suitably modified. The preparation of platinated oligonucleotides from platinated nucleoside monomers was first described in 1996 by two separate groups.64,65 Both aimed at introducing platinum at an internal position of the oligonucleotide chain, and used the platinated H-phosphonate nucleoside derivatives depicted in Figure 9.6A, which shows the differences in the nucleobase involved, as well as in the nature of the ligands and the stereochemistry of the complexes. Cech and coworkers64 argued that H-phosphonates were more robust than phosphoramidites, and that phosphitylation had to precede platination.

NH3

O

A

N DMT O

O

N

Pt O

Bhoc

H3N

O

N

O H P O

O

O

NH N

NH2

Cl NH3

N

NH Fmoc

O

N

NH3 DMT O

Cl

Pt

O H P O

H3N Pt O N HN

B

H3N Cl

O NH

N

O OH

DMT = 4,4'-dimethoxytrityl Bhoc = benzhydryloxycarbonyl Fmoc = 9-fluorenylmethoxycarbonyl

Figure 9.6 Structures of the 2′-deoxythymidine- and 2′-deoxyguanosine H-phosphonate monomers used to assemble platinated oligonucleotides by stepwise solid-phase synthesis (A), and of the platinated PNA monomer used to obtain platinated PNAs (B)


283

H-phosphonate chemistry was used throughout chain elongation. They found that it was more difficult to incorporate the platinated synthon than the natural nucleoside H-phosphonates. Final deprotection with concentrated aqueous ammonia afforded crude oligonucleotides showing rather complex HPLC profiles, and the coordination sphere of the metal was modified. Pyridine had replaced chloride, yielding platinated oligonucleotides with no potential to react with the nucleophiles of a complementary chain. Side reactions during final deprotection and work-up were held to account for the low quality of the crudes. The same sequence of steps was described in Lippard’s study,65 but Hphosphonate chemistry was only used to introduce the platinated monomer, while standard phosphite triester chemistry was used for all the other nucleosides. The authors indicated that after incorporation of the platinated unit the synthesis cycle was modified, but no details were given. Elongation of the full oligonucleotide chain was followed by incubation in water to allow a 1,2-intrastrand G–Pt–G chelate to form, and then by the cleavage and deprotection treatment. A fairly pure platinated oligonucleotide, albeit in low yield, was obtained, but the authors indicated that the protocol still required optimization. All these results indicate that the introduction of platinated nucleoside building blocks within the chain was neither straightforward nor easy to optimize. Although labile substituents seemed to be compatible with oligonucleotide elongation using phosphoramidites,65 subsequent studies from other groups showed that they could react with phosphoramidite derivatives.60,66 The attempts of the groups of van Boom, Reedijk and Lippert to prepare platinated oligonucleotide analogues, in particular peptide nucleic acid chains (PNAs), by solid-phase synthesis are also worth mentioning here.67 The N-Fmoc-N2-Bhoc (Bhoc = benzhydryloxycarbonyl) guanine-containing PNA monomer was reacted with a transplatin analogue to afford an N7-platinated guanine building block with a labile chloride ligand (Figure 9.6B). This monomer was then introduced at the N-terminal of resin-linked, protected PNA chains. Then, following treatment with trifluoroacetic acid, which removes all protecting groups, target-platinated PNAs were obtained. The key point for the success of the synthesis is that the final treatment is applied in conditions that do not block the reactive position at the metal centre. This contrasts with the effect of the ammonia treatment used for the final deprotection of oligonucleotides, which provides an unreactive platinated oligonucleotide.

9.3.4 Regioselective Platination of Partially Protected Oligonucleotides: Use of O6-Guanine Protecting Groups The van Boom and Reedijk groups explored an elegant alternative, in which platinum is directed to the desired guanines while the reaction with other guanines is prevented by the introduction of a bulky protecting group at the O6 position.68,69 The preliminary study68 found that N2-protected guanines could be platinated at the N7 position, unlike O6,N2-protected ones that could not (Figure 9.7). It also

284

Platinated Oligonucleotides Pt

O N O

O

NH

N

N

NH PG

Platinum(II) complex (at least one labile bond)

O

O

N O

O O

N

NH N

NH PG PG = protecting group

N O

O

N O

O O

N

N N

Platinum(II) complex

NH PG

(at least one labile bond)

PG = protecting group

Figure 9.7 N2-protected guanines react with platinum complexes, but protection of the O6 with a bulky group prevents platination at the N7 position

showed that fully protected oligonucleotides reacted very slowly with platinum(II) complexes. The yield of the platination reaction increased when the phosphate protecting groups were removed, which was associated with favourable electrostatic interactions between the positively charged cation and negatively charged phosphates. The proof of concept was obtained when different 2′-deoxyhexanucleotides were selectively platinated at the guanines whose O6-position had been left unprotected (Figure 9.7).69 The oligonucleotide chains were assembled on controlled pore glass (CPG) using an oxalyl linker and either the standard phosphoramidite derivatives (T, ABz, CBz, GiBu, where Bz = benzoyl, iBu = isobutiryl) or the N2-propanoyl,O6diphenylcarbamoyl phosphoramidite derivative of guanine. Cleavage and removal of phosphate-protecting groups afforded the desired partially protected oligonucleotide, which reacted regioselectively with [Pt(en)Cl2] (en = 1,2-diaminoethane). The target platinated oligonucleotide was obtained after removal of nucleobase protecting groups with concentrated aqueous ammonia. Although the authors suggested that this approach might be used for the generation of other site-specifically platinated oligonucleotides, such as trans-Pt-1,3-GNG adducts, this alternative has not yet been explored. Moreover, this alternative seems to be restricted to the preparation of platinated oligonucleotides with no labile platinum–ligand bonds, unless a platinum protecting group is made available that is stable to ammonia, but removable under conditions that do not damage the oligonucleotide.


285

9.3.5 Attachment of Platinated Units to Protected Oligonucleotides A different approach for the preparation of platinated oligonucleotides is the incorporation of the platinated unit to the growing oligonucleotide chain instead of using a platinated nucleoside derivative. In 1999, Ren, Cai and Segal studied the reaction of a [Pt(en)Cl2] analogue containing an appending hydroxyl group with either a phosphitylating reagent or a phosphoramidite derivative of thymine in the presence of tetrazole (or a tetrazole derivative).66 In the two cases the hydroxyl group reacted with the phosphorus(III) derivative (Figure 9.8A), but the evolution of the resulting complex was different depending on whether the reaction yielded a phosphoramidite or a phosphite triester. In the first case, the nucleophilicity of phosphorus prompted quick displacement of one of the chlorides linked to platinum, affording a double chelate. This platinated phosphoramidite could not be activated by tetrazole. Therefore, it did not react with free hydroxyl groups in a nucleoside or in an oligonucleotide chain. The phosphite triester also behaved as a nucleophile, but the resulting complex was a mixture containing a double chelate and the target [Pt(en)Cl2]-derivatized nucleotide. The same group also described70 the synthesis of platinum complexes with no chelating amines, also containing hydroxyl groups to allow attachment to oligonucleotides (Figure 9.8B). However, to the best of our knowledge this methodology has not been explored so far. NiPr2

NiPr2

A CNEO

P

NiPr2

tetrazole

NH

H2N

Pt Cl

Cl iPr2N CNEO

Cl

OH (CH2)n Pt

NH

R-OH

Pt

OCNE

Cl P O tetrazole CNEO NiPr2

Cl

OCNE

tetrazole P O

O

T

O DMT

Cl

H2N

Cl

H2N

NH Pt

Cl

B

O

P

OH

Pt Cl

NH

H2N

Cl

O P O CNEO

O

T

O DMT

P

O

NH Pt

Cl

O

NH2

O

T

O DMT

Cl

CNE = 2-cyanoethyl DMT = 4,4'-dimethoxytrityl

NH2 NH2 (CH2)n OH

Figure 9.8 Intramolecular reaction of phosphorus(III) reagents [(2-cyanoethoxy)bis(diisopropylamino)]phosphine and 5′-phosphoramidite derivative of 3′-O-DMT-thymidine] with derivatized [Pt(en)Cl2] complexes (A). Structure of a cisplatin analogue with two appending hydroxyl groups (B)

286

Platinated Oligonucleotides NH2Me

O N N

Pt O

NH2

(CH2)6

OH

NH2Me activator iPr2N P O TTTT

protecting group

[O]

NH3

NH2Me

aq HCl pH 2,3

OCNE

Cl

Pt

O NH2

(CH2)6 O P O TTTT

NH2Me

O

= solid matrix

CNE = 2-cyanoethyl [O] = oxidation (phosphite to phosphate)

Figure 9.9 Platinated oligothymidines with one labile bond are obtained by reacting the phosphitylated, resin-linked oligothymidine and a masked platinum complex with an appending hydroxyl group, followed by removal of the protecting groups

At the same time, the van Boom, Lippert and Reedijk groups developed a parallel alternative to tether a platinum complex with one labile bond to oligonucleotides.54 Their strategy involved phosphitylation of the 5′-hydroxyl of a resinlinked oligonucleotide chain, and the subsequent reaction of the resulting phosphoramidite with a platinum complex that contained a hydroxyl group (Figure 9.9). After the oligothymidine deprotection step with aqueous ammonia, treatment with aqueous HCl (pH = 2.3) removed the platinum protecting group and introduced a labile chloride group, as required for crosslinking to a complementary sequence. Further development of this approach, in particular optimization of the platinum protecting group to render its elimination compatible with any oligonucleotide sequence (in other words, with no risk of depurination), has not yet been reported. 9.3.6 Regioselective Platination of Base-Modified Oligonucleotides Platinum(II) complexes can react with the nitrogen and sulfur nucleophiles of amino acid side chains, as well as the heterocyclic nitrogen of nucleobases. Since the formation of Pt–S bonds is the kinetically favoured reaction and many sulfur nucleophiles are present in the cells, several research groups have studied the possible transfer of platinum from sulfur to nitrogen and the competition between both nucleophiles.71–81 These studies showed the preference of platinum(II) complexes for sulfur-containing species. Then, depending on the structures of the substrate and the platinum(II) complex (stereochemistry, number of labile bonds), the initially formed Pt–S adduct can evolve into monofunctional, bifunctional and even trifunctional adducts in which Pt–S bonds are usually replaced by the thermodynamically more stable Pt–N linkages. The reaction between several peptide–oligonucleotide conjugates and platinum(II) complexes such as cisplatin, Pt(dien) (dien=1,5-diamino-3-azapentane), Pt(en) and transplatin, revealed that some S–Pt–N adducts, in which the metal linked methionine and either histidine or guanine, were quite stable. In

Synthesis of Platinated Oligonucleotides N S

XS

N O

N XI

H3N

HN

HN N

N

NH3 S Pt

NH

HN

Cl

O

287

oligonucleotide (A, C, G, T)

H3N

Pt

NH3 Cl

N N H

NH

N N

N O

N

O

oligonucleotide (A, C, G, T)

Figure 9.10 Modified oligonucleotides containing the four nucleobases and two cytosine analogues can be selectively platinated at the appending thioether and imidazole tethers. (Adapted from J. Biol. Inorg. Chem., 2007, 12, 901–911.)

addition, platinum preferred the N7 of guanine rather than nitrogens from the imidazole ring. However, thioether–platinum–imidazole adducts could form in the presence of guanines, provided sulfur promotes binding to the imidazole ring rather than the N7 of guanines.78,80 On the basis of these studies, Marchán and Grandas82 envisaged an alternative method for preparing modified oligonucleotides that included, in addition to all the natural nucleobases, the thioether and imidazole functional groups present in the methionine and histidine side chains. Sulfur was expected to direct the platination process and prevent uncontrolled reaction with the nucleobases (Figure 9.10). For this purpose, two 5-methylcytidine analogues with thioether or imidazole groups attached to the 4-position (XS and XI, respectively; see Figure 9.10) were synthesized by reacting protected 4-(1,2,4-triazole)cytidine with the corresponding amines.82 After phosphitylation, the phosphoramidite derivatives were successfully introduced in oligonucleotide chains using the standard phosphite triester chemistry with no protection on either the thioether or the imidazole ring. Several oligonucleotide sequences containing either one or two analogues appended to the 5′-end were synthesized and used in preliminary platination studies. The modified nucleosides were placed at the 5′-end to minimize the destabilization that modified cytosines cause in a DNA duplex, especially in the middle of the chain.82,83 The platination studies were carried out using transplatin, because it is a simple, mononuclear complex whose reactions with single-stranded oligonucleotides have been thoroughly studied, and whose capacity to form crosslinked duplexes is clearly established.49–51 Unlike cisplatin, transplatin offers the advantage that the trans geometry prevents the sulfur ligand from labilizing the trans ammine group. Hence, using transplatin, mixtures of di- and tricoordinate adducts, which are more difficult to handle, are not formed.80 The reaction between transplatin and oligonucleotides containing the thioethermodified nucleobase was fast, but generally afforded unstable adducts and complex reaction mixtures. However, the imidazole-containing oligonucleotides reacted with transplatin much more slowly, in particular in slightly basic solutions, and it was found that imidazole-modified cytosine was less reactive than the natural nucleobases.

288


Figure 9.11 Identification of the platination site in modified oligonucleotides. Use of chemical and enzymatic reactions in combination with mass spectrometric data. (Adapted from J. Biol. Inorg. Chem., 2007, 12, 901–911.)

In contrast, platination of oligonucleotides containing the four natural nucleobases and the two nucleoside analogues XS and XI in neighbouring positions (5′dXXACGTTGAG) afforded fairly stable chelates, in which the trans-Pt(NH3)2 unit linked the thioether and the imidazole ring. The new compound isolated from the reaction of 5′dXSXIACGTTGAG with transplatin had the mass expected for a platinated oligonucleotide. It did not react with H2O2, which indicated coordination to the thioether. It was more stable to digestion with a 5′-exonuclease (calf spleen phosphodiesterase) than the nonplatinated oligonucleotide, which also confirmed coordination to the thioether at the 5′-end. Finally, digestion with a 3′-exonuclease (snake venom phosphodiesterase) removed the natural nucleosides, giving the 5′ dXSXI fragment plus a trans-Pt(NH3)2 unit, confirming coordination with the imidazole-containing nucleoside (Figure 9.11). Four chelates may be formed, because platinum may react with either of the two sulfur nonbonding electron pairs and link to either of the two nitrogens of the imidazole ring. The presence of a single main peak in the HPLC traces indicated that the isomers coeluted in the analysis conditions used. Similar results were obtained with 5′dXIXSACGTTGAG. The two chelates involving XI- and XS-platinum bonds remained unmodified for 24 h in the reaction mixture. However, at longer reaction times platinum migrated from the thioether to the N7 of the closest guanine in the sequence, to give a chelate that was thermodynamically more stable. In no case were monofunctional adducts detected, which indicates that the reaction of platinum with the thioether was immediately followed by coordination with one of the nitrogen atoms in the imidazole ring. Moreover, although the two modified oligonucleotides contained the four nucleobases and a GAG triplet, transplatin (in equimolar amount in relation with the modified oligonucleotide) preferred the thioether and the imidazole groups of the two neighbouring cytosine analogues to any of the natural bases. In other words, these modified, but unprotected, oligonucleotides were regioselectively platinated in spite of the presence of several guanines in the chain. These results opened up the possibility of using these chelates in crosslinking experiments, aiming at the control of gene expression by targeting complementary sequences in either mRNA or miRNAs.84

Platinated Oligonucleotides for Duplex Crosslinking

N

Pt

N

HN

N N

H2 N

NH2

N

O

NH2

H2N

NH3

NH Cl

N N

O

Cl

Pt

NH3 NH3

289

NH3 NH3

N N

O

= oligonucleotide (protected when anchored to resin) = solid matrix

Figure 9.12 Introduction of a chelating group in oligonucleotides using a convertible nucleoside, and regioselective reaction of the polypyrimidine chain with cisplatin.

While studying the effect of platinated aminoalkyl arms on triplex formation, Miller et al. reported the synthesis of platinated oligopyrimidines in which one cytosine was replaced by N4-(2-aminoethyl)cytosine.85 The oligonucleotides were assembled on a solid support, and the incorporation of a 4-triazole-deoxyuridine derivative in the middle of the chain allowed the introduction of the aminoethyl function by treatment with ethylenediamine. After ammonia deprotection and cleavage, the aminoethyl arm of the unprotected oligonucleotide was transformed into a Pt(dien) or cis-chlorodiammineplatinum(II) complex by reaction with [Pt(dien)Cl]Cl or cisplatin, respectively (Figure 9.12).

9.4 Use of Platinated Oligonucleotides for Duplex Crosslinking 9.4.1 Initial Experiments In 1983, Vlasov et al. described for the first time the use of platinum(II) compounds for crosslinking DNA duplexes.86 The first step was the reaction of a binuclear platinum(II) compound incorporating two labile functions with the N7 of an oligonucleotide containing a single guanine. Then, hybridization promoted reaction with the guanine at the 5′ end of the complementary strand and gave the interstrand crosslink (Figure 9.13). Regioselectivity in the platination of the second strand, together with the fact that noncomplementary chains were not crosslinked, showed that complementarity is a key point in these processes. As a result, the authors envisaged the possibility of using platinated oligonucleotides for damaging certain sequences of nucleic acids in vivo. Given the great affinity of phosphorothioate groups for platinum(II) complexes,87 a few years later Orgel et al. introduced a single phosphorothioate group at the 5′-end of an antisense oligonucleotide to mediate crosslinking to longer DNA strands that contained a complementary sequence,88,89 or even to proteins.90 In these

290

Platinated Oligonucleotides 5'

dTCCGCCTTT + Pt Pt

Pt

3'

5'

dTCCGCCTTT Pt

pH 5.0

Pt

dAGGCGG

5'

dTCCGCCTTT Pt

pH 7.1

Pt

3'

dAGGCGG

Pt = binuclear platinum(II) complex

Figure 9.13 Regioselective reaction with a binuclear complex is followed by hybridization to the complementary chain, which triggers formation of the crosslink

3' 3'

5'

S Pt

5'

Figure 9.14 Oligonucleotides with one phosphorothioate group are regioselectively platinated at sulfur. Then duplex formation induces formation of an interchain crosslink. (Adapted from Nucleic Acids Res., 1990, 18, 5163–5171.)

experiments, the platinum(II) compound (cisplatin, K2PtCl4, transplatin or a binuclear platinum complex) was added once hybridization of both strands was achieved, as demonstrated by nondenaturing gel electrophoresis. Then, crosslinking occurred when the single-stranded tail of the template bent around the 5′-terminal phosphorothioate of the antisense oligonucleotide, which placed sulfur and residues 2–5 in the target strand sufficiently close to be crosslinked by the platinum(II) compound (Figure 9.14). The different assays showed that guanine and adenine and, to a lesser extent, cytosine were the preferred sites for crosslinking, but, in the absence of these residues in positions 2–5, crosslinking occurred at thymine. The antisense oligonucleotides contained few guanines and when a guanine was placed close to the phosphorothioate group, an intramolecular ‘suicide’ reaction occurred. Among the platinum(II) compounds, the crosslinking efficiency of transplatin was comparable to that of K2PtCl4, but cisplatin was less efficient. 9.4.2 Crosslinking Experiments Using Platinated Natural Oligonucleotides and PNA Analogues While conducting research into the interactions between cis- and transplatin and DNA, Leng et al. discovered a linkage isomerization reaction between transplatinmodified oligonucleotides and their complementary strands, promoted by the formation of a DNA double helix.52 Although oligonucleotides with trans-Pt(NH3)2 1,3-intrastrand crosslinks at GNG triplets (where N is any nucleoside but G) were kinetically inert as long as they remained single-stranded (except those containing

Platinated Oligonucleotides for Duplex Crosslinking Cl H3N

GNG

Pt

NH3 Cl

GNG

N = A, C, T

+ complementary chain, hybridization

GNG

GNG

C N' C

C N' C

= oligonucleotide

291

,

H3N

=

N': base complementary to N Pt

NH3

Figure 9.15 Hybridization of an oligonucleotide with a 1,3-trans-Pt(NH3)2 chelate to the complementary strand induces a rearrangement that crosslinks the two chains

the CGNG sequence, see Section 9.3.1 and Figure 9.4C),45,53 they became unstable upon pairing with their complementary DNA or RNA strands. This interaction promoted the rearrangement of the 1,3-intrastrand chelate into an interstrand crosslink between the 5′-G of the initially platinated triplet and its complementary C, resulting in the covalent linkage of both strands (Figure 9.15). Since isomerization kinetics were independent of the nature and concentration of salts in the reaction medium, it was suggested that the mechanism involved direct nucleophilic attack of the cytosine complementary to the platinated 5′-G residue. The process was not dramatically affected by the nature of the residue (A, C or T) lying between the chelated guanines. Since the discovery of this reaction, some authors have attempted to optimize the isomerization process and make platinated oligonucleotides efficient therapeutic molecules for the control of gene expression.91 The rate of the interstrand crosslinking reaction varied, depending on the sequence facing the intrastrand crosslink, and was too slow when the complementary oligoribonucleotide sequences were 5′-CN′C (N′ = any nucleotide). However, by using platinated 2′-Omethyloligoribonucleotides and modifying the complementary sequence so that the triplet facing the chelate was replaced by 5′UA or 5′CA doublets, the rearrangement was achieved within a few minutes.92 In an experiment with a 12/11-mer in which a d(AT) doublet replaced the CTC triplet in a regular duplex, an interstrand G-N7– Pt–A-N1 crosslink was formed.93 Although this methodology has given promising results in blocking translation both in vitro and in cultured cells,94,95 its application in therapy is problematic. Again, sequence restriction requires a maximum of two guanines in the GNG triplet in the oligonucleotide to be platinated. The same strategy was used to crosslink an antisense oligo-2′-Omethylribonucleotide to the human telomeric sequence (T2AG3)n.96

292


A different approach to crosslink platinated oligonucleotides with their complementary strands was developed by Lippert et al.97 It was based on generating a transplatin monofunctional adduct in a single-stranded oligonucleotide, which, upon hybridization with the complementary chain, evolved into an interstrand crosslink. To minimize the formation of multiplatinated side products, the reaction was carried out at low pH (3.6) and temperature (4 °C), and the oligonucleotide to be platinated contained a single guanine at the 5′-end. The corresponding N7-Pt(NH3)2Cl adduct reacted with the opposite guanine in the complementary strand to form the final interstrand crosslink. Broad application of this strategy is severely limited by the above-stated requirements and by the fact that monofunctional adducts may be easily deactivated by intracellular nucleophiles before reaching the therapeutic target. As also stated above (Sections 9.3.1 and 9.3.5), use of a platinum protecting group might at least prevent undesired replacement of labile ligands in monofunctional adducts.46 As in the previous strategy, formation of an interstrand crosslink between a PNA and the complementary DNA strand was achieved by introducing a transplatin monofunctional adduct at a guanine of the PNA chain.67 Unlike the previous approaches, in this case there was no sequence restriction because the platinated PNA was synthesized by standard solid-phase methodologies that used a platinated monomer of guanine (see Figure 9.6B). However, the crosslinking reaction was too slow, which was attributed to the low solubility of the platinated PNA in the reaction media.

9.4.3 Crosslinking Experiments Using Oligonucleotides Platinated at Base-Modified Nucleosides As oligonucleotides regioselectively platinated at thioether and imidazole groups were made available,82 studying whether hybridization with the complementary chain would promote platinum migration and generate an interstrand crosslink was the next step. Annealing of platinated oligonucleotide 5′dXIXSACGTTGAG (coordination positions are highlighted in bold) to an equimolar amount of the complementary chain (5′dCTCAACGTGTTTG) triggered a ligand-exchange substitution, affording exclusively an interchain crosslink product, as detected by both anion exchange HPLC and polyacrylamide gel electrophoresis (Figure 9.16).84 The crosslinked duplex was isolated and its structure was inferred from MALDI-TOF mass spectrometric analysis before and after chemical and enzymatic reactions and from NMR. It was confirmed that platinum remained coordinated to imidazole in the modified nucleoside XI, but linked to a new position, the N7 of the guanine opposite XS. Similar results were obtained using platinated 5′dXSXIACGTTGAG. Although the crosslinking reactions were slow, under the hybridization-favouring reaction conditions and in the presence of the complementary chain, the chelates (5′dXIXSACGTTGAG or 5′dXSXIACGTTGAG) remained stable. In other words, under these conditions the intrachain rearrangement to a more stable chelate (5′dXIXSACGTTGAG or 5′dXSXIACGTTGAG) was not observed.

Platinated Oligonucleotides for Duplex Crosslinking

293

Figure 9.16 Duplex formation between oligonucleotides selectively platinated at nucleobases modified with appending thioether and imidazole groups (XS and XI, respectively) and the complementary chain promotes migration of platinum from sulfur to guanine (N7), and crosslinks the two chains. (Adapted from Angew. Chem. Int. Ed., 2006, 45, 8194–8197.)

Figure 9.17 Crosslinked duplexes can be obtained by reacting transplatin, an oligonucleotide incorporating the XS and XI nucleobases at neighbouring positions and the complementary chain. The metal links the imidazole ring and the N7 of the guanine opposite XS. (Adapted from Angew. Chem. Int. Ed., 2006, 45, 8194–8197.)

These results showed that this methodology82,84 allowed targeting and crosslinking of complementary oligonucleotides without sequence restriction, and they opened up the possibility of using such platinated oligonucleotides in experiments aiming at the control of gene expression, targeting either mRNAs or miRNAs. The possibility of obtaining crosslinked duplexes without previous platination of one of the two strands was also explored, as Orgel and collaborators had done earlier.88,89 One-pot crosslinking experiments were carried out by mixing equimolar amounts of the modified oligonucleotide, the complementary chain and transplatin in duplex-forming conditions (Figure 9.17). The crosslinked duplex formed was the same as in the previous experiments, which indicated that previous isolation of the platinated modified oligonucleotide was unnecessary. Moreover, the thioether–Pt– imidazole chelate was always detected in the one-pot reaction mixture, which suggested that its formation preceded that of the crosslinked duplex. The fact that no

294


other chelates were detected again confirmed that the high affinity of sulfur for the metal led to the formation of the intrachain thioether–Pt–imidazole adduct, preventing other oligonucleotide sequences from being platinated, even when they contained various guanine residues. The ligand-exchange substitution process was highly regioselective, since, in crosslinking experiments involving platinated 5′dXIXSACGTTGAG and complementary oligonucleotides differing in the overhang sequence, migration from the thioether to the guanine opposite XS was always preferred, even over migration to the guanine opposite XI. Noncomplementary strands did not become crosslinked, which confirmed that crosslinking was the result of annealing-promoted rearrangement. Hence, the crosslinking process was also sequence selective, since it only took place if the two chains were complementary. Although reaction times were high, particularly when the modified oligonucleotide was not preplatinated, changes in the modified oligonucleotide and the complementary chain could accelerate the crosslinking reaction, as shown by Giraud-Paris and Leng in their experiments with oligonucleotides containing a platinated GNG triplet.92,94

9.5 Use of Platinated Oligonucleotides for Triplex Crosslinking As previously stated (Section 9.1), one of the problems of antigene strategy is the low stability of triple helices in physiological conditions, as the third strand can be displaced by the replication or the transcription machineries. Taking into account the requirements of the parallel motif for triplex formation (polypyrimidine oligonucleotides recognize polypurine-polypyrimidine tracks of DNA), in 1996 Leng and coworkers attempted to covalently attach the third oligonucleotide strand to the duplex by using transplatin as the crosslinking reagent.98 Then, polypyrimidine oligonucleotides containing a single guanine either in the middle or at the 5′-end of the sequence were platinated at acidic pH, which enabled the regioselective formation of the monofunctional adduct at the N7 of guanine. Both platinated oligonucleotides bound the ‘complementary’ duplex and formed interstrand crosslinks that involved almost exclusively guanines within the purine-rich strand. The crosslinking yield was higher when the monofunctional adduct was located at the 5′-end of the third strand because the triplex was more stable than when the platinum adduct was located in the middle of the sequence. In addition, several oligonucleotides containing a monofunctional adduct at cytosine were evaluated in crosslinking experiments and, although the triplex partially dissociated, the interstrand crosslinks were also formed. The rate of the crosslinking reaction depended on the nature of the adduct, but not strongly on its location.99 The problem associated with this strategy is, again, the low stability of monofunctional adducts, which may react with many different compounds (glutathione, proteins) before reaching dsDNA. Moreover, as reported by Leng et al.,98 platinated oligonucleotides with one labile bond can be deactivated by suicide reactions to form intrastrand crosslinks. More recently, Miller et al. examined the effect of oligonucleotides platinated at N4-(2-aminoethyl)cytosines85 (with Pt(dien) or cis-Pt(NH3)2Cl moieties) in

References

295

triplex stability (see Figure 9.12). As in the results reported previously,59,60 the Pt(dien) group sterically interfered with the formation of the triplex when it was located in the middle of the sequence, but triplex stability was restored when this group was tethered to the 5′-end of the triplex-forming oligonucleotide via a 2-aminoethylcarbamate linkage. In crosslinking experiments, the cisplatin monoadduct-containing oligonucleotide did not produce any significant interstrand crosslink with the DNA duplex, which was again attributed to steric hindrance. As indicated in Section 9.3.2, the McLaughlin group had reported the synthesis of oligothymidine sequences tethering a cis-bifunctional platinated complex at the 5′-end, for use in DNA duplex targeting by oligodeoxynucleotide-mediated triplex formation.59,60 These platinated oligonucleotides retained crosslinking ability because of the presence of the two cis chloride ligands attached to the platinum centre. Although a dT8 sequence was not long enough to facilitate triplex formation and hence platinum-mediated crosslinking, the dT15 sequence formed a triplex and crosslinked duplexes containing two noncomplementary overhangs and a GpG sequence at various sites along the overhang. Platinum-mediated crosslinking occurred exclusively with the purine-rich strand of the duplex target directly involved in triplex formation, and replacement of guanines by thymines or 7-deazaguanines in this strand prevented the formation of the interstrand crosslink. Unfortunately, no further progress has been reported.

References 1. Zamecnik, P.C.; Stephenson, M.L.; Inhibition of Rous sarcoma virus replication and cell transformation by a specific oligodeoxynucleotide; Proc. Natl. Acad. Sci. USA, 1978, 75, 280–284. 2. Aboul-Fadl, T.; Antisense oligonucleotides: The state of the art; Curr. Med. Chem., 2005, 12, 2193–2212. 3. Wei-Hua, P.; Clawson, G.A.; Antisense applications for biological control; J. Cell Biochem., 2006, 98, 14–35. 4. Marwick, C.; First ‘antisense’ drug will treat CMV retinitis; J. Am. Med. Assoc., 1998, 280, 871. 5. Potera, C.; Antisense-down, but not out; Nat. Biotechnol., 2007, 25, 497–499. 6. Fichou, Y.; Férec, C.; The potential of oligonucleotides for therapeutic application; Trends Biotechnol.; 2006, 24, 563–570. 7. Kruger, K.; Grabowski, P.J.; Zaug, A.J.; Sands, J.; Gottschling, D.E.; Cech, T.R.; Self-splicing RNA: autoexcision and autocyclization of the ribosomal RNA intervening sequence of Tetrahymena; Cell, 1982, 31, 147–157. 8. Uhlenbeck, O.C.; A small catalytic oligoribonucleotide; Nature, 1987, 328, 596–600. 9. Khan, A.U.; Ribozyme: A clinical tool; Clin. Chim. Acta, 2006, 367, 20–27. 10. Fire, A.; Xu, S.; Montgomery, M.K.; Kostas, S.A.; Driver, S.E.; Mello, C.C.; Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans; Nature, 1998, 391, 806–811. 11. Chakraborty, C.; Potentiality of small interfering RNAs (siRNA) as recent therapeutic targets for gene-silencing; Curr. Drug Targets, 2007, 8, 469–482. 12. Kim, D.H.; Rossi, J.J.; Strategies for silencing human disease using RNA interference; Nat. Rev. Genet.; 2007, 8, 173–184. 13. Ameres, S.L.; Martinez, J.; Schroeder, R.; Molecular basis for target RNA recognition and cleavage by human RISC; Cell, 2007, 130, 101–112.

296


14. Gruenweller, A.; Hartmann, R.K.; RNA interference as a gene-specific approach for molecular medicine; Curr. Med. Chem., 2005, 12, 3143–3161. 15. Ambros, V.; The functions of animal microRNAs; Nature, 2004, 431, 350–355. 16. Zhang, B.; Wang, Q.; Pan, X.; MicroRNAs and their regulatory roles in animals and plants; J. Cell. Physiol., 2007, 210, 279–289. 17. Mack, G.S.; MicroRNA gets down to business; Nat. Biotechnol., 2007, 25, 631–638. 18. Kruetzfeldt, J.; Rajewsky, N.; Braich, R.; Rajeev, K.G.; Tuschl, T.; Manoharan, M.; Stoffel, M.; Silencing of microRNAs in vivo with ‘antagomirs’; Nature, 2005, 438, 685–689. 19. Weiler, J.; Hunziker, J.; Hall, J.; Anti-miRNA oligonucleotides (AMOs): ammunition to target miRNAs implicated in human disease? Gene Ther., 2006, 13, 496–502. 20. Xodo, L.E.; Cogoi, S.; Rapozzi, V.; Antigene strategies to down-regulate gene expression in mammalian cells; Curr. Pharm. Des.; 2004, 10, 805–819. 21. Francois, J.C.; Saison-Behmoaras, T.; Hélène, C.; Sequence-specific recognition of the major groove of DNA by oligodeoxynucleotides via triple helix formation. Footprinting studies; Nucleic Acids Res., 1988, 16, 11431–11440. 22. Vasquez, K.M.; Narayanan, L.; Glazer, P.M.; Specific mutations induced by triplex-forming oligonucleotides in mice; Science, 2000, 290, 530–533. 23. Ellington, A.D.; Szostak, J.W.; In vitro selection of RNA molecules that bind specific ligands; Nature, 1990, 346, 818–822. 24. Tuerk, C.; Gold, L.; Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase; Science, 1990, 249, 505–510. 25. Gragoudas, E.S.; Adamis, A.P.; Cunningham, E.T.; Feinsod, M.; Guyer, D.R.; Pegaptanib for neovascular age-related macular degeneration; N. Eng. J. Med.; 2004, 351, 2805–2816. 26. Bunka, D.H.J.; Stockley, P.G.; Aptamers come of age – at last; Nat. Rev. Microbiol., 2006, 4, 588–596. 27. Famulok, M.; Hartig, J.S.; Mayer, G.; Functional aptamers and aptazymes in biotechnology, diagnostics and therapy; Chem. Rev., 2007, 107, 3715–3743. 28. Crooke, S.T.; Molecular mechanisms of action of antisense oligonucleotides; Biochim. Biophys. Acta, 1999, 1489, 31–44. 29. Eckstein, F.; Nucleoside phosphorothioates; J. Am. Chem. Soc.; 1966, 88, 4292–4294. 30. Summerton, J.E.; Morpholino, siRNA and S-DNA compared: impact of structure and mechanism of action on off-target effects and sequence specificity; Curr. Top. Med. Chem., 2007, 7, 651–660. 31. Singh, S.K.; Nielsen, P.; Koshkin, A.A.; Wengel, J.; LNA (locked nucleic acids): synthesis and high-affinity nucleic acid recognition; J. Chem. Soc. Chem. Commun., 1998, 455–456. 32. Nielsen, P.E.; Egholm, M.; Berg, R.H.; Buchardt, O.; Sequence-selective recognition of DNA by strand displacement by a thymine-substituted polyamide; Science, 1991, 254, 1497–1500. 33. Summerton, J.; Morpholino antisense oligomers: the case for an RNase H-independent structural type; Biochim. Biophys. Acta, 1999, 1489, 141–158. 34. Glick, G.D.; Design, synthesis and analysis of conformationally constrained nucleic acids; Biopolymers, 1998, 48, 83–96. 35. Wilds, C.J.; Noronha, A.M.; Robidoux, S.; Miller, P.S.; Mispair-aligned N3T-alkyl-N3T interstrand cross-linked DNA: synthesis and characterization of duplexes with interstrand cross-links of variable lengths; J. Am. Chem. Soc., 2004, 126, 9257–9265. 36. Noronha, A.M.; Wilds, C.J.; Miller, P.S.; N4C–alkyl–N4C cross-linked DNA: bending deformations in duplexes that contain a –CNC– interstrand cross-link; Biochemistry, 2002, 41, 8605–8612. 37. Letsinger, R.L.; Wu, T.; Use of a stilbenedicarboxamide bridge in stabilizing, monitoring, and photochemically altering folded conformations of oligonucleotides. J. Am. Chem. Soc., 1995, 117, 7323–7328. 38. Stutz, A.; Langenegger, S.M.; Haner, R.; Phenanthrene-derived DNA hairpin mimics; Helv. Chim. Acta, 2003, 86, 3156–3163.

References

297

39. Lewis, F.D.; Helvoigt, S.A.; Letsinger, R.L.; Synthesis and spectroscopy of Ru(II)-bridged DNA hairpins; J. Chem. Soc. Chem. Commun., 1999, 327–328. 40. Bianke, G.; Haner, R.; A metal-coordinating DNA hairpin mimic. ChemBioChem., 2004, 5, 1063–1068. 41. Ali, M.M.; Oishi, M.; Nagatsugi, F.; Mori, K.; Nagasaki, Y.; Kataoka, K.; Sasaki, S.; Intracellular inducible alkylation system that exhibits antisense effects with greater potency and selectivity than the natural oligonucleotide; Angew. Chem. Int. Ed., 2006, 45, 3136–3140. 42. Zhou, Q.; Rokita, S.E.; A general strategy for target-promoted alkylation in biological systems; Proc. Natl. Acad. Sci. USA, 2003, 100, 15452–15457. 43. Hägerlöf, M.; Hedman, H.; Elmroth, S.K.C.; Platination of the siRNA sense-strand modulates both efficacy and selectivity in vitro; Biochem. Biophys. Res. Commun., 2007, 361, 14–19. 44. Baik, M.H.; Friesner, R.A.; Lippard, S.J.; Theoretical study of cisplatin binding to purine bases: Why does cisplatin prefer guanine over adenine? J. Am. Chem. Soc., 2003, 125, 14082–14092. 45. Comess, K.M.; Costello, C.E.; Lippard, S.J.; Identification and characterization of a novel linkage isomerization in the reaction of trans-diamminedichloroplatinum(II) with 5′d(TCTACGCGTTCT); Biochemistry, 1990, 29, 2102–2110. 46. Berghoff, U.; Schmidt, K.; Janik, M.; Schroder, G.; Lippert, B.; Monofunctional transPtII(NH3)2 modification of pyrimidine-rich oligodeoxyribonucleotides: direct platination and use of a protective group; Inorg. Chim. Acta, 1998, 269, 135–142. 47. Lepre, C.A.; Chassot, L.; Costello, C.E.; Lippard, S.J.; Synthesis and characterization of trans-Pt(NH3)2Cl2] adducts of d(CCTC)·(GAGTCTCC) d(GGAGACTCGAGG); Biochemistry, 1990, 29, 811–823. 48. Janik, M.B.L.; Lippert, B.; Trans-(NH3)2PtII-modified deoxyoligonucleotides as potential antisense agents: cross-linking reactions between two 12-mers; J. Biol. Inorg. Chem., 1999, 4, 645–653. 49. Brabec, V.; Leng, M.; DNA interstrand cross-links of trans-diamminedichloroplatinum(II) are preferentially formed between guanine and complementary cytosine residues; Proc. Natl. Acad. Sci. USA, 1993, 90, 5345–5349. 50. Brabec, V.; Sip, M.; Leng, M.; DNA conformational change produced by the site-specific interstrand cross-link of trans-diamminedichloroplatinum(II); Biochemistry, 1993, 32, 11676–11681. 51. Paquet, F.; Boudvillain, M.; Lancelot, G.; Leng, M.; NMR solution structure of a DNA dodecamer containing a transplatin interstrand GN7-CN3 cross-link; Nucleic Acids Res., 1999, 27, 4261–4268. 52. Dalbies, R.; Payet, D.; Leng, M.; DNA double helix promotes a linkage isomerization reaction in trans-diamminedichloroplatinum(II) (transplatin)-modified DNA; Proc. Natl. Acad. Sci. USA, 1994, 91, 8147–8151. 53. Dalbies, R.; Boudvillain, M.; Leng, M.; Linkage isomerization reaction of intrastrand cross-links in trans-diamminedichloroplatinum(II)-modified single-stranded oligonucleotides; Nucleic Acids Res., 1995, 23, 949–953. 54. Schmidt, K.S.; Filippov, D.V.; Meeuwenoord, N.J.; van der Marel, G.A.; Van Boom, J.H.; Lippert, B.; Reedijk, J.; Solid-phase synthesis of a monofunctional trans-a2PtII complex tethered to a single-stranded oligonucleotide; Angew. Chem. Int. Ed., 2000, 39, 375–377. 55. Takahara, P.M.; Rosenzweig, A.C.; Frederick, C.A.; Lippard, S.J.; Crystal structure of double-stranded DNA containing the major adduct of the anticancer drug cisplatin; Nature, 1995, 377, 649–652. 56. Silverman, A.P.; Bu, W.; Cohen, S.M.; Lippard, S.J.; 2.4 Å Crystal structure of the asymmetric platinum complex {Pt(ammine)(cyclohexylamin)}2+ bound to a dodecamer DNA duplex; J. Biol. Chem.; 2002, 277, 49743–49749. 57. Baruah, H.; Wright, M.W.; Bierbach, U.; Solution structural study of a DNA duplex containing the guanine-N7 adduct formed by a cytotoxic platinum-acridine hybrid reagent; Biochemistry, 2005, 44, 6059–6070.

298


58. Coste, F.; Malinge, J-M.; Serre, L.; Shepard, W.; Roth, M.; Leng, M.; Zelwer, C.; Crystal structure of a double-stranded DNA containing a cisplatin interstrand cross-link at 1.63 Å resolution: hydration at the platinated site; Nucleic Acids Res., 1999, 27, 1837–1846. 59. Sharma, S.K.; McLaughlin, L.W.; Cross-linking of a DNA conjugate tethering a cisbifunctional platinated complex to a target DNA duplex; J. Am. Chem. Soc., 2002, 124, 9658–9659. 60. Sharma, S.K.; McLaughlin, L.W.; Triplex mediated delivery of a platinum complex to a specific DNA target site; J. Inorg. Biochem., 2004, 98, 1570–1577. 61. Beaucage, S.L.; Radhakrishnan, P.; Advances in the synthesis of oligonucleotides by the phosphoramidite approach; Tetrahedron, 1992, 48, 2223–2311. 62. Garegg, P.J.; Regbert, T.; Stawinski, J.; Stromberg, R.; Formation of internucleotidic bonds via phosphonate intermediates; Chem. Scr., 1985, 25, 280–282. 63. Froehler, B.C.; Matteucci, M.D.; Nucleoside H-phosphonates: Valuable intermediates in the synthesis of deoxyoligonucleotides; Tetrahedron Lett.; 1986, 27, 469–472. 64. Schliepe, J.; Berghoff, U.; Lippert, B.; Cech, D.; Automated solid phase synthesis of platinated oligonucleotides via nucleoside phosphonates; Angew. Chem. Int. Ed., 1996, 35, 646–648. 65. Manchanda, R.; Dunham, S.U.; Lippard, S.J.; Automated solid-phase synthesis of site-specifically platinated oligodeoxyribonucleotides; J. Am. Chem. Soc.; 1996, 118, 5144– 5145. 66. Ren, S.; Cai, L.; Segal, B.M.; Synthesis and characterization of monodeoxynucleotide tethered platinum-(II) and -(IV) complexes; J. Chem. Soc. Dalton Trans.; 1999, 1413–1422. 67. Schmidt, K.S.; Boudvillain, M.; Schwartz, A.; Van der Marel, G.A.; Van Boom, J.H.; Reedijk, J.; Lippert, B.; Monofunctionally trans-diammine platinum(II)-modified peptide nucleic acid oligomers: a new generation of potential antisense drugs; Chem. Eur. J.; 2002, 8, 5566–5570. 68. Heetebrij, R.J.; Tromp, R.A.; van der Marel, G.A.; Van Boom, J.H.; Reedijk, J.; A novel approach to site-specifically platinated oligonucleotides applying combinations of nucleobase protecting groups; J. Chem. Soc. Chem. Commun., 1999, 1693–1694. 69. Heetebrij, R.J.; de Kort, M.; Meeuwenoord, N.J.; den Dulk, H.; van der Marel, G.A.; Van Boom, J.H.; Reedijk, J.; A versatile approach towards regioselective platinated DNA sequences; Chem. Eur. J., 2003, 9, 1823–1827. 70. Cai, L.; Lim, K.; Ren, S.; Cadena, R.S.; Beck, W.T.; Synthesis and in vitro antitumor activity of oligonucleotide-tethered and related platinum complexes; J. Med. Chem., 2001, 44, 2959–2965. 71. Van Boom, S.S. G.E.; Reedijk, J.; Unprecedented migration of [Pt(dien)]2+ (dien = 1,5diamino-3-azapentane) from sulfur to guanosine-N7 in S-guanosyl-L-homocysteine (sgh); J. Chem. Soc. Chem. Commun., 1993, 1397–1398. 72. Barnham, K.J.; Djuran, M.I.; del Socorro Murdoch, P.; Sadler, P.J.; Intermolecular displacement of S-bound L-methionine on platinum(II) by guanosine 5′-monophosphate: implications for the mechanism of action of anticancer drugs; J. Chem. Soc. Chem. Commun., 1994, 721–722. 73. Fröhling, C.D.W.; Sheldrick, W.S.; Intramolecular competition between histidine and methionine side chains in reactions of dipeptides with [Pt(en)(H2O)2]2+ (en = H2NCH2CH2NH2); J. Chem. Soc. Dalton Trans., 1997, 4411–4420. 74. Fröhling, C.D.W.; Sheldrick, W.S.; Intramolecular migration of [Pt(dien)]2+ (dien = 1,5diamino-3-azapentane) from sulfur to imidazole-N1 in histidylmethionine (his-metH); Chem. Commun., 1997, 1737–1738. 75. Hahn, M.; Wolters, D.; Sheldrick, W.S.; Hulsbergen, F.B.; Reedijk, J.; [Pt(dien)]2+ migrates intramolecularly from methionine S to imidazole Ne 2 in the peptides H-His-Gly-Met-OH and Ac-His-Ala-Ala-Ala-Met-NHPh; J. Biol. Inorg. Chem.; 1999, 4, 412–420. 76. Van Boom, S.S.G.E.; Chen, B.W.; Teuben, J.M.; Reedijk, J.; Platinum-thioether bonds can be reverted by guanine-N7 bonds in Pt(dien)2+ model adducts; Inorg. Chem., 1999, 38, 1450–1455.

References

299

77. Reedijk, J.; Why does cisplatin reach guanine-N7 with competing S-donor ligands available in the cell?; Chem. Rev., 1999, 99, 2499–2510. 78. Marchán, V.; Moreno, V.; Pedroso, E.; Grandas, A.; Towards a better understanding of the cisplatin mode of action; Chem. Eur., J. 2001, 7, 808–815. 79. Hahn, M.; Kleine, M.; Sheldrick, W.S.; Interaction of cisplatin with methionine- and histidine-containing peptides: competition between backbone binding, macrochelation and peptide cleavage; J. Biol. Inorg. Chem., 2001, 6, 556–566. 80. Marchán, V.; Pedroso, E.; Grandas, A.; Insights into the reaction of transplatin with DNA and proteins: Methionine-mediated formation of histidine-guanine trans-Pt(NH3)2 crosslinks; Chem. Eur. J., 2004, 10, 5369–5375. 81. Deubel, D.V.; On the competition of the purine bases, functionalities of peptide side chains, and protecting agents for the coordination sites of dicationic cisplatin derivatives; J. Am. Chem. Soc., 2002, 124, 5834–5842. 82. Algueró, B.; Pedroso, E.; Marchán, V.; Grandas, A.; Incorporation of two modified nucleosides allows selective platination of an oligonucleotide making it suitable for duplex cross-linking; J. Biol. Inorg. Chem., 2007, 12, 901–911. 83. Allerson, C.R.; Chen, S.L.; Verdine, G.L.; A chemical method for site-specific modification of RNA: The convertible nucleoside approach; J. Am. Chem. Soc., 1997, 119, 7423– 7433. 84. Algueró, B.; López de la Osa, J.; González, C.; Pedroso, E.; Marchán, V.; Grandas, A.; Selective platination of modified oligonucleotides and duplex cross-links; Angew. Chem. Int. Ed., 2006, 45, 8194–8197. 85. Campbell, M.A.; Mason, T.M.; Miller, P.S.; Interactions of platinum(II)-derivatized triplexforming oligonucleotides with DNA. Can. J. Chem., 2007, 85, 241–248. 86. Vlasov, V.V.; Gorn, V.V.; Ivanova, E.M.; Kazakov, S.A.; Mamaev, S.V.; Complementary addressed modification of oligonucleotide d(pGpGpCpGpGpA) with platinum derivative of oligonucleotide d(pTpCpCpGpCpCpTpTpT); FEBS Lett., 1983, 162, 286–289. 87. Strothkamp, K.G.; Lippard, S.J.; Platinum binds selectively to phosphorothioate groups in mono- and polynucleotides: a general method for heavy metal staining of specific nucleotides; Proc. Natl. Acad. Sci. USA, 1976, 73, 2536–2540. 88. Chu, B.C.F.; Orgel, L.E.; Optimization of the efficiency of cross-linking PtII oligonucleotide phosphorothioate complexes to complementary oligonucleotides; Nucleic Acids Res., 1990, 18, 5163–5171. 89. Gruff, E.S.; Orgel, L.E.; An efficient, sequence-specific method for crosslinking complementary oligonucleotides using binuclear platinum complexes; Nucleic Acids Res., 1991, 19, 6849–6854. 90. Chu, B.C.F.; Orgel, L.E.; Crosslinking transcription factors to their recognition sequences with platinum(II) complexes; Nucleic Acids Res., 1992, 20, 2497–2502. 91. Escaffre, M.; Chottard, J.C.; Bombard, S.; Rearrangement of a 1,3-trans-{Pt(NH3)2[(GXG)N7G,N7G]} intrastrand cross-link into interstrand cross-links within RNA duplexes; Nucleic Acids Res., 2002, 30, 5222–5228. 92. Boudvillain, M.; Guerin, M.; Dalbies, R.; Saison-Behmoaras, T.; Leng, M.; Transplatinmodified oligo(2′-O-methyl ribonucleotide)s: a new tool for selective modulation of gene expression; Biochemistry, 1997, 36, 2925–2931. 93. Andersen, B.; Bernal-Mendez, E.; Leng, M.; Sletten, E.; NMR solution structure of a DNA 12/11-mer: d(CTCCTGTGTCTC) · d(GAGATA-AGGAG) containing a transplatin interstrand G-N7/A-N1 cross-link; Eur. J. Inorg. Chem., 2000, 6, 1201–1210. 94. Giraud-Panis, M.J.; Leng, M.; Transplatin-modified oligonucleotides as modulators of gene expression; Pharmacol. Ther., 2000, 85, 175–181. 95. Aupeix-Scheidler, K.; Chabas, S.; Bidou, L.; Rousset, J.P.; Leng, M.; Toulme, J.J.; Inhibition of in vitro and ex vivo translation by a transplatin-modified oligo (2′-Omethylribonucleotide) directed against the HIV-1 gag-pol frameshift signal; Nucleic Acids Res., 2000, 28, 438–445.

300


96. Perrier, S.; Seela, F.; Schwartz, A.; Leng, M.; Chottard, J.C.; The human telomeric sequence (T2AG3)n is efficiently cross-linked by AN1 binding to the platinum of a trans-Pt(NH3)2 chelate of an antisense oligo-2′-O-methylribonucleotide; Eur J. Inorg. Chem., 2003, 8, 1641–1644. 97. Muller, J.; Drumm, M.; Boudvillain, M.; Leng, M.; Sletten, E.; Lippert, B.; Parallel-stranded DNA with Hoogsteen base pairing stabilized by a trans-[Pt(NH3)2]2+ cross-link: characterization and conversion into a homodimer and a triplex; J. Biol. Inorg. Chem.; 2000, 5, 603–611. 98. Colombier, C.; Lippert, B.; Leng, M.; Interstrand cross-linking reaction in triplexes containing a monofunctional transplatin-adduct; Nucleic Acids Res., 1996, 24, 4519–4524. 99. Bernal-Mendez, E.; Sun, J.; Gonzalez-Vilchez, F.; Leng, M.; Reactivity of transplatinmodified oligonucleotides in triple-helical DNA complexes. New J. Chem., 1998, 22, 1479–1483.

10 Rhodium– and Tin–DNA Interactions and Applications Kenneth D. Camm and Patrick C. McGowan

10.1 Introduction The ubiquity of transition metals employed in biological processes is a direct consequence of the versatile chemistry such metals exhibit. In particular, the multiple oxidation states and diverse structural motifs accessible to transition metals have resulted in their incorporation into complex ‘bioinorganic’ species in nature. These same properties have facilitated the development of artificial transition metal complexes, designed to modify or regulate many biological processes.1 The Lewis acidity of transition metals, and consequently their affinity for Lewis bases such as N, O or S donor ligands, renders transition metal complexes suitable for interaction with a multitude of biologically relevant species, including nucleic acids (DNA and RNA). Cellular function and replication relies on the chemistry and interactions of nucleic acids, and consequently these species have frequently been targeted in therapeutic protocols for diseases as varied as cancer and arthritis.2 The first, and to-date most successful, inorganic complex employed in the treatment of cancer, [Pt(NH3)2Cl2] (cisplatin, Figure 10.1a)3,4 and its derivatives, is highly effective against testicular and ovarian cancer, and in other malignancies including cervical, bladder, head and neck tumours. The cytotoxic activity of cisplatin is a direct consequence of covalent interactions between the Pt centre and N7s of adjacent guanine residues on DNA, to form a covalent 1,2-intrastrand d(GpG)crosslink.4 Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

302

Rhodium– and Tin–DNA Interactions and Applications

Crosslinking results in inhibition of transcription and DNA replication, eventually leading to cell death. Following the success of cisplatin, a catalogue of research has accumulated, investigating the interactions of other transition metals with DNA.1,2,5–9 Much research has focused on the development of novel therapeutic protocols that may circumvent the toxicity issues associated with cisplatin chemotherapy, and to probe the nature of nucleic acid structure and function in biological processes. To this end, investigations have been performed on metal–DNA interactions covering a large majority of the periodic table, many of which are reviewed in this publication. Herein we describe recent developments focusing on the interactions of rhodium and tin with DNA, and how such work has contributed to our understanding of the role of nucleic acids in biological processes and facilitated the development of novel antitumour agents.

10.2 Metal–DNA Interactions There are three principle interactions that occur between metal complexes and oligonucleotides such as DNA. The first involves covalent interactions between the Lewis acidic metal and nucleic acid Lewis bases, such as nucleophilic guanine N7 residues. The second interaction occurs through intercalation of a small molecule between stacked base pairs of the double helix structure. Both such interactions cause disruption of base stacking in the double helix through interference with van der Waals or hydrogen-bonding interactions, and thus result in destabilization of double-helix conformation. The third interaction is electrostatic, occurring between a cationic complex and the anionic phosphate backbone.2,8,10 Neutralization of this negative charge reduces repulsive forces between adjacent phosphate groups, thus stabilizing the double helix.11,12 Of note is the fact that the former two interactions are often selective, while the electrostatic interactions may occur at any phosphate site along the phosphate backbone.13,14 These interactions can cause changes in the DNA structure, affecting cellular processes and potentially leading to cell death.4,8,10

10.3 Rhodium–DNA Interactions 10.3.1 Rhodium(I) Complexes Encouraged by the promising chemotherapeutic properties of cisplatin, a number of groups in the 1970s began to investigate the potential antineoplastic activity of other platinum group transition metal complexes, including rhodium. Preliminary work investigated the effects of cis-, square planar organometallic Rh(I) complexes of type [Rh(COD)L2]+ (L = 2,2′-bipyridine; 1,10-phenanthroline; piperidine) on growth of a number of tumours.15 In all cases, however, the Rh(I) complexes were less active than cisplatin. Similar observations were made on a series of related

Rhodium–DNA Interactions

303

complexes.16 The reduced activity of such species was attributed to the instability of the Rh(I) centre towards oxidation to the kinetically inert octahedral Rh(III) complex in aqueous solution and therefore in vivo. In addition, the greatly enhanced lability of the leaving groups on Rh(I) versus Pt(II) complexes suggested that the Rh(I) species would be oxidized, and thus deactivated before reaching the target. No advanced studies on the interaction of Rh(I) species with oligonucleotides such as DNA were performed owing to these limitations.

10.3.2 Dirhodium(II) Complexes Dirhodium carboxylate complexes of general formula (Rh2(O2CR)4L2 (R = Me, Et, Pr; L = solvent) (Figure 10.1b) have been found to inhibit DNA replication and protein synthesis with a similar potency to cisplatin.17 In the examples given above, antitumour activity was found to increase with increasing lipophilicity of the R group, presumably through increased cell membrane permeability of the complex. Related complexes bearing various bidentate bridging species such as nitrogencontaining ligands18–20 (Figure 10.1c), phosphine ligands,21 (Figure 10.1d) and formamidinates22 (Figure 10.1e) have been developed, and exhibit antitumour activity comparable to, or higher than, cisplatin.17 Structure–activity relationships are, however somewhat more complex than the two-pole complimentary principle suggested by Yang et al. for organometallic antitumour agents.23 The similarities in antitumour properties between cisplatin derivatives and dirhodium complexes may be unexpected, owing to the obvious structural differences between the two species. Indeed the dirhodium complexes are capable of binding strongly to adenine residues, unlike cisplatin, which has a preference for guanine.17 Model experiments employed nucleo(s)tides to demonstrate that the interaction between rhodium and adenine residues is stabilized by favourable hydrogen bonds between the purine exocyclic NH2(6) group and a carboxylate oxygen atom of the dirhodium complex. In contrast, axial coordination of guanine residues is inhibited by electrostatic repulsions between the ketone O6 and oxygen atoms of the carboxylate ligands present within the complex.17,24 In subsequent studies, the unlikely substitution of two bridging ligands on complexes of type [Rh2(O2CR)4L2] by two guanine residues was directly observed by X-ray crystallographic analysis.25,26 The guanine residues were found to form equatorial bridging interactions, through the N7/O6 sites in a headto-head (H-H) (e.g. cis-[Rh2-(m-O2CCH3)2(9-EtGuaH)2(Me2CO)(H2O)](BF4)2 (Figure 10.1f)26 or head-to-tail (H-T) orientation (e.g. cis-[Rh2-(m-O2CCF3)2(9EtGuaH)2(Me2CO)2](CF3CO2)2 (Figure 10.1g).25 These findings led to further studies on interactions between dirhodium complexes and the dinucleotide d(GpG), where a similar chelate N7/O6 interaction with Rh2(OAc)4 was observed.27,28 Although the dirhodium complexes were found to bind through both axial and equatorial positions, the mechanism of binding is believed to occur through initial interaction through axial sites, followed by rearrangement to a more thermodynamically stable equatorial position.29,30 Indeed, recent studies demonstrated that placing nonlabile ligands in the axial position resulted in a decrease in the ability of the complexes to inhibit transcription in vitro, demonstrating the necessity for labile axial ligands, and

304


Figure 10.1 Dinuclear Rh complexes

supporting the proposed rearrangement mechanism.31 The introduction of labile equatorial ligands may also facilitate ligand rearrangement to the equatorial position, and may therefore provide an explanation of the increased activity observed through introduction of electron-withdrawing groups, such as in Rh2(m-O2CF3)4, that exhibited activity comparable to that of cisplatin.32 In an extension to this work, dirhodium interactions with single-stranded oligonucleotides containing GG sites were investigated. This work further corroborated the correlation between lability of the ligands present and nucleic acid binding affinity, wherein the affinity of the dirhodium complexes were compared to cisplatin derivatives: cis-[Pt(NH3)2(H2O)2]2+ ≈ Rh2(m-O2CCF3)4 > cis-[Pt(NH3)2Cl2] >> cis-[Rh2(m-O2CCH3)2(NCCH3)6](BF4)2 > Rh2(mO2CCH3)4.17,33,34 More recent work has refuted the conclusions of earlier studies that suggested dirhodium complexes did not interact with double-stranded DNA. Indeed, interaction of a number of species demonstrated the presence of covalently linked DNA adducts. Significantly, these studies demonstrated the formation of stable DNA interstrand crosslinks35 Although specific binding has yet to be ascertained, the presence of a mixture of mono- and bifunctional adducts, including the more stable equatorial adducts was observed. Once again, the binding modes determined in model studies corroborated with data from the studies performed on doublestranded DNA, wherein an increased lability of the leaving groups corresponds to an increase in occurrence of interstrand crosslinks.36 As observed for binding of guanine residues to dirhodium complexes, bidentate equatorial binding of adenine residues on dinucleotide d(ApA) was observed, with the d(ApA) fragment


305

spanning the Rh–Rh bond via N7/N6 bridges.33,37 Such observations again demonstrate the potential for dirhodium complexes to form interstrand crosslinks. The ability of dirhodium complexes to bind either adenine or guanine residues on plasmid pUC19 DNA was recently demonstrated, further disproving previous conclusions concerning the inability of such species to bind double-stranded DNA.38

10.3.3 Rhodium Complexes in Photodynamic Therapy Owing to the d-electron availability and interchangeable redox properties of transition metals, such complexes have found application in the area of photodynamic therapy (PDT).39 The DNA toxicity of oxygen-derived intermediates from PDT are well established, and light-induced formation of 1O2, O2− and •OH in the vicinity of DNA results in considerable degradation of DNA structure. The development of novel photoreagents based on bimetallic rhodium complexes has received considerable attention, owing to the propensity of such species to induce metal–metal charge transfer excited states to promote DNA cleavage. Dunbar and coworkers investigated the complex Rh2(m-O2CCH3)4, for potential photocleavage of DNA (Figure 10.2a).40 The d7-d7 Rh(II)-Rh(II) core was found to promote photoinduced cation formation in the presence of an electron acceptor. Subsequent DNA ‘nicking’ then occurred.41 Following these observations, a series of potential photoactivated complexes designed to interact with DNA both through coordination and intercalation were developed (Figure 10.2b).40,42,43 The complexes studied, with the exception of [Rh2(mO2CCH3)2(CH3CN)6](BF4)2,43 consisted of a chelating intercalating ligand on one of the two rhodium centres, with labile cis ligands on the second to facilitate covalent coordination to DNA nucleobases. Such species demonstrated DNA nicking upon photolysis, and cytotoxicity towards human skin cell lines.42 The dual binding capability of the related complex cis-[Rh2(dap)(m-O2CCH3)2(h1O2CCH3)(CH3OH)](O2CCH3) (dap = 1,12-diazaperylene) (Figure 10.2c),44 through both coordination and intercalation interactions with DNA, was elegantly demonstrated using 2D NMR studies.45 Similar studies were recently performed on a complex with two stable formamidinate and two labile trifluoroacetate bridging groups.37 cis-Rh2(DTolF)2(O2CCF3)2(H2O)2 (DTolF = N,N′-di-p-tolylformamidinate) (Figure 10.1e) has exhibited antitumour activity comparable to that of cisplatin against Yoshida ascites and T8 sarcomas, with a significant reduction in toxicity, showing virtually no toxicity in mice.22 Other heterobimetallic photodynamic species, bearing a Rh–Ru core have been developed,46 in which a low energy Ru–Rh metal–metal charge transfer excited state results in direct DNA nicking in the absence of electron acceptors. The extensive and elegant work performed by Dunbar and others has shed light on the potential mechanisms of activity of the structurally diverse dirhodium complexes. Despite the obvious differences in structure and reactivity between mononuclear cisplatin derivatives and the bimetallic rhodium complexes investigated, such research has demonstrated that the mode of activity may indeed be through formation of interstrand crosslinks between the rhodium metal centres and adjacent guanine or adenine residues on DNA. Through increased understanding

306


Figure 10.2 Dinuclear rhodium complexes for photodynamic therapy

of the interactions between dirhodium species and DNA, novel complexes have been designed to bind DNA through both coordinative and intercalative interactions. Dimeric rhodium complexes have often demonstrated antitumour activity in the same order of magnitude as cisplatin, while potentially circumventing the toxicity issues associated with the successful drug. Furthermore, research into the potential of dirhodium complexes in photodynamic therapy may result in replacements for cisplatin derivatives in the foreseeable future.17–38, 40–46 10.3.4 Rhodium(III) Complexes The kinetic inertness of rhodium(III) complexes may be believed to preclude the ability of such species to modify or probe DNA and RNA function through covalent interactions. Indeed, divalent Pt, divalent (and trivalent) Ru species have received considerable attention in this field, directly owing to the lability and substitution chemistry of their ligands. Rhodium(III) complexes are low in the list of potential DNA interacting agents, owing to their inability to undergo substitution reactions on a physiologically relevant timescale.47 Further application is also limited by the inability of Rh(III), in contrast to Ru(III), to undergo activation through reduction to the divalent species. As a consequence, few cytotoxic Rh(III) complexes have been developed.8,10,48 A number of Rh(III) analogues of known antitumour active Ru(III) species have also been investigated. Like fac-[RuCl3(NH3)3], mer[RhCl3(NH3)3] is active, but insoluble.8 Rh(III) analogues of the active Ru(III) species, Na[trans-[RhCl4(Me2SO)(Im)] and (HIm)trans-[RhCl4(Im)2] were essentially inactive, exhibiting modest inhibition on the growth of the primary MCa mammary tumours.49 The limited activity in each case has been attributed to the kinetic stability of the Rh(III) relative to the Ru(III) systems towards ligand substitution.


307

The concept of ‘photochemically’ activated Rh(III) cisplatin analogues has been investigated for complexes of the type cis-[RhCl2(polypyridyl)2]+.50,51 For example, cis-[RhCl2(phen)2]+ (phen 1,-10-phenanthroline) was found to undergo light-induced covalent binding to DNA.52 Although kinetically inert under thermal conditions, irradiation of the complexes results in d-electron promotion into ligandfield bands, thus increasing the rate of ligand exchange by orders of magnitude relative to thermal conditions. The reaction with N7 of guanine residues occurred owing to reductive quenching of the excited state by nucleobase–metal charge transfer, although with minimal biological effects.50 Such promising preliminary results led to the investigation of other complexes that demonstrated oxidative DNA strand scission.51 In particular, work by Barton and coworkers have led to many promising developments in this field. 10.3.5 Site-Specific Rhodium–DNA Interactions Barton and coworkers have produced a large body of work investigating the sitespecific interaction of Rh(III) intercalating complexes with DNA, with a view to introducing selective cleavage of the DNA strand in specific regions towards ‘sitedirected’ photodynamic therapy. Such work has also led to insights into the functions of nucleic acids in biological processes, and the complexes developed have been used both as photoactivated chemotherapeutic agents,51 and as probes for charge transport processes in DNA.53 The pairing of nucleobases in healthy cells occurs through combination of GC and AT residues. DNA base mismatches (i.e. CC, CA or CT combinations) arise owing to enzymatic errors or DNA damage during the process of genetic recombination or DNA replication.54,55 The inability of cells to repair mismatches is associated with cancerous transformations.56 Early work by Barton and coworkers involved the development of phenanthrenequinone diimine (phi) complexes of rhodium, that may intercalate into the major groove of DNA, and subsequently promote photoactivated strand scission.57–59 In an extension of this work, bulky heterocyclic aromatic imine ligands were introduced in complexes such as [Rh(bpy)2(chrysi)3+] (chrysi = 5,6chrysenequinonediimine, Figure 10.3).60 The bulky four-ring chrysene ligand, being too large to intercalate into standard DNA grooves, instead intercalated preferentially into the larger, perturbed site of the thermodynamically destabilized

Figure 10.3 [Rh(bpy)2(chrysi)3+] (chrysi = 5,6-chrysenequinonediimine)

308


mismatch. Indeed, a direct correlation was observed between the extent to which a mismatch is destabilized (which depends on the bases involved) and the binding affinity of the complex for that mismatch.61 The complexes exhibited remarkably specific mismatch recognition, with photoactivated DNA cleavage observed at over 80% of mismatch sites, irrespective of the sequence of the mispaired bases involved.62 Significantly, cleavage was observed at a single-base mismatch in a 2725 base pair plasmid heteroduplex.61 Such high specificity has led to many applications of such complexes, including investigations into the correlation between deficiencies in mismatch repair and cancerous transformations, and towards the development of novel cancer treatments.56 Related complexes have also been employed as probes for mismatch occurrence;63,64 as targeting agents for conventional cisplatin chemotherapeutic protocols;65,66 and as site-specific delivery agents for metal-binding peptide conjugates.67 The complexes developed also inhibit cellular proliferation in mismatch-repairdeficient cells (such as those associated with cancerous growth) in preference to healthy, mismatch-proficient cells.68 Such observations have led to complexes designed for selective chemotherapeutic agents that target mismatches in cancerous, rather than healthy cells.68 Recent work has determined the mode of DNA binding involved in mismatch recognition. The crystal structure of [Rh(bpy)2(chrysi)]3+ bound to a CA mismatch in a DNA oligonucleotide revealed that, in contrast to earlier reports on the mode of intercalation, wherein DNA does not unwind to allow intercalation of the new ligand, the complex inserted into the double helix specifically at the mismatch site, ejecting both mismatched bases, to accommodate the new ligand.69 Such results again corroborate the ability of the bulky complex to specifically recognize mismatches over normal base pairing. Very recently the intercalation of a complex into a DNA cytosine–cytosine mismatch was identified using NMR solution studies.70

10.4 Tin–DNA Interactions 10.4.1 Introduction Tin complexes, in particular organotin species, where a direct C–Sn bond is present within the complex, have found application in many industries. The widespread use of organotin complexes in the food, agricultural, paints and polymer industries has resulted in their accumulative distribution in the environment and in biological systems. The biological toxicity of tin species is well known, and such species exhibit a particular toxicity towards the central nervous system. The ability of inorganic divalent tin to interact with DNA has been established.71–73 Such species were found to enter human white blood cells, and cause extensive damage to DNA through oxidative mechanisms. As a consequence, much research has focused on the biological effects of organotin complexes.2,8,23,74,75 Despite the associated toxicity issues, organotin compounds have found many pharmacological applications, as potential antiinflammatory, antimicrobial and antitumour agents. Understanding the interac-

Tin–DNA Interactions

309

tions of organotin reagents with biological species such as DNA is therefore a principle area of research in preventing, and exploiting, the toxicity of tin complexes.

10.4.2 Organotin Complexes–DNA Interactions Organotin(IV) complexes of the type X2R2Sn and X2R2L2Sn (X = halide, pseudohalide; R = alkyl, aryl; L = nitrogen ligand, e.g. py, bpy) have demonstrated moderate activity against a range of tumours.8 The application of such species in the clinic, however, is limited by their toxicity, which has been found to increase in the order RSn < R2Sn < R3Sn. An increase in the lipophilicity of the R group has been correlated with an increase in toxicity, and alkyl complexes are often more toxic than their aryl analogues. The principle forms of toxicity are neurological, and may occur as a result of the affinity of tin for sulfur, i.e. through coordination to protein thiols.76 The correlation between lipophilicity and toxicity may be related to the cell membrane-permeability of lipophilic species. Maintaining a balance between the hydrophobic and hydrophilic character of the complexes involved, therefore, has been a focus of much research, with the intention of increasing antitumour activity, while reducing toxicity issues. For example, highly lipophilic stannocene-type complexes such as decaphenyl stannocene, [h5-C5(C6H5)5]2Sn, structurally analogous to the antitumour active titanocene dichloride,2 have exhibited antitumour activity against Ehrlich Ascites tumour with cure rates of 68–90%, remarkably with relatively low toxicity.77 The fact that such sterically crowded Sn(II) centres exhibit antitumour activity points towards a different mode of activity than covalent coordination of these complexes to nucleobases on DNA. Although not discussed by the authors, potential intercalation of the planar aromatic substituents on the stannocenes into DNA (see above) may be a potential mechanism of activity, which may also explain why the pentasubstituted cyclopentadienyl ligands exhibit some antitumour activity, albeit less pronounced than the corresponding stannocene. While the low water solubility of the majority of organotin compounds is a significant problem, the introduction of hydrophilic leaving groups8,74,75 or, for example, their inclusion into b-cyclodextrin75 may circumvent such limitations. In contrast to cisplatin, the mechanism of activity of organotin compounds has not been well established. Owing to the structural similarity of such complexes to cisplatin, DNA has been frequently proposed as a potential therapeutic target and, in order to gain further understanding on the antitumour activity of such species, the interactions of organotin complexes with DNA have been investigated by a number of groups. Organotin(IV) complexes have been shown to coordinate to bis(adenine) ligands to form tetrahedral adducts. Reactions of guanine, thymine, cytosine and uracil under similar conditions, however, afforded no coordination products.78 Indeed, the interaction of diorganotin(IV) compounds such as R2SnCl2(phen) (R = Et, n-Bu, Ph) with nucleotide monophosphates in neutral aqueous solution afford ill-defined, polymeric species in which the Sn(IV) centre coordinates only to the phosphate groups, with no observed coordination to the nucleobases present.23,75,79 Interaction of Bu2SnCl2 or Bu3SnCl with 5′-AMP and 5′-GMP exhibited multiple

310


Figure 10.4 Organotin–DNA adduct formation

binding modes to the Sn(IV) centre through 31P NMR analysis, wherein the Sn(IV) centre could be bound in a chelate or nonchelate manner to the phosphate oxygen.80 In 119Sn Mossbauer studies, Barbieri observed the interaction of solvated Sn(IV) species (R2SnCl2(EtOH)2 and R3SnCl(EtOH), R= Me, Et, n-Bu, n-Oct or Ph), in which possible DNA adducts, R2Sn(DNA)2 and R3Sn(DNA) were observed as gel phases (DNA condensation).81–83 Adduct formation was proposed to occur by neutralization of the anionic phosphate groups by the cationic Sn centre, with proposed octahedral and trigonal bipyramidal structures (Figure 10.4). Interestingly, while the active species for cisplatin antitumour activity is known to be the hydrolysis product, ([(NH3)2Pt(OH2)2]2+) analogous Sn species, [Me2Sn(OH)(H2O)n]+, Me2Sn(OH)2 and Me3Sn(OH)(H2O)2 did not interact with native DNA.84 Such observations led the authors to propose structure–activity relationships, wherein the ability of organotin(IV) complexes to bind DNA depends on the lipophilicity of the complex. Therefore, as lipophilicity increases for RnSn(IV) in the series: R = Me < Et < Ph < n-Bu < n-Oct, so the tendency of the complex to form DNA-condensates increases.82 When formation of DNA condensates does occur, the interaction was proposed to occur through electrostatic interaction between ethanol solvated species and oxygen atoms of the phosphodiester groups present. Later theoretical and model studies on organotin(IV)–DNA interactions further supported the conclusions that binding of such species occurs through electrostatic interaction with phosphate O atoms.85,86 Binding of organotin(IV) species to O donor ligands in nucleotides has exhibited pH dependence in a number of studies. For example, coordination of hydrated [Me2Sn(IV)]2+ with a series of nucleotides under acidic conditions (pH < 4) resulted in binding through phosphate oxygen atoms. At intermediate pH (4–9.5), no interaction occurred, while under basic conditions, (pH > 9.5) the hydroxyl groups of the sugar functionality present on the nucleotide were also capable of coordination to the Sn centre, in particular if a second deprotonated hydroxyl group was adjacent to the first, to form a chelate complex.8 Similar observations were made in elegant studies on the interaction of Et2SnCl2(phen) with 5′-dGMP, using [trans-en2Os(nH2)](CF3SO3)2 as 1H-NMR probe.8,75 In all cases, the pH dependence of the Sn–DNA interaction was observed. The chemical and structural differences between organotin complexes and cisplatin would suggest that very different modes of antitumour activity are exhibited by each. The principle sites of interaction for organotin species would appear

References

311

categorically to be oxygen atoms of phosphate and, depending on pH, hydroxyl groups on the ribose or deoxyribose moieties present. Further studies on the interaction of organotin(IV) complexes with calf thymus DNA revealed that no disruption of DNA double helix structure occurs. Neither does any DNA degradation occur through oxidative stress mechanisms.87 The question of how organotin complexes may inhibit tumour growth remains, as yet, unanswered.

