i SYNTHESIS AND BIOLOGICAL EVALUATION OF PYRIMIDINE AND ISOQUINOLINE INHIBITORS AS POTENTIAL ANTIMALARIAL ANTIFOLATES AND TRANSMISSION-BLOCKING AGENTS by Khonzisizwe Somandi (1153883) Degree of Doctor of Philosophy: A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy Supervisor: Professor Amanda Rousseau 24 April 2023 in Johannesburg ii Declaration I declare that this thesis has been written and compiled by myself and confirm that the work presented in this thesis is my own, except for the biological testing, which was performed by colleagues at BIOTEC in Thailand and at the University of Pretoria and the CSIR. The work presented in this thesis was carried out by myself under the supervision of Professor A.L. Rousseau. This thesis is being submitted for the Doctor of Philosophy Degree in Science in the University of the Witwatersrand, Johannesburg and I declare that the work has not be submitted before for any other degree or professional qualification in any other university. ______________________ Khonzisizwe Somandi 24th day of April 2023 at Johannesburg. iii ABSTRACT Malaria continues to be a serious threat, in particular to the African region. According to the World Health Organisation, in 2021, 247 million malaria cases were reported globally of which 95 % and 96 % of malaria related deaths were in the African region. The persistence of the disease, amongst it being difficult to treat and kill, is also attributed to its resistance to currently used antimalarial agents, including class II antifolate drugs such as pyrimethamine, which are used to target the P. falciparum dihydrofolate reductase (PfDHFR) enzyme. However many other drugs have lost activity because of mutations in the active site of the enzyme. The first component of the research described herein has been to synthesise 2,4- diaminopyrimidine analogues that work by disrupting folate metabolism by inhibiting PfDHFR. In a four step synthetic approach we have successfully prepared a series of pyrimidine-2,4-diamines possessing a flexible four atom linker at the 5-position of the pyrimidine ring (in yields of 33-96 %), which could prove advantageous in avoiding clashes with mutant amino acids in the enzyme active site. Enzyme inhibition assays of the compounds have shown successful inhibition of the wild-type (WT) and quadruple mutant (QM) PfDHFR in nM ranges (Ki-WT; 1.27 – 242.72 nM and Ki-QM; 13.01 – 208.23 nM). A moderate antiplasmodial activity in vitro was observed for all compounds assessed against the drug sensitive strain IC50 (TM4/8.2) 0.42 – 28.0 µM and the drug resistant strain IC50(V1S) 3.72 – 53.7 µM. The second component of the research focuses on the synthesis of transmission-blocking analogues that target the sexual stage of the malaria parasite life cycle and work by inhibiting stage IV/V gametocytes which prevents the transmission of the parasite from the human host back to the feeding mosquito. A series of 3-substituted-isoquinolin-1-yl benzamides, derivatives of the hit compound MMV1581558, have been successfully prepared in a synthetic protocol that involves only two steps from relatively simple precursors, in yields ranging between 14 – 68 %. All analogues are undergoing biological assessment against stage IV/V gametocytes and currently we have only received the results of the asexual blood stage activity assay, with most analogues displaying only moderate activity (IC50 1.18 – 7 µM). Additional biological assays are still underway. iv DEDICATION In Loving Memory Mxolisi Somandi 1962 - 2019 This thesis is dedicated to my late father, Rev. Mxolisi Somandi. A man who believed in me when I most doubted myself. In one of our final conversations before you passed on, you urged me to take on this journey and promised to support me in any way until its end. I am where you inspired me to be now-where I desired to be; and although you may not be living amongst us anymore, your spirit lives amongst and your spiritual support has been heartfelt throughout this journey. v ACKNOWLEDGEMENTS AND THANKS First and most importantly, I would like to thank Professor Amanda Rousseau for the major role you have had in my postgraduate journey. From being supervised by you in a semester project in my Honours year to being supervised by you in my Masters and PhD research, I consider myself blessed to have been around someone who is a fountain of knowledge, compassionate and always pushing me to perform at my absolute best. Your words of encouragement and constant catch-up chats have truly made this journey much more fulfilling. To the Wits Organic group, Prof. de Koning, Prof. Bode, Prof. Jo Michael, Dr Kennedy, Dr Zimuwandeyi, Dr Ntsimango, thank you for your support and gruelling questions and suggestions especially during our group presentations, these have truly groomed me into a better scientist and communicator. To my lab mates, Sandile, Lebo, Aneesa, Nafisa, Cecilia, Alex, Matthew, Bianca and both Kabelo’s, thank you for making the lab a fun and exciting place to be and an environment where we can freely share ideas and learn from each other. A special thanks to my colleagues, Dr Kamo Butsi, who co-supervised me in my Honours and worked with me in my Masters and PhD years, although we were working on separate projects your advice and suggestions were invaluable. Matthew Maree, a special thanks to you for your constant availability and assistance with the molecular modelling part of my research, thank you for patience and tremendous help. Similarly to Edward Chavalala and Kabelo Dilebo, thank you as well for also being available to help with the molecular modelling part of my research. To Lebogang Tefu, it was a pleasure working with you in my second project, your hard working and charismatic nature made the project much more enjoyable. A special thanks to the CSIR for allowing me to use the CHPC system for molecular modelling. Dr Zimuwandeyi and Dr Kotzè, a special thanks for the NMR training and also keeping the instruments performing optimally for us to run our experiments. Dr Zimuwandeyi and Dr Selepe (University of Pretoria), my special gratitude goes to you for running the variable temperature NMR for me. Without you, this project would have not been possible. To Dr Eric Morifi, thank you for the HRMS training and ensuring that the HRMS instrument is also in peak condition. vi Professor Fernandes, thank you for running countless crystal structures for me, this truly helped in the progression of my research. To the collaborators who conducted biological assays; Dr S. Kamchonwongpaisan and the team at BIOTEC in Thailand. The University of Pretoria and the CSIR in teams led by Professor L- M. Birkholtz and Dr M. van der Watt. I thank you for your efforts in performing biological assays for my compounds and ensuring that the results reported are of a high standard. A special thanks to the University of the Witwatersrand for providing the facilities to do my research and for offering me the Postgraduate Merit Award which aided me financially and a huge thank you to the National Research Foundation (NRF) for the financial support. To my mother (Nomfundo Somandi) thank you for your constant prayers and support throughout my research and to my siblings, thank you for always encouraging and motivating me. To a very special person in my life, Lutho Zono. Thank you for always been there. You have witnessed this journey first hand and you have supported, inspired and encouraged me throughout. I truly thank you for always being there for me. Most importantly, I am thankful to God, for the constant protection, guidance and wisdom. I am where I am today because of your grace and mercy. vii LIST OF ABBREVIATIONS µM Micromolar ABS Asexual blood stage AcOH Acetic acid ACTs Artemisinin Combination Therapy AP2-G Apetala 2-G ATP Adenosine triphosphate CDI 1,1'-Carbonyldiimidazole CSIR Council for Scientific and Industrial Research CSP Circumsporozoite protein DCC N, N′-Dicyclohexylcarbodiimide DCE Dichloroethane DCM Dichloromethane DDT Dichloro-diphenyl-trichloroethane DHA Dihydroartemisinin DHF Dihydrofolate/ Dihydrofolic acid DHFR Dihydrofolate reductase DHNA Dihydroneopterin aldolase DHPS Dihydropteroate synthase DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid DNDi Drugs for Neglected Diseases Initiative DVI Dengue Vaccine initiative EtOAc Ethyl acetate EtOH Ethanol EVI European Vaccine Initiative FIND Foundation for Innovative New Diagnostics GSK GlaxoSmithKline GTP Guanosine triphosphate HRMS High resolution mass spectrometry IFD Induced fit docking viii ITN Insecticide treated nets IPM International Partnerships for Microbicides IR Infrared IRS Indoor residual spraying KOtBu Potassium tertiary butoxide LAH Lithium aluminium hydride MCR Multicomponent coupling reaction MeOH Methanol MMV Medicines for Malaria Venture MTHFR Methylene tetrahydrofolate reductase n-BuLi n-Butyllithium nM Nanomolar NMR Nuclear magnetic resonance OSM Open Source Malaria p-ABA p-Aminobenzoic acid PDPs Product Development Partnerships Pf Plasmodium falciparum PfHda2 P. falciparum histone deacetylase 2 PfHP1 P. falciparum heterochromatin protein 1 PLP Pyridoxal phosphate PPPK Pyrophosphokinase PRB Pandemic response box PYR Pyrimethamine QM Quadruple mutant RNA Ribonucleic acid SAM S-adenosylmethionine SAR Structure activity relationships SDX Sulfadoxine SHMT Serine hydroxymethyltransferase TBAB Tetrabutylammonium bromide TCAMS Tres Cantos Antimalarial THF (biology) Tetrahydrofolate ix THF (chemistry) Tetrahydrofuran TLC Thin layer chromatography WHO World Health Organisation WT Wild-type XP Extra precision x TABLE OF CONTENTS TITLE PAGE……………………………………………………………………………i DECLARATION………………………………………………………………………..ii ABSTRACT……………………………………………………………………………..iii DEDICATION…………………………………………………………………………..iv ACKNOWLEDGEMENTS AND THANKS…………………………………………..v LIST OF ABBREVIATIONS…………………………………………………………...vii PREFACE…………………………………………………………………………….……1 INTRODUCTION: A brief overview of malaria………………………………….……..1 Malaria distribution and mortality…………………………………………………….…1 Malaria parasite life cycle……………………………………………………………….…3 Preventative measures for malaria……………………………………………………..….4 i) Vector control…………………………………………………………………….5 ii) Vaccines…………………………………………………………………………..6 iii) Chemotherapy for malaria infections……………………………………………..7 iv) Product development partnerships (PDPs) for malaria…………………...……....9 Aims of the project……………………………………………………………………….….11 PART 1-THE REPLICATION STAGE…………………………………………….……..12 CHAPTER 1: Folates and folate metabolism………………………………………….…..12 1.1. Folate metabolism………………...…………………………………………………...…12 1.2. Antifolates and P. falciparum resistance………………………..…………………….17 1.2.1. Class I antifolates and P. falciparum resistance……………………………………...17 1.2.2. Class II antifolates and P. falciparum resistance……………………………………..18 1.3. Synthesis of 6-substituted-2,4-diaminopyrimidine-5-phenethylamines as inhibitors of PfDHFR………………………………………………………….