Metal-based anticancer agents 9781788014069, 1788014065, 9781788016452

Citation preview

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP001

Metal-­based Anticancer Agents

View Online

Metallobiology Series Editor-­in-­chief: Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP001

C. David Garner, University of Nottingham, UK

Series editors:

Stefano L. Ciurli, University of Bologna, Italy Julie Kovacs, University of Washington, USA Emma Raven, University of Bristol, UK Hongzhe Sun, University of Hong Kong, China Anthony Wedd, University of Melbourne, Australia

Titles in the Series:

1: Mechanisms and Metal Involvement in Neurodegenerative Diseases 2: Binding, Transport and Storage of Metal Ions in Biological Cells 3: 2-­Oxoglutarate-­dependent Oxygenases 4: Heme Peroxidases 5: Molybdenum and Tungsten Enzymes: Biochemistry 6: Molybdenum and Tungsten Enzymes: Bioinorganic Chemistry 7: Molybdenum and Tungsten Enzymes: Spectroscopic and Theoretical Investigations 8: Metal Chelation in Medicine 9: Metalloenzymes in Denitrification: Applications and Environmental Impacts 10: The Biological Chemistry of Nickel 11: Gas Sensing in Cells 12: Gasotransmitters 13: Dioxygen-­dependent Heme Enzymes 14: Metal-­based Anticancer Agents

How to obtain future titles on publication:

A standing order plan is available for this series. A standing order will bring delivery of each new volume immediately on publication.

For further information please contact:

Book Sales Department, Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge, CB4 0WF, UK Telephone: +44 (0)1223 420066, Fax: +44 (0)1223 420247, Email: [email protected] Visit our website at www.rsc.org/books

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP001

Metal-­based Anticancer Agents Edited by

Angela Casini

Cardiff University, UK Email: [email protected]

Anne Vessières

Pierre and Marie Curie University, France Email: [email protected] and

Samuel M. Meier-­Menches

University of Vienna, Austria Email: [email protected]

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP001

View Online

Metallobiology Series No. 14 Print ISBN: 978-­1-­78801-­406-­9 PDF ISBN: 978-­1-­78801-­645-­2 EPUB ISBN: 978-­1-­78801-­767-­1 Print ISSN: 2045-­547X Electronic ISSN: 2045-­5488 A catalogue record for this book is available from the British Library © The Royal Society of Chemistry 2019 All rights reserved Apart from fair dealing for the purposes of research for non-­commercial purposes or for private study, criticism or review, as permitted under the Copyright, Designs and Patents Act 1988 and the Copyright and Related Rights Regulations 2003, this publication may not be reproduced, stored or transmitted, in any form or by any means, without the prior permission in writing of The Royal Society of Chemistry or the copyright owner, or in the case of reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK, or in accordance with the terms of the licences issued by the appropriate Reproduction Rights Organization outside the UK. Enquiries concerning reproduction outside the terms stated here should be sent to The Royal Society of ­ Chemistry at the address printed on this page. Whilst this material has been produced with all due care, The Royal Society of Chemistry cannot be held responsible or liable for its accuracy and completeness, nor for any ­ consequences arising from any errors or the use of the information contained in this ­ publication. The publication of advertisements does not constitute any endorsement by The Royal Society of Chemistry or Authors of any products advertised. The views and opinions advanced by contributors do not necessarily reflect those of The Royal Society of Chemistry which shall not be liable for any resulting loss or damage arising as a result of reliance upon this material. The Royal Society of Chemistry is a charity, registered in England and Wales, Number 207890, and a company incorporated in England by Royal Charter (Registered No. RC000524), registered office: Burlington House, Piccadilly, London W1J 0BA, UK, Telephone: +44 (0) 20 7437 8656. For further information see our website at www.rsc.org Printed in the United Kingdom by CPI Group (UK) Ltd, Croydon, CR0 4YY, UK

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP005

Preface Since one of the first reports by Collier and Kraus1 and subsequently by Rosenberg et al.2, metal-­based anticancer agents continue to intrigue us by the myth that surrounds them. Especially, three platinums and one arsenic compound, together with several radiopharmaceuticals, are tremendously successful in the clinical setting and are frequently applied in anticancer therapy. These drugs are essentially able to cure patients suffering from different types of cancer, among which are testicular cancer and acute promyelocytic leukaemia. Despite their success and obvious relevance to oncology, metal-­based anticancer drug discovery programs remain purely academic. This is a phenomenon that can be explained by the perceived risks and challenges associated with them. Very often, metal-­based anticancer agents are reactive molecules that undergo ligand exchange and redox reactions, as well as complex speciation mechanisms. This reactivity represents a double-­edged sword: on the one hand, it contributes to their prodrug character and unprecedented clinical effects, but on the other, it is difficult to control. This leads to a widespread and often categorical preconception of metal-­based anticancer agents being toxic and having unspecific modes of action. Nowadays, novel drug design strategies and holistic analytical approaches warrant a rethinking of these preconceptions, demonstrating that the biological effects of metal complexes are not only due to the metal ion but also to the compound's overall structure and types of ligands. Several unresolved research questions are being intensively investigated in the field, including: (i) relating phenotypic observations to the metallodrug's chemical properties and molecular mechanisms of action; (ii) identifying molecular targets; (iii) tuning the reactivity and developing speciation methods; (iv) efficiently delivering the reactive payloads; (v) identifying reliable

  Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

v

View Online

Preface

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP005

vi

models for screening of anticancer activity in vivo; and, finally, (vi) convincing investors of the promise of novel metal-­based anticancer agents, to fund their clinical development. These points are also illustrative of the interdisciplinary nature of metal-­ based drug discovery, which necessitates combining knowledge across disciplines, from the inorganic and organometallic synthesis of metal-­based anticancer agents and possible radioisotope labelling, to their bioanalytical and pharmacological/toxicological study, ultimately to establish and manage clinical trials. Thus, this book summarizes the recent developments in the field of metal-­based compounds for biomedical applications in cancer treatment and diagnosis, while putting them in the context of the above-­ mentioned challenges. The book is organized into three sections. The first covers the main classes of metal-­based anticancer agents and is divided into eight chapters, including state-­of-­the-­art knowledge about their respective pharmacological activities and modes of action. The second section highlights emerging concepts in metallodrug discovery, essentially discussing both novel anticancer drug delivery and therapeutic strategies, relying on supramolecular metal-­based structures, and enabling analytical approaches for target identification. The third section is devoted to preclinical and clinical considerations for both metal-­based radiopharmaceuticals and anticancer agents, including their potential impact on cancer immune recognition. In summary, this book is a useful tool for experienced researchers, but also for people new to the field of metal-­based anticancer agents, such as undergraduate or postgraduate students. The book also accounts for the inter-­and cross-­disciplinary character of this exciting research area in order to establish a common language among researchers of different backgrounds in chemical sciences, biology and medicine. We would like to thank all the contributing authors for their valuable efforts in tailoring each chapter to reach out to a broad audience and in providing a critical discussion of the various topics. Additionally, we are grateful to Katie Morrey and Drew Gwilliams of the Royal Society of Chemistry for their excellent editorial assistance throughout the entire process, as well as to Prof. Dave Garner (Editor-­in-­chief, Metallobiology Series) who inspired this project. Finally, we would like to acknowledge all the colleagues in the field for sharing our passion for metal-­based anticancer agents and for steadily advancing our knowledge of and insight into these fascinating and medically useful molecules. We keep in mind that the ultimate goal of our efforts is to cure diseases and improve the quality of life for cancer patients. Angela Casini, Anne Vessières and Samuel M. Meier-­Menches

References 1. W. A. Collier and F. Krauss, Z. Krebsforsch., 1931, 34, 526–530. 2. B. Rosenberg, L. Vancamp, J. E. Trosko and V. H. Mansour, Nature, 1965, 205, 698–699.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007

Contents Part I: The Main Classes of Metal-based Anticancer Agents and their Modes of Action Chapter 1 Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents by Incorporating Clinically Approved Drugs as Ligands  Reece G. Kenny and Celine J. Marmion

1.1 Introduction  1.2 Platinum Complexes Incorporating Clinically Approved Drugs or Derivatives Thereof as Ligands  1.2.1 Vorinostat and Belinostat Derivatives as Ligands  1.2.2 Valproic Acid and Phenylbutyric Acid as Ligands  1.2.3 Indomethacin, Ibuprofen and Aspirin as Ligands  1.2.4 Ethacrynic Acid as a Ligand  1.2.5 Dichloroacetate or Dichloroacetate Derivatives as Ligands  1.3 Platinum Complexes Incorporating More Than One Clinically Approved Drug as a Ligand  1.4 Conclusions  Abbreviations  Acknowledgements  References 

  Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

vii

3 3 6 6 9 15 18 19 22 24 25 25 26

View Online

Contents

viii

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007

Chapter 2 Ruthenium, Osmium and Iridium in the Fight Against Cancer  Isolda Romero-­Canelón

2.1 Introduction  2.2 Structural Diversity  2.2.1 Octahedral Coordination Complexes  2.2.2 Organometallic Arene ‘Piano-­stool’ Complexes  2.2.3 Other Relevant Coordination Spheres  2.3 Mechanisms of Action of Ru, Os and Ir Metal Complexes  2.3.1 DNA as a Target  2.3.2 Redox Modulation and Mitochondrial Targeting  2.4 Challenges in the Investigations of Mechanisms of Action at the Cellular Level  2.5 Is There a Bright Future for Ruthenium, Osmium and Iridium Complexes in the Fight Against Cancer?  Abbreviations  Acknowledgements  References  Chapter 3 Iron Compounds as Anticancer Agents  Anne Vessieres



3.1 Introduction  3.2 Study of Ferrocene Complexes  3.2.1 Background  3.2.2 The Ferrocifen Family  3.2.3 Ferrocene Complexes of Natural Products  3.2.4 Ferrocenyl Complexes of Histone Deacetylase Inhibitors (HDACi)  3.2.5 Ferrocenyl Derivatives of Nucleosides  3.2.6 N-­Alkylaminoferrocenes  3.2.7 Ferrocenyl Alkylpyridinium Cations Used for Photodynamic Therapy (PDT)  3.3 Coordination Complexes of Iron(ii) and Iron(iii)  3.4 Molecules Active via Chelation with Iron  3.4.1 Bleomycins (BLMs)  3.4.2 Iron Chelators  3.5 Conclusion  Abbreviations  Acknowledgements  References 

31 31 32 32 34 40 40 41 42 52 53 54 55 55 62 62 65 65 67 69 72 73 74 75 76 78 78 80 82 83 83 83

View Online

Contents

ix

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007

Chapter 4 Recent Advances in Anticancer Copper Compounds  Andrew Kellett, Zara Molphy, Vickie McKee and Creina Slator

4.1 Introduction—Copper Complexes as Redox-­active Cytotoxins  4.2 Copper Enzymes and Transport Proteins: Pathways for Developing Redox-­active Therapeutics  4.3 NCI-­60 Screening of Anticancer Copper Complexes  4.4 Mechanistic Analysis of Cytotoxic Copper Complexes  4.4.1 An Overview of Cell Death Mechanisms  4.4.2 Copper-­mediated ROS Production  4.4.3 Mitochondrial Toxicity  4.4.4 DNA-­targeted Copper Complexes  4.4.5 Oxidative DNA Damage  4.5 Summary and Outlook  Abbreviations  Acknowledgements  References  Chapter 5 Anticancer Gold Compounds  Di Hu, Chun-Nam Lok and Chi-­Ming Che



5.1 The Development of Gold Compounds in Medicine  5.2 Anticancer Gold(i) Complexes  5.2.1 Antiarthritic Gold(i) Drugs with Anticancer Activities  5.2.2 Gold(i)–Phosphane Complexes  5.2.3 Gold(i)–Thiourea Complexes  5.2.4 Gold(i)–NHC Complexes  5.2.5 Gold(i)–Alkynyl Complexes  5.3 Anticancer Gold(iii) Complexes  5.3.1 Gold(iii) Porphyrin  5.3.2 Coordination Gold(iii) Complexes with Various Ligands  5.3.3 Cyclometallated Gold(iii) Complexes  5.4 Nano Formulation of Gold Complexes with Improved Anticancer Potency  5.5 Conclusions and Outlook  Abbreviations  Acknowledgements  References 

91

91 92 97 101 101 104 106 107 111 113 113 114 115 120

120 121 121 122 123 124 126 127 127 129 131 136 137 138 139 139

View Online

Contents

x

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007

Chapter 6 Heterometallic Complexes as Anticancer Agents  Natalia Curado and Maria Contel

6.1 Introduction  6.2 Heterometallic Compounds Containing Ferrocenyl-­derived Molecules  6.3 Heterometallic Compounds Containing Ruthenium(ii)–Arene Fragments  6.3.1 Ruthenium–Gold Compounds  6.3.2 Ruthenium–Platinum Compounds  6.3.3 Ruthenium–Cobalt Compounds  6.4 Heterometallic Compounds Containing Titanocenes  6.4.1 Titanocene–Ruthenium Compounds  6.4.2 Titanocene–Gold Compounds  6.5 Other Heterometallic Compounds  6.5.1 Gold-­containing Compounds (Gold– Platinum, Gold–Ruthenium, Gold–Cobalt, Gold–Silver, and Gold–Copper)  6.5.2 Ruthenium-­containing Compounds (Ruthenium–Platinum, Ruthenium–Nickel, and Ruthenium–Copper)  6.5.3 Theranostic Compounds (Ruthenium–M, M = Gold, Osmium, Rhodium, Gadolinium; Gadolinium–Platinum; and Rhenium–Gold)  6.5.4 Other (Cobalt-­based and Copper–Zinc)  6.6 Conclusion  Acknowledgement  References  Chapter 7 Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects  Debbie C. Crans, Noah E. Barkley, Liliana Montezinho and M. Margarida Castro



7.1 Introduction  7.2 Phosphorylation in Tumorigenesis and Signaling Pathways  7.3 Inhibition of Phosphorylases and Kinases by Vanadate and Vanadium-­containing Compounds  7.3.1 Phosphatases and Their Inhibition by Vanadate and Vanadium Compounds  7.3.2 Kinases and Their Inhibition by Vanadate and Vanadium Compounds 

143 143 144 148 149 151 152 153 153 153 157 157 159 160 162 162 163 163 169

169 172 178 178 180

View Online

Contents



Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007



xi

7.3.3 Additional Phosphorylases and Their Inhibition by Vanadate and Vanadium Compounds  7.3.4 Mitogen-­activated Protein Kinase and Phosphoinsitide 3-­Kinase Signaling Pathways  7.4 The Effects of Vanadate and VCs in Cellular Systems  7.5 In Vivo Studies of Vanadium Compounds in Animal Model Systems  7.6 Conclusions  Acknowledgements  References  Chapter 8 Arsenic-­based Anticancer Agents  Stéphane Gibaud



8.1 Introduction and General Overview of Arsenic Anticancer Drugs  8.2 Mechanism of Action  8.2.1 Transport Across Biomembranes  8.2.2 Reactivity with Thiols  8.2.3 Biological Effects  8.3 Arsenic in the Treatment of Acute Promyelocytic Leukemia  8.4 Arsenical Drugs and Glioma  8.5 Conclusion  Abbreviations  References 

180 181 184 187 188 188 188 196

196 201 201 201 202 207 207 208 209 210

Part II. Enabling Concepts in Metallodrug Discovery Chapter 9 Supramolecular Metal-­based Structures for Applications in Cancer Therapy  Margot N. Wenzel, Benjamin Woods and Angela Casini

9.1 Introduction  9.2 Supramolecular Coordination Complexes  9.2.1 Synthesis of 2D (Metallacycles) and 3D (Metallacages) SCCs  9.2.2 Synthesis of Heteroleptic, Interlocked and Heterometallic Cages  9.2.3 Synthesis of Helicates 

217 217 218 218 220 223

View Online

Contents

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007

xii



9.3 SCCs as Anticancer Agents  9.3.1 Cytotoxic PdII and PtII SCCs  9.3.2 Cytotoxic Ruthenium(ii)–Arene SCCs  9.3.3 DNA-­targeted SCCs  9.4 SCCs as Drug Delivery Systems  9.4.1 SCCs as Drug Delivery Systems for Anticancer Agents  9.4.2 Prodrug-­based SCCs  9.5 In Vivo Studies on Anticancer SCCs  9.6 Conclusions and Perspectives  Abbreviations  Acknowledgements  References 

Chapter 10 Enabling Methods to Elucidate the Effects of Metal-­based Anticancer Agents  D. Kreutz, C. Gerner and S. M. Meier-­Menches

10.1 Introduction  10.2 Assessing Activation Mechanisms of Metal-­based Anticancer Agents  10.2.1 DI-­ESI-­MS  10.2.2 LC-­ESI-­MS  10.2.3 CZE-­ESI-­MS  10.3 Identifying Targets of Metal-­based Anticancer Agents  10.4 Elucidating Modes of Action of Metal-­based Anticancer Agents  10.4.1 Transcriptional Profiling  10.4.2 Proteome Profiling  10.5 Conclusion  Abbreviations  Acknowledgements  References 

224 225 226 227 229 230 234 235 237 239 239 240 246 246 248 249 252 253 254 259 259 260 263 264 265 265

Part III. Preclinical and Clinical Evaluation Chapter 11 Metal-­based Radiotherapeutics  Christian A. Mason, Lukas M. Carter and Jason S. Lewis

273



273 274 280



11.1 Introduction  11.1.1 The Radiotherapeutic Armamentarium  11.1.2 Radiobiologic Comparisons  11.1.3 Selection Criteria for Therapeutic Radionuclides  11.1.4 The Role of Imaging in the Application of Endoradiotherapy 

282 286

View Online

Contents



Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-FP007



xiii

11.2 Advances in Preclinical Radiopharmaceutical Development  11.2.1 Preclinical Developments in Beta Therapies  11.2.2 Preclinical Developments in Alpha Therapies  11.2.3 Preclinical Developments in Auger Electron and Conversion Electron Therapies  11.3 Advances in Clinical Radiopharmaceuticals  11.3.1 Beta Therapies in the Clinic  11.3.2 Alpha Therapies in the Clinic  11.4 Future Perspectives  Abbreviations  Acknowledgements  References 

Chapter 12 Challenges and Chances in the Preclinical to Clinical Translation of Anticancer Metallodrugs  Isabella Pötsch, Dina Baier, Bernhard K. Keppler and Walter Berger

12.1 Introduction  12.2 Strategies and Challenges in the Clinical Development of Novel (Metal) Drugs in Oncology  12.2.1 Factors Causing Failure of Anticancer (Metal) Drug Development  12.2.2 Anticancer Metal Drugs Need a Defined Target  12.3 Strategies for Clinical Development of Novel Anticancer Metal Drugs  12.4 Current Status of Novel Anticancer Metal Drugs in Clinical Evaluation  12.4.1 Platinum  12.4.2 Ruthenium  12.4.3 Gold  12.4.4 Gallium  12.4.5 Other Metals  12.5 Metal Complexes as Immunological Drugs and Possible Partners for Immunotherapy  12.5.1 Metal Drugs and Impact on Cancer Immune Recognition  12.5.2 Clinical Situation and Approvals for Combinations of Metal Drugs with Checkpoint Inhibitors  12.6 Conclusion and Outlook  References  Subject Index 

287 287 291 294 295 295 297 300 301 301 302 308

308 313 313 318 320 322 322 326 328 329 332 334 334 335 336 336 348

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Part I

The Main Classes of Metal-based Anticancer Agents and their Modes of Action

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

View Online

           

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

View Online

Chapter 1

Enhancing the Therapeutic Potential of Platinum-­ based Anticancer Agents by Incorporating Clinically Approved Drugs as Ligands Reece G. Kenny and Celine J. Marmion* Centre for Synthesis and Chemical Biology, Department of Chemistry, Royal College of Surgeons in Ireland, 123 St. Stephen's Green, Dublin 2, Ireland *E-­mail: [email protected]

1.1 Introduction Cancer is a multi-­factorial disease which results from a myriad of genetic and environmental factors. It continues to represent a global health challenge not least because the global population is growing and ageing but also because access to information related to its prevention, early detection and treatment options, particularly in developing countries, is limited. The lack of provision of adequate medical and public health infrastructure are also contributory factors.1 While there are different options to treat cancer, they typically include a combination of surgery, chemotherapy and/ or radiotherapy. Immunotherapy, a form of treatment which stimulates

  Metallobiology Series No. 14 Metal-­based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-­Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

3

View Online

Chapter 1

4

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

2

our immune system to kill cancer cells, and oncolytic virotherapies,3 are also emerging as promising options to complement existing therapeutic regimens. Chemotherapeutic drugs operate either alone or in combination regimens by blocking the unwanted proliferation of cancer cells. Platinum drugs constitute a major chemotherapeutic drug class for the treatment of cancer. In fact, it has been stated that nearly 50% of all cancer treatments involve platinum drugs.4 Despite their clinical success, there are currently only three platinum drugs in worldwide clinical use, namely cisplatin (1), carboplatin (2) and oxaliplatin (3) and three others, nedaplatin (4), heptaplatin (5) and lobaplatin (6) in use in Japan, South Korea and China respectively (see Figure 1.1).5 These square planar platinum(ii) drugs contain ‘non-­leaving’ nitrogen donor ligands and labile chlorido or dicarboxylato ligands. Cisplatin, as a representative example, elicits its anticancer effect by first accumulating in tumour cells, whereupon the labile ligands are displaced by water ligands. It is these resulting aquated platinum(ii) species that can then irreversibly bind DNA (typically to the N7 of guanine nucleobases) leading to the formation of DNA lesions and ultimately triggering apoptosis or programmed tumour cell death. The reader is directed to a recent review which provides a comprehensive account into the development, mechanism of action and clinical utility of these ‘classical’ platinum(ii) drugs.5 It is broadly acknowledged that these drugs are enormously successful against a wide range of cancers. They do however have significant limitations. While the lability of the anionic ‘leaving’ ligands is important for anticancer activity as mentioned earlier, it also means that the complexes are more susceptible to ligand substitution. Cisplatin, for example, can readily react with thiol-­containing biomolecules such as glutathione and metallothioneins. These side reactions not only account for the toxicity profile associated with cisplatin and related complexes, from minor to dose-­limiting, but also play a role in the emergence of resistance against

Figure 1.1 Platinum drugs in clinical use (1–6) with the years in which they received regulatory approval indicated in parentheses.

View Online

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

6

5

platinum drugs. For example, some cancer cells are known to overexpress thiol-­containing biomolecules so even if the platinum drug accumulates successfully in tumour cells, those that contain high concentrations of these thiol-­containing biomolecules compete with DNA for platinum drug binding. This leads to a reduction or elimination in the efficacy of the drug towards those cells. Of the platinum drugs in clinical use, all but cisplatin contain bidentate dicarboxylato ligands. These chelating bidentate ‘leaving’ ligands were chosen on the basis that they would confer greater stability on the complexes (chelate effect), enhancing their half-­lives and thus reducing their susceptibility to these unwanted side reactions. The more stable carboplatin, for example, is significantly less toxic on the body than cisplatin.5 In addition to limiting the formation of platinum-­DNA adducts, cells have also acquired other modes in which to build resistance and reduce drug sensitivity to cancer cells. Such resistance can arise from multiple factors including genetic and epigenetic changes as outlined in a comprehensive review by Gottesman et al.7 The reader is also referred to another relevant review on molecular mechanisms underpinning platinum drug resistance in ovarian cancer by Tapia et al.6 In order to overcome toxicity and resistance issues associated with these ‘classical’ platinum drugs, a number of research approaches have been and continue to be actively explored. Early research focused on developing analogues of cisplatin, i.e. complexes in the cis configuration containing a central platinum(ii) ion, with ammine or substituted amine non-­leaving group(s) and anionic labile ligands but none were found to offer any significant advantage over cisplatin.8 Trans analogues9,10 and polynuclear platinum(ii) drugs11,12 have also been investigated with one trinuclear complex, BBR3464,13 advancing to phase II clinical trials. This trinuclear complex represented the first example of a ‘non-­classical’ platinum drug to advance to the clinical setting. This breakthrough undoubtedly prompted scientists working in this field to explore more ‘non-­traditional’ approaches in their drug design strategies. There now appears to be a tangible shift in research focus which is resulting in a more ‘targeted’ approach in the quest to bring forward an alternative ‘classical’ or ‘non-­classical’ drug class. The exploitation of nanotechnologies to selectively deliver platinum drugs to tumour cells is one such approach, which has received considerable attention of late.5,14,15 An alternative approach is to tether targeting moieties on the platinum(ii) scaffold to generate complexes with greater selectivity for tumour cells and this may reduce unwanted side reactions and thus lower toxic side effects.5 Developing complexes with a mechanism of action different to classical platinum(ii) drugs is another area being actively explored. These complexes may be able to overcome drug resistance issues that have been plaguing many therapeutic regimens including those related to platinum drugs. Alternatively, targeting drug transport mechanisms may lead to new complexes with an improved chemotherapeutic profile.16 Another design strategy is to move from square planar platinum(ii) to the more kinetically inert, octahedral platinum(iv) complexes. Platinum(iv) complexes may offer

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

6

Chapter 1

advantages over platinum(ii) drugs in that the coordinatively saturated platinum(iv) centre is more resistant to ligand substitution. This may prevent the aforementioned unwanted side reactions with biomolecules and may thus increase the safety profile of these drugs. Furthermore, the presence of six coordination sites in octahedral platinum(iv) complexes, as opposed to four in square planar platinum(ii) complexes, allows for greater latitude in terms of adding extra functionality. For example, it is possible to incorporate bioactive ligands or targeting vectors in these axial positions. Upon uptake into the more reducing environment found in tumour cells, the bioactive ligands may be released following reduction of the platinum(iv) complex to its platinum(ii) analogue, a mechanism referred to as ‘activation by reduction’, and these ligands may then work synergistically with the resulting platinum(ii) agent in killing cancer cells.5,17,18 Furthermore, given the rising costs associated with drug development, in addition to the growing timeframe involved in getting a drug ‘from bench to shelf’, academia and industry are also focusing on ways in which to ‘repurpose’ existing drugs – this could potentially lead to an accelerated route for drug discovery.19 In this context and in this chapter, we have endeavoured to showcase how clinically approved drugs or derivatives thereof may be exploited as potential ligands and how their corresponding platinum(ii) and platinum(iv) complexes may form the basis of a new drug class which may offer advantages over existing therapeutic regimens. Interestingly, while there is a sound rationale behind repurposing existing drugs, those to date that have been tethered to platinum have not been chosen for this purpose. Rather, they have been selected because they, in their own right, have exhibited anticancer activities albeit some are in clinical use for other indications.

1.2 Platinum Complexes Incorporating Clinically Approved Drugs or Derivatives Thereof as Ligands 1.2.1 Vorinostat and Belinostat Derivatives as Ligands Over 47% of approved drugs have been shown to target the inhibition of enzymes.20,21 Enzyme inhibition continues to represent an attractive target when designing new drugs including metal-­based anticancer agents.22 Hydroxamic acids constitute an important class of metalloenzyme inhibitors.23,24 For example, Vorinostat (7) and Belinostat (Bel) (8) (see Figure 1.2) are two hydroxamate-­based enzyme inhibitors which are used clinically to treat cancer patients. Vorinostat, also known as suberoylanilide hydroxamic acid (SAHA), received FDA approval in 2009 as a treatment for cutaneous T-­cell lymphomas in patients with progressive, persistent, or recurrent disease on or following two systemic therapeutic regimes.25–27 It is marketed under the name Zolinza and was developed by Merck. Belinostat (Bel) (8) (Figure 1.2) also referred to as Beleodaq or PDX101, is a second generation

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

7

Figure 1.2 Chemical structures of Vorinostat (7), Belinostat (8), Pt-­malSAHA (9)41,45 and Pt-­malBel (10).46

analogue of Vorinostat. It was granted FDA approval in 2014 for the treatment of patients with relapsed or refractory peripheral T-­cell lymphoma, a rare and fast-­growing type of non-­Hodgkin lymphoma (NHL).28–30 It was developed by TopoTarget (now Onxeo) and Spectrum Pharmaceuticals.30,31 Both Vorinostat and Bel are currently in various stages of clinical trials for a number of solid malignancies, including ovarian cancer.32 The anticancer effects of Vorinostat and Bel have been primarily attributed to their ability to inhibit a class of enzymes known as histone deacetylases (HDAC). These HDAC enzymes, together with histone acetyltransferases (HAT), play a key role in chromatin organisation. Chromatin is a highly compact, tense structure made up of small, positively charged histone proteins around which the negatively charged DNA coils.33 Acetylation of these core proteins, mediated by HATs, results in relaxation of the chromatin structure, and this upregulates transcription. Deacetylation, in contrast, causes the chromatin structure to condense which in turn downregulates transcription. Disruption of either enzyme-­catalysed process alters the structure which then impacts on the function of chromatin.34 The hydroxamate functional group of Vorinostat and Bel has been shown to play a key role in their ability to inhibit the HDAC enzymatic function. For example, from crystallographic and molecular modelling studies, Vorinostat has been shown to bind directly to the zinc ion at the enzyme active site. The aliphatic six-­carbon spacer of Vorinostat fits neatly into a narrow channel within the enzyme structure while the phenyl ring of Vorinostat interacts with the enzyme surface via hydrophobic interactions.35 At around the same time that Vorinostat was advancing to the clinic as an anticancer agent, we had been trying to generate some novel platinum(ii)-­ hydroxamato36,37 and ruthenium(iii)-­hydroxamato38–40 complexes. We found that the hydroxamic acid ligand did not readily bind to the platinum(ii) centre, rather it required an ancillary metal binding group in order to successfully coordinate to the platinum ion.36 Inspired by the in vitro and in vivo cytotoxicity profile of Vorinostat and its success in clinical trials, we decided

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

8

Chapter 1

to derivatise Vorinostat in such a way as to facilitate its binding to a cisplatin core while not compromising its HDAC inhibition activity. We did this by incorporating a platinum-­binding malonate linker on the phenyl ring of Vorinostat while leaving the hydroxamate moiety free to interact with the HDAC active site zinc ion. This resulted in the generation of the first example of a dual threat platinum(ii)-­HDAC inhibitor complex, cis-­[PtII(NH3)2­ (malSAHA-­2H)] (herein referred to as Pt-­malSAHA) (9) (see Figure 1.2).41 The idea  was that this Pt-­malSAHA complex would combine, into one drug molecule, the DNA-­binding properties of the platinum(ii) carboplatin-­like scaffold with the HDAC inhibitory properties of Vorinostat. Our design strategy was further motivated by the fact that HDAC inhibitors, including Vorinostat, had been shown to have preferential selectivity for tumour cells over healthy ones.42,43 They were also found to synergistically enhance the anticancer efficacy of existing drugs, including cisplatin.44 This Pt-­malSAHA complex was found to bind DNA, inhibit the HDAC8 enzyme (albeit to a lesser extent than Vorinostat) and, significantly, had potent in vitro cytotoxicity against a range of cisplatin-­sensitive and cisplatin-­resistant tumour cell lines.41,45 Interestingly, the Pt-­malSAHA complex was significantly less toxic (by one order of magnitude) compared to cisplatin against a representative healthy normal human dermal fibroblast (NHDF) cell line.41 A more in depth follow-­up study provided evidence that the complex accumulated better in tumour cells, much more so than cisplatin or Vorinostat but it bound to DNA less readily when compared to cisplatin.45 In hindsight, this is not surprising given that the chelating malonate linker bound to the platinum ion in Pt-­ malSAHA would be expected to be considerably less labile than the chlorido ligands of cisplatin. DNA binding was found to be enhanced in the presence of thiol-­containing molecules such as glutathione and thiourea, and complex activation occurred in cytosolic but not nuclear extracts of human cancer cells.45 We likewise derivatised Bel, as we did with Vorinostat, and complexed this malonate-­substituted Bel to the square planar PtII(NH3)2 framework to generate cis-­[PtII(NH3)2(malBel-­2H)] (Pt-­malBel) (10) the Bel analogue of Pt-­ malSAHA (9) (see Figure 1.2).46 While it exhibited cytotoxicity comparable to that of Pt-­malSAHA (9) against cisplatin-­sensitive A2780 ovarian cells, it was considerably more cytotoxic when compared to Pt-­malSAHA against the cisplatin-­resistant A2780cisR cells. Like Pt-­malSAHA, it too was considerably less toxic against a representative healthy NHDF cell line. These studies provided one of the first examples of how the therapeutic potential of platinum-­based anticancer agents, such as cisplatin, could be enhanced by incorporating clinically approved drugs, or derivatives thereof, as ligands. For example, preliminary evidence suggests that the Pt-­malSAHA complex, when compared with cisplatin, has the advantage of being significantly less toxic to healthy cells while retaining a cytotoxicity profile similar to that of cisplatin – it may thus overcome some of the dose-­limiting toxic side effects associated with cisplatin and therefore be better tolerated by cancer patients, were it to progress to the clinic.

View Online

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

9

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

1.2.2 Valproic Acid and Phenylbutyric Acid as Ligands We mentioned previously the advantages of ‘repurposing’ existing drugs as an accelerated route for drug discovery. Valproic acid (VPA, 2-­propylpentanoic acid) (11) (Figure 1.3) is one such drug which may fall into this category. It is used clinically to treat epilepsy and bipolar disorder and is included in the World Health Organization's List of Essential Medicines.47

Figure 1.3 Chemical structures of valproic acid (VPA) (11), platinum(ii)-­VPA

complexes (12–13),49 platinum(iv)-­VPA complexes (14–18)53–55 and a platinum(iv)-­octanoate complex (19).

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

10

Chapter 1

A number of platinum(ii)-­ and platinum(iv)-­VPA complexes have been reported. The stimulation behind their development was not however to ‘repurpose’ VPA but rather to exploit its ability to inhibit HDACs. It had also been shown to possess antimetastatic and anticancer properties.48 The generation of the resulting complexes has led to some intriguing findings. We reported the first examples of platinum-­VPA complexes, namely trans-­ [Pt(VPA-­1H)2(NH3)(py)] (12) and trans-­[Pt(VPA-­1H)2(py)2] (py = pyridine) (13) (see Figure 1.3).49 The inspiration behind their development was driven by evidence that VPA had been shown to exhibit synergistic cytotoxicity with cisplatin in a range of ovarian carcinoma cells, but also had the capacity to re-­sensitise cells that had acquired resistance to cisplatin.50 We were also cognisant of the elegant work of Farrell and others who revived research into the use of planar trans-­platinum amine (TPA) complexes as potential alternatives to classical cis-­platinum(ii) drugs.10,51 For example, Farrell et al. had previously shown that the cytotoxicity profile of TPA complexes could be markedly enhanced by incorporating carboxylato moieties in place of the ‘traditional’ chlorido leaving ligands.51,52 Other studies had indicated that substitution of the ‘traditional’ non-­leaving ammine or NH3 ligands with N-­donor heterocyclic ligands such as py could markedly enhance the cytotoxicity of the resulting trans complexes. Despite the rationale behind the generation of our complexes, trans-­[Pt(VPA-­1H)2(NH3)(py)] (12) and trans-­ [Pt(VPA-­1H)2(py)2] (13), in which the chlorido ligands in trans-­[PtCl2(py)2] and trans-­[PtCl2(NH3)(py)] had been replaced by VPA ligands, exhibited only marginally enhanced cytotoxicity against cisplatin-­sensitive A2780 and cisplatin-­ resistant A2780cisR ovarian cells, when compared to cisplatin. Interestingly, later studies by other groups, as outlined below, demonstrated that changing the oxidation state of the platinum ion from +2 to +4, while retaining the VPA as a HDAC inhibitor ligand, could markedly enhance the efficacy of the resulting complexes. Several teams have since independently reported platinum(iv)-­VPA complexes. Tang, Shen, et al. developed a platinum(iv) complex incorporating a cisplatin framework with two axial VPA ligands, namely cis,cis,trans-­ diamminedichlorobisvalproato-­platinum(iv) or VAAP (14) (Figure 1.3).53 They packaged it into PEG-­PCL nanoparticles or dispersed into a Tween 80 surfactant in order to promote tumour cell uptake. Both the non-­PEGylated and PEGylated VAAP derivatives were found to be highly cytotoxic, much more so than cisplatin, across multiple tumour cell lines. The ability of VAAP to induce HDAC inhibition was, however, not evaluated.53 In a parallel study, Osella et al. likewise developed the same platinum(iv)-­VPA derivative, i.e. VAAP (14) (Figure 1.3).54 They also found the complex to be highly cytotoxic, in their case against four highly malignant and highly chemoresistant plural mesothelioma cell lines, again more so than cisplatin. The team were intrigued, however, when they assessed the cytotoxicity of an isomer of VAAP (14), namely cis,cis,trans-­diamminedichloridobis(n-­octanoato)platinum(iv) (19), in which the VPA ligands, which had been chosen for their potential to inhibit HDACs, were replaced with ‘innocent’ octanoate (OA)

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

11

ligands. The latter complex exhibited similar and even sometimes greater efficacy than the platinum(iv)-­HDAC inhibitor conjugate (14). This raised the important question as to whether or not the stoichiometric ratio of VPA : platinum (2 : 1) in VAAP (14) was sufficient to induce a synergistic outcome. They ultimately concluded that the improved cytotoxicity observed for the complexes could not be attributed to any synergistic contribution of the VPA ligands as HDAC inhibitors, rather that the VPA and OA ligands were altering the pharmacokinetic profile of the resulting complexes, leading to an increase in their lipophilicities and ultimately enhancing their accumulation into tumour cells.54 Gibson, Brabec, et al. developed two further platinum(iv)-­VPA complexes but, in their case, they employed an oxaliplatin scaffold to which they appended either one or two axial VPA ligands, (15 and 16) (Figure 1.3).55 Their choice of oxaliplatin as a core scaffold was in part driven by the fact that oxaliplatin, when given in combination with the HDAC inhibitor trichostatin A, resulted in an additive cytotoxic effect when tested against gastric tumour cells. Their complexes exhibited greater efficacy against both cisplatin-­sensitive and cisplatin-­resistant cell lines when compared to platinum(iv) analogues without biologically active axial ligands. Interestingly, their oxaliplatin-­t ype platinum(iv)-­VPA complexes were found to markedly downregulate HDACs, leading to a reduction in cellular levels of HDACs rather than direct inhibition of the HDAC enzymes. The enhanced cytotoxicity observed for these complexes was linked to this HDAC downregulation.55 This contrasted with the findings by Osella et al.54 which primarily attributed lipophilicity to the improved cytotoxicity profile of their cisplatin-­ type platinum(iv)-­VPA complex, VAAP (14). Gibson et al. undertook a further study to probe these differing conclusions. They compared the efficacy of platinum(iv)-­VPA complexes bearing a cisplatin equatorial core, (14 and 17) (Figure 1.3), against platinum(iv) analogues bearing ‘innocent’ biologically inert axial acetate ligands.56 Again, the VPA complexes were more cytotoxic than the complexes bearing the ‘innocent’ carboxylate ligands against the cell lines tested. The complexes also appeared to inhibit the expression of the HDAC protein, rather than inhibit HDAC directly. The complexes bearing the VPA ligands interfered with other cellular processes, including interfering with enzymes such as glutathione S-­transferases. The team recognised that the enhanced lipophilicity bestowed on the complexes by the presence of the VPA ligands was also a positive contributory factor, accounting for the improved cytotoxicity observed for these complexes. They did not rule out the possibility that the platinum(iv)-­VPA complexes could also be interacting with other biological targets, an important conclusion to highlight given the complex environment within any given cell.56 Expanding their study, Gibson et al. developed platinum(iv) analogues of cisplatin, (14 and 17), oxaliplatin, (15 and 16), and trans-­[Pt(n-­butylamine) (piperidino-­piperidine)Cl2]+ (18) (see Figure 1.3), incorporating VPA in one or both axial positions. They also included platinum(iv) complexes (21–23) incorporating 4-­phenylbutyrate (PhB) (20) (Figure 1.4) as axial ligands.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

12

Chapter 1

Sodium phenylbutyrate, the sodium salt of PhB, is used to treat urea cycle disorders. It is also undergoing clinical trials as a treatment for cancer, haemoglobinopathies, motor neuron diseases and cystic fibrosis.57 The ligand PhB has also been shown to possess HDAC inhibition properties and it was on this premise that it was included in this study. The cytotoxicities of the resulting complexes (21–23) were compared to platinum(iv) complexes bearing ‘innocent’ carboxylate ligands. This was an interesting systematic study to identify not only the optimal platinum core scaffold but also the optimal axial HDAC inhibitor ligands, to ultimately optimise cytotoxic activity. In all cases, the complexes with a cisplatin core were more cytotoxic than those with an oxaliplatin core. The cis,trans,cis-­[Pt(NH3)2(PhB)2Cl2] (22) (Figure 1.4) exhibited greatest cytotoxicity against all of the human cancer cell lines tested; lung, breast, pancreatic, kidney, prostate and colon carcinoma, along with melanoma. This complex was approximately 100-­fold more cytotoxic than cisplatin and even more so when compared to cisplatin-­t ype platinum(iv) derivatives bearing either two hydroxido, two acetato or two VPA ligands. Again, there appeared to be a positive correlation between lipophilicity, cellular accumulation and cytotoxicity of these complexes. Histone deacetylase inhibition appeared to be enhanced when the VPA or PhB ligands were coordinated to the metal scaffold, with the complexes being more potent HDAC inhibitors relative to VPA or PhB alone. It was clear by the end of this study that the cytotoxicity of these complexes was multi-­ factorial and that a ‘dual-­functional’ view of these complexes was being overly simplistic.58 As key themes began to emerge, two of the main research groups working in this field (Osella et al. and Gibson, Brabec, et al.) began a collaborative study to better elucidate the mechanism of action giving rise to the enhanced cytotoxicity observed for the platinum(iv) complexes incorporating the axial ‘innocent’ OA ligands, over those containing branched isomers like VPA.59 The HDAC inhibition properties of OA are known to be significantly less than those of VPA, yet the platinum(iv) complexes bearing OA axial ligands were more cytotoxic than those containing VPA ligands against numerous cell lines.59 Accounting for this anomaly formed the focus of this study. The study validated that the platinum(iv) complex bearing two axial OA ligands was the most potent across several cell lines, significantly more so compared to cisplatin. This enhanced cytotoxicity could not be attributed to HDAC inhibition alone. Further studies to assess the capacity of the complex to either bind or methylate DNA and its impact on the mitochondria were undertaken to try to ascertain the influence of the OA ligands on the overall cytotoxicity profile of the complex. As anticipated, the presence of the OA ligands endowed the complex with greater lipophilicity which in turn enhanced its accumulation into tumour cells. Upon cellular accumulation, the complex was reduced to platinum(ii) with concomitant release of the OA ligands. The resulting platinum(ii) adduct was shown to bind DNA while the OA ligands were found not only to hypermethylate DNA but also to cause a reduction of

View Online

13

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

Figure 1.4 Chemical structures of 4-­phenylbutyric acid (PhB) (20) and platinum(iv)-­ PhB complexes (21–28).58,61

the mitochondrial membrane potential.59 DNA methylation is known to play a key role in gene expression associated with carcinogenesis, cancer progression and metastasis.60 It is interesting to note that what were perceived as ‘innocent’ or ‘biologically inactive’ OA ligands were not at all ‘innocent’. The platinum(iv) complex bearing two axial OA ligands (19) (Figure 1.3), was significantly more cytotoxic than the complexes incorporating the HDAC inhibitors, VPA and

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

14

Chapter 1

PhB. While the results were unexpected, the methodical approach of the team led to interesting findings as indicated above and highlights the need for latitude in interpreting results. More recently, Erxleben, Montagner, et al. developed a related class of platinum(iv)-­PhB complexes. In their case, they employed a platinum scaffold incorporating the N,N-non-leaving ligand of oxaliplatin and the O,O-leaving ligand of carboplatin, to which they appended the PhB ligand in one or both axial positions.61 Their library consisted of complexes bearing either two axial PhB ligands (24), or one PhB and either hydroxide (25), acetate (26), succinate (27) or benzoate (28) in the other axial position (see Figure 1.4). The complex which demonstrated greatest efficacy turned out to be the derivative containing a PhB in one axial position and benzoate in the other (28) (Figure 1.4), with IC50 values lower than carboplatin against all the cell lines tested. Of the complexes developed, there appeared to be a direct correlation between cytotoxicity and cellular accumulation and HDAC inhibition with the lead complex (28) exhibiting highest accumulation and HDAC inhibition ability.61 A slightly different approach by Gandin, Gibson, et al. involved the development of platinum(iv) analogues of [Pt(1S,2S-­diaminocyclohexane)(5, 6-­dimethyl-­1,10-­phenanthroline)]2+ (Pt56MeSS) (29) (Figure 1.5). Unlike classical platinum(ii) drugs, this Pt56MeSS complex does not contain any labile ligands and is thus not expected to undergo the type of substitution reactions associated with classical platinum(ii) drugs. Rather, it bears two non-­leaving bidentate N,N-­donor ligands. Although it breaks the structure activity relationship associated with classical platinum(ii) drugs, the compound has been shown to be highly cytotoxic towards cisplatin-­resistant and oxaliplatin-­resistant cell lines.62 A proposed target for this complex is the mitochondrion. In this study, both non-­bioactive, lipophilic and bioactive (VPA and PhB), axial ligands (30–35) were tethered to the platinum(iv) base of Pt56MeSS (29) (Figure 1.5).62 This new library of platinum(iv) derivatives had, on average, greater efficacy over cisplatin and possessed either comparable or lower efficacy when compared to Pt56MeSS itself. Interestingly, the presence of the axial HDAC inhibitor ligands appeared to have little or no influence on the cytotoxicity of the resulting complexes. For example, the average IC50 value for Pt56MeSS across seven human tumour cell lines (lung (H157), colon (HCT-­15), breast (MCF-­7), thyroid (BCPAP), ovarian (2008) and pancreatic (BxPC3)) was 1.24 µM as compared to 2.34 µM for the platinum(iv)

Figure 1.5 Chemical structures of platinum(iv) analogues (30–35) of Pt56MeSS (29).62

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

15

complex bearing one OAc and one PhB ligand (33), 1.41 µM for the complex bearing two PhB ligands (34) and 2.00 µM for the complex bearing one OAc and one VPA ligand (35).62 There did not appear to be any correlation between cellular accumulation and cytotoxicity, nor HDAC inhibition, in the case of VPA and PhB axial ligands. Furthermore, these results were in direct contrast with previous reports of platinum(iv) complexes incorporating a cisplatin scaffold and bearing PhB axial ligands, where there was an order of magnitude difference in cytotoxicity recorded. For example, the IC50 value for the platinum(iv) complex bearing two PhB axial ligands and containing a cisplatin core (22) (Figure 1.4), was 0.18 µM against breast (MCF-­7) cancer cells and 0.31 µM against colon (HCT-­15) cells in contrast to IC50 values of 2.22 µM and 0.69 µM for 34 against the same cell lines, respectively. That said, the core frameworks are different, which probably accounts for the differences observed.62

1.2.3 Indomethacin, Ibuprofen and Aspirin as Ligands Cyclooxygenase (COX) enzymes are amongst the most widely studied and best understood of all the mammalian oxygenases. Three isoforms of COX have been identified; COX-­1, COX-­2 and COX-­3.63 They catalyse the conversion (bis-­dioxygenation and subsequent reduction) of arachidonic acid to prostaglandin G2 and H2 (PGG2 and PGH2), mediators of inflammatory and anaphylactic reactions. For this reason, they have been the subject of intense investigation and, already, a number of COX inhibitors are in clinical use to treat pain and inflammation. Non-­steroidal anti-­inflammatory drugs (NSAIDs) are one such class that have proven highly successful in this regard. These NSAIDs include, but are by no means limited to, indomethacin (36), ibuprofen (37) and aspirin (38) (Figure 1.6), the drugs of relevance to this section. Interestingly, there is a significant body of evidence in the literature to support the hypothesis that overexpression of the COX-­2 isoform is a driver of carcinogenesis and, conversely, inhibition of COX-­2 is an attractive drug target for cancer prevention and therapy. For example, COX-­2 has been shown to be constitutively expressed throughout breast cancer development, from the detection stage to cancer progression and metastases. It has also been shown that mammary carcinogenesis (which includes mutagenesis, mitogenesis, angiogenesis, reduced apoptosis, metastasis and immunosuppression) has been linked to COX-­2-­mediated prostaglandin E2 (PGE2) biosynthesis. There is also evidence to support that COX-­2 inhibitors decrease the risk of breast cancer in female patients without this disease and, similarly, decrease the incidence of recurrence risk and mortality in patients with breast cancer.64 Overexpression of COX-­2 is not limited to breast cancer cells. Other tumour types which overexpress COX-­2 include skin, oesophagus, stomach, colorectal, pancreas, and bladder.65 It is also interesting to note that a number of COX inhibitors, when used in combination with established cancer drugs, including cisplatin, paclitaxel and doxorubicin, act in a synergistic

View Online

Chapter 1

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

16

Figure 1.6 Chemical structures of indomethacin (36), ibuprofen (37) and aspirin (38) and their platinum(iv) complexes (39–43).67,68,72–75

manner.66 To the best of our knowledge, Neumann et al. were the first to report platinum(iv)-­NSAID complexes. They rationally designed symmetrical platinum(iv) prodrugs consisting of either a cisplatin (39 and 41, respectively) or oxaliplatin core scaffold (40 and 42, respectively) and incorporating either indomethacin (39 and 40) or ibuprofen (41 and 42) in the axial positions (see Figure 1.6).67,68 Indomethacin is a non-­specific COX inhibitor. It has been found to induce a profound reduction in the ability of breast cancer cells to invade and degrade the extracellular matrix gel.69 The cisplatin-­like indomethacin (39) and ibuprofen (41) derivatives were significantly more cytotoxic compared to cisplatin. They were also highly cytotoxic against cisplatin-­resistant MDA-­ MB-­231 breast cancer cells. When tested against two tumour cells with different levels of COX-­2 expression, namely the cisplatin-­sensitive colorectal HCT 116 carcinoma cells, which do not express COX-­2, and cisplatin-­resistant breast MDA-­MB-­231 adenocarcinoma cells, which exhibit high constitutive COX-­2 expression, the cytotoxicities of the complexes did not differ significantly. Interestingly, despite comparable cytotoxicities, the indomethacin complexes exhibited strong COX inhibitory activity in contrast to the ibuprofen analogues, which exhibited poor COX inhibition. The team concluded that there did not appear to be a correlation between cytotoxicity and COX

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

17

inhibition, nor to their potency as NSAIDs, nor to the levels of COX expression in these cells. The team ultimately concluded that the change in the physicochemical properties of the complexes, resulting from the presence of the lipophilic NSAIDs, facilitated greater uptake of the complexes into the tumour cells and it was this property that accounted for the increase in cytotoxicity relative to cisplatin.67 A follow-­up study by the same team sought to validate their hypothesis. Employing the same NSAID ligands, they developed both platinum(iv)-­ NSAID conjugates, this time with an oxaliplatin base (40 and 42) (Figure 1.6) as well as the analogous platinum(ii) derivatives for comparative purposes. In all cases, the platinum(iv)-­NSAIDs exhibited more potent cytotoxicities relative to the platinum(ii) analogues. The former were also able to overcome cisplatin resistance. The team concluded that COX-­mediated pathways were not responsible for the cytotoxicity of these complexes.68 Aspirin (acetylsalicylic acid) (38) (Figure 1.6) is a hugely successful drug used globally as an analgesic and anti-­pyretic agent. It is also commonly used for cardiovascular prophylaxis. There is considerable evidence in the literature citing its potential as a cancer chemopreventive agent, particularly against colorectal cancer (CRC), with several studies showing a drop in the incidence of CRCs following long-­term treatment with low doses of aspirin.70,71 Dhar et al. developed Platin-­A, an asymmetric platinum(iv) prodrug incorporating one aspirin axial ligand (43) (see Figure 1.6).72 Another group, led by Liu, reported this same complex a few months later, but they referred to it as Asplatin.73 The study of Dhar et al. showed that, upon tumour cell accumulation and following the proposed ‘activation by reduction’ process, Platin-­A released cisplatin and one equivalent of aspirin. It was found to be highly cytotoxic against androgen-­unresponsive prostate PC3 and DU145 cells, with a cytotoxicity profile comparable to that of cisplatin. Aspirin alone was not cytotoxic. The complex was also later shown to possess anti-­inflammatory properties mediated via COX-­2 inhibition.74 The study by Liu et al. screened this same complex, Asplatin (43) (Figure 1.6) against a range of additional tumour cell lines (cervical HeLa, breast MCF-­7, liver HepG2, lung A549, A549R) in addition to normal human fibroblast cells.73 They found Asplatin to be highly cytotoxic across all the tumour cells, including chemoresistant cells, and more so than cisplatin. Asplatin was found to retain its cytotoxic activity, better than cisplatin, when tested in vivo in mice bearing lung A549 tumours, while also exhibiting a better safety profile. They also demonstrated that the complex could be reduced to its cisplatin core and aspirin following treatment with equimolar concentrations of ascorbic acid, supporting the ‘activation by reduction’ hypothesis proposed by Dhar et al. While the team demonstrated that Asplatin could bind DNA, they did not assess its ability to inhibit COX activity.73 They did, however, in a later study, report evidence suggesting that the complex modulated the cellular response to the platinum cytotoxic agent. Through apoptosis analysis and gene expression studies, the complex was shown to promote apoptosis via the BCL-­2-­associated mitochondrial pathway. While BCL-­2 was shown to be downregulated, BAX and

View Online

Chapter 1

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

18

BAK were found to be upregulated, and this combination caused an increase in the permeability of the mitochondrial outer membrane. This change in permeability facilitated the release of cytochrome C into the cytosol, promoting apoptosis mediated via caspase activation processes.75 This is one of only a few studies reported to date in which gene expression analysis was exploited in order to more fully elucidate the mechanisms underpinning the anticancer properties of platinum complexes.

1.2.4 Ethacrynic Acid as a Ligand Ethacrynic acid (Edecrin) (44) (Figure 1.7) is in clinical use as a loop diuretic which, when administered, leads to prompt and excessive diuresis. Its primary mode of action has been attributed to its ability to inhibit the activity of the Na+-­K+-­2Cl− symporter in the thick ascending limb of the loop of Henle.76 Ethacrynic acid has also been shown to possess potent glutathione S-­transferase (GST) inhibition activity.77,78 Glutathione S-­transferases are considered one of the most important classes of detoxification enzymes that work to remove harmful chemicals from the body via phase II biotransformations.78 They have been specifically shown to play a part in the detoxification of platinum anticancer drugs and, in fact, certain cisplatin-­resistant tumours have been shown to overexpress these enzymes.79 These GST enzymes catalyse the nucleophilic S-­conjugation between the thiol group of glutathione and xenobiotics, including cisplatin, which facilitates their elimination from the body via the mercapturic acid pathway.80 Given the affinity of the platinum ion for ‘soft’ nucleophiles, including those containing thiol groups, it is not surprising that they represent a viable target for GST-­mediated detoxification. Dyson et al. developed ethacraplatin (45) (Figure 1.7), a platinum(iv) complex consisting of a cisplatin core with

Figure 1.7 Chemical structures of ethacrynic acid (44), ethacraplatin (45)81 and

another platinum(iv)-­ethacrynate complex bearing only one ethacrynate ligand (46).82

View Online

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

19

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

81

ethacrynate ligands in both axial positions. They proposed that the presence of the ethacrynate ligands would endow the complex with greater lipophilic properties and that this would enhance its uptake into tumour cells, more so than cisplatin. They also suggested that, upon tumour cell entry, the complex would be reduced which would result in the release of its GST inhibitor ligands, thus inhibiting any platinum drug detoxification. Cisplatin, which would be simultaneously released, would therefore be free to interact with its biological target, DNA.81 When screened against a series of tumour cell lines, including cisplatin-­ resistant breast, lung and colon carcinomas, ethacraplatin was indeed found to be significantly more cytotoxic compared to cisplatin alone. Follow-­up biochemical and structural studies were conducted by the same group in an attempt to better understand the mode of action of ethacraplatin.81 The group specifically chose GST P1-­1 as their protein target for this study, given its importance in the mercapturic acid detoxification pathway. Ethacraplatin was indeed shown to bind to this target at the dimer interface, with the ethacrynate ligands interacting at both active sites. Interestingly, the cisplatin scaffold was found to be sandwiched between two bridging cysteine residues at the dimer interface, suggesting that it would remain bound and therefore not free to bind its target, DNA. While this study demonstrated evidence of strong and irreversible enzymatic inhibition by ethacraplatin (45), it also revealed that the cytotoxicity could not be attributed to platinum-­DNA-­ binding interactions as originally anticipated.81 It was suggested more recently that cisplatin release from ethacraplatin was not sufficiently fast enough, due to its low reduction rate, to facilitate any platinum-­DNA binding. With this in mind, Ang, Montagner, Nowak-­ Sliwinska, Dyson, et al. recently developed an ethacraplatin analogue in which they replaced one of the ethacrynate ligands with a hydroxido ligand (46) (Figure 1.7).82 This monofunctionalised complex was shown to overcome the aforementioned limitation of ethacraplatin. The ethacrynate ligand was readily released in vitro from the platinum(iv) framework, with concomitant release of the cytotoxic platinum(ii) agent. The complex was shown to inhibit GST in a non-­competitive way. Despite losing one of the ethacrynate ligands, the complex retained its cytotoxicity against both cisplatin-­ sensitive and cisplatin-­resistant tumour cells. Its potential as a new class of dual functional anticancer agent was validated using an in vivo study which demonstrated that the complex induced ∼80% inhibition of tumour growth in a human ovarian carcinoma tumour model.82

1.2.5 Dichloroacetate or Dichloroacetate Derivatives as Ligands 2,2-­Dichloroacetic acid (DCA) (47) (Figure 1.8) is an intriguing small molecule with various therapeutic applications. For example, DCA is not only used to treat inherited mitochondrial disorders that result in lactic acidosis, but also pulmonary hypertension.83 It is also under investigation as an

View Online

Chapter 1

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

20

Figure 1.8 Chemical structures of 2,2-­dichloroacetic acid (DCA) (47) and platinum(ii) complexes bearing DCA derivatives (48–53)88,89 and the platinum(iv)-­DCA complex, mitaplatin (54).90

anticancer agent given its capacity to reverse the ‘Warburg effect’.83 This effect refers to the over-­reliance of cancer cells on cytosolic aerobic glycolysis to generate energy, in contrast to healthy cells which primarily rely on mitochondrial oxidative phosphorylation.84 Tumour cells can therefore be targeted while leaving normal cells unharmed. This simple DCA molecule works by modulating carbohydrate metabolism at the level of the multi-­enzyme mitochondrial pyruvate dehydrogenase complex (PDC). This complex, which is present in the mitochondrial matrix, acts as a gatekeeper by linking cytoplasmic glycolysis to the mitochondrial tricarboxylic acid (TCA) cycle and oxidative phosphorylation (OXPHOS).83 Pyruvate dehydrogenase kinases (PDK 1–4) act by inhibiting the activity of this complex while pyruvate dehydrogenase phosphatases (PDP 1 and 2), in contrast, maintain PDC in its unphosphorylated, active state. Certain cancer

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

21

cells overexpress these PDK isoforms and this leads to a reduction in PDC activity.85 Dichloroacetate, which is a structural analogue of pyruvate, acts by inhibiting PDKs, thus stimulating PDC and OXPHOS activity.86,87 Several studies have indicated that this increase in OXPHOS activity leads to the production of reactive oxygen species by the mitochondrial respiratory chain which, following other downstream changes in mitochondrial function, leads to selective apoptosis of tumour cells. This effect has also been observed in rodents bearing tumour xenografts.86 Li et al. developed two interesting platinum(ii) complexes bearing ligands which contain a DCA derivative as part of their framework. Specifically they tethered the DCA fragment to cyclobutane-­1,1-­dicarboxylate via an ester bond linkage which resulted in the formation of 3-­dichoroacetoxylcyclo­ butane-­1,1-­dicarboxylate. Cyclobutane-­1,1-­dicarboxylate was an interesting  choice given that it is the O,O-­bidentate leaving ligand in carboplatin. It was this derivatised ligand that they then complexed to platinum(ii) centres. Two novel mixed-­ammine/amine platinum(ii) complexes (48 and 49) (Figure 1.8) were formed.88 Although the complexes exhibited marked cytotoxic selectivity towards cancer cells (including cisplatin-­resistant SK-­OV-­3 cells) over BEAS-­2B normal cells, their further development was limited due to their poor aqueous solubility.88 In order to address this shortcoming, the same group developed four additional diam(m)ine platinum(ii) complexes. These complexes retained the original carboplatin-­like dicarboxylate DCA-­containing leaving group but the N-­donor non-­leaving ligand(s) were modified (50–53) (Figure 1.8). Following cytotoxicity screening against A549, SK-­OV-­3 and SK-­OV-­3/DDP (cisplatin resistant) cell lines, the oxaliplatin analogue, 51, was found to be not only the most potent but also ∼10 times more water soluble than the complexes reported previously. It was found to be 60 times more cytotoxic compared to carboplatin against the A549 cells, six times more so against SK-­OV-­3 cells and >15 times more cytotoxic against the SK-­OV-­3/DDP cells which are cisplatin resistant. This complex was also shown to release its moiety via hydrolysis of the ester bond under physiological conditions.89 There is one example to date of a platinum(iv)-­DCA complex, namely mitaplatin (54) (see Figure 1.8).90 This complex, which was developed by Dhar and Lippard, contains a cisplatin equatorial base with two axial DCA ligands. Cytotoxicity screening against eight cancerous cell lines revealed that, in addition to having comparable cytotoxicity to cisplatin, the complex exhibited equal or improved cytotoxicity compared to many of the platinum(iv) complexes that had been reported at that time. It was also found to be more cytotoxic than DCA alone.90 The fact that mitaplatin (54) was shown to have low toxicity against human fibroblast cells suggested that the presence of the DCA ligands may be impacting on its cytoselectivity profile. This cytoselectivity provides preliminary evidence to support their drug design strategy that cells dependent on glycolysis would be targeted. The complex, 54, induced DNA damage analogous to the type of damage induced by cisplatin, which supports an ‘activation by reduction’ mode of action. Liang et al. provided

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

22

Chapter 1

evidence that mitaplatin specifically targets the mitochondria. They demonstrated that mitaplatin activated a downstream mitochondrial-­dependent cell death in tumour cells resistant to cisplatin. It was found to act as a metabolic modulator by downregulating the phosphorylation of pyruvate dehydrogenase. This led to a glycolytic shift back to oxidative phosphorylation by increasing the uptake of acetyl-­CoA into the mitochondria.91 Platinum(iv) complexes, as stated earlier, are expected to be kinetically inert. This, however, was found not to be the case for mitaplatin (54) (Figure 1.8). Gibson et al. revealed that mitaplatin was not very stable under biological conditions, including cell culture media. They showed that 50% of mitaplatin had undergone hydrolytic degradation after only 2 hours.92 Given the cytotoxicity profile of mitaplatin, the group concluded that the uptake of mitaplatin into tumour cells must be faster than its rate of hydrolysis and that mitaplatin hydrolysis products, in addition to mitaplatin, may be responsible for the cytotoxicity observed. Encapsulation of mitaplatin in a nanoparticle formulation was later shown by Lippard et al. to increase its efficacy in an in vivo mouse xenograft model of triple-­ negative breast cancer.93

1.3 Platinum Complexes Incorporating More Than One Clinically Approved Drug as a Ligand While there is plenty of evidence now in the literature to support a multi-­ targeted approach in the quest to bring forward a new drug class, Gibson et al. brought this to another level in that they rationally designed and developed a dinuclear ‘quadruple action’ platinum(iv) prodrug. This complex contains two platinum(iv) centres, which, upon reduction, release simultaneously four different bioactive moieties; namely, cisplatin, Pt56MeSS (29) (Figure 1.5), and the clinically approved DCA (47) (Figure 1.8) and PhB (20) (Figure 1.4) ligands.94 One platinum(iv) centre is contained within a cisplatin scaffold with one monodentate DCA ligand and one bridging dicarboxylate ligand. The DCA ligand was chosen for its PDK inhibitory properties. The other platinum(iv) centre is derived from a non-­DNA-­binding Pt56MeSS core (29) (Figure 1.5) previously shown to act on the mitochondria. The coordination sphere around this platinum(iv) centre is completed by one axial monodentate PhB HDAC inhibitor ligand and the bridging dicarboxylate ligand. A number of compounds were also synthesised serving as reference standards for biological testing. These included the platinum(ii) complexes cisplatin, oxaliplatin and Pt56MeSS (29) (see Figure 1.5), the monomeric dual action platinum(iv) complexes (56) and (57) and the dimer (58) (Figure 1.9). The activity of the dinuclear ‘quadruple action’ platinum(iv) prodrug (55) was compared to these complexes across a range of human cancer cell lines, including ovarian, cervical, lung, colon and pancreatic cancer cells.94 Of the complexes assessed, the ‘quadruple action’ complex (55) was significantly more cytotoxic compared to the dual functioning or single

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

23

Figure 1.9 Chemical structures of a ‘quadruple action’ platinum(iv) prodrug (55)94

and its mononuclear (56 and 57) and dinuclear (58) adducts (with DCA highlighted in green and PhB in blue).

agents, suggesting that the bioactive ligands present within the complex framework were contributing to the enhanced cytotoxicity observed. It was also significantly more cytotoxic compared to cisplatin alone and oxaliplatin alone. The ‘quadruple action’ complex, in particular, exhibited remarkable cytotoxic activity against KRAS-­mutated pancreatic (MIAPaCa-­2) and colon (LoVo) cancer cell lines. For example, the complex had an IC50 value of 0.06 ± 0.01 µM against MIAPaCa-­2 cells and 0.02 ± 0.005 µM against LoVo cells. These IC50 values were between 200-­ and 450-­fold lower than those for cisplatin against the same cell lines (where the IC50 value for cisplatin against MIAPaCa-­2 cells was 13.45 ± 2.45 µM and against LoVo cells was 9.12 ± 0.005 µM). The ‘quadruple action’ complex was also 40‐fold more selective towards KRAS-­mutated cells compared to non‐cancerous cells. KRAS represents a major oncogene associated with aggressive cancers.95 These cancers tend to have poor prognosis and to date there does not appear to be any effective drug treatment. The activity of 55 was also assessed against HCT-­15, BxPC3 and KRAS-­mutated PSN1 spheroid models. These 3-­D cell cultures are used to more closely mimic the in vivo environment of tumour cells. In all cases,

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

24

Chapter 1

the complex was more cytotoxic compared to cisplatin and compared favourably to oxaliplatin in the non-­KRAS-­mutated cells. Given the sound rationale behind the development of this quadruple action complex, the team went on to establish that, upon cellular activation, it had the capacity to covalently modify DNA following release of its cisplatin core. It was also shown to impact on mitochondrial function and inhibit HDACs, validating the original design strategy.

1.4 Conclusions Cancer is a multi-­factorial disease which continues to represent a global health challenge. Platinum drugs remain a cornerstone in cancer drug treatment regimens but they have shortcomings, including toxic side effects and resistance issues. The rational design and development of new platinum drugs that can overcome these drawbacks remains a thriving field of research. In recent years, there has been a tangible shift in focus towards designing platinum drugs that contain either vectors to target tumour cells (thus reducing unwanted side reactions and consequently dose-­limiting toxic side effects), or platinum drugs that can target more than one cellular entity. The latter approach may lead to new ‘multi-­targeted’ drug candidates with a mechanism of action different from clinically used platinum drugs and may thus overcome resistance issues that can plague some cancer treatment regimens. This chapter provides an overview of recent developments in the design and development of a relatively new type of ‘multi-­targeted’ platinum drug complex. These complexes, in addition to targeting DNA, incorporate clinically approved drugs, or derivatives thereof, as ligands which target other cellular entities. Interestingly, while there is a sound rationale behind repurposing existing drugs as outlined in the introduction to this chapter, those drugs to date that have been tethered to platinum have not been chosen for this purpose. Rather, they have been selected because they have, in their own right, exhibited anticancer activities, albeit that some are in clinical use for other indications. There is no doubt that there are advantages to this approach. Knowing the pharmacokinetic profile of these drug ligands which have already been approved for clinical use can better inform the drug design strategy. However, due cognizance should also be given to the therapeutic advantages of employing drug molecules as ligands over their known toxicity profiles. Secondly, incorporating two drug entities into one drug molecule allows for greater pharmacokinetic control, including delivery of a single drug entity to its target site. Having one drug instead of two may also potentially reduce drug costs as well as increase patient compliance. We have also highlighted numerous examples of platinum drug complexes in which the incorporation of the drug, or drug derivative, has clearly

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

25

endowed the resulting complexes with enhanced in vitro cytotoxicity compared with cisplatin, one of the gold standards in chemotherapeutic regimens. Many of the complexes highlighted in this chapter have demonstrated great potential, particularly against tumour cells that are known to be resistant to platinum drug treatments. Whether the potent cytotoxic properties of these promising agents can translate to in vivo cancer xenograft models remains to be explored, in most cases. In conclusion, there is no doubt that with due cognisance, one can carefully select a clinically approved drug to serve as a ligand to enhance the therapeutic efficacy of existing platinum drugs. In vivo studies will play a critical role in determining the potential of these and any new complexes of this type if they are to advance from pre-­clinical to clinical development. The literature is rich with examples of drugs that can be further exploited as drug ligands. Combining these drugs into a metallodrug framework may well form the basis of a new drug class which may offer advantages over existing therapeutic regimens.

Abbreviations Bel Belinostat COX Cyclooxygenase VAAP cis,cis,trans-­Diamminedichlorobisvalproato-­platinum(iv) DCA Dichloroacetate HAT Histone acetyltransferases HDAC Histone deacetylases GST Glutathione S-­transferase TCA Mitochondrial tricarboxylic acid NHL Non-­Hodgkin lymphoma NSAIDs Non-­steroidal anti-­inflammatory drugs NHDF Normal human dermal fibroblast OA Octanoato OXPHOS  Oxidative phosphorylation PhB 4-­Phenylbutyrate py Pyridine SAHA Suberoylanilide hydroxamic acid SubH Suberoyl-­bis-­hydroxamic acid TPA trans-­Platinum planar amine VPA Valproic acid, 2-­propylpentanoic acid

Acknowledgements This material is based upon works supported by the Science Foundation Ireland under Grant Nos. [11/RFP.1/CHS/3095] and [12/TIDA/B2384] and [17/ TIDA/5009]. This work has also been funded by the RCSI under the Apjohn Scholarship programme. Funding under the Programme for Research in

View Online

26

Chapter 1

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Third-­Level Institutions and co-­funding under the European Regional Development fund (BioAT programme) is also acknowledged. The authors would also like to acknowledge COST CM1105 and CA13135 for providing a platform to progress fruitful collaborations.

References 1. https://www.cancer.org/health-­care-­professionals/our-­global-­health-­ work/global-­cancer-­burden.html, (accessed 26/07/2018). 2. T. Pham, S. Roth, J. Kong, G. Guerra, V. Narasimhan, L. Pereira, J. Desai, A. Heriot and R. Ramsay, Ann. Surg. Oncol., 2018, 25, 3404. 3. A. T. Baker, C. Aguirre-­Hernandez, G. Hallden and A. L. Parker, Cancers, 2018, 10, 201. 4. T. C. Johnstone, G. Y. Park and S. J. Lippard, Anticancer Res., 2014, 34, 471. 5. T. C. Johnstone, K. Suntharalingam and S. J. Lippard, Chem. Rev., 2016, 116, 3436. 6. G. Tapia and I. Diaz-­Padilla, Molecular Mechanisms of Platinum Resistance in Ovarian Cancer, Ovarian Cancer -­ A Clinical and Translational Update, Available from: https://www.intechopen.com/books/ovarian-­cancer-­a-­ clinical-­and-­translational-­update/molecular-­mechanisms-­of-­platinum-­ resistance-­in-­ovarian-­cancer, 2013. 7. D.-­W. Shen, L. M. Pouliot, M. D. Hall and M. M. Gottesman, Pharmacol. Rev., 2012, 64, 706. 8. L. Kelland, Nat. Rev. Cancer, 2007, 7, 573. 9. M. Coluccia and G. Natile, Anti-­Cancer Agents Med. Chem., 2007, 7, 111. 10. N. Farrell, Cancer Invest., 1993, 11, 578. 11. V. Brabec, J. Kasparkova, V. Menon and N. P. Farrell, Met. Ions Life Sci., 2018, 18, 43–68. 12. N. P. Farrell, Chem. Soc. Rev., 2015, 44, 8773. 13. D. I. Jodrell, T. R. Evans, W. Steward, D. Cameron, J. Prendiville, C. Aschele, C. Noberasco, M. Lind, J. Carmichael, N. Dobbs, G. Camboni, B. Gatti and F. De Braud, Eur. J. Cancer, 2004, 40, 1872. 14. X. Han, J. Sun, Y. Wang and Z. He, Med. Res. Rev., 2015, 35, 1268. 15. J. P. Parker, Z. Ude and C. J. Marmion, Metallomics, 2016, 8, 43. 16. H. H. Chen, W. C. Chen, Z. D. Liang, W. B. Tsai, Y. Long, I. Aiba, S. Fu, R. Broaddus, J. Liu, L. G. Feun, N. Savaraj and M. T. Kuo, Expert Opin. Ther. Targets, 2015, 19, 1307. 17. R. G. Kenny, S. W. Chuah, A. Crawford and C. J. Marmion, Eur. J. Inorg. Chem., 2017, 2017, 1596. 18. V. Venkatesh and P. J. Sadler, Met. Ions Life Sci., 2018, 18, 69–108. 19. Y. Y. Li and S. J. M. Jones, Genome Med., 2012, 4, 27. 20. A. L. Hopkins and C. R. Groom, Nat. Rev. Drug Discovery, 2002, 1, 727. 21. R. R. Ramsay and K. F. Tipton, Molecules, 2017, 22, E1192.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

27

22. D. Griffith, J. P. Parker and C. J. Marmion, Anti-­Cancer Agents Med. Chem., 2010, 10, 354. 23. D. Griffith, M. Devocelle and C. J. Marmion, in Amino Acids, Peptides and Proteins in Organic Chemistry, 2010, vol. 2, p. 93. 24. C. J. Marmion, J. P. Parker and K. B. Nolan, in Comprehensive Inorganic Chemistry II (Second Edition): From Elements to Applications, 2013, vol. 3, p. 683. 25. W. K. Kelly, P. Marks and V. M. Richon, Clin. Cancer Res., 2015, 21, 2198. 26. B. S. Mann, J. R. Johnson, M. H. Cohen, R. Justice and R. Pazdur, Oncologist, 2007, 12, 1247. 27. J. J. McClure, X. Li and C. J. Chou, Adv. Cancer Res., 2018, 138, 183. 28. H. Z. Lee, V. E. Kwitkowski, P. L. Del Valle, M. S. Ricci, H. Saber, B. A. Habtemariam, J. Bullock, E. Bloomquist, Y. Li Shen, X. H. Chen, J. Brown, N. Mehrotra, S. Dorff, R. Charlab, R. C. Kane, E. Kaminskas, R. Justice, A. T. Farrell and R. Pazdur, Clin. Cancer Res., 2015, 21, 2666. 29. A. J. Moskowitz and S. M. Horwitz, Leuk. Lymphoma, 2017, 58, 1306. 30. R. M. Poole, Drugs, 2014, 74, 1543. 31. J. Vitfell-­Rasmussen, I. Judson, A. Safwat, R. L. Jones, P. B. Rossen, M. Lind-­Hansen, P. Knoblauch and A. Krarup-­Hansen, Sarcoma, 2016, 2016, 2090271. 32. H. J. Mackay, H. Hirte, T. Colgan, A. Covens, K. MacAlpine, P. Grenci, L. Wang, J. Mason, P. A. Pham, M. S. Tsao, J. Pan, J. Zwiebel and A. M. Oza, Eur. J. Cancer, 2010, 46, 1573. 33. K. Luger, M. L. Dechassa and D. J. Tremethick, Nat. Rev. Mol. Cell Biol., 2012, 13, 436. 34. A. Peserico and C. Simone, J. Biomed. Biotechnol., 2011, 2011, 371832, 10 pages. 35. M. S. Finnin, J. R. Donigian, A. Cohen, V. M. Richon, R. A. Rifkind, P. A. Marks, R. Breslow and N. P. Pavletich, Nature, 1999, 401, 188. 36. D. Griffith, K. Lyssenko, P. Jensen, P. E. Kruger and C. J. Marmion, Dalton Trans., 2005, 956. 37. C. Mulcahy, F. M. Dolgushin, K. A. Krot, D. Griffith and C. J. Marmion, Dalton Trans., 2005, 1993. 38. D. Griffith, S. Cecco, E. Zangrando, A. Bergamo, G. Sava and C. J. Marmion, JBIC, J. Inorg. Biol. Chem., 2008, 13, 511. 39. D. Griffith, E. Zangrando, E. Alessio and C. Marmion, Inorg. Chim. Acta, 2004, 357, 3770. 40. C. J. Marmion, T. Murphy, J. R. Docherty and K. B. Nolan, Chem. Commun., 2000, 1153. 41. D. Griffith, M. P. Morgan and C. J. Marmion, Chem. Commun., 2009, 6735. 42. O. A. Botrugno, F. Santoro and S. Minucci, Cancer Lett., 2009, 280, 134. 43. G. R. Leggatt and B. Gabrielli, Immunol. Cell Biol., 2012, 90, 33.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

28

Chapter 1

44. K. L. Jin, J. Y. Park, E. J. Noh, K. L. Hoe, J. H. Lee, J. H. Kim and J. H. Nam, J. Gynecol. Oncol., 2010, 21, 262. 45. V. Brabec, D. M. Griffith, A. Kisova, H. Kostrhunova, L. Zerzankova, C. J. Marmion and J. Kasparkova, Mol. Pharm., 2012, 9, 1990. 46. J. P. Parker, H. Nimir, D. M. Griffith, B. Duff, A. J. Chubb, M. P. Brennan, M. P. Morgan, D. A. Egan and C. J. Marmion, J. Inorg. Biochem., 2013, 124, 70. 47. http://www.who.int/medicines/publications/essentialmedicines/en/ (accessed 30/07/2018). 48. T. Eckschlager, J. Plch, M. Stiborova and J. Hrabeta, Int. J. Mol. Sci., 2017, 18( 7 ), 1414. 49. D. M. Griffith, B. Duff, K. Y. Suponitsky, K. Kavanagh, M. P. Morgan, D. Egan and C. J. Marmion, J. Inorg. Biochem., 2011, 105, 793. 50. C. T. Lin, H. C. Lai, H. Y. Lee, W. H. Lin, C. C. Chang, T. Y. Chu, Y. W. Lin, K. D. Lee and M. H. Yu, Cancer Sci., 2008, 99, 1218. 51. B. T. Benedetti, S. Quintal and N. P. Farrell, Dalton Trans., 2011, 40, 10983. 52. E. S. Ma, W. D. Bates, A. Edmunds, L. R. Kelland, T. Fojo and N. Farrell, J. Med. Chem., 2005, 48, 5651. 53. J. Yang, X. Sun, W. Mao, M. Sui, J. Tang and Y. Shen, Mol. Pharm., 2012, 9, 2793. 54. M. Alessio, I. Zanellato, I. Bonarrigo, E. Gabano, M. Ravera and D. Osella, J. Inorg. Biochem., 2013, 129, 52. 55. V. Novohradsky, L. Zerzankova, J. Stepankova, O. Vrana, R. Raveendran, D. Gibson, J. Kasparkova and V. Brabec, J. Inorg. Biochem., 2014, 140, 72. 56. V. Novohradsky, L. Zerzankova, J. Stepankova, O. Vrana, R. Raveendran, D. Gibson, J. Kasparkova and V. Brabec, Biochem. Pharmacol., 2015, 95, 133. 57. T. Iannitti and B. Palmieri, Drugs R&D, 2011, 11, 227. 58. R. Raveendran, J. P. Braude, E. Wexselblatt, V. Novohradsky, O. Stuchlikova, V. Brabec, V. Gandin and D. Gibson, Chem. Sci., 2016, 7, 2381. 59. V. Novohradsky, I. Zanellato, C. Marzano, J. Pracharova, J. Kasparkova, D. Gibson, V. Gandin, D. Osella and V. Brabec, Sci. Rep., 2017, 7, 3751. 60. M. Szyf, Curr. Oncol., 2008, 15, 72. 61. A. R. Z. Almotairy, V. Gandin, L. Morrison, C. Marzano, D. Montagner and A. Erxleben, J. Inorg. Biochem., 2017, 177, 1. 62. B. W. J. Harper, E. Petruzzella, R. Sirota, F. F. Faccioli, J. R. Aldrich-­Wright, V. Gandin and D. Gibson, Dalton Trans., 2017, 46, 7005. 63. C. A. Rouzer and L. J. Marnett, J. Lipid Res., 2009, 50(Suppl), S29. 64. R. E. Harris, B. C. Casto and Z. M. Harris, World J. Clin. Oncol., 2014, 5, 677. 65. K. Muller-­Decker and G. Furstenberger, Mol. Carcinog., 2007, 46, 705.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

Enhancing the Therapeutic Potential of Platinum-­based Anticancer Agents

29

66. N. Ghosh, R. Chaki, V. Mandal and S. C. Mandal, Pharmacol. Rep., 2010, 62, 233. 67. W. Neumann, B. C. Crews, L. J. Marnett and E. Hey-­Hawkins, ChemMedChem, 2014, 9, 1150. 68. W. Neumann, B. C. Crews, M. B. Sarosi, C. M. Daniel, K. Ghebreselasie, M. S. Scholz, L. J. Marnett and E. Hey-­Hawkins, ChemMedChem, 2015, 10, 183. 69. E. Ackerstaff, B. Gimi, D. Artemov and Z. M. Bhujwalla, Neoplasia, 2007, 9, 222. 70. A. M. Fajardo and G. A. Piazza, Am. J. Physiol.: Gastrointest. Liver Physiol., 2015, 309, G59. 71. A. Grancher, P. Michel, F. Di Fiore and D. Sefrioui, Bull. Cancer, 2018, 105, 171. 72. R. K. Pathak, S. Marrache, J. H. Choi, T. B. Berding and S. Dhar, Angew. Chem., Int. Ed. Engl., 2014, 53, 1963. 73. Q. Cheng, H. Shi, H. Wang, Y. Min, J. Wang and Y. Liu, Chem. Commun., 2014, 50, 7427. 74. U. Basu, B. Banik, R. Wen, R. K. Pathak and S. Dhar, Dalton Trans., 2016, 45, 12992. 75. Q. Cheng, H. Shi, H. Wang, J. Wang and Y. Liu, Metallomics, 2016, 8, 672. 76. J. Molnar and J. C. Somberg, Am. J. Ther., 2009, 16, 86. 77. R. N. Armstrong, Chem. Res. Toxicol., 1997, 10, 2. 78. D. Sheehan, G. Meade, V. M. Foley and C. A. Dowd, Biochem. J., 2001, 360, 1. 79. E. E. Ramsay and P. J. Dilda, Front. Pharmacol., 2014, 5, 181. 80. C. Peklak-­Scott, P. K. Smitherman, A. J. Townsend and C. S. Morrow, Mol. Cancer Ther., 2008, 7, 3247. 81. W. H. Ang, I. Khalaila, C. S. Allardyce, L. Juillerat-­Jeanneret and P. J. Dyson, J. Am. Chem. Soc., 2005, 127, 1382. 82. K. G. Z. Lee, M. V. Babak, A. Weiss, P. J. Dyson, P. Nowak-­Sliwinska, D. Montagner and W. H. Ang, ChemMedChem, 2018, 13, 1210. 83. M. O. James, S. C. Jahn, G. Zhong, M. G. Smeltz, Z. Hu and P. W. Stacpoole, Pharmacol. Ther., 2017, 170, 166. 84. M. G. Vander Heiden, L. C. Cantley and C. B. Thompson, Science, 2009, 324, 1029. 85. J. Li, M. Kato and D. T. Chuang, J. Biol. Chem., 2009, 284, 34458. 86. S. Kankotia and P. W. Stacpoole, Biochim. Biophys. Acta, 2014, 1846, 617. 87. P. W. Stacpoole, J. Natl. Cancer Inst., 2017, 109, djx071. 88. W. Liu, J. Su, J. Jiang, X. Li, Q. Ye, H. Zhou, J. Chen and Y. Li, Sci. Rep., 2013, 3, 2464. 89. W. Liu, J. Jiang, Y. Xu, S. Hou, L. Sun, Q. Ye and L. Lou, J. Inorg. Biochem., 2015, 146, 14. 90. S. Dhar and S. J. Lippard, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 22199.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00001

30

Chapter 1

91. X. Xue, S. You, Q. Zhang, Y. Wu, G. Z. Zou, P. C. Wang, Y. L. Zhao, Y. Xu, L. Jia, X. Zhang and X. J. Liang, Mol. Pharm., 2012, 9, 634. 92. E. Wexselblatt, E. Yavin and D. Gibson, Angew. Chem., Int. Ed. Engl., 2013, 52, 6059. 93. T. C. Johnstone, N. Kulak, E. M. Pridgen, O. C. Farokhzad, R. Langer and S. J. Lippard, ACS Nano, 2013, 7, 5675. 94. E. Petruzzella, J. P. Braude, J. R. Aldrich-­Wright, V. Gandin and D. Gibson, Angew. Chem., Int. Ed. Engl., 2017, 56, 11539. 95. A. D. Cox, S. W. Fesik, A. C. Kimmelman, J. Luo and C. J. Der, Nat. Rev. Drug Discovery, 2014, 13, 828.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Chapter 2

Ruthenium, Osmium and Iridium in the Fight Against Cancer Isolda Romero-­Canelón School of Pharmacy, Institute of Clinical Sciences, University of Birmingham, Birmingham B15 2TT, UK *E-­mail: [email protected]

2.1  Introduction Precious metals such as ruthenium, osmium and iridium are at the forefront of metal-­based anticancer drug research and they are of particular relevance in the fight against platinum resistance. Although cisplatin and its derivatives are used in more than 50% of all antineoplastic regimes, alone or in combination therapy, there is a critical requirement for treatments that are able to overcome inherent and acquired Pt-­resistance.1 Although targeted chemotherapeutics, based on organic small molecules with novel mechanisms of action (for example, kinase inhibitors), have been developed more recently, these often have a much narrower clinical utility than Pt-­based drugs and generally suffer from rapid onset of resistance, hence there is still a large unmet clinical need. Unlike the DNA-­based mechanism of action (MoA) of Pt-­based drugs, Ru, Os and Ir complexes, can exert their anticancer activity by the activation of a wide array of cellular pathways, which offer the prospect of combating   Metallobiology Series No. 14 Metal-­based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-­Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

31

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

32

Chapter 2

Pt-­resistance, reducing deleterious side effects and increasing the range of treatable malignancies. The development of bioinorganic chemistry, and in general the research into the use of metals in medicine, has already rendered two Ru complexes as clinical candidates for antineoplastic treatment. This has sparked great interest in similar structures with other metal centres such as Os and Ir. This chapter looks into the wide structural diversity that can be achieved with coordination complexes centred on Ru, Os and Ir. It further examines some of the most prominent aspects of the MoAs responsible for the antiproliferative activity and explores some of the physicochemical challenges of these investigations at the cellular level.