References 1. Bertini, I.; Gray, H.B.; Lippard, S.J.; Valentine, J.S.; Bioinorganic Chemistry, University Science Books, Sausalito, CA, 1994. 2. Guo, Z.; Sadler, P.J.; Metals in medicine; Ang. Chem. Int. Ed., 1999, 38(11), 1512–1531. 3. Rosenberg, B.; VanCamp, L.; Trosko, J.E.; Mansour, V.H.; Platinum compounds: a new class of potent antitumor agents; Nature, 1969, 222(5191), 385–386. 4. Wang, D.; Lippard, S.J.; Cellular processing of platinum anticancer drugs. Nat. Rev. Drug Discov., 2005, 4(4), 307–320. 5. Anastassopoulou, J.; Metal–DNA interactions; J. Mol. Struct., 2003, 651–653, 19–26. 6. Theophanides, T.; Anastassopoulou, J.; Demunno, G.; Metal–DNA interactions. In: Metal– Ligand Interactions, Russo, N., Salahub, D.R., Witko, M., Eds, NATO Science Series, II: Mathematics, Physics and Chemistry, Vol. 116, 2003, 285–300. 7. Guo, Z.; Sadler, P.J.; Medicinal inorganic chemistry; Adv. Inorg. Chem., 2000, 49, 183–306. 8. Clarke, M.J.; Zhu, F.; Frasca, D.R.; Non-platinum chemotherapeutic metallopharmaceuticals; Chem. Rev., 1999, 99(9), 2511–2533. 9. Boerner, L.J.K.; Zaleski, J.M.; Metal complex–DNA interactions: from transcription inhibition to photoactivated cleavage; Curr. Opin. Chem. Biol., 2005, 9(2), 135–144. 10. Clarke, M.J.; Limauro, S.; Metal-containing anticancer agents. In: Handbook of MetalLigand Interactions in Biological Fluids: Vol. 2, Bioinorganic Chemistry, Berthon, G., Ed., Marcel Dekker, New York, 1995, 1010–1018. 11. Sigel, A.; Sigel, H., Eds.; Metal Ions in Biological Systems, Vol. 32: Interactions of Metal Ions with Nucleotides, Nucleic Acids, and Their Constituents, Marcel Dekker, New York, 1996. 12. Duguid, J.G.; Bloomfield, V.A.; Benevides, J.M.; Thomas, G.J., Jr.; Raman spectroscopy of DNA-metal complexes. II. The thermal denaturation of DNA in the presence of Sr2+, Ba2+, Mg2+, Ca2+, Mn2+, Co2+, Ni2+, and Cd2+; Biophys. J., 1995, 69(6), 2623–2641. 13. Gil, E.D.S.; Serrano, S.H.P.; Ferreira, E.I.; Kubota, L.T.; Electrochemical evaluation of rhodium dimer-DNA interactions; J. Pharm. Biomed. Anal., 2002, 29(4), 579–584. 14. Tselepi-Kalouli, E.; Katsaros, N.; The interaction of rhodium(II) and rhodium(III) with DNA; J. Inorg. Biochem.; 1990, 40(2), 95–102. 15. Giraldi, T.; Zassinovich, G.; Mestroni, G.; Antitumor action of planar, organometallic rhodium(I) complexes; Chem. Biol. Inter., 1974, 9(6), 389–394. 16. Zassinovich, G.; Mestroni, G.; Camus, A.; Diolefinic complexes of rhodium(I) and iridium(I) with nitrogen-containing ligands; J. Organometal. Chem., 1975, 91(3), 379–388. 17. Chifotides, H.T.; Dunbar, K.R.; Interactions of metal–metal-bonded antitumor active complexes with DNA fragments and DNA; Acc. Chem. Res., 2005, 38(2), 146–156. 18. Pruchnik, F.P.; Jakimowicz, P.; Ciunik, Z.; Zakrzewska-Czerwinska, J.; Opolski, A.; Wietrzyk, J. et al.; Rhodium(III) complexes with polypyridyls and pyrazole and their antitumor activity; Inorg. Chim. Acta, 2002, 334, 59–66. 19. Pruchnik, F.P.; Kluczewska, G.; Wilczok, A.; Mazurek, U.; Wilczok, T.; Rhodium(II) complexes with phenanthrolines and their metabolic action on synchronized cell culture; J. Inorg. Biochem., 1997, 65(1), 25–34.

312


20. Pruchnik, F.; Dus, D.; Properties of rhodium(II) complexes having cytostatic activity. J. Inorg. Biochem., 1996, 61(1), 55–61. 21. Pruchnik, F.P.; Starosta, R.; Ciunik, Z.; Opolski, A.; Wietrzyk, J.; Wojdat, E. et al.; Tetraacetatodirhodium(II) complexes with tris(methoxyphenyl) phosphines, their reactivity, structure, and antitumor activity; Can. J. Chem., 2001,79(5/6), 868–877. 22. Fimiani, V.; Ainis, T.; Cavallaro, A.; Piraino, P.; Antitumor effect of the new rhodium(II) complex: Rh2(Form)2(O2CCF3)2(H2O)2 (Form = N,N′-di-p-tolyl formamidinate); J. Chemother., 1990, 2(5), 319–326. 23. Yang, P.; Guo, M.; Interactions of organometallic anticancer agents with nucleotides and DNA; Coord. Chem. Rev., 1999, 185–186, 189–211. 24. Aoki, K.; Salam, M.A.; Interligand interactions affecting specific metal bonding to nucleic acid bases. A case of [Rh2(OAc)4], [Rh2(HNOCCF3)4], and [Rh2(OAc)2(HNOCCF3)2] toward purine nucleobases and nucleosides; Inorg. Chim. Acta, 2002, 339, 427– 437. 25. Dunbar, K.R.; Matonic, J.H.; Saharan, V.P.; Crawford, C.A.; Christou, G.; Structural evidence for a new metal-binding mode for guanine bases: Implications for the binding of dinuclear antitumor agents to DNA; J. Am. Chem. Soc., 1994, 116(5), 2201–2202. 26. Crawford, C.A.; Day, E.F.; Saharan, V.P.; Folting, K.; Huffman, J.C.; Dunbar, K.R. et al.; N7,O6 bridging 9-ethylguanine (9-EtGH) groups in dinuclear metal-metal bonded complexes with bond orders of one, two or four; J. Chem. Soc. Chem. Comm., 1996, (10), 1113–1114. 27. Chifotides, H.T.; Koshlap, K.M.; Perez, L.M.; Dunbar, K.R.; Novel binding interactions of the DNA fragment d(pGpG) cross-linked by the antitumor active compound tetrakis(mcarboxylato)dirhodium(II,II); J. Am. Chem. Soc., 2003, 125(35), 10714–1024. 28. Chifotides, H.T.; Koshlap, K.M.; Perez, L.M.; Dunbar, K.R.; Unprecedented head-to-head conformers of d(GpG) bound to the antitumor active compound tetrakis(m-carboxylato )dirhodium(II,II); J. Am. Chem. Soc., 2003, 125(35), 10703–10713. 29. Perlepes, S.P.; Huffman, J.C.; Matonic, J.H.; Dunbar, K.R.; Christou, G.; Binding of 2,2′bipyridine to the dirhodium(II) tetraacetate core: unusual structural features and biological relevance of the product Rh2(OAc)4(bpy); J. Am. Chem. Soc., 1991, 113(7), 2770–2771. 30. Crawford, C.A.; Matonic, J.H.; Streib, W.E.; Huffman, J.C.; Dunbar, K.R.; Christou, G.; Reaction of 2,2′-bipyridine (bpy) with dirhodium carboxylates: mono-bpy products with variable chelate binding modes and insights into the reaction mechanism; Inorg. Chem., 1993, 32(14), 3125–3133. 31. Aguirre, J.D.; Lutterman, D.A.; Angeles-Boza, A.M.; Dunbar, K.R.; Turro, C.; Effect of axial coordination on the electronic structure and biological activity of dirhodium(II,II) complexes; Inorg. Chem., 2007, 46(18), 7494–7502. 32. Reibscheid, E.M.; Zyngier, S.; Maria, D.A.; Mistrone, R.J.; Sinisterra, R.D.; Couto, L.G. et al.; Antitumor effects of rhodium (II) complexes on mice bearing Ehrlich tumors; Braz. J. Med. Biol. Res., 1994, 27(1), 91–94. 33. Chifotides, H.T.; Koomen, J.M.; Kang, M.; Tichy, S.E.; Dunbar, K.R.; Russell, D.H.; Binding of DNA purine sites to dirhodium compounds probed by mass spectrometry; Inorg. Chem., 2004, 43(20), 6177–6187. 34. Asara, J.M.; Hess, J.S.; Lozada, E.; Dunbar, K.R.; Allison, J.; Evidence for binding of dirhodium bis-acetate units to adjacent GG and AA sites on single-stranded DNA; J. Am. Chem. Soc., 2000, 122(1), 8–13. 35. Dunham Shari, U.; Chifotides Helen, T.; Mikulski, S.; Burr Amity, E.; Dunbar, Kim R.; Covalent binding and interstrand cross-linking of duplex DNA by dirhodium(II,II) carboxylate compounds; Biochemistry, 2005, 44(3), 996–1003. 36. Chifotides, H.T.; Fu, P.K.L.; Dunbar, K.R.; Turro, C.; Effect of equatorial ligands of dirhodium(II,II) complexes on the efficiency and mechanism of transcription inhibition in vitro; Inorg. Chem.; 2004, 43(3), 1175–1183. 37. Chifotides, H.T.; Dunbar, K.R.; Unprecedented head-to-head right-handed cross-links between the antitumor bis(m-N,N′-di-p-tolylformamidinate) dirhodium(II,II) core and

References

38. 39. 40. 41. 42. 43. 44. 45.

46. 47.

48. 49.

50. 51. 52.

53. 54. 55. 56. 57.

313

the dinucleotide d(ApA) with the adenine bases in the rare imino form; J. Am. Chem. Soc.; 2007, 129(41), 12480–12490. Rahman, M.M.; Yasuda, H.; Katsura, S.; Mizuno, A.; Inhibition of endonuclease cleavage and DNA replication of E. coli plasmid by the antitumor rhodium(II) complex; Arch. Biochem. Biophys., 2007, 464(1), 28–35. Detty, M.R.; Gibson, S.L.; Wagner, S.J.; Current clinical and preclinical photosensitizers for use in photodynamic therapy; J. Med. Chem.; 2004, 47(16), 3897–3915. Bradley, P.M.; Angeles-Boza, A.M.; Dunbar, K.R.; Turro, C.; Direct DNA photocleavage by a new intercalating dirhodium(II/II) complex: comparison to Rh2(m-O2CCH3)4; Inorg. Chem., 2004, 43(8), 2450–2452. Fu, P.K.L.; Bradley, P.M.;Turro, C.; DNA cleavage by photogenerated Rh2(O2CCH3)4(H2O)2+; Inorg. Chem.; 2001, 40(11), 2476–2477. Angeles-Boza, A.M.; Bradley, P.M.; Fu, P.K.L.; Wicke, S.E.; Bacsa, J.; Dunbar, K.R. et al.; DNA binding and photocleavage in vitro by new dirhodium(II) dppz complexes: correlation to cytotoxicity and photocytotoxicity; Inorg. Chem.; 2004, 43(26), 8510–8519. Lutterman, D.A.; Fu, P.K.L.; Turro, C.; cis-[Rh2(m-O2CCH3)2(CH3CN)6]2+ as a photoactivated cisplatin analog; J. Am. Chem. Soc., 2006, 128(3), 738–739. Chouai, A.; Wicke, S.E.; Turro, C.; Bacsa, J.; Dunbar, K.R.; Wang, D. et al.; Ruthenium(II) complexes of 1,12-diazaperylene and their interactions with DNA; Inorg. Chem., 2005, 44(17), 5996–6003. Kang, M.; Chouai, A.; Chifotides, H.T.; Dunbar, K.R.; 2D NMR spectroscopic evidence for unprecedented interactions of cis-[Rh2(dap)(m-O2CCH3)2(H1-O2CCH3)(CH3OH)] (O2CCH3) with a DNA oligonucleotide: combination of intercalative and coordinative binding; Angew. Chem., Int. Ed., 2006, 45(37), 6148–6151. Swavey, S.; Brewer, K.J.; Visible light induced photocleavage of DNA by a mixed-metal supramolecular complex: [{(bpy)2Ru(dpp)}2RhCl2]5+; Inorg. Chem., 2002, 41(24), 6196–6198. Barry, C.G.; Turney, E.C.; Day, C.S.; Saluta, G.; Kucera, G.L.; Bierbach, U.; Thermally inert metal ammines as light-inducible DNA-targeted agents. synthesis, photochemistry, and photobiology of a prototypical rhodium(III)-intercalator conjugate; Inorg. Chem., 2002, 41(26), 7159–7169. Pieper, T.; Borsky, K.; Keppler, B.K.; Non-platinum antitumor compounds. Topics Biol. Inorg. Chem., 1999, 1(Metallopharmaceuticals I),171–199. Mestroni, G.; Alessio, E.; Sessanta o Santi, A.; Geremia, S.; Bergamo, A.; Sava, G. et al.; Rhodium(III) analogs of antitumor-active ruthenium(III) compounds: the crystal structure of [ImH][trans-RhCl4(Im)2] (Im = imidazole); Inorg. Chim. Acta, 1998, 273(1,2), 62–71. Billadeau, M.A.; Morrison, H.; Photolytic covalent binding of metal complexes to DNA. In: Metal ions in Biological Systems, Vol. 33, Sigel, A. and Sigel, H., Eds., Marcel Dekker, New York, 1996, 269–296. Erkkila, K.E.; Odom, D.T.; Barton, J.K.; Recognition and reaction of metallointercalators with DNA; Chem. Rev., 1999, 99(9), 2777–2795. Mahnken, R.E.; Billadeau, M.A.; Nikonowicz, E.P.; Morrison, H.; Development of photo cis-platinum reagents. Reaction of cis-dichlorobis(1,10-phenanthroline)rhodium(III) with calf thymus DNA, nucleotides and nucleosides; J. Am. Chem. Soc., 1992, 114(24), 9253–9265. O’Neill, M.A.; Barton, J.K.; DNA-mediated charge transport chemistry and biology; Topics Curr. Chem., 2004, 236 (Long-Range Charge Transfer in DNA I), 67–115. Modrich, P.; Mechanisms and biological effects of mismatch repair; Ann. Rev. Genet., 1991, 25, 229–253. Kolodner, R.; Biochemistry and genetics of eukaryotic mismatch repair; Genes Devel., 1996, 10(12), 1433–1442. Kunkel, T.A.; Erie, D.A.; DNA mismatch repair; Ann. Rev. Biochem., 2005, 74, 681–710. David, S.S.; Barton, J.K.; NMR evidence for specific intercalation of D-Rh(phen)2phi3+ in [d(CTCGAC)2]; J. Am. Chem. Soc., 1993, 115(7), 2984–2985.

314


58. Hudson, B.P.; Dupureur, C.M.; Barton, J.K.; 1H NMR structural evidence for the sequence-specific design of an intercalator: ∆-a-[Rh[(R,R)-Me2trien]phi]3+ bound to d(GAGTGCACTC)2; J. Am. Chem. Soc., 1995, 117(36), 9379–9380. 59. Sitlani, A.; Long, E.C.; Pyle, A.M.; Barton, J.K.; DNA photocleavage by phenanthrenequinone diimine complexes of rhodium(III): shape-selective recognition and reaction; J. Am. Chem. Soc., 1992, 114(7), 2303–2312. 60. Jackson, B.A.; Barton, J.K.; Recognition of DNA base mismatches by a rhodium intercalator; J. Am. Chem. Soc., 1997, 119(52), 12986–12987. 61. Jackson, B.A., Barton, J.K.; Recognition of base mismatches in DNA by 5,6-chrysenequinone diimine complexes of rhodium(III): A proposed mechanism for preferential binding in destabilized regions of the double helix; Biochemistry; 2000, 39(20), 6176–6182. 62. Jackson, B.A.; Alekseyev, V.Y.; Barton, J.K.; A versatile mismatch recognition agent: specific cleavage of a plasmid DNA at a single base mispair; Biochemistry, 1999 38(15), 4655–4662. 63. Rueba, E.; Hart, J.R.; Barton, J.K.; [Ru(bpy)2(L)]Cl2: Luminescent metal complexes that bind DNA base mismatches; Inorg. Chem., 2004, 43(15), 4570–4578. 64. Zeglis, B.M.; Barton, J.K.; A mismatch-selective bifunctional rhodium-oregon green conjugate: a fluorescent probe for mismatched DNA; J. Am. Chem. Soc., 2006, 128(17), 5654–5655. 65. Schatzschneider, U.; Barton, J.K.; Bifunctional rhodium intercalator conjugates as mismatch-directing DNA alkylating agents; J. Am. Chem. Soc., 2004, 126(28), 8630–8631. 66. Petitjean, A.; Barton, J.K.; Tuning the DNA reactivity of cis-platinum: Conjugation to a mismatch-specific metallointercalator; J. Am. Chem. Soc., 2004, 126(45), 14728–14729. 67. Copeland, K.D.; Fitzsimons, M.P.; Houser, R.P.; Barton, J.K.; DNA hydrolysis and oxidative cleavage by metal-binding peptides tethered to rhodium intercalators; Biochemistry, 2002, 41(1), 343–356. 68. Hart, J.R.; Glebov, O.; Ernst, R.J.; Kirsch, I.R.; Barton, J.; DNA mismatch-specific targeting and hypersensitivity of mismatch-repair-deficient cells to bulky rhodium(III) intercalators; Proc. Natl. Acad. Sci. USA, 2006, 103(42), 15359–15363. 69. Pierre, V.C.; Kaiser, J.T.; Barton, J.K.; Insights into finding a mismatch through the structure of a mispaired DNA bound by a rhodium intercalator; Proc. Natl. Acad. Sci. USA, 2007, 104(2), 429–434. 70. Cordier, C.; Pierre, V.C.; Barton, J.K.; Insertion of a bulky rhodium complex into a DNA cytosine-cytosine mismatch: An NMR solution study; J. Am. Chem. Soc., 2007, 129(40), 12287–12295. 71. McLean, J.R.N.; Birnboim, H.C.; Pontefact, R.; Kaplan, J.G.; The effect of tin chloride on the structure and function of DNA in human white blood cells; Chem. Biol. Inter., 1983, 46(2), 189–200. 72. McLean, J.R.N.; Blakey, D.H.; Douglas, G.R.; Kaplan, J.G.; The effect of stannous and stannic (tin) chloride on DNA in Chinese hamster ovary cells; Mut. Res. Lett., 1983, 119(2), 195–201. 73. Dantas, F.J.S., Moraes, M.O.; Carvalho, E.F.; Valsa, J.O.; Bernardo-Filho, M.; Caldeira-DeAraujo, A.; Lethality induced by stannous chloride on Escherichia coli AB1157: participation of reactive oxygen species; Food Chem. Toxicol., 1996, 34(10), 959–962. 74. De Vos, D.; Willem, R.; Gielen, M.; Van Wingerden, K.E.; Nooter, K.; The development of novel organotin antitumor drugs: structure and activity; Metal-Based Drugs. 1998, 5(4), 179–188. 75. Pellerito, L.; Nagy, L.; Organotin(IV)n+ complexes formed with biologically active ligands: equilibrium and structural studies, and some biological aspects; Coord. Chem. Rev., 2002, 224(1–2), 111–150. 76. Penninks, A.H.; Seinen, W.; Mechanisms of dialkyltin-induced immunopathology; Vet. Quart., 1984, 6(4), 209–215. 77. Kopf-Maier, P.; Janiak, C.; Schumann, H.; Antitumor properties of organometallic metallocene complexes of tin and germanium; J. Cancer Res. Clin. Oncol., 1988, 114(5), 502–506.

References

315

78. Cardin, C.J.; Roy, A.; Anticancer activity of organotin compounds. 2. Interaction of diorganotin dihalides with nucleic acid bases and nucleosides; the synthesis of adenine, adenosine and 9-methyladenine adducts; Inorg. Chim. Acta, 1985, 107(1), 57–61. 79. Barbieri, R.; Silvestri, A.; Piro, V.; Tin-119 Moessbauer titration of dimethyl- and trimethyltin(IV) hydroxides with model ligands mimicking nucleic acid phosphate sites, and with deoxyribonucleic acid; J. Chem. Soc., Dalton Trans.: 1990, (12), 3605–3609. 80. Atkinson, A.; Rodriguez, M.D.; Shewmaker, T.E.; Walmsley, J.A.; Synthesis and characterization of compounds of di- and tributyltin chloride with adenine and guanine mononucleotides; Inorg. Chim. Acta, 1999, 285(1), 60–69. 81. Barbieri, R.; Silvestri, A.; Giuliani, A.M.; Piro, V.; Di Simone, F.; Madonia, G.; Organotin compounds and deoxyribonucleic acid; J. Chem. Soc., Dalton Trans.: 1992, (4), 585–590. 82. Piro, V.; Di Simone, F.; Madonia, G.; Silvestri, A.; Giuliani, A.M.; Ruisi, G. et al.; The interaction of organotins with native DNA; Appl. Organomet. Chem., 1992, 6(6), 537–542. 83. Barbieri, R.; Ruisi, G.; Silvestri, A.; Giuliani, A.M.; Barbieri, A.; Spina, G. et al.; Dynamics of tin nuclei in alkyltin(IV)-deoxyribonucleic acid condensates by variable-temperature tin-119 Moessbauer spectroscopy; J. Chem. Soc., Dalton Trans.: 1995, (3), 467–475. 84. Barbieri, R.; Silvestri, A.; The hydrolysis of Me2SnIV and Me3SnIV moieties monitored through tin-119 Moessbauer spectroscopy; Inorg. Chim. Acta, 1991, 188(1), 95–98. 85. Barone, G.; Ramusino, M.C.; Barbieri, R.; La Manna, G.; Semiempirical calculations on the interaction between dimethyltin(IV) and DNA model system; Theochem., 1999, 469, 143–149. 86. Barbieri, R.; Huber, F.; Silvestri, A.; Ruisi, G.; Rossi, M.; Barone, G. et al.; The interaction of S,N-coordinated dimethyltin(IV) derivatives with deoxyribonucleic acid: structure and dynamics by 119Sn Mossbauer spectroscopy; Appl. Organomet. Chem., 1999, 13(8), 595–603. 87. Barone, G.; Barbieri, R.; La Manna, G.; Koch, M.H.J.; The interaction of deoxyribonucleic acid with methyltin(IV) moieties in solution studied by small-angle X-ray scattering, circular dichroism and UV spectroscopy; Appl. Organomet. Chem. 2000, 14(4), 189–196.

Part C DNA Recognition: Nucleases and Sensors


11 Groove-Binding Ruthenium(II) Complexes as Probes of DNA Recognition Jayden A. Smith, J. Grant Collins and F. Richard Keene

11.1 Introduction The application of polypyridyl transition metal complexes to the structural elucidation of biological molecules had its origins in the observations of Dwyer et al. in the 1950s.1 The different biological activity of tris(phenanthroline) complexes of Fe(II), Ni(II), Ru(II), Os(II), Co(II) and Zn(II) in laboratory mice was reported, with selectivity of the D (right-handed) in preference to the L (left-handed) enantiomer noted in the case of inert species. They also noted a curariform activity and significant antibacterial characteristics, and a significant enhancement of activity on methyl substitution on the ligands in the cases of iron and ruthenium species. As such complexes were chemically inert, Dwyer et al. reasoned that any effect they had must be based primarily on their physical interactions with biological systems. Ruthenium has often been the metal of choice when studying the interactions of polypyridyl transition metal complexes with nucleic acids, owing to the wellestablished chemistry and rich photophysical properties of this genre.2 Polypyridylruthenium complexes are typically chemically inert, meaning their interactions with nucleic acids are usually noncovalent and hence reversible, and being octahedral tris(bidentate) species they possess an inherent chirality which can be exploited in their interactions with chiral polynucleotides. Furthermore, these complexes exhibit strong metal-to-ligand charge transfer (MLCT)-induced absorbances in the visible Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

320

Groove-Binding Ruthenium(II) Complexes

region of the spectrum and notable luminescent emissions – perturbations of these spectral characteristics upon nucleic-acid binding provide a convenient means to appraise the extent or nature of the interaction.

11.2 Mononuclear Complexes [Ru(phen)3]2+ (phen = 1,10-phenanthroline; see Figure 11.1d) has received particular attention as an inert metalloprobe of biological molecules. Its DNA binding ability has been most extensively investigated by Barton and coworkers, who in the early 1980s, elaborated upon the work of Dwyer et al. by investigating the in vitro enantioselectivity of the complex.3 On the basis of hypochromicity measurements it was proposed that the complex bound via intercalation of a single phenanthroline ligand. Preferential association between the D form of the complex and right-handed B-DNA was confirmed. The observed enantioselectivity is believed to be due to the unfavourable steric interactions between the nonintercalated ligands of the L enantiomer with the sugar-phosphate backbone of the DNA. Alternatively, the nonintercalating phenanthroline ligands of the D isomer fit easily within the complementary right-handed helical groove. No enantioselectivity was observed in experiments conducted with left-handed Z-DNA, presumably due to the altered groove dimensions. The intercalative model proposed by Barton et al. was supported by experiments utilizing the analogous complex with 4,7-diphenyl-1,10-phenanthroline (DIP; Figure 11.1o) as the ligand. As expected, the phenyl groups heightened the observed enantioselectivity, purportedly by exaggerating the steric interactions between the L enantiomer and the phosphate backbone, as well as enhancing the extent of intercalation.4 Again, both enantiomers were found to bind to Z-DNA with equal affinity. By contrast, [Ru(bpy)3]2+ (bpy = 2,2′-bipyridine; Figure 11.1a), lacking sufficiently elongated polyaromatic ligands to participate in intercalation, associated with DNA only weakly in what was believed to be primarily an electrostatic interaction.5 However, the intercalative binding proposal has been the subject of considerable controversy. Viscosity measurements undertaken by Chaires et al.6 and scanning force microscopy investigations by Williams, Bottomley et al.7 both suggested a nonintercalative mode of binding. Subsequent NMR experiments of Collins et al.8 gave some indication that [Ru(phen)3]2+ may indeed be partially intercalating, a result confirmed by more recent linear dichroism studies by Nordén and Lincoln:9 they claimed that both enantiomers of the complex may bind with one phenanthroline ligand partially (semi- or quasi-) intercalated via the minor groove, and their proposed binding geometry is in reasonable agreement with the available NMR data. This partial intercalation model is commonly seen amongst complexes which have polyaromatic ligands that either have insufficient span to insert fully between stacked base pairs, or in cases where the intercalating ligand deviates from planarity, inhibiting complete insertion.10 Indeed, a circular and linear dichroism study by Nordén and coworkers eventually dismissed an intercalative binding mode for [Ru(DIP)3]2+.11

Mononuclear Complexes

H3C

321

CH3

N

N

(a) bpy

N

N

(d) phen

(c) tBu2bpy

(b) Me2bpy

H3C

CH3

N

N

N

N

CH3

H3C

H3C

CH3

CH3

H3C N

N

H3C

N

CH3

N

N

N

N

(g) 4,7-Me2phen

(e) 2,9-Me2phen

N

(h) TMP

(f) 5,6-Me2phen

N

N

N

N

N

N

N

N

N

N

N

N

N

N

(l) dppx

(j) dppz

CH3

H3C

N H

N

(k) dpqC

(i) dpq

H2N

N

N H

N

NH

(o) DIP

NH2

(m) (R,R)-Me2trien

N

N N

N

(n) IP

HN

H3C N

(q) picchxnMe2

HN (p) phi

N

N

N H3C

N

(r) 2-appt

N H2N

Figure 11.1 Bidentate and tetradentate ligands

N N

N H

322


(s) bbtb N

S N

S N

N

H N

N

N

Y

(t) PIP

N

N

N

NH

N

N

NH

(v) chrysi

(u) tactp

Figure 11.1 Continued

The controversy surrounding the exact mode of binding adopted by [Ru(phen)3]2+, heightened by the relatively low binding affinity of the complex (∼6 × 103 M−1, dependent upon DNA sequence),3,12 was exacerbated by debate over whether the complex was binding in the major or the minor groove. Two-dimensional NMR experiments have revealed that both enantiomers of [Ru(phen)3]2+ bind in the minor groove of DNA.13 The numerous conflicting results regarding the site and orientation of binding have been attributed to variations in technique, DNA sequence and even metal complex/DNA/salt concentrations.14,15 A minor-groove binding mode was consistent with the trend observed amongst the bulk of other polypyridylruthenium(II) complexes studied.16–18 The binding nature of the related complex [Ru(phen)2(dppz)]2+ (dppz = dipyrido[3,2-a:2′,3′-c] phenazine; Figure 11.1j) has also invoked vigorous debate.19 Complexes based on the dppz ligand function as ‘molecular light switches’: in aqueous solution [Ru(bpy)2(dppz)]2+ and [Ru(phen)2(dppz)]2+ are effectively quenched by the proton-donating solvent and show no luminescence, but when intercalated to DNA the hydrophobic environment allows them to luminesce strongly.20,21 Additionally, the facile electron transfer capabilities of intercalated ruthenium dppz complexes has seen their frequent use in studies of long-range DNA-mediated electron transfer.20,22–26 Unlike phen, the intercalative ability of the dppz ligand was unambiguous and undisputed, having been proven using a variety of techniques and methodologies.16,21,27,28 [Ru(phen)2(dppz)]2+ has been found to bind to DNA with an association constant in excess of 106 M−1:27,28 the extended planar aromatic ring system of the dppz ligand lends itself to intercalative stacking between base pairs much more readily than its less-extended phen counterpart.29–31 Nevertheless, there still exists a significant electrostatic contribution to the binding of dppz complexes.27

Mononuclear Complexes

323

Despite universal support for the intercalative binding mode of dppz-based complexes, the groove via which the complexes intercalated became a point of contention, but present understanding is that it binds from the minor groove. NMR studies have suggested that both D-[Ru(phen)2(dpq)]2+ (dpq = dipyrido[3,2-d:2′,3′-f] quinoxaline; Figure 11.1i) and D-[Ru(phen)2(dpqC)]2+ (dpqC = dipyrido[3,2-a:2′,3′-c](6,7,8,9-tetrahydro)phenazine; Figure 11.1k) bind via the minor groove.8,32 The dpq and dpqC complexes do not bind to DNA as strongly as do their dppz or dppx analogues (dppx = 7,8-dimethyldipyridophenazine; Figure 11.1l), presumably as they cannot intercalate as effectively as dppz or dppx, which possess larger aromatic areas that may potentially stack between DNA bases. Additional NMR experiments conducted by Collins et al. on [Ru(Me2phen)2(dpq)]2+ and [Ru(Me2phen)2(dppz)]2+ (2,9-Me2phen = 2,9-dimethyl-1,10-phenanthroline; Figure 11.1e) also suggested intercalation via the minor groove (methylated phenanthroline ligands were used to simplify the NMR spectra and to provide strong NOE signals in NOESY spectra).17,18 Recently, Nordén and coworkers conceded the possibility of intercalation from either groove when the other is blocked; this concession was based in part on spectroscopic investigations into the properties of DNA-bound [Ru(phen)2(dppz)]2+ in the presence and absence of the minor-groove binder 4′,6diamidino-2-phenylindole (DAPI).33,34 Negligible enantioselectivity is observed in the DNA-binding of [Ru(phen)2(dppz)]2+ 17,28 and its 2,2′-bipyridine analogue [Ru(bpy)2(dppz)]2+,34 although in each case differential luminescence and binding rates were observed between the D and L isomers of each complex, suggesting different binding geometries for each enantiomer. The bulk of studies conducted into polypyridyl complex–nucleic acid interactions have concerned themselves with intercalating mononuclear complexes, typically of the form [Ru(ancillary)2(intercalator)]2+, where there is a single dedicated intercalating ligand and a pair of ancillary ligands that occupy the groove (to a variable extent). Due to the three-dimensional nature of the association between these octahedral complexes and their chiral polynucleotide targets, the nature of all the ligands − ancillary or intercalating − influences the binding affinity of the complex although a larger intercalative surface area generally correlates with a greater binding affinity.35 Molecular modelling suggests that increasing the length of the intercalating ligand favours binding via the minor groove, whereas increases to the width of this ligand promote binding from the major groove.36 These findings are corroborated by experimental results obtained with Ru(II) and Rh(III) complexes featuring the relatively wide intercalating ligand phi (phenanthrenequinone diimine; Figure 11.1p). NMR experiments suggest that D-cis-a-[Ru(RR-picchxnMe2)(dpq)]2+ (picchxnMe2 = N,N′-dimethyl-N,N′-di(2-picolyl)-1,2-diaminocyclohexane; Figure 11.1q) intercalates via the minor groove of DNA, whereas the analogous complex with a phi intercalator, D-cis-a-[Ru(RR-picchxnMe2)(phi)]2+, binds via the major groove.37 The phi ligand has seen particularly extensive use in octahedral rhodium complexes, and such species almost invariably intercalate via the major groove.38 Rhodium complexes containing phi, or derivatives thereof, have demonstrated an impressive degree of selectivity for features such as base sequences,38,39 sequence-dependent

324


twisting,40,41 base mismatches42,43 and base triples.44 The usual enantioselectivity favouring the D form is observed in DNA-binding experiments with these complexes.45 The role of ancillary ligands is therefore important, as the steric bulk of the nonintercalating ligands directly influences how deeply the complex may intercalate. In addition, hydrophobic interactions, van der Waals contacts, and hydrogen bonding brought about by the ancillary ligands are believed to be the major driving forces behind many DNA-complex associations.18,27 A good illustration of this is the complex [Ru(IP)2(dppz)]2+ (IP = imidazole[4,5-f][1,10]phenanthroline; Figure 11.1n), which, with its bulkier aromatic ancillary ligands, binds to DNA with an affinity several times that of its parent complex [Ru(bpy)2(dppz)]2+.46 Increased hydrophobicity in the form of strategically placed methyl groups can also strengthen the DNA-complex interaction: [Ru(5,6-Me2phen)2(dppz)]2+ (5,6-Me2phen = 5,6dimethyl-1,10-phenanthroline; Figure 11.1f) binds more strongly than does [Ru(phen)2(dppz)]2+.47 Ultimately, how well the ancillary ligands complement the shape of the groove in which they are binding (which itself depends on the nature of the target nucleic acid) may be the most significant factor governing binding.48 A detailed survey of DNA recognition by metallointercalators is beyond the scope of this review; however, an excellent review from the Barton laboratory has recently been published.49 Furthermore, rhodium(III)-based metallointercalators have been more extensively utilized for DNA recognition than ruthenium(II) complexes. However, the modification of the intercalating and ancillary ligands in [Ru(ancillary)2(intercalator)]2+ complexes has allowed some modulation of the binding selectivity to achieve specific DNA recognition. At the most basic level, ruthenium complexes can be used to distinguish between the A, B and Z-forms of DNA. For example, both the D and L enantiomers of [Ru(DIP)3]2+ bind to Z-form (left-handed) DNA, whereas only the D isomer binds to B-DNA;50 however, this was subsequently disputed by Norden and coworkers.11 Alternatively, [Ru(TMP)3]2+ (TMP = 3,4,7,8-tetramethyl-1,10-phenanthroline; Figure 11.1h) has been shown to specifically bind A-form DNA.51 More recently, it has been demonstrated that ruthenium-based metallointercalators can be used to target DNA mismatch sequences. These noncomplementary base pairs can occur naturally, or through the interaction of genotoxic chemicals or ionizing radiation, and can lead to a range of diseases.52 Barton and coworkers have reported that [Ru(bpy)2(tactp)]2+ (tactp = 4,5,9,18tetraazachrysenol(9,10-b)-triphenylene; Figure 11.1u) preferentially binds a CC mismatch.49 The tactp intercalating ligand combines the DNA light-switch effect of dppz-based complexes with the chrysi (5,6-chrysene quinone diimine; Figure 11.1v) ligand that is selective for mismatch sites. The sterically bulky chrysi ligand is too wide to intercalate at matched DNA sites, but is suited to thermodynamically destabilized sites.49

11.3 Dinuclear Complexes The utility of mononuclear metal complexes as probes for structural recognition in polynucleotides is limited by their relatively small size, as a typical mononuclear

Dinuclear Complexes

325

complex may, at best, span only four to six base pairs.53 Additionally, mononuclear complexes generally possess low binding affinities (typically ∼104–105 M−1, with the exception of strong intercalators like the dppz complexes) and consequently they are displaced from nucleic acids at relatively low ionic strengths, limiting their application in environments encountered in vivo.54 To mimic the selectivity of nucleic-acid binding proteins, larger species (spanning 10–16 base pairs) will be required: di-, tri- and oligonuclear metal complexes not only have a larger span, but they also possess a greater stereochemical diversity, allowing them to more effectively probe the shape- and structure-recognition mechanisms of nucleic acids than their mononuclear counterparts. Furthermore, dinuclear species possess a larger cationic charge; they may potentially have greater nucleic-acid binding affinities than mononuclear complexes; the presence of two metal centres amplifies any possible chiral discrimination effects; and the larger size of the complex results in considerably slower DNA dissociation rates (a property that is advantageous to both the biological activity and NMR studies of such complexes).55 The structures of bridging ligands discussed in this chapter are shown in Figure 11.2. The potential appeal of a dinuclear binding species can be exemplified by an early study of bis-intercalating organic molecules in which two intercalating chromophores (such as the ethidium cation) were joined via a flexible polyamine/alkane linker.56 Such dyes bound double-stranded DNA with an affinity several orders of magnitude higher than their monomeric analogues. As an example of a DNA-binding dinuclear metal complex, Aldrich-Wright et al. have investigated a similar flexible linker dimer, [{Ru(dpq)2}2(m-phen-x–SOS– x-phen)]4+ (SOS = 2-mercaptoethyl ether; the attachment position of the linker, x, is 4 or 5; dpq = dipyrido[3,2-d:2′,3′-f] quinoxaline; see Figure 11.3), featuring intercalating ancillary (terminal) ligands on each of the metal centres.53 The complex was found to have a binding affinity of approximately 6 × 107 M−1, far in excess of its mononuclear counterpart [Ru(dpq)2(phen)]2+ (K = 5.4 × 104 M−1).57 Furthermore, the complex demonstrates some degree of specificity for purine-purine sequences and has the potential for hydrogen-bonding interactions with the S–O–S bridge. There have also been a number of reports of complexes that bind via an intercalating bridging ligand. An early example of this was the 2,3-bis(2-pyridyl)benzo [g]quinoxaline (dpb; Figure 11.2d) bridged species described by Carlson et al.58 It was found that [{Ru(NH3)4}2(m-dpb)]4+ bound with an affinity similar to that of a mononuclear dppz-based intercalator, whereas [{Ru(bpy)2}2(m-dpb)]4+ bound rather weakly. This discrimination between the two complexes was attributed to the increased steric bulk of the bpy ancillary ligands interfering with complete insertion of the intercalating dpb moiety in the latter case and potential affinity-enhancing hydrogen bonding from the NH3 ligands in the former. Further studies on complexes in which the bridging ligand was extended – and potentially intercalating − revealed very interesting binding dynamics. The stereoisomers of the complex [{m-dppz(11,11′)dppz}(pp)4Ru2]4+ (dppz(11,11′)dppz = 11,11′bis(dipyrido[3,2-a:2′,3′-c]phenazine; Figure 11.2g; pp = bpy or phen) initially assumed a metastable major-groove-bound state at slightly different orientations for the various stereoisomers before threading through the double helix so that there was one Ru(phen)2 moiety in each groove, with the bridge intercalated.59

326

Groove-Binding Ruthenium(II) Complexes N

N

N

N

(f) pztp N

N

(a) bpm

N N N

N

N N

N

N

N

N

N

(b) HAT

N

N N

N

N

N

N

N

N

N

N

(g) dppz(11,11')dppz

N

(c) dppm N

N

N

N

CN

N N

O HN

N

(h) C4(cpdppz)2

N

(d) dpb

N

HN N

N

N

N

O

N

N (e) 2,3-dpp

N

Figure 11.2 Bridging ligands

CN

Dinuclear Complexes

327

N

N

NH

HN N

N

(i) bipp N

N N

N

H3C

N

N

N

N

N

N

N H2C

H3C

(j) tpphz

N

N

n

(k) bbn

N

Figure 11.2 Continued

4+ S

N (dpq)2Ru

N

S

N

O N

Ru(dpq)2

Figure 11.3 Representation of [{Ru(dpq)2}2(m-phen-5–SOS–5-phen)]4+.53

The introduction of some greater degree of flexibility between the two bridging dppz moieties was found to result in more expedient threading. The complex [mC4(cpdppz)2(phen)4Ru2]4+ (C4(cpdppz)2 = N,N′-bis(12-cyano-12,13-dihydro-11H-8cyclopenta[b]dipyrido[3,2-h:2′,3′-j]phenazine-12-carbonyl)-1,4-diaminobutane; Figure 11.2h) threaded through the DNA helix such that the ancillary phen ligands sat in the minor groove while the flexible linker occupied the major groove.60,61 Due to the bis-intercalative nature of the bridging ligand, the complex bound with an affinity several orders of magnitude greater than the analogous mononuclear species [Ru(phen)2(dppz)]2+ (∼1010 M−1 versus 106 M−1).61 The threading process is believed to require extensive conformational changes to the DNA helix, including the breaking and re-forming of base pairs. This is supported by the observation that, as with

328


the dppz(11,11′)dppz-bridged species, AT-threading was faster than that for GC-rich sequences due to the stronger base-pairing of the latter. Kelly and coworkers have compared the binding behaviour of the dinuclear complex [(bpy)2Ru{bbn}Ru(bpy)2]4+ (bbn: n = 5 {1,5-bis[4(4′-methyl-2,2′bipyridyl)]pentane; n = 7 {1,7-bis[4(4′-methyl-2,2′-bipyridyl)]heptane}; Figure 11.2k) with its mononuclear analogues [Ru(bpy)3]2+ and [Ru(bpy)2(Me2bpy)]2+ (Me2bpy = 4,4′-dimethyl-2,2′-bipyridine; Figure 11.1b) and found that the bimetallic species had a much higher binding affinity, was more efficient at photosensitising DNA strand breaks and its binding affinity was less sensitive to ionic strength.62,63 Additional investigations were undertaken with [(phen)2Ru{bbn}Ru(phen)2]4+ (where n = 5, 7, or 10).63 Again, the dinuclear complex was found to have a stronger binding affinity than its mononuclear counterpart, [Ru(phen)2(Me2bpy)]2+, and that the binding affinity was dependent upon the linker chain length. The most effective binding was observed at n = 7, somewhat shorter than the optimal chain length (n > 8) of the classical intercalators with polymethylene chains.56 This shorter chain length is consistent with partial intercalation of a phenanthroline ligand on each metal centre.54 Other intercalative bridging moieties used in the study of DNA-binding dinuclear ruthenium complexes include phenanthroline/imidazole derivatives 64,65 and porphyrins,66 each of which yielded only moderate binding affinities akin to those observed in mononuclear species which bind via intercalation. One particularly interesting species utilized the intercalative bridge bipp (bipp = 2,9-bis(2-imidazo[4,5f][1,10]phenanthroline)-1,10-phenanthroline; Figure 11.2i) which features vacant chelating sites well suited to the binding of a third metal ion, specifically Cu2+. The emissive properties of the complex were quenched upon binding of the copper ion, whereas the DNA-bound complex exhibited a large increase in emission. Subsequent addition of copper ions to the complex–DNA system had no effect because the chelating site was blocked upon intercalation.67,68 This is in contrast with the socalled ‘molecular nut and bolt’ system of [Ru(bpy)2(tpphz)]2+ (tpphz = tetrapyrido[3,2a:2′,3′-c:3′′,2′′-h:2′′,3′′-j]phenazine; Figure 11.2j), which possesses a similar chelating site on the tpphz ligand. Upon intercalation with DNA, the chelating site projects out of the opposite side of the helix where it can coordinate a copper ion, quenching the emission of the complex and locking it in place.69 Many multinuclear species incorporating mixed-metal polypyridyl systems have shown considerable promise as photoactivated drugs: trinuclear Ru–Rh species (essentially two Ru(L)2 moieties linked to a RhCl2 core via 2,3-dpp bridges (2,3-dpp = 2,3-bis(2-pyridyl)pyrazine; Figure 11.2e)), for instance, have demonstrated visible light-induced photocleavage of DNA.70 This has been attributed to a Ru → Rh metal-to-metal charge transfer excited state and is dependent upon the nature of the bridging ligand (the analogous complex with 2,2′-bipyrimidine bridges (bpm; Figure 11.2a) was inactive).71 Ru/Pt mixed species, bridged by 2,3-dpp, dpq and dpb ligands, have been found to bind DNA covalently via the square-planar platinum moiety.72,73 Several nonintercalative, semi-rigid dinuclear complexes have also been investigated. Complexes bridged by the asymmetric phenanthroline derivative pztp ((3pyrazin-2-yl)-as-triazino[5,6-f]-1,10-phenanthroline; Figure 11.2f) have been found to bind via electrostatic/groove binding interactions, whereas the mononuclear

Dinuclear Complexes

329

equivalent intercalates.74 Furthermore, the groove complementarity necessary for intercalation means that the mononuclear complex exhibits enantioselectivity, whereas the (relatively) loosely associating dinuclear species does not. Jiang and coworkers have surveyed a number of dinuclear complexes featuring bridges derived from bpy.75–78 They all bind via a nonintercalative mode with modest binding affinities, however in a number of cases they exhibit intriguing enantiopreferences in which the L enantiomer appears to be the stronger binding.75,78 11.3.1 Stereoisomers and Their Separation A critical issue in the investigation of the interaction of the dinuclear species with DNA has been the ability to control the stereochemistry of the component metal centres. This has been achieved by a combination of techniques – stereoselective synthesis and chromatography.79,80 The stereoselective synthesis has utilized the chiral mononuclear precursors D/L-[Ru(pp)2(py)2]2+ 81 and D/L-[Ru(pp)2(CO)2]2+/ [Ru(pp)(pp′)(CO)2]2+ 82 (pp and pp′ are bidentate polypyridyl ligands; pp ≠ pp′), which can undergo substitution under conditions which allow the retention of the chiral integrity of the metal centre. The chromatographic technique has involved cation exchange, in which the counteranion has been chosen so that it differentially associates with the stereoisomers to be separated.83–85 The technique has allowed the routine separation of diastereoisomers,86–91 the resolution of enantiomers,86,89,90 the separation of geometric isomers89,92 and the separation of chiral helical forms.93 The combination of the stereoselective and chromatographic techniques has been used for the separation of stereoisomers of dinuclear and trinuclear species.92,94,95 11.3.2 Interaction of Dinuclear Complexes with Duplex DNA 1

H NMR experiments with the DL, DD and LL stereoisomers of [{Ru(Me2bpy)2}2(mbpm)]4+ (see Figure 11.4) demonstrated that the ruthenium(II) complexes bound duplex DNA in the minor groove, but with relatively weak affinity (K ≈ 103 M−1).90 Given the relative dimensions of the metal complex and the DNA grooves, it was somewhat surprising that the ruthenium complex bound in the minor groove: the

M

M

meso {Λ∆}

M

M

{ΛΛ}

M

rac

M

{∆∆}

Figure 11.4 A schematic representation of the three stereoisomers of a symmetric dinuclear ruthenium complex

330


ruthenium complex is approximately 16 Å long and 8.0 Å wide, while for canonical form B-type DNA the width of the major and minor grooves are 11.6 Å and 6.0 Å, respectively. Although the ruthenium complexes bound duplex DNA weakly, the NMR results demonstrated that the DD- and LL-enantiomers interacted differently with the dodecanucleotides used in the study, with the DD-enantiomer having a slightly higher affinity. Mononuclear polypyridyl ruthenium(II) complexes that bind DNA nonintercalatively generally associate in the minor groove with a strong preference for A/T rich sequences.96 Distinctively, the stereoisomers of [{Ru(Me2bpy)2}2(m-bpm)]4+ were shown to bind d(CAATCCGGATTG)2 at the central CCGG sequence and at the CA/GT terminal residues, rather than at the expected A/T rich regions. As the minor groove is particularly narrow at A/T rich sequences, it is probable that the relatively bulky ruthenium complex could not be easily accommodated at the AAT/TTA sites. Consequently, it was proposed that the ruthenium complex binds at the central CCGG site, where the minor groove is more open, and at the terminal CA/GT site where the minor groove would also be considerably widened due to the fraying of the terminal base pair. 11.3.3 Interaction of Dinuclear Complexes with DNA Bulge Sequences As the bulky ruthenium complexes only bound DNA where the minor groove was relatively wide, it was proposed that they would bind DNA secondary structures that contain a more open or flexible groove with significantly higher affinity than duplex DNA. This proposal was consistent with the results from Kirsch-De Mesmaeker, Keene and coworkers,97 who showed that [{Ru(phen)2}2(m-HAT)]4+ (HAT = 1,4,5,8,9,12-hexaazatriphenylene; Figure 11.2b) binds partially-denatured DNA significantly more strongly than duplex DNA. One example of an open DNA structure is a sequence that contains a so-called ‘bulge’ – the inclusion of one or more bases on one strand that have no base(s) on the complementary strand with which to form a base pair (see Figure 11.5). DNA bulge sites can be created during recombination between imperfectly homologous sequences, and are thought to play an important role in frame-shift mutagenesis and be specific recognition sites for some DNA binding proteins. Consequently, there is considerable interest in developing small molecules that target bulge sites. NMR experiments demonstrated that DD-[{Ru(Me2bpy)2}2(m-bpm)]4+ bound an oligonucleotide containing a single adenine bulge significantly more strongly than the corresponding control, nonbulged, oligonucleotide.98 Furthermore, NOESY experiments indicated that the ruthenium complex bound specifically at the bulge site in the self-complementary tridecanucleotide d(CCGAGAATTCCGG)2 (where the A in bold represents the adenine bulge). Simple binding models, consistent with

Figure 11.5 A structural tridecanucleotide

representation

of

the

adenine

bulge

(A4)

containing

Dinuclear Complexes

331

Figure 11.6 Molecular model of DD-[{Ru(bpy)2}(m-bpm){Ru(Me2bpy)2}]4+ bound to d(CCGAGAATTCCGG)2.98

the NOE data, showed that the adenine bulge site was intrahelical and base-stacked with the adjacent guanine residues (see Figure 11.6). The binding model also indicated that the DNA minor groove had significantly widened at the bulge site to accommodate the ruthenium complex. As the minor groove at the bulge site in the free tridecanucleotide was not significantly wider than ‘canonical’ DNA, it was concluded that the bulge site introduces an increased flexibility into the local DNA structure that allows the specific metal complex binding. Interestingly, while the DD-enantiomer of [{Ru(Me2bpy)2}2(m-bpm)]4+ bound relatively strongly at the bulge site of d(CCGAGAATTCCGG)2, the corresponding LL-isomer did not bind at the bulge site, with weak binding to the terminal nucleotide residues being observed. The study provided a rare example of total enantioselectivity in the binding of an inert transition metal complex with DNA.99,100 The meso-(DL)-diastereoisomer bound with an affinity intermediate between the DDand LL-enantiomers. While it had been demonstrated that dinuclear ruthenium complexes could target single base bulges in DNA, it was not known if these metal complexes have potential for the larger bulge sites (three or more bases) often found in DNA and

332


RNA. Molecular modelling suggested that the insertion of a three or more base bulge into duplex DNA should induce a bend towards the major groove that subsequently decreases the amount of helical twist at the bulge site.101 Consequently, the minor groove should be significantly straighter at the bulge site than in standard canonical form DNA, and therefore, a potentially good binding site for the rigid dinuclear ruthenium complexes. DD-[{Ru(phen)2}2(m-dppm)]4+ (dppm = 4,6-bis(2-pyridyl)pyrimidine; Figure 11.2c) was shown to bind at the A3-bulge site of the nonself-complementary dodecanucleotide duplex d(GCATCGAAAGCTACG)•d(CGTAGCCGATGC) with high affinity (K = 3 × 105 M−1) and selectivity.102 Interestingly, the NMR results suggested that at least one of the phen ligands is bound by partial intercalation. Upfield shifts of similar magnitude were observed for the phen resonances in a study of the binding of the well-known intercalating agent [Pt(en)(phen)]2+ (en = ethylenediamine) to a hexanucleotide.103 While the minor groove is no wider at the A3-bulge site compared to the normal duplex regions, molecular models showed that the smaller helical twist coupled with the 50 ° bend towards the major groove results in the minor groove being straighter at the A3-bulge sequence. This allows the bridging dppm ligand to follow the groove, and hence, fit more fully within the groove at the bulge site. Furthermore, possible steric clashes with the oligonucleotide backbone are also reduced for the two phenanthroline ligands that are almost parallel with the bridging dppm ligand. The remaining two phenanthroline ligands are then ideally placed to be inserted deeply within the duplex structure, as shown in Figure 11.7.

11.3.4 Interaction of Dinuclear Complexes with DNA Hairpin Sequences DNA hairpin sequences, or stem-loop structures – shown in Figure 11.8 – are secondary structural motifs known to arise naturally in base sequences possessing self-complementarity. Although not as common as in RNA, DNA hairpins and cruciforms (two opposed hairpins) are believed to play a role in a variety of fundamental biological processes, including the control of gene expression and mutagenic events.104–109 Many potential hairpin or cruciform structures have been identified in genomic DNA in regions associated with the regulation of transcription.105,106 Additionally, a number of DNA-binding proteins have been found to preferentially target such sequences.110–113 Conversely, a number of neurological diseases have been attributed to the disadvantageous formation of hairpin structures during the replication of trinucleotide repeat sequences.114,115 Hence, there is also significant interest in the design of small molecules that specifically target hairpin structures. The complexes meso-[{Ru(phen)2}2(m-HAT)]4+ and meso-[{Ru(4,7-Me2phen)2}2 (m-HAT)]4+ (4,7-Me2phen = 4,7-dimethyl-1,10-phenanthroline; Figure 11.1g) were shown to bind an icosanucleotide that contains a seven-base-pair stem section and a six-base loop region with greater affinity than the corresponding four-base loop octadecanucleotide and three control duplex dodecanucleotides.116 Based on the results from one- and two-dimensional NMR spectra, models of meso-[{Ru(phen)2}2 (m-HAT)]4+ and meso-[{Ru(4,7-Me2phen)2}2(m-HAT)]4+ bound to the six-base hairpin were produced. Each complex bound on the minor-groove side of the

Dinuclear Complexes

333

Figure 11.7 Two alternate orientations of an energy-minimized model of DD-[{Ru(phen)2}2(mdppm)]4+ bound at the bulge site of the A3-bulge oligonucleotide.102 The dinuclear ruthenium complex is coloured black

Figure 11.8 Schematic representation of a six-base hairpin loop sequence

334


icosanucleotide, with the HAT ligand positioned at the stem-loop interface with one set of terminal phenanthroline ligands projecting into the groove of the stem, and the other projecting into the loop region. The groove at the stem-loop interface is wider, or more flexible, than the duplex stem of the icosanucleotide. It was suggested that the binding was stabilized by energetically favourable van der Waals and hydrophobic interactions with the bulky dinuclear complexes. Of the two metal complexes, meso-[{Ru(4,7-Me2phen)2}2(m-HAT)]4+ was found to bind the six-base hairpin-containing icosanucleotide more strongly; however, it also bound the control duplex sequences with significant affinity. Secretion of the hydrophobic methyl groups away in the minor groove, and the subsequent displacement of solvent, was proposed as the driving force behind this greater binding ability of the methylated species. Alternatively, while meso-[{Ru(phen)2}2(m-HAT)]4+ bound the six-base hairpin sequence less strongly than the corresponding methyl analogue, it showed greater selectivity in its binding when comparing the affinity to the hairpin sequence and control duplex structures. Metal complexes possessing phen or 4,7-Me2phen terminal ligands bound the icosanucleotide hairpin structure with greater affinity than the ruthenium complexes containing corresponding bpy and Me2bpy terminal ligands.116 Presumably, the larger surface areas of the phenanthroline ligand compared to bipyridine results in better van der Waals contacts with the walls of the minor groove, or more favourable hydrophobic interactions, or allows a semi-intercalative association with DNA bases, or a combination of all three. With regards to bridging ligands, the various studies demonstrated that ruthenium complexes containing a bpm bridge are generally selective for bulge sites, while a HAT bridging ligand is preferable for targeting hairpin structures. It was also noted that the meso diastereoisomer gives the strongest binding, particularly in those complexes possessing phenanthroline ligands. Simple computer models illustrated that the terminal ligands of a meso isomer are able to follow the contours of the minor groove, whereas the DD- and LL-enantiomeric forms of the complex encounter steric clashes at either end of the complex.116

11.3.5 Interaction of Dinuclear Complexes with RNA Sequences Although it has been clearly established that inert dinuclear ruthenium(II) complexes preferentially bind to nonduplex DNA structures, little is know about their interaction with RNA. Nonduplex structures occur in both DNA and RNA, but are much more prevalent in the latter. The formation of such structures is generally transient and/or disadvantageous in DNA. Alternatively, RNA exhibits a large diversity of nonduplex structures, which have an important role for recognition sites for proteins, RNA splicing and RNA folding, making them promising targets for potential diagnostic and therapeutic agents. Given the differences in the conformations of DNA and RNA, it is not possible to extrapolate the results obtained for DNA to RNA. Standard form DNA adopts a B-type helix, whereas RNA adopts an A-type structure. This results in significant differences in the groove dimensions – DNA contains a wide major groove and relatively narrow minor groove, whereas, RNA has a deep and very narrow major groove and a wide shallow minor groove (see Figure 11.9).117

Dinuclear Complexes

335

Figure 11.9 Comparative models of B- (left) and A-form (right) nucleic acids. The major and minor grooves are labelled

In order to make a direct comparison with the results obtained with DNA, the binding of the two enantiomers of [{Ru(Me2bpy)2}2(m-bpm)]4+ to the RNA analogue of the adenine bulge-containing DNA sequence d(CCGAGAATTCCGG)2, and the corresponding bulge-free control RNA dodecanucleotide, were examined.118 Both enantiomers bound the control RNA duplex weakly (K = 1 × 103 M−1), but bound the r(CCGAGAAUUCCGG)2 structure with considerably greater affinity (K = 6 × 104 M−1). From NOESY spectra, it was established that the ruthenium complex did selectively bind at the bulge site. A binding model that was consistent with all the NMR was obtained that demonstrated that DD-[{Ru(Me2bpy)2}2(mbpm)]4+ could effectively associate at the bulge site in the RNA minor groove. It was proposed that the insertion of the bulge into the tridecanucleotide formed a metal complex binding pocket by partially unwinding the duplex at the unpaired adenine. This perturbation to the RNA duplex caused a reduction in the curvature of the minor groove that may have aided the binding of the ruthenium complex. Alternatively, the minor groove structure at the bulge site may be more flexible, due to the reduced thermodynamic stability produced by the unpaired adenine, thereby allowing a better induced-fit binding site. The results of the study demonstrated that inert dinuclear ruthenium complexes have excellent potential as probes for nonduplex RNA structures, in a similar manner to that previously established for DNA. 11.3.6 Dinuclear Ruthenium Complexes with Flexible Ligands Although the dinuclear ruthenium complexes that are bridged by rigid linking ligands, such as bpm and HAT, exhibit relatively strong binding to nonduplex DNA

336


and RNA structures, it is generally only possible for one metal centre to bind deeply within the minor groove. The rigid ruthenium complex cannot follow the curvature of the groove in nonduplex structures that are not dramatically altered from the standard duplex, and hence the second metal centre must project out of the groove. It was proposed that dinuclear ruthenium complexes linked by a flexible chain could overcome this limitation, and potentially allow both metal centres to bind optimally in the minor groove. The stereoisomers of a series of dinuclear ruthenium(II) complexes [{Ru(phen)2}2(m-bbn)]4+ with flexible bridging ligands (bbn; Figure 11.2k) – bb2 (1,2bis[4(4′-methyl-2,2′-bipyridyl)]ethane), bb5 (1,5-bis[4(4′-methyl-2,2′-bipyridyl)]pentane), bb7 (1,7-bis[4(4′-methyl-2,2′-bipyridyl)]heptane) and bb10 (1,10-bis[4(4′-methyl2,2′-bipyridyl)]decane} – were synthesized and their binding to the single adenine bulge-containing tridecanucleotide, d(CCGAGAATTCCGG)2, studied.119 Based on the results from a fluorescent DAPI-displacement assay (DAPI = 4′,6-diamidino-2phenylindole; see Section 11.3.7), equilibrium dialysis and affinity binding chromatography experiments it was determined that the DD-enantiomer of the bb7-bridged complex bound the bulge-containing oligonucleotide with the highest affinity. NMR experiments demonstrated that the DD-bb7-bridged complex still selectively targeted the bulge site, while molecular modelling showed the ability of the flexibly linked complex to follow the curvature of the DNA minor groove, as shown in Figure 11.10. It was concluded that flexibly linked dinuclear complexes will still exhibit the selectivity for nonduplex DNA sites observed for the rigid complexes,

Figure 11.10 Comparison of the binding of DD-[{Ru(phen)2}2(m-bb7)]4+(left) and DD[{Ru(phen)2}2(m-bpm)]4+ (right) to the adenine-bulged tridecanucleotide duplex. The complexes are coloured black, and the arrows indicate the location of the second of the ruthenium metals centres (after the first one has positioned itself in the bulge region) showing how in the flexibly linked complex the linker follows the curve in the groove119

Dinuclear Complexes

337

but they should be able to bind with higher affinity. No conclusive reason was proposed for why DD-[{Ru(phen)2}2(m-bb7)]4+ bound more strongly than the corresponding complexes containing two, five or ten methylene groups. It was considered likely that the bb2-bridged complex contains insufficient degrees of freedom to bind tightly, whereas the preference of the bb7-bridged complex over the bb5- and bb-10 bridged complexes may rely on the differential binding potential of the DNA sequence where the second ruthenium centre binds. 11.3.7 Fluorescent DAPI-Displacement Assay While NMR spectroscopy provides the most detailed information on the nature of the binding of the ruthenium complexes with any of the DNA or RNA structures, it is a relatively time-consuming and expensive technique (in terms of the cost of the oligonucleotides). A suitable method is required to screen the relative binding affinity of a large number of ruthenium complexes, in terms of variable ligands and stereoisomers, against a range of DNA and RNA structures. Boger et al.120 developed a now widely used fluorescent intercalator displacement (FID) assay to screen organic compounds that bind, by intercalation or groove association, with oligonucleotides. The assay is based upon the loss of fluorescence resulting from the displacement of an intercalating dye such as ethidium bromide (EthBr) or thiazole orange (TO) from a DNA sequence by the molecule of interest. While the assay appears to be widely applicable, some anomalies were observed with dinuclear ruthenium complexes between the results obtained with the FID method and observations on the same oligonucleotide–metal complex combinations using other techniques. As the disparity may have been due to differences in binding modes between the intercalating dyes and the minor groove-binding metal complexes, a modified fluorescence assay was proposed that used the minor groovebinding compound DAPI. It was suggested that there may be a direct displacement of the fluorescent DAPI dye by the minor groove-binding ruthenium complexes.121 In a systematic evaluation of the DNA-binding affinities of a range of dinuclear ruthenium complexes with various oligonucleotides, some notable inconsistencies were observed between the results of the FID fluorescence assay using an intercalating dye, compared with DAPI. The metal complex-oligonucleotide combinations that gave inconsistent results in the fluorescent displacement assays were further examined by other techniques, including NMR, affinity chromatography and equilibrium dialysis. The results from the DAPI assay were totally consistent with the relative metal complex-oligonucleotide affinities determined by the other techniques. It was concluded that the difference in the binding mode between the intercalating dye and the groove-binding ruthenium complexes were responsible for the discrepancies.121 11.3.8 Affinity Chromatography The interactions of DNA with metal complexes may be utilized in the ‘reverse’ sense to develop a remarkably efficient affinity chromatographic technique for the

338


separation of stereoisomers of dinuclear species.122 The attachment of selected biotinylated oligonucleotides to a stationary phase comprising streptavidin immobilized on Sepharose has allowed separation over very short distances (2–5 cm) of mixtures of different complexes, the separation of geometric isomers and diastereoisomers of the same complex, and resolution of chiral forms – separations which have only been achieved by the conventional cation-exchange column techniques with an effective column length in excess of 30 metres!

11.4 Potential Biological Significance While it has been clearly demonstrated that dinuclear ruthenium complexes have considerable potential as probes for nonduplex DNA and RNA, this potential will only be realized if: (1) the metal complexes can be functionalized so that they selectively bind specific nonduplex structures that are biologically significant; and/or (2) the metal complexes can be transported into human cells, and thereby elicit some biological response, e.g. they are cytotoxic. Given the great diversity of organic ligands that can be readily synthesized, it is probable that once specific nonduplex sites are identified, hydrogen bond donors and acceptors and/or van der Waals interaction recognition groups can be incorporated into the metal complex. An excellent example of this type of approach was reported by Barton and coworkers, who specifically designed the DNA-intercalating complex D-a-[Rh{(R,R)-Me2trien}(phi)]3+ ((R,R)-Me2trien = 2R,9R-diamino-4,7diazadecane; Figure 11.1m) to bind to the sequence 5′-TGCA-3′.123 The intercalating ligand phi locks the rhodium complex into the DNA duplex. Additional recognition elements, hydrogen bonding NH2 groups and methyl groups that can make van der Waals contacts with the TMe groups, were then incorporated into the metal complex in the correct three-dimensional position to interact with their DNA counterpart.123 Although there has been one recent study,124 little is known about the cellular transport and cytotoxicity of dinuclear ruthenium complexes. However, the biological properties of a range of mononuclear ruthenium complexes have been reported. Puckett and Barton125 examined the cellular transport of a series of DNA-intercalating ruthenium complexes, and demonstrated that those with greater lipophilicity exhibit highest uptake, while charge and size were not particularly important. Furthermore, Liu et al.126 reported that [Ru(phen)2(PIP)]2+ (where PIP = R-phenylimidazo[4,5-f] [1,10]phenanthroline; Figure 11.1t), which binds DNA by intercalation, exhibited substantial cytotoxicity in a range of cell lines. Given that the cytotoxicity of many organic compounds is due to their ability to intercalate into the DNA helix, it is not surprising that ruthenium complexes that bind DNA in a similar mode also exhibit reasonable cytotoxicity. Similarly, examples of DNA groove-binding ruthenium complexes that are cytotoxic have been reported. Ma et al.127 demonstrated that although [Ru(tBu2bpy)2(2-appt)]2+ (tBu2bpy = 4,4′bis(tert-butyl)-2,2′-bipyridine; Figure 11.1c; 2-appt = 2-amino-4-phenylamino-6-(2-

Abbreviations of Ligands

339

pyridyl)-1,3,5-triazine; Figure 11.1r) bound DNA with only modest affinity, it also exhibited moderate cytotoxicity in several human cancer cell lines. Additionally, Spillane et al.128 reported that the DNA minor groove binding metal complex [Ru(bpy)2(bbtb)]2+ (bbtb = 4,4′-bis(benzothiazol-2-yl)-2,2′-bipyridine; Figure 11.1s) showed significant cytotoxicity in a range of human cancer cell lines. Due to the success (and associated clinical problems) of cisplatin as an anticancer agent, there has been intense interest in transition-metal-based compounds that bind DNA. Through the early studies by the research groups of Barton and Nordén, ruthenium complexes that can bind DNA and RNA by intercalation are now well-established probes of nucleic acids. Over the last five or so years, dinuclear ruthenium complexes that can only bind nucleic acids through association in the minor groove have also become recognized as valuable probes for nonduplex nucleic acid structures. As nonduplex nucleic acid structures have been shown to be important in the regulation of gene expression, di- and oligonuclear ruthenium complexes may have significant potential as therapeutic agents for a range of diseases. Furthermore, given the vast amount of classical inorganic chemistry research that has involved ruthenium(II) complexes, it is probable that this class of nucleic acidbinding metal complex can be readily tuned to achieve the desired biological outcomes.

Abbreviations of Ligands 2-appt bbn

bbtb bipp bpm bpy t Bu2bpy C4(cpdppz)2

chrysi DIP dpb 2,3-dpp dppm dppx dppz dppz(11,11′)dppz

2-amino-4-phenylamino-6-(2-pyridyl)-1,3,5-triazine, Figure 11.1r bb2 {1,2-bis[4(4′-methyl-2,2′-bipyridyl)]ethane}, bb5 {1,5bis[4(4′-methyl-2,2′-bipyridyl)]pentane}, bb7 {1,7-bis[4(4′methyl-2,2′-bipyridyl)]heptane}, and bb10 {1,10-bis[4(4′-methyl-2,2′-bipyridyl)]decane}, Figure 11.2k 4,4′-bis(benzothiazol-2-yl)-2,2′-bipyridine, Figure 11.1s 2,9-bis(2-imidazo[4,5-f][1,10]phenanthroline)-1,10phenanthroline, Figure 11.2i 2,2′-bipyrimidine, Figure 11.2a 2,2′-bipyridine, Figure 11.1a 4,4′-bis(tert-butyl)-2,2′-bipyridine, Figure 11.1c N,N′-bis(12-cyano-12,13-dihydro-11H-8-cyclopenta[b]dipyrido[ 3,2-h:2′,3′-j]phenazine-12-carbonyl)-1,4-diaminobutane, Figure 11.2h 5,6-chrysene quinone diimine, Figure 11.1v 4,7-diphenyl-1,10-phenanthroline, Figure 11.1o 2,3-bis(2-pyridyl)benzo[g]quinoxaline, Figure 11.2d 2,3-bis(2-pyridyl)pyrazine, Figure 11.2e 4,6-bis(2-pyridyl)pyrimidine, Figure 11.2c 7,8-dimethyldipyridophenazine, Figure 11.1l dipyrido[3,2-a:2′,3′-c]phenazine, Figure 11.1j 11,11′-bis(dipyrido[3,2-a:2′,3′-c]phenazine, Figure 11.2g

340


dpq dpqC HAT IP Me2bpy 2,9-Me2phen 5,6-Me2phen 4,7-Me2phen (R,R)Me2trien phen phi picchxnMe2 PIP pztp tactp TMP tpphz

dipyrido[3,2-d:2′,3′-f]quinoxaline, Figure 11.1i dipyrido[3,2-a:2′,3′-c](6,7,8,9-tetrahydro)phenazine, Figure 11.1k 1,4,5,8,9,12-hexaazatriphenylene, Figure 11.2b imidazole[4,5-f][1,10]phenanthroline, Figure 11.1n 4,4′-dimethyl-2,2′-bipyridine, Figure 11.1b 2,9-dimethyl-1,10-phenanthroline, Figure 11.1e 5,6-dimethyl-1,10-phenanthroline, Figure 11.1f 4,7-dimethyl-1,10-phenanthroline, Figure 11.1g 2R,9R-diamino-4,7-diazadecane, Figure 11.1m 1,10-phenanthroline, Figure 11.1d phenanthrenequinone diimine, Figure 11.1p N,N′-dimethyl-N,N′-di(2-picolyl)-1,2-diaminocyclohexane, Figure 11.1q R-phenylimidazo[4,5-f][1,10]phenanthroline, Figure 11.1t 3-pyrazin-2-yl)-as-triazino[5,6-f]-1,10-phenanthroline, Figure 11.2f 4,5,9,18-tetraazachrysenol(9,10-b)-triphenylene, Figure 11.1u 3,4,7,8-tetramethyl-1,10-phenanthroline, Figure 11.1h tetrapyrido[3,2-a:2′,3′-c:3′′,2′′-h:2′′,3′′-j]phenazine, Figure 11.2j

References 1. Shulman, A.; Dwyer, F.P.; Metal chelates in biological systems. In: Chelating Agents and Metal Chelates, Dwyer, F.P. and Mellor, D.P., Eds, Academic, New York, 1964; and references therein. 2. Juris, A.; Barigelletti, F.; Campagna, S.; Balzani, V.; Belser, P.; von Zelewsky, A.; Ru(II) polypyridine complexes: photophysics, photochemistry, electrochemistry, and chemiluminescence; Coord. Chem. Rev., 1988, 84, 85–277. 3. Barton, J.K.; Danishefsky, A.T.; Goldberg, J.M.; Tris(phenanthroline)ruthenium(II): stereoselectivity in binding to DNA; J. Am. Chem. Soc., 1984, 106, 2172–2176. 4. Barton, J.K.; Basile, L.A.; Danishefsky, A.; Alexandrescu, A.; Chiral probes for the handedness of DNA helices: enantiomers of tris(4,7-diphenylphenanthroline)ruthenium (II); Proc. Natl. Acad. Sci. USA, 1984, 81, 1961–1965. 5. Kumar, C.V.; Barton, J.K.; Turro, N.J.; Photophysics of ruthenium complexes bound to double helical DNA; J. Am. Chem. Soc., 1985, 107, 5518–5523. 6. Satyanarayana, S.; Dabrowiak, J.C.; Chaires, J.B.; Neither D- nor L-tris(phenanthroline) ruthenium(II) binds to DNA by classical intercalation; Biochemistry, 1992, 31, 9319–9324. 7. Coury, J.E.; Anderson, J.R.; McFaillsom, L.; Williams, L.D.; Bottomley, L.A.; Scanning force microscopy of small ligand-nucleic acid complexes: tris(o-phenanthroline) ruthenium(II) as a test for a new assay; J. Am. Chem. Soc., 1997, 119, 3792–3796. 8. Collins, J.G.; Sleeman, A.D.; Aldrich-Wright, J.R.; Greguric, I.; Hambley, T.W.; A 1H NMR study of the DNA binding of ruthenium(II) polypyridyl complexes; Inorg. Chem., 1998, 37, 3133–3141. 9. Lincoln, P.; Nordén, B.; DNA binding geometries of ruthenium(II) complexes with 1,10phenanthroline and 2,2′-bipyridine ligands studied with linear dichroism spectroscopy. Borderline cases of intercalation; J. Phys. Chem. B, 1998, 102, 9583–9594. 10. Ji, L.-N.; Zou, X.-H.; Liu, J.-G.; Shape- and enantioselective interaction of Ru(II)/Co(III) polypyridyl complexes with DNA; Coord. Chem. Rev., 2001, 216–217, 513–536.

References

341

11. Kim, H.-K.; Nordén, B.; Absence of chiral discrimination in the interaction of tris(diphenylphenanthroline)ruthenium(II) with DNA; Chem. Commun., 1997, 2375– 2376. 12. Barton, J.K.; Goldberg, J.M.; Kumar, C.V.; Turro, N.J.; Binding modes and base specificity of tris(phenanthroline)ruthenium(II) enantiomers with nucleic acids: tuning the stereoselectivity; J. Am. Chem. Soc.; 1986, 108, 2081–2088. 13. Eriksson, M.; Leijon, M.; Hiort, C.; Nordén, B.; Gräslund, A.; Minor groove binding of [Ru(phen)3]2+ to [d(CGCGATCGCG)]2 evidenced by two-dimensional NMR; J. Am. Chem. Soc., 1992, 114, 4933–4934. 14. Ottaviani, M.F.; Ghatlia, N.D.; Bossmann, S.H.; Barton, J.K.; Dürr, H.; Turro, N.J.; Nitroxide-labeled ruthenium(II)-polypyridyl complexes as EPR probes to study organized systems. 2. Combined photophysical and EPR investigations of B-DNA; J. Am. Chem. Soc., 1992, 114, 8946–8952. 15. Coggan, D.Z.M.; Haworth, I.S.; Bates, P.J.; Robinson, A.; Rodger, A.; DNA binding of ruthenium tris(1,10-phenanthroline): Evidence for the dependence of binding mode on metal complex concentration; Inorg. Chem., 1999, 38, 4486–4497. 16. Lincoln, P.; Broo, A.; Nordén, B.; Diastereomeric DNA-binding geometries of intercalated ruthenium(II) trischelates probed by linear dichroism: [Ru(phen)2DPPZ]2+ and [Ru(phen)2BDPPZ]2+; J. Am. Chem. Soc., 1996, 118, 2644–2653. 17. Collins, J.G.; Aldrich-Wright, J.R.; Greguric, I.D.; Pellegrini, P.A.; Binding of the D- and L-Enantiomers of [Ru(dmphen)2dpq]2+ to the Hexanucleotide d(GTCGAC)2; Inorg. Chem., 1999, 38, 5502–5509. 18. Greguric, A.; Greguric, I.D.; Hambley, T.W.; Aldrich-Wright, J.R.; Collins, J.G.; Minor groove intercalation of D-[Ru(Me2phen)2dppz]2+ to the hexanucleotide d(GTCGAC)2; J. Chem. Soc., Dalton Trans., 2002, 849–855. 19. Metcalf, C.; Thomas, J.A.; Kinetically inert transition metal complexes that reversibly bind to DNA; Chem. Soc. Rev., 2003, 32, 215–224. 20. Jenkins, Y.; Barton, J.K.; A sequence-specific molecular light switch: tethering of an oligonucleotide to a dipyridophenazine complex of ruthenium(II); J. Am. Chem. Soc., 1992, 114, 8736–8738. 21. Olson, E.J.C.; Hu, D.; Hormann, A.; Jonkman, A.M.; Arkin, M.R.; Stemp, E.D.A.; Barton, J.K.; Barbara, P.F.; First observation of the key intermediate in the ‘light-switch’ mechanism of [Ru(phen)2dppz]2+; J. Am. Chem. Soc., 1997, 119, 11458–11467. 22. Murphy, C.J.; Arkin, M.R.; Jenkins, Y.; Ghatlia, N.D.; Bossmann, S.H.; Turro, N.J.; Barton, J.K.; Long-range photoinduced electron transfer through a DNA helix; Science, 1993, 262, 1025–1029. 23. Stemp, E.D.A.; Arkin, M.R.; Barton, J.K.; Electron transfer between metallointercalators bound to DNA: Spectral identification of the transient intermediate; J. Am. Chem. Soc., 1995, 117, 2375–2376. 24. Arkin, M.R.; Stemp, E.D.A.; Holmlin, R.E.; Barton, J.K.; Hormann, A.; Olson, E.J.C.; Barbara, P.F.; Rates of DNA-mediated electron transfer between metallointercalators; Science, 1996, 273, 475–480. 25. Lincoln, P.; Tuite, E.; Nordén, B.; Short-circuiting the molecular wire: cooperative binding of D-[Ru(phen)2dppz]2+ and D-[Rh(phi)2bipy]3+ to DNA; J. Am. Chem. Soc., 1997, 119, 1454–1455. 26. Schulman, L.S.; Bossmann, S.H.; Turro, N.J.; Analysis of luminescence quenching on calf thymus DNA; J. Phys. Chem., 1995, 99, 9283–9292. 27. Haq, I.; Lincoln, P.; Suh, D.C.; Nordén, B.; Chowdhry, B.Z.; Chaires, J.B.; Interaction of D- and L-[Ru(phen)2DPPZ]2+ with DNA: A calorimetric and equilibrium binding study; J. Am. Chem. Soc., 1995, 117, 4788–4796. 28. Hiort, C.; Lincoln, P.; Nordén, B.; DNA binding of D- and L-[Ru(phen)2DPPZ]2+; J. Am. Chem. Soc., 1993, 115, 3448–3454. 29. Moucheron, C.; Kirsch-De Mesmaeker, A.; Choua, S.; Photophysics of Ru(phen)2(PHEHAT)2+: A novel ‘light switch’ for DNA and photo-oxidant for mononucleotides; Inorg. Chem., 1997, 36, 584–592.