……..24 xi CHAPTER 2: RESULTS AND DISCUSSION……………………………………………27 2.1. In silico molecular modelling – induced fit docking of proposed analogues (28)…………………………………………………………………………………………...29 2.2. In silico molecular modelling – qikprop pharmacokinetic property predictions of proposed analogues (28)………………………….………………………………………...34 2.3. Approaches to the synthesis of 2,4-diaminopyrimidine core………………………..39 2.4. Approaches to the synthesis of 2,4-diaminopyrimidine-5-carbonitriles……………41 2.4.1. Synthesis of 2,4-diamino-6-phenylpyrimidine-5-carbonitrile 30a……………..……...45 2.4.2. Synthesis of 2,4-diamino-6-cyclohexylpyrimidine-5-carbonitrile 30b…………..…....47 2.4.3. Synthesis of 2,4-diamino-6-cyclopropylpyrimidine-5-carbonitrile 30c………....….....48 2.5. Synthesis of 2,4-diaminopyrimidine-5-carbaldehydes (31)…………………………..49 2.5.1. Synthesis of 2,4-diaminopyrimidine-6-phenylpyrimidine-5-carbaldehyde 31a……….50 2.5.2. Synthesis of 2,4-diaminopyrimidine-6-cyclohexylpyrimidine-5-carbaldehyde 31b.….52 2.5.3. Synthesis of 2,4-diaminopyrimidine-6-cyclopropylpyrimidine-5-carbaldehyde 31c….53 2.6. Attempted synthesis of 6-substituted-2,4-diaminopyrimidine-5-phenethylamines (28) using reductive amination methodology………………………………..………...…..54 2.7. Synthesis of 6-substituted-2,4-diaminopyrimidine-5-phenethylimines (51)………..56 2.7.1 Synthesis of ((E)-6-phenyl-5-{[(2-methoxyphenethyl)imino]methyl}pyrimidine-2,4 diamine 51a and analogues………………………….……………………………………..…57 2.8. Synthesis of 6-substituted-2,4-diaminopyrimidine-5-phenethylamines (28)………..61 2.8.1. Synthesis of 6-phenyl-5-{[(2-methoxyphenethyl)amino]methyl}pyrimidine-2,4-diamine 28a and analogues…………………………………………………………………………….61 2.9. Biological assessment of pyrimidine analogues (28)………………………………….67 2.9.1. Enzyme inhibition and whole cell P. falciparum antiplasmodial studies……………...68 xii CHAPTER 3: SUMMARY OF RESULTS, FUTURE DIRECTIONS AND PRELIMINARY MODELLING STUDIES ………………………………………………75 3.1. Summary of results………...…………………………………………………………...75 3.2. Future directions………………………………………………………………………..77 3.2.1. In silico molecular modelling – ligand docking studies of modified analogues 54 – 56……………………………………………………………………….....………….....78 3.2.2. In silico molecular modelling – qikprop pharmacokinetics property predictions of modified analogues 54 – 56………………………………………………………….…….....82 3.2.3. Proposed synthetic route to modified analogues 54 – 56……………………………….85 CHAPTER 4: EXPERIMENTAL-PART 1………………………………………………..87 4.1. Molecular modelling…………………………………………………………………....87 4.2. General Procedures………………………………………………………………….…87 4.3. General procedure for synthesis of 6-substituted-2,4-diamino-5-carbonitrile- pyrimidines and analogues (30)………………………………………………………….....88 4.3.1. Synthesis of 2,4-diamino-6-phenylpyrimidine-5-carbonitrile 30a…………………….89 4.3.2. Synthesis of 2,4-diamino-6-cyclohexylpyrimidine-5- carbonitrile 30b………………………………………………………………………………89 4.3.3. Synthesis of 2,4-diamino-6-cyclopropylpyrimidine-5-carbonitrile 30c…………..…...90 4.4. Preparation of Raney Nickel……………………………………………………….….90 4.5. General procedure for synthesis of 6-substituted-2,4-diaminopyrimidine-5- carbaldehydes (31)…………………………………………………………………………..91 4.5.1. Synthesis of 2,4-diaminopyrimidine-6-phenylpyrimidine-5-carbaldehyde31a……….91 4.5.2. Synthesis of 2,4-diaminopyrimidine-6-cyclohexylpyrimidine-5-carbaldehyde 31b…..92 4.5.3. Synthesis of 2,4-diaminopyrimidine-6-cyclopropylpyrimidine-5-carbaldehyde 31c….92 4.6. General procedure for the synthesis of 2,4-diaminopyrimidine-5-phenethylimines (51)………………………………………………………………………………………...…93 xiii 4.6.1. Synthesis of (E)-6-phenyl-5-{[(2-methoxyphenethyl)imino]methyl}pyrimidine-2,4 diamine 51a………………………………………………………………………………..…93 4.6.2. Synthesis of (E)-6-phenyl-5-{[(3-methoxyphenethyl)imino]methyl}pyrimidine-2,4 diamine 51b………………………………………………………………………………..…94 4.6.3. Synthesis of (E)-6-phenyl-5-{[(4-methoxyphenethyl)imino]methyl}pyrimidine-2,4 diamine 51c…………………………………………………………………………………..95 4.6.4. Synthesis of (E)-6-phenyl-5-{[(2-fluorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51d…………………………………………………………………………………..95 4.6.5. Synthesis of (E)-6-phenyl-5-{[(3-fluorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51e…………………………………………………………………………………..96 4.6.6. Synthesis of (E)-6-phenyl-5-{[(4-fluorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51f…………………………………………………………………………………...97 4.6.7. Synthesis of (E)-6-phenyl-5-{[(2-chlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51g………………………………………………………………………………......97 4.6.8. Synthesis of (E)-6-phenyl-5-{[(3-chlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51h…………………………………………………………………………………..98 4.6.9. Synthesis of (E)-6-phenyl-5-{[(4-chlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51i…………………………………………………………………………………...99 4.6.10. Synthesis of (E)-6-phenyl-5-{[(2,4-dichlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51j…………………………………………………………………………………...99 4.6.11. Synthesis of (E)-6-phenyl-5-{[(3,4-dichlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51k…………………………………………………………………………………100 4.6.12. Synthesis of (E)-6-cyclohexyl-5-{[(2-methoxyphenethyl)imino]methyl}pyrimidine- 2,4 diamine 51l……………………………………………………………………………...101 4.6.13. Synthesis of (E)-6-cyclohexyl-5-{[(3-methoxyphenethyl)imino]methyl}pyrimidine- 2,4 diamine 51m……………………………………………………………………………101 4.6.14. Synthesis of (E)-6-cyclohexyl-5-{[(4-methoxyphenethyl)imino]methyl}pyrimidine- 2,4 diamine 51n……………………………………………………………………………102 xiv 4.6.15. Synthesis of (E)-6-cyclohexyl-5-{[(2-fluorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51o…………………………………………………………………………………103 4.6.16. Synthesis of (E)-6-cyclohexyl-5-{[(3-fluorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51p…………………………………………………………………………………103 4.6.17. Synthesis of (E)-6-cyclohexyl-5-{[(4-fluorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51q………………………………………………………………………………....104 4.6.18. Synthesis of (E)-6-cyclohexyl-5-{[(2-chlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51r……………………………………………………………………………...….105 4.6.19. Synthesis of (E)-6-cyclohexyl-5-{[(3-chlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51s………………………………………………………………………………….105 4.6.20. Synthesis of (E)-6-cyclohexyl-5-{[(4-chlorophenethyl)imino]methyl}pyrimidine-2,4 diamine 51t……………………………………………………………………………….....106 4.6.21. Synthesis of (E)-6-cyclohexyl-5-{[(2,4-dichlorophenethyl)imino]-methyl}- pyrimidine-2,4 diamine 51u……………………………….………………………………..107 4.6.22. Synthesis of (E)-6-cyclohexyl-5-{[(3,4-dichlorophenethyl)imino]-methyl}- pyrimidine-2,4 diamine 51v………………………………………………………………....107 4.6.23. Synthesis of (E)-6-cyclopropyl-5-{[(2-methoxyphenethyl)-imino]methyl}- pyrimidine-2,4-diamine 51w……………………………………………………………..…108 4.6.24. Synthesis of (E)-6-cyclopropyl-5-{[(3-methoxyphenethyl)imino]-methyl}- pyrimidine-2,4-diamine 51x………………………………………………………………...109 4.6.25. Synthesis of (E)-6-cyclopropyl-5-{[(4-methoxyphenethyl)imino]-methyl}- pyrimidine-2,4-diamine 51y………………………………………………………………...109 4.6.26. Synthesis of (E)-6-cyclopropyl-5-{[(2-fluorophenethyl)imino]methyl}pyrimidine-2,4- diamine 51z………………………………………………………………………………….110 4.6.27. Synthesis of (E)-6-cyclopropyl-5-{[(3-fluorophenethyl)imino]methyl}pyrimidine-2,4- diamine 51a’……………………………………………………………………………...…110 4.6.28. Synthesis of (E)-6-cyclopropyl-5-{[(4-fluorophenethyl)imino]methyl}pyrimidine-2,4- diamine 51b’………………………………………………………………………………...111 xv 4.6.29. Synthesis of (E)-6-cyclopropyl-5-{[(2-chlorophenethyl)imino]methyl}pyrimidine-2,4- diamine 51c’………………………………………………………………………………...112 4.6.30. Synthesis of (E)-6-cyclopropyl-5-{[(3-chlorophenethyl)imino]methyl}pyrimidine-2,4- diamine 51d’……………………………………………………………………………...…112 4.6.31. Synthesis of (E)-6-cyclopropyl-5-{[(4-chlorophenethyl)imino]methyl}pyrimidine-2,4- diamine 51e’…………………………………………………………………………….…..113 4.6.32. Synthesis of (E)-6-cyclopropyl-5-{[(2,4-dichlorophenethyl)imino]-methyl}- pyrimidine-2,4-diamine 51f’………………………………………………………………..113 4.6.33. Synthesis of (E)-6-cyclopropyl-5-{[(3,4-dichlorophenethyl)imino]-methyl}- pyrimidine-2,4-diamine 51g’………………………………………………………………..114 4.7. General procedure for synthesis of 2,4-diaminopyrimidine-5-phenethylamine (28)……………………………………………………………………………………….....115 4.7.1. Synthesis of 6-phenyl-5-{[(2-methoxyphenethyl)amino]methyl}pyrimidine-2,4-diamine 28a…………………………………………………………………………………………..115 4.7.2. Synthesis of 6-phenyl-5-{[(3-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28b…………………………………………………………………………………116 4.7.3. Synthesis of 6-phenyl-5-{[(4-methoxyphenethyl)amino]methyl}pyrimidine-2,4-diamine 28c………………………………………………………………………………………..…117 4.7.4. Synthesis of 6-phenyl-5-{[(3-fluorophenethyl)amino]methyl}pyrimidine-2,4-diamine 28e…………………………………………………………………………………………..117 4.7.5. Synthesis of 6-phenyl-5-{[(3-chlorophenethyl)amino]methyl}pyrimidine-2,4-diamine 28h…………………………………………………………………………………………..118 4.7.6. Synthesis of 6-cyclohexyl-5-{[(2-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28l……………………………………………………………………………….....119 4.7.7. Synthesis of 6-cyclohexyl-5-{[(3-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28m…………………………………………………………………………..…….119 4.7.8. Synthesis of 6-cyclohexyl-5-{[(4-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28n………………………………………………………………………………....120 xvi 4.7.9. Synthesis of 6-cyclohexyl-5-{[(2-fluorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28o…………………………………………………………………………………121 4.7.10. Synthesis of 6-cyclohexyl-5-{[(3-fluorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28p…………………………………………………………………………………121 4.7.11. Synthesis of 6-cyclohexyl-5-{[(4-fluorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28q………………………………………………………………………………....112 4.7.12. Synthesis of 6-cyclohexyl-5-{[(2-chlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28r………………………………………………………………………………...123 4.7.13. Synthesis of 6-cyclohexyl-5-{[(3-chlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28s………………………………………………………………………………....124 4.7.14. Synthesis of 6-cyclohexyl-5-{[(4-chlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28t………………………………………………………………………………….124 4.7.15. Synthesis of 6-cyclohexyl-5-{[(2,4-dichlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28u………………………………………………………………………………....125 4.7.16. Synthesis of 6-cyclohexyl-5-{[(3,4-dichlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28v…………………………………………………………………………………125 4.7.17. Synthesis of 6-cyclopropyl-5-{[(2-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28w………………………………………………………………………………...126 4.7.18. Synthesis of 6-cyclopropyl-5-{[(3-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28x…………………………………………………………………………………127 4.7.19. Synthesis of 6-cyclopropyl-5-{[(4-methoxyphenethyl)amino]methyl}pyrimidine-2,4- diamine 28y………………………………………………………………………………....127 4.7.20. Synthesis of 6-cyclopropyl-5-{[(2-fluorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28z………………………………………………………………………………….128 4.7.21. Synthesis of 6-cyclopropyl-5-{[(3-fluorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28a’………………………………………………………………………………...128 4.7.22. Synthesis of 6-cyclopropyl-5-{[(4-fluorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28b’………………………………………………………………………………..129 4.7.23. Synthesis of 6-cyclopropyl-5-{[(2-chlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28c’………………………………………………………………………………...130 xvii 4.7.24. Synthesis of 6-cyclopropyl-5-{[(3-chlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28d’………………………………………………………………………………...130 4.7.25. Synthesis of 6-cyclopropyl-5-{[(4-chlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28e’………………………………………………………………………………...131 4.7.26. Synthesis of 6-cyclopropyl-5-{[(2,4-dichlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28f’………………………………………………………………………………...131 4.7.27. Synthesis of 6-cyclopropyl-5-{[(3,4-dichlorophenethyl)amino]methyl}pyrimidine-2,4- diamine 28g’………………………………………………………………………………..132 4.8. Crystal structure and refinement……………………………………………………132 4.9. Enzyme preparation and inhibition assays………………………………………….133 4.10. Parasite culture, antiplasmodial and cytotoxicity testing in vitro........................................................................................................................................133 Appendix………………………………………………………………………………...