2.2  Structural Diversity The molecular diversity offered by Ru, Os and Ir complexes is vast and their flexibility should enable translation into a wide pool of drug candidates. In general, biological activities of complexes depend on the nature of the metal centre, in addition to its oxidation state, coordination number and geometry, and fine-­tuning of their biological activity can be achieved by small variations in the organic ligands. Taken together, these features influence the complexes' reactivity, as well as the kinetics and thermodynamics of ligand substitution that are so crucial for antiproliferative activity. Properties of ligand substitution were the reason behind the original development of Ru complexes as they are highly similar to those of Pt square planar complexes. Further advantages also include (a) an octahedral geometry, which allows for a wider range of reaction mechanisms involving the metal centre, (b) a more varied range of accessible oxidation states within biologically relevant conditions and (c) the ability of Ru to mimic Fe in protein-­ binding processes.2 Ru and Os complexes can often have almost identical structures; with many basic scaffolds also being able to accommodate Ir centres. However, ligand substitution reactions in the heavier congeners are usually orders of magnitude slower. Hence, with the aim of having wider control over kinetic and thermodynamic properties, the field has expanded to incorporate Os and Ir complexes as viable alternatives to the Pt-­drugs in the clinic.3 The following sections will explore the most common coordination spheres found in Ru, Os and Ir complexes that are being developed as anticancer agents.

2.2.1  Octahedral Coordination Complexes Octahedral complex fac-­[Ru(NH3)3Cl3] (Figure 2.1a), reported by Clarke in 1976, initiated the research into Ru anticancer agents after observations that the complex had the ability to inhibit the division of Escherichia coli.4 However, it was not until 1989 that the field achieved its first breakthrough when Keppler and co-­workers reported KP1019 5 (Figure 2.1b). This complex of molecular formula trans-­[RuCl4(Ind)2]IndH, where Ind = indazole, is isoelectronic with Clarke's complex but avoids its shortfalls regarding chemical stability

View Online

33

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

Figure 2.1  Octahedral  ruthenium(iii) complexes: (a) fac-­[Ru(NH3)3Cl3], (b) KP1019, (c) NAMI-­A, (d) [Ru(bpy)2(ab-­PBI-­R)] 2+,10 (e) [Ru(phen)2(dppz)]2+,11 (f) dinuclear Ru azo-­derivative reported by Chao and co-­workers.12

under biologically relevant conditions. The second breakthrough came in the form of NAMI-­A (Figure 2.1c), of molecular formula trans-­[RuCl4(DMSO) (Im)]ImH, where Im = imidazole, which was synthesised by the groups of Alessio and Sava.6,7 Other interesting octahedral ruthenium(ii) complexes are those that include biologically active ligands in their structure and use the metal centre as more than a scaffold for drug delivery, using it instead as a synergistic

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

34

Chapter 2

unit that would improve the biological profile of the antineoplastic agent. Such is the case for a bipyridine (bpy) complex [Ru(5-­FU)(PPh3)2(bpy)]PF6 that includes the clinically used anticancer drug 5-­fluorouracil (5-­FU),8 or the complex cis-­[Ru(bpy)2(L)](PF6)2 that combines the metal centre with an EGFR-­inhibiting 4-­anilinoquinazoline (L) derivative.9 Great effort has been made regarding the synthesis and characterisation of other octahedral polypyridyl complexes. Such structures are highly tuneable and offer great scope for the improvement of photo-­physical properties; this makes them ideal for the development of photosensitisers for photodynamic therapy.13,14 Examples of this are the complexes studied by the groups of Gasser and Chao. They include those of formula [Ru(bpy)2(ab-­ PBI-­R)](PF6)2, where ab-­PBI = azabenz-­annulated perylene bis-­imide and R = 2,6-­diisopropylphenyl or 3-­pentyl (Figure 2.1d), which localise in mitochondria and can achieve nanomolar IC50 values,10 and phenazine derivatives [Ru(phen)2(dppz)]2+ (Figure 2.1e).11 There are also examples of dinuclear azo derivatives (Figure 2.1f) that can be exploited in two-­photon photodynamic therapy after glutathione (GSH) activation.12

2.2.2  Organometallic Arene ‘Piano-­stool’ Complexes Organometallic piano-­stool complexes, which include in their structure an arene moiety, were originally reported by Tocher in 1992 and subsequently developed by the groups of Sadler and Dyson.15 These ruthenium complexes were initially designed to include a monodentate Cl ligand as an activation site that would, like cisplatin, hydrolyse prior to DNA binding. Nowadays, these complexes are of great interest in the bioinorganic field due to the chemical flexibility in their structure, and their potential is recognised beyond DNA interactions. Based on five main building blocks and of general formula [(arene)M(X)(Y)(Z)]n+, there is almost unlimited scope for design using the basic molecular structure shown in Figure 2.2. In the cases of interest in this chapter, M represents d6 ions ruthenium(ii), osmium(ii) or iridium(iii). The arene unit is used to adjust the lipophilicity of the complex and the monodentate ligand X still functions as an activation site. Interestingly, ligands Y and Z can be either monodentate or structurally linked to offer a bidentate ligand Y–Z.16 In either case, these ligands allow for the modification of thermodynamic and kinetic parameters of the activation of the M–X bond. The biological activity and potential for drug development of such complexes results from an intricate relationship between all these building blocks.3

2.2.2.1 The Arene Unit Frequently used arene units include p-­cymene and biphenyl in ruthenium(ii) and osmium(ii) complexes, while Cp* and its derivatives are used for iridium(iii) (Figure 2.2), as the negative charge of the arene ligand is relevant in stabilising the oxidation state of the metal centre. The relationship between

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

35

Figure 2.2  General  structure of organometallic ‘piano-­stool’ complexes. this structural unit and the water solubility of the complexes is well established; however, its influence on the biological behaviour and antineoplastic activity is still under discussion.17 In principle, lipophilic complexes, obtained by extension of the arene unit, should be able to accumulate in intracellular compartments, and, if concentration is directly related to biological activity, then these three parameters (lipophilicity, accumulation and activity) should correlate linearly. Hence, it would be expected that, for example, a ruthenium(ii) biphenyl complex, should be more potent than its p-­cymene analogue, as long as all other structural units remain unchanged. One such correlation can be observed in a series of iridium(iii) complexes of general formula [(arene) Ir(phen)Cl]+, where phen = 1,10-­phenanthroline. The Cp* derivative (Figure 2.3a) is inactive towards ovarian cancer cells. Nonetheless, sequential increments in the arene unit from Cp* to Cpph and then to Cpbiph renders complexes with IC50 values of 6.7 and 0.7 µM, respectively (Figure 2.3b and c). The elongation of the aromatic moiety increases the extent of hydrolysis of the Ir–Cl bond and although the reaction slows down, the acidity of the aqua species increases.18 Such a relationship between increased lipophilicity and higher anticancer potency is not always straightforward, as there is more complexity to the biological activity of a metal-­based complex than simple cellular accumulation.

2.2.2.2 The Monodentate Ligand X Even though the impact of the arene unit and the nature of the Y, Z ligands have been widely investigated, the repercussions for modifications of the monodentate X ligand are still in their infancy. As per the original design concept, the monodentate ligand X was only relevant based on its ability to

View Online

Chapter 2

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

36

Figure 2.3  Influence  of the different building blocks in organometallic ‘piano-­ stool’ complexes: influence of the arene unit extension (a–c), variations in the monodentate ligand X (d–i) and variations in the Y,Z ligands ( j–o).

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

37

generate an aqua-­ or a hydroxo-­derivative. Only such species could then generate the coordinate vacancy that would allow for DNA binding. Nonetheless, deeper understanding in the varied MoAs of these complexes has revealed that inclusion of a hydrolysable monodentate ligand X is no longer an indispensable requirement. Remarkably, in the case of the Ru kinase inhibitors developed by Meggers, the non-­lability of the monodentate –CO ligand is key in its MoA, as it allows the complex to reach intracellular targets structurally unchanged (Figure 2.3d).19 In the case of ruthenium(ii) complexes, variations in the monodentate ligand include: halogens, acetonitrile,20 P-­bound sugar derivatives21 and carbenes,22,23 as well as derivatives in which the monodentate ligand X is really an extension arm of the arene unit.24,25 The X ligand exchange in osmium(ii) complexes has been studied less to date. Nonetheless, drawing similarities in the behaviour of Ru and Os complexes is still possible. Imino pyridine derivatives [(p-­cymene)Ru(NMe2-­impy)X]PF6 (Figure 2.3e), where X varies from Cl to I, show contrasting cellular uptake pathways, with the iodide derivative exhibiting a higher percentage of passive diffusion into A2780 ovarian cancer cells.26 Interestingly, a comparison between these complexes and their arylazo analogues (Figure 2.3f), where M = RuII or OsII and X = Cl or I, showed that the monodentate changes not only influence hydrolysis and activation of the metal complexes but also induce differences in cellular accumulation and compartmentalisation with consequences in cellular behaviour in relation to activity, selectivity and p53  dependence.27 Pyridine derivative ZL105 [(Cpbiph)Ir(ppy)(py)]+, where ppy = 2-­phenyl­pyridine, (Figure 2.3g) outperforms its chlorido parent complex [(Cpbiph)Ir(ppy)Cl]  (Figure 2.3h) both in potency and selectivity towards cancer cells. A possible reason for this relies in the variation in the hydrolysis rate of the Ir–Cl vs. Ir–py bond and its influence on the complex's ability to reach its intracellular  target (see Section 2.2.3).28 The use of substituted pyridine derivatives has also been reported.29 Interestingly, iridium hydrosulfide derivatives with general formula [(arene)Ir(N,N-­)(SH)]PF6 (Figure 2.3i) are also highly active towards cancer cells.30

2.2.2.3 The Y and Z Ligands Variations in the donor atoms of the Y and Z ligands, and, in general, in their electronic properties have strong repercussions on the M–X bond and the hydrolysis of the complex. At one end of the spectrum, aryl azopyridine derivatives exhibit a slow rate of hydrolysis and a low pKa of the aqua-­derivative, while at the other end, acetyl acetonate derivatives show fast hydrolysis rates and high pKa values. Such observations are transferable between ruthenium(ii) and osmium(ii) derivatives, as shown in Figure 2.4.31 A highly studied ruthenium(ii) ‘piano-­stool’ complex bearing two independent monodentate Y and Z ligands is RAPTA-­C (Figure 2.3j) developed by the Dyson group. This complex contains two Cl atoms and a P-­bound 

View Online

Chapter 2

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

38

Figure 2.4  Comparison  of the rate of hydrolysis and pKa of the aqua-­derivatives in ruthenium(ii) and osmium(ii) ‘piano-­stool’ complexes.

1,3,5-­triaza-­7-­phosphaadamantane (PTA) unit as the third monodentate ligand, which is highly relevant for achieving water solubility. The basic structure of  this complex has been systematically investigated resulting in highly active complexes such as RAPTA-­B and RAPTA-­T. In the former, the arene unit is a benzene and in the latter it is a toluene. Other reports include ‘piano-­stools’ that keep the PTA moiety constant and change Y and Z as monodentate or bidentate ligands.32 Probably the most well-­known ‘piano-­stool’ ruthenium(ii) complex which involves a Y–Z chelating ligand is RM175 developed by the Sadler group (Figure 2.3m).33 This complex, which progressed to pre-­clinical trials, exhibited great activity towards mammary carcinomas in particular. Taking into consideration such potential, a similar derivative was synthesised using osmium(ii) as the metal centre which was found to be inactive.34 This observation can be justified taking into consideration that the MoA of RM175 is based on DNA interactions after hydrolysis and the marked differences in the kinetics

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

39

of aquation of both derivatives. Such differences in ligand exchange reactions have also been observed in other series of arene complexes.35 Chelating N,N-­ligands include structurally simple saturated amines,36,37 aromatic polypyridyl rings,38 as well as, imino-­ and azo-­pyridine39,40 derivatives. In particular, N-­dimethylphenyl azopyridine has been used with the three metal centres of interest, generating highly potent complexes.40–42 Osmium(ii) complexes where the bidentate ligand Y–Z changes between an iminopyridine and an arylazo pyridine derivative43 (Figure 2.3d and f) have shown contrasting chemical reactivity. Unlike the latter complexes, the former binds to the nucleotide base guanine after undergoing aquation, and oxidises the coenzyme nicotinamide adenine dinucleotide (NADH).44 N,O-­ ligands involve ketoamines or picolinic acids45,46 while O,O-­chelators incorporate β-­dicarbonyl derivatives. Such series of osmium(ii) derivatives show how variations in chemical stability regarding hydrolysis and acidity of the aqua adduct influence anticancer activity (Figure 2.3k and l).36,47,48 Other relevant coordination spheres are N,S-­, N,C-­,49 P,P-­50 and even C,C-­ using carbene type ligands.51 N,S-­coordination complexes include N-­ substituted 2-­pyridinecarbothioamide derivatives, [(p-­cymene)Ru/Os(N,S-­) Cl]Cl (Figure 2.3n) which offer promise for further research into orally active complexes,52,53 while N,C-­coordinations comprise cyclometalated complexes such as [(arene)Ru/Ir(N,C-­)Cl] (Figure 2.3o) reported by Ruiz and co-­workers.54 Variation from N,N-­ to N,C-­chelators, has a great impact on the anticancer potency of iridium(iii) complexes. For example, the N,N-­phenanthroline complex [(Cp*)Ir(phen)Cl]+ is inactive towards A2780 ovarian cancer cells, while the highly structurally related N,C-­bound phenyl pyridine complex [(Cp*)Ir(ppy)Cl] has an IC50 of 10.8 µM in the same cell line.55,56 The high flexibility of the scaffold of organometallic ‘piano-­stools’ is further exploited in multinuclear complexes in which the Y–Z ligand includes more than one bidentate coordination position. An example of this strategy is the O,O-­derivatives of general formula [(p-­cymene)Ru(O,O-­C6H5O2N(CH2)nNC6H5O2-­ O,O)Ru(p-­cymene)] reported by Hartinger and co-­workers. Another example of the use of a ‘piano-­stool’ base structure to generate multinuclear structures is the synthesis of metallacages. These types of complex have the advantage that they can be further developed and used as Trojan horses for the delivery of highly hydrophobic cargos into cancer cells.4,57,58 The chemical space of ‘piano-­stool’ complexes has also been used to incorporate ligands, which per se exhibit biologically relevant activities. For example, the group of Meggers has used staurosporine derivatives in which the carbohydrate unit has been exchanged for the fragment containing the metal centre, more specifically ruthenium(ii), in order to design kinase inhibitors.59 Multidrug-­resistance modulators derived from phenoxazine and anthracene moieties have also been included, with the aim of developing complexes that would be capable of overcoming drug resistance by P-­glycoprotein inhibition.4

View Online

40

Chapter 2

2.2.3  Other Relevant Coordination Spheres

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

2.2.3.1 Cyclometalated Octahedral Complexes Exploiting their outstanding luminescent properties, cyclometalated complexes, in particular with ruthenium(ii) and iridium(iii) centres, are being developed as viable alternatives for photodynamic and photo-­activated chemotherapy, as well as for theranostic and imaging agents.60 These types of metal complexes include structures such as those reported by the groups of Mao,61 Huang62 and Ruiz.63

2.2.3.2 Ferrocifen Derivatives Ru and Os analogues of ferrocifen are very promising.64,65 In such structures, the metal sandwich complex is linked to a tamoxifen derivative. The parent compound, ferrocifen, which has been extensively investigated by Jaouen's group, acts as a selective oestrogen receptor modulator and is highly active towards cancer cells. Remarkably, the complex exhibits antiproliferative activity towards both hormone-­dependent and hormone-­independent breast cancer cells, with good in vitro and in vivo potency.66

2.2.3.3 Carbene Complexes and Their Derivatives Metal-­based complexes that include in their structures a metal–carbene bond are known for their promising antimicrobial activity, but they are also being investigated for their potential use as anticancer agents. In particular, Ru derivatives related to Grubbs and Hoveyda–Grubbs catalysts have shown moderate inhibitory enzymatic activity.67 Further examples are being developed by the groups of Metzler-­Nolte and Ott.68 The former include mono-­ and bis-­NHC–Ir(i) complexes, where NHC = N-­heterocyclic carbene.69–71

2.3  M  echanisms of Action of Ru, Os and Ir Metal Complexes As mentioned in previous sections, two lead Ru(iii) complexes have entered clinical trials, KP1019 and NAMI-­A (Figure 2.1b and c, respectively). Although these two complexes have great structural similarities with regard to oxidation state, coordination number and geometry, they exhibit very different biological properties. NAMI-­A is directed to control tumour metastasis, while KP1019 and its sodium salt NKP1339 are both active towards primary tumours. It is believed that ruthenium(iii) centres are activated by intracellular reduction to ruthenium(ii) by reaction with ascorbic acid or GSH. However, the similarities in the cellular fate of these complexes ends there. There is no evidence to suggest that KP1019 or NKP1339 have DNA as their cellular target, and it is accepted that their MoA is tightly linked to alterations

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

41

in the cellular redox state. In particular, by causing protein damage and ER stress, which leads to G2/M cell cycle arrest and apoptosis.72 Meanwhile, NAMI-­A can inhibit angiogenesis induced by vascular endothelial growth factor. Also, this complex is believed to interact with actin-­t ype proteins on the cell surface which leads to reduced mobility of invasive cancer cells and hence the observed antimetastatic activity.7,73–75 Such contrast in the MoA of highly chemically related ruthenium(iii) complexes shows the wide diversity of cellular events that can be activated by small structural changes. Hence, research into the cellular basis of the MoA of Ru, Os and Ir complexes is of great relevance when developing potential drug candidates. Although, intense efforts have been directed to understanding the structure activity relationship of such complexes, extrapolation of general rules is not always straightforward.

2.3.1  DNA as a Target As previously mentioned, ‘piano-­stool’ complexes were originally designed to have cellular behaviour similar to that of cisplatin. The monodentate ligand would allow hydrolysis activation prior to DNA binding, and an extended arene unit would provide further scope for double helix intercalation. Hence, early studies were highly focused on DNA interactions, and although the field has advanced to include other cellular targets and research has diversified into a variety of MoAs, DNA interaction is still an attractive goal. A prime example of such complexes is RM175 (Figure 2.3m). This complex undergoes activation by hydrolysis followed by the generation of an aqua-­ or hydroxo-­derivative (Figure 2.5a), and then selectively binds to the N7 of guanine (Figure 2.5b), as well as intercalating into the minor groove of the double helix, given the extended arene unit. These DNA interactions have been extensively investigated by the groups of Sadler20 and Brabec. More importantly research has been extended to other ruthenium(ii), osmium(ii)

Figure 2.5  DNA  as a target: hydrolysis (a) and subsequent nucleobase binding (b) of ruthenium(ii) arene complex RM175.

View Online

42

Chapter 2

and iridium(iii) ‘piano-­stool’ complexes bearing N,N-­, N,O-­ and O,O-­chelating ligands in their structure (Figure 2.3k and l).36,38,40,47,76–83

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

2.3.2  Redox Modulation and Mitochondrial Targeting Another highly relevant target for metal-­based complexes, with antineoplastic activity, is the mitochondria in cancer cells. Although they do not represent a classical or long-­studied target in metals in the medical field, they are certainly gaining momentum and recent developments are highly promising. The resurgence of the Warburg effect in other aspects of cancer research84,85 and the prospect of cellular selectivity have undoubtedly sparked a remarkable interest in the development of complexes that can exploit mitochondrial dysfunction by taking advantage of the redox properties not only of metal centres per se, but of the complexes as a whole.86 It is well established that metal Pt-­drugs currently in clinical use target DNA. Their most prominent means of selectivity is the fast duplication rate of tumour cells, which as a consequence has pronounced deleterious side effects on other rapid proliferating cells.87 Targeting mitochondria instead offers the possibility to minimise this drawback. Mitochondria in cancer cells, as the machinery responsible for keeping up with the high energy demands of rapid proliferation, are highly susceptible to redox changes. Such changes, even small, place an excess burden on an organelle that is, in most cancer cases, defective and already dealing with elevated levels of intrinsic reactive oxygen species (ROS).88 Hence, designing a metal-­based complex that can maximise the redox imbalance would, in principle, allow for targeted disruption of cancer cell metabolism while providing selectivity towards normal cells.89 This strategy takes advantage of mitochondrial dysfunction, a frequently occurring event in cancer cells that may involve oncogenic stress, mitochondrial DNA mutations and redox hyperactivity.90–92 This redox imbalance, and manipulation of ROS levels in general,86,93 can be achieved by several means, such as directly generating an excess of ROS, lowering the levels of ROS-­controlling molecules such as GSH or even disrupting delicate equilibria like NAD(P)+/NAD(P)H, GSH/GSSG or even pyruvate/ lactate, amongst others.94 Even more rewarding would be the development of a multi-­targeted agent that could modify several aspects of this cellular redox balance.95 Such a strategy would reduce the possibilities of resistance developing in a clinical setting. Sadler and co-­workers have pioneered this research, starting from the development of ruthenium(ii) complexes with O,O-­bidentate ligands that could catalyse GSH dimerisation and did not have DNA as their cellular target.42 This research has expanded to include analogue structures bearing osmium(ii) and iridium(iii) metal centres. Osmium(ii) complexes with general formula [(arene)Os(N,N-­)X]PF6 are of particular interest. These compounds are more active in vitro than cisplatin and oxaliplatin in A2780 ovarian and HCT116 colorectal cancers and are equally potent in Pt-­resistant and

View Online

Ruthenium, Osmium and Iridium in the Fight Against Cancer

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

27,43

43

p53-­deficient cell lines. They are also more selective than Pt-­based drugs towards cancer cells over normal lung fibroblasts and breast epithelial cells, suggesting a potential reduction in off-­target toxicity.96 Such positive results have lead to further studies, particularly based on the lead compound FY26 [(p-­cymene)Os(NMe2-­azpy)I]PF6 (Figure 2.3m). Complex FY26, as an iodido derivative does not readily hydrolyse unlike its chlorido analogue. In addition, it is stable in phosphate-­buffered saline and blood serum. Nonetheless, exposure of MCF7 cancer cells to the complex leads to a quick efflux of iodine. This observation was made after the synthesis of a 131I (β−/γ emitter, t1/2 8.02 d) radiolabelled derivative and investigations of its cellular accumulation, with independent Os/I monitoring. Efflux of the monodentate X ligand, while the metal centre remained in intracellular compartments, indicated that the complex was activated upon entering the cancer cells. Deeper investigations revealed that the osmium(ii) complex can form hydroxido derivatives (X = –OH) upon activation by intracellular GSH and subsequently generate thiolato and sulfenato adducts. These reactions – in a test tube – have been proven to produce radical species (monitored by electron paramagnetic resonance, EPR) that are consistent with observations of ROS generation in vitro.97 Once in the intracellular space, complex FY26 localises preferentially in mitochondria98 allowing for a MoA that is highly related to perturbation of mitochondrial function and can exploit mutations present in cancer cells. Hence, the proposed MoA of this complex is based on the induction of ROS, more particularly on a burst of superoxide. Under similar conditions, MRC5 fibroblasts do not generate relevant ROS levels and the low induction of oxidative stress observed is rapidly balanced by the fibroblasts – which may be underpinning the selectivity of the osmium complex. In vivo activity is also very promising. A single injection of FY26 in mice at either 0.25x or 1x its maximum tolerated dose induced statistically significant tumour growth delay in HCT116 xenograft models, with the latter showing a mean tumour doubling time per day of 6.2, compared to 6.4 for cisplatin and 3.9 for the untreated control. Achieving efficacy similar to cisplatin is an impressive result for a single iv bolus of an un-­optimised drug, and, based on these data, it is envisaged that repeat dosing of osmium drugs over a period of time (i.e. akin to a chemotherapy regime) would have a significant and lasting anti-­ tumour effect.99 FY26 was also used to demonstrate that it is possible to use cutting-­edge systems biology approaches to investigate the MoA of metal-­based complexes. This report was the first study in which whole-­cell RNA-­sequencing was used to establish a time-­dependent overview of the genetic response of ovarian cancer cells to an osmium chemotherapeutic agent.100 Both transcriptomic and proteomic studies suggest an attack on glycolysis which switches energy production towards oxidative phosphorylation. It also demonstrates the crucial relationship between a burst of cellular superoxide and the potency of metal complexes. In particular, complex FY26 is highly lipophilic

View Online

Chapter 2

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

44

Figure 2.6  NRF2-­  pathway mediated oxidative stress response activated by FY26 in A2780 ovarian cancer cells.

and positively charged, hence its potential to accumulate in mitochondria.98 The RNA-­sequencing studies have highlighted that the NFR2 oxidative stress response pathway is activated within 4 h of drug exposure (Figure 2.6). Analysis of the differential up-­ and downregulation of components in this pathway, shows that some antioxidant enzymes are downregulated, highlighting the deficiencies in stress response in A2780 ovarian cancer cells. Furthermore, the observed upregulation of superoxide dismutase (SOD) suggests that both superoxide and hydrogen peroxide play an important role in the anticancer activity of this metal-­based agent. The raised ROS levels, and consequent mitochondrial dysfunction, initiate a multi-­targeted cytotoxicity, leading to apoptotic cell death. Even though this work was ground breaking in the field, there is still much to be explored regarding the pathways activated by such complexes.100 These RNA-­sequencing results and their impact on the investigation of the MoA of complex FY26 paved the way for similar analyses, such as in the case of complex ZL109 [(Cpph)Ir(NMe2-­azpy)I]PF6 which bears the same N,N-­ chelating ligand.101 Pathway analysis showed similar activation of the NRF2 transcription factor, and its antioxidant response pathway. CAT (catalase) and EPHX (epoxide hydrolase) genes are downregulated, while GSR (glutathione reductase) and NQO (NADH(P) quinone oxidoreductase) are both upregulated. Treatment with this complex also showed the characteristic burst of superoxide and other ROS in A2780 cancer cells. Comparison of the cellular response to FY26 and ZL109, with regard to apoptosis induction, showed that both complexes induce high levels of inhibitor of apoptosis proteins.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

45

Other comparisons between these two complexes, FY26 and ZL109, involved analysis of data from the National Cancer Institute NCI-­60 screen and the 916-­cell line Sanger Screening. The former data set revealed that both compounds are particularly active in cell lines COLO25 and MDA-­ MB-­468, which is consistent with both cell lines having reduced levels of GSTP1 (glutathione S-­transferase P1) and as a consequence being redox deficient.101,102 Analysis of the antiproliferative activity in the cell lines of the Sanger Screening, which are part of the Sanger Cancer Genome Project, showed that ZL109 has selectivity towards malignancies of testicular, urogenital and cervical origin. In previous investigations, it has been shown that the ZL109 MoA is independent of p53 status and in this particular breast cancer such a mutation is present, which may be related to the complex's high potency. This is also consistent with p53 being mutated in MDA-­MB-­468.101 A growing strategy to indirectly target mitochondrial function by altering the redox state in cells relies on the inhibition of redox-­related enzymes such as thioredoxin reductase (TrxR), superoxide dismutase (SOD), catalase, peroxidase, glyceraldehyde 3-­phosphate dehydrogenase (GAPDH), to name but a few. Examples of such enzyme inhibition include the use of polypyridyl complexes, this is the case for [Ru(diimine)3](ClO4)2 (Figure 2.7a) which exerts its antiproliferative activity towards A375 human melanoma cells by inhibition of TrxR; it also induces G0/G1 arrest and upregulation of oncogene p53.103 The inhibition of such a relevant target has also been investigated using osmocenyl derivatives that contain tamoxifen-­like moieties in their structure (Figure 2.7b).104 In the case of the inhibition of glutathione S-­transferase, osmium(ii) complexes have been developed using ethacrynic acid-­modified ligands. Figure 2.7c shows one of these complexes, in which ligands Y and Z are monodentate and X is a PTA.105 In comparison, the complex in Figure 2.7d utilises a bidentate ethacrynic derivative in positions Y and Z while monodentate ligand X is a chloride atom.106 Octahedral cyclometalated complexes have also been intensively studied as mitochondria-­targeting agents. An example of this is the complex [(ppy)2Ir(BTPC)]PF6, where BTPC = 2-­byclyclo[2.2.1]hept-­5-­en-­yl-­1H-­1,3,7,8-­ tetraazacyclopenta[l]phenanthrene (Figure 2.7e) which can increase the levels of intracellular ROS and induce a decrease in the mitochondrial membrane potential in SGC-­7901 cells, as well as causing cell cycle arrest in the G0/G1 phase and subsequent activation of autophagy. Research has also indicated that the complex upregulates Bcl-­2, while downregulating Bcl-­x and procaspase 7.107 A similar chemical scaffold has been investigated in depth by Mao and co-­workers to obtain coumarin-­108 and guanidinium-­109 derived complexes (Figure 2.7f and g, respectively). They not only induce mitochondrial damage by increasing ROS levels but in some cases can also be used as theranostic agents. Such is the case for a series of fluorinated phenanthroline derivatives (Figure 2.7h), which allow for real-­time tracking of mitochondrial damage by confocal microscopy.110

View Online

Chapter 2

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

46

Figure 2.7  Examples  of complexes that target mitochondria.

2.3.2.1 In-­cell Catalysis: A Prime Example of Redox Modulation Redox modulation in cancer cells, as a means of antiproliferative activity, could also be achieved by exploiting catalytic reactions. Metal-­based complexes are well known as catalysts, and taking into consideration the wealth of knowledge around their use in industry, exploiting such catalytic properties for rational drug design may just be a natural progression. In both cases the thermodynamics and kinetics of ligand substitution have to be tightly

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

47

controlled; hence, the field of metals in medicine could have a lot to learn from material chemists.31,111 One early report suggesting a link between the catalytic activity of a complex and its antiproliferative activity used derivatives of general formula [(arene)Ru(azpy)I]PF6, where the arene unit varied between p-­cymene (Figure 2.3f) and biphenyl, while the bidentate ligand included several aryl azopyridine derivatives. Although these complexes did not undergo hydrolysis, they were active against A549 lung cancer cells, which excluded the DNA-­targeting theory that had been previously established. Their antiproliferative activity was justified with the help of two key observations (a) the complexes were able to – in a test tube – catalyse the dimerisation of GSH into GSSG and (b) ROS were detected in A549 cells. Based on this, the authors proposed a catalytic cycle in which GSH dimerises and an equivalent of H2O2 is produced (Figure 2.8a).42 Later research exploited a catalytic transfer hydrogenation reaction to generate NADH from NAD+ using sodium formate as a hydride source; interestingly, other reports are concerned with the reverse reaction. In this regard ruthenium(ii) complexes with N,N-­chelating ligands 2,2′-­bypyrimidine or phenanthroline are used to observe the generation of a key Ru-­hydride species (Figure 2.8b). Iridium(iii) analogue complexes were further used to determine the turnover number (TON) and turnover frequency (TOF) of the catalytic reaction. When administered to A2780 cancer cells, the complex [(Cpxph)Ir(phen)Cl]PF6 doubled the ratio of cellular NAD+/NADH, varying from 7.95 to 14.87. This change in the redox couple was then associated with the observed antiproliferative activity.112 A close derivative, [(Cpxbiph)Ir(phen)Cl]PF6, is also catalytically active and shows submicromolar activity in vitro. This complex has been investigated through the NCI-­60 cancer cell screen, rendering a mean GI50 value of 2.34 µM and a mean LD50 of 31.62 µM (corresponding values for cisplatin, 1.49 µM and 44.00 µM, respectively). Although the COMPARE algorithm analysis of these data confirms polypharmacology, it is most interesting that drug-­ exposed cells reveal changes in cellular morphology consistent with mitochondrial damage and activation of apoptosis.41 The antiproliferative activity of iridium(iii) complexes is greatly improved by changes to the monodentate ligand X. Such is the case for complexes  [(Cpxbiph)Ir(ppy)X]PF6, where X varies from Cl to pyridine (Figure 2.3g and h) the  halogenated derivative exhibits an IC50 value of 0.7 µM in A2780 ovarian cancer cells and a four-­fold selectivity between normal and cancer cells. In contrast, the pyridine derivative is more potent, with a corresponding IC50 value of 0.12 µM and a 13-­fold selectivity. The two complexes generate high levels of ROS and superoxide in drug-­exposed cancer cells, and this induction is proportional to the differences in their antiproliferative activities. Although at very different rates, both complexes undergo hydrolysis and – in a test tube – both can catalyse the generation of H2O2 by reaction with NADH. Interestingly, there are also marked variations in the reaction rate, with GSH quenching the generation of peroxide (Figure 2.9). Such differences in the reaction

View Online

Chapter 2

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

48

Figure 2.8  In-­  cell catalysis: a prime example of redox modulation as a means of antiproliferative activity. Catalytic dimerisation of GSH (a), ruthenium and iridium transfer hydrogenation catalysts (b), Noyori-type transfer hydrogenation catalysts (c) and asymmetric catalysts (d).

kinetics, because of ligand lability, have a great influence on the antiproliferative activity. The less reactive pyridine derivative reaches the intracellular target site intact and it is then that ROS are produced. The chloride complex, in contrast, is rapidly deactivated by reaction with GSH.28 In these reports concerning the MoA of iridium(iii) complexes, there are two separate bodies of experimental data. On the one hand, the characterisation of the catalytic systems in a test tube is very well documented, and

View Online

49

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

Figure 2.9  Generation  of H2O2 by iridium(iii) complexes and their reaction with GSH.

on the other hand, the biological evidence supports involvement of redox modulation in the antiproliferative activity. However, up to this point there was no proof of a link between the two. The first evidence for intracellular or in-­cell catalysis stems from research into Noyori-­t ype ruthenium(ii) catalysts, which were originally used to reduce ketones by transfer hydrogenation reactions. Such reactions involve the donation of hydride from a sacrificial donor and can target a wide range of substrates. Regioselective catalytic reduction of NAD+ was observed using complexes of general formula [(arene)Ru(R-­en)Cl] where the arene unit varied between p-­c ymene and o-­therphenyl, and the R-­en ligands were all sulfonamides with N,N-­ coordination sites (Figure 2.8c). Catalytic conversion of NAD+ to NADH was conducted – in a test tube – using sodium formate as a hydride source and corresponding values for TON and TOF were determined. In principle, all

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

50

Chapter 2

complexes showed antiproliferative activity towards A2780 cancer cells. However, the most relevant data were obtained when the cells were exposed to the drug by co-­administration with sodium formate. In all cases, the potency of the complexes markedly increased. Importantly, potentiation of the complexes was not observed when co-­administered with sodium acetate, as it cannot act as a hydride source. Metal cellular accumulation remained unchanged regardless of the co-­administration of either formate or acetate. Further studies also demonstrated that the variation in the NAD+/NADH ratio exposed to a fixed concentration of the ruthenium(ii) complex was linearly related to changes in the formate concentration (non-­toxic doses ranging from 0 to 2 mM). Such experiments provided the link between the catalytic efficiency in the test tube and the changes in the antiproliferative activity of the complex.113 Noyori catalysts, with ruthenium(ii) metal centres, have been developed for asymmetric transfer hydrogenations and their catalytic cycle involves the formation of an unstable 16e− intermediary. Nonetheless, such an intermediary becomes stable and can be isolated when the metal centre is changed to the heavier congener osmium(ii).114 Isomers R,R-­ and S,S-­ of catalysts of general formula [(arene)Os(TsDPEN)], where TsDPEN = N-­(p-­toluenesulfonyl)-­ 1,2-­diphenylethylenediamine have been used to further corroborate that in-­cell catalysis can be exploited as a means of antiproliferative activity. The catalytic properties of both isomers were investigated using the reduction of acetophenone, their TOFs were determined and, most importantly, the enantiomeric excess generated was confirmed to be above 94% in all cases. Although acetophenone might not seem to be a biologically relevant substrate of choice to evaluate these catalysts, it is indeed a standard test for the investigation of transfer hydrogenation reactions in the materials field. Hence its relevance in confirming the activity and robustness of the osmium(ii) complexes. It was then hypothesised that, using both catalytic isomers, it was possible to follow the generation of lactate from pyruvate after catalytic conversion. This reaction was followed – in a test tube – using NMR studies. The stereochemistry and enantiomeric excess of the conversion, which is highly dependent on the concentration of formate, was then determined using enzymatic enantioselective assay kits. The next step was to investigate this reaction in A2780 ovarian cancer cells. Critically, as only the R,R-­isomer would have been able to generate a non-­natural form of lactate, the detection of d-­lactate in cell lysates corroborated beyond doubt that an in-­cell catalytic reaction of transfer hydrogenation had taken place (Figure 2.8d).115 Although these are prime examples of catalysis at the cellular level as a means for antiproliferative activity, there is alternative research that pursues other examples of in-­cell catalysis.116 Particularly relevant work has been carried out by Ward looking into artificial metallo-­enzyme activity and in vivo catalysed new-­to-­nature reactions.117–120 Metallo-­enzyme activity investigated includes reactions of hydrogenation, transfer hydrogenation, bond formation, C–H activation, sulfoxidation and C–H oxidation.120 In contrast,

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

51

new-­to-­nature reactions are being exploited in the medicinal chemistry and synthetic biology fields. In this case, abiotic metals allow the catalytic reactions of allyloxycarbonyl and propargyloxycarbonyl cleavage in cellular compartments. They can also intervene in amide bond formation, Heck's reactions or even in Suzuki–Miyaura cross-­couplings.117

2.3.2.2 Intelligent Combination Therapy: A Way of Maximising the Potential of Metal-­based Therapy? Achieving antiproliferative activity by means of cellular redox modulation opens up great possibilities for developing intelligent combination therapy. The majority of current combination regimes are determined by the choice of two or more drugs, which preferably do not overlap in terms of their MoA. This approach results in a regime that may show very good additive effects in comparison to the individual components. However, such choices do not fully maximise the potential of either drug. In contrast, intelligent combination therapy is guided not by additive, but by synergistic interactions in which the MoA of one component aids, enables or enhances the MoA of the second drug. An example of this strategy is the use of redox modulators in combination with metal-­based complexes, which not only potentiate the antiproliferative activity, but also dramatically increase the selectivity for cancer cells over normal cells. In addition to providing novel therapies for parental cancers, manipulation of the cellular levels of ROS may also provide a highly effective strategy for treating Pt-­resistant cancers. Organometallic arene complexes for which the induction of ROS is part of their MoA may benefit from co-­administration with chemicals that can allow the reduction of cellular redox quenchers, such as l-­buthionine sulfoxamine (l-­BSO). It selectively inhibits γ-­glutamylcysteine synthetase, an enzyme required in the first step of GSH synthesis, hence allowing the reduction of tripeptide levels in cancer cells.95 The arene osmium(ii) complex FY26 is highly active in vitro and its MoA involves bursts of superoxide (see Section 3.2). Its IC50 in A2780 ovarian cancer cells is 160 nM. However, this value can be reduced to 69 nM when co-­administered with a non-­toxic dose of 5 µM l-­BSO. Importantly, the selectivity between normal and cancer cells is improved from 28-­ to 63-­fold, which could potentially be translated into a reduction of side effects in the clinical setting; the corresponding value for cisplatin under the same conditions is a 10-­fold improvement. Deeper investigations into the cellular basis for this observation revealed that the co-­administration of the redox modulator does not cause modification in the cell cycle of cancer cells, neither does it further induce higher levels of ROS or an earlier onset of apoptosis. On the contrary, 5 µM l-­BSO reduces 50% of the GSH pool available, hence, although it does not increase the damage caused by FY26, it removes the possibility of the cancer cell repairing such damage.96 Similar results have been obtained using structurally related ruthenium(ii), osmium(ii) and iridium(iii) arene complexes.95

View Online

52

Chapter 2

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

2.4  C  hallenges in the Investigations of Mechanisms of Action at the Cellular Level Metal-­based complexes with antineoplastic properties are, in the majority of cases, multi-­targeted and, although advantageous with regard to the reduced possibility of developing resistance, that creates a major challenge when investigating their behaviour at the cellular level. Such affirmation is supported by the seemingly high reactivity and promiscuity of metal complexes; hence, defining a single target for a given complex and subsequently defining the MoA based on the sole interaction with the target of choice may only give a partial story of what occurs at the cellular level. An example that illustrates this point can be drawn from the widely accepted MoA of cisplatin. Although, the bioinorganic field considers this platinum(ii) square planar complex as a DNA-­targeting agent, there is evidence that only 1% of the total Pt concentration in a given cell is localised in the nuclei. Therefore, it is acceptable to ask where the other 99% of Pt is localised, and what other relevant biological pathways is this high percentage of metal drug activating. Even more relevant, would be to ask whether any of the activated pathways are contributing to the observed anticancer activity. The rapid development of novel analytical and physicochemical techniques is enhancing our ability to investigate the fate of metal complexes once they have entered intracellular compartments. Nonetheless, there is still a long way to go in order to define a cell-­wide influence of these metal agents. No single technique holds the answer, hence a systems biology approach to investigating the biological processes that culminate in cell death may be a very attractive strategy.121 Some of the more novel techniques used involve the use of state-­of-­the-­art instrumentation such as synchrotron radiation, nanoscale secondary ion mass spectrometry (NanoSIMS) analysis and advanced (metallo) proteomic approaches. An example of the use of synchrotron radiation is the use of an X-­ray fluorescence nanoprobe (SXRFN) to reveal the target site and subcellular distribution of the osmium(ii) organometallic arene complex FY26 (Figure 2.3f) in A2780 ovarian cancer cells. The investigation showed that the metal selectively accumulates in mitochondria and this is accompanied by a Ca flux from the endoplasmic reticulum as a first step in activation of cell death pathways.98 Figure 2.10 (left) shows the cellular distribution of osmium compared to that of endogenous Zn, P and Ca, and Figure 2.10 (right) demonstrates the co-­localisation of the elements in a 500 nm thick section of A2780 cells. The same osmium(ii) complex was further investigated using microfocus synchrotron X-­ray fluorescence (SXRF) in order to establish drug penetration in 3D cancer cell models. These experiments provided insights into the efficient distribution of FY26 within 3D ovarian cancer spheroids.122 Synchrotron radiation has also been key in the investigations into the in vivo activation by reduction of ruthenium(iii) centres to ruthenium(ii), primarily using KP1019 (Figure 2.1b). This research was focused on the variation of the characteristic edge energies in the X-­ray absorption spectra of the

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

53

Figure 2.10  XRF  maps of FY26 distribution in 500 nm sections of A2780 ovarian cancer cells. (Left) Cellular distribution of Os, Zn, P and Ca. (Right) Co-­localisation of Os (red), Zn (green) and Ca (blue). Reproduced from ref. 98 with permission from John Wiley and Sons, © 2017 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA.

metal centre upon changes in the oxidation state and coordination sphere of the complex in liver samples of mice.72,123 NanoSIMS can be used, for example, to investigate changes in the coordination sphere of the metal centres once they are localised in intracellular compartments. Such is the case for the ruthenium(ii) centre in RAPTA-­C (Figure 2.3j), which localises mainly in the membrane of cisplatin-­resistant A2780Cis ovarian cancer cells, loses the arene unit and remains P-­bound to the PTA moiety.124 Another advanced mass spectrometry technique, laser ablation-­inductively coupled plasma-­mass spectrometry (LA-­ICP-­MS) has also been used to determine the spatial distribution of the complex [(p-­c ymene)Ru(N-­fluorophenyl-­2-­p yridine-­ carbo thioamide)Cl]Cl and its osmium(ii) analogue. This study used samples derived from mice bearing a CT-­26 tumour after a single administration of the metal drugs, and revealed that the highest metal concentration was localised in the animal liver, followed by kidney, lung and tumour tissue. Interestingly, comparing the penetration of both metal drugs, it is remarkable that the Ru derivative is capable of deeper localisation in mice organs.125

2.5  I s There a Bright Future for Ruthenium, Osmium and Iridium Complexes in the Fight Against Cancer? Recent advances in the research into Ru, Os and Ir complexes with anticancer applications show, beyond any doubt, that there is great potential in this field. Proof of this, is the vast number of examples and publications that include complexes with outstanding activity in vitro and even in vivo. This

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

54

Chapter 2

has not, however, been translated into a high number of complexes undergoing clinical trials, and, to date, only two ruthenium(iii) complexes have been used in humans. Such statistics should be a call for reflection. Nonetheless, great progress is evident, particularly in understanding the MoAs at cellular level. Currently, targeting strategies go beyond DNA and include a wide range of cellular events such as inhibition of thioredoxin reductase,126 cathepsin B,127 plectin,53 histone deacetylases (HDAC),128 histone demethylases (JMJD2),129 bromodomain-­containing protein 4 (BRD4),130 or the AKT/mTor pathway,131 as well as, disruption of protein–protein interactions in the H-­Ras/Raf-­1 complex.132 Furthermore, the development of in-­cell catalysis introduces important possibilities to manipulate the redox activity of cancer cells as a means for antiproliferative activity. This progress has been markedly aided by the increase in inter-­ and multi-­disciplinary teams that are tackling challenging issues from multiple perspectives, as well as the developments in novel analytical techniques. In due course, such new knowledge will help to optimise new generations of complexes, which will hopefully include that elusive alternative to platinum-­ based complexes in the clinic.

Abbreviations 5-­FU 5-­Fluorouracil acac Acetylacetonate bpy Bipyridine Cp* Pentamethylcyclopentadienyl ligand en Ethylendiamine ER Endoplasmic reticulum GAPDH Glyceraldehyde 3-­phosphate dehydrogenase GSH Glutathione reduced GSSG Glutathione oxidised HDAC Histone deacetylases Im Imidazole Ind Indazole l-­BSO l-­Buthionine sulfoxamine MoA Mechanism of action mtDNA Mitochondrial DNA NADH/NAD+  Coenzyme nicotinamide adenine dinucleotide phen Phenanthroline ppy 2-­Phenylpyridine PTA 1,3,5-­Triaza-­7-­phosphaadamantane ROS Reactive oxygen species SOD Superoxide dismutase TrxR Thioredoxin reductase TOF Turnover frequency TON Turnover number

View Online

Ruthenium, Osmium and Iridium in the Fight Against Cancer

55

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Acknowledgements Economic support from the School of Pharmacy and Institute of Clinical Sciences at the University of Birmingham are acknowledged. Many thanks also to PJS, JPPC and HKB for their helpful comments.

References 1. L. Kelland, Nat. Rev. Cancer, 2007, 7, 573–584. 2. M. Jakupec, M. Galanski, V. B. Arion, C. G. Hartinger and B. K. Keppler, Dalton Trans., 2008, 183–194. 3. S. M. Meier-­Menches, C. Gerner, W. Berger, C. G. Hartinger and B. K. Keppler, Chem. Soc. Rev., 2018, 47, 909–928. 4. G. Süss-­Fink, Dalton Trans., 2010, 39, 1673–1688. 5. C. G. Hartinger, M. Jakupec, S. Zorbas-­Seifried, M. Groessl, A. Egger, W. Berger, H. Zorbas, P. J. Dyson and B. K. Keppler, Chem. Biodiversity, 2008, 5, 2140–2155. 6. A. Bergamo, R. Gagliardi, V. Scarcia, A. Furlani, E. Alessio, G. Mestroni and G. Sava, J. Pharmacol. Exp. Ther., 1999, 289, 559–564. 7. E. Alessio, Eur. J. Inorg. Chem., 2017, 1549–1560. 8. V. R. Silva, R. S. Corrêa, L. D. S. Santos, M. B. P. Soares, A. A. Batista and D. P. Bezerra, Sci. Rep., 2018, 8, 1–13. 9. Y. Zhang, W. Zheng, Q. Luo, Y. Zhao, E. Zhang, S. Liu and F. Wang, Dalton Trans., 2015, 44, 13100–13111. 10. C. Mari, H. Huang, R. Rubbiani, M. Schulze, F. Wurthner, H. Chao and G. Gasser, Eur. J. Inorg. Chem., 2017, 1745–1752. 11. J. Hess, H. Huang, A. Kaiser, V. Pierroz, O. Blacque, H. Chao and G. Gasser, Chem. -­Eur. J., 2017, 23, 9888–9896. 12. L. Zeng, S. Kuang, G. Li, C. Jin, L. Ji and H. Chao, Chem. Commun., 2017, 53, 1977–1980. 13. L. Salassa, Eur. J. Inorg. Chem., 2011, 4931–4947. 14. G. Shi, S. Monro, R. Hennigar, J. Colpitts, J. Fong, K. Kasimova, H. Yin, R. DeCoste, C. Spencer, L. Chamberlain, A. Mandel, L. Lilge and S. A. McFarland, Coord. Chem. Rev., 2015, 282, 127–138. 15. G. Gasser, I. Ott and N. Metzler-­Nolte, J. Med. Chem., 2011, 54, 3–25. 16. P. C. Bruijnincx and P. J. Sadler, Adv. Inorg. Chem., 2009, 1–62. 17. C. A. Puckett, R. J. Ernst and J. K. Barton, Dalton Trans., 2010, 39, 1159–1170. 18. Z. Liu, A. Habtemariam, A. M. Pizarro, S. A. Fletcher, A. Kisova, O. Vrana, L. Salassa, P. C. A. Bruijnincx, G. J. Clarkson, V. Brabec and P. J. Sadler, J. Med. Chem., 2011, 54, 3011–3026. 19. J. Maksimoska, L. Feng, K. Harms, C. Yi, J. Kissil, R. Marmorstein and E. Meggers, J. Am. Chem. Soc., 2008, 130, 15764–15765. 20. L. Ronconi and P. J. Sadler, Coord. Chem. Rev., 2007, 251, 1633–1648. 21. M. Hanif, S. M. Meier, W. Kandioller, A. Bytzek, M. Hejl, C. G. Hartinger, A. Nazarov, V. B. Arion, M. Jakupec, P. J. Dyson and B. K. Keppler, J. Inorg. Biochem., 2011, 105, 224–231.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

56

Chapter 2

22. F. Schmitt, K. Donnelly, J. K. Muenzner, T. Rehm, V. Novohradsky, V. Brabec, J. Kasparkova, M. Albrecht, R. Schobert and T. Mueller, J. Inorg. Biochem., 2016, 163, 221–228. 23. K. J. Kilpin, S. Crot, T. Riedel, J. A. Kitchen and P. J. Dyson, Dalton Trans., 2014, 43, 1443–1448. 24. A. M. Pizarro, M. Melchart, A. Habtemariam, L. Salassa, F. Fabbiani, S. Parsons and P. J. Sadler, Inorg. Chem., 2010, 49, 3310–3319. 25. F. Martínez-­Peña and A. M. Pizarro, Chem. -­Eur. J., 2017, 23, 16231–16241. 26. I. Romero-­Canelón, A. M. Pizarro, A. Habtemariam and P. J. Sadler, Metallomics, 2012, 4, 1271–1279. 27. I. Romero-­Canelón, L. Salassa and P. J. Sadler, J. Med. Chem., 2013, 56, 1291–1300. 28. Z. Liu, I. Romero-­Canelón, B. Qamar, J. M. Hearn, A. Habtemariam, N. P. E. Barry, A. M. Pizarro, G. J. Clarkson and P. J. Sadler, Angew. Chem., Int. Ed., 2014, 53, 3941–3946. 29. Z. Liu, I. Romero-­Canelón, A. Habtemariam, G. J. Clarkson and P. J. Sadler, Organometallics, 2014, 33, 5324–5333. 30. P. Štarha, A. Habtemariam, I. Romero-­Canelón, G. J. Clarkson and P. J. Sadler, Inorg. Chem., 2016, 55, 2324–2331. 31. A. L. Noffke, A. Habtemariam, A. M. Pizarro and P. J. Sadler, Chem. Commun., 2012, 48, 5219–5246. 32. A. K. Renfrew, A. D. Phillips, A. E. Egger, C. G. Hartinger, S. S. Bosquain, A. A. Nazarov, B. K. Keppler, L. Gonsalvi, M. Peruzzini and P. J. Dyson, Organometallics, 2009, 28, 1165–1172. 33. R. E. Morris, R. E. Aird, P. D. S. Murdoch, H. Chen, J. Cummings, N. D. Hughes, S. Parsons, A. Parkin, G. Boyd, D. I. Jodrell and P. J. Sadler, J. Med. Chem., 2001, 44, 3616–3621. 34. A. Bergamo, A. Masi, A. Peacock, A. Habtemariam, P. J. Sadler and G. Sava, J. Inorg. Biochem., 2010, 104, 79–86. 35. A. Casini, F. Edafe, M. Erlandsson, L. Gonsalvi, A. Ciancetta, N. Re, A. Ienco, L. Messori, M. Peruzzini and P. J. Dyson, Dalton Trans., 2010, 39, 5556–5563. 36. A. Peacock, S. Parsons and P. J. Sadler, J. Am. Chem. Soc., 2007, 129, 3348–3357. 37. Y. K. Yan, M. Melchart, A. Habtemariam and P. J. Sadler, Chem. Commun., 2005, 4764–4776. 38. A. Peacock, A. Habtemariam, S. Moggach, A. Prescimone, S. Parsons and P. J. Sadler, Inorg. Chem., 2007, 46, 4049–4059. 39. Y. Fu, A. Habtemariam, A. M. B. H. Basri, D. Braddick, G. J. Clarkson and P. J. Sadler, Dalton Trans., 2011, 40, 10553–10562. 40. S. J. Dougan, M. Melchart, A. Habtemariam, S. Parsons and P. J. Sadler, Inorg. Chem., 2006, 45, 10882–10894. 41. J. M. Hearn, I. Romero-­Canelón, B. Qamar, Z. Liu, I. Hands-­Portman and P. J. Sadler, ACS Chem. Biol., 2013, 8, 1335–1343. 42. S. J. Dougan, A. Habtemariam, S. E. McHale, S. Parsons and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 11628–11633.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

57

43. Y. Fu, A. Habtemariam, A. M. Pizarro, S. H. van Rijt, D. J. Healey, P. A. Cooper, S. D. Shnyder, G. J. Clarkson and P. J. Sadler, J. Med. Chem., 2010, 53, 8192–8196. 44. Y. Fu, M. J. Romero, A. Habtemariam, M. E. Snowden, L. Song, G. J. Clarkson, B. Qamar, A. M. Pizarro, P. R. Unwin and P. J. Sadler, Chem. Sci., 2012, 3, 2485–2493. 45. S. H. Van Rijt, A. Mukherjee, A. M. Pizarro and P. J. Sadler, J. Med. Chem., 2010, 53, 840–849. 46. S. H. van Rijt, A. Peacock, R. D. L. Johnstone, S. Parsons and P. J. Sadler, Inorg. Chem., 2009, 48, 1753–1762. 47. M. Melchart, A. Habtemariam, S. Parsons and P. J. Sadler, J. Inorg. Biochem., 2007, 101, 1903–1912. 48. R. Fernández, M. Melchart, A. Habtemariam, S. Parsons and P. J. Sadler, Chem. -­Eur. J., 2004, 10, 5173–5179. 49. A. J. Millett, A. Habtemariam, I. Romero-­Canelón, G. J. Clarkson and P. J. Sadler, Organometallics, 2015, 34, 2683–2694. 50. J. Li, M. Tian, Z. Tian, S. Zhang, C. Yan, C. Shao and Z. Liu, Inorg. Chem., 2018, 57, 1705–1716. 51. C. Wang, J. Liu, Z. Tian, M. Tian, L. Tian, W. Zhao and Z. Liu, Dalton Trans., 2017, 46, 6870–6883. 52. S. M. Meier, M. Hanif, Z. Adhireksan, V. Pichler, M. Novak, E. Jirkovsky, M. Jakupec, V. B. Arion, C. Davey, B. K. Keppler and C. G. Hartinger, Chem. Sci., 2013, 1837–1846. 53. S. M. Meier, D. Kreutz, L. Winter, M. H. M. Klose, K. Cseh, T. Weiss, A. Bileck, B. Alte, J. C. Mader, S. Jana, A. Chatterjee, A. Bhattacharyya, M. Hejl, M. A. Jakupec, P. Heffeter, W. Berger, C. G. Hartinger, B. K. Keppler, G. Wiche and C. Gerner, Angew. Chem., Int. Ed., 2017, 56, 8267–8271. 54. G. S. Yellol, A. Donaire, J. G. Yellol, V. Vasylyeva, C. Janiak and J. Ruiz, Chem. Commun., 2013, 49, 11533–11535. 55. Z. Liu, L. Salassa, A. Habtemariam, A. M. Pizarro, G. J. Clarkson and P. J. Sadler, Inorg. Chem., 2011, 50, 5777–5783. 56. Z. Liu, A. Habtemariam, A. M. Pizarro, G. J. Clarkson and P. J. Sadler, Organometallics, 2011, 30, 4702–4710. 57. N. P. E. Barry, O. Zava, P. J. Dyson and B. Therrien, J. Organomet. Chem., 2012, 705, 1–6. 58. N. P. E. Barry, F. Edafe, P. J. Dyson and B. Therrien, Dalton Trans., 2010, 39, 2816–2820. 59. S. P. Mulcahy and E. Meggers, Top. Organomet. Chem., 2010, 32, 141–153. 60. A. Zamora, G. Vigueras, V. Rodríguez, M. D. Santana and J. Ruiz, Coord. Chem. Rev., 2018, 360, 34–76. 61. N. Wu, J.-­J. Cao, X.-­W. Wu, C.-­P. Tan, L.-­N. Ji and Z.-­W. Mao, Dalton Trans., 2017, 46, 13482–13491. 62. W. Y. Zhang, Q. Y. Yi, Y. J. Wang, F. Du, M. He, B. Tang, D. Wan, Y. J. Liu and H. L. Huang, Eur. J. Med. Chem., 2018, 151, 568–584. 63. J. Ruiz, C. Vicente, C. De Haro and D. Bautista, Inorg. Chem., 2013, 52, 974–982.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

58

Chapter 2

64. V. Scalcon, M. Salmain, A. Folda, S. Top, P. Pigeon, H. Z. Shirley Lee, G. Jaouen, A. Bindoli, A. Vessières and M. P. Rigobello, Metallomics, 2017, 9, 949–959. 65. H. Z. S. Lee, O. Buriez, F. Chau, E. Labbé, R. Ganguly, C. Amatore, G. Jaouen, A. Vessières, W. K. Leong and S. Top, Eur. J. Inorg. Chem., 2015, 4217–4226. 66. G. Jaouen, A. Vessières and S. Top, Chem. Soc. Rev., 2015, 44, 8802–8817. 67. L. Oehninger, R. Rubbiani and I. Ott, Dalton Trans., 2013, 42, 3269–3284. 68. W. Streciwilk, A. Terenzi, X. Cheng, L. Hager, Y. Dabiri, P. Prochnow, J. E. Bandow, S. Wölfl, B. Keppler and I. Ott, Eur. J. Med. Chem., 2018, 156, 148–161. 69. Y. Gothe, T. Marzo, L. Messori and N. Metzler-­Nolte, Chem. -­Eur. J., 2016, 22, 12487–12494. 70. Y. Gothe, T. Marzo, L. Messori and N. Metzler-­Nolte, Chem. Commun., 2015, 51, 3151–3153. 71. Y. Gothe, I. Romero-­Canelón, T. Marzo, P. J. Sadler, L. Messori and N. Metzler-­Nolte, Eur. J. Inorg. Chem., 2018, 2018, 2461–2470. 72. R. Trondl, P. Heffeter, C. R. Kowol, M. A. Jakupec, W. Berger and B. K. Keppler, Chem. Sci., 2014, 5, 2925–2932. 73. A. Bergamo and G. Sava, Dalton Trans., 2011, 40, 7817–7823. 74. G. Sava and A. Bergamo, in Platinum and Other Heavy Metal Compounds, ed. A. Bonetti, R. Leone, F. M. Muggia and S. B. Howell, Humana Press, Totowa, NJ, 2009, pp. 57–66. 75. A. Bergamo and G. Sava, Chem. Soc. Rev., 2015, 44, 8818–8835. 76. M. Melchart, A. Habtemariam, O. Novakova, S. Moggach, F. P. Fabbiani, S. Parsons, V. Brabec and P. J. Sadler, Inorg. Chem., 2007, 46, 8950–8962. 77. A. Peacock, M. Melchart, R. J. Deeth, A. Habtemariam, S. Parsons and P. J. Sadler, Chem. -­Eur. J., 2007, 13, 2601–2613. 78. A. Habtemariam, M. Melchart, R. Fernandez, S. Parsons, I. D. Oswald, A. Parkin, F. P. Fabbiani, J. E. Davidson, A. Dawson, R. E. Aird, D. I. Jodrell and P. J. Sadler, J. Med. Chem., 2006, 49, 6858–6868. 79. F. Wang, A. Habtemariam, E. van der Geer, R. Fernández, M. Melchart, R. J. Deeth, R. Aird, S. Guichard, F. P. Fabbiani, P. Lozano-­Casal, I. D. H.   Oswald, D. I. Jodrell, S. Parsons and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 18269–18274. 80. O. Novakova, H. Chen, O. Vrana, A. Rodger, P. J. Sadler and V. Brabec, Biochemistry, 2003, 42, 11544–11554. 81. T. Bugarcic, O. Nováková, A. Halámiková, L. Zerzánková, O. Vrána, J. Kaspárková, A. Habtemariam, S. Parsons, P. J. Sadler and V. Brabec, J. Med. Chem., 2008, 51, 5310–5319. 82. Z. Liu, A. Habtemariam, A. M. Pizarro, S. Fletcher, A. Kisova, O. Vrana, L. Salassa, P. C. Bruijnincx, G. J. Clarkson, V. Brabec and P. J. Sadler, J. Med. Chem., 2011, 54, 3011–3026. 83. O. Nováková, A. Nazarov, C. G. Hartinger, B. K. Keppler and V. Brabec, Biochem. Pharmacol., 2009, 77, 364–374.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

59

84. Z. Chen, W. Lu, C. Garcia-­Prieto and P. Huang, J. Bioenerg. Biomembr., 2007, 39, 267–274. 85. P. P. Hsu and D. M. Sabatini, Cell, 2008, 134, 703–707. 86. U. Jungwirth, C. R. Kowol, B. K. Keppler, C. G. Hartinger, W. Berger and P. Heffeter, Antioxid. Redox Signaling, 2011, 15, 1085–1127. 87. D. Tennant, R. V. Durán and E. Gottlieb, Nat. Rev. Cancer, 2010, 10, 267–277. 88. G. Chen, F. Wang, D. Trachootham and P. Huang, Mitochondrion, 2010, 10, 614–625. 89. E. Oldham and H. Æ. Jinsong, Cancer Chemother. Pharmacol., 2004, 209–219. 90. L. Galluzzi, N. Larochette, N. Zamzami and G. Kroemer, Oncogene, 2006, 25, 4812–4830. 91. G. Kroemer, Oncogene, 2006, 25, 4630–4632. 92. P. Pinton and G. Kroemer, Nat. Chem. Biol., 2014, 10, 89–90. 93. J. Watson, Open Biol., 2013, 3, 120144. 94. J. D. Pennington, T. Jau, C. Wang, P. Nguyen, L. Sun, K. Bisht, D. Smart and D. Gius, Drug Resist. Updates, 2005, 8, 322–330. 95. I. Romero-­Canelón and P. J. Sadler, Inorg. Chem., 2013, 52, 12276–12291. 96. I. Romero-­Canelón, M. Mos and P. J. Sadler, J. Med. Chem., 2015, 58, 7874–7880. 97. R. J. Needham, C. Sanchez-­Cano, X. Zhang, I. Romero-­Canelón, A. Habtemariam, M. S. Cooper, L. Meszaros, G. J. Clarkson, P. J. Blower and P. J. Sadler, Angew. Chem., Int. Ed., 2017, 56, 1017–1020. 98. C. Sanchez-­Cano, I. Romero-­Canelón, Y. Yang, I. J. Hands-­Portman, S. Bohic, P. Cloetens and P. J. Sadler, Chem. -­Eur. J., 2017, 23, 2512–2516. 99. S. D. Shnyder, Y. Fu, A. Habtemariam, S. H. Van Rijt, P. A. Cooper, P. M. Loadman, P. J. Sadler, S. H. van Rijt, P. A. Cooper, P. M. Loadman and P. J. Sadler, MedChemComm, 2011, 2, 666–668. 100. J. M. Hearn, I. Romero-­Canelón, A. F. Munro, Y. Fu, A. M. Pizarro, M. J. Garnett, U. McDermott, N. O. Carragher and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, E3800–E3805. 101. J. M. Hearn, G. M. Hughes, I. Romero-­Canelón, A. F. Munro, B. Rubio-­ Ruiz, Z. Liu, N. O. Carragher and P. J. Sadler, Metallomics, 2018, 10, 93–107. 102. D. Sidler, A. Brockmann, J. Mueller, U. Nachbur, N. Corazza, P. Renzulli, A. Hemphill and T. Brunner, Oncogene, 2012, 31, 4095–4106. 103. Z. Luo, L. Yu, F. Yang, Z. Zhao, B. Yu, H. Lai, K.-­H. Wong, S.-­M. Ngai, W. Zheng and T. Chen, Metallomics, 2014, 6, 1480–1490. 104. V. Scalcon, S. Top, H. Z. S. Lee, A. Citta, A. Folda, A. Bindoli, W. K. Leong, M. Salmain, A. Vessières, G. Jaouen and M. P. Rigobello, J. Inorg. Biochem., 2016, 160, 296–304. 105. G. Agonigi, T. Riedel, M. P. Gay, L. Biancalana, E. Oñate, P. J. Dyson, G. Pampaloni, E. Păunescu, M. A. Esteruelas and F. Marchetti, Organometallics, 2016, 35, 1046–1056.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

60

Chapter 2

106. E. Păunescu, M. Soudani, P. Martin, R. Scopelliti, M. Lo Bello and P. J. Dyson, Organometallics, 2017, 36, 3312–3321. 107. C. Zhang, S. H. Lai, C. C. Zeng, B. Tang, D. Wan, D. G. Xing and Y. J. Liu, J. Biol. Inorg. Chem., 2016, 21, 1047–1060. 108. R.-­R. Ye, C.-­P. Tan, L.-­N. Ji and Z.-­W. Mao, Dalton Trans., 2016, 45, 13042–13051. 109. X. D. Song, X. Kong, S. F. He, J. X. Chen, J. Sun, B. B. Chen, J. W. Zhao and Z. W. Mao, Eur. J. Med. Chem., 2017, 138, 246–254. 110. M. Ouyang, L. Zeng, H. Huang, C. Jin, J. Liu, Y. Chen, L. Ji and H. Chao, Dalton Trans., 2017, 46, 6734–6744. 111. G. Süss-­Fink, J. Organomet. Chem., 2014, 751, 2–19. 112. S. Betanzos-­Lara, Z. Liu, A. Habtemariam, A. M. Pizarro, B. Qamar and P. J. Sadler, Angew. Chem., Int. Ed., 2012, 51, 3897–3900. 113. J. J. Soldevila-­Barreda, I. Romero-­Canelón, A. Habtemariam and P. J. Sadler, Nat. Commun., 2015, 6, 6582. 114. J. P. C. Coverdale, C. Sanchez-­Cano, G. J. Clarkson, R. Soni, M. Wills and P. J. Sadler, Chem. -­Eur. J., 2015, 8043–8046. 115. J. P. C. Coverdale, I. Romero-­Canelón, C. Sanchez-­Cano, G. J. Clarkson, A. Habtemariam, M. Wills and P. J. Sadler, Nat. Chem., 2018, 10, 347–354. 116. M. Tomás-­Gamasa, M. Martínez-­Calvo, J. R. Couceiro and J. L. Mascareñas, Nat. Commun., 2016, 7, 12538. 117. J. G. Rebelein and T. R. Ward, Curr. Opin. Biotechnol., 2018, 53, 106–114. 118. Y. M. Wilson, M. Dürrenberger, E. S. Nogueira and T. R. Ward, J. Am. Chem. Soc., 2014, 136, 8928–8932. 119. Y. Okamoto, V. Köhler, C. E. Paul, F. Hollmann and T. R. Ward, ACS Catal., 2016, 6, 3553–3557. 120. F. Schwizer, Y. Okamoto, T. Heinisch, Y. Gu, M. M. Pellizzoni, V. Lebrun, R. Reuter, V. Köhler, J. C. Lewis and T. R. Ward, Chem. Rev., 2018, 118, 142–231. 121. I. Romero-­Canelón and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 4187–4188. 122. C. Sanchez-­Cano, I. Romero-­Canelón, K. Geraki and P. J. Sadler, J. Inorg. Biochem., 2018, 185, 26–29. 123. A. Hummer, P. Heffeter, W. Berger, M. Filipits, D. Batchelor, G. E. Büchel,  M. Jakupec, B. K. Keppler and A. Rompel, J. Med. Chem., 2013, 56, 1182–1196. 124. R. F. S. Lee, S. Escrig, M. Croisier, S. Clerc-­Rosset, G. W. Knott, A. Meibom, C. A. Davey, K. Johnsson and P. J. Dyson, Chem. Commun., 2015, 51, 16486–16489. 125. M. H. M. Klose, S. Theiner, C. Kornauth, S. M. Meier-­Menches, P. Heffeter, W. Berger, G. Koellensperger and B. K. Keppler, Metallomics, 2018, 10, 388–396. 126. W. H. Ang, A. Casini, G. Sava and P. J. Dyson, J. Organomet. Chem., 2011, 696, 989–998.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00031

Ruthenium, Osmium and Iridium in the Fight Against Cancer

61

127. A. Mitrović, J. Kljun, I. Sosič, S. Gobec, I. Turel and J. Kos, Dalton Trans., 2016, 45, 16913–16921. 128. R.-­R. Ye, C.-­P. Tan, L. He, M.-­H. Chen, L.-­N. Ji and Z.-­W. Mao, Chem. Commun., 2014, 50, 10945–10948. 129. L. J. Liu, L. Lu, H. J. Zhong, B. He, D. W. J. Kwong, D. L. Ma and C. H. Leung, J. Med. Chem., 2015, 58, 6697–6703. 130. H.-­J. Zhong, L. Lu, K.-­H. Leung, C. C. L. Wong, C. Peng, S.-­C. Yan, D.-­ L. Ma, Z. Cai, H.-­M. David Wang and C.-­H. Leung, Chem. Sci., 2015, 6, 5400–5408. 131. B. Tang, D. Wan, Y. J. Wang, Q. Y. Yi, B. H. Guo and Y. J. Liu, Eur. J. Med. Chem., 2018, 145, 302–314. 132. L.-­J. Liu, W. Wang, S.-­Y. Huang, Y. Hong, G. Li, S. Lin, J. Tian, Z. Cai, H.-­ M. D. Wang, D.-­L. Ma and C.-­H. Leung, Chem. Sci., 2017, 8, 4756–4763.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Chapter 3

Iron Compounds as Anticancer Agents ANNE Vessieres* Sorbonne Université, CNRS, Institut Parisien de Chimie Moléculaire, UMR CNRS 8232, 4, Place Jussieu, F-­75005 Paris, France *E-­mail: anne.vessieres@sorbonne-­universite.fr

3.1  Introduction Iron is the most abundant transition metal on earth, and it is likely to be the last one remaining when all the rest have been extracted and consumed. Iron is also an essential element in the functioning of the human body, which contains an average of approximately 6 g for a 70 kg male.1 It is a key component of heme and as such plays an essential role in transporting dioxygen in the blood, as well as being involved at the cellular level in the production of energy and the synthesis of DNA. Indeed, ribonucleotide reductase (RR), an enzyme that catalyses an essential step in the synthesis and repair of DNA, namely the transformation of ribonucleotides into deoxyribonucleotides, has two iron atoms in its active site.2 Iron is also present in hemoproteins such as the catalases and cytochromes P450, as well as in non-­heme iron enzymes that play an important role in the activation of dioxygen.3,4 These biological activities are associated with its different degrees of oxidation, essentially iron(ii) and iron(iii), and its ability to cycle between these two stable states (ferrous and ferric). Iron is also a component of transferrin and ferritin, metalloproteins which are responsible, respectively, for the transport of   Metallobiology Series No. 14 Metal-­based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-­Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

62

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

63

free iron in biological fluids and its storage in the body. The change in state of iron from iron(ii) to iron(iii) in the presence of dioxygen is accompanied by production of reactive oxygen species (ROS) which are potentially toxic for cells. The homeostasis of iron in the cell is a key factor in their behaviour and has to be well controlled to avoid this toxic effect. It should also be noted that cancer cells require more iron than healthy cells due in part to their rapid rate of DNA synthesis.5 Of all the metals used in the composition of metallodrugs discussed in this book, iron is the only element present in the body in the gram range, which gives it the not insignificant advantage of being a metal that, except in disorders that involve an excess of iron, is considered non-­toxic to humans. For this reason, molecules that contain iron are more readily accepted by the medical community than metallodrugs based on heavy metals such as platinum. An inventory of iron complexes that have been the subject of extended study in medicinal chemistry to date reveals only three organometallic complexes of ferrocene (Scheme 3.1). These are: (1) ferrocerone, currently the only molecule of this type to be commercialized,6 (2) ferroquine, a ferrocene analogue of chloroquine which is active on chloroquine-­resistant strains of Plasmodium falciparum,7–9 currently in phase II clinical development and recently shown to possess antitumor activity10 and (3) ferrocifens, ferrocene analogues of the reference antiestrogen tamoxifen, the anticancer effects of which have been studied for many years by our research group in Paris.11 Indeed, the study of these complexes has contributed substantially to the development of the field of bioorganometallic chemistry, defined as “the synthesis and study of organometallic species of biological and medicinal interest”.12–14 Molecules of this family are currently in development at the start-­up company Ferroscan. Interest in iron as an anticancer agent leads to the study of three very different categories of complexes. The first is the family of organometallic complexes of ferrocene that have been the subject of a number of review articles11,15–18 and book chapters.19–21 The second family, that of the inorganic coordination complexes of iron(ii) and iron(iii), is less well developed and is discussed only in one recently published review.22 A third family of

Scheme 3.1  Structures  of ferrocerone, ferroquine and ferrocifens.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

64

Chapter 3

molecules that has not yet to our knowledge appeared in any reviews of iron-­based metallodrugs, namely iron chelating agents, will be presented here. These molecules earn their place in this chapter through their ability to use the iron present in the body to form cytotoxic complexes of iron. The most representative molecules of this family are the bleomycins, antibiotics produced by the bacterium Streptomyces verticillus that are prescribed in the treatment of certain cancers (testicular cancer, certain types of lymphoma).23 To this can be added molecules originally intended to treat iron overload which have also been shown to have antiproliferative effects on cancer cells.24,25 A number of bimetallic mixed (organometallic/coordination) complexes containing a ferrocene and another transition metal (Au, Cu, Pd, Pt, Rh, Ru) have been synthesized with the aim of studying their anticancer properties. They are the subject of Chapter 6. A large number of the publications of interest cited in this chapter concern work at the interface of chemistry and biology. They generally fall into two categories. The first category is focused on the synthesis of novel complexes while the second concerns the study of the antiproliferative effects of the complexes in vitro on one or more cancer cell lines. The goal is determination of IC50 values for the new complexes to provide a rough-­ and-­ready assessment of their potential as anticancer agents. The most promising molecules are then subjected to biological studies, some more extended than others, to attempt to determine their mechanism of action, their effect on the cell cycle, the type of cell death they induce, their effect on signaling pathways and possibly other parameters. In vivo studies of the effects of these new molecules on the growth of xenografted tumours are rarely undertaken at this stage, although it is important to bear in mind that in vitro testing is far from sufficient to predict the behaviour of the molecules in vivo. Recent articles underlined this evident point, in particular for medications given orally.26,27 In addition, the Lipinski rule of five28 which is used in the pharmaceutical industry to select the best candidates for in vivo study is not suitable for organometallic complexes, which are generally lipophilic molecules, and the quantitative estimate drug-­likeness rule,29 which proposes another approach, is not yet widely used.30 For in vitro studies these molecules are dissolved in DMSO then added to biological media to reach a final maximal percentage of 1% that is not toxic for cells. In vitro, these lipophilic compounds easily pass through the cellular membrane. It has thus been shown that the addition of a lipophilic substituent onto a potentially cytotoxic but hydrophilic synthon facilitates its entry into the cell and enables expression of its cytotoxicity.31 Finally, basing the selection of potentially interesting complexes on the criterion of their IC50 values does not take account of the fact that in vivo an inactive molecule may be metabolized to form an active metabolite. This is the case for tamoxifen which is metabolized in the liver to hydroxytamoxifen, its active form. It also overlooks the molecules that have a strong antimetastatic effect in vivo while showing only a weak cytotoxic effect in vitro.