342


30. Bolger, J.; Gourdon, A.; Ishow, E.; Launnay, J.-P.; Mononuclear and binuclear tetrapyrido[3,2-a:2′,3′-c:3′′,2′′-h:2′′′,3′′′-j]phenazine (tpphz) ruthenium and osmium complexes; Inorg. Chem., 1996, 35, 2937–2944. 31. Zhen, Q.-X.; Ye, B.-H.; Liu, J.-G.; Zhang, Q.-L.; Ji, L.-N.; Wang, L.; Synthesis, characterization and DNA-binding properties of [Ru(phen)2taptp]2+ and [Ru(phen)2dptatp]2+; Inorg. Chim. Acta, 2000, 303, 141–147. 32. Greguric, I.; Aldrich-Wright, J.R.; Collins, J.G.; A 1H NMR study of the binding of D[Ru(phen)2DPQ]2+ to the hexanucleotide d(GTCGAC)2. Evidence for intercalation from the minor groove; J. Am. Chem. Soc., 1997, 119, 3621–3622. 33. Yun, B.H.; Kim, J.-O.; Lee, B.W.; Lincoln, P.; Nordén, B.; Kim, J.-M.; Kim, S.K.; Simultaneous binding of ruthenium(II) [(1,10-phenanthroline)2dipyridophenazine]2+ and minor groove binder 4′,6-diamidino-2-phenylindole to poly[d(A-T)2] at high binding densities: Observation of fluorescence resonance energy transfer across the DNA stem; J. Phys. Chem. B, 2003, 107, 9858–9864. 34. Lee, B.W.; Moon, S.J.; Youn, M.R.; Kim, J.H.; Jang, H.G.; Kim, S.K.; DNA to [Ru(1,10phenanthroline)2L]2+; Biophys. J., 2003, 85, 3865–3871. 35. Ji, L.-N.; Zou, X.-H.; Liu, J.-G.; Shape- and enantioselective interaction of Ru(II)/Co(III) polypyridyl complexes with DNA; Coord. Chem. Rev., 2001, 216–217, 513–536. 36. Han, D.; Wang, H.; Ren, N.; Molecular modeling of B-DNA site recognition by Ru intercalators: molecular shape selection; J. Mol. Model., 2004, 10, 216–222. 37. Proudfoot, E.M.; Mackay, J.P.; Karuso, P.; Probing site specificity of DNA binding metallointercalators by NMR spectroscopy and molecular modeling; Biochemistry, 2001, 40, 4867–4878. 38. Terbrueggen, R.H.; Johann, T.W.; Barton, J.K.; Functionalized rhodium intercalators for DNA recognition; Inorg. Chem., 1998, 37, 6874–6883. 39. David, S.S.; Barton, J.K.; NMR evidence for specific intercalation of D-[Rh(phen)2(phi)]3+ in [d(GTCGAC)2]; J. Am. Chem. Soc., 1993, 115, 2984–2985. 40. Campisi, D.; Morii, T.; Barton, J.K.; Correlations of crystal structures of DNA oligonucleotides with enantioselective recognition by Rh(phen)2phi3+: Probes of DNA propeller twisting in solution; Biochemistry, 1994, 33, 4130–4139. 41. Terbrueggen, R.H.; Barton, J.K.; Sequence-specific DNA binding by a rhodium complex: Recognition based on sequence-dependent twistability; Biochemistry, 1995, 34, 8227–8234. 42. Jackson, B.A.; Barton, J.K.; Recognition of DNA base mismatches by a rhodium intercalator; J. Am. Chem. Soc.; 1997, 119, 12986–12987. 43. Rüba, E.; Hart, J.R.; Barton, J.K.; [Ru(bpy)2(L)]Cl2: Luminescent metal complexes that bind DNA base mismatches; Inorg. Chem., 2004, 43, 4570–4578. 44. Chow, C.S.; Barton, J.K.; Shape-selective cleavage of tRNAPhe by transition metal complexes; J. Am. Chem. Soc., 1990, 112, 2839–2841. 45. Hudson, B.P.; Barton, J.K.; Solution structure of a metallointercalator bound site specifically to DNA; J. Am. Chem. Soc., 1998, 120, 6877–6888. 46. Liu, J.-G.; Ye, B.-H.; Li, H.; Ji, L.-N.; Li, R.-H.; Zhou, J.-Y.; Synthesis, characterization and DNA-binding properties of novel dipyridophenazine complex of ruthenium (II): [Ru(IP)2(DPPZ)]2+; J. Inorg. Biochem.; 1999, 73, 117–122. 47. Maheswari, P.U.; Rajendiran, V.; Palaniandavar, M.; Parthasarathi, R.; Subramanian, V.; Synthesis, characterization and DNA-binding properties of rac-[Ru(5,6-dmp)2(dppz)]2+ – enantiopreferential DNA binding and co-ligand promoted exciton coupling; J. Inorg. Biochem., 2006, 100, 3–17. 48. Pyle, A.M.; Rehmann, J.P.; Meshoyrer, R.; Kumar, C.V.; Turro, N.J.; Barton, J.K.; Mixedligand complexes of ruthenium(II): factors governing binding to DNA; J. Am. Chem. Soc., 1989, 111, 3051–3058. 49. Zeglis, B.M.; Pierre, V.C.; Barton, J.K.; Metallo-intercalators and metallo-insertors; Chem. Commun., 2007, 4565–4579. 50. Goldstein, B.M.; Barton, J.K.; Berman, H.M.; Crystal and molecular structure of a chiralspecific DNA-binding agent: tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II); Inorg. Chem., 1986, 25, 842–847.

References

343

51. Mei, H-Y.; Barton, J.K.; Chiral probe for A-form helices of DNA and RNA: tris(tetram ethylphenanthroline)ruthenium(II); J. Am. Chem. Soc., 1986, 108, 7414–7416. 52. Ruba, E.; Hart, J.R.; Barton, J.K.; [Ru(bpy)2(L)]Cl2: Luminescent metal complexes that bind DNA base mismatches; Inorg. Chem., 2004, 43, 4570–4578. 53. Aldrich-Wright, J.R.; Brodie, C.; Glazer, E.C.; Luedtke, N.W.; Elson-Schwaband, L.; Tor, Y.; Symmetrical dinuclear complexes with high DNA affinity based on [Ru(dpq)2(phen)]2+; Chem. Commun., 2004, 1018–1019. 54. O’Reilly, F.M.; Kelly, J.M.; Binding of bimetallic 1,10-phenanthroline ruthenium(II) complexes to DNA; New J. Chem., 1998, 22, 215–217. 55. Muller, W.; Crothers, D.M.; Studies of the binding of actinomycin and related compounds to DNA; J. Mol. Biol., 1968, 35, 251–290. 56. Glazer, A.N.; Rye, H.S.; Stable dye-DNA intercalation complexes as reagents for highsensitivity fluorescence detection; Nature, 1992, 359, 859–861. 57. Delaney, S.; Pascaly, M.; Bhattacharya, P.K.; Han, K.; Barton, J.K.; Oxidative damage by ruthenium complexes containing the dipyridophenazine ligand or its derivatives: a focus on intercalation; Inorg. Chem., 2002, 41, 1966–1974. 58. Carlson, D.L.; Huchital, D.H.; Mantilla, E.J.; Sheardy, R.D.; Murphy, W.R.; A new class of DNA metallobinders showing spectator ligand size selectivity: binding of ligandbridged bimetallic complexes of ruthenium(II) to calf thymus DNA; J. Am. Chem. Soc., 1993, 115, 6424–6425. 59. Wilhelmsson, L.M.; Esbjörner, E.K.; Westerlund, F.; Nordén, B.; Lincoln, P.; Meso stereoisomer as a probe of enantioselective threading intercalation of semirigid ruthenium complex [m(11,11′-bidppz)(phen)4Ru2]4+; J. Phys. Chem. B, 2003, 107, 11784–11793. 60. Önfelt, B.; Lincoln, P.; Nordén, B.; A molecular staple for DNA: Threading bis-intercalating [Ru(phen)2dppz]2+ dimer; J. Am. Chem. Soc., 1999, 121, 10846–10847. 61. Önfelt, B.; Lincoln, P.; Nordén, B.; Enantioselective DNA threading dynamics by phenazine-linked [Ru(phen)2dppz]2+ dimers; J. Am. Chem. Soc., 2001, 123, 3630–3637. 62. O’Reilly, F.; Kelly, J.; Kirsch-De Mesmaeker, A.; Interaction of a series of bimetallic ruthenium(II) bipyridyl complexes with DNA; Chem. Commun., 1996, 1013–1014. 63. O’Reilly, F.M.; Kelly, J.M.; Photophysical study of DNA-bound complexes containing two covalently linked [Ru(2,2′-bipyridine)3]2+-like centers; J. Phys. Chem. B, 2000, 104, 7206–7213. 64. Chao, H.; Yuan, Y-X.; Zhou, F.; Ji, L.-N.; Zhang, J.; DNA interactions of a dinuclear ruthenium(II) complex bridged by 1,3-bis(1,10-phenanthroline[5,6-d]imidazol-2yl)benzene; Transition Metal Chem., 2006, 31, 465–469. 65. Liu, F.; Wang, K.; Bai, G.; Zhang, Y.; Gao, L.; The pH-induced emission switching and interesting DNA-binding properties of a novel dinuclear ruthenium(II) complex; Inorg. Chem., 2004, 43, 1799–1806. 66. Narra, M.; Elliot, P.; Swavey, S.; Synthesis, characterization and DNA interactions of 5,15-(4-pyridyl)-10,20-(pentafluorophenyl)porphyrin coordinated to two [Ru(bipy)2Cl]+ groups; Inorg. Chim. Acta, 2006, 359, 2256–2262. 67. Wu, J.-Z.; Yuan, L.; Yu, Y.; Effect of copper ion on DNA intercalating behavior of a chelation unsaturated diruthenium polypyridyl complex; Inorg. Chim. Acta, 2006, 359, 718–720. 68. Wu, J.-Z.; Yuan, L.; Synthesis and DNA interaction studies of a binuclear ruthenium(II) complex with 2,9-bis(2-imidazo[4,5-f][1,10]phenanthroline)-1,10-phenanthroline as bridging and intercalating ligand; J. Inorg. Biochem., 2004, 98, 41–45. 69. Tysoe, S.A.; Kopelman, R.; Schelzig, D.; Flipping the molecular light switch off: Formation of DNA-bound heterobimetallic complexes using Ru(bpy)2tpphz2+ and transition metal ions; Inorg. Chem., 1999, 38, 5196–5197. 70. Holder, A.A.; Swavey, S.; Brewer, K.J.; Design aspects for the development of mixedmetal supramolecular complexes capable of visible light induced photocleavage of DNA; Inorg. Chem., 2004, 43, 303–308. 71. Swavey, S.; Brewer, K.J.; Visible light induced photocleavage of DNA by a mixedmetal supramolecular complex: [{(bpy)2Ru(dpp)}2RhCl2]5+; Inorg. Chem., 2002, 41, 6196–6198.

344


72. Milkevitch, M.; Storrie, H.; Brauns, E.; Brewer, K.J.; Shirley, B.W.; A new class of supramolecular, mixed-metal DNA-binding agents: the interaction of RuII,PtII and OsII,PtII bimetallic complexes with DNA; Inorg. Chem., 1997, 36, 4534–4538. 73. Williams, R.L.; Toft, H.N.; Winkel, B.; Brewer, K.J.; Synthesis, characterization, and DNA binding properties of a series of Ru, Pt mixed-metal complexes; Inorg. Chem., 2003, 42, 4394–4400. 74. Zou, X.-H.; Ye, B.-H.; Li, H.; Liu, J.-G.; Xiong, Y.; Ji, L.-N.; Mono- and bi-nuclear ruthenium(II) complexes containing a new asymmetric ligand 3-(pyrazin-2-yl)-as-triazino[5,6-f]1,10-phenanthroline: synthesis, characterization and DNA-binding properties; J. Chem. Soc., Dalton Trans., 1999, 1423–1428. 75. Jiang, C.-W.; Chao, H.; Hong, X.-L.; Li, H.; Mei, W.-J.; Ji, L.-N.; Enantiopreferential DNA-binding of a novel dinuclear complex [(bpy)2Ru(bdptb)Ru(bpy)2]4+; Inorg. Chem. Commun., 2003, 6, 773–775. 76. Jiang, C.-W.; Syntheses, characterization and DNA binding of mono- and diruthenium(II) complexes containing 2,2′-bis(1,2,4-triazino[5,6-f]phenanthren-3-yl)-4,4′-bipyridine and 2,2′-bipyridine ligands; Inorg. Chim. Acta, 2005, 358, 1231–1236. 77. Jiang, C.-W.; Chiral diruthenium(II) complexes containing a modified 4,4′-bipyridine bridging ligand: syntheses, characterization and DNA binding; Inorg. Chim. Acta, 2004, 357, 3403–3406. 78. Jiang, C.-W.; Dimeric 2,2-bipyridylruthenium(II) complexes containing 2,2-bis(1,2,4triazin-3-yl)-4,4-bipyridine-like bridging ligands: Syntheses, characterization and DNA binding; Eur. J. Inorg. Chem., 2004, 2277–2282. 79. Keene, F.R.; Stereochemistry and polymetallic ligand-bridged molecular assemblies; Coord. Chem. Rev., 1997, 166, 121–159. 80. Keene, F.R.; Isolation and characterisation of stereoisomers in di- and tri-nuclear complexes; Chem. Soc. Rev.; 1998, 27, 185–193. 81. Hua, X.; von Zelewsky, A.; Enantiomerically pure chiral RuII(L-L)2 building blocks for coordination compounds; Inorg. Chem.; 1995, 34, 5791–5797. 82. Rutherford, T.J.; Quagliotto, M.G.; Keene, F.R.; Chiral [Ru(pp)2(CO)2]2+ species (pp = bidentate polypyridyl ligand) and their use in the stereoselective synthesis of ligandbridge dinuclear complexes; Inorg. Chem., 1995, 34, 3857–3858. 83. Rutherford, T.J.; Pellegrini, P.A.; Aldrich-Wright, J.R.; Junk, P.C.; Keene, F.R.; Isolation of enantiomers of a range of tris(bidentate)ruthenium(II) species using chromatographic resolution and stereoretentive synthetic methods; Eur. J. Inorg. Chem., 1998, 1677–1688. 84. Fletcher, N.C.; Junk, P.C.; Reitsma, D.A.; Keene, F.R.; Chromatographic separation of stereoisomers of dinuclear ligand-bridged polypyridyl complexes of ruthenium(II); J. Chem. Soc., Dalton Trans., 1998, 133–138. 85. Fletcher, N.C.; Keene, F.R.; Anion interactions with (polypyridyl)ruthenium complexes, and their importance in the cation-exchange chromatographic separation of stereoisomers of dinuclear species; J. Chem. Soc., Dalton Trans., 1999, 683–689. 86. Kelso, L.S.; Reitsma, D.A.; Keene, F.R.; Spectral and electrochemical properties of the diastereoisomeric forms of azobis(2-pyridine)-bridged diruthenium species; Inorg. Chem., 1996, 35, 5144–5153. 87. Richardson, C.; Steel, P.J.; D’Alessandro, D.M.; Junk, P.C.; Keene, F.R.; Mono- and dinuclear complexes of the ligands 3,4-di(2-pyridyl)-1,2,5-oxadiazole and 3,4-di(2-pyridyl)1,2,5-thiadiazole; new bridges allowing unusually strong metal-metal interactions; J. Chem. Soc., Dalton Trans., 2002, 2775–2785. 88. Zampese, J.A.; Keene, F.R.; Steel, P.J.; Diastereoisomeric dinuclear ruthenium complexes of 2,5-di(2-pyridyl)thiazolo[5,4d]thiazole; Dalton Trans. 2004, 4124–4129. 89. Patterson, B.T.; Foley, F.M.; Richards, D.; Keene, F.R.; Synthetic and stereochemical studies of dinuclear ruthenium(II) with mixed terminal ligands; Dalton Trans., 2003, 709–714. 90. Foley, F.M.; Keene, F.R.; Collins, J.G.; The DNA binding of the DD-, DL- and LLstereoisomers of [{Ru(Me2bpy)2}2(m-bpm)]4+; J. Chem. Soc., Dalton Trans., 2001, 2968– 2974.

References

345

91. D’Alessandro, D.M.; Keene, F.R.; Diastereoisomers as probes for solvent reorganizational effects on IVCT in dinuclear ruthenium complexes; Chem. Phys., 2006, 324, 8–25. 92. Patterson, B.T.; Keene, F.R.; Isolation of geometric isomers within diastereoisomers of dinuclear ligand-bridged complexes; Inorg. Chem., 1998, 37, 645–650. 93. Rapenne, G.; Patterson, B.T.; Sauvage, J.-P.; Keene, F.R.; Resolution, X-ray structure and absolute configuration of a double-stranded helical diiron(II) bis(terpyridine) complex; Chem. Commun., 1999, 1853–1854. 94. Rutherford, T.J.; Keene, F.R.; Stereoisomers in heterometallic (Ru2Os) and heteroleptic homometallic (RuRu′Ru′′) trinuclear complexes incorporating the bridging ligand HAT (1,4,5,8,9,12-hexaazatriphenylene); J. Chem. Soc., Dalton Trans., 1998, 1155–1162. 95. Rutherford, T.J.; Van Gijte, O.; Kirsch-De Mesmaeker, A.; Keene, F.R.; Stereoisomers of mono-, di- and tri-ruthenium(II) complexes containing the bridging ligand 1,4,5,8,9,12hexaazatriphenylene (HAT), and studies of their photophysical and redox properties; Inorg. Chem., 1997, 36, 4465–4474. 96. Aldrich-Wright, J.R.; Greguric, I.D.; Lau, C.H.Y.; Pellegrini, P.; Collins, J.G.; Transition metal complexes as probes for DNA; Recent Res. Devel. Inorg. Chem., 1998, 1, 13–36. 97. Broadkorb, A.; Kirsch-De Mesmaeker, A.; Rutherford, T.J.; Keene, F.R.; Stereoselective interaction and photo-electron transfer between mononucleotides or DNA and the stereoisomers of a dinuclear Ru(II) HAT complex; Eur. J. Inorg. Chem., 2001, 2151–2160. 98. Patterson, B.T.; Collins, J.G.; Foley, F.M.; Keene, F.R.; Dinuclear ruthenium(II) complexes as probes for DNA bulge sites; J. Chem. Soc., Dalton Trans., 2002, 4343–4350. 99. Smith, J.A.; Collins, J.G.; Patterson, B.T.; Keene, F.R.; Total enantioselectivity in the DNA binding of the dinuclear ruthenium(II) complex [{Ru(Me2bpy)2}2(m-bpm)]4+ {bpm = 2,2′bipyrimidine; Me2bpy = 4,4′-dimethyl-2,2′-bipyridine}; Dalton Trans., 2004, 1277– 1283. 100. Morgan, J.L.; Buck, D.P.; Turley, A.G.; Collins, J.G.; Keene, F.R.; Meso-[{Ru(phen)2}2(mbpm)]4+: a high-affinity DNA bulge probe {bpm = 2,2′-bipyrimidine; phen = 1,10-phenanthroline}; Inorg. Chim. Acta, 2006, 359, 888–898. 101. Rosen, M.A.; Live, D.; Patel, D.J.; Comparative NMR study of AN-bulge loops in DNA duplexes: intrahelical stacking of A, A-A, and A-A-A bulge loops; Biochemistry, 1992, 31, 4004–4014. 102. Morgan, J.L.; Buck, D.P.; Turley, A.G.; Collins, J.G.; Keene, F.R.; Selectivity at a three-base bulge site in the DNA binding of DD-[{Ru(phen)2}2(m-dppm)]4+ {dppm = 4,6-bis(2pyridyl)pyrimidine; phen = 1,10-phenanthroline}; J. Biol. Inorg. Chem., 2006, 11, 824–834. 103. Collins, J.G.; Rixon, R.M.; Aldrich-Wright, J.R.; Interaction of [Pt(en)(phen)]2+ and [Pt(en)(phi)]2+ with the hexanucleotide d(GTCGAC)2: Evidence for minor groove binding; Inorg. Chem., 2000, 39, 4377–4379. 104. Batey, R.T.; Rambo, R.P.; Doudna, J.A.; Tertiary motifs in RNA structure and folding; Angew. Chem. Int. Ed., 1999, 38, 2326–2343. 105. Varani, G.; Exceptionally stable nucleic acid hairpins; Annu. Rev. Biophys. Biomol. Struct., 1995, 24, 379–404. 106. Mizuuchi, K.; Mizuuchi, M.; Gellert, M.; Cruciform structures in palindromic DNA are favored by DNA supercoiling; J. Mol. Biol., 1982, 156, 229–243. 107. Dai, X.; Greizerstein, M.B.; Nadas-Chinni, K.; Rothman-Denes, L.B.; Supercoil-induced extrusion of a regulatory DNA hairpin; Proc. Natl. Acad. Sci. USA, 1997, 94, 2174–2179. 108. Spiro, C.; McMurray, C.T.; Switching of DNA secondary structure in proenkephalin transcriptional regulation; J. Biol. Chem., 1997, 272, 33145–33152. 109. Glucksmann, M.A.; Markiewicz, P.; Malone, C.; Rothman-Denes, L.B.; Specific sequences and a hairpin structure in the template strand are required for N4 virion RNA polymerase promoter recognition; Cell, 1992, 70, 491–500. 110. Zazopoulos, E.; Lalli, E.; Stocco, D.M.; Sassone-Corsi, P.; DNA binding and transcriptional repression by DAX-1 blocks steroidogenesis; Nature, 1997, 390, 311–315.

346


111. Froelich-Ammon, S.J.; Gale, K.C.; Osheroff, N.; Site-specific cleavage of a DNA hairpin by topoisomerase II. DNA secondary structure as a determinant of enzyme recognition/cleavage; J. Biol. Chem., 1994, 269, 7719–7725. 112. Hao, Q.; Peumans, W.J.; van Damme, E.J.M.; Type-1 ribosome-inactivating protein from iris (Iris hollandica var. Professor Blaauw) binds specific genomic DNA fragments; Biochem. J., 2001, 357, 875-880. 113. Lilley, D.M.; The inverted repeat as a recognizable structural feature in supercoiled DNA molecules; Proc. Natl. Acad. Sci. USA, 1980, 77, 6468–6472. 114. Sutherland, G.R.; Richards, R.I.; Simple tandem DNA repeats and human genetic disease; Proc. Natl. Acad. Sci. USA, 1995, 92, 3636–3641. 115. Pearson, C.E.; Edamura, K.N.; Cleary, J.D.; Repeat instability: mechanisms of dynamic mutations; Nat. Rev. Genet., 2005, 6, 729–742. 116. Smith, J.A.; Morgan, J.L.; Turley, A.G.; Collins, J.G.; Keene, F.R.; Meso-[{Ru(phen)2}2(mHAT)]4+: a high-affinity DNA hairpin probe (HAT = 1,4,5,8,9,12-hexaazatriphenylene; phen = 1,10-phenantroline); Dalton Trans., 2006, 3179–3187. 117. Neidle, S.; DNA Structure and Recognition, IRL Press, Oxford, 1994. 118. Spillane, C.B.; Smith, J.A.; Buck, D.P.; Collins, J.G.; Keene, F.R.; Dinuclear ruthenium(II) complexes as potential probes for RNA bulge sites; Dalton Trans., 2007, 5290–5296. 119. Morgan, J.L.; Spillane, C.B.; Smith, J.A.; Buck, D.P.; Collins, J.G.; Keene, F.R.; Dinuclear ruthenium(II) complexes with flexible bridges as DNA bulge-selective probes; Dalton Trans. 2007, 4333–4342. 120. Boger, D.L.; Fink, B.E.; Brunette, S.R.; Tse, W.C.; Hedrick, M.P.; A simple, high-resolution method for establishing DNA binding affinity and sequence selectivity; J. Am. Chem. Soc., 2001, 123, 5878–5891. 121. Spillane, C.B.; Smith, J.A.; Morgan, J.L.; Keene, F.R.; DNA affinity binding studies using a fluorescent dye displacement technique – the dichotomy of the binding site; Chem., 2007, 12, 819–824. 122. Smith, J.A.; Keene, F.R.; Separation of stereoisomers of dinuclear metal complexes by binding affinity chromatography using non-duplex DNA; Chem. Commun., 2006, 2583–2585. 123. Hudson, B.P.; Dupureur, C.M.; Barton, J.K.; 1H NMR structural evidence for the sequence-specific design of an intercalator: D-a-[Rh[(R,R)-Me2trien]phi]3+ bound to d(GAGTGCACTC)2; J. Am. Chem. Soc., 1995, 117, 9379–9380. 124. McDonnell, U.; Kerchoffs, J.M.C.A.; Castineiras, R.P.M.; Hicks, M.R.; Hotz, A.C.G.; Hannon, M.J.; Rodger, A.; Synthesis and cytotoxicity of dinuclear complexes containing ruthenium(II) bipyridyl units linked by a bis(pyridylimine) ligand; Dalton Trans., 2008, 667–675. 125. Puckett, C.A.; Barton, J.K.; Methods to explore cellular uptake of ruthenium complexes; J. Am. Chem. Soc., 2007, 129, 46–47. 126. Liu, J.; Zheng, W.; Shi, S.; Tan, C.; Chen, J.; Zheng. K.; Ji, L.; Synthesis, antitumor activity and structure–activity relationships of a series of Ru(II) complexes; J. Inorg. Biochem., 2008, 102, 193–202. 127. Ma, D.-L.; Che, C.-M.; Siu, F.-M.; Yang, M.; Wong, K.-Y.; DNA binding and cytotoxicity of ruthenium(II) and rhenium(I) complexes of 2-amino-4-phenylamino-6-(2-pyridyl)1,3,5-triazine; Inorg. Chem., 2007, 46, 740–749. 128. Spillane, C.B.; Fletcher, N.C.; Roundtree, S.; van den Berg, H.; Chanduloy, S.; Morgan, J.L.; Keene, F.R.; Benzothiazole bipyridine complexes of ruthenium(II) with cytotoxic activity; J. Biol. Inorg. Chem., 2007, 12, 797–807.

12 DNA Recognition and Binding by Peptide–Metal Complex Conjugates Alexandra Myari and Nick Hadjiliadis

12.1

Introduction

The design and synthesis of chemical agents capable of binding specifically to DNA have gained great research interest in recent years.1 The discovery of novel sequence specificities, beyond the sequences recognized by the known natural restriction endonucleases, may possibly offer extended possibilities in DNA manipulation.2 Such specific DNA-binding agents could find several applications in molecular biology (sequencing, cloning) and clinical applications (gene regulation, gene therapy, chemotherapy).2 X-ray crystallography3,4 has revealed many molecular details regarding protein– DNA recognition. These involve the formation of a hydrogen-bonding network between peptide donor groups (coming from the peptide backbone and amino acid side chains) and the edges of the bases, consisting of what is called direct readout. In addition, formation of specific complexes between the protein and DNA is often followed by conformational changes of both protein and DNA molecules as well as release of water molecules part of the DNA hydration spine.3–5 Peptides are low molecular weight DNA-binding agents that can effectively mimic the bulky proteins. Peptides have the same donor groups utilized by proteins in the process of DNA recognition. They are also easily synthesized and their amino-acid composition and chirality can be tuned in order to achieve the desired DNA-binding properties. Binding of peptides has been thoroughly investigated and Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

348

DNA Recognition and Binding by Peptide–Metal Complex Conjugates

sequence selectivity has been demonstrated in many cases.6 However, short peptides lack the DNA-binding affinity of the native proteins. This is why dimeric peptides have been used for such studies, instead.7 Peptides can become powerful tools in our hands if a specific protein–DNA recognition code is determined. There is a big debate between different research groups as to whether a ‘DNA recognition code’ exists.8 As many structures of DNA–transcription factor complexes became known in atomic detail, it was recognized that some DNA-binding proteins share the same structural framework, such as helix-turn-helix, zinc finger and the basic domain leucine zipper motifs. Suzuki et al.8 demonstrated that a simple code for DNA recognition can be applied in the case of transcription factors. According to Suzuki et al.,8 the major part of the DNA recognition code consists of two types of rules; chemical and stereochemical. The chemical rules are based on the intrinsic chemical ability of a given residue and a base to form chemical contacts, either a hydrogen bond or hydrophobic interaction in the major groove. The chemical rules making up the DNA recognition code are given in Table 12.1. In addition, the main target for hydrophobic interaction in the major groove of DNA is the single methyl group of the T base, being recognized by hydrophobic residues such as Ala, Val, Ile, Leu, Met, Phe, Tyr, Trp and Thr. Stereochemical rules are summarized in a stereochemical chart which is deduced from crystal and NMR structures of DNA–protein complexes.8 This stereochemical chart indicates which

Table 12.1 Single amino acid–single base contacts (chemical rules). The ‘specific’ residue partners are shown in bold, while nonspecific partners are in plain text. Chemical merit points, semi-arbitrary numbers associated with particular contacts, are used to quantify the energy and specificity of a pairing between an amino acid residue and a nucleotide base. For example, the interaction of Arg (to G), which is particularly favourable and specific, with the residue receives 15 merit points, while the interaction between Ser (to any base), which is less specific, is given 10 points. (M. Suzuki, N. Yagi, DNA-recognition code of transcription factors in the helix-turn-helix,probe helix,hormone receptor, and zinc finger families. Proc. Natl. Acad. Sci, 91, 12357–12361. Copyright 1994, National Academy of Sciences, USA.) small

medium

large

aromatic

A

Cys, Ser, Thr

10

Asn Asp His

15 9 8

Gln Glu Arg, Lys Met

15 9 3 5

Tyr Trp

5 5

T

Ala Cys,Ser, Thr

10

Val, Ile Asn His

12 10 8

Leu, Met Gln Arg, Lys

12 10 5

Tyr, Phe Trp

G

Cys, Ser, Thr

10

His Asn

12 10

Arg, Lys Gln

15 10

Tyr

5

C

Val Cys, Ser, Thr

8 10

Asp Asn His Ile

12 10 8 8

Glu Gln Leu, Met

12 10 8

Tyr, Phe Trp

8 8

12 12

Transition Metal Complex–Peptide Conjugates

349

positions of the protein specifically contact bases and shows what residue sizes are compatible with these positions, resulting in a unique pattern shared by all members of each family of DNA-binding proteins.9 DNA recognition codes were also proposed for other families of DNA-binding proteins such as homeodomains,10 the EGR (early growth response factor) family of zinc-finger transcription factors11 and zinc-finger proteins12 However, it was realized that there is no clear one-to-one correspondence between amino acids and bases.13 Consequently, the rules cannot be generalized for all DNA-binding proteins. Furthermore, a rules-based approach is likely to fail in cases where the same pair interacts using a variety of geometries or a protein recognizes specific DNA sequences indirectly, i.e. via water-mediated contacts, via specific sequence-dependent conformational features and/or via binding-induced distortion of DNA.13 In many cases, peptides have been combined with a metal centre in order to increase their thermodynamic stability and DNA-binding affinity. In this context, research has followed mainly two directions. The first approach involves coordination of a metal ion to a peptide ligand forming metal–peptide complexes. The metal ion may play a structural role imparting structure to an unstructured peptide and/or conferring nucleolytic activity. Many excellent reviews are available with information about the DNA interactions of several types of metal–peptide complexes, such as metal complexes of small peptides,14 lanthanide-based metallopeptides,15 zincfinger peptides,16 chemical nuclease–protein conjugates.17 On the other hand, the second approach focuses on appending the peptide either to an aromatic ring or the backbone of another ligand participating in the coordination of a metal ion, thus forming metal complex–peptide chimeras or conjugates. Parent metal complexes are molecules with established DNA-binding properties. Tethering the peptide to such metal complexes is in most cases anticipated to result in sequence-specificity upon DNA-binding. Transition metal complex–peptide conjugates, discussed herein, offer a new potential in DNA binding and cleavage. They combine the DNA-binding properties of metal complexes accomplished via electrostatic and/or van der Waals interactions, or covalent binding with the ability to form specific contacts with the DNA bases provided by the peptide backbone and side chain donor groups. Additionally, an appropriately designed metal complex may confer its properties to the chimeric molecule resulting in DNA conformational, chemiluminescent or fluorescent probes and chemical nucleases.

12.2 Transition Metal Complex–Peptide Conjugates It is generally known that proteins often use a significant percentage of their amino acids to form the nonspecific contacts that provide the affinity to DNA, while they use an appreciably smaller number of amino acids to make direct contacts with the base pairs. Based on this concept, when constructing metal complex–peptide conjugates, the metal complex provides nonspecific affinity and less steric bulk and complexity than the full DNA binding protein, while the essential recognition site of the original peptide motif is maintained.

350


12.2.1

Dipeptides

Jan Reedijk and his collaborators have undertaken the construction of a library of peptide-tethered platinum complexes using automated solid-phase synthesis.18 The general structure of the target library is shown in Figure 12.1 and encompasses a dipeptide tethered to an ethylenediamine moiety (en), which, in turn, serves as a platinum chelating ligand. As constituents of the appended dipeptide, six amino acids were selected: glycine, phenylalanine, lysine, arginine, serine and glutamate. Incorporating these amino acids in both positions of the appended dipeptide resulted in a final library of 36 compounds. Subsequently, each compound was screened for in vitro anticancer activity and apparently no important cytotoxicity was observed compared to cisplatin. In continuation of this work, M. S. Robillard et al.19 used the linear amplification system in order to investigate the DNA damage intensity and sequence specificity of adduct formation by six peptide–platinum complexes, cisplatin (1) and Pt(en)Cl2 (2) in pUC19 plasmid as well as in intact Hela cells. The six peptide–platinum complexes were members of the previously described18 library of peptide–dichloroplatinum complexes and differed only in their terminal pendant amino acid: Gly, Phe, Lys, Arg, Ser and Glu. Interestingly, the appended dipeptides did not alter the DNA sequence specificity (runs of consecutive guanines) and did not increase the reactivity with DNA of the complexes compared to the unfunctionalized parent compounds, cisplatin and Pt(en)Cl2. On the other hand, the nature of the terminal amino acid affected the reactivity of the peptide–platinum complexes towards plasmid DNA, with the glycine-tethered complex and the phenylalanine-tethered complex causing the highest damage, followed by (in decreasing order) lysine-tethered, arginine-tethered, serine-tethered and glutamate-tethered. The fact that the negatively charged Glu displayed negligible activity implied that electrostatic interactions may play a role in the DNA binding. 12.2.2 Tripeptides Gly–His–Lys. The tripeptide glycyl-l-histidyl-l-lysine, GHK, has been isolated as a square-planar complex with copper and iron from human plasma and was found to act synergistically with these metals to alter the growth and metabolism of cultured hepatoma cells and hepatocytes.20 The Cu(II) atom is tetracoordinated with the Cl

Cl

O

Pt H2N H2C

NH CH2

C C H2

AA2 N H

O

H N C O

C AA1

NH2

AA = Amino Acid Figure 12.1 The general structure of the dichloroplatinum(II) peptide complexes


H5 H 3C 4 m-bpy Lys O H 2N

His O

H CH N

α

C

C

H βC H

O H CH N

α β

CH2

γ

H C H H δC H H εC H

H3

H5 HN 1

C

Gly

H3'

H CH N

C

α

H

O

H5

2 2'

4'

H5'

H6

H6

bpy

H4 N

H3 H3'

2 2'

N

N Ru

2+

H6' H6'

N

N H6'

H6

H5

N

H4

351

2'

H3'

2

H3

H4' H 5' H5' H 4'

bpy

N3 H2

NH2

Figure 12.2 Structure of the complex D-[Ru(bpy)2(m-bpy-GHK)]2+ with atom numbering

imidazole nitrogen, a peptide nitrogen, the terminal glycyl amino group, and an oxygen atom from a water molecule.21,22 In order to investigate its possible sequence specificity upon DNA binding, N. Hadjiliadis and collaborators23,24 conjugated peptide GHK to the bpy ligand of the parent complex [Ru(bpy)3]2+ (bpy = 2,2′-bipyridine) (Figure 12.2) which is known to bind DNA reversibly25,26 and induce photoactivated cleavage.27 A 1H NMR study of the interaction of the resulting diastereomeric complexes L- and D-[Ru(bpy)2(mbpy-GHK)]2+ (m-bpy = 4-carboxy-4′-methyl-2,2′-bipyridine) with two different DNA sequences, the d(CGCGAATTCGCG)223 and d(CGCGATCGCG)2,24 revealed that minor changes in the DNA sequence can result in different binding characteristics. In the case of the system D-[Ru(bpy)2(m-GHK)]2+–d(CGCGAATTCGCG)2 1:1, molecular models based on the NOE contacts (Figure 12.3) showed that the Disomer bound to the DNA major groove with the two bpy ligands oriented towards the double helix. The L-isomer, on the other hand, approached the DNA minor groove using the m-bpy ligand placing the lysine ε-amino group close to the DNA phosphates (Figure 12.3).23 Furthermore, significant chemical shifts of peptide and cytosine’s exocyclic amino group protons, as well as several intermolecular NOEs demonstrated that the peptide GHK was oriented towards the DNA major groove close to the central AATT sequence forming possibly specific contacts with the DNA bases. However, no significant interaction was observed for the D-[Ru(bpy)2(mGHK)]2+–d(CGCGATCGCG)2 system at ratio 1:1. Interestingly, increase of the metal complex to DNA duplex ratio resulted in binding to the major groove and strand opening of the DNA duplex. At 2:1 ratio, the majority of the intermolecular NOEs took place between the peptide Gly–His–Lys and sugar protons located in the major groove. This observation implied an important role for the peptide upon

352


Figure 12.3 Molecular models of the interactions between [d(CGCGAATTCGCG)2] and D[Ru(bpy)2(m-GHK)]Cl2 (A) and L-[Ru(bpy)2(m-GHK)]Cl2 (B) reflecting the average binding of each complex to the oligonucleotide. (A. Myari, N.Hadjiliadis, A.Garoufis, Synthesis and characterization of the diastereomers L- and D-[Ru(bpy)2(m-bpy.Gly-His-L-His-L-Lys)]Cl2. 1H NMR studies on their interactions with the deoxynucleotide duplex d[5′-CGCGAATTCGCG3′], Eur. J. Inorg. Chem., 2004, 7, 1435–1436. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

interaction of the chimera with DNA, but the absence of significant shifts of the peptide functional groups (or their neighbouring protons) indicated that probably no specific contact between peptide and the DNA bases was formed, rather an electrostatic type of interaction was assumed. Photocleavage studies were also conducted with the plasmid pUC19 and a 158 bp restriction fragment showing that both diastereomers cleave DNA with similar efficiency, attacking mainly the guanines of the sequence probably by generating active oxygen species.24 In an attempt to combine the notable cytotoxicity observed for chloro(polypyridyl) ruthenium complexes28 with the possible DNA recognition specificity provided by the peptide moiety, K. Karidi et al. synthesized the chloro-complex of the tethered peptide [Ru(terpy)(4-CO2H-4′-Mebpy-Gly-l-His-l-LysCONH2)Cl](PF6).29 The interactions of the ruthenium peptide complex with CT-DNA and plasmid DNA were studied with various spectroscopic techniques, showing coordinative binding of the ruthenium complex to DNA, after hydrolysis of the coordinated chloride, and a preference for the bases guanine and cytosine over the bases thymine and adenine. Electrostatic interactions between the complex cation and the polyanionic DNA chain assisted this binding whereas DNA unwinding occurred in all cases with high binding ratio (Ru/base) values (r > 0.3). Moreover, interaction of the peptide with DNA was evidenced by 31P NMR spectroscopy for the one of the two positional isomers of the ruthenium complex.


353

Numerous ruthenium nitrosyl complexes containing polypyridyl ligands have been synthesized, with possible uses as regulators of blood pressure30,31 or as antitumour agents releasing NO within tumour cells.32,33 Ruthenium nitrosyls may photorelease the coordinated NO in aqueous media.34,35 The complex [RuII(terpy)(4COGHK-4′-Mebpy)(NO)] was found to be stable only in nonaqueous solvents, and in contact with the water it is transformed into the nitro derivative.36 Photoactivation with visible light of the nitrosyl complex in dry MeCN resulted in the release of NO, while there was partial release of NO2− from the nitro complex. A DNA-binding study was performed in order to investigate the possible role of the peptide GHK using the nitro complex, since it was the only one stable in aqueous media. Interaction of the nitro complex with short fragments (70–300 bp) of calf thymus DNA induced slight shortening of the apparent polynucleotide length, while the peptide GHK was found to interact with the DNA helix in a synergistic way with the whole complex.36 Gly–Gl–Ser. Recently, M. J. Hannon et al.37 explored the possibility of attaching peptides to metallosupramolecular cylinders in order to create sequence-specific DNA binding agents targeted to the less-exploited DNA major groove. The parent cylinders were tetracationic triple stranded (Fe(II)) and dicationic double-stranded (Cu(I) or Ag(I)), based on two bis-pyridylimine ligands with the peptide functionality placed at position 5 of the pyridine ring (termed Lp1 and Lp2, Figure 12.4). The peptide selected for conjugation to the cylinder was the tripeptide Gly-Gly-Ser-CoNH2. The two glycine residues were selected in order to provide the minimal steric bulk close to the cylinder, whereas the serine residue at the Cterminal was chosen based on its known frequent occurrence in protein motif–DNA base contacts. The parent tetracationic triple-stranded iron(II) cylinder binds with CT-DNA in the major groove, inducing dramatic intramolecular DNA coiling38 in the heart of three-way and other Y-shaped junctions.39a,b The parent Cu(I) dicationic cylinder also binds CT-DNA and in the presence of hydrogen peroxide induces H2 C H C H2NOC-Ser-Gly-Gly

H N

C

N

N

N

N

Lp1 CH3 H C

C O

N

H3C

5

C

H N

Gly-Gly-Ser-CONH2

CH2 N H2 C

Lp2

6

4

H3 C

N CH2

3

O

H2 C

H2 C

H2NOC-Ser-Gly-Gly

2 1

O

H N

H C

H C

2 1

CH3

N

3

6

4 5

C

H N

Gly-Gly-Ser-CONH2

O

Figure 12.4 The bis-pyridylimine based, bis-peptide end functionalised ligands, Lp1 and Lp2 with atom numbering at the peptide-substituted pyridine ring

354


DNA cleavage.39c DNA-binding experiments of the peptide-functionalized cylinders indicated that the Fe(II) cylinder retained its ability to coil DNA, albeit to less extent compared to the unsubstituted analogue, whereas the Cu(I) cylinder cleaved DNA, with a tendency for double-stranded cleavage.37 Apparently, this is a promising strategy for the construction of sequence-specific major-groove-binding molecules by attaching peptides able to confer sequence-specificity upon DNA binding. 12.2.3

Bioactive Peptides

J. Brunner and J.K. Barton have addressed the problem of insufficient cellular uptake of drugs by attaching a cell-penetrating peptide onto the bpy (2,2′bipyridine) ligand of the rhodium intercalator [Rh(chrysi)(phen)(bpy)]3+ (chrysi = chrysenequinone diimine, phen = 1,10-phenanthroline).40 d-octaarginine, a well-known cell-penetrating peptide for delivery to the nucleus,41 was appended to the rhodium complex. Rhodium intercalators bearing the sterically demanding ligand chrysi bind preferentially to single-base mismatched sites and, upon photoactivation, the complexes furthermore promote strand breakage neighbouring the mismatched site.42 The presence of the oligoarginines was found to increase the nonspecific binding affinity (at least two orders of magnitude) of the complexes for both matched and mismatched DNA, while photocleavage remained selective for the mismatched site. Confocal microscopy experiments confirmed the nuclear localization of the metal–peptide conjugates demonstrating that the delivery properties of the appended cell-penetrating peptide are not affected by the presence of the bulky octahedral metal complex. 12.2.4

De novo Designed Peptides

The variety of the different amino acid side chain functionalities offers a great tool in designing peptides of diverse composition in order to elucidate peptide–DNA interactions. K.D. Copeland et al. used metallointercalator–peptide conjugates as a model system for studying DNA–protein crosslinking.43 The ruthenium intercalator, [Ru-(phen)(bpy′)(dppz)]2+ (bpy′ = 4-(butyric acid)-4′-methyl-2,2′-bipyridine), binds with high affinity to DNA and initiates crosslinking through a flash-quench reaction. Pentamer peptides were designed in order to be long enough to allow interactions between the peptide and DNA, even when the dppz ligand is intercalated into the base stack, and short enough to remain unstructured. An Ala was added at the Ntermini intended as a flexible spacer, and a Gly at the C-termini simplified synthesis and was intended to adsorb end effects. Finally, the three central residues were varied. Two types of peptides were considered: peptides with positively charged residues (Lys or Arg) at positions 2 and 4 and peptides with aliphatic residues (Ala and Gly) at positions 2 and 4. The outcome of this study was that the peptide composition does impact binding and crosslinking. The conjugates bearing lysine residues generally showed greater affinity for DNA. Furthermore, the Lys–X–Lys or Arg–X–Arg peptides gave the highest levels of crosslinking. This difference in crosslinking activity may be attributed to increased interaction between the DNA


355

and the positively charged peptide or to the reactivity of the Lys residues. Variation of the central residue did produce some change in binding affinity. The binding constants increased in the order Ru-KWK < Ru-KAK < Ru-KYK. Even though the Trp residue was expected to contribute more in binding affinity through intercalation, probably it was not ideally positioned since tethered peptides are constrained. The differences observed in crosslinking were quite small as the central position in the peptide was varied. 12.2.5

Protein Fragments

In some cases, intact DNA-binding protein fragments with individual recognition properties were chosen for conjugation to the transition metal complexes. The aim was to construct a sequence-specific chimera in which the peptide recognizes its cognate DNA sequence. M.Y. Ogawa and his coworkers44 prepared a chimeric metallo-bZIP protein (Figure 12.5), (GBR-CC)Ru, in which [Ru(bpy)2(phen-IA)]2+ (phen-IA = N-

Figure 12.5 Computer generated model of the synthetic metalloprotein of the designed photoactive DNA binding peptide developed from the crystal structure of cFos-cJun bound to DNA. (Reprinted from Inorg. Chim. Acta, 300–302, Lasey R.C, Banerji S.S., Ogawa M.Y., Synthesis and characterization of a sequence-specific DNA-binding protein that contains ruthenium polypyridyl centres, 822–828. Copyright 2000, with permission from Elsevier.)

356


iodoacetyl-5-amino-1,10-phenanthroline) was attached to the coiled-coil (CC) dimerization region based on an artificial sequence developed by Hodges and coworkers (Figure 12.5).45 The DNA-binding domain of the protein was taken from the basic region of the GCN-4 transcription factor (GBR). The photoactive metal complex was attached to Cys29, the most heavily solvent-exposed position of the coiled-coil closest to the DNA-binding domain, and, as indicated by its emission spectrum, the ruthenium centre remained exposed to solvent when coupled to the bZIP peptide. Electrophoretic mobility shift assays showed that the chimeric metalloprotein retained the essential DNA recognition properties of the native GCN-4 transcriptional activator. Peptide–DNA complexes were formed with the AP1 and CRE sequences, known recognition sites of GCN-4, but not the divergent Sp1 sequence. Peptide titration studies indicated that the affinities of (GBR-CC)Ru for the AP1 and CRE sites were comparable. (GBR-CC)Ru did not produce observable photoinduced DNA damage, probably due to the separation distance between the ruthenium sites and the DNA bases.44 The DNA recognition properties of the a3-helix of the DNA-binding phage P22 repressor were investigated by attachment onto the metallointercalating [Rh(phi)2(phen′)]3+ (phi = 9,l0-phenanthrenequinone diimine; phen′ = 5(amidoglutaryl)-1,10-phenanthroline).46 Remarkably, a single glutamate at position 10 was found to be essential in directing DNA site-recognition to the sequence 5′-CCA-3′. It is noteworthy that conservative modifications of this glutamate, including aspartate substitution or derivatization to the glutamate methyl ester, abolished the 5′-CCA-3′ recognition. On the other hand, circular dichroism indicated significant a-helical content in the peptide, which depended upon the presence of the glutamate. The experimental evidence indicated that the glutamate may play a dual role, folding the metal–peptide complexes into a unique conformation and also interacting directly with DNA. However, the metal–peptide complexes did not preferentially target the putative operator sequence for protein P22R. This fact may arise because in the native protein, intraprotein interactions determine the orientation of the a3 recognition helix, whereas in the metal–peptide complexes, the peptide conformation may be directed instead by the metal centre and the solvent. Furthermore, moving the glutamate at position 10 in the sequence of the appended peptide to position 6 changed the sequence preference of the metallointercalator–peptide conjugate to 5′-ACA-3′, whereas replacing Glu6 with Arg resulted in a change in the apparent consensus sequence from 5′-ACA-3′ to 5′-(T/G)CA-3′.47 These results clearly illustrate that with a single amino-acid modification, the recognition sequence and the photocleavage characteristics of the metallointercalator–peptide conjugate can be perturbed. A similar work by N.Y. Sardesai and J.K. Barton focused on the DNA recognition properties of metal–peptide complexes containing peptides derived from the sequence of the a3 recognition helix of the phage 434 repressor protein (434R) appended to the metallointercalator [Rh(phi)2(phen′)]3+.48 A series of peptide sequences was designed, based upon 434R and criteria for maximizing peptide ahelicity. The three Gln residues were left unchanged because they make base-specific contacts with the 434R consensus sequence 5′-ACAA-3′. Therefore, a family of peptides aA, aB (aA–Q7R), aC (aA–Q3R), and aD (aA–Q3A) was formed, in which peptides were different from one another by a single amino acid at one of the three


357

Gln residues implicated in DNA recognition. The 5′-ACAA-3′ sites were cleaved strongly by all metal–peptide complexes. However, in contrast to 434R, they showed a slight 5′-ACGA-3′ > 5′-ACAA-3′ preference. Single-site mutants at Q7 or Q3 did not affect the relative specificity of the metal–peptide complexes, but changed the affinity. Two issues were considered to contribute in the reduction in specificity. The first issue involved the intrinsic site selectivity of the metallointercalator resulting in preference also for the sites 5′-ACA-3′ or 5′-ATG-3′. The second issue addressed the conformational flexibility of appended peptides resulting in decrease of the thermodynamic advantage to be gained from the ensemble of contacts available in the DNA-binding protein or even promote nonspecific interactions with the phosphate backbone, so that the major contribution to DNA site specificity would be derived instead from the metal complex.48 Another interesting study involved a peptide (7p) derived from the recognition loop of the restriction endonuclease MunI from Mycoplasma unidentified, which was appended to the groove-binding complex [Ru(bpy)2(m-bpy)]2+ (m-bpy = 4methyl-4′-carboxylate-2,2′-bipyridine).49 MunI recognizes the palindromic hexanucleotide sequence C/AATTG and cleaves as indicated by the slash. All amino acid residues involved in sequence-specific interactions lie within a single short region (Arg115-Gly116-Asn117-Ala118-His119-Glu120-Arg121 = 7p).50 A 1H NMR study of the interactions of the diastereomeric complexes L- and D-[Ru(bpy)2(m-bpy-7p)]2+ with the oligonucleotide sequences d(5′-CGCGATCGCG-3′)2 and d(5′-GCGCTTAAGCGC-3′)2 indicated binding of both isomers to the ends of both sequences. Furthermore, the L-isomer exhibited additional NOE contacts to the major groove protons of the GCT/CGA sequence of the decanucleotide and the central TT/AA part of the dodecanucleotide. No sequence-specific contacts were indicated by the experimental results, albeit intermolecular NOEs showed that the peptide was oriented towards the DNA major groove. Thus, the appended peptide did not reproduce the recognition characteristics of the endonuclease that it was derived from. This is another case in which the conformational flexibility of the peptide moiety did not allow formation of sequence-specific contacts, whereas in the case of d(5′GCGCTTAAGCGC-3′)2 the resultant conformers were bound to DNA via many different orientations.49

12.2.6

Summary

Tethering of a peptide to a transition metal complex, especially to a metallointercalator, increases the overall DNA-binding affinity. Many studies have shown that minimalist peptide domains are insufficient for tight binding and dimerization has been utilized in order to enhance affinity.7 N.Y. Sardesai et al. have demonstrated that a 100-fold excess of free peptide was incapable of competing with the corresponding rhodium metallointercalator-peptide conjugate in binding to the DNA target.46 On the other hand, the peptide contributes substantially to the DNAbinding affinity, resulting in tighter binding compared to the metal complex.48 Simple coordination metal complexes often lack hydrogen-bonding donors or acceptors, thus undergoing their site-specific reactions solely on the basis of shapes and symmetries.51 Thus, in most cases, the appended peptide contributed

358


significantly in the sequence specificity of the conjugates, being able to discriminate between different DNA sequences. Furthermore, the sequence selectivity was in many cases perturbed by a single amino acid modification,46,47 demonstrating the versatility of peptides as systems utilized for DNA recognition. However, in some cases there is a possibility that the intrinsic preference of the metal complex for a specific DNA sequence may interfere with sequence specificity conferred by the peptide moiety. Hence, the resulting sequence specificity may differ from that expected in cases where the peptide sequence is derived from a DNA-binding protein.48 Although, small peptides are composed of readily available building blocks and it is facile to tune the peptide composition to achieve the desired properties and variations, they lack the structural constraints and control that may be needed for predictive design. Indeed, their conformational flexibility accounts for the failure of recognizing the same sequence as the native protein they are derived from.48,49

12.3

Metallointercalator–Metallopeptide Conjugates

Design of tethered peptides has also been extended to mimicking the properties of certain proteins, such as DNA hydrolases, zinc-binding proteins and other important cell proteins which interact with DNA. Barton and coworkers tethered a peptide designed de novo, based on the active sites of metal-containing hydrolases to the metallointercalator [Rh(phi)2(bpy′)]3+.52 Two histidines were placed in positions 7 and 11 along a 16-residue peptide (P1) to create a zinc coordination site on one face of an a-helix. A glutamate was included at position 4, and glutamate–lysine salt bridges, as well as space-filling alanines were added to increase a-helicity (Figure 12.6). Both helical content and cleavage activity of plasmid DNA reached a maximum at stoichiometric amounts of Zn2+ using micromolar concentrations of the chimeric complex, owing to the strong binding affinity of the intercalating moiety. Furthermore, direct evidence was obtained for a hydrolytic cleavage reaction.52 Interestingly, the same peptide, when conjugated to ethidium, in the presence of Zn(II), forms a closed chelate structure which prevents binding to DNA and concomitant cleavage.53 The importance of several key residues of P1 was further assessed by determining the DNA cleavage efficiency of mutations in plasmid cleavage assays.54 Not surprisingly, the two histidine residues at positions 7 and 11 were found to be important for zinc coordination and cleavage activity. On the other hand, the cleavage rate constant was not affected upon mutation of the glutamic acid at position 4.54 Also, the selection of the metallointercalator seemed to be critical for the hydrolytic activity, since the attachment of P1 to [Rh(phen)(bpy)(chrysi)]3+, a metallointercalator which binds to DNA mismatches, did not produce Zn2+-promoted hydrolysis.54 In addition, zinc(II)-promoted cleavage of supercoiled plasmid has also been demonstrated for another metallointercalator–metallopeptide conjugate bearing a short b-hairpin derived from the BamHI restriction endonuclease, incorporating the three important catalytic residues (Asp94, Glu111 and Glu113) responsible for the hydro-

A Critical Survey and Future Perspectives phi

359

3+

NH HN

H N phi

Rh N H

N N

bpy H N Asp Pro Asp Glu Leu Glu His Ala Ala Lys His Glu Ala Ala Ala Lys CONH2 O HN

NH N

N Zn 2+

Figure 12.6 The Rh(III) complex-peptide conjugate, Rh(phi)2bpy′-Peptide (Rh-P1) with the Zn(II) coordination site being illustrated

lytic activity of the native enzyme.54 It must be emphasized that both appended peptides did not contribute to the recognition properties of the sequence-neutral metallointercalator [Rh(phi)2(bpy′)]3+ but efficiently delivered the metal ion to the sugar-phosphate backbone.

12.4 A Critical Survey and Future Perspectives Research on the development of artificial nucleases aims at high cleavage efficiency, DNA affinity and sequence selectivity. Attempting to compare different transition metal complex–peptide conjugates regarding their DNA cleavage efficiency becomes a rather difficult task due to the large variety of reaction conditions used. Therefore, comparisons can be made tentatively under the concept that in most cases, reaction conditions are optimized in order to give the best results. Furthermore, there is no available data sufficient for comparisons of kinetic parameters. Experimental data derived from the studies already described are summarized in Table 12.2. First of all, DNA binding by intercalation increases DNA binding affinity by at least two orders of magnitude compared to DNA groove binding. For example, for the series [Ru(bpy)3]2+, [Ru(bpy)2(phen)]2+, [Ru(bpy)2(DIP)]2+ and [Ru(bpy)2(phi)]2+, Kb values of 0.7 × l03, 0.7 × l03, 1.7 × l03, and 1.6 × l05 M−1, have been determined respectively.55 Given that increased DNA binding affinity results, in most cases, in high DNA cleavage efficiency, it is clearly seen from Table 12.2 (cleavage efficiency is compared as % formation of nicked and linear DNA from plasmid DNA

hydrolytic oxidative hydrolytic hydrolytic – hydrolytic hydrolytic hydrolytic oxidative oxidative oxidative oxidative

[Rh(phi)2(bpy′)]-P1 + Zn(II) 1:1 [Rh(phi)2(bpy′)]-P1 + Cu(II) 1:1 [Rh(phi)2(bpy′)]-P1 + Cd(II) 1:1 [Rh(phi)2(bpy′)]-P1 + Fe(II) 1:1 [Rh(phi)2(bpy′)]-Bamb [Rh(phi)2(bpy′)]-Bam + 100 eq Zn(II) [Rh(phi)2(bpy′)]-Bam + 100 eq Cd(II) [Rh(phi)2(bpy′)]-Bam + 100 eq Mg(II) [Rh(phi)2(bpy′)]-Bam + 1 eq Cu(II) [Rh(phi)2(bpy′)]-Bam + 1 eq Fe(II) [Rh(phi)2(phen′)]-aAc [Rh(phi)2(phen′)]-PAd 44 45 50 20 38 54 57 41 62 35

(18) (18) (18) (18) (26) (26) (26) (26) (26) (26) – –

65 (10) 40 (18)

% nicked plasmid DNAe

† Cleavage sites are underlined. * Samples have been irradiated. nd Sequence selectivity has not been determined. a P1 is a de novo designed peptide incorporating histidine residues for Zn2+ coordination. b Bam is a peptide derived from the active center of BamHI. c aA is a peptide derived from the recognition helix of phage 434 repressor protein. d PA is a peptide derived from the a3 helix of the phage P22 repressor. e The control experiment (plasmid only) is given in parenthesis.

oxidative –

[Ru(bpy)2(m-GHK)]2+ [Rh(phi)2(bpy′)]-P1a

Mechanism of cleavage

38 (0) – – – – 2 (0) 4 (0) – 3 (0) 2 (0) – –

– 4 (0)

% linear plasmid DNAe

5′-NACAA-3′* 5′-CCA-3′*

nd

nd

nd

nd

nd

nd

nd

nd

Mainly G* Modest selectivity for 5′-Pu-Py-Pu-3′* 5′-Py-Py-Pu-Pu-Pu-3′ 5′-Py-Py-Pu-Pu-Pu-3′

Sequence selectivity†

Table 12.2 Nuclease activity of transition metal complexes-peptide conjugates on super-coiled plasmid

54 54 54 54 54 54 54 54 54 54 48 46

24 54

Reference

360 DNA Recognition and Binding by Peptide–Metal Complex Conjugates

A Critical Survey and Future Perspectives

361

degradation) that the [Ru(phi)2(bpy′)]-P1 + Zn(II) 1:1 [54] (where P1 is a de novo designed a-helical peptide containing histidine residues for Zn2+ coordination) is the most efficient DNA cleaving metal complex with totally 82% formation of nicked plasmid and linear DNA (64% increase of modified plasmid DNA compared to control sample). On the other hand, high sequence selectivity is obtained in the case of appended peptides derived from DNA-binding proteins, especially when these peptides are either already structured (e.g. a3 helix of the phage P22 repressor46) or initially designed to be a-helical, with a-helicity further increased upon metal coordination (e.g. coordination of Zn(II) to appropriately designed peptides54). However, it is important to keep in mind that sequence selectivity and cleavage efficiency of intercalating metal complexes tethered to peptides can be further improved and optimized by the selection of the appropriate ancillary ligands, intercalating ligands, linker and appended peptides. Metal–peptide complexes show higher cleavage ability of DNA compared to metal complex–peptide conjugates. As seen from Table 12.2, it is obvious that 100% cleavage of plasmid DNA is not reported for any of the metal complex–peptide conjugates (82% is the highest percentage of cleavage observed in the case of [Rh(phi)2(bpy′)]-P1+Zn(II) 1:1). On the other hand, 100% conversion of supercoiled plasmid DNA (form I) to nicked circular DNA (form II) and linear DNA (form III) is reported for most metal-coordinated peptides.56–58 Furthermore, Ni(II) metallopeptides of the type Xaa–Xaa–His show remarkably high DNA-binding constants,56 comparable with those of intercalating metal complexes tethered to peptides (107–108 M−1).46,47 Since metal complex–peptide conjugates and Ni(II) metallopeptides bind to DNA with similar affinities, then the reason for the higher cleavage efficiency observed for Ni(II) metallopeptides probably lies in the different mechanisms of DNA cleavage. A possible explanation is the shape selectivity upon DNA binding of Ni(II) metallopeptides, reminiscent of the known minor-groove binding drugs, that enables them to fit into the DNA minor groove bringing the metal centre in close proximity to the deoxyribose moiety (cleavage takes place commonly by C4′-H abstraction). Even higher cleavage efficiency can be accomplished when metallopeptides are conjugated to an intercalating ligand, such as in the case of (bis-Ni(II)·GGH-NDI), in which two Ni(II)·GGH units are conjugated to napthalene diimide (NDI).59 The resultant nuclease is the most efficient reported so far among metallopeptides (100 times more efficient than Ni(II)·GGH).59 In addition, sequence selectivity and cleavage efficiency of metal–peptide complexes can conveniently be modulated either by changing the position of amino-acids bearing the appropriate side chain functionalities within the peptide sequence, thus discriminating between different mixed A/T sequences or changing amino acid chirality resulting in totally different sequence recognition (e.g. 5′-CCT-3′ in the case of Ni(II)·Gly-D-Asn-His56). Another important issue concerning the design of applicable artificial nucleases is the mechanistic pathway followed. Formation of diffusible active oxygen species results in base-directed reactivity, with guanines being the most vulnerable residues toward oxidation. Consequently, any sequence selectivity observed upon DNA binding cannot be reproduced after initiation of cleavage. This is obvious in the case of peptides conjugated to the parent complex [Ru(bpy)3]2+.24 Hence, formation of

362


diffusible radicals upon DNA cleavage is undesirable when designing an artificial nuclease bearing high sequence selectivity. Often, a hydrolytic rather than an oxidative cleavage mechanism is desirable for several reasons. Production of diffusible free radicals by oxidative cleavage of DNA results in strand ends that cannot be enzymatically religated in molecular biology applications. For clinical applications, oxidative cleavage can cause indiscriminate peripheral damage to the cell while radical diffusion may affect significantly the specificity of cleavage that can be achieved.15 Rh(III) metallointercalator– Zn(II) metallopeptide conjugates are the only chimeric molecules utilizing a hydrolytic mechanism of cleavage (Table 12.2). The pseudo first-order rate constant of [Rh(phi)2(bpy′)-P1+Zn(II) (1 × 10−5 s−1)54 determined for plasmid DNA degradation is of the same magnitude to the bimetallic Zn(II)2-heptapeptide developed by P. Scrimin et al.58 and to lanthanide metallopeptides.60 Therefore, conjugation of a Rh(III) metallointercalator to a hydrolytically active Zn(II) metallopeptide appears to be an effective system for the hydrolysis of DNA, displaying moderate sequence selectivity.54 The versatility of such systems could probably meet the requirements for obtaining the desired hydrolytic activity in the near future.

12.5

Conclusion

By careful design of transition metal complex–peptide conjugates, specific recognition of a DNA site may be achieved, which in some cases can be perturbed by a single amino acid modification. However, there are some limitations for the recognition of a DNA-binding protein cognate sequence by a peptide model of the native protein. In this case, issues of peptide flexibility need to be addressed. Indeed, highly structured appended peptides provide a higher degree of specificity for DNA recognition. DNA specific recognition is often accompanied by significant nucleolytic activity. Transition metal complex–peptide conjugates are not as efficient in DNA cleavage as metal-coordinated peptides. The reason probably lies in the different mechanism employed in each case. However, appropriately designed appended peptides may result in significant DNA hydrolytic activity by delivering Zn(II) ions to the DNA phosphate backbone. Importantly, metal complex–peptide conjugates are able to interact with the DNA major groove, commonly exploited upon DNA recognition by DNA-binding proteins. The versatility of these systems has become evident, thus further enhancement of their DNA recognition and cleavage properties may be achieved by appropriate selection of the ancillary ligands, the peptide composition and the metal ion.

Abbreviations AP1 bpy

Activator protein 2,2′-Bipyridine

References

bZip chrysi CRE CT-DNA DIP dppz en GGH GHK KAK KWK KYK m-bpy NDI NOE Phen-IA terpy

363

Basic leucine zipper Chrysenequinone diimine Cyclization recombination Calf thymus deoxyribose nucleic acid 4,7-Diphenyl-1,10-phenanthroline Dipyridophenazine Ethylene diamine Glycyl–glycyl–histidine Glycyl–histidyl–lysine Lysyl–alanyl–lysine Lysyl–tryptophanyl–lysine Lysyl–tyrosyl–lysine 4-Carboxy-4′-methyl-2,2′-bipyridine Napthalene diimide Nuclear Overhauser Effect N-iodoacetyl-5-amino-1,10-phenanthroline 2,2′:6′,2″-Terpyridine

References 1. (a) Dervan, P.B.; Molecular recognition of DNA by small molecules; Bioorg. Med. Chem., 2001, 9, 2215–2235. (b) Erkkila, K.E.; Odom, D.T.; Barton, J.K.; Recognition and reaction of metallointercalators with DNA; Chem. Rev., 1999, 99, 2777–2795. (c) Metcalfe, C.; Thomas, J.A.; Kinetically inert transition metal complexes that reversibly bind to DNA; Chem. Soc. Rev., 2003, 32, 215–224. 2. (a) Hegg, E.L.; Burstun, J.N.; Toward the development of metal-based synthetic nucleases and peptidases: a rationale and progress report in applying the principles of coordination chemistry; Coord. Chem. Rev., 1998, 173, 133–165. (b) Sigman, D.S.; Bruice, T. W.; Mazumder, A.; Sutton, C.L.; Targeted chemical nucleases; Acc. Chem. Res., 1993, 26, 98–104. 3. Pingoud, A.; Jeltsch, A.; Structure and function of type II restriction endonucleases; Nucleic Acid Res., 2001, 9, 3705–3727. 4. (a) Narayan, V.A.; Kriwacki, R.W.; Caradonna, J.P.; Structures of zinc finger domains from transcription factor Sp1. Insights into sequence-specific protein–DNA recognition; J. Biol. Chem., 1997, 272, 7801–7809. (b) Pavletich, N.P.; Pabo, C.O.; Zinc finger–DNA recognition: crystal structure of a Zif268-DNA complex at 2.1 Å; Science, 1991, 252, 809–817; (c) Elrod-Erickson, M.; Rould, M.A.; Nekludova, L.; Pabo, C.O.; Zif268 protein–DNA complex refined at 1.6 Å: a model system for understanding zinc finger– DNA interactions; Structure, 1996, 4, 1171–1180; (d) Nolte, R.T.; Conlin, R.M.; Harrison, S.C.; Brown, R.S.; Differing roles for zinc fingers in DNA recognition: Structure of a six-finger transcription factor IIIA complex; Proc. Natl. Acad. Sci. USA, 1998, 95, 2938–2943. 5. Neidle, S.; Principles of ligand–DNA recognition. In: DNA Structure and Recognition; Rickwood, D. and Male, D., Eds, Oxford University Press Inc.; New York, 1994. 6. (a) Ueno, M.; Murakami, A.; Makino, K.; Morri, T.; Arranging quaternary structure of peptides by cyclodextrin-guest inclusion complex: sequence-specific DNA binding by a peptide dimer with artificial dimerization module; J. Am. Chem. Soc., 1993, 115, 12575– 12576. (b) Ueno, M.; Sawada, M.; Makino, K.; Morii T.; Recognition of nonpalindromic DNA sequence by a peptide heterodimer with artificial dimerization module; J. Am.

364

7.

8.

9. 10. 11. 12. 13. 14. 15. 16.

17. 18. 19. 20.

21.

DNA Recognition and Binding by Peptide–Metal Complex Conjugates Chem. Soc., 1994, 116, 11137–11138. (c) Morii, T.; Yamane, J.; Aizawa, Y.; Makino, K.; Sugiura, Y.; Cooperative oligomerization enhances sequence-selective DNA binding by a short peptide; J. Am. Chem. Soc., 1996, 118, 10011–10017. (a) Talanian, R.V.; McKnight, C.J.; Kim, P.S.; Sequence-specific DNA binding by a short peptide dimer; Science, 1990, 249, 769–771. (b) Morii, T.; Simomura, M.; Morimoto, S.; Saito, J.; Sequence-specific DNA binding by a geometrically constrained peptide dimer; J. Am. Chem. Soc., 1993, 115, 1150–1151. (a) Benos, P.V.; Lapedes, A.S.; Stormo, G.D.; Is there a code for protein–DNA recognition? Probab (ilistical)ly …; BioEssays, 2002, 24, 466–475. (b) Suzuki, M.; Brenner, S.E.; Gerstein, M.; Yagi, N.; DNA recognition code of transcription factors; Prot. Eng., 1995, 8, 319–328. Suzuki, M.; Yagi, N.; DNA recognition code of transcription factors in the helix-turnhelix, probe helix, hormone receptor and zinc finger families; Proc. Natl Acad. Sci. USA, 1994, 91, 12357–12361. Damante, G; Pellizzari, L.; Esposito, G.; Fogolari, F.; Viglino, P.; Fabbro, D.; Tell, G.; Formisano, S.; Lauro, R.D.; A molecular code dictates sequence-specific DNA recognition by homeodomains; EMBO J., 1996, 15, 4992–5000. Benos, P.V.; Lapedes, A.S.; Stormo, G.D.; Probabilistic code for DNA recognition by proteins of the EGR family; J. Mol. Biol., 2002, 323, 701–727. Sera, T.; Uranga, C.; Rational design of artificial zinc-finger proteins using a nondegenerate recognition code table; Biochemistry, 2002, 41, 7074–7081. Sarai, A.; Kono, H.; Protein–DNA recognition patterns and predictions; Annu. Rev. Biophys. Biomol. Struct., 2005, 34, 379–398. (a) Long, E.C.; Ni(II)·Xaa-Xaa-His metallopeptide-DNA/RNA interactions; Acc. Chem. Res., 1999, 32, 827–836. (b) Mancin, F.; Scrimin, P.; Tecilla, P.; Tonellato, U.; Artificial metallonucleases; J. Chem. Soc. Chem. Commun., 2005, 2540–2548. (a) Franklin, S.J.; Lanthanide-mediated DNA hydrolysis; Curr. Opin. Chem. Biol., 2001, 5, 201–208. (b) Franklin, S.J.; Welch, J.T.; The helix-turn-helix as a scaffold for chimeric nuclease design; Comm. Inorg. Chem., 2005, 26, 127–164. (a) Beerli, R.R.; Barbas, III, C.F.; Engineering polydactyl zinc-finger transcription factors; Nat. Biotechnol., 2002, 20, 135–141 and refs therein. (b) Pabo, C.O.; Peisech, E.; Grant, R.A.; Design and selection of novel Cys2His2 zinc finger proteins; Ann. Rev. Biochem., 2001, 70, 313–340. (c) Jantz, D.; Amann, B.T.; Gatto, Jr, G.J.; Berg, J.M.; The design of functional DNA-binding proteins based on zinc finger domains; Chem. Rev., 2004, 104, 789–800. (d) Dhanasekaran, M.; Negi, S.; Sugiura, Y.; Designer zinc finger proteins: tools for creating artificial DNA-binding functional proteins; Acc. Chem. Res., 2006, 39, 45–52, (e) Durai, S.; Mani, M.; Kandarelou, K.; Wu, J.; Porteus, M.H.; Chandrasegaran, S.; Zinc finger nucleases: custom-designed molecular scissors for genome engineering of plant and mammalian cells; Nucl. Acids Res., 2005, 33, 5978–5990. Sigman, D.S.; Bruice, T.W.; Mazumder, A.; Sutton, C.; Targeted chemical nucleases; Acc. Chem. Res., 1993, 26, 98–104. Robillard, M.S.; Bacac, M.; Elst, H.; Flamigni, A.; van der Marel, G.A.; Van Boom, J.H.; Reedijk, J.; Automated parallel solid-phase synthesis and anticancer screening of a library of peptide-tethered platinum(II) complexes; J. Comb. Chem., 2003, 5, 821–825. Robillard, M.S.; Davies, N.P.; van der Marel, G.A.; Van Boom, J.H.; Reedijk, J.; Murray, V.; The interaction of peptide-tethered platinum(II) complexes with DNA; J. Inorg. Biochem., 2003, 96, 331–338. (a) Pickart, L.; Thaler, M.M.; Tripeptide in human serum which prolongs survival of normal liver cells and stimulates growth in neoplastic liver; Nature New Biol., 1973, 243, 85–87. (b) Schlesinger, D.; Pickart, L.; Thaler, M.M.; Growth modulating serum tripeptide is glycyl-histidyl-lysine; Experientia, 1977, 33, 324–325. (c) Pickart, L.; Thaler, M.M.; Millard, M.M.; Effect of transition metals on recovery from plasma of the growthmodulating tripeptide glycylhistidyllysine; J. Chromatogr., 1979, 175, 65–73. Freedman, J.H.; Pickart, L.; Weinstein, B.; Mims, W.B.; Peisach, J.; Structure of the glycylL-histidyl-L-lysine-copper(II) complex in solution; Biochemistry, 1982, 21, 4540–4544.

References

365

22. Laussac, J.P.; Haran, R.; Sarkar, B.; N.m.r. and e.p.r. investigation of the interaction of copper(II) and glycyl-L-histidyl-L-lysine, a growth-modulating tripeptide from plasma; Biochem. J., 1983, 209, 533–539. 23. Myari, A.; Hadjiliadis, N.; Garoufis, A.; Synthesis and characterization of the diastereomers L- and D-[Ru(bpy)2(m-bpy-Gly-L-His-L-Lys)]Cl2 – 1H NMR studies on their interactions with the deoxynucleotide duplex d[(5′-CGCGAATTCGCG-3′)2]; Eur. J. Inorg. Chem., 2004, 1427–1439. 24. Myari, A.; Hadjiliadis, N.; Garoufis, A.; Malina, J.; Brabec, V.; NMR analysis of duplex d(CGCGATCGCG)2 modified by L- and D-[Ru(bpy)2(m-GHK)]Cl2 and DNA photocleavage study; J. Biol. Inorg. Chem., 2006, 12, 279–292. 25. Yamagishi, A.; Evidence for stereospecific binding of tris(1,10-phenanthroline)ruthenium(II) to DNA is provided by electronic dichroism; J. Chem. Soc. Chem. Commun., 1983, 572–573. 26. Yamagishi, A.; Electric dichroism evidence for stereospecific binding of optically active tris chelated complexes to DNA; J. Phys. Chem., 1984, 88, 5709–5713. 27. Sigman, D.S.; Mazumder, A.; Perrin, D.M.; Chemical nucleases; Chem. Rev., 1993, 93, 2295–2316. 28. Novakova, O.; Kasparkova, J.; Vrana, O.; Vanvliet, P.M.; Reedijk, J.; Brabec, V.; Correlation between cytotoxicity and DNA binding of polypyridyl ruthenium complexes; Biochemistry, 1995, 34, 12369–12378. 29. Karidi, K.; Garoufis, A.; Hadjiliadis, N.; Reedijk, J.; Solid-phase synthesis, characterization and DNA binding properties of the first chloro(polypyridyl)ruthenium conjugated peptide complex; J. Chem. Soc., Dalton Trans., 2005, 728–734. 30. Lopes, L.G.F.; Wieraszko, A.; El-Sherif, Y.; Clarke, M.J.; The trans-labilization of nitric oxide in RuII complexes by C-bound imidazoles; Inorg. Chim. Acta, 2001, 312, 15–22. 31. Lang, D.R.; Davis, J.A.; Lopes, L.G.F.; Ferro, A.A.; Vasconcellos, L.C.G.; Franco, D.W.; Tfouni, E.; Wieraszko, A.; Clarke, M.J.; A controlled NO-releasing compound: synthesis, molecular structure, spectroscopy, electrochemistry, and chemical reactivity of R,R,S,Strans-[RuCl(NO)(cyclam)]2+(1,4,8,11-tetraazacyclotetradecane); Inorg. Chem., 2000, 39, 2294–2300. 32. Karidi, K.; Garoufis, A.; Tsipis, A.; Hadjiliadis, N.; den Dulk, H.; Reedijk, J.; Synthesis, characterization, in vitro antitumor activity, DNA-binding properties and electronic structure (DFT) of the new complex cis-(Cl,Cl)[RuIICl2(NO+)(terpy)]Cl; J. Chem. Soc., Dalton Trans., 2005, 1176–1187. 33. Slocik, J.M.; Shepherd, R.E.; Coordination of Ru(NO)Cl3 to the tripeptides gly–gly–gly and gly–gly–his: N-terminal amine–amide and C-terminal imidazole–amide functionalities in bidentate chelation; Inorg. Chim. Acta, 2000, 311, 80–94. 34. Patra, A.K.; Mascharak, P.K.; A ruthenium nitrosyl that rapidly delivers NO to proteins in aqueous solution upon short exposure to UV light; Inorg. Chem., 2003, 42, 7363– 7365. 35. Bordini, J.; Hughes, D.L.; Neto, J.D.D.; Da Cunha, C.J.; Nitric oxide photorelease from ruthenium salen complexes in aqueous and organic solutions; Inorg. Chem., 2002, 41, 5410–5416. 36. Karidi, K.; Garoufis, A.; Hadjiliadis, N.; Lutz, M.; Spek, A.L.; Reedijk, J.; Synthesis, characterization, and DNA-binding studies of nitro(oligopyridine)ruthenium(II) complexes; Inorg. Chem., 2006, 45, 10282–10292. 37. Cardo, L.; Hannon, M.J.; Design and DNA-binding of metallo-supramolecular cylinders conjugated to peptides; Inorg. Chim. Acta, 2008, doi; 10.1016/j.ica.2008.02.050. 38. (a) Hannon, M.J.; Moreno, V.; Prieto, M.J.; Molderheim, E.; Sletten, E.; Meistermann, I.; Isaac, C.J.; Sanders, K.J.; Rodger, A.; Intramolecular DNA coiling mediated by a metallo supramolecular cylinder; Angew. Chem., Intl. Ed., 2001, 40, 879–884. (b) Uerpmann, C.; Malina, J.; Pascu, M.; Clarkson, G.J.; Moreno, V.; Rodger, A.; Grandas, A.; Hannon, M. J.; Design and DNA binding of an extended triple-stranded metallo-supramolecular cylinder; Chem., Eur. J., 2005, 11, 1750–1756. (c) Meistermann, I.; Moreno, V.; Prieto, M.J.; Moldrheim, E.; Sletten, E.; Khalid, S.; Rodger, P.M.; Peberdy, J.C.; Isaac, C.J.; Rodger, A.,

366

39.