…134 PART 2-THE TRANSMISSION STAGE………………………………………………..143 CHAPTER 5: Gametocytes and transmission-blocking agents………………………...146 5.1. Gametocyte development……………………………………………………………..146 5.1.1. Gametocyte commitment…………………………………………………....147 5.1.2. Gametocyte maturation…………………………………………..…….……148 5.1.3. Gametocyte metabolism………………………………………………...…..151 5.2. Known transmission-blocking drugs…………………………………………..…….152 5.3. Potential gametocytocidal drugs under development………………………..…..…154 5.4. Approach to the synthesis of transmission-blocking hit compound MMV1581558 and derivatives……………………………………………………………………….….….159 5.4.1. Synthesis of substituted isoquinolines as potential transmission-blocking compounds……………………………………………………………...…..160 CHAPTER 6: RESULTS AND DISCUSSION………………………………….….…....162 6.1. Properties of MMV1581558 and preliminary SAR data……………………….…..162 6.2. Approaches to the synthesis of the isoquinoline core……………………………..…164 xviii 6.3. Synthesis of 3-substituted-1-amino-isoquinolines (80)……………………………....167 6.3.1. Synthesis of 3-(pyridin-2-yl)isoquinolin-1-amine and halogenated analogues…….....168 6.3.2. Synthesis of 3-(pyridin-3-yl)isoquinolin-1-amine 80e…..…………………………....171 6.3.3. Synthesis of 3-(pyridin-4-yl)isoquinolin-1-amine 80f……………………………….172 6.4. Synthesis of 2,6-dimethyl-N-3-substituted-isoquinolin-1-yl-benzamides………….174 6.4.1. Synthesis of 2,6-dimethyl-N-(3-(pyridin-2-yl)isoquinolin-1-yl)benzamide 77a…….176 6.4.2. Synthesis of 3-methyl-N-(3-(pyridin-2-yl)isoquinolin-1-yl)picolinamide 77g………182 6.4.3. Attempted synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)picolinamide 77h……..183 6.4.4. Attempted synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)picolinamide 77h using DCC protocol……………………………………………………………………………….185 6.4.5. Synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)picolinamide 77h……..………….188 6.4.6. Synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)nicotinamide 77i………………….190 6.4.7. Synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)isonicotinamide 77j……………….191 6.5. Synthesis of 2-(3-(pyridin-2-yl)isoquinolin-1-yl)isoindoline-1,3-dione (122)……....194 6.6. Biological Assay Data…………………………………………………………………195 CHAPTER 7: CONCLUSIONS AND FUTURE WORK………………………………..200 7.1. Conclusions………………………………………………………………………….....200 7.2. Future Work…………………………………………………………………………...201 CHAPTER 8: EXPERIMENTAL-PART 2……………………………………………....203 8.1. General procedure for synthesis of 3-substituted-1-amino-isoquinoline and analogues (80)……………………………………………………………………………………….…203 8.1.1. Synthesis of 3-(pyridin-2-yl)isoquinolin-1-amine 80a……………………………….203 8.1.2. Synthesis of 7-fluoro-3-(pyridin-2-yl)isoquinolin-1-amine 80b……………..………204 8.1.3. Synthesis of 6-chloro-3-(pyridin-2-yl)isoquinolin-1-amine 80c……………………..204 8.1.4. Synthesis of 8-chloro-3-(pyridin-2-yl)isoquinolin-1-amine 80d……………………..205 xix 8.1.5. Synthesis of 3-(pyridin-3-yl)isoquinolin-1-amine 80e………………………………..205 8.1.6. Synthesis of 3-(pyridin-4-yl)isoquinolin-1-amine 80f………………………………..206 8.2. General procedure for synthesis of substituted-isoquinolin-1-yl benzamides (77)………………………………………………………………………………………….206 8.2.1. Synthesis of 2,6-dimethyl-N-(3-(pyridin-2-yl)isoquinolin-1-yl)benzamide 77a……..207 8.2.2. Synthesis of N-(7-fluoro-3-(pyridin-2-yl)isoquinolin-1-yl)-2,6-dimethylbenzamide 77b…………………………………………………………………………………………..208 8.2.3. Synthesis of N-(6-chloro-3-(pyridin-2-yl)isoquinolin-1-yl)-2,6-dimethylbenzamide 77c………………………………………………………………..………………………....209 8.2.4. Synthesis of N-(8-chloro-3-(pyridin-2-yl)isoquinolin-1-yl)-2,6-dimethylbenzamide 77d………………………………………………………………………………………….210 8.2.5. Synthesis of 2,6-dimethyl-N-(3-(pyridin-3-yl)isoquinolin-1-yl)benzamide 77e…….210 8.2.6. Synthesis of 2,6-dimethyl-N-(3-(pyridin-4-yl)isoquinolin-1-yl)benzamide 77f……..211 8.2.7. Synthesis of 3-methyl-N-(3-(pyridin-2-yl)isoquinolin-1-yl)picolinamide 77g………212 8.2.8. Synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)picolinamide 77h…………………214 8.2.9. Synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)nicotinamide 77i…………………..214 8.2.10. Synthesis of N-(3-(pyridin-2-yl)isoquinolin-1-yl)isonicotinamide 77j……………...215 8.3. Synthesis of 2-(3-(pyridin-2-yl)isoquinolin-1-yl)isoindoline-1,3-dione (122).……...216 8.4. Crystal structure and refinement…………………………………………………….217 Appendix………………………………………………..……………………………….…218 References………………………………………………………………………………….230 1 PREFACE INTRODUCTION-A brief overview of malaria Malaria distribution and mortality Malaria continues to be a growing health burden in Africa. The Plasmodium parasite causes malaria, and transmission to humans occurs by inoculation from a female Anopheles mosquito that is infected.1 There are several known parasite species that cause malaria in humans, including, P. vivax, P. malariae, P. ovale, P. knowlesi and P. falciparum.2 P. falciparum is prevalent in Africa, causes the most fatalities,1 and is the focus of this research project. The malaria disease distribution is very climate and socio-economic specific. The disease is common in developing regions where the humidity and temperatures are high. Some of these regions include Africa, South/South East Asia, Central/South America, Dominican Republic and Oceania to name a few.3 The African region, however, continues to have disproportionally high malaria infections.1 According to the World Health Organisation, in 2021 there were 247 million malaria cases recorded globally, with 95% of those cases being caused by P. falciparum infections in the African region. Over 50% of all malaria deaths globally occur in four African countries, namely, Nigeria, the Democratic Republic of the Congo, United Republic of Tanzania, and Mozambique.1 Other African countries that are also a concern in malaria related cases and deaths include Angola, Uganda and Burkina Faso.3 Additionally, 80% of global malaria deaths were of children under the age of 5 years.1 These alarming statistics clearly indicate that this disease is largely an African problem and is lethal amongst very young children, despite the fact that malaria is treatable. There is therefore a need to find more effective ways of treating malaria. Over the past 20 years, scientists have been dedicated to reducing malaria deaths in affected African populations. A steady decline in child (below 5 years) and population mortality has been observed since 2000, Figure 1 and Figure 2 are illustrations of this.3,4 Notably, the global annual deaths have decreased from about 900 000 to 630 000 and of this, the African region has decreased its annual death toll from about 840 000 to 602 000, Figure 1.4 Child mortality has also declined, as illustrated in Figure 2, where some regions have eliminated child mortality due to malaria infections. 2 Figure 1: Global annual malaria deaths from 2000 to 2020 in regions with high malaria infections.4 Figure 2: Mortality rate of children under the age of 5 years from 2000 to 2019 in the African region.4 3 This is promising progress for the eradication of malaria as this highlights how preventative measures for malaria, which is discussed below, are key components to the eradication of the disease. However, it is also observed that there has been a stagnant period from about the year 2015, with an increase in malaria related deaths in 2020 which could be attributed to disruptions caused by the Covid-19 pandemic,3 Figure 1. This is indicative that more still needs to be done to alleviate the malaria burden and eventually eradicate the disease. Malaria parasite life cycle An in-depth knowledge of the malaria parasite life cycle remains a key component in the potential success of drug design and the development of antimalarial agents, as understanding the life cycle provides much needed information regarding potential novel targets within the parasite which can be exploited to synthesise a desirable and potentially effective drug. An overview of the life cycle of the malaria parasite is given in Figure 3. Figure 3: Illustration of the malaria parasite life cycle.5 4 The malaria parasite utilises two hosts to complete its life cycle: the human and the mosquito. When taking a blood meal, an infected female Anopheles mosquito injects sporozoites into the human host (Figure 3, annotation 1). The liver cells are infected (Figure 3, annotation 2) and mature into schizonts (Figure 3, annotation 3) that rupture and release merozoites (Figure 3, annotation 4), which infect red blood cells (Figure 3, annotation 5). Note that the species P. ovale and P. vivax have a dormant stage [hypnozoites] that can stay in the liver for weeks or years and causes relapse by delayed bloodstream invasion,5 however this is absent from our species of interest, P. falciparum. The initial replication of the parasite in the liver (exo- erythrocytic schizogony (Figure 3, annotation A)) is followed by an asexual multiplication in the erythrocytes (erythrocytic schizogony (Figure 3, annotation B)). Immature trophozoites develop and form schizonts that rupture and release more merozoites (Figure 3, annotation 6). Some evolve and form gametocytes (Figure 3, annotation 7), which are the sexual erythrocytic stages of the parasite.5 During a blood meal on a human host, an Anopheles mosquito then ingests these female (macrogametocytes) and male (microgametocytes) gametocytes (Figure 3, annotation 8). Male gametocytes penetrate the female gametocytes forming zygotes in the mosquito (Figure 3, annotation 9) which then become ookinetes that are elongated and motile (Figure 3, annotation 10) and invade the mosquito’s midgut wall, developing into oocysts (Figure 3, annotation 11).6 The oocysts develop, rupture, and release sporozoites (Figure 3, annotation 12) that migrate to the salivary glands of the mosquito. Injecting these sporozoites (Figure 3, annotation 1) into a new human host continues the malaria parasite life cycle. The multiplication of the parasite which occurs within the mosquito is called the ‘Sporogonic cycle’ (Figure 3, annotation C).5 As sexual reproduction occurs in the mosquito, the mosquito should be called the “host” and humans the vector. However, throughout literature, the mosquito is referred to as the vector, and humans, the host, and this is the approach adopted in this thesis. Preventative measures for malaria Malaria prevention strategies are key components in the fight against the disease. These strategies can either prevent malaria infections entirely, i.e., (i) vector control (in this case of the Anopheles mosquito), or these strategies can help cure malaria and reduce mortality and transmission, i.e. (ii) vaccines and (iii) chemotherapy. 