View Online

Iron Compounds as Anticancer Agents

65

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

The best known example is that of NAMI-­A , a complex of ruthenium developed by G. Sava, who has given a clear account of this problem in recent reviews.32,33

3.2  Study of Ferrocene Complexes 3.2.1  Background Ferrocene was synthesized in 1951 by P. L. Pauson34 and S. A. Miller,35 and its structure determined in 1952 by E. O. Fischer36 and G. Wilkinson37 who received the Nobel Prize in Chemistry for this work in 1973. Ferrocene is the archetype of organometallic sandwich compounds. It is air-­stable and behaves like an aromatic, providing access to a library of extremely varied molecules. Ferrocene is characterized by facile and reversible oxidation of iron(ii) to iron(iii). Its redox potential is 0.4 V, a value that is particularly interesting for medicinal chemistry as it is compatible with the redox potentials, in the range +0.4 to −0.44 V, found in the cell and thus favourable to in cellulo electron transfer and redox cycling.38 This is not the case for its ruthenocene and osmocene analogues which have redox potentials of 0.8 and 0.6 V, respectively. It is this property of ferrocene that has recently been advanced as an explanation for the superior activity against cancer cells in vitro that has been observed for the ferrocene complex of hydroxytamoxifen compared to ruthenocene and osmocene.39 The first attempts to use this complex in medicinal chemistry have shown that ferrocene alone has limited toxicity even when administered in high dose.40 Study of its metabolism has shown that it was hydroxylated in the liver by cytochrome P450 to hydroxyferrocene,41 an unstable compound in physiological media leading ultimately to the release of solvated iron atoms. The first ferrocene complexes to have been studied for their anticancer effects were the ferrocenium salts 1 and 2, prepared in the mid-­1970s by Köpf and P. Köpf-­Maier (Scheme 3.2).42 Their idea was to find complexes with mechanisms different from that of cisplatin, a drug that at that period had recently come to market43 prompting considerable progress in the treatment of incurable cancers.44 Intraperitoneal injection of these complexes produced a reduction in the development of ascites in mice. However these molecules have the disadvantage of being unstable in solution. D. Osella and co-­workers later completed the series by synthesizing decamethyl ferrocenium tetrafluoroborate 3 then

Scheme 3.2  Ferrocenium  salts.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

66

Chapter 3

testing this series of molecules in vitro on MCF-­7 (hormone-­dependent breast cancer) cells.45 The complexes described by Köpf-­Maier have high IC50 values (IC50 > 300 µM). The most active complex is 3 (IC50 = 37 µM). Its mechanism of action is linked to the production of ROS (superoxide and hydroxy radical) leading, via a Fenton reaction, to oxidative damage of DNA. The study of these complexes was not continued, while the same period saw the development of ferrocene as a bioisostere of aromatic substituents present in already established drugs, without achieving any notable progress. The first examples to appear in the literature date from 1975 and concern antibiotic derivatives, in particular ferrocene derivatives of penicillin, but these studies were unsuccessful.46,47 The first ferrocene complexes to represent a real innovation in the domain of anticancer drugs appeared only in 1996, when our research group synthesized ferrocene complexes of hydroxytamoxifen (Fc-­OH-­Tam, 4; see Scheme 3.3).48,49 Since that pioneering study a large number of ferrocene

Scheme 3.3  Selected  members of the ferrocifen family.

View Online

Iron Compounds as Anticancer Agents

67

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

complexes derived from molecules known for their anticancer effects have been prepared, as cited in a number of recent reviews.11,50,51 I have chosen to present here only a representative selection of these complexes.

3.2.2  The Ferrocifen Family Scheme 3.3 represents the four most studied hits of this family of around 300 molecules.11,51,52 These are the tamoxifen-­like complex, 4, mono and diphenolic complexes, 5 and 6, and the ansa-­diphenolic complex, 8, in which the two cyclopentadienyl rings are linked by an aliphatic chain.

3.2.2.1 Tamoxifen-­like Ferrocifens The first complexes of this family to be synthesized were the ferrocene analogues of hydroxytamoxifen (OH-­Tam), the active metabolite of tamoxifen, a drug used in adjuvant treatment of hormone-­dependent breast cancers (see Scheme 3.1).53 The original intent was to study the effect of replacing the aromatic β ring of OH-­Tam with a ferrocenyl substituent. These complexes are obtained via a McMurry coupling of the corresponding ketones.49,54 Complexes with amino chains of various lengths were synthesized. Complex 4 (Scheme 3.3) has been studied the most owing to its ease of synthesis and its strong antiproliferative effect (IC50 around 0.5 µM) on hormone-­ dependent (MCF-­7) and hormone-­independent (MDA-­MB-­231) breast cancer cells, while OH-­Tam itself is only active on hormone-­dependent cells.49,54 This effect is the result of an antiestrogenic component linked to its tamoxifen-­like structure plus a cytotoxic effect specific to the ferrocene complex. It has in fact been clearly established that it is the ferrocenyl-­ double bond-­phenol linkage pattern that is the source of this high cytotoxicity.11,51,52 Complex 4 produces ROS immediately upon entering MCF-­7 and MDA-­MB-­231 cells, while OH-­Tam does not.55,56 At low concentrations (1 µM), 4 induces senescence but not apoptosis of MCF-­7 and MDA-­MB-­231 cells.56 Interestingly, complex 4 induces this type of cell death on cells that are not susceptible to apoptosis (glioblastomas, triple negative breast cancer cells, etc.), and are associated with the solid tumours that remain the most difficult to treat. In Jurkat cells (leukaemia) that are susceptible only to apoptosis, a high concentration of 4 (15 µM) induces inhibition of mitochondrial thioredoxin reductase (TrxR) as well as lowering the mitochondrial membrane potential (MMP) leading to apoptosis of the cells.39 The interaction of 4 at the level of the mitochondria can be explained by the protonation, at physiological pH, of its dimethylamino group, giving it the character of a lipophilic cation. Its presence in these organelles has been confirmed by the high amount of iron found by ICP-­OES in the mitochondria of Jurkat cells (36% of total iron in mitochondria versus 10% in cytosol and 64% in crude nuclei).39 Owing to its lipophilicity, complex 4 cannot be administered in vivo without encapsulation in lipid nanocapsules (LNCs).57,58 These are prepared with a

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

68

Chapter 3

lipophilic core (Labrafac®) that allows solubilization of ferrocene complexes. This is then covered by a polyethylene glycol (PEG-­HS®, solutol HS15) surfactant. Intraperitoneal injection of LNCs loaded with 4 showed significant regression of xenografted tumours of MDA-­MB-­321 cells in severe combined immunodeficient (SCID) mice.59 The recently reported ansa-­tamoxifen-­like complex, 9,60 is more toxic than 4 on MDA-­MB-­231 cells (IC50 = 0.18 µM), but its low solubility in DMSO hinders its development.

3.2.2.2 Mono and Diphenol Ferrocifens The mono and diphenolic complexes 5 and 6 have been synthesized.61,62 These complexes have a strong cytotoxic effect on MDA-­MB-­231 cells (IC50 = 0.6 and 1.5 µM, respectively).63 This effect is linked to the unique redox properties of the ferrocifens and involves the formation of the corresponding quinone methides 5-­QM and 6-­QM, according to a reaction scheme in which all steps have been confirmed by electrochemistry and electron paramagnetic resonance (EPR).64,65 These QMs are thought to be the source of the toxicity of ferrocifens 5 and 6 which appear to be prodrugs whose activation in the cells leads to their cytotoxicity. These QMs can be obtained chemically by oxidation in the presence of Ag2O. They are reactive species that give rise to 1,8-­Michael additions with nucleophilic thiols such as glutathione.66 They are also good inhibitors in vitro of purified cytosolic TrxR (IC50 around 2.5 µM) while 5 and 6 are not. On the other hand, 6, unlike 4, does not inhibit TrxR within Jurkat cells.67 This result is explained by the intracellular transformation of 6-­QM to the corresponding indene, 7, a complex that cannot give rise to Michael additions.67 Experiments on human glioma cell lines (Hs683 and U373) have shown that 5 induced senescence.68 This is linked to the expression of SA-­β-­galactosidase, and to the secretion of various interleukins and TNF-­α. In vivo, intravenous injection into rats of stealth LNCs loaded with 5 induces significant regression of ectopic tumours of 9L cells (rat glioblastoma).57,69,70 Long-­term survivals (animals still living after 100 days of experimentation) have occurred in rats bearing orthotopic tumours of 9L cells when the injection of LNCs loaded with 5 is accompanied by radiotherapy.71

3.2.2.3 Ansa-­ferrocifens Ansa-­ferrocifens in which the two cyclopentadienyl rings are linked by an aliphatic chain of variable length were prepared and studied by our group and by D. Plazuk.72–74 The most cytotoxic molecule of this series is ansa-­Fc-­diOH, 8. Its toxicity on MDA-­MB-­231 cells is very high (IC50 = 0.089 µM), i.e. about ten times higher than that of 5, the corresponding acyclic complex.63 The analysis performed with 5 and 8 (NSC identifiers: 748141/1, 750285/1), the acyclic and cyclic diphenols, on the NIH NCI collection of 60 cancer cell lines confirmed the strong toxicity of these complexes (average IC50 values 0.52 and 0.18 µM, respectively). These two complexes are more active on cancer cell lines that are resistant to apoptosis and have no effect on healthy cells.57

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

69

Unlike in the acyclic series, chemical oxidation of 8 by Ag2O does not permit formation of the corresponding quinone methide. However, the organic radical [8-­QM]• has been observed by EPR during the enzymatic oxidation of 8 by the HRP/H2O2 system.75 The relative stabilization of the latter may explain the high toxicity of 8. Quinone methides can again be formed by chemical oxidation when the chain linking the two rings is lengthened, but the toxicity of these complexes is lower than that of 8 (IC50 = 1.85 µM for a five-­carbon chain).72 Study of the biological effects of 8 show that at low doses (50 nM) it induces senescence of 9L cells, while a higher dose (0.5 µM) induces apoptosis.76 In vivo studies undertaken with stealth LNCs reveal strong regression of ectopic tumours of 9L cells implanted in rats.76

3.2.2.4 The New Generation of Ferrocifens A new series of ferrocifens with a hydroxypropyl side chain were recently synthesized.66 The most active product is the diphenolic complex 13 which on MDA-­MB-­231 cells has a higher toxicity than 5, the reference diphenolic complex (IC50 = 0.11 µM and 0.6 µM, respectively). Chemical oxidation of 13 by Ag2O leads to formation of an unprecedented tetrahydrofuran quinone methide, 14, via internal cyclization of the hydroxy alkyl chain and may be the source of its high toxicity.77 Oxidation of the corresponding organic molecule leads to the formation of the acyclic vinyl quinone methide, an analogue of that formed with the other ferrocifens. This difference seems to be linked to the stabilization of the α-­carbenium ion by the ferrocenyl group, which facilitates tetrahydrofuran ring formation. Addition of thiols onto 14 leads to 1,6-­adducts rather than the 1,8-­adducts observed previously. This series of molecules with novel properties opens up new perspectives for the ferrocifen family.

3.2.2.5 Dimethylamino Ferrocifens The antiproliferative effects of the diamino complexes of the ansa and acyclic series (10,11) have been compared to the effect of 12, the corresponding organic molecule.78 In this case the ferrocene and organic complexes have similar and strong antiproliferative effects (IC50 around 0.40 µM on MDA-­ MB-­231 cells). The cytotoxicity of these molecules is thus not linked to the presence of the metal entity but to that of the two amino chains that have a detergent effect at the level of the cell membrane. These complexes are also the only ferrocifens to show a significant antibacterial effect.79

3.2.3  Ferrocene Complexes of Natural Products A number of ferrocene complexes of natural products have been prepared and are the topic of a recent review.16 A selection of these molecules with biologically interesting properties is presented here. Plazuk and co-­workers

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

70

Chapter 3

Scheme 3.4  Ferrocenyl  derivatives of taxanes. have been particularly productive in this area, for example having synthesized a number of ferrocene complexes of taxanes, paclitaxel and docetaxel (Scheme 3.4).80–82 These molecules, which have a complex chemical structure, were originally extracted from the bark or needles of yew. They have become anticancer agents of the first rank due to their strong antimitotic effect, which is associated with inhibition of tubuline depolymerisation. Taxanes bearing a ferrocenyl group in the 2′O position as well as the derivative of paclitaxel with a substitution of the 3′N-­benzoyl group were therefore prepared (15–17).82 These complexes were tested on colon cancer cell lines (SW620) or lines resistant to various drugs (doxorubicin, etoposide, methotrexate). The paclitaxel complex 15 is more active than its precursor on colon cancer cell lines resistant to methotrexate (SW620M; IC50 = 0.70/6.38 µM),80 while the docetaxel complex 16 is more active on cell lines resistant to doxorubicin (SW620D; IC50 = 16.5/35.16 µM) and etoposide (SW620E; IC50 1.56/21.27 µM). Complex 17 is ten times more cytotoxic than paclitaxel on SW620 cells (IC50 = 0.11/1.1 µM). Its mechanism of action appears to be the same as that of its precursor, since it has been shown that it too can stabilize microtubules. This behaviour can be explained by the fact that the complex has a good fit in the taxane-­binding site of tubulin, even in the presence of the bulky ferrocene. Ferrocene derivatives of podophyllotoxin, a highly toxic molecule isolated from the roots and rhizomes of the American Mayapple (Podophyllum peltatum), have recently been prepared (Scheme 3.5).83,84 This molecule represents an interesting situation. In fact its toxicity is so high that it is considered a poison, with its use limited to topical application. The aim therefore is to lower the toxicity to render the molecule compatible with use in chemotherapy. Etoposide and teniposide, podophyllotoxin derivatives that inhibit topoisomerase were thus prepared and are used as anticancer medications. Ferrocene complexes were synthesized that replicate the functionalization of the 4-­hydroxy group of these molecules, now substituted by esters, amides and ferrocene 1,2,3-­triazoles. The most interesting products are the esters (18–20). Their cytotoxicity is high (IC50 in the range 0.2–0.3 µM on SW620 cells) but at around ten times lower than that

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

71

Scheme 3.5  Podophyllotoxin  derivatives.

Scheme 3.6  Ferrocenyl  derivatives of artemisinin. of podophyllotoxin (IC50 = 0.02 µM) represents the desired outcome. Their mechanism of action remains similar to that of podophyllotoxin (blockage of cells in the G2/M phase). Artemisinin is a molecule in the Chinese pharmacopeia, extracted from the annual plant Artemisia annua L. This is one of the standard molecules in the treatment of malaria, and Tu Youyou, who identified its chemical structure, was awarded the Nobel Prize in Physiology and Medicine in 2015. Interestingly, its ferrocene derivatives were synthesized by Tsogoeva and Efferth (21 and 22, Scheme 3.6).85,86 The aim was to include in the molecule a ferrocene fragment capable of producing ROS that could facilitate the oxidative cleavage of the endoperoxide bond of the trioxane fragment. These complexes are very active on wild type human leukaemia (CCRF-­CEM) and multidrug-­resistant (CEM/ ADR5000) cell lines (IC50 = 0.25 µM and 0.57 µM for 21, 0.07 µM and 1.8 µM for 22), while artemisinin has no antiproliferative effect on these cells (IC50 = 37 and 27 µM). Although these are innovative and potentially interesting molecules their mechanism of action has not been studied.

View Online

72

Chapter 3

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

3.2.4  F  errocenyl Complexes of Histone Deacetylase Inhibitors (HDACi) The processes of acetylation and deacetylation of the lysines of chromatin are fundamental to its remodelling and gene transcription, and belong to the area of epigenetics. The histone deacetylases (HDACs) play an important role in this process since they catalyse the removal of acetyl groups leading to transcriptional repression via a condensed DNA. The inhibitors of these enzymes are considered to be potential anticancer agents. Indeed, suberoylanilide hydroxamic acid (SAHA), the archetype of these inhibitors, is prescribed for the treatment of cutaneous T-­cell lymphoma. Ferrocene derivatives of SAHA and TCH106, another HADAC inhibitor, have been prepared by Spencer and by our group (23–29) (Scheme 3.7).87–92 These complexes are obtained by substitution of the aromatic ring of the molecules, thus preserving the Zn-­binder fragment which is essential for the expression of their inhibitory HDAC properties. Complexes 23–26, synthesized by Spencer, are all inhibitors of various HDAC enzymes, which are grouped into eight different classes

Scheme 3.7  Ferrocenyl  complexes of HDACi.

View Online

Iron Compounds as Anticancer Agents

73

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

89–92

(HDAC1–HDAC8). The complexes are often more active than SAHA on some of these enzyme classes. Complex 25, for example, has IC50 values lower than that for SAHA on four of the HDAC classes studied, with a remarkably low value of 90 pM on HDAC6 (490 pM for SAHA). The in vitro cytotoxicity of these complexes on MCF-­7 cells is, however, slightly inferior to that of SAHA (IC50 = 3.1 µM versus 0.73 µM).91 A molecular docking study of 23 in HDAC8 has shown that its ferrocene substituent may overlap the aryl cap of SAHA. Study of its mechanism of action showed that it induced rapid production of ROS in the cells, followed by perturbation of the MMP and inhibition of autophagy.89 The ferrocene complexes 27 and 29, prepared in our laboratory, are themselves tamoxifen/SAHA hybrids in which the dimethylamino chain of Fc-­Tam is replaced by the hydroxyamino chain of SAHA (Scheme 3.7).87,88 The idea was to evaluate the effect of the combination of these two molecules on their biological activity. The results show potentiation of the cytotoxicity of 35 on MDA-­MB-­231 cells, with this complex showing higher toxicity than its precursors, Fc-­Tam and SAHA (IC50 = 0.7 µM for 27 versus 2.6 and 3.6 µM for Fc-­Tam and SAHA), while this is not the case for 28, its organic hybrid (IC50 = 8.6 µM). Complex 29 is equally cytotoxic on these cells (IC50 = 1.8 µM and 0.5 µM). The study of the mechanism of action of 27 and 28 shows that, like SAHA, they induce expression of the p21 gene, but are not, however, HDAC inhibitors. Their cytotoxicity thus seems more likely to be linked to an effect of the ferrocene entity.

3.2.5  Ferrocenyl Derivatives of Nucleosides Ferrocenyl derivatives of nucleosides were the subject of a recent review,93 although only a few of them have been studied, in the laboratories of Schmalz and Tucker, for their in vitro cytotoxic effect (Scheme 3.8).94–96 In these complexes, the ferrocene entity mimics the ribose fragment of the nucleosides cytidine, adenosine or thymidine. Complex 30 derived from cytidine has a cytotoxic effect on Burkitt-­like lymphoma cells (BJAB) and acute lymphoblastic leukaemia (ALL) (LD50 10–20 µM) and leads to apoptosis, while 32, the corresponding complex of thymidine, has no effect.94 As for ferrocene complexes bearing a hydroxyalkyl chain (33–37), the derivatives of adenosine (31) and thymidine (33) have comparable cytotoxicities on leukaemia cells (CEM cells; IC50 0.35–0.90 µM).96 Studies of a series of complexes with aliphatic chains of variable length have shown that 33, the complex with the longest chain, is cytotoxic on human osteosarcoma cells (IC50 = 4.4 µM) while those with shorter chains are not (34–37; IC50 = 58–86 µM).95 This complex stands out from the others for its higher lipophilicity (log P = 0.96 for 33, 0.03–0.05 for 34–37) and for a weaker redox potential of the ferrocene (E1/2 = 424 mV for 33 and 518 mV for 35). These two factors are thought to be responsible for the observed differences in cytotoxicity.95

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

74

Chapter 3

Scheme 3.8  Ferrocenyl  derivatives of nucleosides.

Scheme 3.9  Chemical  reactivity and selection of pro-­DLCs.

3.2.6  N-­Alkylaminoferrocenes The potential of N-­alkylaminoferrocenes (NAAFs) as anticancer prodrugs has recently been studied by Mokhir and co-­workers.97–100 Three molecules of interest (38–40) are shown in Scheme 3.9. The idea is to couple NAAFs, bearing a variable R substituent, to the ROS-­ sensitive group N-­carbonyloxymethyl phenylboronic acid pinacol ester, with the aim of creating pro-­DLCs (delocalized lipophilic cations, Scheme 3.9). These molecules can then be activated specifically in the cancer cells, which contain higer levels of ROS (essentially H2O2) than normal cells. Generation of this cation is accompanied by intracellular production of ROS that are more active than H2O2 (superoxide and hydroxy radical) and by formation of a quinone methide that is also toxic.99 These molecules are capable of inducing the death of cancer cells while sparing healthy cells, whose level of intracellular ROS is insufficient to generate DLCs. The three

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

75

DLCs (38–40) are distributed differently in the cells. Complex 39 is concentrated predominantly in lysosomes and 40 in mitochondria, while 38, the first to be synthesized, has no particular target. Studies of the antiproliferative effects of these molecules on leukaemia cells (HL-­60 and BL-­2) have shown that 38 is cytotoxic on these cells (IC50 = 11 and 25 µM) but not on normal cells (fibroblasts).98,99 This cytotoxicity is independent of the p53 status of the cells (HL-­60 is p53 negative while BL-­2 is wild type p53). In vivo studies on BDF1 mice bearing leukaemia L1210 cells have shown that injection of 38 resulted in a 28% increase in survival time in these animals (i.p. injection for 6 days, 26 mg kg−1).97 The prolonged survival was accompanied by increased oxidative stress and membrane damage in L1210 cells isolated from the ascites fluid of animals. Complex 39 causes a disruption of the lysosomes in DU-­145 (prostate cancer) cells, which is probably the principal cause of its high cytotoxicity (IC50 = 6.5 µM).98 Injection of 39 into mice bearing Nk/Ly lymphoma (i.p. injection every 2 days for 15 days, 40 mg kg−1) caused significant reduction of the tumours in six of seven treated animals. The synthesis of 40, a DLC complex of carboplatin, one of the three derivatives of platinum widely used as an anticancer drug, was recently published.100 In A2780 cells, 40 is more concentrated in the mitochondria compared to carboplatin alone (5.7 more Pt in the mitochondria of cells treated with 40 than in cells treated with carboplatin alone), resulting in a significant increase in its cytotoxicity (IC50 = 20 µM for 40 versus 82 µM for carboplatin alone).100

3.2.7  F  errocenyl Alkylpyridinium Cations Used for Photodynamic Therapy (PDT) Chakravarty and co-­workers recently prepared the first ferrocenyl conjugates active in PDT (Scheme 3.10).101 These are ferrocenyl alkylpyridinium cations 41 and 42 which show a strong absorption band in the visible range. Molecule 43 also possesses a phosphonium bromide, whose characteristics as a

Scheme 3.10  Ferrocenyl  alkylpyridinium cations used for PDT.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

76

Chapter 3

lipophilic cation it is hoped to enhance so as to favour its localization in the mitochondria and enable better expression of its effect in PDT. Complexes 41–43 show high photocytotoxicity after irradiation with visible light (400–700 nm). The most active complex after irradiation is 43 (IC50 = 1.3 µM on HeLa cells). It induces apoptosis of the cells by production of ROS. It is much less toxic on healthy cells (IC50 = 27.5 µM) as are non-­irradiated complex 3 (IC50 = 27.5 µM) and 44, the organic product (IC50 > 50 µM).

3.3  Coordination Complexes of Iron(ii) and Iron(iii) The inventory of the inorganic complexes of iron(ii) and iron(iii) synthesized as anticancer agents appeared in a recent review, apparently the only one yet published on this subject.22 It lists many complexes, but most often these are compounds that have been the subject of only a single paper reporting their synthesis and the evaluation of their cytotoxicity in vitro on cancer cell lines. We have chosen to focus here on just a few examples that have received more detailed study (Scheme 3.11). The first complexes in this series (45, 46) were published in 2005 by Chi-­ Ming Che.102,103 They are iron(ii) complexes of pentadentate N-­donor ligand that are intended to mimic the effects of bleomycin. These complexes have high cytotoxicity (IC50 = 0.6–3.4 µM), inducing apoptosis of the cells associated with damage to cellular DNA, cell cycle arrest, induction of mitogen-­ activated protein kinase signalling pathways and p53 tumour suppressor protein. The changes in gene expression induced by these complexes were found to be highly similar to that induced by ciclopirox, an iron chelator with antifungal properties.102 Salophen complexes of iron(ii) and iron(iii) (47–49) were prepared with the aim of mimicking the mechanism of action of cisplatin.104–106 The first two complexes (47 and 48) block the growth of solid tumour cells (MCF-­7, MDA-­ MB-­231, HT-­29) as well as those of lymphoma and leukaemia. They cause a strong release of Cu/Zn superoxide dismutase and induce apoptosis. The corresponding manganese complexes and the ligand alone are much less active. The effects of the complexes are independent of the oxidation state of the particular metal.104,105 The last complex of this family, 49, is highly cytotoxic for ovarian cancer cells SKOV-­3 (IC50 = 300 nM)106 and neuroblastoma107 in which it induces apoptosis of the cells, blocking them in the S phase of the cell cycle. Polypyridine complexes of various metals have been much studied for their potential as fluorescent tracers108,109 and molecules for use in PDT.110 The anticancer effects of a number of metal complexes of the polypyridines, in particular those of Pt,111 Cu,112 Ru113 and La,114,115 were described several years ago, and those of Fe more recently.116,117 Complexes 50–52 have strong antiproliferative effects on several cell lines (HeLa, U87, C6, SiHa) most particularly on Caski cells (cervical cancer carcinoma) with IC50 values between 0.25 and 0.75 µM. Complex 52 is the one with the best therapeutic index (highest ratio of IC50 values on cancer cells versus healthy cells). These

View Online

77

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

Scheme 3.11  Coordination  complexes of iron(ii) and iron(iii). lipophilic complexes are transported to the interior of the cell by the transferrin receptor. Their antiproliferative effects induce apoptosis, with blockage of the cells in the sub-­G1 phase, and could be linked to an inhibition of thioredoxin reductase (TrxR).118,119 Simultaneously with their work on the ferrocenyl conjugates of alkylpyridinium salts described above, Chakravarty and his group also synthesized catecholates of iron(iii) (53–60) active in PDT.120–122 These complexes

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

78

Chapter 3

include a polyaromatic, “DNA-­intercalating” part plus a bidentate dianionic catechol ligand which plays a dual role: first, it is crucial for the stabilization of the metal in the +3 oxidation state, and second it furnishes the ligand-­to-­metal charge transfer band essential in PDT. These complexes only become cytotoxic when irradiated by visible range (400–700 nm) or red (600–750 nm) light. Complex 55 (R = tBu) shows elevated toxicity on HeLa cells after irradiation (IC50 = 2.2 µM) although it is only slightly toxic without irradiation (IC50 = 85.4 µM). After 4 h incubation, 55 is localized essentially in the nucleus of cells where after irradiation it induces DNA cleavage (formation of DNA ladder) and cell death by apoptosis. In complexes 57 and 58, the cation triphenylphosphonium has been added to permit targeting of mitochondria.122 These complexes are effectively found in the mitochondria, but their cytotoxicity is less than that of 55 (IC50 around 8–10 µM on HeLa cells after irradiation) although higher than that of the non-­cationic complexes 59 and 60.

3.4  Molecules Active via Chelation with Iron The molecules of this family all possess a metal-­binding region that binds in the cell to iron(ii) or iron(iii) to give cytotoxic complexes.

3.4.1  Bleomycins (BLMs) The example that best illustrates this type of molecule is that of the bleomycins, a family of glycopeptide antibiotics discovered in 1966 and first isolated from Streptomyces verticillus.23 The bleomycins are used in association with other drugs in the treatment of a number of cancers (lymphomas, head and neck cancers and germ-­cell tumours). Their chemical structure is shown in Scheme 3.12. Studies of the mechanism of action of bleomycin (BLM) have been the subject of many publications and reviews.23,123 Bleomycin administered intravenously in its metal-­free form is rapidly complexed in blood plasma by CuII. This complex is then taken up by transporters which introduce it into the cell. After reduction of CuII to CuI, the latter is exchanged with iron(ii). This in turn is oxidized to iron(iii) in the presence of one-­electron reductant and dioxygen. The bleomycin is thus transformed into its activated species (bleomycin-­iron(iii)-­OOH), an iron octahedral complex in which five nitrogen atoms are donated by the bleomycin (Scheme 3.13). This intermediate has a half-­life of several minutes at 4 °C and is responsible for initiating single-­stranded and double-­stranded DNA damage. The damage is similar to that generated by ionizing radiation. The X-­ray structure of this intermediate, which has a half-­life of just a few minutes, has never been determined. However, the structure of the bleomycin/CuII complex has been solved, making it possible to visualize the organization of the ligands of the bleomycin around the metal (Scheme 3.12).23

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

79

Scheme 3.12  (Left)  Chemical structure of bleomycins, the five nitrogens in red

are involved in the formation of an iron octahedral complex, giving the activated bleomycin that is responsible for DNA cleavage in cells. (Right) X-­ray structure of the metal-­binding domain of bleomycin bound with Cu(ii), N: blue; O: red; C: green; Cu(ii) ion: magenta; chloride ion: cyan sphere (adapted from ref. 23 with permission from Springer Nature, Copyright 2005).

Scheme 3.13  Proposed  mechanisms for generation of activated bleomycin in vivo. Adapted from ref. 23 with permission from Springer Nature, Copyright 2005.

The role played in the toxicity of BLMs by the other bimetallic cations present in cells (CuII, ZnII, CoII) has been widely discussed. This has led to the synthesis of synthetic mimics of the metal-­binding domains of bleomycins which has made it possible to better understand their mechanism of action in cell cultures and cell-­free systems.124,125 This is the case for iron(ii)-­N4-­Py which has been particularly well studied and allowed it to be shown that, within cells, the metals of the other complexes exchange with iron(ii). It is thus the iron(ii)-­N4-­Py complex which appears to be the source of oxidative damage to the cells irrespective of the initial metal complex (MnII, CuII, ZnII) that was originally added to the cell culture medium.126

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

80

Chapter 3

Moreover, Osella and co-­workers have made the interesting observation that a 24 h preincubation of MCF-­7 cells in the presence of the ferrocenium salt 3 (Scheme 3.2) significantly increased the cytotoxicity of bleomycin. This translates to a cell survival percentage of 65% for incubation in the presence of bleomycin alone and 13% after preincubation with 3.45 This synergistic effect is due to the generation of iron cations within the cell arising from the degradation of 3 and confirms the role played by iron in the cytotoxicity of bleomycin.

3.4.2  Iron Chelators The iron chelator family of molecules was initially prepared for the treatment of pathologies associated with iron overload (e.g. beta-­thalassemia).127 These molecules also have anticancer effects which have been the subject of many in vitro and in vivo studies, as reported in a number of reviews.24,25,126,128 A perusal of the literature shows that several of them have undergone numerous clinical trials, some showing encouraging results.25 A selection of ligands of interest is shown in Scheme 3.14. These ligands can form octahedral iron complexes. They can be divided into three categories depending on the number of donor atoms in the ligand (2, 3 or 6) and give rise to bidentate, tridentate or hexadentate complexes with stoichiometry (1 : 3, 1 : 2, 1 : 1).

Scheme 3.14  Chemical  structure of iron chelators.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

81

The first molecule of this family to be studied was desferrioxamine (DFO). DFO is a naturally occurring hexadentate siderophore from the bacterium Streptomyces pilosus. This iron chelator is considered the gold standard for treatment of iron overload in patients suffering from beta-­ thalassemia although the need for repeated intravenous injections is a constraint. DFO is a hexadentate ligand which forms a complex with iron of stoichiometry 1 : 1. This complex possesses antiproliferative effects in vitro on neuroblastoma and leukaemia cells.129,130 Clinical trials carried out on neuroblastoma patients with DFO either alone or in combination with other anticancer molecules (carboplatin, etoposide, cyclophosphamide) produced contrasting results.25 One of the studies carried out on 57 patients proved particularly conclusive (24 complete response, 26 partial response, 3 minor response and 4 non-­responses) while the study on children proved negative. The clinical development of this molecule has however been limited by its poor bioavailability and short plasma half-­life (approximately 12 minutes). Deferasirox (DFX) is a synthetic molecule used to treat iron overload that has the advantages of being orally administered and having few secondary effects. This molecule is cytotoxic on hepatic cancer cells (HUH7) but not on primary hepatocyte cultures,131 raising the prospect of a specificity of action on cancer cells. Its mechanism of action involves the inactivation of the oncogenic nuclear factor-­κB (NF-­κB) and a modulation of the metabolism of the polyamines.132 DFX is also active in vivo in a mouse model of human myeloid leukaemia (U937 cells).133 Deferiprone (DFP), a third molecule approved for the treatment of thalassemias, also has an antiproliferative effect on tumour cells in culture.134 It is a bidentate ligand in the hydroxypyridone family which forms iron complexes of stoichiometry 3 : 1 in concentrated medium. In physiological medium, the latter tends to disassociate to complexes 2 : 2 and 2 : 1. This incomplete coordination of the iron favours the production of ROS and gives the complex high toxicity. Nevertheless it is currently under clinical trials for treatment of Parkinson's disease.135 The thiosemicarbazones are iron chelators that have long been studied as anticancer agents as well as for their antibacterial, antimalarial and antiviral effects.136 They remain of great interest in the scientific community and are the subject of a review published recently.128 This family, headed by triapine (3-­AP), has been much studied, particularly by Richardson's laboratory.25 3-­AP is a N,N,S tridentate ligand capable of chelating intracellular iron(ii) and iron(iii). It was initially developed as a potent RR inhibitor. The mechanism of action of 3-­AP is still not fully understood although it is clearly linked to the formation of an iron chelate. It may be related to the production of ROS and/or depletion of cellular iron pools. 3-­AP has undergone numerous phase I and II clinical trials for treatment of metastatic cancer, alone or in association with gemcitabine.137 These trials reveal limited efficacy but high toxicity (thrombocytopenia

View Online

Chapter 3

82

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

138–140

and neutropenia) for 3-­AP. However, recent encouraging results in triapine-­cisplatin-­radiation radiochemotherapy for advanced stage cervical cancer have led to the establishment of two clinical trials supported by the NCI.141 Structure–activity relationship studies of 3-­AP have led to the synthesis of many molecules some of which show particular promise, especially the molecules Dp44mT142,143 and DpC144 which have a N,N,S system of coordination identical to that of 3-­AP while being more active than 3-­AP in vitro and in vivo. It has been shown that Dp44mT does not lead to overall iron depletion of the tumour in vivo in animal models but acts to sequester intracellular iron to generate cytotoxic iron complexes. Thus its mechanism of action is different from that of typical chelators such as DFO that act solely to deplete tumour cell iron stores and inhibit cancer growth by that mechanism. DpC has recently entered phase I clinical trial evaluation. Finally, the innovative nature of the molecule COTI-­2, which has also recently entered a clinical trial, should be mentioned. Its structure was actually found via in silico computer-­aided drug design on CHEMAS, a computational platform.145 It seems reasonable to believe that the potential of this research topic is far from having been fully explored. But judging by what we have seen so far, the future appears to be very open.

3.5  Conclusion In the family of iron anticancer drugs presented here, the ferrocene complexes, emblematic molecules of organometallic chemistry, figure prominently. However, the other members of this family also bring to it a wide diversity of structure and mechanism of action. Among these are the coordination complexes of iron whose recently discovered potential for use in PDT appears promising. As for the iron chelates, their use as anticancer agents has been the subject of a number of clinical trials, without, however, leading to prescription use as yet. The discovery of thiosemicarbazones, ribonucleotide reductase inhibitors, provides new approaches that are showing concrete developments with the launch of new clinical trials. Unsurprisingly, the molecules attracting most attention are the prodrugs. The best example of these is bleomycin, which in the presence of a reducing agent and dioxygen, draws from the intracellular iron pool to generate in situ a highly reactive iron(iii) complex. The ferrocifens, for their part, are activated within cancer cells to quinone methides, reactive electrophilic species that are the source of their cytotoxicity, while the N-­alkylaminoferrocenes are activated by ROS that are present in greater amounts in cancer cells compared to healthy cells. The spectrum of iron anticancer drugs encompasses a vast field of research from both the synthetic and the mechanistic point of view, and augurs well for providing further discoveries that, at this stage of development, have no discernible limit.

View Online

Iron Compounds as Anticancer Agents

83

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Abbreviations BLM Bleomycin DMSO Dimethylsulfoxide Fc Ferrocene; η5-­(C5H5)2Fe GSH Glutathione HDAC Histone deacetylase HDACi Histone deacetylase inhibitor HRP Horseradish peroxidase ICP-­OES Inductively coupled plasma optical emission spectroscopy i.p. Intra-­peritoneally Jurkat cells Human leukaemic lymphoblastoid cells LNC Lipid nanocapsule MCF-­7 Hormone-­dependent breast cancer cells MDA-­MB-­231 Hormone-­independent breast cancer cells MMP Mitochondrial membrane potential PDT Photodynamic therapy PEG Polyethylene glycol PC3 Prostate cancer cells QM Quinone methide RR Ribonucleotide reductase ROS Reactive oxygen species SAHA Suberoylanilide hydroxamic acid SCID mice Severe combined immunodeficient mice SOD Superoxide dismutase SW620 Colon cancer cells TrxR Thioredoxin reductase TLMs Tamoxifen-­like metallocifens TNBC Triple negative breast cancer

Acknowledgements I wish to thank Gérard Jaouen, Michèle Salmain, Vincent Corcé and M. J. McGlinchey for helpful discussion. I would like to acknowledge the invaluable help provided by B. McGlinchey in the translation of the manuscript.

References 1. H. Haraguchi, J. Anal. At. Spectrom., 2004, 19, 5–14. 2. M. Kolberg, K. R. Strand, P. Graff and K. K. Andersson, Biochim. Biophys. Acta, Proteins Proteomics, 2004, 1699, 1–34. 3. W. Nam, Acc. Chem. Res., 2007, 40, 522–531. 4. W. Nam, Y.-­M. Lee and S. Fukuzumi, Acc. Chem. Res., 2014, 47, 1146–1154. 5. J. C. Kwok and D. R. Richardson, Crit. Rev. Oncol. Hematol., 2002, 42, 65–78.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

84

Chapter 3

6. A. N. Nesmeyanov, L. G. Bogomolova, V. Viltchevskaya, et al., US Pat., 119356, 1971. 7. C. Biot, W. Daher, N. Chavain, T. Fandeur, J. Khalife, D. Dive and E. De Clercq, J. Med. Chem., 2006, 49, 2845–2849. 8. F. Dubar and C. Biot, in Bioorganometallic Chemistry: Applications in Drug Discovery, Biocatalysis and Imaging, ed. G. Jaouen and M. Salmain, Wiley-­VCH, Weinheim, 2015, ch. 5, pp. 141–164. 9. C. Roux and C. Biot, Future Med. Chem., 2012, 4, 783–797. 10. A. Kondratskyi, K. Kondratska, F. V. Abeele, D. Gordienko, C. Dubois, R.-­ A. Toillon, C. Slomianny, S. Lemiere, P. Delcourt, E. Dewailly, R. Skryma, C. Biot and N. Prevarskaya, Sci. Rep., 2017, 7, 15896. 11. G. Jaouen, A. Vessieres and S. Top, Chem. Soc. Rev., 2015, 44, 8802–8817. 12. G. Jaouen and A. Vessieres, Pure Appl. Chem., 1985, 57, 1865–1874. 13. G. Jaouen, A. Vessieres and I. S. Butler, Acc. Chem. Res., 1993, 26, 361–369. 14. G. Jaouen, W. Beck and M. J. McGlinchey, in Bioorganometallics: Biomolecules, Labeling, Medicines, ed. G. Jaouen, Wiley-­VCH, Weinheim, 2006, ch. 1, pp. 1–37. 15. S. S. Braga and A. M. S. Silva, Organometallics, 2013, 32, 5626–5639. 16. K. Kowalski, Coord. Chem. Rev., 2018, 366, 91–108. 17. C. Ornelas, New J. Chem., 2011, 35, 1973–1985. 18. M. Patra and G. Gasser, Nat. Rev. Chem., 2017, 1, 0066. 19. N. Metzler-­Nolte and M. Salmain, in Ferrocenes, Ligands, Materials and Biomolecules, ed. P. Stepnicka, John Wiley & Sons, Chichester, UK, 2008, pp. 499–639. 20. E. A. Hillard, A. Vessieres and G. Jaouen, in Medicinal Organometallic Chemistry, ed. G. Jaouen and N. Metzler-­Nolte, 2010, vol. 32, pp. 81–117. 21. G. Jaouen and S. Top, in Advances in Organometallic Chemistry and Catalysis, ed. A. J. L. Pombeiro, Wiley, 2014, ch. 42, pp. 561–580. 22. W. A. Wani, U. Baig, S. Shreaz, R. A. Shiekh, P. F. Iqbal, E. Jameel, A. Ahmad, S. H. Mohd-­Setapar, M. Mushtaque and L. T. Huna, New J. Chem., 2016, 40, 1063–1090. 23. J. Y. Chen and J. Stubbe, Nat. Rev. Cancer, 2005, 5, 102–112. 24. V. Corce, S. G. Gouin, S. Renaud, F. Gaboriau and D. Deniaud, Bioorg. Med. Chem. Lett., 2016, 26, 251–256. 25. Y. Yu, E. Gutierrez, Z. Kovacevic, F. Saletta, P. Obeidy, Y. S. Rahmanto and D. R. Richardson, Curr. Med. Chem., 2012, 19, 2689–2702. 26. M. P. Gleeson, A. Hersey, D. Montanari and J. Overington, Nat. Rev. Drug Discovery, 2011, 10, 197–208. 27. S. M. Meier-­Menches, C. Gerner, W. Berger, C. G. Hartinger and B. K. Keppler, Chem. Soc. Rev., 2018, 47, 909–928. 28. C. A. Lipinski, F. Lombardo, B. W. Dominy and P. J. Feeney, Adv. Drug Delivery Rev., 1997, 23, 3–25. 29. G. R. Bickerton, G. V. Paolini, J. Besnard, S. Muresan and A. L. Hopkins, Nat. Chem., 2012, 4, 90–98.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

85

30. S. M. Meier, M. Hanif, Z. Adhireksan, V. Pichler, M. Novak, E. Jirkovsky, M. A. Jakupec, V. B. Arion, C. A. Davey, B. K. Keppler and C. G. Hartinger, Chem. Sci., 2013, 4, 1837–1846. 31. S. Clede, F. Lambert, R. Saint-­Fort, M. A. Plamont, H. Bertrand, A. Vessieres and C. Policar, Chem. -­Eur. J., 2014, 20, 8714–8722. 32. A. Bergamo and G. Sava, Chem. Soc. Rev., 2015, 44, 8818–8835. 33. A. Bergamo, P. J. Dyson and G. Sava, Coord. Chem. Rev., 2018, 360, 17–33. 34. T. J. Kealy and P. L. Pauson, Nature, 1951, 168, 1039–1040. 35. S. A. Miller, J. A. Tebboth and J. F. Tremaine, J. Chem. Soc., 1952, 632–635. 36. E. O. Fischer and W. Pfab, Z. Naturforsch., 1952, 7b, 377. 37. G. Wilkinson, M. Rosenblum, M. C. Whiting and R. B. Woodward, J. Am. Chem. Soc., 1952, 74, 2125–2126. 38. P. Kovacic, Med. Hypotheses, 2007, 69, 510–516. 39. V. Scalcon, M. Salmain, A. Folda, S. Top, P. Pigeon, H. Z. S. Lee, G. Jaouen, A. Bindoli, A. Vessieres and M. P. Rigobello, Metallomics, 2017, 9, 949–959. 40. R. A. Yeary, Toxicol. Appl. Pharmacol., 1969, 15, 666–676. 41. R. P. Hanzlik and W. H. Soine, J. Am. Chem. Soc., 1978, 100, 1290–1291. 42. P. Köpf-­Maier, H. Köpf and E. W. Neuse, Angew. Chem., Int. Ed. Engl., 1984, 23, 456–457. 43. B. Rosenberg, L. Vancamp and T. Krigas, Nature, 1965, 205, 698–699. 44. B. Rosenberg and L. VanCamp, Cancer Res., 1970, 30, 1799–1802. 45. G. Tabbi, C. Cassino, G. Cavigiolio, D. Colangelo, A. Ghiglia, I. Viano and D. Osella, J. Med. Chem., 2002, 45, 5786–5796. 46. E. I. Edwards, R. Epton and G. Marr, J. Organomet. Chem., 1975, 85, C23–C25. 47. E. I. Edwards, R. Epton and G. Marr, J. Organomet. Chem., 1976, 122, C49–C53. 48. S. Top, J. Tang, A. Vessières, D. Carrez, C. Provot and G. Jaouen, Chem. Commun., 1996, 955–956. 49. S. Top, A. Vessières, G. Leclercq, J. Quivy, J. Tang, J. Vaissermann, M. Huché and G. Jaouen, Chem. -­Eur. J., 2003, 9, 5223–5236. 50. M. Patra and G. Gasser, Nat. Rev. Chem., 2017, 1, 12. 51. A. Vessières, J. Organomet. Chem., 2013, 734, 3–16. 52. A. Nguyen, A. Vessières, E. A. Hillard, S. Top, P. Pigeon and G. Jaouen, Chimia, 2007, 61, 716–724. 53. V. C. Jordan, Br. J. Pharmacol., 2006, 147, S269–S276. 54. S. Top, A. Vessières, C. Cabestaing, I. Laios, G. Leclercq, C. Provot and G. Jaouen, J. Organomet. Chem., 2001, 637, 500–506. 55. C. Lu, J. M. Heldt, M. Guille-­Collignon, F. Lemaitre, G. Jaouen, A. Vessieres and C. Amatore, ChemMedChem, 2014, 9, 1286–1293. 56. A. Vessieres, C. Corbet, J. M. Heldt, N. Lories, N. Jouy, I. Laios, G. Leclercq, G. Jaouen and R. A. Toillon, J. Inorg. Biochem., 2010, 104, 503–511. 57. E. Allard, C. Passirani, E. Garcion, P. Pigeon, A. Vessières, G. Jaouen and J. P. Benoit, J. Controlled Release, 2008, 130, 146–153.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

86

Chapter 3

58. N. T. Huynh, C. Passirani, P. Saulnier and J. P. Benoit, Int. J. Pharm., 2009, 379, 201–209. 59. A. L. Laine, E. Adriaenssens, A. Vessieres, G. Jaouen, C. Corbet, E. Desruelles, P. Pigeon, R. A. Toillon and C. Passirani, Biomaterials, 2013, 34, 6949–6956. 60. M. Gormen, P. Pigeon, Y. Wang, A. Vessieres, S. Top, F. Martial, C. Gros, M. J. McGlinchey and G. Jaouen, Eur. J. Inorg. Chem., 2017, 454–465. 61. E. A. Hillard, A. Vessières, S. Top, P. Pigeon, K. Kowalski, M. Huché and G. Jaouen, J. Organomet. Chem., 2007, 692, 1315–1326. 62. G. Jaouen, S. Top, A. Vessières, G. Leclercq, J. Quivy, L. Jin and A. Croisy, C. R. Acad. Sci., Ser. IIc: Chim., 2000, 89–93. 63. M. Görmen, P. Pigeon, S. Top, E. A. Hillard, M. Huché, C. G. Hartinger, F. de Montigny, M.-­A. Plamont, A. Vessières and G. Jaouen, ChemMedChem, 2010, 5, 2039–2050. 64. E. Hillard, A. Vessieres, L. Thouin, G. Jaouen and C. Amatore, Angew. Chem., Int. Ed., 2006, 45, 285–290. 65. P. Messina, E. Labbe, O. Buriez, E. A. Hillard, A. Vessieres, D. Hamels, S. Top, G. Jaouen, Y. M. Frapart, D. Mansuy and C. Amatore, Chem. -­Eur. J., 2012, 18, 6581–6587. 66. Y. Wang, M. A. Richard, S. Top, P. M. Dansette, P. Pigeon, A. Vessieres, D. Mansuy and G. Jaouen, Angew. Chem., Int. Ed., 2016, 55, 10431–10434. 67. A. Citta, A. Folda, A. Bindoli, P. Pascal Pigeon, S. Top, A. Vessières, M. Salmain, G. Jaouen and M. P. Rigobello, J. Med. Chem., 2014, 57, 8849–8859. 68. C. Bruyère, V. Mathieu, A. Vessières, P. Pigeon, S. Top, G. Jaouen and R. Kiss, J. Inorg. Biochem., 2014, 141, 144–151. 69. N. T. Huynh, M. Morille, J. Bejaud, P. Legras, A. Vessières, G. Jaouen, J. P. Benoit and C. Passirani, Pharm. Res., 2011, 28, 3189–3198. 70. N. T. Huynh, C. Passirani, E. Allard-­Vannier, L. Lemaire, J. Roux, E. Garcion, A. Vessières and J. P. Benoit, Int. J. Pharm., 2012, 423, 55–62. 71. E. Allard, D. Jarnet, A. Vessieres, S. Vinchon-­Petit, G. Jaouen, J.-­P. Benoit and C. Passirani, Pharm. Res., 2010, 27, 56–64. 72. M. Beauperin, S. Top, M. A. Richard, D. Plazuk, P. Pigeon, S. Toma, V. Polackova and G. Jaouen, Dalton Trans., 2016, 45, 13126–13134. 73. D. Plazuk, A. Vessières, F. Le Bideau, G. Jaouen and J. Zakrzewski, Tetrahedron Lett., 2004, 45, 5425–5427. 74. D. Plazuk, A. Vessières, E. A. Hillard, O. Buriez, E. Labbé, P. Pigeon, M. A. Plamont, C. Amatore, J. Zakrzewski and G. Jaouen, J. Med. Chem., 2009, 52, 4964–4967. 75. V. Scalcon, A. Citta, A. Folda, A. Bindoli, M. Salmain, I. Ciofini, S. Blanchard, J. D. Cazares-­Marinero, Y. Wang, P. Pigeon, G. Jaouen, A. Vessieres and M. P. Rigobello, J. Inorg. Biochem., 2016, 165, 146–151. 76. A.-­L. Lainé, A. Clavreul, A. Rousseau, C. Tétaud, A. Vessières, E. Garcion, G. Jaouen, L. Aubert, M. Guilbert, J. P. Benoit, R. A. Toillon and C. Passirani, Nanomedicine, 2014, 10, 1667–1677.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

87

77. Y. Wang, P. M. Dansette, P. Pigeon, S. Top, M. J. McGlinchey, D. Mansuy and G. Jaouen, Chem. Sci., 2018, 9, 70–78. 78. P. Pigeon, S. Top, A. Vessieres, M. Huche, M. Gormen, M. El Arbi, M. A. Plamont, M. J. McGlinchey and G. Jaouen, New J. Chem., 2011, 35, 2212–2218. 79. M. El Arbi, P. Pigeon, S. Top, A. Rhouma, S. Aifa, A. Rebai, A. Vessieres, M. A. Plamont and G. Jaouen, J. Organomet. Chem., 2011, 696, 1038–1048. 80. D. Plazuk, A. Wieczorek, A. Blauz and B. Rychlik, MedChemComm, 2012, 3, 498–501. 81. D. Plazuk, A. Wieczorek, W. M. Ciszewski, K. Kowalczyk, A. Blauz, S. Pawledzio, A. Makal, C. Eurtivong, H. J. Arabshahi, J. Reynisson, C. G. Hartinger and B. Rychlik, ChemMedChem, 2017, 12, 1882–1892. 82. A. Wieczorek, A. Blauz, A. Zal, H. J. Arabshahi, J. Reynisson, C. G. Hartinger, B. Rychlik and D. Plazuk, Chem. -­Eur. J., 2016, 22, 11413–11421. 83. M. Beauperin, D. Polat, F. Roudesly, S. Top, A. Vessieres, J. Oble, G. Jaouen and G. Poli, J. Organomet. Chem., 2017, 839, 83–90. 84. A. Wieczorek, A. Blauz, A. Makal, B. Rychlik and D. Plazuk, Dalton Trans., 2017, 46, 10847–10858. 85. C. Reiter, A. C. Karagoez, T. Froehlich, V. Klein, M. Zeino, K. Viertel, J. Held, B. Mordmueller, S. E. Ozturk, H. Anil, T. Efferth and S. B. Tsogoeva, Eur. J. Med. Chem., 2014, 75, 403–412. 86. C. Reiter, T. Froehlich, M. Zeino, M. Marschall, H. Bahsi, M. Leidenberger, O. Friedrich, B. Kappes, F. Hampel, T. Efferth and S. B. Tsogoeva, Eur. J. Med. Chem., 2015, 97, 164–172. 87. J. Cazares Marinero, M. Lapierre, V. Cavailles, R. Saint-­Fort, A. Vessieres, S. Top and G. Jaouen, Dalton Trans., 2013, 42, 15489–15501. 88. J. Cazares-­Marinero, S. Top, A. Vessieres and G. Jaouen, Dalton Trans., 2014, 43, 817–830. 89. M. Librizzi, A. Longo, R. Chiarelli, J. Amin, J. Spencer and C. Luparello, Chem. Res. Toxicol., 2012, 25, 2608–2616. 90. C. A. Ocasio, S. Sansook, R. Jones, J. M. Roberts, T. G. Scott, N. Tsoureas, P. Coxhead, M. Guille, G. J. Tizzard, S. J. Coles, H. Hochegger, J. E. Bradner and J. Spencer, Organometallics, 2017, 36, 3276–3283. 91. J. Spencer, J. Amin, M. H. Wang, G. Packham, S. S. S. Alwi, G. J. Tizzard, S. J. Coles, R. M. Paranal, J. E. Bradner and T. D. Heightman, ACS Med. Chem. Lett., 2011, 2, 358–362. 92. J. Spencer, J. Amin, R. Boddiboyena, G. Packham, B. E. Cavell, S. S. S. Alwi, R. M. Paranal, T. D. Heightman, M. H. Wang, B. Marsden, P. Coxhead, M. Guille, G. J. Tizzard, S. J. Coles and J. E. Bradner, MedChemComm, 2012, 3, 61–64. 93. K. Kowalski, Coord. Chem. Rev., 2016, 317, 132–156. 94. P. James, J. Neudoerfl, M. Eissmann, P. Jesse, A. Prokop and H.-­G. Schmalz, Org. Lett., 2006, 8, 2763–2766. 95. J. L. Kedge, H. V. Nguyen, Z. Khan, L. Male, M. K. Ismail, H. V. Roberts, N. J. Hodges, S. L. Horswell, Y. Mehellou and J. H. R. Tucker, Eur. J. Inorg. Chem., 2017, 466–476.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

88

Chapter 3

96. H. V. Nguyen, A. Sallustrau, J. Balzarini, M. R. Bedford, J. C. Eden, N. Georgousi, N. J. Hodges, J. Kedge, Y. Mehellou, C. Tselepis and J. H. R. Tucker, J. Med. Chem., 2014, 57, 5817–5822. 97. S. Daum, V. F. Chekhun, I. N. Todor, N. Y. Lukianova, Y. V. Shvets, L. Sellner, K. Putzker, J. Lewis, T. Zenz, I. A. M. de Graaf, G. M. M. Groothuis, A. Casini, O. Zozulia, F. Hampel and A. Mokhir, J. Med. Chem., 2015, 58, 2015–2024. 98. S. Daum, M. S. V. Reshetnikov, M. Sisa, T. Dumych, M. D. Lootsik, R. Bilyy, E. Bila, C. Janko, C. Alexiou, M. Herrmann, L. Sellner and A. Mokhir, Angew. Chem., Int. Ed., 2017, 56, 15545–15549. 99. H. Hagen, P. Marzenell, E. Jentzsch, F. Wenz, M. R. Veldwijk and A. Mokhir, J. Med. Chem., 2012, 55, 924–934. 100. V. Reshetnikov, S. Daum, C. Janko, W. Karawacka, R. Tietze, C. Alexiou, S. Paryzhak, T. Dumych, R. Bilyy, P. Tripal, B. Schmid, R. Palmisano and A. Mokhir, Angew. Chem., Int. Ed., 2018, 57, 11943–11946. 101. B. Balaji, B. Balakrishnan, S. Perumalla, A. A. Karande and A. R. Chakravarty, Eur. J. Inorg. Chem., 2015, 1398–1407. 102. W.-­L. Kwong, C.-­N. Lok, C.-­W. Tse, E. L.-­M. Wong and C.-­M. Che, Chem. -­Eur. J., 2015, 21, 3062–3072. 103. E. L. M. Wong, G. S. Fang, C. M. Che and N. Y. Zhu, Chem. Commun., 2005, 4578–4580. 104. A. Hille, I. Ott, A. Kitanovic, I. Kitanovic, H. Alborzinia, E. Lederer, S. Wolfl, N. Metzler-­Nolte, S. Schafer, W. S. Sheldrick, C. Bischof, U. Schatzschneider and R. Gust, J. Biol. Inorg. Chem., 2009, 14, 711–725. 105. A. Hille, T. Wolf, P. Schumacher, I. Ott, R. Gust and B. Kircher, Arch. Pharm., 2011, 344, 217–223. 106. T. S. Lange, K. K. Kim, R. K. Singh, R. M. Strongin, C. K. McCourt and L. Brard, PLoS One, 2008, 3, 2303. 107. K. K. Kim, R. K. Singh, R. M. Strongin, R. G. Moore, L. Brard and T. S. Lange, PLoS One, 2011, 6, 10. 108. E. Licandro, M. Panigati, M. Salmain and A. Vessieres, in Bioorganometallic Chemistry: Applications in Drug Discovery, Biocatalysis and Imaging, ed. G. Jaouen and M. Salmain, Wiley-VCH, Weinheim, 2015, pp. 339–392. 109. K. K. W. Lo, Acc. Chem. Res., 2015, 48, 2985–2995. 110. F. Heinemann, J. Karges and G. Gasser, Acc. Chem. Res., 2017, 50, 2727–2736. 111. S. Roy, K. D. Hagen, P. U. Maheswari, M. Lutz, A. L. Spek, J. Reedijk and G. P. van Wezel, ChemMedChem, 2008, 3, 1427–1434. 112. M. Devereux, D. O. Shea, A. Kellett, M. McCann, M. Walsh, D. Egan, C. Deegan, K. Kgdziora, G. Rosair and H. Mulller-­Bunz, J. Inorg. Biochem., 2007, 101, 881–892. 113. L. L. Zeng, Y. Chen, J. P. Liu, H. Y. Huang, R. L. Guan, L. N. Ji and H. Chao, Sci. Rep., 2016, 6, 19449. 114. P. Heffeter, M. A. Jakupec, W. Korner, S. Wild, N. G. von Keyserlingk, L. Elbling, H. Zorbas, A. Korynevska, S. Knasmuller, H. Sutterluty, M. Micksche, B. K. Keppler and W. Berger, Biochem. Pharmacol., 2006, 71, 426–440.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

Iron Compounds as Anticancer Agents

89

115. P. Heffeter, M. A. Jakupec, W. Korner, P. Chiba, C. Pirker, R. Dornetshuber, L. Elbling, H. Sutterluty, M. Micksche, B. K. Keppler and W. Berger, Biochem. Pharmacol., 2007, 73, 1873–1886. 116. J. J. Chen, Z. D. Luo, Z. N. Zhao, L. N. Xie, W. J. Zheng and T. F. Chen, Biomaterials, 2015, 71, 168–177. 117. L. N. Xie, Z. D. Luo, Z. N. Zhao and T. F. Chen, J. Med. Chem., 2017, 60, 202–214. 118. C. Gabbiani, G. Mastrobuoni, F. Sorrentino, B. Dani, M. P. Rigobello, A. Bindoli, M. A. Cinellu, G. Pieraccini, L. Messori and A. Casini, MedChemComm, 2011, 2, 50–54. 119. E. Schuh, C. Pfluger, A. Citta, A. Folda, M. P. Rigobello, A. Bindoli, A. Casini and F. Mohr, J. Med. Chem., 2012, 55, 5518–5528. 120. U. Basu, I. Khan, A. Hussain, P. Kondaiah and A. R. Chakravarty, Angew. Chem., Int. Ed., 2012, 51, 2658–2661. 121. U. Basu, I. Pant, I. Khan, A. Hussain, P. Kondaiah and A. R. Chakravarty, Chem. -­Asian J., 2014, 9, 2494–2504. 122. U. Basu, I. Pant, P. Kondaiah and A. R. Chakravarty, Eur. J. Inorg. Chem., 2016, 1002–1012. 123. R. M. Burger, J. Peisach and S. B. Horwitz, J. Biol. Chem., 1981, 256, 1636–1644. 124. Q. A. Li, T. A. van den Berg, B. L. Feringa and G. Roelfes, Dalton Trans., 2010, 39, 8012–8021. 125. Q. Li, M. G. P. van der Wijst, H. G. Kazernier, M. G. Rots and G. Roelfes, ACS Chem. Biol., 2014, 9, 1044–1051. 126. A. Geersing, N. Segaud, M. G. P. van der Wijst, M. G. Rots and G. Roelfes, Inorg. Chem., 2018, 57, 7748–7756. 127. J. B. Porter, Blood Rev., 2009, 23, S3–S7. 128. P. Heffeter, V. F. S. Pape, E. A. Enyedy, B. K. Keppler, G. Szakacs and C. R. Kowol, Antioxid. Redox Signaling, 2019, 30, 1062–1082. 129. J. Blatt and S. Stitely, Cancer Res., 1987, 47, 1749–1750. 130. Z. Estrov, A. Tawa, X. H. Wang, I. D. Dube, H. Sulh, A. Cohen, E. W. Gelfand and M. H. Freedman, Blood, 1987, 69, 757–761. 131. K. Chantrel-­Groussard, F. Gaboriau, N. Pasdeloup, R. Havouis, H. Nick, J.-­L. Pierre, P. Brissot and G. Lescoat, Eur. J. Pharmacol., 2006, 541, 129–137. 132. E. Messa, S. Carturan, C. Maffe, M. Pautasso, E. Bracco, A. Roetto, F. Messa, F. Arruga, I. Defilippi, V. Rosso, C. Zanone, A. Rotolo, E. Greco, R. M. Pellegrino, D. Alberti, G. Saglio and D. Cilloni, Haematologica, 2010, 95, 1308–1316. 133. J. H. Ohyashiki, C. Kobayashi, R. Hamamura, S. Okabe, T. Tauchi and K. Ohyashiki, Cancer Sci., 2009, 100, 970–977. 134. J. F. Zeidner, J. E. Karp, A. L. Blackford, B. D. Smith, I. Gojo, S. D. Gore, M. J. Levis, H. E. Carraway, J. M. Greer, S. P. Ivy, K. W. Pratz and M. A. McDevitt, Haematologica, 2014, 99, 672–678. 135. A. Martin-­Bastida, R. J. Ward, R. Newbould, P. Piccini, D. Sharp, C. Kabba, M. C. Patel, M. Spino, J. Connelly, F. Tricta, R. R. Crichton and D. T. Dexter, Sci. Rep., 2017, 7, 1398.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00062

90

Chapter 3

136. H. Beraldo and D. Gambino, Mini-­Rev. Med. Chem., 2004, 4, 31–39. 137. D. S. Kalinowski and D. R. Richardson, Pharmacol. Rev., 2005, 57, 547–583. 138. F. N. Alrefaie, B. Wonke and A. V. Hoffbrand, Eur. J. Haematol., 1994, 53, 298–301. 139. Z. Estrov, A. Cohen, E. W. Gelfand and M. H. Freedman, Toxicol. In Vitro, 1988, 2, 131–134. 140. J. A. Walker, R. A. Sherman and R. P. Eisinger, Am. J. Kidney Dis., 1985, 6, 254–256. 141. C. A. Kunos and S. P. Ivy, Front. Oncol., 2018, 8, 149. 142. M. Whitnall, J. Howard, P. Ponka and D. R. Richardson, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 14901–14906. 143. J. Yuan, D. B. Lovejoy and D. R. Richardson, Blood, 2004, 104, 1450–1458. 144. Z.-­L. Guo, D. R. Richardson, D. S. Kalinowski, Z. Kovacevic, K. C. Tan-­Un and G. C.-­F. Chan, J. Hematol. Oncol., 2016, 9, 98. 145. K. Y. Salim, S. M. Vareki, W. R. Danter and J. Koropatnick, Oncotarget, 2016, 7, 41363–41379.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Chapter 4

Recent Advances in Anticancer Copper Compounds Andrew Kellett*a, Zara Molphya, Vickie McKeea,b and Creina Slatora a

School of Chemical Sciences and National Institute for Cellular Biotechnology, Dublin City University, Glasnevin, Dublin 9, Ireland; bDepartment of Physics, Chemistry and Pharmacy, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark *E-­mail: [email protected]

4.1  I ntroduction—Copper Complexes as Redox-­ active Cytotoxins Copper complexes are emerging as an attractive chemotype for the treatment of human cancer. Part of this interest stems from their biologically accessible redox properties, wide structural variability, and bioavailability. A wide range of therapeutically relevant copper complexes are known and several in-­depth reviews have summarised their potential.1–3 Here, we focus on the activity profiles of several anticancer copper complexes that have undergone 60-­cancer cell line screening by the National Cancer Institute's (NCI) Developmental Therapeutics Program (DTP). This analysis provides  comprehensive pre-­clinical evaluation in a variety of cells including leukae­ mia, non-­small cell lung (NSCL), colon, central nervous system, melanoma, ovarian, renal, prostate, and breast. A second advantage of this approach

  Metallobiology Series No. 14 Metal-­based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-­Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

91

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

92

Chapter 4

is that the spectrum of results can be analysed using the COMPARE algorithm4 to provide a comprehensive insight into potential mechanisms of action. This is achieved by measuring correlations in activity with other NCI compounds—including those of standard agents,5 marketed drugs, mechanistic, and diversity datasets. Our interest in developing chemotherapeutic copper compounds stems from seminal work by David Sigman who identified the first synthetic chemical nuclease [Cu(1,10-­phenanthroline)2]2+ (Cu-­phen).6,7 This complex is known to semi-­intercalate and oxidise DNA from the minor groove but requires the  presence of a reductant to release the active Cu(i) form. Several biochemical applications of Cu-­phen have been identified so far, including DNA footprinting8 and recombinant protein production.9 Significant research efforts  to develop therapeutically relevant copper 1,10-­phenanthroline complexes have uncovered several lead agents that have undergone NCI-­60 analysis. In several cases, the COMPARE algorithm tells us these agents are mechanistically unique and analysis conducted in our laboratories has implicated reactive oxygen species (ROS), mitochondrial poisoning, and double strand break (DSB) formation within genomic DNA in their cellular death mechanisms. In addition to NCI-­60 screened copper complexes, a number of other structurally relevant compounds are discussed, along with mechanistic aspects of their chemotherapeutic activity.

4.2  C  opper Enzymes and Transport Proteins: Pathways for Developing Redox-­active Therapeutics Copper is an essential element required chiefly for its redox properties; it is second only to iron in electron transport metalloproteins and in enzymes concerned with dioxygen transport, utilisation, and disposal.10–12 Since copper can generate ROS, it is not found as the free ion inside cells, and a system of copper transport and storage proteins and other ligands (such as glutathione) is required to maintain homeostasis. The copper transport system comprises a passive membrane-­bound transporter hCtr1 which delivers copper to chaperone proteins such as the human Atox1 (Hah1), which in turn transfers copper to the ATP-­driven membrane transporters ATP7A and ATP7B for incorporation into proteins or for removal from the cell.13,14 Atox1 has also been shown to deliver copper into the nucleus and act as a transcription factor. The copper-­binding sites in these proteins are sulfur-­rich; similarly, Atox1, ATP7A, and ATP7B have conserved Cys-­X-­X-­ Cys copper-­binding motifs, while hCtr1 includes methionine-­rich copper-­ binding sites. Given the soft HSAB (hard and soft (Lewis) acid and base) nature of these donors, it is probable that transport occurs in the Cu(i) state. The Cys-­X-­X-­Cys sites in Atox1 lead to a two-­coordinate copper site in monomeric systems or a four-­coordinate site in dimeric assemblies, the

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

93

latter suggesting a possible mechanism for copper ion transfer between proteins. The copper transport systems are also implicated in the mechanisms of resistance to platinum drugs, as well as the route by which cisplatin can enter cells.14,15 Certainly, cisplatin and oxaliplatin interact with Atox1 in solution15,16 and platinum ions can bind at the copper-­binding sites,16–18 but transport of [Pt(NH3)2]2+ has not been unambiguously established. A structure showing a [Pt(NH3)2]2+ unit bound at the Cu-­binding site of Atox1 has been published, but later revisions suggest the data may be better interpreted as a single copper ion (or a single platinum ion).15–19 Nevertheless, the initial structure indicates how cisplatin might “piggy-­back” on this transport system. Recent studies suggest glutathione-­mediated simultaneous cisplatin and copper binding to Atox1 is possible, forming a Cu/ Pt/S(Cys) cluster.15,20 This may present a better scenario for [Pt(NH3)2]2+ binding to the protein with retention of square planar geometry and the two ammine ligands. The +i and +ii states of copper are readily accessible under biological conditions (and +iii is also achievable) but, for each of the biologically active copper metalloproteins, control of the redox potential is critical. In synthetic chemistry, adjustment of the redox potential is generally approached by modification of the ligands (e.g. HSAB properties) but in metalloproteins the available donor set is restricted and additional tuning of redox potential is achieved through control of coordination geometry. Copper(i) is a d10 system and insensitive to coordination geometry but the d9 copper(ii) ion has a geometric preference for square-­based geometry (tetragonal or square pyramidal) and significant deviations from this lead to an increase in redox potential making it easier to reduce the copper(ii) ions. A second requirement of electron transfer metalloproteins is that the electron transfer should be fast. The Franck–Condon principle suggests that this will be the case where there is no change in geometry between oxidation states, so that the reorganisational energy is minimised. In bioinorganic chemistry, these ideas were first applied to the type 1 copper proteins (specifically plastocyanin, Figure 4.1A), where the reduced and oxidised geometries are essentially identical, the coordination sphere includes soft and medium HSAB donors, and the coordination geometry is imposed by the protein.21 Similar considerations, however, apply to any redox-­active copper system. The broad range of mono-­, di-­, and multi-­nuclear active sites in copper metalloproteins has been thoroughly reviewed recently10–12,22 so only selected examples will be briefly described here. Of particular relevance to the rest of this chapter are a range of type 2 copper enzymes where dioxygen is activated by binding to copper in a mononuclear active site. These include monooxygenases, one dioxygenase, galactose oxidase, and amine oxidases, which all have ping-­pong reaction mechanisms, where the copper oxidation state cycles between I and II. Spectroscopic, kinetic, and theoretical data, along

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

View Online

Figure 4.1  Active  sites of: (A) Cu(ii) poplar plastocyanin (Protein Data Bank (https://www.rcsb.org/); PDB 4DP9); (B) PHM with dioxygen bound (PDB 1SDW); (C) quercetinase complexed with quercetin (PDB 1H1I); (D) amine oxidase with TPQ cofactor (PDB 1IVX); (E) galactose oxidase with crosslinked Tyr-­Cys (PDB 1GOG); (F) “Cu(ii) dioxo” intermediate fungal polysaccharide monooxygenase (PDB 5TKH); (G) Cu(ii) form of human Cu,Zn SOD1(PDB 1hl5); (H) Cu,Zn SOD ping-­pong mechanism; (I) copper monooxygenase mechanism modified from ref. 11. Copper ions shown as green or brown spheres for Cu(ii) and Cu(i), respectively.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

95

with model studies, suggest the key intermediate is the superoxide adduct Cu(ii)–O2•−, although species such as Cu(ii)–O• have also been postulated.10,11 The geometry and ligation at the copper ion tune the redox potential to allow generation of the Cu(ii)–O2•− intermediate. Representative examples of the active site structures are shown in Figure 4.1; all contain at least two histidine ligands and a vacant (or labile) site where dioxygen can bind. The variation in other ligation, neighbouring redox-­active species, and geometry tunes the copper centre for the particular reactions required. The three well-­studied monooxygenases, peptidylglycine α-­hydroxylating monooxygenase (PHM), dopamine β-­monooxygenase (DβM), and tyramine β-­monooxygenase (TβM) catalyse regio-­ and stereospecific hydroxylations of secondary C–H bonds in the synthesis of neurotransmitters and hormones. Each contains non-­coupled binuclear active sites, i.e. there is a second copper electron transfer site about 11 Å from the reaction centre (Figure 4.1B) but there is no bridge or electronic coupling between the two. All have similar ping-­pong mechanisms (Figure 4.1I): first both Cu(ii) ions of the fully oxidised state are reduced to Cu(i) by ascorbate; next the dioxygen binds to the Cu(i) site and the substrate binds nearby; the Cu(ii)–O2•− intermediate promotes hydrogen atom transfer from the substrate, forming Cu(ii)–OOH− and substrate radical. The reaction proceeds either by: (1) homolytic cleavage of the O–O bond, forming a cupric oxyl intermediate (Cu(ii)–O•) which reacts with the substrate radical to give the product, or (2) direct reaction of the hydroperoxide with the substrate, giving hydroxylated product and Cu(ii)–O•. Both cycles are completed by addition of an external proton and electron.10,23 Quercetinase is the only copper dioxygenase known and one of only two copper proteins known to bind a carboxylate residue. A structure of the enzyme–substrate complex (Figure 4.1C) shows the five-­coordinate Cu(ii) ion with the deprotonated quercetin bound as a monodentate ligand, and also hydrogen bonded to the glutamine. The carboxylate assists in deprotonating the quercetin substrate and also tunes the redox potential of the copper ion; unusually, this species interacts with dioxygen without needing a reducing agent. It seems the non-­innocent nature of the quercetin (Q) allows equilibrium between Cu(ii)–Q and Cu(i)–Q•, in which an electron has transferred from the ligand to the copper, with the Cu(ii)–Q complex as the major component. The enzyme mechanism is not yet fully understood; the dioxygen probably reacts with the minor Cu(i)–Q• component to generate a Cu(ii)– O2•− intermediate (although reaction with the substrate radical, forming an organic peroxide, may also be possible). Amine oxidases and galactose oxidase achieve a two-­electron oxidation of substrates (amine to aldehyde and alcohol to aldehyde, respectively) at a single copper centre.10,22 This requires an additional redox-­active cofactor in the active sites, generated as a post-­translational modification catalysed by the copper centre via a Cu(ii)–O2•− intermediate. In amine oxidases a tyrosine residue is oxidised (by six electrons) to 2,4,5-­trihydroxylphenylalanine

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

96

Chapter 4

quinone (TPQ) that sits in the active site but is not bound to the copper ion (Figure 4.1D); in galactose oxidase a coordinated tyrosine is crosslinked to cysteine (Figure 4.1E). Copper zinc superoxide dismutase, Cu,ZnSOD (Figure 4.1G), is among the best-­studied copper enzymes and its function is to remove superoxide ions via a very fast ping-­pong mechanism (Figure 4.1H) in which the copper(ii) site is first reduced and then re-­oxidised by superoxide. The step in which Cu(ii) is reduced to Cu(i) and O2•− oxidised to O2 is thought to occur via a Cu(ii)–O2•− intermediate (i.e. the reverse of the processes described above), but the second O2•− ion is reduced by an outer sphere mechanism and does not bind to the Cu(i) centre.11,24 Once again, the redox potential of the copper site is important for the enzyme function: while superoxide is a strong oxidising agent, it is a weaker reductant and the catalytic cycle depends on both halves of the process operating efficiently. In this case, the breaking and reforming of the histidine bridge helps to control the redox potential; in turn, the pK a of this histidine is tuned by the zinc ion. Direct structural evidence for the active intermediate in any of these enzymes is difficult to obtain given its reactive nature and the susceptibility of copper metalloproteins to photoreduction by X-­rays. In the case of PHM an X-­ray structure is available of the Cu(ii)–O2•− intermediate with the slow-­ reacting substrate N-­acetyl-­3,5-­diiodotyrosyl-­d-­threonine also bound near the copper (Figure 4.1F).25 It shows the expected end-­on bonding of the dioxygen in a suitable orientation to abstract the relevant substrate hydrogen atom, though the bond length is a little short for superoxide (O–O = 1.23 Å) and the oxidation states are ambiguous. A similar site has been characterised for bacterial and fungal lytic polysaccharide monooxygenases (LPMO),26,27 that cleave polysaccharides via oxygenation rather than hydrolysis. Again, the nature of the bound dioxygen species is not unambiguously established, but favours a peroxide ion rather than superoxide. Nonetheless, there is a convincing weight of evidence for the formation of a Cu(ii)–O2•− intermediate from spectroscopic studies, theoretical calculations, and from model complexes.10–12,22,28 These enzymes illustrate the controlled generation and use of ROS at copper centres. Suitably designed copper metallodrugs may generate the same ROS and some will be decomposed by enzymes such as SOD (O2•−) or catalase (H2O2). However, if significant ROS are produced cellular defences can become overwhelmed, leading to oxidative damage and cell death.2 The enzyme-­active sites suggest a starting point for DNA-­targeted metallodrug development and in our laboratory we employ phenanthroline or bipyridine ligands to supply two pyridyl donors (modelling two histidine residues) and other O or N donors to tune the properties. For example, the quercetinase structure suggests a carboxylate ligand might tune the redox potential/geometry such that a reducing agent is not required for reduction of the copper.

View Online

Recent Advances in Anticancer Copper Compounds

97

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

4.3  N  CI-­60 Screening of Anticancer Copper Complexes A ‘rite of passage’ for many developmental anticancer agents is the 60-­cancer cell line screen provided by the National Cancer Institute's (NCI) Developmental Therapeutics Program (DTP). This resource determines cancer cell sensitivity in a variety of tissues including leukaemia, NSCL, colon, central nervous system, melanoma, ovarian, renal, prostate, and breast and is typically beyond the capability of most research laboratories. NCI-­60 analysis involves an initial drug-­induced cytotoxicity measurement at 10 µM with growth inhibition measured by sulforhodamine B that binds stoichiometrically to proteins and so provides a proxy for cell mass or proliferation. Lead agents with sufficient activity at 10 µM can progress to 5-­dose testing where 50% growth inhibition (GI50), total growth inhibition (TGI), and 50% lethal dose concentrations (LC50) are calculated.29 In recent years, a number of developmental copper complexes have been screened by the NCI-­60 panel (Figure 4.2A). Inspired by the stability of cytotoxic carbene–silver complexes derived from 4,5-­substituted imidazoles,30 Tacke and co-­workers developed a lipophilic benzyl-­substituted copper(i) carbene complex WBC4. Nanomolar activity was identified for almost all NCI-­60 cell lines, including leukaemic, along with excellent sensitivity to triple negative breast cancer (MDA-­MB-­468).31 The least sensitive cancer to WBC4 was multi-­drug-­resistant ovarian (NCI/ADR-­RES) where low micromolar dosage (1.45 µM) was required to inhibit 50% growth. By examining GI50 values, CAKI-­1 renal cells were selected for xenograft models in nude mice (NMRI:nu/nu) with the complex reducing tumour volume with a treatment to control (T/C) value of 0.38 on day 32. In vivo mechanistic evidence involving CD31 immunohistochemistry then pointed to anti-­angiogenic effects where a reduction in microvessel number, area, and ratio was determined in tumours treated with the agent (10 mg kg−1). Copper(ii) 1,10-­phenanthroline (phen) complexes have generated attention as ROS-­active antitumoral candidates. Although [Cu(phen)2]2+ (Cu-­phen) is an excellent DNA oxidant and is applied in high-­fidelity DNA footprinting and for recombinant protein production,9 we determined that Cu-­phen poorly discriminates DNA from protein targets and is non-­selectively toxic to a diversity of cell lineages.32 Cu-­phen complexes with coordinated ancillary O,O′ or N,O′ ligands, however, are less promiscuous and so are attractive in anticancer drug design. This attraction stems from the role of coordinated carboxylates in copper enzymes and it was recently established that ligated carboxylates in the Cu-­phen chemotype ([Cu(phen)2(RCOO)]+(Cu-O,O′-phen); Figure 4.2B) enhance double-­stranded (ds) DNA recognition, reduce serum albumin binding, and afford control of toxicity towards human cancer cells.32 Although the cytotoxicity of bis-­phenanthroline complexes is generally higher, Cu(ii) mono-­phen complexes are also interesting developmental  agents—particularly when ancillary O,O′ ligands are present.33 This was

Published on 05 April 2019 on https://pubs.rsc.org | doi:

98 plexes examined in our laboratories. (C) Molecular structures of cytotoxic Casiopeína copper(ii) complexes. (D) Selected copper complexes with anticancer properties.