40. 41.

42. 43. 44. 45. 46. 47. 48. 49.

50. 51. 52. 53. 54.

DNA Recognition and Binding by Peptide–Metal Complex Conjugates Hannon, M.J.; Intramolecular DNA coiling mediated by metallo-supramolecular cylinders: differential binding of P and M helical enantiomers; Proc. Natl. Acad. Sci. USA., 2002, 99, 5069–5074. (a) Cerasino, L.; Hannon, M.J.; Sletten, E.; DNA three-way junction with a dinuclear iron(II) supramolecular helicate at the center: A NMR structural Study; Inorg. Chem., 2007, 46, 6245–6251. (b) Malina, J.; Hannon, M.J.; Brabec, V.; Recognition of DNA threeway junctions by metallosupramolecular cylinders: gel electrophoresis studies; Chem. Eur. J., 2007, 13, 3871–3877. (c) Childs, L.J.; Malina, J.; Rolfsnes, B.E.; Pascu, M.; Prieto, M.J.; Broome, M.J.; Rodger, P.M.; Sletten, E.; Moreno, V.; Rodger, A.; Hannon, M.J.; A DNA-binding copper(I) metallo-supramolecular cylinder that can act as an artificial nuclease; Chem. Eur. J., 2006, 12, 4919–4927. Brunner, J.; Barton, J.K.; Targeting DNA mismatches with rhodium intercalators functionalized with a cell-penetrating peptide; Biochemistry, 2006, 45, 12295–12302. Wender, P.A.; Mitchell, D.J.; Pattabiraman, K.; Pelkey, E.T.; Steinman, L.; Rothbard, J.B.; The design, synthesis, and evaluation of molecules that enable or enhance cellular uptake: Peptoid molecular transporters; Proc. Natl. Acad. Sci. USA, 2000, 97, 13003–13008. Jackson, B.A.; Alekseyev, V.Y.; Barton, J.K.; A versatile mismatch recognition agent: specific cleavage of a plasmid DNA at a single base mispair; Biochemistry, 1999, 38, 4655–4662. Copeland, K.D.; Lueras, A.M.K.; Stemp, E.D.A.; Barton, J.K. DNA cross-linking with metallointercalator-peptide conjugates; Biochemistry, 2002, 41, 12785–12797. Lasey, R.C.; Banerji, S.S.; Ogawa, M.Y.; Synthesis and characterization of a sequencespecific DNA-binding protein that contains ruthenium polypyridyl centers; Inorg. Chim. Acta, 2000, 300–302, 822–828. Hodges, R.S.; De novo design of alpha-helical proteins: basic research to medical applications; Biochem. Cell Biol., 1996, 74, 133–154. Sardesai, N.Y.; Zimmermann, K.; Barton, J.K.; DNA recognition by peptide complexes of rhodium(III): Example of a glutamate switch; J. Am. Chem. Soc., 1994, 116, 7502–7508. Hastings, C.A.; Barton, J.K.; Perturbing the DNA sequence selectivity of metallointercalator-peptide conjugates by single amino acid modification; Biochemistry, 1999, 38, 10042–10051. Sardesai, N.Y.; Barton, J.K.; DNA recognition by metal-peptide complexes containing the recognition helix of the phage 434 repressor; J. Biol. Inorg. Chem., 1997, 2, 762– 771. Myari, A.; Hadjiliadis, N.; Garoufis, A.; Synthesis and characterization of the diastereomers L- and D-[Ru(bpy)2(m-bpy-L-Arg-Gly-L-Asn-L-Ala-L-His-L-Glu-LArg)]Cl2.1H NMR studies on their interactions with the deoxynucleotide duplex d[(5′GCGCTTAAGCGC-3′)2] and d[(5′-CGCGATCGCG-3′)2]; J. Inorg. Biochem., 2005, 99, 616–626. Deibert, M.; Grazulis, S.; Janulaitis, A.; Siksnys, V.; Huber, R.; Crystal structure of MunI restriction endonuclease in complex with cognate DNA at 1.7 Å resolution; EMBO J., 1999, 18, 5805–5816. Barton, J.K.; Metals and DNA: Molecular left-handed complements; Science, 1986, 233, 727–734. Fitzsimons, M.P.; Barton, J.K.; Design of a synthetic nuclease: DNA hydrolysis by a zinc-binding peptide tethered to a rhodium intercalator; J. Am. Chem. Soc., 1997, 119, 3379–3380. Houser, R.P.; Fitzsimons, M.P.; Barton, J.K.; Metal-dependent intramolecular chiral induction: The Zn2+ complex of an ethidium-peptide conjugate; Inorg. Chem., 1999, 38, 1368–1370. Copeland, K.D.; Fitzsimons, M.P.; Houser, R.P.; Barton, J.K.; DNA hydrolysis and Oxidative Cleavage by Metal-Binding Peptides Tethered to Rhodium intercalators; Biochemistry, 2002, 41, 343–356.

References

367

55. Pyle, A.M.; Rehmann, J.P.; Meshoyrer, R.; Kumar, C.V.; Turro, N.J.; Barton, J.K.; Mixedligand complexes of ruthenium(II): Factors governing binding to DNA; J. Am. Chem. Soc., 1989, 111, 3051–3058. 56. (a) Liang, Q.; Eason, P.D.; Long, E.C.; Metallopeptide-DNA interactions: Site-selectivity based on amino acid composition and chirality; J. Am. Chem. Soc., 1995, 117, 9625–9631. (b) Huang, X.; Pieczko, M.E.; Long, E.C.; Combinatorial optimization of the DNA cleaving Ni(II)·Xaa-Xaa-His metallotripeptide domain; Biochemistry, 1999, 38, 2160–2166. (c) Fang, Y.-Y.; Ray, B.D.; Claussen, C.A.; Lipkowitz, K.B.; Long, E.C.; Ni(II)·Arg-Gly-HisDNA interactions: Investigation into the basis for minor-groove binding and recognition; J. Am. Chem. Soc., 2004, 126, 5403–5412. (d) Fang, Y.-Y.; Claussen, C.A.; Lipkowitz, K.B.; Long, E.C.; Diastereoselective DNA cleavage recognition by Ni(II)·Gly-Gly-Hisderived metallopeptides; J. Am. Chem. Soc., 2006, 128, 3198–3207. (e) Jin, Y.; Lewis, M.A.; Gokhale, N.H.; Long, E.C.; Cowan, J.A.; Influence of stereochemistry and redox potentials on the single- and double-strand DNA cleavage efficiency of Cu(II) and Ni(II)·LysGly-His-derived ATCUN metallopeptides; J. Am. Chem. Soc., 2007, 129, 8353–8361. 57. (a) Ananias, D.C.; Long E.C.; DNA strand scission by dioxygen + light-activated cobalt metallopeptides; Inorg. Chem., 1997, 36, 2469–2471. (b) Ananias, D.C.; Long E.C.; Highly selective DNA modification by ambient O2-activated Co(II)·Lys-Gly-His metallopeptides; J. Am. Chem. Soc., 2000, 122, 10460–10461. 58. Sissi, C.; Rossi, P.; Felluga, F.; Formaggio, F.; Palumbo, M.; Tecilla, P.; Toniolo, C.; Scrimin, P.; Dinuclear Zn2+ complexes of synthetic heptapeptides as artificial nucleases; J. Am. Chem. Soc., 2001, 123, 3169–3170. 59. Shullenberger, D.F.; Eason, P.D.; Long, E.C.; Design and synthesis of a versatile DNAcleaving metallopeptide structural domain; J. Am. Chem. Soc., 1993, 115, 11038–11039. 60. (a) Welch, J.T.; Sirish, M.; Lindstrom, K.M.; Franklin, S.J.; De novo nucleases based on HTH and EF-hand chimeras; Inorg. Chem., 2001, 40, 1982–1984. (b) Sirish, M.; Franklin, S.J.; Hydrolytically active Eu(III) and Ce(IV) EF-hand peptides; J. Inorg. Biochem., 2002, 91, 253–258.

13 Artificial Restriction Agents: Hydrolytic Agents for DNA Cleavage Fabrizio Mancin and Paolo Tecilla

13.1 Introduction The inertness of DNA toward hydrolytic cleavage is tremendous. Indeed, the estimated half-life times for a single P–O bond cleavage at 25 °C and pH 7 range from hundreds of thousands to hundreds of millions of years.1 Such hydrolytic resistance contributes to the preservation of the genetic information, but also represents a serious obstacle, since several biological processes involving DNA manipulation, e.g. expression and duplication, repair of damages and elimination of foreign DNA, require hydrolytic reactions. The solution of this problem is provided by a wealth of hydrolytic enzymes, nucleases and topoisomerases, which efficiently catalyse DNA scission. Many such enzymes contain metal ions in their active sites, mainly Ca(II), Mg(II) and Zn(II), that play a fundamental role in their catalytic action.1b,2 As a matter of fact, the ability of metal ions to promote the hydrolytic cleavage of esters has been well known by chemists for decades. However, the development of synthetic agents capable to accelerate the hydrolysis of DNA has attracted substantial research efforts only in the last 20 years.2a,3 There are several reasons that justify the interest in such systems. First, the mechanistic information obtained in the study of artificial agents could lead to a better understanding of the chemistry of the corresponding hydrolytic enzymes. Second, they could be employed for detoxification of pesticides and chemical weapons, which often have phosphate Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

370

Artificial Restriction Agents

ester-like structures. Third, efficient hydrolytic agents for phosphate diesters could be employed as artificial restriction enzymes for molecular biology. Such systems would be highly valuable because they could, in principle, be designed to cleave DNA with sequence selectivity different from that of the natural enzymes. The latter, in fact, recognize DNA sequences of only 4, 6 or 8 bases, and such selectivity may be inadequate for the precise control of the manipulation of large DNA of higher species. Finally, the realization of antiDNA drugs can be envisaged in a more distant future. Since the first example of an artificial hydrolitic agent for DNA cleavage, reported by J. K. Barton in 1987,4 several promising systems have been proposed up to the description of the first studies on DNA manipulation with artificial agents, reported in very recent years by Komiyama and coworkers.5 This chapter highlights the progress reported during these two decades toward obtaining synthetic nucleases. Particular attention will be devoted to the strategies that can be pursued to increase the efficiency and the sequence selectivity of the systems.

13.2 DNA Hydrolysis Information available on the mechanism of spontaneous hydrolysis of DNA is obviously scarce due to its virtually absent reactivity. The commonly accepted pathway for the hydrolysis of deoxynucleotide phosphates involves the nucleophilic attack of a water (or hydroxide) oxygen at the phosphorus to give a five-coordinate phosphate intermediate or transition state (Figure 13.1).3o,p,6 Subsequent cleavage of either the P-O3′ or P-O5′ (usually P-O3′ cleavage in the enzymatic process) causes a strand scission, yielding the R–OH and R–O–PO3(H2) termini. As mentioned before, such a reaction is exceedingly slow in physiological conditions (i.e, aqueous solution at 25 °C and pH 7) and reaction rate constants of about 1 × 10−15 s−1 have

Figure 13.1 Reaction pathway for DNA hydrolysis. The enzymatically promoted P-03′ scission is shown

DNA Hydrolysis

371

been estimated.1a As a consequence, spontaneous DNA degradation occurs mainly through different pathways, involving either C–O cleavage, nucleobase ring opening or radical attack (Figure 13.2).1a Metal ions can assist the P–O hydrolytic cleavage of phosphodiesters in different ways (Figure 13.3). Acting as Lewis acids, they can activate the phosphate group toward the attack of the nucleophile, increase the leaving group ability of the departing alcohol and activate a metal-coordinated water molecule as a nucleophile.1b Moreover, indirect activation modes are possible, with metal bound hydroxides or water molecules performing general base or acid catalysis. As observed in many enzymatic processes, several metal centers may act cooperatively to accelerate these mechanistic steps. Accordingly, multinuclear complexes are often more active than their mononuclear counterparts.1b The exceptional high resistance of DNA toward hydrolysis makes even a simple kinetic investigation of the cleavage process quite challenging. However, supercoiled plasmid DNA is a more accessible substrate and, because of this, very popular

Figure 13.2 Alternative DNA degradation pathways

Figure 13.3 Possible activation modes provided by metal ions for acceleration of the hydrolysis of phosphate diesters

372


Figure 13.4 The three forms of plasmid DNA

for such studies.3r This particular form of DNA, which is commonly found in bacteria, is a cyclic supercoiled double strand made by several thousands of base pairs (Figure 13.4). One single-strand scission unravels the supercoiled DNA (form I) to a relaxed circular one (nicked, form II), while a second scission on the complementary strand, within about twelve base pairs from the first one, generates a linear DNA form (form III). These three DNA forms can be easily separated and quantified by gel electrophoresis, thus allowing a simple and rapid kinetic analysis. Moreover, supercoiled DNA is somehow more reactive than a short linear DNA fragment and this makes the study of the reaction easier, even in the presence of relatively poor catalysts. Such an increased reactivity is due: (a) to the obvious reason that the cleavage of a single phosphodiester bond in a large DNA molecule is statistically more likely than in a smaller DNA fragment and (b) to the internal strain resulting in ground-state destabilization.3r Finally, two other peculiar and relevant features of plasmid DNA are: (i) the possibility to discriminate between a single-strand or a double-strand cleavage simply by a statistical analysis of the relative abundance of forms II and III produced in the reaction; (ii) the possibility to enzymatically religate form III or form II giving both a closed circular form and concatemer DNA which can be easily detected by gel electrophoresis.3r Religation experiments are important in the mechanistic investigation of DNA cleavage because only fragments that derive from a hydrolytic process can be religated by the enzyme. Therefore, the success or the failure of this experiment may give important support to or seriously question the occurrence of a clean hydrolytic process. A careful mechanistic investigation becomes extremely important when metal ions with known redox chemistry are used for the development of artificial nucleases. For example, it is well known that Cu(II) and Fe(III) complexes, in the proper conditions, cleave DNA by an oxidative pathway.7 This process is so efficient that it always competes with the hydrolytic cleavage. Because the reactivity of oxidative agents depends usually on cofactors such as H2O2, molecular oxygen and reducing agents, and may produce diffusible radicals, the absence of externally added cofactors, the insensitivity to radical scavengers and the preservation of reactivity under anaerobic reactivity are generally considered sufficient evidence of a hydrolytic

Free Ions and Mononuclear Complexes 373

Figure 13.5 The [Zn2(pyrimol)2]2+ complex: the first Zn-based artificial oxidative nuclease

reaction. However, oxidative agents that cleave DNA through nondiffusible radicals have been known for a long time and, more recently, several complexes able to perform oxidative cleavage in the absence of coreactants and in anaerobic conditions have been reported.8 Remarkably, in a very recent example by Reedijk and coworkers,9 such behaviour has been described for the first time, even in the case of Zn(II) complexes, a metal ion so far considered only capable of promoting hydrolytic chemistry. This peculiar reactivity is probably related to the pirymol ligand employed (Figure 13.5), but highlights the importance of a carefully performed mechanistic investigation and indicates that any claim of hydrolytic reactivity, particularly in the case of Cu(II) or iron complexes, not supported by clear-cut evidence, such as enzymatic religation or fragments identification, should be considered with caution.

13.3 Free Ions and Mononuclear Complexes At the beginning of the last decade it was demonstrated that aqueous trivalent lanthanide ions accelerate the hydrolysis of DNA.10 Subsequent papers by Schneider and coworkers reported a systematic investigation of such ions aimed at determining the relevant kinetic parameters for the nicking of plasmid DNA.11 The different ions show similar maximum reactivities under saturation conditions (kmax 0.7–1.0 × 10−4 s−1, 37 °C, pH 7.0), but different affinities for DNA. The latter increase along the series from La(III) (Ka = 5.0 × 102 M−1) to Er(III) (Ka = 1.7 × 104 M−1). The combination of moderate reactivity with low affinity for the substrate is such that substantial degradation of DNA can be obtained only at high metal ion concentrations. In contrast, late lanthanides, such as Tm(III), Yb(III) and Lu(III), display a remarkable reactivity, even at low concentrations, due to their high affinities for DNA. However, a decrease of the cleavage rate is observed at high metal ion concentrations, probably as a consequence of their aggregation, with formation of less active species. Among lanthanides, cerium is peculiar in its ability to reach a stable tetravalent oxidation state. Ce(IV) is much more efficient in hydrolysing DNA than the

374


trivalent lanthanide ions. At pH 7.0 and 50 °C, the half-life of the dinucleotide TpT is reduced to 3.6 hours, which amounts to more than 11 orders of magnitude rate acceleration over the spontaneous hydrolysis reaction.10a,12 As a result, Ce(IV) is 20 to 1000 times more efficient than any other trivalent lanthanide, with comparable high efficacy toward single- and double-stranded linear DNA.13 The source of this remarkable reactivity has been attributed by Komiyama and coworkers to the great electron-withdrawing ability of this tetravalent ion.3m Surprisingly, free Ce(IV) ions are inactive in promoting plasmid DNA hydrolysis, probably due to steric hindrance between the Ce(IV) hydroxide gels, which form at neutral pH, and the substrate.14 In fact, when converted into homogeneous solutions by the addition of solubilizing agents (dextran, polyvinylpirrolidone), Ce(IV) recovers activity toward plasmid DNA and is about ten times more reactive than trivalent lanthanide ions.15 Aqueous lanthanide ions are toxic for biological systems and, more generally, free metal ions are not suitable candidates for preparing appealing catalysts. The formation of stable complexes with proper ligands is therefore mandatory. However, this task turns out to be quite difficult in the case of lanthanide ions for two reasons: (i) they undergo very facile ligand exchange; (ii) the catalytic activity requires an unsaturated coordination sphere to allow interaction with water (nucleophile) and the substrate. Stable complexes can be obtained by using polyaminocarboxylate derivatives, crown ethers or azacrown ligands.3i In many cases, however, the formation of the complex results in a lower reactivity than that observed with the free metal ions, particularly in the case of polyaminocarboxylate complexes, due to the decrease in the overall charge of the complex and to the saturation of the metal ion coordination sphere. Neutral ligands, such as azacrowns or crown ethers, are thus more suitable, in order to retain the reactivity of the free metal ion, although their coordination strength is not so high and the complexing agent must be used in large excess to ensure complete formation of soluble complexes.16 In the case of Ce(IV), the reactivity decrease brought about by the use of anionic ligands leads to an interesting substrate selectivity. In fact, the Ce(IV)EDTA complex is able to cleave single-stranded DNA (ssDNA), but not doublestranded DNA (dsDNA).17 The reaction is slow, but highly selective and, as a consequence, gap sites or bulges present in a dsDNA that expose short sequences of ssDNA are cleaved without touching the rest of the macromolecule.18 Such behaviour is the basis of the ARCUT system proposed by Komiyama and coworkers,5 which will be described in detail later. Metal ions other than lanthanides are scarcely active in promoting DNA hydrolysis. Nevertheless, a few mononuclear metal complexes have been reported to act as artificial nucleases. The Co(III)-TAMEN complex (Figure 13.6) cleaves plasmid DNA, at 37 °C and pH 7.6, with a first-order rate constant of 5 × 10−5 s−1.19 Although the concentration of complex required, 1mM, is relatively high, suggesting a poor interaction with the substrate, the mechanism is fully hydrolytic, as the nicked DNA can be religated. This complex is also active toward single-stranded DNA. Zn(II) is one of the metal ions most frequently found in hydrolytic metalloenzymes, but the activities reported for simple mononuclear complexes are usually low. One remarkable exception is the Zn(II)-BHTDE complex (BHDTE = N,N′-bis(benzylhistidyl)diethylenetriamine, Figure 13.7) reported by Ichikawa and

Free Ions and Mononuclear Complexes 375

Figure 13.6 The TAMEN ligand

Figure 13.7 [Zn(H2BHTDE)OH]3+ and [Zn(quercetin)2] complexes

coworkers,20 which is quite reactive toward plasmid DNA: at pH 7, 50 °C and 3 mM complex concentration, the supercoiled form of plasmid DNA is completely cleaved after 24 h, leading to both nicked and linear products. From this data a rate constant of 4 × 10−5 s−1 can be roughly estimated. The cleavage efficiency shows a maximum at pH 6.5–7.0, suggesting the possibility that the protonated primary amino groups play a role in increasing DNA affinity. Another interesting example is the [Zn(quercetin)2] complex,21a (Figure 13.7), which at pH 7.2, 37 °C and 100 mM complex concentration cleaves supercoiled plasmid DNA with a rate constant of 1.7 × 10−4 s−1, which corresponds to a half-life of about 1 hour. Linearized DNA could be religated with the enzyme T4 ligase, ruling out the possibility of oxidative mechanisms involving phenoxyl radicals and confirming a pure hydrolytic cleavage. Remarkably, the similar Mn(II)-quercetin complex also shows hydrolytic activity very close to that of the Zn(II) complex (k = 1.3 × 10−4 s−1, at pH 7.2, 37 °C and 100 mM complex concentration).21b Because of their redox properties, Cu(II) complexes have been frequently used in the development of agents for the oxidative cleavage of DNA,7 but there are also several examples of Cu(II) complexes reported to cleave DNA with a hydrolytic mechanism. Burstyn and coworkers published the first example of a Cu(II)-based hydrolytic agent in 1996.22a It was shown that at a 25 mM concentration the Cu(II)-TACN complex (Figure 13.8) cleaves supercoiled DNA at 50 °C and pH 7.8 with an

376


Figure 13.8 Structures of some polyamine ligands used for the formation of mononuclear Cu(II) complexes active in DNA cleavage

estimated rate constant of ca. 1.5 × 10−5 s−1. A decrease of reactivity was observed at higher complex concentrations probably due to the formation of unreactive mhydroxo dimers.22b The complex is also active in the degradation of single-stranded DNA. However, anaerobic experiments showed that the activity in the absence of oxygen is reduced by 30%, indicating at least the simultaneous occurrence of both hydrolytic cleavage and oxidative cleavage. Later on, Fujii and coworkers reported that the Cu(II)-TACH complex (Figure 13.8) promotes the oxygen-independent cleavage of plasmid DNA.23a Saturation kinetics were observed with a kmax value of 1.2 × 10−3 s−1 and a Ka value of 1.3 × 104 M−1 at 35 °C and pH 8.1. Such reactivity is among the highest reported so far and corresponds to a half-life time for the supercoiled form of 20 minutes in the presence of 75 mM Cu(II) complex. Several other complexes of macrocyclic and linear triamine ligands were tested and found to be scarcely reactive. On the basis of these results and of the relatively good affinity of the Cu(II)-TACH complex for DNA, the authors suggest the possibility of specific binding of the catalyst to the substrate. A detailed analysis of complex speciation and pH reactivity profiles indicate that the monohydroxo complex is the active species at pH values above neutrality, as observed in the hydrolytic cleavage of model phosphate diesters.23b However, linearized plasmid DNA could not be enzymatically religated.23b Similar results have also been obtained for the related Cu(II)-TACI complex (Figure 13.8), which again shows an oxygen-independent activity (k = 2.3 × 10−3 s−1 at 48 mM complex concentration), but fails the enzymatic religation test.24 A pure hydrolytic mechanism, demonstrated both by oxygen-independent activity and enzymatic religation, characterizes the activity of the Cu(II) complexes of natural aminoglycosides, such as neamine (Figure 13.8) studied by Cowan and coworkers.25 With the Cu(II)-neamine complex, saturation kinetics were observed with a kmax value of 5.2 × 10−4 s−1 and a Ka value of 2.4 × 105 M−1 at 37 °C and pH 7.3. The very high affinity of the complex for DNA is attributed to the tight binding of the positively charged aminoglycoside ligand to DNA; this apparently makes the system very effective even at low concentrations, reducing the half-life time for the

Bimetallic Complexes

377

supercoiled form, in the presence of 25 mM complex concentration, to only 26 minutes. In the last few years several other reports on copper complexes able to cleave DNA have appeared in the literature.26 In most cases, the authors suggest a hydrolytic mechanism, basing their proposal only on the observation of reactivity in anaerobic conditions and on insensitivity to radical scavengers. However, although these behaviours are compatible with a hydrolytic mechanism, they have also been reported in the case of oxidative systems (see Section 13.2) and, therefore, they cannot be taken as clear-cut mechanistic evidence. Particularly when a metal ion such as Cu(II) is involved, the elucidation of the mechanism requires more stringent evidence, such as, for example, enzymatic religation or chemical identification of the cleaved DNA fragments. In the absence of such experiments the final decision on the true nature of the DNA cleavage remains uncertain.

13.4 Bimetallic Complexes The extraordinary catalytic efficiency of natural metallonucleases most often relies on the cooperative action of two or more metal ions. Available X-ray structures of these proteins indicate that most of these metal ions are placed within a narrow range of distances from each other.2 Accordingly, in a synthetic catalyst a precise spatial localization of the ions appears to be mandatory to ensure multiple interactions with the substrate and, consequently, to take advantage simultaneously of all the different activation modes that metal ions can provide for the hydrolysis of phosphate esters (substrate, leaving group and nucleophile activation).1b On the basis of these guidelines, a series of bimetallic complexes have been synthesized and investigated as catalysts for the hydrolytic cleavage of DNA. In 1996, Schneider and coworkers studied a 30-membered azacrown macrocycle (Figure 13.9a) that can bind two Eu(III) or Pr(III) ions.27 The Pr(III) complex was able to promote plasmid DNA nicking with a kmax = 2.8 × 10−4 s−1 and Ka = 3.0 × 103 M−1 at 37 °C and pH 7.0. However, the activity of the binuclear complex was only twofold that of the free metal ion. More effective are the Er(III)2 complexes of a Schiff-base-containing macrocycle (Figure 13.9b), proposed by Zhu and coworkers, which degrades supercoiled DNA at 37 °C, pH 7.0 with a maximum rate constant of 1.0 × 10−3 s−1.28 Unfortunately

Figure 13.9 Macrocyclic ligands capable of binding two lanthanide ions reported by Schneider (a) and Zhu (b)

378


Figure 13.10 The HXTA ligand

such high reactivity is counterbalanced by a low affinity for the substrate (Ka = 1.0 × 103 M−1) so that high concentrations of complex are required to obtain high reaction rates. A very efficient dicerium complex of the polyaminocarboxylate ligand HXTA (Figure 13.10) has been reported by Que and coworkers.29 At a concentration of 10 mM, pH 8 and 37 °C, the Ce2-HXTA complex cleaves plasmid DNA with a rate constant of 1.4 × 10−4 s−1, which corresponds to a half-life of 1.4 hours for the nicking process. Interestingly, double-strand scission is preferred to single-strand scission and the complex also cleaves linear double-stranded DNA with high regioselectivity (more than 90%) favouring the scission of the P-O3′ bond. Fe(III) is present in the active site of some phosphatases, but it has been rarely employed in artificial nucleases. Interestingly, the few Fe(III) complexes reported in the literature with hydrolytic activity in DNA cleavage are all binuclear. The first example, described by Schnaith, Que and coworkers in 1994,30 is based on a tetrabenzimidazolylmethyl derivative of 1,3-diamino-2-hydroxypropane (Figure 13.11, HPTB) and shows a surprising reactivity: the Fe(III)2-HPTB complex requires ‘oxidative’ conditions, i.e. the presence of H2O2 or O2 and a reductant (dithiothreitol or ascorbate), but yields DNA fragments that are consistent with hydrolytic cleavage. In fact, the linearized plasmid DNA can be enzymatically religated with a quantitative conversion, indicating that only the products of P-O3′ bond scissions are formed. This high regioselectivity is also supported by 3′- and 5′-end labelling studies. The cleavage of plasmid DNA is so efficient that, in the presence of 10 mM complex, at 25 °C and pH 8.0, the supercoiled form is completely degraded immediately after mixing. The extent of cleavage depends on the concentration of the complex, but, surprisingly, it does not increase with extended reaction times. To justify this unusual reactivity, the authors proposed a mechanism involving nucleophilic attack of a metal-ion-bound peroxide to the phosphodiester bond.31 Similar behaviour has been recently reported for the dinuclear complex [CpFe(CO)2]2 which hydrolytically cleaves DNA at very low concentrations in the presence of H2O2 or other peroxides.32 Another, again very efficient, diiron(III) complex, based on a benzimidazolylmethyl derivative of 1,4,7-triazaheptane (Figure 13.11, DTPB), has been investigated by Liu and coworkers.33 In the presence of 100 mM of Fe(III)2-DTPB complex, at 37 °C and pH 7.0, the supercoiled DNA is degraded with a rate constant of 2.1 × 10−3 s−1, which corresponds to a half-life of 5 minutes. This is the current ‘world record’ for the cleavage of plasmid DNA under ‘hydrolytic conditions’. The extent of cleavage does not depend on the presence of O2 and the linearized DNA can be

Bimetallic Complexes

Figure 13.11 (DTPB)

379

Bimetallic ligand for Fe(III) reported by Schnaith and Que (HPTB), and Liu

quantitatively religated, thus confirming the occurrence of a hydrolytic mechanism with the sole cleavage of the P-O3′ bond. The complex shows a remarkable affinity for plasmid DNA (Ka = 1 × 105 M−1), probably related to the high positive charge of its bimetallic structure. The comparison with other bimetallic complexes indicates that a m-oxo bridge between the two metal ions and the presence of a labile BF4− ligand play a key role in determining the reactivity: the first by ensuring the proper intermetallic distance and the second by allowing a facile substitution to yield free coordination sites for interaction with the substrate. Surprisingly, examples of bimetallic systems based on Cu(II) or Zn(II) are scarce and show a disappointingly low reactivity.34 The only important exception is a binuclear Zn(II)-binding heptapeptide (Figure 13.12), described by Scrimin and coworkers,35 which cleaves plasmid DNA with a first-order rate constant of 1.0 × 10−5 s−1 at pH 7.0, 37 °C and 3.6 mM complex concentration. Due to the presence of several a-disubstituted amino acids, the peptide is folded to a relevant extent in a 310-helical conformation, and the two metal chelating TACN moieties face each other at a distance of about 6 Å (the pitch of the 310-helix). Even if such distance is larger than that found in bimetallic hydrolytic enzymes, the system is about 20 times more reactive than its monometallic counterpart and the cooperativity between the two metal centres has been clearly demonstrated with Zn2+-concentration-dependent experiments. The high cooperativity observed has been justified by assuming the formation of a sandwich-like complex in which the two Zn(II)-TACN complexes interact with a single DNA phosphate group, taking full advantage of their complementary roles for the hydrolytic cleavage.

380


Figure 13.12

Chemical structure of the Scrimin’s ditopic Zn(II)-binding heptapeptide

An impressive high reactivity was recently reported by Yu and coworkers,36a who studied dimetallic Zn(II) complexes made by linking two cyclen ligands with different spacers bearing alcoholic, phenolic or pyridinic moieties. The authors claim complete nicking of plasmid DNA in only five minutes at pH 7.1, 37 °C and 25 mM complex concentration. Longer times of incubation result in further fragmentation and DNA smearing on the electrophoresis gel. However, the complexes are active only in the presence of vitamin C as reducing agent. Therefore, the occurrence of some concurrent oxidative degradation pathway, involving the ligand backbone or adventitious redox active metal ions, is possible. The same type of reactivity was also observed in the case of the Zn(II) complexes of uracil-cyclen conjugate ligands.36b These results, although impressive if the mechanism proves to be really hydrolytic, stress the importance of using caution when studying the mechanism of DNA cleavage, even when Zn(II) is employed as metal ion. Indeed this natural substrate is really stable toward hydrolysis, but is also very delicate from other points of view and competitive cleavage pathways are always possible.

13.5 Conjugation with DNA-affine Subunits Simple considerations, supported by several examples reported in the case of oxidative DNA cleaving agents,37 suggest that the reactivity of hydrolytic systems should be strongly enhanced when their structures comprise groups with high DNA affinity. Furthermore, such groups could also provide sequence specificity to the cleaving agents and thus open the way toward the obtainment of real artificial restriction enzymes. The obvious choice among the different families of compounds able to increase DNA affinity is the utilization of intercalators. Quite surprisingly, examples of metal complexes appended to intercalating groups as hydrolytic agents are rare, and the effects of such elements on the reactivity of the systems are not always straightforward. For example, in the case of the Scrimin’s peptide discussed above,35 the introduction of an acridine moiety at the N-terminus of the heptapeptide resulted in a slightly higher activity at low catalyst concentration compensated by a decrease at higher concentrations.

Conjugation with DNA-affine Subunits 381

Figure 13.13 Artificial nucleases based on ruthenium and rhodium intercalators reported by Burton and coworkers

In 1987, Barton and coworkers published the first example of an artificial system with DNA hydrolytic cleavage activity.4 It was based on a ruthenium intercalator with two arms having the role of metal-binding moieties (Figure 13.13a). The Zn(II) and the Cd(II) binuclear complexes of this ligand cleaved plasmid DNA with high efficiency. At 37 °C and pH 8.5, 40% of the supercoiled form was degraded in the presence of a 7 mM concentration of the complex after 5 hours of incubation (a first-order rate constant of 3 × 10−5 s−1 can be roughly estimated). Religation experiments showed that the hydrolysis occured randomly at the P-O3′ and P-O5′ bonds. However, the affinity of the ligand for the metal ions seemed to be quite low and a large excess of metal ion was required. Unfortunately, no evidence concerning the role played by the intercalator was reported. Later on, Barton and coworkers studied the reactivity of a mononuclear Zn(II)binding peptide, tethered to a rhodium complex as major groove intercalator (Figure 13.13b).38 The system promotes plasmid DNA cleavage with a rate constant of 2.5 × 10−5 s−1 at pH 6, 37 °C and 5 mM complex concentration, and shows similar activity toward linear double-stranded DNA. Analysis of the fragments produced showed that the cleavage occurs only at the P-O3′ bonds with a modest sequence selectivity for 5′-Pu-Py-Pu-Py-3′sites (with cleavage at Py), presumably the result of the Rh complex binding selectivity. The reactivity decreases as the pH approaches neutrality. The peptide is scarcely structured and this is probably the source of the weak binding of the Zn(II) ion. Although the presence of the rhodium complex intercalator is of paramount importance for the activity of the complex, its substitution with a different rhodium-based intercalator, which preferentially binds at mismatches, or with the ethidium intercalator led to unreactive systems.38b Schneider and coworkers appended two naphthalene groups to an azacrown ligand via C6 alkyl spacers (Figure 13.14a) and used it as a cofactor for Eu(III)promoted DNA cleavage.39 The results are puzzling because conjugation with the intercalating units led to an increase in the intrinsic reactivity with regard to the

382


Figure 13.14 Intercalator-metal binding unit conjugates reported by Schneider (a), Nakamura (b) and Tonellato (c)

free metal ion, but to a decrease in the DNA affinity. Finally, an earlier report by Nakamura and Hashimoto,40 which investigated the reactivity of a hydroxamic acid linked to a phenanthridine intercalator (Figure 13.14b) in the presence of different lanthanide ions, points attention to the length of the spacer that tethers the intercalating unit to the catalytic group as a key element for cleavage activity. This result is confirmed by the reactivity of a series of Zn(II)-cis-cis-triaminocyclohexane complex–anthraquinone intercalator conjugates, linked by alkyl spacers of different length, reported by Tonellato and coworkers (Figure 13.14c).41 At a concentration of 5 mM, the complex of the derivative with a C8 alkyl spacer cleaves supercoiled DNA with a rate of 4.6 × 10−6 s−1 at pH 7 and 37 °C. Saturation kinetics have been observed with a binding constant (Ka) of about 1 × 104 M−1, in agreement with the reported DNA affinity of anthraquinone. Thus, the conjugation of the metal complex with the intercalating group led to a 15-fold increase in the cleavage efficiency when compared with the Zn-triaminocyclohexane complex lacking the anthraquinone moiety. Comparison of the reactivity of the different complexes showed a remarkable increase in DNA cleaving efficiency due to the increase in the spacer length. In the case of the shortest spacer (C4) no cleavage was observed, indicating that the advantages derived by the increased DNA affinity may be cancelled out by incorrect positioning of the reactive group. A recent example of a DNA-intercalating Zn(II) complex has been described by Yang and coworkers,42 who studied the reactivity of the Zn[(phen)(dione)Cl]ClO4 complex (Figure 13.15) toward pBR322 plasmid DNA. In this case the ligand itself is able to intercalate DNA with a binding constant of 2.4 × 104 M−1 and the cleavage proceeds with a rate constant of 5.8 × 10−5 s−1 at pH 8.1, 37 °C and a 3.0 mM complex concentration. The observed rate constant is relatively high, but it has to be taken in account that it has been obtained at high complex concentration. Due to the nature of the ligand, a direct comparison with a nonbinding complex is not possible, but the authors report that other complexes made by Zn2+ and phenanthroline are not active.

Conjugation with DNA-affine Subunits 383

Figure 13.15 The [Zn(phen)(dione)]2+ complex

+

+N

N N N

+N

Figure 13.16

N

2+

Zn

OH

N N +

Proposed structure of the active form of the Zn(II) complex studied by Mao

DNA affinity elements other than intercalators have also been used. Positively charged ammonium or peralkylammonium groups have a good affinity toward the polyanionic DNA. Schneider and coworkers43 studied the reactivity of the Co(III) complexes of cyclen derivatives bearing side chains of different lengths and terminating with a trimethyl ammonium group. At 37 °C and pH 7.0, they observed saturation kinetics, both with the Co(III)-cyclen complex and with the complexes of the alkylammonium conjugates. Interestingly, while the maximum reactivity remains constant within the series (kcat ca. 2 × 10−4 s−1), the binding constant increases with the length of the spacer (from 1.0 × 103 M−1 in the case of Co(III)-cyclen to 5.5 × 103 M−1 for the complex with a C6 spacer). Further increments of activity could be obtained by adding a second alkylammonium or Co(III)-cyclen groups. Again, peralkylammium groups were employed as affinity elements by Mao and coworkers.44 They synthesized two Zn(II) complexes of bipyridyl ligands bearing ethylammonium or butylammonium groups in the positions 5 and 5′. The former Zn(II) complex cleaves DNA at pH 7.2, 37 °C and 50 mM complex concentration with a rate constant of 3.8 × 10−5 sec−1. The pH reactivity profile shows a maximum at pH 7.5. On the basis of the crystal structure of the complex, the authors suggest that the quaternary ammonium groups (Figure 13.16) are placed in the right position to interact with the two negatively charged phosphate groups neighbouring the one that is activated by the metal ion. More sophisticated and biologically minded systems have been reported by the groups of Franklin and Sugiura. The first designed a 33-mer peptide incorporating a DNA recognition sequence and a metal-ion binding site.45 The peptide folds in a

384


helix-turn-helix (HTH) motif where the two helices have been derived from the DNA-binding engrailed homeodomain and the turn reproduces the sequence of the calcium binding site of calmodulin. Addition of Ce(IV) or Eu(III) leads to the formation of a peptide/lanthanide ion complex which retains the HTH tertiary structure. The Ce(IV) complex (and to a minor extent also the Eu(III) complex) promotes the cleavage of plasmid and linear double-stranded DNA and the reactivity of the complex is similar to that of the free ion. However, while the uncomplexed metal ion cleaves randomly both the P-O3′ and P-O5′ bonds, the peptide complex is regioselective for P-O3′ bonds. Furthermore, modest sequence selectivity for T/C rich sequences was also observed. In a similar approach, Sugiura and Nomura have succeeded in the preparation of zinc-finger peptides with hydrolytic ability toward phosphate esters, by mutating some amino acid residues coordinated to the zinc ion.46a This mutant is based on the sequence of the second finger in the three-tandem zinc finger protein Sp1, a DNA binding protein that binds specifically to GC boxes. Tandem alignment of three zinc finger mutants in a way similar to natural proteins resulted in efficient and site selective cleavage of plasmid DNA pUC19GC, and of a 37 bp DNA duplex, both containing a GC box. In another example, two zinc-finger peptides were connected to a peptide linker able to bind lanthanide ions.46b In the presence of Ce(IV) ions, the system cleaves linear dsDNA with high sequence selectivity.

13.6 Conjugation with Sequence-Selective Elements A more exciting and fascinating application of the DNA hydrolytic catalysts is the obtainment of artificial restriction enzymes with higher or different sequence specificity than natural systems. However, such a goal requires highly specific sequence recognition while most of the systems so far described produce random cleavage of DNA or show only modest sequence selectivity. A much higher selectivity is required and, in principle, this can be obtained by metal complexes conjugated with DNA oligonucleotides (for antisense recognition or triple helix formation) or PNA fragments. At the present time, only a very few examples of such systems are known. Komiyama and coworkers47 first reported, in 1994, a conjugate in which an iminodiacetate metal-binding group was appended to a 19-mer DNA oligonucleotide. In the presence of Ce(IV) ions, the DNA– iminodiacetate conjugate efficiently cleaved single-stranded 40-residue DNA at the linkage between residues 30 and 31, according to the predicted selectivity, with scission of the P-O5′ bond. Noticeably, the cleavage of the single-stranded DNA by this conjugate is much more efficient than hydrolysis of the dinucleotide TpT by Ce(IV)iminodiacetate, suggesting that the DNA moiety also has a role in delivering the metal ion close to cleavable phosphate groups. Krämer and coworkers described more recently a family of PNA–metal chelating group conjugates whose Zr(IV) complexes selectively cleave single-stranded DNA oligonucleotides.48 The cleavage is less efficient than in the case of Komiya-

The ARCUT System 385

ma’s DNA conjugate, probably due to the use of the less reactive Zr(IV) ion. Moreover, the employment, under the conditions used, of a large amount of free Zr(IV), due to the low metal affinity of the ligand (TRIS) subunit, leads to a substantial nonspecific random cleavage. At any rate, the fragments formed are consistent with the anticipated selectivity and the scission occurrs with remarkable regiospecificity at the P-O3′ bond.

13.7 The ARCUT System Very recently, Komiyama and coworkers described a new approach to the selective hydrolytic cleavage of DNA, based on the selectivity of the Ce(IV)-EDTA complex.5 As mentioned before, this complex cleaves ssDNA with such preference that bulges or gap-sites in dsDNA are selectively cleaved.17,18 In the ARCUT protocol, gap-like structures are formed at the desired sites of dsDNA or plasmid DNA by invasion of two pseudo-complementary PNA strands (pcPNA), whose sequences are selected in order to bind laterally shifted base sequences in the target (Figure 13.17).49 The two single-stranded portions formed in the invaded DNA are hence cleaved by incubation with Ce(IV)-EDTA. The sequence selectivity can be programmed without limitations and cleavage does not occur even in the presence of a single base mismatch.50 Cleaving efficiency is greatly enhanced by attaching a serine phosphate monoester at the pcPNA termini close to the gap site, since the dianionic terminal phosphate induces the accumulation of the Ce(IV) complex.50 However the reaction still requires, in many cases, high temperature (50 °C) and prolonged reaction times (64 h). DNA fragments obtained can be religated and connected with foreign double-stranded DNA by using DNA ligase and a ligation joint to provide recombinant DNA (Figure 13.17).18,49,51 The method has been applied by Komiyama and coworkers to the site-selective scission of genomic DNA of Escherichia coli (4.6 × 106 bases),50 to the site-selective scission of enzymatically methylated DNA (which is resistant to restriction enzymes)50, to the preparation of recombinant DNA18,49,51 and even of fusion proteins.52

gap sites Ligation joint

pcPNAs Ce (IV)/EDTA

Figure 13.17 binant DNA

Ligase

Schematic representation of the ARCUT protocol for the preparation of recom-

386


13.8 A Critical Survey and New Perspectives Twenty years have passed since the first report by Barton and coworkers of an artificial agent for the hydrolytic cleavage of DNA.4 Since then, the realization of such systems has attracted great interest and has led, with Komiyama’s ARCUT,5 to the first examples of DNA manipulation performed with synthetic agents. Still, several important goals remain unachieved and the recognition of the most promising guidelines is still not easy. The ideal features of an artificial hydrolytic agent are high activity, high DNA affinity and sequence selectivity. Other requirements may vary depending on the final application: for an artificial restriction agent, double-stranded cleavage at the P-O3′ site is a requirement for subsequent religation, while for an antisense drug the absence of cofactors, activity at very low concentrations and at physiological pH and temperature are surely important. For all the possible applications, the occurrence of an oxidative cleavage, either as main or concurrent pathway, is highly undesirable. The last point is of particular relevance with synthetic agents. Metal ions used by nature in nucleases and phosphatases, e.g. Ca(II), Mg(II), Fe(II), Fe(III) and Zn(II), do not have, with exception of the Fe(II)/Fe(III) couple, relevant redox properties. Chemists, however, can pick up the most suitable candidate over the whole periodic table. The ideal metal ion for the realization of a hydrolytic system should have a hard character (to bind phosphate oxygen atoms), high Lewis acidity (to polarize both the phosphate group and the nucleophile) and a fast rate of ligand exchange (to ensure catalytic turnover).3o Several metal ions not used by nature best fulfil such requisites, particularly trivalent lanthanide ions, Ce(IV) and Cu(II). In the case of Cu(II), however, oxidative cleavage pathways have been detected for most of the hydrolytically active Cu(II)-based systems so far reported, both in the presence and in the absence of reducing agents and, on some occasions, even of oxygen. The adventitious presence of reducing agents either in DNA samples obtained from natural sources or in the in vivo environment, makes the use of Cu(II)-based systems as hydrolytic agents rather discouraging, notwithstanding their excellent activity. Moreover, the recent results form Rediijk and coworkers,9 who discovered oxidative DNA cleavage with a Zn(II) complex containing a phenol unit, put further emphasis on the need to assess clearly the reaction mechanism, even in the case of metal ions without a relevant redox activity. After the cleavage pathway, the first problem to face is that of the hydrolytic efficiency of the artificial system. A comparison of the activity of the different agents reported is rather difficult because of the large variety of reaction conditions used. However, a rough evaluation of the most reactive agents is tentatively shown in Table 13.1, which reports, where possible, the plasmid DNA degradation rate at a fixed agent concentration (5 mM, kψ′ ) and agent to DNA (base pair) ratio (0.125, kψ′′), and the Michaelis–Menten parameters. A first glance at the kψ′ and kψ′′ data reported, indicates that simple metal ions and monometallic complexes can hardly achieve relevant reactivity. Such low activity appears to be related mainly to a low affinity for DNA. Indeed, when the kcat

2.6·104 1.7·104 – – 1.3·104 3.0·103 1.0·103 – – – 5.5·103 – 2.5·105 1.7·103 – 1.0·104

7.0·10−5 1.0·10−4 – – 1.2·10−3 2.8·10−4 1.0·10−3 – – – 2.7·10−4 – 5.2·10−4 5.8·10−4 – –

Eu(III) Er(III) Zn(II)-pseudopeptide (50 °C) Zn-quercetin Cu(II)-TACH (35 °C) Pr(III)2-azacrown Er(III)2-shiff macrocycle Ce(IV)2-HXTA Fe(III)2-DTPB Zn(II)2-peptide Co(III)-cyclenC6ammonium Zn(II)-bipy-ammonium Cu(II)-neamine Eu(III)-naphtocrown Zn(II)-Rh-peptide Zn(II)-TACH-anthraquinone

8.0·10−6 7.8·10−6 ∼4·10−5 (3 m M) 1.7·10−4 (100 µ M) 7.3·10−5 4.1·10−6 5.0·10−6 1.4·10−4 (10 m M) 2.1·10−3 (100 m M) 1.0·10−5 (3.6 m M) 7.2·10−6 3.8·10−5 (50 m M) 2.8·10−4 4.9·10−6 2.5·10−5 (5.0 m M) 4.6·10−6 (5.0 m M)

kψ′ (s−1)b 4.1·10−6 3.8·10−6 – – 1.3·10−5 2.0·10−6 – 1.4·10−4 (r = 0.07) – 1.0·10−5 (r = 0.3) 3.5·10−6 – 3.1·10−4 2.3·10−6 2.5·10−5 1.9·10−6

kψ′′ (s−1)c

7.3 7.0 6.0 7.0

7.0 7.0 7.0 7.2 8.1 7.0 7.0 8.0 7.0 7.0 7.0

pH

c

b

19 19 12 40 7 19 – 150 77 12 19 38 50 19 40 12

[bp]d (mM)

Reactivity at 37 °C unless otherwise stated. Pseudo first-order rate constant for the cleavage of plasmid DNA at a fixed concentration of metal complex (5 µM). Pseudo-first-order rate constant for the cleavage of plasmid DNA at a fixed metal complex to DNA (base pair) ratio (0.125), unless otherwise stated. d Concentration of plasmid DNA reported as base pairs.

a

Ka (M−1)

kcat (s−1)

Hydrolytic Agenta

Table 13.1 Kinetic parameters for some of the most active artificial metallonucleases

39 11 20 21a 23 27 28 29 33 35 43 44 25 39 38 41

reference

A Critical Survey and New Perspectives 387

388


values reported are examined, several agents appear have the potential to reach a good reactivity provided their affinity for DNA is increased. As inspired by natural enzymes, the development of bimetallic systems should be a promising strategy in order to increase the activity of the artificial hydrolytic agents. In these systems, not only the intrinsic reactivity may increase as a result of the cooperative action of the two metal centres, but also the DNA affinity should increase due to the larger positive charge of the complex and to the possibility of multipoint interaction with the DNA phosphate backbone. As a matter of fact, two of the most reactive agents so far reported, Liu’s Fe(III)2-DTPB33 and Que’s Ce(IV)2HXTA,29 are bimetallic. This strategy is particularly important, both in the case of Zn(II)-based systems, as the intrinsic hydrolytic activity of this metal is somewhat lower that that of the other ions, and in the case of lanthanide systems, since the obtainment of nonlabile metal complexes with free coordination sites usually leads to the use of polycarboxylic ligands which, as a drawback, reduce the reactivity of these ions. Nevertheless, the binuclear agents so far reported are relatively few and, in many cases, the reactivity gains are not so impressive when compared to the corresponding mononuclear complexes. Studies on model phosphate esters clearly indicate that the design of such systems is more delicate than the simple synthesis of ligands capable of binding two metal ions. The intermetallic distance and the rigidity of the complex are crucial features in order to allow full cooperation between the metal centres. Moreover, at least in the case of Cu(II)- and Zn(II)based agents, the formation of m-hydroxo bridges between the two ions can dramatically decrease the reactivity and must, accordingly, be strictly avoided.24a,53 Besides the realization of bimetallic systems, the main route toward the obtainment of efficient artificial agents requires, as mentioned before, an increase in the affinity of the hydrolytic complexes for DNA. The mechanism of the metal-ioncatalysed hydrolysis of phosphate esters involves, as a crucial initial step, the coordination of the substrate to the metal ion. This coordination is an essential requisite in order to deliver the metal-ion-coordinated hydroxide nucleophile close to the phosphate group, thus offsetting the electrostatic repulsion between the two negatively charged species, and allowing the activation of the phosphate toward nucleophilic attack. The metal-ion binding ability of phosphate diesters is rather poor and it is only slightly favoured by the polyanionic nature of DNA. It is not accidental that all the most reactive systems so far reported feature DNA affinity elements, such as two metal ions, intercalators, positively charged ammonium groups and aminosugars. Even in the cases where DNA binding has not been investigated, such as the Zn-(quercetin)2 complex,21 an effect of the poliphenolic ligand to increase the substrate affinity is probably present. However, the results reported indicate that the simple conjugation of a metal complex to a DNA-binding unit is not sufficient, particularly in the case of intercalating groups, and that a very careful design of the catalyst structure is required. In fact, in order to obtain the desired activity, the binding must lead to the correct geometry favouring a close contact between the metal complex and the DNA phosphate backbone. In this context, it appears that the design of the linker that connects the metal ion complex to the DNA-binding subunit is of great importance and, in particular, its length and flexibility. This is not an easy task since these features may easily change depending on the DNA binding site (major groove, minor groove, mismatches) and need to be tailored accordingly. In

Acknowledgements

389

contrast, the results obtained by Ichikawa20 and Mao44 with Zn(II) complexes and by Schneider43 with Co(III) complexes bearing ammonium groups indicate that the use of such affinity elements may be much less demanding in terms of structural requirements. Short DNA oligonucleotides (for both antisense recognition and triple helix formation) or PNA fragments are probably the most promising DNA-binding subunits available to chemists. Indeed, such systems conjugate tight binding to DNA with high selectivity, and such features are ideally suited for the obtainment of a truly useful artificial nuclease. Such systems have been extensively studied for RNA3b but, up to today, very few agents have been studied with DNA and they have only been tested on short single strand fragments, while the problem of double-strand cleavage of DNA has still to be addressed.

13.9 Conclusions Since the first report on the ability of metal ions to promote DNA hydrolysis, tremendous progress toward the obtainment of efficient synthetic DNA hydrolytic agents has been attained. At present, chemists have at their disposal catalysts that cleave DNA in a few tens of minutes and, in some cases, can display interesting selectivity. However, several other aspects need to be addressed before a catalyst may reach the stage of practical application. Among the most important are: selectivity, double-strand scission and cleavage efficiency. The last point is of particular importance in the case of intrinsically less reactive, but biologically relevant metal ions, such as, for example, Zn(II). The use of multinuclear complexes as catalysts and of selective DNA-binding subunits supported by careful design of the system appear to be the main routes that have been successfully followed to date. Now, the problem is the integration of these different features into a single system in order to obtain really effective metallonuclease models. The Komiyama ARCUT system5 outlines an alternative strategy to that followed by the majority of the researchers in the field. In this case, selective cleavage of ssDNA over dsDNA has been exploited trough an extensive investigation of the reactivity of a single, very simple, and even moderately efficient cleaving agent. Such selectivity, conjugated with a smart design for the working protocol, has led to the realization of the only artificial restriction agent so far reported. Also in this case higher reactivity would be desirable in order to achieve shorter reaction times and higher cleavage yields, but the success obtained represents strong encouragement toward a renewed research effort in this field.

Acknowledgements We are deeply indebted to Prof. Umberto Tonellato, Prof. Paolo Scrimin, Prof. Manlio Palumbo, Dr Claudia Sissi and Prof. Stefano Moro for their productive collaboration and numerous stimulating discussions.

390


References 1. a) Schroeder, G.K.; Lad, C.; Wyman, P.; Williams, N.H.; Wolfenden, R.; The time required for water attack at the phosphorus atom of simple phosphodiesters and of DNA; Proc. Natl. Acad. Sci. USA, 2006, 103, 4052–4055. b) Williams, N.H.; Takasaki, B.; Wall, M.; Chin, J.; Structure and nuclease activity of simple dinuclear metal complexes: Quantitative dissection of the role of metal ions; Acc. Chem. Res., 1999, 32, 485–493. 2. a) Weston, J.; Mode of action of bi- and trinuclear zinc hydrolases and their synthetic analogues; Chem. Rev., 2005, 105, 2151–2174; b) Lowther, W.T.; Matthews, B.W.; Metalloaminopeptidases: Common functional themes in disparate structural surroundings; Chem. Rev. 2002, 102, 4581–4607; c) Jedrzejas, M.J.; Setlow, P.; Comparison of the binuclear metalloenzymes diphosphoglycerate-independent phosphoglycerate mutase and alkaline phosphatase: Their mechanism of catalysis via a phosphoserine intermediate; Chem. Rev., 2001, 101, 607–618; d) Cowan, J.A.; Metal activation of enzymes in nucleic acid biochemistry; Chem. Rev., 1998, 98, 1067–1087; e) Wilcox, D.E.; Binuclear metallohydrolases; Chem. Rev., 1996, 96, 2435–2458; f) Sträter, N.; Lipscomb, W.N.; Klabunde, T.; Krebs, B.; Two-metal ion catalysis in enzymatic acyl- and phosphoryl-transfer reactions; Angew. Chem. Int. Ed. Engl., 1996, 35, 2024–2055. 3. a) Mancin, F.; Tecilla, P.; Zinc(II) complexes as hydrolytic catalysts of phosphate diester cleavage: from model substrates to nucleic acids; New J. Chem., 2007, 31, 800–817; b) Niittymäki, T.; Lönnberg, H.; Artificial ribonucleases; Org. Biomol. Chem., 2006, 4, 15–25; c) Mancin, F.; Scrimin, P.; Tecilla, P.; Tonellato, U.; Artificial metallonucleases; J. Chem. Soc., Chem. Commun., 2005, 2540–2548; d) Yatsimirsky, A.K.; Metal ion catalysis in acyl and phosphoryl transfer: Transition states as ligands; Coord. Chem. Rev., 2005, 249, 1997–2011; e) Liu, C.; Wang, M.; Zhang, T.; Sun, H.; DNA hydrolysis promoted by di- and multi-nuclear metal complexes; Coord. Chem. Rev., 2004, 248, 147–168; f) Morrow, J.R.; Iranzo, O.; Synthetic metallonucleases for RNA cleavage; Curr. Opin. Chem. Biol., 2004, 8, 192–200; g) Cowan, J.A.; Chemical nucleases; Curr. Opin. Chem. Biol., 2001, 5, 634–642; h) Sreedhara, A.; Cowan, J.A.; Catalytic hydrolysis of DNA by metal ions and complexes; J. Biol. Inorg. Chem., 2001, 6, 337–347; i) Franklin, S.J.; Lanthanide-mediated DNA hydrolysis; Curr. Opin. Chem. Biol., 2001, 5, 201–208; j) Kimura, E.; Dimetallic hydrolases and their models; Curr. Opin. Chem. Biol., 2000, 4, 207–213; k) Kimura, E.; Kikuta, E.; Why zinc in zinc enzymes? From biological roles to DNA base-selective recognition; J. Biol. Inorg. Chem., 2000, 5, 139–155; l) Krämer, R.; Bioinorganic models for the catalytic cooperation of metal ions and functional groups in nuclease and peptidase enzymes; Coord. Chem. Rev., 1999, 182, 243–261; m) Komiyama, M.; Takeda, N.; Shigekawa, H.; Hydrolysis of DNA and RNA by lanthanide ions: mechanistic studies leading to new applications; J. Chem. Soc., Chem. Commun., 1999, 1443–1451; n) Ott, R.; Krämer, R.; DNA hydrolysis by inorganic catalysts; Appl. Microbiol. Biotechnol., 1999, 52, 761–767; o) Hegg, E.L.; Burstyn, J.N.; Toward the development of metal-based synthetic nucleases and peptidases: a rationale and progress report in applying the principles of coordination chemistry; Coord. Chem. Rev., 1998, 173, 133–165; p) Komiyama, M.; Sumaoka, J.; Progress towards synthetic enzymes for phosphoester hydrolysis; Curr. Opin. Chem. Biol., 1998, 2, 751–757; q) Chin, J.; Developing artificial hydrolytic metalloenzymes by a unified mechanistic approach; Acc. Chem. Res., 1991, 24, 145–152; r) Barton, J.K.; Metal/Nucleic-Acid Interactions. In: Bioinorganic Chemistry; Bertini I.; Gray H.B.; Lippard S.J.; Valentine, J.S., Eds., University Science Books; Mill Valley, 1994, 455–504. 4. Basile, L.A.; Barton, J.K.; Design of a double-stranded DNA cleaving agent with two polyamine metal-binding arms: Ru(DIP)2Macron+; J. Am. Chem. Soc., 1987, 109, 7548–7550. 5. Sumaoka, J.; Yamamoto, Y.; Kitamura, Y.; Komiyama, M.; Artificial restriction DNA cutters (ARCUT) for future biotechnology; Curr. Org. Chem., 2007, 11, 463–475. 6. Takeda, N.; Shibata, M.; Tajima, N.; Hirao, K.; Komiyama, M.; Kinetic and theoretical studies on the mechanism of alkaline hydrolysis of DNA; J. Org. Chem., 2000, 65, 4391–4396.

References

391

7. a) Jiang, Q.; Xiao, N.; Shi, P.; Zhu, Y.; Guo, Z.; Design of artificial metallonucleases with oxidative mechanism; Coord. Chem. Rev., 2007, 251, 1951–1972; b) Sigman, D.S.; Mazumder, A.; Perrin, D.M.; Chemical nucleases; Chem. Rev., 1993, 93, 2295–2316. 8. Maheswari, P.U.; Roy, S.; den Dulk, H.; Barends, S.; van Wezel, G.; Kozlevcar, B.; Gamez, P.; Reedijk, J.; The square-planar cytotoxic [CuII(pyrimol)Cl] complex acts as an efficient DNA cleaver without reductant; J. Am. Chem. Soc., 2006, 128, 710–711. 9. Maheswari, P.U.; Barends, S.; Özalp-Yaman, S.; de Hoog, P.; Casellas, H.; Teat, S.J.; Massera, C.; Lutz, M.; Spek, A.L.; van Wezel, G.P.; Gamez, P.; Reedijk, J.; Unique ligand-based oxidative DNA cleavage by zinc(II) complexes of Hpyramol and Hpyrimol; Chem. Eur. J., 2007, 13, 5213–5222. 10. a) Komiyama, M.; Takeda, N.; Takahashi, Y.; Uchida, H.; Shiiba, T.; Kodama, T.; Yashiro, M.; Efficient and oxygen-independent hydrolysis of single-stranded-DNA by cerium(IV) ion; J. Chem. Soc., Perkin 2, 1995, 269–274; b) Takasaki, B.K.; Chin, J.; Cleavage of the phosphate diester backbone of DNA with cerium(III) and molecular-oxygen; J. Am. Chem. Soc., 1994, 116, 1121–1122; c) Komiyama, M.; Shiiba, T.; Kodama, T.; Takeda, N.; Sumaoka, J.; Yashiro, M.; DNA hydrolysis by cerium(IV) does not involve either molecular-oxygen or hydrogen-peroxide; Chem. Lett., 1994, 1025–1028. 11. Roigk, A.; Hettich, R.; Schneider, H.J.; Unusual catalyst concentration effects in the hydrolysis of phenyl phosphate esters and of DNA: A systematic investigation of the lanthanide series; Inorg. Chem., 1998, 37, 751–756. 12. Sumaoka, J.; Azuma, Y.; Komiyama, M.; Enzymatic manipulation of the fragments obtained by cerium(IV)-induced DNA scission: Characterization of hydrolytic termini; Chem. Eur. J., 1998, 4, 205–209. 13. The reactivity of Ce(IV) ions is further enhanced by the use of mixtures with Pr(III) or Nd(III): Takeda, N.; Imai, T.; Irisawa, M.; Sumaoka, J.; Yashiro, M.; Shigekawa, H.; Komiyama, M.; Unprecedentedly fast DNA hydrolysis by the synergism of the cerium(IV)praseodymium(III) and the cerium(IV)-neodymium(III) combinations; Chem. Lett., 1996, 599–600. 14. Sumaoka, J.; Igawa, T.; Furuki, K.; Komiyama, M.; Homogeneous Ce(IV) complexes for efficient hydrolysis of plasmid DNA; Chem. Lett., 2000, 56–57. 15. Igawa, T.; Sumaoka, J.; Komiyama, M.; Hydrolysis of oligonucleotides by homogeneous Ce(IV)/EDTA complex; Chem. Lett., 2000, 356–357. 16. a) Roigk, A.; Schneider, H.J.; Noncovalently bound cofactors for chemical nucleases; Eur. J. Org. Chem., 2001, 205–209; b) Chand, D.K.; Bharadwaj, P.K.; Schneider, H.J.; Cryptands and related tripodal ligands: Interaction with nucleic acids and nuclease activity of their Eu(III) complexes; Tetrahedron, 2001, 57, 6727–6732; c) Roigk, A.; Yescheulova, O.V.; Fedorov, Y.V.; Fedorova, O.A.; Gromov, S.P.; Schneider, H.J.; Carboxylic groups as cofactors in the lanthanide-catalyzed hydrolysis of phosphate esters. Stabilities of europium(III) complexes with aza-benzo-15-crown-5 ether derivatives and their catalytic activity vs bis(p-nitrophenyl)phosphate and DNA; Org. Lett., 1999, 1, 833–835; d) Berg, T.; Simeonov, A.; Janda, K.D.; A combined parallel synthesis and screening of macrocyclic lanthanide complexes for the cleavage of phospho di- and triesters and double-stranded DNA; J. Comb. Chem., 1999, 1, 96–100. 17. Kitamura, Y.; Sumaoka, J.; Komiyama, M.; Hydrolysis of DNA by cerium(IV)/EDTA complex; Tetrahedron, 2003, 59, 10403–10408. 18. Kitamura, Y.; Komiyama, M.; Preferential hydrolysis of gap and bulge sites in DNA by Ce(IV)/EDTA complex; Nucleic Acids Res., 2002, 30, e102. 19. Dixon, N.E.; Geue, R.J.; Lambert, J.N.; Moghaddas, S.; Pearce, D.A.; Sargeson, A.M.; DNA hydrolysis by stable metal complexes; J. Chem. Soc., Chem. Commun., 1996, 1287–1288. 20. Ichikawa, K.; Tarnai, M.; Uddin, M.K.; Nakata, K.; Sato, S.; Hydrolysis of natural and artificial phosphoesters using zinc model compound with a histidine-containing pseudopeptide; J. Inorg. Biochem., 2002, 91, 437–450. 21. a) Tan, J.; Wang, B.; Zhu, L.; Hydrolytic cleavage of DNA by quercetin zinc(II) complex; Bioorg. Med. Chem. Lett., 2007, 17, 1197–1199; b) Tan, J.; Wang, B.; Zhu, L.; Hydrolytic

392

22.

23.

24. 25.

26.

27. 28. 29.

30.

Artificial Restriction Agents cleavage of DNA by quercetin manganese(II) complexes; Colloids Surf. B, 2007, 55, 149–152. a) Hegg, E.L.; Burstyn, J.N.; Copper(II) macrocycles cleave single-stranded and doublestranded DNA under both aerobic and anaerobic conditions; Inorg. Chem., 1996, 35, 7474–7481; b) Deck, K.M.; Tseng, T.A.; Burstyn, J.N.; Triisopropyltriazacyclononane copper(II): An efficient phosphodiester hydrolysis catalyst and DNA cleavage agent; Inorg. Chem., 2002, 41, 669–677. a) Itoh, T.; Hisada, H.; Sumiya, T.; Hosono, M.; Usui, Y.; Fujii, Y.; Hydrolytic cleavage of DNA by a novel copper(II) complex with cis,cis-1,3,5-triaminocyclohexane; J. Chem. Soc., Chem. Commun., 1997, 677–678; b) Kobayashi, T.; Tobita, S.; Kobayashi, M.; Imajyo, T.; Chikira, M.; Yashiro, M.; Fujii, Y.; Effects of N-alkyl and ammonium groups on the hydrolytic cleavage of DNA with a Cu(II)TACH (1,3,5-triaminocyclohexane) complex. Speciation, kinetic, and DNA-binding studies for reaction mechanism; J. Inorg. Biochem., 2007, 101, 348–361. Sissi, C.; Mancin, F.; Gatos, M.; Palumbo, M.; Tecilla, P.; Tonellato, U.; Efficient plasmid DNA cleavage by a mononuclear copper(II) complex; Inorg. Chem., 2005, 44, 2310–2317. a) Sreedhara, A.; Cowan, J.A.; Efficient catalytic cleavage of DNA mediated by metalloaminoglycosides; J. Chem. Soc., Chem. Commun., 1998, 1737–1738; b) Sreedhara, A.; Freed, J.D.; Cowan, J.A.; Efficient inorganic deoxyribonucleases. Greater than 50-millionfold rate enhancement in enzyme-like DNA cleavage; J. Am. Chem. Soc., 2000, 122, 8814–8824. a) Rajendiran, V.; Karthik, R.; Palaniandavar, M.; Stoeckli-Evans, H.; Periasamy, V.S.; Akbarsha, M.A.; Srinag, B.S.; Krishnamurthy, H.; Mixed-ligand copper(II)-phenolate complexes: Effect of coligand on enhanced DNA and protein binding, DNA cleavage, and anticancer activity; Inorg. Chem., 2007, 46, 8208–8221; b) Selvakumar, B.; Rajendiran, V.; Maheswari, P.U.; Stoeckli-Evans, H.; Palaniandavar, M.; Structures, spectra, and DNAbinding properties of mixed ligand copper(II) complexes of iminodiacetic acid: The novel role of diimine co-ligands on DNA conformation and hydrolytic and oxidative double strand DNA cleavage; J. Inorg. Biochem., 2006, 100, 316–330; c) Fernandes, C.; Parrilha, G.L.; Lessa, J.A.; Santiago, L.J.M.; Kanashiro, M.M.; Boniolo, F.S.; Bortoluzzi, A.J.; Vugman, N.V.; Herbst, M.H.; Horn, A. Jr.; Synthesis, crystal structure, nuclease and in vitro antitumor activities of a new mononuclear copper(II) complex containing a tripodal N3O ligand; Inorg. Chim. Acta, 2006, 359, 3167–3176. d) Rossi, L.M.; Neves, A.; Bortoluzzi, A. J.; Hörner, R.; Szpoganicz, B.; Terenzi, H.; Mangrich, A.S.; Pereira-Maia, E.; Castellano, E.E.; Haase, W.; Synthesis, structure and properties of unsymmetrical m-alkoxodicopper(II) complexes: biological relevance to phosphodiester and DNA cleavage and cytotoxic activity; Inorg. Chim. Acta, 2005, 358, 1807–1822. e) Scarpellini, M.; Neves, A.; Horner, R.; Bortoluzzi, A.J.; Szpoganics, B.; Zucco, C.; Silva, R.A.N.; Drago, V.; Mangrich, A.S.; Ortiz, W.A.; Passos, W.A.C.; de Oliveira, M.C.B.; Terenzi, H.; Phosphate diester hydrolysis and DNA damage promoted by new cis-aqua/hydroxy copper(II) complexes containing tridentate imidazole-rich ligands; Inorg. Chem., 2003, 42, 8353–8365. Ragunathan, K.G.; Schneider, H.J.; Binuclear lanthanide complexes as catalysts for the hydrolysis of bis(p-nitrophenyl)phosphate and double-stranded DNA; Angew. Chem. Int. Ed. Engl., 1996, 35, 1219–1221. Zhu, B.; Zhao, D.Q.; Ni, J.Z.; Zeng, Q.H.; Huang, B.Q.; Wang, Z.L.; Binuclear lanthanide complexes as catalysts for the hydrolysis of double-stranded DNA; Inorg. Chem. Commun., 1999, 2, 351–353. a) Branum, M.E.; Tipton, A.K.; Zhu, S.R.; Que, L.; Double-strand hydrolysis of plasmid DNA by dicerium complexes at 37 °C; J. Am. Chem. Soc., 2001, 123, 1898–1904; b) Branum, M.E.; Que, L.; Double-strand DNA hydrolysis by dilanthanide complexes; J. Biol. Inorg. Chem., 1999, 4, 593–600. Schnaith, L.M.T.; Hanson, R.S.; Que, L.; Double-stranded cleavage of PBR322 by a diiron complex via a hydrolytic mechanism; Proc. Natl. Acad. Sci. USA, 1994, 91, 569– 573.

References

393

31. A similar mechanism has been proposed for the cleavage of a dinucleotide by La(III) and H2O2: Takasaki, B.K.; Chin, J.; Synergistic effect between La(III) and hydrogen-peroxide in phosphate diester cleavage; J. Am. Chem. Soc., 1993, 115, 9337–9338. 32. Shell, T.A.; Glass, J.E.; Mackey, M.A.; Layman, K.A.; Mohler, D.L.; Formal DNA hydrolysis by mono- and dinuclear iron complexes; Inorg. Chem., 2007, 46, 8120–8122. 33. Liu, C.; Yu, S.; Li, D.; Liao, Z.; Sun, X.; Xu, H.; DNA hydrolytic cleavage by the diiron(III) complex Fe2(DTPB)(m-O(m-Ac)Cl(BF4)2: Comparison with other binuclear transition metal complexes; Inorg. Chem., 2002, 41, 913–922. 34. a) Zhu, L.; dos Santos, O.; Koo, C.W.; Rybstein, M.; Pape, L.; Canary, J.W.; Geometrydependent phosphodiester hydrolysis catalyzed by binuclear copper complexes; Inorg. Chem., 2003, 42, 7912–7920; b) Aka, F.N.; Akkaya, M.S.; Akkaya, E.U.; Remarkable cooperative action of two zinc centers in the hydrolysis of plasmid DNA; J. Mol. Catal. A, 2001, 165, 291–294. 35. Sissi, C.; Rossi, P.; Felluga, F.; Formaggio, F.; Palumbo, M.; Tecilla, P.; Toniolo, C.; Scrimin, P.; Dinuclear Zn2+ complexes of synthetic heptapeptides as artificial nucleases; J. Am. Chem. Soc., 2001, 123, 3169–3170. 36. a) Xiang, Q.X.; Zhang, J.; Liu, P.Y.; Xia, C.Q.; Zhou, Z.Y.; Xie, R.G.; Yu, X.Q.; Dinuclear macrocyclic polyamine zinc(II) complexes: Syntheses, characterization and their interaction with plasmid DNA; J. Inorg. Biochem., 2005, 99, 1661–1669; b) Wang, X.Y.; Zhang, J.; Li, K.; Jiang, N.; Chen, S.Y.; Lin, H.H.; Huang, Y.; Ma, L.J.; Yu, X.Q.; Synthesis and DNA cleavage activities of mononuclear macrocyclic polyamine zinc(II), copper(II), cobalt(II) complexes which linked with uracil; Bioorg. Med. Chem., 2006, 14, 6745–6751. 37. a) Chen, C.H.B.; Milne, L.; Landgraf, R.; Perrin, D.M.; Sigman, D.S.; Artificial nucleases; Chembiochem., 2001, 2, 735–740; b) Erkkila, K.E.; Odom, D.T.; Barton, J.K.; Recognition and reaction of metallointercalators with DNA; Chem. Rev., 1999, 99, 2777–2795. 38. a) Fitzsimons, M.P.; Barton, J.K.; Design of a synthetic nuclease: DNA hydrolysis by a zinc-binding peptide tethered to a rhodium intercalator; J. Am. Chem. Soc., 1997, 119, 3379–3380; b) Copeland, K.D.; Fitzsimons, M.P.; Houser, R.P.; Barton, J.K.; DNA hydrolysis and oxidative cleavage by metal-binding peptides tethered to rhodium intercalators; Biochemistry, 2002, 41, 343–356. 39. The values of kcat and Ka are respectively 7.0 × 10−5 s−1 and 2.6 × 104 M−1 for the free Eu(III), and 5.8 × 10−4 s−1 and 1.7 × 103 M−1 for the crown ether complex: a) Rammo, J.; Schneider, H.J.; Ligand and cosubstrate effects on the hydrolysis of phosphate esters and DNA with lanthanoids; Liebigs Ann., 1996, 1757–1767; b) Rammo, J.; Hettich, R.; Roigk, A.; Schneider, H.J.; Catalysis of DNA cleavage by lanthanide complexes with nucleophilic or intercalating ligands and their kinetic characterization; J. Chem. Soc., Chem. Commun., 1996, 105–107. 40. a) Hashimoto, S.; Nakamura, Y.; Characterization of lanthanide-mediated DNA cleavage by intercalator-linked hydroxamic acids: Comparison with transition systems; J. Chem. Soc., Perkin Trans. 1, 1996, 2623–2628; b) Hashimoto, S.; Nakamura, Y.; Nuclease activity of a hydroxamic acid-derivative in the presence of various metal-ions; J. Chem. Soc., Chem. Commun., 1995, 1413–1414. 41. Boseggia, E.; Gatos, M.; Lucatello, L.; Mancin, F.; Moro, S.; Palumbo, M.; Sissi, C.; Tecilla, P.; Tonellato, U.; Zagotto, G.; Toward efficient Zn(II)-based artificial nucleases; J. Am. Chem. Soc., 2004, 126, 4543–4549. 42. Yuan, C.X.; Wei, Y.B.; Yang, P.; DNA-binding and cleavage studies of zinc(II) mixedpolypyridyl complex; Chin. J. Chem., 2006, 24, 1006–1012. 43. Hettich, R.; Schneider, H.J.; Cobalt(III) polyamine complexes as catalysts for the hydrolysis of phosphate esters and of DNA. A measurable 10 million-fold rate increase; J. Am. Chem. Soc., 1997, 119, 5638–5647. 44. An, Y.; Lin, Y.Y.; Wang, H.; Sun, H.Z.; Tong, M.L.; Ji, L.N.; Mao, Z.W.; Cleavage of doublestrand DNA by zinc complexes of dicationic 2,2′-dipyridyl derivatives; J. Chem. Soc., Dalton Trans., 2007, 1250–1254. 45. Kovacic, R.T.; Welch, J.T.; Franklin, S.J.; Sequence-selective DNA cleavage by a chimeric metallopeptide; J. Am. Chem. Soc., 2003, 125, 6656–6662.