5 i) Vector control The two main vector control interventions are indoor residual spraying (IRS) of insecticides and the use of insecticide treated nets (ITN).1 These strategies can be used together to achieve the best malarial control results.6 Both these strategies employ the use of insecticides to kill the mosquito that bites or rests indoors.7 One of the earliest used synthetic insecticides was DDT 1 (dichloro-diphenyl-trichloroethane), developed in the 1940’s, Figure 4.8 This insecticide achieved great success in the prevention and control of malaria, however it had to be discontinued in most countries because of its negative impact on wildlife, its adverse environmental effects, its persistence in the environment and because of its risk as a probable human carcinogen.8 In some developing countries, however, DDT is still used for essential public health purposes, e.g. its use in African countries for malaria vector control. Currently, pyrethroids, of the general structure 2, are the most commonly used insecticides, Figure 4.7 These are synthetic derivatives of naturally occurring pyrethrin insecticides found in the extract from the flowers Chrysanthemum cinerariaefolium.9 They work by binding to and disrupting the voltage-gated sodium channels of insect nerves.9 Unfortunately, several species of Anopheles mosquito have been reported to have developed resistance against this class of insecticides.3 The WHO also reports that 78 countries have found insecticide resistant mosquitoes to at least 1 of the 4 commonly-used insecticide classes in the period 2010–2019, these include, pyrethroids 2, carbamates 3, organophosphates 4, and organochlorines, e.g. DDT 1, Figure 4. Additionally, 29 countries have reported mosquito resistance to all main types of insecticides.1,10 Figure 4: Structure of DDT (1) and general structure of commonly used insecticides pyrethroids (2), carbamates (3) and organophosphates (4). 6 ii) Vaccines One of the most noticeable interventions to help control malaria has been an effective malaria vaccine. Very few vaccine candidates have demonstrated sufficient efficacy. The most notable vaccine to date is the RTS,S/AS01 vaccine produced by GlaxoSmithKline (GSK).11 This is the only vaccine that has gone to phase III trials and has reproducible efficacy in different populations. RTS,S/AS01 is a recombinant protein vaccine that targets the circumsporozoite protein (CSP) of P. falciparum, expressed at the pre-erythrocytic stage of infection.3 Although considered efficacious, the vaccine displayed a low efficacy in phase III trials with 36 % efficacy against clinical malaria.11 This was observed in children 5 and 17 months old within 4 years after being vaccinated and receiving a booster 21 months later.11 It was also reported that the RTS,S/AS01 efficacy peaked at ~60–70% in the initial 6 months of receiving a vaccination, but rapidly declined in efficacy, and had limited or non-significant efficacy by 18 months.3 The WHO and partners aim to achieve a malaria vaccine with efficacy >75% by the year 2030.3 Despite the lower than anticipated efficacy, the vaccine significantly reduces the severity of malaria in young children. As a result of this, in October 2021 the WHO recommended that the RTS,S/AS01 malaria vaccine be used more broadly among children living in regions with moderate to high P. falciparum malaria transmission.1 Another promising vaccine candidate is the novel R21/Matrix-M pre-erythrocytic malaria vaccine candidate, originally designed and developed at the University of Oxford.12,13 Similar to RTS,S/AS01, R21 is also a recombinant protein vaccine.12 Both vaccines possess the protein HBsAg, which is fused to the C-terminus and central repeats of the CSP, which self assemble into virus-like particles in yeast. A key difference however, is that R21 does not have excess HBsAg, whereas RTS,S does.12 Additionally, in R21 only fusion protein moieties are present, whereas RTS,S comprises 20% with 80% being HBsAg monomers expressed individually, this therefore likely diminishes CSP coverage of the virus-like particle surface.12 The R21/Matrix-M vaccine displayed high immunogenicity in preclinical studies and therefore was selected for clinical development. Phase IIb trials have been conducted by Datoo et al. for the R21/Matrix-M vaccine for children aged 5–17 months in Nanoro, Burkina Faso, which is a malaria hotspot.12 The vaccine was administered and the efficacy monitored after 6 months and 12 months respectively, after which, a booster was administered and similarly the efficacy analysed over a 6 month and 12 month period. Essentially, two groups (group 1 and group 2) 7 were treated with different compositions of adjuvant Matrix-M of the vaccine, i.e. for group 1, 5 µg R21/25 µg Matrix-M and for group 2, 5 µg R21/50 µg Matrix-M. The third group (group 3) was a control and received Rabivax-S. Each group had 150 participants. The findings reported that for the primary vaccine administration, group 1 showed an efficacy of 74 % in a period of 6 months and 71 % after 12 months. Group 2 showed 76 % (6 months) and 76 % (12 months).12 After a booster shot was administered, results showed that for group 1 the efficacy was unchanged at 74 % in a period of 6 months and 71 % after 12 months. For group 2 however, an improvement of 78 % (6 months) and 80 % (12 months). Furthermore, a Cox regression model that compares group 1 and group 3 resulted in vaccine efficacy of 74% and a comparison of group 2 with group 3 resulted in 77% efficacy.12 Only mild adverse effects were observed with the most common being a fever. These results therefore render the R21/Matrix-M vaccine a promising candidate. In April 2023, Oxford University reported that the R21/Matrix-M vaccine has ongoing phase III clinical trials in Burkina Faso, Kenya, Mali and Tanzania with 4800 children enrolled.13 Reports also state that the recent phase III trial data is also displaying high-levels of efficacy by the vaccine and a good safety profile.13 These phase III trial results are expected to be reported later in the year.13 iii) Chemotherapy for malaria infections Antimalarial agents are an essential tool to combat the effects of the growing malaria disease burden. The current antimalarial drugs used are split into three categories. These include aryl aminoalcohol compounds (such as quinine 5 and chloroquine 6), antifolate compounds (such as pyrimethamine 7 and proguanil 8) and artemisinin compounds (such as dihydroartemisinin (DHA) 9 and artesunate 10), Figure 5.2 Although most of these drugs are effective, in some regions the resistance towards each of these classes of drugs has been observed.2 Artemisinin resistance is well characterised in Southeast Asia 14,15 and more recently in some regions outside Southeast Asia, including the Greater Mekong Subregion which extends from Cambodia to parts of Vietnam, Laos, Thailand, Myanmar, and China.14 Chloroquine-resistant P. falciparum initially occurred in the 1950s in three to four areas in Southeast Asia, South America and Oceania. To date, the resistance to chloroquine has spread to almost every area worldwide where P. falciparum malaria is present.15 Resistance to antifolates was initially observed in South America and Southeast Asia in the 1970s, shortly after their introduction. 8 Antifolate resistance in Africa was noted in the early 1980s in East Africa and in the late 1980s for West Africa. 16 Figure 5: Antimalarial agents quinine (5), chloroquine (6), pyrimethamine (7), proguanil (8), dihydroartemisinin (9) artesunate (10). The use of transmission-blocking agents as a more novel approach to eliminating the malaria disease by blocking the transmission from the infected human host to the feeding Anopheles mosquito is also being investigated.17 There have been no clinically approved transmission- blocking drugs yet, however, some drug candidates are undergoing preclinical (e.g. MMV183 11); human volunteer (SJ733 12); and human exploratory (e.g. MMV390048 13) phases (Figure 6).17 Figure 6: Some transmission-blocking candidates under development MMV183 (11), SJ733 (12), MMV390048 (13). 