Chapter 4

Figure 4.2  (A)  Copper complexes screened by the NCI-­60 panel. (B) Selected copper phenanthrene, bipyridine, and polypyridyl com-

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

99

identified during the development of cytotoxic square planar Cu-­ph-­phen and Cu-­ph-­bipy complexes (ph = phthalate; bipy = 2,2′-­bipyridine; Figure 4.2B). In a preliminary screen, Cu-­ph-­phen was more cytotoxic than Cu-­ ph-­bipy and showed low micromolar activity towards cisplatin-­recalcitrant ovarian cancer cells (SKOV3). In that work, we suggested Cu-­ph-­bipy was attenuated due to the size and flexibility of the bipy ligand that impeded ROS catalysis and DNA intercalation.33 Cu-­ph-­phen was then submitted to the NCI-­60 panel where low micromolar activity (LC50) was observed in the majority of solid epithelial cell lines—with values ranging from 2.70 µM to 12.70 µM in nearly all NSCL, colon, CNS, melanoma, ovarian, renal, prostate, and breast cell lines—along with inactivity towards leukaemic cells (>100 µM).34 Further progress in therapeutic mono-­phen complex design was made by the Khoo group where the incorporation of amino acid (N,O′) analogues led to the progression of [Cu(dmg)(phen)(H2O)]+ (where dmg = α-­ dimethylglycine) through the NCI-­60 panel.35 In that study GI50 values were found at the low micromolar concentration range (0.42–3.03 µM) for almost all cell types; however, LC50 analysis showed that drug concentrations >100 µM were required in leukaemic cell lines and selected breast cancer cells (HS 578T and T-­47D).36 Although mononuclear complexes display encouraging activity, the introduction of a second Cu-­phen centre was recently found to heighten NCI-­60 potency between 10-­ and 100-­fold.37 In that study we selected two metallo­ drug leads (Cu-­oda and Cu-­terph; oda = octanedioate and terph = terephthalate) where in each case two Cu-­phen units were bridged by a dicarboxylate linker but the length and rigidity of the linkers differed distinctly. Interestingly, the more rigid Cu-­terph complex had the highest overall potency and demonstrated < 450 nM GI50 values in all cell lines and hypersensitivity in the melanoma line SK-­MEL-­5.37 While Cu-­oda displayed a wide range of GI50 values, its enhanced solubility made it a more promising therapeutic lead. When growth inhibition fails to identify selectivity, LC50 values (i.e. cytotoxicity) can sometimes serve as a useful alternative for NCI-­60 drug sensitivity. This can be understood when observing the spectrum of activity from resistant to sensitive cells across the panel, which abrogates broad-­scale cytotoxicity. Using this approach, Cu-­oda and Cu-­terph displayed a broader range of activities; Cu-­terph had enhanced cytotoxic effects across the majority of cell lines within the tested panel (LC50 < 1 µM) with excellent activity towards CNS, colon, and ovarian cancers and was particularly active against the melanoma line. Cu-­oda, however, displayed a broader range of activities and was most effective against melanoma, renal, and prostate cancers. The LC50 values of both di-­Cu(ii) complexes indicated inactivity towards leukaemic and breast cancer cells (HS 578T and T-­47D)—a trend also observed for the aforementioned [Cu(dmg)(phen)(H2O)]+ complex. One of the most significant features of the NCI-­60 is the COMPARE algorithm that provides an insight into potential mechanisms of drug action.4 The algorithm measures correlations in activity profiles for compounds populated in the NCI database—including those of standard agents,5 marketed drugs, mechanistic, and diversity datasets. Results from the COMPARE

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

100

Chapter 4

algorithm are ranked by Pearson's correlation coefficients (PCC or r values) and can range from −1 (negative correlation) to +1 (positive correlation), with 0 representing no correlation. We employed COMPARE to identify the mechanism of Cu-­oda and Cu-­terph complexes against established agents. Significantly, no correlation was observed with GI50 values greater  than 0.5 (a threshold used in other developmental studies37) including clinically employed metallodrugs.35,38–40 In order to guide mechanistic experiments for Cu-­oda and Cu-­terph, PCCs with lower LC50 values (r > 0.2)  identified some correlation with clinical standards and marketed agents. Interestingly, the highest number of correlations occurred for DNA-­damaging agents and topoisomerase poisons with signalling and hormonal agents showing a degree of overlap.37 To correlate plausible modes of mechanistic action by anticancer copper complexes, we conducted COMPARE analysis using the GI50 profiles of publicly accessible Cu(ii) agents and compared the data to clinical metallodrugs, including platinum drugs, bleomycin, and arsenic trioxide—the results of this analysis are reported in Figure 4.3 and discussed herein. Copper(ii) phen carboxylate agents (Cu-­terph, Cu-­oda, and Cu-­ph-­phen) display a positive correlation with one another but a stronger correlation was observed between Cu-ph-phen and Cu-oda complexes than between the two di-­nuclear complexes (Cu-­terph and Cu-­oda). The carboxylate complexes showed no significant overlap with proteasome inhibitor Cu-­pip-­nap41 (r ≤ 0.33; pip-­nap = 1-­piperdinyl-­N-­methyl-­2-­naphthalenolate) and are negatively correlated with Cu-­elesclomol and all platinum-­based agents. In agreement

Figure 4.3  COMPARE  analysis of GI50 values with positive (r > 0) and negative

correlations (r < 0) for a range of copper, platinum and other metal-­ based anticancer agents with respective NSC numbers (NCI identifier code). Matrix overlap of r = 1.0 is omitted for clarity with r = 0.5 intercept highlighted. Standard metallo-­drugs include arsenic trioxide  (NSC92859), cisplatin (NSC119875), bleomycin (NSC125066), carboplatin (NSC241240) and oxaliplatin (NSC266046). Developmental copper(ii) agents are Cu-­terph (NSC767441), Cu-­oda (NSC767442), Cu-­ph-­phen (NSC767443), Cu-­pip-­nap (NSC109268) and clinically advanced Cu-­ elesclomol (NSC766922).

View Online

Recent Advances in Anticancer Copper Compounds

101

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

42

with established mechanistic data, cisplatin and carboplatin are strongly correlated (r = 0.851) but do not overlap significantly with oxaliplatin. Further analysis with alternative metal-­based agents (arsenic trioxide and bleomycin) demonstrated negative correlation with the carboxylate complexes, however, positive PCCs were found with Cu-­pip-­nap and Cu-­elesclomol. The combined analysis in Figure 4.3 provides an important overview of the mechanistic relationship between developmental anticancer copper complexes and metallodrugs in clinical use. The data clearly show that copper compounds have unique modes of action that do not overlap with platinum(ii) drugs or with either bleomycin or arsenic trioxide. Another important observation is that although the Cu(ii)-­phen complexes show homology, this chemotype appears to be distinct from other copper complexes (Cu-­elesclomol and Cu-­pip-­nap) examined within the NCI-­60 panel.

4.4  M  echanistic Analysis of Cytotoxic Copper Complexes 4.4.1  An Overview of Cell Death Mechanisms Many cytotoxic Cu-­phen complexes discussed so far have also been examined in vitro within our laboratories against a selection of cancer cell lines, some of which lack p53 function or contain multi-­drug-­resistant genetic mutations (Table 4.1). Accordingly, some cell lines have resistance to classical platinum therapy and, although a connection between Pt(ii) uptake and active copper transport is known (see Section 4.2), COMPARE analysis with publicly available copper complexes (Figure 4.2A), together with direct comparisons with cisplatin (Table 4.1), suggests that copper metallodrugs are mechanistically unique. To help understand the reasons for this, several studies probing copper complex-­mediated cell death were conducted. One of the best-­studied cell death mechanisms is apoptosis—a cell suicide process generally used to discard superfluous cells. Apoptosis is regulated by caspases that are found in pro-­enzymatic form but become active through dimerisation or via proteolytic cleavage.43,44 Apoptotic activation arises from: (1) an intrinsic pathway originating in the mitochondria whereby BH-­3 translocation initiates apoptogenic factors that release caspase 9, or (2) an extrinsic process involving extracellular stress signals that target membrane-­bound TNFα or death receptors to activate caspase 8 (Figure 4.4).45 The induction of apoptosis by Cu-­ph-­phen, Cu-­oda, and Cu-­terph complexes was identified by annexin V quantification.34,37 Further analysis using flow cytometry then  showed the internal mitochondrial pathway (caspase 9) was activated rather than the external caspase 8. Once activated, caspase 9 leads to proteolytic and organelle degradation mediated by caspase 3/7 and is characterised by cytological alterations such as DNA fragmentation, contraction of the cytoskeleton, membrane blebbing, and the formation of apoptotic bodies46—all of which were observed with Cu-­ph-­phen, Cu-­oda, and Cu-­terph  complexes.

Published on 05 April 2019 on https://pubs.rsc.org | doi:

102

Table 4.1  IC  50 values (µM) for selected Cu(ii) complexes and antitumour agents cisplatin and mitoxantrone against a range of cell lines, including ovarian (SKOV3), prostate (DU145), breast (MCF7) and colon (HT29) cancer cell lines tested over 24 and 96 h. IC50 (µM) SKOV3

DU145

MCF7

HT29

Complex

24 h

96 h

24 h

96 h

24 h

96 h

24 h

96 h

References

Cu-­ph-­phen2 Cu-­isoph-­phen2 Cu-­ph-­phen Cu-­ph-­bipy Cu-­sal-­phen Cu-­dips-­phen Cu(3MeOsal)(phen) Cu-­terph Cisplatin Mitoxantrone

— — 6.2 ± 0.9 79.3 ± 2.9 7.1 ± 0.5 37.4 ± 2.0 9.2 ± 0.6 6.7 ± 0.4 >100 54.5 ± 2.6

— — 5.6 ± 0.3 62.3 ± 3.1 4.8 ± 0.5 3.8 ± 0.4 2.9 ± 0.6 — 37.1 ± 10.1 —

11.6 ± 4.5 10.6 ± 2.2 8.0 ± 0.0 >100 7.8 ± 1.2 28.3 ± 2.1 8.4 ± 0.9 5.7 ± 0.2 >100 27.3 ± 5.6

— — 5.4 ± 0.3 68.6 ± 2.2 6.2 ± 0.3 6.0 ± 0.2 4.5 ± 0.4 — 8.6 ± 0.7 —

44.9 ± 7.0 41.2 ± 1.4 51.8 ± 1.8 >100 49.3 ± 2.1 32.6 ± 4.3 47.5 ± 3.0 7.9 ± 0.4 90.8 ± 3.5 7.5 ±0.2

— — 5.6 ± 0.3 82.7 ± 3.4 5.3 ± 1.6 6.6 ± 1.1 5.8 ± 0.9 — 72.5 ± 3.3 —

6.0 ± 0.4 5.8 ± 0.2 8.4 ± 0.5 >100 11.2 ± 2.1 47.7 ± 5.2 12.9 ± 3.4 5.4 ± 0.3 >100 8.0 ± 0.6

— — 3.4 ± 0.4 51.8 ± 2.3 3.07 ± 0.5 6.1 ± 1.3 2.8 ± 1.1 — 72.1 ± 1.5 —

97 97 33 33 61 61 61 60, 97 33 60, 97

Chapter 4

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

103

Figure 4.4  Proposed  mechanism of action for copper(ii)-­phenanthroline antican-

cer agents. Complexes are incorporated into the cell through copper transport protein (Ctr1). Copper(ii) complexes take one of two pathways: (1) nuclear localisation and intercalative binding at the minor groove of duplex DNA or (2) mitochondrial accumulation due to MMP (ΔΨm), facilitated by the cationic complex charge and lipophilicity of phenanthroline groups. Intracellular generation of radical species such as superoxide (O2•−), singlet oxygen (1O2), hydrogen peroxide (H2O2) and hydroxyl radicals (•OH), detected through radical-­specific antioxidants, induce significant DNA degradation through the formation of DSBs which can be quantified by γH2AX and the comet assay. Enhanced levels of superoxide within the mitochondria (MitoSOX) causes collapse of the transmembrane potential measured through the extent of mitochondrial depolarisation. This results in the release of cytochrome c and activation of the intrinsic pathway through apoptosome formation and activation of initiator caspase 9 and executioner caspase 3 or 7. Apoptosis can be quantitatively measured through Annexin V and visualised by confocal microscopy.

Members of the Casiopeína family—of general formula [Cu(N,N′)(N,O)]+ and [Cu(N,N′)(O,O′)]+ (where N,N′ = bipy or phen; N,O = gly; and O,O′ = acac; Figure 4.2C)—were recently investigated, with the majority of complexes triggering apoptosis by the intrinsic mitochondrial pathway. Casiopeína II-­gly (CasII-­gly), for example, was found to target multiple mitochondrial sites47 and consequently activated caspase 9, cell cycle arrest (via BAX), along with p53 and p21 pathways.48 Interestingly, CasII-­gly also induced  caspase-­independent cell death upon low-­dose exposure to malignant glioma cells. The cell death mechanism in this case appears to be linked to ROS-­ mediated nuclear translocation of mitochondrial apoptogenic factors.49 One other member of the Casiopeína family (CasIII-­ia), however, was found to activate caspase 8 within this same cell line. Further investigation with CasIII-­ia found that apoptosis was mediated in a caspase-­independent manner along

View Online

Chapter 4

104

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

50

with the induction of autophagy —an alternative self-­digestion cell death process rendered by autophagic vacuole formation.51,52 There are relatively few developmental metallodrugs that induce autophagy; however, one interesting example that we recently identified is Mn-­oda,53 a di-­Mn(ii) complex that is structurally similar to the pro-­apoptotic Cu-­oda. Cell death mechanisms associated with apoptosis have been identified for mono-­ and di-­Cu(ii) complexes containing amino acid ancillary ligands. Here, Ramakrishnan and co-­workers reported mitotic catastrophe—a recently classified cell death mode during metaphase characterised by micro-­ or multi-­nucleation54—in H460 NSCL cancer when treated with [Cu(benzamide)(5,6-­dmp)(ClO4)]2 (benzamide = 2-­hydroxy-­N-­[2-­(methylamin o)-­ethyl]benzamide and dmp = dimethyl phen).56 Another study by this group reported that the tyrosine analogue [Cu(l-­t yr)(5,6-­dmp)(H2O)] caused necrosis upon 48 h exposure to HEp-­2 human larynx cancer cells that underwent structural changes such as the swelling and formation of uniform nuclei devoid of fragmentation.55 Interestingly, in both of these studies, complexes containing poly-­aromatic chelators such as bipy, phen, and dpq (dpq = dipyridoquinoxaline) all initiated classical cell death mechanisms associated with intrinsic apoptosis.

4.4.2  Copper-­mediated ROS Production As illustrated in Section 4.2, copper enzymes can activate dioxygen by efficiently generating Cu(ii)–O2•−, or (potentially) Cu(ii)–O•, for reaction with particular substrates. Some leakage of ROS occurs however, and enzymes such as SOD (O2•−) or catalase (H2O2) are required to protect cell structures from damage. Copper metallodrugs that generate excess ROS by Fenton (eqn (4.1)) or Haber–Weiss (eqn (4.2)) mechanisms are known to overcome these endogenous protection mechanisms to induce ROS-­mediated cell death.57 Fenton:    

H2O2 + Cu(i) → •OH + OH− + Cu(ii)



(4.1)

   

Haber–Weiss: O2•− + Cu(ii) → Cu(i) + O2

H2O2 + Cu(i) → Cu(ii) + •OH + OH−

   



Net: O2•− + H2O2 → •OH + OH− + O2

(4.2)

   

A wealth of research has focused on ROS activity of copper metallodrugs and in characterising (putative) intracellular free radicals or those formed at the chemical–biology interface.2,58 One probe routinely used to quantify cytosolic ROS levels is 2,7-­dichlorofluorescein diacetate (DCFH-­DA). In the presence of endogenously generated ROS or RNS, DCFH-­DA is oxidised to release the fluorophore 2,7-­dichlorofluorescein (DCF) and although a variety

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

105

of free radicals can liberate DCF, it is an important entry point for examining oxidative stress (Figure 4.4). Using this approach, ROS production in HT29 (colon)59 and A549 (lung)60 cancer cells was characterised upon nanomolar drug exposure to Cu-­oda and Cu-­terph complexes. We designed these experiments over a short timeframe (15–300 min) to avoid indirectly measuring oxidative stress and results were compared to a positive control of H2O2 (0.5 µM) and cisplatin, which showed no activity. In another study, DCF  measurements were taken from an SKOV3 cell line exposed to Cu-­ph-­phen and Cu-­ph-­bipy.33 A significant difference (∼four-­fold) in the time-­course of ROS production by Cu-­ph-­phen relative to Cu-­ph-­bipy was found. Interestingly, this study also identified oxidative stress induction by the free phen ligand  (but not bipy). Further examples of cytotoxic ROS-­active square planar phen complexes were generated with salicylate (sal), 3,5-­diisopropylsalicylate (dips), and 3-­methoxysalicylate (3-­MeOsal) (Figure 4.2B and Table 4.1).61 These compounds are potent SOD mimetics capable of catalytically scavenging the superoxide radical, but are inactive as catalase mimetics. The complexes effectively bind and cleave DNA but do not inhibit COX-­2 or COX-­1 activity, suggesting that interference in the arachidonic acid cascade is not significant to their mode of action. Casiopeína complexes (cf. Figure 4.2C) are important pro-­oxidants with clinical potential.62 Part of their cytotoxic mechanism is connected to the redox cycling of glutathione (GSH) which, in turn, elevates intracellular H2O2 and O2•− levels; neuroblastoma cells (SK-­N-­SH and CHP-­212) are particularly responsive with the complexes causing ROS mitochondrial damage63 and forming 8-­oxo-­dG oxidative DNA lesions.64 Since the identification of specific ROS is required to gain a deeper understanding of cytotoxic action, it is often desirable to use probes with high selectivity. To help narrow this range, cell lines of interest can be pre-­exposed to antioxidants that specifically trap (or detoxify) ROS or RNS before a known lethal dose of the metallodrug is introduced. Two recent studies with SKOV3 cells help to demonstrate the poten­tial of this approach; in the first example, a superoxide trap (tiron, 4,5-­dihy- ­  droxy-­1,3-­benzenedisulfonic acid disodium salt) was identified to attenuate Cu-­ph-­phen cytotoxicity and this radical species was subsequently found to mediate mitochondrial depolarisation.34 In the second example, superoxide (tiron) and singlet oxygen (l-­histidine) traps significantly impeded Cu-­oda and Cu-­terph toxicity. The presence of both radicals was later confirmed by flow cytometric and confocal analysis with exomarkers for mitochondrial (MitoSOX) and nuclear (dihydroethidium; DHE) superoxide along with singlet oxygen (singlet oxygen sensor green; SOSG).37 The bis-­thiosemicarbazone complex Cu-­ATSM (Figure 4.2D) has entered into phase I and II clinical trials for the treatment of cervical and non-­small cell lung cancers.65 Using structure-­activity-­functionalisation, both the lipophilicity and reduction potential were tuned by double alkylation of the ligand backbone. This change resulted in an attractive redox potential range for hypoxic tumour selectivity, characterised by low O2 levels.66 Cu-­ATSM has a specific mechanism based on reduction of Cu(ii) to Cu(i) in the cytosol or microsome where, in low oxygen environments, the unstable species

View Online

106

Chapter 4

becomes intracellularly trapped. In the presence of O2, the complex is oxidised to Cu(ii) and can diffuse back out of the cell.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

4.4.3  Mitochondrial Toxicity Unambiguous targeting of the mitochondria has generated attention for the development of novel chemotherapeutic agents and witnessed progression of ROS-­modulating agents into clinical trials.67,68 Elevated ROS levels are known to alter the chemical environment of the mitochondria, most notably causing depletion of mitochondrial membrane potential (ΔΨm)—the H+ electrochemical gradient that drives ATP synthesis (Figure 4.4; mitochondrial pathway). The delivery of mito-­targeted anticancer agents can be achieved with penetrating peptide sequences, liposomes, nanocarriers, lipohilic cations including triphenylphosphonium (TPP), or small molecules that traverse pH gradients. A Cu(ii) complex functionalised with TPP [Cu(ttpy-­tpp)(Br)2]+ (where ttpy-­tpp = 4′-­p-­tolyl-­(2,2′:6′,2″-­terpyridyl)triphenylphosphonium bromide) was found to dissipate the mitochondrial transmembrane potential by a factor of 20-­fold over a non-­targeted terpyridine complex.69 Significantly, metal-­free 1,10-­phen can initiate mitophagy, the selective degradation of the mitochondria leading to fragmentation, extensive depolarisation, and loss of mitochondrial mass.70 In combination with the redox capabilities of Cu-­phen systems—and their lipophilic cationic structures—the targeted accumulation of this chemotype within the mitochondria is an attractive therapeutic  strategy. Both Cu-­oda and Cu-­terph demonstrated selective dissipation of  ΔΨm in SKOV3 cells and displayed 25-­fold higher activity than the known proton ionophore carbonyl cyanide m-­chlorophenyl hydrazine. In that study, we reasoned a net complex charge (2+) and the combined lipophilicity of four phen ligands was significant for the heightened accumulation. Further examination identified elevated levels of mitochondrial superoxide production with MitoSOX that were considerably higher when compared to nuclear superoxide generation by DHE.37 CasII-­gly, CasIII-­ia, and CasIII-­ea complexes also trigger mitochondrial superoxide production and induce depolarisation in neuroblastoma cells.63 Additionally, the amino acid derivative (CasII-­gly) was found to decrease mitochondrial DNA (mtDNA) copy number and ATP synthesis, which imbalanced mitochondrially encoded respiratory proteins in isolated mitochondria.47,71 Interestingly, Suntharalingam and co-­workers recently coupled 1,10-­phenanthroline to a mitochondria-­penetrating peptide (MPP) comprising alternating d-­arginine and l-­cyclohexylalanine amino acids to yield the metallopeptide [Cu(phen-­MPP)Cl2] (Figure 4.2D).72 This complex showed selective cytotoxicity in breast cancer stem cells (CSC; IC50 = 8.1 ± 0.6 µM) over non-­CSCs (IC50 = 17.3 ± 0.7 µM) and 3D cell culture studies demonstrated mammosphere reduction to a greater extent than the CSC control salinomycin. Cytotoxicity, in this case, appears to be dependent on mitochondrial accumulation followed by Cu-­mediated ROS generation that precipitates mitochondrial dysfunction to activate JNK and p38 apoptotic pathways. The stability of the Cu(ii)-­phen-­MPP complex was identified extracellularly

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

107

using mass spectrometry but, since Cu(i)-­phen has protease properties,73 an important aspect for clinical application would be to measure stability in a reducing environment (e.g. with either ascorbate or suitable thiols) to match the hypoxic tumour environment. Copper(ii) complexes with mono-­ phenanthroline ligands have, in their own right, mitochondrial-­targeting properties and earlier work has established that depletion of ΔΨm is linked to superoxide production.34 Prior to this, seminal work by McCann and Kavanagh using fungal and bacterial cell lines established that metal-­free phen (and derivatives), along with several metal complexes, damages mitochondrial function to uncouple respiration (reviewed in ref. 74). A Cu(ii)-­binding ligand, elesclomol (Figure 4.2A), was entered into phase I and II clinical trials for myeloid leukaemia, melanoma, and various solid tumours, including NSCL cancer, prostate, ovarian, and fallopian tube cancer for lone and/or combination therapy.65,67 Phase III studies were discontinued, however, as certain patients receiving the agent, in combination with paclitaxel, were identified to have a higher mortality rate.75 The cytotoxic mechanism of elesclomol was initially connected with ROS evolution but more recently it was identified as a potent copper chelating agent.76 Further work demonstrated that elesclomol-­induced ROS generation is dependent on the chelation and redox cycling of copper in the mitochondria, whereby the ligand chelates extracellular copper and shuttles Cu(ii)–elesclomol into the cell. The complex rapidly transports copper to mitochondria and, once dissociated, the free ligand is rapidly effluxed from cells where it continues its shuttling cycle. Further mechanistic work employed a comprehensive screen of Saccharomyces cerevisiae yeast cells that identified target genes in the electron transport chain (ETC).77 As mtDNA encodes for 13 protein subunits of complexes regulating the ETC, melanoma HBL-­ρ0 cells depleted of mtDNA were used to identify whether complex-­mediated ROS production was dependent on ETC.

4.4.4  DNA-­targeted Copper Complexes The process of molecular recognition of a drug molecule by DNA often involves conformational changes to DNA structure. Ultimately, there are two classes of binding to DNA, covalent and non-­covalent interactions. Covalent binding results in non-­reversible changes to DNA through adduct formation; for example, cisplatin predominantly crosslinks DNA strands by binding two neighbouring guanine bases, inducing unwinding and kinking of duplex DNA.78,79 In contrast, non-­covalent drug interactions with DNA are more reversible, resulting in temporary structural changes. Understanding small molecule interactions with DNA is of major importance in the rational design of more powerful and sequence-­specific anticancer agents. Copper complexes have been found to interact non-­covalently with DNA rather than forming covalent adducts,80 therefore this discussion is focused on non-­covalent interactions. Non-­covalent interactions between DNA and small molecules are dependent on their structural features; the coordination geometry of the copper complex, along with ligand planarity and size,

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

108

Chapter 4

control DNA-­binding affinity and specificity. These interactions have been identified through UV and fluorescence-­based biochemical and biophysical, photophysical, crystallographic, and NMR techniques. Classical modes for  reversible binding of molecules to duplex DNA include: (1) groove binding (major and minor), (2) intercalation (including partial intercalation), and (3) insertion.81 Intercalation requires a planar aromatic ligand to be present in the complex system and can be further divided into semi-­intercalation, symmetrical intercalation, and canted intercalation.82,83 It is also worth noting that non-­covalent drug-­DNA interactions can involve a combination of binding modes. One such example is bleomycin (BLM) which contains a positively charged bithiazole unit responsible for mixed binding modes, including partial intercalation and binding within the minor groove.84

4.4.4.1 Mononuclear Agents The discovery of the first synthetic chemical nuclease, Cu-­phen, sparked intensive efforts towards the development of new artificial metallonucleases that mediate DNA cleavage through the production of ROS.6 The Cu-­phen complex predominantly binds to DNA within the minor groove and abstracts a hydrogen atom from the C1′ deoxyribose position, resulting in strand scission.85 In addition, this chemotype has shown promising antitumoral, antifungal, and antimicrobial properties.32,86 Cu-­phen binds to both DNA and protein biomolecules without specificity, and can be considered a “promiscuous agent”. Accordingly, modulation of the Cu-­phen structure to find more suitable drug candidates represents an interesting challenge. A number of structural modifications to the chemotype have been explored, including inner-­sphere modification with oxygen donor ligands such as carboxylates,32 systematic extension of the ligated phenazine ligand,86,87 incorporation of either N,N′ and O,O′ or N,N′ and O,N′ ligands,2,61,62,88,89 or the introduction of a second Cu(ii) centre.37,59,60,90,91 We developed a series of cytotoxic bis-­chelate Cu(ii)-­phenanthroline-­ phenazine complexes based on the Cu-­phen chemotype in which one phen ligand was replaced by a phenazine diimine ligand (Cu-­N,N′-­phen, where N,N′ = dpq, dppz, or dppn; Figure 4.2B). These each contain a larger aromatic heterocyclic ligand for stacking or partial intercalation between base pairs (bp).86 These complexes were found to show significantly enhanced DNA-­binding properties when compared directly to Cu-­phen, with apparent DNA-­binding constants of the series in the range of ∼107 M(bp)−1, among the highest DNA-­ binding constants reported in the literature for Cu(ii) complexes (Figure 4.5A). These data highlight the influence of an extended phenazine π-­framework and also a prominent role for the ancillary chelated phenanthroline in nucleotide-­binding affinity. The influence of systematically extending the ligated phenazine ligand was also identified by hydrodynamic changes in solution. The complexes were found to increase the polymer length of salmon testes dsDNA in proportion to the amount of drug bound, with the trend following Cu-­dpq-­phen > Cu-­dppz-­phen > Cu-­phen.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

109

Figure 4.5  (A)  DNA-­binding properties of selected mono and dinuclear Cu(ii) com-

plexes to EtBr saturated calf thymus DNA solution (where Kapp = Ke × 12.6/C50, where Ke = 9.5 × 106 M(bp)−1). (B) Mean intensity fluorescence of γH2AX fluorescence in SKOV3 cells within selected copper systems: (i) dinuclear Cu-­oda and Cu-­terph (24 h treatment at approximate IC50 concentration of 1.0 µM);37 (ii) Cu-­phen, Cu-­for-­phen, Cu-­ace-­phen, Cu-­pro-­phen, Cu-­iso-­phen and Cu-­piv-­phen (24 h exposure at IC25 of 2.9, 2.9, 3.0, 3.0, 2.9, 3.6 µM, respectively);61 and (iii) Cu-­ph-­phen (5.6 µM) and Cu-­bipy-­phen (62.3 µM) at corresponding IC50 over 96 h.33 (C) Confocal imaging of γH2AX foci (green) within nuclei (red) of SKOV3 cells after 96 h treatment with cisplatin (control), Cu-­ph-­phen and Cu-­ph-­bipy.

Since spectroscopic analysis of these Cu(ii) systems is limited to indirect ethidium bromide displacement assays (loss of fluorescence due to ejection of bound ethidium bromide upon displacement by the complex), further displacement studies were performed on synthetic co-­polymers poly[d(G-­C)2] and poly[d(A-­T)2] due to their unique structural features mimicking the major and minor groove, respectively. The Cu-­N,N′-­phen complexes displayed effective quenching on both polymers, with only a slight preference shown for the G-­C rich polymer. To further probe the nucleotide-­binding specificity of the complex series, thermal melting studies were carried out on both synthetic long-­chain co-­polymers. The complexes were found to have little or no interaction with the disfavoured poly[d(A-­T)2] polymer while significant stabilisation—in line with that of control intercalating agent actinomycin D (ΔTM + 12 °C)—was obtained on the poly[d(G-­C)2] polymer, with the trend following Cu-­dpq-­phen > Cu-­dppz-­phen > Cu-­phen.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

110

Chapter 4

Two square planar copper(ii) mononuclear complexes, Cu-­ph-­phen and Cu-­bipy-­phen, extensively discussed in earlier sections also have binding capability to circulating tumour (ct) DNA (Figure 4.5A) and can displace bound EtBr from long nucleotide polymers in the order poly[d(A-­T)2] > ctDNA > poly[d(G-­C)2], with mechanistic studies suggesting partial intercalation of phen in the minor groove.33,34 Examples of complexes with enhanced DNA binding relative to Cu-­phen (Kapp = 6.67 × 105 M−1 (bp)) include the cationic series [Cu(RCOO)(phen)2]+ (R = –H, –CH3, –C2H5, –CH(CH3)2, and –C(CH3)3),61 where Kapp values were significantly enhanced (by at least ten-­fold) relative to Cu-­phen and increase with the steric bulk of the ancillary ligand (cf. [Cu(pivalate)(phen)2]2+ (Cu-­piv-­phen)). Further evidence for intercalation was identified by viscosity analysis. Recent advances in molecular dynamics simulation methods applied to DNA have also extended our knowledge in the field of drug-­DNA interactions. Galindo-­Murillo et al. manually orientated a series of 21 Casiopeínas into the minor groove of the Dickerson Drew dodecamer d(CGCGAATTCGCG)2 (next to the central A6-­T7 step), since previous work suggested the minor groove was the preferred binding site.92 A total of 35 simulations were carried out in this study revealing five individual modes of complex binding interactions: (1) stacking on the terminal base pairs of the DNA chain (66%), (2) minor groove binding (9%), (3) intercalation with base-­pair displacement/opening (11%), (4) minor groove binding with stacking of one of the terminal frayed bases (6%), and (5) intercalation near the end of the DNA chain (9%).93 Interestingly, the most prevalent binding mode involving stacking with terminal base pairs was not as energetically favoured as minor groove binding. This study highlights a new set of possible interaction modes which complexes can undergo when binding to short sequences of DNA.

4.4.4.2 Dinuclear Agents Cu-­oda and Cu-­terph are structurally similar dicopper(ii) complexes containing two Cu-­phen units bridged by dicarboxylate linkers. In the solid state, the two metal centres in the Cu-oda complex are 9.7 Å apart, while the presence of the rigid terph linker in the Cu-­terph scaffold results in a Cu–Cu intermetal distance of 11.0 Å in the solid state. The topoisomerase I-­mediated DNA relaxation assay was employed to investigate whether the dicarboxylate linker influences intercalation on a supercoiled (SC) plasmid substrate. Interestingly, both di-­nuclear systems efficiently intercalated and unwound the negatively SC plasmid to a relaxed open circular conformation prior to the introduction of positive supercoiling and nicking (SSBs). Cu-­oda is not only selective against solid epithelial cancer cells from the NCI-­60 human cell line panel, but was also found to discriminate between TA/TA and AT/AT steps within 12-­mer palindromic sequences differing only in their central step.37 Circular dichroism spectroscopy revealed site-­ specific intercalation at the central TA/TA site in one oligonucleotide, while a strong minimum in this same spectral region indicated localised Z-­like DNA formation in the oligonucleotide with a central AT/AT step.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

111

Intercalation was also observed in two other sequences containing a central TATA step which only differed by the flanking bases. When thymine is substituted with uracil, the bulky methyl group is removed from the major groove, which becomes more accessible. Significantly, the elliptical signal at 288 nm increased on binding the dinuclear complex to the uracil-­ substituted sequence, suggesting some degree of intercalation occurs at the major groove. Thermal melting analysis on the 12-­mer sequence containing the central TA/TA step resulted in significant thermal stabilisation of +8.77 °C, and enhanced duplex stability of 4.34 kJ mol−1, further highlighting the intercalative ability of the complex. Similar sequence bias was displayed by the Λ-­enantiomer of the ruthenium(ii) “light switch” complex, [Ru(phen)2(dppz)]2+, where a crystallographic study by Cardin and co-­workers showed the dppz ligand intercalated symmetrically and perpendicularly from the minor groove of a 10-­mer palindromic oligonucleotide with a central TA/TA step, but not in the second sequence which differed only in the central step switched to AT/ AT.94 Not only is it possible for the dppz ligand to interact with the d(CCGGTACCGG)2 by perpendicular intercalation at T5A6/T5A6, but canted or angled intercalation can also occur at C1C2/G9G10 and semi-­intercalation at G3G4/ C6C7. In both d(CCGGTACCGG)2 and d(CCGGATCCGG)2 sequences, angled intercalation by the dppz ligand is accompanied by semi-­intercalation of one phen ligand.

4.4.5  Oxidative DNA Damage A number of groups have worked towards designing copper complexes that damage DNA and promote cellular toxicity by catalytic ROS generation. A clinically relevant example is BLM which requires the presence of a reduced transition metal (e.g. Fe(ii), Cu(i)), molecular oxygen, and a one-­electron reductant in order to generate the active metallo-­bleomycin species which can catalyse single-­stranded (ss) and double-­stranded (ds) DNA damage, with the latter thought to be the source of cytotoxicity. The cleavage chemistry of BLM is distinct from that of Cu-­phen as the major site of oxidative attack involves C-­4′ hydrogen atom abstraction rather than C-­1′.95 Casiopeínas also exert DNA damage by: (1) binding to DNA through intercalative or non-­ intercalative mechanisms, (2) generating ROS resulting in DNA oxidation and degradation, as well as (3) inducing mitochondrial toxicity.96 In our laboratories we have employed agarose gel electrophoresis, microfluidic ‘on-­chip’ analysis, and the formation of 8-­oxo-­dG lesions to identify ROS formation (including superoxide (O2•−), hydrogen peroxide (H2O2), hydroxyl radical (•OH), metal-­oxo, and metal-­hydroxo species) as critical components in the oxidative DNA-­damaging profile of a number of copper(ii) complexes.32,60,86,87 Bis-­phen Cu(ii) phthalate complexes [Cu(ph)(phen)2], [Cu(isophthalate)(phen)2], and Cu-­terph were the first ‘self-­activating’ chemical nucleases of their class reported.97 They showed concentration-­dependent relaxation of supercoiled (SC) plasmid DNA to open circular and linear (L) conformations in the absence of exogenous reductant. These complexes, in

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

112

Chapter 4

particular Cu-­terph, represent a significant development in the field of artificial chemical nucleases, as DNA cleavage was not limited to the presence of an exogenous reducing environment.97 Further analysis then established the water-­soluble di-­nuclear Cu-­oda complex as a ‘self-­activating’ complex also capable of mediating oxidative DNA cleavage in the absence of exogenous reductant.59 Recently, the in vitro and intracellular DNA oxidation profile of Cu-­ph-­phen was investigated in the presence of a number of ROS-­specific scavengers.34 Tiron was determined to significantly impede the cleavage of SC plasmid DNA since dsDNA damage was prevented. In SKOV3 cells, pre-­treatment with mannitol (•OH scavenger), pyruvate (H2O2), or tiron (O2•−) prior to treatment with a toxic dose of Cu-­ph-­phen (3 µM) resulted in enhanced cell survival by 12, 16, and 23%, respectively. In vitro and intracellular results were in excellent agreement and identified the generation of superoxide as the principal species mediating the copper-­catalysed Haber–Weiss reaction via Fenton-­ type chemistry. In a previous study, immunodetection of γH2AX foci (formed in response to DSBs, see Figure 4.4) by confocal microscopy and flow cytometry was carried out on SKOV3 cells exposed to an IC50 dose of Cu-­ph-­phen over 96 hours and compared directly to the Cu-­ph-­bipy derivative and cisplatin.33 By quantifying the mean intensity fluorescence (MIF) of γH2AX, it was possible to identify a large number of DSBs in comparison to the bipyridyl derivative or cisplatin, where DSBs observed in the latter were most likely due to DNA repair mechanisms and excision of platinated adducts. Statistical analysis of the MIF values, in order to quantify the average number of DSBs, gave results following the trend Cu-­ph-­phen > cisplatin ≥ Cu-­ph-­bipy. As noted above, extending the π-­system from bipy to phen results in significant (partial) intercalative effects. The Cu-­ph-­phen complex also produced  superior numbers of DSBs in SKOV3 cells, potent cytotoxicity across a number of cell lines including MCF-­7, DU145, SKOV3, and HT29 lines (Table 4.1), greater DNA cleavage on SC plasmid DNA, and stronger binding to ctDNA than the bipy derivative, further highlighting the importance of the rational design of anticancer agents (Figure 4.5B and C). A similar study was undertaken with bis-­phen Cu(ii) complexes incorporating sterically modified carboxylate ligands (e.g. Cu-­piv-­phen, Figure 4.2B). 32 Here, the overall production of DSBs was enhanced at lower concentrations when compared to Cu-­ph-­phen and Cu-­ph-­bipy but did not match either of the di-­Cu(ii) complexes Cu-­oda and Cu-­terph (Figure 4.5B). The ability of the dinuclear Cu-­terph and Cu-­oda complexes to cause DSB formation in SKOV3 cells was also quantified by immunodetection of γH2AX, and this method was used to probe the cellular processes previously identified by the COMPARE algorithm.37 Both complexes were found to induce concentration-­dependent DSB formation with maximal MIF detection comparable to doxorubicin exposure (Figure 4.5B). This result highlights the influence of nuclearity in the complex scaffold since both di-­nuclear species produced an order of magnitude higher MIF of γH2AX than Cu-­ph-­phen under identical experimental conditions. ROS-­selective antioxidants were employed to identify the species responsible for intracellular DNA damage.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

113

The presence of tiron increased cellular viability by ∼30% for Cu-­oda and ∼50% for Cu-­terph, while the overall trend in antioxidant-­promoted SKOV3 survival followed the trend: tiron >> histidine > lipoic acid > pyruvate > methionine > mannitol. Interestingly, this study identified higher intracellular O2•− and 1O2 production by the more structurally rigid complex Cu-­terph.

4.5  Summary and Outlook There have been significant advances in developing copper compounds that show selective chemotherapeutic potential towards human cancers. This progress has extended to bioactive copper 1,10-­phenanthroline complexes and several of these mono-­ and dinuclear agents, incorporating either carboxylate, glycinate, or acetylacetonate ligands, have shown considerable clinical potential. Some of these agents were identified to have unique activity in the National Cancer Institute's 60-­cancer cell line screen (NCI-­ 60) and have alternative mechanisms to clinically established agents such as cisplatin and bleomycin. Although DNA is an important cellular target for these developmental copper metallodrugs,98 it appears that cell death is  triggered by metal-­catalysed pro-­apoptotic ROS and RNS that damage cytoplasmic, mitochondrial, and genome function. Work in our group is now focused on the application of polypyridyl copper(ii) complexes  (cf. Cu-­tpma-­dppz, Figure 4.2B) that display excellent solution stabilities and form unique DNA oxidative lesions;99 dual acting copper(ii) phenanthrene complexes containing the clinically established suberoylanilide hydroxamate (saha) histone deacetylase inhibitor (cf. Cu-­saha-­dppz, Figure 4.2B);100 and the newly identified phosphate binding di-­nuclear copper(ii) complex Cu2TPNap (Figure 4.2B) that combines major groove binding and oxidative DNA cleavage.

Abbreviations acac acetylacetonato ace acetate ATSM diacetyl-­bis(N4-­methylthiosemicarbazone) benzamide  2-­hydroxy-­N-­[2-­(methylamino)ethyl]benzamide bipy 2,2′-­bipyridine BLM bleomycin DSBs double strand breaks DHE dihydroethidium dips 3,5-­diisopropylsalicylate dmg α-­dimethylglycine dppn benzo[i]dipyrido[3,2-­a:2′,3′-­h]phenazine dppz dipyrido[3,2-­a:2′,3′-­c]phenazine dpq dipyrido[3,2-­f:2′3′-­h]quinoxaline ds double-­stranded DTP Developmental Therapeutics Program

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

114

Chapter 4

EtBr ethidium bromide ETC electron transport chain GI50 50% growth inhibition gly glycinato for formate Fx l-­cyclohexylalanine H2O2 hydrogen peroxide IC50 50% inhibition concentration half maximal inhibitory concentration iso isobutyrate LC50 50% lethal concentration MIF mean intensity fluorescence MPP mitochondrial penetrating peptide composed of rFxrFxrFx NCI National Cancer Institute oda octanedioate ph phthalate pip-­nap 1-­piperdinyl-­N-­methyl-­2-­naphthalenolate PHM peptidylglycine α-­hydroxylating monooxygenase piv pivalate phen 1,10-­phenanthroline pro propionate ROS reactive oxygen species saha suberoylanilide hydroxamate sal salicylate SC supercoiled SOD superoxide dismutase SOSG singlet oxygen sensor green ss single-­stranded SSBs single strand breaks terph terephthalate TGI total growth inhibition tpma tris-­(2-­pyridylmethyl)amine TPNap tetra-­(2-­pyridyl)-­NMe-­naphthalene tpp triphenylphosphonium ttpy-­tpp 4′-­p-­tolyl-­(2,2′:6′,2″-­terpyridyl)triphenylphosphonium bromide TPQ 2,4,5-­trihydroxylphenylalanine quinone l-­t yr l-­t yrosine 3MeOSal  3-­methoxysalicylate 5,6-­dmp 5,6-­dimethyl-­1,10-­phenanthroline ΔΨm mitochondrial membrane potential γH2AX phosphorylated H2AX

Acknowledgements Funding from Science Foundation Ireland Career Development Award (SFI-­ CDA) [15/CDA/3648] is gratefully acknowledged.

View Online

Recent Advances in Anticancer Copper Compounds

115

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

References 1. F. Tisato, C. Marzano, M. Porchia, M. Pellei and C. Santini, Med. Res. Rev., 2010, 30, 708–749. 2. C. Santini, M. Pellei, V. Gandin, M. Porchia, F. Tisato and C. Marzano, Chem. Rev., 2014, 114, 815–862. 3. M. Wehbe, A. W. Y. Leung, M. J. Abrams, C. Orvig and M. B. Bally, Dalton Trans., 2017, 46, 10758–10773. 4. Databases & Tools | DTP, https://dtp.cancer.gov/databases_tools/compare.htm, accessed 23 July 2018. 5. S. L. Holbeck, J. M. Collins and J. H. Doroshow, Mol. Cancer Ther., 2010, 9, 1451–1460. 6. D. S. Sigman, D. R. Graham, V. D'Aurora and A. M. Stern, J. Biol. Chem., 1979, 254, 12269–12272. 7. D. S. Sigman, A. Mazumder and D. M. Perrin, Chem. Rev., 1993, 93, 2295–2316. 8. S. Basak and V. Nagaraja, Nucleic Acids Res., 2001, 29, e105. 9. R. Larragy, J. Fitzgerald, A. Prisecaru, V. McKee, P. Leonard and A. Kellett, Chem. Commun., 2015, 51, 12908–12911. 10. E. I. Solomon, D. E. Heppner, E. M. Johnston, J. W. Ginsbach, J. Cirera, M. Qayyum, M. T. Kieber-­Emmons, C. H. Kjaergaard, R. G. Hadt and L. Tian, Chem. Rev., 2014, 114, 3659–3853. 11. D. A. Quist, D. E. Diaz, J. J. Liu and K. D. Karlin, J. Biol. Inorg. Chem., 2017, 22, 253–288. 12. C. E. Elwell, N. L. Gagnon, B. D. Neisen, D. Dhar, A. D. Spaeth, G. M. Yee and W. B. Tolman, Chem. Rev., 2017, 117, 2059–2107. 13. A. K. Boal and A. C. Rosenzweig, Chem. Rev., 2009, 109, 4760–4779. 14. S. B. Howell, R. Safaei, C. A. Larson and M. J. Sailor, Mol. Pharmacol., 2010, 77, 887–894. 15. F. Arnesano, M. I. Nardella and G. Natile, Coord. Chem. Rev., 2018, 374, 254–260. 16. B. D. Belviso, A. Galliani, A. Lasorsa, V. Mirabelli, R. Caliandro, F. Arnesano and G. Natile, Inorg. Chem., 2016, 55, 6563–6573. 17. I. Shabalin, Z. Dauter, M. Jaskolski, W. Minor and A. Wlodawer, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2015, 71, 1965–1979. 18. A. K. Boal and A. C. Rosenzweig, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2015, 71, 1984–1986. 19. A. K. Boal and A. C. Rosenzweig, J. Am. Chem. Soc., 2009, 131, 14196–14197. 20. N. V. Dolgova, C. Yu, J. P. Cvitkovic, M. Hodak, K. H. Nienaber, K. L. Summers, J. J. H. Cotelesage, J. Bernholc, G. A. Kaminski, I. J. Pickering, G. N. George and O. Y. Dmitriev, Biochemistry, 2017, 56, 3129–3141. 21. H. C. Freeman and J. M. Guss, in Handbook of Metalloproteins, Wiley, 2001, vol. 2, pp. 1153–1169. 22. E. M. Shepard and D. M. Dooley, Acc. Chem. Res., 2015, 48, 1218–1226. 23. J. P. Klinman, J. Biol. Chem., 2006, 281, 3013–3016.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

116

Chapter 4

24. Y. Sheng, I. A. Abreu, D. E. Cabelli, M. J. Maroney, A.-­F. Miller, M. Teixeira and J. S. Valentine, Chem. Rev., 2014, 114, 3854–3918. 25. S. T. Prigge, B. A. Eipper, R. E. Mains and L. M. Amzel, Science, 2004, 304, 864–867. 26. W. B. O'Dell, P. K. Agarwal and F. Meilleur, Angew. Chem., Int. Ed., 2017, 56, 767–770. 27. J.-­P. Bacik, S. Mekasha, Z. Forsberg, A. Y. Kovalevsky, G. Vaaje-­Kolstad, V. G. H. Eijsink, J. C. Nix, L. Coates, M. J. Cuneo, C. J. Unkefer and J. C.-­H. Chen, Biochemistry, 2017, 56, 2529–2532. 28. M. Bhadra, J. Y. C. Lee, R. E. Cowley, S. Kim, M. A. Siegler, E. I. Solomon and K. D. Karlin, J. Am. Chem. Soc., 2018, 140, 9042–9045. 29. R. H. Shoemaker, Nat. Rev. Cancer, 2006, 6, 813–823. 30. D. A. Medvetz, K. M. Hindi, M. J. Panzner, A. J. Ditto, Y. H. Yun and W. J. Youngs, Met.-­Based Drugs, 2008, 2008, 1–7, Article ID 384010. 31. W. Walther, I. Fichtner, F. Hackenberg, W. Streciwilk and M. Tacke, Lett. Drug Des. Discovery, 2014, 11, 825–832. 32. A. Prisecaru, V. McKee, O. Howe, G. Rochford, M. McCann, J. Colleran, M. Pour, N. Barron, N. Gathergood and A. Kellett, J. Med. Chem., 2013, 56, 8599–8615. 33. A. Kellett, O. Howe, M. O'Connor, M. McCann, B. S. Creaven, S. McClean, A. Foltyn-­Arfa Kia, A. Casey and M. Devereux, Free Radic. Biol. Med., 2012, 53, 564–576. 34. C. Slator, N. Barron, O. Howe and A. Kellett, ACS Chem. Biol., 2016, 11, 159–171. 35. C. H. Ng, S. M. Kong, Y. L. Tiong, M. J. Maah, N. Sukram, M. Ahmad and A. S. B. Khoo, Metallomics, 2014, 6, 892–906. 36. M. Ahmad, S.-­N. Suhaimi, T.-­L. Chu, N. A. Aziz, N.-­K. M. Kornain, D. S. Samiulla, K.-­W. Lo, C.-­H. Ng and A. S.-­B. Khoo, PLoS One, 2018, 13, e0191295. 37. C. Slator, Z. Molphy, V. McKee, C. Long, T. Brown and A. Kellett, Nucleic Acids Res., 2018, 46, 2733–2750. 38. J. M. Hearn, I. Romero-­Canelón, B. Qamar, Z. Liu, I. Hands-­Portman and P. J. Sadler, ACS Chem. Biol., 2013, 8, 1335–1343. 39. Y. Fu, C. Sanchez-­Cano, R. Soni, I. Romero-­Canelon, J. M. Hearn, Z. Liu, M. Wills and P. J. Sadler, Dalton Trans., 2016, 45, 8367–8378. 40. J. J. Gills, S. Holbeck, M. Hollingshead, S. M. Hewitt, A. P. Kozikowski and P. A. Dennis, Mol. Cancer Ther., 2006, 5, 713–722. 41. K. G. Daniel, P. Gupta, R. H. Harbach, W. C. Guida and Q. P. Dou, Biochem. Pharmacol., 2004, 67, 1139–1151. 42. L. Kelland, Nat. Rev. Cancer, 2007, 7, 573–584. 43. M. O. Hengartner, Nature, 2000, 407, 770–776. 44. A. Degterev, M. Boyd and J. Yuan, Oncogene, 2003, 22, 8543–8567. 45. S. Fulda and K.-­M. Debatin, Oncogene, 2006, 25, 4798–4811. 46. J. F. Kerr, A. H. Wyllie and A. R. Currie, Br. J. Cancer, 1972, 26, 239–257. 47. A. Marín-­Hernández, I. Gracia-­Mora, L. Ruiz-­Ramírez and R. Moreno-­ Sánchez, Biochem. Pharmacol., 2003, 65, 1979–1989.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

117

48. A. I. Valencia-­Cruz, L. I. Uribe-­Figueroa, R. Galindo-­Murillo, K. Baca-­ López, A. G. Gutiérrez, A. Vázquez-­Aguirre, L. Ruiz-­Azuara, E. Hernández-­ Lemus and C. Mejía, PLoS One, 2013, 8, e54664. 49. C. Trejo-­Solís, G. Palencia, S. Zuñiga, A. Rodríguez-­Ropon, L. Osorio-­ Rico, S. Torres Luvia, I. Gracia-­Mora, L. Marquez-­Rosado, A. Sánchez, M. E. Moreno-­García, A. Cruz, M. E. Bravo-­Gómez, L. Ruiz-­Ramírez, S. Rodríquez-­Enriquez and J. Sotelo, Neoplasia, 2005, 7, 563–574. 50. C. Trejo-­Solís, D. Jimenez-­Farfan, S. Rodríguez-­Enríquez, F. Fernandez-­ Valverde, A. Cruz-­Salgado, L. Ruiz-­Azuara and J. Sotelo, BMC Cancer, 2012, 12, 156. 51. B. Levine and G. Kroemer, Cell, 2008, 132, 27–42. 52. T. Shintani and D. J. Klionsky, Science, 2004, 306, 990–995. 53. C. Slator, Z. Molphy, V. McKee and A. Kellett, Redox Biol., 2017, 12, 150–161. 54. G. Kroemer, L. Galluzzi, P. Vandenabeele, J. Abrams, E. S. Alnemri, E. H. Baehrecke, M. V. Blagosklonny, W. S. El-­Deiry, P. Golstein, D. R. Green, M. Hengartner, R. A. Knight, S. Kumar, S. A. Lipton, W. Malorni, G. Nuñez, M. E. Peter, J. Tschopp, J. Yuan, M. Piacentini, B. Zhivotovsky and G. Melino, Cell Death Differ., 2009, 16, 3–11. 55. S. Ramakrishnan, V. Rajendiran, M. Palaniandavar, V. S. Periasamy, B. S. Srinag, H. Krishnamurthy and M. A. Akbarsha, Inorg. Chem., 2009, 48, 1309–1322. 56. S. Ramakrishnan, D. Shakthipriya, E. Suresh, V. S. Periasamy, M. A. Akbarsha and M. Palaniandavar, Inorg. Chem., 2011, 50, 6458–6471. 57. B. Halliwell and J. Gutteridge, Free Radicals in Biology and Medicine, Oxford University Press, 5th edn, 2015. 58. A. T. Dharmaraja, J. Med. Chem., 2017, 60, 3221–3240. 59. A. Kellett, M. O'Connor, M. McCann, O. Howe, A. Casey, P. McCarron, K. Kavanagh, M. McNamara, S. Kennedy, D. D. May, P. S. Skell, D. O'Shea and M. Devereux, MedChemComm, 2011, 2, 579–584. 60. A. Prisecaru, M. Devereux, N. Barron, M. McCann, J. Colleran, A. Casey, V. McKee and A. Kellett, Chem. Commun., 2012, 48, 6906–6908. 61. M. O'Connor, A. Kellett, M. McCann, G. Rosair, M. McNamara, O. Howe, B. S. Creaven, S. McClean, A. F.-­A. Kia, D. O'Shea and M. Devereux, J. Med. Chem., 2012, 55, 1957–1968. 62. L. Ruiz-­Azuara and M. E. Bravo-­Gómez, Curr. Med. Chem., 2010, 17, 3606–3615. 63. J. C. García-­Ramos, A. G. Gutiérrez, A. Vázquez-­Aguirre, Y. Toledano-­ Magaña, A. L. Alonso-­Sáenz, V. Gómez-­Vidales, M. Flores-­Alamo, C. Mejía and L. Ruiz-­Azuara, BioMetals, 2017, 30, 43–58. 64. A. Rivero-­Müller, A. De Vizcaya-­Ruiz, N. Plant, L. Ruiz and M. Dobrota, Chem.-­Biol. Interact., 2007, 165, 189–199. 65. Home -­ ClinicalTrials.gov, https://clinicaltrials.gov/, accessed 25 August 2017. 66. J. L. J. Dearling, J. S. Lewis, D. W. McCarthy, M. J. Welch and P. J. Blower, Chem. Commun., 1998, 2531–2532.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

118

Chapter 4

67. D. Trachootham, J. Alexandre and P. Huang, Nat. Rev. Drug Discovery, 2009, 8, 579–591. 68. S. Fulda, L. Galluzzi and G. Kroemer, Nat. Rev. Drug Discovery, 2010, 9, 447–464. 69. W. Zhou, X. Wang, M. Hu, C. Zhu and Z. Guo, Chem. Sci., 2014, 5, 2761–2770. 70. S. J. Park, J. H. Shin, E. S. Kim, Y. K. Jo, J. H. Kim, J. J. Hwang, J. C. Kim and D.-­H. Cho, FEBS Lett., 2012, 586, 4303–4310. 71. R. Kachadourian, H. M. Brechbuhl, L. Ruiz-­Azuara, I. Gracia-­Mora and B. J. Day, Toxicology, 2010, 268, 176–183. 72. K. Laws, G. Bineva-­Todd, A. Eskandari, C. Lu, N. O'Reilly and K. Suntharalingam, Angew. Chem., Int. Ed., 2018, 57, 287–291. 73. J. Gallagher, O. Zelenko, A. D. Walts and D. S. Sigman, Biochemistry, 1998, 37, 2096–2104. 74. M. McCann, A. Kellett, K. Kavanagh, M. Devereux and A. L. S. Santos, Curr. Med. Chem., 2012, 19, 2703–2714. 75. S. J. O'Day, A. M. M. Eggermont, V. Chiarion-­Sileni, R. Kefford, J. J. Grob, L. Mortier, C. Robert, J. Schachter, A. Testori, J. Mackiewicz, P. Friedlander, C. Garbe, S. Ugurel, F. Collichio, W. Guo, J. Lufkin, S. Bahcall, V. Vukovic and A. Hauschild, J. Clin. Oncol., 2013, 31, 1211–1218. 76. M. Nagai, N. H. Vo, L. Shin Ogawa, D. Chimmanamada, T. Inoue, J. Chu, B. C. Beaudette-­Zlatanova, R. Lu, R. K. Blackman, J. Barsoum, K. Koya and Y. Wada, Free Radic. Biol. Med., 2012, 52, 2142–2150. 77. R. K. Blackman, K. Cheung-­Ong, M. Gebbia, D. A. Proia, S. He, J. Kepros, A. Jonneaux, P. Marchetti, J. Kluza, P. E. Rao, Y. Wada, G. Giaever and C. Nislow, PLoS One, 2012, 7, e29798. 78. S. E. Sherman and S. J. Lippard, Chem. Rev., 1987, 87, 1153–1181. 79. M. V. Keck and S. J. Lippard, J. Am. Chem. Soc., 1992, 114, 3386–3390. 80. A. Erxleben, Coord. Chem. Rev., 2018, 360, 92–121. 81. H. L. Mi, I. H. Lau and J. K. Barton, Inorg. Chem., 2007, 46, 9528–9530. 82. J. P. Hall, K. O'Sullivan, A. Naseer, J. A. Smith, J. M. Kelly and C. J. Cardin, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 17610–17614. 83. C. J. Cardin and J. P. Hall, in DNA-­targeting Molecules as Therapeutic Agents, ed. M. J. Waring, Royal Society of Chemistry, Cambridge, 2018, pp. 198–227. 84. J. Chen and J. Stubbe, Nat. Rev. Cancer, 2005, 5, 102–112. 85. M. Kuwabara, C. Yoon, T. Goyne, T. Thederahn and D. S. Sigman, Biochemistry, 1986, 25, 7401–7408. 86. Z. Molphy, A. Prisecaru, C. Slator, N. Barron, M. McCann, J. Colleran, D. Chandran, N. Gathergood and A. Kellett, Inorg. Chem., 2014, 53, 5392–5404. 87. Z. Molphy, C. Slator, C. Chatgilialoglu and A. Kellett, Front. Chem., 2015, 3, 1–9. 88. L. Becco, J. C. Garcia-­Ramos, L. Ruiz-Azuara, D. Gambino and B. Garat, Biol. Trace Elem. Res., 2014, 161, 210–215.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00091

Recent Advances in Anticancer Copper Compounds

119

89. J. C. García-­Ramos, A. G. Gutiérrez, A. Vázquez-­Aguirre, Y. Toledano-­ Magaña, A. L. Alonso-­Sáenz, V. Gómez-­Vidales, M. Flores-­Alamo, C. Mejía and L. Ruiz-­Azuara, BioMetals, 2017, 30, 43–58. 90. A. Kellett, M. McCann, O. Howe, M. O'Connor and M. Devereux, Int. J. Clin. Pharmacol. Ther., 2012, 50, 79–81. 91. Z. Molphy, D. Montagner, S. S. Bhat, C. Slator, C. Long, A. Erxleben and A. Kellett, Nucleic Acids Res., 2018, 46(19), 9918–9931. 92. R. Galindo-­Murillo, L. Ruíz-­Azuara, R. Moreno-­Esparza and F. Cortés-­ Guzmán, Phys. Chem. Chem. Phys., 2012, 14, 15539–15546. 93. R. Galindo-­Murillo, J. C. García-­Ramos, L. Ruiz-­Azuara, T. E. Cheatham and F. Cortés-­Guzmán, Nucleic Acids Res., 2015, 43, 5364–5376. 94. H. Niyazi, J. P. Hall, K. O'Sullivan, G. Winter, T. Sorensen, J. M. Kelly and C. J. Cardin, Nat. Chem., 2012, 4, 621–628. 95. J. Stubbe and J. W. Kozarich, Chem. Rev., 1987, 87, 1107–1136. 96. J. Serment-­Guerrero, M. E. Bravo-­Gomez, E. Lara-­Rivera and L. Ruiz-­ Azuara, J. Inorg. Biochem., 2017, 166, 68–75. 97. A. Kellett, M. O'Connor, M. McCann, M. McNamara, P. Lynch, G. Rosair, V. McKee, B. Creaven, M. Walsh, S. McClean, A. Foltyn, D. O'Shea, O. Howe and M. Devereux, Dalton Trans., 2011, 40, 1024–1027. 98. A. Kellett, Z. Molphy, C. Slator, V. McKee, N.P. Farrell, Chem. Soc. Rev., 2019, 48, 971–988. 99. N. Fantoni, Z. Molphy, C. Slator, G. Menounou, G. Toniolo, G. Mitrikas, V. McKee, C. Chatgilialoglu and A. Kellett, Chem. -­ Eur. J., 2018, 25(1), 221–237. 100. T. J. P. McGivern, C. Slator, A. Kellett and C. J. Marmion, Mol. Pharm., 2018, 15(11), 5058–5071.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Chapter 5

Anticancer Gold Compounds DI Hu, CHUN-NAM Lok and CHI-MING Che* Department of Chemistry and State Key Laboratory of Synthetic Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, China *E-­mail: [email protected]

5.1  T  he Development of Gold Compounds in Medicine Gold was regarded as a precious metal effective against disease in ancient times. In the early 20th century, [Au(CN)2]− and gold(i) thiolates were used for the treatment of tuberculosis. Later, the anti-­rheumatic properties of gold complexes were recognized, leading to their clinical application in the treatment of rheumatoid arthritis (RA). Anti-­rheumatic gold drugs are a class of disease-­modifying anti-­rheumatic drugs, which can reduce inflammation and disease progression in patients with RA.1 Auranofin (a gold(i)– phosphane–thiolate complex, see Figure 5.1) is one such oral anti-­rheumatic drug. In recent decades, there has been growing interest in the development of gold compounds for anticancer applications due to their cytotoxicity against cancer cells.2,3 Inspired by the clinical success of platinum(ii)-­based anticancer agents, research has been focused on establishing new metal complexes with promising anticancer properties.4–6 For traditional platinum-­based therapeutics, the anticancer activity is generally attributed to non-­reparable interaction with DNA. Nonetheless, the design of anticancer agents that are able to target non-­DNA molecules associated with cancer development has been   Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

120

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

121

Figure 5.1  Antiarthritic  gold(i) drugs with anticancer properties. considered an attractive strategy. A number of gold(i) and gold(iii) complexes have been reported to display antitumour properties against a wide range of cancers through different anticancer mechanisms, specifically the selective inhibition of target proteins and enzymes. Gold(i) ions generally display a strong binding affinity to thiol-­containing proteins, while gold(iii) compounds have shown affinity also for N-­donor ligands, including the side-­ chains of histidine residues. The square planar coordination geometry and electronic properties of gold(iii) complexes are their key features, which afford unique structural scaffolds for binding to proteins, leading to functional inhibition, as well as to enhanced redox activity. In the past several decades, research has demonstrated that both families of gold complexes are attractive candidates for applications in cancer therapy.

5.2  Anticancer Gold(i) Complexes The discovery of the anticancer efficacy of auranofin triggered ongoing efforts in the development of gold(i) complexes in cancer therapy. Gold(i) complexes undergo ligand displacement under physiological conditions. The auxiliary ligands can modulate the lipophilicity, stability and binding affinity of gold(i) towards target proteins, thus affecting the intracellular transformation of gold(i) species and their anticancer activity. Previous research has described a number of gold(i) complexes, bearing thiolate, phosphane, N-­heterocyclic carbene (NHC), alkynyl and thiourea ligands, showing promising anticancer potential both in vitro and in vivo.7

5.2.1  Antiarthritic Gold(i) Drugs with Anticancer Activities Starting in the 1970s, the anticancer properties of antiarthritic gold(i) complexes (Figure 5.1) have been extensively studied. Auranofin was found to inhibit cancer cell growth in vitro and effectively increase the lifespan of mice inoculated with lymphocytic leukemia P388 cells in a concentration and dose dependent manner.8 Mechanistic studies have revealed that auranofin can induce oxidation of cysteine (Cys)-­containing peptides, exemplified by the inhibition of thioredoxin reductase (TrxR).9 In addition, formation of reactive oxygen species (ROS), activation of p38 mitogen-­activated protein kinase (MAPK) and the inhibition of proteasome-­associated deubiquitinases (DUBs) have been demonstrated to be associated with auranofin-­induced apoptosis

View Online

Chapter 5

122

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

10–12

and tumour growth inhibition. However, auranofin was ineffective in several subcutaneously implanted solid tumour models and inactive when administered intravenously.13 The rapid covalent binding of auranofin with Cys-­34 of serum albumin subsequently decreases the active gold species and slows cellular uptake, thus resulting in a low in vivo cytotoxicity. The antiarthritic gold(i) drugs, aurothiomalate and aurothioglucose, can significantly increase the survival time of mice inoculated with syngenic cancer cells and inhibit tumour growth.14 These gold complexes block the interaction of protein kinase C iota (PKCι) with its downstream effector, the Par6 scaffold protein, and exert cytostatic effects in cancer cells.15 A search for current clinical trials (http://www.clinicaltrials.gov, August 2018) yielded several ongoing or recent studies for auranofin and aurothiomalate, including several cancers or leukemia (e.g. lung cancer, recurrent ovarian epithelial cancer, chronic lymphocytic leukemia).

5.2.2  Gold(i)–Phosphane Complexes Gold(i)–phosphane complex-­based anticancer agents have been extensively studied for over two decades.16 Among the most established families is that of two-­coordinated gold(i) complexes bearing phosphane ligands, which is represented by auranofin. Mirabelli and co-­workers performed a structure– activity relationship study on these types of compound, which revealed the influence of both the phosphane and the thiolate ligands on anticancer activity.17 Barrios and co-­workers demonstrated how variations in the size of phosphane ligands can markedly influence the cellular uptake process and biodistribution of auranofin analogues, as well as affecting the binding affinity of these complexes to target proteins and their inhibition of enzymes, such as the cathepsin family of lysosomal cysteine proteases.18 In addition to the linear coordination structures, a number of tetrahedral bis-­chelated gold(i)–phosphane complexes have also been reported to display promising anticancer properties. Sadler and Berners-­Price first reported the antitumoral activity of [Au(dppe)2]Cl (dppe = bis(diphenylphosphino)ethane) (Figure 5.2).19a [Au(dppe)2]Cl is inert in the presence of glutathione (GSH) under physiological conditions. This complex has been reported to significantly inhibit tumour growth in different animal models, including leukemia and solid tumours. However, preclinical toxicological studies identified severe toxicity to the heart, liver and lung in dogs and rabbits.20 The toxicity of [Au(dppe)2]Cl is related to its high lipophilicity and stability, which cause non-­selective accumulation of gold species in mitochondria, leading to mitochondrial dysfunction.19b Subsequent studies focused on developing anticancer gold(i)–phosphane complexes with decreased lipophilicity and higher selectivity in killing cancer cells over normal cells. The lipophilic/hydrophilic balance of gold(i) complexes can be modulated through phosphane ligand design. Several such examples have indeed revealed that modifications in the phosphane ligand could slightly reduce the lipophilicity while improving the selectivity for cancer cells.21 By replacing the phenyl substituents with pyridyl

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

123

Figure 5.2  Examples  of gold(i)–phosphane complexes with anticancer properties. groups, a series of structural analogues of [Au(dppe)2]Cl have been synthesized with hydrophilic/lipophilic character spanning a wide range (Figure 5.2). [Au(d2pype)2]Cl (d2pype = 1,2-­bis(di-­2-­pyridylphosphino)ethane), containing an ethyl-­bridged 2-­pyridyl phosphane ligand, displays intermediate lipophilicity and promising antitumour activity. The lipophilic/hydrophilic balance was further fine-­tuned by using the propyl-­bridged 2-­pyridyl phosphane ligand (d2pypp). [Au(d2pypp)2]Cl induced mitochondrial dysfunction, inhibited TrxR activity and displayed selective cytotoxicity against breast cancer cells but not against normal breast cells.22,23 Additional four-­coordinated gold(i) complexes containing phosphane ligands with promising anticancer activity have been developed. Katti and co-­workers reported that [Au[P(CH2OH)3]4]Cl (Figure 5.2) is cytotoxic against various cancer cell lines and significantly prolonged the survival time of mice inoculated with meth A sarcoma cells.24 Some gold(i)–phosphane complexes, such as [Au(PPh3)]Cl, [Au2(dppe)]Cl2 and [Au3(dpmp)] Cl3 (Figure 5.2), have been reported to induce autophagy and subsequently cause cell death.25 Interestingly, Contel and co-­workers reported the synthesis, characterization and stability studies of new heterometallic titanocene–gold complexes [(η-­C5H5)2TiMe(µ-­mba)Au(PR3)] as potential chemotherapeutics for renal cancer.26 The compounds are stable in physiological media and highly cytotoxic against human cancer renal cell lines. In vivo study of Caki-­1 renal cancer xenograft in mice showed a marked tumour reduction (67%) after treatment for 28 days (3 mg kg−1 every other day).

5.2.3  Gold(i)–Thiourea Complexes Thiourea and the related thiosemicarbazone ligands can stabilize gold(i).27 Gold(i)–thiourea complex (Figure 5.3) can potently inhibit TrxR activity via a tight binding mode involving modification of the redox active selenocysteine (Sec)/Cys residue via gold(i) coordination. Gold(i)–thiourea is cytotoxic

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

124

Chapter 5

Figure 5.3  A  gold(i)–thiourea complex with anticancer property. towards cancer cell lines with IC50 values in the low micromolar range. In vivo experiments on mice bearing NCI-­H460 non-­small cell lung cancer cells, intraperitoneally administered with 100 mg kg−1 of gold(i)–thiourea twice a week, resulted in a reduction of tumour size of ca. 38% after a 28-­day treatment, with no body weight loss and mouse death observed during the course of examination.27

5.2.4  Gold(i)–NHC Complexes Gold(i) ion can be stabilized by strong electron-­donating NHC ligands against reduction to gold(0). Gold(i)–NHC complexes are usually stable in ambient and physiological conditions, and provide unique structural scaffolds for specific interactions with biomolecules.28,29 The NHC is also a relatively non-­toxic ligand. In recent years, the anticancer properties of a number of gold(i)–NHC complexes have been extensively studied.30 Modifications on the imidazolium ring or N-­substituents can be facile, and influence the lipophilicity and reactivity of gold(i)–NHC complexes, resulting in tuneable anticancer activity. Berners-­Price and co-­workers first reported the anticancer properties of a panel of cationic linear [Au(NHC)2]+ complexes (Figure 5.4).31,32 These [Au(NHC)2]+ complexes bear different N-­substituents and display a wide range of lipophilicity. Some of these complexes display higher cytotoxicity in breast cancer cells over normal breast cells. [Au(NHC)2]+ complexes are delocalized lipophilic cations (DLCs) and mechanistic studies have revealed that [Au(NHC)2]+ complexes can accumulate in the mitochondria, trigger Ca2+-­ sensitive mitochondrial swelling and subsequently lead to mitochondria-­ dependent apoptosis. [Au(NHC)2]+ complexes were proposed to selectively inhibit TrxR activity through a two-­step reaction mechanism involving the successive substitution of the two NHC ligands to form either [Au(Cys)2] or [Au(Sec)2], with the formation rate of the latter being much faster.33 The thiol reactivity of gold(i)–NHC complexes is influenced by the N-­substituted groups. Those gold(i)–NHC complexes with bulkier groups have the gold(i) centre shielded from thiol attack. Ott and co-­workers described a series of gold(i)–NHC chloride complexes, [Au(NHC)Cl], bearing benzimidazole-­derived NHC ligands (Figure 5.4). The

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

125

Figure 5.4  Examples  of gold(i)–NHC complexes with anticancer properties. anti-­mitochondrial effects and TrxR inhibitory activity have been shown to contribute to the mechanisms of anticancer action of [Au(NHC)Cl].34 The binding of gold(i)–NHC to Cys and Sec has recently been characterized.35 Exchanging the anionic chloride ligand with other neutral ligands, such as NHC or PPh3, produces cationic [Au(NHC)L]+ species. Compared with the neutral chloride derivative, the cationic complexes displayed less inhibition of TrxR activity due to lower reactivity with thiol-­containing proteins. However, the cationic gold species exhibited increased cellular uptake and greater accumulation in mitochondria, as well as enhanced anti-­mitochondrial activity.36 To improve the anticancer potency of the gold(i)–NHC complexes, various functional groups are conjugated to the trans-­ligand or the NHC.37,38 For example, functionalization of gold(i) complexes to a leukemia-­specific DNA aptamer via the NHC ligand resulted in markedly enhanced and selective cytotoxicity in leukemia cells.39 Another example is the incorporation of a DNA intercalating naphthalimide moiety onto the NHC ligand, [Au(NHCNap) Cl] (Figure 5.4), which showed both TrxR inhibition and DNA intercalation properties.40 Appending structures derived from bioactive natural products to the NHC ligand favours the targeting of gold(i)–NHC complexes. For example, Casini and co-­workers showed that [Au(9-­methylcaffein-­8-­ylidene)2]+ could tightly and selectively bind to G-­quadruplexes and is moderately cytotoxic against a wide range of cancer cell lines while non-­toxic towards non-­ tumorigenic ones (Figure 5.4).41,42

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

126

Chapter 5

Che and co-­workers reported a dinuclear gold(i) complex [Au2(dcpm) (bisNHC2C4)]2+ (dcpm = bis(dicyclohexylphosphane)methane) (Figure 5.4).43 This complex utilizes a bridging bis(NHC) ligand and a diphosphane ligand to attain appropriate chemical stability and reactivity towards protein thiols. [Au2(dcpm)(bisNHC2C4)](PF6)2 could selectively react with thiol-­ containing enzymes, but displays sufficient stability towards GSH and blood thiols. The compound potently inhibits the activity of TrxR with the IC50 value being 38 nM.43 The molecular interaction between the gold species and TrxR has been elucidated, showing that the dinuclear [Au2(dcpm)]+ unit coordinates to S(Cys) and Se(Sec) residues by release of the bis(NHC) ligand.43 This type of dinuclear gold(i) complex could inhibit cancer stem-­ like cell activity and significantly suppress tumour growth. In vivo experiments on mice bearing HeLa cells, intraperitoneally administered with 5 mg kg−1 of [Au2(dcpm)(bisNHC2C4)](PF6)2 once every three days, resulted in a reduction of tumour size of ca. 79% and 81% after 6 and 9 days of treatment, respectively. No mouse death or body weight loss was observed after treatment with [Au2(dcpm)(bisNHC2C4)](PF6)2 at this dosage. Inhibition of angiogenesis was also observed in tumour models of mice treated with [Au2(dcpm)(bisNHC2C4)]2+.

5.2.5  Gold(i)–Alkynyl Complexes A series of [Au(PPh3)(alkynyl)] complexes (Figure 5.5) exhibit anticancer properties.44 These complexes can selectively inhibit TrxR activity over the structurally related enzyme glutathione reductase. Treatment of MCF-­7 breast cancer cells with [Au(PPh3)(alkynyl)] was observed to affect tumour cell metabolism and mitochondrial respiration. [Au(PPh3)(alkynyl)] can significantly inhibit the formation of blood vessels in a zebrafish embryo model. A dinuclear gold(i) species, [Au2(PPh3)2(bis-­alkynyl)] with a diethynylfluorene linkage (Figure 5.5), is cytotoxic towards different cancer cell lines with IC50 values at low micromolar levels.45 Interestingly, intraperitoneal administration of mice bearing Hep3B tumours with [Au2(PPh3)2(bis-­alkynyl)] at 2.5 mg per kg per day for nine successive days elicited a significant inhibition of tumour growth with limited adverse effects on vital organs, including the liver and kidney.

Figure 5.5  Gold(i)–alkynyl  complexes with anticancer properties.

View Online

Anticancer Gold Compounds

127

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

5.3  Anticancer Gold(iii) Complexes Gold(iii) complexes have long been regarded as potential alternatives to platinum(ii)-­based anticancer agents, since the two types of complexes possess similar electronic configuration (d8) and structural similarity. However, the instability and reactivity of the gold(iii) ion under physiological conditions hampers the potential applications of these gold complexes in medicine. Gold(iii) can be reduced to gold(i) or gold(0) via intracellular redox reactions.46 Therefore, it is important to stabilize the gold(iii) ion using strong electron-­donating and/or chelating ligands. Over the past few decades, gold(iii) complexes containing ligands including porphyrin, C^N^C, N^N^N, C^N^N, N^N, C^N and dithiocarbamate have been reported to possess good stability under physiological conditions and display promising anticancer potency against various cancer cells.