394


46. a) Nomura, A.; Sugiura, Y.; Sequence-selective and hydrolytic cleavage of DNA by zinc finger mutants; J. Am. Chem. Soc., 2004, 126, 15374–15375; b) Nakatsukasa, T.; Shiraishi, Y.; Negi, S.; Imanishi, M.; Futaki, S.; Sugiura, Y.; Site-specific DNA cleavage by artificial zinc finger-type nuclease with cerium-binding peptide; Biochem. Biophys. Res. Commun., 2005, 330, 247–252. 47. Komiyama, M.; Sequence-specific and hydrolytic scission of DNA and RNA by lanthanide complex-oligoDNA hybrids; J. Biochem., 1995, 118, 665–670. 48. Zelder, F.H.; Mokhir, A.A.; Krämer, R.; Sequence selective hydrolysis of linear DNA using conjugates of Zr(IV) complexes and peptide nucleic acids; Inorg. Chem., 2003, 42, 8618–8620. 49. Yamamoto, Y.; Uehara, A.; Tomita, T.; Komiyama, M.; Site-selective and hydrolytic twostrand scission of double-stranded DNA using Ce(IV)/EDTA and pseudo-complementary PNA; Nucleic Acids Res., 2004, 32, e153. 50. Yamamoto, Y.; Mori, M.; Aiba, Y.; Tomita, T.; Chen, W.; Zhou, J.M.; Uehara, A.; Ren, Y.; Kitamura, Y.; Komiyama, M.; Chemical modification of Ce(IV)/EDTA-based artificial restriction DNA cutter for versatile manipulation of double-stranded DNA; Nucleic Acids Res., 2007, 35, e53. 51. Chen, W.; Kitamura, Y.; Zhou, J.M.; Sumaoka, J.; Komiyama, M.; Site-selective DNA hydrolysis by combining Ce(IV)/EDTA with monophosphate-bearing oligonucleotides and enzymatic ligation of the scission fragments; J. Am. Chem. Soc., 2004, 126, 10285–10291. 52. Yamamoto, Y.; Uehara, A.; Watanabe, A.; Aburatani, H.; Komiyama, M.; Chemical-reaction-based site-selective DNA cutter for PCR-free gene manipulation; Chembiochem., 2006, 7, 673–677. 53. Mancin, F.; Rampazzo, E.; Tecilla, P.; Tonellato, U.; Dinuclear metal complexes based on all-cis-2,4,6-triaminocyclohexane1,3,5-triol as catalysts for cleavage of phosphate esters; Eur. J. Org. Chem., 2004, 281–288.

14 New Metallo-DNAzymes: Fundamental Studies of Metal–DNA Interactions and Metal Sensing Applications Zehui Cao and Yi Lu

14.1 Introduction DNA was long considered only a carrier of genetic information. In 1994, Breaker and Joyce discovered that DNAs isolated using a combinatorial chemistry approach called in vitro selection,1–5 were also able to carry out catalytic activities.1 Since then, many catalytic DNA molecules, termed deoxyribozymes or DNAzymes, have been reported. Being the latest member of the enzyme family after protein and RNA enzymes, DNAzymes have rapidly demonstrated their capability of catalysing a variety of reactions previously known to only ribozymes or protein enzymes,6–9 including RNA/DNA cleavage,1,10–16 ligation,17–19 phosphorylation,18,20 cleavage of phosphoramidate bonds21 and porphyrin metallation.22 Figure 14.1 shows a few examples of DNAzymes. In addition, DNAzymes have shown catalytic efficiencies comparable to those of ribozymes or protein enzymes. For example, the ‘10-23’ DNAzyme identified by Santoro and Joyce has an observed catalytic efficiency of 109 M−1 min−1.15 Unlike RNA and protein enzymes, DNAzymes have not been found in nature, and in vitro selection remains the primary way to isolate and identify new DNAzymes. Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

396

New Metallo-DNAzymes

3' N N N N N N Y rR N N N N N N N 5' 5' N N N N N N R N N N N N N N 3' A G G G C C T A A A G C CTA

3' N N N N N N G rN N N N N N N N 5' N N N N N N N 3' 5' N N N N N N T AA C C A G GG T G C G C C

5' Y Y Y Y A A T A C G N N N N N 3' R R R R A C 3' YYYY

C

CGG

C N N N N N 5' T G

Figure 14.1 DNAzymes for Mg(II) (top), Pb(II) (middle) and Cu(II) (bottom). In each figure, ‘N’ represents any nucleotide provided they abide by Watson–Crick base pairing. All the DNAzymes contain an enzyme strand (bottom strand) and a substrate strand (top strand). Cleavage and ligation sites are marked by an arrow. Y denotes pyrimidine and R denotes purine. (Yi Lu, Prof; New Transition-Metal-Dependent DNAzymes as Efficient Endonucleases and as Selective Metal Biosensors. Chemistry – A European Journal, 2002, 8, 4589–4596. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

14.2 Metal Ions as Important Cofactors of DNAzymes Protein-based enzymes are constructed from 20 amino acids, whereas the variety of structural features of nucleic acid enzymes is limited to only four different nucleotides. The lack of the 2′-OH functional group in DNA compared to RNA further reduces the diversity of conformations that DNAzymes can take for effective target recognition. Nature’s solution to low efficiency and diversity in biomolecular interactions is the employment of cofactors. For example, protein enzymes are known to recruit cofactors, such as NADH, porphyrins and especially metal ions, to broaden the scope of reactions catalysed, increase catalytic efficiency and fine-tune reactivity.23–25 Therefore, to compete favourably with other families of enzymes, cofactors are even more critical for the functions of DNAzymes. Among common cofactors, metal ions are arguably the most important because of their capability to broaden and fine-tune the activity of the emzymes.23–25 In fact, other than a few reports that some DNA/RNAzymes are active independent of the presence of metal ions,13,14 for the majority of DNA/RNAzymes, metal ions, including Mg2+, Mn2+ and Ca2+, are

Selection of DNAzymes Using in vitro Evolution

397

essential for catalytic function at physiological conditions. However, most DNA/ RNAzymes found to date cannot match protein enzymes in terms of either the diversity of metal ions they employ or the specificity of the metal-binding sites. For example, protein enzymes use a wide variety of metal ions, including even the second and third-row transition-metals such as Mo and W, while metal ions used by DNA/RNAzymes are often limited to Mg2+, Ca2+ and Mn2+. Furthermore, it is well known that many protein enzymes possess remarkable metal-binding affinity and specificity, and they are commonly classified by the metal ions they specifically bind (e.g., copper proteins or zinc proteins). DNA/RNAzymes have also been shown to bind certain metal ions selectively, although the number of different metal ions that DNA/RNAzymes can bind is much fewer and the corresponding metal binding affinity is generally lower than those in metalloproteins.26–32 These weaknesses in DNA/RNAzymes are being overcome by using in vitro selection (see below).

14.3 Selection of DNAzymes Using in vitro Evolution Since DNAzymes are almost exclusively generated via in vitro selection, the key to more diverse metal ion cofactors, as well as higher binding affinities and selectivities is improving the selection process. Illustrated in Figure 14.2, generally the selection starts with a pool containing 1014 to 1015 random DNA sequences. Besides a random region in the middle, every DNA in the pool also contains two common priming regions at two ends for polymerase chain reaction (PCR). Under given conditions and cofactors, the DNA sequences that can carry out desired catalytic functions are separated from inactive ones. These active sequences are then amplified by PCR for the next round of reaction and selection. The stringency of the selection can be

Figure 14.2 (Plate 12) Schematic representation of the in vitro selection process (See colour plate section)

398


enhanced in each round by reducing reaction time and cofactor concentration, so that highly active DNAs can be enriched gradually. This process is repeated for a certain number of cycles until DNA sequences with very high activity and selectivity can be isolated and identified. The same strategy has also been successfully employed in the field of aptamer research, where DNAs with high binding affinities and selectivities for the targets of interest, instead of high catalytic activities, are produced. The selection process has allowed the discovery of aptamers for a wide variety of targets, ranging from metal ions33 and small organic molecules,34,35 to proteins36 and even cancer cells.37,38 Theoretically, the selection of DNA/RNAzymes can also be carried out in the presence of many different cofactors, thus allowing the isolation of DNA/RNAzymes with very specific requirements for various cofactors, including any metal ions of interest. Indeed, efforts in this direction have yielded DNA/ RNAzymes that are catalytically active in the presence of some less common metal 43 ion cofactors, such as Pb2+,1,39 Cu2+,12,17 Zn2+,40,41 Co2+,42 and UO2+ It is expected 2 . that DNA/RNAzymes dependent on other metal ions can be generated by a similar approach with well-controlled selection conditions. An inherent problem with the in vitro selection strategy described above is that even though aptamers or DNA/RNAzymes can be obtained under controlled conditions, it does not exclude the possibility that these selected products can also function under different conditions or with different cofactors. This problem obviously results in lower-than-ideal selectivity of the aptamers or DNA/RNAzymes. In some cases, the activity of the nucleic acid enzymes in the presence of the target metal ion is even lower than with certain other metal ions. For example, the Zn(II)-binidng aptamer binds some other metal ions equally well, including Ni(II) and Co(II).44–46 Another DNAzyme, termed ‘10-23’ DNAzyme was selected with Mg(II) present, but its activity with Mn(II) is even higher.15,47,48 Furthermore, the same ‘8-17’ motif of a DNAzyme was obtained by four different laboratories with different cofactor conditions.15,49,40,50 It has a metal-dependent activity in the order Zn(II) >> Ca(II) > Mg(II) under similar conditions. Interestingly, further evaluation of this enzyme showed even higher activity with Pb(II).39 The apparent Kd values for Pb(II), Zn(II) and Mg(II) are 13.5 mM (at pH 6.0), 0.97 mM (at pH 6.0) and 10.5 mM (at pH 7.0), respectively.39 The lack of specificity in these DNAzymes poses major problems in studying metal DNA/RNA interactions and in applying DNAzymes as metal ion sensors. To solve this problem and improve metal ion specificity, a negative selection strategy can be introduced into the selection process by incubating the DNA pool with nontarget metal ions following positive selection with the target ion. In this step, any DNA sequences that have activity in the presence of other metal ions will be discarded, leaving the sequences that only function with the target ion present. This negative selection can be conducted multiple times as needed to greatly eliminate nonspecific sequences. As a demonstration of improved selectivity by negative selection, two parallel selections of Co(II)-dependent DNAzymes were carried out with and without negative selections (shown in Figure 14.3).42 In the absence of the negative selection, the resulting DNAzymes were more active with Pb(II) and Zn(II) than with Co(II). In comparison, when negative selection was incorporated, the obtained DNAzymes were much more active with Co(II) than with other metal ions. No detectable cleavage activity was observed with several other metal ions

Nucleic Acid Enzyme–Metal Ion Interactions 399 Initial pool of DNA/RNA Selections using only Co2+ DNA/RNA with Co2+ activity Positive Selection Selection using only Co2+

Negative Selection

Selection with a 'metal soup' containing competing metal ions Remove sequences that are active with other metal ions

High Co2+ activity, but with low Co2+ selectivity

Continue Co2+ selection

High Co2+ activity with high Co2+ selectivity

Figure 14.3 Schemes of parallel selections of Co(II)-dependent DNAzymes with and without negative selection. (Yi Lu, Prof; New Transition-Metal-Dependent DNAzymes as Efficient Endonucleases and as Selective Metal Biosensors. Chemistry – A European Journal, 2002, 8, 4589–4596. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

investigated including Ca(II), Mg(II) and Cd(II), and cleavage activity in the presence of Mn(II) was minimum. This result shows the effectiveness of negative selection in improving specificity of DNAzymes and further demonstrates that in vitro selection is a powerful and versatile method for generating DNA/RNAzymes with the desired properties by carefully designed strategies and well-controlled conditions.

14.4 Understanding Nucleic Acid Enzyme–Metal Ion Interactions Compared to ribozymes and protein enzymes, the understanding of the detailed interaction between metal ions and DNAzymes and how metal ions affect the activities of the DNAzymes remains very limited. For instance, while many threedimensional structures of ribozymes and protein enzymes are available,51–54 only one DNAzyme structure has been reported.55 Yet this structure is not the enzyme’s active conformation, and thus provides limited insight into the enzymatic activity. Furthermore, whereas a great deal of knowledge has been available on the roles of metal ions in protein enzymes, much less is known about that in nucleic acids. As a result, understanding of the structure and function of metal ions in DNA/RNAzymes

400


can bridge the knowledge gap between protein and nucleic acid enzymes, and also provide critical information for metal sensor development in sensing applications. Many techniques can be used to probe nucleic acid–metal ion interactions. Most of the approaches were initially designed and applied to study functions and mechanisms of ribozymes, because these catalytic RNAs are believed to play an important role in the early stage of life on Earth. One of the most intriguing objectives of nucleic acid enzyme research is the elucidation of structures of the active enzymes bound to metal ion cofactors. This goal was achieved by means of Xray crystallography. Crystal structures of metal ion–ribozyme complexes were obtained.56,57 These structures show that catalytic RNAs have complex three-dimensional structures and may contain multiple metal ion binding sites required for proper folding and activity. Even though X-ray crystallography provides direct visualization of nucleic acid enzyme structures, its application is limited to the solid crystal phase and often affected by the lack of diffraction quality crystals. Several other techniques, including nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) can be employed to study enzyme–metal ion interactions in the solution phase in close to physiological conditions. They often take advantage of paramagnetic metal ions or atoms on the enzyme, and monitor the changes in their spin states in magnetic fields during metal–enzyme interactions. For example, DeRose and colleagues used paramagnetic Mn2+ to substitute native Mg2+ in a hammerhead model ribozyme.58 The EPR signal from Mn2+ was sensitive to the environment surrounding the metal ion, thus providing small but measurable changes in the EPR signal during interaction with the enzyme. Specifically, binding of Mn2+ to a biomolecule induced a small ligand field asymmetry and increased its rotational correlation time, leading to considerable reduction in the amplitude of the six-line EPR signal from the bound Mn2+. This correlation was utilized to differentiate free and bound Mn2+ in the presence of the ribozyme. Binding affinity and the number of bound Mn2+ ions were determined using titration curves based on EPR signals. They found that the ribozyme contained four high-affinity binding sites for Mn2+ with a Kd of ∼4 mM and five other binding sites with a Kd of ∼460 mM. In a more recent report, Britt, Derose and coworkers employed pulsed EPR techniques of electron spin-echo envelope modulation (ESEEM) and electron spin-echo electron nuclear double resonance (ESE-ENDOR) for structural analysis of the paramagnetic metal ion Mn(II) bound to nucleotides and nucleic acids.59 Spin-echo EPR techniques allow the analysis of paramagnetic metal ion interactions with magnetic nuclei of ligands and nearby regions of bound macromolecules.59 A combined analysis of ESEEM and ESE-ENDOR data provided information on the number, type and spatial distribution of magnetic nuclei in the neighbourhood of the Mn2+ ion.59 Additionally, the exact metal ion ligation site on the nucleotide and the hydration number of Mn2+ in the complex were also determined. In another related report, Schiemann and coworkers employed a strategy that replaced Mn2+ ions bound to a Diels–Alder ribozyme with higher-affinity, but EPR-silent, Cd2+ ions. In this way, they were able to characterize each of the five Mn2+ binding sites individually using EPR signals.60 Given these results, it is clear that EPR and related techniques provide a powerful approach to highly precise characterization of the structural features of nucleic acid–metal ion complexes.

Nucleic Acid Enzyme–Metal Ion Interactions 401

Other methods have also been developed to study nucleic acid–metal ion interactions. Pyle and coworkers used terbium(III) to identify metal ion binding sites of a family of group II intron ribozymes.61 Tb3+ and related ions are able to cleave RNA efficiently under physiological conditions. In this study, various concentrations of Tb3+ were incubated with the ribozymes in the presence of Mg2+. The cleaved fragments of the enzymes were analysed by polyacrylamide gel electrophoresis (PAGE). Fragments at low Tb3+ concentrations revealed strong metal ion binding sites, while fragments produced by high concentrations of Tb3+ corresponded to lower-affinity binding sites. Comprehensive analysis of the gel data was successfully used to map most metal ion binding sites of the ribozymes. These Tb3+-binding sites were found to colocalize with Mg2+-binding sites based on two results. First, addition of Mg2+ at low pH effectively eliminated Tb3+ cleavage. Second, Mg2+ was shown to cleave at the same Tb3+ cleavage sites, only at much slower rates. In addition to mapping metal ion binding sites by metal-ion-mediated hydrolysis, NMR and related techniques have been useful in studying nucleic acid enzyme– metal ion interactions. Sigel, Pyle and coworkers employed NMR and two-dimensional nuclear overhauser enhancement spectroscopy (NOESY) to reveal the solution structure of the central core of a group II intron ribozyme and its interactions with Mg2+.62 It was found that binding to Mg2+ had most impact on the nucleotides in the bulge region of Domain 5. Using NMR and other techniques, Sigel and coworkers reported the solution structure of a catalytically active Domain 6 (D6) of a selfsplicing Group II intron ribozyme. A single unpaired adenosine was found to reside within the helix of D6 and was partially stacked between two flanking GU wobble pairs. A novel prominent Mg2+ binding site was identified in the major groove of this site. In a recent report, Sigel and coworkers performed paramagnetic linebroadening experiments with Mn2+ and titration studies with Mg2+ using NMR to determine the affinity constants for Mg2+ binding to five distinct sites in Domain 6 of Group II introns.63 Even though most of the aforementioned methods have been developed and used for ribozymes and RNAs, adaptation of these techniques to DNAzymes is not likely to cause major difficulties. Many efforts are underway and possibilities are being explored along this line of research. Compared to many other techniques, methods based on fluorescence resonance energy transfer (FRET) have been shown to be able to provide unique information on structural alterations of the DNA/RNAzymes upon metal ion binding in real time with minimum perturbation of the interactions. FRET is the energy transfer between two fluorophores when they have inherent spectrum overlap and are in close proximity (80 times better to Pb(II) than to the next competitive metal ion (Co(II) and Zn(II)) and hundreds times better to other metal ions tested. One problem with this sensor was that it required 4 °C for reasonably low fluorescence background and effective detection. To overcome this limitation, a new sensor design was proposed, which added an additional quencher at the 5′-end of the substrate strand.69 This internal quencher ensured a low background in the absence of lead even when the substrate was not efficiently hybridized to the enzyme. Indeed, a much improved signal-to-background ratio was obtained and sensitive lead detection could be conducted at room temperature. Two additional advantages of the fluorescent DNAzyme sensor are worth mentioning. First, since the detection is based on the kinetics of fluorescence enhancement, it is less affected by background fluorescence in environmental samples. Additionally, because catalytic reactions are responsible for signal generation, a single Pb2+ ion can catalyse the turnover of multiple fluorescent DNAzymes when even higher sensitivity is desired. This successful sensing strategy has been applied to other DNAzymes and their metal ion cofactors. One notable example is the sensor for UO2+ 2 recently developed in our lab (Figure 14.7).43 A DNAzyme specific for UO2+ was first identified via in 2 vitro selection. Following that, the same one-fluorophore-two-quencher approach was adapted to the new DNAzyme and UO2+ 2 detection was achieved with excellent sensitivity and selectivity. Most sensors based on molecular probes are not expected

406


Figure 14.7 (Plate 14) Design and performance of DNAzyme-based UO22+ sensor. Competing metals were tested from 10 mM to 1 mM, and UO22+ (the last three bars) was tested from 0 to 10 nM. Inset: Sequence, schematics and fluorescent detection of the DNAzyme-based UO22+ sensor. (Juewen Liu et al., A catalytic beacon sensor for uranium with parts-per-trillion sensitivity and millionfold selectivity, Proc. Natl. Acad. Sci. USA, 104, 2056–2061. Copyright 2007, National Academy of Sciences, USA.) (See colour plate section)

to compete or exceed instrumental analysis. However, the UO2+ 2 DNAzyme sensor was able to realize a detection limit of about 45 pM, even lower than that obtained on ICP-MS (420 pM). In addition, it has a selectivity of more than one million-fold over any other metal ions tested. Development of fluorescent sensors for paramagnetic metal ions (i.e., those metal ions with unpaired electrons) such as Cu2+ has long been a challenge, due to their intrinsic fluorescence quenching properties that often cause decreased emission upon metal binding. On the other hand, DNAzyme-based sensors are less restricted by this problem since the metal-recognition core of the enzyme can be separated from the fluorescent signalling moiety, thus minimizing the contactinduced quenching. Therefore our laboratory developed a fluorescent DNAzyme sensor for Cu(II) based on a previously reported Cu(II) DNAzyme by Breaker and coworkers.12,70,71 This DNA-cleaving DNAzyme was tagged with a fluorophore and two quenchers similar to the sensors above.72 With this sensor, Cu2+ was readily detected down to 35 nM at room temperature, and the selectivity was 2000-fold better than other tested ions. We have recently designed a modified DNAzyme sensor for Hg2+ detection, using the dual quencher method to construct a catalytic beacon. Five T-T mismatches replaced the double-strand in the loop region of the UO2+ 2 DNAzyme that is required for binding UO2+ , thus inactivating the enzyme. Addition of Hg2+ could 2 73–75 stabilize the T-T mismatches, resulting in restored activity. This sensor has a detection limit of 2.4 nM, the lowest among known Hg2+ sensors based on small or macromolecules. The examples shown above all require labelling with fluorophores and quenchers on the termini of the DNA strands. However, the ideal positions to place the fluorophore and quencher pair would be on the two nucleotides closest to the cleavage site on the substrate strand. In this way, the quencher on the enzyme strand would no longer be necessary since the quencher and fluorophore on the substrate could stay close enough for minimum background. Cleavage of the substrate would then separate this pair and generate enhanced fluorescence. One concern for this design is possible activity decrease due to these two modifications near the active

DNAzymes as Metal Ion Sensors

407

site. To investigate the feasibility of internal fluorophore/quencher labelling on DNAzyme, Li and colleagues conducted systematic studies by placing different dye/quencher pairs at various locations on the substrate.76 The results indicated that generally, the closer the dye/quencher to the cleavage site, the higher the fluorescence enhancement upon lead-induced enzymatic cleavage. However, the reaction rates were also lowered as a result. Given these observations, internal dye/quencher labelling for making DNAzyme sensors is feasible, but the design needs to be optimized. Instead of developing sensing strategies based on known DNAzymes, one appealing alternative would be directly selecting DNAzymes with a built-in signal transduction mechanism. Li and coworkers have realized this experimentally by modifying the general in vitro selection scheme.16,77 Compared to conventional selection, these authors added a few extra steps to each round to introduce a short nucleic acid strand to the DNAs in the pool before reaction in the presence of metal ions. This short strand contained an adenine ribonucleotide (rA) as the cleavage site, and a fluorophore and quencher labelling the two nucleotides closest to rA. When incubated with target metal ions, only those sequences that could cleave the rA site were collected and enriched. The final products of this selection were DNAzymes that were able to function normally with the internal labels and generate greatly enhanced fluorescence upon addition of target metal ions.

14.5.2 Colorimetric Sensors Based on DNAzymes Fluorescence-based DNAzyme sensors are excellent for metal ion detection due to their high sensitivity and selectivity. They also provide a solution to on-site metal analysis when equipped with portable fluorometers. However, to further reduce the cost and simplify the process of metal sensing, thus making it readily accessible to a much wider range of fields, including clinics, battlegrounds and even households, new strategies need to be employed. We have made progress in this direction with the development of colorimetric metal ion sensors. In a colorimetric sensor, a colour change is observed in the presence of target analytes. Compared to other techniques, colorimetric sensors can minimize or even eliminate the need for analytical instruments, making on-site detection much more convenient. Some common colour-reporting groups include small organic dyes, conjugated polymers and metallic nanoparticles. Among them, metallic nanoparticles display strong distance-dependent optical properties and very high extinction coefficients.78 Therefore, we designed our first-generation colorimetric lead sensor based on DNAzymes and DNA-functionalized gold nanoparticles (AuNPs).79 Briefly, the substrate of ‘8-17’ DNAzyme was extended at both ends with a short DNA strand of the same sequence. In this way, when AuNPs functionalized with the complementary DNA (cDNA) of the short extension were introduced, the AuNPs became literally crosslinked by the extended substrate to produce aggregated nanoparticle clusters (Figure 14.8). Such AuNP aggregates display a blue colour, due to their strong absorption of light with wavelengths around 700 nm. The ‘8-17’ DNAzyme was then introduced to hybridize with the remaining sequence of substrate. When Pb(II) was present, the enzymatic reaction was initiated to cleave the

408


Figure 14.8 DNAzyme and AuNP-based colorimetric Pb2+ detection: (A) Pb2+-directed assembly of DNAzyme-linked AuNPs aligned in a head-to-tail manner; (B) colour of the AuNPs in the presence of different divalent metal ions. (Reprinted with permission from J. Am. Chem. Soc., 2003, 125, 6642–6643. Copyright 2003, American Chemical Society.)

substrate sequence, leaving a broken linkage between the AuNPs. Consequently, the dispersed AuNPs displayed a deep red colour due to a shifted absorption to around 522 nm. With increasing level of lead ions, a blue-to-red colour shift was clearly discernible by the naked eye. The detection of lead ions could also be realized more accurately by UV/Vis spectroscopy measurements. The ratio of extinction at 522 nm over 700 nm was used to quantify Pb(II) since it rose with increasing concentration of lead ions. A detection limit of ∼100 nM was achieved and the sensor was able to detect lead extracted from paint. An important benefit of the AuNP-based DNAzyme sensor is the easy implementation of tuneable dynamic ranges. A sensor design that allows simple tuning of dynamic range is highly desirable to match the requirements of different applications. For example, the US EPA defines the toxic threshold for lead in paint as ∼2 mM, while that in drinking water is 72 nM. Additionally, regulations on sensitivity of metal ion detection can change over time. It is often not economical or even practical to redesign and remake new sensors to meet these regulation changes, thus giving extra significance to tuneable sensors in practical applications. Taking advantage of the knowledge obtained from biochemical assays and the multiple turnover property of the DNAzyme, we have demonstrated a general scheme for tuning the dynamic range of the colorimetric Pb2+ sensor system.79 DNAzyme activity was abolished by changing the G•T wobble pair to a G-C Watson–Crick pair (Figure 14.9, left).80 This mutated DNAzyme could still assemble AuNPs similarly to the native DNAzyme. When a ratio of 5 : 95 was adopted for the active enzyme (17E) versus inactive enzyme (17Ec), the Pb2+-sensitive range shifted to ∼one order of magnitude higher Pb2+ concentrations (Figure 14.9, right). This tuning property is unique and useful because it allows detection of Pb2+ over a wide concentration range without redesigning the sensor.

DNAzymes as Metal Ion Sensors

409

Figure 14.9 (Left) Secondary structure of the 8-17 DNAzyme for lead ion detection; (Right) dynamic ranges of the colorimetric sensors with different enzyme compositions. (Reprinted with permission from J. Am. Chem. Soc., 2003, 125, 6642–6643. Copyright 2003, American Chemical Society.)

In the first generation colorimetric sensor for lead, an additional heating and cooling step was needed before the DNA-functionalized AuNPs could be effectively crosslinked by the extended substrate strand. In fact, detection of lead ions was also conducted after this annealing step, where absence of lead produced a blue product and lead ions gave a red colour. This problem most likely arose from the alignment of the AuNPs. Based on the design, the substrate hybridized at both ends with two AuNPs tagged with the same cDNA sequence. This means these two AuNPs had to take a head-to-tail alignment, which induced a significant steric effect to prevent the crosslinking. To solve this problem, a tail-to-tail design was adopted and aggregation of AuNP was observed without the annealing step.81 Additionally, the size of the AuNP was changed from 13 nm to 42 nm, which greatly reduced the time needed to see a colour change to 5 min.81 Later, a true ‘turn-on’ colorimetric sensor was developed in which addition of lead ions could directly break up preaggregated AuNPs to produce a distinctive blue-to-red colour shift.82,83 To accelerate Pb2+induced disassembly, small pieces of DNA (called invasive DNA) complementary to the cleaved substrate strands were employed to facilitate AuNP release. Under optimized conditions, a blue-to-red colour shift was observed in 5 min at room temperature in the presence of Pb2+.82 Recently, we further optimized the disassembly conditions by designing asymmetric substrate binding arms so that disassembly occurred right after cleavage without the use of invasive DNAs.83,84 14.5.3 Other DNAzyme-Based Metal Ion Sensors The DNA backbone of the DNAzymes allows convenient modifications to the enzyme to tag a signalling moiety, facilitate immobilization of the enzyme or add additional functionality. Therefore, a variety of other detection strategies can be developed for metal ion sensing based on DNAzymes. In one such attempt, Plaxco and coworkers immobilized the lead-dependent 8-17 DNAzyme onto a gold electrode through the 5′-end of the enzyme strand and attached a methylene-blue (MB) group to the 3′-end.85 The DNAzyme remained rigid when hybridized with the substrate strand, separating the MB group from the surface and giving a low rate of electron transfer. In the presence of Pb2+, the substrate was cleaved and released,

410


making the immobilized enzyme more flexible. Therefore, enhanced electron transfer was observed. Similar to other lead DNAzyme-based sensors, this electrochemical sensor was highly selective for Pb2+. A detection limit of 300 nM was reported and the selectivity was comparable to those of the fluorescent and colorimetric sensors based on the same DNAzyme. By immobilizing DNAzymes on a solid surface, an alternative sensing platform can be built. Notably, this scheme will allow sensor regeneration, as well as long-term storage of the sensor. In one report, the lead-specific 8-17 DNAzyme was covalently attached to gold surfaces.86 In the presence of Pb2+, a fluorophore-labelled substrate fragment was cleaved and released into solution for detection. Thanks to a very low background, a detection limit of 1 nM was achieved. In another report, surfaceimmobilized DNAzyme retained its activity even after storing in a dried state for 30 days at room temperature.87 A detailed characterization of this system was later carried out.88 These studies will facilitate the application of DNAzymes in sensor arrays for analysing complex targets.

14.6 Summary Studies of DNAzymes represent an actively developing interdisciplinary field involving many aspects of chemistry research, including biological, inorganic, physical and analytical chemistry. In some ways, this field has similarities to the study of ribozymes. However, it also holds many new possibilities in exploring structural mechanisms of functional nucleic acids and their biosensing applications. Despite recent progresses, there is still a clear lack of understanding of how DNAzymes carry out their catalytic functions, in comparison with our understanding of protein and RNA enzymes. The three-dimensional structure of the DNAzyme in its active form containing its cofactors has yet to be obtained. The structure features of the DNAzyme will greatly help elucidate the underlying mechanism of DNAzyme–cofactor interactions. In addition, analysis of the function of DNAzyme and its interactions with the surrounding microenvironment at the molecular level using EPR, NMR, smFRET, etc., is also in great need because they can provide valuable information toward the ultimate understanding of DNAzyme activities. On the other hand, given the current level of knowledge of DNAzymes, we have already been able to develop DNAzymes with the desired properties in the presence of specific cofactors using versatile in vitro selection. It will be of great significance to actively extend the family of DNAzymes in pursuit of a wider range of catalytic reactions that they can carry out and an increasing number of cofactors that they can recruit. Due to the fact that DNAs are inherently more stable than RNAs, these newly developed DNAzymes are more suitable to be engineered into practical biosensors for many different targets of interest. Additionally, with a large quantity of DNAzyme-based sensors available, comprehensive sensor arrays can be built for highly reliable analysis of complex samples. Given the increasing demand for convenient, economical yet effective sensors, we will continue to witness rapid development and advancment of the DNAzyme sensing field.

References

411

Acknowledgements The Lu group research described in this review has been supported by the Department of Energy (DE-FG02-08ER64568) and the National Science Foundation (CTS-0120978, DMR-0117792, and DMI-0328162).

References 1. Breaker, R.R.; Joyce, G.F.; A DNA enzyme that cleaves RNA; Chem. Biol., 1994, 1(4), 223–229. 2. Ellington, A.D.; Szostak, J.W.; In vitro selection of RNA molecules that bind specific ligands; Nature, 1990, 346(6287), 818–822. 3. Tuerk, C.; Gold, L.; Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase; Science, 1990, 249(4968), 505–510. 4. Beaudry, A.A.; Joyce, G.F.; Directed evolution of an RNA enzyme; Science, 1992, 257(5070), 635–641. 5. Bartel, D.P.; Szostak, J.W.; Isolation of new ribozymes from a large pool of random sequences; Science, 1993, 261(5127), 1411–1418. 6. Breaker, R.R.; DNA enzymes; Nat. Biotechnol., 1997, 15(5), 427–431. 7. Sen, D.; Geyer, C.R.; DNA enzymes; Curr. Opin. Chem. Biol., 1998, 2(6), 680–687. 8. Breaker, R.R.; Catalytic DNA: in training and seeking employment; Nat. Biotechnol., 1999, 17(5), 422–423. 9. Breaker, R.R.; Making catalytic DNAs; Science, 2000, 290(5499), 2095–2096. 10. Breaker, R.R.; Joyce, G.F.; A DNA enzyme with Mg2+-dependent RNA phosphoesterase activity; Chem. Biol., 1995, 2(10), 655–660. 11. Faulhammer, D.; Famulok, M.; Characterization and divalent metal-ion dependence of in vitro selected deoxyribozymes which cleave DNA/RNA chimeric oligonucleotides; J. Mol. Biol., 1997, 269(2), 188–202. 12. Carmi, N.; Shultz, L.A.; Breaker, R.R.; In vitro selection of self-cleaving DNAs; Chem. Biol., 1996, 3(12), 1039–1046. 13. Geyer, C.R.; Sen, D.; Evidence for the metal-cofactor independence of an RNA phosphodiester-cleaving DNA enzyme; Chem. Biol., 1997, 4(8), 579–593. 14. Roth, A.; Breaker, R.R.; An amino acid as a cofactor for a catalytic polynucleotide; Proc. Natl. Acad. Sci. USA, 1998, 95(11), 6027–6031. 15. Santoro, S.W.; Joyce, G.F.; A general purpose RNA-cleaving DNA enzyme; Proc. Natl. Acad. Sci. USA, 1997, 94(9), 4262–4266. 16. Liu, Z.; Mei, S.H.J.; Brennan, J.D.; Li, Y.; Assemblage of signaling DNA enzymes with intriguing metal-ion specificities and pH dependences; J. Am. Chem. Soc., 2003, 125(25), 7539–7545. 17. Cuenoud, B.; Szostak, J.W.; A DNA metalloenzyme with DNA ligase activity; Nature, 1995, 375(6532), 611–614. 18. Li, Y.; Breaker, R.R.; In vitro selection of kinase and ligase deoxyribozymes; Methods, 2001, 23(2), 179–190. 19. Coppins, R.L.; Silverman, S.K.; Rational modification of a selection strategy leads to deoxyribozymes that create native 3′–5′ RNA linkages; J. Am. Chem. Soc., 2004, 126(50), 16426–16432. 20. Li, Y.; Breaker, R.R.; Phosphorylating DNA with DNA; Proc. Natl. Acad. Sci. USA, 1999, 96(6), 2746–2751. 21. Burmeister, J.; von Kiedrowski, G.; Ellington, A.D.; Cofactor-assisted self-cleavage in DNA libraries with a 3′–5′-phosphoramidate bond; Angew. Chem. Int. Ed., 1997, 36(12), 1321–1324.

412


22. Li, Y.; Sen, D.; A catalytic DNA for porphyrin metalation; Nat. Struct. Biol., 1996, 3(9), 743–747. 23. Bertini, I.; Gray, H.B.; Lippard, S.J.; Valentine, J.S., Eds; Bioinorganic Chemistry, University Science Books, Sausalito, CA, 1994, 611. 24. Lippard, S.J.; Berg, J.M., Eds, Principles of Bioinorganic Chemistry, University Science Books, Mill Valley, CA, 1994, 411. 25. Holm, R.H.; Kennepohl, P.; Solomon, E.I.; Structural and functional aspects of metal sites in biology; Chem. Rev., 1996, 96(7), 2239–2314. 26. Yarus, M.; How many catalytic RNAs? Ions and the Cheshire cat conjecture; FASEB J., 1993, 7(1), 31–39. 27. Pyle, A.M.; Ribozymes: a distinct class of metalloenzymes; Science, 1993, 261(5122), 709–714. 28. Pan, T.; Long, D.M.; Uhlenbeck, O.C.; Divalent metal ions in RNA folding and catalysis. In The RNA World, Gesteland, R.F. and Atkins, J.F., Eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1993, 271–302. 29. Feig, A.L.; Uhlenbeck, O.C.; The role of metal ions in RNA biochemistry. In The RNA World, 2 edn.; Gesteland, R.F., Cech, T.R. and Atkins, J.F., Eds, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1999, 287–319. 30. Lu, Y.; New transition metal-dependent DNAzymes as efficient endonucleases and as selective metal biosensors; Chem. Eur. J., 2002, 8, 4588–4596. 31. Sigel, R.K.O.; Pyle, A.M.; Alternative roles for metal ions in enzyme catalysis and the implications for ribozyme chemistry; Chem. Rev., 2007, 107(1), 97. 32. Freisinger, E.; Sigel, R.K.O.; From nucleotides to ribozymes – A comparison of their metal ion binding properties; Coord. Chem. Rev., 2007, 251(13–14), 1834–1851. 33. Rajendran, M.; Ellington, A.D.; Selection of fluorescent aptamer beacons that light up in the presence of zinc; Anal. Bioanal. Chem., 2007, 390(4), 1067–1075. 34. Huizenga, D.E.; Szostak, J.W., A DNA aptamer that binds adenosine and ATP; Biochem., 1995, 34(2), 656–665. 35. Stojanovic, M.N.; de Prada, P.; Landry, D.W.; Fluorescent sensors based on aptamer selfassembly; J. Am. Chem. Soc., 2000, 122(46), 11547–11548. 36. Bock, L.C.; Griffin, L.C.; Latham, J.A.; Vermaas, E.H.; Toole, J.J.; Selection of singlestranded DNA molecules that bind and inhibit human thrombin; Nature, 1992, 355(6360), 564–566. 37. Cerchia, L.; Duconge, F.; Pestourie, C.; Boulay, J.; Aissouni, Y.; Gombert, K.; Tavitian, B.; de Franciscis, V.; Libri, D.; Neutralizing aptamers from whole-cell SELEX inhibit the RET receptor tyrosine kinase; PLoS Biol, 2005, 3(4), e123. 38. Shangguan, D.; Li, Y.; Tang, Z.; Cao, Z.C.; Chen, H.W.; Mallikaratchy, P.; Sefah, K.; Yang, C.J.; Tan, W.; Aptamers evolved from live cells as effective molecular probes for cancer study; Proc. Natl. Acad. Sci. USA, 2006, 103(32), 11838–11843. 39. Li, J.; Lu, Y.; A highly sensitive and selective catalytic DNA biosensor for lead ions; J. Am. Chem. Soc., 2000, 122(42), 10466–10467. 40. Li, J.; Zheng, W.; Kwon, A.H.; Lu, Y.; In vitro selection and characterization of a highly efficient Zn(II)-dependent RNA-cleaving deoxyribozyme; Nucleic Acids Res., 2000, 28(2), 481–488. 41. Santoro, S.W.; Joyce, G.F.; Sakthivel, K.; Gramatikova, S.; Barbas, C.F., III; RNA cleavage by a DNA enzyme with extended chemical functionality; J. Am. Chem. Soc., 2000, 122(11), 2433–2439. 42. Bruesehoff, P.J.; Li, J.; Augustine, A.J.; Lu, Y.; Improving metal ion specificity during in vitro selection of catalytic DNA; Comb. Chem. High T. Scr., 2002, 5(4), 327–335. 43. Liu, J.; Brown, A.K.; Meng, X.; Cropek, D.M.; Istok, J.D.; Watson, D.B.; Lu, Y.; A catalytic beacon sensor for uranium with parts-per-trillion sensitivity and millionfold selectivity; Proc. Natl. Acad. Sci. USA, 2007, 104(7), 2056. 44. Ciesiolka, J.; Gorski, J.; Yarus, M.; Selection of an RNA domain that binds Zn2+; RNA, 1995, 1(5), 538–550. 45. Ciesiolka, J.; Yarus, M.; Small RNA-divalent domains; RNA, 1996, 2(8), 785–793.

References

413

46. Kawakami, J.; Imanaka, H.; Yokota, Y.; Sugimoto, N.; In vitro selection of aptamers that act with Zn2+; J. Inorg. Biochem., 2000, 82(1–4), 197–206. 47. Santoro, S.W.; Joyce, G.F.; Mechanism and utility of an RNA-cleaving DNA enzyme; Biochem., 1998, 37(38), 13330–13342. 48. Sugimoto, N.; Okumoto, Y.; Ohmichi, T.; Effect of metal ions and sequence of deoxyribozymes on their RNA cleavage activity; J. Chem. Soc., Perkin Trans.2, 1999, (7), 1381–1386. 49. Faulhammer, D.; Famulok, M.; The Ca2+ ion as a cofactor for a novel RNA-cleaving deoxyribozyme; Angew. Chem., Int. Ed., 1996, 35(23/24), 2837–2841. 50. Cruz, R.P.G.; Withers, J.B.; Li, Y.; Dinucleotide junction cleavage versatility of 8-17 deoxyribozyme; Chem. Biol., 2004, 11(1), 57–67. 51. Yeh, J.I.; Charrier, V.; Paulo, J.; Hou, L.H.; Darbon, E.; Claiborne, A.; Hol, W.G.J.; Deutscher, J.; Structures of enterococcal glycerol kinase in the absence and presence of glycerol: Correlation of conformation to substrate binding and a mechanism of activation by phosphorylation; Biochem., 2004, 43(2), 362–373. 52. Guo, F.; Gooding, A.R.; Cech, T.R.; Structure of the Tetrahymena ribozyme: base triple sandwich and metal ion at the active site; Mol. Cell, 2004, 16(3), 351–362. 53. Ke, A.; Zhou, K.; Ding, F.; Cate, J.H.; Doudna, J.A.; A conformational switch controls hepatitis delta virus ribozyme catalysis; Nature, 2004, 429(6988), 201–205. 54. Golden, B.L.; Kim, H.; Chase, E.; Crystal structure of a phage Twort group I ribozymeproduct complex; Nat. Struct. Mol. Biol., 2005, 12(1), 82–89. 55. Nowakowski, J.; Shim, P.J.; Prasad, G.S.; Stout, C.D.; Joyce, G.F.; Crystal structure of an 82-nucleotide RNA-DNA complex formed by the 10-23 DNA enzyme; Nat. Struct. Biol., 1999, 6(2), 151–156. 56. Scott, W.G.; Murray, J.B.; Arnold, J.R.P.; Stoddard, B.L.; Klug, A.; Capturing the structure of a catalytic RNA intermediate: the hammerhead ribozyme; Science, 1996, 274(5295), 2065–2069. 57. Cate, J.H.; Hanna, R.L.; Doudna, J.A.; A magnesium ion core at the heart of a ribozyme domain; Nat. Struct. Biol., 1997, 4(7), 553–558. 58. Horton, T.E.; Clardy, D.R.; DeRose, V.J.; Electron paramagnetic resonance spectroscopic measurement of Mn2+ binding affinities to the hammerhead ribozyme and correlation with cleavage activity; Biochem., 1998, 37(51), 18094–18101. 59. Hoogstraten, C.G.; Grant, C.V.; Horton, T.E.; DeRose, V.J.; Britt, R.D.; Structural analysis of metal ion ligation to nucleotides and nucleic acids using pulsed EPR spectroscopy; J. Am. Chem. Soc., 2002, 124(5), 834–842. 60. Kisseleva, N.; Kraut, S.; Jaschke, A.; Schiemann, O.; Characterizing multiple metal ion binding sites within a ribozyme by cadmium-induced EPR silencing; HFSP J., 2007, 1(2), 127–136. 61. Sigel, R.K.; Vaidya, A.; Pyle, A.M.; Metal ion binding sites in a group II intron core; Nat. Struct. Biol., 2000, 7(12), 1111–1116. 62. Sigel, R.K.O.; Sashital, D.G.; Abramovitz, D.L.; Palmer, A.G.; Butcher, S.E.; Pyle, A.M.; Solution structure of domain 5 of a group II intron ribozyme reveals a new RNA motif; Nat. Struct. Mol. Biol., 2004, 11(2), 187–192. 63. Erat, M.C.; Sigel, R.K.; Determination of the intrinsic affinities of multiple site-specific Mg(2+) ions coordinated to domain 6 of a group II intron ribozyme; Inorg. Chem., 2007, 46(26), 11224–11234. 64. Liu, J.; Lu, Y.; FRET study of a trifluorophore-labeled DNAzyme; J. Am. Chem. Soc., 2002, 124(51), 15208–15216. 65. Kim, H.-K.; Liu, J.; Li, J.; Nagraj, N.; Li, M.; Pavot, C.M.-B.; Lu, Y.; Metal-dependent global folding and activity of the 8-17 DNAzyme studied by fluorescence resonance energy transfer; J. Am. Chem. Soc., 2007, 129(21), 6896–6902. 66. Kim, H.K.; Rasnik, I.; Liu, J.; Ha, T.; Lu, Y.; Dissecting metal ion-dependent folding and catalysis of a single DNAzyme; Nat. Chem. Biol., 2007, 3(12), 763–768. 67. Allen, L.H.; Anemia and iron deficiency: effects on pregnancy outcome; Am. J. Clin. Nutr., 2000, 71(5), 1280S–1284S.

414


68. Emerit, J.; Beaumont, C.; Trivin, F.; Iron metabolism, free radicals, and oxidative injury; Biomed. Pharmacother., 2001, 55(6), 333–339. 69. Liu, J.; Lu, Y.; Improving sluorescent DNAzyme biosensors by combining inter- and intramolecular quenchers; Anal. Chem., 2003, 75(23), 6666–6672. 70. Carmi, N.; Balkhi, H.R.; Breaker, R.R.; Cleaving DNA with DNA; Proc. Natl. Acad. Sci. USA, 1998, 95(5), 2233–2237. 71. Carmi, N.; Breaker, R.R.; Characterization of a DNA-cleaving deoxyribozyme; Bioorg. Med. Chem., 2001, 9(10), 2589–2600. 72. Liu, J.; Lu, Y.; A DNAzyme catalytic beacon sensor for paramagnetic Cu2+ ions in aqueous solution with high sensitivity and selectivity; J. Am. Chem. Soc., 2007, 129(32), 9838–9839. 73. Ono, A.; Togashi, H.; Molecular sensors: Highly selective oligonucleotide-based sensor for mercury(II) in aqueous solutions; Angew. Chem., Int. Ed., 2004, 43(33), 4300–4302. 74. Miyake, Y.; Togashi, H.; Tashiro, M.; Yamaguchi, H.; Oda, S.; Kudo, M.; Tanaka, Y.; Kondo, Y.; Sawa, R.; Fujimoto, T.; Machinami, T.; Ono, A.; MercuryII-mediated formation of thymine-HgII-thymine base pairs in DNA duplexes; J. Am. Chem. Soc., 2006, 128(7), 2172. 75. Tanaka, Y.; Oda, S.; Yamaguchi, H.; Kondo, Y.; Kojima, C.; Ono, A.; 15N-15N J-coupling Across HgII: Direct observation of HgII-mediated T-T base pairs in a DNA duplex; J. Am. Chem. Soc., 2007, 129(2), 244. 76. Chiuman, W.; Li, Y.; Efficient signaling platforms built from a small catalytic DNA and doubly labeled fluorogenic substrates; Nucleic Acids Res., 2007, 35(2), 401–405. 77. Mei, S.H.J.; Liu, Z.; Brennan, J.D.; Li, Y.; An efficient RNA-cleaving DNA enzyme that synchronizes catalysis with fluorescence signaling; J. Am. Chem. Soc., 2003, 125(2), 412–420. 78. Mirkin, C.A.; Letsinger, R.L.; Mucic, R.C.; Storhoff, J.J.; A DNA-based method for rationally assembling nanoparticles into macroscopic materials; Nature, 1996, 382(6592), 607–609. 79. Liu, J.; Lu, Y.; A colorimetric lead biosensor using DNAzyme-directed assembly of gold nanoparticles; J. Am. Chem. Soc., 2003, 125(22), 6642–6643. 80. Brown, A.K.; Li, J.; Pavot, C.M.B.; Lu, Y.; A lead-dependent DNAzyme with a two-step mechanism; Biochem., 2003, 42(23), 7152–7161. 81. Liu, J.; Lu, Y.; Accelerated color change of gold nanoparticles assembled by DNAzymes for simple and fast colorimetric Pb2+ detection; J. Am. Chem. Soc., 2004, 126(39), 12298–12305. 82. Liu, J.; Lu, Y.; Stimuli-responsive disassembly of nanoparticle aggregates for light-up colorimetric sensing; J. Am. Chem. Soc., 2005, 127(36), 12677–12683. 83. Liu, J.; Lu, Y.; Design of asymmetric DNAzymes for dynamic control of nanoparticle aggregation states in response to chemical stimuli; Org. Biomol. Chem., 2006, 4(18), 3435–3441. 84. Liu, J.; Wernette, D.P.; Lu, Y.; Proofreading and error removal in a nanomaterial assembly; Angew. Chem., Int. Ed., 2005, 44(44), 7290–7293. 85. Xiao, Y.; Rowe, A.A.; Plaxco, K.W.; Electrochemical detection of parts-per-billion lead via an electrode-bound DNAzyme assembly; J. Am. Chem. Soc., 2007, 129(2), 262. 86. Swearingen, C.B.; Wernette, D.P.; Cropek, D.M.; Lu, Y.; Sweedler, J.V.; Bohn, P.W.; Immobilization of a catalytic DNA molecular beacon on Au for Pb(II) detection; Anal. Chem., 2005, 77(2), 442–448. 87. Wernette, D.P.; Swearingen, C.B.; Cropek, D.M.; Lu, Y.; Sweedler, J.V.; Bohn, P.W.; Incorporation of a DNAzyme into Au-coated nanocapillary array membranes with an internal standard for Pb(II) sensing; Analyst, 2006, 131(1), 41. 88. Wernette, D.P.; Mead, C.; Bohn, P.W.; Lu, Y.; Surface immobilization of catalytic beacons based on ratiometric fluorescent DNAzyme sensors: A systematic study; Langmuir, 2007, 23(18), 9513–9521.

15 Two-Metal-Ion-Dependent Catalysis in Nucleic Acid Enzymes Wei Yang

15.1 Chemistry of Nucleic Acid Synthesis, Cleavage and Strand Transfer DNA and RNA are synthesized by polymerases in the 5′ to 3′ direction. The hydroxyl group at the 3′ end of a primer strand attacks the a-phosphate of an incoming (deoxy)nucleoside triphosphate ((d)NTP) to incorporate the (d)NMP into the primer strand and release a pyrophosphate molecule (Figure 15.1A). Therefore, nucleic acid synthesis transfers a phosphodiester bond from a (d)NTP to the 3′-OH (the nucleophile) of a growing primer strand. All polymerases have been shown to require Mg2+ ions for catalysis.1,2 The same phosphoryl transfer reaction occurs during DNA or RNA cleavage except that the phosphate being attacked is the backbone of a nucleic acid and the nucleophile is a hydroxyl group from water, protein or ribose (Figure 15.1B,C). When a water molecule serves as a nucleophile, the nucleic acid is hydrolysed. When the nucleophile is a hydroxyl group of protein sidechains, e.g. Ser and Tyr, as in DNA recombination,3–5 a 2′-OH of ribose, as in RNA cleavage6 or a 3′-OH of RNA or DNA, as in splicing and integration,7,8 the phosphorylated product is covalently linked to the protein or nucleotide (Figure 15.1B). The chemistry of a phosphoryl transfer reaction begins with deprotonation and activation of a nucleophile and finishes with protonation of the leaving group. The mechanism for phosphoryl transfer reactions in nucleic acids has been shown to be the SN2 type, involving a pentacovalent phosphate intermediate (phosphate anion) and inversion of the stereo Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

416

Two-Metal-Ion-Dependent Catalysis in Nucleic Acid Enzymes

Figure 15.1 Phosphoryl transfer reactions in nucleic acid synthesis, cleavage and recombination: (A) DNA and RNA synthesis. An incoming (d)NTP is incorporated into the primer strand. The 3′-OH of the primer strand is the nucleophile and attacks the a-phosphate of (d)NTP (highlighted in red). Pyrophosphate (PPi) is the leaving group. (Reproduced from Yang, W. et al., Making and breaking nucleic acids: two-Mg2+-ion catalysis and substrate specificity, Mol. Cell, 2006, 22, 5–13); (B) Cleavage. Scissile phosphate (highlighted in red) is transferred to water (hydrolysis), a nucleotide (as in splicing), a Tyr or Ser protein side chain (as in DNA recombination), or another DNA or RNA strand (as in RNA splicing or DNA recombination) depending on the nucleophile. The products are 3′-OH and 5′-PO4. (C) Same as (B) except the products are 5′-OH and 3′-PO4

configuration at the phosphorus.9–11 A nucleophile can attack the scissile phosphate from either side of the phosphate, and thus either the 3′- or 5′- OH can be the leaving group4,5,12 (Figure 15.1B,C). To date, except for catalysis that results in a 3′-phosphate and a 5′-OH,4,6 one or two metal ions are required for the phosphoryl transfer reaction. 15.1.1 Basic Properties of Mg2+ and Divalent Cations Mg2+ and Ca2+ are the most abundant divalent cations in living organisms,13,14 and Mg2+ is the most abundant divalent cation inside cells.15,16 Others, like Fe2+, and Zn2+ and Cu2+ are widespread, and Mn2+ and Ni2+ are essential, but found at low concentrations.14 The foremost quality of these divalent cations is the high density of positive charge, which makes them efficient for charge neutralization of phospholipids and nucleic acids. The second common property is hydration and the specific ligand requirement. By the atomic or covalent radius Ca2+ is the largest (1.9 Å), Mg2+ the second (1.6 Å) and the rest are similarly smaller (1.3 Å). However, when hydrated, Mg2+ is larger than Ca2+ because Mg2+ is associated with multiple shells of ligand water molecules.16 In biological systems these metal ions never exist without water

DNA and RNA Polymerases

417

Figure 15.2 Preferred coordination geometry of divavlent cations (Me2+): (A) Mg2+ prefers octahedral coordination with six ligands; (B) Zn2+ prefers tetrahedral configuration with four ligands. In protein structures, both Mg2+ and Zn2+ have been observed with five ligands in nonstandard geometry

or ligand. Therefore ionic (Shannon) radii17 are inapplicable. The most common ligand coordination geometry is octahedral or tetrahedral18 (Figure 15.2). Mg2+ prefers to have six inner-sphere ligands arranged in an octahedral configuration, and so does Fe2+.18 Ca2+ and Zn2+ can have octahedral coordination, but Zn2+ is frequently coordinated by four ligands in a tetrahedron,19 and Ca2+ by seven, eight or even nine ligands.20 The number and length of ligand bonds are empirically determined and shown to vary within a very narrow range.18 These metal ions are far more sensitive to the ligand type and geometry than hydrogen bonds or salt bridges between oxygen and nitrogen atoms. Each of these divalent cations has been found to be essential for catalysis by at least one enzyme, e.g. Fe2+ in nonheme iron enzymes (oxygenases and demethylases),21 Zn2+ in deacetylase and protease,22,23 Ni2+ in urease,24 but the Mg2+ ion is most frequently associated with nucleic acid enzymes.13 This is perhaps because of its abundance and solubility, stability in redox potentials compared with Mn2+, Fe2+ and Cu2+, its small size relative to Ca2+ and its rigid coordination geometry compared to the transition metals Fe2+, Cu2+, Ni2+ and Zn2+.16 The ligand bond of the Mg2+ ion is empirically determined to be 2.07 Å.18 The unusual hydration property of Mg2+ ions may also be a factor for its catalytic role. Mg2+ ions exhibit extremely slow exchange rates of inner-shell water molecules16,25 and often retain a couple of water ligands when coordinated by organic or macromolecules.18 The pKa value of a water molecule is reduced from 16.0 in free solution to 11.4 when associated with Mg2+.26

15.2 All DNA and RNA Polymerases Require Two Mg2+ Ions for Catalysis To date, six families of DNA polymerases have been identified, and each contains a number of conserved sequence motifs.27,28 The first four families, A to D,

418


Figure 15.3 (Plate 15) Three types of catalytic centres found in DNA and RNA polymerases; (A) Phage T7 DNA polymerase represents the pol I-like active centre, which is found in the A, B and Y-family DNA polymerases, retroviral reverse transcriptases, phage RNA polymerases and viral RNA-dependent RNA polymerases. The catalytic residues (highlighted in blue and red) include two Asps and one Glu on two adjacent antiparallel strands (highlighted in cyan); (B) Human DNA pol b represents the pol b-like catalytic centre, which is found in the C and X-family DNA polymerases, poly-rA polymerases and CCA-adding enzymes. The catalytic residues include three conserved Asps on two adjacent parallel strands; (C) S. cerevisiae RNA pol II represents all bacterial and eukaryotic DNA-dependent RNA polymerases. The catalytic core consists of a b-barrel with three catalytically essential Asps on a loop (highlighted in cyan). All of these polymerases contain two absolutely conserved Asps, which together with the a-phosphate of (d)NTP jointly coordinate the two Mg2+ (shown as green spheres) (See colour plate section)

encompass all polymerases for DNA replication and a few for repair, while the remaining polymerases in the X and Y families are involved in DNA repair exclusively. Crystal structures of A, B, C, X and Y family polymerases have been determined.1,29–33 The structures of the catalytic domains and the active sites of these DNA polymerases fall into two classes, the pol I-like, which includes the A, B and Y family polymerases (Figure 15.3A), and the pol b-like, which includes the X and C family polymerases (Figure 15.3B). The catalytic domains of both classes contain a b-sheet and several a-helices, but the topology and locations of the catalytic residues differ. In the pol I-like polymerases, the central b-sheet is antiparallel and the catalytic residues are located on two adjacent strands, 1 and 3 (Figure 15.3A). In the pol b-like polymerases, the b-sheet is mixed parallel and antiparallel and the catalytic residues are located on two adjacent parallel strands, 2 and 5 (Figure 15.3B). Interestingly, the two classes of catalytic domain appear to have evolved in parallel and don’t segregate with the catalytic efficiency and accuracy, or type of DNA synthesis (replicative versus repair). Furthermore, reverse transcriptases (RT) and RNA-dependent RNA polymerases encoded by viruses possess pol I-like catalytic domains,29,34 and template-independent poly-A polymerases and CCA-adding enzymes are homologous to the pol b-like polymerases.35–37 RNA polymerases I, II and III, which transcribe DNA to RNA, share a conserved catalytic domain quite different from the DNA polymerases, CCA-adding enzymes, and poly-A and RNA-dependent RNA polymerases. The signature amino

DNA and RNA Polymerases

419

acid sequence of RNA polymerases, NADFDGD, is conserved among RNA polymerases from bacteria to high eukaryotes.38 This motif is located on a loop connecting two strands in a b-barrel domain, as observed in Taq and yeast RNA pol II38,39 (Figure 15.3C). This forms the third class of polymerase catalytic centres. Regardless of the differences in the tertiary structure and catalytic residue locations, all three classes of DNA and RNA polymerases appear to require two metal ions in the active site for substrate binding and the phosphoryl transfer reaction. In the best-studied DNA polymerase, pol b (X family), the two Mg2+ ions are jointly coordinated by two conserved Asps and the a-phosphate of an incoming dNTP30 (Figure 15.3B). Both Mg2+ are coordinated by six ligands in the octahedral geometry. One metal ion (conventionally known as the A metal ion) is close to the 3′-OH of the primer strand and suspected to activate it for nucleophilic attack, and the other metal ion (known as the B metal ion) is coordinated by the b and g phosphates (the leaving group) and is suspected to facilitate product formation. The metal ions are separated by ∼4 Å in the enzyme–substrate complex and by ∼3.4 Å in the enzyme transition-state complex.40 perhaps drawn close together by the transitional highly negatively charged phosphate anion. Perhaps not surprisingly, the same metal ion requirement and coordination geometry have been observed in pol l and m of the same X family30 and in the related CCA-adding enzymes.37 Interestingly, similar coordination of the two metal ions have been observed in pol I-like polymerases, including T7, Taq and Bst DNA pol of the A family,41–43 RB69 DNA pol of the B family,44 HIV RT,45 and Dpo4, pol k and Rev1 of the Y family.31 For the RNA pol II-like catalytic centre, in the currently available structures of RNA pol II complexed with a nucleic acid substrate, the metal ions are not coordinated as in DNA pol I or pol b, and are further apart than 4 Å. This may be due to the double conformations of the incoming UTP46 or the absence of the 3′-OH in the RNA strand.47 Given the three conserved Asp residues in the active site, it is likely that these carboxylates and the a-phosphate of the incoming NTP jointly coordinate the two Mg2+ ions for catalysis as with the other two classes of polymerases.

15.2.1 Alignment of Two Metal Ions and Fidelity of DNA Synthesis Displacement of two metal ions from the canonical binding sites in fact has been observed in the active site of several DNA polymerases. For example, Dpo4, which is capable of translesion synthesis, has a large, preformed and solvent-exposed active site.48 An omission of the 3′-OH in the primer strand, which is commonly used to crystallize polymerase–substrate ternary complexes, leads to ∼1 Å displacement of metal ion A.41,47,49,50 Although a variety of modified template bases (lesions) and mismatched replicating base pairs can bind in the active site of Dpo4,48 the catalytic efficiency differs with different substrates, and efficient nucleotide incorporation appears to correlate with the canonical configuration of two metal ions.50 Similar to Dpo4, DNA pol l also has a preformed active site, and one of the two metal ions is frequently missing from the active site in correlation with the misalignment of reaction groups in the crystal structures.51,52 Such correlation between the metal ion

420


coordination and the preference of substrates may explain why polymerases exhibit strong substrate specificity that surpasses selection by base pairing. The rate-limiting step in polymerization is postulated to be the key step for rejecting a wrong incoming nucleotide and ensuring replication fidelity.53 A conformational step prior to the chemical reaction was previously suggested to be the rate-limiting step in DNA synthesis.54 Because of the structural observation of a large finger-domain movement induced by binding of a correct incoming nucleotide in Taq and T7 DNA pol and in HIV RT that transforms the active site from an ‘open’ to the ‘closed’ state,41,45,55 the fidelity of DNA polymerases was originally attributed to such an induced-fit conformational change. In recent years, this hypothesis has been challenged by presteady-state kinetic analyses of A, B and X family DNA polymerases.53 It has been shown that the finger domain movement or closing of the active site is much faster than the phosphoryl transfer reaction and therefore is not the rate-limiting step.56–60 Furthermore, analogous domain movement is not observed in all polymerases, for example, the X family pol l and all Y family members, which still exhibit respectable fidelity.27 These observations lead to the verdict that a step after dNTP binding and the finger-domain movement is the rate-limiting step. A rate-limiting conformational change that differentiates the correct versus incorrect incoming nucleotide has been observed for all polymerases, including Dpo4,61 whose active site doesn’t undergo finger-domain movement. The conformational step therefore has to be other than the domain movement and is likely subtle. The displacement of metal ions observed in the Dpo4–substrate complexes when the 3′-OH nucleophile is absent or the replicating base pair is a mismatched, suggests that binding of two Mg2+ ions in the active site is insufficient for nucleotide incorporation. The proper positioning and alignment of the two metal ions with regard to the catalytic residues and substrate may be the rate-limiting conformational changes and the key to converting small differences in binding energy to large differences in catalytic efficiency in all polymerases.

15.3 Nucleases That Require Two Mg2+ Ions in the Active Site RNase H and MutH represent two large superfamilies of nucleic acid enzymes that break backbone phosphodiester bonds using two-Mg2+ ion catalysis. The RNase Hlike superfamily includes the transposases and retroviral integrases with a DDE motif, argonaute endonuclease and the Holliday junction resolvase RuvC.12,62 The catalytic domains feature a central mixed b-sheet surrounded by a-helices (Figure 15.4A). Like DNA and RNA polymerases, two absolutely conserved Asps are the hallmark of the RNase H-like nucleases and transposases.63,64 They are located on two adjacent parallel b-strands, one long and one short (Figure 15.4A). One or two additional carboxylate residues (Asp or Glu) located on the surrounding secondary structures may be required for the catalysis. The MutH-like superfamily includes the type IIP restriction endonucleases (REases, see the comprehensive review65) and phage lambda Exo I.66 The catalytic centre marked by the DEK motif is located

Nucleases That Require Two Mg2+ Ions in the Active Site

RNase H

421

MutH

Figure 15.4 (Plate 16) Two families of nucleases that have been shown to use two-metal-ion catalysis: (A) RNase H is a prototype of the superfamily, which includes nucleases, retroviral integrases, RuvC Holliday junction resolvase and dicer argonaute. The two catalytically essential Asps are located on two adjacent parallel strands (highlighted in cyan). Two additional carboxylates may be required for catalysis. Two Mg2+ are jointly coordinated by the carboxylates and the scissile phosphate; (B) MutH represents another superfamily, which includes type IIP restriction endonucleases, T7 endonuclease I (a Holliday junction resolvasae) and T4 exonuclease I. The catalytic residues, Asp (D), Glu (E) and Lys (K) are located on the diverging point of a b-hairpin (highlighted in cyan). The loop that contains the first Asp, which plays the central role in Mg2+ coordination, is often disordered in the MutH-family nucleases in the absence of a properly aligned substrate or divalent cations. A diagram of the metal ion coordination in each case is shown below the actual structure (See colour plate section)

on the diverging point of a b-hairpin (Figure 15.4B). The single conserved Asp residue is located on the first strand of the hairpin. This strand is flexible and often disordered in the crystal structures. The remaining Glu–X–Lys (X is a hydrophobic residue) located on the second strand (Figure 15.4B) are not absolutely conserved. The Glu is occasionally substituted by Gln, and Lys by Glu or Gln.67 It has long been debated whether one, two or three metal ions are required for the phosphoryl transfer reaction catalysed by RNase H- and MutH-like nucleases and transposases. Disagreements typically stem from two general observations. Firstly, among the high-resolution structures of apo-RNase H, transposase and retroviral integrase, one or two metal binding sites have been identified and the locations of metal ions vary from structure to structure.68–71 Secondly, even in the presence

422


of a DNA substrate, the number and location of metal ions vary among MutH-like type IIP REases,65 e.g. no metal ion in the Eco RI–substrate complex72 and three possible binding sites in Eco RV.73 Accompanying the metal ion variations, the active site composition of type IIP REases differs in the number of carboxylates required (two, three or four), and Lys of the DEK motif, catalytically essential for most of MutH-like endonucleases, is replaced by Glu in BamHI and Gln in BglII.67 These observations have raised speculation that REases may have different catalytic mechanisms using varying numbers of divalent cations.65,74 The crystal structures of a bacterial RNase H complexed with RNA/DNA hybrid substrates show two Mg2+ ions bound in the active site at a magnesium concentration below 2.5 mM.63 In the absence of a substrate, two Mn2+ ions were found at similar positions, but only at metal-ion concentrations much higher than are physiologically relevant.68 In the enzyme–substrate complexes, the two metal ions are jointly coordinated by the conserved Asp and scissile phosphate, each of which functions as a bidentate chelate (Figure 15.4A). By convention, the metal ion liganded to the nucleophile is called A, and the one liganded to the leaving group is B. A similar substrate-dependent two-metal-ion coordination is observed in crystal structures of the RNase H-like Tn5 transposase-DNA complexes.64 Among MutH-like nucleases, in the presence of their respective cognate substrates, MutH, BamHI and BglI bind two metal ions in the active site, in spite of the catalytic residue differences with Lys in MutH and BglI and Glu in BamHI.67 The coordination of the two metal ions by the scissile phosphate and a conserved Asp resembles the metal coordination observed in polymerases and RNase H (Figures 15.3 and 15.4), which is hereby referred to as canonical. It appears that in all cases at least one Asp is needed for coordination of the two metal ions. Asp is probably preferred because it has fewer rotamer conformations than Glu and costs less entropy to be fixed in one conformation. The metal ions observed in type IIP REases, PvuII, EcoRI and EcoRV, however, differ from the canonical ones in number and location.73,75,76 Such differences were originally attributed to different compositions of catalytic residues and perhaps different catalytic mechanisms. The conserved active site residues and tertiary structures of MutH, EcoRI PvuII and EcoRV suggest that these enzymes most probably share the same catalytic mechanism. The question is how to reconcile the metal ion differences observed in these related enzyme–substrate complexes.

15.3.1 Binding of Two Metal Ions in the Active Site Is Substrate Dependent Recently, crystal structures of human RNase H complexes with a RNA/DNA hybrid have provided new evidence on the substrate dependence of metal ion binding.77 One of the crystal forms contains 12 protein molecules in one asymmetric unit. Depending on the crystal lattice contacts, in some protein molecules the scissile phosphate is shifted away from the active residues by 1–2 Å, and these molecules contain only a single instead of two Ca2+ (Mg2+ mimics), and the Ca2+ binding site is neither the canonical A or B. This shows that not only are the number and location of divalent cations in the active site dependent on the presence of substrate,


423

but they are also correlated to positioning of the scissile phosphate. Only when the scissile phosphate is appropriately bound in the active site are two divalent cations jointly coordinated by the phosphate and carboxylates for catalysis, as shown in Figure 15.4A. For type IIP REases, both the conformation of the active site residues and the binding of divalent cations depend on the DNA substrate. A cognate DNA sequence is required for two-metal-ion binding in the canonical positions, and, in return, divalent cations are required for high-affinity binding of the DNA substrate.78–80 Without divalent cations, EcoRV binds DNA with low sequence specificity, and in the presence Ca2+, which mimics Mg2+ for substrate binding, but inhibits hydrolysis, EcoRV binds the cognate recognition sequence specifically.81 Structural investigation revealed that when BamHI is associated with the sequence GAATCC (one base pair substitution of the cognate GGATCC sequence), although the binding is tight enough for crystallization, the DNA is shifted away from the binding site of the cognate sequence and metal ions are absent.78 Moreover, the conserved Asp,65 which coordinates both metal ions in MutH, BamHI and BglI, is located on a loop that is disordered in the absence of a cognate DNA, partially ordered in the presence of DNA, and fully ordered only when both metal ions are bound.67,78 Without divalent cations, the catalytic residues of EcoRI in all DNA-complex structures, although ordered, are oriented in such a way that they are incapable of catalysis.72 MutH cleaves the unmethylated strand in the hemimethylated GATC sequence 10–20-fold better than the unmethylated duplex. In the crystal structures of the MutH–DNA complexes, the active site is clearly better ordered and better oriented for catalysis with the hemimethylated GATC than the unmethylated.67 Both the active site conformation and the metal ion binding in MutH-like nucleases are consistently shown to depend on binding of a cognate DNA substrate. The dependence of the nucleases on the cognate DNA substrate for metal-ion binding provides an explanation as to why the number and location of metal ions in the structures of homologous type IIP REases vary. To crystallize nuclease– substrate complexes, cleavage has to be prevented by artificial means, for example, substitution of catalytic residues, use of inert substrate analogues, or divalent cations that do not support catalysis. Strategies used to capture enzyme–substrate complexes with the purpose of preventing chemistry from taking place inevitably perturb the protein–DNA interactions. In addition, crystal lattice contacts may distort macromolecular interactions. One has to interpret crystal structures and particularly the active site conformation, including the number and location of metal ions, with caution. The exquisite sensitivity of metal ions to alterations in coordination environment probably underlies the specificity of these enzymes.

15.3.2 Two Metal Ions Enhance Catalytic Specificity When two divalent cations were first observed in the active site of alkaline phosphatase and the exonuclease of the Klenow fragment,82,83 they appeared to mediate binding of nucleic acids in the carboxylate-rich active site. The two metal ions are located roughly in parallel to the phosphosugar backbone on the opposite side of

424


Figure 15.5 Alignment of metal ions in the two-metal-ion mechanism. Two metal ions, preferably Mg2+, are jointly coordinated by at least one absolutely conserved Asp and the scissile phosphate, which is the a-phosphate of (d)NTP in the case of polymerase. Metal ion A is directly coordinated to the nucleophile and metal ion B to the leaving group. The reaction products are 3′-OH and 5′-PO4. (Reproduced from Yang, W. et al., Making and breaking nucleic acids: two-Mg2+-ion catalysis and substrate specificity, Mol. Cell, 2006, 22, 5–13)

the bases and bisected by the plane of the trigonal core of the pentacovalent intermediate (Figure 15.5). Based on the particular coordination geometry of the two metal ions, Steitz and Steitz proposed a generally applicable two-metal-ion mechanism, by which metal ion A acts like a Lewis base to reduce the pKa of a nucleophile and facilitate its deprotonation and metal ion B stabilizes the pentacovalent intermediate and the 3′ oxyanion leaving group.84 In the last 14 years, many structures of polymerases and nucleases using two-metal-ion catalysis have been determined.1,29–31 In addition to facilitating catalysis, the metal ions appear to play a critical role in enhancing substrate recognition and catalytic specificity for both polymerases and type IIP REases. By being sensitive to the coordination environment, the metal ions induce misalignment between a nonpreferential substrate and catalytic residues, thus greatly reducing the catalytic efficiency. Catalysis occurs efficiently only with a preferred substrate. The enhancement of substrate recognition is not limited to the fidelity of DNA and RNA polymerases, and sequence and methylation specific nucleases, but also extends to structure-specific nucleases, e.g. RNase H and argonaute. Although the latter nucleases have no explicit sequence specificity, RNase H only cleaves the RNA strand in a RNA/DNA hybrid and not in dsRNA, while argonaute functions in tandem with a duplex ruler and generates cleavage products of a specific length.85


425

15.3.3 Movement of the Two Metal Ions During Transition State Formation In all polymerases and nucleases reported to date, the two metal ions are jointly coordinated by a conserved Asp and the scissile phosphate (Figures 15.3 and 15.4), and their general positions conform to the 1993 proposal84 (Figure 15.5). The precise position and separation distance between the two metal ions, however, may vary at each reaction step according to the changes in the coordination environment. Tuning of the two metal ions is evident in both DNA pol b and RNase H. In the crystal structure of pol b–substrate complexes, the two metal ions are 4.0 Å apart, and in the structure of pol b-transition state-like complexes, they become 3.4 Å apart.40 Similarly, the separation of the two Mg2+ ions in the RNase H–substrate complexes vary from 4.0 Å to 4.5 Å, depending on which one of the catalytically essential carboxylates was replaced by carboxylamide.63 When the separation is 4.5 Å, the mutant protein (D132N) is completely inactive. But the mutation (D192), in which the metal ions are 4.0 Å apart, has detectable nuclease activity in the presence of Mn2+.63 The 4.0 Å separation allows the two metal ions to fit snugly on each side of the scissile phosphate, as depicted by Steitz and Steitz84 (Figure 15.5). However, the potential nucleophile coordinated by metal ion A is 3.5 Å or further away from the target phosphorus atom in the nuclease–substrate complexes (Figure 15.5). For nucleophilic attack to occur, the hydroxide ion probably needs to be within 2.5 Å of the phosphorus. The nucleophile is unlikely to dissociate from metal ion A and move over 1 Å by itself, owing to the high energy cost of removing an inner-shell metal-ion ligand and the repulsion between the negatively charged nucleophile and the scissile phosphate. We hypothesize that the 4.0 Å separation of the two metal ions represents a ‘resting’ state, and that metal ion A moves towards metal ion B and brings the nucleophile within striking distance for phosphoryl bond formation. Bringing the two metal ions closer could also better neutralize the developing negative charge on the pentacovalent intermediate. A 3.5 Å separation between two Mg2+ ions has been observed with DNA pol b and RNase H, when complexed with transition state mimics.86 Interestingly, such close juxtaposing is not observed when an Asp coordinating both metal ions is replaced by Asn (D132N).86 In addition, two Mg2+ ions are reported to be as close as 3 Å in the T7 RNA polymerase–substrate complex structure determined at 2.88 Å49 (Figure 15.3A). It is possible to bring two Mg2+ ions closer than 4.0 Å because its interatomic distance is 3.197 Å.87 Ca2+ has been widely used to substitute for Mg2+ during enzyme–substrate complex formation because it usually supports substrate binding. Catalysis of all polymerases and the nucleases mentioned in this chapter is inhibited by Ca2+.67,88–90 Comparisons of crystal structures of nucleases and polymerases reveal little difference between Ca2+ and Mg2+ coordination geometries.67,91,92 Interestingly, the interatomic distance of Ca is 3.94 Å,87 and the closest approach between two Ca2+ ions observed to date is 3.8 Å.93,94 This could be one reason that Ca2+ is unable to support catalysis of two-metal-ion-dependent phosphoryl transfer reactions. In contrast, the size and charge density of Zn2+, Co2+, Cd2+ and Mn2+ are comparable to Mg2+, and these divalent cations can substitute for Mg2+ in catalysis.