9 iv) Product Development Partnerships (PDPs) for malaria PDP’s are international, not-for-profit organisations that focus on the development of vaccines, drugs and diagnostics for neglected and poverty-related diseases, making them available at affordable costs.18 As research and development for new medicinal therapies is time consuming, expensive and risky, typically diseases affecting low and middle income countries are neglected as developing products for these diseases does not a have financial incentive for the pharmaceutical sector. The consequences of this are the gaps and restrictions in the global health innovation pipeline.18 Fortunately, the support of PDPs from bodies such as the German Federal Ministry of Education and Research have helped in bridging those gaps and restrictions in the global health innovation pipeline.18 PDPs work similarly to the pharmaceutical sector in the way in which they develop drugs and vaccines. However, differences in PDPs from private companies is in their funding sources and partners involved, that is, funding will typically come from either governments or philanthropic donors.18,19 Furthermore, PDPs promote and coordinate the collaborative research between partners from the public, private, academic, and philanthropic sector. Some PDPs include Medicines for Malaria Venture (MMV), PATH Malaria, Dengue Vaccine initiative (DVI), Drugs for Neglected Diseases Initiative (DNDi), European Vaccine Initiative (EVI), Foundation for Innovative New Diagnostics (FIND) and International Partnerships for Microbicides (IPM). The MMV and PATH organisations have played a major role in finding new antimalarial treatments. The MMV, founded in 1999, is based in Geneva, Switzerland.19 MMV pave the way for innovation by allowing open collaborations with standard research and development processes which include compound screening and hit-to-lead identification, lead optimization, preclinical development and candidate selection, clinical phase 1, 2, and 3 trials, and drug registration and launch. MMV has collaborated with pharmaceutical companies such as Novartis, in the development and launch of a combination treatment for malaria, Coartem® Dispersible at an affordable cost.19 This treatment is typically used for adults, however it is highly effective and well accepted by children as well.19 PATH was founded in the mid-1970s and has become one of the world's largest non-profit technology organizations that focuses on health concerns in developing countries.19 PATH has 10 performed research on vaccine work and in the malaria field, the PATH Malaria Vaccine Initiative has collaborated with GlaxoSmithKline (GSK) Biologicals in developing a malaria vaccine which was in Phase 3 trials in African countries in 2011.19 There have also been open source efforts for malaria research. Open Source Drug Discovery is governed primarily by six laws; i) all data are open and all ideas are shared, ii) anyone can take part at any level, iii) there will be no patents, iv) suggestions are the best form of criticism, v) public discussion is much more valuable than private email and vi) an open project is bigger than and is not owned by any lab.20 Pharmaceutical research organisiations such as GSK Tres Cantos and Novartis St. Jude’s Children’s Research Hospital have contributed to the open source movement and have released large data sets of antimalarial compounds derived from phenotypic high throughput screening. The GSK Tres Cantos Antimalarial (TCAMS) data set contained numerous hit compounds and provided many starting points for various research groups, which included independent groups, academic groups in collaboration with GSK or internal groups at GSK. Highly potent antimalarial compounds have been discovered by this Open Source Malaria (OSM) initiative, which include the potential antimalarials shown in Figure 7 as starting points since they possess desirable pharmacokinetic properties and high ligand efficiency.20 Figure 7: Potent hit compounds from open source malaria research efforts.20 11 Open source has also greatly inspired a part of this research project as the hit compound MMV1581558, was identified from the MMV Pandemic Response Box and screened by researchers at the University of Pretoria. These efforts found this compound to be active against stage IV/V gametocytes and as part of a collaborative effort, we will be developing the MMV1581558 compound. This will be explored more in detail in the chapters that follow. This open source effort has the potential to increase the efficiency of drug discovery and will allow for more cost effective treatments in diseases associated with low-income regions. Aims of the project This project will look at two aspects of the malaria parasite life cycle for intervention strategies: i) the liver and blood replication stage and ii) the gametocyte transmission stage from human to mosquito. In Part 1, the focus will be on the replication stage and disruption of folate metabolism in the parasite using antifolates.21 Antifolate drugs are antimetabolites with antagonistic properties, which largely target the asexual stages of the malaria parasite.22 Antimalarial antifolates work by inhibiting the folate metabolic enzymes dihydrofolate reductase (DHFR) and/or dihydropteroate synthase (DHPS).21 In this project, we will focus on inhibition of DHFR, which will stop the parasite’s cellular replication. Pyrimidine-based antifolates with this potential will be synthesised. In Part 2, the focus will be on the human to host transmission stage, by inhibition of the stage IV/V gametocytes. To this end, isoquinoline derivatives with potential transmission-blocking ability based on a hit molecule, MMV1581558 identified from the MMV pandemic response box, will be synthesised.22 These compounds target the sexual stage gametocytes and will potentially display transmission- blocking activity, by death or inactivation of the gametocytes.22 12 PART 1-THE REPLICATION STAGE CHAPTER 1: Folates and folate metabolism In the malaria parasite life cycle in humans, replication occurs at the liver stage and at the blood stage. This replication is essentially dependant on folates and folate metabolism. Therefore, understanding the mechanism of action of antifolates is best achieved when understanding the important roles folates play in mammalian cells. Folates are responsible for the synthesis of nucleotides and certain amino acids. They achieve this by their important role in acceptance, redox processing, and transfer of one-carbon units. 23,24 The inability of humans to synthesise folate derivatives 24 makes it essential to include folates in our diet, along with other important nutrients such as vitamin B12 and B6.23 Therefore, without folate derivatives, cell replication cannot take place. 1.1. Folate metabolism Folate metabolism in the malaria parasite is a complex set of enzymatic processes in which folate derivatives are synthesised: the de novo pathway; or in which folates derivatives are recycled: the salvage pathway.24 The metabolic processes differ between humans and the Plasmodium parasite in that humans can only salvage folic acid from nutrient sources. By comparison, malaria parasites are able to perform the de novo synthesis of folate cofactors, as well as recycle them via the salvage pathway. 24 One-carbon metabolism reactions are folate requiring and include the synthesis of pyrimidines and purines. Additionally, these reactions include the metabolism of amino acids and the formation of the key methylating agent, S-adenosylmethionine (SAM), Figure 8. The polyglutamyl form of tetrahydrofolate (THF) has the role, in the one-carbon cycle, as the central folate acceptor.23 This folate coenzyme principally functions to donate one-carbon units in key metabolic pathways. 13 Figure 8: Main reactions of one carbon metabolism and the transsulfuration pathway in folate metabolism.23 The central folate acceptor, THF is highlighted. The cycle begins by converting THF to 5,10-methylene THF and glycine and the main carbon source of this reaction is the 3-carbon of serine. The enzyme that is important for this one carbon transfer is pyridoxal phosphate (PLP)-dependent serine hydroxymethyltransferase (SHMT).23 A portion of the resulting 5,10-methylene-THF produced undergoes conversion, forming 5-methyl-THF via an irreversible enzymatic reduction, catalysed by the enzyme methylene tetrahydrofolate reductase (MTHFR).25 The 5-methyl-THF has an N-5 group that is solely metabolically utilised for transfer to homocysteine. A removal of the methyl group from the 5-methyl-THF occurs by a methionine synthase reaction and transfers it to the vitamin B12 coenzyme before homocysteine, consequently forming methionine.25 The methionine produced also has a role in protein synthesis and additionally it functions by donating a methyl group via conversion to SAM,24 which takes part in over a hundred methyltransferase reactions with a diverse range of acceptor molecules.25 14 The methionine synthase reaction is also responsible for the regeneration of the THF needed for forming 5,10-methylene-THF and 10-formyl-THF which have direct utilisation in thymidylate and purine synthesis.23 The thymidylate synthase reaction involves the 5,10- methylene-THF donating its CH2 unit to become the thymidine methyl group, forming dihydrofolate (DHF), (Figure 9).26 The DHF is used as a substrate by DHFR, the enzyme responsible for the reduction of DHF to THF in a NADPH-dependent reaction, Figure 9. The resulting THF is then reused in one-carbon metabolic processes already described. As described above, DHFR is a key enzyme in folate metabolism, and it is ubiquitously expressed in all organisms.27 Fortunately there is sufficient difference in amino acid sequences in the tertiary structures of DHFR enzymes in different organisms to get selectivity, for example, in the malaria parasite DHFR exists as a bifunctional dimer with thymidylate synthase. Due to the important role in nucleoside biosynthesis, DHFR has become a drug target in disease classes including malaria, cancer and for the treatment of bacterial infections. Inhibition of this enzyme with folate antagonists or antifolates, stops cell replication.28 15 Figure 9: Representation of the key components in the folate salvage pathway. SHMT: serine hydroxymethyltransferase; THF: tetrahydrofolate; DHFR: dihydrofolate reductase; DHF: dihydrofolate; NADP+/(H): nicotinamide adenine dinucleotide phosphate. The de novo synthesis of folates present in some organisms utilises GTP in the initial step, Figure 10. This is followed by the removal of a triphosphate group prior to conversion to 6- hydroxymethyl-7,8-dihydropterin catalysed by dihydroneopterin aldolase (DHNA). This is thought to occur via non-enzymatic loss of pyrophosphate and subsequent removal of the last phosphate via non-specific phosphatase activity.26 The host’s salvage pathway is responsible for the synthesis of p-aminobenzoic acid (p-ABA), a substrate for dihydropteroate synthase, 16 however it can also be synthesised via the shikimate pathway.26 Preceding complex enzyme specific reactions seen in Figure 8, the enzyme DHPS catalyses the condensation of p-ABA with dihydropteridine pyrophosphate in the de novo folate biosynthesis pathway, hence forming 7,8-dihydropteroate, which is a precursor of DHF.24 The enzymes in the folate pathway prefer the polyglutamated forms of the substrates. There is still uncertainty as to whether the new folate that is synthesized in P. falciparum is polyglutamated, at the THF or DHF stage or both, however Yuthavong et al.24, for simplicity’s sake, displayed the process happening at the THF stage (see large brackets) 24 Figure 10. Figure 10: Illustration of the use of the key enzymes in the de novo pathway. 