5.3.1  Gold(iii) Porphyrin In 2003, Che and co-­workers first introduced the anticancer potential of a gold(iii) meso-­tetraphenylporphyrin (TPP) complex (denoted gold-­1a or [Au(TPP)]Cl, Figure 5.6A).47 Since then, several gold(iii) complexes with diverse modifications on the porphyrin ligands have been developed.48 Among these complexes, gold-­1a is one of the most promising anticancer candidates for future development. Various studies have proven the stability of gold-­1a under physiological conditions; this complex is stable against reduction by the biological reductant GSH. Gold-­1a, at concentrations in the low micromolar or even nanomolar levels, displays in vitro cytotoxicity in a wide range of cancer cell lines derived from solid tumour. In vivo studies have also demonstrated that gold-­1a can inhibit tumour growth in different animal cancer models.7 Gold-­1a has demonstrated the capacity to inhibit the self-­renewal ability of cancer stem-­like cells.49 Compared to the clinically used cisplatin, gold-­1a shows a higher anticancer potency, with IC50 values significantly lower than those of cisplatin. More importantly, gold-­1a is equally active against both cisplatin-­sensitive and multidrug-­resistant cancer cells (Figure 5.6B).50 The lack of cross-­resistance to cisplatin suggests that gold-­1a exerts anticancer activity via mechanisms that are different from those of cisplatin. The mechanisms of anticancer action of gold-­1a have been extensively studied using various approaches, including biochemical analyses, transcriptomics and proteomics. Several lines of evidence have revealed that gold-­1a is an active anti-­mitochondrial agent. Treatment of cancer cells with gold-­1a results in a rapid depletion of mitochondrial transmembrane potential, with concomitant suppression of Bcl-­2 protein levels, and subsequent induction of both the caspase-­dependent and caspase-­independent mitochondrial apoptotic pathways. Cellular oxidative stress and changes in the balance between pro-­apoptotic and anti-­apoptotic proteins are crucial events in gold-­1a-­induced apoptosis.51,52 Additional studies have also revealed that

Published on 05 April 2019 on https://pubs.rsc.org | doi:

128

Figure 5.6  (A)  Chemical structure of gold-­1a. (B) Gold-­1a significantly inhibited tumour growth of both cisplatin-­sensitive (A2780) and

Chapter 5

cisplatin-­resistant (A2780cis) ovarian cancer cells without body weight loss in the course of treatment. Reproduced from ref. 50 with permission from The Royal Society of Chemistry. (C) A trifunctional, clickable photoaffinity probe of gold-­1a was used to identify HSP60 as one of the direct molecular targets of gold(iii) porphyrins. (D) Gold-­1a inhibited the chaperone activity of HSP60. Reproduced from ref. 55 with permission from John Wiley and Sons, Copyright © 2016 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

129

Figure 5.7  Examples  of gold(iii) porphyrin complexes with different peripheral substituents on the porphyrin ligands.

gold-­1a can induce cell cycle arrest at the G0/G1 phase, activate p38 MAPK and inhibit TrxR.53,54 However, identification of the direct molecular targets that account for the anticancer actions of metal complexes has remained a difficult task. A trifunctional, clickable photoaffinity probe of gold-­1a and verification experiments by cellular thermal shift, saturation-­transfer difference NMR, protein fluorescence quenching, and protein chaperone assays were employed to demonstrate that the mitochondrial chaperone, heat shock protein 60 (HSP60), is one of the direct molecular targets of gold(iii) porphyrins (Figure 5.6C).55 The ease of structural modification provides ample opportunities for optimizing the anticancer activities of gold(iii) porphyrin complexes by modifying the peripheral substituents on the porphyrin ligand.48 An analogue of gold-­1a with a meso-­hydroxyl phenyl group (gold-­2a, Figure 5.7) displays improved solubility in aqueous media.56 Gold-­2a is highly cytotoxic against breast carcinoma and significantly suppresses breast tumour growth in nude mice. Inhibition of the activity of class I histone deacetylase via aberrant Wnt/β-­catenin signalling contributes to the anticancer action of gold-­2a. Conjugation of other functional groups onto the porphyrin ligand can influence the mechanisms of anticancer action of gold(iii) complexes. For example, [Au (4-­glucosyl-­TPP)]+ (Figure 5.7), a gold(iii) porphyrin bearing a saccharide conjugation, displays cytostatic properties and induced S-­phase cell cycle arrest.48

5.3.2  Coordination Gold(iii) Complexes with Various Ligands Messori and co-­workers reported a panel of dioxo-­bridged dinuclear gold(iii) complexes, such as [Au2(µ-­O)2(N^N)2]2+ (wherein N^N = 2,2′-­bipyridine or a substituted 2,2′-­bipyridine) and [Au2(µ-­O)2(phen2Me)2]2+ (wherein phen2Me = 2,9-­dimethyl-­1,10-­phenanthroline) (Figure 5.8).57–59 These dinuclear gold(iii) complexes displayed anticancer activity against a wide range of cancer cell

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

130

Chapter 5

Figure 5.8  Chemical  structures of dioxo-­bridged dinuclear gold(iii) complexes, gold(iii) dithiocarbamato peptidomimetics, and gold(iii) pyrrolidinedithiocarbamato (PDT) complexes.

lines, including both cisplatin-­sensitive and -­resistant cancer cells. Superoxide dismutase (SOD) can react with the [Au2(µ-­O)2(phen2Me)2]2+ complex, breaking down the dimetallic gold(iii) complex and reducing gold(iii) to two gold(i) ions, with the free phen2Me ligands released.57 Conjugation of biological functional groups, such as amino acids and oligopeptides, to dithiocarbamate ligands has emerged as an attractive strategy for the development of anticancer gold(iii) dithiocarbamato compounds.60 Several gold(iii) dithiocarbamato peptidomimetics (Figure 5.8) have shown potential in targeted anticancer applications.61 These complexes displayed anticancer properties in vitro and in vivo against human breast cancer cells. The proteasome was shown to be a major target of this class of gold(iii) complexes.61 The gold(iii) pyrrolidinedithiocarbamato (PDT) complex, [AuBr2(PDT)] (Figure 5.8), has been reported to be active against a panel of cancer cell lines by affecting cellular components that play crucial roles in the survival of cancer cells.62 Zinc-­based biodegradable metal-­organic frameworks (MOFs) were used as drug carriers. Incorporating the gold(iii) pyrrolidinedithiocarbamato complex in a zinc-­based MOF significantly enhanced the anticancer activity against a cisplatin-­resistant ovarian cancer cell line (A2780cis).63 Casini and co-­workers reported a series of gold(iii) complexes containing nitrogen donor ligands, including 1,10-­phenanthroline, 2,2′-­bipyridine, 4,4′-­dimethyl-­2.2′-­bipyridine, and 4,4′-­diamino-­2,2′-­bipyridine, with anticancer properties (Figure 5.9).64 Of note, the inhibition of aquaporins (AQPs) has been shown to be associated with the mechanisms of anticancer action of these complexes.64–66 AQPs are membrane water/glycerol channels and emerge as targets for cancer therapy.67

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

131

Figure 5.9  Examples  of anticancer gold(iii) complexes with nitrogen donor ligands. A water-­soluble gold(iii) corrole (Figure 5.9) was cytotoxic and cytostatic towards cisplatin-­resistant cancer cell lines.68 Compared with other metal analogues, gold(iii) corrole had a lower binding affinity to human serum albumin (HSA), which might account for its cell killing ability. A class of cytotoxic pyrrole-­based gold(iii) macrocycles (Figure 5.9) were reported to be catalytic inhibitors of topoisomerase IB (Topo I).69 The gold(iii) centre was incorporated with two pyrrole-­imine units linked to a quinozaline moiety on one side and an alkyl chain bridge on the opposite side. Macromolecular simulations demonstrated the essential role of gold(iii) for both DNA intercalation and enzyme inhibition. The gold(iii) macrocycle intercalates DNA at the enzyme's 5′-­TA-­3′ dinucleotide target sequence via π–π stacking, and an Au⋯O electrostatic interaction involving a thymine carbonyl group accounts for the base pair specificity. Thus, this class of gold(iii) macrocycles blocks substrate recognition by Topo I through steric repulsion.69

5.3.3  Cyclometallated Gold(iii) Complexes 5.3.3.1 Pincer-­t ype Gold(iii) Complexes Apart from tetradentate ligands, coordination of gold(iii) with a dianionic tridentate ligand and a neutral auxiliary ligand also produces stable gold(iii) complexes. Che and co-­workers reported the syntheses and anticancer activities of a series of cyclometallated gold(iii) complexes bearing a tridentate C-­deprotonated C^N^C ligand: [Aum(C^N^C)mL]n+ (wherein m = 1–3; n = 0–3; H2C^N^C = 2,6-­diphenylpyridine; Figure 5.10).70 The dianionic tridentate [C^N^C]2− ligand scaffold stabilizes the electrophilic gold(iii) ion and improves the stability of gold(iii) complexes under physiological conditions. The neutral auxiliary ligand, L, can modulate the mechanisms of action, thus influencing the anticancer potency of gold(iii) complexes. With triphenylphosphane as auxiliary ligand, [Au (C^N^C)(PPh3)]+ (Figure 5.10) is

View Online

Chapter 5

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

132

Figure 5.10  Examples  of anticancer [Aum(C^N^C)mL]n+ complexes with different neutral auxiliary ligands L.

cytotoxic against nasopharyngeal carcinoma (NPC). With a bridging bis(diphenylphosphino)propane (µ-­dppp) ligand, the dinuclear gold(iii) complex [Au2(C^N^C)2(µ-­dppp)]2+ (Figure 5.10) displays enhanced cytotoxicity towards a panel of cancer cell lines with IC50 values down to the submicromolar range.71 Importantly, [Au2(C^N^C)2(µ-­dppp)]2+ was also found to elicit a significant inhibition of tumour growth in animal models. In vivo experiments on mice bearing PLC hepatoma cells, intraperitoneally administered with 10 mg kg−1 of [Au2(C^N^C)2(µ-­dppp)]2+ twice per week, resulted in a reduction of tumour size of 77% after four weeks of treatment. [Au2(C^N^C)2(µ-­dppp)]2+ did not result in mortality, significant change in body weight or other apparent adverse effects. Mechanistic studies indicated that TrxR inhibition and induction of endoplasmic reticulum stress are associated with the anticancer action of this complex. In addition to the toxic phosphane ligands, employing a relatively non-­ toxic, electron-­donating NHC ligand as auxiliary ligand produces a series of [Au(C^N^C)(NHC)]+ complexes (Figure 5.11) that are highly cytotoxic against a wide range of cancer cell lines. These compounds displayed much higher cytotoxicity against non-­small cell lung cancer cells than normal lung fibroblast cells.72 Administration of [Au(C^N^C)(NHC2Me)]OTf (NHC2Me = 1,3-­dimethylimidazol-­2-­ylidene) at 10 mg kg−1 week−1 in nude mice bearing PLC hepatoma cells for 28 days significantly suppressed tumour growth with no apparent toxic side effects observed during the entire course of examination.72 The [Au(C^N^C)(NHC)]+ complexes are able to intercalate into DNA and subsequently prevent Topo I-­mediated relaxation of supercoiled DNA. Recently, with the use of clickable photoaffinity probes of the [Au(C^N^C) (NHC)]+ complex, Che and co-­workers have revealed the specific engagement of the gold(iii) complexes with multiple cellular targets, including HSP60, vimentin, nucleophosmin, and YB-­1, accompanied by the expected downstream consequences (Figure 5.11).73

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

133

Figure 5.11  (A)  Chemical structure of the gold(iii)–NHC complex. (B) Chemical

probes of [Au(C^N^C)(NHC)]+ for target identification; green: photoaffinity unit; blue: clickable moiety. (C) Schematic of procedure to identify cellular protein targets using the chemical probe. (D) Fluorescence scanning of 2D gels of lysates from HeLa cells treated with the probe and then clicked with azide-­Cy5. Proteins (1–6) were identified by MALDI-­TOF as the molecular targets of the gold(iii)–NHC complex. Reproduced from ref. 73 with permission from John Wiley and Sons, Copyright © 2017 Wiley‐VCH Verlag GmbH & Co. KGaA, Weinheim.

Cyclometallated gold(iii)–NHC complexes [Au(N^N^N)(NHC)]+ with 2,6-­bis(imidazole-­2-­yl)pyridine or 2,6-­bis(benzimidazol-­2-­yl)pyridine ligands were reported to serve as a switch-­on probe for thiols in biological systems (Figure 5.12).74 In cancer cells, [Au(N^N^N)(NHC)]+ complexes were reduced to gold(i) species with the release of fluorescent H2N^N^N ligands. Upon gold(iii) reduction, the anticancer gold(i)–NHC species were formed and these [Au(N^N^N)(NHC)]+ complexes could significantly suppress tumour growth in mice bearing HeLa xenografts.74 The gold(iii) centre can be stabilized by other tridentate ligand systems, such as 2,2′,2′-­terpyridine (terpy), which was first described by Lippard and co-­workers (Figure 5.13).75 Guo and co-­workers reported a panel of gold(iii) cations with aminoquinoline ligands (Figure 5.13). These complexes are highly cytotoxic against melanoma and lung cancer cell lines.76,77 Messori and co-­workers reported [Au(bipydmb-­H)(OH)]+ and [Au(bipydmp-­H) (2,6-­x ylidine)]+ (bipydmb-­H = 6-­(1,1-­dimethylbenzyl)-­2,20-­bipyridine) (Figure 5.14), both displaying promising cytotoxicity against both cisplatin-­ sensitive and -­resistant ovarian carcinoma cell lines, with inhibition of

View Online

Chapter 5

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

134

Figure 5.12  Gold(iii)–NHC  complexes with fluorescent N^N^N ligands as switch-­on probe for thiols.

Figure 5.13  Examples  of anticancer gold(iii) complexes with tridentate N^N^N ligands.

dmb Figure 5.14  [Au(bipy  -­H)(OH)]+ and [Au(bipydmp-­H)(2,6-­x ylidine)]+ complexes

with anticancer properties.

mitochondrial thioredoxin reductase and mitochondrial respiration.78 Interestingly, the compound [Au(bipydmb-­H)(OH)]+ turned out to be more reactive with nucleobases than with amino acids and proteins.79

View Online

Anticancer Gold Compounds

135

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

5.3.3.2 Bidentate C^N-­t ype Gold(iii) Complexes Che and co-­workers have developed cyclometallated gold(iii) complexes bearing a C-­deprotonated C^N (HC^N = 2-­phenylpyridine) ligand. With biguanide as the auxiliary ligand, a water-­soluble [Au(nBuC^N)(biguanide)]Cl (HC^N = 2-­phenylpyridine) was prepared (Figure 5.15).80 ESI-­MS revealed that this complex is able to form a gold(iii)–GSH adduct and showed cytotoxicity towards different cancer cell lines in association with endoplasmic reticulum stress. [Au(nBuC^N)(dedt)]Cl, with the same C^N ligand and a dithiocarbamate ligand (dedt) (Figure 5.15), showed selective inhibition towards breast cancer MCF-­7 cells but was less toxic to non-­tumorigenic immortalized liver cells (MIHA).81 ESI-­MS experiments revealed that this complex could form covalent adducts with cysteine-­containing peptides and proteins such as deubiquitinases. Transcriptomic and connectivity map analysis, together with the cell-­based deubiquitinase inhibition assay, indicated that deubiquitinases could be cellular targets of this anticancer gold(iii) complex. A series of C^N cyclometallated gold(iii) compounds with the general formula [Au(pyb-­H)L1L2]n+ (Pyb-­H = C^N cyclometallated 2-­benzylpyridine, L1 and L2 = chloride, phosphane or glucosethiolato lignads) (Figure 5.16) have been reported.82 Among them, the derivative bearing one phosphane ligand

+ Figure 5.15  [Au(nBuC^N)(biguanide)]  and [Au(nBuC^N)(dedt)]+ complexes with

anticancer properties.

b Figure 5.16  [Au(py  -­H)L1L2]n+ (Pyb-­H = C^N cyclometallated 2-­benzylpyridine, L1

and L2 = chloride, phosphane or glucosethiolato ligands) complexes with anticancer properties.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

136

Chapter 5

displays the most cytotoxicity towards the HCT116 cancer cell line overexpressing p53. The C^N cyclometallated gold(iii) compounds inhibited the zinc finger enzyme PARP-­1 at nanomolar concentration levels. The binding of gold(iii) compounds to the zinc finger domain of PARP-­1 was studied by a hyphenated mass spectrometry approach combined with quantum mechanics/molecular mechanics studies.83 Compared with other cyclometallated gold(iii) compounds, such as [Au(phen)Cl2]+ (phen = 1,10-­phenanthroline; Figure 5.9), the Au–C^N compound is the most selective candidate for disrupting the PARP-­1 zinc finger domain, forming distinct adducts via replacement of the auxiliary ligands of the gold complex with two cysteinato groups which coordinate to the Au3+ ions. Upon binding to the Cys residues of PARP-­ 1, the Au–C^N compound is able to preserve the oxidation state of gold(iii) and retain the C^N ligand, resulting in alterations of the DNA recognition domain and protein inhibition.

5.4  N  ano Formulation of Gold Complexes with Improved Anticancer Potency Although a great number of metal complexes have been reported to display anticancer potency in various human cancer cell lines, applications of metal complexes in clinical use are still lacking due to their high toxicity in normal cells and tissues, rapid inactivation and frequent occurrence of resistance. Thus, the development of a rational delivery system to increase metal complex bioavailability with fewer side effects is needed. Nano formulation is a potential solution to the challenges facing clinical translation of anticancer metal complexes. Encapsulation of anticancer metal complexes in nanoparticles can increase drug circulation time in the blood, improve biodistribution and reduce toxic side effects.84 Encapsulation of anticancer gold(iii) complexes bearing porphyrin or Schiff-­base ligands by gelatin–acacia microcapsules improved the solution stability and/or in vivo efficacy compared with unencapsulated complexes alone.85 Mesoporous silica nanoparticle (MSN) is also a suitable delivery vehicle for anticancer gold complexes. MSN has the advantages of high drug loading efficiency, ease of surface modification and low toxicity. In order to improve the biocompatibility and targeting efficiency, MSNs were coated with biodegradable chitosan linked with cancer cell targeting arginylglycylaspartic acid peptide. Encapsulation of gold-­1a in this way significantly improved the cytotoxicity of gold-­1a towards cancer cells but only slightly increased the cytotoxicity towards non-­cancerous L02 cells.86 A nanoparticle formulation made of cetyl alcohol and Brij 78 surfactant has been employed to incorporate gold-­1a to reduce the side effects.87,88 Such encapsulated gold-­1a nanoparticles had an average diameter of around 164 nm. This nanoparticle formulation enhanced the preferential uptake of gold-­1a into tumour tissue, rather than major organs, thereby leading to an increased in vivo antitumour effect. Poly(ethylene glycol) (PEG) is a hydrophilic polymer commonly applied to

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

137

Figure 5.17  (A)  Chemical structure of a multifunctional gold(iii) porphyrin–PEG

conjugate. (B) [Au(C^N^C)(4-­dpt)](CF3SO3), which can form a self-­ assembled supramolecular polymer.

drug formulation and an excellent candidate for the development of amphiphilic cytotoxic metal complexes. A multifunctional gold(iii) porphyrin–PEG conjugate [Au(TPP-­COO-­PEG5000-­OCH3)]Cl has been reported (Figure 5.17).89 The gold(iii) porphyrin–PEG conjugates undergo self-­assembly into nanostructures in aqueous media, and the active gold(iii) porphyrin moieties are released from the conjugates in cancer cells without the formation of additional toxic side products. These gold(iii) porphyrin–PEG conjugates display a higher selectivity for cancer cell lines over non-­tumorigenic cells. In vivo experiments have shown that the gold(iii) porphyrin–PEG conjugates can significantly inhibit tumour growth in nude mice bearing human colon cancer HCT116 xenografts. The enhanced permeability and retention effect, which is a characteristic of some solid tumours, presumably promotes the accumulation of gold(iii) porphyrin–PEG conjugates in tumour tissue rather than in normal organs, leading to low systemic toxicity.89 Moreover, the self-­assembly properties render gold(iii) porphyrin–PEG conjugates as nanocarriers to encapsulate other chemotherapeutics, which is advantageous for anticancer treatment. Self-­assembled supramolecular polymer formed by [Au(C^N^C)(4-­dpt)](CF3SO3) (4-­dpt = 2,4-­diamino-­6-­(4-­pyridyl)-­1, 3,5-­triazine) (1-­SP) (Figure 5.17) shows distinctive physical features, including concentration-­dependent specific viscosity and formation of nanofibrillar networks.90 Polymer 1-­SP displayed sustained cytotoxicity and selective cytotoxicity towards cancer cell lines. It can also encapsulate other cytotoxic agents such as gold-­1a to achieve sustained-­release behaviour.

5.5  Conclusions and Outlook The potential of gold complexes as anticancer agents has been extensively studied and much effort has been made in the development of gold-­based antitumour drugs. A number of studies have demonstrated that gold complexes may overcome the problem of cisplatin resistance. The lack of cross-­resistance to cisplatin suggests that gold complexes exert their

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

138

Chapter 5

anticancer activity via different mechanisms from the traditional DNA-­ targeting, platinum-­based anticancer agents. Thiol-­containing enzymes are generally considered the major molecular targets of anticancer gold complexes due to the high binding affinity of gold(i) ions to thiols. Targeting of mitochondria and of the Trx/TrxR system have also been reported to play important roles in gold pharmacology. In addition, various other cancer-­related proteins have been identified as molecular targets of gold complexes. These include, for example, the membrane transporters, aquaporins, zinc finger proteins, and the heat shock protein, HSP60. Development of new techniques, particularly exploiting “omics” technologies, makes it possible to elucidate both the molecular targets and the pathways affected by anticancer metal complexes, thus providing a relatively comprehensive insight into the mechanisms of action of anticancer gold complexes. In the development of bioactive gold(iii) complexes, reduction to gold(0) or gold(i) species in the presence of extracellular thiols is an important consideration. One strategy to improve the anticancer potency of gold complexes and to control their reactivity in physiological conditions is to optimize the coordinated ligand structures; for example making use of organometallic scaffolds. A great number of rationally designed ligands that may afford increased potency have been reported. Selectivity for tumour cells is a key issue that remains to be resolved in the development of anticancer gold compounds. One approach to tackle this is to exploit drug nanocarriers to improve stability and anticancer efficacy. Furthermore, with conjugation of new targeting ligands and/or functional groups, further development of anticancer gold complexes and their translation into clinical applications are feasible.

Abbreviations NHC N-­heterocyclic carbene Cys cysteine Sec selenocysteine TrxR thioredoxin reductase ROS reactive oxygen species MAPK mitogen-­activated protein kinase DUB deubiquitinase GSH glutathione DLC delocalized lipophilic cation SOD superoxide dismutase HSA human serum albumin Topo I topoisomerase I ER endoplasmic reticulum MSN mesoporous silica nanoparticle PEG poly(ethylene glycol)

View Online

Anticancer Gold Compounds

139

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Acknowledgements We acknowledge the support from State Key Laboratory of Synthetic Chemistry and Innovation Technology Fund (ITS/130/14FP). We thank Dr Taotao Zou and Dr Jiesheng Huang for their constructive comments.

References 1. C. F. Shaw III, Chem. Rev., 1999, 99, 2589. 2. S. Jurgens, F. E. Kuhn and A. Casini, Curr. Med. Chem., 2018, 25, 437. 3. A. Casini, R. W. Sun and I. Ott, Met. Ions Life Sci., 2018, 18, 199–218. 4. P. C. A. Bruijnincx and P. J. Sadler, Curr. Opin. Chem. Biol., 2008, 12, 197. 5. N. Muhammad and Z. Guo, Curr. Opin. Chem. Biol., 2014, 19, 144. 6. S. H. van Rijt and P. J. Sadler, Drug Discovery Today, 2009, 14, 1089. 7. T. Zou, C. T. Lum, C.-­N. Lok, J.-­J. Zhang and C.-­M. Che, Chem. Soc. Rev., 2015, 44, 8786. 8. C. Roder and M. J. Thomson, Drugs R&D, 2015, 15, 13. 9. Y. M. Go, J. R. Roede, D. I. Walker, D. M. Duong, N. T. Seyfried, M. Orr, Y. Liang, K. D. Pennell and D. P. Jones, Mol. Cell. Proteomics, 2013, 12, 3285. 10. S.-­J. Park and I.-­S. Kim, Br. J. Pharmacol., 2005, 146, 506. 11. X. Chen, X. Shi, C. Zhao, X. Li, X. Lan, S. Liu, H. Huang, N. Liu, S. Liao, D. Zang, W. Song, Q. Liu, B. Z. Carter, Q. P. Dou, X. Wang and J. Liu, Oncotarget, 2014, 5, 9118. 12. N. Liu, X. Li, H. Huang, C. Zhao, S. Liao, C. Yang, S. Liu, W. Song, X. Lu, X. Lan, X. Chen, S. Yi, L. Xu, L. Jiang, C. Zhao, X. Dong, P. Zhou, S. Li, S. Wang, X. Shi, P. Q. Dou, X. Wang and J. Liu, Oncotarget, 2014, 5, 5453. 13. P. J. Sadler and R. E. Sue, Met.-­Based Drugs, 1994, 1, 107. 14. M. Stallings-­Mann, L. Jamieson, R. P. Regala, C. Weems, N. R. Murray and A. P. Fields, Cancer Res., 2006, 66, 1767. 15. A. M. Butler, M. L. Scotti Buzhardt, E. Erdogan, S. Li, K. S. Inman, A. P. Fields and N. R. Murray, Oncotarget, 2015, 6, 15297. 16. L. Joao Carlos and R. Laura, Anticancer Agents Med. Chem., 2011, 11, 921. 17. C. K. Mirabelli, R. K. Johnson, D. T. Hill, L. F. Faucette, G. R. Girard, G. Y. Kuo, C. M. Sung and S. T. Crooke, J. Med. Chem., 1986, 29, 218. 18. S. S. Gunatilleke and A. M. Barrios, J. Med. Chem., 2006, 49, 3933. 19. (a) S. J. Berners-Price, C. K. Mirabelli, R. K. Johnson, M. R. Mattern, F. L. McCabe, L. F. Faucette, C. M. Sung, S. M. Mong, P. J. Sadler and S. T. Crooke, Cancer Res., 1986, 46, 5486–5493; (b) P. J. Barnard and S. J. Berners-­Price, Coord. Chem. Rev., 2007, 251, 1889–1902. 20. S. J. Berners-­Price and A. Filipovska, Metallomics, 2011, 3, 863. 21. M. J. McKeage, S. J. Berners-­Price, P. Galettis, R. J. Bowen, W. Brouwer, L. Ding, L. Zhuang and B. C. Baguley, Cancer Chemother. Pharmacol., 2000, 46, 343.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

140

Chapter 5

22. A. S. Humphreys, A. Filipovska, S. J. Berners-­Price, G. A. Koutsantonis, B. W. Skelton and A. H. White, Dalton Trans., 2007, 4943. 23. O. Rackham, S. J. Nichols, P. J. Leedman, S. J. Berners-­Price and A. Filipovska, Biochem. Pharmacol., 2007, 74, 992. 24. N. Pillarsetty, K. K. Katti, T. J. Hoffman, W. A. Volkert, K. V. Katti, H. Kamei and T. Koide, J. Med. Chem., 2003, 46, 1130. 25. S. Tian, F.-­M. Siu, S. C. F. Kui, C.-­N. Lok and C.-­M. Che, Chem. Commun., 2011, 47, 9318. 26. J. Fernández-­Gallardo, B. T. Elie, T. Sadhukha, S. Prabha, M. Sanaú, S. A. Rotenberg, J. W. Ramos and M. Contel, Chem. Sci., 2015, 6, 5269. 27. K. Yan, C.-­N. Lok, K. Bierla and C.-­M. Che, Chem. Commun., 2010, 46, 7691. 28. P. Zhang and P. J. Sadler, J. Organomet. Chem., 2017, 839, 5. 29. L. Oehninger, R. Rubbiani and I. Ott, Dalton Trans., 2013, 42, 3269. 30. T. Zou, C. N. Lok, P. K. Wan, Z. F. Zhang, S. K. Fung and C. M. Che, Curr. Opin. Chem. Biol., 2018, 43, 30. 31. P. J. Barnard, M. V. Baker, S. J. Berners-­Price and D. A. Day, J. Inorg. Biochem., 2004, 98, 1642. 32. M. V. Baker, P. J. Barnard, S. J. Berners-­Price, S. K. Brayshaw, J. L. Hickey, B. W. Skelton and A. H. White, Dalton Trans., 2006, 30, 3708. 33. J. L. Hickey, R. A. Ruhayel, P. J. Barnard, M. V. Baker, S. J. Berners-­Price and A. Filipovska, J. Am. Chem. Soc., 2008, 130, 12570. 34. R. Rubbiani, I. Kitanovic, H. Alborzinia, S. Can, A. Kitanovic, L. A. Onambele, M. Stefanopoulou, Y. Geldmacher, W. S. Sheldrick, G. Wolber, A. Prokop, S. Wölfl and I. Ott, J. Med. Chem., 2010, 53, 8608. 35. A. Pratesi, C. Gabbiani, E. Michelucci, M. Ginanneschi, A. M. Papini, R. Rubbiani, I. Ott and L. Messori, J. Inorg. Biochem., 2014, 136, 161. 36. R. Rubbiani, S. Can, I. Kitanovic, H. Alborzinia, M. Stefanopoulou, M. Kokoschka, S. Mönchgesang, W. S. Sheldrick, S. Wölfl and I. Ott, J. Med. Chem., 2011, 54, 8646. 37. Y. F. Mui, J. Fernández-­Gallardo, B. T. Elie, A. Gubran, I. Maluenda, M. Sanaú, O. Navarro and M. Contel, Organometallics, 2016, 35, 1218. 38. N. Estrada-­Ortiz, F. Guarra, I. A. M. de Graaf, L. Marchetti, M. H. de Jager, G. M. M. Groothuis, C. Gabbiani and A. Casini, ChemMedChem, 2017, 12, 1429. 39. W. Niu, X. Chen, W. Tan and A. S. Veige, Angew. Chem., Int. Ed., 2016, 55, 8889. 40. A. Meyer, L. Oehninger, Y. Geldmacher, H. Alborzinia, S. Wölfl, W. S. Sheldrick and I. Ott, ChemMedChem, 2014, 9, 1794. 41. C. Bazzicalupi, M. Ferraroni, F. Papi, L. Massai, B. Bertrand, L. Messori, P. Gratteri and A. Casini, Angew. Chem., Int. Ed. Engl., 2016, 55, 4256. 42. D. Wragg, A. de Almeida, R. Bonsignore, F. E. Kuhn, S. Leoni and A. Casini, Angew. Chem., Int. Ed. Engl., 2018, 57, 14524–14528. 43. T. Zou, C. T. Lum, C.-­N. Lok, W.-­P. To, K.-­H. Low and C.-­M. Che, Angew. Chem., Int. Ed., 2014, 53, 5810. 44. A. Meyer, C. P. Bagowski, M. Kokoschka, M. Stefanopoulou, H. Alborzinia, S. Can, D. H. Vlecken, W. S. Sheldrick, S. Wölfl and I. Ott, Angew. Chem., Int. Ed., 2012, 51, 8895.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

Anticancer Gold Compounds

141

45. C.-­H. Chui, R. S.-­M. Wong, R. Gambari, G. Y.-­M. Cheng, M. C.-­W. Yuen, K.-­W. Chan, S.-­W. Tong, F.-­Y. Lau, P. B.-­S. Lai, K.-­H. Lam, C.-­L. Ho, C.-­W. Kan, K. S.-­Y. Leung and W.-­Y. Wong, Bioorg. Med. Chem., 2009, 17, 7872. 46. S. J. Berners-­Price and P. J. Sadler, Coord. Chem. Rev., 1996, 151, 1. 47. C.-­M. Che, R. W.-­Y. Sun, W.-­Y. Yu, C.-­B. Ko, N. Zhu and H. Sun, Chem. Commun., 2003, 1718. 48. R. W.-­Y. Sun, C. K.-­L. Li, D.-­L. Ma, J. J. Yan, C.-­N. Lok, C.-­H. Leung, N. Zhu and C.-­M. Che, Chem. -­Eur. J., 2010, 16, 3097. 49. C. T. Lum, A. S.-­T. Wong, M. C. M. Lin, C.-­M. Che and R. W.-­Y. Sun, Chem. Commun., 2013, 49, 4364. 50. C. T. Lum, R. W.-­Y. Sun, T. Zou and C.-­M. Che, Chem. Sci., 2014, 5, 1579. 51. Y. Wang, Q. Y. He, C. M. Che and J. F. Chiu, Proteomics, 2006, 6, 131. 52. Y. Wang, Q.-­Y. He, R. W.-­Y. Sun, C.-­M. Che and J.-­F. Chiu, Cancer Res., 2005, 65, 11553. 53. S. Tu, R. Wai-­Yin Sun, M. C. Lin, J. Tao Cui, B. Zou, Q. Gu, H. F. Kung, C. M. Che and B. C. Wong, Cancer, 2009, 115, 4459. 54. Y. Wang, Q. Y. He, R. W. Sun, C. M. Che and J. F. Chiu, Eur. J. Pharmacol., 2007, 554, 113. 55. D. Hu, Y. Liu, Y. T. Lai, K. C. Tong, Y. M. Fung, C. N. Lok and C. M. Che, Angew. Chem., Int. Ed., 2016, 55, 1387. 56. K. H.-­M. Chow, R. W.-­Y. Sun, J. B. B. Lam, C. K.-­L. Li, A. Xu, D.-­L. Ma, R. Abagyan, Y. Wang and C.-­M. Che, Cancer Res., 2010, 70, 329. 57. M. A. Cinellu, L. Maiore, M. Manassero, A. Casini, M. Arca, H. H. Fiebig, G. Kelter, E. Michelucci, G. Pieraccini, C. Gabbiani and L. Messori, ACS Med. Chem. Lett., 2010, 1, 336. 58. C. Gabbiani, A. Casini, L. Messori, A. Guerri, M. A. Cinellu, G. Mlnghetti, M. Corsini, C. Rosani, P. Zanello and M. Arca, Inorg. Chem., 2008, 47, 2368. 59. A. Casini, M. A. Cinellu, G. Minghetti, C. Gabbiani, M. Coronnello, E. Mini and L. Messori, J. Med. Chem., 2006, 49, 5524. 60. V. Milacic, D. Fregona and Q. P. Dou, Histol. Histopathol., 2008, 23, 101. 61. C. Nardon, S. M. Schmitt, H. Yang, J. Zuo, D. Fregona and Q. P. Dou, PLoS One, 2014, 9, e84248. 62. C. Nardon, F. Chiara, L. Brustolin, A. Gambalunga, F. Ciscato, A. Rasola, A. Trevisan and D. Fregona, ChemistryOpen, 2015, 4, 183. 63. R. W. Sun, M. Zhang, D. Li, M. Li and A. S. Wong, J. Inorg. Biochem., 2016, 163, 1. 64. A. P. Martins, A. Ciancetta, A. de Almeida, A. Marrone, N. Re, G. Soveral and A. Casini, ChemMedChem, 2013, 8, 1086. 65. A. Serna, A. Galán-­Cobo, C. Rodrigues, I. Sánchez-­Gomar, J. J. Toledo-­ Aral, T. F. Moura, A. Casini, G. Soveral and M. Echevarría, J. Cell. Physiol., 2014, 229, 1787. 66. A. P. Martins, A. Marrone, A. Ciancetta, A. Galan Cobo, M. Echevarria, T. F. Moura, N. Re, A. Casini and G. Soveral, PLoS One, 2012, 7, e37435. 67. B. Aikman, A. de Almeida, S. M. Meier-­Menches and A. Casini, Metallomics, 2018, 10, 696.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00120

142

Chapter 5

68. R. D. Teo, H. B. Gray, P. Lim, J. Termini, E. Domeshek and Z. Gross, Chem. Commun., 2014, 50, 13789. 69. K. J. Akerman, A. M. Fagenson, V. Cyril, M. Taylor, M. T. Muller, M. P. Akerman and O. Q. Munro, J. Am. Chem. Soc., 2014, 136, 5670. 70. C. K.-­L. Li, R. W.-­Y. Sun, S. C. F. Kui, N. Y. Zhu and C.-­M. Che, Chem. -­Eur. J., 2006, 12, 5253. 71. R. W.-­Y. Sun, C.-­N. Lok, T. T.-­H. Fong, C. K.-­L. Li, Z. F. Yang, T. Zou, A. F.-­M. Siu and C.-­M. Che, Chem. Sci., 2013, 4, 1979. 72. J. J. Yan, A. L.-­F. Chow, C.-­H. Leung, R. W.-­Y. Sun, D.-­L. Ma and C.-­M. Che, Chem. Commun., 2010, 46, 3893. 73. S. K. Fung, T. Zou, B. Cao, P.-­Y. Lee, Y. M. E. Fung, D. Hu, C.-­N. Lok and C.-­M. Che, Angew. Chem., Int. Ed., 2017, 56, 3892. 74. T. Zou, C. T. Lum, S. S.-­Y. Chui and C.-­M. Che, Angew. Chem., Int. Ed., 2013, 52, 2930. 75. L. S. Hollis and S. J. Lippard, J. Am. Chem. Soc., 1983, 105, 4293. 76. P. F. Shi, Q. Jiang, Y. M. Zhao, Y. M. Zhang, J. Lin, L. P. Lin, J. Ding and Z. J. Guo, J. Biol. Inorg. Chem., 2006, 11, 745. 77. T. Yang, C. Tu, J. Y. Zhang, L. P. Lin, X. M. Zhang, Q. Liu, J. Ding, Q. Xu and Z. J. Guo, Dalton Trans., 2003, 3419. 78. M. P. Rigobello, L. Messori, G. Marcon, M. A. Cinellu, M. Bragadin, A. Folda, G. Scutari and A. Bindoli, J. Inorg. Biochem., 2004, 98, 1634. 79. S. M. Meier, C. Gerner, B. K. Keppler, M. A. Cinellu and A. Casini, Inorg. Chem., 2016, 55, 4248. 80. J.-­J. Zhang, R. W.-­Y. Sun and C.-­M. Che, Chem. Commun., 2012, 48, 3388. 81. J.-­J. Zhang, K.-­M. Ng, C.-­N. Lok, R. W.-­Y. Sun and C.-­M. Che, Chem. Commun., 2013, 49, 5153. 82. B. Bertrand, S. Spreckelmeyer, E. Bodio, F. Cocco, M. Picquet, P. Richard, P. Le Gendre, C. Orvig, M. A. Cinellu and A. Casini, Dalton Trans., 2015, 44, 11911. 83. M. N. Wenzel, S. M. Meier-­Menches, T. L. Williams, E. Ramisch, G. Barone and A. Casini, Chem. Commun., 2018, 54, 611. 84. N. P. E. Barry and P. J. Sadler, ACS Nano, 2013, 7, 5654. 85. J. J. Yan, R. W. Sun, P. Wu, M. C. Lin, A. S. Chan and C. M. Che, Dalton Trans., 2010, 39, 7700. 86. L. He, T. Chen, Y. You, H. Hu, W. Zheng, W.-­L. Kwong, T. Zou and C.-­M. Che, Angew. Chem., Int. Ed., 2014, 53, 12532. 87. P. Lee, R. Zhang, V. Li, X. Liu, R. W. Sun, C. M. Che and K. K. Wong, Int. J. Nanomed., 2012, 7, 731. 88. P. Lee, Y. Zhu, J. J. Yan, R. W. Y. Sun, W. Hao, X. Liu, C.-­M. Che and K. K. Y. Wong, Nanotechnol., Sci. Appl., 2010, 3, 23. 89. C. Y.-­S. Chung, S.-­K. Fung, K.-­C. Tong, P.-­K. Wan, C.-­N. Lok, Y. Huang, T. Chen and C.-­M. Che, Chem. Sci., 2017, 8, 1942. 90. J. J. Zhang, W. Lu, R. W. Sun and C. M. Che, Angew. Chem., Int. Ed. Engl., 2012, 51, 4882.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Chapter 6

Heterometallic Complexes as Anticancer Agents NATALIA Curadoa and MARIA Contel*a-e a

Department of Chemistry, Brooklyn College, The City University of New York, Brooklyn, NY, 11210, USA; bChemistry PhD Program, The Graduate Center, The City University of New York, 365 Fifth Avenue, New York, NY, 10016, USA; cBiochemistry PhD Program, The Graduate Center, The City University of New York, 365 Fifth Avenue, New York, NY, 10016, USA; dBiology PhD Program, The Graduate Center, The City University of New York, 365 Fifth Avenue, New York, NY, 10016, USA; eCancer Biology Program, University of Hawaii Cancer Center, University of Hawaii at Manoa, Honolulu, USA *E-­mail: [email protected]

6.1  Introduction Heterometallic compounds can be designed to harness chemotherapeutic traits of distinct metals into a single molecule. This chapter will focus solely on heterometallic compounds as anticancer agents reported until April 2018.1–83 To the best of our knowledge there are no reviews on this topic besides a 2018 review from Gimeno et al. on heterobimetallic complexes for theranostic applications.2 The hypothesis (first described by A. Casini and co-­workers)1 is that the incorporation of two different biologically active metals in the same molecule may improve their antitumor activity as a result of metal-­specific interactions with distinct biological targets (cooperative effect) or by the improved chemicophysical properties of the resulting heterometallic compound (synergism). Approximately 57 articles have been published and one US patent   Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

143

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

144

Chapter 6

issued in this topic since the seminal work from Casini et al. in 2010.1–3,8,9,17– 20,24,25,27–29,31–33,35–40,42–63,65–67,69–83 We also include in this chapter articles and one US patent on the combination of ferrocene-­derived,4–7,11–16,21–23,26,30,34,41 ruthenium-­derived64,68 and copper-­derived82 molecules with a second metal that were published prior to 2010 (and that, in most cases, did not per se focus on the hypothesis of the dual effect of the two metals). There are, however, fewer articles of compounds documenting anticancer activity comparisons of the heteronuclear compounds with respect to the monometallic fragments (alone or in combination)1,15,29,42–44,46–48,50–54,58,62–67,70,79–81 or describing more detailed mechanistic22,23,42–44,47,48,50–54,63,65,66,72,79,80,83 and/or in vivo21,22,53,54,83 studies. We have divided this chapter into four different sections with the first three sections based on families of compounds containing metallic fragments already known as being biologically active against cancer (ferrocenyl, p-­ cymene ruthenium and titanocenes) and one section on ‘other heterometallic compounds’ that covers molecules based on diverse biologically active metals (mostly gold, platinum, ruthenium, rhenium, and gadolinium). In some cases, we placed a particular compound in one specific category despite overlap with another, if we thought it was more relevant to its properties or it allowed for a better classification. We only mention examples of compounds with cytotoxicities and/or selectivities that are an improvement with respect to the currently used metallodrug cisplatin. We focus mainly on examples describing mechanistic studies and those providing data on in vivo efficacy trials.

6.2  H  eterometallic Compounds Containing Ferrocenyl-­derived Molecules It has been demonstrated that introduction of the organometallic ferrocene motif into molecules improves their anticancer properties. This is mainly attributed to the low toxicity, stability in aqueous and aerobic media, high lipophilicity, ease of functionalization and unique electrochemical behaviour of ferrocene.84 While ferrocene itself is not cytotoxic, its ferrocenium cation (one-­electron oxidized product) is known to show cytotoxicity.84 The anticancer activity attributed to the ferrocene motif comes partly from the possibility of forming singlet oxygen (Fenton chemistry).85 The elegant work of Jaouen and co-­workers on the synthesis of ferrocifens (tamoxifen-­like drugs) is a great example in this area.86 Ferrocene has also been introduced as a second metallic fragment to generate heterometallic complexes with potential anticancer properties.4–41 Thus, ferrocene has been functionalized to provide ferrocenes containing: (a) amines4–9 and pyrazoles,10 (b) phosphanes11–25 or phosphane derivatives,26–28 including iminophosphoranes,29 (c) N-­heterocyclic carbenes,30,31 (d) vinyl groups,32 (e) diketonato groups,33,34 (f) hydrazone Schiff-­ base complexes,35 (f) biological compounds (such as amino acid esters36,37 or guanidines38), and even (h) first-­and second-­generation dendrimer ligands.39,40 Ferrocene has also been directly coordinated to a metallic ruthenium(ii) arene fragment.41 The derivatization of ferrocene allowed for the presence of a functional group able to coordinate other metals (platinum(ii),5,9,10,16–20,26,27

View Online

Heterometallic Complexes as Anticancer Agents

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

5,21–23,27,28

12–14,16,17,19,24,28,30,32,36

145 29,36

palladium(ii), gold(i) and gold(iii), ruthenium(ii)6,7,25,32,39,40 and ruthenium(iii),15 silver(i),16,24 copper(i)11,16,35 and copper(ii),8 rhodium(i)5,33,34 and rhodium(iii)40 and iridium(i)5 and iridium(iii)).40 The variety of compounds reported is very large. For two-­thirds of the reports the cytotoxicity found was worse or just similar to that of known drugs such as cisplatin. Chart 6.1 depicts examples of compounds that were more active (and/or selective) than cisplatin. For most of these active compounds, the biological effects seem to come mostly from the second metal, but in selected cases there is a clear correlation between the ferrocene-­based fragment and an improvement of cytotoxicity as we describe next. Compound 1 is an example of a ferrocene-­containing terpyridyl copper(ii) phenanthroline complex for which the incorporation of the ferrocene fragment improves its photocytotoxicity with respect to the non-­ferrocene-­containing (phenyl) complex.8 The synergistic effect for this compound is due to the Fc+– Fc and Cu(ii)–Cu(i) redox couples, which are able to form reactive hydroxyl radicals in the presence of both oxidizing and reducing reagents. The radicals cause considerable DNA cleavage activity and enhanced cellular activity of 1 in

Chart 6.1  

View Online

Chapter 6

146

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

8

HeLa cells with respect to the phenyl analogue. The trimetallic platinum(ii) compound 2, coordinated to two secondary amines containing ferrocene fragments, is another example, for which the electrochemical behaviour of the ferrocene units greatly influences its anticancer activity.9 For this compound, while there is a lack of electronic communication between the two units, there is a single redox process with oxidation of the Fe centers. This compound is more cytotoxic than cisplatin in a colon cancer cell line (WiDr). Cell cycle studies indicated that the mode of action of 2 is quite different from that of cisplatin, with a possible role for the ferrocene moiety.9 Cycloplatinated derivative 3, containing a ferrocenyl-­pyrazole, was more active than cisplatin in the colon cancer cell line HCT116. However, the activity seems to be more related to the well-­known effects of cycloplatinated species.87 A number of metallic compounds containing phosphanes incorporating ferrocene units (mostly diphenylphosphinoferrocene (MPPF) and 1,1′-­bis-­diphe-­ nylphosphinoferrocene (DPPF)) have been described.11–25 In most cases, while part of the effects found could be attributed to the ferrocene units, the cytotoxicity was similar or lower than that of cisplatin. Two exceptions are the ruthenium(iii) coordination compound 4 and the cyclopalladated compound 5. The ruthenium derivative 4 was found to have cytotoxicity to the triple negative breast cancer cell line MDA-­MB-­231 of ca. 10 µM, which is much lower than that of cisplatin, the phosphane DPPF and the starting material [RuCl3NO·2H2O].15 The chiral cyclopalladated compound 5 derived from N,N-­dimethyl-­1-­ phenethylamine with a bridging DPPF ligand (also known as BCP: biphosphinic palladacycle complex) has been studied in vitro and in vivo.21–23 This compound showed cytotoxicity in leukemia cell lines (HL60, Jurkat and K562) but not in normal human lymphocytes. The compound is apoptotic and induces the inhibition of caspase-­3 and -­6. Compound 5 induces chromatin condensation, apoptotic bodies and DNA fragmentation.22,23 It also produces lysosomal-­membrane permeabilization and intralysosomal accumulation. The lysosomal pathway suggested was new at the time and involved cathepsin B as a death mediator.22,23 Importantly, 5 did not produce lesions in liver or kidney 14 days after drug administration (100 mg kg−1, i.p.). White and red blood cells of 5-­treated mice presented normal morphological features.22 In a different study on Walker tumor-­bearing rats,21 a 90% growth inhibition of solid tumors (one dose of 2.0 mg kg−1) was observed. The treatment in rats with 5 (100 mg kg−1, i.p.) for 14 days did not produce alterations in white or red blood cells or hepatic, kidney and spleen tissues as observed in mice.21 While synergistic studies were not carried out, it was suggested that the ferrocene motifs had an effect on the phosphane coordinated to the cyclopalladated fragments.21 Gold(i) thiolates containing non-­cytotoxic ferrocenyl-­amide phosphanes, such as 6 in Chart 6.1, have good antiproliferative activity in human cell lines A549 and Hep-­G2.24 Similarly, cationic gold(i) compounds with two N-­ heterocyclic carbene (NHC) ligands incorporating ferrocenyl were found to be active in Jurkat and MCF-­7 cell lines while being more selective to normal cells than cisplatin.30 While not mentioned in these articles, it seems that the biological activity is mostly related to the well-­known effects of gold(i) compounds containing lipophilic phosphanes and N-­heterocyclic carbene ligands.88,89

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

147

Compounds like 7, with two or four ferrocenyl motifs coordinated to the N atoms (NHC), have been described more recently.31 These compounds were designed to alter the redox environment within cancer cells. It was hypothesized that a hybrid compound able to inhibit thioredoxin reductase (gold part) and inducing non-­specific reactive oxygen species (ROS) (ferrocene part) could override the ROS regulatory pathway in tumor models.31 These compounds were indeed effectively able to regulate ROS via multiple mechanisms in A549 lung cancer cells, improving their cytotoxicity considerably with respect to Auranofin (an old anti-­rheumatic drug currently being ‘repurposed’ as a chemotherapeutic for different diseases, including cancer) and cisplatin. From RNA microarray gene expression studies it was revealed that 7 was able to induce endoplasmic reticulum stress response pathways as a result of ROS increase.31 Phosphanes like MPPF and DPPF can also be functionalized to generate iminophosphorane (IM) ligands (containing a N=P bond).29 Bimetallic and trimetallic coordination compounds of gold(iii) and palladium(ii) containing these IM ligands have been reported (8 and 9 in Chart 6.1). In this case a synergistic effect was demonstrated as the compounds were much more cytotoxic than the ligands themselves and gold or palladium compounds containing analogous IM ligands based on PPh3. Trimetallic compounds 8 and 9 (more stable than the mononuclear and dinuclear precursors) were significantly more cytotoxic than cisplatin in the resistant A2780R and MCF7 cell lines, supporting the idea of a different mechanism of action. Preliminary mechanistic studies indicated that neutral palladium(ii) compounds (like 9) showed strong interactions with DNA, while cationic gold(iii) derivatives (like 8) were ineffective. However, the gold(iii) compounds were good inhibitors of the zinc-­finger protein PARP-­1, a possible target enzyme for anticancer metal compounds.29 This study supported the idea that while the incorporation of ferrocene units may modulate and improve the anticancer properties of the resulting heterometallic complexes, the mechanisms are in most cases more influenced by the second metal incorporated. Ferrocene has also been functionalized to incorporate β-­diketonato ligands able to coordinate metals such as rhodium(i) (like 10 in Chart 6.1).33,34 For this compound there is an easy oxidation of the rhodium(i) core and the ferrocenyl unit. Compound 10 is more cytotoxic than the free ferrocenyl ligand. This compound is not as cytotoxic in HeLa as cisplatin (one order of magnitude less responsive)34 but has cytotoxicity similar to that of cisplatin for prostate tumor 1542 cells, while being more selective toward prostate normal cells.33 In addition, the cell pathway is very different from that of cisplatin, giving rise to necrosis and abnormal morphology.33 Catalytically generated ferrocene-­containing guanidines can be coordinated to platinum(ii) centers to generate complexes like 11 in Chart 6.1 that are active against a variety of human cancer cell lines (GI50 values in the range 1.4–2.6 µM) and more cytotoxic than cisplatin in the resistant T-­47D and WiDR cell lines.38 The direct attachment of the guanidine group to the ferrocene moiety allows for a very easy one-­electron oxidation making this type of ligand attractive for the synthesis of heteronuclear complexes with anticancer properties. Ferrocene has also been introduced in supramolecular entities containing a second metal. First-­ and second-­generation heterometallic dendrimers

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

148

Chapter 6

Chart 6.2   containing ferrocenyl and a second metallic motif (ruthenium(ii)–p-­cymene, ruthenium(ii)–hexamethylbenzene (η6-­C6Me6), rhodium(iii) or iridium(iii)– cyclopentadienyl) have been synthesized39,40 (Chart 6.2 shows the most cytotoxic compounds that contain ruthenium(ii)–arene fragments). Compounds like 12 and 13 slowed the growth of both A2780 cisplatin and A2780 cisplatin-­resistant ovarian cancer cell lines by more than 50% (IC50 < 5 µM). The electrochemical studies performed revealed that these types of complex result in two irreversible redox processes (oxidation of the iron(ii) and ruthenium(ii) centers). The second-­generation ruthenium(ii)–η6-­C6Me6 metallodendrimer (15) was the most cytotoxic with respect to A2780 and especially A2780 cisplatin-­ resistant ovarian cancer cell lines, and more selective than cisplatin.40 DNA-­ binding experiments of 14 and 15 revealed that the mode of action involved non-­covalent interactions with DNA.40 The authors claimed that the presence of ferrocene in these metallodendrimers improved the cytotoxicity with respect to previously reported non-­ferrocenyl analogues.39 As indicated earlier, the cytotoxicity exerted and the mode of action for most of the heterometallic compounds described seems to come mainly from the second metallic fragment. Ferrocene acts a modulator that benefits the heterometallic molecule due to its redox behaviour (as demonstrated in some cases). However, the lack of more detailed mechanistic studies prevents making broader generalizations. One problem with a number of these heterometallic compounds is their lack of stability in physiological media, which may have deterred their development as anticancer agents.

6.3  H  eterometallic Compounds Containing Ruthenium(ii)–Arene Fragments Ruthenium(ii)–arene complexes, especially those based on the arene p-­ cymene, have emerged as potential anticancer agents90–92 due to two main relevant properties: (a) the amphiphilic properties of the ruthenium–arene

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

149

system (hydrophobic arene versus hydrophilic ruthenium metal fragment), and (b) the ease of synthetic modification of the arene moiety, which may facilitate targeted chemotherapy.90 While some compounds are active against primary tumors, others are only active against metastases.90–92 In this section we discuss the introduction of a second metallic fragment to ruthenium(ii)– arene systems.

6.3.1  Ruthenium–Gold Compounds In 2015 our group in Brooklyn College (in collaboration with the group of Messori in Florence) reported on the first heterometallic compounds (besides derivatized ferrocenes) containing a p-­cymene ruthenium(ii) motif (Chart 6.3).42 Neutral compounds like 17 and 18, incorporating a [Ru(p-­cymene) Cl2(dppm-­kP)] fragment (16, R) with cytotoxic and potential antimetastatic properties and a gold(i) metallic moiety (containing either a chloride or a thiolate), were synthesized and found to have relevant in vitro properties against the HCT116 colon cancer cell line.42 The new compounds manifested a more favourable in vitro pharmacological profile toward cancer cells than individual ruthenium and gold species, being either more cytotoxic or more selective. Studies of the interactions of the compounds with plasmid DNA (pBR322) indicated that these bimetallic compounds act through a mechanism where nucleic acids are not the only, or primary, targets. Compound 17 inhibited purified cathepsin B in the micromolar range.42 As an improvement, our group developed second-­generation cationic compounds

Chart 6.3  

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

150

Chapter 6

like 19–22, incorporating gold(i)–N-­heterocyclic (NHC) carbene ligands.43 Cationic gold NHC complexes display strong antimitochondrial and thioredoxin reductase effects.88,89 Indeed, the combination of the ruthenium monometallic compound 16 (R) with fragments [Au(NHC)]+ gave rise to cationic Ru–Au compounds that displayed very high stability under physiological conditions, notable cytotoxicity in renal and colon cancer cell lines, and cytotoxic selectivity. We were able to demonstrate the synergistic effect of the combination of the two metallic entities.43 Preliminary mechanistic studies indicated that these compounds behaved unlike cisplatin and more like other Au(i) derivatives containing phosphane (such as Auranofin) and NHC ligands.88,89 The compounds did not interact with DNA but inhibited mitochondrial thioredoxin reductase in a human clear cell renal carcinoma cell line (Caki-­1).43 More recently, we have reported on the mechanism in vitro in the Caki-­1 cell line of a selected compound of this family of cationic derivatives, [Ru(p-­ cymene)Cl2(η1-­dppm)Au(IMes)]ClO4 (22, RANCE-­1).44 For assessing synergistic effects, we studied the monometallic Ru(ii) 16 (R) and Au(i) 23 (ANCE-­1) compounds and made comparisons with the well-­known Auranofin. In addition to being cytotoxic and apoptotic, RANCE-­1 (22) displays relevant inhibition of migration (82%), invasion (66%) and angiogenesis, while also inhibiting molecular pathways associated with these processes.44 The molecular targets inhibited include different interleukins, metalloproteases and cathepsins, which are involved in tumor metastasis and angiogenesis, and to an even higher degree the angiogenic vascular endothelial growth factor (VEGF). It is noteworthy that the inhibition observed is in general better than that of the individual monometallic fragments present in the heterometallic compound. In some cases, this inhibition can be correlated with a particular metallic fragment, either gold or ruthenium, of the bimetallic compound (reinforcing the idea of the positive synergistic effect caused by the two distinct metals). During this study we found that Auranofin had a very similar effect to RANCE-­1 (22) in the renal cancer cell line Caki-­1 in terms of its antiproliferative and antimetastatic properties and some of the sets of targets inhibited. RANCE-­1 (22), however, was a better VEGF inhibitor than Auranofin and a much better pan-­MMP and pan-­cathepsin inhibitor. Moreover, RANCE-­1 (22) blocked all angiogenic formation, while Auranofin merely induced branching disturbances of de novo angiogenesis in an in vitro model (Figure 6.1). Due to the excellent results in vitro, the bimetallic Ru–Au compound 22 (RANCE-­1) (a potential cytotoxic, antimetastatic and antiangiogenic agent) was selected for an in vivo study on the biological response and organ distribution in NOD.CB17-­Prkdc SCID/J (n = 6 mice total (males n = 3 and females n = 3)) bearing subcutaneous clear cell renal carcinoma (Caki-­1).45 We found that this compound precludes all tumor growth (tumor burden increase not significant) in a 21 day efficacy trial with doses of 22 every 72 hours (10 mg kg−1). Very importantly, pharmacokinetic and histopathology

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

151

Figure 6.1  Induction  of endothelial cell reorganization into 3D vessel structures by

bimetallic Ru–Au 22 (RANCE-­1), monometallic Ru 16 (R), Au 23 (ANCE-­ 1), and Auranofin. Human umbilical vein endothelial cells (HUVEC) were seeded in plates coated with Geltrex® matrix using complete HUVEC media and incubated at 37 °C and 5% CO2. At post-­seeding IC10 concentrations of RANCE-­1, R, ANCE-­1, or Auranofin, 0.1% DMSO was added. (Left) Representative phase-­contrast images captured 24 h after the compounds were added. (Right) Quantitation of tube formation; TN: tube nodes, TL: tube length. Reproduced from ref. 44 with permission from Springer Nature, Copyright 2018.

studies indicated a remarkable lack of systemic toxicity for this bimetallic compound. Mechanistically, for the most part the in vivo data can be reconciled with in vitro data already reported.45 The groups of Casini and Bodio also reported on cationic ruthenium(ii)–p-­ cymene derivatives containing gold(i)–NHC fragments like 24 and 25 (Chart 6.3).46 In this case the compounds were much less cytotoxic on ovarian carcinoma (A2780, SKOV-­3) and human lung cancer cells (A549),46 indicating that the nature of both starting p-­cymene ruthenium(ii) and gold(i) species is crucial in developing efficacious Ru–Au derivatives.

6.3.2  Ruthenium–Platinum Compounds Bimetallic compounds containing the p-­cymeme ruthenium(ii) motif and a platinum(iv) moiety have been developed by the group of Zhu and co-­workers.47,48 Platinum(iv) compounds are known to be reduced preferentially inside cancer cells, while being more inert kinetically toward biomolecules than platinum(ii) derivatives.87 The bimetallic compounds named ruthplatins (26–29 in Chart 6.4) are water soluble and stable after hydration for up to 4 days.47 They display sub-­micromolar and nanomolar cytotoxicity in most of the human cancer cell lines tested, including cisplatin-­resistant ovarian cancer A2780cisR and lung cancer A549cisR (8-­ to 107-­fold more cytotoxic than cisplatin) while having better selectivity.47 Bimetallic Ru–Pt compound 26 is also more effective than cisplatin in killing breast cancer

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

152

Chapter 6

Chart 6.4  

Chart 6.5   MCF-­7 spheroids.47 Mechanistic studies revealed a synergistic effect of the bimetallic compounds and that compound 26 induces larger fractions of both apoptotic and necrotic cells than cisplatin (higher suppression of DNA replication and a robust S phase for cell cycle). The potential antimetastatic properties of 26 were demonstrated in triple negative breast cancer MDA-­ MB-­231 and lung cancer A549 cell lines (2D scratch assay) with an inhibition of migration of 54% after 24 hours.47 More recently,48 this group has also reported a novel bimetallic Ru–Pt compound (30) that is also stable and soluble in water with a much better selectivity profile than cisplatin (while having similar cytotoxicity) and that prevented invasion in the triple negative breast cancer cell line MDA-­MB-­231 (3D trans-­well assay). Importantly, 30 displayed lower toxicity than cisplatin in developing zebrafish embryos.48

6.3.3  Ruthenium–Cobalt Compounds A cobaltocene derivative (η5-­Cp)Co[C4-­trans-­Ph2(4-­Py)2] (L) has been used to generate three cobalt–ruthenium heterometallic molecular rectangles (31– 33 in Chart 6.5) by coordination-­driven self-­assembly of the cobalt derivative L and one of three dinuclear ruthenium precursors containing the p-­cymene Ru(ii) motif (A1–A3).49

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

153

The autophagic activities and cell death ratios of the heterometallic compounds were increased with respect to the cobaltocene derivative. Compounds 31–33 induced gastric cancer cell death by modifying autophagy and apoptosis. While the synergistic effect was not demonstrated, it was shown that incorporation of the Ru dinuclear complexes increased cytotoxicity dramatically with respect to the cobaltocene L.49 From the results described in this section we conclude that highly stable ruthenium(ii)–platinum(iv) and ruthenium(ii)–gold(i) compounds display synergistic biological effects and modes of action due to the combination of both metallic fragments. These bimetallic complexes are very promising candidates for further preclinical evaluation in metastatic cancer models.

6.4  H  eterometallic Compounds Containing Titanocenes Titanocene dichloride (TDC, [(η5-­C5H5)2TiCl2]) was the first organometallic complex to enter clinical trials in 1993.92 This compound exhibited considerable antitumor activity in both in vitro and in vivo experimental models.89–92 However, phase II clinical trials with patients with metastatic renal cell carcinoma or metastatic breast cancer showed very low efficacy.89–92 Derivatization of titanocene has afforded monometallic compounds with relevant in vivo properties in renal cancer.89–92 In this section, we will describe heterometallic compounds with the titanocene motif.

6.4.1  Titanocene–Ruthenium Compounds In 2010 Casini and co-­workers reported on the first heterometallic compounds based on titanocene dichloride (TDC) and the Ru–p-­cymene derivative RAPTA-­C (a non-­toxic and highly selective compound with promising antimetastatic in vivo activity).1 They described a series of bimetallic complexes with phosphanes (differences in the length of the linker and the substituents at the phosphorous atoms), both neutral (like 34 and 35 in Chart 6.6) and cationic. These compounds displayed relevant cytotoxicity on human ovarian cancer cells (A2780 and A2780cisR). Bimetallic Ti–Ru compounds were markedly more active than their Ti or Ru monometallic analogues (TDC and RAPTA-­C). The compounds inhibited cathepsin B (with a correlation with inhibition and the length of the alkyl chain between the metal centres). In addition, ESI-­MS provided evidence for binding of a ruthenium(ii) fragment to proteins.1

6.4.2  Titanocene–Gold Compounds A number of titanocene-­containing gold(i) fragments (either with phosphanes or N-­heterocyclic carbenes) have been described since 2011 by the groups of Casini50 and our group at Brooklyn College51–57 (Chart 6.6).

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

154

Chapter 6

Chart 6.6   Cyclopentadienyl rings in TDC were modified, with the inclusion of phosphane fragments able to coordinate gold(i) fragments.50,51 Chart 6.6 shows examples of the resulting compounds that were more active against ovarian50 or ovarian and prostate51 cancer cell lines, namely, cationic 36 50 and neutral 37.51 For these types of compounds (termed as zero generation in Chart 6.6), a synergistic effect due to the presence of both metallic fragments on the same molecule was found.50,51 It was demonstrated that the compounds did not interact with plasmid (pBR322) DNA50 and that the interaction with Calf Thymus (CT) DNA was electrostatic in nature, indicating a mode of action different from cisplatin and more related to gold(i) fragments.51 While the compounds were less acidic than titanocene dichloride and more stable than titanocene dichloride at physiological pH,51 the stability of the complexes was still far from optimal. It has been proposed that TDC breaks down at neutral pH due to hydrolysis generating titanium oxo species and cyclopentadiene. Zero generation Ti–Au compounds, like 36 and 37, could therefore potentially break into monometallic species in physiological media, or in vivo, before reaching the tumors, defeating the purpose of the single molecule, multimetallic approach. In order to generate more stable compounds, our group envisioned that attaching the gold fragment via a different linker that could coordinate more strongly to titanium (e.g. carboxylate groups) would generate more stable complexes.52 Ti–O bonds are known to be considerably stronger (ΔHf298 = 662(16) kJ mol−1) than Ti–C (ΔHf298 = 439 kJ mol−1) or Ti–Cl (ΔHf298 = 494 kJ mol−1). Compounds like 38 (first-­generation Ti–Au) were prepared and characterized crystallographically. They were more stable than TDC and the zero

View Online

Heterometallic Complexes as Anticancer Agents

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

52

155

generation compounds. Compound 38 was found to be markedly more cytotoxic in renal cancer cell lines (nanomolar range) than cisplatin or the titanocene benchmark compound for renal cancer, Titanocene Y. Compound 38 was found to be apoptotic and not to interact with plasmid (pBR322) DNA. Preliminary mechanistic studies in Caki-­1 renal cancer cells, coupled with studies of their inhibitory properties on a panel of 35 kinases of oncological interest, indicate that these compounds inhibit protein kinases of the AKT and MAPKAPK families with a higher selectivity toward MAPKAPK3 (nanomolar range).52 The first-­generation compounds were found to be selective in vitro and have a very favourable preliminary toxicity profile on C57black6 mice. In order to improve the pharmacological profile of the Ti–Au derivatives by incorporation of gold(i) thiolates containing phosphanes (like Auranofin), a bifunctional ligand (derived from 4-­mercaptobenzoic acid), including both carboxylate and thiolate groups, was chosen.53,54 Monometallic gold(i) compounds with this linker (like 39 53,54 or the more recently prepared 41,55 containing the AuPEt3 fragment) can be coordinated to the titanocene dimethyl, generating even more stable second-­generation bimetallic compounds like 40 (Titanocref)53,54 and 42 55 (Titanofin) (Chart 6.6). Bimetallic compounds 40 (Titanocref)53,54 and 42 55 (Titanofin) are more cytotoxic than the gold monometallic derivatives (39 and 41) and significantly more cytotoxic than titanocene dichloride, while being quite selective.52–55 We have performed some preliminary mechanistic studies with Titanocref (40) and found that this compound does not interact with plasmid (pbR322) DNA but that it inhibits thioredoxin reductase and protein kinases that drive cell migration (such as AKT, p90-­RSK and MAPKAPK3).53,54 We reported on the co-­localization of both titanium and gold metals (1 : 1 ratio) in Caki-­1 renal cancer cells for Titanocref (40) which was a further demonstration of the robustness of these second-­generation bimetallic compounds in vitro (Figure 6.2A).53,54 More recently,55 we have found that Titanocref (40) and Titanofin (42) inhibit migration, invasion, and angiogenic assembly along with molecular markers associated with these processes such as prometastatic IL(s), MMP(s), TNF-­α, and proangiogenic VEGF, FGF-­basic. The bimetallic compounds also strongly inhibit the mitochondrial protein TrxR (thioredoxin reductase), often overexpressed in cancer cells evading apoptosis, and also inhibits FOXC2, PECAM-­1, and HIF-­1α, whose overexpression is linked to resistance to genotoxic chemotherapy.55 In all these studies Auranofin was used as control and found to have behavior similar (not identical) to the bimetallic complexes but a different cell cycle. Auranofin induced cell arrest at G1/G0, while Titanocref (40) and Titanofin (42) induced arrest at G2/M.55 Importantly, studies in vivo for first-­generation 38 and second-­generation 40 (Caki-­1 xenografts in NOD.CB17-­Prdkdc SCID/J mice) showed that, while Titanocref (40) shrank tumors by 67% after 28 days of treatment (3 mg per

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

156

Chapter 6

Figure 6.2  (A)  Cellular uptake of compound 40 in CAKI-­1 cells. The concen-

tration of compound 5 calculated based on Ti and Au content in the cell lysates is similar, suggesting that the compound is robust and both the elements are co-­localized in the cells. (B) Percentage reduction in tumor burden in a cohort of 12 female NOD.CB17-­ Prkdc SCID/J mice inoculated subcutaneously with 8 × 106 Caki-­1 cells. The treatment started when tumors were palpable (6 mm diameter). Six mice were treated with compound 40 (red bars), and 6 were treated with the vehicle, 100 µL normal saline (0.9% NaCl) (black bars). Compound 40 was administered as 3 mg kg−1 every 48 h. Reproduced from ref. 53 with permission from The Royal Society of Chemistry.

kg every 48 hours, Figure 6.2B), first-­generation heterometallic compound 38 was non-­inhibitory, stressing the relevance of the structural modifications on the linker scaffold.53,54 While these compounds had a favourable preliminary toxicity profile, the pharmacokinetic data were not very accurate.53,54 We hypothesized that this was due to the frequency of the dosage. More recently, we have performed more detailed in vivo efficacy trials56 for bimetallic compounds 40 (Titanocref) and 42 (Titanofin) in the same mice model but this time sex matched (three male and three female per compound). The design of the efficacy in vivo trial was performed after pharmacokinetic analyses and it was found that the optimal dosage was every 72 hours (5 mg per kg for 40 and 10 mg per kg for 42). Both compounds showed an impressive tumor size reduction of 51% (Titanocref, 40) and 60% (Titanofin, 42) after 21 days of treatment. Detailed histopathological studies showed a remarkable lack of systemic toxicity for these compounds.56 As in the case of the Ru–Au derivative (RANCE-­1, 22) described in Section 6.3.1, mechanistic studies performed on tissues and tumor point to a correlation between the effects observed in vitro with those found in vivo (for the most part).56 With the aim to improve the pharmacological profile of the bimetallic Ti– Au second-­generation compounds even further, we exchanged the gold(i) phosphane with an N-­heterocyclic carbene ligand (Au–NHC compounds are usually more stable and display strong antimitochondrial effects) to

View Online

Heterometallic Complexes as Anticancer Agents

157

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

57

obtain third-­generation complexes (Chart 6.6). Compounds 44a–d (with four different NHC ligands) were, however, found not to be so active for the renal cancer cell line Caki-­1. The highest activity and selectivity, and a synergistic effect of the resulting bimetallic compounds 44a–d, were found for prostate and colon cancer cell lines.57 Preliminary mechanistic studies showed a very similar mode of action to the second-­generation compounds in PC3 prostate cancer cells. The compounds were found to be highly apoptotic, exhibited strong antimigratory effects, and showed inhibition of TrRx. Again, these compounds were as stable as the second-­generation compounds and the co-­localization of both titanium and gold metals (1 : 1 ratio) was demonstrated in PC3 cancer cells, indicating the robustness of the bimetallic species.57 For titanocene-­based heterometallic complexes, it seems that the mode of action is mainly due to the second metal (either ruthenium or gold). However, in these heterometallic complexes the titanocene fragment seems to have a non-­negligible role in the modification of some of the chemicophysical properties of the compounds, with remarkable improvement of their pharmacological profile.

6.5  Other Heterometallic Compounds This section covers a diverse body of compounds based on different combinations of metallic fragments known to be active against tumors and/or displaying optical imaging properties.

6.5.1  G  old-­c ontaining Compounds (Gold–Platinum, Gold–Ruthenium, Gold–Cobalt, Gold–Silver, and Gold–Copper) In addition to previously described gold compounds containing ferrocene, p-­cymene ruthenium, or titanocene fragments (see Sections 6.2, 6.3.1, and 6.4.2) rationally designed gold bimetallic compounds aiming to incorporate modes of action of a second metallic core have been reported. This subsection collects reports on gold(i) compounds containing platinum(ii),58,59 ruthenium(ii),60 cobalt(ii),61 silver(i),62 and copper(i).63 The most active compounds are depicted in Chart 6.7. The group of Cinellu et al. reported on bimetallic gold(i)–platinum(ii) complexes like 45, in Chart 6.7. This compound was markedly more cytotoxic than cisplatin and the monometallic platinum starting material in ovarian cancer cell lines (including cisplatin-­resistant cell lines).59 Interestingly, it was found that the biological effect was not due to synergism but to an additive effect instead. Compound 45 had acceptable stability and solubility profiles in the aqueous environment. A joint analysis of ESI-­MS and crystallographic

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

158

Chapter 6

Chart 6.7  

data of adducts formed by interaction of this bimetallic compound with proteins showed that it cleaves. Only monometallic fragments were found to be bound to the biomolecules.59 Metallocene-­based compounds containing ruthenocenyl60 and cobaltocene61 incorporating gold(i) fragments have been described (e.g. 46 and 47). While their cytotoxicity in ovarian cancer cell lines was similar to that of cisplatin, it was demonstrated that the Co–Au compound 47 inhibited thioredoxin reductase (TrxR).61 While silver is better known for its antimicrobial properties, there are some relevant examples of silver anticancer agents.87 Gold(i) compounds containing silver(i) fragments (such as 48) have been described.62 It was demonstrated that heterobimetallic gold–silver compounds were more effective (submicromolar range) than previously described gold–gold analogues and cisplatin in ovarian cell lines (including cisplatin-­resistant cell line A2780cis).62 Laguna et al. reported a series of heterometallic gold(i)–copper(i) complexes containing the ligand 2-­propargylthiopyridine.63 In vitro studies against colon cancer cell lines Caco-­2-­PD7 and Caco2-­TC7 showed values in the nanomolar range (2 × 10−4 µM) for the bimetallic complex 49, remarkably lower than those found for its monometallic counterparts or for cisplatin or

View Online

Heterometallic Complexes as Anticancer Agents

159

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Auranofin. The compounds were highly apoptotic but more detailed mechanistic studies were not performed.

6.5.2  R  uthenium-­containing Compounds (Ruthenium– Platinum, Ruthenium–Nickel, and Ruthenium–Copper) Ruthenium(ii) polypiridyl complexes were investigated to explore their DNA interactions (supramolecular) but have emerged more recently as a new class of therapeutics, luminescent imaging agents, biological probes, and theranostics.93 Compounds containing ruthenium–piridyl motifs and incorporating platinum(ii),64,65 nickel(ii),66 and copper(ii)66,67 have been described. Chart 6.7 depicts one example of this type of compound, the ruthenium–platinum derivative 50.65 Some compounds displayed relevant non-­intercalative DNA binding64 and, in some cases, DNA photocleavage,66 and significant cell growth inhibition.65–67 Compound 50, while less cytotoxic on ovarian cell lines than cisplatin, was found to be cytostatic and to inhibit cell proliferation by upregulating the cyclin-­dependent kinase inhibitor p27KIP1 and cell growth by a gain of function at the G1 restriction point (resulting in G1 cell cycle arrest). This compound did not show covalent binding to DNA.65 Ruthenocene compounds (previously described by the group of Garcia and co-­workers) have been modified with polypyridine ligands to generate heterometallic complexes containing copper(i), such as 51 in Chart 6.7.67 While 51 exhibited relevant cytotoxicity in ovarian A2780 and breast MCF7 cancer cell lines, a synergistic effect could not really be demonstrated (cytotoxicity of the resulting bimetallic compound was higher than that of the copper cationic precursor and very similar to that of the ruthenium precursor).67 Heterometallic compounds based on ruthenium(iii) fragments resembling well-­known ruthenium(iii) anticancer derivatives NAMI-­A and KP1019, and containing platinum(ii) moieties resembling cisplatin, have been reported (such as trinuclear Ru2Pt compound 52 in Chart 6.7).68,69 Water-­soluble 52 was found to bind tightly to DNA and had an intermediate toxicity (between that of cisplatin and NAMI-­A) in selected lines from the US National Cancer Institute NCI Developmental Therapeutics Program in vitro screening. However, the profile of 52, when submitted to the COMPARE program from NCI, indicated a rather novel and unique mode of action not based on that of cisplatin and NAMI-­A.68 A related binuclear hybrid Ru–Pt was found to have stronger interactions with DNA (stronger even than cisplatin).69 Both trinuclear and dinuclear derivatives displayed antimigratory effects in highly metastatic lung cancer A549 and triple negative breast cancer (MDA-­MB231) cell lines (with the effects greater for 52).69 The binuclear derivative had a mode of action (COMPARE assay) that could be correlated with cisplatin, indicating a strong influence on the number of ruthenium metallic fragments on the hybrid.69

View Online

160

Chapter 6

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

6.5.3  T  heranostic Compounds (Ruthenium–M, M = Gold, Osmium, Rhodium, Gadolinium; Gadolinium– Platinum; and Rhenium–Gold) Theranostic compounds combine therapeutic and imaging properties in a single molecule. Heterometallic complexes with potential theranostic applications have been recently reviewed.2 In this section we will highlight relevant examples. Heterometallic compounds with antitumor and optical imaging activity based on luminescent polypyridyl ruthenium(ii) fragments (“Ru(bpy)2”) and a second metal like gold(i),70,71 osmium(ii),71 rhodium(i),71 or luminescent gadolinium(iii)72 have been described. Ruthenium(ii) polypyridyl complexes containing gold(i) fragments70,71 (such as 53 in Chart 6.7) have been found to be cytotoxic in cancer cells, while their photophysical properties were used to evaluate their localization inside tumor cells or organelles by confocal70 or fluorescence71 microscopy. Compound 53, containing a gold(i) fragment resembling Auranofin, was found to have a cytotoxicity comparable with or higher than that of cisplatin

Chart 6.8  

View Online

Heterometallic Complexes as Anticancer Agents

161

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

71

in ovarian and lung cancer cell lines (A2780S, A2780R, and A549). The uptake of 53 by cells at 37 °C (fluorescence microscopy) was demonstrated with targeting of organelles in the cytoplasm or the nuclei.71 This compound (like most gold compounds) did not show interaction with DNA. It was demonstrated that the two metals do not interfere with one another (electronically independent) and the authors claimed that the improvement in cytotoxicity may come from the chemicophysical properties of the resulting compound.71 Incorporation of gadolinium(iii) in the ruthenium(ii) polypyridyl scaffolds has also been explored (54 in Chart 6.7).72 Gadolinium(iii) compounds have shown a great potential in magnetic resonance imaging (MRI) and are also used as radiosensitizers and in synchrotron stereotactic radiotherapy.72–74,93 They commonly display functionalized chelator macrocycle ligands like DOTA (1,4,7,10-­tetraazacyclododecane-­1,4,7, 10-­tetraacetic acid) or DOTA-­TOC (DOTA-­Tyr3-­octreotide).94 The ruthenium(ii)–gadolinium(iii) compound 54 was designed as a potential bimodal probe for MRI and optical imaging.72 It was demonstrated that the compound was cytotoxic and apoptotic in HeLa cells and that it had a good cell permeability (54 could be visualized in the cells). Gadolinium(iii) compounds incorporating platinum(ii) centers73,74 (via a thiol modification of linkers) have been found to be cytotoxic and to interact with DNA. They are robust during cell uptake and accumulate preferentially in the nuclei.73,74 Sessler and co-­workers have used texaphyrin ligands containing gadolinium,75 known as tumor-­localizing agents, to coordinate platinum(ii)76,77 and platinum(iv)78 complexes. In this way they synthesized trimetallic and bimetallic species (see compound 55 as an example in Chart 6.8).78 They demonstrated an increased cellular uptake for the platinum(ii) derivatives and that some mechanisms of resistance usual for cisplatin were overcome.76,77 The goal of using and coordinating platinum(iv) compounds was to control the release of platinum(ii) fragments in the cellular milieu avoiding side reactions of the platinum center with other biomolecules.78 In vitro studies toward wild type 2780CP and platinum-­resistant A2780 ovarian cancer cell lines showed better resistance factor values of the complexes when compared with cisplatin (8.32 for compound 55 vs. 18.25). In addition, controlled exposure to light or a reducing environment were used to successfully prove the slow release of platinum moieties.78 Gimeno and co-­workers have developed bimetallic theranostic compounds based on the well-­known d6 tricarbonyl-­bis-­imine rhenium(i) as emissive fragment and gold(i)–phosphane moieties as the cytotoxic component (like more active compounds 56–59 in Chart 6.8).2,79,80 It was demonstrated that the cytotoxicity on lung cancer cell line A549 came solely from the gold(i) fragment and that the optical properties of the rhenium fragment were not compromised.2,79,80 The biodistribution by fluorescence imaging showed that, while the rhenium monometallic derivative

View Online

162

Chapter 6

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

localized in cytoplasm and to some extent in the mitochondria, the bimetallic compounds accumulated preferentially in the nucleus and nucleolus.2,79,80 Preliminary studies showed that DNA could be one possible target and that the compounds entered the cell by passive diffusion.80

6.5.4  Other (Cobalt-­based and Copper–Zinc) There are three reports in this last subsection that do not fit in any other classification within this chapter.81–83 Cobalt(ii) octahedral complexes with pyridine-­based ligands were used to prepare trinuclear or binuclear heterometallic complexes containing zinc(ii), cadmium(ii), and mercury(ii) which did not display improved cytotoxicity (in human brain tumor cancer cell line U87) with respect to the monometallic cobalt starting material.81 Based on the relevant anticancer activities exhibited by copper and organotin derivatives, a trinuclear copper–tin species CuSn2(Trp) containing chiral l-­tryptophan-­derived bis(1,2-­diaminobenzene)copper(ii) was synthesized (compound 60 in Chart 6.8).82 Compound 60 exerted cell death via apoptosis.82 More detailed studies of 60 in vitro demonstrated improved cytotoxicity in liver Hep-­G2 and ovarian cancer PA-­1 cells and confirmed the apoptotic behavior.83 These findings, as well as low DNA damage, encouraged exploration of in vivo toxicity.83 The oxidative stress caused by free radicals and ROS species was studied in kidney, liver, and brain (by means of histopathological procedures). No systemic toxicity (genotoxicity, nephrotoxicity, hepatotoxicity, and neurotoxicity) for 60 was found, when compared with cisplatin.83

6.6  Conclusion To summarize, a large number of heterometallic compounds with anticancer properties have been reported in the last decade. The compounds described are extremely diverse in terms of metals and ligand systems. In this chapter, we have aimed to classify them into families of compounds with similar biologically active motifs. Many of the studies described herein support the hypothesis of an improved pharmacological profile of heterometallic compounds due to cooperative and/or synergistic effects of the different metals. The synergistic effect has been demonstrated in vitro for a significant number of compounds (or families of compounds), while a cooperative effect has been found in fewer cases. The lack of detailed mechanistic studies (except for specific sets of compounds) prevents making broader generalizations. The mechanisms in some cases are mostly based on the biological effects of one of the metals, while the second metal acts a modulator of electronic or other chemicophysical properties. In other cases (like

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

163

ruthenium(ii)–arene-­based compounds) the observed effects seem to truly come from the two different metals. Two families of compounds have emerged from these studies as extremely promising potential candidates against metastatic cancers (such as renal and triple negative breast cancer): ruthenium(ii)–arene-­based compounds containing gold(i) and platinum(iv) centers and titanocene(iv)–gold(i) compounds. The low systemic toxicity found for these compounds, coupled with relevant in vitro and in vivo efficacy, and significant metastasis and angiogenesis inhibition correlated to that of related targets, warrant advanced preclinical studies to better understand the mode of action of these bimetallic compounds, to study their pharmacology and toxicology, and to improve their formulation. Lastly, some heterometallic compounds have also emerged as potential anticancer theranostics with rhenium(i)–gold(i) compounds as very promising candidates for further development. Overall, this is a field of much growing interest within the metallodrugs community. Heterometallic compounds can now be truly envisioned as multitargeted therapies designed to avert target specific-­induced resistance. We should expect to find relevant advanced preclinical studies in the very near future. This will help with elucidation of targets and modes of action of heterometallic compounds and will allow further drug design optimization.