426


15.3.4 Metal Ions A and B Together Destabilize the Substrate and Assist Product Formation Metal ion A has been suggested to play an essential role in nucleophile formation in ribozymes and enzymes that lack a conventional general base.84 Experimental observations accumulated over the years suggest that nucleophile formation cannot be achieved by metal ion A alone. In the absence of an apparent general base, a potential nucleophile is often coordinated by the nucleic acid substrate, in addition to metal ion A. For example, in RNase H, the potential nucleophilic water molecule interacts with metal ion A and the pro-Rp nonbridging oxygen of the phosphate 3′ to the scissile bond,63 and in MutH, a Lys sidechain supplies the third ligand67 (Figure 15.4B). Neither phosphate oxygen nor Lys is a conventional general base. Interestingly, although the pro-Rp oxygen coordinates the nucleophile in both RNase H and MutH, its stereospecific replacement with sulfur has the opposite effect. The Rp sulfur enhances the cleavage activity of MutH,67 but it inhibits RNase H.95 MutH, whose active site contains a Lys, likely benefits from the extra negative charge brought by the Rp sulfur (due to the nature of the single P–S).96 But RNase H, whose active site contains four carboxylates, may be inhibited by the extra negative charge and steric hindrance due to the sulfur substitution (larger atomic radius than oxygen). Parallel to the diverse ways of substrate-assisted nucleophile formation, the coordinating carboxylate for metal ion A can be replaced by carboxylamide or histidine in the RNase H family.63,97 It appears that the net charge environment contributed by both the catalytic residues and the substrate is the key for twometal-ion catalysis. Metal ion B may promote the phosphoryl transfer reaction instead of passively stabilizing the transition state, as originally proposed.84 Even though the A metal ion seemingly initiates the chemical reaction by directly coordinating to the nucleophile, replacement of carboxylates that coordinate metal ion A often do not eliminate nuclease activities as effectively as mutating the B metal ion ligands.63,98 In addition, the B site also exhibits a stronger metal ion preference than the A site. In the 3′ to 5′ exonuclease of DNA polymerase I, the B site normally binds a Mg2+ ion,99 but it rejects Mg2+ when the nonbridging oxygen of the scissile phosphate is replaced by sulfur.100 In the crystal structures of RNase H–substrate complexes, the Mg2+ ion at the B site is coordinated irregularly by five ligands, none of which is a water molecule (Figure 15.6A).63 The nonideal coordination geometry and ligands devoid of water suggest that metal ion B in the enzyme–substrate complex is in a high-energy state and may strain the scissile phosphodiester bond. The deviation from ideal octahedral coordination geometry may explain why Mg2+ is preferred over Ca2+ for catalysis. Ca2+ readily accepts six, seven, eight or nine ligands in various coordination geometries, and promiscuity regarding ligand type and coordination geometry renders Ca2+ less likely to destabilize the enzyme–substrate complex and promote product formation. This may be another reason that Ca2+ does not usually replace Mg2+ in phosphoryl transfer reactions. In a mutant RNase H–product complex, the B-site Mg2+ shifts by 0.9 Å and becomes coordinated in near octahedral geometry by six ligands, two of which are water molecules (Figure 15.6B).86 The change from five-ligand coordination of metal


427

Figure 15.6 (Plate 17) Coordination of the two metal ions changes during RNA cleavage by RNase H: (A) Enzyme–substrate complex; (B) Enzyme–product complex (See colour plate section)

ion B in the substrate complex to a favourable six-ligand coordination in the product complex indicates that the two metal ions also play a role in product formation. The two metal ions jointly coordinated by the scissile phosphate may also facilitate product release. After splitting into 5′-phosphate and 3′-OH, the cleaved scissile phosphate can no longer coordinate the two metal ions simultaneously. At least one metal ion has to be displaced or released, as observed in product complexes of DNA and RNA polymerases and nucleases.49,50,73,86,91 The RNase H–product complexes have been trapped in the crystal form by a key residue for product release (E188) being substituted by Ala. In the product complex the cleaved scissile phosphate moves out of the active site, and concomitantly metal ion A moves by 2 Å and retains only one ligand from RNase H.86 Presumably eventual dissociation of the metal ion leads to the release of the cleavage product. Instead of having separate responsibilities, as initially proposed, the two metal ions appear to work together throughout phosphoryl transfer reactions. It is widely accepted that enzymes facilitate chemical reactions by proper alignment of substrates relative to each other and to the catalytic residues, thereby lowering the energy barrier between the substrate and product states. As such, the key element of two-metal-ion catalysis may be the proper alignment of the two metal ions with regard to the conserved carboxylates and substrate. The strong conservation of twometal-ion coordination in polymerases, nucleases and group I intron (in which a backbone phosphate replaces the conserved Asp to coordinate both metal ions)101 certainly supports this metal-ion centric view. Once the proper alignment is formed and the enzyme–substrate complex is destabilized, a nucleophile can form and be activated in various ways without a conventional general base. Similarly, after the reaction proceeds past the pentacovalent intermediate, protonation of a leaving group is energetically favourable, even without a general acid. This view is supported by the biochemical and computational studies of metal-ion-dependent proton transfer in DNA and RNA polymerases.102–104

428


15.4 Advantages of Two-Metal-Ion Catalysis: Specificity and Versatility The exquisite sensitivity of metal ions to the local environment16,18,105 provides a means for nucleic acid enzymes to convert small conformational differences into large catalytic rate changes. Three features of two-metal-ion mechanisms ensure the substrate recognition and catalytic specificity: (i) interdependence of substrate recognition and two-metal-ion binding, as illustrated in RNase H,63 type II restriction endonucleases,67,78,81 ribozyme106 and DNA polymerases50,107; (ii) usage of metal ions with highly stringent coordination requirements to destabilize the enzyme–substrate complex and promote product formation, as observed with RNase H86 and (iii) reliance on the electrostatic environment created by both the catalytic residues and the nucleic acid substrate for nucleophile formation, and for metal ion movement during the course of the phosphoryl transfer reaction. Mg2+, known for its exquisite octahedral coordination geometry and narrow distribution of coordination distance,18 is most often used in two-metal catalysis. Ca2+, because of its large atomic radius and indifference to ligand number and coordination distance, usually does not support phosphoryl transfer reactions. Mn2+, which is most similar to Mg2+ in chemical nature, but has relaxed coordination requirements, can replace Mg2+ and allow phosphoryl transfer reactions to occur with altered catalytic residues, nonideal substrates and reduced specificity.50,108–112 The natural abundance of cellular Mg2+ over Mn2+ thus furthers the substrate specificity of two-metal-ion catalysis. The second advantage of two-metal-ion catalysis is its versatility and reversibility. The versatility is clearly demonstrated in DNA and RNA synthesis, where polymerases have to incorporate various nucleotide building blocks ((d)NTPs) with roughly equal efficiency and yet in each reaction cycle select only the right one according to the template base. Two-metal-ion catalysis, which allows the active site to align correctly only upon association with a Watson–Crick base pair, enables DNA and RNA polymerases to achieve high specificity with a broad range of substrates. Regarding reversibility, since two-metal-ion catalysis has no specific requirement for a general base or acid, it supports both making and breaking of DNA and RNA. During DNA transposition and RNA splicing, a single active site with two metal ions can carry out at least two consecutive phosphoryl transfer reactions.63,84 In the first reaction, DNA or RNA is cleaved to generate a 3′-OH (product), and in the second reaction the 3′-OH is the nucleophile (substrate) and results in strand transfer (Figure 15.1). The symmetric arrangement of the two metal ions in the active site of Tn5 transposase and group I ribozyme (Figure 15.7) allows the metal ions to alternate their roles in activating a nucleophile and stabilizing a pentacovalent intermediate in consecutive phosphoryl transfer reactions without releasing the 3′-OH.64,101 15.4.1 One-Metal-Ion Catalysis Exceptions to two-metal-ion catalysis are one-metal-ion and metal-ion-independent catalysis utilized by certain nucleases and recombinases. To date, only reactions resulting in 5′-OH and 3′-phosphate are known to be catalysed by enzymes that

Advantages of Two-Metal-Ion Catalysis

429

Figure 15.7 (Plate 18) Coordination of the two metal ions in Group I intron. The two metal ions have similar coordination geometry (hence symmetric) and can reverse their roles as ‘A’ and ‘B’ metal ions in the successive steps of RNA splicing reaction. (Reproduced from Yang, W. et al., Making and breaking nucleic acids: two-Mg2+-ion catalysis and substrate specificity, Mol. Cell, 2006, 22, 5–13) (See colour plate section)

don’t require divalent cations, for example, hepatitis virus delta ribozyme,113 RNase A, type IB topoisomerase, phage l integrase and g d resolvase.114 Reactions resulting in 5′-phosphate and 3′-OH usually require one or two metal ions. Nucleases with the bba-metal motif, for example His-Cys and CIY-YIC homing and HNH endonucleases74,115–118 have no conserved catalytically essential carboxylates in the active site and require one single divalent cation for catalysis. The single metal ion occupies the same position relative to the scissile phosphate as metal ion B of the twometal-ion mechanism. These nucleases often depend on a conserved His to activate a nucleophile. Without the requirement for specifically aligning two metal ions, a variety of divalent cations are able to support one-metal-ion catalysis.118 These types of nucleases are either sequence nonspecific or specific. If sequence specific, the cleavage and binding specificity are equal and specificity is achieved by noncatalytic domains that recognize 20–40 bps, much longer than the 4–8 bps typically recognized by type II REases.119 15.4.2 Prediction of Two-Metal-Ion Catalysis Two-metal-ion catalysis, because of its unique specificity and versatility, can be readily predicted based on the composition of catalytic residues, the nature of

430


nucleophile and substrate specificity. If all, or the majority of, the catalytically essential residues are carboxylates, with at least one invariable Asp, and if the enzyme exhibits stricter catalytic specificity than binding specificity, catalysis almost certainly depends on two metal ions. Structure-specific FEN-1 and XPF endonucleases,120,121 RNase III and dicer endonucleases,97,122 and Group II ribozymes catalysing RNA splicing with high substrate specificity, clearly fulfil the criteria for a twometal-ion catalysis. A noted variation of two-metal-ion catalysis is the three metal ions found in homodimeric LAGLIDADG homing endonucleases,123 where a single metal ion B is shared between two neighbouring active sites. Nevertheless, each phosphoryl transfer reaction requires two metal ions coordinated by a conserved Asp.

15.5 Concluding Remarks In this chapter we have summarized the following properties of two-metal-ion catalysis: (i) participation of the two metal ions in substrate recognition; (ii) the role of metal ions, particularly at the B site, in destabilizing the enzyme–substrate complex and driving the reaction forward; (iii) dependence on a specific electrostatic environment rather than metal ion A per se for the nucleophile formation and nucleophilic attack; (iv) movement of two metal ions towards each other to stabilize transition state and facilitate product formation; (v) movement of two metal ions away from each other to release products. The key to two-metal-ion catalysis is the proper alignment of the metal ions with regard to the catalytic residues and the nucleic acid substrate. The extensive involvement of the nucleic acid substrate in metal-ion binding and the sensitivity of metal ions to the coordination environment enable nucleic acid enzymes that utilize two-metal-ion catalysis to convert small thermodynamic differences into large changes in catalytic rate and achieve extraordinary specificity.

Acknowledgements I thank Dr R. Craigie for reviewing the manuscript and the Intramural Research Program of the NIH, NIDDK for funding the research. Figures 15.1A, 15.5 and 15.7 are reproductions from Mol. Cell, 22, 5–13 Figures 2A, 1B and 3C, respectively.

References 1. Jager, J.; Pata, J.D.; Getting a grip: polymerases and their substrate complexes; Curr. Opin. Struct. Biol., 1999, 9, 21–28. 2. Armache, K.J.; Kettenberger, H.; Cramer, P.; The dynamic machinery of mRNA elongation; Curr. Opin. Struct. Biol., 2005, 15, 197–203.

References

431

3. Radman-Livaja, M.; Biswas, T.; Ellenberger, T.; Landy, A.; Aihara, H.; DNA arms do the legwork to ensure the directionality of lambda site-specific recombination; Curr. Opin. Struct. Biol., 2006, 16, 42–50. 4. Grindley, N.D.; Whiteson, K.L.; Rice, P.A.; Mechanisms of site-specific recombination; Annu. Rev. Biochem., 2006, 75, 567–605. 5. Koonin, E.V.; Ilyina, T.V.; Computer-assisted dissection of rolling circle DNA replication; Biosystems, 1993, 30, 241–268. 6. Fedor, M.J.; Williamson, J.R.; The catalytic diversity of RNAs; Nat. Rev. Mol. Cell Biol., 2005, 6, 399–412. 7. Stahley, M.R.; Strobel, S.A.; RNA splicing, group I intron crystal structures reveal the basis of splice site selection and metal ion catalysis; Curr. Opin. Struct. Biol., 2006, 16, 319–326. 8. Esposito, D.; Craigie, R.; HIV integrase structure and function; Adv. Virus Res., 1999, 52, 319–333. 9. Burgers, P.M.; Eckstein, F.; Stereochemistry of internucleotide bond formation by polynucleotide phosphorylase from Micrococcus luteus; Biochem., 1979, 18, 450–454. 10. Brody, R.S.; Frey, P.A.; Unambiguous determination of the stereochemistry of nucleotidyl transfer catalyzed by DNA polymerase I from Escherichia coli; Biochem., 1981, 20, 1245–1252. 11. Mizuuchi, K.; Nobbs, T.J.; Halford, S.E.; Adzuma, K.; Qin, J.; A new method for determining the stereochemistry of DNA cleavage reactions: application to the SfiI and HpaII restriction endonucleases and to the MuA transposase; Biochem., 1999, 38, 4640–4648. 12. Yang, W.; Steitz, T.A.; Recombining the structures of HIV integrase, RuvC and RNase H; Structure, 1995, 3, 131–134. 13. Cowan, J.A.; Structural and catalytic chemistry of magnesium-dependent enzymes; Biometals, 2002, 15, 225–235. 14. Lyons, T.J.; Eide, D.J.; Transport and storage of metal ions in biology. In Biological Inorganic Chemistry: Structure, and Reactivity, Bertini, I.; Gray, H.B.; Stiefel, E.I.; and Valentine, J.S.; Eds, University Science Books, Herndon, VA, 2006, 57–77. 15. Romani, A.; Scarpa, A.; Regulation of cell magnesium; Arch. Biochem. Biophys., 1992, 298, 1–12. 16. Maguire, M.E.; Cowan, J.A.; Magnesium chemistry and biochemistry; Biometals, 2002, 15, 203–210. 17. Pauling, L.; The Nature of the Chemical Bond, Cornell University Press, Ithaca, 1961. 18. Harding, M.M.; The geometry of metal-ligand interactions relevant to proteins; Acta Crystallogr. D Biol. Crystallogr., 1999, 55(8), 1432–1443. 19. Christianson, D.W.; Structural biology of zinc; Adv. Protein Chem., 1992, 42, 281–355. 20. Pidcock, E.; Moore, G.R.; Structural characteristics of protein binding sites for calcium and lanthanide ions; J. Biol. Inorg. Chem., 2001, 6, 479–489. 21. Lange, S.J.; Que, L., Jr; Oxygen activating nonheme iron enzymes; Curr. Opin. Chem. Biol., 1998, 2, 159–172. 22. Hernick, M.; Fierke, C.A.; Zinc hydrolases: the mechanisms of zinc-dependent deacetylases; Arch. Biochem. Biophys., 2005, 433, 71–84. 23. Kim, D.H.; Mobashery, S.; Mechanism-based inhibition of zinc proteases; Curr. Med. Chem., 2001, 8, 959–965. 24. Jabri, E.; Carr, M.B.; Hausinger, R.P.; Karplus, P.A.; The crystal structure of urease from Klebsiella aerogenes; Science, 1995, 268, 998–1004. 25. Diebler, H.; Eigen, M.; Ilgenfritz, G.; Maass, G.; Winkler, R.; Kinetics and mechanisms of reactions of main group metal ions with biological carriers; Pure Appl. Chem., 1969, 20, 93–115. 26. Pontius, B.W.; Lott, W.B.; von Hippel, P.H.; Observations on catalysis by hammerhead ribozymes are consistent with a two-divalent-metal-ion mechanism; Proc. Natl. Acad. Sci. USA, 1997, 94, 2290–2294. 27. Bebenek, K.; Kunkel, T.A.; Functions of DNA polymerases. Adv. Protein. Chem., 2004, 69, 137–165.

432


28. Cann, I.K.; Ishino, Y.; Archaeal DNA replication: identifying the pieces to solve a puzzle; Genetics, 1999, 152, 1249–1267. 29. Steitz, T.A.; Yin, Y.W.; Accuracy, lesion bypass, strand displacement and translocation by DNA polymerases; Philos. Trans. R. Soc. Lond. B Biol. Sci., 2004, 359, 17–23. 30. Moon, A.F.; Garcia-Diaz, M.; Batra, V.K.; Beard, W.A.; Bebenek, K.; Kunkel, T.A. et al.; The X family portrait: Structural insights into biological functions of X family polymerases; DNA Repair (Amst.), 2007, 6, 1709–1725. 31. Yang, W.; Woodgate, R.; What a difference a decade makes: insights into translesion DNA synthesis; Proc. Natl. Acad. Sci. USA, 2007, 104, 15591–15598. 32. Bailey, S.; Wing, R.A.; Steitz, T.A.; The structure of T. aquaticus DNA polymerase III is distinct from eukaryotic replicative DNA polymerases; Cell, 2006, 126, 893–904. 33. Lamers, M.H.; Georgescu, R.E.; Lee, S.G.; O’Donnell, M.; Kuriyan, J.; Crystal structure of the catalytic alpha subunit of E. coli replicative DNA polymerase III; Cell, 2006, 126, 881–892. 34. Ferrer-Orta, C.; Arias, A.; Escarmis, C.; Verdaguer, N.; A comparison of viral RNAdependent RNA polymerases; Curr. Opin. Struct. Biol., 2006, 16, 27–34. 35. Hall, T.M.; Poly(A) tail synthesis and regulation: recent structural insights; Curr. Opin. Struct. Biol., 2002, 12, 82–88. 36. Martin, G.; Keller, W.; Doublie, S.; Crystal structure of mammalian poly(A) polymerase in complex with an analog of ATP; EMBO J., 2000, 19, 4193–4203. 37. Xiong, Y.; Steitz, T.A.; A story with a good ending: tRNA 3′-end maturation by CCAadding enzymes; Curr. Opin. Struct. Biol., 2006, 16, 12–17. 38. Kornberg, R.D.; The molecular basis of eucaryotic transcription; Cell Death Differ., 2007, 14, 1989–1997. 39. Murakami, K.S.; Darst, S.A.; Bacterial RNA polymerases: the whole story; Curr. Opin. Struct. Biol., 2003, 13, 31–39. 40. Batra, V.K.; Beard, W.A.; Shock, D.D.; Krahn, J.M.; Pedersen, L.C.; Wilson, S.H.; Magnesium-induced assembly of a complete DNA polymerase catalytic complex; Structure, 2006, 14, 757–766. 41. Doublié, S.; Tabor, S.; Long, A.M.; Richardson, C.C.; Ellenberger, T.; Crystal structure of a bacteriophage T7 DNA replication complex at 2.2 A resolution; Nature, 1998, 391, 251–258. 42. Li, Y.; Dutta, S.; Doublie, S.; Bdour, H.M.; Taylor, J.S.; Ellenberger, T.; Nucleotide insertion opposite a cis-syn thymine dimer by a replicative DNA polymerase from bacteriophage T7; Nat. Struct. Mol. Biol., 2004, 11, 784–790. 43. Johnson, S.J.; Taylor, J.S.; Beese, L.S.; Processive DNA synthesis observed in a polymerase crystal suggests a mechanism for the prevention of frameshift mutations; Proc. Natl. Acad. Sci. USA, 2003, 100, 3895–3900. 44. Franklin, M.C.; Wang, J.; Steitz, T.A.; Structure of the replicating complex of a pol alpha family DNA polymerase; Cell, 2001, 105, 657–667. 45. Huang, H.; Chopra, R.; Verdine, G.L.; Harrison, S.C.; Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug resistance; Science, 1998, 282, 1669–1675. 46. Wang, D.; Bushnell, D.A.; Westover, K.D.; Kaplan, C.D.; Kornberg, R.D.; Structural basis of transcription: role of the trigger loop in substrate specificity and catalysis; Cell, 2006, 127, 941–954. 47. Westover, K.D.; Bushnell, D.A.; Kornberg, R.D.; Structural basis of transcription: nucleotide selection by rotation in the RNA polymerase II active center; Cell, 2004, 119, 481–489. 48. Yang, W.; Portraits of a Y-family DNA polymerase; FEBS Lett., 2005, 579, 868– 872. 49. Yin, Y.W.; Steitz, T.A.; The structural mechanism of translocation and helicase activity in T7 RNA polymerase; Cell, 2004, 116, 393–404. 50. Vaisman, A.; Ling, H.; Woodgate, R.; Yang, W.; Fidelity of Dpo4: effect of metal ions, nucleotide selection and pyrophosphorolysis; EMBO J., 2005, 24, 2957–2967.

References

433

51. Garcia-Diaz, M.; Bebenek, K.; Krahn, J.M.; Pedersen, L.C.; Kunkel, T.A.; Role of the catalytic metal during polymerization by DNA polymerase lambda; DNA Repair (Amst.), 2007, 6, 1333–1340. 52. Garcia-Diaz, M.; Bebenek, K.; Krahn, J.M.; Kunkel, T.A.; Pedersen, L.C.; A closed conformation for the Pol lambda catalytic cycle; Nat. Struct. Mol. Biol., 2005, 12, 97–98. 53. Showalter, A.K.; Tsai, M.D.; A reexamination of the nucleotide incorporation fidelity of DNA polymerases; Biochem., 2002, 41, 10571–10576. 54. Wong, I.; Patel, S.S.; Johnson, K.A.; An induced-fit kinetic mechanism for DNA replication fidelity: direct measurement by single-turnover kinetics; Biochem., 1992, 30, 526–537. 55. Li, Y.; Korolev, S.; Waksman, G.; Crystal structures of open and closed forms of binary and ternary complexes of the large fragment of Thermus aquaticus DNA polymerase I: structural basis for nucleotide incorporation; EMBO J., 1998, 17, 7514–7525. 56. Shah, A.M.; Li, S.X.; Anderson, K.S.; Sweasy, J.B.; Y265H mutator mutant of DNA polymerase beta. Proper teometric alignment is critical for fidelity; J. Biol. Chem., 2001, 276, 10824–10831. 57. Zhong, X.; Patel, S.S.; Werneburg, B.G.; Tsai, M.D.; DNA polymerase beta: multiple conformational changes in the mechanism of catalysis; Biochem., 1997, 36, 11891–11900. 58. Bakhtina, M.; Lee, S.; Wang, Y.; Dunlap, C.; Lamarche, B.; Tsai, M.D.; Use of viscogens, dNTPalphaS, and rhodium(III) as probes in stopped-flow experiments to obtain new evidence for the mechanism of catalysis by DNA polymerase beta; Biochem., 2005, 44, 5177–5187. 59. Rothwell, P.J.; Mitaksov, V.; Waksman, G.; Motions of the fingers subdomain of klentaq1 are fast and not rate limiting: implications for the molecular basis of fidelity in DNA polymerases; Mol. Cell, 2005, 19, 345–355. 60. Zhang, H.; Cao, W.; Zakharova, E.; Konigsberg, W.; De La Cruz, E.M.; Fluorescence of 2-aminopurine reveals rapid conformational changes in the RB69 DNA polymeraseprimer/template complexes upon binding and incorporation of matched deoxynucleoside triphosphates; Nucleic Acids Res., 2007, 35, 6052–6062. 61. Fiala, K.A.; Suo, Z.; Mechanism of DNA polymerization catalyzed by Sulfolobus solfataricus P2 DNA polymerase IV; Biochem., 2004, 43, 2116–2125. 62. Song, J.J.; Smith, S.K.; Hannon, G.J.; Joshua-Tor, L.; Crystal structure of Argonaute and its implications for RISC slicer activity; Science, 2004, 305, 1434–1437. 63. Nowotny, M.; Gaidamakov, S.A.; Crouch, R.J.; Yang, W.; Crystal structures of RNase H bound to an RNA/DNA hybrid: substrate specificity and metal-dependent catalysis; Cell, 2005, 121, 1005–1016. 64. Steiniger-White, M.; Rayment, I.; Reznikoff, W.S.; Structure/function insights into Tn5 transposition; Curr. Opin. Struct. Biol., 2004, 14, 50–57. 65. Pingoud, A.; Fuxreiter, M.; Pingoud, V.; Wende, W.; Type II restriction endonucleases: structure and mechanism; Cell Mol. Life Sci., 2005, 62, 685–707. 66. Kovall, R.A.; Matthews, B.W.; Structural, functional, and evolutionary relationships between lambda-exonuclease and the type II restriction endonucleases; Proc. Natl. Acad. Sci. USA, 1998, 95, 7893–7897. 67. Lee, J.Y.; Chang, J.; Joseph, N.; Ghirlando, R.; Rao, D.N.; Yang, W.; MutH complexed with hemi- and unmethylated DNAs: coupling base recognition and DNA cleavage; Mol. Cell, 2005, 20, 155–166. 68. Goedken, E.R.; Marqusee, S.; Co-crystal of Escherichia coli RNase HI with Mn2+ ions reveals two divalent metals bound in the active site; J. Biol. Chem., 2001, 276, 7266–7271. 69. Tsunaka, Y.; Takano, K.; Matsumura, H.; Yamagata, Y.; Kanaya, S.; Identification of single Mn(2+) binding sites required for activation of the mutant proteins of E.coli RNase HI at Glu48 and/or Asp134 by X-ray crystallography; J. Mol. Biol., 2005, 345, 1171–1183. 70. Maignan, S.; Guilloteau, J.P.; Zhou-Liu, Q.; Clement-Mella, C.; Mikol, V.; Crystal structures of the catalytic domain of HIV-1 integrase free and complexed with its metal

434

71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93.

Two-Metal-Ion-Dependent Catalysis in Nucleic Acid Enzymes cofactor: high level of similarity of the active site with other viral integrases; J. Mol. Biol., 1998, 282, 359–368. Bujacz, G.; Jaskolski, M.; Alexandratos, J.; Wlodawer, A.; Merkel, G.; Katz, R.A. et al.; High-resolution structure of the catalytic domain of avian sarcoma virus integrase; J. Mol. Biol., 1995, 253, 333–346. Sapienza, P.J.; Rosenberg, J.M.; Jen-Jacobson, L.; Structural and thermodynamic basis for enhanced DNA binding by a promiscuous mutant EcoRI endonuclease; Structure, 2007, 15, 1368–1382. Horton, N.C.; Perona, J.J.; DNA cleavage by EcoRV endonuclease: two metal ions in three metal ion binding sites; Biochem., 2004, 43, 6841–6857. Galburt, E.A.; Stoddard, B.L.; Catalytic mechanisms of restriction and homing endonucleases; Biochem., 2002, 41, 13851–13860. Kim, Y.C.; Grable, J.C.; Love, R.; Greene, P.J.; Rosenberg, J.M.; Refinement of EcoRI endonuclease crystal structure: a revised protein chain tracing; Science, 1990, 249, 1307–1309. Horton, J.R.; Cheng, X.; PvuII endonuclease contains two calcium ions in active sites; J. Mol. Biol., 2000, 300, 1049–1056. Nowotny, M.; Gaidamakov, S.A.; Ghirlando, R.; Cerritelli, S.M.; Crouch, R.J.; Yang, W.; Structure of human RNase H1 complexed with an RNA/DNA hybrid: insight into HIV reverse transcription; Mol. Cell, 2007, 28, 264–276. Viadiu, H.; Aggarwal, A.K.; Structure of BamHI bound to nonspecific DNA: a model for DNA sliding; Mol. Cell, 2000, 5, 889–895. Bowen, L.M.; Muller, G.; Riehl, J.P.; Dupureur, C.M.; Lanthanide spectroscopic studies of the dinuclear and Mg(II)-dependent PvuII restriction endonuclease; Biochem., 2004, 43, 15286–15295. Xu, Q.S.; Kucera, R.B.; Roberts, R.J.; Guo, H.C.; An asymmetric complex of restriction endonuclease MspI on its palindromic DNA recognition site; Structure, 2004, 12, 1741–1747. Vipond, I.B.; Halford, S.E.; Specific DNA recognition by EcoRV restriction endonuclease induced by calcium ions; Biochem., 1995, 34, 1113–1119. Kim, E.E.; Wyckoff, H.W.; Reaction mechanism of alkaline phosphatase based on crystal structures. Two-metal ion catalysis; J. Mol. Biol., 1992, 218, 449–464. Beese, L.S.; Steitz, T.A.; Structural basis for the 3′–5′ exonuclease activity of Escherichia coli DNA polymerase I: a two metal ion mechanism; EMBO J., 1992, 10, 25–33. Steitz, T.A.; Steitz, J.A.; A general two-metal-ion mechanism for catalytic RNA; Proc. Natl. Acad. Sci. USA, 1993, 90, 6498–6502. Tolia, N.H.; Joshua-Tor, L.; Slicer and the argonautes; Nat. Chem. Biol., 2007, 3, 36–43. Nowotny, M.; Yang, W.; Stepwise analyses of metal ions in RNase H catalysis from substrate destabilization to product release; EMBO J., 2006, 25, 1924–1933. Pauling, L.; Atomic radii and interatomic distances in metals; J. Am. Chem. Soc., 1947, 69, 542–553. Kunkel, T.A.; Tcheng, J.E.; Meyer, R.R.; Purification and properties of DNA polymerasebeta from guinea pig liver; Biochim. Biophys. Acta, 1978, 520, 302–316. Ling, H.; Boudsocq, F.; Plosky, B.S.; Woodgate, R.; Yang, W.; Replication of a cis-syn thymine dimer at atomic resolution; Nature, 2003, 424, 1083–1087. Vipond, I.B.; Baldwin, G.S.; Halford, S.E.; Divalent metal ions at the active sites of the EcoRV and EcoRI restriction endonucleases; Biochem., 1995, 34, 697–704. Viadiu, H.; Aggarwal, A.K.; The role of metals in catalysis by the restriction endonuclease BamHI; Nat. Struct. Biol., 1998, 5, 910–916. Ling, H.; Boudsocq, F.; Woodgate, R.; Yang, W.; Snapshots of replication through an abasic lesion; structural basis for base substitutions and frameshifts; Mol. Cell, 2004, 13, 751–762. Weis, W.I.; Drickamer, K.; Hendrickson, W.A.; Structure of a C-type mannose-binding protein complexed with an oligosaccharide; Nature, 1992, 360, 127–134.

References

435

94. Soriano-Garcia, M.; Padmanabhan, K.; de Vos, A.M.; Tulinsky, A.; The Ca2+ ion and membrane binding structure of the Gla domain of Ca-prothrombin fragment 1; Biochem., 1992, 31, 2554–2566. 95. Haruki, M.; Tsunaka, Y.; Morikawa, M.; Iwai, S.; Kanaya, S.; Catalysis by Escherichia coli ribonuclease HI is facilitated by a phosphate group of the substrate; Biochem., 2000, 39, 13939–13944. 96. Frey, P.A.; Sammons, R.D.; Bond order and charge localization in nucleoside phosphorothioates; Science, 1985, 228, 541–545. 97. Rivas, F.V.; Tolia, N.H.; Song, J.J.; Aragon, J.P.; Liu, J.; Hannon, G.J. et al.; Purified Argonaute2 and an siRNA form recombinant human RISC; Nat. Struct. Mol. Biol., 2005, 12, 340–349. 98. Derbyshire, V.; Grindley, N.D.; Joyce, C.M.; The 3′–5′ exonuclease of DNA polymerase I of Escherichia coli: contribution of each amino acid at the active site to the reaction; EMBO J., 1992, 10, 17–24. 99. Brautigam, C.A.; Steitz, T.A.; Structural principles for the inhibition of the 3′–5′ exonuclease activity of Escherichia coli DNA polymerase I by phosphorothioates; J. Mol. Biol., 1998, 277, 363–377. 100. Brautigam, C.A.; Sun, S.; Piccirilli, J.A.; Steitz, T.A.; Structures of normal single-stranded DNA and deoxyribo-3′-S-phosphorothiolates bound to the 3′–5′ exonucleolytic active site of DNA polymerase I from Escherichia coli; Biochem., 1999, 38, 696–704. 101. Stahley, M.R.; Strobel, S.A.; Structural evidence for a two-metal-ion mechanism of group I intron splicing; Science, 2005, 309, 1587–1590. 102. Castro, C.; Smidansky, E.; Maksimchuk, K.R.; Arnold, J.J.; Korneeva, V.S.; Gotte, M. et al.; Two proton transfers in the transition state for nucleotidyl transfer catalyzed by RNA- and DNA-dependent RNA and DNA polymerases; Proc. Natl. Acad. Sci. USA, 2007, 104, 4267–4272. 103. Lin, P.; Pedersen, L.C.; Batra, V.K.; Beard, W.A.; Wilson, S.H.; Pedersen, L.G.; Energy analysis of chemistry for correct insertion by DNA polymerase beta; Proc. Natl. Acad. Sci. USA, 2006, 103, 13294–13299. 104. Wang, L.; Yu, X.; Hu, P.; Broyde, S.; Zhang, Y.; A water-mediated and substrate-assisted catalytic mechanism for Sulfolobus solfataricus DNA polymerase IV; J. Am. Chem. Soc., 2007, 129, 4731–4737. 105. Glusker, J.P.; Katz, A.K.; Bock, C.W.; Two-metal binding motifs in protein crystal structrures; Struct. Chem., 2001, 12, 323–341. 106. Rangan, P.; Woodson, S.A.; Structural requirement for Mg2+ binding in the group I intron core; J. Mol. Biol., 2003, 329, 229–238. 107. Gangurde, R.; Modak, M.J.; Participation of active-site carboxylates of Escherichia coli DNA polymerase I (Klenow fragment) in the formation of a prepolymerase ternary complex; Biochem., 2002, 41, 14552–14559. 108. Allingham, J.S.; Haniford, D.B.; Mechanisms of metal ion action in Tn10 transposition; J. Mol. Biol., 2002, 319, 53–65. 109. Kunkel, T.A.; Loeb, L.A.; On the fidelity of DNA replication. Effect of divalent metal ion activators and deoxyrionucleoside triphosphate pools on in vitro mutagenesis; J. Biol. Chem., 1979, 254, 5718–5725. 110. Tabor, S.; Richardson, C.C.; Effect of manganese ions on the incorporation of dideoxynucleotides by bacteriophage T7 DNA polymerase and Escherichia coli DNA polymerase I; Proc. Natl. Acad. Sci. USA, 1989, 86, 4076–4080. 111. Vermote, C.L.; Halford, S.E.; EcoRV restriction endonuclease: communication between catalytic metal ions and DNA recognition; Biochem., 1992, 31, 6082–6089. 112. Cirino, N.M.; Cameron, C.E.; Smith, J.S.; Rausch, J.W.; Roth, M.J.; Benkovic, S.J. et al.; Divalent cation modulation of the ribonuclease functions of human immunodeficiency virus reverse transcriptase; Biochem., 1995, 34, 9936–9943. 113. Ke, A.; Zhou, K.; Ding, F.; Cate, J.H.; Doudna, J.A.; A conformational switch controls hepatitis delta virus ribozyme catalysis; Nature, 2004, 429, 201–205. 114. Wigley, D.B.; Teaching a new dog old tricks?; Structure, 1998, 6, 543–548.

436


115. Flick, K.E.; Jurica, M.S.; Monnat, R.J., Jr; Stoddard, B.L.; DNA binding and cleavage by the nuclear intron-encoded homing endonuclease I-PpoI; Nature, 1998, 394, 96–101. 116. Friedhoff, P.; Franke, I.; Meiss, G.; Wende, W.; Krause, K.L.; Pingoud, A.; A similar active site for non-specific and specific endonucleases; Nat. Struct. Biol., 1999, 6, 112–113. 117. Kuhlmann, U.C.; Moore, G.R.; James, R.; Kleanthous, C.; Hemmings, A.M.; Structural parsimony in endonuclease active sites: should the number of homing endonuclease families be redefined? FEBS Lett., 1999, 463, 1–2. 118. Hsia, K.C.; Li, C.L.; Yuan, H.S.; Structural and functional insight into sugar-nonspecific nucleases in host defense; Curr. Opin. Struct. Biol., 2005, 15, 126–134. 119. Chevalier, B.S.; Stoddard, B.L.; Homing endonucleases: structural and functional insight into the catalysts of intron/intein mobility; Nucleic Acids Res., 2001, 29, 3757–3774. 120. Hosfield, D.J.; Frank, G.; Weng, Y.; Tainer, J.A.; Shen, B.; Newly discovered archaebacterial flap endonucleases show a structure-specific mechanism for DNA substrate binding and catalysis resembling human flap endonuclease-1; J. Biol. Chem., 1998, 273, 27154–27161. 121. Newman, M.; Murray-Rust, J.; Lally, J.; Rudolf, J.; Fadden, A.; Knowles, P.P. et al.; Structure of an XPF endonuclease with and without DNA suggests a model for substrate recognition; EMBO J., 2005, 24, 895–905. 122. Gan, J.; Tropea, J.E.; Austin, B.P.; Court, D.L.; Waugh, D.S.; Ji, X.; Structural insight into the mechanism of double-stranded RNA processing by ribonuclease III; Cell, 2006, 124, 355–366. 123. Chevalier, B.; Sussman, D.; Otis, C.; Noel, A.J.; Turmel, M.; Lemieux, C. et al.; Metaldependent DNA cleavage mechanism of the I-CreI LAGLIDADG homing endonuclease; Biochem., 2004, 43, 14015–14026.

Part D Toxicological Aspects


16 Structural Studies on the MercuryII-Mediated T-T Base-Pair Using NMR Spectroscopy Yoshiyuki Tanaka and Akira Ono

16.1 The History of the MercuryII-Mediated Thymine-Thymine Base Pair Studies on the mercuryII–DNA interaction began with the paper by Katz in 1952.1 In that report, Katz performed light scattering and viscosity measurements on calf thymus DNA in a complex with mercury ions (HgII), and found that HgII can bind to DNA in a reversible manner. In 1954, Thomas followed Katz’s experiments with HgII titration experiments on calf thymus DNA using UV spectra.2 Yamane, Davidson and Dove performed extensive HgII titration experiments with UV spectra, using E. coli DNA3,4(1960, 1961), M. lysodeikticus DNA4(1961), poly-d(AT)4(1961), poly-rA5(1962), poly-U5(1962) and tobacco mosaic virus RNA5(1962). The results they obtained from these studies showed: (i) upon one HgII binding to DNA, two protons are released in the initial stage of HgII-titrations; (ii) in the HgII titration experiments of natural DNA (E. coli and M. lysodeikticus DNA), UV spectra showed an isosbestic point until the molar ratio of [HgII]/[mole of nucleotides of the DNA] became 0.5; (iii) in the case of poly-d(AT) (adenine (A)–thymine (T) alternating polymer), no isosbestic point was observed for the UV spectra of HgII titrations; (iv) in the case of poly-d(AT), the number of released protons upon HgII-binding (∆[H+]/∆[HgII]) decreased at a molar ratio ([HgII]/[mole of nucleotides of the DNA]) of 0.25; (v) in the HgII titration of poly-uridine(U), UV spectra showed an isosbestic point until the molar ratio of [HgII]/[mole of Metal Complex–DNA Interactions Edited by Nick Hadjiliadis and Einar Sletten © 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17629-3

440

Structural Studies on the MercuryII-Mediated T-T Base Pair

R

N

O

O

N

HgII N

O

O

N R

Scheme 16.1 O H II N N Hg

N R

N N

N

O

N R

Scheme 16.2 nucleotides of the DNA] became 0.5; (vi) a possible HgII binding site might be N3 of T/U. Based on the above results, Katz reconsidered the stoichiometry of HgII binding to DNA molecules6,7(1962), and proposed a chemical structure of the HgII mediated T-T base pair (T-HgII-T) as a possible interaction mode in DNA molecules8(1963) (Scheme 16.1). Furthermore, he presented a possible mechanism for DNA strand displacements based on the T-HgII-T pair. To obtain quantitative data, thermodynamic studies on HgII–nucleoside complexes were performed by Eichhorn and Clark9(1963), and Simpson10(1964). Simpson determined association constants of HgII to nucleosides under various pH values. In 1965, Nandi et al. proposed a unique interaction mode for an A-HgII-T base pair that is different from the T-HgII-T base pair11 (Scheme 16.2). In 1966, Gruenwedel and Davidson reported UV-monitored titration experiments on natural DNA molecules (calf thymus DNA, bacteriophage T4 DNA and M. lysodeikticus DNA) with methymercuric hydroxide (CH3HgOH).12 Based on the Katz model of the T-HgII-T pair, they porposed that methymercuric ions can denature DNA molecules by HgII binding to N3 of T and N1 of guanosine (G). These pioneering works stimulated many structural studies of the HgII-mediated T-T base pair in DNA duplexes, with crystallography, vibrational spectroscopy and NMR spectroscopy being applied.

16.2 Crystallographic Studies on HgII–Nucleobase Complexes There are two kinds of crystal structures for an HgII–complex with uracil (U) or thymine (T) (Figure 16.1). In 1971, Carrabine and Sundaralingam determined the structure of a HgII–uracil (U) complex13 (Figure 16.1a). In this structure, two uracil molecules are bound to HgII at their O4-carbonyl oxygen [O4(U)]; this structure was different, however, from the putative structure of the T-HgII-T pair proposed by Katz. In 1974, Kosturko, Folzer and Stewart reported a crystal structure of the

UV, UVCD and Vibrational Spectral Studies 441 (a)

H O

O

N N H

O

HgII

H N

O N

H

(b)

O N

N O

O HgII

N

N

O

Figure 16.1 Experimentally observed U-HgII-U13 and T-HgII-T14 base pairs in crystals. Solution conditions for crystallization: (a) 9.72 mM uracil and 4.86 mM HgCl2 at pH 4.2; (b) 20 mM 1-methylthymine and 10 mM HgO

1-methylthymine–HgII (2 : 1) complex14 (Figure 16.1b). Interestingly, this structure essentially reflects Katz’s model. This crystal structure demonstrated for the first time that N3 of T can also be an HgII binding site. Although these two crystal structures do not directly indicate that the binding modes observed in the crystals correspond to those observed in natural DNA, the determined structures were important experimental data for considering an HgII binding mode. On the other hand, there are studies exploring other types of HgII–U complexes and HgII complexes with other nucleobases15–28 (Figure 16.2). Beauchamp and coworkers reported crystal structures of methylmercury complexes with adenine and cytidine (Figures 16.2a–g).15–19 In 1988, Sheldrick and Gross also reported methylmercury complexes with 7-methylguanine (Figures 16.2h,i).20 Lippert and coworkers reported HgII complexes with 1,3-dimethyluracil at the C5-position (Figures 16.2j– l),23–25 an HgII complex with 9-methyladenine at the N7-position (Figure 16.2m)26 and bimetallic HgII,PtII complexes with 1-methylcytosine (Figures 16.2n,o).21,27 These data indicated the intrinsic potential of nucleobases to form various types of metal complexes.

16.3 UV, UVCD and Vibrational (IR/Raman) Spectral Studies As mentioned above, UV absorption spectra were employed from the early stages of studies on the HgII–DNA interaction. UV spectral studies12,29 were continued by Gruenwedel’s group, which developed their study into UVCD spectra.30–32 They presented comprehensive UVCD spectra of HgII–DNA complexes under various conditions (see references for details). In the case of vibrational (IR/Raman) spectra, complexations between methylmercuric ions and nucleosides were studied (Figure 16.3).33–44 In 1974, Mansy, Tobias and coworkers reported Raman spectra of a CH3HgII–uracil complex with an N3-HgII bond (Figure 16.3a),33 a CH3HgII–5′-GMP (guanosine-5′monophosphate) complex with an N1-HgII bond at neutral pH in H2O solution (Figure 16.3b)34 and a CH3HgII–5′-GMP complex with an N7-HgII bond at acidic pH in H2O solution (Figure 16.3c).34 Proton–HgII exchange processes are required to form the complexes (Figures 16.3a,b), and the derived Raman spectra provide reference data for the T-HgII-T base pair. There are also Raman spectra of CH3HgII–cytidine and CH3HgII–5′-AMP (adenosine-5′-monophosphate) complexes in H2O solution (pH 3, 5 and 7) (Figures 16.3d,e),36 the HgII–uracil complex with O4-HgII

442


Figure 16.2 Experimentally observed HgII complexes in crystals. References for the crystal structures and crystallization conditions are as follows: (a) and (b): 15; (c): 16,17; (d): 18; (e), (f) and (g): 19; (h) and (i): 20; (j): 24; (k): 23; (l): 25; (m): 26; (n): 21; (o): 27; (p): 22. Solution conditions for crystallization: (a) aqueous mixture of CH3HgNO3 (1 M aqueous solution) and adenine (67–100 mM aqueous solution) in 1 : 1 ratio; (b) aqueous mixture of CH3HgClO4 (1 M aqueous solution), NaOH (1 M aqueous solution) and 9-methyladenine in 2 : 1 : 1 ratio; (c) crystallization details are not given; (d) aqueous mixture of CH3HgNO3 or CH3HgClO4 (1 M aqueous solution) and adenine in 1 : 1 ratio; (e) CH3HgOH, CH3HgNO3 (∼0.5 M aqueous solution) and 1-methylcytosine (methanol) at 2 : 1 : 1 ratio; (f) CH3HgOH (1 M aqueous solution) and 9-methyladenine (acetonitrile) at 2 : 1 ratio; (g) aqueous mixture of CH3HgNO3 or CH3HgClO4 (1 M aqueous solution) and adenine (67–100 mM aqueous solution) in 3 : 1 ratio; (h) 54 mM CH3HgOH and 54 mM 7-methylguanine at pH 9–12; (i) 54 mM CH3HgOH and 18 mM 7-methylguanine at pH 1–3; (j) 18 mM Hg(O2CCH3)3 and 18 mM 1,3-dimethyluracil at pH 3.5; (k) 20 mM Hg(O2CCH3)(1,3-dimethyluracil-C5) and 20 mM KCN in water; (l) 47 mM Hg(O2CCH3)(1,3-dimethyluracil-C5) and 47 mM 9methyladenine at pH 1.2; (m) 44 mM Hg(NO3)2 and 88 mM 9-methyladenine at pH 2.5; (n) 6.8 mM HgO and 6.8 mM trans-[Pt(NH3)2(1-methylcytosine)2](NO2)2 at pH 5.8; (o) 9.9 mM Hg(O2CCF3)2 and 10 mM trans-[Pt(NH3)2(1-methylcytosine-N4)2](NO2)2, raising pH 1.8 to 2.9

bonds in the crystal phase (Figure 16.3f),37 the HgII–5′-UTP (uridine-5′triphosphate) complex with C5-HgII bonds in H2O solution (Figure 16.3g)37 and CH3HgII–inosine complexes with N1-HgII bond (pH 8) and an N7-HgII bond (pH < 2) in H2O solution (Figures 3h,i).38Also,Raman spectra of a CH3HgII–5′-TMP (thymidine-5′monophosphate) complex in H2O (Figure 16.3j), complexes between CH3HgII and 5′-TMP, 5′-AMP, 5′-GMP and 5′-CMP (cytidine-5′-monophosphate) mixtures in

UV, UVCD and Vibrational Spectral Studies 443

Figure 16.3 Compounds for which Raman/FTIR spectra were recorded. References for these compounds are as follows. (a): 33; (b) and (c): 34; (d) and (e): 36; (f) and (g): 37; (h) and (i): 38; (j), (k) and (l): 39; (m): 39,42; (n), (o), (p) and (q): 40; (r), (s) and (t): 41; (u): 39; (v) 43

444


H2O (Figure 16.3k), a CH3HgII–1-methylthymine complex in H2O (Figure 16.3l) and a crystalline HgII–1-methylthymine complex (Figure 16.3m) were presented.39 In addition, they presented IR spectra from solid samples of organomercury–guanosine/ thymidine complexes (Figures 16.3n–q).40 In 1982, Beauchamp, Savoie and coworkers also reported IR/Raman spectra of methylmercury–adenine complexes (Figures 16.3r–t).41 Interestingly, in 1977, Mansy, Tobias and coworkers recorded Raman spectra of calf thymus DNA in the presence of CH3HgIIOH (Figure 16.3u).39 This data provided Raman spectroscopic information on HgII and natural DNA molecules. As mentioned in Section 16.2, there are three types of HgII–T/U complexes, namely N3-, O4- and C5-mercurated ones. In the above noted vibrational studies, all types of complexes in water solutions were characterized by Raman spectra and several marker bands for detecting each HgII–U/T complex were found. Notably, not only solution samples, but also crystal samples prepared by means of Kosturko’s method14 (with N3-HgII bonds) and Carrabine’s method13(with O4-HgII bonds) were employed. This fact implied that the derived FT-IR/Raman spectra surely originate from the crystal structure and are not from any other structure. Therefore, the data obtained from the studies provides rigid and reliable basal data for the assignment of Raman bands from other systems. In 2001, Morzyk-Ociepa and Michalska reported very fine FT-IR and Raman spectra of solid samples of HgII complexes with 1-methylthymine.42 Furthermore, they performed quantum chemical calculations of vibrational spectra, and extensively assigned IR/Raman bands.42 In 2003, they reported FTIR and FT-Raman spectra of AgI-1-methyluracil complexes (U-AgI-U complex) (Figure 16.3v) that could be used as reference data for the T-HgII-T base pair.43 In 2004, they recalculated the theoretical frequency of vibrational spectra, and refined the assignments of Raman and IR bands.44 Therefore, these data can be applied to the interpretation of FT-IR/Raman spectra in relation to specific vibrational modes. For vibrational spectroscopy, sufficient amounts of accurate spectral data of HgII–nucleoside complexes have been collected, along with data on neutral thymine.45 Especially, in terms of the structural studies on T-HgII-T base pairs, HgII–ligand interactions within THgII-T base pairs can be precisely analysed.

16.4 NMR Spectral Studies 16.4.1 Nucleoside–Metal Ion Systems NMR studies on nucleobase–metal ion(ZnII) interactions in dimethyl sulfoxide (DMSO) began in the 1960s.46,47 These studies extended to nucleoside–HgII interactions.48,49 At that time, NMR chemical shifts and their perturbations upon metal salt titrations were used to understand the nucleoside–metal cation/counteranion interactions.46–52 In 1975, Sohma and coworkers extensively studied chemical shift perturbations of 1H and 13C nuclei in cytidine upon the addition of various metal cations and counteranions to study how they affect 1H and 13C chemical shifts.52

NMR Spectral Studies

445

In 1978, Marzilli and coworkers monitored the HgII binding to cytidine using C NMR spectroscopy. In their report, higher-field shifts of carbon atoms (C2, C4) were reported (Table 16.1).53 Next, they reported 1H and 13C NMR spectra of complexes between HgII and nucleosides (uridine, cytosine and adenosine) in dimethylsulfoxide (DMSO).54,55 In that study, they found that cytidine bound to HgII, probably at N3, but that uridine did not bind to HgII through its N3 atom in DMSO.54,55 This observation was somewhat controversial, since proton–HgII exchange reactions in H2O had already been proposed for uridine. In 1982, however, they found that proton–HgII exchange reactions in DMSO can occur at the N1 of guanosine ‘in the presence of triethylamine base’ (Table 16.1).56 Through the 13C NMR experiments, moderate but significant chemical shift perturbations were observed for the resonances from the carbon atoms adjacent to the metal-binding nitrogen atoms (Table 16.1). For the HgII–thymidine system, Buncel and coworkers in 1981, 1985 and 1986 reported 13C NMR spectra of the complexes (Table 16.1).57–59 In these reports, they isolated the complexes of HgII–thymidine and HgII–guanosine, and dissolved these complexes in DMSO.57–61 The derived solutions were used for NMR measurements. According to these data, significant lower-field shifts of the 13C nuclei adjacent to HgII-binding sites (N3) were observed for C2 and C4 nuclei (Table 16.1) (see references for details).58,59 These data strongly supported the HgII-T complexation through N3, as proposed by Katz.8 Interestingly, Buncel and coworkers reported that proton– HgII exchange reaction occur at guanine to give a HgII–guanosine covalent complex through N1. This is because they observed the lower-field shifts of the 13C chemical shifts of C2 and C6 (N1-neighbours) (see references for details).57–59 In 1982, 1983 and 1986, Buchanan and coworkers monitored HgII binding processes to guanosine, cytosine, adenosine and inosine in DMSO by using 15N NMR spectroscopy (Table 16.2).62–64 From their experiments, very large chemical shift perturbations were observed for N7 of guanosine (∼20 ppm higher field), which meant that the HgII binding site in guanosine was not N1 but N7 (Table 16.2)62 Under their conditions, proton–HgII exchange did not occur. However, this appeared to be because NMR spectra were recorded in the DMSO solution without any base, which is indispensable for a deprotonative metallization of guanosine at N1 in DMSO, based on the observations by Marzilli and coworkers.56 Next, Buchanan and coworkers studied the methylmercuryII complexes with 5′-GMP and 3′-CMP using 15 N NMR spectroscopy.64 It was pointed out that at pH 8 in H2O, HgII binding sites became the N1 of 5′-GMP with proton–HgII exchange processes (Table 16.2).64 In another water solution system, Polak and Plavec, in 1999, reported 5N NMR spectra of HgII complexes with methyl esters of 5′-GMP and 3′-GMP (guanosine3′-monophosphate) (Table 16.2).65 In their report, both compounds were seen to bind to HgII through their N7 atoms. They postulated that chelation of HgII with N7 and the phosphate group makes HgII bind to N7. With these data, however, HgII binding sites of various nucleosides were determined with 15N NMR spectroscopy. In a later section (Section 16.4.2), 15N NMR data on the proton–HgII exchange systems for a DNA duplex are described. In respect of another interacting point, 199Hg NMR spectra of HgII–nucleoside complexes were recorded by Norris and Kumar, in 1984.66,67 They reported 199Hg 13

Hg (0.5 M) CH3HgII (0.2 eq.)

−2.2 ∼−0.5

N3 −2.7 – ∼0.7

−1.24 – −0.66 −1.25

+0.5 +0.3 +0.9 +1.1

−0.2 +0.5 +0.1 +0.9

−1.8

−0.7

+1.7 ∼+1.5

–

– – – – –

−2.63 −2.76 −1.27 −2.60

n.a. n.a.

n.a. n.a. n.a. n.a. – – – –

n.a. n.a. n.a. n.a.

+0.5 n.a. −1.78 n.a. – n.a. −0.37 n.a. −1.42 n.a. +2.1 n.a. +0.2 n.a. n.a. ∼+1

5-CH3

C5

+0.4 +1.04 – +0.24 +0.98 −0.5 −2.0 ∼0

n.a. n.a. n.a. n.a.

C8

+0.9 ∼−0.2

n.a. n.a.

+0.13 +0.07 +0.48 +0.32 +2.5 – +2.3 +1.1

−1.04 +1.61 +1.75 −0.37 +0.67 −0.85 +1.88 −0.03 N.D. −0.11 −0.04 – – – –

–

+4.6 +2.75 +3.04 +2.58 −0.40 +5.6 +3.4 ∼+3

0.0 −0.3 +0.4 +0.3

C6

53 64

65 65 65 65 103 103 76, (105) 104

56 56 56 56

58, (59) 56 56 56 56 58, (57) 57 64

58, (59) 58 80 80

Reference

a Chemical shift changes are listed in ppm. Positive and negative values represent lower-field shifts and higher-field shifts, respectively. MepdG: methylester of 5′-GMP; dGpMe: methylester of 3′-GMP; n.a.: not applicable; N.D.: not detected; –: not recorded. b For DNA/RNA oligomers, the chemical shift changes of the underlined residues are indicated.

cytidine (0.2 M)/DMSO/3′-CMP/H2O (pH 6.0)/-

II

N7 N7 N7 N7 N7 N7 N7 N7

HgII (2 mM) HgII (2 mM) ZnII (20 mM) ZnII (20 mM) ZnII (13.0 mM) ZnII (7.5 mM) CdII (5 eq.) PtII (15 mM)

MepdG (10 mM)/H2O (pH 7.5)/dGpMe (10 mM)/H2O (pH 7.5)/MepdG (10 mM)/H2O (pH 7.5)/dGpMe (10 mM)/H2O (pH 7.5)/d(ATGGGTACCCAT)2b (1.7 mM)/ H2O (pH 6)/d(ATGGGTACCCAT)2b (1.7 mM)/ H2O (pH 6)/r(GGACGAGUCC)2b/ H2O (pH 6.0)/d(TGGT) (15 mM)/ H2O (pH 6)/-

+0.69 – +0.88 +0.54

+0.8 −0.48 – −0.18 −0.65 0.0 +0.9 ∼−0.5

+2.4 +2.6 N.D. N.D.

C4

– – – – – – – −0.5

– – – –

HgII (0.7 M) HgII HgII (0.7 M) HgII (0.7 M)

guanosine (0.2 M)/DMSO/2′-deoxyguanosine/DMSO/inosine/DMSO/1-methylguanosine (0.2 M)/DMSO/-

+4.0 +3.90 +4.07 +5.69 +0.23 +3.9 +2.6 ∼+3

+2.7 +2.6 N.D. N.D.

C2

– – – – – – – –

N1 N1 N1 N1 N1 N1 N1 N1

HgII (0.5 eq.) HgII (0.7 M) HgII HgII (0.7 M) HgII (0.7 M) CH3HgII (1.0 eq.) CH3HgII (1.0 eq.) CH3HgII (1.0 eq.)

guanosine (isolated G-HgII-G)/DMSO/guanosine (0.2 M)/DMSO/Et3N (0.4 M) 2′-deoxyguanosine/DMSO/Et3N inosine/DMSO/Et3N 1-methylguanosine (0.2 M)/DMSO/Et3N (0.4 M) guanosine (isolated CH3HgII-G)/DMSO/inosine (isolated CH3HgII-I)/DMSO/5′-GMP (isolated CH3HgII-GMP)/H2O (pH 8.0)/-

N3 N3 N3 N3

HgII (0.5 eq.) CH3HgII (1.0 eq.) HgII (2.0 mM) HgII (2.0 mM)

Site

thymidine (isolated T-HgII-T)/DMSO/thymidine (isolated CH3HgII-T)/DMSO/d(GCGCTTTTGCGC)b (2 mM)/ H2O (pH 6.0)/d(GCGCTTTTGCGC)b (2 mM)/ H2O (pH 6.0)/-

Metal

C chemical shift perturbations upon metal-ion bindinga

13

Sample (M)/solvent/additive (M)

Table 16.1

446 Structural Studies on the MercuryII-Mediated T-T Base Pair

(2.0 mM)/H2O (2.0 mM)/H2O (2.0 mM)/H2O (2.0 mM)/H2O

(4.0 mM) (4.0 mM) (4.0 mM) (4.0 mM)

N1

N7 N7 N7 N7 N7 N7 N7 N7 N7 N7 N7 N7 N7 N7 N7

HgII (0.5 M) HgII (2 mM) HgII (2 mM) HgII (0.75 eq.) CdII (2.5 mM) CdII (12 mM) ZnII (0.7 eq.) ZnII (0.5 M) ZnII (20 mM) ZnII (20 mM) ZnII (4.9 mM) ZnII (4.9 mM) ZnII (12 mM) ZnII (1 eq.) ZnII (0.5 M)

guanosine (0.5 M)/DMSO MepdG (10 mM)/H2O (pH 7.5) dGpMe (10 mM)/H2O (pH 7.5) inosine/DMSO r(GGACGAGUCC)2b (0.5 mM)/ H2O (pH 6) r(CGGAGUUGGC) · r(GCCAAAGCCG)b (3 mM)/ H2O (pH 6.8) inosine/DMSO guanosine (0.5 M)/DMSO MepdG (10 mM)/H2O (pH 7.5) dGpMe (10 mM)/H2O (pH 7.5) d(GGTACCGGTACC)2b (0.29 mM)/H2O (pH 5.0) d(GGTACCGGTACC)2b (0.29 mM)/H2O (pH 5.0) r(CGGAGUUGGC) · r(GCCAAAGCCG)b (3 mM)/ H2O (pH 6.8) guanine (DFT calculation) adenosine (0.5 M)/DMSO

∼+1 −18.0 −11.9 ∼−22 ∼−0.5 – – ∼0 +2.7 – ∼0 ∼0 – – – – – – ∼0

−9.5 −1.0 +1.5 ∼−1 ∼0 – – ∼0 −0.6 – ∼0 ∼0 – – – – – – −1.6

∼+2.5 – – n.a. +3.1 – n.a. ∼+2 – – – – – – +1.4

+7.4 +1.2 ∼+10

∼+2

N.D. N.D.

−3.3 +2.4

+49.9 +15.3 +9.4

– – – –

NH2

∼+4

+30.3 +35.3 +29.9 +30.8

N3

∼+1

– – – –

N1

−1.2

−0.5 −0.7

∼+2

∼−20.5 −1.5 −0.5 −4.8 −19.6 −20.5 ∼−15 ∼−20 −8.4 −4.2 −4.0 −14.8 −20.0 −14.8 −7.5

n.a. n.a. n.a.

– – – –

N7

∼+4 N.D. N.D. ∼+2 +2.5 – ∼+1 ∼+3 +0.9 +1.0 7

+6 O

-O -O

pH < 7

HCrO4-

O

Chromate L

L

+6

CrO42-

Hydrogen Chromate

L

Cr

+3

Cr

L L

L

CrL6 Chromium(III) Figure 17.1 The major forms of physiologically relevant chromium species

Introduction

465

are also able to cross the cell membrane, but rely solely on passive diffusion and are therefore taken up approximately three orders of magnitude more slowly than Cr(VI) ions.24 Once in the cell, Cr(VI) is quickly reduced by a variety of intracellular reductants including ascorbic acid and reduced thiols, such as cysteine and the ubiquitous tripeptide glutathione.25 Ascorbic acid is the most kinetically favoured of these reductants26 and is thought to be the major intracellular reductant of Cr(VI). However, the exact contribution of the different reductants towards intracellular chromate reduction and genotoxicity is not known. Irrespective of the exact nature of the reductant, the ensuing rapid intracellular reduction of Cr(VI) to Cr(V), Cr(IV) and finally to Cr(III) creates a concentration gradient across the cell membrane that favours the continual uptake of Cr(VI) from the extracellular matrix. To adequately understand the mechanism of how chromium interacts with DNA it is first necessary to define the DNA reactive intermediates that are formed during the intracellular reduction process of Cr(VI). 17.1.3 Reactive Intermediates Formed During Reduction of Chromate The determination of a mechanistic pathway of cancer formation by Cr(VI) has been complicated by the wide variety of reactive species formed during the intracellular metabolism of this metal. These pathways are further complicated by the fact that many of these reactive intermediates may give rise to similar DNA-damage products, making it a challenge to associate specific reactants with the final premutagenic DNA lesion. As stated above, there are three intracellular chromate reductants that have been widely studied and are considered to be the primary players in chromate metabolism and the ensuing DNA damage. These are ascorbate, glutathione and, to a lesser degree based on low physiological concentration, hydrogen peroxide. Reduction of chromate by all three of these compounds forms both radical species and high valent metal species. While most research has focused on the radical-mediated pathways, it is becoming clear that a direct or metal-mediated pathway may play as great or greater role in DNA damage induced by chromate. Cr(VI) Reactions with Ascorbate The in vitro reaction of Cr(VI) with ascorbate has been extensively studied. Ascorbate is kinetically one of the best physiological reductants of chromate and was found to be the major reductant in rat liver, lung and kidney ultrafiltrates.27 Cr(VI) reduction by ascorbate has been found to form a number of different radical species including the ascorbyl radical and the carbon dioxide radical cation (Figure 17.2; I and II respectively) as well as high valent ascorbate-ligated Cr(V) and Cr(IV) species (Figure 17.3A). The relative yield of each of these reactive species is dependent upon ascorbate to Cr(VI) ratios.28,29 However, since Cr(VI) uptake is essentially steady-state from the extracellular milieu, and ascorbate concentrations in human cells are variable depending on dietary status, defining a Cr(VI):ascorbate ratio and the likely reactive intermediates that will be formed is unpredictable at best. Gross DNA-damage studies using plasmid DNA however, have implicated the high-valent metal species generically shown in Figure 17.3A as the likely reactive intermediates

466

Cr-Induced DNA Damage and Repair Carbon-centred radicals from ascorbate O

O

+ O

CO2

OH

O OH

I

II

Oxygen-centred radicals from hydrogen peroxide +

1

O2

+

OH

O2

IV

III

V

Sulfur-centred radical from glutathione H N

O

O O

NH

O

S

H3N

O

VI

O

Figure 17.2 Compilation of reductant-based radical species observed upon reduction of chromate

O

O

+5

O

Cr

A)

+4

O O

O

O O

O

O S

B)

N

O

Cr

O

O

+5 Cr

+4

S

N

N

Cr N

S

S O

O O

C)

O

O Cr

O

O

+5 O O

Figure 17.3 Putative high-valent intermediates of chromium arising during the intracellular reduction of chromate by: (A) ascorbic acid, ( B) glutathione and (C) hydrogen peroxide

Introduction

467

responsible for lesion formation.30–32 Perhaps more interesting is current research suggesting that ascorbate, classically thought of as an antioxidant, may serve as a pro-oxidant with regard to intracellular metabolism of Cr(VI).33 There is also recent evidence that ascorbate-ligated high-valent chromium species may have multiple roles in both oxidative DNA damage and in the formation of Cr-DNA adducts, as detailed in later sections of this review. Cr(VI) Reactions with Glutathione Glutathione, a tripeptide of g -glutamylcysteinylglycine, is likely to be another major player in the intracellular reduction of Cr(VI). Reduction of Cr(VI) in the presence of glutathione is significantly slower than that observed with ascorbate, but glutathione is generally present in cells at considerably higher (millimolar) levels than ascorbate.34 A number of reactive intermediates have been associated with the Cr(VI) metabolism process by glutathione.35–37 Among these are the glutathionyl radical (Figure 17.2; VI), which may react with molecular oxygen to produce the superoxide anion38 (Figure 17.2; V). Similar to the reactions observed with ascorbate metabolism of Cr(VI), glutathione metabolism of this metal also forms high-valent Cr(V) species (Figure 17.3B) that have been observed by electron spin resonance spectroscopy, ESR.35,36 Once again, the ultimate species that may be responsible for DNA damage during the glutathione-dependent metabolism process is unknown. Much like ascorbate, glutathione’s normal role of maintaining cellular redox homeostasis may be altered to one of a pro-oxidant with regard to Cr(VI) metabolism. Cr(VI) Reactions with Hydrogen Peroxide The reactions of Cr(VI) with hydrogen peroxide have been extensively studied, although this may be due more to the ability of this reaction to induce significant DNA modifications than with any cellular reality. Hydrogen peroxide is highly reactive with metal species and as such is kept at very low levels in the cell, which would seem to preclude any significant metabolic reduction of Cr(VI) by this species.39,40 However, a case could be made for reactions of this nature in the mitochondria where a steady-state level of hydrogen peroxide is produced during respiration accounting for up to 5% of total cellular oxygen consumption.41 In general, the reaction of Cr(VI) with hydrogen peroxide generates the classical reactive oxygen species, ROS, that have been observed with the Fenton metals of iron and copper. These include singlet oxygen, the hydroxyl radical and the superoxide anion (Figure 17.2; III, IV and V respectively).42,43 A high-valent Cr(V) tetraperoxochromate species (Figure 17.3C) has also been observed by ESR that is highly reactive with DNA.44 This species is only formed in extremely alkaline solutions and is probably not physiologically relevant. The DNA reactivity of ROS species have been extensively reviewed elsewhere and will not be discussed to any length in this review. However, it should be noted that many of the same DNA lesions formed by these classical ROS are now being found to arise from a direct or metal-mediated type of oxidation reaction with DNA.

468

Cr-Induced DNA Damage and Repair

17.1.4 General Types of DNA Damage Associated with Chromate A wide variety of genetic damage that may initiate cancer formation has been observed following Cr(VI) exposure. This wide array of genetic damage has confounded the chromium research community and there is still little general consensus on the mechanism(s) of DNA damage by Cr(VI), even though it is one of the most extensively studied carcinogenic metals. This damage includes DNA base and sugar oxidation, Cr-DNA adducts, DNA–protein crosslinks, DNA–DNA crosslinks and DNA strand scission.45 The downstream effects of these lesions, if not outright toxicity, can be lesion accumulation, mutation and apoptosis resistance. Cells that do not undergo apoptosis due to the accumulation of chromium-induced genetic damage will develop gene mutations,46 chromosomal aberrations47,48 and sister chromatid exchanges.47–49 This genetic damage caused by hexavalent chromium exposure is of considerable concern for both industrial chrome workers and the general population. Even with improved industrial practices, some workers are still exposed to as much as several hundred µg of Cr(VI) m−3 of air.47 The general public is also at risk due to environmental chromium pollution from a legacy of 150 years of often illconceived landfill practices, as well as pollution stemming from the mining, smelting and wood treatment industries. There are 719 active superfund sites in the United States with chromium listed as at least one of the major contaminants present.50 Automobile emissions and cigarette smoke are another source of nonoccupational exposure to Cr(VI).47 Understanding the mechanisms and types of chromiuminduced genotoxicity is essential for combating the effects of occupational and environmental exposure to Cr(VI) compounds. The remainder of this review focuses on specific types of DNA damage that Cr(VI) compounds are known to form, the current knowledge of how endogenous repair processes play a role in the removal of these adducts and, the induction of downstream cellular effects from these DNAdamaging processes.

17.2 Nucleobase Oxidation 17.2.1 Cr(VI)-Induced Damage at DNA Nucleobases Once Cr(VI) has been taken up by a cell, it is rapidly reduced.25 The highly reactive, metastable intermediate oxidation states of Cr(V) and Cr(IV) are created as Cr(VI) is reduced to the more stable Cr(III).51–53 This reduction scenario sets up two major pathways that can lead to oxidation of the nucleobases. One pathway involves the formation of radical species as detailed above, but with primary emphasis placed on reactive oxygen species (ROS). Through Fenton-like reactions with H2O2, both Cr(III)53 and Cr(VI)54,55 have been shown to be capable of producing hydroxyl radicals that can attack the electron-rich aromatic rings of the nucleobases. The ROS pathway is favourable in the presence of high concentrations of H2O2 (0.5 mM– 27 mM),53–55 which have been used for many in vitro studies. However as stated previously, under normal physiological conditions the concentration of H2O2 in a

Nucleobase Oxidation

469

cell is much lower than 1 mM,56 indicating that oxidation of nucleic acids by ROS is likely to play a small role in the overall nucleobase oxidative damage that has been observed in cells exposed to Cr(VI). A second pathway that is thought to account for the majority of nucleobase oxidation induced by Cr(VI) exposure involves the direct abstraction of an electron from nucleic acids by the Cr(V) and/or Cr(IV) intermediate species formed during Cr(VI) reduction.57 The vast majority of nucleobase oxidation observed during the Cr(VI) reduction process occurs at guanine residues. Guanine is preferentially oxidized by ROS and by high-valent chromium intermediates because of its lower one-electron reduction potential relative to that of the other three bases (Table 17.1).58–60 The reduction potential becomes even lower for 5′ guanine residues when a run of consecutive guanine residues is present in a DNA sequence (Table 17.1).58–60 The most recognized lesion formed by a single electron abstraction mechanism from a guanine residue is 7,8-dihydro-8oxo-2′-deoxyguanosine (8-oxoG) (Figure 17.4). In fact, 8-oxoG is often viewed as the primary genetic marker of oxidative stress23 and if not repaired is mutagenic, as 8-oxoG can mispair with adenine during DNA replication leading to G:C→T:A transversion mutations.61 Table 17.1 Reduction potentials for nucleosides and nucleotide sequences58–60 Nucleosidea

E (V vs NHE)

DNA Sequence

E (V vs NHE)

Guanine (G) Adenine (A) Cytosine (C) Thymine (T) 8-oxoG

1.29 1.42 1.6 1.7 0.7

GGG GG GA GC GT

0.64 0.82 1.00 1.15 1.16

a

The reduction as listed is for the radical cation of the corresponding nucleoside or nucleotide.

O N N R

NH N

O

O Crn

NH2

Crn-1

N N R

OH

NH N

NH2

H

N

HO

N R

NH N

NH2

- e-

Guanine

- H+ O H N

NH

O N R

N

NH2

8-oxoG

Figure 17.4 Sequential single electron oxidation mechanism of a guanine residue that gives rise to the premutagenic 8-oxoG lesion

470

Cr-Induced DNA Damage and Repair O

N N R

O NH

N Guanine

LxCrn NH2

O L Crn x

N

NH

O N R

O

LxCrn-2 O

N

NH2

H N N R

NH N

NH2

8-oxoG

Figure 17.5 Two-electron oxidation mechanism that gives rise to 8-oxoG

High-valent metal species such as those observed to arise from Cr(VI) reduction (Figure 17.3) have the potential to generate the same 8-oxoG lesion, but through a different mechanism than that of the single-electron oxidation shown in Figure 17.4. This mechanism involves the transfer of the axial oxo atom (a twoelectron oxidation mechanism) of the Cr(V) or Cr(IV) complex to the substrate through a discrete innersphere interaction (Figure 17.5). While this mechanism has yet to be shown to form 8-oxoG from guanine, it has been shown to occur for a number of other substrates.62,63 Validation of this mechanism is possible using isotopically labelled oxygen in the oxo-atom-containing chromium complex, but is problematic due to the rapid exchange of the oxo-atom with the surrounding water.62 However, a similar mechanism for 8-oxoG formation has been observed with highvalent ruthenium-oxo complexes.64 For years 8-oxoG was considered the terminal (and most relevant) oxidation product of guanine in DNA. However, the reduction potential of 8-oxoG is considerably lower than that of the parent guanine residue (Table 17.1) making it a prime target for further oxidative events. While it may seem unlikely for two discrete oxidative events to occur at the same DNA base (without some sort of a cage effect) recent studies have shown the formation of further oxidized products of 8-oxoG both in vitro65 and in cellular systems.66 A mechanism referred to as charge transfer has been proposed to explain this effect. When an oxidative event occurs at one base in a DNA sequence it can be translated through the DNA to an electron sink – a nucleoside with a lower reduction potential than the neighbouring nucleosides.67 The stacked bases in B-form DNA create an array of overlapping π-orbitals that facilitate charge transfer through the DNA sequence.68 8-OxoG lesions can act as a sink for charge transfer because of their lowered reduction potential relative to that of the four undamaged nucleosides. When an 8-oxoG lesion is present within approximately 10–20 base pairs (presumably from an initial oxidative insult and greatly dependent on the DNA sequence)68 of the site of a secondary electron abstraction from a nucleobase, the resultant charge can be transferred to an 8-oxoG residue, causing a further oxidative event at that lesion. The low reduction potential of 8-oxoG combined with its ability to act as a sink for charge transfer in a DNA sequence suggests that 8-oxoG is not likely to be the end product of an oxidative attack on the nucleobases. One of the further oxidized lesions that has been shown to form both in vitro65 and in cellular systems66 is spiroiminodihydantoin (Sp). Sp lesions can be generated by a single electron abstraction from 8-oxoG (Figure 17.6),65 which can be achieved by both Cr(V) and Cr(IV)

Nucleobase Oxidation O H N

- e-

NH

O N

N R

NH2

8-oxoG

N

H N

- H+

NH

O N R

O

O H N

471

- e-

N

O N R

NH2

N

NH2

H2O H

O O HN O

NH N R

N H

acyl shift NH

Spiroiminodihydantoin, Sp

H N

O

O N

O N R

N

NH2

5-OH-8-oxoG

Figure 17.6 Single-electron oxidation mechanism of 8-oxoG to give the Sp lesion65

intermediates during reduction by ascorbate. In this mechanism, the initial oneelectron abstraction forms an 8-oxoG radical cation intermediate that is a substrate for nucleophilic addition of water to form the 5-hydroxy-8-oxoG intermediate that undergoes rearrangement via an acyl shift to give Sp (Figure 17.6). It should be noted that while the overall reaction is a two-electron oxidation, the mechanism invokes two discrete one-electron abstraction processes. Recently, an alternative mechanism has been proposed for the formation of the further oxidized product Sp from 8-oxoG.69 This mechanism is similar to that for the formation of 8-oxoG from guanine through the oxo atom transfer mechanism proposed in Figure 17.5. In this mechanism, an inner-sphere oxidation complex is formed with the 8-oxoG lesion, resulting in transfer of the oxo atom from the Cr(V) species to the C5 of 8-oxoG (Figure 17.7).69 This mechanism was confirmed using isotopic labelling, although it is still unclear whether it is viable in cellular systems since it would require a direct interaction between the chromium complex and the 8-oxoG lesion. The formation of Sp helps to explain the high level of mutations, and the varying types of mutations, observed in Cr(VI) treated systems, that cannot be accounted for by the formation of 8-oxoG lesions alone. Sp, like 8-oxoG, can produce G:C→T:A transversion mutations, but at a much higher level than 8-oxoG. In addition to the G:C→T:A transversion mutations, Sp can also produce G:C→C:G transversion mutations and polymerase arrest both in vitro and in cellular systems.70–72 There are potentially a wide variety of nucleobase lesions that have the ability to be formed from multiple chromium-induced oxidation events at guanine in DNA. 8-OxoG and Sp are the only two that have been observed in cellular systems and as such are the main focus in this part of the review. A lesion similar to Sp, guanidinohydantoin (Gh), has been shown to form with Cr(V) treatment in vitro from the 8-oxoG nucleoside,73,74 but has yet to be identified as arising in cellular systems. Figure 17.8 shows a more comprehensive picture of the wide variety of nucleobase


472

H N

X NH

O N R

O

O

O

N

RN

Cr X

+5 X

X

-2e-

NH

N

O O

NH2

N H

8-oxoG

O

O

X

Cr X 5+

N

O

H N N R

HO-

HN O

N N R

N H

H

acyl shift

HO

O

H O N

NH2

NH

N R

H2O

O

H N

O N

NH

O N R

OH-

NH2

NH2

N

nucleophilic attack ring opening

OO

NH

N

NH2

OH

Spiroiminodihydantoin, Sp Figure 17.7 Formation of Sp from 8-oxoG via a two-electron, oxo atom transfer mechanism69 O

O O

H N

HN

NH

O N R

NH2

N 8-oxoG

N R

O

N H

NH2

Spiroiminodihydantoin, (Sp)

O

HN

N

O

HN

NH

O

NH

O N R

N H

NH2

N R

N

NH2

Guanidinohydantoin, (Gh) Oxidized Guanidinohydantoin, (Ghox) O

O N

RHN

O

NH2 NH2

NH2 N

RHN

Oxazolone (Oz)

Imidazalone (Iz) O

H N

O

N

O

O NH

RN O Parabanic Acid (Pa)

HO

RHN

O

Oxaluric Acid (Oa)

Figure 17.8 Oxidative nucleobase lesions considered possible to arise from Cr(VI) exposure. Only 8-oxoG and Sp have thus far been observed in cellular systems

Nucleobase Oxidation

473

lesions that may be induced by Cr(VI) exposure. It should be noted that, with the exception of Sp and Gh, none of these other further oxidized lesions have yet been identified as arising from Cr(VI) treatment of DNA in vitro or in cellular systems. They have, however, been identified as arising in vitro in a number of other oxidizing systems. 17.2.2 Repair of Nucleobase Damage Repair of oxidized nucleobases is necessary to prevent the perpetuation of mutations in the genetic code. Oxidized DNA lesions such as 8-oxoG and Sp are typically repaired via the base excision repair (BER) pathway. There are several glycosylases that have been shown to repair the oxidized nucleobase lesions induced by Cr(VI) exposure. Table 17.2 provides a list of those glycosylases and some of the lesions that they are known to recognize. OGG1 is a mammalian glycosylase which is a homologue of the bacterial BER enzyme MutM and functions to repair 8-oxoG lesions.75 OGG1-deficient mice have been shown to accumulate 8-oxoG lesions over their wild-type counterparts.76 Another bacterial BER enzyme, endonuclease VIII (Nei), has been recently characterized77 and shown to repair a number of further oxidized DNA lesions including Sp.78 Nei deficient Escherichia coli have been shown to accumulate Sp lesions, but not 8-oxoG lesions66 similar to that reported in OGG1deficient mice indicating that the Nei repair enzyme is directed at the repair of these secondary oxidation products. Recently, three mammalian homologues of the Nei gene were identified and labelled Neil1, Neil2 and Neil3 (Nei-like).79,80 The NEIL1 and NEIL2 proteins were initially found to recognize cytosine-derived lesions80 and 8-oxoG lesions present in DNA bubble structures generated during transcription or replication.81 Recently NEIL1 and NEIL2 have been shown to efficiently recognize the oxidized lesions Sp and Gh, but have very poor, to no, affinity for 8-oxoG in normal B-form DNA.82 NEIL1 was shown to cleave Sp and Gh in both singleand double-stranded oligonucleotides, while NEIL2 showed cleavage of Gh in both substrates, but could not cleave Sp in double-stranded oligonucleotides.82

Table 17.2 BER glycosylases known to repair Cr(VI) induced nucleobase lesions. Some of these glycosylases are known to repair additional lesions not listed here but those additional lesions have not yet been associated with Cr(VI) induced genetic damage Prokaryotic BER glycosylase

Known nucleobase substrates (for prokaryotic BER only)

Mammalian BER homologuesa

MutM (Fpg)

8oxoG:C, Sp:C, Gh:C, 8oxoG:G, Sp:G, Gh:G 8oxoG:A, G:A Sp:A, Gh:A, 8oxoG:C, Sp:C, Gh: C, 8oxoG:G, Sp:G, Gh:G

OGG1

MutY Nei

MYH NEIL1, NEIL2

a The mammalian BER homologues generally do not have as wide a range of nucleobase substrates as that of the prokaryotic BER enzymes shown.