24 GTPC: GTP cyclohydrolase I; DHNA: dihydroneopterin aldolase; PPPK: hydroxymethyldihydropterin pyrophosphokinase; DHFS: dihydrofolate synthase; FPGS: folylpolyglutamate synthase; TS: thymidylate synthase.24 17 Displayed above in Figure 10 are essential substrates and enzymes required in the folate pathway to form THF and its use in the thymidylate cycle for the malaria parasite, using both the DHPS enzyme in the de novo pathway and the DHFR enzyme in the salvage pathway.24 Antifolates target enzymes in these pathways and fall under two classes: class I antifolates, which inhibit dihydropteroate synthase (DHPS), and class II antifolates, which inhibit dihydrofolate reductase (DHFR).29 Using both DHPS and DHFR inhibitors is synergistic, therefore they are often used in combination in the treatment of malaria. The class I antifolate, sulfadoxine (SDX) and its target DHPS as well as the class II antifolate, pyrimethamine (PYR) and its target DHFR are shown in Figure 10 above.24 1.2. Antifolates and P. falciparum resistance With the acquired knowledge of folate metabolism and the function of folate derivatives in cells, antifolates were developed. Antifolates have been studied for more than 50 years.21 They are a class of anti-metabolites that act as antagonists and competitive inhibitors of the DHFR and DHPS enzymes.23 By targeting folate metabolism, antifolates stop DNA/RNA synthesis and repair, cell division and protein synthesis.24 Antimalarial drug resistance, however, is a grave concern as it has resulted in an increase in the mortality and morbidity rate of the disease in recent years.30 Drug resistance is observed globally for P. falciparum, P. vivax, and P. malariae parasites. P. falciparum has been reported to be resistant to various antimalarial drug classes, including class I antifolates. 1.2.1. Class I antifolates and P. falciparum resistance Class I antifolates are classified as DHPS inhibitors. Currently used class I antifolates are sulfa- containing compounds.24 The discovery that sulfa-drugs disrupt the de novo pathway led to the use of this class of compound as antimalarial agents.29 These sulfa-drugs mimic p-ABA (para- aminobenzoic acid) 14 and inhibit DHPS by forming sulfadihydropteroate products which further disrupt folate metabolism.24 Examples of class I antifolates include sulfadoxine 15 and dapsone 16, Figure 11.24,29 Dapsone 16, is the most potent DHPS inhibitor synthesised. Sulfadoxine 15 and dapsone 16 differ from p-ABA 14 by the replacement of the carboxylic acid (red) functional group with a sulfonamide or sulfoxide functional group (blue), Figure 11. Previous attempts to use class I antifolates as a monotherapy were unsuccessful due to their low efficacy and unacceptable toxicity.29 The low efficacy could be attributed to the fact that the folate salvage pathway ensures adequate levels of folate derivatives, despite inhibition of 18 the de novo pathway by these drugs. Interest in this class of antifolates remained however, since it was demonstrated that they acted synergistically with DHFR inhibitors. This prompted their use in antifolate combination therapies.29 The long-acting class I antifolate, sulfadoxine 15, has been used with pyrimethamine 7 (will be discussed later), a DHFR inhibitor in a combination therapy which is known as Fansidar®. The short acting class I antifolate, dapsone 16, is also used with pyrimethamine 7 in a combination therapy which is known as Maloprim®. The use of Fansidar® is very common in most malaria endemic regions for malaria prevention in pregnancy.31 Figure 11: Structure of p-ABA (14) and class I antifolates; sulfadoxine (15) and dapsone (16). Recent studies have shown that point mutations S436A, A437G, K540E, A581G, and A613S cause resistance to class I antifolates and therefore reduce the effectiveness of these drugs. 30,32 1.2.2. Class II antifolates and P. falciparum resistance Class II antifolates are a class of compounds that target DHFR.29 Aminopterin (4-aminopteroyl- glutamic acid), (17, Figure 12), was the first clinically accepted folate analogue. It was first clinically used in 1947 by Sydney Farder and co-workers, in the treatment of children with leukaemia.29,33 Aminopterin’s success was short-lived however, due to a developed toxicity 29 and it was also observed that within a few months of treatment, leukaemia cells grew again, and the newly growing leukaemia cells were resistant to aminopterin 17.33 This influenced the development of methotrexate (4-amino-10-methylpteroyl-glutamic acid) (18, Figure 12), which resulted in a better therapeutic index, hence receiving huge success in the treatment of certain cancers and autoimmune disorders.21 Both aminopterin 17 and methotrexate 18 mimic the natural substrate of DHFR, dihydrofolic acid (DHF) 19. Both compounds, aminopterin 17 and methotrexate 18, respectively, replace the carbonyl functional group (red) of DHF 19 with an amine group (blue). In methotrexate 18, the amine group (green) of DHF 19 is substituted with a methyl group (purple), Figure 12. 19 Figure 12: Structural comparison of aminopterin (17), methotrexate (18) and dihydrofolic acid (DHF) (19). The first reported class II antimalarial antifolate was proguanil 8, which was discovered in 1945, Figure 13. Its better activity and higher therapeutic index than quinine (5, Figure 14) against avian malaria in animal models prompted its use in humans.29 Proguanil 8 is a prodrug which inhibits the DHFR enzyme after being metabolised in vivo to its dihydrotriazine form, cycloguanil 20 (Figure 13).34 Proguanil 8 has been used as a prophylactic agent in monotherapy and synergistically with atovaquone 21, Figure 14 an inhibitor of electron- transport to the cytochrome bc1 complex (coenzyme Q) in a combination therapy called Malarone®. Figure 13: Structures of proguanil (8) and cycloguanil (20) Figure 14: Structures of quinine (5) and atovaquone (21) 20 The success of proguanil 8 initiated the search for dihydrotriazine analogues. These findings made way for the discovery of chlorproguanil 22, which cyclizes to its dihydrotriazine form chlorcycloguanil 23 in vivo (Figure 15). Chlorproguanil 22 comprises of a dichlorinated phenyl ring in contrast with the monochlorinated phenyl ring of proguanil and as previously mentioned, is also metabolised to the active metabolite, chlorcycloguanil 23. The higher potency of chlorproguanil 22 compared to proguanil 8 has prompted its recommended use as a prophylactic agent at lower doses.29 However, research has later demonstrated an inadequate prophylactic protection from this drug. This antifolate has been used in combination therapy with dapsone 16 under the name Lapdap®.29 Figure 15: Structures of chlorproguanil (22) and chlorcycloguanil (23). Clociguanil (BRL 50216) 24, Figure 16, was also discovered. This is a derivative of chlorcycloguanil 23 and differs by possessing a methylene-oxy linker between the dihydrotriazine and the phenyl ring. The clinical evaluation of this compound, conducted in Africa, demonstrated that clociguanil showed good antimalarial properties, however it had an erratic bioavailability and was short acting,29,35 which was not seen as an advantage over pyrimethamine 7 and proguanil 8, hence the drug was abandoned.29,35 Other discoveries included BRL 6231 (WR 99210) 25, Figure 16. This is a triazine derivative and possesses a five-atom linker between the dihydrotriazine ring and the chlorinated phenyl ring. In the early 1980s, WR 99210 25 showed selective inhibition of parasite growth in vitro,36 however, because of its low bioavailability, the biguanide pro-drug of WR 99210 25, PS-15 26 (Figure 16) was developed. PS-15 26 demonstrated improved bioavailability and a better potency than WR 99210 25 in an in vivo model, which revived interest in this molecule for clinical use. 21 Figure 16: Structures of flexible class II antifolates; clociguanil (24) WR 99210 (25), and PS-15 (26). Researchers attributed the increase in the potency of the cycloguanil analogues to the increase in chlorines in the phenyl rings as well as the increase in the length of the linker between the dihydrotriazine ring and the phenyl ring, which later become important in attempting to combat P. falciparum resistance.29,37 Pyrimethamine (7, Figure 17) is a class II antifolate belonging to the 2,4-diaminopyrimidine family.38 Interest was initiated in these compounds when they were synthesised as folic acid analogues and evaluated in the treatment of tumours.29 Falco et al.39 observed the similarity of these compounds to proguanil (8, Figure 13) and hypothesised their potential antimalarial activity. A successful screening of these 2,4-diaminopyrimidines led to the identification of the antimalarial pyrimethamine 7. It is the most utilised antimalarial antifolate agent to date. As previously mentioned, pyrimethamine 7 is used mostly in combination with sulfadoxine 15 or dapsone 16, Figure 11, known as Fansidar® and Maloprim® respectively.29,38 However it has also been used in monotherapy, but to a lesser extent and is sold under the brand name Daraprim.29 Figure 17: Structure of pyrimethamine (7) 22 As discussed previously, class II antifolates target the PfDHFR enzyme. Research has shown that the following point mutations; N51I, C59R, S108N, and I164L cause drug resistance.30,32 These mutations in the PfDHFR enzyme have caused previously potent drugs such as pyrimethamine 7 and cycloguanil 20 to lose activity with IC50 values of > 90 µM against parasite strains bearing mutant forms of DHFR.40 Mutations, S108N and C59R cause steric clashes with inhibitors in the DHFR active site, while mutations, N51I and I164L shift the chains around the active site, causing the binding affinity of small analogues to be reduced.40 When the P. falciparum resistant strain bearing quadruple mutant DHFR was observed and studied by researchers, findings showed that the N108 side chain of the mutant enzyme had steric clashes with the p-Cl atom of pyrimethamine 7, causing decreased affinity to the mutant enzyme. However, on the contrary, this mutant strain was sensitive to WR 99210 25.41 This can be attributed to the structure of WR99210 25. Unlike smaller molecules, such as pyrimethamine 7 or cycloguanil 20 that bear a rigid bi-aryl axis that restricts their conformation, WR 99210 25 possesses a flexible five-atom linker chain. This is important because the inherent flexibility of WR 99210 25 allows the molecule to avoid steric clashes due to mutant amino acids and the flexible side chain adopts a conformation which fits well in the active site, therefore contributing to binding.42 This is illustrated in Figure 18 (A-D) by a molecular model representation of pyrimethamine 7 and WR99210 25 binding to the active sites of both the wild type and quadruple mutant PfDHFR. Furthermore, Figure 18 (E-F) shows how the binding position of pyrimethamine 7 changes poses in the active site of wild- type PfDHFR (blue) and quadruple mutant PfDHFR (red), this change in position may attribute to a reduced binding affinity in the quadruple mutant PfDHFR. As for WR99210 25, the triazine core maintains its position and only the flexible linker changes conformation for facilitate favourable binding in the respective active sites. 