Acknowledgement We thank the support of the National Cancer Institute and the National Institute for General Medical Sciences (NIGMS) for grants 1SC1CA182844 and 2SC1 GM127278-­05A1 (M.C.). N.C. thanks Fundación Alfonso Martín Escudero (Spain) for a postdoctoral fellowship.

References 1. F. Pelletier, V. Comte, A. Massard, M. Wenzel, S. Toulot, P. Richard, M. Picquet, P. Le Gendre, O. Zava, F. Edafe, A. Casini and P. J. Dyson, J. Med. Chem., 2010, 53(19), 6923. 2. V. Fernandez-­Moreira and M. C. Gimeno, Chem. - Eur. J., 2018, 24(14), 3345. 3. C. G. Hartinger, A. D. Phillips and A. A. Nazarov, Curr. Top. Med. Chem., 2011, 11(21), 2688. 4. R. W. Mason, K. McGrouther, P. R. R. Ranatunge-­Bandarage, B. H. Robinson and J. Simpson, Appl. Organomet. Chem., 1999, 13(3), 163. 5. J. Rajput, J. R. Moss, A. T. Hutton, D. T. Hendricks, C. E. Arendse and C. Imrie, J. Organomet. Chem., 2004, 689(9), 1553. 6. M. Auzias, B. Therrien, G. Süss-­Finka, P. Stepnickab, W. H. Ang and P. J. Dyson, Inorg. Chem., 2008, 47(2), 578.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

164

Chapter 6

7. M. Auzias, J. Gueniat, B. Therrien, G. Süss-­Fink, A. K. Renfrew and P. J. Dyson, J. Organomet. Chem., 2009, 694(6), 855. 8. B. Maity, M. Roy, B. Banik, R. Majumdar, R. R. Dighe and A. R. Chakravarty, Organometallics, 2010, 29(16), 3632. 9. D. Nieto, A. M. González-­Vadillo, S. Bruña, C. J. Pastor, C. Ríos-­Luci, L. G. León, J. M. Padrón, C. Navarro-­Ranninger and I. Cuadrado, Dalton Trans., 2012, 41(2), 432. 10. E. Guillen, A. Gonzalez, P. K. Basu, A. Ghosh, M. Font-­Bardia, T. Calvet, C. Calvis, R. Messeguer and C. Lopez, J. Organomet. Chem., 2017, 828, 122. 11. G. Pilloni, R. Graziani, B. Longato and B. Corain, Inorg. Chim. Acta, 1991, 190(2), 165. 12. M. Viotte, B. Gautheron, M. M. Kubicki, I. Ninfant'ev and S. P. Fricker, Met.-­Based Drugs, 1995, 2(6), 311. 13. M. Viotte, B. Gautheron, M. M. Kubicki, Y. Mugnier and R. V. Parish, Inorg. Chem., 1995, 34(13), 3465. 14. M. Viotte, B. Gautheron, I. Ninfant'ev and L. G. Kuz'mina, Inorg. Chim. Acta, 1996, 253(1), 71. 15. G. Von Poelhsitz, A. L. Bogado, M. Peres de Araujo, H. S. Selistre-­de-­ Araujo, J. Ellena, E. E. Castellano and A. A. Batista, Polyhedron, 2007, 26(16), 4707. 16. H. Kodama, K. Ohto, N. Oohara, K. Nakatsui and Y. Kaneda, US Pat. 7390915 B1, 06/24/2008. 17. E. Fourie, E. Erasmus, J. C. Swarts, A. Jakob, H. Lang, G. K. Joone and C. E. J. van Rensburg, Anticancer Res., 2011, 31(3), 825. 18. H. Samouei, M. Rashidi and F. W. Heinemann, J. Organomet. Chem., 2011, 696(23), 3764. 19. H. Bjelosevic, I. A. Guzei, L. C. Spencer, T. Persson, F. H. Kriel, R. Hewer, M. J. Nell, J. Gut, C. E. J. van Rensburg, P. J. Rosenthal, J. Coates, J. Darkwa and S. K. C. Elmroth, J. Organomet. Chem., 2012, 720, 52. 20. P. Chellan, K. M. Land, A. Shokar, A. Au, S. H. An, C. M. Clavel, P. J. Dyson, C. de Kock, P. J. Smith, K. Chibale and G. S. Smith, Organometallics, 2012, 31(16), 5791. 21. C. Bincoletto, I. L. S. Tersariol, C. R. Oliveira, S. Dreher, D. M. Fausto, M. A. Soufen, F. D. Nascimento and A. C. F. Caires, Bioorg. Med. Chem., 2005, 13(8), 3047. 22. C. M. V. Barbosa, C. R. Oliveira, F. D. Nascimiento, M. C. M. Smith, D. M. Fausto, M. A. Soufen, E. Sena, R. C. Araujo, I. L. S. Tersariol, C. Bincoletto and A. C. F. Caires, Eur. J. Pharmacol., 2006, 542(1–3), 37. 23. C. R. Oliveira, C. M. V. Barbosa, F. D. Nascimento, C. S. Lanetzki, M. B. Meneghin, F. E. G. Pereira, E. J. Paredes-­Gamero, A. T. Ferreira, T. Rodrigues, M. L. S. Queiroz, A. C. F. Caires, I. L. S. Tersariol and C. Bincoletto, Chem.-­Biol. Interact., 2009, 177(3), 181. 24. H. Goitia, Y. Nieto, M. D. Villacampa, C. Kasper, A. Laguna and M. C. Gimeno, Organometallics, 2013, 32(20), 6069. 25. H. Charvatova, T. Riedel, I. Cisarova, P. J. Dyson and P. Stepnicka, J. Organomet. Chem., 2016, 802, 21.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

165

26. W. Henderson and S. R. Alley, Inorg. Chim. Acta, 2001, 322(1–2), 106. 27. J. Schulz, A. K. Renfrew, I. Cisarova, P. J. Dyson and P. Stepnicka, Appl. Organomet. Chem., 2010, 24(5), 392. 28. J. Schulz, J. Tauchman, I. Cisarova, T. Riedel, P. J. Dyson and P. Stepnicka, J. Organomet. Chem., 2014, 751, 604. 29. N. Lease, V. Vasilevski, M. Carreira, A. de Almeida, M. Sanaú, P. Hirva, A. Casini and M. Contel, J. Med. Chem., 2013, 56(14), 5806. 30. U. E. I. Horvath, G. Bentivoglio, M. Hummel, H. Schottenbergm, K. Wurst, M. J. Nell, C. E. J. van Rensburg, S. Cronje and H. G. Raubenhemier, New J. Chem., 2008, 32(3), 533. 31. J. F. Arambula, R. McCall, K. J. Sidoran, D. Magda, N. A. Mitchell, C. W. Bielawski, V. M. Lynch, J. L. Sessler and K. Arumugam, Chem. Sci., 2016, 7(2), 1245. 32. I. Ott, K. Kowalski, R. Gust, J. Maurer, P. Mucke and R. F. Winter, Bioorg. Med. Chem. Lett., 2010, 20(3), 866. 33. B. Weber, A. Serafin, J. Michie, C. van Rensburg, J. C. Swarts and L. Bohm, Anticancer Res., 2004, 24(2B), 763. 34. J. Conradie and J. C. Swarts, Dalton Trans., 2011, 40(22), 5844. 35. P. Sathyadevi, P. Krishnamoorthy, R. R. Burotac, A. H. Cowley and N. Dharmaraj, Metallomics, 2012, 4(5), 498. 36. M. C. Gimeno, H. Goitia, A. Laguna, M. E. Luque, M. D. Villacampa, C. Sepulveda and M. Meireles, J. Inorg. Biochem., 2011, 105(11), 1373. 37. J. Tauchman, G. Süss-­Fink, P. Stepnicka, O. Zavac and P. J. Dyson, J. Organomet. Chem., 2013, 723, 233. 38. D. Nieto, S. Bruña, A. M. González-­Vadillo, J. Perles, F. Carrillo-­Hermosilla, A. Antiñolo, J. M. Padrón, G. B. Plata and I. Cuadrado, Organometallics, 2015, 34(22), 5407. 39. P. Govender, H. Lemmerhirt, A. T. Hutton, B. Therrien, P. J. Bednarski and G. S. Smith, Organometallics, 2014, 33(19), 5535. 40. P. Govender, T. Riedel, P. J. Dyson and G. S. Smith, Dalton Trans., 2016, 45(23), 9529. 41. C.-­H. Wu, D.-­H. Wu, X. Liu, G. Guoyiqibayi, D.-­D. Guo, G. Lv, X.-­M. Wang, H. Yan, H. Jiang and Z.-­H. Lu, Inorg. Chem., 2009, 48(6), 2352. 42. L. Massai, J. Fernández-­Gallardo, A. Guerri, A. Arcangeli, S. Pillozzi, M. Contel and L. Messori, Dalton Trans., 2015, 44(24), 11067. 43. J. Fernández-­Gallardo, B. T. Elie, M. Sanaú and M. Contel, Chem. Commun., 2016, 52(15), 3155. 44. B. T. Elie, Y. Pechenyy, F. Uddin and M. Contel, J. Biol. Inorg. Chem., 2018, 23(3), 399. 45. B. T. Elie, K. Hubbard, M. A. Cornejo, B. Layek, S. Prabha and M. Contel, To be submitted. 46. B. Bertrand, A. Citta, I. L. Franken, M. Picquet, A. Folda, V. Scalcon, M. P. Rigobello, P. Le Gendre, A. Casini and E. Bodio, J. Biol. Inorg. Chem., 2015, 20(6), 1005. 47. L. Ma, R. Ma, Z. Wang, S.-­M. Yiu and G. Zhu, Chem. Commun., 2016, 52(71), 10735.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

166

Chapter 6

48. L. Ma, X. Lin, C. Li, Z. Xu, C.-­Y. Chan, M.-­K. Tse, P. Shi and G. Zhu, Inorg. Chem., 2018, 57(5), 2917. 49. N. Singh, S. Jang, J. H. Jo, D. H. Kim, D. W. Park, I. H. Kim, H. Kim, S. C. Kang and K.-­W. Chi, Chem. -­Eur. J., 2016, 22(45), 16157. 50. M. Wenzel, B. Bertrand, M.-­J. Eymin, V. Comte, J. A. Harvey, P. Richard, M. Groessl, O. Zava, H. Amrouche, P. D. Harvey, P. Le Gendre, M. Picquet and A. Casini, Inorg. Chem., 2011, 50(19), 9472. 51. J. F. González-­Pantoja, M. Stern, A. A. Jarzecki, E. Royo, E. Robles-­ Escajeda, A. Varela-­Ramírez, R. J. Aguilera and M. Contel, Inorg. Chem., 2011, 50(21), 11099. 52. J. Fernández-­Gallardo, B. T. Elie, F. J. Sulzmaier, M. Sanaú, J. W. Ramos and M. Contel, Organometallics, 2014, 33(22), 6669. 53. J. Fernández-­Gallardo, B. T. Elie, T. Sadhukha, S. Prabha, M. Sanaú, S. A. Rotenberg, J. W. Ramos and M. Contel, Chem. Sci., 2015, 6(9), 5269. 54. M. Contel, J. Fernández-­Gallardo, B. T. Elie and J. W. Ramos, US Pat. 9315531, 04/19/2016. 55. B. T. Elie, J. Fernández-­Gallardo, N. Curado, M. A. Cornejo, J. W. Ramos and M. Contel, Eur. J. Med. Chem., 2019, 161, 310. 56. B. T. Elie, Doctoral Dissertation Thesis, Biology PhD Program, City University of New York, NY, USA, 2018. 57. Y. F. Mui, J. Fernández-­G allardo, B. T. Elie, A. Gubran, I. Maluenda, M. Sanaú, O. Navarro and M. Contel, Organometallics, 2016, 35(9), 1218. 58. M. Wenzel, E. Bigaeva, P. Richard, P. Le Gendre, M. Picquet, A. Casini and E. Bodio, J. Inorg. Biochem., 2014, 141, 10. 59. M. Serratrice, L. Maiore, A. Zucca, S. Stoccoro, I. Landini, E. Mini, L. Massai, G. Ferraro, A. Merlino, L. Messori and M. A. Cinellu, Dalton Trans., 2016, 45(2), 579. 60. H. Bjelosevic, I. A. Guzei, L. C. Spencer, T. Persson, F. H. Kriel, R. Hewer, M. J. Nell, J. Gut, C. E. J. van Rensburg, P. J. Rosenthal, J. Coates, J. Darkwa and S. K. C. Elmroth, J. Organomet. Chem., 2012, 720, 52. 61. S. Vanicek, H. Hopacka, K. Wurst, S. Vergeiner, S. Kankowski, J. Schur, B. Bildstein and I. Ott, Dalton Trans., 2016, 45(4), 1345. 62. E. Barreiro, J. S. Casas, M. D. Couce, A. Sánchez, J. Sordo and E. M. Vázquez-­López, J. Inorg. Biochem., 2014, 131, 68. 63. E. García-­Moreno, S. Gascón, M. J. Rodriguez-­Yoldi, E. Cerrada and M. Laguna, Organometallics, 2013, 32(13), 3710. 64. K. Van der Schilden, F. Garcia, H. Kooijman, A. L. Spek, J. G. Haasnoot and J. Reedijk, Angew. Chem., Int. Ed., 2004, 43(42), 5668. 65. V. Ramu, M. R. Gill, P. J. Jarman, D. Turton, J. A. Thomas, A. Das and C. Smythe, Chem. - Eur. J., 2015, 21(25), 9185. 66. Y. Liu, T. Chen, Y.-­S. Wong, W.-­J. Mei, X.-­M. Huang, F. Yang, J. Liu and W.-­J. Zheng, Chem.-­Biol. Interact., 2010, 183(3), 349. 67. J. Lopes, D. Alves, T. S. Morais, P. J. Costa, M. F. M. Piedade, F. Marques, M. J. Villa de Brito and H. M. Garcia, J. Inorg. Biochem., 2017, 169, 68.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

Heterometallic Complexes as Anticancer Agents

167

68. A. Herman, J. M. Tanski, M. F. Tibbetts and C. M. Anderson, Inorg. Chem., 2008, 47(1), 274. 69. C. M. Anderson, I. R. Taylor, M. F. Tibbetts, J. Philpott, Y. Hu and J. M. Tanski, Inorg. Chem., 2012, 51(23), 12917. 70. L. Boselli, M. Carraz, S. Mazères, L. Paloque, G. González, F. Benoit-­Vical, A. Valentin, C. Hemmert and H. Gornitzka, Organometallics, 2015, 34(6), 1046. 71. M. Wenzel, A. Almeida, E. Bigaeva, P. Kavanagh, M. Picquet, P. Le Gendre, E. Bodio and A. Casini, Inorg. Chem., 2016, 55(5), 2544. 72. A. Nithyakumar and V. Alexander, Dalton Trans., 2015, 44(40), 17800. 73. J. M. Fenton, M. Busse and L. M. Rendina, Aust. J. Chem., 2015, 68(4), 576. 74. E. L. Crossley, J. B. Aitken, S. Vogt, H. H. Harris and L. M. Rendina, Angew. Chem., Int. Ed., 2010, 49(7), 1231. 75. R. A. Miller, K. Woodburn, Q. Fan, M. Renschler, J. L. Sessler and J. A. Koutcher, Int. J. Radiat. Oncol., 1999, 45(4), 981. 76. J. F. Arambula, J. L. Sessler and Z. H. Siddik, Bioorg. Med. Chem., 2011, 21, 1701. 77. F. Arambula, J. L. Sessler and Z. H. Siddik, MedChemComm, 2012, 3, 1275. 78. G. Thiabaud, J. F. Arambula, Z. H. Siddik and J. Sessler, Chem. - Eur. J., 2014, 20(29), 8942. 79. V. Fernández-­Moreira, I. Marzo and M. C. Gimeno, Chem. Sci., 2014, 5(11), 4434. 80. A. Luengo, V. Fernández-­Moreira, I. Marzo and M. C. Gimeno, Inorg. Chem., 2017, 56(24), 15159. 81. N. K. Kaushik, A. Mishra, A. Ali, J. S. Adhikari, A. K. Verma and R. Gupta, J. Biol. Inorg. Chem., 2012, 17(8), 1217. 82. M. Chauhan, K. Banerjee and F. Arjmand, Inorg. Chem., 2007, 46(8), 3072. 83. Y. Zaidi, F. Arjmand, N. Zaidi, J. A. Usmani, H. Zubair, K. Akhtar, M. Hossain and G. G. H. A. Shadab, Metallomics, 2014, 6(8), 1469. 84. C. Ornelas, New J. Chem., 2011, 35(10), 1973. 85. E. Hillard, A. Vessières, F. Le Bideau, D. Plazuk, D. Spera, M. Huche and G. Jaouen, ChemMedChem, 2006, 1(5), 551 and references therein. 86. G. Jaouen, G. A. Vessières and S. Top, Chem. Soc. Rev., 2015, 44(24), 8802 and references therein. 87. M. Frik, J. Fernández-­Gallardo, O. Gonzalo, V. Mangas-­Sanjuan, M. González-­Alvarez, A. Serrano del Valle, C. Hu, I. González-­Alvarez, M. Bermejo, I. Marzo and M. Contel, J. Med. Chem., 2015, 58(15), 5825 and references therein. 88. B. Benoît and A. Casini, Dalton Trans., 2014, 43(11), 4209 and references therein. 89. S. Medici, M. Peana, V. M. Nurchi, J. I. Lachowicz, G. Crisponi and M. A. Zoroddu, Chem. Soc. Rev., 2015, 284, 329. 90. P. Zhang and P. J. Sadler, J. Organomet. Chem., 2017, 839, 5.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00143

168

Chapter 6

91. U. Ndagi, N. Mhlongo and M. E. Soliman, Drug Des., Dev. Ther., 2017, 11, 599. 92. W. E. Berdel, H. J. Dchmoll, M. E. Scheulen, A. Korfel, M. F. Knoche, A. Harstrick, F. Bach and G. Sa, J. Cancer Res. Clin. Oncol., 1994, 120, R172. 93. F. E. Poynton, S. A. Bright, S. Blasco, D. C. Williams and J. M. Kelly, Chem. Soc. Rev., 2017, 46(24), 7706. 94. P. Hermann, J. Kotek, V. Kubicek and I. Lukes, Dalton Trans., 2008, 0(23), 3027.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Chapter 7

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects Debbie C. Crans*a,b, Noah E. Barkleyc, Liliana Montezinhod and M. Margarida Castro*e,f a

Colorado State University, Department of Chemistry, Fort Collins, CO 80525, USA; bColorado State University, Cell and Molecular Biology, Fort Collins, CO 80525, USA; cColorado State University, Molecular and Cellular Integrative Neuroscience Program, Fort Collins, CO 80525, USA; dCenter for Investigation Vasco da Gama (CIVG), Department of Veterinary Medicine, Escola Universitária Vasco da Gama, Coimbra, Portugal; eUniversity of Coimbra, Department of Life Sciences, Faculty of Science and Technology, 3000-­456 Coimbra, Portugal; fUniversity of Coimbra, Coimbra Chemistry Center, 3000-­456 Coimbra, Portugal *E-­mail: [email protected]; [email protected]

7.1  Introduction Vanadium salts and vanadium compounds (VCs) are known inhibitors for a wide range of different classes of enzymes.1–4 However, the anticancer and antidiabetic effects have been attributed to a few classes of enzymes associated with cellular regulation, such as the signal transduction enzymes, protein tyrosine phosphatases.1 Thus, although many classes of enzymes are inhibited by vanadium salts and VCs, some selectivity is reported. Some   Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

169

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

170

Chapter 7

enzymes such as phosphatases are generally believed to be inhibited by vanadate but it is often not recognized that, because some are inhibited very potently and others are barely inhibited, this affinity difference results in the appearance of a difference in selectivity. The enzymes that are known to be most potently inhibited by vanadium salts and compounds are the Na+, K+-­ATPases,5–7 ribonucleases8 and protein tyrosine phosphatases.1 Protein tyrosine phosphatases are particularly strongly inhibited by vanadate and other VCs because the vanadium can take on the geometry of the transition state of the hydrolysis of the phosphate ester.1,9,10 Our data-­mining studies showed that the trigonal bipyramidal (TBP) geometry is that found bound to the phosphatases in the V–protein phosphatase complexes, regardless of which form was present in solution1,9,11 (see Figure 7.1). Vanadium(v) is a first-­row transition metal in group V of the periodic table, with an interesting and diverse chemistry in solution. Focusing on oxidation states from +3 to +5 that are present under biological conditions, different anionic and cationic species form depending on the total concentration, pH and ionic strength of the medium.12–16 It is a trace element in the environment, present in the Earth's crust, soil, fossils, water reservoirs, the atmosphere and food.17 It is essential in some organisms, where it has an important biological role such as the metal center of V-­haloperoxidases in marine algae, fungi,

Figure 7.1  The  active site of the PTP1B showing a vanadium atom (in pink) bound

with many H-­bonds (in green) to the vanadate oxygen atoms documenting the tight fit in the second transition state. The structure has a coordination geometry approaching TBP for the vanadium. Atom color code: oxygen, red; nitrogen, blue; carbon, grey; vanadium, pink. Selected residues are labeled and the following sections of the protein define the active site. (1) The P-­loop (residues 215–222), which includes the cysteine nucleophile (Cys 215) and important groups such as Ser 222 and Arg 221 which stabilize the transition state. (2) The flexible loop with a WPD sequence (residues 179–187) containing the general acid/ base Asp 181. (3) The Q-­loop (residues 261–262) including the Gln 262 residue important for transition state 2 shown. (4) The lysine-­loop (residues 119–126) containing Lys 120, a conserved residue that is involved in key interactions with the WPD-­loop. (5) The Tyr-­P recognition loop (residues 47–49) involved in substrate recognition and binding.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

171

lichens, Streptomyces bacteria and cyanobacteria; as a cofactor of V-­dependent nitrogenases in some nitrogen-­fixing bacteria; in mushrooms of genus Amanita muscaria in the form of a non-­oxidovanadium(iv) complex named amavadin; and also in Ascideans, marine invertebrates that accumulate vanadium from seawater in vanadocytes.12,18 It does not seem to be an essential micronutrient for humans, but somehow interferes with physiological and metabolic functions causing positive and negative effects depending on its concentration in the human body.19,20 Interest has been devoted to this metal, particularly in the last couple of decades, due to its demonstrated pharmacological properties as antidiabetic, anticancer, antimicrobial, antiparasitic (Chagas, Leishmaniasis, amoebiasis) and anti-­inflammatory agents,1,17,20–24 as well as its cardiovascular, neuroprotective,25,26 osteogenic27,28 and, more recently, immunomodulatory effects.29 The structure and speciation of the vanadium salts or VCs are important to their action.30–36 Vanadium in oxidation state V is known to act as a structural and electronic phosphate analog and found to bind to enzymes at phosphate-­binding sites.30–35 Vanadium in oxidation state iv is known to act as a bivalent metal ion, being able to replace cations such as Zn2+ and Ca2+.36 Vanadium in the form of vanadate is known to oligomerize when its concentration reaches millimolar levels, and these oligomeric species have properties different from monomeric vanadate.12,30,37,38 Vanadium can also be reduced in the presence of some of the reducing agents that are present in biological systems.39 Knowledge of the mechanism of action of each VC at the molecular, cellular and organism levels will help to better understand its therapeutic activity toward different pathologies. The biological effects of vanadium coordination compounds depend on the administered form and on the biological system under investigation.4 Therefore, it is critical to specify the biological system and the experimental conditions under investigation, so that the details in the observed effects can be related to the chemical speciation involved.3,30,33,38 Many VCs have been synthesized and tested, from vanadium(iv,v) salts to complexes of vanadium (iii,iv,v) with a wide range of organic ligands. Some complexes have ligands that are natural products and some improve absorption and specificity and minimize the dose and the toxic effects of the metal.40,41 Representative complexes reported in the literature are listed in Figure 7.2. The VCs were structurally characterized in the solid state and in solution, particularly under physiological conditions. Their therapeutic activity and cytotoxicity have been evaluated in different cellular models and some have been tested in vivo, in a range of different animal models.14,42 Bis(ethylmaltolato)oxovanadium(iv) (BEOV) was subjected to phase 1 and 2 clinical trials as a potential antidiabetic agent40 but went off-­patent on 30 September 2011,34 which deterred future use of this compound. Despite the many promising activities that have been reported in the literature for the medicinal properties of VCs,23,24,43–46 issues in overcoming the problems of specificity and off-­target toxicity for long-­term use in vivo have been difficult.

View Online

Chapter 7

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

172

Figure 7.2  Schematic  representation of the chemical structures of some vanadium

compounds subjected to cellular studies as anticancer agents (see also Table 7.1). Abbreviations used in the figure are: ema (ethyl maltol or 2-­ethyl-­3-­hydroxy-­4H-­pyran-­4-­one); pic (picolinate or pyridine-­2-­ carboxylate); acac (acetylacetonate); dipic (dipicolinate); cyst (cysteine); chrys (chrysin); salen (2,2′-­ethylenebis(nitrilomethylidene) diphenol or N,N′-­ethylenebis(salicylimine)).

Furthermore, VCs have antagonist effects which may explain the reported controversies regarding some of the biological responses. Such properties complicate any future potential clinical application.17,19,47 Here, we focus on summarizing the enzyme inhibitory effects and aim to associate them with the anticancer properties of VCs.46,48–50 We describe in vitro cell studies and in vivo studies reported in the literature (see Table 7.1 for summary). This information provides clarification of the chemical and biochemical properties as well as the mechanism of systemic, cellular and molecular anticancer effects of VCs. However, many questions remain for ongoing and future investigation.

7.2  P  hosphorylation in Tumorigenesis and Signaling Pathways The enzymes involved in molecular mechanisms in cell regulation, particularly those whose deregulation gives rise to uncontrolled cellular proliferation and ultimately to tumorigenesis, are characteristic of families of cancer diseases.27–29,45–47,51 Possible targets for VC action will be identified, enabling

Published on 05 April 2019 on https://pubs.rsc.org | doi:

anticancer metallodrugs.a

Biological system Vanadium compound

Human cell lines

Vanadate salts (ortho-­and meta-­vanadate)

↑Cyclin D1; ↑caspase-­3; Esophageal squamous EC109; ↓Cell proliferation; cell cycle arrest at G2/M and S ↓PIWIL2; ↓Bcl2; DNA neuroblastoma SH SY5Y; phase; ↑apoptosis; fragmentation; activation hepatocarcinoma HCC; thy↓viability: ↓ψm; ↑ROS of RET/PTC1 and PI3K/ roid papillary carcinoma; AKT; ↑p53, ↑Bax; ↓Bcl2; breast MCF-­7; HCC HepG2 ↓Na/K-­ATPase activation; and Huh7 cells ↓exp HIF-­1α, HIF2α, VEGF, LDHA, Glut1; ↑cyclin B1 and CDK1 HaCaT ↓Cell growth Suppression of c-­fos proto-­ oncogene expression; induction of p15INKB expression Ovarian A2780; breast MCF7; Inhibition of cell pro— prostate PC3 liferation (increased cytotoxicity) Malignant melanoma cell lines Cell cycle arrest, ↑apoptosis — A375 and CN-­mel; fibroblast cell line BJ (normal control) Hepatocarcinoma HepG2; ↑Cytotoxicty; ↑apoptoDirect interaction with colon cancer HTC-­116, sis; ↓cell proliferation; DNA – oxidative cleavage Caco-­2, HT-­29 cell lines ↑ROS; ↓ψm; ↑cytotoxicty (control myofibroblast on cancer cells CCD18Co); calf thymus DNA Osteosarcoma MG-­63 ↑Cytotoxicty; ↑apoptosis ↑Caspases, ↑Bcl-­x, ↑DAPK; ↓PKB, AKT, AKT1

Vanadyl salts

VO(2-­hydroxy-­ naphthylaldimine); VO(polypyridyl) Vanadate; VO(dhp)2; VO(mpp)2; VO(ppp)2 VO(galactomannan); VO(Schiff bases) [VO(sal) and derivatives]

Molecular target

Reference 45, 98, 100–102, 106, 134

99

117 116 109, 119, 131, 143

46, 121 (continued)

173

V(iv,v)(8-­quinolinol-­­­ derivatives); VO(nta)

Antitumor effects

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

Table 7.1  Summary  of literature data resulting from cell and in vivo studies with vanadium salts and vanadium compounds as promising

Published on 05 April 2019 on https://pubs.rsc.org | doi:

174

Table 7.1  (continued) Biological system Vanadium compound

Human cell lines

VO(salicylaldimine); VO(N,N-­polypyridyl)

Ovarian A2780; breast MCF7; prostate PC3; Kidney HGK293; leukemia HL-­60

Antitumor effects

Molecular target

117, 124

118, 122, 145

132 110, 112, 138, 157–159

113 120 125 115

Chapter 7

↑Cell death; morphologDNA, nucleus, membrane ical changes (nuclear atypic and membrane alterations) VO(acac)2; VO(acac Pancreatic AsPC-­1; hepatocar- Cell cycle arrest at G2/M and PI3K/AKT; MAPK/ER; ↑ERK derivatives); Metavacinoma HepG2 G1/S; ↑ROS; ↓cancer cell and ↓pRb phosphorylanadate; VO(malato-­ growth; ↓viability tion; ↓cyclins and ↑p21 derivatives); vanadyl salt expression V(iv)(l-­cyst-­methyl ester) Murine tumor model ↓Tumor growth; ↑ROS; ↑Bcl-­2; ↑p53, ↑Bax; DNA damage; apoptosis ↑caspases; ↓VEGF-­A and induction MMP-­9 Alteration in expression ↓Cell proliferation; ↓cell VO(chrysin); VO(silibinin); Osteosarcoma MG-­63; colon levels of PKB/AKT, PAK, viability; ↓GSH/GSSG; VO(luteolin); adenocarcinoma HT-­29 DAPK, CdK4,6,7, FADD, cell cycle arrest at G2/M; VO(curcumin); and caco-­2; breast cancer AP2, NAK, JNK; ↑AKT1 ↑ROS; apoptosis inducVO(hesperidin) MDAMB231; lung cancer and ↓FAK phosphorylation; antioxidant and A549; prostate cancer tion; ↑caspase 3; ↓NF-­κB; anticancer activity LNCaP; breast cancer topoisomerase IB inhibiMCF-­7 tion; ↑peroxidase activity VO(alginate, Hepatoma BEL-­7402; (hepatic ↓Cell proliferation PTP1B inhibition polysaccharides) normal cells HL 7702) VO(pyridoxal) Melanoma A375; lung carci↑Cancer cell death; apopto- — noma A549 sis induction; ↑ROS; ↓ψm VO(salen) K562 ↓Cell proliferation; cell cycle — arrest at G2/M V(v)(peroxido-­betaine) Lung adenocarcinoma cells; ↓Cancer cell viability; apop- ↓H-­ras; ↓MMP2 expression breast cancer epithelial tosis induction; ↑ROS cells

Reference

Published on 05 April 2019 on https://pubs.rsc.org | doi:

VO(monosaccharide); VO(disaccharide); VO(ascorbate)

Metvan

Vanadocene dichloride

VO(indenoisoquinolines); VO(dibenzonaphthyridines) VO(phenanthroline/­ quinolone) Orthovanadate

Mitochondria from rat liver

↓Cell proliferation; ↑ROS; Cytochrome c release; ERK ↓ψm; apoptosis induction, activation morphology.changes; disruption of cytoskeleton fibers ↑ERKs phosphorylation; Non-­transformed osteoblast ↓Proliferation, ↑ROS and ↑GS3K MC3T3E1; (rat) osteosarapoptosis induction in coma UMR106 cells tumoral osteoblasts (antitumoral effect); collagen synthesis, ↑mitogenesis, ↑glucose consumption in non-­transformed osteoblasts(osteogenic activation) Leukemia, multiple myeloma, Apoptosis induction; ↓ψm; ↓MMP-­9 and 2 expression breast, glioblastoma, ovar↑ROS; ↓GSH ian, prostate and testicular cancer cells HepG2; EAT cells ↓Cell proliferation; apopto- Metallocene-­DNA complex sis induction; cell cycle formation arrest at G1/S and G2/M; ↑cytotoxicity — — ↓Topoisomerase I activity

103

Pancreatic adenocarcinoma PANC-­1 (control: normal pancreas duct epithelial) hTERT-­HPNE Mice with MDAY-­D2 tumors male BALB/c-­nu/nu mice

Cell cycle arrest at G2/M; — ↑ROS; cell death (necrosis and/or apoptosis)

130

↓Tumor growth; ↓tumor size; ↓tumoral microvessel density

45, 98

148

123, 154, 161, 162

147

(continued)

175

↓Tumoral expression HIF-­1α, HIF2α, VEGF, procaspases 3,9, cleaved PARP

27, 28, 144, 160

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

V(v,iv)salts; VO(acac)2; VO(cit)2; VO(salicylaldehyde semicarbazone)

Biological system

176

Published on 05 April 2019 on https://pubs.rsc.org | doi:

Table 7.1  (continued) Vanadium compound

Human cell lines

Antitumor effects

Metavanadate

Male Sprague–Dawley rats – induced hepatocarcinoma; Female Sprague–Dawley rats – induced mammary preneoplasia Mice with MDAY-­D2 tumors

Anti-­clastogenic; ↓inci↓8OHdGs, ssBs and/or dence; repair hyperplastic DPCs lesions; ↓tumor size

97, 100

↓Tumor growth; tumor suppression ↓Tumor progression; favorable pharmacokinetics; lack of toxicity

PTEN inhibition

149, 150

—

148

↓Number and size of hyper- — plastic nodules; ↑citP450 levels; ↓glutathione; ↓GST and GGT levels ↓Tumor growth rate; ↑sur— vival time; no toxic effects

156

Bis-­peroxovanadium analogs Metvan [VO(bis(4,7-­ dimethyl-­1,10-­ phenanthroline)) sulfato] V(v)-­1,2,5-­dihydroxy vit D3 V(iii)(cyst)

Mouse xenograft model of glioblastoma and breast cancer Rats (DEN induced hepatocarcinogenesis) Leiomyosarcoma-­induced Wistar rats

Molecular target

Reference

155, 163

a

Chapter 7

Abbreviations: acac: acetylacetonate; AKT: also known as PKB – protein kinase B; AP-­2: activating protein 2; Bax: apoptosis regulator Bax; Bcl-­2: B-­cell lymphoma 2-­family of proteins that regulate apoptosis; cyst: cysteine; cit: citrate; CDK: cyclin-­dependent kinase; DAPK: death associated protein kinase; dhp: 1,2-­dimethyl-­3-­hydroxy-­4(1H)-­pyridinonate; DPCs: DNA-­protein crosslinks; EAT cells: Ehrlich ascites tumor cells; ERK: extracellular signal-­regulated kinase; FADD: Fas-­associated protein with death domain; FAK: focal adhesion kinase; GLUT1: glucose transporter 1; GGT: gamma-­glutamyl transpeptidase; GS3K: glycogen synthase-­3-­kinase; GST: glutathione S-­transferase; HIF: hypoxia-­inducible factor; H-­ras: also known as transforming protein p21 – small G protein of the ras subfamily; JNK: c-­Jun N-­terminal kinase; LDHA: lactate dehydrogenase-­A; MDAY: highly malignant and metastatic tumor; MAPK: mitogen-­activated protein kinase; MMP: matrix metalloproteinases; mpp: 1-­methyl-­3-­hydroxy-­4(1H)-­pyridinonate; NF-­κB: nuclear factor kappa-­ light-­chain-­enhancer of activated B cells; nta: nitrilotriacetate; 8-­OHdGs: 8-­hydroxy-­2′-­deoxyguanosines (modified DNA guanines); p15INKB: cyclin-­ dependent kinase (CDK) inhibitor; p21: also known as cyclin-­dependent kinase inhibitor 1 or CDK-­interacting protein 1; p53: also known as TP53 – tumor protein p53; PAK: p21 activated kinase; PARP: poly (ADP-­ribose) polymerase; PI3K: phosphatidylinositol-­4,5-­bisphosphate 3-­kinase; PIWIL2: PIWI-­like protein 2; PKB: protein kinase B; ppp: 1-­phenyl-­2-­methyl-­3-­hydroxy-­4(1H)pyridinonate; PTC1: protein phosphatase 2C homolog 1 (or short name PP2C-­1); PTP: protein tyrosine phosphatase; PTEN: phosphatase and tensin homolog – tumor suppressor phosphatase; RET proto-­oncogene: “rearranged during transfectioni” – encodes a receptor tyrosine kinase for members of the glial cell line-­derived neurotrophic factor (GDNF); ROS: reactive oxygen species; salen: 2,2′-­ethylenebis(nitrilomethylidene)diphenol (or N,N′-­ethylenebis(salicylimine)); ssBs: single-­strand-­breaks; VEGF: vascular endothelial growth factor; Ψm: mithocondrial membrane potential.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

177

better comparison of the reported studies devoted to investigating the cellular anticancer properties of VCs and understanding ongoing and future antitumor strategies. The discovery of new molecular targets for these metal-­ based drugs remains a continuing challenge but could allow rational design of more effective therapeutic VCs. Cancer research over the last decade has identified several patterns that provide a basis to understand cancer biology. These hallmarks represent specific and complementary environments that support tumor growth, transformation in cancer and metastatic dissemination. These hallmarks include sustaining proliferative signaling, evading growth suppressors, resisting cell death, enabling replicative immortality, inducing angiogenesis and activating invasion and metastasis (Figure 7.3a).51 Protein kinases and phosphatases, by the processes of phosphorylation and dephosphorylation, regulate cellular functions such as cell proliferation, cell differentiation, survival,

Figure 7.3  (a)  Different hallmarks demonstrated to be involved in the development

and progression of cancer. Outstanding progress toward understanding the mechanisms underlying each hallmark have been achieved over the last decades.51 (b) Representation of some of the intracellular signaling pathways activated by the RTKs. The phosphorylation status and signaling activity of RTK is determined by the kinase activity of the RTKs and by the activity of PTPs (shaded blue). Several kinases (shaded yellow) are shown at the end of each pathway and play a major role in phosphorylating and changing the activity of several target proteins. After phosphorylation some of the target proteins are activated, whereas others are inhibited, resulting in complex cellular changes.

View Online

Chapter 7

178

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

27–29,45–47

adhesion, angiogenesis and mobility. The action of protein kinases and phosphatases plays a major role in cellular functions; however, its deregulation has been implicated in cancer, and thus represents an important target for the development of potential antitumor agents. These pathways are initiated upon activation of the receptor tyrosine kinases (RTKs), either through phosphorylation or dephosphorylation. The general representation of some of the intracellular signaling pathways activated by RTKs is depicted in Figure 7.3b.52,53 Many secreted growth factors and hormones act through RTKs such as: epithelial growth factor, platelet-­ derived growth factor, fibroblast growth factors, hepatocyte growth factor, insulin, insulin-­like growth factor-­1 (IGF-­1), vascular endothelial growth factor (VEGF), macrophage-­colony-­stimulating factor (M-­CGF) and neurotrophins, among others. RTKs exhibit intrinsic tyrosine kinase activity. The binding of a ligand activates the kinase domains which cross-­phosphorylate the two monomers of the dimeric receptor. Phosphorylation first occurs at a regulatory site which causes conformational changes that allow the kinase domain to phosphorylate other tyrosine residues in the receptor and also in signal transduction proteins. The phosphorylation status and signaling activity of RTKs is determined by a combination of the kinase activity of the RTK and by the activity of protein tyrosine phosphatases (PTPs). Several kinases exist at the end of each pathway and are responsible for phosphorylating different target proteins. After phosphorylation, some of the target proteins are activated, whereas others are inhibited when phosphorylated, resulting in cellular alterations. RTKs have been shown to be key regulators of normal cellular processes, while their deregulation has been associated with the development and progression of cancer. In addition, impaired downregulation of RTK activity by PTPs is also a mechanism that has been proposed to be involved in cancer.52,53 Below we will describe the action of vanadate and VCs on enzymes and two specific signaling pathways.

7.3  I nhibition of Phosphorylases and Kinases by Vanadate and Vanadium-­containing Compounds 7.3.1  P  hosphatases and Their Inhibition by Vanadate and Vanadium Compounds Phosphatases are a group of hydrolases (class EC3.1.3) that cleave organic phosphate bonds to form inorganic phosphate and an organic compound with a hydroxyl group. This group of enzymes is very diverse with regard to structure and substrate specificity. The sub-­group that contains a tyrosine in the active site is referred to as the protein tyrosine phosphatases (PTPases), and they are generally signal transduction and regulatory proteins in cells. PTPases are inhibited by vanadate and vanadium(iv) compounds although, in contrast to antidiabetic ex vivo and in vivo studies which demonstrate that vanadium(iv) compounds tend to be more potent,

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

179

X-­ray crystallography and immunotherapy applications favour vanadium(v) systems.1,2,4,9–11,17,29,41,42,47 Inhibition by vanadate is generally attributed to the vanadium being able to form the five-­coordinate transition state geometry or a high energy intermediate.1,9,11,54,55 The coordination environment around the vanadium can be VO4O, VO4N or VO4S depending on the phosphatase.1,9,10,54,55 In addition, there are some phosphatases that have metal ions in the active site and the mechanism will involve, for example, a metal-­bound OH group.56 A number of V–phosphatase X-­ray structures have been characterized and extensive studies have been carried out with these systems documenting the details of their reaction mechanism. The mechanism for the protein tyrosine phosphatase type 1B (abbreviated as PTP1B) is particularly well known because of its role in signal transduction and the fact that the antidiabetic effects of vanadate and other antidiabetic compounds are generally attributed to this phosphatase.54,55,57–61 In addition to the effect of vanadate, the Zn2+ cation is reported to bind with nanomolar affinities to PTP1B, facilitating a non-­trivial response by this phosphatase.32 Yersinia PTP1B is the only phosphatase for which an X-­ray structure exists, showing vanadate covalently coordinated to tyrosine in the PTP enzyme.1,9,54 A corresponding structure also exists for a phosphodiesterase (see Section 7.3.3). The specific mechanism involved in the dephosphorylation reaction varies depending on the amino acid nucleophile in the active site of the phosphatase. These subtle changes in the active site result in differences in the mechanisms of enzyme-­catalyzed reactions, ranging from dissociative, associative or concerted mechanisms as shown in Figure 7.4.54 The most well-­known mechanism is the associative one in which the nucleophile on the enzyme is coordinated to the vanadium in the first step, forming a five-­coordinate transition state complex (Figure 7.4, top). A number of phosphatases pass through a concerted mechanism involving an exploded transition state in which the transition state has longer bonds to the incoming nucleophile and the outgoing electrophile. Since vanadate is isoelectronic with phosphate its physical properties are similar and it can readily replace a phosphate when binding to the enzyme. Thus, the vanadate would take the place of the phosphate group that is being cleaved off, as shown in Figure 7.4. Since each phosphatase binds vanadate differently, there are small differences in the specific binding and mechanisms, resulting in varying affinities of vanadate for the phosphatase. X-­ray structures have been characterized for a number of different proteins, including protein–vanadate structures, and the systems have been compared in detail using data-­mining methods, as shown in Figure 7.1 for VO4S. Inhibition of enzymes using associative or concerted mechanisms is particularly relevant for the development of cancer because the regulatory phosphatases are crucial to the signal transduction pathways. The cellular regulatory mechanisms that govern cellular development and responses exhibit one of these mechanisms.9,54 This chemistry and these enzymes are therefore important for cancer progression.51,62,63

View Online

Chapter 7

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

180

Figure 7.4  The  three limiting enzyme-­catalyzed mechanisms for phosphate ester

hydrolysis: associative (top), concerted (middle) and dissociative (bottom). The addition of vanadate leads to replacement of the phosphate group that is being cleaved with vanadate, leading to formation of a stable complex. Adapted from ref. 9 with permission from American Chemical Society, Copyright 2015.

7.3.2  K  inases and Their Inhibition by Vanadate and Vanadium Compounds Enzymes of class EC2, called transferases, have also been suggested to be inhibited by vanadate.64,65 Results obtained from the cellular studies described in Section 7.3.1 suggest that kinases are also inhibited by vanadate.66,67 Upon isolation of some protein kinases insignificant inhibition was observed with the enzyme in purified form.9,66,68,69 Since kinases catalyze the phosphorylation reaction, which is the reverse of the dephosphorylation reaction catalyzed by phosphatases, the effects reported in cells could similarly be attributed to inhibition of phosphatases.

7.3.3  A  dditional Phosphorylases and Their Inhibition by Vanadate and Vanadium Compounds Additional related hydrolases such as Na,K-­ATPase (EC3.6.3), ATP synthase (EC3.6.3), Ca2+-­ATPase (E3.6.3)70 as well as phosphodiesterases (EC3.1.4) are also inhibited by vanadate. These enzymes are generally inhibited through mechanisms that are different from those described above for the phosphatases.71 For the Na,K-­ATPase, which is present in the cell membrane

View Online

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

181

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

+

and couples H transport with the hydrolysis of ATP, vanadate binds to the phosphate-­binding site, as a substrate analog, mimicking phosphate binding. As a result, the Na,K-­ATPase undergoes a conformational change which locks the protein, and the enzyme is no longer able to turn over.6,72 The inhibition of the ATP synthase (also referred to as the F1-­ATPase),73 and Ca2+-­ATPase,70 similarly involves binding of vanadate as a phosphate analog followed by conformational changes. Phosphodiesterase enzymes catalyze the breaking of the phosphodiester backbone of DNA or RNA. In addition to small molecule phosphodiesterases, this class of enzymes includes cAMP-­selective hydrolases and cGMP-­selective hydrolases such as phospholipases C and D, autotaxin, sphingomyelin phosphodiesterase, DNases, RNases and restriction endonucleases. Because these enzymes are very diverse and change from tissue to tissue they are popular pharmaceutical targets,74 as exemplified by common drugs used for purposes other than cancer such as sildenafil (currently used as Viagra),75 which is a PDE5 inhibitor, and Rolipram (previously used as an antidepressant), which is a PDE4 inhibitor.76 Examples of two phosphodiesterases implemented for protection against cancer development are the rod outer segment phosphodiesterase and tyrosyl-­DNA phosphodiesterase (Tdp1). The rod outer segment phosphodiesterase, interestingly, is activated by vanadate thus representing one of these enzymes in which the effect of vanadate does not lead to inhibition but to activation.77 Tdp1 is a DNA repair enzyme that catalyzes the hydrolysis of a phosphodiester bond between a tyrosine residue and a DNA 3′-­phosphate. Human Tdp1 is believed to be responsible for repairing lesions that occur when topoisomerase I stalls in DNA and for repair of free radical-­mediated DNA double-­strand breaks. The three-­dimensional X-­ray structures of human Tdp1 bound to vanadate have been reported, supporting a transient link between human type IB topoisomerase and DNA.78 This structure is related to the structure observed for the PTPase, PTP1B, the enzyme to which many of the observed antidiabetic effects is attributed.

7.3.4  M  itogen-­activated Protein Kinase and Phosphoinositide 3-­Kinase Signaling Pathways In addition to the isolated enzymes described above, other pathways have been identified and found to affect cancer progression that are sensitive to vanadium salts and VCs.79,80 Some of these cellular pathways were deduced based on cell responses to vanadate and VCs in a number of studies providing specific examples of the pathways relating to RTK summarized in Figure 7.3. These pathways are abbreviated as the RTK-­MAP kinase (mitogen-­activated protein kinase) (Figure 7.5) and the RTK-­PI3K pathway (phosphoinositide 3-­kinase pathway) and are shown in Figure 7.6. RTK-­MAP kinase and RTK-­PI3K pathways lead to protein activity and gene expression changes that facilitate cellular processes, such as cell proliferation, differentiation, survival, adhesion, angiogenesis and motility, and are

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

182

Chapter 7

Figure 7.5  General  overview of the RTK-­mitogen-­activated protein (MAP) kinase

signaling pathway. The signaling is initiated when a ligand binds to the extracellular ligand-­binding domain of two RTK subunits, bringing them together so that the RTKs are close enough to phosphorylate each other on tyrosine residues. Grb2 is an adaptor protein that binds both to phosphotyrosines on the RTK and to Son of sevenless (Sos). Once Sos is brought to the membrane, it causes the small G protein Ras to dissociate from GDP and bind to GTP, thereby becoming active. GTP-­bound Ras will bind to MAP-­KKK (Raf), activating it. Activated MAP-­KKK (Raf) activates MAP-­KK (Mek) by phosphorylation of select Ser/Thr residues. Activated MAP-­KK (Mek) activates MAP-­K (Erk). Activated MAP-­K (Erk) regulates the activation of many proteins by phosphorylation.

therefore involved in the development and progression of cancer. RTK-­MAP kinase and RTK-­PI3K pathways are cascades that can mediate both physiological and pathological responses in mammalian cells when deregulated, and they are the most studied and well characterized cascades in cancer cell biology. Administration of vanadate and VCs to cells elicits a cellular anticancer response. The RTK-­MAP kinase activates extracellular-­signal-­regulated kinases (ERKs) by phosphorylation, see Figure 7.5. The proteins are activated by dual phosphorylation on their regulatory Tyr/Thr residues located within the Thr–Xaa–Tyr motif.81 Upon phosphorylation of the regulatory Tyr/Thr

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

183

Figure 7.6  Schematic  representation of the phosphoinositide 3-­kinase (PI3K)

pathway. PI3K consists of the catalytic subunit, p110, and the regulatory subunit, p85. PI3K phosphorylates PIP2 (phosphatidylinositol 3,4-­bisphosphate) and produces PIP3 (phosphatidylinositol 3,4,5-­trisphosphate). PIP3 then activates 3-­phosphoinositide-­dependent kinase 1 (PDK1) and its major downstream effector, protein kinase B (PKB), also known as AKT. Phosphorylation of AKT leads to the phosphorylation of different cytosolic and nuclear target proteins. After phosphorylation some of the target proteins are activated, whereas others are inhibited, as can be observed in this figure. Stimulatory events are indicated by arrows and inhibitory events are indicated by lines ending in flat lines. Phosphatase and tensin homolog (PTEN) (shaded yellow) dephosphorylates PIP3 and inhibits activation of AKT. PTEN is inactive in different types of cancers.

residues, the ERK proteins phosphorylate a variety of substrates on their consensus sites, mostly Pro–Xaa–Ser/Thr–Pro,82 resulting in changes in protein activity and gene expression causing cellular changes. Inactivation of ERK proteins by treatment with vanadate or VCs is mediated by either tyrosine phosphatases (PTPs)83 or by protein serine/threonine phosphatases, such as PP2A.84 In general, PTPs are more potently affected by vanadate than protein serine/threonine phosphatases, so the former are presumed to have a greater impact with treatment by vanadate. Some of the substrates phosphorylated by ERKs are localized in the cytoplasm of the cell, while others are in the nucleus. ERKs phosphorylate and activate different transcription factors in the nucleus, such as Elk1,85 c-­Fos,86 p53 87 and c-­Jun,88 among others. ERKs mediate the phosphorylation and activation of the 90 kDa ribosomal S6 kinases (RSKs),89 which can independently translocate

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

184

Chapter 7

into the nucleus and phosphorylate other substrates. Additional substrates for ERKs are the mitogen-­ and stress-­activated kinase (MSK)90 and the MAPK-­interacting kinases (MNKs),91 and several apoptotic regulatory molecules, such as BCL2-­associated X protein (Bax), B-­cell lymphoma 2 (Bcl-­2), Bim, Bcl-­2-­associated death promoter (Bad) and myeloid cell leukemia 1 (Mcl-­1) (Figure 7.5). The RTK-­PI3K-­AKT pathway ultimately results in the activation of protein kinase B (PKB), also known as AKT, which leads to the phosphorylation of different cytosolic and nuclear target proteins. AKT activates and inhibits a large number of effectors of different cytosolic and nuclear target proteins via phosphorylation. AKT targets include mammalian rapamycin (mTOR), IκB kinase (IKK), mouse double minute 2 homolog (MDM2), BAD, p27, p21, glycogen synthase kinase‐3β (GSK3β), cAMP response element-­binding protein, the tumor suppressor genes, tuberous sclerosis complex 2 (TSC2) and the forkhead family of transcription factors (FOXO). As a result, the PI3K pathway promotes cell proliferation, survival, migration, differentiation and ribosome biogenesis and translation. AKT signaling has been found to be altered in cancer,75,92,93 and therefore it has been widely investigated as a target for anticancer therapies.94 The tumor suppressor, phosphatase and tensin homolog (PTEN) constitutes an important negative regulator of the PI3K-­AKT pathway, as it dephosphorylates phosphatidylinositol 3,4,5-­trisphosphate (PIP3) back to phosphatidylinositol 3,4-­bisphosphate (PIP2) rapidly and thus limits the downstream activation of PDK1 and AKT.95 The PTEN tumor suppressor is regularly found to be mutated, resulting in activity changes, in a wide range of neoplastic diseases.96 The presence of vanadate or VCs results in activity changes which, when induced in cells in this pathway, cause changes in cellular regulation.

7.4  T  he Effects of Vanadate and VCs in Cellular Systems Knowledge about where and how VCs act opens up avenues for a therapeutic strategy focused on their possible molecular targets, thus developing a rational design for improved medicinal efficacy of VCs and predicted structure– activity relationships. Studies have been conducted in vitro in a number of malignant cell lines to evaluate the effects of VCs on cell morphology, physiology and metabolism. Many studies have focused on vanadium as a regulator of cell growth, differentiation and death, these being the first evidence of V salts and VCs as potential anticancer agents. Cellular models have been of great importance in the discovery of the molecular mechanisms underlying the pharmacological actions of VCs. The inhibition of cell proliferation, decrease of cell viability, increase of cytotoxicity and/or genotoxicity, induction of apoptosis and increase in cell death, selectively in cancer cells, accounts for its antineoplastic activity (summarized in Table 7.1).

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

185

Vanadium(iv,v) salts (vanadyl and vanadate) were first investigated for their pharmacological properties, both in vitro and in vivo, and still continue to be studied in cellular models because of the need to compare them with the effects of simple salts, for speciation consideration.97–103 However, due to their poor gastrointestinal absorption, high doses in vivo are required, creating severe side effects. Thus, vanadium complexes with organic ligands have attracted much interest to overcome this problem, favoring improved properties relating to solubility, biodistribution, targeting and excretion. The effects of VCs on cancer cell proliferation and differentiation, cell cycle arrest and induction of apoptosis seems to be related to their direct interaction with DNA and/or key proteins/enzymes of intracellular signaling pathways. Potential signaling proteins are depicted in Figures 7.3, 7.5 and 7.6, and they induce the observed deregulation and tumorigenesis directly through inhibition of PTPases47,54,55,58–61,104 or indirectly through production of reactive oxygen species (ROS).105 Furthermore, reported epidemiological studies suggest that incidences of liver, pancreatic, and endometrial cancers are associated with diabetes even though the molecular mechanisms are not understood.23,57 Orthovanadate showed antiproliferative action and arrested the cell cycle,98,100–102,106 and the same was observed with vanadyl sulfate,107,108 in several cancer cell lines. VCs with organic molecules, like VIVO(2-­hydroxynaphthylaldimine), VIVO(dhp)2, VIV,V(galactomannan), IV IV V IV V O(chrysin)2, V O(Schiff bases), V (tpy)2, V O(acac)2, VIVO(alginato)2, VIVO(silibinin)2, VIVO(malato)2, VIVO(pyridoxal)2, VIVO(bis(4,7-­dimethyl-­ 1,10-­phenanthroline))2 (also designated as metvan) and VIVO(polypyridil) have shown anti-­proliferative properties50,103,109–122 by arresting the cell cycle, mainly in the G2/M and G1 phases, and thus increasing tumor cell death.101,102,106,110,122–129 Anticancer properties are recognized when cytotoxicity is higher in tumor cells compared with normal cells, discarding the toxic effects of the compounds.116,118,120,128,130,131 VCs induce apoptosis in cancer cells, another indication of antineoplastic activity.100–102,106,109–112,116,119,123,125,132–134 Evidence of apoptosis is demonstrated by molecular features such as upregulated expression of caspase 3 and other caspases,46,110–112,134 cytochrome c release from mitochondria103 and morphological changes in the cell such as atypical changes in the nucleus and membrane alterations.103,129,133,135 Apoptosis occurs as a consequence of several deleterious factors, one of them is ROS increase (favored in cancer cells, and produced by Fenton reaction), reactions with atmospheric oxygen or redox cycling between V(iv) and V(v). Thus, VCs can induce tumor cell apoptosis as a result of ROS production.49,103,109,111,112,114,115,120,122,128,130,132–134,136,137 Disruption of the mitochondrial membrane potential (ψm) was observed with some VCs,102,103,109,111,119,120,134 often resulting from ROS production, causing impairment of mitochondrial function109 and, ultimately, apoptosis.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

186

Chapter 7

In cancer cells the enzymatic antioxidant systems, like superoxide dismutase, catalase, glutathione peroxidase and heme oxygenase, show increased expression and activity.109,133,138 In contrast, the reduced glutathione (GSH)/oxidized glutathione (GSSG) ratio is decreased110–112 and protects the cells against ROS production. Interestingly, the antioxidant action of flavonoids (like polyphenols, curcumin, silibinin, chrysin, hesperidin) as free radical scavengers is improved when they are complexed with vanadium(iv). Therefore, under some conditions VCs act as anti-­stress agents, cancer chemo-­preventatives and chemo-­protectors to mitigate adverse effects of chemotherapy.133,139 The biological effects of VCs as anticancer or antitumor preventatives depends on their ligands, the complexation, the dose and the type of cells. Other contradictory effects of VCs were also demonstrated. In bone tissue and bone-­related cells some VCs show osteogenic activity (proliferation, differentiation and mineralization of the extracellular matrix) in normal bone cells (mice bone marrow cells, normal osteoblast cell line MC3T3-­E1)27,139,140 and others present antitumoral properties in osteosarcoma cell lines like G-­63 or UMR/06.46,103,111,112,121,141 Vanadium exerts antagonistic biochemical and pharmacological responses through a complicated interplay of mechanisms. Some VCs such as VIV(polypyridyl), VIVO(phenanthroline)2, metvan and others containing aromatic ligands interact directly with DNA23,103,124 by intercalation135,142 or through binding to phosphate ester residues.17 These interactions are responsible for DNA damage119,132 and/or fragmentation,49,111,134,143 ending up in cell apoptosis. Signaling cascades regulate cellular events by maintaining the normal metabolic and physiological conditions of the cell, including cell proliferation, differentiation and death (see Section 7.2). Abnormal behavior of specific transduction pathways deregulates these biological processes, causing pathological features, with the proteins/enzymes/transcription factors/ growth factors of these cascades being the most plausible targets for the direct action of V species. Studies have demonstrated that V(v) and V(iv) salts and some VCs such as VIVO(acac)2, VIVO(cit)2 and VIVO(saccharides) induce regulation of MAPK/ERK and PI3K/AKT cascades,122,128,144–146 as demonstrated by the changes in protein phosphorylation levels blocking cell cycle progress, and thus inhibiting tumor cells proliferation. PTPase inhibition (or PTP phosphorylase activation) by VCs99,113,145,146 seems to be the common molecular event to explain these effects. VCs and V salts also regulate the expression of key proteins of intracellular signaling. They promote the increase of cyclin D1 expression originating cell cycle arrest at S phase,106 inhibit the expression of P-­element-­induced wimpy testis (PIWIL2) resulting in apoptosis induction, decreased percentage of viable cells and promote cell cycle arrest at G2/M and S phases.101 It was suggested that protein tyrosine kinase 2β (PTK2B), focal adhesion kinase (FAK) and protein kinase C (PKC) families are targets for VIVO(chrys)2 antitumor activity.141 The expression of epidermal growth

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

187

factor receptor, phospholipase C (PLC) and Src are inhibited by V(v,iv) salts, as well as the MAPK cascades.146 In addition, they decrease cyclin D1, E and A145 and bcl-­2 100,125,129,132 expression, while increased expression of a cyclin-­dependent kinase inhibitor p21,145 p53 100,132 and bax100,125,129,132 was observed. VIVO(silibinin)2 110 and V(v,iv) salts146 reduce the activation of the nuclear factor kappa B (NF-­κB) transcription factor, thus increasing cell apoptosis. Vanadyl salt suppresses the expression of c-­fos proto-­oncogene which is associated with cell growth inhibition,99 whereas VIVO(acac)2 reduces phosphorylation of retinoblastoma tumor suppressor protein pRb, resulting in cell cycle arrest at G1/S phase.145 It was reported that VCs are inhibitors of topoisomerase I110,126,147 giving rise to DNA strand break and apoptosis. There is also evidence that they decrease small G protein in the Ras subfamily (H-­ras), MMP2 and 9 expression115,132,148 and inhibit the activity and expression of PTPases involved in proliferation, differentiation and migration of tumor cells.104 Peroxovanadates showed anti-­angiogenic activity, as well as preventing vascular restenosis by inhibiting endothelins and immunomodulator secretion, and modulate the behavior of adhesive molecules, thus demonstrating their anticarcinogenic and antimetastatic capacity.49

7.5  I n Vivo Studies of Vanadium Compounds in Animal Model Systems There are numerous studies in the literature dedicated to investigating the anticancer effects of VCs on a variety of tumor cell lines to elucidate the mechanism underlying their therapeutic properties, but only a few report animal studies. Vanadium salts, either in the form of vanadate(v) or vanadyl(iv) have been tested in rats and mice and exerted antineoplastic effects.97,98,100,102,103 These effects were evidenced by the inhibition of tumor cell growth in mice containing the highly metastatic MDAY murine lymphoma,98 repair of hyperplastic lesions and significant size reduction of mammary tumor induced in female Sprague–Dawley rats100 and anticlastogenic effects in diethylnitrosamine (DEN)-­induced hepatocarcinoma in Sprague–Dawley rats.97 To overcome the problem of administering the effective high doses of vanadium(iv,v) salts, in vivo studies with VCs such as bisperoxovanadium analogs,149–151 metvan,148 vanadocene dichloride,152–154 VIII(cyst)3 155 and VV(1,2,5-­dihydroxyvitD3)156 demonstrated that they decrease tumor growth, delay tumor growth/progression, prolong survival time in mouse xenograft models of human malignant glioblastoma and breast cancer, presenting favorable pharmacokinetics and no observed toxicity.148 A decrease in size and number of hyperplastic nodules, increased levels of hepatic microsomal cytochrome P450, decreased levels of GSH and decrease in glutathione S transferase (GST) and γ-­glutamyl transpeptidase (GGT) activity were also observed in hepatocarcinomatous-­ induced rats.156

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

188

Chapter 7

A novel concept for combatting cancer reported the enhancement of oncolytic viruses by vanadate, vanadyl sulfate and some VCs.29 These oncolytic viruses induce antitumor immunity through selective replication in a range of resistant tumor cells. However, the presence of vanadate will enhance oncolytic virus infection in vitro and ex vivo in resistant tumor cell lines. This was evidenced by measuring titer count as a means of assessing viral replication in vitro. The ex vivo evidence was obtained by measuring the expression of green fluorescent protein and its associated fluorescence. Furthermore, in vivo studies in several syngeneic tumor models demonstrated a systemic, durable response exhibiting T-­cell dependence after injection of a suspension of combined virus and vanadate. Overall, these studies demonstrate the ability of vanadium compounds to enhance viral oncolysis and systematic anticancer immunity, facilitating development of new and improved strategies for immunotherapy.29

7.6  Conclusions Several excellent reviews describing the anticancer effect of vanadium have been reported.2,3,10,17,21,34,48–50,66 Here we focus on associating the enzyme inhibition with the observed anticancer effects. Studies investigating the antidiabetic effects of V salts and VCs have been emerging, showing some overlap with the molecular mechanisms of anticancer agents,23,34,122 suggesting that the reported lack of VC specificity and selectivity leading to adverse side effects can be overcome. In addition, the recent report demonstrating potentiation of oncolytic virotherapy by vanadium compounds suggests that there is some potential for vanadium compounds in the future.29 However, continued studies into the mechanism of action of various systems remains of the utmost importance in order to properly evaluate whether potential VCs are possible future therapeutic anticancer agents.

Acknowledgements The authors thank Colorado State University, Universitária Vasco da Gama and University of Coimbra for support. The authors also thank Prof. Craig McLauchlan for helping us to prepare Figure 7.1.

References 1. C. C. McLauchlan, B. J. Peters, G. R. Willsky and D. C. Crans, Coord. Chem. Rev., 2015, 301–302, 163. 2. J. C. Pessoa, S. Etcheverry and D. Gambino, Coord. Chem. Rev., 2015, 301, 24. 3. C. Gabriel, E. Kioseoglou, J. Venetis, V. Psycharis, C. Raptopoulou, A. Terzis, G. Voyiatzis, M. Bertmer, C. Mateescu and A. Salifoglou, Inorg. Chem., 2012, 51, 6056.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

189

4. G. R. Willsky, L.-­H. Chi, M. Godzala III, P. J. Kostyniak, J. J. Smee, A. M. Trujillo, J. A. Alfano, W. Ding, Z. Hu and D. C. Crans, Coord. Chem. Rev., 2011, 255, 2258. 5. L. C. Cantley and P. Aisen, J. Biol. Chem., 1979, 254, 1781. 6. L. C. Cantley, L. G. Cantley and L. Josephson, J. Biol. Chem., 1978, 253, 7361. 7. P. North and R. L. Post, J. Biol. Chem., 1984, 259, 4971. 8. R. N. Lindquist, J. L. Lynn and G. E. Lienhard, J. Am. Chem. Soc., 1973, 95, 8762. 9. D. C. Crans, J. Org. Chem., 2015, 80, 11899. 10. L. P. Lu and M. L. Zhu, Anti-­Cancer Agents Med. Chem., 2011, 11, 164. 11. D. C. Crans, M. L. Tarlton and C. C. McLauchlan, Eur. J. Inorg. Chem., 2014, 4450. 12. D. C. Crans, J. J. Smee, E. Gaidamauskas and L. Yang, Chem. Rev., 2004, 104, 849. 13. D. Rehder, Bioinorganic Vanadium Chemistry, John Wiley & Sons, Ltd, 2008. 14. D. Rehder, M. Bashirpoor, S. Jantzen, M. Schmidt, M. Farahbakhsh and H. Nekola, in Vanadium Compounds Chemistry, Biochemistry, and Therapeutic Applications, American Chemical Society, Michigan, 1st edn, 1998, ch. 4, vol. 711, p. 60. 15. K. Elvingson, A. D. Keramidas, D. C. Crans and L. Pettersson, J. Am. Chem. Soc., 1998, 119, 7005. 16. K. A. Doucette, K. N. Hassell and D. C. Crans, J. Inorg. Biochem., 2016, 165, 56. 17. D. Rehder, Future Med. Chem., 2016, 8, 325. 18. D. Rehder, Metallomics, 2015, 7, 730. 19. K. Gruzewska, A. Michno, T. Pawelczyk and H. Bielarczyk, J. Physiol. Pharmacol., 2014, 65, 603. 20. D. Rehder, Met. Ions Life Sci., 2013, 13, 139. 21. D. A. Barrio and S. B. Etcheverry, Curr. Med. Chem., 2010, 17, 3632. 22. J. Korbecki, I. Baranowska-­Bosiacka, I. Gutowska and D. Chlubek, Acta Biochim. Pol., 2012, 59, 195. 23. X.-­G. Yang and K. Wang, Curr. Top. Med. Chem., 2016, 16, 675. 24. D. Rehder, Future Med. Chem., 2012, 4, 1823. 25. M. S. Bhuiyan and K. Fukunaga, J. Pharmacol. Sci., 2009, 110, 1. 26. N. Shioda, F. Han and K. Fukunaga, Int. Rev. Neurobiol., 2009, 85, 375. 27. D. A. Barrio and S. B. Etcheverry, Can. J. Physiol. Pharmacol., 2006, 84, 677. 28. A. M. Cortizo, M. S. Molinuevo, D. A. Barrio and L. Bruzzone, Int. J. Biochem. Cell Biol., 2006, 38, 1171. 29. M. Selman, C. Rousso, A. Bergeron, H. H. Son, R. Krishnan, N. A. El-­ Sayes, O. Varette, A. Chen, F. Le Boeuf, F. Tzelepis, J. C. Bell, D. C. Crans and J. S. Diallo, Mol. Ther., 2018, 26, 56. 30. D. C. Crans, K. A. Woll, K. Prusinskas, M. D. Johnson and E. Norkus, Inorg. Chem., 2013, 52, 12262.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

190

Chapter 7

31. Y. Yoshikawa, H. Sakurai, D. C. Crans, G. Micera and E. Garribba, Dalton Trans., 2014, 43, 6965. 32. E. Bellomo, K. Birla Singh, A. Massarotti, C. Hogstrand and W. Maret, Coord. Chem. Rev., 2016, 327–328, 70–83. 33. A. Levina, D. C. Crans and P. Lay, Coord. Chem. Rev., 2017, 352, 473. 34. D. Crans, L. Yang, A. Haase and X. Yang, Met. Ions Life Sci., 2018, 18, 251. 35. N. D. Chasteen, Struct. Bonding, 1983, 53, 105. 36. N. D. Chasteen, R. J. DeKoch, B. L. Rogers and M. W. Hanna, J. Am. Chem. Soc., 1973, 95, 1301. 37. D. C. Crans, E. M. Willging and S. K. Butler, J. Am. Chem. Soc., 1990, 165, 56. 38. N. Samart, Z. Arhouma, S. Kumar, H. A. Murakami, D. C. Crick and D. C. Crans, Front. Chem., 2018, 6, 519. 39. D. C. Crans, B. Zhang, E. Gaidamauskas, A. D. Keramidas, G. R. Willsky and C. R. Roberts, Inorg. Chem., 2010, 49, 4245. 40. K. H. Thompson, J. Lichter, C. LeBel, M. C. Scaife, J. H. McNeill and C. Orvig, J. Inorg. Biochem., 2009, 103, 554. 41. K. H. Thompson and C. Orvig, J. Inorg. Biochem., 2006, 100, 1925. 42. K. H. Thompson and O. Chris, Coord. Chem. Rev., 2001, 219–221, 1033. 43. D. A. Barrio, P. A. Williams, A. M. Cortizo and S. B. Etcheverry, J. Biol. Inorg. Chem., 2003, 8, 459. 44. V. Conte and G. Licini, Coord. Chem. Rev., 2015, 301–302, 1. 45. W. Jiang, G. Li, W. Li, P. Wang, P. Xiu, X. Jiang, B. Liu, X. Sun and H. Jiang, Sci. Rep., 2018, 8, 9706. 46. I. E. Leon, P. Diez, E. J. Baran, S. B. Etcheverry and M. Fuentes, Metallomics, 2017, 9, 891. 47. T. Scior, J. A. Guevara-­Garcia, Q. T. Do, P. Bernard and S. Laufer, Curr. Med. Chem., 2016, 23, 2874. 48. A. Bishayee, A. Waghray, M. A. Patel and M. Chatterjee, Cancer Lett., 2010, 294, 1. 49. A. M. Evangelou, Crit. Rev. Oncog. Hematol., 2002, 42, 249. 50. E. Kioseoglou, S. Petanidis, C. Gabriel and A. Salifoglou, Coord. Chem. Rev., 2015, 301–302, 87. 51. D. Hanahan and R. A. Weinberg, Cell, 2011, 144, 646. 52. J. Avruch, Biochim. Biophys. Acta, 2007, 1773, 1150. 53. G. Pearson, F. Robinson, T. Beers Gibson, B. E. Xu, M. Karandikar, K. Berman and M. H. Cobb, Endocr. Rev., 2001, 22, 153. 54. T. A. S. Brandao, A. C. Hengge and S. J. Johnson, J. Biol. Chem., 2010, 285, 15874. 55. T. A. S. Brandão, S. J. Johnson and A. C. Hengge, Arch. Biochem. Biophys., 2012, 525, 53. 56. V. Lopez, T. Stevens and R. N. Lindquist, Arch. Biochem. Biophys., 1976, 175, 31. 57. P. Vigneri, F. Frasca, L. Sciacca, G. Pandini and R. Vigneri, Endocr.-­Relat. Cancer, 2009, 16, 1103. 58. X. Niu, R. Xiao, N. Wang, Z. Wang, Y. Zhang, Q. Xia and X. Yang, Curr. Top. Med. Chem., 2016, 16, 811.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects



191

59. K. Fukunaga, Saibo Kogaku, 2011, 30, 631. 60. P. Heneberg, Curr. Med. Chem., 2009, 16, 706. 61. F. Nxumalo, N. R. Glover and A. S. Tracey, J. Biol. Inorg. Chem., 1998, 3, 534. 62. T. Motiwala and S. T. Jacob, Prog. Nucleic Acid Res. Mol. Biol., 2006, 81, 297. 63. C. M. Krejsa and G. L. Schieven, Environ. Health Perspect., 1998, 106(Suppl 5), 1179. 64. J. Korbecki, I. Baranowska-­Bosiacka, I. Gutowska and D. Chlubek, Int. J. Mol. Sci., 2015, 16, 12648. 65. A. B. Goldfine, D. C. Simonson, F. Folli, E. Patti and C. R. Kahn, Mol. Cell. Biochem., 1995, 153, 217. 66. A. Bishayee, A. Waghray, M. A. Patel and M. Chatterjee, Cancer Lett., 2010, 294, 1. 67. N. Sekar, J. P. Li and Y. Shechter, Crit. Rev. Biochem. Mol. Biol., 1996, 31, 339. 68. S. Pugazhenthi, J. F. Angel and R. L. Khandelwal, Metab., Clin. Exp., 1991, 40, 941. 69. H. Faneca, V. Figueiredo, I. Tomaz, G. Gonçalves, F. Avecilla, M. P. de Lima, C. F. Geraldes, J. C. Pessoa and M. M. C. Castro, J. Inorg. Biochem., 2009, 103, 601. 70. M. Aureliano, F. Henao, T. Tiago, R. O. Duarte, J. J. G. Moura, B. Baruah and D. C. Crans, Inorg. Chem., 2008, 47, 5677. 71. Z. Yuan, T. Cai, J. Tian, A. V. Ivanov, D. R. Giovannucci and Z. Xie, Mol. Biol. Cell, 2005, 16, 4034. 72. R. Bruck, Z. Halpern, H. Aeed, Y. Shechter and S. J. Karlish, Pflugers Arch., 1998, 435, 610. 73. M. A. Biachet, J. Hullihen, P. L. Pedersen and L. M. Amzel, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 11065. 74. B. Weiss and W. N. Hait, Annu. Rev. Pharmacol. Toxicol., 1977, 17, 441. 75. K. M. Nicholson and N. G. Anderson, Cell. Signalling, 2002, 14, 381. 76. J. Zhu, E. Mix and B. Winblad, CNS Drug Rev., 2001, 7, 387. 77. L. V. Del Priore and A. Lewis, Biochim. Biophys. Acta, 1985, 845, 81. 78. D. R. Davies, H. Interthal, J. J. Champoux and W. G. J. Hol, J. Mol. Biol., 2002, 324, 917. 79. A. Zorzano, M. Palacin, L. Marti and S. Garcia-­Vicente, J. Inorg. Biochem., 2009, 103, 559. 80. M. J. Pereira, E. Carvalho, J. W. Eriksson, D. C. Crans and M. Aureliano, J. Inorg. Biochem., 2009, 103, 1687. 81. D. M. Payne, A. J. Rossomando, P. Martino, A. K. Erickson, J. H. Her, J. Shabanowitz, D. F. Hunt, M. J. Weber and T. W. Sturgill, EMBO J., 1991, 10, 885. 82. F. A. Gonzalez, D. L. Raden and R. J. Davis, J. Biol. Chem., 1991, 266, 22159. 83. R. Pulido, A. Zuniga and A. Ullrich, EMBO J., 1998, 17, 7337. 84. D. R. Alessi, N. Gomez, G. Moorhead, T. Lewis, S. M. Keyse and P. Cohen, Curr. Biol., 1995, 5, 283. 85. H. Gille, A. D. Sharrocks and P. E. Shaw, Nature, 1992, 358, 414.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

192

Chapter 7

86. L. O. Murphy, S. Smith, R. H. Chen, D. C. Fingar and J. Blenis, Nat. Cell Biol., 2002, 4, 556. 87. D. M. Milne, D. G. Campbell, F. B. Caudwell and D. W. Meek, J. Biol. Chem., 1994, 269, 9253. 88. S. Morton, R. J. Davis, A. McLaren and P. Cohen, EMBO J., 2003, 22, 3876. 89. T. W. Sturgill, L. B. Ray, E. Erikson and J. L. Maller, Nature, 1988, 334, 715. 90. M. Deak, A. D. Clifton, L. M. Lucocq and D. R. Alessi, EMBO J., 1998, 17, 4426. 91. A. J. Waskiewicz, A. Flynn, C. G. Proud and J. A. Cooper, EMBO J., 1997, 16, 1909. 92. I. Hers, E. E. Vincent and J. M. Tavare, Cell. Signalling, 2011, 23, 1515. 93. P. S. Mundi, J. Sachdev, C. McCourt and K. Kalinsky, Br. J. Clin. Pharmacol., 2016, 82, 943. 94. L. Willems, J. Tamburini, N. Chapuis, C. Lacombe, P. Mayeux and D. Bouscary, Curr. Oncol. Rep., 2012, 14, 129. 95. C. Blanco-­Aparicio, O. Renner, J. F. Leal and A. Carnero, Carcinogenesis, 2007, 28, 1379. 96. D. Bonneau and M. Longy, Hum. Mutat., 2000, 16, 109. 97. T. Chakraborty, N. Pandey, A. Chatterjee, B. Ghosh, B. Rana and M. Chatterjee, Mutat. Res., 2006, 609, 117. 98. T. F. Cruz, A. Morgan and W. Min, Mol. Cell. Biochem., 1995, 153, 161. 99. A. Evangelou, E. Kolettas, M. Tenopoulou, D. Galaris, E. S. Gonos and G. Manos, Met. Ions Biol. Med., 2002, 7, 154. 100. R. S. Ray, B. Ghosh, A. Rana and M. Chatterjee, Int. J. Cancer, 2007, 120, 13. 101. X. Tian, J. Fan, W. Hou, S. Bai, Q. Ao and H. Tong, Mol. Med. Rep., 2016, 13, 874. 102. Y. Wu, Y. Ma, Z. Xu, D. Wang, B. Zhao, H. Pan, J. Wang, D. Xu, X. Zhao, S. Pan, L. Liu, W. Dai and H. Jiang, Cancer Lett., 2014, 351, 108. 103. Y. Zhao, L. Ye, H. Liu, Q. Xia, Y. Zhang, X. Yang and K. Wang, J. Inorg. Biochem., 2010, 104, 371. 104. E. Irving and A. W. Stoker, Molecules, 2017, 22, 2269. 105. C. M. Krejsa and G. L. Schieven, Environ. Health Perspect. Suppl., 1998, 106, 1179. 106. J. Yang, Z. Zhang, S. Jiang, M. Zhang, J. Lu, L. Huang, T. Zhang, K. Gong, S. Yan, Z. Yang and G. Shao, J. Toxicol. Environ. Health A, 2016, 79, 864. 107. P. Holko, J. Ligeza, J. Kisielewska, A. M. Kordowiak and A. Klein, Pol. J. Pathol., 2008, 59, 3. 108. A. M. Kordowiak, A. Klein, A. Goc and W. Dabros, Pol. J. Pathol., 2007, 58, 51. 109. M. M. Cunha-­de Padua, S. M. Suter Correia Cadena, C. L. de Oliveira Petkowicz, G. R. Martinez, M. E. Merlin Rocha, A. L. R. Merce and G. R. Noleto, Carbohydr. Polym., 2017, 173, 665. 110. I. E. Leon, J. F. Cadavid-­Vargas, I. Tiscornia, V. Porro, S. Castelli, P. Katkar, A. Desideri, M. Bollati-­Fogolin and S. B. Etcheverry, J. Biol. Inorg. Chem., 2015, 20, 1175.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