474


Repair of isolated nucleobase lesions by BER glycosylases involves a baseflipping mechanism. A damaged guanine residue will be located by a glycosylase scanning the DNA. It will then be flipped out from the interior regions of the DNA strand and into the active site pocket of the attached glycosylase.83 This base-flipping mechanism allows the glycosylase to gain access to the glycosidic bond of the nucleobase lesion. Once the glycosidic bond has been cleaved, the base is released and the glycosylase forms a Schiff base at the abasic site.83 This Schiff base will remain in place until it is displaced by other proteins needed to further process the abasic site. These proteins include apurinic/apyrimidinic endonuclease 1 (APE1) which cleaves the phosphate backbone just to the 5′ side of the abasic site, DNA polymerases, which then fill in the gap in the DNA polymer, and finally DNA ligase, which repairs the nick in the phosphate backbone. Many BER glycosylases harbour intrinsic AP lyase activity, which would allow them to further process the abasic site. This activity is very weak in most glycosylases and they do not complete the belimination, but instead this cleavage step is typically carried out by APE1. NEIL1 and NEIL2 however, have strong lyase activity and are thought to complete b/delimination at the abasic site.84 Recently, NEIL1 and NEIL2 have shown to have considerably different roles with regard to replication associated repair (RAR). NEIL1, but not NEIL2, shows strong cell cycle dependence with the S-phase.85 NEIL1, but not NEIL2, has also shown enhanced activity in the presence of proliferating cell nuclear antigen (PCNA), a processivity factor for eukaryotic polymerase d. The NEIL1/PCNA-stimulated activity suggests that NEIL1 may play an active role in oxidized base lesion removal during replication.86 Recent modelling and dynamic studies on NEIL1 have shed light on why this BER enzyme is specific for the nonplanar, further oxidized lesions, such as Sp and other modified pyrimidines.80,81 These studies have shown that the binding pocket is somewhat flexible and is more accommodating of the pyrimidine-like further oxidized products, such as Sp and has complementary hydrogen-bond donor–acceptor properties.87 In contrast, the planar 8-oxoG lesion binds only shallowly into the pocket, has few hydrogen-bond donor–acceptor pairs, and has a solvent-exposed six-membered ring.87 Taken together, the recent data on the formation of further oxidized guanine lesions in DNA and the identification of specific repair mechanisms suggest that these lesions may be important in the genotoxicity of a variety of carcinogens, including Cr(VI).

17.3 Sugar Oxidation 17.3.1 Cr(VI)-Induced Damage at Deoxyribose Highly reactive Cr(V) and Cr(IV) species are capable of directly abstracting electrons from the conjugated nucleobases, as discussed above. Similarly, these chromium species can also oxidize the deoxyribose sugar by direct abstraction of hydrogen atoms. Studies have shown that chromium preferentially attacks the C4′

Sugar Oxidation

475

and C1′ hydrogens.57,88 This preference stems from the accessibility of the C4′ which, along with the C5′ hydrogens, project into the minor groove and are the most solvent accessible of the deoxyribose hydrogens in normal B-form DNA.86 Cr(V) is considered to preferentially attack the C4′hydrogen and not the C5′ hydrogen of deoxyribose because it is easier to remove a hydrogen from a tertiary rather than a secondary carbon despite the fact that the C5′ hydrogens are slightly more solvent accessible.88 While the tertiary C1′ hydrogen is also oriented towards the minor groove, it is not as readily accessible for oxidation by chromium, as H1′ abstraction requires that an oxidant bind to the minor groove and then orient itself directly towards the H1′ position.89 In the presence of hydrogen peroxide, (1 mM) abstraction of the C1′ hydrogen has been observed.90 Reactions with hydrogen peroxide are considered of little significance in the overall scheme of chromium-induced genetic damage, since physiological concentrations of hydrogen peroxide are much lower than 1 mM.56 However, a model metastable Cr(V) complex with the bis-tris buffer designated Cr(V)-BT has shown the formation of the deoxyribose oxidation product 5-methylene-2-furanone (5-MF) that would indicate a classical C1′ hydrogen abstraction mechanism (Figure 17.9).57 It is proposed that the correct orientation for hydrogen abstraction at C1′ by this Cr(V) complex is generated through interaction with the phosphate backbone of DNA.57 This Cr(V) complex also showed a smaller amount of C4′ oxidation product. A different model Cr(V) complex, bis(2-ethyl-2-hydroxybutyrato)oxochromat e(V) or Cr(V)-ehba (which is proposed to resemble the Cr(V)-ascorbate complexes formed during ascorbate-induced Cr(VI) reduction; Figure 17.3A), has been shown to abstract the C4′ hydrogen of deoxyribose. This C4′ hydrogen abstraction mechanism

OR O

OR O-

P

O

O H

O

H P OR

H H

O O

B

O-

Cr(V)-BT

B O-

P

+

O

B

O H

H H1'

H O

H

O

H

O P

O

O2

O-

5-methylene-2-furanone (5-MF)

OR + OR O

P

O-

OH free 3'-phosphate termini

Figure 17.9 Abstraction of the C1′ hydrogen by Cr(V)-BT leads to formation of the free base, the 5-methylene-2-furanone sugar product and frank strand scission57

476

Cr-Induced DNA Damage and Repair OR

O

P

OR O-

O

O H

O

H4' H P OR

O-

Cr(V)-ehba

P

O-

O H

H

O O

P

O2

H

H

H

B

O

H H

O O

B

B

H

H

H O

O-

Base propenal

OR

+ OR O

P

O-

O O

O

glycolic acid termini

Figure 17.10 Abstraction of the C4′ hydrogen by Cr(V)-ehba leads to formation of a base propenals and frank strand scission88,91

was assigned based on the formation of deoxyribose oxidation products of base propenals, glycolic acid (Figure 17.10) and, like the C1′ mechanism, gives rise to frank DNA strand breaks57,88,89,91 Also, much like the Cr(V)-BT model complex, Cr(V)-ehba has shown the ability to interact with phosphates to position the oxidizing metal towards the site of oxidation (the C4′ hydrogen atom of deoxyribose).92 Single-strand breaks, such as those initiated by abstraction of a C4′ or C1′ hydrogen by Cr(V) and/or Cr(IV) species, are one of the most common genetic lesions reported following exposure to Cr(VI).45 Single-strand breaks have been shown to cause base deletions,93 as well as gross chromosomal damage94 including sister chromatid exchanges47–49 and micronuclei formation.45 The gross chromosomal damage that has been observed in cells exposed to Cr(VI) has led to the labelling of hexavalent chromium compounds as clastogenic in addition to their classification as carcinogens. Another lethal downstream effect of single-strand breaks are doublestrand breaks created when a replication fork encounters a single-strand break and collapses.96 Based on the literature of Cr(VI)-induced single-strand breaks, it is apparent that rapid repair of these breaks is necessary to avoid genotoxicity. 17.3.2 Repair of Deoxyribose Oxidation Products Single-strand breaks occur frequently, even when a cell has not been exposed to a genotoxin and are in fact thought to be the most common type of genetic lesion.95

Chromium–DNA Binding

477

Cells have developed a repair system for single-strand breaks that employs the same proteins whether the damage is generated endogenously or exogenously.95 There are four steps involved in repair of single-strand breaks: detection, end processing, gap filling and ligation. Single-strand breaks are initially detected by poly (ADPribose) polymerases 1 and 2, which bind to the breakage site and recruit endprocessing proteins.91 Breaks induced by high-valent chromium species have a 3′ sugar fragment that can be removed by APE1 with the help of X-ray repair crosscomplementing protein 1 (XRCC1).93,95 Gap filling is accomplished by insertion of a single nucleotide followed by DNA ligase sealing the nick in the sugar backbone.

17.4 Chromium–DNA Binding 17.4.1 Formation of Chromium Adducts with DNA Chromium, whether it be in the +V, +IV or +III oxidation state, can be attracted to the negatively charged DNA–phosphate backbone. Chromium ions may associate themselves with DNA through initial electrostatic interactions with the phosphate backbone that can lead to the formation of a metal–ligand complex with the phosphate oxygens.92 However, the metal–ligand complex formed is generally transient in nature, as the electron lone pair necessary for coordination is shared between the two oxygens attached to the phosphate (Figure 17.11a). This type of chromium– DNA adduct has shown to be facile to exchange and nonmutagenic in cellular systems.97 As such, it is likely of little importance with regard to the overall mechanism of DNA damage by chromium that gives rise to carcinogenesis. Formation of covalent bonds between chromium and the DNA nucleobases is a more controversial proposal. A number of studies have suggested that chromium may interact with guanine in DNA at the N7 in much the same manner as that seen with cis-platinum.98 However, the lone pair of electrons at the N7 of guanine are delocalized into the purine ring and as such, coordination at this site would not be favoured for chromium. The unique acid/base properties, square planar geometry and Jahn–Teller distortions associated with cis-platinum that allow binding at the N7 of guanine in DNA99 do not exist with chromium. Despite DNA being a poor ligand, some studies have shown that a portion of reduced chromium appears to associate in a covalent manner with DNA. This binding appears only when the reduced chromium is formed during the reduction with Cr(VI) and not through interaction with the final, stable reduced chromium oxidation state, Cr(III). This suggests that the intermediate Cr(V) and/or Cr(IV) oxidation states of chromium may play a role in this binding mechanism. The DNA binding interaction with chromium has been proposed to be through the formation of ternary adducts, (Figures 17.11b,d).100 These ternary adducts involve bidentate chelation between the DNA phosphate backbone and the N7 group of a guanine residue. One study has shown that following treatment of DNA with Cr(VI) and a reductant, a large amount of chromium was initially associated with the DNA. However, a high salt wash removed all but 20% of the initial associated chromium. What chromium remained


478

O L L

R

L O

Cr L

N

P

L

NH

O N

O

O

O

L

N

O

O

a

NH

L NH2

N

O

H H

N

Cr

P

NH2

N

O

L

L

R

H

H

R

H

H H

R

H

H

H

b

H L R L O

P

L Cr

O L

O

O

H N

O N

O O H c

Figure 17.11 adducts100,102

H

R

N

L

R

NH

H

NH2

P

Cr

H

O

N

NH

L

O

N

O

NH2

N

O

H H

O

L

H

H

H

H

R

H

H

d

Proposed structures for covalent Cr(III)–DNA bound binary and ternary

associated with the DNA was chelatable with EDTA indicating that it was likely covalently bound.101,102 It should be noted, however, that no detailed structure has yet shown the veracity of these adducts. The ability of chromium to covalently bind to DNA would allow for the formation of chromium-mediated DNA–DNA and protein–DNA crosslinks that have been proposed to arise in Cr(VI)-treated cellular systems. This will be discussed in more detailed in subsequent sections. 17.4.2 MMR Repair and Toxicity Associated with Chromium–DNA Adducts Recent studies have indicated that the small ternary chromium–DNA adducts proposed in Figures 17.11b and d are recognized and repaired by the mismatch repair (MMR) system.103 Inactivating any of the four major MMR proteins (MSH6, MSH2, MLH1, PMS2) from human or mouse cells creates a Cr(VI) resistant phenotype.103,104 Cells proficient in MMR should undergo apoptosis following exposure to low doses of Cr(VI), as MMR plays a role in both repair and controlled cellular toxicity.103 MMR is proposed to recognize DNA bending caused by Cr–DNA adducts105 and then cleave a segment of DNA – up to 1000 base pairs – surrounding these adducts.106–108 This repair process is now thought to act as a signal for the activation of cell cycle arrest and eventually apoptosis by creating single-strand breaks,93 which are manifested into double-strand breaks following a round of replication.107 Because a round of replication is required before the double-strand breaks will be manifested, MMR-induced double-strand breaks appear most frequently in the G2 phase of the cell cycle.102,107 The ability of MMR to trigger G2 cell cycle arrest as well as

Chromium–DNA Binding

479

apoptosis is enhanced by supplementation of cell cultures with ascorbate.104,109 As mentioned earlier, ascorbate is thought to be the major intracellular reductant of Cr(VI) creating high-valent, highly reactive chromium intermediates. These studies further the hypothesis that ascorbate, normally viewed as an antioxidant, may act as a pro-oxidant in cells treated with Cr(VI). One specific characteristic of lung cancers caused by Cr(VI) – unlike other lung carcinogens – indicates that cells carrying MMR deficiencies will be prone to becoming cancerous. This characteristic is called microsatellite instability110 and is associated with a complete loss of MMR activity.111,112 Another unique characteristic of Cr(VI)-induced cancers is that lung tumours from Cr(VI) cancer subjects retain a wild-type p53 tumour suppressor gene.113 Therefore, constant exposure to Cr(VI) compounds may not be necessary for tumour formation, since one would thus expect the p53 gene to accumulate a large number of chromium-induced mutations. Cells lacking MMR show resistance to an initial pulse of Cr(VI). This would lead to accumulation of a small number of lesions and adducts, and then the lack of an MMR system would allow those cells to continue to generate spontaneous mutations without exposure to additional chromium. A model has been developed by Salnikow and Zhitkovich to explain how MMR deficient cells may actually be selected following exposure to hexavalent chromium compounds (Figure 17.12).113 Cr(VI) Exposure

Formation of Cr-DNA Adducts

MMR Proficient Cells

MMR Deficient Cells

G2 Cell Cycle Arrest

Apoptosis Resistance

Apoptosis

Cell Replication

Tumour Formation

Figure 17.12 Selection scheme for developing a pool of MMR-deficient cells that may explain the induction of a cancerous phenotype from chromate exposure113

480


Those cells that are MMR deficient will be apoptosis resistant following Cr(VI) exposure. While the normal MMR proficient cells undergo apoptosis, the now initiated MMR-deficient cells will continue to grow and divide creating a population of cells that harbour a host of genetic mutations. To date, these studies are the most complete that show a correlation between chromium adduct formation, the role of DNA repair and the downstream cellular effects that can lead to cancer formation.

17.5 Cr(VI)-Induced DNA Crosslinks 17.5.1 Formation of Crosslinks As mentioned earlier, another class of lesions that can form following exposure of DNA to high-valent chromium species are crosslinks. Crosslinking can occur between DNA and a protein, between two adjacent sites on the same DNA strand (intrastrand crosslink) or between neighbouring DNA strands (interstrand crosslink). Small amounts of chromium can covalently bind to DNA through interactions with the phosphate backbone116,117 and form ternary adducts as described above.100,102 Chromium(III)-mediated ternary adducts may also involve a linkage between the phosphate backbone and protein, or with adjacent or opposite N7 groups of guanine residues. These lesions would be the chromium-mediated protein and DNA crosslinks. Since the DNA phosphate backbone is a fairly weak ligand, crosslink formation involving covalently bound chromium atoms cannot fully explain the extensive formation of crosslinks observed in many chromium-treated systems. These type of coordinative adducts are the least well defined and, once again, no structural characterization have been carried out on these systems. An alternative proposal to explain the formation of protein and DNA crosslinks is an oxidative pathway. Formation of single-strand breaks due to deoxyribose oxidation57,88 and accumulation of 8-oxoG and Sp due to guanine oxidation82 following Cr(VI) exposure demonstrate that chromium can be a powerful DNA oxidant. Oxidation of the guanine residue, as well as the 8-oxoG residue, forms a radical cation, as depicted in Figures 17.4 and 17.6. The radical cation presents an excellent site for nucleophilic attack and, although water is not a strong nucleophile, the vast amounts of water present surrounding the DNA favours the formation of lesions such as 8-oxoG and Sp. A number of much better nucleophiles are also present around the DNA, but are substantially less abundant than water. These include amino acids, such as histidine and cysteine, either as small peptides or within complete proteins, as well as the phosphate backbone of neighbouring DNA strands. When these strong nucleophiles are in close proximity to a chromium-generated radical cation, they may out-compete water and a DNA adduct with that nucleophile will be formed rather than an oxidized base.69 Therefore, DNA–DNA and protein–DNA crosslinks could be formed without the discrete coordination of a chromium atom, but as a result of a single-electron oxidation induced by Cr(V) or Cr(IV) ions (Figure 17.13). While no structural evidence for this type of reaction in

Cr(VI)-Induced DNA Crosslinks 481 H N O N

O NH

N

N

dR Guanine

O

O

Cr(V)/Cr(IV)

H N

NH

O NH2

N

O

N

H N

Cr(V)/Cr(IV)

NH

O

NH2

dR

NH2

N

dR

8-oxoG

Protein or other nucleophile

R2

C

dR

N HN

NH2

Spiroiminodihydantoin, Sp

N

O

N

O H2 > 7 pH

O

HN CH CH2

R1 H N

H N

N

NH

dR

N O

O

O N

N H

NH2

Guanidinohydantoin, Gh

N

O N

H pH 2 O < 7

N

NH2

dR

DNA-ProteinCrosslink

Figure 17.13 Putative oxidative crosslinking mechanism showing the formation of a DNA– protein crosslink from the guanine oxidation pathway

DNA has yet been shown, the validity of this reaction with chromium has recently been demonstrated by Cr(V)-induced adduction of a lysine to an oxidized 8-oxoG nucleoside.69 17.5.2 NER and Recombination Repair of Crosslinks Regardless of how the crosslinks are formed, they greatly alter the normal conformation of B-form DNA. Such bulky lesions can prove to be lethal for the cell as they lea,d not only to premutagenic deletions, but also to replication and transcriptional blockage.45 Nucleotide excision repair (NER) and recombination repair are required to restore the DNA template to its undamaged form. Many protein crosslinks recruit NER mechanisms because they only form a minor distortion of the DNA phosphate backbone.117,118 This is because a large proportion of the protein crosslinks do not involve a complete protein, just individual amino acids or short peptide chains,119,120 creating a relatively compact lesion. These types of crosslinks create enough distortion to recruit NER repair proteins,45 but are not bulky enough to require recombination repair. There are five steps involved in NER: lesion recognition, unwinding of the damaged region, incision, excision and synthesis/ligation.121 High-valent chromiuminduced DNA–protein crosslinks are initially identified by Xeroderma pigmentosum group A (XPA)122 and replication protein A (RPA),123,124 which interact with one another and with the DNA lesion to enhance the specificity of lesion recognition.125 The XPA protein will then recruit Xeroderma pigmentosum group B (XPB) and

482


Xeroderma pigmentosum group D (XPD) helicases to unwind the DNA surrounding the damaged site. XPB will unwind the DNA in the 3′ to 5′ direction126 and XPD will unwind the DNA in the 5′ to 3′ direction127 from the site of the lesion. Once the DNA has been unwound, a patch of DNA ∼25–30 base pairs in length can be cleaved. The first incision is made on the 3′ side of the lesion by Xeroderma pigmentosum group G (XPG) endonuclease.128,129 This incision is made ∼3–5 base pairs to the 3′ side of the lesion130,131 and is required to be completed before a 5′ incision can be made.131 Excision repair cross-complementing 1 (ERCC1) and 4 (ERCC4) proteins will then form a complex and cleave the DNA phosphate backbone ∼22–25 base pairs to the 5′ side of the lesion.129,130,132 RPA remains bound throughout the incision step because it is required for facilitating release of the excised oligonucleotide.120 Following release of the oligonucleotide, proliferating cell nuclear antigen (PCNA) displaces the incision proteins,130 clearing the way for polymerase d or polymerase e to synthesize a new segment of DNA.131 The nicks in the phosphate backbone are then sealed by DNA ligase. Chromium-induced DNA crosslinks, as well as protein crosslinks involving a complete protein are thought to be the primary lesions obstructing DNA replication past guanine residues.132–134 Treatment of mammalian cells with nonlethal concentrations of Cr(VI) has been shown to greatly enhance recombination,135 lending support to the proposal that these types of crosslinks do form in cellular systems following Cr(VI) exposure. When a polymerase encounters one of these bulky chromiuminduced lesions, it will stall, causing the replication fork to collapse.136 Collapse of a replication fork creates a single-ended double-strand break.137 A variation of homologous recombination repair called replication fork repair is used to mend doublestrand breaks created by the collapse of a replication fork.138 As a side note, replication fork collapse can also be induced by the presence of an unrepaired single-strand break139 or by the presence of a rigid oxidized nucleobase lesion.96 Oxidation of the deoxyribose by high-valent chromium ions, as well as formation of secondary oxidative guanine lesions like Sp could thus trigger replication fork collapse if the subsequent single-strand breaks are not promptly repaired. Homologous recombination repair is also referred to as error-free repair,137 making it the most favourable form of recombination for repair of exogenous DNA damage. However, in the replication fork repair version of homologous recombination, sister chromatid exchanges can potentially arise if the leading strand template becomes covalently attached to the daughter lagging strand.139,140 Chromium compounds have been shown to induce sister chromatid exchanges,47–49 which may be partially attributed to homologous recombination repair following the collapse of replication forks at chromium-induced DNA and protein crosslinks.

17.6 Conclusions Even though chromium has been one of the most extensively researched metal carcinogens, there is still much that is not known about its fundamental mechanism of interaction with DNA. This current lack of understanding is due to the complexity

References

483

of the metabolic process of Cr(VI) that gives rise to the myriad of DNA reactive intermediates and the wide variety of putative DNA lesions that have been identified from this process. Recent studies focusing on DNA repair have helped expand our knowledge of how many of these lesions may be processed and how, if not repaired, may lead to the cancerous phenotype in cellular systems. However, there still remains much work that needs to be done in order to understand the structural nature of these lesions and the fundamental processes that give rise to them.

References 1. Nriagu, J.O.; Neiboer, E., Eds; Chromium in Natural and Human Environments, Vol. 20, John Wiley & Sons, New York, 1998, 1–20. 2. Katz, S.A.; Salem, H.; The Biological and Environmental Chemistry of Chromium, VCH Publishers, New York, 1994, 32–37. 3. Cohen, M.D.; Kargacin, B.; Klein, C.B.; Costa, M.; Mechanisms of chromium carcinogenicity and toxicity; Crit. Rev. Toxicol., 1993, 23, 255–281. 4. Pfeil, R.; Lungentumoren als Berufskrankung in Chromatbetrieben; Dtsch. Med. Wochensch., 1935, 61, 1197–1202. 5. Baetjer, A.M.; Pulmonary carcinoma in chromate workers I. A review of the literature and report of cases; Arch. Indust. Hyg. Occup. Med., 1950, 2, 487–504. 6. Baetjer, A.M.; Pulmonary carcinoma in chromate workers II. Incidence on basis of hospital records; Arch. Indust. Hyg. Occup. Med., 1950, 2, 505–516. 7. Hayer, R.B.; Lilienfeld, A.M.; Snell, L.M.; Mortality in chromium chemical production workers. A prospective study; Int. J. Epidemiol., 1979, 8, 365–374. 8. Bidstrup, P.L.; Carcinoma of the lung in chromate workers; Br. J. Ind. Med., 1951, 8, 302–305. 9. Bidstrup, P.L.; Case, R.A.M.; Carcinoma in the lung of workmen in the bichromateproducing industry in Great Britain; Br. J. Ind. Med., 1956, 13, 260–264. 10. Royle, H.; Toxicity of chromic acid in the chrome plating industry (1); Environ. Res., 1975, 10, 39–53. 11. Royle, H.; Toxicity of chromic acid in the chrome plating industry (2); Environ. Res., 1975, 10, 141–163. 12. Frentzel-Beyme, R.; Lung cancer mortality of workers employed in chromate pigment factories. A multicentric European epidemiological study; Cancer Res. Clin. Oncol., 1983, 105, 183–188. 13. Franchini, L.; Magnani, F.; Mutti, A.; Mortality experiences among chrome plating workers; Scand. J. Environ. Health, 1983, 9, 247–252. 14. Satoh, K.; Fukuda, Y.; Torii, K.; Katsuno, N.; Epidemiological study of workers engaged in the manufacture of chromium compounds; J. Occ. Med., 1981, 23, 835–838. 15. IARC monograph on the evaluation of carcinogenic risk to humans, chromium, nickel and welding; Lyon, France, 1990. 16. US Environmental Protection Agency; EPA Locating and Estimating Air Emissions from Sources of Chromium, Office of Air Quality Planning and Standards, Research Triangle Park, NC, 1984. 17. US Environmental Protection Agency; EPA Health Assessment Document for Chromium, Office of Air Quality Planning and Standards, Research Triangle Park, NC, 1984. 18. US Environmental Protection Agency; EPA Health Effects Assessment for Hexavalent Chromium, Office of Emergency and Remedial Response, Cincinnati, OH, 1984. 19. US Environmental Protection Agency; National Priorities List Basic Site Query Results, www.oaspub.epa.gov/oerrpage/basicqry (accessed 9 Nov 2007).

484


20. Occupational Safety and Health Administration; Corporate-Wide Settlement Agreements, www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=CWSA&p_id=883 (accessed 9 Nov 2007). 21. Barceloux, D.G.; Chromium; Clin. Tox., 1999, 37, 173–194. 22. Salnikow, K.; Zhitkovich, A.; Costa, M.; Analysis of the binding sites of chromium to DNA and protein in vitro and in intact cells; Carcinogenesis, 1992 13, 2341–2346. 23. Arslan, P.; Beltrame, M.; Tomasi, A.; Intracellular chromium reduction; Biochim. Biophys. Acta., 1987, 931, 10–15. 24. Kortenkamp, A.; Beyersmann, D.; O’Brien, P.; Uptake of chromium (III) complexes by erythrocytes; Toxicol. Environ. Chem., 1987, 14, 23–32. 25. Connett, P.H.; Wetterhahn, K.E.; Metabolism of the carcinogen chromate by cellular constituents; Struct. Bond., 1983, 54, 93–124. 26. Quievryn, G.; Peterson, E.; Messer, J.; Zhitkovich, A.; Genotoxicity and mutagenicity of chromium(VI)/ascorbate-generated DNA adducts in human and bacterial cells; Biochem., 2003, 42, 1062–1070. 27. Standeven,A.M.;Wetterhahn, K.E.;Ascorbate is the principle reductant of chromium(VI) in rat liver and kidney ultrafiltrates; Carcinogenesis, 1991, 12, 1733–1737. 28. Stearns, D.M.; Wetterhahn, K.E.; Reaction of chromium(VI) with ascorbate produces chromium(V), chromium(IV), and carbon-based radicals; Chem. Res. Toxicol., 1994, 7, 219–230. 29. Goodgame, D.M.L.; Joy, A.M.; EPR study of the Cr(V) and radical species produced in the reduction of Cr(VI) by ascorbate; Inorg. Chim Acta, 1987, 135, 115–118. 30. Stearns, D.M.; Kennedy, L.J.; Courtney, K.D.; Giangrande, P.H.; Phieffer, L.S.; Wetterhahn, K.E.; Reduction of chromium(VI) by ascorbate leads to chromium-DNA binding and DNA strand breaks in vitro; Biochem., 1995, 34, 910–919. 31. Da Cruz Fresco, P.; Shacker, F.; Kortenkamp, A.; The reductive conversion of chromium(VI) by ascorbate gives rise to apurinic/apyrimidinic sites in isolated DNA; Chem. Res. Toxicol., 1995, 8, 884–890. 32. Lay, P.A.; Levina, A.; Activation of molecular oxygen during the reactions of chromium(VI/V/IV) with biological reductants: Implications for chromium-induced genotoxicities; J. Am. Chem. Soc., 1998, 120, 6704–6714. 33. Martin, B.D.; Schoenhard, J.A.; Hwang, J.-M.; Sugden, K.D.; Ascorbate is a pro-oxidant in chromium-treated human lung cells; Mutat. Res., 2006, 610, 74–84. 34. Connett, P.H.; Wetterhahn, K.E.; In vitro reaction of the carcinogen chromate with cellular thiols and carboxylic acids; J. Am. Chem. Soc., 1985, 107, 4282–4288. 35. Bose, R.N.; Moghaddas, S.; Gelerinter, E.; Long-lived chromium(IV) and chromium(V) metabolites in the chromium(VI)-glutathione reaction: NMR, ESR, HPLC, and kinetic characterization; Inorg. Chem., 1992, 31, 987–1994. 36. Borges, K.M.; Boswell, J.S.; Liebross, R.H.; Wetterhahn, K.E.; Activation of chromium(VI) by thiols results in chromium(V) formation, chromium binding to DNA and altered DNA conformation; Carcinogenesis, 1991, 12, 551–561. 37. Kortenkamp, A.; Casadevall, M.; Faux, S.P.; Jenner, A.; Shayer, R.O.J.; Woodbridge, N.; O’Brien, P.; A role for molecular oxygen in the formation of DNA damage during the reduction of the carcinogen chromium(VI) by glutathione; Arch. Biochem. Biophys., 1996, 329, 199–207. 38. Sevilla, M.D.; Becker, D.; Swarts, S.; Herrington, J.; Sulfinyl radical formation from the reaction of cysteine and glutathione thiyl radicals with molecular oxygen; Biochem. Biophys. Res. Commun., 1987, 144, 1037–1042. 39. Tomaszewski, K.E.; Agarwal, D.K.; Melnick, R.L.; In vitro steady-state levels of hydrogen peroxide after exposure of male F344 rats and female B6C3F1 mice to hepatic peroxisome proliferators; Carcinogenesis, 1986, 7, 1871–1876. 40. Oshino, N.; Chance, B.; Sies, H.; Bucher, T.; The role of H2O2 generation in perfused rat liver and the reaction of catalase compound I and hydrogen donors; Arch. Biochem. Biophys., 1973, 154, 117–131.

References

485

41. Chance, B.; Sies, H.; Boveris, A.; Hydroperoxide metabolism in mammalian organs; Physiol Rev., 1979, 69, 527–605. 42. Shi, X.; Dalal, N.S.; The role of superoxide radical in chromium(VI) generated hydroxyl radical: The Cr(VI) Haber-Weiss cycle; Arch. Biochem. Biophys., 1992, 292, 323–327. 43. Zhang, L.; Lay, P.A.; EPR spectroscopic studies on the formation of chromium(V) peroxo complexes in the reaction of chromium(VI) with hydrogen peroxide; Inorg. Chem., 1998, 37, 1729–1733. 44. Shi, X.; Dalal, N.S.; Generation of hydroxyl radical by chromate in biologically relevant systems: Role of Cr(V) complexes versus tetraperoxochromate(V); Environ. Health Perspect., 1994, 102(3), 231–236. 45. O’Brien, T.J.; Ceryak, S.; Patierno, S.R.; Complexities of chromium carcinogenesis: role of cellular response, repair and recovery mechanisms; Mutat. Res., 2003, 533, 3–36. 46. Chen, J.; Thilly, W.G.; Mutational spectrum of chromium(VI) in human cells; Mutat. Res., 1994, 323, 21–27. 47. Sen, P.; Costa, M.; Incidence and localization of sister chromatid exchanges induced by nickel and chromium compounds; Carcinogenesis, 1986, 7, 1527–1533. 48. Sarto, F.; Cominato, V.; Bianchi, V.; Levis, A.G.; Increased incidence of chromosomal aberrations and sister chromatid exchanges in workers exposed to chromic acid (CrO3) in electroplating factories; Carcinogenesis, 1982, 3, 1011–1016. 49. Elias, Z.; Schneider, O.; Aubry, F.; Daniere, M.C.; Poirot, O.; Sister chromatid exchanges in Chinese hamster V79 cells treated with the trivalent chromium compounds chromic chloride and chromic oxide; Carcinogenesis, 1983, 4, 605–611. 50. US Environmental Protection Agency; Consumer Factsheet on: Chromium. 21 Feb. 2006. EPA Ground Water and Drinking Water, www.epa.gov/safewater/contaminants/dw_contamfs /chromium.html (accessed 9 Nov 2007). 51. Liebross, R.H.; Wetterhahn, K.E.; In vivo formation of chromium(V) in chick embryo red blood cells; Chem. Res. Toxicol., 1990, 3, 401–403. 52. Rossi, S.C.; Gorman, N.; Wetterhahn, K.E.; Mitochondrial reduction of the carcinogen chromate: formation of chromium(V); Chem. Res. Toxicol., 1988, 1, 101–107. 53. Tsou, T.C.; Yang, J.L.; Formation of reactive oxygen species and DNA strand breakage during interaction of chromium(III) and hydrogen peroxide in vitro: evidence for a chromium(III)-mediated Fenton-like reaction; Chem. Biol. Interact., 1996, 102, 133–153. 54. Aiyar, J.; Berkovits, H.J.; Floyd, R.A.; Wetterhahn, K.E.; Reaction of chromium(VI) with hydrogen peroxide in the presence of glutathione: reactive intermediates and resulting DNA damage; Chem. Res. Toxicol., 1990, 3, 595–603. 55. Aiyar, J.; Berkovits, H.J.; Floyd, R.A.; Wetterhahn, K.E.; Reaction of chromium(VI) with glutathione or with hydrogen peroxide: identification of reactive intermediates and their role in chromium(VI)-induced DNA damage; Environ. Health Perspect., 1991, 92, 53–62. 56. Mueller, S.; Riedel, H.D.; Stremmel, W.; Determination of catalase activity at physiological hydrogen peroxide concentrations; Anal. Biochem., 1997, 245, 55–60. 57. Bose, R.N.; Fonkeng, B.S.; Moghaddas, S.; Stroup, D.; Mechanisms of DNA damage by chromium(V)-DNA carcinogens; Nucl. Acids Res., 1998, 26, 1588–1596. 58. Steenken, S.; Jovanovic, S.; How easily oxidizable is DNA? One-electron reduction potentials of adenosine and guanosine radicals in aqueous solution; J. Am. Chem. Soc., 1997, 119, 617–618. 59. Yanagawa, H.; Ogawa, Y.; Ueno, M.; Redox ribonucleotides. Isolation and characterization of 5-hydroxyuridine, 8-hydroxyguanosine, and 8-hydroxyadenosine from Torula yeast RNA; J. Biol. Chem., 1992, 267, 13320–13326. 60. Saito, I.; Takayama, M.; Sugiyama, H.; Nakatani, K.; Photoinduced DNA cleavage via electron transfer: demonstration that guanine residues located 5’ to guanine are the most electron-donating sites; J. Am. Chem. Soc., 1995, 117, 6406–6407. 61. Shibutani, S.; Takeshita, A.; Grollman, A.P.; Insertion of specific bases during DNA synthesis past the oxidation-damaged base 8-oxodG; Nature, 1991, 349, 431–434.

486


62. Sugden, K.D.; Wetterhahn, K.E.; Reaction of chromium(V) with the EPR spin traps 5,5-dimethylpyrroline N-oxide and phenyl-N-tertbutylnitrone resulting in direct oxidation; Inorg. Chem., 1996, 35, 651–657. 63. Groves, J.T.; Kruper, W.J.; Preparation and characterization of an oxoporphinatochromium(V) complex; J. Am. Chem. Soc., 1979, 101, 7613–7615. 64. Neyhart, G.A.; Grover, N.; Smith, S.S.; Kalsbeck, W.A.; Fairley, T.A.; Cory, M.; Thorp, H.H.; Binding and kinetic studies of oxidation of DNA by oxoruthenium(IV); J. Am. Chem. Soc., 1993, 115, 4423–4428. 65. Luo, W.; Muller, J.G.; Rachlin, E.M.; Burrows, C.J.; Characterization of spiroiminodihydantoin as a product of one electron oxidation of 8-oxo-7,8 dihydroguanosine; Org. Lett., 2000, 2, 613–616. 66. Hailer, M.K.; Slade, P.G.; Martin, B.D.; Sugden, K.D.; Nei deficient Escherichia coli are sensitive to chromate and accumulate the oxidized guanine lesion spiroiminodihydantoin; Chem. Res. Toxicol., 2005, 18, 1378–1383. 67. Steenken, S.; Jovanovic, S.V.; Bietti, M.; Bernhard, K.; The trap depth (in DNA) of 8oxo-7,8-dihydro-2’deoxyguanosine as derived from electron-transfer equilibria in aqueous solution; J. Am. Chem. Soc., 2000, 122, 2373–2374. 68. Delaney, S.; Barton, J.K.; Long range DNA charge transport; J. Org. Chem., 2003, 68, 6475–6483. 69. Slade, P.G.; Priestley, N.P.; Sugden, K.D.; Spiroiminodihydantoin as an oxo-atom transfer product of 8-oxo-2’-deoxyguanosine oxidation by chromium(V); Org. Lett., 2007, 9, 4411–4414. 70. Duarte, V.; Muller, J.G.; Burrows, C.J.; Insertion of dGMP and dAMP during in vitro DNA synthesis opposite an oxidized form of 7,8-dihydro-8-oxoguanine; Nucleic Acids Res., 1999, 27, 496–502. 71. Henderson, P.T.; Delaney, J.C.; Muller, J.G.; Neeley, W.L.; Tannenbaum, S.R.; Burrows, C.J.; Essigmann, J.M.; The hydantoin lesions formed from oxidation of 7,8-dihydro-8oxoguanine are potent sources of replication errors in vivo; Biochem., 2003, 42, 9257–9262. 72. Kornyushyna, O.; Berges, A.M.; Muller, J.G.; Burrows, C.J.; In vitro nucleotide misinsertion opposite the oxidized guanosine lesions spiroiminodihydantoin and guanidinohydantoin and DNA synthesis past the lesions using Escherichia coli DNA polymerase I (Klenow fragment); Biochem., 2002, 41, 15304–15314. 73. Sugden, K.D.; Campo, C.K.; Martin, B.D.; Direct oxidation of guanine and 7,8-dihydro8-oxoguanine in DNA by a high valent chromium complex: A possible mechanism for chromate genotoxicity; Chem. Res. Toxicol., 2001, 14, 1315–1322. 74. Sugden, K.D.; Martin, B.D.; Guanine and 7,8-dihydro-8-oxo-guanine-specific oxidation in DNA by chromium(V); Environ. Health Perspect., 2002, 110(5), 725–728. 75. Aburantani, H.; Hippou, Y.; Ishida, T.; Takashima, R.; Matsuba, C.; Kodama, M.T.; Yasui, A.; Yamamoto, K.; Asano, M.; Fukasawa, K.; Yoshinari, T.; Inoue, H.; Ohtsuka, E.; Nishimura, S.; Cloning and characterization of mammalian 8-hydroxyguanine-specific DNA glycosylase/apurinic, apyrimidinic lyase, a functional mutM homologue; Cancer Res., 1997, 57, 2151–2156. 76. Minowa, O.; Arai, T.; Hirano, M.; Monden, Y.; Nakai, S.; Fukuda, M.; Itoh, M.; Takano, H.; Hippou, Y.; Abutantani, H.; Masumura, K.; Nohmi, T.; Nishimura, S.; Noda, T.; Mmg/ Ogg1 gene inactivation results in accumulation of 8-hydroxyguanine in mice; Proc. Natl. Acad. Sci. USA, 2000, 97, 4156–4161. 77. Jiang, D.; Hatahet, Z.; Melamede, R.J.; Kow, Y.W.; Wallace, S.S.; Characterization of Escherichia coli endonuclease VIII; J. Biol. Chem., 1997, 272, 32230–32239. 78. Hazra, T.K.; Muller, J.G.; Manuel, R.C.; Burrows, C.J.; Lloyd, S.R.; Mitra, S.; Repair of hydantoins, one electron oxidation product of 8-oxo-guanine, by DNA glycosylases of Escherichia coli; Nucleic Acids Res., 2001, 29, 967–1974. 79. Takao, M.; Kanno, S.; Kobayashi, K.; Zhang, Q.; Yonei, S.; van der Horst, G.; Yasui, A.; A back-up glycosylase in Nth1 knock-out mice is a functional Nei (endonuclease VIII) homologue; J. Biol. Chem., 2002, 277, 42205–42213.

References

487

80. Hazra, T.K.; Kow, Y.W.; Hatahet, Z.; Imhoff, B.; Boldogh, I.; Mokkapati, S.K.; Mitra, S.; Izumi, T.; Identification and characterization of a novel human DNA glycosylase for repair of cytosine derived lesions; J. Biol. Chem., 2002, 277, 30417–30420. 81. Dou, H.; Mitra, S.; Hazra, T.K.; Repair of oxidized bases in DNA bubble structures by human DNA glycosylases NEIL1 and NEIL2; J. Biol. Chem., 2003, 278, 49679–49684. 82. Hailer, M.K; Slade, P.G.; Martin, B.D.; Rosenquist, T.A.; Sugden, K.D.; Recognition of the oxidized lesions spiroiminodihydantoin and guanidinohydantoin in DNA by the mammalian base excision repair glycosylases NEIL1 and NEIL2; DNA Repair, 2005, 4, 41–50. 83. Bruner, S.D.; Norman, D.P.; Verdine, G.L.; Structural basis for recognition and repair of the endogenous mutagen 8-oxoguanine in DNA; Nature, 2000, 403, 859–866. 84. Mokkapati, S.K.; Wiederhold, L.; Hazra, T.K.; Mitra, S.; Stimulation of DNA glycosylase activity of OGG1 by NEIL1: functional collaboration between two human DNA glycosylases; Biochem., 2004, 43, 11596–11604. 85. Hazra, T.K.; Izumi, T.; Boldogh, I.; Imhoff, B.; Kow, Y.W.; Jaruga, P.; Dizdaroglu, M.; Mitra, S.; Identification and characterization of a human DNA glycosylase for repair of modified bases in oxidatively damaged DNA; Proc. Natl. Acad. Sci. USA, 2002, 99, 3523–3528. 86. Dou, H.; Theriot, C.A.; Das, A.; Hedge, M.L.; Matsumoto, Y.; Boldogh, I.; Hazra, T.K.; Bhakat, K.K.; Mitra, S.; Interaction of the human DNA glycosylase NEIL1 with proliferating cell nuclear antigen: The potential for replication-associated-repair of oxidized bases in mammalian genomes; J. Biol. Chem., 2008, 283, 3130–3140. 87. Jia, L.; Shafirovich, V.; Geacintov, N.E.; Broyde, S.; Lesion specificity in the base excision repair enzyme hNeil1: Modeling and dynamoics studies; Biochem., 2007, 46, 5305–5314. 88. Sugden, K.D.; Wetterhahn, K.E.; Direct and hydrogen peroxide-induced chromium(V) oxidation of deoxyribose in single-stranded and double-stranded calf thymus DNA; Chem. Res. Toxicol., 1997, 10, 1397–1406. 89. Pogoselski, W.K.; Tullius, T.D.; Oxidative strand scission of nucleic acids: routes initiated by hydrogen abstraction from the sugar moiety; Chem. Rev., 1998, 98, 1089–1107. 90. Breen, P.A.; Murphy, J.A.; Reactions of oxyl radicals with DNA; Free Rad. Biol. Med., 1995, 18, 1033–1077. 91. Sugden, K.D.; Formation of modified cleavage termini from the reaction of chromium(V) with DNA; J. Inorg. Biochem., 1999, 77, 177–183. 92. Sugden, K.D.; Wetterhahn, K.E.; EPR evidence for chromium(V) binding to phosphate and pyrophosphate: implications for chromium(V)-DNA interactions; Inorg. Chem., 1996, 35, 3727–3728. 93. Caldecott, K.W.; XRCC1 and DNA strand break repair; DNA Repair, 2003, 2, 955–969. 94. Wise, Sr. J.P.; Wise, S.S.; Little, J.E.; The cytotoxicity and genotoxicity of particulate and soluble hexavalent chromium in human lung cells; Mutat. Res., 2002, 517, 221–229. 95. Damia, G., D’Incalci, M.; Targeting DNA repair as a promising approach in cancer therapy; Eur. J. Cancer, 2007, 43, 1791–1801. 96. Caldecott, K.W.; DNA single-strand breaks and neurodegeneration; DNA Repair, 2004, 3, 875–882. 97. Blankert, S.A.; Coryell, V.H.; Picard, B.T.; Wolf, K.K.; Lomas, R.E.; Stearns, D.M.; Characterization of nonmutagenic Cr(III)-DNA Interactions; Chem. Res. Toxicol., 2003, 16, 847–854. 98. Gelasco, A.; Lippard, S.J.; NMR solution structure of a DNA dodecamer duplex containing a cis-diammineplatinum(II) d(GpG) intrastrand crosslink, the major adduct of the anticancer drug Cisplatin; Biochem., 1998, 37, 9230–9239. 99. Lippard, S.J.; New chemistry of an old molecule: cis-[Pt(NH3)2Cl2]; Science, 1982, 218, 1075–1082. 100. Zhitkovich, A.; Song, Y.; Quievryn, G.; Voitkun, V.; Non-oxidative mechanisms are responsible for the induction of mutagenesis by reduction of Cr(VI) with cysteine: role

488

101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122.

Cr-Induced DNA Damage and Repair of ternary DNA adducts in Cr(III)-dependent mutagenesis; Biochem., 2001, 40, 549–560. Snow, E.T.; Xu, L.S.; Chromium(III) bound to DNA templates promotes increased polymerase processivity and decreased fidelity during replication in vitro; Biochem., 1991, 30, 11238–11245. Zhitkovich, A.; Importance of chromium-DNA adducts in mutagenicity and toxicity of chromium(VI); Chem. Res. Toxicol., 2005, 18, 3–11. Peterson-Roth, E.; Reynolds, M.; Quievryn, G.; Zhitkovich, A.; Mismatch repair proteins are activators of toxic responses to chromium-DNA damage; Mol. Cell. Biol., 2005, 25, 3596–3607. Reynolds, M.; Zhitkovich, A.; Cellular vitamin C increases chromate toxicity via a death program requiring mismatch repair but not p53; Carcinogenesis, 2007, 28, 1613–1620. Strauss-Soukup, J.K.; Vaghefi, M.M.; Hogrefe, R.I.; Maher III, L.J.; Effects of neutralization pattern and stereochemistry on DNA bending by methylphosphonate substitutions; Biochem., 1997, 36, 8692–8698. Jiricny, J.; Nyström-Lahti, M.; Mismatch repair defects in cancer; Curr. Opin. Genet. Dev., 2000, 10, 157–161. Modrich, P.; Lahue, R.; Mismatch repair in replication fidelity, genetic recombination, and cancer biology; Annu. Rev. Biochem., 1996, 65, 101–133 Kolodner, R.D.; Marsischky, G.T.; Eukaryotic DNA mismatch repair; Curr. Opin. Genet. Dev., 1999, 9, 89–96. Reynolds, M.; Stoddard, L.; Bespalov, I.; Zhitkovich, A.; Ascorbate acts as a highly potent inducer of chromate mutagenesis and clastogenesis: linkage to DNA breaks in G2 phase by mismatch repair; Nucleic Acids Res., 2007, 35, 465–476. Hirose, T.; Kondo, K.; Takahashi, Y.; Ishikura, H.; Fujino, H.; Tsuyuguchi, M.; Hashimoto, M.; Yokose, T.; Mukai, K.; Kodama, T.; Moden, Y.; Frequent microsatellite instability in lung cancer from chromate-exposed workers; Mol. Carcinog., 2002, 33, 172–180. Kunkel, T.; Erie, D.A.; DNA mismatch repair; Annu. Rev. Biochem., 2005, 74, 681–710. Modrich, P.; Mechanisms in eukaryotic mismatch repair; J. Biol. Chem., 2006, 281, 30305–30309. Salnikow, K.; Zhitkovich, A.; Genetic and epigenetic mechanisms in metal carcinogenesis and cocarcinogenesis: nickel, arsenic, and chromium; Chem. Res. Toxicol., 2008, 21, 28–44. Strauss, J.K.; Maher III, J.L.; DNA bending in asymmetric phosphate neutralization; Science, 1994, 266, 1829–1834. Crothers, D.M.; Upsetting the forces of balance in DNA; Science, 1994, 266, 1819–1820. Voitkun, V.; Zhitkovich, A.; Costa, M.; Complexing of amino acids to DNA by chromate in intact cells; Environ. Health Perspect., 1994, 102(3), 251–255. Zhitkovich, A.; Voitkun, V.; Costa, M.; Formation of amino acid-DNA complexes by hexavalent and trivalent chromium in vitro: importance of trivalent chromium and the phosphate group; Biochem., 1996, 35, 7575–7582. Nickoloff, J.A.; Heokstra, M.F.; DNA Damage and Repair, Vol. II: DNA Repair in Higher Eukaryotes, Humana Press Inc., Totowa, NJ, 1998, 350–355. Andrews, A.D.; Barrett, S.F.; Yoder, F.W.; Robbins, J.H.; Xeroderma pigmetosum neurological abnormalities correlate with colony-forming ability after ultraviolet radiation; Proc. Natl. Acad. Sci. USA, 1978, 75, 1984–1988. Burns, J.L.; Guzder, S.N.; Sung, P.; Prakash, S.; Prakash, L.; An affinity of human replication protein A for ultraviolet-damaged DNA; J. Biol. Chem., 1996, 271, 11607–11610. Clugston, C.K.; McLaughlin, K.; Kenny, M.K.; Brown, R.; Binding of human singlestranded DNA binding protein to DNA damaged by the anticancer drug cisdiamminedichloroplatinum (II); Cancer Res., 1992, 52, 6375–6379. Matsuda, T.; Saijo, M.; Kuraoka, I.; Kobayashi, T.; Nakatsu, Y.; Nagai, A.; Enjoji, T.; Masutani, C.; Sugasawa, K.; Hanaoka, F.; Yasui, A.; Tanaka, K.; DNA repair protein XPA binds replication protein A (RPA); J. Biol. Chem., 1992, 270, 4152–4157.

References

489

123. Schaeffer, L.; Roy, R.; Humbert, S.; Moncollin, V.; Vermeulen, W.; Hoeijmakers, J.H.J.; Chambon, P.; Egly, J.M.; DNA repair helicase: a component of BTF2 (TFIIH) basic transcription factor; Science, 1993, 260, 58–63. 124. Sung, P.; Bailly, V.; Weber, C.; Thompson, L.H.; Prakash, L.; Prakash, S.; Human Xeroderma pigmentosum group D gene encodes a DNA helicase; Nature, 1993, 365, 852–855. 125. O’Donovan, A.; Scherly, D.; Clarkson, S.G.; Wood, R.G.; Isolation of active recombinant XPG protein, a human DNA repair endonuclease; J. Biol. Chem., 1994, 269, 15965–15968. 126. Matsunaga, T.; Mu, D.; Park, C.H.; Reardon, J.T.; Sancar, A.; Human DNA repair excision nuclease; J. Biol. Chem., 1995, 270, 20862–20869. 127. Huang, J.C.; Svoboda, D.L.; Reardon, J.T.; Sancar, A.; Human nucleotide excision nuclease removes thymine dimmers from DNA by incising the 22nd phosphodiester bond 5’ and the 6th phosphodiester bond 3’ to the photodimer; Proc. Natl. Acad. Sci. USA, 1992, 89, 3664–3668. 128. Mu, D.; Hsu, D.S.; Sancar, A.; Reaction mechanism of human DNA repair excision nuclease; J. Biol. Chem., 1996, 271, 8285–8294. 129. Park, C.H.; Bessho, T.; Matsunaga, T.; Sancar, A.; Purification and characterization of the XPF-ERCC1 complex of human DNA repair excision nuclease; J. Biol. Chem., 1995, 270, 22657–22660. 130. Nichols, A.F.; Sancar, A.; Purification of PCNA as a nucleotide excision repair protein; Nucleic Acids Res., 1992, 20, 2441–2446. 131. Hubscher, U.; Spadari, S.; DNA replication and chemotherapy; Physiol. Rev., 1994, 74, 259–304. 132. O’Brien, T.; Mandel, H.G.; Pritchard, D.E.; Patierno, S.R.; Critical role of chromium (Cr)-DNA interactions in the formation of Cr-induced polymerase arresting lesions; Biochem., 2002, 41, 12529–12537. 133. Bridgewater, L.C.; Manning, F.C.; Patierno, S.R.; Base-specific arrest of in vitro DNA replication by carcinogenic chromium: relationship to DNA interstrand crosslinking; Carcinogenesis, 1994, 15, 2421–2427. 134. Bridgewater, L.C.; Manning, F.C.; Patierno, S.R.; Arrest of replication by mammalian DNA polymerases a and b caused by chromium-DNA lesions; Mol. Carcinog., 1998, 23, 201–206. 135. Helleday, T.; Nilsson, R.; Jenssen, D.; Arsenic(III) and heavy metal ions induce intrachromosomal homologous recombination in the hprt gene of V79 Chinese hamster cells; Environ. Mol. Mutagen., 2000, 35, 114–122. 136. O’Brien, T.J.; Fornsaglio, J.L.; Ceryak, S.R.; Patierno, S.R.; Effects of hexavalent chromium on the survival and cell cycle distribution of repair-deficient S. cerevisiae; DNA Repair, 2002, 1, 617–627. 137. Helleday, T.; Lo, J.; van Gent, D.C.; Engelward, B.P.; DNA double-strand break repair: from mechanistic understanding to cancer treatment; DNA Repair, 2007, 6, 923–935. 138. Strumberg, D.; Pilon, A.A.; Smith, M.; Hickey, R.; Malkas, L.; Pommier, Y.; Conversion of topoisomerase I cleavage complexes on the leading strand of ribosomal DNA into 5’-phosphorylated DNA double-strand breaks by replication runoff; Mol. Cell. Biol., 2000, 20, 3977–3987. 139. Richardson, C.; Moynahan, M.E.; Jasin, M.; Double-strand break repair by interchromosomal recombination: suppression of chromosomal translocations; Genes Dev., 2000, 12, 3831–3842. 140. Richardson, C.; Jasin, M.; Frequent chromosomal translocations induced by DNA double-strand breaks; Nature, 2000, 405, 676–700.

18 Arsenic-Induced Carcinogenicity: New Insights in Molecular Mechanism Andrea Hartwig and Tanja Schwerdtle

18.1 Introduction Arsenic is a well-documented human carcinogen after both oral exposure and inhalation. While environmentally relevant arsenic species include inorganic as well as organic arsenicals, elevated cancer incidences have been attributed to inorganic arsenic, and research has focused on identifying the underlying mechanisms. A variety of inorganic arsenates or arsenites occur in water, soil and food, while the most common inorganic arsenical in the air is arsenic trioxide (As2O3). In general, trivalent arsenites tend to be more toxic than pentavalent arsenates; however, in many cases mixed exposure towards trivalent and pentavalent inorganic arsenic occurs and the precise chemical speciation is usually not known.1,2 Regarding worldwide public health, the most important medium for inorganic arsenic exposure is drinking water, causing many types of cancer and a wide range of further adverse health effects, including neurotoxicity, liver injury and the endemic ‘blackfoot disease’.3–6 Furthermore, occupational exposure is still relevant, even though the use of arsenicals has been restricted in industrialized countries considerably during recent years.


492

Arsenic-Induced Carcinogenicity

18.2 Carcinogenicity and Cocarcinogenicity Predominant tumours in humans exposed chronically via inhalation or the gastrointestinal tract are lung and skin cancers, respectively. Additionally, case reports and epidemiological studies indicate increased incidences of tumours in the skin, the liver and the digestive tract after chronic arsenic inhalation, as well as increased risks of internal tumours (mainly bladder, lung, and to a lesser extent liver, kidney and prostate) after chronic arsenic ingestion (reviewed in references 3–6). Numerous scientific commissions and regulatory agencies have classified arsenic as a human carcinogen; among these are the International Agency for Research on Cancer (IARC), the German Commission for the Investigation of Health Hazards of Chemical Compounds in the Work Area (MAK commission) and the United States Environmental Protection Agency (USEPA).2,7–11 In contrast to humans, where the carcinogenic potential is clearly evident, experimental animals usually fail to demonstrate increased tumour incidences, with the exception of studies in mice demonstrating transplacental carcinogenesis.2,12–14 Nevertheless, cocarcinogenic effects have been observed in several experimental settings. Thus, early animal studies have shown that lung adenoma incidences were increased by benzo[a]pyrene and arsenic cotreatment, as compared to treatment with each chemical alone.15 More recently, arsenite has been demonstrated to enhance the carcinogenicity of UV radiation in mice, thus combined exposure to arsenite-enriched drinking water and solar UV radiation strongly increased skin tumour incidences, as compared to mice given UV alone.16,17 Similar cocarcinogenic effects were also seen in humans. The results from several epidemiological studies suggest more than additive effects on respiratory tract cancer incidences in smokers that were exposed to arsenic occupationally or via drinking water.18,19 Additionally, epidemiological studies provide some evidence that arsenite acts as a cocarcinogen in skin cancer (reviewed in reference 20). With respect to organic methylated arsenic metabolites, oral treatment of animals with dimethylarsinic acid (DMAV) has been shown to act as a tumour promoter in several organs and as a complete carcinogen in the rat bladder; however it is questionable whether the applied DMAV concentrations could be attained in vivo after exposure to nonacutely toxic inorganic arsenic doses (summarized in references 21 and 22).

18.3 Metabolism of Inorganic Arsenic After inhalation or ingestion of inorganic arsenic, humans and many other mammals have the ability to metabolize inorganic arsenic into organic arsenic forms that are more readily excreted via urine. After reduction of arsenate, arsenite is converted to its trivalent and pentavalent methylated metabolites, monomethylarsonous (MMAIII) and dimethylarsinous (DMAIII) acid, monomethylarsonic (MMAV) and dimethylarsinic (DMAV) acid. To date, the individual steps of biomethylation, which

Metabolism of Inorganic Arsenic Arsenite

Arsenate

OH O

AsV

OH

O-

HO

2e-

ATG

SG

OH

PNP/ GSTw

AsIII

GSH

OH

AsIII

GS

AS3MT SAHC

SAHC

AsV

O OH

SG

AsIII

DMAIII

GSH

HO

MADG

OH

AsIII

CH3

O

O-

MMAIII

CH3

AsV

MMAV

SAM

AS3MT

CH3 A

SAHC AS3MT

SAHC

AsIII

CH3

2e

SAM

HO

AsIII

GS

OH

-

MMAIII

CH3

CH3

O- MMAV

GSTw

B

SAM AS3MT

OH

HO

SG

SAM

A

493

O

GSTw 2e-

AsV O-

CH3 DMAV

GS

AsIII

CH3

CH3

CH3

CH3

CH3

DMAG

GSH

HO

AsIII DMAIII

CH3

O

AsV O-

CH3 DMAV

CH3

Figure 18.1 Biotransformation of inorganic arsenic into methylated metabolites. Shown are schemes of the originally proposed, generally accepted, pathway with successive oxidative methylation and reduction steps (A) and the recently proposed alternative pathway, involving formation of arsenic-glutathion complexes (B). AS3MT, arsenic (+3 oxidation state) methyltransferase; ATG, arsenite triglutathione; DMAG, dimethylarsinic glutathione; GSTw, glutathione S-transferase w; MADG, monomethylarsonic diglutathione; PNP, purine nucleoside phosphorylase; SAHC, S-adenosylhomocysteine; SAM, S-adenosylmethionine (summarized from references 2, 23 and 28)

catalyse conversion of inorganic arsenic to the methylated metabolites enzymatically, are still not fully understood (reviewed in references 23–26). Two possible metabolic pathways have been postulated, yielding the same methylated metabolites. First a pathway has been proposed with successive oxidative methylation and reduction steps.27 Recently, alternative metabolization steps have been discussed, focusing on glutathione (GSH) conjugation and subsequent methylation28 (Figure 18.1). Furthermore, recent work suggested that cysteine residues in cellular proteins might serve an analogous function in complexing the arsenic-containing substrate like GSH.29 Whereas biomethylation has long been thought to be a detoxification process, nowadays it is reasonable to conclude that some adverse health effects seen in humans chronically exposed to inorganic arsenic are in fact caused by these metabolites. With respect to the trivalent metabolites MMAIII and DMAIII, numerous recent

494


studies have shown that they are as toxic, or even more toxic, compared to inorganic arsenic. Thus, with respect to molecular mechanisms, special emphasis is given to the biomethylated arsenicals, as they may at least contribute to inorganic arsenicinduced genotoxicity and presumably carcinogenicity (reviewed in references 30– 33). Whereas the ability of arsenic to induce carcinogenicity is beyond dispute, the mechanisms of how arsenic induces cancer in humans have not been clearly identified. Various modes of action have been proposed to be involved, which are discussed based on recent reviews22,26,30,32,34–37 and numerous original papers. Since evidence of inorganic arsenic-induced carcinogenicity in laboratory animals is mostly negative, modes of action studies are mostly conducted in cell culture systems.

18.4 Modes of Action 18.4.1 Induction of Genetic Damage Unlike many other carcinogens, arsenite neither directly interacts with DNA nor induces point mutations in bacterial or mammalian test systems (reviewed in references 4 and 10). However, arsenite has been shown to induce large deletion (multilocus) mutations in hamster human hybrid cells,38 micronuclei (MN) and chromosomal aberrations (CA), aneuploidy and sister-chromatid exchanges (SCE) in various mammalian cells (recently summarized in reference 2). Also in peripheral lymphocytes, oral mucosa or bladder urothelial cells of humans exposed towards elevated levels of arsenic in drinking water, numerous studies observed increases in chromosomal aberrations, micronuclei and sister chromatid exchanges (Table 18.1). In vitro studies in mammalian cells demonstrated the induction of DNA damage (strand breaks, oxidative base modifications, apurinic/apyrimidinic sites, DNA– protein crosslinks) and inhibition of DNA repair (see below). Thus, chromosomal alterations may be a secondary result of arsenite-induced DNA damage and DNA repair inhibition.32 Furthermore Li and Broome proposed a model, in which arsenite crosslinks tubulin and inhibits GTP binding, resulting in disturbed tubulin polymerization and mitosis,39 which may contribute to micronuclei formation. Additionally, arsenite and arsenate can cause gene amplification in mouse 3T6 cells.40 Since oncogenes are amplified in several animal and human tumours, the ability of arsenic to induce gene amplification may relate to its carcinogenic effects. Regarding the methylated species, MMAIII and DMAIII are generally genotoxic at lower concentrations than inorganic arsenic, while effects of the pentavalent metabolites MMAV and DMAV are either absent or restricted to much higher concentrations (e.g., references 41–45). In this context Kligermann and coworkers have recently summarized evidence that trivalent methylated arsenicals are potent clastogens and, due to spindle disruption, are aneugens as well.32 While arsenic is not a classical mutagen, it increases the genotoxicity, mutagenicity and clastogenicity of other DNA damaging agents, among others UV light,

Modes of Action

495

Table 18.1 Genotoxicity in humans after exposure towards elevated arsenic drinking water levels Individuals (exposed, controls) Country

Arsenic in drinking water

Effects in exposed individuals

Ref.

24 (11, 13) Mexico 437 (282, 155) Argentina 36 (18, 18) Nevada, USA 44 (22, 22) Argentina 69 (35, 34) Mexico 125 (70, 55) Chile 74 (42, 32) Finland 32 (19, 13) Inner Mongolia, China 66 (45, 21) West Bengal, India 95 (59, 36) West Bengal, India 317 (163, 154) West Bengal, India 320 (165, 155) West Bengal, India 217 (106, 111) North Chile 206 (105, 102) North Chile 306 (204, 102) West Bengal, India

controls: 26 µg/l exposed: 390 µg/l controls: 20 µg/l exposed: 130 µg/l controls: 16 µg/l exposed: 1313 µg/l controls: 8.4 µg/l exposed: 205 µg/l controls: 30 µg/l exposed: 408 µg/l controls: 15 µg/l exposed: 600 µg/l controls: 7 µg/l exposed: 410 µg/l controls: 4.5 µg/l exposed: 527.5 µg/l controls: 5.5 µg/l exposed: 368.1 µg/l controls: 6.4 µg/l exposed: 211.7 µg/l controls: 9.2 µg/l exposed: 214.7 µg/l controls: 9.1 µg/l exposed: 215 µg/l controls: 2.0 µg/l exposed: 750 µg/l controls: 2.0 µg/l exposed: 750 µg/l controls: 7.2 µg/l exposed: 220 µg/l

L: CA (↑), SCE (-), mutations in HPRT-locus ↑ L: SCE ↑

53 54

UC: MN (↑)

55

L: MN ↑, trisomy, SCE (-) L: CA (↑), OC,BC: MN (↑) UC: MN (↑)

56

58

L: CA ↑

59

OC,UC: MN ↑

60

L, OC, UC: MN ↑

61

L: SCE ↑, CA ↑

62

L,OC,UC: MN ↑

63

L: CA ↑, SCE ↑

64

L: MN ↑

65

OC: MN (↑)

66

L,OC,UC: MN ↑ L: CA ↑

67

57

CA, chromosomal aberrations; L, lymphocytes; MN, micronuclei; OC, oral mucosa cells; SCE, sister chromatid exchanges; UC, urothelial cells; ↑, significant increase; (↑), increase but not significant, (-), no effect

benzo[a]pyrene and alkylating agents,46–51 which may be explained by the interference with DNA repair processes.52 This is consistent with the comutagenic effect of arsenic, resulting in arsenic cocarcinogenesis, which has been shown in vivo (see below). 18.4.2 Involvement of Reactive Oxygen (ROS) and Nitrogen Species (RNS) in Arsenic Response Numerous studies provide strong evidence that oxidative stress mediated by increased levels of ROS and RNS is an important molecular mechanism contributing to arsenic-induced genotoxicity and carcinogenicity (reviewed in references

496


35, 68 and 69). Thus, in diverse cellular systems, arsenite increases the generation of superoxide anions ( iO−2 ) and hydrogen peroxide (H2O2), and modulates the level of nitric oxide (NO). Subsequently these reactive species can be converted to other, more damaging reactive species, such as the hydroxyl radical (·OH) and peroxynitrite (ONOO−). Several origins of elevated levels of reactive species have been suggested. They include interactions with the respiratory chain,70,71 their generation during metabolism of inorganic arsenic, such as the formation of intermediary dimethylarsine and radical arsenic species,72–79 the release of iron from ferritin by dimethylated arsenic species,80,81 and modulation of NO synthases.31,68,82 All of these mechanisms may lead to oxidative stress, resulting from an imbalance between free radical generation and cellular antioxidant defence systems.83 Moreover, arsenic has not only been shown to increase the generation of reactive species, but also to interact with cellular protection mechanisms. Thus arsenic is believed to change cellular redox homeostasis by decreasing cellular GSH. This might be due to the ability of trivalent arsenicals to complex with thiol groups, resulting in GSH binding and depletion, consumption of GSH during arsenic metabolism, as well as effects of trivalent methylated arsenicals on glutathione-related enzymes.84,85 With respect to genotoxicity, the application of radical scavengers revealed the involvement of arsenite-induced ROS and RNS in the induction of lipid peroxidation, protein oxidation, DNA damage (summarized in reference 35) and DNA repair inhibition.86 Furthermore chronic low-dose arsenic alters gene expression and protein levels that are associated with oxidative stress and inflammation (e.g., references 87–89), which may in part be due to oxidation of major transcriptional redox sensitive regulators (e.g., Nrf2, nuclear factor-erythroid 2-related factor 2) of altered genes. Thus the induced reactive species are known to activate signal cascades such as the mitogen-activated protein kinases (MAPKs) cascade and the transcription factors AP-1 (activator protein-1) and NFkB (nuclear factor-kB) (summarized in references 31 and 90). Finally, the fact that arsenite-induced oxidative stress is also evident in populations exposed to arsenic-contaminated drinking water underscores its role in arsenic-induced carcinogenicity (see, for example, references 91–97).

18.4.3 Epigenetic Effects In addition to the induction of genetic damage, altered DNA methylation may also contribute to carcinogenicity. Thus, in tumours global methylation is typically reduced, but some gene-specific promoter methylation is increased, thereby affecting the expression of protooncogenes and tumour suppressor genes, resulting in cell transformation and upregulated cell growth.98,99 Over the last 10 years accumulating evidence from cell culture studies, experimental animals and also from arsenicexposed humans suggests that epigenetic changes contribute to arsenic-induced carcinogenicity. Thus both hypo- and hypermethylation have been observed after arsenic exposure. For instance, increased cytosine methylation in the p53 promoter was detected in A549 human lung cells by arsenite and arsenate, but not by DMAv 100

Modes of Action

497

(further effects on p53 are discussed below). After exposure of A549 human lung and UOK human kidney cells to arsenite for several weeks, Zhong and Mass observed both hypo- and hypermethylation of different genes,101 whereas in a rat liver cell line102 and human HaCaT keratinocytes103 global DNA hypomethylation was found following chronic exposure to low levels of arsenite. In A/J mice, which were chronically orally exposed to arsenate and developed lung tumours, hypermethylation was observed with the consequence of diminished p16INK4a and RASSF1A expression occurring in these tumours.104 With respect to humans, dosedependent hypermethylation of the p53 and p16 genes was observed in blood samples of arsenic-exposed skin cancer patients in West Bengal.105 Additionally, in a population-based study of human bladder cancer, arsenic exposure measured as toenail arsenic was associated with promoter hypermethylation of the tumour suppressor proteins RASSF1A and PRSS3, but not p16INK4a.106 While hypomethylation may be due to inhibition of DNA methyltransferases or depletion of the methyl donor SAM,102,103 which is a common cofactor in DNA and arsenic methylation, hypermethylation is not easily understood and further studies are required to resolve this question.

18.4.4 Inhibition of DNA Repair To ensure faithful duplication and inheritance of genetic material, organisms have evolved to efficiently detect and remove DNA damage resulting from either endogenous sources (cellular metabolic processes) or exogenous sources (environmental factors). Dysregulation of DNA damage-response processes, including cell cycle arrest, DNA repair, senescence or apoptosis, contribute to genomic instability on the cellular level, predisposing the organism to cancer.107 Therefore, it is absolutely essential for cells to efficiently respond to DNA damage through coordinated and integrated cell cycle checkpoints and DNA repair pathways.108–110 Concerning DNA repair inhibition as one underlying mechanism of arsenic-induced carcinogenicity, several studies point to an interaction of arsenic with various DNA repair pathways, including nucleotide (NER) and base excision repair (BER) (reviewed in references 30 and 34). NER is capable of removing a wide variety of bulky, DNA-helix-distorting lesions, caused by UV radiation, chemotherapeutic drugs or environmental mutagens. During the last few years there has been accumulating evidence that low, noncytotoxic concentrations of arsenite inhibit nucleotide excision repair (NER),47,86,111–115 which is even more pronounced with the trivalent methylated metabolites MMAIII and DMAIII.42,116 Thus, arsenic has been shown to decrease removal of bulky DNA adducts induced by UV radiation47,86,111–113,115 or benzo[a]pyrene in cultured cells and laboratory animals.42,114,116,117 NER works through a ‘cut-andpatch’ mechanism; after DNA damage recognition, incisions at sites flanking the lesion occur and an oligonucleotide (24–32 nucleotides) containing the lesion is excised, followed by subsequent restoration of the original DNA sequence by polymerisation/ligation using the nondamaged strand as a template. In total, about

498


30 proteins are involved (recently summarized in references 108 and 118). In the first stage, arsenite seems to disturb the DNA damage recognition/incision step, but effects on DNA ligation are observed as well113 One possible mechanism of arsenic-induced toxicity may lie in its ability to react with thiols, for example in zinc-binding structures prevalent in many transcription factors, cell cycle control and DNA repair proteins (recently reviewed in reference 119). In case of the central NER DNA damage recognition protein XPA, trivalent arsenicals have been shown to release zinc from a 37 amino acid XPAzf peptide, representing the zinc-finger domain of the human XPA protein.42 Thus equimolar MMAIII released ZnII easily, forming mono- and diarsenical derivatives of XPAzf.120 In contrast, a 10-fold excess of arsenite was required to partially oxidize the zinc-finger structure of XPAzf. Thus, zinc-binding structures may be sensitive targets for arsenicals, even though the actual species involved in the specific interaction differ. With respect to XPA function, subcellular studies observed no or only a slight decrease on XPA binding to UVC-121 or MMC-damaged122 oligonucleotides by arsenite. To date the most sensitive target related to DNA repair is inhibition of poly(ADP-ribosyl)ation by trivalent arsenicals at nanomolar concentrations, with MMAIII and DMAIII showing the strongest effects.123,124 Poly(ADP-ribosyl)ation, which is mediated mainly by the poly(ADP-ribose) polymerase 1 (PARP-1), plays a major role in DNA damage signalling and the recruitment of DNA repair proteins to sites of damage. Although the role of PARP-1 is not fully understood as yet, there is strong evidence that PARP-1 contributes to base excision repair (BER),125 which among others is responsible for the repair of oxidative DNA damage, and perhaps also to NER. In this context, Poonepalli et al. reported that in mouse embryonic fibroblasts, a lack of PARP-1 gene product enhances cellular sensitivity to arsenite.126 With respect to BER, arsenite has also been shown to inhibit DNA ligase III activity in cells and nuclear extracts, most probably indirectly by altering cellular redox levels or by affecting signal transduction pathways related to ligase activity.127,128 As mentioned before, effects on the tumour suppressor protein p53 may also contribute to arsenic-induced carcinogenicity. Following activation by posttranslational modifications such as phosphorylation, acetylation and poly(ADP-ribosyl)ation, p53 plays a guarding role in maintaining genome integrity by activating the transcription of genes involved in cell cycle check points, cellular senescence, apoptosis and DNA repair (e.g., XPC, p48, hOGG1). At present effects of arsenic on p53 are not well understood, since responses are different depending on cell type and incubation times. Investigations in human fibroblasts and human lymphoblastoid cell lines have identified that arsenite could induce p53 accumulation through an ATMdependent pathway.129,130 On the other hand, arsenite and MMAIII have been shown to inhibit p53 phosphorylation and to reduce overall p53 level, as well as p53 DNAbinding activity.116,131 Furthermore a recent study has postulated increased formation of conformationally mutated p53 by arsenite.132 Due to the loss of its zinc-containing folded wild-type conformation, this conformationally mutated p53 is no longer capable of function as a transcription factor;133,134 many cancer-associated mutations cause loss of this conformation.135

Acknowledgements

499

Besides direct and indirect interaction with repair or repair-associated proteins, arsenic compounds may also diminish DNA repair by altering the expression of specific DNA repair genes. Thus Hamadeh et al.136 and Andrew et al.137 reported a decreased gene expression of various repair genes in arsenite-exposed human normal keratinocytes and cultured human bronchial epithelial cells, respectively. Our recent work demonstrates that arsenite strongly decreased expression and protein level of XPC, which is believed to be the principle initiator of global genome NER. This phenomenon was even more pronounced with MMAIII. This led to diminished association of XPC to sites of local UVC damage in human fibroblasts, resulting in decreased recruitment of further NER proteins and therefore disrupted DNA repair.138 In humans, arsenic exposure via drinking water was correlated in a dosedependent manner to decreased expression of NER and BER genes and diminished repair of lesions in lymphocytes.139,140 Additionally, a recent study showed that upon induction of DNA damage by gamma irradiation, the repair capacity in whole blood of arsenic-exposed individuals with premalignant hyperkeratosis was significantly less, compared to that of individuals without skin lesions.141 Table 18.2 summarizes studies providing evidence for inhibition of nucleotide and base excision repair by inorganic and methylated arsenicals.

18.5 Conclusions and Research Needs Taken together, the carcinogenicity of inorganic arsenic is well documented, but underlying mechanisms are not completely understood. Current evidence suggests that indirect mechanisms play a predominant role: rather than direct binding to DNA, interactions with proteins are of major importance, leading to oxidative stress and modulating the cellular response to DNA damage, which may in turn decrease genomic integrity (Figure 18.2). This would also explain the enhancement of benzo[a]pyrene and UV-induced carcinogenicity in animals and exposed humans, which may be due to the disturbance of DNA repair processes. However, in spite of many advances in arsenic-related toxicology, several aspects need to be addressed in future research. They include further insights into arsenic metabolism in humans and the relevance of metabolites in arsenic-induced genotoxicity and carcinogenicity. Furthermore, the impact of genetic polymorphisms, both in arsenic biomethylation and in DNA repair genes, needs to be clarified. Finally, the toxicological impact of organic arsenic compounds such as arsenosugars and arsenolipids warrants special attention.

Acknowledgements This work was supported by the DFG grant number Ha 2372/3–4 and SCHW 903/3–2.