23 Figure 18: Molecular model illustrating the rigid binding of pyrimethamine (7) and the flexible binding nature of WR99210 (25) in both the wild-type and quadruple mutant PfDHFR active sites. A-B: Comparison of pyrimethamine (A) and WR99210 (B) binding to wild-type PfDHFR. C-D: Comparison of pyrimethamine (C) and WR99210 (D) binding to quadruple mutant PfDHFR bearing the following point mutations, N51I, C59R, 24 S108N, and I164L.43 E-F: Overlay of pyrimethamine (E) and WR99210 (F) in active site of wild-type PfDHFR (blue) and quadruple mutant PfDHFR (red). The emerging antimalarial drug-resistant parasite threatens the malaria control efforts achieved thus far. This has prompted researchers to look for new and improved treatment regimens to treat and control the disease. 1.3. Synthesis of 6-substituted-2,4-diaminopyrimidine-5-phenethylamines as inhibitors of PfDHFR The first component of this research project aims to synthesise class II antifolates bearing a pyrimidine core. This part of the research is a continuation of a bigger project which was initiated by Prof. Rousseau while working at the CSIR. Active antimalarial dihydrotriazine analogues of general structure 27 were synthesised and shown to be inhibitors of PfDHFR, Figure 19. These triazine compounds also mimic the natural substrate DHF 19. Key functional groups that are important for activity have been highlighted. These include: the nitrogen backbone (purple), which is key for binding to the active site; the four atom linker chain (blue), which is essential for flexibility of the compound and finally the substituted phenyl ring (orange), which is important for additional binding in the active site (Figure 19). Figure 19: Substituted dihydrotriazine (27) prepared previously alongside DHF (19).44 A series of analogues bearing flexible linkers ranging in length of two up to six atoms between the dihydrotriazine and the phenyl ring was synthesised. The rationale behind the synthesis of compounds 27 bearing flexible linkers is that this allows the compound to adopt conformations 25 that can avoid steric clashes in mutant amino acids in the active site, due to its inherent flexibility. This flexibility is also advantageous in that the compound 27 can adopt conformations that can find new binding sites in the active site of the enzyme, similar to what is observed for WR99210 (25), Figure 18. Rousseau et al. reported that analogues bearing the four atom linker showed the best antiplasmodial activity.44 However, the presence of a chiral centre in the resulting product gave racemic mixtures, which was not desirable, and an enantioselective synthesis was not possible.44 Our work over the past three years has provided a solution to the latter by synthesising pyrimidine derivatives (28, Figure 20), which lack the chiral centre and hence allow for a single product to be synthesised. These compounds are structural derivatives of the earlier antifolate, pyrimethamine 7, which has lost activity against drug resistant malaria strains. We proposed a synthetic route to compounds of general structure 28 as shown in Scheme 1. The difference in in vitro antiplasmodial activity when R is a phenyl, cyclohexyl or cyclopropyl group and X is OMe, F or Cl, will be determined. In this manner we hope to establish structure-activity relationships (SAR) for antiplasmodial antifolates bearing a pyrimidine core substituted with a flexible side chain. Figure 20: General structure of pyrimidine analogues proposed in this work. Scheme 1 shows the disconnection from 6-substituted-2,4-diaminopyrimidine-5- phenethylamines which are of interest in this project. As mentioned previously, these compounds are structural derivatives of pyrimethamine 7, and structural modifications include replacing the rigid bi-aryl bond of pyrimethamine 7 with a more flexible 4-atom linker chain in the C-5 position of the pyrimidine ring and varying the R groups. This characteristic has been proven to be advantageous in that when point mutations occur in the PfDHFR active site, the flexibility of the linker chain allows the compound to position itself to avoid steric clashes with mutant amino acid residues, or alternatively find additional binding regions. As represented in the disconnection in Scheme 1, a multicomponent coupling reaction between suitably substituted aldehydes 29, malononitrile and guanidine hydrochloride under basic conditions forms a substituted 2,4-diaminocyanopyrimidine 30. Compound 30 can be subjected 26 to a reduction reaction to afford the aldehyde 31. Compound 31 can then undergo a reductive amination reaction which will give the desired substituted amine 28. Scheme 1: Disconnection from target compound (28) The analogues synthesised will be subjected to PfDHFR enzyme inhibition assays against both the wild type and quadruple mutant PfDHFR to test for inhibitory activity. Selectivity will be assessed by determining inhibition of human DHFR in an enzymatic assay. Analogues yielding promising inhibitory results will then be assessed in whole cell P. falciparum assays in vitro against both drug sensitive and drug resistant strains to assess the antiplasmodial activity. Compounds will also be assessed in standard in vitro cytotoxicity assays. 27 CHAPTER 2: RESULTS AND DISCUSSION The first component of the research project focuses on the synthesis of 2,4-diaminopyrimidines 28, Scheme 2, as potential antiplasmodial antifolates. The synthesis will be achieved following the protocol shown in Scheme 2. Scheme 2: Reagents and conditions: a) Malononitrile, guanidine hydrochloride, NaOAc, H2O/EtOH, reflux, N2, 2.5-3h; b) Raney Ni, 81% aq. Formic acid, reflux, N2, 1 h; c) NaBH(OAc)3, 1,2-DCE, glacial AcOH, substituted phenethylamine, reflux, N2, 3h. Previous work has been done by Professor Amanda Rousseau, while working at the CSIR, on a similar set of analogues, where active dihydrotriazine compounds 27, Figure 21, were synthesised. These dihydrotriazine compounds 27 showed antiplasmodial activity in vitro against both the wild-type and drug resistant parasite strains.44 The drug resistant parasite strain that compounds 27 were assessed against was the P. falciparum cycloguanil resistant mutant (Pf FCR-3). These compounds 27 were highly active against the Pf FCR-3 strain showing IC50 values in the range of 2.66-171 nM, some examples of which are included in Figure 21, with the most active compound of the series being 27 when X = 2,4-diCl (IC50, 3 nM). The compounds were also shown to be potent inhibitors of both wild type (Ki 2.3 – 29.5 nM) and quadruple mutant (Ki 4.9 – 29.1 nM) PfDHFR. Although these compounds showed promising activity, they contain a chiral centre, which gave rise to racemic mixtures. This was not desirable as a stereoselective synthesis was not readily accessible, and purification of the racemates proved challenging resulting in low yields of product. To avoid these issues, pyrimidine analogues 32, Figure 21, of these active dihydrotriazines 27 were synthesised in an earlier project by Seanego et al.45 These pyrimidine analogues 32 eliminated the chiral centre at C-6 (denoted as an * in 27) and hence solved the issues mentioned previously. Unfortunately, these pyrimidines 32 were less potent than the dihydrotriazines 27 against the same strain of the parasite in vitro and displayed IC50 values in the low micromolar range (0.9-27 µM, Pf 28 FCR-3),45 with compound 32 where X = 3,5-diCl demonstrating the best activity of the series, Figure 21. This significant drop in activity could potentially be attributed to compounds 32 possessing different pharmacokinetic properties, or the introduction of a rigid sp2 hybridised biaryl axis now present at C-6 between the phenyl substituent and the pyrimidine ring in 32, which replaced the sp3 hybridised C-6 observed in the dihydrotriazines 27. Enzyme inhibition activity of these compounds 32 against PfDHFR, however was not assessed. Figure 21: Selected general structures and IC50 values of previously prepared 2,4- diaminopyrimidines (32) and dihydrotriazine compounds (27).44,45 As a continuation of the previous work, the effect of the 6-phenyl substituent of compounds 32 on inhibitory activity against PfDHFR is being determined in a separate project, where compounds 33, Figure 22, where the ether linker chain is maintained and R varies by having more flexible alkyl substituents, are being synthesised and assessed. This research project however, will focus on synthesising 2,4-diaminopyrimidines of general structure 28, Figure 22 that possess an amine (-NH) group in the linker chain of compounds 28, which may have a role in the binding of the compounds. The free NH may improve the pharmacokinetic properties of the compounds and may have a positive influence on LogD values of these compounds. We will synthesise analogues with a 6-phenyl substituent to determine if the change in linker has any effect in the biological activity; but will also introduce an sp3 hybridised cyclohexyl substituent and a smaller cyclopropyl substituent at R to determine the effect of these cycloalkane substituents on the inhibitory activity against PfDHFR. These substituents will provide information about the effects of having a rigid aromatic group, that is, the phenyl ring, in comparison to an sp3 hybridised alkyl group like the cyclohexyl ring on the activity of the compound. Furthermore, having the smaller cyclopropyl group will assist in 29 comparing how effective smaller cycloalkanes in comparison to larger cyclohexyl and phenyl groups are when it comes to the compound’s activity. Should these variations be feasible, as a continuation of the work, we might consider the use of non-classical bioisosteres such as cubane as an R substitient.46 Bioisosteres are reported to have an advantage over aromatic substituents, such as phenyls, in that they are reported to improve solubility.46 To this end, this thesis will report our efforts towards the synthesis and biological evaluation of compounds 28. Figure 22: General structures of diaminopyrimidines (33) and (28) being prepared. 