193

111. I. E. Leon, A. L. Di Virgilio, V. Porro, C. I. Muglia, L. G. Naso, P. A. Williams, M. Bollati-­Fogolin and S. B. Etcheverry, Dalton Trans., 2013, 42, 11868. 112. I. E. Leon, V. Porro, A. L. Di Virgilio, L. G. Naso, P. A. Williams, M. Bollati-­ Fogolin and S. B. Etcheverry, J. Biol. Inorg. Chem., 2014, 19, 59. 113. S. Liu, G. Liu and Y. Yi, Carbohydr. Polym., 2015, 121, 86. 114. L. G. Naso, E. G. Ferrer, N. Butenko, I. Cavaco, L. Lezama, T. Rojo, S. B. Etcheverry and P. A. Williams, J. Biol. Inorg. Chem., 2011, 16, 653. 115. S. Petanidis, E. Kioseoglou, M. Hadzopoulou-­Cladaras and A. Salifoglou, Cancer Lett., 2013, 335, 387. 116. C. Rozzo, D. Sanna, E. Garribba, M. Serra, A. Cantara, G. Palmieri and M. Pisano, J. Inorg. Biochem., 2017, 174, 14. 117. G. Scalese, M. F. Mosquillo, S. Rostan, J. Castiglioni, I. Alho, L. Perez, I. Correia, F. Marques, J. Costa Pessoa and D. Gambino, J. Inorg. Biochem., 2017, 175, 154. 118. S. Sgarbossa, E. Diana, D. Marabello, A. Deagostino, S. Cadamuro, A. Barge, E. Laurenti, M. Gallicchio, V. Boscaro and E. Ghibaudi, J. Inorg. Biochem., 2013, 128, 26. 119. A. Sinha, K. Banerjee, A. Banerjee, A. Sarkar, M. Ahir, A. Adhikary, M. Chatterjee and S. K. Choudhuri, Biomed. Pharmacother., 2017, 92, 509. 120. M. Strianese, A. Basile, A. Mazzone, S. Morello, M. C. Turco and C. Pellecchia, J. Cell. Physiol., 2013, 228, 2202. 121. A. Tesmar, D. Wyrzykowski, R. Kruszynski, K. Niska, I. Inkielewicz-­ Stepniak, J. Drzezdzon, D. Jacewicz and L. Chmurzynski, BioMetals, 2017, 30, 261. 122. J. X. Wu, Y. H. Hong and X. G. Yang, J. Biol. Inorg. Chem., 2016, 21, 919. 123. J. Aubrecht, R. K. Narla, P. Ghosh, J. Stanek and F. M. Uckun, Toxicol. Appl. Pharmacol., 1999, 154, 228. 124. J. Benitez, L. Becco, I. Correia, S. M. Leal, H. Guiset, J. C. Pessoa, J. Lorenzo, S. Tanco, P. Escobar, V. Moreno, B. Garat and D. Gambino, J. Inorg. Biochem., 2011, 105, 303. 125. A. Meshkini and R. Yazdanparast, Exp. Mol. Pathol., 2010, 89, 334. 126. X. M. Mo, Z. F. Chen, X. Qi, Y. T. Li and J. Li, Bioinorg. Chem. Appl., 2012, 2012, 756374. 127. C. S. Navara, A. Benyumov, A. Vassilev, R. K. Narla, P. Ghosh and F. M. Uckun, Anti-­Cancer Drugs, 2001, 12, 369. 128. Q. Wang, T. T. Liu, Y. Fu, K. Wang and X. G. Yang, J. Biol. Inorg. Chem., 2010, 15, 1087. 129. J. Y. Yang and Z. R. Wang, Arch. Pharm. Res., 2006, 29, 859. 130. S. Kowalski, S. Hac, D. Wyrzykowski, A. Zauszkiewicz-­Pawlak and I. Inkielewicz-­Stepniak, Oncotarget, 2017, 8, 60324. 131. N. A. Lewis, F. Liu, L. Seymour, A. Magnusen, T. R. Erves, J. F. Arca, F. A. Beckford, R. Venkatraman, A. Gonzalez-­Sarrias, F. R. Fronczek, D. G. Vanderveer, N. P. Seeram, A. Liu, W. L. Jarrett and A. A. Holder, Eur. J. Inorg. Chem., 2012, 2012, 664.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

194

Chapter 7

132. A. Basu, A. Bhattacharjee, P. Ghosh, A. Samanta and S. Bhattacharya, Biomed. Pharmacother., 2017, 93, 816. 133. S. B. Etcheverry, E. G. Ferrer, L. Naso, J. Rivadeneira, V. Salinas and P. A. Williams, J. Biol. Inorg. Chem., 2008, 13, 435. 134. A. P. Goncalves, A. Videira, P. Soares and V. Maximo, Life Sci., 2011, 89, 371. 135. G. Scalese, I. Correia, J. Benitez, S. Rostan, F. Marques, F. Mendes, A. P. Matos, J. Costa Pessoa and D. Gambino, J. Inorg. Biochem., 2017, 166, 162. 136. S. Matsugo, H. Sugiyama, Y. Nishimoto, H. Misu, T. Takamura, S. Kaneko, Y. Kubo, R. Saito and K. Kanamori, Inorg. Chim. Acta, 2014, 420, 53. 137. R. K. Narla, Y. Dong, P. Ghosh, K. Thoen and F. M. Uckun, Drugs Future, 2000, 25, 1053. 138. A. Hamidi, L. Hassani, F. Mohammadi, P. Jahangoshayi and K. Mohammadi, J. Enzyme Inhib. Med. Chem., 2016, 31, 1124. 139. A. Basu, P. Ghosh, A. Bhattacharjee, A. R. Patra and S. Bhattacharya, Mutagenesis, 2015, 30, 509. 140. S. B. Etcheverry and D. A. Barrio, ACS Symp. Ser., 2007, 974, 204. 141. I. E. Leon, P. Diez, S. B. Etcheverry and M. Fuentes, Metallomics, 2016, 8, 739. 142. L. H. Abdel-­Rahman, A. M. Abu-­Dief, M. O. Aboelez and A. A. Hassan Abdel-­Mawgoud, J. Photochem. Photobiol., B, 2017, 170, 271. 143. R. Paulpandiyan and N. Raman, Bioorg. Chem., 2017, 73, 100. 144. D. A. Barrio, E. R. Cattaneo, M. C. Apezteguia and S. B. Etcheverry, Can. J. Physiol. Pharmacol., 2006, 84, 765. 145. Y. Fu, Q. Wang, X. G. Yang, X. D. Yang and K. Wang, J. Biol. Inorg. Chem., 2008, 13, 1001. 146. J. Korbecki, I. Baranowska-­Bosiacka, I. Gutowska and D. Chlubek, Int. J. Mol. Sci., 2015, 16, 12648. 147. K. Gokduman, Curr. Drug Targets, 2016, 17, 1928. 148. O. J. D'Cruz and F. M. Uckun, Expert Opin. Invest. Drugs, 2002, 11, 1829. 149. A. C. Schmid, R. D. Byrne, R. Vilar and R. Woscholski, FEBS Lett., 2004, 566, 35. 150. P. J. Scrivens, M. A. Alaoui-­Jamali, G. Giannini, T. Wang, M. Loignon, G. Batist and V. A. Sandor, Mol. Cancer Ther., 2003, 2, 1053. 151. C. L. Walker, M. J. Walker, N. K. Liu, E. C. Risberg, X. Gao, J. Chen and X. M. Xu, PLoS One, 2012, 7, e30012. 152. P. Kopf-­Maier and P. Funke-­Kaiser, Toxicology, 1986, 38, 81. 153. P. Kopf-­Maier and D. Krahl, Chem.-­Biol. Interact., 1983, 44, 317. 154. P. Kopf-­Maier, W. Wagner and H. Kopf, Cancer Chemother. Pharmacol., 1981, 5, 237. 155. R. Liasko, A. Themistoclis, S. Karkabounas, M. Malamas, A. J. Tasiopoulos, D. Stefanou, P. Collery and A. Evangelou, Anticancer Res., 1998, 18, 3609.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00169

Vanadium Compounds as Enzyme Inhibitors with a Focus on Anticancer Effects

195

156. R. Basak, M. Basu and M. Chatterjee, Chem.-­Biol. Interact., 2000, 128, 1. 157. H. Du, J. Xiang, Y. Zhang, Y. Tang and G. Xu, J. Inorg. Biochem., 2008, 102, 146. 158. I. E. Leon, J. F. Cadavid-­Vargas, A. Resasco, F. Maschi, M. A. Ayala, C. Carbone and S. B. Etcheverry, J. Biol. Inorg. Chem., 2016, 21, 1009. 159. L. G. Naso, L. Lezama, M. Valcarcel, C. Salado, P. Villace, D. Kortazar, E. G. Ferrer and P. A. Williams, J. Inorg. Biochem., 2016, 157, 80. 160. M. S. Molinuevo, D. A. Barrio, A. M. Cortizo and S. B. Etcheverry, Cancer Chemother. Pharmacol., 2004, 53, 163. 161. P. Kopf-­Maier, W. Wagner and E. Liss, J. Cancer Res. Clin. Oncol., 1983, 106, 44. 162. D. Sanna, M. Serra, V. Ugone, L. Manca, M. Pirastru, P. Buglyo, L. Biro, G. Micera and E. Garribba, Metallomics, 2016, 8, 532. 163. A. Evangelou, S. Karkabounas, G. Kalpouzos, M. Malamas, R. Liasko, D. Stefanou, A. T. Vlahos and T. A. Kabanos, Cancer Lett., 1997, 119, 221.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Chapter 8

Arsenic-­based Anticancer Agents Stéphane Gibaud EA3452 Cibles thérapeutiques, formulation et Evaluation Préclinique du Médicament (CITHEFOR), Université de Lorraine, Faculté de Pharmacie, 5, rue Albert Lebrun, 54000 Nancy, France *E-­mail: stephane.gibaud@univ-­lorraine.fr

8.1  I ntroduction and General Overview of Arsenic Anticancer Drugs In western countries, arsenic was initially proposed for the treatment of infectious diseases. The history of the treatment of malaria, fevers, trypanosomiasis, syphilis and other bacteria and parasites has been described in a previous review.1 The first use of arsenic in the treatment of cancer likely coincides with the first accurate descriptions that have been made of cancer cells in the 17th century. In 1664, Pierre Aliot published an opuscule on how to treat cancer “without surgery or a red-­hot iron” (Nuntius profligate sine ferro et igne carcinomatis, missus, ducibus itineris Hippocrato et Galeno ad chirurgiae studiosos). The first edition of this article has not been found, but it was republished in the Traité du cancer printed in 1698 in Paris. Pierre Aliot described clinical observations on cancer. He noted that if the cancer (i.e., the tumor) is removed by surgery, the patient is not cured. In contrast, the patient experienced many clinical complications and extreme pain. In fact, Pierre Aliot   Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

196

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

197

was a famous physician in Lorraine (which now belongs to France) who was called to treat Queen Anne of Austria for a breast cancer. Pierre Aliot's “recipe” was presented as follows:    “Take a pound of fine realgar powder; pour the powder into a large flask and add a very strong lexivial until it creates a supernatant of about four fingers width. Place the vessel in a sand bath for twenty-­four hours at a quite hot temperature. Pour the extract solution into another vessel and repeat the extraction with the same amount of lexivial on the remaining product; at the same time, frequently shake the flask. Add this extract solution to the first one and pour another lexivial on the remaining realgar to finish the extraction; also add this extract in the other vessel. You will repeat this operation until the realgar is almost completely dissolved. I say ‘almost’ since it will always remain a metallic powder that cannot be dissolved in the alkali. Filter all the solutions through a paper-­stoup in a vessel; start the precipitation by adding Vinegar of Saturn until a precipitation occurs at the bottom. Let the raw material stand for ten to twelve hours; afterwards, pour out the contents of the vessel by tilting it and remove all the useless solution. Prepare ten to fifteen lotions with this powder with warm water: the more you wash it, the better it will be. Once the last wash is done, dry the material, calcine it by burning highly purified spirit of wine. By the end, instead of pure spirit of wine, it is possible to burn a filtrated opium-­ laden spirit of wine.”    Pierre Aliot uses realgar (As4S4) to prepare this remedy. Realgar is an arsenic sulfide mineral that is also known as “ruby of arsenic” or “ruby sulfur”. It is orange-­red in color, melts at 320 °C, and burns with a bluish flame releasing fumes of arsenic and sulfur. The recipe of Pierre Aliot is not very accurate, but a paper published in 2004 reported the main components that can be found in alkali extracts of realgar.2 For example, an alkali extract contained 85% sodium arsenite (AsIII) and 15% sodium arsenate (AsV) in 0.1 M NaOH. The addition of “vinegar of Saturn” allowed the formation of lead arsenide and lead arsenate. The application of the product led to the mortification of the tissue that can be removed with a blade. Based on the theories of Hippocrates and Galen, the author qualified his remedy of “excoriating and adsorbing”. In fact, this adventurous application led only to tissue necrosis, and we can imagine now how dangerous it was to the patient. In the 18th century, arsenic was still considered a very dangerous compound, but Thomas Fowler braved the danger when he reintroduced arsenous acid as a solution. Arsenic trioxide was very poorly soluble in water, but it was possible to obtain a “ready to use” solution at higher pH with potassium bicarbonate (i.e., potassium arsenite). The route of administration changed from generally external to oral. In his medical reports on “the effects of arsenic in the cure of agues, remittent fevers and periodic headaches” (1786),3 Fowler explained how he tried to imitate the “tasteless ague drop” that was prescribed in the Stanford

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

198

Chapter 8

Hospital. This “solutio mineralis” became the well-­known “Fowler's solution” and was used during years as a substitute for ague drops and costly Peruvian bark. The therapeutic indication had nothing to do with cancer. In fact, arsenic returned to the field of cancer only a century later based on the results Cutler and Bradford demonstrating its effect on leukemia.4 Fowler's solution was then used to treat chronic myelogenous leukemia but was discarded again with the advent of cytotoxic chemotherapy in the 20th century. It was then considered more carcinogenic than a therapeutic treatment! Nevertheless, in the 1970s, a group of scientists at the Harbin Medical Institute (China) reported encouraging results in the treatment of malignant lymphoma (e.g., acute promyelocytic leukemia (APL) and chronic myeloid leukemia), lymphoma and esophageal carcinoma.5 The administered drug was a solution of As2O3 with a trace amount of mercury and was called Ailing-­1 (i.e., old Chinese medicine). Ailing-­1 induced complete remissions in two-­thirds of APL patients.5 When these impressive results were noted in western countries, they were confirmed with pure As2O3 in APL. Hence, arsenic treatment has re-­emerged in the west, and the results of clinical trials using arsenic-­based drugs in cancer have been extensively reviewed.6–8 Of note, in numerous papers, the drug is currently called “arsenic trioxide”, which is actually used for its production. In aqueous solution, the compound is in fact arsenite9 (Scheme 8.1). Arsenous acid cannot be obtained from aqueous solutions of As2O3, where it reprecipitates. However, in aqueous solution As2O3 exists, and is in equilibrium with, various forms of arsenite. Arsenite also tends to polymerize when it crystallizes, and various crystal shapes have been described.10 Specifically, As2O3 can take the form of meta-­arsenites or ortho-­arsenites (Scheme 8.1).9,11 The pKa values of As(OH)3 are 9.23, 12.13 and 13.40.11 To date, arsenic is marketed as the drug Trisenox®, and the international non-­ proprietary (INN) name is merely “arsenic”. Trisenox®, is formulated as a solution of arsenic trioxide in water-­for-­injection using sodium hydroxide and hydrochloric acid to adjust to pH 8. To avoid misinterpretation, this is the solution we refer to as ATO throughout this paper.

Scheme 8.1  Dissociation  products of arsenous acid. Arsenous acid is amphoteric, dissociating via two mechanisms. The dissociation constants are 6 × 10−10 and 1 × 10−14.

View Online

Arsenic-based Anticancer Agents

199

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

−1

Trisenox® (solution, 1 mg ml for infusion) received a marketing authorization in the USA (2000) and in Europe (2002) for the treatment of APL. It was indicated for induction of remission and consolidation in adult patients with:    ●● Newly diagnosed low-­to-­intermediate risk APL (white blood cell count, ≤ 10 × 103 µl−1) in combination with all-­trans retinoic acid (ATRA) ●● Relapsed/refractory APL (previous treatment should have included a retinoid and chemotherapy) characterized by the presence of the t(15;17) translocation and/or the presence of the promyelocytic leukaemia/retinoic acid receptor-­α (PML/RARα) gene (i.e., fusion of the promyelocytic leukemia protein gene with the retinoic acid receptor α gene).    Trisenox® (Teva) must be administered intravenously over 1–2 hours. The infusion duration may be extended up to 4 hours if vasomotor reactions are observed. Phenasen® is a similar drug marketed in Australia (Phebra Pty Ltd). Trisenox® and Phenasen® contain the same active ingredient, in the same total concentration: 10 mg arsenic trioxide in 10 ml volume. Whereas Phenasen® is available as a vial, Trisenox® is available as an ampoule. Taking into account the impressive results obtained with Trisenox®, the need to identify improved arsenic compounds with reduced side-­effects emerged, and some of the “old antiparasitic compounds” became worth testing. These compounds were in fact organo-­arsenicals and, by definition, contain a chemical bound between arsenic and carbon. The first significant organoarsenical drug was synthesized by Pierre Antoine Béchamp (1816– 1908) by chemically reacting arsenic acid with aniline. Paul Ehrlich did additional experiments on arsenic and he obtained Salvarsan in 1910. After the Second World War, E.A.H. Friedheim also greatly improved the treatment of trypanosomiasis by melaminophenyl arsenicals. Among these drugs, melarsoprol (Scheme 8.2), which is the only organoarsenical drug still in use,12 was assessed in the field of cancer research. The in vitro experiments offer some hope for this compound,13,14 and a clinical trial was also performed in patients with refractory or resistant leukemia.15 Nevertheless, the author concluded that the doses and schedules were associated with excessive CNS toxicity. To date, no other clinical trial has been performed.

Scheme 8.2  Chemical  structure of melarsoprol.

View Online

Chapter 8

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

200

Melarsoprol inhibits growth and induces apoptosis, but it does not affect PML and/or PML/RAR nuclear localization.16 Hence, Trisenox® was confirmed to be the drug of choice for a particular type of leukemia, namely, APL. From a chemical point of view, melarsoprol is probably not the best organoarsenical: it bears a melaminyl group,17 which is typically associated with the antiparasitic activity of the compound. This opened a series of studies on the anticancer properties of other dithia-­arsanes without a melaminyl group.18–23 In 2006, a wide series of dithiarsolanes (Scheme 8.3) were assessed in two leukemia cell lines (K562 myelogenous leukemia and U937 histiocytic lymphoma).18 Their systemic toxicity was estimated by the corresponding LD50 in mice. The cytotoxic activity of each derivative was significantly enhanced compared with arsenic trioxide, and the therapeutic index (TI = LD50/IC50) was also improved. For example, TI was 279.17 ± 41.32 for arsthinol (Scheme 8.3, R1 = –H, R2 = –NH–CO–CH3, R3 = OH) vs. 3.41 ± 0.32 for ATO. Among the experimental arsenic-­based anticancer complexes, arsenoplatins have been proposed to provide synergistic activity of ATO and cisplatin. For example, arsenoplatin 1 (Scheme 8.4) has significant biological activity in several cancer cell lines (ovarian cisplatin resistant A2780; colon HCT-­116; glioblastoma U87) and preliminary data are consistent with the ability of arsenoplatins to overcome cisplatin resistance mechanisms.88

Scheme 8.3  General  formula of dithia-­arsanes subject to assessment of anticancer properties.89

Scheme 8.4  Chemical  structure of arsenoplatin 1, synthesized by heating cisplatin

with As2O3 in acetonitrile/water mixture (9 : 1, v/v) at 90 °C for three days.88

View Online

Arsenic-based Anticancer Agents

201

8.2  Mechanism of Action

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

8.2.1  Transport Across Biomembranes Several studies have been done on cellular uptake of ATO. At physiological pH, the dominating species is As(OH)3, which can cross the plasma membrane readily and appear in the cytosol.24 It was also demonstrated that arsenic could bind to the membrane and induce changes in the fluidity of membrane lipids and in the negative charge density in the outer surface of the membrane.24 Although it has been assumed that As(OH)3 enters cells only by passive diffusion, it emerged in the 2000s that As(OH)3 transport might be facilitated by aquaglyceroporins. The mammalian aquaglyceroporins GlpF facilitate glycerol uptake,25,26 but it was demonstrated that GlpF also transports polyols, urea27 and unionized forms of antimony (Sb(OH)3)28 and arsenic (As(OH)3).29 This family of membrane proteins was first identified as a trivalent metalloid transporter by Sanders and coworkers28 in Escherichia coli. Such a mechanism can be explained by the similarity in the charge distribution of As(OH)3 to glycerol molecules.30 Given the ubiquity of these transporters, mammalian aquaglyceroporins AQP7 and AQP9 were also studied; the results suggested that AQP7 and AQP9 are a major route of arsenic uptake into mammalian cells.29 It was also demonstrated that the treatment of HL60 leukemia cells with vitamin D results in an increase in expression of AQP9 and hypersensitivity to Trisenox®.31 Hence, it might be possible to design other compounds that have the potential to increase the expression or activity of the AQP9 specifically in leukemia cells to increase the efficacy of ATO.32

8.2.2  Reactivity with Thiols Until the 1990s, the mechanism of the anticancer effect of trivalent arsenical derivatives was only partially elucidated. After numerous years, and especially given the work of Friedheim on dithiarsolanes, it is well known that trivalent arsenic binds to thiolated molecules.33 This property has been used to propose an antidote: the British anti-­Lewisite (i.e., 2,3-­dimercaptopropanol or BAL or dimercaprol) for intoxication cases:34 dimercaprol competes with the thiol groups for binding the metal ion, which is then excreted in the urine. A solution of BAL is still proposed for this indication.35,36 Pentavalent arsenic is less toxic, and some authors noted that its effects are not similar.37 Nevertheless, various experiments on the metabolism and fate of arsenical compounds reported their possible interconversion to the corresponding trivalent compound and vice versa. We can deduce that the main mechanism of action of arsenical compounds is the fixation of thiols. Several papers have also demonstrated that the tight binding on vicinal thiols is more stable. These observation were first reported using “small”

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

202

Chapter 8

dithiols, such as 3,4-­dimercaptopropanol, which can lead to dithiarsolanes.38 The hypothesis that arsenic has a very high affinity for vicinal thiols became very popular (Scheme 8.5). Modern biotechnology methods allow us to obtain additional evidence, but the facts are slightly more complex than a mere binding to vicinal thiols,39 given that proteins have complex three-­dimensional structures. For example, Ordonez et al. studied the fixation of arsenic (i.e., sodium arsenite) on proteins produced by Escherichia coli (ArsR repressor encoded by Escherichia coli plasmid R773). The authors proposed that Cys 55 and either Cys 15 or Cys 16 first bind arsenic (AsIII) in a weak complex. This action is followed by unraveling the beginning of the α1-­helix, which exposes the third thiolate and completes the high-­affinity binding of arsenic (AsIII) (Figure 8.1).40 Luedtke and coworkers studied the interaction with proteins of pro-­ fluorescent biarsenical reagents (fluorescein arsenical hairpin binder–ethanedithiol; FlAsH-­EDT2 and resorufin arsenical hairpin binder–ethanedithiol; ReAsH-­EDT2, Scheme 8.5).41 This strategy termed “bipartite tetracysteine display” enables the detection of protein–protein interactions. The fluorescence here proves the attachment of two arsenic atoms on two pairs of vicinal thiols. These experiments have been performed on various polypeptides and some years later by another team on PML. We will revisit this later in this chapter. In vivo, and especially in humans, thiols may be present as protein-­bound (PSH) or non-­protein-­bound (NPSH) SH groups.42 The highest concentration of NPSH is accounted for by glutathione. The overall concentration of glutathione ranges between 0.5 and 15 mM.43 In the organism, all the thiolated molecules have the potential to link to trivalent arsenic (e.g., R–As=O, arsenite ions). Nevertheless, the affinity may differ as a function of the accessibility of these thiols.33 Because of the large concentration of thiols in the cell, with the majority of them in the reduced state, arsenic has a wide array of targets. The majority of proteins that bind arsenic have spatially close Cys residues that bind arsenic through the sulfhydryl side-­chain. In cancer cells, there is a more limited set of interactions that are thought to exert the antitumor effect of arsenous acid, as discussed below.

8.2.3  Biological Effects Given that trivalent arsenic targets many endogenous proteins, many papers have been published in the last 20 years, and the spectrum of “possible targets” is starting to be better understood. Various cellular models have demonstrated the potency of arsenic to alter the function of various enzymes and signaling molecules. For example, on HeLa S3 cells (cells from a cervical adenocarcinoma subclone) arsenic stimulates JNK activity44 via inhibition of a constitutive dual-­specificity JNK phosphatase. This phosphatase maintains low basal JNK activity in non-­stimulated cells, and its inhibition may lead to tumor promotion through induction of proto-­oncogenes, such as c-­jun

View Online

203

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

Scheme 8.5  Theoretical  fixation of various arsenical compounds on cysteine resi-

dues. (A) Arsenous acid (fixation of 3 Cys). (B) Monomethylarsonous acid (fixation of 2 Cys). (C) Dimethylarsonous acid (fixation of 1 Cys). (D) p-­ Aminophenyl arsenoxide (fixation of 2 Cys). (E) FIAsH–EDT2 a fluorescent, biarsenical compound that binds covalently to tetracysteine sequences (fixation of 4 Cys). (F) ReAsH–EDT2 a fluorescent, biarsenical compound that binds covalently to tetracysteine sequences (fixation of 4 Cys).

View Online

Chapter 8

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

204

Figure 8.1  Binding  of arsenic on an ArsR “arsenic-­resistance regulatory protein”

from Ordonez et al. Binding is proposed to occur via two steps: (1) As(iii) (blue sphere) binds first to the thiolates of Cys 55 of one subunit and Cys 15 or Cys 16 (shown here for only Cys 15); (2) the end of α1 unravels to allow the adjacent cysteine residue to become the third ligand to As(iii), forming a high-­affinity site. Reproduced from ref. 40 with permission from American Society for Biochemistry and Molecular Biology, Copyright 2008.

and c-­fos, and stimulation of AP-­1 activity.44 The same phosphatase may also regulate p38/Mpk2 activity. Another example involves the translocation of several protein kinase C (PKC) isoforms from the cytosol to the plasma membrane after exposure to arsenite.45 Most of these experiments were first performed to understand the toxicity and the carcinogenic effects of arsenic, but our understanding of their anticancer activity truly started during clinical trials with ATO (i.e., As(OH3) when it is dissolved). Concerning ATO's anticancer activity, the drug affected biological processes including apoptosis, autophagy, differentiation, cell cycle, reactive oxygen species production and/or neovascularization as described in Sections 8.2.3.1–8.2.3.3. Furthermore, arsenic exhibits some effects on telomerase activity and telomere length.46

8.2.3.1 Apoptosis/Cell Differentiation Apoptosis is considered an important biological process involved in normal cell turnover, proper functioning and development of the immune system, embryonic development, chemical-­induced cell death and hormone-­ dependent atrophy. Cancer can occur if too little apoptosis occurs, resulting in malignant cells that will not die. In 1996, Andre and coworkers47 proposed apoptosis as a mechanism for ATO in the induction of remission in APL.48,49 A few years later, exposure to

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

205

ATO was demonstrated to activate caspases. In their clinical trial, Soignet et al. demonstrated that responses to ATO were accompanied by enhanced expression of proforms of various caspases (i.e., caspases 2 and 3) and activation of caspases 1 and 2.7 At this time, an accurate mechanism is not known, and caspase activation might be direct or indirect. Other explanations were put forward when Lallemand-­Breitenbach and coworkers reported on the synergy between retinoic acid and ATO.50 Both drugs induced catabolism of the oncogenic and APL-­specific PML/RARα fusion protein. These authors also demonstrated that combining ATO with retinoic acid accelerates tumor regression through enhanced differentiation and apoptosis in a mouse model [Leukemic cells from PML/RARα transgenic mice (leukemia 935) were propagated by injecting blasts into the tail vein of syngenic FVB-­NICO mice]. Each drug targets a specific moiety of the fusion protein to the proteasome: ATO interacts with PML and retinoic acid with RARα. ATO elicits the sumoylation of PML proteins.51 This process is a post-­ translational modification corresponding to the addition of SUMOs (small ubiquitin-­like modifiers). It was suggested that two consecutive steps are involved because mutants that do not associate with the nuclear matrix are not sumoylated, whereas nuclear matrix-­associated PML is consistently and completely polysumoylated.51 A sequential model has been proposed for the formation of PML nuclear bodies: a progressive association of PML to the matrix leading to the sumoylation.51 It was proposed that arsenic-­induced oligomerization of PML facilitates subsequent protein modification by sumoylation and ubiquitination and enhances its degradation. All these findings are interesting from a chemical point of view. Actually, arsenous acid (i.e., ATO) is likely to bind to PML protein, and its interaction has been studied by a comparison between three compounds that could exhibit various associations with cysteine residues: arsenous acid (3 SH links possible), phenylarsine oxide (2 SH links possible), and FlAsH-EDT2 (4 SH links possible, see Scheme 8.5).41,51 The authors ultimately concluded that the three compounds exhibited different biological effects. FlAsH-EDT2 (4 SH links) binds two coiled-­coil pairs and can initiate PML sumoylation and degradation more efficiently than arsenous acid.51 Conversely, phenylarsine oxide (2 SH links) only binds a single coiled-­coil pair. Phenylarsine oxide promotes PML disulfide formation (presumably through generation of ROS), yet it paradoxically antagonizes both basal PML matrix attachment and sumoylation.51 Arsenous acid (i.e., ATO) could form As2O3-­mediated intramolecular bridges41,51 but the authors concluded that studies should define the arrangement of arsenic atoms within PML multimers and determine the functional consequences of binding. Although APL cells are very sensitive to treatment with ATO, there are important ATO targets relevant to other cancer cells. For example, ATO exposure can also promote apoptosis by activating caspase 3 in neuroblastoma cell lines and myeloid leukemia cells.52

View Online

206

Chapter 8

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

8.2.3.2 Oxidative Stress Pathway Many papers have demonstrated that reactive oxygen species are generated by arsenic.49,53,54 This oxidative stress can play a role in toxic and therapeutic effects. Hydrogen peroxide mediates apoptosis by reducing the mitochondrial membrane potential and activating the caspase pathway.55 As previously noted, arsenic may inactivate many enzymes via binding to sulfhydryl groups. Some of these enzymes belong to the endogenous thioredoxin (Trx) and glutathione system [e.g., glutathione reductase, glutathione peroxidase, glutathione S-­transferase and thioredoxin reductase (TrxR)].56 Consequently, the intracellular redox potential can be affected, leading to apoptosis. Of note, is that Trx is an oxidoreductase that contains a dithiol active site. It performs a number of signaling and homeostasis activities related to the cellular oxidation state and can also modulate the levels of adventitious reactive oxygen species.57 The dithiol active site is maintained in the reduced state by another enzyme, TrxR. TrxR is an NADPH-­dependent flavoenzyme that utilizes an active site composed of a redox-­active Cys pair and a redox-­ active selenocysteine–Cys pair. ATO has been shown to interact strongly with selenocysteine residues in TrxR.58 Recently, the selenocysteine-­dependent TrxR enzyme has emerged as an important molecular target for anticancer drug development.59,60 Finally, other redox-­sensitive signaling molecules include AP-­1, NF-­κB, IκB and Tp53.44,61,62 These proteins can further modify cell signaling and gene expression.

8.2.3.3 Neovascularization In cancer, new growth of the vascular network is important given that proliferation of the primary tumor and metastatic spread require an adequate supply of oxygen and nutrients. ATO can induce angiogenesis63 via oxygen species (ROS) generation, which activates AKT and ERK1/2 signaling pathways and increases the expression of vascular endothelial growth factor (VEGF) and hypoxia-­inducible factor 1 (HIF-­1). On the other hand, the novel organoarsenical GSAO, 4-­(N-­(S-­glutathionylacetyl)amino) phenylarsonous acid (Scheme 8.6)64 exhibits potential anti-­angiogenic capability via applications in cancer where tumor metastasis relies on neovascularization. GSAO is a substrate of γGT and the resulting metabolite [4-­(N-­(S-­cysteinylglycylacetyl)amino) phenylarsonous acid, GCAO] is transported across the plasma membrane by an organic anion transporter. Furthermore, GCAO is processed in the cytosol by dipeptidases to [4-­(N-­(S-­cysteinylacetyl)amino) phenylarsonous acid; CAO]. This metabolite reacts with the mitochondrial adenine nucleotide translocase of the inner mitochondrial membrane of endothelial cells, which leads to proliferation arrest.65 The use of GSAO as a novel anti-­neovascular agent could exhibit promise in human oncology therapeutics given the prospect of its endothelial selectivity and limited toxicity.64

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

207

Scheme 8.6  GSAO,  4-­(N-­(S-­glutathionylacetyl)amino) phenylarsonous acid.

8.3  A  rsenic in the Treatment of Acute Promyelocytic Leukemia In APL patients, retinoic acid triggers differentiation, whereas ATO induces both partial differentiation and apoptosis. Although their mechanisms of action are believed to be distinct, these two drugs both induce catabolism of the oncogenic promyelocytic leukemia (PML)/RARα fusion protein. Combining ATO with retinoic acid can accelerate tumor regression through enhanced differentiation and apoptosis.50 Although retinoic acid or ATO alone only prolongs survival two-­ to three-­fold, the two drugs lead to tumor clearance after a nine-­month relapse-­free period. ATO is currently associated with retinoic acid and proposed as standard therapy in the treatment of APL.

8.4  Arsenical Drugs and Glioma The most common malignant glioma is glioblastoma, which is associated with a median survival of 12–15 months. The EORTC (European Organization for Research and Treatment of Cancer) has published a study on the concomitant use of radiation therapy and adjuvant temozolomide, offering slightly improved mortality (overall survival: 9.8% after 5 years).66 This treatment has been adopted as the new standard treatment, but it remains unsatisfactory.67 In addition to temozolomide, other compounds [bevacizumab68 and VEGF receptor tyrosine kinase inhibitors (pazopanib,69,70 lapatinib,70,71 erlotinib72,73)] have been tested in clinical trials, but only phase II studies have been published. Low molecular weight kinase inhibitors may offer advantages in terms of drug delivery, whereas monoclonal antibodies exhibit increased specificity but face delivery restrictions. To date, few molecularly targeted therapies have demonstrated significant antineoplastic activity, possibly due to tumor heterogeneity. ATO has also been shown to be effective in vitro and in mouse models especially, because it can induce both autophagy and apoptosis.74 Autophagy has

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

208

Chapter 8

Scheme 8.7  Chemical  structure of arsthinol. been suggested as a possible mechanism for non-­apoptotic cell death. The mechanism has been mostly studied in glioma cells.74–77 ATO induces autophagic cell death in malignant glioma cells by upregulation of the mitochondrial cell death protein BNIP3 76 and downregulation of survivin.78 The drug was recently administered with temozolomide and radiotherapy in a phase II clinical trial, and the authors concluded that adding ATO was feasible without increased side-­effects.79 Nevertheless, ATO did not improve overall survival compared with historic data, but it identified glioma as a potential indication for future developments concerning arsenic-­based drugs. Hence, organoarsenical compounds attracted interest for the treatment of glioma. These compounds are typically more lipophilic and able to pass through the blood–brain barrier. A good example of lipophilic organoarsenicals is arsthinol (Scheme 8.7). Given its high lipophilicity and its very low solubility in water, arsthinol was first used only orally as an antiprotozoal agent.80 The compound was synthesized for the first time in 1949 by Friedheim via the formation of a complex including acetarsol and 2,3-­dimercaptopropanol (British anti-­Lewisite).81 It was marketed few years later by Endo Products (Balarsen, tablets, 0.1 g) for the treatment of amoebiasis and yaws.82,83 Among trivalent organoarsenicals, arsthinol was very well tolerated.84,85 Solubilization of arsthinol was recently achieved by complexation with cyclodextrins.12,21 This formulation opened up new fields for the use of lipophilic arsenical drugs, especially in cancer treatments. This complex was assessed in a mouse model of glioma (heterotopic graft of U87MG; groups of 5 animals; administration 5 days per week)23 and compared with ATO at 65% of the maximum tolerated dose. The median survival after ATO i.p. treatment was 15 days, which was similar to that of the control group (14 days). In contrast, survival after arsthinol/cyclodextrin complex i.p. treatment reached 21 days (p < 0.05).

8.5  Conclusion Arsenic-­based drugs mostly belong to history. Nevertheless, the good efficacy of ATO in acute promyelocytic leukemia raised questions about its mechanism of action. To date, we know that arsenous acid exhibits unique activity in acute promyelocytic leukemia given its binding to the PML/RARα fusion protein. Its efficacy on glioma cells is particularly promising and is likely to exhibit specificity, given that ATO induces both apoptosis and autophagy.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

209

Thus, the efficacy of ATO on glioma has recently been tested in a phase II clinical trial, but it did not improve overall survival compared with historical data. Nevertheless, these results suggest that glioma could serve as a potential indication for future developments of arsenic-­based drugs. ATO has also been tested in the treatment of autoimmune diseases (i.e., systemic lupus erythematosus, SLE). The main results have been obtained in vitro on splenocytes of MRL/lpr mice and peripheral blood mononuclear cells of SLE patients: the mRNA and protein expression levels are lower when these cells are incubated with ATO.86 A phase II clinical trial has been done (unpublished data) and the drug has also obtained an orphan drug designation from the European Medicines Authority and the Federal Drug Administration for graft-­versus-­host disease.87

Abbreviations γGT γ-­Glutamyltransferase APL Acute promyelocytic leukemia AQP Aquaglyceroporin AsIII Trivalent arsenic AsV Pentavalent arsenic ATO Arsenic trioxide BAL British anti-­Lewisite (2,3-­dimercaptopropanol) CNS Central nervous system Cys Cysteine FlAsH-­EDT2 Fluorescein arsenical hairpin binder–ethanedithiol GlpF Glycerol facilitator (aquaglyceroporin) GCAO 4-­(N-­(S-­Cysteinylglycylacetyl)amino) phenylarsonous acid GSAO 4-­(N-­(S-­Glutathionylacetyl)amino) phenylarsonous acid EMA European Medicines Agency EORTC European Organization for Research and Treatment of Cancer FDA US Food and Drug Administration IC50 Inhibiting concentration 50 INN International non-­proprietary name LD50 Lethal dose 50 NPSH Non-­protein-­bound SH groups PML/RARα Fusion of the promyelocytic leukemia protein gene with the retinoic acid receptor α gene PML/RARα Fusion of the promyelocytic leukemia protein with the retinoic acid receptor α PKC Protein kinase C PBMC Peripheral blood mononuclear cells PSH Protein-­bound SH groups ReAsH-­EDT2 Resorufin arsenical hairpin binder–ethanedithiol RARα Retinoic acid receptor alpha

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

210

Chapter 8

ROS Reactive oxygen species SLE Systemic lupus erythematosus SUMO Small ubiquitin-­like modifier Trx Thioredoxin TrxR Thioredoxin reductase VEGF Vascular endothelial growth factor TI Therapeutic index

References 1. S. Gibaud and G. Jaouen, in Topics in Organometallic Chemistry, ed. G. Jaouen and N. M. N. Metzler-­Nolte, 2010, vol. 32, p. 1. 2. J. Z. Wu and P. C. Ho, J. Chromatogr. A, 2004, 1026, 261. 3. T. Fowler, Lond. Med. J., 1786, 7(Pt 2), 192. 4. E. G. Cutler and E. H. Bradford, Am. J. Med. Sci., 1878, 74. 5. H. D. Sun, L. Ma, X. Hu and T. Zhang, Chin. J. Comb. Trad. Chin. Med. West Med., 1992, 12, 170. 6. D. Douer and M. S. Tallman, J. Clin. Oncol., 2005, 23, 2396. 7. S. Soignet, P. Maslak, Z. Wang, S. Jhanwar, E. Calleja, L. Dardashti, D. Corso, A. DeBlasio, J. Gabrilove, D. A. Scheinberg, P. Pandolfi and R. Warrell, N. Engl. J. Med., 1998, 339, 1341. 8. P. J. Dilda and P. J. Hogg, Cancer Treat. Rev., 2007, 33, 542. 9. N. Harihara Subramanian, M. R. Hariganesh, R. Ganesh Jeevan and G. Radhakrishnan, J. Chromatogr. Sci., 2011, 628. 10. C. Lee and W. T. Harrison, Acta Crystallogr., C, 2004, 60, m215. 11. P. A. O'Day, Elements, 2006, 2, 77. 12. S. Gibaud, S. B. Zirar, P. Mutzenhardt, I. Fries and A. Astier, Int. J. Pharm., 2005, 306, 107. 13. R. Rivi, E. Calleja, A. Konig, L. Lai, C. Gambacorti-­Passerini, D. Scheinberg, J. L. Gabrilove, R. P. Warrell and P. P. Pandolfi, Blood, 1996, 88, 68a. 14. A. Konig, L. Wrazel, R. P. Warrell Jr, R. Rivi, P. P. Pandolfi, A. Jakubowski and J. L. Gabrilove, Blood, 1997, 90, 562. 15. S. L. Soignet, W. P. Tong, S. Hirschfeld and R. P. Warrell Jr, Cancer Chemother. Pharmacol., 1999, 44, 417. 16. Z. G. Wang, R. Rivi, L. Delva, A. Konig, D. A. Scheinberg, C. Gambacorti-­ Passerini, J. L. Gabrilove, R. P. Warrell Jr and P. P. Pandolfi, Blood, 1998, 92, 1497. 17. E. A. H. Friedheim, US Pat., 2422724, 1947. 18. S. Gibaud, R. Alfonsi, P. Mutzenhardt, I. Fries and A. Astier, J. Organomet. Chem., 2006, 691, 1081. 19. S. Ben Zirar, S. Gibaud, A. Camut and A. Astier, J. Organomet. Chem., 2007, 692, 1348. 20. S. Gibaud and A. Astier, Ann. Pharm. Fr., 2007, 65, 162. 21. S. Ben Zirar, A. Astier, M. Muchow and S. Gibaud, Eur. J. Pharm. Biopharm., 2008, 70, 649.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

211

22. I. Ajana, A. Astier and S. Gibaud, Oncologie, 2009, 11, S34. 23. S. Becherirat, M.-­C. Lanhers, M. Socha, M. Yemloul, A. Astier, C. Loboda, N. Aniceto and S. Gibaud, Eur. J. Pharm. Biopharm., 2013, 85, 560. 24. T. L. Zhang, Y. X. Gao, J. F. Lu and K. Wang, J. Inorg. Biochem., 2000, 79, 195. 25. G. Sweet, C. Gandor, R. Voegele, N. Wittekindt, J. Beuerle, V. Truniger, E. C. Lin and W. Boos, J. Bacteriol., 1990, 172, 424. 26. K. B. Heller, E. C. Lin and T. H. Wilson, J. Bacteriol., 1980, 144, 274. 27. M. Borgnia, S. Nielsen, A. Engel and P. Agre, Annu. Rev. Biochem., 1999, 68, 425. 28. O. I. Sanders, C. Rensing, M. Kuroda, B. Mitra and B. P. Rosen, J. Bacteriol., 1997, 179, 3365. 29. Z. Liu, J. Shen, J. M. Carbrey, R. Mukhopadhyay, P. Agre and B. P. Rosen, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 6053. 30. A. Porquet and M. Filella, Chem. Res. Toxicol., 2007, 20, 1269. 31. H. Bhattacharjee, J. Carbrey, B. P. Rosen and R. Mukhopadhyay, Biochem. Biophys. Res. Commun., 2004, 322, 836. 32. G. P. Bienert, M. D. Schussler and T. P. Jahn, Trends Biochem. Sci., 2008, 33, 20. 33. S. Shen, X. F. Li, W. R. Cullen, M. Weinfeld and X. C. Le, Chem. Rev., 2013, 113, 7769. 34. A. Heyman, J. B. Pfeiffer Jr, R. W. Willett and H. M. Taylor, N. Engl. J. Med., 1956, 254, 401. 35. R. A. Peters, L. A. Stocken, R. H. S. Thompson, F. N. Woodward, A. F. Millidge and E. J. Gasson, US Pat., 2432797, 1947. 36. E. A. H. Friedheim, US Pat., 2664432, 1953. 37. H. Minden and R. Rothe, Int. Arch. Arbeitsmed., 1968, 24, 359. 38. B. Berger and A. H. Fairlamb, Antimicrob. Agents Chemother., 1994, 38, 1298. 39. S. Shen, X. F. Li, W. R. Cullen, M. Weinfeld and X. C. Le, Chem. Rev., 2013, 113, 7769. 40. E. Ordonez, S. Thiyagarajan, J. D. Cook, T. L. Stemmler, J. A. Gil, L. M. Mateos and B. P. Rosen, J. Biol. Chem., 2008, 283, 25706. 41. N. W. Luedtke, R. J. Dexter, D. B. Fried and A. Schepartz, Nat. Chem. Biol., 2007, 3, 779. 42. O. Vos, Adv. Space Res., 1992, 12, 201. 43. M. Deponte, Biochim. Biophys. Acta, 2013, 1830, 3217. 44. M. Cavigelli, W. W. Li, A. Lin, B. Su, K. Yoshioka and M. Karin, EMBO J., 1996, 15, 6269. 45. N. Y. Chen, W. Y. Ma, C. Huang, M. Ding and Z. Dong, J. Environ. Pathol. Toxicol. Oncol., 2000, 19, 297. 46. Y. Cheng, Y. Li, C. Ma, Y. Song, H. Xu, H. Yu, S. Xu, Q. Mu, H. Li, Y. Chen and G. Zhao, Oncotarget, 2016, 7, 12682. 47. C. Andre, M. C. Guillemin, J. Zhu, M. H. Koken, F. Quignon, L. Herve, M. K. Chelbi-­Alix, D. Dhumeaux, Z. Y. Wang, L. Degos, Z. Chen and H. de The, Exp. Cell Res., 1996, 229, 253.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

212

Chapter 8

48. Y. Akao, H. Mizoguchi, S. Kojima, T. Naoe, N. Ohishi and K. Yagi, Br. J. Haematol., 1998, 102, 1055. 49. Y. Sun, C. Wang, L. Wang, Z. Dai and K. Yang, Cell. Mol. Biol. Lett., 2018, 23, 13. 50. V. Lallemand-­Breitenbach, M. C. Guillemin, A. Janin, M. T. Daniel, L. Degos, S. C. Kogan, J. M. Bishop and H. de The, J. Exp. Med., 1999, 189, 1043. 51. M. Jeanne, V. Lallemand-­Breitenbach, O. Ferhi, M. Koken, M. Le Bras, S. Duffort, L. Peres, C. Berthier, H. Soilihi, B. Raught and H. de The, Cancer Cell, 2010, 18, 88. 52. Y. Akao, Y. Nakagawa and K. Akiyama, FEBS Lett., 1999, 455, 59. 53. G. Kroemer and H. de The, J. Natl. Cancer Inst., 1999, 91, 743. 54. C. Berndt, C. H. Lillig and A. Holmgren, Am. J. Physiol.: Heart Circ. Physiol., 2007, 292, H1227. 55. Y. C. Chen, S. Y. Lin-­Shiau and J. K. Lin, J. Cell. Physiol., 1998, 177, 324. 56. E. P. Swindell, P. L. Hankins, H. Chen, D. U. Miodragovic and T. V. O'Halloran, Inorg. Chem., 2013, 52, 12292. 57. C. H. Lillig and A. Holmgren, Antioxid. Redox Signaling, 2007, 9, 25. 58. J. Lu, E. H. Chew and A. Holmgren, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 12288. 59. G. Powis, D. Mustacich and A. Coon, Free Radicals Biol. Med., 2000, 29, 312. 60. V. Scalcon, A. Bindoli and M. P. Rigobello, Free Radicals Biol. Med., 2018, 127, 62. 61. M. E. El-­Sabban, R. Nasr, G. Dbaibo, O. Hermine, N. Abboushi, F. Quignon, J. C. Ameisen, F. Bex, H. de The and A. Bazarbachi, Blood, 2000, 96, 2849. 62. J. Mathieu and F. Besancon, Int. J. Cancer, 2006, 119, 1723. 63. L. Z. Liu, Y. Jiang, R. L. Carpenter, Y. Jing, S. C. Peiper and B. H. Jiang, PLoS One, 2011, 6, e20858. 64. M. A. Elliott, S. J. Ford, E. Prasad, L. J. Dick, H. Farmer, P. J. Hogg and G. W. Halbert, Int. J. Pharm., 2012, 426, 67. 65. N. Donoghue, P. T. Yam, X. M. Jiang and P. J. Hogg, Protein Sci., 2000, 9, 2436. 66. G. Dresemann, OncoTargets Ther., 2010, 3, 139. 67. R. Stupp, M. E. Hegi, W. P. Mason, M. J. van den Bent, M. J. Taphoorn, R. C. Janzer, S. K. Ludwin, A. Allgeier, B. Fisher, K. Belanger, P. Hau, A. A. Brandes, J. Gijtenbeek, C. Marosi, C. J. Vecht, K. Mokhtari, P. Wesseling, S. Villa, E. Eisenhauer, T. Gorlia, M. Weller, D. Lacombe, J. G. Cairncross, R. O. Mirimanoff, European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups and National Cancer Institute of Canada Clinical Trials Group, Lancet Oncol., 2009, 10, 459. 68. M. R. Gilbert, J. J. Dignam, T. S. Armstrong, J. S. Wefel, D. T. Blumenthal, M. A. Vogelbaum, H. Colman, A. Chakravarti, S. Pugh, M. Won, R. Jeraj, P. D. Brown, K. A. Jaeckle, D. Schiff, V. W. Stieber, D. G. Brachman, M. Werner-­Wasik, I. W. Tremont-­Lukats, E. P. Sulman, K. D. Aldape, W. J. Curran Jr and M. P. Mehta, N. Engl. J. Med., 2014, 370, 699.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00196

Arsenic-based Anticancer Agents

213

69. F. M. Iwamoto, K. R. Lamborn, H. I. Robins, M. P. Mehta, S. M. Chang, N. A. Butowski, L. M. Deangelis, L. E. Abrey, W. T. Zhang, M. D. Prados and H. A. Fine, Neuro-­Oncology, 2010, 12, 855. 70. D. A. Reardon, M. D. Groves, P. Y. Wen, L. Nabors, T. Mikkelsen, S. Rosenfeld, J. Raizer, J. Barriuso, R. E. McLendon, A. B. Suttle, B. Ma, C. M. Curtis, M. M. Dar and J. de Bono, Clin. Cancer Res., 2013, 19, 900. 71. V. Karavasilis, V. Kotoula, G. Pentheroudakis, D. Televantou, S. Lambaki, S. Chrisafi, M. Bobos and G. Fountzilas, J. Neurol., 2013, 260, 1469. 72. A. Broniscer, S. J. Baker, C. F. Stewart, T. E. Merchant, F. H. Laningham, P. Schaiquevich, M. Kocak, E. B. Morris, R. Endersby, D. W. Ellison and A. Gajjar, Clin. Cancer Res., 2009, 15, 701. 73. M. D. Prados, K. R. Lamborn, S. Chang, E. Burton, N. Butowski, M. Malec, A. Kapadia, J. Rabbitt, M. S. Page, A. Fedoroff, D. Xie and S. K. Kelley, Neuro-­Oncology, 2006, 8, 67. 74. H.-­W. Chiu, Y.-­S. Ho and Y.-­J. Wang, J. Mol. Med., 2011, 89, 927. 75. T. Kanzawa, Y. Kondo, H. Ito, S. Kondo and I. Germano, Cancer Res., 2003, 63, 2103. 76. T. Kanzawa, L. Zhang, L. Xiao, I. M. Germano, Y. Kondo and S. Kondo, Oncogene, 2005, 24, 980. 77. M. Pastorek, P. Gronesova, D. Cholujova, L. Hunakova, Z. Bujnakova, P. Balaz, J. Duraj, T. C. Lee and J. Sedlak, Neoplasma, 2014, 61, 700. 78. H.-­W. Chiu, Y.-­S. Ho and Y.-­J. Wang, J. Mol. Med., 2011, 89, 927. 79. P. Kumthekar, S. Grimm, J. Chandler, M. Mehta, M. Marymont, R. Levy, K. Muro, I. Helenowski, K. McCarthy, L. Fountas and J. Raizer, J. Neuro-­ Oncol., 2017, 133, 589. 80. E. H. Loughlin, A. A. Joseph and W. G. Mullin, Antibiot. Chemother., 1954, 4, 647. 81. E. A. H. Friedheim, Am. J. Trop. Med. Hyg., 1949, 29, 173. 82. E. Friedheim, Am. J. Trop. Med. Hyg., 1949, s1-­29, 185. 83. Anonymous, JAMA, 1953, 152, 531. 84. C. Brown, W. Gebhart and A. Reich, JAMA, 1956, 160, 360. 85. Anonymous, WHO Chron., 1956, 10, 26. 86. H. Hu, E. Chen, Y. Li, X. Zhu, T. Zhang and X. Zhu, Biol. Trace Elem. Res., 2018, 184, 391. 87. Z. X. Jiao, Y. Leng, J. J. Xia, H. Q. Wu, N. Jin, J. Z. Fu, L. N. Cheng, J. H. Wang, S. B. Ni and Z. Q. Qi, Xenotransplantation, 2016, 23, 237. 88. D. U. Miodragovic, J. A. Quentzel, J. W. Kurutz, C. L. Stern, R. W. Ahn, I. Kandela, A. Mazar and T. V. O'Halloran, Angew. Chem., Int. Ed. Engl., 2013, 52, 10749. 89. A. Astier and S. Gibaud, WO Pat., 2005/034935, 2005.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

Part II

Enabling Concepts in Metallodrug Discovery

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

View Online

           

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

View Online

Chapter 9

Supramolecular Metal-­based Structures for Applications in Cancer Therapy Margot N. Wenzel, Benjamin Woods and Angela Casini* School of Chemistry, Cardiff University, Main Building, Park Place, Cardiff CF10 3AT, United Kingdom *E-­mail: [email protected]

9.1 Introduction In the last few decades, the integrative self-­sorting observed in nature, whereby the combination and assembly of multi-­component biomolecules occurs in an optimized, regulated and controlled manner, has encouraged the development of artificial metal-­based supramolecular systems with increased diversities and functionalities. In this field, two main types of nanoarchitectures have been described: Metal–Organic Frameworks (MOFs) and Supramolecular Coordination Complexes (SCCs). While MOFs are porous polymers formed by coordination bonds between metal ions or clusters and organic linkers,1,2 SCCs are discrete two-­ (2D) or three-­dimensional (3D) structures.3 Coordination self-­ assembly represents one of the most successful approaches designed by chemists to obtain libraries of SCCs, whose size and shape can be carefully controlled through the judicious choice of metal centers and   Metallobiology Series No. 14 Metal-based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

217

View Online

Chapter 9

218

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

4–7

complementary multidentate ligands. In particular, the 3D family of SCCs defined by metallacages and chiral helicates has received great attention. In general, metallacages feature an internal cavity accessible to guest encapsulation, and thus, exploitable for various functions and applications involving host–guest chemistry.8–10 Although having been described for several decades for the variety of their structures, SCCs have been exploited for diverse applications only in recent years. For example, among the most attractive areas of applicability, catalysis (stabilization of reactive intermediates or solubilization of insoluble precursors within the host cavity),10 sensing and molecular recognition9,11,12 are certainly the most explored. Recent thematic reviews have summarized the exponential progress made in the design, synthesis and application of these discrete nanostructures.3,5,6,13 Thus, this chapter is not intended to detail the design perspective of such architectures, but rather to summarize the various approaches used to obtain different families of SCCs and to focus on the most recent examples, reporting their potential in cancer therapy and anticancer drug delivery.

9.2 Supramolecular Coordination Complexes 9.2.1 Synthesis of 2D (Metallacycles) and 3D (Metallacages) SCCs The synthesis of SCCs has been described for several decades and has been shown to be easily predictable and reproducible through the wise choice of a combination of multidentate ligands (Lewis base ‘donors’) and metal ions (Lewis acid ‘acceptors’), based on exploitation of their coordination geometry. In most cases, the self-­assembly proceeds under mild conditions to form metallacycles and metallacages of general formula [MxLy]z (M = metal, L = ligand, z = charge) in high yields. Various synthetic strategies to obtain SCCs of different shapes have been developed, the earliest being gathered under the theme of ‘directional bonding’ and including edge-­ and face-­directed approaches, whereby a metal acceptor and a ligand donor are mixed in specific ratios to form highly symmetrical polygons and polyhedra (metallacycles and metallacages, respectively).5 Other methodologies to form SCCs include symmetry-­adapted and weak-­link approaches.14 A schematic representation of the main types of ‘directional bonding’ methodologies is given in Figure 9.1A–D. In the edge-­directed self-­assembly approach, the polyhedra are formed using stoichiometric ratios of ligand (often bidendate) to metal precursors (Figure 9.1A). The symmetry-­adapted method relies on the same principle but using multibranched chelating ligands (often at least tridentate), thus leading to the generation of highly symmetric architectures (Figure 9.1B). In the face-­directed approach, developed by Fujita in the late 1990s and also referred to as the ‘paneling method’, rigid pre-­formed ligands act as ‘panels’ to guide coordination complex precursors to form the SCCs (Figure 9.1C).15 In this strategy, modifying the availability of the

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

Supramolecular Metal-based Structures for Applications in Cancer Therapy

219

Figure 9.1 Schematic representations of the main synthetic approaches to form

supramolecular complexes by self-­assembly. Homometallic: (A) edge-­ directed self-­assembly, with a Pd2L4 example;104 (B) symmetry interaction, with a Ag3L2 example;141 (C) face-­directed self-­assembly, with the example of a capsule containing six RuII ions;21 (D) weak-­link method, with an example of interchangeable Rh-­based metallacycle.142 Heterometallic: (E) example of a heterometallic M1–M2 rectangle (M1 = Rh, Ir; M2 = Zn, Ni, Cu);41 (F) example of a 3D Ni6–Fe8 heterometallic cube.45

coordination sites of the metal forces an increased degree of directionality to form the final 3D-­cage. Finally, the weak-­link method uses a hemilabile ligand to first form a ‘weak link’ (weak metal–heteroatom bond) with a metal precursor, generating a condensed metallacycle, which can then be opened by selective introduction of an ancillary ligand with a higher binding affinity (Figure 9.1D).16 The two main synthetic approaches, namely edge-­ (mainly developed by Stang et al.4,17) and face-­directed (first reported by Fujita and coworkers18), have been mainly described to form 2D and 3D nanoarchitectures based on PdII, PtII and RuII centres. For example, in the early 2000s, Fujita reported on a [Pd(N1,N1,N2,N2-­tetramethylethane-­1,2-­diamine)]6L4 (L = ligand: 2,4,6-­tri(pyridin-­4-­yl)-­1,3,5-­triazine) cage accessible by face-­ directed self-­assembly between four equivalents of the planar tridentate ligand (acting as a panel) and six equivalents of the PdII precursor.19 Therrien and coworkers also used a similar type of tripyridyl-­triazine ligand to form RuII-­based capsules of general formula [Ru2]3L2 by face-­directed

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

220

Chapter 9

self-­assembly from a dinuclear 2,5-­dihydroxy-­1,4-­benzoquinonato complex ([Ru2(p-­iPrC6H4Me)2(C6H2O4)Cl2]).20,21 Edge-­directed approaches to form metallacages have been described mainly based on bidendate ‘banana-­shaped’ pyridyl-­based ligands and PdII and PtII metal precursors.7 In general, the simplest examples of 3D metallacages are based on M2L4-­t ype scaffolds, but numerous other structures have been reported (M2L3,22 M4L4,23 M4L6,22,24,25 M8L12,24 M12L24,26 M24L48,27 etc.).7 The largest discrete self-­assembled polyhedron obtained so far by edge-­ directed synthesis has been described by Fujita and coworkers, whereby a Pd48L96 scaffold was observed.28

9.2.2 Synthesis of Heteroleptic, Interlocked and Heterometallic Cages In recent years, thanks to a better understanding of the structure and formation of metallacages and helicates, the development of 3D architectures with increased levels of complexity has risen exponentially. Until recently, only one type of ligand and metal ion have been used in various ratios to form single and highly symmetrical 3D structures. However, very recently, several types of ligands and/or metal ions have been used to generate more complex structures with new properties. In particular, research has focused on the development of both heteroleptic and interlocked (or interpenetrated) metallacages. The controlled synthesis of heteroleptic cages, whereby at least one of the ligands used to form the cage is different than the others (cage of type MxLyL′z) would be attractive to various fields of application, in particular biological applications, as this would allow the insertion of multiple components on a single structure (for instance the concomitant attachment of multiple targeting groups or imaging tools). However, a clear synthetic methodology to predict the self-­assembly of two or more types of ligands with metal ions to generate clean heteroleptic structures has not been described yet.29–31 Inherently, the 3D cages obtained by the face-­directed synthetic approach can be described as heteroleptic structures as the panel, usually formed by an organometallic linker, is formed first, followed by the cage self-­assembly. Thus, the final architecture possesses a single type of metal ion, but features two different organic components (a ligand and a paneling linker). However, this approach has not yet been applied to the insertion of functional groups of interest for use in biological applications. In 2016, Crowley and coworkers reported on the study of Pd2(L1)2(L2)2-­t ype heteroleptic cage formation using the edge-­directed synthetic approach, although, rather than mixing two types of tripyridyl-­based ligands in various ratios with metal ions, the heteroleptic systems were obtained by ligand substitution from homoleptic cages (Figure 9.2A).32 Using Density Functional Theory calculations (DFT), the cis isomer was found to be more stable than the trans, and the heteroleptic metallacages were found to be kinetically metastable intermediates rather than the thermodynamic products of the ligand exchange reaction.32

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

Supramolecular Metal-based Structures for Applications in Cancer Therapy

221

Figure 9.2 (A) Example of a heteroleptic Pd2(L1)2(L2)24+ cage obtained by ligand

exchange from a homoleptic Pd2L44+ cage (the cis isomer was found to be more stable than the trans).32 (B) Schematic representation and corresponding X-­ray structure (CCDC no. 1035264) of interlocked cages [3BF4@Pd4L8](BF4−)5.38 The X-­ray structure is depicted using the Discovery Studio software.

Another type of recent and more sophisticated system based on SCCs is the interpenetrated (or interlocked) double (or more) metallacage. These are usually formed using less bulky and lengthier ligands than in the classical homoleptic metallacage, but the ligands are still banana-­shaped.33 In this category, the first example was described by Kuroda and coworkers, who reported on a double and interpenetrated cage of the general formula [Pd4L8], in which L is a conformationally flexible bidentate ligand.34 However, it was shown that this dimeric structure was only very slightly favored thermodynamically compared to the monomer. Afterwards, Clever and coworkers reported on the formation of interlocked double cages based on a more rigid bidentate ligand bearing a central dibenzosuberone.35 The investigation showed that the self-­assembly first proceeds to form a monomeric cage as a kinetically favored intermediate, which then fully converts, upon heating

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

222

Chapter 9

for a few hours, to form an interpenetrated dimer whereby three BF4 anions (coming from the PdII precursor) are sandwiched between the four metal centres.35 The dimer was also found to be a strong receptor for halide anions, with a significant preference for chlorides.36 More recently, the same group reported on modifications of the ligand backbone (including the use of acridone-­based ligands, Figure 9.2B) to change the size and steric hindrance of the internal cavity between the two monomers, and showed that the affinity of the sandwiched anions can be predicted and adjusted with judicious ligand design.33,37,38 Finally, several recent advances have been made in the synthesis of discrete heterometallic self-­assemblies.39 In general, the synthetic strategies towards such heterometallic architectures, either metallacycles or metallacages, can be summarized into two categories: (1) the metalloligand (metal-­ containing ligand) approach and (2) the self-­sorting approach. While the former is based on the step-­wise assembly of metalloligands used as building blocks and offers precise control over the placement of appended functional groups, the latter can generate systems with more complexity, but with scarce control on their resultant architecture. Compared with homometallic architectures, these assemblies possess large internal cavities that have the potential for exploitation as reactors, host–guest frameworks or as platforms for the design of materials with tailored properties, including controlled redox reactivity. It should be noted that a coordination complex with two appended groups is the simplest metalloligand, and this type of scaffold has been widely used to build heterometallic assemblies. For example, bidentate metalloligands with 4-­pyridyl groups are known to generally favor the self-­assembly of heterometallic macrocycles. Using this approach, Lees and coworkers prepared a series of heterometallic square complexes of different transition metals, including Pd–Ru and Re–M (M = Fe, Ru, or Os) multinuclear complexes.40 The same strategy was used for the preparation of hexanuclear heterometallic metallarectangles by the reaction of the metalloligands [M(L)2]2+ (M = ZnII, NiII and CuII, L = 4′-­(4-­pyridyl)-­2,2′:6′,2″-­terpyridine (4-­pyterpy)) with half-­sandwich organometallic units [Cp*2M2(µ-­DHNA)Cl2] (M = Ir, Rh; DHNA = 6,11-­dihydroxy-­5,12-­naphthacenedione; Cp* = pentamethylcyclopentadienyl ligand) (Figure 9.1E).41 Exploiting the self-­sorting approach, in an effort to prepare metal–organic metallacycles and cages with higher complexity and diversity, the groups of Nitschke42,43 and Schmittel44 have published a series of supramolecular heterometallic structures with a range of elegant architectures. As an example of this strategy, a single heterotopic ligand that can coordinate to FeII (as a tris(pyridylimine) complex) as well as PtII/PdII ions (through its terminal pyridine group), was used to form soluble heterometallic [8Fe+6 Pt/Pd] cubic structures in one-­pot reactions.42,43 The energy-­minimized molecular model shows the square planar PtII ion residing in the middle of each face, with the octahedral FeII ions defining the vertices of the cube. Cubic cages employing labile FeII ions and pyridylimine ligands were also synthesized,

View Online

Supramolecular Metal-based Structures for Applications in Cancer Therapy

223

II

which incorporate Ni ions immobilized by TAPP (tetrakis(4-­aminophenyl) porphyrin) moieties (Figure 9.1F).45

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

9.2.3 Synthesis of Helicates Within the 3D family of SCCs, helicates represent a chiral example of assemblies between rigid organic ligands (typically formed of binding units separated by flexible spacers) and metal ions (Figure 9.3A). The term ‘helicate’ (inorganic double helices) was given to these polymetallic helical multi-­ stranded complexes by Nobel Laureate Jean-­Marie Lehn in the late 1980s, for their resemblance to α-­helices in terms of their diameter, charge and chirality.46 The first reported helicates were composed of di-­ and trinuclear structures of CuI ions with two (Figure 9.3A) and three bipyridine-­based ligands, respectively.47 The formation of such structures was found to be spontaneous, thus in a similar timeframe to self-­assembly processes in

Figure 9.3 (A) Schematic representation of the formation of a double-­stranded helicate and possible configurations, with an example of a trinuclear CuI helicate.143 (B) Schematic representation and corresponding X-­ray structure (CCDC no. 722438) of a Ni2L34+ cylinder.58 The X-­ray structure is depicted using the Discovery Studio software.

View Online

Chapter 9

224

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

48

biological systems. Furthermore, the trinuclear architecture was found to have a 1 : 1 ratio of helicates with opposite chirality (P-­chirality (right) and M-­chirality (left)). The various synthetic methodologies and examples of helicates reported in the literature have been gathered in a review by Hopfgartner et al. in 1997,48 and later on by Tsukube and Albrecht.49,50 Briefly, to obtain optically pure helicates, two main types of approaches have been developed: in the first case, a rigid and rather short linker allows mechanical coupling of the metal configurations (thus, the enantiomers are chemically different); in the second option, an optically pure ligand and a rigid linker lead to a diastereomeric excess of one helicity based on the increased thermodynamic stability of one isomer compared to the other(s).51 Several complementary strategies have been reported to isolate optically pure helicates, including the resolution of racemic helicates,51 either by crystallisation52 or chromatography,53,54 and the use of chiral counter ions, to introduce an enantiomeric enrichment in the architectures.55 Actually, the generation of multinuclear helicates produces metal-­based macrocyclic (two-­stranded helicate) or macrobicyclic (three-­stranded helicate) cavities, which are in most cases chiral. This leads to selective molecular recognition processes for small binding partners, which in turn may serve as templates to maintain the structure of the helicates and/or the chirality, by blocking the kinetic racemization process.48,56,57 Within the helicates family, Hannon and coworkers developed the synthesis of ‘cylinders’, dinuclear triple-­helical compounds which are prepared in a single step from a pyridyl-­aldehyde, a diamine and an octahedral metal (usually FeII or NiII) (Figure 9.3B).58 The cylinders also differ from earlier helicates since they are endowed with a certain amount of rigidity along the length of the structure, due to π-­stacking interactions between the rings of the diphenylmethane ‘spacer’. By contrast, the Lehn helicate systems comprise bipyridine ligands linked by flexible alkyl or alkyl ether chains, introducing a higher degree of flexibility into the helical structure.

9.3 SCCs as Anticancer Agents Several small-­molecule transition metal complexes have been studied over the years for their anticancer properties, but only some of them have been approved for use as anticancer agents. Among them, the US FDA-­approved drug cisplatin is used to treat a range of cancers such as carcinomas and germ cell tumors, among others.59,60 Taking inspiration from the clinical success of cisplatin, supramolecular metal-­containing complexes are also under investigation as experimental cytotoxic anticancer agents. This section summarizes some of the most investigated systems, including coordination and organometallic supramolecular scaffolds, highlighting their main features and design principles.

View Online

Supramolecular Metal-based Structures for Applications in Cancer Therapy

225

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

9.3.1 Cytotoxic PdII and PtII SCCs Palladium(ii) and platinum(ii) ions are both d8 transition metals which adopt a square planar geometry. In SCCs, the metal precursor can be cis-­capped to allow only two coordination sites available for complexation to multidentate ligands to form discrete SCC architectures.13 As representative examples, dinuclear PtII and PdII metallacycles, featuring both amide-­based dipyridyl ligands and 1,1′-­bis(diphenylphosphino) ferrocene ligands, have been studied as anticancer agents in vitro in comparison with cisplatin.61 Interestingly, both metallacycles displayed increased antiproliferative effects (IC50 values between 4 and 36 µM depending on the cell line tested) compared to their metal precursor (IC50 values between 33 and 144 µM) and organic ligand (IC50 > 100 µM), suggesting that the structure of the intact metallacycle is essential for the observed activity.61 Notably, the platinum(ii)-­based metallacycle was found to be a more potent inhibitor of cell proliferation against brain (IC50 = 4.5 µM), head and neck (IC50 ca. 13 µM) and thyroid (IC50 ca. 12 µM) cancer cell lines than cisplatin (IC50 ca. 70, 73 and 50 µM, respectively), and yet had a lower effect against non-­cancerous cells (IC50 ca. two-­fold higher than cisplatin for the Pt-­metallacycle).61 The mechanism of action of both metallacycles has been further investigated in vitro against the brain tumor cell line T98G.61 The results suggest that both compounds induce oxidative stress and cell death by apoptosis. In vitro fluorescence microscopy studies showed that the cancer cells readily internalized both metallacycles, and that upon cell uptake the fluorescent ligand was released. As well as 2D metallacycles, 3D supramolecular architectures using PtII have been studied for their anticancer properties. For example, a highly charged [Pt6L4]12+ metallacage was studied for its cytotoxic behavior and mechanism of action in human ovarian cancer cells.62 The metallacage displayed a similar cytotoxicity range as cisplatin towards a range of human cancer cell lines, including A2780, HT-­29, A549 and MCF-­7. However, the cage was ca. seven-­fold less toxic than cisplatin towards normal lung cells. Using atomic absorption spectroscopy to evaluate the intracellular Pt content, the 3D SCCs were found to be localized inside the cell nucleus.62 In fact, it was shown that the mechanism of action involved non-­covalent binding of the compound to DNA via intercalation. [Pd2L4]4+ supramolecular helicates are also promising anticancer agents. Studies by Crowley and coworkers of the effect of different ligands (functionalized trispyridyl scaffolds with rigid alkene linker vs. benzotriazoles, hexane-­triazoles and PEG-­triazoles-­bisfunctionalised phenyl rings) on the biological activity of Pd2L4 helicates have been carried out, and showed a direct correlation between the stability of the helicate in biological media and its cytotoxicity.63,64 Similar [Pd2L4]4+ (L = 1,3-­bis-­hexanetriazole phenyl) helicates were also found to be up to seven-­fold more toxic (IC50 ca. 6 µM) than cisplatin against the cisplatin-­resistant MDA-­MB-­231 cell line.63 In this case, the [Pd2L4]4+ helicates were shown to induce cell death by disruption of the cell membrane.63 Finally, examples of 3D metallacages, of the type Pd2L4,

View Online

226

Chapter 9

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

have also been reported as potential cytotoxic agents in cancer cells, either alone or by encapsulation of a cytotoxic agent. These latter examples are described in Section 9.4.1.

9.3.2 Cytotoxic Ruthenium(ii)–Arene SCCs Typically, ruthenium(ii) ions (d6 electronic configuration) adopt a hexa-­ coordinated octahedral geometry. Hence, in order to produce discrete supramolecular entities, rather than extended coordination polymers or MOFs, an auxiliary ligand is needed to block some of the several RuII coordination sites. In many cases this is achieved by forming dinuclear ruthenium(ii)–arene ‘clip’ complexes before linking these clips together via pyridine-­containing multidentate ligands to form a range of 3D and 2D supramolecular polyhedra.65–71 Biologically active Ru-­based coordination and organometallic complexes72 have recently prompted analogous studies of ruthenium–arene SCCs with a particular focus on their anticancer properties. Ruthenium(ii) metallacycles had already reported in the late 1990s and displayed properties, such as water solubility and stability, that render them suitable for biological applications. It was initially postulated that cytotoxicity associated with supramolecular RuII complexes was due to their intracellular dissociation, and subsequent binding of the released ruthenium cations to proteins and DNA, causing extensive cell damage and apoptosis.66 However, these complexes have been shown to cause cell death also by triggering excessive autophagy, the controlled process of recycling dysfunctional or destroyed proteins and organelles via lysosome digestion.73 Within a series of ruthenium(ii)–arene metallarectangles with different paneling linkers, one has been shown to be more potent in vitro against multidrug-­resistant human colon cancer cells (HCT-­15/CLO2, IC50 ca. 16.5 µM) compared to cisplatin and doxorubicin,65 suggesting that the mechanism of cytotoxic action of these supramolecular structures is different from those of classical anticancer metallodrugs and requires further investigation. Similarly to metallacycles, ruthenium(ii)–arene metallabowls have also been developed and tested for their antiproliferative properties in vitro against a range of cancer cell lines (colorectal, gastric and liver cancer cells).71,73,74 Among them, a metallabowl featuring 8-­dihydroxy-­1,4-­naphthaquinonato ligands was two-­fold more active than both cisplatin and doxorubicin against HCT-­15 cells (IC50 ca. 7 µM for the metallabowl; IC50 ca. 13 µM for cisplatin; IC50 ca. 16 µM for doxorubicin).71 Further investigations showed that, upon metallabowl exposure, the expression of two known colorectal cancer suppressors, p53 and the adenomatous polyposis coli (APC) gene, increased in HCT-­15 cells.71,74 Three-­dimensional ruthenium–arene SCCs have also been developed as anticancer agents. By introducing tridentate, planar ligands to the binuclear arene RuII ‘clips’, a range of hexanuclear RuII metallacages have been

View Online

Supramolecular Metal-based Structures for Applications in Cancer Therapy

227

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

i

reported by variation of the paneling linker ([Ru2(p-­PrC6H4Me)2(OO∩OO)] [CF3SO3]2 (OO∩OO = 2,5-­dioxydo-­1,4-­benzoquinonato [dobq], 5,8-­dihydro  xy-­1,4-­naphthaquinonato (donq) and 6,11-­dihydroxy-­5,12-­naphthacenedio nato [dotq], etc.)).75–78 Of these 3D systems, the dinuclear ‘clip’ containing donq as the bridging ligand was found to be the most potent inhibitor of cell viability against a range of cancer cell lines (IC50 values between 25 and 90 µM, whereas all the other tested metallacages were non-­cytotoxic).66,79 Alongside ruthenium–arene complexes exhibiting antiproliferative effects per se, examples have been reported whereby hexaruthenium(ii) metallaprisms encapsulate molecules, and the resulting host–guest systems possess antiproliferative effects.80

9.3.3 DNA-­targeted SCCs Most of the numerous cytotoxic SCCs mentioned in previous sections are non-­selective for cancer cells, thus, leading to possible side effects. With the aim of developing tumor-­directed SCCs, the initial strategy exploited the chirality of some of these scaffolds to direct molecular recognition of specific biological targets, namely nucleic acids. A first study investigated the mode of binding of a binuclear FeII triple-­stranded helicate (cylinder) to a DNA model by NMR spectroscopy and computational modelling techniques.81 The results obtained suggested that the helicate binds to the DNA major groove. It was also revealed that although a racemic mixture of the chiral helicate was introduced to the double-­stranded oligonucleotide, only the M-­enantiomer was able to bind DNA, causing a change in its conformation.81 The mode of binding of the FeII cylinder was further studied and revealed that the helix preferentially binds to short (8–10 base pair) purine– pyrimidine tracts within the DNA sequence.82 The affinity for specific DNA sequences proved a promising feature to target cancer cells via binding of the helicates to oncogenes.82 Later on, it was also discovered that FeII cylinders have a high binding specificity for RNA three-­way junctions,72 as well as for certain non-­canonical secondary DNA structures, such as DNA bulges83,84 and G-­quadruplex DNA (G4).85 In particular, targeting telomeric G4s and stabilization of these structures has been shown to inhibit telomerase activity, leading to cell death.86 To that end, a pair of enantiomeric FeII helicates, which were soluble in aqueous media, were synthesized and their affinity for human telomeric G4s was assessed.85 The P-­enantiomer FeII helicate was found to bind strongly and selectively to the G4, whereas the M-­enantiomer showed no association. Furthermore, the strong binding affinity to G-­quadruplex DNA translated into strong inhibition of telomerase activity. Another non-­canonical DNA structure which has been reported as a promising target for binuclear metallahelicates (FeII and RuII based) are triple-­stranded ‘Y-­shaped’ junctions.82,87–92 These secondary structures form during DNA transcription and replication, and as such are promising targets to achieve cell cycle control. It was also shown that their stabilization

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

228

Chapter 9

severely inhibits the function of polymerase enzymes, and, indeed, this has been postulated as the mode of action accounting for the cytotoxicity of the helicates.89 Further studies of DNA binding of FeII helicates revealed that, to facilitate strong binding to the major groove of duplex DNA, a rigid helicate is preferred over the analogous flexible helicates.93 Moreover, the rigid FeII helicate under investigation was a potent cytotoxic agent against cisplatin-­ resistant human ovarian carcinoma cells.93 In addition to Fe-­based helicates, the self-­assembled platinum(ii) molecular square [Pt(en)(4,4′-­dipyridyl)]4 (en = ethylenediamine, Figure 9.4A) has been reported to be an efficient G-­quadruplex binder and telomerase inhibitor.94 Molecular modeling studies suggested that the square arrangement of the four bipyridyl ligands and the highly electropositive nature of the overall complex, as well as hydrogen-­bonding interactions between the ethylenediamine ligands and phosphates of the DNA backbone, all contribute to the observed strong binding affinity to G4. More recently, a supramolecular [Pt2L2]6+ binuclear metallacycle with large, planar 2,7-­diaza-­p yrene-­based ligands has been explored for its DNA-­ binding properties.95 This interaction caused DNA bending, which in turn prevented DNA processing and replication. Moreover, the metallacycle exhibited antiproliferative effects in cancer cells and a different spectrum of activity with respect to cisplatin.95 More recently, supramolecular PtII quadrangular boxes, featuring L-­shaped 4,4′-­bipyridine ligands, were shown to bind native and G-­quadruplex DNA motifs in a size-­dependent fashion.96 Specifically, three dinuclear Pt molecular squares of distinct size (ranging between 110 and 220 Å) showed biological activity against cancer cells and heavily influenced the expression of genes known to form G-­quadruplexes in their promoter regions. Interestingly, the smallest

Figure 9.4 (A) Schematic representation of a multinuclear PtII metallacycle studied

as a quadruplex binder and telomerase inhibitor.94 (B) Schematic representation of a Ru8 cage bearing porphyrin ligands studied as a nucleic acid binder.97

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

Supramolecular Metal-based Structures for Applications in Cancer Therapy

229

Pt-­box displays less activity, but enhanced selectivity for the G4 promoter c-­Kit, as shown by FRET (Fluorescence Resonance Energy Transfer) DNA melting assays.96 Finally, 3D SCCs, including boxes and cubes, have been reported to be able to interact with nucleic acids. For example, porphyrin-­based scaffolds were designed and linked via ruthenium complexes used as bridging blocks able to connect two porphyrin units and create octa-­ruthenium supramolecular cubes (Figure 9.4B).97 The G4-­binding properties of the cubes were studied by different techniques, including fluorescence intercalation displacement (FID) and surface plasmon resonance, and the results obtained showed strong interactions with different G4 models, but also scarce selectivity with respect to duplex DNA.