Arsenite (100 µM) MMAIII (10 µM) DMAIII (10 µM) MMAV (5 mM) DMAV (no effect up to 10 mM)

Arsenite (5 µM) MMAIII (2.5 µM) DMAIII (2.5 µM) MMAV (250 µM) DMAV (250 µM) Arsenite (5 µM) MMAIII (1 µM) DMAIII (1 µM) Arsenite (500 µM)

Arsenite (10 mg/kg) Arsenite (1 µM)

Arsenite (2.5 µM) Arsenite (1 µM)

Arsenite (80 µM) Arsenite (100 µM)

As2O3 (1 mg/l) Arsenate (6 mg/l) As2O3 (0.5 mg/l) Arsenate (5 mg/l) Arsenite (20 µM)

Arsenic compound (lowest effective dose)

XPAfz (peptide representing the XPA zinc finger domain)

XPA

h. CRL2522 fibroblasts h. colon HCT116 cancer cells

h. A549 adenocarcinoma lung cell line

h. lymphocytes SV40-transformed fibroblasts (MRC5CV1) h. fibroblast CHO cells h. fibroblast Sprague-Dawley rats h. TK6 lymphoblastoid cells

CHO-K1 cells

h. fibroblast (SF24, XP6KO, XP2SA) V79 cells

Species (test system)

removal of BPDE-induced DNA adducts ↓ impairment of p53 induction in response to BPDE treatment binding of XPA to the MMC C-DNA interstrand crosslink ↓ release of zinc from XPAzf

NER NER

Removal of B[a]P-induced DNA adducts ↓ removal of UV-induced photolesions ↓ comutagenicity with UV removal of BPDE-induced DNA adducts ↓

NER

NER

NER

NER

NER NER

NER

NER

NER

Pathway

Removal of UV-induced photolesions ↓ Removal of UV-induced photolesions ↓

removal of UV-induced photolesions ↓ coclastogenicity with UV Removal of B[a]P-induced lesions ↓

comutagenicity with UV

removal of UV-induced photolesions ↓

End Point

Table 18.2 Studies providing evidence for inhibition of nucleotide and base excision repair by arsenicals

42, 120

122

116

42

114 115

113 86

117

111, 112

47

Ref.

500 Arsenic-Induced Carcinogenicity

h. BEAS-2B bronchial epithelial cells h. lymphocytes (in vivo) h. lymphocytes (in vivo) (in vitro) h. VH10hTert fibroblasts

Arsenite (5 µM)

H2O2-stimulated poly(ADP-ribosyl)ation ↓

PARP-1 activity ↓

h. Molt-3 T-cell lymphoma cells

h. HeLa S3 cervic carcinoma cells

h. HeLa S3 cervic carcinoma cells

PARP-1

Arsenite (0.01 µM)

MMAIII (0.001 µM) DMAIII (0.001 µM) MMAV/DMAV (no effect up to 250 µM) Arsenite (10 µM) MMAIII (10 µM) DMAIII (10 µM)

BER (NER)

BER (NER) BER (NER) BER (NER)

BER

BER BER

NER

NER BER NER BER NER NER

Pathway

124

124

123

142

89

127 128

138

140 139

137

136

Ref.

2-AAAF, 2-acetoxyacetylaminofluorene; APE1, AP-endonuclease 1, BER endonuclease; B[a]P, benzo[a]pyrene; BER, base excision repair; BPDE, benzo[a]pyrene diolepoxide; ERCC1, excision repair cross-complementing rodent repair deficiency complementation group 1; FEN1, FLAP endonuclease 1; h, human; LIG1, DNA-Ligase 1; MMC, mitomycin C; NER, nucleotide excision repair; PARP-1, poly(ADP-ribose)polymerase-1; PCNA, proliferating cell nuclear antigen; p48, subunit of the DDB2 NER damage recognition protein; Pold/e, DNA polymerase d/e; RFC, replication factor C; UDG2, uracil DNA glycosylase 2; XP, Xeroderma Pigmentosum group proteins, NER damage recognition / incision proteins; XPAzf, peptide representing the zinc finger domain of the human XPA protein.

H2O2-stimulated poly(ADP-ribosyl)ation ↓

BER incision activity (cell extracts) ↓ APE1↓, Polb↓, Polb↓, PARP-activity ↓

Arsenite (2.5 µM)

Arsenite (≥5 µM)

Chinese V79 hamster cells h. HOS osteosarcoma cells h. AG06 keratinocytes LIG1 h. HaCaT keratinocytes

XPC↓, p48↓, p53↓ FEN1↓, Pold↓, Pole↓, UDG2↓, p53↓ XPC ↓, XPD ↓, RFC ↓, PCNA ↓, LIG1↓ APE1↓, LIG1↓ ERCC1↓, XPF ↓, XPB↓ ERCC1↓, ERCC1 protein ↓ repair of 2-AAAF-induced lesions ↓ Disturbed assembly of NER proteins after UV-irradiation XPC ↓, p48↓, XPC protein ↓ Nuclear DNA ligase activity ↓ Ligase activity in nuclear extracts of cells ↓

End Point

Arsenite (10 µM) Arsenite (10 µM) (>2 mM)

Arsenite (5 µM) MMAIII (0.1 µM)

As in drinking water As in drinking water

h. normal epidermal keratinocytes

Species (test system)

Arsenite (0.5 µM)

Arsenic compound (lowest effective dose)

Acknowledgements 501

502

Arsenic-Induced Carcinogenicity Exposure towards arsenicals Metabolism in vivo

Inorganic and organic trivalent and pentavalent arsenicals

Oxidative stress, protein binding, altered DNA methylation

• DNA damage, protein oxidation, lipid peroxidation

• Alteration of gene expression

• Inhibition of DNA repair

• Modulation of signal transduction pathways

Persistence of DNA lesions, mutations, chromosomal damage, upregulation of stress proteins and protooncogenes, downregulation of tumour suppressor genes, modulation of cell cycle

Accumulation of mutations, enhanced cell poliferation

Carcinogenicity, cocarcinogenicity

Figure 18.2 Major mechanisms in arsenic induced carcinogenicity

Abbreviations 2-AAAF AP-1 APE1 AS3MT ATG B[a]P BER BPDE CA DMAV DMAIII DMAG GSH GSTw ERCC1

2-acetoxyacetylaminofluorene activator protein-1 AP-endonuclease 1 arsenic (+3 oxidation state) methyltransferase arsenite triglutathione benzo[a]pyrene base excision repair benzo[a]pyrene diolepoxide chromosomal aberrations dimethylarsinic acid dimethylarsinous acid dimethylarsinic glutathione glutathione glutathione S-transferase w excision repair cross-complementing rodent repair deficiency complementation group 1

References

FEN1 h IARC L LIG1 MADG MAK commission MMAV MMAIII MN MMC NER NFkB Nrf2 OC PARP-1 PCNA p48 PNP Pold/e RFC RNS ROS SAHC SAM SCE UC UDG2 USEPA XP XPAzf

503

FLAP endonuclease 1 human International Agency for Research on Cancer lymphocytes DNA ligase 1 monomethylarsonic diglutathione German Commission for the Investigation of Health Hazards of Chemical Compounds in the Work Area monomethylarsonic acid monomethylarsonous acid micronuclei mitomycin C nucleotide excision repair nuclear factor-kB factor-erythroid 2-related factor 2 oral mucosa cells poly(ADP-ribose)polymerase-1 proliferating cell nuclear antigen subunit of the DDB2 NER damage recognition protein purine nucleoside phosphorylase DNA polymerase d/e replication factor C reactive nitrogen species reactive oxygen species S-adenosylhomocysteine S-adenosylmethionine sister-chromatid exchanges urothelial cells uracil DNA glycosylase 2 United States Environmental Protection Agency Xeroderma pigmentosum group proteins, NER damage recognition/incision proteins peptide representing the zinc finger domain of the human XPA protein

References 1. ATSDR; Toxicological Profile For Arsenic; Atlanta, Georgia: U.S. Department of Health and Human Services, Public Health Service 2000. 2. ATSDR; Toxicological Profile for Arsenic; Atlanta, Georgia: U.S. Department of Health and Human Services, Public Health Service 2007. 3. Chou, C.H.; De Rosa, C.T.; Case studies-arsenic; Int. J. Hyg. Environ. Health 2003, 206, 381–386. 4. National Research Council; Arsenic in drinking water; Washington D.C.; National Academy Press; 1999.

504


5. National Research Council; Arsenic in drinking water 2001 update; Washington D.C.; National Academy Press; 2001. 6. Yoshida, T.; Yamauchi, H.; Fan Sun, G.; Chronic health effects in people exposed to arsenic via the drinking water; dose-response relationships in review; Toxicol. Appl. Pharmacol. 2004, 198, 243–252. 7. IARC; Arsenic and Arsenic compounds; Lyon 1980. 8. IARC; Arsenic and Arsenic compounds (Group 1) 1987. 9. Greim, H. (ed); The MAK-Collection for Occupational Health and Safety; MAK Value Documentations; Arsenic and its inorganic compounds (except arsine); Wiley-VCH; Weinheim; 2002, 21, 49–106. 10. WHO; Arsenic and arsenic compounds; Genf 2001. 11. DFG; Deutsche Forschungsgemeinschaft; List of MAK and BAT values 2008; Report No. 44; Wiley-VCH; Weinheim; 2008. 12. Waalkes, M.P.; Liu, J.; Ward, J.M.; Diwan, B.A.; Animal models for arsenic carcinogenesis; inorganic arsenic is a transplacental carcinogen in mice; Toxicol. Appl. Pharmacol. 2004, 198, 377–384. 13. Wang, J.P.; Qi, L.; Moore, M.R.; Ng, J.C.; A review of animal models for the study of arsenic carcinogenesis; Toxicol. Lett. 2002, 133, 17–31. 14. Waalkes, M.P.; Liu, J.; Diwan, B.A.; Transplacental arsenic carcinogenesis in mice; Toxicol. Appl. Pharmacol. 2007, 222, 271–280. 15. Pershagen, G.; Nordberg, G.; Bjorklund, N.E.; Carcinomas of the respiratory tract in hamsters given arsenic trioxide and/or benzo[a]pyrene by the pulmonary route; Environ. Res. 1984, 34, 227–241. 16. Rossman, T.G.; Uddin, A.N.; Burns, F.J.; Bosland, M.C.; Arsenite is a cocarcinogen with solar ultraviolet radiation for mouse skin; an animal model for arsenic carcinogenesis; Toxicol. Appl. Pharmacol. 2001, 176, 64–71. 17. Burns, F.J.; Uddin, A.N.; Wu, F.; Nadas, A.; Rossman, T.G.; Arsenic-induced enhancement of ultraviolet radiation carcinogenesis in mouse skin; a dose-response study; Environ. Health Perspect. 2004, 112, 599–603. 18. Chiou, H.Y.; Hsueh, Y.M.; Liaw, K.F.; Horng, S.F.; Chiang, M.H.; Pu, Y.S.; et al., Incidence of internal cancers and ingested inorganic arsenic; a seven-year follow-up study in Taiwan; Cancer Res. 1995, 55, 1296–1300. 19. Tsuda, T.; Babazono, A.; Yamamoto, E.; Kurumatani, N.; Mino, Y.; Ogawa, T.; et al., Ingested arsenic and internal cancer; a historical cohort study followed for 33 years; Am. J. Epidemiol. 1995, 141, 198–209. 20. Rossman, T.G.; Uddin, A.N.; Burns, F.J.; Evidence that arsenite acts as a cocarcinogen in skin cancer; Toxicol. Appl. Pharmacol. 2004, 198, 394–404. 21. Kenyon, E.M.; Hughes, M.F.; A concise review of the toxicity and carcinogenicity of dimethylarsinic acid; Toxicology 2001, 160, 227–236. 22. Kitchin, K.T.; Recent advances in arsenic carcinogenesis; modes of action, animal model systems, and methylated arsenic metabolites; Toxicol. Appl. Pharmacol. 2001, 172, 249–261. 23. Pott, W.A.; Benjamin, S.A.; Yang, R.S.; Pharmacokinetics, metabolism, and carcinogenicity of arsenic; Rev. Environ. Contam. Toxicol. 2001, 169, 165–214. 24. Thomas, D.J.; Molecular processes in cellular arsenic metabolism; Toxicol. Appl. Pharmacol. 2007, 222, 365–373. 25. Aposhian, V.H.; Zakharyan, R.A.; Avram, M.D.; Sampayo-Reyes, A.; Wollenber,g M.L.; A review of the enzymology of arsenic metabolism and a new potential role of hydrogen peroxide in the detoxication of the trivalent arsenic species; Toxicol. Appl. Pharmacol. 2004, 198, 327–335. 26. Aposhian, H.V.; Aposhian, M.M.; Arsenic toxicology: five questions; Chem. Res. Toxicol. 2006, 19, 1–15. 27. Cullen, W.R.; Reimer, K.J.; Arsenic speciation in the environment; Chemical Reviews (Washington DC, United States) 1989, 89, 713–764.

References

505

28. Hayakawa, T.; Kobayashi, Y.; Cui, X.; Hirano, S.; A new metabolic pathway of arsenite: arsenic-glutathione complexes are substrates for human arsenic methyltransferase Cyt19; Arch. Toxicol. 2005, 79, 508–518. 29. Naranmandura, H.; Suzuki, N.; Suzuki, K.T.; Trivalent arsenicals are bound to proteins during reductive methylation; Chem. Res. Toxicol. 2006, 19, 1010–1018. 30. Hartwig, A.; Blessing, H.; Schwerdtle, T.; Walter, I.; Modulation of DNA repair processes by arsenic and selenium compounds; Toxicology 2003, 193, 161–169. 31. Kumagai, Y.; Sumi, D.; Arsenic; signal transduction, transcription factor, and biotransformation involved in cellular response and toxicity; Annu. Rev. Pharmacol. Toxicol. 2007; 47, 243–262. 32. Kligerman, A.D.; Tennant, A.H.; Insights into the carcinogenic mode of action of arsenic; Toxicol. Appl. Pharmacol. 2007, 222, 281–288. 33. Hughes, M.F.; Arsenic toxicity and potential mechanisms of action; Toxicol. Lett. 2002, 133, 1–16. 34. Huang, C.; Ke, Q.; Costa, M.; Shi, X.; Molecular mechanisms of arsenic carcinogenesis; Mol. Cell. Biochem. 2004, 255, 57–66. 35. Shi, H.; Shi, X.; Liu, K.J.; Oxidative mechanism of arsenic toxicity and carcinogenesis; Mol. Cell. Biochem. 2004, 255, 67–78. 36. Schoen, A.; Beck, B.; Sharma, R.; Dube, E.; Arsenic toxicity at low doses; epidemiological and mode of action considerations; Toxicol. Appl. Pharmacol. 2004, 198, 253–267. 37. Gebel, T.W.; Genotoxicity of arsenical compounds; Int. J. Hyg. Environ. Health 2001, 203, 249–262. 38. Hei, T.K.; Liu, S.X.; Waldren, C.; Mutagenicity of arsenic in mammalian cells; role of reactive oxygen species; Proc. Natl. Acad. Sci. USA 1998, 95, 8103–8107. 39. Li, Y.M.; Broome, J.D.; Arsenic targets tubulins to induce apoptosis in myeloid leukemia cells; Cancer Res. 1999, 59, 776–780. 40. Lee, T.C.; Tanaka, N.; Lamb, P.W.; Gilmer, T.M.; Barrett, J.C.; Induction of gene amplification by arsenic; Science 1988, 241, 79–81. 41. Schwerdtle, T.; Walter, I.; Mackiw, I.; Hartwig, A.; Induction of oxidative DNA damage by arsenite and its trivalent and pentavalent methylated metabolites in cultured human cells and isolated DNA; Carcinogenesis 2003, 24, 967–974. 42. Schwerdtle, T.; Walter, I., Hartwig, A.; Arsenite and its biomethylated metabolites interfere with the formation and repair of stable BPDE-induced DNA adducts in human cells and impair XPAzf and Fpg; DNA Repair (Amst) 2003, 2, 1449–1463. 43. Wang, T.S.; Chung, C.H.; Wang, A.S.; Bau, D.T.; Samikkannu, T.; Jan, K.Y.; et al., Endonuclease III, formamidopyrimidine-DNA glycosylase, and proteinase K additively enhance arsenic-induced DNA strand breaks in human cells; Chem. Res. Toxicol. 2002, 15, 1254–1258. 44. Mass, M.J.; Tennant, A.; Roop, B.C.; Cullen, W.R.; Styblo, M.; Thomas, D.J.; et al., Methylated trivalent arsenic species are genotoxic; Chem. Res. Toxicol. 2001, 14, 355–361. 45. Kligerman, A.D.; Doerr, C.L.; Tennant, A.H.; Harrington-Brock, K.; Allen, J.W.; Winkfield, E.; et al., Methylated trivalent arsenicals as candidate ultimate genotoxic forms of arsenic: induction of chromosomal mutations but not gene mutations; Environ. Mol. Mutagen. 2003, 42, 192–205. 46. Rossman, T.G.; Molina, M.; Klein, C.B.; Comutagens in E; Coli and chinese hamster cells with special attention to arsenite; 1985. 47. Okui, T.; Fujiwara, Y.; Inhibition of human excision DNA repair by inorganic arsenic and the comutagenic effect in V79 Chinese hamster cells; Mutat. Res. 1986, 172, 69–76. 48. Li, J.H.; Rossman, T.G.; Comutagenesis of sodium arsenite with ultraviolet radiation in Chinese hamster V79 cells; Biol. Met. 1991, 4, 197–200. 49. Maier, A.; Schumann, B.L.; Chang, X.; Talaska, G.; Puga, A.; Arsenic co-exposure potentiates benzo[a]pyrene genotoxicity; Mutat. Res. 2002, 517, 101–111.

506


50. Lewinska, D.; Arkusz, J.; Stanczyk, M.; Palus, J.; Dziubaltowska, E.; Stepnik, M.; Comparison of the effects of arsenic and cadmium on benzo(a)pyrene-induced micronuclei in mouse bone-marrow; Mutat. Res. 2007, 632, 37–43. 51. Wiencke, J.K.; Yager, J.W.; Varkonyi, A.; Hultner, M.; Lutze, L.H.; Study of arsenic mutagenesis using the plasmid shuttle vector pZ189 propagated in DNA repair proficient human cells; Mutat. Res. 1997, 386, 335–344. 52. Beyersmann, D.; Hartwig, A.; Carcinogenic metal compounds: recent insight into molecular and cellular mechanisms; Arch. Toxicol. 2008, 82, 493–512. 53. Ostrosky-Wegman, P.; Gonsebatt, M.E.; Montero, R.; Vega, L.; Barba, H.; Espinosa, J.; et al., Lymphocyte proliferation kinetics and genotoxic findings in a pilot study on individuals chronically exposed to arsenic in Mexico; Mutat. Res. 1991, 250, 477–482. 54. Lerda, D.; Sister-chromatid exchange (SCE) among individuals chronically exposed to arsenic in drinking water; Mutat. Res. 1994, 312, 111–120. 55. Warner, M.L.; Moore, L.E.; Smith, M.T.; Kalman, D.A.; Fanning, E.; Smith, A.H.; Increased micronuclei in exfoliated bladder cells of individuals who chronically ingest arsenic-contaminated water in Nevada. Cancer Epidemiol. Biomarkers Prev. 1994, 3, 583– 590. 56. Dulout, F.N.; Grillo, C.A.; Seoane, A.I.; Maderna, C.R.; Nilsson, R.; Vahter, M.; et al., Chromosomal aberrations in peripheral blood lymphocytes from native Andean women and children from northwestern Argentina exposed to arsenic in drinking water; Mutat. Res. 1996, 370, 151–158. 57. Gonsebatt, M.E.; Vega, L.; Salazar, A.M.; Montero, R.; Guzman, P.; Blas, J.; et al., Cytogenetic effects in human exposure to arsenic; Mutat. Res. 1997, 386, 219–228. 58. Moore, L.E.; Smith, A.H.; Hopenhayn-Rich, C.; Biggs, M.L.; Kalman, D.A.; Smith, M.T.; Micronuclei in exfoliated bladder cells among individuals chronically exposed to arsenic in drinking water; Cancer Epidemiol. Biomarkers Prev. 1997, 6, 31–36. 59. Maki-Paakkanen, J.; Kurttio, P.; Paldy, A.; Pekkanen, J.; Association between the clastogenic effect in peripheral lymphocytes and human exposure to arsenic through drinking water; Environ. Mol. Mutagen. 1998, 32, 301–313. 60. Tian, D.; Ma, H.; Feng, Z.; Xia, Y.; Le, X.C.; Ni, Z.; et al., Analyses of micronuclei in exfoliated epithelial cells from individuals chronically exposed to arsenic via drinking water in inner Mongolia, China. J. Toxicol. Environ. Health A 2001, 64, 473–484. 61. Basu, A.; Mahata, J.; Roy, A.K.; Sarkar, J.N.; Poddar, G.; Nandy, A.K.; et al., Enhanced frequency of micronuclei in individuals exposed to arsenic through drinking water in West Bengal, India; Mutat. Res. 2002, 516, 29–40. 62. Mahata, J.; Basu, A.; Ghoshal, S.; Sarkar, J.N.; Roy, A.K.; Poddar, G.; et al., Chromosomal aberrations and sister chromatid exchanges in individuals exposed to arsenic through drinking water in West Bengal, India; Mutat. Res. 2003, 534, 133–143. 63. Basu, A.; Ghosh, P.; Das, J.K.; Banerjee, A.; Ray, K.; Giri, A.K.; Micronuclei as biomarkers of carcinogen exposure in populations exposed to arsenic through drinking water in West Bengal, India: a comparative study in three cell types; Cancer Epidemiol. Biomarkers Prev. 2004, 13, 820–827. 64. Mahata, J.; Chaki, M.; Ghosh, P.; Das, L.K.; Baidya, K.; Ray, K.; et al., Chromosomal aberrations in arsenic-exposed human populations; a review with special reference to a comprehensive study in West Bengal, India; Cytogenet. Genome. Res. 2004, 104, 359–364. 65. Martinez, V.; Creus, A.; Venegas, W.; Arroyo, A.; Beck, J.P.; Gebel, T.W.; et al., Evaluation of micronucleus induction in a Chilean population environmentally exposed to arsenic; Mutat. Res. 2004, 564, 65–74. 66. Martinez, V.; Creus, A.; Venegas, W.; Arroyo, A.; Beck, J.P.; Gebel, T.W.; et al., Micronuclei assessment in buccal cells of people environmentally exposed to arsenic in northern Chile; Toxicol. Lett. 2005, 155, 319–327. 67. Ghosh, P.; Basu, A.; Mahata, J.; Basu, S.; Sengupta, M.; Das, J.K.; et al., Cytogenetic damage and genetic variants in the individuals susceptible to arsenic-induced cancer through drinking water; Int. J. Cancer 2006, 118, 2470–2478.

References

507

68. Kumagai, Y.; Pi, J.; Molecular basis for arsenic-induced alteration in nitric oxide production and oxidative stress; implication of endothelial dysfunction; Toxicol. Appl. Pharmacol. 2004, 198, 450–457. 69. Hughes, M.F.; Kitchin, K.T.; Arsenic, oxidative stress and carcniogenesis; In; Singh HH, editor; Oxidative Stress, Disease and Cancer; London; Imperial College Press; 2006, 825–850. 70. Lynn, S.; Gurr, J.R.; Lai, H.T.; Jan, K.Y.; NADH oxidase activation is involved in arseniteinduced oxidative DNA damage in human vascular smooth muscle cells; Circ. Res. 2000, 86, 514–519. 71. Corsini, E.; Asti, L.; Viviani, B.; Marinovich, M.; Galli, C.L.; Sodium arsenate induces overproduction of interleukin-1alpha in murine keratinocytes: role of mitochondria; J. Invest. Dermatol. 1999, 113, 760–765. 72. Yamanaka, K.; Hoshino, M.; Okamoto, M.; Sawamura, R.; Hasegawa, A.; Okada, S.; Induction of DNA damage by dimethylarsine, a metabolite of inorganic arsenics, is for the major part likely due to its peroxyl radical; Biochem. Biophys. Res. Commun. 1990, 168, 58–64. 73. Yamanaka, K.; Kato, K.; Mizoi, M.; An, Y.; Takabayashi, F.; Nakano, M.; et al., The role of active arsenic species produced by metabolic reduction of dimethylarsinic acid in genotoxicity and tumorigenesis; Toxicol. Appl. Pharmacol. 2004, 198, 385–393. 74. Yamanaka, K.; Hasegawa, A.; Sawamura, R.; Okada, S.; Cellular response to oxidative damage in lung induced by the administration of dimethylarsinic acid, a major metabolite of inorganic arsenics, in mice; Toxicol. Appl. Pharmacol. 1991, 108, 205–213. 75. Yamanaka, K.; Okada, S.; Induction of lung-specific DNA damage by metabolically methylated arsenics via the production of free radicals; Environ. Health Perspect. 1994, 102 Suppl 3, 37–40. 76. Yamanaka, K.; Hayashi, H.; Tachikawa, M.; Kato, K.; Hasegawa, A.; Oku, N.; et al., Metabolic methylation is a possible genotoxicity-enhancing process of inorganic arsenics; Mutat. Res. 1997, 394, 95–101. 77. Yamanaka, K.; Katsumata, K.; Ikuma, K.; Hasegawa, A.; Nakano, M.; Okada, S.; The role of orally administered dimethylarsinic acid, a main metabolite of inorganic arsenics, in the promotion and progression of UVB-induced skin tumorigenesis in hairless mice; Cancer Lett. 2000, 152, 79–85. 78. Yamanaka, K.; Takabayashi, F.; Mizoi, M.; An, Y.; Hasegawa, A.; Okada, S.; Oral exposure of dimethylarsinic acid, a main metabolite of inorganic arsenics, in mice leads to an increase in 8-Oxo-2′-deoxyguanosine level, specifically in the target organs for arsenic carcinogenesis; Biochem. Biophys. Res. Commun. 2001, 287, 66–70. 79. Yamanaka, K.; Mizoi, M.; Tachikawa, M.; Hasegawa, A.; Hoshino, M., Okada, S.; Oxidative DNA damage following exposure to dimethylarsinous iodide; the formation of cis-thymine glycol; Toxicol. Lett. 2003, 143, 145–153. 80. Ahmad, S.; Kitchin, K.T.; Cullen, W.R.; Arsenic species that cause release of iron from ferritin and generation of activated oxygen; Arch. Biochem. Biophys. 2000, 382, 195–202. 81. Ahmad, S.; Kitchin, K.T.; Cullen, W.R.; Plasmid DNA damage caused by methylated arsenicals, ascorbic acid and human liver ferritin; Toxicol. Lett. 2002, 133, 47–57. 82. Li, B.; Sun, Y.; Sun, X.; Wang, Y.; Li, X.; Kumagai, Y.; et al., Monomethylarsonous acid induced cytotoxicity and endothelial nitric oxide synthase phosphorylation in endothelial cells; Bull. Environ. Contam. Toxicol. 2007, 78, 455–458. 83. Epe, B.; Role of endogenous oxidative DNA damage in carcinogenesis; what can we learn from repair-deficient mice? Biol. Chem. 2002, 383, 467–475. 84. Chouchane, S.; Snow, E.T.; In vitro effect of arsenical compounds on glutathione-related enzymes; Chem. Res. Toxicol. 2001, 14, 517–522. 85. Styblo, M.; Serves, S.V.; Cullen, W.R.; Thomas, D.J.; Comparative inhibition of yeast glutathione reductase by arsenicals and arsenothiols; Chem. Res. Toxicol. 1997, 10, 27–33. 86. Bau, D.T.; Gurr, J.R.; Jan, K.Y.; Nitric oxide is involved in arsenite inhibition of pyrimidine dimer excision; Carcinogenesis 2001, 22, 709–716.

508


87. Snow, E.T.; Sykora, P.; Durham, T.R.; Klein, C.B.; Arsenic, mode of action at biologically plausible low doses; What are the implications for low dose cancer risk? Toxicol. Appl. Pharmacol. 2005, 207, 557–564. 88. Li, M.; Cai, J.F.; Chiu, J.F.; Arsenic induces oxidative stress and activates stress gene expressions in cultured lung epithelial cells; J. Cell. Biochem. 2002, 87, 29–38. 89. Sykora, P.; Snow, E.T.; Modulation of DNA polymerase beta-dependent base excision repair in cultured human cells after low dose exposure to arsenite; Toxicol. Appl. Pharmacol. 2008, 228, 385–394. 90. Leonard, S.S.; Harris, G.K.; Shi, X.; Metal-induced oxidative stress and signal transduction; Free. Radic. Biol. Med. 2004, 37, 1921–1942. 91. Kubota, R.; Kunito, T.; Agusa, T.; Fujihara, J.; Monirith, I.; Iwata, H.; et al., Urinary 8-hydroxy-2′-deoxyguanosine in inhabitants chronically exposed to arsenic in groundwater in Cambodia; J. Environ. Monit. 2006, 8, 293–299. 92. Mo, J.; Xia, Y.; Wade, T.J.; Schmitt, M.; Le, X.C.; Dang, R.; et al., Chronic arsenic exposure and oxidative stress, OGG1 expression and arsenic exposure, nail selenium, and skin hyperkeratosis in Inner Mongolia; Environ. Health Perspect. 2006, 114, 835–841. 93. Basu, A.; Som, A.; Ghoshal, S.; Mondal, L.; Chaubey, R.C.; Bhilwade, H.N.; et al., Assessment of DNA damage in peripheral blood lymphocytes of individuals susceptible to arsenic induced toxicity in West Bengal, India; Toxicol. Lett. 2005, 159, 100–112. 94. Fujino, Y.; Guo, X.; Liu, J.; Matthews, I.P.; Shirane, K.; Wu, K.; et al., Chronic arsenic exposure and urinary 8-hydroxy-2’-deoxyguanosine in an arsenic-affected area in Inner Mongolia, China; J. Expo. Anal. Environ. Epidemiol. 2005, 15, 147–152. 95. Pi, J.; Yamauchi, H.; Kumagai, Y.; Sun, G.; Yoshida, T.; Aikawa, H.; et al., Evidence for induction of oxidative stress caused by chronic exposure of Chinese residents to arsenic contained in drinking water; Environ. Health Perspect. 2002, 110, 331–336. 96. Wu, M.M.; Chiou, H.Y.; Wang, T.W.; Hsueh, Y.M.; Wang, I.H.; Chen, C.J.; et al., Association of blood arsenic levels with increased reactive oxidants and decreased antioxidant capacity in a human population of northeastern Taiwan; Environ. Health Perspect. 2001, 109, 1011–1017. 97. Breton, C.V.; Kile, M.L.; Catalano, P.J.; Hoffman, E.; Quamruzzaman, Q.; Rahman, M.; et al., GSTM1 and APE1 genotypes affect arsenic-induced oxidative stress: a repeated measures study; Environ. Health 2007, 6, 39. 98. Baylin, S.B.; Belinsky, S.A.; Herman, J.G.; Aberrant methylation of gene promoters in cancer-concepts, misconcepts, and promise; J. Natl. Cancer Inst. 2000, 92, 1460– 1461. 99. Baylin, S.B.; Herman, J.G.; DNA hypermethylation in tumorigenesis; epigenetics joins genetics; Trends. Genet. 2000, 16, 168–174. 100. Mass, M.J.; Wang, L.; Arsenic alters cytosine methylation patterns of the promoter of the tumor suppressor gene p53 in human lung cells: a model for a mechanism of carcinogenesis; Mutat. Res. 1997, 386, 263–277. 101. Zhong, C.X.; Mass, M.J.; Both hypomethylation and hypermethylation of DNA associated with arsenite exposure in cultures of human cells identified by methylationsensitive arbitrarily-primed PCR; Toxicol. Lett. 2001, 122, 223–234. 102. Zhao, C.Q.; Young, M.R.; Diwan, B.A.; Coogan, T.P.; Waalkes, M.P.; Association of arsenic-induced malignant transformation with DNA hypomethylation and aberrant gene expression; Proc. Natl. Acad. Sci. USA 1997, 94, 10907–10912. 103. Reichard, J.F.; Schnekenburger, M.; Puga, A.; Long term low-dose arsenic exposure induces loss of DNA methylation; Biochem. Biophys. Res. Commun. 2007, 352, 188–192. 104. Cui, X.; Wakai, T.; Shirai, Y.; Hatakeyama, K.; Hirano, S.; Chronic oral exposure to inorganic arsenate interferes with methylation status of p16INK4a and RASSF1A and induces lung cancer in A/J mice; Toxicol. Sci. 2006, 91, 372–381. 105. Chanda S, Dasgupta UB, Guhamazumder D, Gupta M, Chaudhuri U, Lahiri S, et al., DNA hypermethylation of promoter of gene p53 and p16 in arsenic-exposed people with and without malignancy. Toxicol. Sci. 2006; 89: 431–437.

References

509

106. Marsit, C.J.; Karagas, M.R.; Danaee, H.; Liu, M.; Andrew, A.; Schned, A.; et al., Carcinogen exposure and gene promoter hypermethylation in bladder cancer; Carcinogenesis 2006, 27, 112–116. 107. Huen, M.S.; Chen, J.; The DNA damage response pathways: at the crossroad of protein modifications; Cell. Res. 2008, 18, 8–16. 108. Friedberg, E.C.; Walker, G.C.; Siede, W.; Wood, R.D.; Schultz, R.A.; Ellenberger, T.; DNA Repair and Mutagenesis; Second Edition; Washington; ASM Press; 2006. 109. Friedberg, E.C.; A brief history of the DNA repair field. Cell. Res. 2008, 18, 3–7. 110. Hakem, R.; DNA-damage repair; the good, the bad, and the ugly; EMBO J. 2008, 27, 589–605. 111. Lee-Chen, S.F.; Yu, C.T.; Jan, K.Y.; Effect of arsenite on the DNA repair of UVirradiated Chinese hamster ovary cells; Mutagenesis 1992, 7, 51–55. 112. Wang, T.C.; Huang, J.S.; Yang, V.C.; Lan, H.J.; Lin, C.J.; Jan, K.Y.; Delay of the excision of UV light-induced DNA adducts is involved in the coclastogenicity of UV light plus arsenite; Int. J. Radiat. Biol. 1994, 66, 367–372. 113. Hartwig, A.; Gröblinghoff, U.D.; Beyersmann, D.; Natarajan, A.T.; Filon, R.; Mullenders, L.H.; Interaction of arsenic(III) with nucleotide excision repair in UV- irradiated human fibroblasts; Carcinogenesis 1997, 18, 399–405. 114. Tran, H.P.; Prakash, A.S.; Barnard, R.; Chiswell, B.; Ng, J.C.; Arsenic inhibits the repair of DNA damage induced by benzo(a)pyrene; Toxicol. Lett. 2002, 133, 59–67. 115. Danaee, H.; Nelson, H.H.; Liber, H.; Little, J.B.; Kelsey, K.T.; Low dose exposure to sodium arsenite synergistically interacts with UV radiation to induce mutations and alter DNA repair in human cells; Mutagenesis 2004, 19, 143–148. 116. Shen, S.; Lee, J.; Weinfeld, M.; Le, X.C.; Attenuation of DNA damage-induced p53 expression by arsenic; A possible mechanism for arsenic co-carcinogenesis; Mol. Carcinog. 2008, 47, 508–518. 117. Hartmann, A.; Speit, G.; Effect of arsenic and cadmium on the persistence of mutageninduced DNA lesions in human cells; Environ. Mol. Mutagen. 1996, 27, 98–104. 118. Fousteri, M.; Mullenders, L.H.; Transcription-coupled nucleotide excision repair in mammalian cells; molecular mechanisms and biological effects; Cell. Res. 2008, 18, 73–84. 119. Kitchin, K.T.; Wallace, K.; The role of protein binding of trivalent arsenicals in arsenic carcinogenesis and toxicity; J. Inorg. Biochem. 2008, 102, 532–539. 120. Piatek, K.; Schwerdtle, T.; Hartwig, A.; Bal, W.; Monomethylarsonous Acid destroys a tetrathiolate zinc finger much more efficiently than inorganic arsenite; mechanistic considerations and consequences for DNA repair inhibition; Chem. Res. Toxicol. 2008, 21, 600–606. 121. Asmuss, M.; Mullenders, L.H.; Eker, A.; Hartwig, A.; Differential effects of toxic metal compounds on the activities of Fpg and XPA, two zinc finger proteins involved in DNA repair; Carcinogenesis 2000, 21, 2097–2104. 122. Mustra, D.J.; Warren, A.J.; Wilcox, D.E.; Hamilton, J.W.; Preferential binding of human XPA to the mitomycin C-DNA interstrand crosslink and modulation by arsenic and cadmium; Chem. Biol. Interact. 2007, 168, 159–168. 123. Hartwig, A.; Pelzer, A.; Asmuss, M.; Bürkle, A.; Very low concentrations of arsenite suppress poly(ADP-ribosyl)ation in mammalian cells; Int. J. Cancer 2003, 104, 1–6. 124. Walter, I.; Schwerdtle, T.; Thuy, C.; Parsons, J.L.; Dianov, G.L.; Hartwig, A.; Impact of arsenite and its methylated metabolites on PARP-1 activity, PARP-1 gene expression and poly(ADP-ribosyl)ation in cultured human cells; DNA Repair (Amst) 2007, 6, 61–70. 125. Bürkle, A.; Brabeck, C.; Diefenbach, J.; Beneke, S.; The emerging role of poly(ADPribose) polymerase-1 in longevity; Int. J. Biochem. Cell. Biol. 2005, 37, 1043–1053. 126. Poonepalli, A.; Balakrishnan, L.; Khaw, A.K.; Low, G.K.; Jayapal, M.; Bhattacharjee, R.N.; et al., Lack of poly(ADP-ribose) polymerase-1 gene product enhances cellular sensitivity to arsenite; Cancer Res. 2005, 65, 10977–10983. 127. Li, J.H.; Rossman, T.G.; Inhibition of DNA ligase activity by arsenite; a possible mechanism of its comutagenesis; Mol. Toxicol. 1989, 2, 1–9.

510


128. Hu, Y.; Su, L.; Snow, E.T.; Arsenic toxicity is enzyme specific and its affects on ligation are not caused by the direct inhibition of DNA repair enzymes; Mutat. Res. 1998, 408, 203–218. 129. Yih, L.H.; Lee, T.C.; Arsenite induces p53 accumulation through an ATM-dependent pathway in human fibroblasts; Cancer Res. 2000, 60, 6346–6352. 130. Menendez, D.; Mora, G.; Salazar, A.M.; Ostrosky-Wegman, P.; ATM status confers sensitivity to arsenic cytotoxic effects; Mutagenesis 2001, 16, 443–448. 131. Tang, F.; Liu, G.; He, Z.; Ma, W.Y.; Bode, A.M.; Dong, Z.; Arsenite inhibits p53 phosphorylation, DNA binding activity, and p53 target gene p21 expression in mouse epidermal JB6 cells; Mol. Carcinog. 2006, 45, 861–870. 132. Chai, C.Y.; Huang, Y.C.; Hung, W.C.; Kang, W.Y.; Chen, W.T.; Arsenic salt-induced DNA damage and expression of mutant p53 and COX-2 proteins in SV-40 immortalized human uroepithelial cells. Mutagenesis 2007, 22, 403–408. 133. Butler, J.S.; Loh, S.N.; Structure, function, and aggregation of the zinc-free form of the p53 DNA binding domain; Biochemistry 2003, 42, 2396–2403. 134. Wang, P.L.; Sait, F.; Winter, G.; The ‘wildtype’ conformation of p53: epitope mapping using hybrid proteins; Oncogene. 2001, 20, 2318–2324. 135. Martin, A.C.; Facchiano, A.M.; Cuff, A.L.; Hernandez-Boussard, T.; Olivier, M.; Hainaut, P.; et al., Integrating mutation data and structural analysis of the TP53 tumor-suppressor protein; Hum. Mutat. 2002, 19, 149–164. 136. Hamadeh, H.K.; Trouba, K.J.; Amin, R.P.; Afshari, C.A.; Germolec, D.; Coordination of altered DNA repair and damage pathways in arsenite-exposed keratinocytes; Toxicol. Sci. 2002, 69, 306–316. 137. Andrew, A.S.; Warren, A.J.; Barchowsky, A.; Temple, K.A.; Klei, L.; Soucy, N.V.; et al., Genomic and proteomic profiling of responses to toxic metals in human lung cells; Environ. Health Perspect. 2003, 111, 825–835. 138. Nollen, M.; Ebert, F.; Moser, J.; Mullenders, L.H.; Hartwig, A.; Schwerdtle, T.; Impact of arsenic on nucleotide excision repair; XPC function, protein level and gene expression; Molecular Nutrition and Food Research, 2008, submitted. 139. Andrew, A.S.; Burgess, J.L.; Meza, M.M.; Demidenko, E.; Waugh, M.G.; Hamilton, J.W.; et al., Arsenic exposure is associated with decreased DNA repair in vitro and in individuals exposed to drinking water arsenic; Environ. Health Perspect., 2006, 114, 1193–1198. 140. Andrew, A.S.; Karagas, M.R.; Hamilton, J.W.; Decreased DNA repair gene expression among individuals exposed to arsenic in United States drinking water; Int. J. Cancer, 2003, 104, 263–268. 141. Banerjee, M.; Sarma, N.; Biswas, R.; Roy, J.; Mukherjee, A.; Giri, A.K.; DNA repair deficiency leads to susceptibility to develop arsenic-induced premalignant skin lesions; Int. J. Cancer, 2008, 123, 283–287. 142. Yager, J.W.; Wiencke, J.K.; Inhibition of poly(ADP-ribose) polymerase by arsenite; Mutat. Res., 1997, 386, 345–351.

Index

Page numbers in italics refer to figures and diagrams. adenine dirhodium complexes, interaction with, 303, 304–305 metal ion affinities, 3 metal-DNA supramolecular chemistry 1D coordination polymers, 110– 113, 112, 115, 115–116, 116, 117 2D coordination polymers, 118– 119, 119 3D coordination polymers, 119– 121, 120, 121 chelates, 124, 125 H-bonding patterns, 123, 124 triangular supramolecules, 100–101 organotin complexes, interaction with, 309 platinum complexes, interaction with, 138, 138–139, 217, 220–222, 222 A-DNA, 335 affinity chromatography, 337–338 ammonium DNA affinity element, use as, 383 G-quadruplexes competition with Na+ and K+, 75–76

coordination in d[(G4T4G4)]2, 74 movement within, 79–80 residence times, 78–79 role in, 59 structure, effect on, 77–78 aptamers, 55, 275–276, 398 ARCUT system, 385, 385 arsenic carcinogenicity and cocarcinogenicity, 492, 502 exposure pathways, 491 genotoxicity, 495 metabolism of inorganic arsenic, 492–494, 493 modes of cancer induction, proposed, 502 epigenetic effects, 496–497 induction of genetic damage, 494–495 inhibition of DNA repair, 497–499, 500–501 oxidative stress, 495–496 ascorbate, 465–467, 466, 479 barium, 59, 60 base excision repair, 178, 179, 180, 193


512

Index

effected by arsenic, 498 repair of Cr(VI) induced nucleobase lesions, 473, 473–474 B-DNA, 22–23, 238, 239, 335 biomethylation of inorganic arsenic, 492–494, 493 cadmium metal-DNA supramolecular chemistry 1D coordination polymers, 113, 114, 114–116, 115, 116 supramolecular squares, 102, 102– 103, 104 caesium, 59 calcium basic properties of Ca2+ cations, 416–417 Ca2+ in substrate binding, 422–423 G-quadruplexes coordination in d(TG4T), 71, 72 role in, 59–60 polymerases and nucleases, inhibition of, 425 preferred coordination geometry of Ca2+ cations, 417 cancer, modes of action, 236 arsenic-induced cancers, 494–499, 502 base mismatches, 307 chromium-induced cancers, 464–465 telomere dysfunction, 209, 225 cancer therapies, 236. see also photodynamic therapy (PDT) antitelomerase drugs, 213, 223 chemotherapy, 176 oligonucleotide therapies, 273–276 platinum drugs, 135–137, 136, 200, 216–217, 224–225 potential G-quadruplex targets for antitumour drugs, 215–216 ruthenium complexes, 339 telomerase inhibition, 213, 225 carboplatin, 135, 136, 188 catalysis in nucleic acid enzymes. see nucleic acid enzyme catalysis cerium Ce2-HXTA complex in DNA hydrolysis, 378 Ce(IV) complexes in DNA hydrolysis, 373–374, 383–384

Ce(IV)-EDTA complex in ARCUT system, 385, 385 chemical shifts C chemical shifts, 446 calculation, 452–453 N chemical shifts, 447, 450, 452 in T-HgII-T base pair formation, 452–454 chromium adduct formation, 477–478, 478 carcinogenic classification and limits, 463–464 carcinogenic properties, 464–465, 479, 479–480 chromate structure, 464 chromium(III) structure, 464 clastogenic classification, 476 Cr(VI) reductions, intracellular by ascorbate, 465–467, 466 by glutathione, 466, 467 by hydrogen peroxide, 466, 467 Cr(VI)-induced DNA crosslinks formation, 480–481, 481 repair, 481–482 DNA damage, 468–473, 480–482 repair, 473–474, 476–477, 478–480, 479, 481–482 exposure sources, 468 hydrogen chromate structure, 464 industrial uses, 464 lesions formed, 472 8-oxoG, 469, 469–470, 470 crosslinks, 480–482 guanidinohydantoin lesions, 471, 472 spiroiminodihydantoin lesions, 470–471, 471, 472 nucleobase oxidation electron abstraction pathway, 469–473 reactive oxygen species pathway, 468–469 reduction potentials, 469 repair, 473–474 sugar oxidation C1′ hydrogen abstraction mechanism, 475, 475 C4′ hydrogen abstraction mechanism, 475–476, 476 repair, 476–477

Index cisplatin (and related platinum complexes), 135, 136. see also platinum drugs binding patterns, 18 conformer distribution in cisplatin adducts of G derivatives adducts with 1-Me-5′-GMP, 154–155 adducts with 3′-GMP, 152–154, 153 adducts with 5′-GMP, 154 conformer distribution signature, 152, 153 quantification of atropisomer distribution, 156 cytotoxic activity, 301–302 DNA adducts repair, 193–200 repair, inhibition of, 183 structure, 188–189 types, 139, 139–140 DNA adducts with tethered guanine bases chirality, 158 flexibility, 161–163, 162 interligand interactions, 159 multiple conformers, 159–160 structure of d(GpG), 157 wrapping of platinum binding site, 160–161, 161 G-quadruplex structures, interaction with human telomeric sequence AG3(T2AG3)3, 220, 220–221 known classes, 217–218, 218 platinum-acridine complex, 221– 222, 222 platinum-quinacridine complex, 221–222, 222 Tetrahymena sequence (T2G4)4, 218–219, 219 method of anticancer activity, 216–217, 301 models of G/G crosslinks, 140–141 platinated oligonucleotides, 279, 280, 287 retro modelling, 141–145, 142, 144 structure, 136 telomerase, interaction with, 223–224 telomeric DNA, interaction with, 223, 224

513

cobalt Co(II)-dependent DNAzymes, selection of, 398–399, 399 Co(III) complexes in DNA hydrolysis, 374, 383 G-quadruplexes, role in, 59–60 oligodeoxynucleotides, interaction with, 7, 15 copper copper complex-peptide conjugate, 353–354 Cu-(GMP) complex, 4, 5 Cu(II) complexes in DNA hydrolysis, 375–377, 376 DNA melting temperature, effect on, 4, 16 DNAzyme for Cu(II), 396 DNAzyme-based Cu2+ sensor, 406 metal-DNA supramolecular chemistry, 107 1D coordination polymers, 110– 113, 112 2D coordination polymers, 118, 118 3D porous coordination polymer, 119–120, 120 oligodeoxynucleotides, adducts with, 15–16 crosslinking of oligonucleotides, 276–278, 277, 278, 289–295 cytosine depurination, 46–47 double-proton transfer, 44–46, 45 metal ion affinities, 3 metal-DNA supramolecular chemistry 1D coordination polymers, 110, 110, 113 2D coordination polymers, 118, 118 chelates, 124 H-bonding patterns to metals, 122 large supramolecules, 107–108, 108 square supramolecules, 104 triangular supramolecules, 101, 101–102 platinum binding sites, possible, 138, 138, 140 point defects, 48 depurination, 46–47, 48 dirhodium complexes, 303–305, 304

514

Index

DNA. see also B-DNA; M-DNA catalytic properties, 31 control of metal complexes, 124, 125 DNA binding. see also metal ion-DNA interactions, binding chromium, 477–478, 478 groove binding, 22–26 intercalation vs. groove binding, 359–361 mismatch recognition, 308 nucleobase-metal binding angles, 97 photobinding, 259–260 platinum, 138, 138–140, 139 ruthenium, 325–329. see also ruthenium, dinuclear complexes as DNA probes; ruthenium, mononuclear complexes as DNA probes stability constants, 34–38, 35, 36, 37, 38 transition metal polyazine complexes, 257–258 DNA bulge sequences, 330, 330–331, 331, 333 DNA cleavage degradation pathways, 370, 370–371 hydrolysis. see DNA hydrolysis oxidative vs. hydrolytic, 362 phosphoryl transfer reaction, 415–416, 416 photocleavage, 6, 7–9, 260–262, 305, 308, 352 DNA damage. see also DNA photomodification and photochemical degradation agents of damage, 175, 176 arsenic-induced, 494–502 chromium-induced, 468–473, 480–482 cisplatin-induced, 140 deletion, 48 double-strand breaks, 182, 199 mismatches, 307–308 oxidation, 308 persistence, factors contributing to, 185 photosensitizer damage, 255 platinum adducts, 176–177 types of damage, 175–176 DNA grooves intercalation vs. groove binding, 359–361 sequence specific binding divalent cations, 24–26 general binding trends, 26 monovalent cations, 23–24 structure, 22–23

DNA hairpin sequences, 332–334, 333 DNA hydrolysis ARCUT system, 385, 385 metallic hydrolytic agents bimetallic complexes, 377–380 conjugated with DNA affinity subunits, 380–384 conjugated with sequence-selective elements, 384–385 free ions, 373–374 metal ions, 371, 386 mononuclear complexes, 374–377 reaction pathway, 370, 370–371 research aims, 369–370 research considerations bimetallic systems, development of, 388 DNA affinity, increasing, 388–389 hydrolytic efficiency, 386–388, 387 ideal features of hydrolytic agents, 386 DNA photomodification and photochemical degradation mechanism of photomodification, 257–258 mechanisms of photochemical degradation photobinding DNA through a coordinated ligand, 260 photobinding DNA through a metal centre, 259 photocleavage of DNA, 260–262, 262 DNA polymerases, 179, 418, 418 DNA recognition code, 348, 348–349 DNA repair arsenic-induced inhibition, 494, 497–499 defined, 176 double-stranded breaks, repair of, 182–184, 184, 184–185, 186 implications for design of platinum drugs, 200 repair proteins and antitumour platinum compounds, 189 base excision repair system, 189 high-mobility group proteins, 193 homologous recombination system, 192 mismatch repair system, 191–192 nonhomologous end-joining system, 192

Index nucleotide excision repair system, 189–191 translesion synthesis DNA polymerases, 192–193 single-strand breaks, repair of, 476–477 types base excision repair, 178, 179, 180, 193, 473, 473–474 direct reversal of damage, 177–178 DNA damage bypass (translesion synthesis), 185–188, 187 homologous recombination, 184– 185, 186, 199–200 mismatch repair, 180–182, 183, 198–199, 478–480 nonhomologous end-joining, 199 nucleotide excision repair, 178–180, 181, 193–198, 481–482 recombination repair, 482 translesion synthesis (DNA damage bypass), 185–188, 187 DNA synthesis, 415, 416 DNA-metal interactions. see metal complexDNA interactions; metal ionDNA interactions DNAzymes catalytic activities, 395 cofactors, 396–398 future research needs, 410 as metal ion sensors colorimetric sensors, 407–409, 408, 409 fluorescence-based sensors, 404– 407, 405, 406 other sensors, 409–410 nucleic acid enzyme-metal ion interactions, 399–403 selection using in vitro evolution, 397, 397–399 double-proton transfer (DPT), 31–32, 43–46, 45, 47–48 europium Eu(III) complexes in DNA hydrolysis, 377–378, 381–382, 383–384 G-quadruplexes, role in, 59 excited state light absorbers applications in photodynamic therapy (PDT) excited state electron transfer theory, 252–253

515

excited state energy transfer theory, 254 Type I photooxidation reactions, 253–254 Type II photooxidation reactions, 254–255 groups of bimolecular interactions, 251–252 fluorescent DAPI-displacement assays, 337 gene promoters in G-quadruplex-forming sequences, 64 c-kit promotor sequence, 65, 66 c-myc-1245 sequence, 65, 66 c-myc-2345 sequence, 65, 65–66 c-myc-23456 sequence, 65, 66 glutathione reduction of Cr(VI), 466, 467 role in arsenic-induced cancers, 493, 493, 496 G-quadruplexes cation competition, 75–77 cation coordination and effect on stability, 59–61, 80–81 cation localization, 76 cation locations and coordination geometries, 69–70, 70 cation movement, 78–80 coordination of cations within d[(G4T4G4)2] crystal structures, 73–75 d(G4T4G4), 73–75 NMR studies, 75–78 coordination of cations within d[(TG4T)4], 69–73 d(TG4T), 71, 71–72 inhibition of cell proliferation, 213–215, 215 metallo-organic complexes, 215–216, 216 molecular switches, 61 numbers of bound cations, 77 Oxytricha telomere repeat analogues, 67, 67–69 platinum complexes, interaction with, 217–223, 218 promoter regions, adoption by, 64–66 structure basic, 56, 56–57 in human telomeric DNA, 61–64, 63

516

Index

loops, 58, 58–59 stoichiometry, 57, 57–58 structure of r(UG4U), 72–73 structure of telomeric DNA sequences, 213, 214 four-repeat sequences, 62–64, 63 single repeat sequences, 62 two-repeat sequences, 62 guanine. see also G-quadruplexes cis amine interactions in M-DNA adducts, 149–150 depurination, 46–47 dirhodium complexes, binding with, 303, 305 double-proton transfer, 44–46, 45 G/G crosslinks, orientations in, 140–141, 141 G-H8 resonances affected by Mn2+ adducts, 10–14, 11, 12, 13, 14 interacting sites, strength of, 6 metal ion affinities, 3 metal-DNA supramolecular chemistry 1D coordination polymers, 110– 111, 112, 114, 114, 115, 115 chelates, 124 cyclic hexanuclear platinumguanine derivative, 105–106, 106 H-bonding patterns to metals, 123, 124 square supramolecules, 102, 102– 103, 103 oxidation, 7, 469–473 platinum binding sites, possible, 138, 139–140 point defects, 48 retro models with untethered guanine bases, 143–145 rhodium(III) complexes, binding with, 307 telomeric DNA, 209, 223 highest occupied molecular orbital (HOMO), 6–7, 7 HOMO-LUMO interactions, 6, 127, 239–240, 240, 453, 454 hydrogen peroxide reduction of Cr(VI), 466, 467 intercalation, 320, 322–323 intercalator-metal conjugates, 380–382, 381, 382

ionization potentials (IP), 6, 8, 8 iridium, 100–101 iron Fe(III) complexes in DNA hydrolysis, 378–379 iron complex-peptide conjugate, 353–354 oligodeoxynucleotides adducts, 14–15 preferred coordination geometry of Fe2+ cations, 417 lanthanides, 373–374 lead DNA binding models, 34–35, 35 DNA binding parameters, 35 DNAzyme for Pb(II), 396 DNAzyme-based Pb2+ sensors, 405, 405, 407–410, 408, 409 G-quadruplexes, role in, 59 lesions. see also DNA damage; DNA repair chromium-induced, 469, 469–470, 470, 471, 472, 480 growth, 236 Lewis acidity, 301 ligands. see also nucleobases Cu(II) complexes in DNA hydrolysis, use with, 376 Fe(III) complexes in DNA hydrolysis, use with, 379 G-quadruplexes, 215, 215, 216, 218 guanine cis amine interactions, 149–150 interligand interactions, 151 lanthanide ions in DNA hydrolysis, use with, 374, 375, 377, 378 metal-DNA binding sites, role in determining, 22 metallosupramolecular-peptide conjugates, used in, 353, 353 platinum complexes, 137, 142, 142–145, 144 polyazine, used in PDT, 218, 244 ruthenium(II) probes of DNA, used in, 321–322 ancillary, 324 bridging, 325–329, 326–327, 334 flexible, 335–337, 336 intercalating, 320, 322, 323–324 solid-state structures of dynamic nucleotides, 150–151 lipophilicity, 309, 310

Index lithium G-quadruplexes coordination in d(TG4T), 71 role in, 59 lowest unoccupied molecular orbital (LUMO), 6 HOMO-LUMO interactions, 6, 127, 239–240, 240, 453, 454 magnesium basic properties of Mg2+ cations, 416–417 DNA grooves, occupancy in, 25, 25 DNAzyme for Mg(II), 396 G-quadruplexes coordination in d(TG4T), 72 role in, 59–60 polymerase catalysis Mg2+ requirement, 417–419 ion alignment, 419–420 preferred coordination geometry of Mg2+ ions, 417, 417 manganese DNA grooves, occupancy in, 25 G-H8 resonances in oligodeoxynucleotide adducts, effect on, 10–14, 11, 12, 13, 14 G-quadruplexes, role in, 59–60 Mn(II) complexes in DNA hydrolysis, 375 as relaxation probe, 9 M-DNA, 125–127, 126 M-DNA adducts internucleotide interactions, 145–147, 147, 148 solid-state structures, 150–151 mercury DNAzyme-based Hg2+ sensor, 406 metal-DNA supramolecular chemistry, 107–108, 108, 115 T-HgII-T base pair applications, 454–455 biological relevance, 455–457 crystallographic studies, 440–441, 441, 442 HgII-complex with uracil, 440–441, 441 NMR studies, 444–452, 449, 450, 451, 452 proposed structures, 440 reaction pathway, 453 research history, 439–440

517

types of HgII-T/U complexes, 444 UV, UVCD and vibrational studies, 441–444, 443 metal complex-DNA interactions. see also specific metals covalence index, correlation with, 38, 38 modes of interaction with oligonucleotides, 302 photochemical reactions, 255–257 types of DNA used in experiments, 256, 256–257 metal complex-peptide conjugates. see also specific metals binding, 347–349 bioactive peptides, 354 cleavage, 359–361 compared to metal-peptide complexes, 361 de novo designed peptides, 354–355 dipeptides, 350, 350 DNA-binding affinity, 357–358 nuclease activity on super-coiled plasmid, 360 protein fragments, 355, 355–357 sequence selectivity, 361 tripeptides glycyl-L-histidyl-L lysine (GHK), 350–352, 351, 352 Gly-Gly-Ser-CoNH2, 353, 353–354 ruthenium-peptide complexes, 351–353 metal ion-DNA interactions. see also specific metals binding, 3–4, 4, 5, 32, 38–43 DNA hydrolysis, role in, 371, 371 mechanistic investigations, importance of, 372–373 G-quadruplexes effect on stability, 59–61 locations and geometries, 69–70, 74 movement within, 78–80 role in determining structure, 61 ion reactions with hydrated electrons, 32, 32 metal-induced point defects depurination, 46–47 double proton transfer, 43–46, 45 formation, 47–48 NMR studies with oligonucleotides [Pt(dien)]+ ions, 18–21 Co2+ ions, 15

518

Index

Cu2+ ions, 15–16 Fe2+ ions, 14–15 Mn2+ ions, 10–14 Ni2+ ions, 15 Zn2+ ions, 16–18 sequence selectivity, 4 ab initio calculation studies, 6–9 photo-cleavage studies, 6, 7–9 stability constants, 34–38, 35, 36, 37 thermodynamic adsorption model of complex formation, 33–34 metal-DNA supramolecular chemistry building blocks and structural possibilities, 96–98, 97, 98 discrete architectures larger supramolecules, 105–108, 106, 107, 108 square supramolecules, 100, 102, 102–105, 103, 104, 105 triangular supramolecules, 99–102, 100, 101 H-binding patterns, 122, 122–124, 123 infinite architectures 1D coordination polymers, 109, 109–117, 110, 111, 112, 117 2D coordination polymers, 118, 118–119, 119 3D coordination polymers, 119– 121, 120, 121 potential applications, 108, 109 known structures, 95, 96 M-DNA, 125–127 metal coordination sites and base-base interactions chelates, 124 H-bonding, 121–124 metallo-DNAzymes Cu(II), 396 Mg(II), 396 nucleic acid enzyme-metal ion interactions studies fluorescence resonance energy transfer (FRET), 401–402, 402 nucleic acid enzyme-metal ion interactions, 401 nucleic acid-metal ion interactions, 401 single molecule FRET (smFRET), 402–403, 403 structural studies, 400 Pb(II), 396

metallointercalator-metallopeptide conjugates, 358–359, 359 metallonucleases, artificial, 387 metal-oxalate frameworks, 116, 117 mismatch repair, 180–182, 183, 198–199, 478–480 molecular wires, 113, 125–127 mutations metal-induced point defects depurination, 46–47 double-proton transfer (DPT), 43–46, 45 formation of point defects, 47–48 transition-type, 31, 47–48 transversion-type, 31 nickel chemical shift probe, 9 G-H8 and A-H8 resonances, effect on, 16 oligodeoxynucleotides, adducts with, 15, 16 UV spectra of various DNAs and polynucleotides, influence on, 41, 41–43, 42, 43 NMR studies cation coordination in d[(G4T4G4)2], 75–78 G-quadruplexes cation coordination and effect on stability, 59–61 cation movement, 78–80 coordination of cations within d[(G4T4G4)2], 75–78 human telomeric sequences, structure of, 62–63, 63 loops, 58–59 Oxytricha telomeric sequences, structure of, 67, 67–69 sequence selectivity in DNA-metal binding, 9–21 methodology, 9–10 model systems, 10 oligodeoxynucleotides-transition metal adducts, 10–21 structure of the T-HgII-T base pair 15 N-HgII-15N covalent linking, 450, 450, 451 J-coupling values, 451 nucleoside-metal ion systems, 444–448

Index oligonucleotide-metal systems, 448–449, 449 T-HgII-T base pair, 449–452 nucleases, artificial mechanistic pathways, 361–362 metallonucleases, 387 research aims, 359 nucleic acid enzyme catalysis catalytic centres, types of, 418 Mg2+ ions, requirement for two DNA and RNA polymerases, 417–419 RNase H and MutH, 420–427, 421 one-metal-ion mechanism, 428–429 two-metal-ion mechanism advantages, 428, 429 catalytic specificity enhancement, 423–424 effect of substrate on binding of metal ions, 422–423 ion alignment, 419–420, 424, 427 Mg2+ ions, 417–419, 420–422 movement of two metal ions, 425 prediction, 429–430 separate and joint functions of metal ions, 426–427, 427 types of catalytic centres, 418–419 nucleobases. see also adenine; cytosine; guanine; thymine; uracil as bridging ligands 1D coordination polymers, 109, 110, 110–113 3D coordination polymers, 119– 121, 120 as bridging ligands with anions 1D coordination polymers, 114–117 2D coordination polymers, 118–119 Cr(VI)-induced damage and repair base excision repair (BER) pathway, 473 lesion pathways, 468–473, 469, 470, 471, 472 reduction potentials, 469 formation of chromium-DNA adducts, 477–478, 478 metal-DNA supramolecular chemistry building blocks, 96–97, 97 chelating, 124, 125 self-pairing, 122, 122–124, 123 preferred metal ion binding sites, 3

519

nucleotide excision repair, 178–180, 181, 193–198, 481–482, 497–498 oligonucleotides formation of G-quadruplexes, 57 metal complexes conjugated with, 384 modes of interactions with, 302 NMR studies with metal complexes, 448–449, 449 NMR studies with metal ions [Pt(dien)]+ ions, 18–21 Co2+ ions, 15 Cu2+ ions, 15–16 Fe2+ ions, 14–15 Mn2+ ions, 10–14 Ni2+ ions, 15 Zn2+ ions, 16–18 platinum complexes, interaction with, 160–163, 161, 162 oligonucleotides, platinated crosslinking, 276–278, 277, 278 duplex, 289–294, 290, 291, 293 triplex, 294–295 drug requirements, 276 sulfur-platinum-nitrogen adducts, 286–287, 287 synthesis base-modified oligonucleotides, 286–289, 289 partially protected oligonucleotides, 283–284, 284 platination site, identification of, 288 protected oligonucleotides, 285, 285–286, 286 solid-phase synthesis, 281–283, 282 unmodified oligonucleotides, 279–281, 280 unprotected oligonucleotides, 281, 281 therapeutic applications antisense oligonucleotides, 273– 274, 274 aptamers, 275–276 microRNA targeting, 275 ribozymes, 274–275 RNA interference, 275 triplex-forming oligonucleotides, 275

520

Index

organotin industrial uses, 308 lipophilicity and toxicity, correlation between, 309 organotin complex-DNA interactions, 309–311, 310 toxicity, 308 osmium polyazine complexes used in PDT electronic excitation and unimolecular decay, 242, 243, 247–248, 248 phototoxicity, 265–266 oxaliplatin, 135–136, 136, 155, 155–156, 188 p53 tumour suppressor protein, 496, 498 palladium, 103, 103 paramagnetic relaxation, 9–10 peptides. see metal complex-peptide conjugates; metallointercalatormetallopeptide conjugates peraklyammonium groups, 383–384 pH effect on H-conformer preferences in platinum complexes, 149, 150 influence on organotin-DNA interactions, 310 phosphoryl transfer reactions, 415–416, 416 photo-cleavage studies, 6, 7–8, 8 photodynamic therapy (PDT) cell studies with metal complexes cellular uptake, 263–265 complexes used, 263 phototoxicity, 265, 265–266 clinical treatments, 236–237, 237 definition, 236 designing polyazines for DNA photomodification design considerations, 238, 257 ground state interactions, 257–258 photochemical degradation mechanisms, 258–262 DNA type used, influence of, 256, 256–257 PDT agents, 237–238 research needed, 267 rhodium complexes, 305–306, 306 photophysical processes of interest in PDT research

bimolecular excited state interactions, 251–255 electronic excitation, 239–240, 240, 241 unimolecular electronic excited state decay, 240–242, 242 picoplatin, 136, 136 plasmid DNA, 371–372, 372 platination DNA, effects on, 140 enhancement of biological effects, 276–278, 278 oligonucleotides base-modified, 286–289 duplex crosslinking, 289–294 partially protected, 283–284, 284 protected, 285, 285–286, 286 solid-phase synthesis, 281–283, 282 triplex crosslinking, 294–295 unmodified, 279–281, 280 unprotected, 281, 281 telomeric DNA, 225 platinum. see also cisplatin (and related platinum complexes); platination; platinum drugs; transplatin binding to DNA, 138, 138–140 crosslinking DNA duplexes, use in, 289–294, 290 DNA damage and repair, 176–177 metal-DNA supramolecular chemistry 1D coordination polymers, 109– 110, 110, 111 2D coordination polymers, 118, 118 large structures, 105–107, 106, 107, 108 metallocalixarenes, 104, 105, 105, 105, 105 supramolecular triangles and squares, 100, 101, 101–102 oligodeoxynucleotides, adducts with, 18–21, 19, 20, 21 peptide-tethered complexes, 350, 350 telomerase, interaction with, 223–225 telomeric DNA duplexes, interaction with, 223 platinum drugs. see also cisplatin (and related platinum complexes) activity series, 137 delivery to tumours, 136 design factors to consider, 200 drugs in use, 135–136, 136

Index effects of DNA repair, 177 repair of DNA damage base excision repair, 193 homologous recombination repair, 199–200 mismatch repair, 198–199 nonhomologous end-joining, 199 nucleotide excision repair, 193–198 structure-activity relationships (SARs), 137 types, 135–136, 136 polyazine complexes, transition metal cellular uptake, 263–265 common electronic transitions, 243 complexes used in PDT, 263 electronic excitation and unimolecular decay osmium complexes, 242, 243, 247– 248, 248 rhodium complexes, 242, 248–250, 249, 250 ruthenium complexes, 242, 243– 247, 247, 248 ruthenium-rhodium mixed complexes, 250–251, 251 ligands, 244 photophysical properties, 245–246 toxicity studies, 265–266 potassium G-quadruplexes competition with Na+ , 77 coordination in d[(G4T4G4)]2, 74 coordination in d(G4T4G4), 73–74 human telomeric sequences, 63, 63–64 melting temperatures, 60 residence times, 78 role in, 59, 61 praseodymium, 377 proteins repair of DNA modified by antitumour platinum compounds AAG, 189 base excision repair, 189 DNA-PK, 192 high-mobility group proteins, 193 hMutLα, 191–192 hMutSα, 191–192 homologous recombination repair, 192 mismatch repair, 191–192

521

nonhomologous end-joining, 192 nucleotide excision repair, 189–191 RNA polymerase II, 191 RPA, 190, 191 translesion synthesis DNA polymerases, 192–193 XPA, 190 XPC-HR23B, 190–191 in telomeres, 210–211, 211 proton tunneling, 44–46, 45 purines metal-DNA supramolecular chemistry cyclic trimeric complexes, 100–101 large structures, 105–106 molecular squares, 100, 102, 102– 103, 104 molecular triangles, 99–101, 100, 101 tetrameric complexes, 100–101 structure, 97 pyrimidines metal-DNA supramolecular chemistry large structures, 106–108 metallocalixarenes, 103–105 molecular triangles, 101–102 structure, 97 reactive nitrogen species, 495–496 reactive oxygen species, 495–496 recombination repair, 482 retro modelling cisplatin analogues, 141–143, 142 untethered guanine bases, 143, 143–145, 144 rhodium complexes used in PDT, 305–306, 306 dirhodium (II) complexes, 303–305, 304 intercalator-metal conjugates, 381, 381 metal-DNA supramolecular chemistry, 100–101, 113, 114 polyazine complexes used in PDT cellular uptake, 264 DNA photochemical degradation, 259–260 electronic excitation and unimolecular decay, 242, 248– 251, 249, 250, 251 phototoxicity, 265, 265–266 rhodium (I) complexes, 302–303 rhodium (III) complexes, 306–307

522

Index

rhodium complex-peptide conjugate, 354, 358–359, 359 site-specific rhodium-DNA interactions, 307, 307–308 ribozymes, 274–275, 400 RNA dinuclear ruthenium, interaction with, 334–335 Okazaki fragments, 209–210, 210 RNA cleavage, 415–416, 416, 426–427, 427 RNA interference, 275 RNA polymerases, 418, 418–419 RNA repair, 177 RNA synthesis, 415, 416 rubidium, 23, 23, 59 ruthenium advantages of, as DNA probe, 319–320 dinuclear complexes as DNA probes advantages over mononuclear species, 325 affinity chromatography, 337–338 applications, potential, 338–339 binding behavior, 325–329 DNA bulge sequences, interaction with, 330–332, 331, 333 DNA hairpin sequences, interaction with, 332–334 duplex DNA, interaction with, 329, 329–330 flexible ligands, 335–337, 336 fluorescent DAPI-displacement assays, 337 research needs, 338 RNA sequences, interaction with, 334–335 S-O-S bridge, 325, 327 stereoisomers, 329 intercalator-metal conjugates, 381, 381 metal-DNA supramolecular chemistry, 100–101, 101 mononuclear complexes as DNA probes binding site and orientation, 322–324 enantioselectivity, 320 influence of nature of ligands on binding, 323–324 intercalation, 320, 322 polyazine complexes used in PDT cellular uptake, 264

electronic excitation and unimolecular decay, 242, 243– 247, 247, 248, 250–251, 251 photobinding of DNA, 259, 260 phototoxicity, 265, 265–266 ruthenium complex-peptide conjugate de novo designed peptides, 354–355 protein fragments, 355–357 satraplatin, 136, 136 shelterin (telosome), 210–211, 211 silver metal-DNA supramolecular chemistry 1D coordination polymers, 109– 110, 110, 111 2D coordination polymers, 118– 119, 119 3D coordination polymers, 120– 121, 121 sodium DNA grooves, occupancy in, 24 effect on stability constants of metal ions, 36, 36–37 G-quadruplexes competition with K+, 77 competition with NH4+, 75–76 coordination in d[(G4T4G4)]2, 74 coordination in d[(TG4T)4], 70 coordination in d(G4T4G4), 73 coordination in d(TG4T), 71, 71, 72 localization, 70, 76 melting temperatures, 60 residence times, 78–79 role in, 59, 61 spine of hydration, 23, 24 stability constants of metal-DNA complexes covalence index, correlation with, 38, 38 determination, 34–36, 35, 36 DNA G-C content, effect of, 37, 37–38 ionic strength, effects of, 36, 36–37 strontium G-quadruplexes coordination in r(UG4U), 72–73 role in, 59, 60 sugar oxidation, Cr(VI)-induced, 474–477, 475, 476 tautomeric transformations, 44, 47 telomerase, 212, 212–213, 223–225 telomeres, 209–211, 210

Index telomeric DNA G-quadruplexes, 213, 214 human, 61–64, 63 Oxytricha, 67, 67–69 platinum complexes, interaction with, 223, 225 telosome (shelterin), 210–211, 211 thallium G-quadruplexes coordination in d[(G4T4G4)]2, 74 coordination in d(G4T4G4), 75 coordination in d(TG4T), 71, 72 localization, 76–77 residence times, 78 role in, 59 thermodynamic adsorption model of complex formation, 33–34 thymine loop coordination in G-quartets d(G3T4G3), 68 d(G3T4G3C), 61 d(G3T4G4), 69 d(G4T4G3), 68–69 d(G4T4G4), 67–68, 73–75 metal ion affinities, 3 metal-DNA supramolecular chemistry 1D coordination polymers, 109 chelates, 124, 125 H-bonding patterns to metals, 122 large supramolecules, 106–107, 107 platinum binding sites, possible, 138 thymine-thymine base pair, HgII-mediated applications, 454–455 biological relevance, 455–457 structural studies crystallographic studies, 440–441, 441, 442 HgII-complex with uracil, 440–441, 441 NMR studies, 444–452, 449, 450, 451, 452 proposed structures, 440 reaction pathway, 453 research history, 439–440 types of HgII-T/U complexes, 444 UV, UVCD and vibrational (IR/Raman) spectral studies, 441–444, 443

523

tin organotin industrial uses, 308 lipophilicity and toxicity, correlation between, 309 organotin complexes-DNA interactions, 309–311, 310 toxicity, 308 transplatin, 136, 136, 140 platinated oligonucleotides crosslinking experiments, 279–280, 280, 292–295, 293 isomerization reaction, 290–291, 291 regioselective platination, 286–288 uracil HgII-complex, 440–441, 441 metal-DNA supramolecular chemistry 1D coordination polymers, 109, 109–110 chelates, 124 H-bonding patterns to metals, 122, 124 large supramolecules, 106–107, 107 square supramolecules, 105, 105 uranium, 405–406, 406 UV studies HgII-nucleobase complexes, 441 study of metal-DNA interactions bathochromic shift, 39, 39–41, 40 double-proton transfer (DPT) measurements, 43–46 influence of GC composition, 41–43 influence of Ni2+ on spectra of DNAs and polymerases, 41, 41–43, 42, 43 vibrational studies with HgII-nucleobase complexes, 441–444, 443 X-ray crystallography studies cation-groove interaction studies, 22 G-quadruplexes coordination of cations within d[(G4T4G4)2], 73–75 coordination of cations within d[(TG4T)4], 69–73 metal ion–ribozyme complexes, 400

524

Index

zinc DNA cleavage, 358 G-quadruplexes, role in, 59–60 oligodeoxynucleotides, adducts with, 16–18 chemical shift variation for G-H8 resonances, 16–17, 17 effect of ZnCl2 concentrations, 17 preferred coordination geometry of Zn2+ cations, 417, 417

zinc-finger peptides in DNA hydrolysis, 384 Zn(II) complexes in DNA hydrolysis, 374–375, 375, 379–380, 380, 383 intercalator-metal conjugates, 381, 382, 383 Zn(II) complexes in DNA oxidative cleavage, 373, 373 zirconium, 384–385

Metal Complex - DNA Interactions

DNA-Protein Interactions: Principles and Protocols

DNA INTERACTIONS WITH POLYMERS AND SURFACTANTS

Drug-DNA Interactions: Structures and Spectra

Metal-DNA Chemistry (Acs Symposium Series)

DNA-Protein Interactions. Principles and Protocols

DNA-Protein Interactions Principles and Protocols

Macromolecules Containing Metal and Metal-Like Elements, Nanoscale Interactions of Metal-Containing Polymers

DNA

DNA Computing - DNA 9

DNA

DNA

DNA

Bio-Geo Interactions in Metal-Contaminated Soils (Soil Biology, 31)

Modern Supramolecular Gold Chemistry: Gold-Metal Interactions and Applications

Methods in Enzymology Vol 208: Protein-DNA Interactions

Nucleic Acid Metal Ion Interactions (RSC Biomolecular Sciences)

Metal

Metal Complex Catalysts Supercritical Fluid Polymerization Supramolecular Architecture

Reagents for Transition Metal Complex and Organometallic Syntheses, Inorganic Syntheses

Macromolecules Containing Metal and Metal-Like Elements, Metal-Coordination Polymers

DNA Repair

DNA Prospector

Mitochondrial DNA

DNA Topology

Recombinant DNA

Dna Prospector

DNA Computing

Junk DNA

DNA Repair

DNA Topology

Metal Complex - DNA Interactions

DNA-Protein Interactions: Principles and Protocols

DNA INTERACTIONS WITH POLYMERS AND SURFACTANTS

Drug-DNA Interactions: Structures and Spectra

Metal-DNA Chemistry (Acs Symposium Series)

DNA-Protein Interactions. Principles and Protocols

DNA-Protein Interactions Principles and Protocols

Macromolecules Containing Metal and Metal-Like Elements, Nanoscale Interactions of Metal-Containing Polymers

DNA

DNA Computing - DNA 9

DNA

DNA

DNA

Bio-Geo Interactions in Metal-Contaminated Soils (Soil Biology, 31)

Modern Supramolecular Gold Chemistry: Gold-Metal Interactions and Applications

Methods in Enzymology Vol 208: Protein-DNA Interactions

Nucleic Acid Metal Ion Interactions (RSC Biomolecular Sciences)

Metal

Metal Complex Catalysts Supercritical Fluid Polymerization Supramolecular Architecture

Reagents for Transition Metal Complex and Organometallic Syntheses, Inorganic Syntheses

Macromolecules Containing Metal and Metal-Like Elements, Metal-Coordination Polymers

DNA Repair

DNA Prospector

Mitochondrial DNA

DNA Topology

Recombinant DNA

Dna Prospector

DNA Computing

Junk DNA

DNA Repair

DNA Topology

Recommend Documents