9.4 SCCs as Drug Delivery Systems The focus of this chapter is on SCCs that show promise for biomedical applications not only due to their intrinsic anticancer potential, but also for their favorable properties as drug delivery systems. This is particularly relevant to cancer chemotherapy, where the success rate remains limited, primarily due to the scarce selectivity of drugs for the tumor tissue, often resulting in severe toxicity and in the development of drug resistance. So far, lipid nanosystems, such as liposomes and micelles, along with virus-­inspired vectors and polymeric particles, dendrimers as well as inorganic nanoparticles, have been studied to deliver bioactive compounds to tumor sites. However, such targeted constructs have several limitations: for example, polymers and dendrimers often require considerable synthetic effort and can be plagued by low yields and largely amorphous final structures, while nanoparticles often present issues of toxicity and lack of biodegradability.98 In this context, supramolecular metallacages feature a number of properties that make them attractive candidates for future drug delivery systems. For example, the rigid, porous structure offers a secure cavity for small drug molecules, to protect them from metabolism, and the ability to modify the ligand structure both pre-­ and post-­self-­assembly allows for the properties of the resulting cage to be improved. Furthermore, since metallacages, in contrast to MOFs, are discrete chemical entities, the issues of solubility in an aqueous environment can be potentially overcome. Despite these attractive features, SCC drug delivery is still in its infancy.99 So far, SCC drug delivery systems have been based on: (1) both encapsulation of a drug, driven by hydrophobicity of the cargo drug molecule and the host cavity, and non-­covalent interactions within the host cavity, as well as, (2) bonding of a prodrug species to the SCC architecture. In the latter case, the active part of the prodrug can then be cleaved from the SCC via external chemical stimuli to deliver the drug in a controlled manner. In the next sections, representative examples of both strategies will be provided.

View Online

230

Chapter 9

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

9.4.1 SCCs as Drug Delivery Systems for Anticancer Agents Pioneering work on encapsulation-­based SCC drug delivery systems has focused on the study of hexaruthenium metallacages (metalla-­prisms),21 based on previous studies on ruthenium(ii) metallacycles,100 as they were found to encapsulate lipophilic molecules. Specifically, the cationic hexanuclear metalla-­prism [(p-­c ymene)6Ru6(tpt)2(dhbq)3]6+ (tpt = 2,4,6-­trispy ridyl-­1,3,5-­triazine; dhbq = 2,5-­dihydroxy-­1,4-­benzoquinonato) was used to encapsulate the hydrophobic PdII and PtII complexes [M(acac)2] (M = metal, acac = acetylacetonato) (Figure 9.5A).21 The metalla-­prism is water soluble and moderately cytotoxic (IC50 ca. 23 µM) against human ovarian A2780 cancer cells, while the [M(acac)2] complexes are completely inactive due to their inherent lack of solubility in water. Interestingly, the encapsulated [Pd(acac)2] ([Pd(acac)2]⊂[( p-­c ymene)6Ru6(tpt)2(dhbq)3]6+) was 20-­ fold more cytotoxic (IC50 ca. 1 µM) than the empty metalla-­prism (IC50 ca. 23 µM), while the effect was markedly less pronounced with the encapsulated [Pt(acac)2] complex (IC50 ca. 12 µM). This initial study provided the proof-­of-­concept for what was defined as ‘the Trojan horse strategy’ of hiding a cytotoxic agent in the cavity of a metallacage until, after internalization within the diseased cells, the drug can be released and perform its cell-­killing activity. Subsequently, a hexaruthenium metallacage of the type [Ru6( p-­iPrC6  H4Me)6(tpt)2(C6H2O4)3]6+ was investigated for the release mechanism of encapsulated fluorescent pyrene derivatives and for its anticancer properties in vitro.20 The fluorescence of the pyrene derivative is quenched upon encapsulation, allowing for the release of the molecule to be monitored by fluorescence spectroscopy.20 Concerning the antiproliferative properties, while the free pyrene derivative and the cage complex alone were scarcely cytotoxic (IC50 >20 µM and 16 µM, respectively), the host–guest complex was considerably more active (IC50 ca. 6 µM).20 Fluorescence microscopy data suggested that the increased cytotoxicity was due to an increased uptake of the poorly soluble pyrene derivative into the cancer cell after being delivered by the water-­soluble cage complex. The encapsulation properties of the hexaruthenium metallacage with a series of functionalized fluorescent pyrene derivatives was characterized using NMR (1H, 2D, DOSY) spectroscopy and electrospray ionization mass spectrometry (ESI-­MS).101 The synthesis of the host–guest system proceeded via a two-­step process: first the di-­ruthenium half-­sandwich molecular ‘clip’ reacts with silver triflate to produce a reactive intermediate; afterwards, a 2 : 1 solution of the tridentate tpt ligand and the pyrene derivative is introduced to form the host–guest complex via self-­assembly.101 The antiproliferative properties of the vacant cage and the pyrene-­cage complexes were studied in A2780 ovarian cancer cells, and the host–guest complexes showed the lowest IC50 values.101 The study also demonstrated that the hexaruthenium cage complexes can improve the efficacy of insoluble inhibitors in vitro.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

View Online

Figure 9.5 (A) Schematic representation and corresponding X-­ray structure (CCDC no.

673229) of a [[Ru2L′]3L2]6+ cage encapsulating [Pt(acac)2].21 (B) Schematic representation and corresponding X-­ray structure (CCDC no. 853227) of an exo-­functionalized [Pd2L4]4+ metallacage encapsulating two equivalents of cisplatin.104 (C) Schematic representation and corresponding X-­ray structure (CCDC no. 902397) of a [Pd2L4]4+ cage encapsulating two equivalents of corannulene.112 (D) Schematic representation of a [Pt3L3]6+ hexagon exo-­functionalized with three equivalents of a PtIV prodrug.123 The X-­ray structures are depicted using the Discovery Studio software.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

232

Chapter 9

The effect of the portal size of the hexaruthenium metallacage complex on the retention of the planar guest molecules, [Pd(acac)2] complex and 1-­(4,6-­ dichloro-­1,3,5-­triazin-­2-­yl)pyrene, was also investigated.102 Thus, three hexaruthenium cages were prepared by extending the polycyclic aromatic system in the di-­ruthenium bridging ligands, using the 1,4-­naphthoquinonato, 1,4-­anthraquinonato and 5,12-­naphthacenedionato analogues, which progressively decreased the portal size of the cage, while the internal cavity remained largely the same. The host–guest properties of these water-­soluble supramolecular drug delivery systems were studied in solution by NMR and fluorescence spectroscopy. The results showed that the complex with the largest pore size (estimated to be approximately 7.4 × 10.2 Å by molecular modeling) is more stable, suggesting that a larger pore size facilitates the entrance of the guest molecule into the cage, while the smaller pore size retains the guest molecule more effectively.102 Inductively Coupled Plasma Mass Spectrometry (ICP-­MS) and fluorescence microscopy allowed assessment that all cages deliver the host to intracellular organelles and the mechanisms of uptake involve endocytosis/macropinocytosis rather than passive diffusion across the cell membrane.102 Other examples of SCCs as drug delivery systems, based on other transition metals, include surface-­functionalized porous coordination nanocages of CuII and 5-­(prop-­2-­ynyloxy)isophthalic acid (pi), bearing a water-­solubilizing polymer (PEG5k), which have been successfully synthesized using a ‘click chemistry’ approach.103 The scaffold is composed of 12 di-­copper paddlewheel clusters and 24 isophthalate moieties, with eight triangular and six square windows that are roughly 8 and 12 Å across, respectively. The internal cavity has a diameter of ca. 15 Å and the cage has high stability in aqueous medium. In addition, the drug loading and release capacity of the cages has been evaluated using the anticancer drug 5-­fluorouracil (5-­FU).103 Drug-­release experiments were carried out by dialyzing the drug-­loaded Cu(pi)–PEG5k against phosphate-­buffered saline (PBS) solution at room temperature. Interestingly, around 20% of the loaded drug was released during the first 2 hours, while a flatter release curve can be observed up to 24 hours. The latter slow release has been associated with the slow diffusion rate of 5-­FU caused by the strong interaction between the Lewis acid sites in Cu(pi) and the basic site of 5-­FU. Within the M2L4 cage family, Crowley and coworkers designed a cationic [Pd2L4]4+ cage using (2,6-­bis(pyridin-­3-­ylethynyl)pyridine) as the bidentate ligand,104 based on previous work by Fujita and coworkers.105 Various methods, including 1H NMR, ESI-­MS and XRD, showed that a quadruple-­stranded cage was formed, with the internal cavity lined with the nitrogen atoms from the central pyridine of the ligand. Interestingly, the encapsulation of the anticancer drug cisplatin within the metallacage cavity was demonstrated by XRD studies, revealing that two molecules of the drug could be contained (Figure 9.5B).104 The release of cisplatin was facilitated by the introduction of competing ligands (4-­dimethylaminopyridine or Cl−) to disassemble the cage, and the disassembly process was monitored via 1H NMR and ESI-­MS.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

Supramolecular Metal-based Structures for Applications in Cancer Therapy

233

More recently, Casini and coworkers explored similar cationic [Pd2L4]4+ systems and developed the exo-­functionalization of the ligand scaffold to add different functionalities (e.g. fluorescent tags).106,107 Structural studies by 1H NMR and XRD were performed demonstrating encapsulation of cisplatin.108,109 Furthermore, the cytotoxicity of the [Pd2L4]4+ cages have been tested in vitro against a small panel of human cancer cells, showing scarce or moderate antiproliferative activities depending on the ligand scaffold.109 The activity of encapsulated cisplatin in the benzyl alcohol-­exo-­functionalized PdII cage was evaluated against SKOV-­3 ovarian cancer cells, showing that the encapsulated cisplatin displayed an important increase in cytotoxic potency (IC50 = 1.9 ± 0.5 µM) compared to free cisplatin (IC50 = 15.4 ± 2.2 µM) and the vacant cage complex (IC50 = 11.6 ± 1.7 µM).109 The [Pd2L4]4+ metallacages possess fluorescence properties and fluorescence microscopy studies allowed monitoring of their uptake in cancer cells. Notably, most of the reported metallacages and their precursors were non-­toxic in healthy rat liver tissue ex vivo, making them suitable for application as drug delivery systems.109,110 Water solubility and stability under physiological conditions are both crucial for the biological application of SCCs. Unfortunately, [Pd2L4]4+ cages of this type are scarcely soluble in water, despite their positive charge. However, increasing the hydrophilic character of these systems has been demonstrated via the introduction of water-­soluble moieties in their scaffold, including PEG.111 Control of the host–guest properties of the cavity defined by the SCC is another essential feature to implement for drug encapsulation. For example, in contrast to the previously mentioned metallacages, anthracene-­ based PtII-­ and PdII-­linked coordination capsules provide a characteristic spherical cavity closely surrounded by polyaromatic frameworks (Figure 9.5C).112 The isolated cavity features a diameter of ca. 1 nm and a volume of ca. 600 Å3.113 These systems can accommodate various neutral molecules in the confined cavity through hydrophobic interactions, but also π-­stacking, in aqueous solution. Cages of this type were recently reported to be able to encapsulate spherical (paracyclophanes, adamantanes and fullerene), planar (pyrenes, triphenylene and caffeine) and bowl-­shaped molecules (corannulene),112–114 while fluorescence microscopy studies allowed investigation of the intracellular accumulation of these systems.114 It is worth mentioning that the capsules – even without their guest molecules – manifest very pronounced cytotoxic effects, which may prevent their application as ‘pure’ drug delivery systems. The observed trends in the anticancer activity of the capsules and their host–guest complexes correlate with their different stabilities towards glutathione, estimated by NMR-­based kinetic experiments.114 The data suggest the glutathione-­triggered disassembly of the capsular structures in cells as a potential activation pathway for their cytotoxic activity. Selective accumulation of metallacages in tumors was initially hypothesized to occur via the enhanced permeability and retention (EPR) effect, which has been widely explored in cancer therapy for delivery via passive

View Online

Chapter 9

234

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

78

targeting. In fact, the EPR effect targeting solid tumors has been predominantly shown to be involved in the passive targeting of drugs with a molecular weight of more than 40 kDa (20–200 nm in diameter) and for low molecular weight drugs tethered to or encapsulated in drug carriers such as polymeric conjugates, liposomes and polymeric nanoparticles.115 However, for supramolecular metallacages, such as [Pd2L4]4+, with a molecular weight of ca. 2–3 kDa and diameter of ca. 10–15 Å, the EPR effect is not likely to influence their delivery. Furthermore, it should be noted that the success in preclinical in vivo studies of drug accumulation in tumors due to the EPR effect has so far not translated into success in clinical trials.116 Therefore, active tumor-­t argeting mechanisms are necessary to improve selectivity for cancerous cells, for example via the conjugation of cancer cell-­specific ligands to the outside of a SCC drug delivery system. However, this concept has scarcely been explored so far. For example, one study has shown non-­covalent peptide coating on self-­assembled M12L24 coordination spheres,117 while encapsulation of a protein within a Pd12L14 cage has been achieved by appropriate endo-­functionalization of the ligands.118 In this latter case, ligands were first tethered to the protein and then the cage was reconstituted via self-­assembly upon addition of other ligands and metal precursors. Notably, this was the first example of encapsulation of a protein within synthetic host molecules and may reveal novel strategies to deliver proteins at specific sites and to control their function.118 To achieve targeting of SCCs, the first example of bioconjugation of self-­ assembled [Pd2L4]4+ cages to a model linear peptide via amide bond formation of a –COOH (or –NH2) exo-­functionalized ligand/cage was recently reported by Casini and coworkers.119 In the future, other types of exo-­ functionalization, other than amide bond formation, for tethering metallacages to peptides or antibodies (e.g. via click chemistry120) deserves further investigation.

9.4.2 Prodrug-­based SCCs In order to generate a supramolecular drug delivery system delivering a prodrug, Lippard and coworkers used a tridentate ligand and a PtII precursor (Pt(ethane-­1,2-­diamine)) to obtain a cationic [Pt4L6]12+ cage.121 The metallacage was found to form a stable host–guest complex with an adamantyl PtIV prodrug in a 1 : 4 ratio in D2O upon sonication, followed by heating at 80 °C.121 The hydrophobic adamantyl moiety of the prodrug molecule was postulated to be securely encapsulated within the hydrophobic cavity of the hexanuclear cage, as suggested by 1D and 2D NMR spectroscopy. The cage encapsulating the PtIV complex showed antiproliferative activities (micromolar range) against a small panel of human cancer cell lines and exhibited higher cytotoxicity than the PtIV prodrug and of the hexanuclear PtII cage alone.121 The mechanistic hypothesis is

View Online

Supramolecular Metal-based Structures for Applications in Cancer Therapy

235

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

IV

that the Pt complex can be reduced intracellularly by ascorbic acid, thus, releasing cisplatin, 1-­adamantylamine and succinic acid, as suggested by NMR spectroscopy and mass spectrometry. In fact, the cytotoxic effect of the host–guest system showed traits characteristic of cisplatin-­induced cell damage. Further studies on a similar water-­soluble [Pt4L6]12+ metallacage revealed that fluorescein, a highly emissive fluorophore, could also be encapsulated within the cavity.122 Furthermore, upon encapsulation of fluorescein in PBS a color change from green (associated with free fluorescein, λabs = 490 nm) to pink (λabs = 503 nm) was observed. This observation facilitated the characterization of the encapsulation process by UV–Visible spectroscopy, which revealed a stoichiometry of 1 : 1 host : guest and an experimental dissociation constant (Kd) of 2.5 µM. Afterwards, the fluorescein moiety was conjugated to a PtIV prodrug and encapsulated in the metallacage, and the cellular uptake and release of the guest prodrug could be studied in HeLa cells in vitro by fluorescence microscopy.122 Apart from encapsulation strategies of drugs within the cavity of 3D metallacages, a recent example of conjugation of prodrugs to the surface of a 2D SCC has been described as an alternative drug delivery approach. Specifically, a supramolecular [Pt3L3]6+ hexagon was formed by self-­assembly between a dinuclear PtIV precursor and a bidentate ligand conjugated to organoplatinum species so that the resulting supramolecular hexagons would deliver three equivalents of cisplatin upon reduction of the PtIV prodrug in the intracellular environment (Figure 9.5D).123 The antiproliferative effects of the hexagonal SCC were tested towards a range of cancer cell lines, sensitive or resistant to cisplatin, and the results showed that the supramolecular hexagon was more potent than cisplatin against all the tested cell lines,123 even though no control with a PtIV prodrug alone was reported. Mechanistic studies suggested the prodrug induced apoptosis by causing DNA damage due to the release of cisplatin upon reduction of the PtIV complex in the intercellular environment. Interestingly, quantification of the intracellular metal content suggested that the increased potency of the supramolecular hexagon, compared to cisplatin, was due to its higher cellular uptake.123

9.5 In Vivo Studies on Anticancer SCCs Although the field of SCCs as anticancer therapeutics is still in its infancy, a few preliminary in vivo experiments have been carried out. Thus, the anticancer activity of two ruthenium(ii)–arene metallacycles, one with a 2D rectangular geometry and one featuring a metallabowl geometry (Figure 9.6A), were studied in vitro against human gastric carcinoma cells (AGS) and human colon cancer cells (HCT-­15).73 The results showed that both metallacycles had comparable antiproliferative activities with respect to cisplatin and doxorubicin. Following these promising results, a hollow

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

236

Chapter 9

Figure 9.6 Schematic representations of (A) two arene–ruthenium(ii) metallacycles (left: 2D rectangular geometry; right: ‘metallabowl’ geometry)73 and of (B) a rhomboidal platinum(ii) metallacycle,124 studied in vivo for their anticancer properties.

fiber assay was conducted, whereby a semipermeable fiber impregnated with the HCT-­15 cells was implanted into the intraperitoneal and subcutaneous compartments of nude mice.73 The two ruthenium–arene metallacycles were then administered to the impregnated nude mice, and the animals were left for 7 days before the hollow fibers were removed and the tumors examined. The study revealed that the metallabowl-­t ype metallacycle was a more potent inhibitor of cancer cell growth than the metallarectangle. However, both these ruthenium–arene scaffolds were not as effective inhibitors of cell proliferation as cisplatin in the hollow fibers located in the intraperitoneal and subcutaneous regions of the host mice.73 The mechanism of induced cell death was investigated and the study revealed that both metallacycles induced autophagy in HCT-­15 cells, and again the metallabowl was more potent than the metallarectangle, in line with the observed anticancer activity.73 A luminescent, 2D platinum(ii) metallacycle of rhomboidal geometry has also been studied in vitro and in vivo for its anticancer activity (Figure 9.6B).124

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00215

Supramolecular Metal-based Structures for Applications in Cancer Therapy

237

Of note, the compound remained intact upon cellular internalization and did not photobleach under the conditions of the confocal microscopy experiment. Preliminary in vitro studies against lung (A549) and cervical carcinoma (HeLa) cells confirmed rapid cellular uptake of the platinum metallacycle.124 Afterwards, a mouse tumor xenograft model, using nude mice injected in the subcutaneous region with human breast cancer cells (MDA-­MB-­231 cells), was used for the in vivo study. The tumors were allowed to reach a volume of 200 mm3 before drug administration. Mice were treated with a solution (300 µL) of Pt-­metallacycle at a concentration of 0.6 mg mL−1, administered via intraperitoneal injection every 3 days over 30 days. The study revealed that after 30 days, a 64% median tumor volume reduction was observed in treated mice with respect to controls. Furthermore, the tumor growth inhibition, measured by the change in volume of the tumor throughout the length of the experiment and defined by the T/C ratio (in %, corresponds to the ratio between the Treatment (T) over the Control (C)),125 was calculated as 36%, well below the National Cancer Institute standard (as the lower threshold for tumor inhibition of HSA > dGMP, where HSA is human serum albumin and dGMP is 2′-­deoxyguanosine monophosphate.30 A comparison of cisplatin and transplatin analogues exposed simultaneously to a model peptide containing adjacent Cys/His and an oligonucleotide, revealed a binding preference of cisplatin towards the oligonucleotide, especially to adjacent guanine residues, while the transplatin analogues tended to form adducts with Cys from the peptide.43 This may indicate the molecular manifestation of the distinct biological effects of cisplatin and their transplatin analogues. Moreover, experiments were performed by incubating a protein mixture of ub, cyt c and superoxide dismutase simultaneously with cisplatin and RAPTA-­C to investigate metallodrug competition for similar binding sites.44 While both compounds were found to form adducts with ub in isolation, RAPTA-­C preferentially coordinated to ub in this competitive assay and no platinum adducts were observed, suggesting that they bear the same binding site on the protein. In contrast, both ruthenium and platinum adducts were detected on cyt c, while no adducts were observed on superoxide dismutase. A fusion molecule between RAPTA-­C and chlorambucil connected via the arene ligand was investigated with respect to the competitive reactivity towards 9-­ethylguanine. The nucleobase may be alkylated by chlorambucil via nucleophilic attack on the aziridium or by coordination to ruthenium.45 Interestingly, under kinetic control, 9-­ethylguanine coordinated quickly to ruthenium in the absence of other biomolecules. This was not observed when His was co-­incubated, which itself is more efficient in coordinating to ruthenium. Consequently, His occupied the ligand sites of ruthenium in a

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

252

Chapter 10

bidentate manner, which forced alkylation of 9-­ethylguanine via chlorambucil. Under thermodynamic control, alkylation of 9-­ethylguanine via the chlorambucil moiety acted as an irreversible sink and this is indicative of the molecular difference between metallodrugs and organic reactive electrophiles (i.e. coordinative vs. covalent bonding).

10.2.2  LC-­ESI-­MS The LC-­ESI-­MS setup was originally applied to competitive model studies of metallo-­prodrugs with small molecules. Compared to DI-­ESI-­MS, LC-­ESI-­MS benefits from a higher dynamic range and can deal with an increased sample complexity. However, the ligand exchange kinetics must be well above the required time for sample preparation and analysis to reduce ligand scrambling. In a study in 2005, RM175 (Figure 10.1) was reacted with GSH and cyclic guanosine monophosphate (cGMP) under competitive conditions. By a combined analysis of LC-­ESI-­MS and NMR it was found that under aerobic conditions and at physiological pH, RM175 readily formed an adduct with GSH by coordinating to the cysteine thiol and then oxidizing the cysteine to the sulfenato moiety, which in turn facilitated its release from the metal.46 Accordingly, even an excess of 200-­fold of GSH with respect to cGMP, did not inhibit the metallodrug from coordinating to N7 of cGMP, which provided an intriguing perspective for circumventing cellular detoxification and resistance. Although examples are few, LC-­ESI-­MS is nowadays most successfully applied to sample the binding preferences of metallo-­prodrugs towards mixtures of proteins or protein domains, which are difficult to analyse by DI-­ ESI-­MS. In such a setting, the binding preference of cyclometallated gold(iii) organometallics was investigated under competitive conditions towards two zinc finger domains.47 In earlier studies, the investigated compounds were shown to interact with zinc fingers and to inhibit the DNA damage recognition protein poly(ADP-­ribose)-­polymerase-­1 (PARP-­1), which contains a zinc finger as well.48,49 Zinc fingers are crucial DNA recognition domains in proteins and regulate cellular functions.48 They can be classified according to their Zn coordination environment and peptide sequence into at least five families when ignoring higher order structures.50 In general, the Cys2His2 coordination environment for Zn2+ is characteristic for transcription factors, while the Cys2HisCys environment is present in PARP-­1.47 Thus, the N-­terminal zinc finger protein domains of PARP-­1 and a generic transcription factor domain were used to assess the binding preference of the compounds towards PARP-­1. While the investigated compounds interacted with both zinc finger domains individually, by expulsing Zn2+ and forming a gold finger, they clearly displayed a binding preference for the PARP-­1 zinc finger under competitive conditions.47 Intriguingly, the selectivity for PARP-­1 interaction was completely abrogated by substituting the 2-­benzylpyridine chelating ligand for N-­phenylpyridin-­2-­amine, i.e. a –CH2 for an –NH in the ligand backbone, emphasizing the importance of secondary interactions as a prerequisite for selectivity. This work also provided a means to assess the binding selectivity towards PARP-­1 as a binding preference ratio.

View Online

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

253

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

10.2.3  CZE-­ESI-­MS CZE represents an orthogonal separation technique compared to LC, because the separation principle of electrophoresis is dictated by the ion radius and charge.51 Separation according to radius and charge is a powerful tool to account for ligand exchange reactions of metallo-­prodrugs and to separate interaction products that differ in charge, but only slightly in mass (Figure 10.3A).51 Compared to the previously described experimental approaches, CZE-­ESI-­MS is well suited to determine the binding preferences of metallo-­ prodrugs towards complex mixtures of intact proteins and oligonucleotides, which may possess overall opposite charges depending on the pI of the protein and the pH of the solution. Determining the DNA vs. protein-­binding preference of metallo-­prodrugs under competitive conditions is of fundamental interest to medicinal inorganic chemistry. In addition to the NCP mentioned above, which provides a static thermodynamic picture of metallodrug selectivity,15 CZE-­ESI-­MS offers the possibility for time-­dependent analysis of binding preferences under kinetic control. The drawback of the reduced complexity compared to the NCP is partly compensated by the short measurement time of roughly 10 minutes per sample.52,53 In this setting using a mixture of ub and an 8-­mer oligonucleotide containing adjacent guanines, the binding preferences of cisplatin (Figure 10.3A), RM175 and an organometallic ruthenium(ii) maltolato metallodrug were evaluated.52 While both cisplatin and RM175 showed preferred binding to guanines of the oligonucleotide as revealed by online top-­down analysis, the activated ruthenium(ii) maltolato derivative selectively interacted with ub by binding to Met1.52 Moreover, RAPTA-­C also preferentially interacted with the oligonucleotide over the protein

Figure 10.3  (A)  Analysis of the reaction mixture containing the protein ubiq-

uitin (ub), an 8-­mer oligonucleotide (DNA), and cisplatin revealed a binding preference towards the oligonucleotide and highlights sequential hydrolysis of cisplatin. EIE = Extracted ion electropherogram. (B) Online fragmentation experiments of the [DNA + Pt(NH3)2] adduct by collision-­induced dissociation revealed the location of the platinum-­binding sites on the oligonucleotide with single nucleotide resolution, as illustrated in bold letters in the inserted sequence. Reproduced from ref. 52 with permission from the Royal Society of Chemistry.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

254

Chapter 10

under the conditions used, but, intriguingly, coordinated mainly via adenine and cytosine to the oligonucleotide.53 The binding sites of the metallodrugs on the protein and oligonucleotide were resolved with single-­residue resolution in some cases (Figure 10.3B). The binding preference was quantified by means of analysing identical samples by CZE-­ICP-­MS and subsequent integration of the respective metal traces that correlated with the phosphorous traces of the DNA phosphate backbone.53 Hyphenated methods based on ICP-­MS in addition to ESI-­MS provide excellent opportunities to study the distribution of metals in the biological context because they yield quantitative information on adduct formation or stability of a metallodrug.54–56

10.3  I dentifying Targets of Metal-­based Anticancer Agents With the NCP and the CZE-­ESI-­MS approaches described above, a metallo-­ prodrug may be evaluated with respect to its tendency of targeting either the genome or the proteome.17,52 Being model systems these assays provide a suitable starting point for identifying protein targets if protein binding was evidenced.9 For example, microwestern arrays were used to monitor changes in protein expression of 259 proteins of which 66 displayed significant regulation upon treatment with cisplatin.57 The protein 3-­phosphoinositide-­dependent protein kinase-­1 (PDK1) was strongly regulated and inhibition of PDK1 showed synergistic effects as a combination treatment with cisplatin. However, UniProt lists over 20 000 reviewed proteins in the human proteome58 and such a number requires even more powerful, high-­performance instrumentation for a comprehensive analysis. It is thus not surprising that validated protein targets are known for only a handful of metallo(-­pro)drugs and systematic approaches to identify protein targets are still rare. Moreover, the previously discussed ligand exchange reactions as well as biologically induced redox processes add to the challenge of target identification of non-­platinum metallo(-­pro) drugs.59,60 Emerging strategies for determining potential protein-­binding partners are based on metallomic and proteomic approaches, including photo-­reactive probes, fluorophores and affinity purification strategies as the most prominent examples, often in an interdisciplinary setting (Figure 10.4). These approaches rely mainly on gel-­based or gel-­free MS-­based techniques. One early study reported on a photo-­reactive DNA–cisplatin adduct to identify DNA–protein cross-­links that provided insight into the cellular processes of dealing with Pt-­induced DNA lesions.61 Nuclear extracts of cancer cells were exposed to the DNA–drug complex and proteins were crosslinked by photo-­activation. After isolation of these adducts, interacting proteins were identified by western blot and/or mass spectrometric analysis, revealing especially DNA repair and high mobility group-­domain proteins.61,62

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

255

Figure 10.4  Structures  of chemical probes employed for target identification.

Proteomics-­based target identification was carried out with (A) bismuth, copper and nickel probes containing a fluorophore and an azide, (B) N-­heterocyclic carbene gold(iii) containing a diazirine and an alkyne, (C) gold(iii) porphyrin containing a benzophenol and an alkyne and (D) an organometallic ruthenium prodrug containing a biotin moiety.

ICP-­MS is a powerful metallomics tool for the identification and quantification of metal ions in biological systems. The harsh ionization conditions atomize the samples leading to the loss of molecular information; however, this allows detecting the distribution of metals in the cellular system. For example, outer membrane protein A was identified as a target of cisplatin in Escherichia coli after the platinum content in bands of a 1D polyacrylamide gel was determined by laser ablation (LA)-­ICP-­MS. The band with the highest platinum content was subsequently digested and analysed by ESI-­MS.63 Size exclusion chromatography (SEC), in combination with ICP-­MS, did not identify targets of the ruthenium-­based compound KP1019; however, it indicated a cytosolic location and binding partners of high molecular weight.64 This finding was confirmed by another SEC-­ICP-­MS investigation and the binding pattern of the drug was further examined by applying capillary HPLC combined with ESI triple quadrupole MS.65 The latter suggested 15 proteins that potentially bind to the ruthenium anticancer agent in the cytosol. Combining ICP-­MS with continuous-­flow gel electrophoresis (GE-­ICP-­MS) enables the determination of both the metal and the interacting protein partners, as exemplified for the anti-­ulcer compound, colloidal bismuth subcitrate (CBS), acting on Helicobacter pylori.66 A T-­connection enabled simultaneous online metal detection by ICP-­MS and fraction collection. Seven

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

256

Chapter 10

bismuth-­binding proteins were detected by GE-­ICP-­MS and proteins in the corresponding fraction were identified by matrix-­assisted laser desorption/ ionization time-­of-­flight MS (MALDI-­TOF-­MS). One of the first proteomics-­based methods for target identification using molecular MS was based on hyphenating two-­dimensional liquid chromatography (2D-­LC) to ESI-­MS. This approach was termed multidimensional protein identification technology (MudPIT) and showed a higher dynamic range and more protein identifications compared to classical gel-­based proteomics at the time.67 MudPIT analysis requires that the metal displays a characteristic isotopic distribution and that the metallodrug–protein adducts are kinetically inert and hence not affected by the digestion and the diverse chromatographic conditions. The method was applied to E. coli cells treated with the organometallic [(η6-­p-­cymene)RuCl2(DMSO)]68 (where DMSO = dimethylsulfoxide) or cisplatin,69 identifying 5 and 31 potential protein targets, respectively, among which were stress-­regulated proteins as well as helicases. Notably, there is no need for prior derivatization of the compounds with this approach. Lai et al. employed a fluorescent probe to visualize metal-­binding proteins in living cells and as a means of sampling the metalloproteome.70 Interestingly, fluorescence was induced primarily when Ni, Bi, Co, Cu or Fe were complexed on the probe and the remaining ligand-­binding site was occupied by either His or Cys. Photo-­activation induced covalent anchoring to the metalloprotein and enabled subsequent proteomic analysis. Fluorescent spots separated by two-­dimensional gel electrophoresis (2-­DE) were identified by MALDI-­TOF-­MS revealing potential protein targets of the investigated metal. Potential targets were further examined with bioinformatic analyses, e.g. gene ontology (GO) enrichment or protein–protein interaction networks. The method was tested specifically using Ni2+ and Cu2+, identifying 44 and 54 binding proteins, respectively. Proteins identified in H. pylori by the Ni2+-­tracer as Ni binding included HpUreB (urease), HpUreG (GTPase) and HpHspA (chaperone). Furthermore, the results revealed an involvement in the homeostatic process, transition metal ion binding, oxidation–reduction processes and antioxidant activity. The Cu2+-­tracer was examined in HeLa cells and the binding partners were assigned to GO terms, among which were nuclear mRNA splicing, nucleotide binding, cell redox homeostasis, translation factor activity and energy generation. Comparative proteomics and immobilized-­metal affinity chromato-­ graphy (IMAC) for bismuth ions were performed to investigate CBS in H. pylori employing 2-­DE and MALDI-­TOF-­MS analysis.71 Eight proteins were regulated upon treatment and four of them were determined to bind Bi3+ ions, namely HSPA, HSPB, NAPA and TSAA, indicating a direct targeting of these proteins. To investigate interactions of bismuth-­based drugs in H. pylori, Wang et al. further employed a fluorescent Bi3+-­tracer, 2-­DE-­based proteomics and GE-­ICP-­MS.72 To quantitatively assess protein regulation upon exposure to CBS, an iTRAQ (isobaric tags for relative and absolute

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

257

quantitation) proteome-­profiling approach using LC-­MS/MS analysis was chosen. Next to 63 binding proteins, 116 regulated ones were identified. Several potential targets were selected for validation by cellular thermal-­ shift assay (CETSA) or western blotting. Furthermore, bioinformatic analysis, including interaction networks and GO annotations, revealed the influence of bismuth on reactive oxygen species (ROS), pH buffering and other pathways essential for H. pylori survival, and supported a multi-­ targeted mechanism of action. A photo-­affinity–click chemistry approach, combined with a quantitative proteomics experiment using stable isotope labelling by amino acids in cell culture (SILAC), identified HSP60 as the major molecular target of the anticancer-­active gold(iii) meso-­tetraphenylporphyrin.73 Target validation was performed by a biotin pull-­down with immunoblot detection using the HSP60 antibody, direct incubation with the protein followed by photo-­affinity labelling and click reaction, as well as saturation-­transfer difference NMR measurements. Intracellular effects were monitored by CETSA, showing a higher stability of HSP60 upon drug treatment, mitochondrial substrate activity measurements and protein fluorescence quenching upon exposure to the metallodrug. A pincer gold(iii) N-­heterocyclic carbene (NHC) complex with anticancer activity was modified with a photo-­active diazirine group and a clickable alkyne moiety as a probe for target identification in cancer cells.60 The implementation of such clickable photo-­affinity probes enabled covalent binding to the interacting protein upon UV irradiation and subsequent anchoring to a reporter, e.g. biotin or a fluorophore. Similarly, 2-­DE combined with MALDI-­TOF-­MS analysis was used for protein identification. Six potential targets were identified for the above-­mentioned Au–NHC compound in HeLa cells. A pull-­down experiment with streptavidin beads of the biotinylated proteins and subsequent HPLC-­Orbitrap MS analysis confirmed five of these. Probe distribution and co-­localisation with the potential targets were examined with cellular imaging experiments and immunofluorescence, respectively. Further validation strategies included determining protein functionalities after treatment with the gold(iii)–NHC compound, including substrate turnovers and degradation of the respective proteins. A combination of shotgun proteomics and bioinformatic analysis revealed the modulation of the eukaryotic initiation factor 2α signalling pathway as the main regulatory effect of the gold(iii) complex, which can result from the proteomically indicated and functionally verified degradation of vimentin, among others. Moreover, target profiling of a RAPTA derivative was investigated by a chemical proteomics approach based on affinity purification.74 The biotin-­ modified compound was immobilized on streptavidin beads and exposed separately to untreated and drug-­treated whole-­cell lysates of an ovarian cancer cell line. In principle, this differential procedure allowed the elimination of non-­specific interactors since the binding sites of potential targets are already saturated in the lysates of treated cells and cannot interact

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

258

Chapter 10

with the immobilized drug anymore, thus revealing drug-­specific interaction partners. In the case of the RAPTA derivative, 18 potential protein interactors were identified, including cancer-­related, cell cycle-­regulating, histone-­related and ribosomal proteins, as well as extracellular growth factors. Recently, an integrated proteomics approach combining target and response profiling revealed the first protein target for a ruthenium(arene) pyridinecarbothioamide, termed plecstatin, i.e. a metallo-­prodrug.59 Target hypotheses were derived from affinity purification in which a biotin-­ modified plecstatin was immobilized on streptavidin beads. A differential strategy of pre-­saturation was followed, similarly to that for the RAPTA compound, using whole-­cell lysates of a colon carcinoma cell line. A target profiling plot was constructed to visualize potential targets (Figure 10.5A). The fold-­change of normal versus competitive pull-­down represents the y-­axis, and thus the target enrichment, while the significance of the enrichment is expressed as a p-value on the x-­axis. Furthermore, protein intensity and specificity are represented by bubble size and colour, respectively. Of several hundred identified species, only outer dense fiber protein 2 (ODF2) and plectin (PLEC) were characterized by high enrichment, significance and specificity. In a second dimension, protein regulation upon drug treatment at sub-­cytotoxic concentrations was determined by response profiling. Target–response networks were then generated using the results of the two proteomic dimensions, linking potential targets to cellular effects upon drug treatment, and giving a strong indication that plectin may be a potential target (Figure 10.5B). Immunofluorescence-­staining experiments with wild-­ type and plectin-­deficient mouse keratinocytes validated plectin as the main

Figure 10.5  (A)  Visualization of a differential target profiling experiment depicting

probable potential targets for plecstatins. (B) Combination of target profiling and response profiling yields target–response networks for each probable target as shown here for plectin. Adapted from ref. 59 with permission from John Wiley and Sons, Copyright © 2017 Wiley-­ VCH Verlag GmbH & Co. KGaA, Weinheim.

View Online

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

259

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

cellular target for plecstatin with significant inhibitory effects on cancer invasiveness. Importantly, plecstatin-­1 was tumour inhibiting when administered orally in the invasive B16 melanoma tumour model.59

10.4  E  lucidating Modes of Action of Metal-­based Anticancer Agents Disease states can be considered as out-­of-­equilibrium conditions and cells attempt to restore homeostasis by adapting RNA and protein expression levels, which can be specific to stress responses. Similarly, drugs aim at restoring homeostasis or killing diseased cells by initiating changes of biologically active molecules in cells on their own.75 Therefore, the analysis of protein expression changes can help to elucidate a plausible mode of action of a drug.76 The difference between the terms drug effect and mode of action should be noted in this context. For example, the mode of action of cisplatin involves DNA binding followed by the induction of apoptosis, while drug effects may include the regulation of proteins of the Nrf2 pathway or glutathione. The­ potential benefits of systems pharmacology approaches to investigate metallo­ drug effects in cells have recently been emphasized.77 This section briefly­ covers recent investigations on transcriptional profiling before discussing the most promising methods of MS-­based proteome profiling to elucidate metallo(-­pro)drug effects on cells.

10.4.1  Transcriptional Profiling Transcription is the first step of gene expression and the resulting mRNA is translated into proteins. Thus, the study of transcriptional changes upon drug treatment can provide evidence of cellular adaptions to drug exposure.78–80 Specifically, investigations using gene chips were performed on pentadentate pyridine–iron(ii) complexes81 and on dinuclear C^N^C cyclometallated gold(iii)–phosphine complexes.82 In combination with connectivity map analysis, the latter highlighted the compound's similarity to inhibitors of thioredoxin reductase and inducers of ER stress. Western blots and reverse transcription polymerase chain reaction (RT–PCR) analysis confirmed ER stress induction. Moreover, direct incubation of the gold(iii) compound with rat TrxR1 inhibited the catalytic function of the protein. Furthermore, an up-­ regulation of the death receptor 5 (DR5) was detected by transcriptomic analysis. Its involvement in the mode of action was verified through treatment in the presence and absence of DR5 activation by siRNA knock-­down experiments. These analyses enabled a proposal for the anti-­tumour mechanism of this gold(iii) compound indicating involvement of the death receptor pathway and ER stress induction as well as TrxR inhibition. Recently, the iridium(iii) complex [Ir(CpXph)(azpyNMe2)Cl]PF6 (CpXph = phenyl-­ tetramethylcyclopentadienyl, azpyNMe2 = N,N-­dimethylphenylazopyridine) was investigated by a pharmaco-­genomic study.83 The whole transcriptome

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

260

Chapter 10

of ovarian cancer cells was analysed by next-­generation sequencing (RNA-­ seq) after 4, 12, 24 and 48 hours of drug exposure. Not only did the number of differentially expressed genes vary between the different time points, but also did their regulation, e.g. 12 hours of treatment resulted in a higher amount of down-­regulated transcripts, whereas a similar proportion of up-­ and down-­ regulation was observed after 24 hours. Antioxidant response pathways were already induced after a short time of exposure. Even though important transcription factors for antioxidant reaction were up-­regulated, the downstream genes were not, which may indicate the inability of these cells to cope with the ROS induced by this metallodrug.

10.4.2  Proteome Profiling Changes in protein abundance upon drug treatment result from upstream transcriptional and translational regulation and enable direct evaluation of cellular responses to stress, including drug treatments.76 Mechanistic investigations on the proteome are usually performed by 2-­DE in combination with MALDI-­TOF-­MS or by shotgun proteomics. The latter involves tryptic digestion of the entire protein sample, analysis by liquid chromatography coupled to tandem mass spectrometry (LC-­MS/MS) and subsequent bioin­ formatic evaluation (Figure 10.6A). A proteomics study using 2-­DE and MALDI-­TOF-­MS revealed that a gold(iii) porphyrin regulated several proteins involved in stress response, translation, protein degradation, proliferation and induced apoptosis.84,85 The results were supported by evaluating the mitochondrial transmembrane potential and caspase activity, immunocytochemistry staining, cell cycle and western blot analysis and indeed suggested a multi-­targeted mechanism leading to apoptosis. A further study was reported on cisplatin-­sensitive and -­resistant cancer cells treated with the gold-­based Auoxo6 and auranofin.86,87 The cisplatin-­ sensitive cell line showed only a few regulated proteins upon treatment and indicated that both compounds followed similar modes of action. Among the regulated proteins were those involved in cellular redox homeostasis (up-­ and down-­regulation of peroxiredoxin 1 and 6, respectively), activating caspase 3 (up-­regulation of heterogeneous ribonucleoprotein H) and triggering apoptosis (up-­regulation of ezrin).87 In cisplatin-­resistant cells, on the other hand, auranofin mainly influenced proteins related to protein degradation, while Auoxo6 affected proteins involved in trafficking, as well as mRNA splicing and stability.86 Both compounds down-­regulated thioredoxin-­like protein 1, which is involved in oxidative stress defence, and thioredoxins are considered probable targets for many therapeutic gold compounds.10 A C^N^N cyclometallated gold(iii) drug candidate was investigated with respect to effects in ovarian cancer cells by a similar gel-­based proteomics approach.88 Out of 87 differentially expressed proteins, 29 proteins were up-­ regulated upon treatment, including representatives of protein synthesis,

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

261

Figure 10.6  (A)  Depiction of the workflow for analysing tryptic digests by shotgun

proteomics. The final protein list usually contains several thousand proteins. (B) Venn diagram of regulated proteins after treating SW480 cancer cells with representative gallium(iii), lanthanum(iii), ruthenium(iii) and platinum(ii) drug candidates. (C) The treatment effect of arsenic trioxide on the colon carcinoma cell lines HCT116 and SW480 can be represented in a spider web by the average regulation of grouped proteins according to DNA mismatch repair (DMM), ROS protection (ROS protc.), ER stress protection (ER protc.), endocytosis, cell adhesion and mitochondrial proteins. Adapted from ref. 101 with permission from John Wiley and Sons, Copyright © 2017 Wiley-­VCH Verlag GmbH & Co. KGaA, Weinheim.

stress response and the cytoskeleton. Bioinformatic processing, including network analysis, GO and pathway enrichment, revealed that the overall protein alterations mainly concerned mitochondrial processes and glucose metabolism, as validated by metabolic and western blot analysis. Several 2-­DE MALDI-­TOF-­MS approaches also investigated regulation upon cisplatin treatment giving a better insight into the DNA damage response, down-­regulation of NF-­κB (mediated by TRAF2) and apoptosis induction via death receptor as well as mitochondria.89,90 Moreover, tissue samples from patients suffering from cervical cancer were analysed before and after combinatory treatment with paclitaxel and cisplatin.91 The 13 proteins found to be differentially expressed after chemotherapy indicated a down-­regulation of chaperones and energy production, as well as induction of apoptosis. Cisplatin resistance was also studied thoroughly using 2-­DE MALDI-­TOF-­MS, but is not discussed here.92–95

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

262

Chapter 10

Drug delivery may be advantageous for toxic anticancer compounds like cisplatin. Protein regulation in the human liver cancer cell line HepG2 exposed to cisplatin alone, as well as cisplatin bound to human serum transferrin, were compared by 2-­DE and MALDI-­TOF/TOF-­MS analysis, which revealed distinct mechanisms of apoptosis induction.96 Selected regulated proteins were verified by western blots and RT–PCR, such as the proteasome activator PA28γ. Furthermore, two-­dimensional difference in gel electrophoresis (2D-­DIGE), where fluorescent-­labelled protein samples are mixed, separated by 2-­DE on the same gel and then examined by laser scanners, overcoming the problem of gel-­to-­gel variations, was applied to investigate metallodrug-­induced proteome alterations. MALDI-­TOF-­MS and ESI-­ion trap MS/MS were used for protein identification.97 Ovarian cancer cells were treated with a dithiocarbamate gold(iii) dibromide or a phenanthroline-­derived dinuclear oxo-­bridged gold(iii) compound in this setting.98 Significant alterations were detected for eight and ten spots for the two drugs, respectively, compared to controls. Additionally, eight proteins were regulated by both treatments, including the up-­regulation of Ubiquilin-­1, which interacts with the proteasome. Thus, the results indicate an influence of the gold compounds on protein degradation pathways. Similarly, the two ruthenium compounds NAMI-­A and RAPTA-­T were investigated, revealing related modes of action while displaying distinct differences to cisplatin treatment.97 A SILAC-­based (phospho-­)proteomics approach indicated alterations in DNA damage response pathways and cytoskeleton organisation in embryonic stem cells upon cisplatin treatment.99 Additional transcriptomic analysis via quantitative RT–PCR revealed correlation of mRNAs, proteins and phosphoproteins involved in the same pathways, while the relative abundance of transcripts and the corresponding proteins hardly correlated otherwise. Importantly, phosphoproteomics revealed activated signalling pathways by genotoxic stress. Protein regulation in breast cancer cells upon exposure to the ruthenium(ii)-­based drug candidates RAPTA-­T and RAPTA-­EA, where EA = etacrynic acid, were investigated by the so-­called functional identification of target by expression proteomics.100 LC-­MS/MS-­based shotgun proteomics of at least two cell lines treated with various drugs were compared, ranking proteins according to their regulation in an attempt to connect drug targets and mechanisms by proteome-­profiling experiments. The observed regulation upon RAPTA-­T or RAPTA-­EA exposure suggested a broad mechanism of action, including suppression of tumorigenicity and metastasis for RAPTA-­T, while the EA derivative seemed to cause an oxidative stress response. Moreover, LC-­MS/MS-­based shotgun proteomics was utilized to obtain response profiles of the two colon carcinoma cell lines HCT116 and SW480 to different metallo(-­pro)drugs, i.e. arsenic trioxide, [tris(8-­quinolinolato)gallium(iii)], [tris(1,10-­phenanthroline)lanthanum(iii)] trithiocyanate, sodium trans-­[tetrachlorobis(1H-­indazole)ruthenate(iii)] (IT-­139) and [{(1R,2R,4R)-­4-­ methyl-­1,2-­cyclohexanediamine}oxalatoplatinum(ii)].101 Cells were fractionated into cytoplasmic and nuclear extracts. Trypsin-­digested samples were

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

263

analysed by shotgun proteomics (Figure 10.6A). Each experiment generated around 30 000 fragment spectra of peptides that comprised information for matching against protein sequences. Up to 5100 proteins were identified per run with a false discovery rate of 0.01 and under the condition that at least one unique peptide for a given protein was found. Several hundred proteins were regulated in the cytoplasmic and nuclear fractions and allowed to completely reconstruct the mechanism of action of arsenic trioxide in the drug-­resistant SW480 cancer cell line, verifying the applicability of such an approach. Specifically, heme oxygenase 1 was up-­regulated upon treatment, while the protein target PML was down-­regulated, i.e. degraded. In contrast, the HCT116 cancer cells showed a truncated response to arsenic trioxide treatment, probably through differences in the basal protein abundances in the cell­ lines. Moreover, the regulatory effects observed by these novel metallodrugs­ were surprisingly diverse and only five protein regulations were common in SW480 cancer cells (Figure 10.6B). This suggested specific modes­ of action of these representatives from different classes of metallodrugs. Regulated proteins of the different metallodrugs were subsequently categorized into six functional groups; namely, protection from oxidative stress, endocytosis, mitochondrial function, DNA repair, protection from ER stress and cell adhesion. The average regulation of proteins of each group was then represented in a spider web to visually compare global metallodrug effects in cells, as exemplified for the colon carcinoma cell lines HCT116 and SW480 treated with arsenic trioxide (Figure 10.6C). The central hexagon denotes zero regulation, whereas the outer hexagon and the central point represent an average four-­fold up-­or down-­regulation, respectively. For example, the induction of ROS protection upon treatment with arsenic trioxide is stronger in SW480 cancer cells than in HCT116. The findings from the proteomic experiments were further validated by cell adhesion assays and glutathione measurements. Such approaches have considerable value in generating initial mechanistic hypotheses on phenotypically discovered therapeutic agents.

10.5  Conclusion Mass spectrometry has significantly contributed to the understanding of the reactivity and biological effects of metallo(-­pro)drugs designed as anticancer agents. Methods that detect metallo(-­pro)drugs directly are used to evaluate reaction products under competitive conditions by direct infusion as well as by coupling to separation techniques. Such model systems allow a detailed molecular picture to be drawn about the activation pathways of metallo-­ prodrugs and the binding preferences of their activated species at different levels of complexity. Compared to these methods, MS-­based proteomic techniques are able to sample metallodrug effects or potential targets in the cellular context and in an indirect manner, i.e. not through sampling the metal species. These investigations challenged the generalising preconception that metallo(-­pro)drugs

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

264

Chapter 10

are poorly selective at a molecular level. In fact, they provided evidence that at least some of these drug candidates display an unexpected binding preference and target selectivity towards proteins, in contrast to the established platinum(ii) class. Such global approaches build an exclusive path to generate a hypothesis on potential molecular targets of phenotypically discovered next-­generation metal-­based anticancer agents for further validation. All together, these techniques can provide vital information for translating investigational compounds into the clinical setting. However, the successful application of these “omics” techniques is crucially dependent on biologically meaningful data interpretation using robust dimension-­reducing processes.

Abbreviations 2-­DE Two-­dimensional gel electrophoresis 2D-­DIGE Two-­dimensional difference in gel electrophoresis azpyNMe2 N,N-­Dimethylphenylazopyridine CBS Colloidal bismuth subcitrate CETSA Cellular thermal-­shift assay cGMP Cyclic guanosine monophosphate CpXph Phenyl-­tetramethylcyclopentadienyl Cys l-­Cysteine Cyt c Cytochrome c CZE Capillary zone electrophoresis dGMP 2′-­Deoxyguanosine monophosphate DI Direct infusion DMSO Dimethylsulfoxide DR5 Death receptor 5 ER Endoplasmic reticulum ESI Electrospray ionization GE Gel electrophoresis GO Gene ontology GRP78 Endoplasmic reticulum chaperone BiP GSH Glutathione HSA Human serum albumin His l-­Histidine ICP Inductively coupled plasma IMAC Immobilized-­metal affinity chromatography iTRAQ Isobaric tags for relative and absolute quantitation LA Laser ablation LC Liquid chromatography LC-­MS/MS Liquid chromatography coupled to tandem mass spectrometry MALDI-­TOF Matrix-­assisted laser desorption/ionization time-­of-­flight Met l-­Methionine

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

265

MS Mass spectrometry MudPIT Multidimensional protein identification technology NCP Nucleosome core particle NER Nucleotide-­excision repair NHC N-­heterocyclic carbene PARP-­1 Poly(ADP-­ribose)-­polymerase-­1 PDK1 3-­Phosphoinositide-­dependent protein kinase-­1 PML Protein PML ROS Reactive oxygen species RT–PCR Reverse transcription polymerase chain reaction SEC Size exclusion chromatography SeCys l-­Selenocysteine SeMeCys Methyl-­l-­selenocysteine SILAC Stable isotope labelling by amino acids in cell culture TrxR1 Thioredoxin reductase-­1 Ub Ubiquitin

Acknowledgements The authors are grateful to the Faculty of Chemistry, University of Vienna, for financial support.

References 1. D. Wang and S. J. Lippard, Nat. Rev. Drug Discovery, 2005, 4(4), 307. 2. L. B. Travis, S. D. Fossa, S. J. Schonfeld, M. L. McMaster, C. F. Lynch, H. Storm, P. Hall, E. Holowaty, A. Andersen, E. Pukkala, M. Andersson, M. Kaijser, M. Gospodarowicz, T. Joensuu, R. J. Cohen, J. D. Boice, G. M. Dores and E. S. Gilbert, J. Natl. Cancer Inst., 2005, 97(18), 1354. 3. N. A. Bowden, Cancer Lett., 2014, 346(2), 163. 4. J. Zhu, Z. Chen, V. Lallemand-­Breitenbach and H. de The, Nat. Rev. Cancer, 2002, 2(9), 705. 5. H. de The, Nat. Rev. Cancer, 2018, 18(2), 117. 6. H. A. Burris, S. Bakewell, J. C. Bendell, J. Infante, S. F. Jones, D. R. Spigel, G. J. Weiss, R. K. Ramanathan, A. Ogden and D. Von Hoff, ESMO Open, 2016, 1(6), e000154. 7. https://clinicaltrials.gov/ct2/show/NCT03053635, accessed 2018-­07-­03. 8. C. Roder and M. J. Thomson, Drugs R&D, 2015, 15(1), 13. 9. S. M. Meier-­Menches, C. Gerner, W. Berger, C. G. Hartinger and B. K. Keppler, Chem. Soc. Rev., 2018, 47(3), 909. 10. O. Karaca, S. M. Meier-­Menches, A. Casini and F. E. Kuhn, Chem. Commun., 2017, 53(59), 8249. 11. S. Jurgens and A. Casini, Chimia, 2017, 71(5), 308. 12. T. T. Zou, C. T. Lum, C. N. Lok, J. J. Zhang and C. M. Che, Chem. Soc. Rev., 2015, 44(24), 8786.

View Online

266

13. G. Jaouen, A. Vessieres and S. Top, Chem. Soc. Rev., 2015, 44(24), 8802. 14. E. Meggers, Chem. Commun., 2009, (9), 1001. 15. B. Wu, P. Droge and C. A. Davey, Nat. Chem. Biol., 2008, 4(2), 110. 16. S. M. Meier, M. Hanif, Z. Adhireksan, V. Pichler, M. Novak, E. Jirkovsky, M. A. Jakupec, V. B. Arion, C. A. Davey, B. K. Keppler and C. G. Hartinger, Chem. Sci., 2013, 4(4), 1837. 17. Z. Adhireksan, G. E. Davey, P. Campomanes, M. Groessl, C. M. Clavel, H. J. Yu, A. A. Nazarov, C. H. F. Yeo, W. H. Ang, P. Droge, U. Rothlisberger, P. J. Dyson and C. A. Davey, Nat. Commun., 2014, 5, 3462. 18. C. G. Hartinger, M. Groessl, S. M. Meier, A. Casini and P. J. Dyson, Chem. Soc. Rev., 2013, 42(14), 6186. 19. M. Wenzel and A. Casini, Coord. Chem. Rev., 2017, 352, 432. 20. D. C. Crans, Coord. Chem. Rev., 2017, 352, 398. 21. S. M. Meier, M. V. Babak, B. K. Keppler and C. G. Hartinger, ChemMedChem, 2014, 9(7), 1351. 22. M. Groessl and C. G. Hartinger, Anal. Bioanal. Chem., 2013, 405(6), 1791. 23. C. R. Centerwall, K. A. Tacka, D. J. Kerwood, J. Goodisman, B. B. Toms, R. L. Dubowy and J. C. Dabrowiak, Mol. Pharmacol., 2006, 70(1), 348. 24. J. Lokich and N. Anderson, Ann. Oncol., 1998, 9(1), 13. 25. J. Reedijk, Platinum Met. Rev., 2008, 52(1), 2. 26. D. M. Templeton and H. Fujishiro, Coord. Chem. Rev., 2017, 352, 424. 27. E. Kansanen, S. M. Kuosmanen, H. Leinonen and A. L. Levonen, Redox Biol., 2013, 1(1), 45. 28. W. H. Ang, I. Khalaila, C. S. Allardyce, L. Juillerat-­Jeanneret and P. J. Dyson, J. Am. Chem. Soc., 2005, 127(5), 1382. 29. T. Peleg-­Shulman and D. Gibson, J. Am. Chem. Soc., 2001, 123(13), 3171. 30. S. M. Meier, M. Hanif, W. Kandioller, B. K. Keppler and C. G. Hartinger, J. Inorg. Biochem., 2012, 108, 91. 31. A. Rathgeb, A. Bohm, M. S. Novak, A. Gavriluta, O. Domotor, J. B. Tornmasino, E. A. Enyedy, S. Shova, S. Meier, M. A. Jakupec, D. Luneau and V. B. Arion, Inorg. Chem., 2014, 53(5), 2718. 32. S. M. Meier, M. Novak, W. Kandioller, M. A. Jakupec, V. B. Arion, N. Metzler-­Nolte, B. K. Keppler and C. G. Hartinger, Chem. -­ Eur. J., 2013, 19(28), 9297. 33. S. M. Meier, Y. O. Tsybin, P. J. Dyson, B. K. Keppler and C. G. Hartinger, Anal. Bioanal. Chem., 2012, 402(8), 2655. 34. M. Hanif, H. Henke, S. M. Meier, S. Martic, M. Labib, W. Kandioller, M. A. Jakupec, V. B. Arion, H. B. Kraatz, B. K. Keppler and C. G. Hartinger, Inorg. Chem., 2010, 49(17), 7953. 35. S. M. Meier, C. Gerner, B. K. Keppler, M. A. Cinellu and A. Casini, Inorg. Chem., 2016, 55(9), 4248. 36. M. V. Babak, M. Pfaffeneder-­Kmen, S. M. Meier-­Menches, M. S. Legina, S. Theiner, C. Licona, C. Orvain, M. Hejl, M. Hanif, M. A. Jakupec, B. K. Keppler, C. Gaiddon and C. G. Hartinger, Inorg. Chem., 2018, 57(5), 2851.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Chapter 10

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

267

37. F. Y. Wang, J. Bella, J. A. Parkinson and P. J. Sadler, J. Biol. Inorg. Chem., 2005, 10(2), 147. 38. T. Zhao and F. L. King, J. Biol. Inorg. Chem., 2011, 16(4), 633. 39. R. F. S. Lee, L. Menin, L. Patiny, D. Ortiz and P. J. Dyson, Anal. Chem., 2017, 89(22), 11985. 40. W. Kandioller, E. Balsano, S. M. Meier, U. Jungwirth, S. Goschl, A. Roller, M. A. Jakupec, W. Berger, B. K. Keppler and C. G. Hartinger, Chem. Commun., 2013, 49(32), 3348. 41. M. Knipp, A. V. Karotki, S. Chesnov, G. Natile, P. J. Sadler, V. Brabec and M. Vasak, J. Med. Chem., 2007, 50(17), 4075. 42. A. Casini, A. Karotki, C. Gabbiani, F. Rugi, M. Vasak, L. Messori and P. J. Dyson, Metallomics, 2009, 1(5), 434. 43. L. Cubo, M. Groessl, P. J. Dyson, A. G. Quiroga, C. Navarro-­Ranninger and A. Casini, ChemMedChem, 2010, 5(8), 1335. 44. A. Casini, C. Gabbiani, E. Michelucci, G. Pieraccini, G. Moneti, P. J. Dyson and L. Messori, J. Biol. Inorg. Chem., 2009, 14(5), 761. 45. A. A. Nazarov, S. M. Meier, O. Zava, Y. N. Nosova, E. R. Milaeva, C. G. Hartinger and P. J. Dyson, Dalton Trans., 2015, 44(8), 3614. 46. F. Y. Wang, J. J. Xu, A. Habtemariam, J. Bella and P. J. Sadler, J. Am. Chem. Soc., 2005, 127(50), 17734. 47. M. N. Wenzel, S. M. Meier-­Menches, T. L. Williams, E. Ramisch, G. Barone and A. Casini, Chem. Commun., 2018, 54(6), 611. 48. F. Mendes, M. Groessl, A. A. Nazarov, Y. O. Tsybin, G. Sava, I. Santos, P. J. Dyson and A. Casini, J. Med. Chem., 2011, 54(7), 2196. 49. U. A. Laskay, C. Garino, Y. O. Tsybin, L. Salassa and A. Casini, Chem. Commun., 2015, 51(9), 1612. 50. S. M. Quintal, Q. A. dePaula and N. P. Farrell, Metallomics, 2011, 3(2), 121. 51. H. Holtkamp, G. Grabmann and C. G. Hartinger, Electrophoresis, 2016, 37(7–8), 959. 52. C. Artner, H. U. Holtkamp, W. Kandioller, C. G. Hartinger, S. M. Meier-­ Menches and B. K. Keppler, Chem. Commun., 2017, 53(57), 8002. 53. C. Artner, H. U. Holtkamp, C. G. Hartinger and S. M. Meier-­Menches, J. Inorg. Biochem., 2017, 177, 322. 54. M. Groessl, M. Terenghi, A. Casini, L. Elviri, R. Lobinski and P. J. Dyson, J. Anal. At. Spectrom., 2010, 25(3), 305. 55. G. Hermann, P. Heffeter, T. Falta, W. Berger, S. Hann and G. Koellensperger, Metallomics, 2013, 5(6), 636. 56. S. Theiner, M. Grabarics, L. Galvez, H. P. Varbanov, N. S. Sommerfeld, M. Galanski, B. K. Keppler and G. Koellensperger, Dalton Trans., 2018, 47(15), 5252. 57. A. L. Stark, A. G. Madian, S. W. Williams, V. Chen, C. Wing, R. J. Hause Jr, L. A. To, A. L. Gill, J. L. Myers, L. K. Gorsic, M. F. Ciaccio, K. P. White, R. B. Jones and M. E. Dolan, J. Proteome Res., 2017, 16(11), 4227. 58. A. Bateman, M. J. Martin, C. O'Donovan, M. Magrane, E. Alpi, R. Antunes, B. Bely, M. Bingley, C. Bonilla, R. Britto, B. Bursteinas, H.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

268

Chapter 10

Bye-­Ajee, A. Cowley, A. Da Silva, M. De Giorgi, T. Dogan, F. Fazzini, L. G. Castro, L. Figueira, P. Garmiri, G. Georghiou, D. Gonzalez, E. Hatton-­ Ellis, W. Li, W. Liu, R. Lopez, J. Luo, Y. Lussi, A. MacDougall, A. Nightingale, B. Palka, K. Pichler, D. Poggioli, S. Pundir, L. Pureza, G. Qi, S. Rosanoff, R. Saidi, T. Sawford, A. Shypitsyna, E. Speretta, E. Turner, N. Tyagi, V. Volynkin, T. Wardell, K. Warner, X. Watkins, R. Zaru, H. Zellner, I. Xenarios, L. Bougueleret, A. Bridge, S. Poux, N. Redaschi, L. Aimo, G. ArgoudPuy, A. Auchincloss, K. Axelsen, P. Bansal, D. Baratin, M. C. Blatter, B. Boeckmann, J. Bolleman, E. Boutet, L. Breuza, C. Casal-­Casas, E. De Castro, E. Coudert, B. Cuche, M. Doche, D. Dornevil, S. Duvaud, A. Estreicher, L. Famiglietti, M. Feuermann, E. Gasteiger, S. Gehant, V. Gerritsen, A. Gos, N. Gruaz-­Gumowski, U. Hinz, C. Hulo, F. Jungo, G. Keller, V. Lara, P. Lemercier, D. Lieberherr, T. Lombardot, X. Martin, P. Masson, A. Morgat, T. Neto, N. Nouspikel, S. Paesano, I. Pedruzzi, S. Pilbout, M. Pozzato, M. Pruess, C. Rivoire, B. Roechert, M. Schneider, C. Sigrist, K. Sonesson, S. Staehli, A. Stutz, S. Sundaram, M. Tognolli, L. Verbregue, A. L. Veuthey, C. H. Wu, C. N. Arighi, L. Arminski, C. Chen, Y. Chen, J. S. Garavelli, H. Huang, K. Laiho, P. McGarvey, D. A. Natale, K. Ross, C. R. Vinayaka, Q. Wang, Y. Wang, L. S. Yeh and J. Zhang, Nucleic Acids Res., 2017, 45(D1), D158. 59. S. M. Meier, D. Kreutz, L. Winter, M. H. M. Klose, K. Cseh, T. Weiss, A. Bileck, B. Alte, J. C. Mader, S. Jana, A. Chatterjee, A. Bhattacharyya, M. Hejl, M. A. Jakupec, P. Heffeter, W. Berger, C. G. Hartinger, B. K. Keppler, G. Wiche and C. Gerner, Angew. Chem., Int. Ed., 2017, 56(28), 8267. 60. S. K. Fung, T. Zou, B. Cao, P. Y. Lee, Y. M. Fung, D. Hu, C. N. Lok and C. M. Che, Angew. Chem., Int. Ed., 2017, 56(14), 3892. 61. E. R. Guggenheim, D. Xu, C. X. Zhang, P. V. Chang and S. J. Lippard, ChemBioChem, 2009, 10(1), 141. 62. C. X. Zhang, P. V. Chang and S. J. Lippard, J. Am. Chem. Soc., 2004, 126(21), 6536. 63. C. S. Allardyce, P. J. Dyson, F. R. Abou-­Shakra, H. Birtwistle and J. Coffey, Chem. Commun., 2001, (24), 2708. 64. M. Groessl, O. Zava and P. J. Dyson, Metallomics, 2011, 3(6), 591. 65. M. Matczuk, M. Kupiec, J. Legat, K. Pawlak, A. R. Timerbaev and M. Jarosz, Analyst, 2015, 140(10), 3492. 66. L. Hu, T. Cheng, B. He, L. Li, Y. Wang, Y. T. Lai, G. Jiang and H. Sun, Angew. Chem., Int. Ed., 2013, 52(18), 4916. 67. D. A. Wolters, M. P. Washburn and J. R. Yates, Anal. Chem., 2001, 73(23), 5683. 68. J. Will, A. Kyas, W. S. Sheldrick and D. Wolters, J. Biol. Inorg. Chem., 2007, 12(6), 883. 69. J. Will, W. S. Sheldrick and D. Wolters, J. Biol. Inorg. Chem., 2008, 13(3), 421. 70. Y. T. Lai, Y. Yang, L. G. Hu, T. F. Cheng, Y. Y. Chang, M. Koohi-­Moghadam, Y. C. Wang, J. Xia, J. W. Wang, H. Y. Li and H. Z. Sun, Metallomics, 2017, 9(1), 38.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

Enabling Methods to Elucidate the Effects of Metal-based Anticancer Agents

269

71. R. Ge, X. Sun, Q. Gu, R. M. Watt, J. A. Tanner, B. C. Wong, H. H. Xia, J. D. Huang, Q. Y. He and H. Sun, J. Biol. Inorg. Chem., 2007, 12(6), 831. 72. Y. Wang, L. Hu, F. Xu, Q. Quan, Y. T. Lai, W. Xia, Y. Yang, Y. Y. Chang, X. Yang, Z. Chai, J. Wang, I. K. Chu, H. Li and H. Sun, Chem. Sci., 2017, 8(6), 4626. 73. D. Hu, Y. Liu, Y. T. Lai, K. C. Tong, Y. M. Fung, C. N. Lok and C. M. Che, Angew. Chem., Int. Ed., 2016, 55(4), 1387. 74. M. V. Babak, S. M. Meier, K. V. M. Huber, J. Reynisson, A. A. Legin, M. A. Jakupec, A. Roller, A. Stukalov, M. Gridling, K. L. Bennett, J. Colinge, W. Berger, P. J. Dyson, G. Superti-­Furga, B. K. Keppler and C. G. Hartinger, Chem. Sci., 2015, 6(4), 2449. 75. A. Bileck, D. Kreutz, B. Muqaku, A. Slany and C. Gerner, J. Proteome Res., 2014, 13(12), 5989. 76. N. L. Anderson and N. G. Anderson, Electrophoresis, 1998, 19(11), 1853. 77. I. Romero-­Canelon and P. J. Sadler, Proc. Natl. Acad. Sci. U. S. A., 2015, 112(14), 4187. 78. K. K. Jovanović, M. Tanić, I. Ivanović, N. Gligorijević, B. P. Dojčinović and S. Radulović, J. Inorg. Biochem., 2016, 163, 362. 79. A. Bergamo, M. Gerdol, M. Lucafò, C. Pelillo, M. Battaglia, A. Pallavicini and G. Sava, Metallomics, 2015, 7(10), 1439. 80. L. von Stechow, A. Ruiz-­Aracama, B. van de Water, A. Peijnenburg, E. Danen and A. Lommen, PLoS One, 2013, 8(10), e76476. 81. W. L. Kwong, C. N. Lok, C. W. Tse, E. L. M. Wong and C. M. Che, Chem. -­Eur. J., 2015, 21(7), 3062. 82. R. W. Y. Sun, C. N. Lok, T. T. H. Fong, C. K. L. Li, Z. F. Yang, A. F. M. Siu and C. M. Che, Chem. Sci., 2013, 4(5), 1979. 83. J. M. Hearn, G. M. Hughes, I. Romero-­Canelón, A. F. Munro, B. Rubio-­ Ruiz, Z. Liu, N. O. Carragher and P. J. Sadler, Metallomics, 2018, 10(1), 93. 84. Y. Wang, Q. Y. He, C. M. Che and J. F. Chiu, Proteomics, 2006, 6(1), 131. 85. Y. Wang, Q. Y. He, R. W. Y. Sun, C. M. Che and J. F. Chiu, Cancer Res., 2005, 65(24), 11553. 86. F. Guidi, I. Landini, M. Puglia, F. Magherini, C. Gabbiani, M. A. Cinellu, S. Nobili, T. Fiaschi, L. Bini, E. Mini, L. Messori and A. Modesti, Metallomics, 2012, 4(3), 307. 87. F. Magherini, A. Modesti, L. Bini, M. Puglia, I. Landini, S. Nobili, E. Mini, M. A. Cinellu, C. Gabbiani and L. Messori, J. Biol. Inorg. Chem., 2010, 15(4), 573. 88. T. Gamberi, L. Massai, F. Magherini, I. Landini, T. Fiaschi, F. Scaletti, C. Gabbiani, L. Bianchi, L. Bini, S. Nobili, G. Perrone, E. Mini, L. Messori and A. Modesti, J. Proteomics, 2014, 103, 103. 89. G. Zhang, L. Sun, X. Lu, Z. Chen, P. J. Duerksen-­Hughes, H. Hu, X. Zhu and J. Yang, Mutat. Res., 2012, 746(1), 66. 90. E. K. Yim, K. H. Lee, C. J. Kim and J. S. Park, Int. J. Gynecol. Cancer, 2006, 16(2), 690. 91. H. Liu, Y. Han, R. Mi, Y. Zhang, G. Su, H. Wang, X. Zhou, X. Liu and B. Zhu, Int. J. Gynecol. Cancer, 2011, 21(8), 1452.

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00246

270

Chapter 10

92. V. Paulitschke, V. Haudek-­Prinz, J. Griss, W. Berger, T. Mohr, H. Pehamberger, R. Kunstfeld and C. Gerner, J. Proteome Res., 2013, 12(7), 3264. 93. X. D. Yan, L. Y. Pan, Y. Yuan, J. H. Lang and N. Mao, J. Proteome Res., 2007, 6(2), 772. 94. K. Le Moguen, H. Lincet, E. Deslandes, M. Hubert-­Roux, C. Lange, L. Poulain, P. Gauduchon and B. Baudin, Proteomics, 2006, 6(19), 5183. 95. A. Castagna, P. Antonioli, H. Astner, M. Hamdan, S. C. Righetti, P. Perego, F. Zunino and P. G. Righetti, Proteomics, 2004, 4(10), 3246. 96. L. Z. Luo, H. W. Jin and H. Q. Huang, J. Proteomics, 2012, 77, 237. 97. F. Guidi, A. Modesti, I. Landini, S. Nobili, E. Mini, L. Bini, M. Puglia, A. Casini, P. J. Dyson, C. Gabbiani and L. Messori, J. Inorg. Biochem., 2013, 118, 94. 98. F. Guidi, M. Puglia, C. Gabbiani, I. Landini, T. Gamberi, D. Fregona, M. A. Cinellu, S. Nobili, E. Mini, L. Bini, P. A. Modesti, A. Modesti and L. Messori, Mol. BioSyst., 2012, 8(4), 985. 99. A. Pines, C. D. Kelstrup, M. G. Vrouwe, J. C. Puigvert, D. Typas, B. Misovic, A. de Groot, L. von Stechow, B. van de Water, E. H. J. Danen, H. Vrieling, L. H. F. Mullenders and J. V. Olsen, Mol. Cell. Biol., 2011, 31(24), 4964. 100. R. F. S. Lee, A. Chernobrovkin, D. Rutishauser, C. S. Allardyce, D. Hacker, K. Johnsson, R. A. Zubarev and P. J. Dyson, Sci. Rep., 2017, 7(1), 1590. 101. D. Kreutz, A. Bileck, K. Plessl, D. Wolrab, M. Groessl, B. K. Keppler, S. M. Meier and C. Gerner, Chem. -­Eur. J., 2017, 23(8), 1881.

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00271

Part III

Preclinical and Clinical Evaluation

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00271

View Online

           

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00271

View Online

Chapter 11

Metal-­based Radiotherapeutics Christian A. Masona, Lukas M. Cartera and Jason S. Lewis*a,b,c a

Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, 10065, USA; bMolecular Pharmacology Program, Memorial Sloan Kettering Cancer Center, New York, NY, 10065, USA; cWeill Cornell Medical College, New York, NY, 10065, USA *E-­mail: [email protected]

11.1 Introduction The fields of nuclear medicine and radiology have only recently shifted from their foundations in recognizing basic anatomic and metabolic signatures, toward precision imaging and therapy. These evolving paradigms include targeted delivery of molecular contrasts for probing tissue phenotype and function with molecular imaging modalities, or delivery of cytotoxic radiation for specific ablation of extirpative cell populations in cancer or infectious disease. Recognition of molecular signatures of disease ultimately requires harmonization of the biochemical aspects of the vector/target interaction while simultaneously aligning the decay aspects of the radionuclide for successful imaging or therapy. The wide variation in disease pathologies, relevant targeting strategies, vector structure, and tolerance to modification/ radiolabeling chemistry thus demands flexibility in the choice of radionuclide for each purpose.

  Metallobiology Series No. 14 Metal-­based Anticancer Agents Edited by Angela Casini, Anne Vessières and Samuel M. Meier-­Menches © The Royal Society of Chemistry 2019 Published by the Royal Society of Chemistry, www.rsc.org

273

View Online

Published on 05 April 2019 on https://pubs.rsc.org | doi:10.1039/9781788016452-00271

274

Chapter 11

Most of the progress in expansion of the arsenal of radionuclides for biomedical application has involved the radiometals. This is in part due to statistical considerations (there is simply a large number of metallic radionuclides with nuclear properties suitable for imaging/therapy in comparison to the non-­metals), but one cannot overstate the importance of developments in radiometal production routes, targeting, and concomitant advancements in chelator chemistry and bioconjugation strategies for incorporation of these into biological vectors.1 The fruit of these efforts includes a suite of radiometals with various physical half-­lives for adequately matching the pharmacokinetic profile of the targeting vector, various charge states and physicochemical properties for alteration of vector pharmacokinetics and target interactions, and various modes of radioactive decay suited to imaging, therapy, or theranostics. The principal aim of this chapter is to provide an overview of the current state-­of-­the-­art in metallo-­radiotherapeutics. Within this scope, examples of radionuclide therapeutics which have been deployed clinically will be discussed, but much of the discussion will highlight novel and promising radiotherapeutic strategies in various stages of preclinical development. A brief discourse outlining the types of therapeutic radiation and relevant radiobiology at play is included, to provide an adequate background for this discussion. Finally, as nuclear imaging strategies are nearly always pre-­ or co-­requisites of effective radionuclide therapies (e.g. for therapy selection, pre-­therapy planning, dosimetry, and assessment of therapeutic response), the role of imaging is discussed as well.

11.1.1 The Radiotherapeutic Armamentarium The primary classifications of radiation useful for targeted radionuclide therapies are “particle” radiations and include beta, alpha, and Auger electrons.† Each type is distinguished by multiple features, including nuclear/ extranuclear origin, typical particle range, emission energy profile, linear energy transfer (LET), and radiobiologic profile, with the relevance of each feature ultimately dictated by the particular radionuclide in consideration (see Table 11.1). Indeed, rational selection of a suitable therapeutic radionuclide for incorporation into a molecular vector requires extensive consideration of these factors and their interplay, in addition to the associated labeling chemistry, biochemistry/pharmacokinetics of the vector/target interaction, and its compatibility with the lifetime of the radionuclide.

†

Unless otherwise specified, for this discussion we broadly classify all mono-­energetic electron emissions as “Auger electrons”, including true Auger electrons, Coster–Kronig and Super Coster–Kronig transitions, and conversion electrons.

Published on 05 April 2019 on https://pubs.rsc.org | doi:

Table 11.1 Properties of selected therapeutically relevant radionuclides. Primary therapeutic emissions

Nuclide t1/2

Emission

Net yield   per decay Energy [%] [MeV]avg

225

77

Ac

9.92 d

α β

400 200

5.8–8.4 0.198–0.660

47–86 µm 0.43–2.5 mm

218 (11.4%), 440 DOTA (25.9%)

PET: 68Ga, 86Y, 44Sc; SPECT: 67Ga, 111In

As Bi

38.8 h 60.6 min 45.6 min 12.7 h

β− α β− β− α β+

100 100 100 200 100 17.6

0.226 6.05 0.56–0.77 0.198–0.660 5.9–8.4 0.278

0.53 mm 51 µm 2.0–3.1 mm 0.43–2.5 mm 49–86 µm 0.73 mm

N/A 727 (6.67%)

Thiol-­based

440 (25.9%)

β−

38.5

0.191

0.41 mm

DOTA, NETA, DTPA Diamsar, NOTA, DOTA, TETA

PET: 72As PET: 68Ga, 86Y, 44Sc; SPECT: 67Ga, 111In PET: 68Ga, 86Y, 44Sc; SPECT: 67Ga, 111In

212 213

Bi

64

CDSA rangea

Main gamma emissions [keV]b Recommended Theranostic imaging (intensity) chelatorsc companion(s) Production

Cu

511 (35.2%)

67

Cu

61.8 h

β−

100

0.141

0.25 mm

67

Ga

3.26 d

Auger e− CE

— 33

>0.0075 ∼0.08–0.09

2.80 d

Auger e− CE

— 14

>0.019 ∼0.14–0.22