2.1. In silico molecular modelling - induced fit docking of proposed analogues (28) In silico molecular modelling of compounds 28 in the active site simulation of PfDHFR was performed using Schrodinger software and the crystal structure of our target enzyme DHFR. The model used (1J3K) had the known compound WR99210 25 and a co-factor bound in the active site and were removed prior to docking. Figure 23 shows the compounds whose predicted binding ability was assessed using induced fit docking (IFD) on Schrodinger. An induced fit docking (IFD) study was performed on compounds 28 bearing 6-phenyl, 6- cyclohexyl and 6-cyclopropyl substituents respectively, with X = OMe, F, Cl (from readily commercially available phenethylamines). This study assessed the binding ability of compounds 28 against the quadruple mutant PfDHFR (PDB structure 1J3K47) which possesses the following mutations: N51I, C59R, S108N, and I164L. This study incorporated docking constraints in the form of forced hydrogen bonds to the backbone of residues Ile14 and mutant Ile164 and to the Asp54 side chain. This was done to ensure the ligands adhered to the known conserved binding mode. For the initial Glide docking and Prime refinement steps, the default settings were used, while for the redocking step, the more rigorous Glide XP algorithm was used for the glide redocking step. 30 Figure 23: Compounds that were subject to IFD studies using Schrodinger. Compounds 28 were then subjected to the IFD experiments and each compound generated several docking scores, including the XP GScore (calculated for the final glide XP docking step) the Prime energy (approximates the overall energy of the binding pose) and the IFDScore (combines the Prime energy and docking score into a single term). For all docking scores, the lower the calculated energy is, the better the binding of the compound. Table 1 shows the results produced by the study. The IFD results shown in Table 1 predicted a high binding ability (low energies) of the series of compounds 28 to the target enzyme with IFD scores for compounds 28a-g’ ranging from -2471.36 to -2464.66 cal.mol-1. These results compared well with the known ligands pyrimethamine (PYR) 7, WR99210 25 and the natural substrate dihydrofolic acid (DHF) 19 with IFD scores of -2466.26, -2462.52 and -2470.90 cal.mol-1 respectively. Encouragingly, our series of compounds mostly generated better IFD scores than WR99210 which was reported to be active against P. falciparum strains bearing (quadruple mutant) DHFR.41 31 Table 1: Induced Fit Docking results of compounds 28a-g’ into the active site of PfDHFR (1J3K). [a] All docking scores given in units cal.mol-1. When comparing the binding ability of the compounds 28, a trend in the binding ability of the substituents X was observed. Compounds bearing methoxy substituents generally displayed better IFD docking scores in the ranges of (-2471.58 to -2468.92 cal.mol-1). Compounds possessing 2,4- and 3,4-dichloro substituents (-2469.84 to -2466.36 cal.mol-1) followed, then the chloro substituents (-2469.43 to -2464.66 cal.mol-1) and finally the fluoro substituents in the ranges of -2469.43 to -2467.95 cal.mol-1. Additionally IFD results predicted that the position of the X-substituent on the phenyl ring had a significant influence on binding, with the 3-position being clearly favoured to the 4-position and 2-position respectively. No clear trend was observed for the 6-R substituents, however, on average the 6-cyclopropyl series 28w-g’ displayed the best results with an IFD score average of -2468.97 cal.mol-1 followed by the 6- phenyl series 28a-k (IFD score average of -2468.89 cal.mol-1) and the 6-cyclohexyl series 28l- v with an average IFD score of -2468.62 cal.mol-1. Surprisingly, the 6-phenyl series produced better IFD data than we had anticipated, especially for compounds 28a-c, which had the best IFD scores for our series of compounds. The reported Entry XP GScore[a] Prime energy[a] IFD Score [a] Entry XP GScore[a] Prime energy[a] IFD Score [a] 28a -13.027 -49146.7 -2470.36 28s -11.239 -49167.7 -2469.62 28b -13.874 -49154.1 -2471.58 28t -10.425 -49162.0 -2468.53 28c -11.004 -49166.1 -2469.31 28u -8.926 -49148.6 -2466.36 28d -11.890 -49132.3 -2468.50 28v -11.196 -49155.4 -2468.96 28e -12.331 -49142.0 -2469.43 28w -11.028 -49159.7 -2469.01 28f -10.758 -49155.7 -2468.54 28x -10.993 -49167.4 -2469.36 28g -7.883 -49135.6 -2464.66 28y -11.231 -49159.6 -2469.21 28h -12.331 -49142.0 -2469.43 28z -10.998 -49157.5 -2468.87 28i -12.091 -49128.8 -2468.53 28a’ -11.356 -49131.9 -2467.95 28j -11.019 -49150.1 -2468.52 28b’ -10.122 -49168.2 -2468.53 28k -11.991 -49139.3 -2468.96 28c’ -10.192 -49164.5 -2468.42 28l -9.542 -49187.5 -2468.92 28d’ -11.386 -49154.8 -2469.13 28m -10.945 -49162.6 -2469.08 28e’ -10.726 -49168.0 -2469.13 28n -11.692 -49149.3 -2469.16 28f’ -11.483 -49167.2 -2469.84 28o -10.769 -49161.1 -2468.82 28g’ -10.967 -49165.8 -2469.26 28p -11.283 -49144.6 -2468.51 DHF -15.891 -49100.1 -2470.90 28q -10.171 -49160.2 -2468.18 WR99210 -9.257 -49065.3 -2462.52 28r -10.358 -49167.0 -2468.71 PYR -8.827 -49148.7 -2466.26 32 biological assay results on a similar set of compounds 32, Figure 21, discussed above, gave us the expectation that the 6-phenyl series may display much weaker binding than the 6- cyclohexyl and 6-cyclopropyl series. We would have attributed this to the sp2 hybridised 6- phenyl substituent which we expected to have unfavourable clashes in the active site, however this was not confirmed by the IFD scores in our modelling study. Conserved binding interactions observed between our series of compounds 28 and the active site included strong pi-cation and pi-pi interactions with Phe58. Hydrogen bond interactions were also observed for Cys15, Ile14, mutant Leu164 and mutant Asn108, with Asn108 behaving as both a hydrogen bond donor and acceptor from the secondary amine present in the linker chain, (Figure 24 B). The NH in the linker region however does not seem to be essential in terms of interactions with the protein as interactions observed for the compound in Figure 24 B are not commonly observed with the other compounds, as displayed in Figure 24 A and Figure 24 C. Hydrogen bonds to Cys15 were commonly observed for compounds bearing the 6-cyclopropyl substituent (Figure 24 C, compound 28 w). The close proximity to this residue is most likely owed to the small cyclopropyl group possessed by these compounds. Further interactions were observed with the oxygen of the methoxy group of 28a acting as a hydrogen bond acceptor to the Tyr170 residue, (Figure 24 A) and the chlorine of the dichloro-phenyl substituent of 28v having halogen bonds to Ser167, (Figure 24 B), which highlights the advantage offered by the different substituents on the phenyl ring. Furthermore, the IFD results demonstrated that at physiological pH, the nitrogen in position-1 (Figure 24 A and B) in the pyrimidine ring can undergo protonation and engage in electrostatic interactions with nearby residues, with a salt-bridge interaction to the charged residue Asp54 being especially significant, Figure 24 demonstrates these findings. 33 Figure 24: An illustration of A: 28a; B: 28v, and C: 28w bound in the active site of PfDHFR (PDB: 1J3K) and the corresponding 2D interaction diagram for each compound. 34 Compounds 28l-v and 28w-g’ bearing the 6-cyclohexyl and 6-cyclopropyl groups respectively both displayed a tightly conserved binding mode of the pyrimidine ring, with the position of the pyrimidine core maintained between ligands while the linker chain and the terminal phenyl rings displayed more variability in position, Figure 25A. Notably, when comparing with the binding mode of the 6-phenyl substituted compounds 28a-k, it is observed that these compounds occupy a slightly offset binding position with slightly different binding interactions, Figure 25B. These findings are likely caused by the rigid biaryl axis at the C-6 position present in these compounds, and may be the reason for the reduced potency of compounds 32 bearing a 6-phenyl substituent (Figure 21). Figure 25: (A) 6-Cyclohexyl (28l-v) and 6-cyclopropyl (28w-g’) analogues superimposed to show the conserved binding mode. (B) 6-Phenyl (28i) and 6-cyclopropyl (28x) analogues superimposed to demonstrate offset binding mode of the 6-phenyl analogues.40 2.2. In silico molecular modelling – qikprop pharmacokinetic property predictions of proposed analogues (28) QikProp prediction of the pharmacokinetic properties of compounds 28a-g’ was also assessed for this series of analogues and compared to the known antifolates PYR 7, WR99210 25 and the natural substrate DHF 19. Table 2 summarises the predicted pharmacokinetic properties. 35 Table 2: Qikprop predictions for pharmacokinetic properties of 28a-g’ and known antifolates PYR (7), WR99210 (25) and DHF (19). Entry #stars [a] CNS [b] MW[c] dipole [d] donor HB[e] accept HB[f] QPlogPo/w [g] QPlogS[ h] QPPCaco[i] QPlogBB[j] #metab[k] QPlogKhsa[l] HOA[m] %HOA[n] PSA[o] RO5 [p] RO3 [q] 28a 0 0 349.435 2.151 5.000 5.250 2.223 -1.861 187.489 -0.613 4 -0.074 3 80.643 88.591 0 0 28b 0 -1 349.435 1.618 5.000 5.250 2.483 -2.812 131.575 -0.899 4 0.039 3 79.416 91.732 0 0 28c 0 -2 349.435 4.480 5.000 5.250 2.306 -2.887 75.682 -1.168 4 0.038 3 74.079 93.904 0 0 28d 0 -1 337.399 2.403 5.000 4.500 2.417 -2.974 81.473 -0.992 3 0.029 3 75.303 87.500 0 0 28e 0 -2 337.399 1.016 5.000 4.500 2.550 -3.269 78.844 -1.008 3 0.067 3 75.825 87.872 0 0 28f 0 -2 337.399 3.805 5.000 4.500 2.549 -3.267 78.843 -1.008 3 0.067 3 75.820 87.874 0 0 28g 0 -1 353.853 3.296 5.000 4.500 2.707 -3.012 121.468 -0.715 3 0.075 3 80.104 83.219 0 0 28h 0 0 353.853 3.033 5.000 4.500 2.814 -3.125 133.799 -0.640 3 0.094 3 81.481 82.889 0 0 28i 0 -1 353.853 3.798 5.000 4.500 2.799 -3.625 78.833 -0.969 3 0.133 3 77.284 87.875 0 0 28j 0 -1 388.299 1.558 5.000 4.500 3.431 -4.371 107.614 -0.720 3 0.275 3 83.404 84.184 0 0 28k 0 0 388.299 3.883 5.000 4.500 3.230 -3.699 133.906 -0.516 3 0.189 3 83.924 82.848 0 0 28l 0 0 355.48 1.499 5.000 5.250 2.678 -3.079 170.171 -0.682 5 0.214 3 82.552 88.10 0 0 28m 0 0 355.482 2.719 5.000 5.250 2.212 -1.624 215.902 -0.385 5 0.082 3 81.674 90.023 0 0 28n 0 -2 355.482 1.810 5.000 5.250 2.583 -3.694 100.075 -1.047 5 0.224 3 77.871 93.469 0 0 28o 0 -1 343.446 1.327 5.000 4.500 2.691 -3.519 122.930 -0.748 4 0.216 3 80.102 84.421 0 0 28p 1 0 343.446 0.805 5.000 4.500 2.579 -3.208 117.647 -0.674 4 0.176 3 79.107 83.042 0 0 28q 0 0 343.446 1.979 5.000 4.500 2.739 -3.554 132.052 -0.677 4 0.215 3 80.938 84.378 0 0 28r 0 -1 359.901 4.650 5.000 4.500 2.669 -3.265 113.959 -0.708 4 0.221 3 79.387 81.804 0 0 28s 1 0 359.901 0.808 5.000 4.500 2.828 -3.555 118.076 -0.630 4 0.240 3 80.590 83.026 0 0 28t 0 0 359.901 2.003 5.000 4.500 2.983 -3.901 131.909 -0.634 4 0.280 3 82.363 84.323 0 0 28u 0 0 394.346 3.434 5.000 4.500 3.158 -3.986 113.996 -0.565 4 0.327 3 82.253 81.757 0 0 28v 0 0 394.346 6.385 5.000 4.500 2.825 -2.446 137.151 -0.267 4 0.274 3 81.739 83.990 0 0 28w 0 0 313.402 1.898 5.000 5.250 1.465 -0.956 137.631 -0.518 4 -0.162 2 73.803 88.195 0 0 28x 0 -2 313.402 2.091 5.000 5.250 1.781 -2.828 94.967 -1.059 4 -0.096 3 72.771 93.151 0 0 28y 1 -2 313.402 0.664 5.000 5.250 1.788 -2.855 95.743 -1.062 4 -0.096 3 72.875 93.612 0 0 36 Table 2: Qikprop predictions for pharmacokinetic properties of 29a-g’ and known antifolates PYR (7), WR99210 (25) and DHF (19) continued. Entry #stars [a] CNS[b] MW[c] dipole[d ] donor HB[e] accept HB[f] QPlogPo/w [g] QPlogS[h] QPPCaco[i] QPlogBB[j] #meta b[k] QPlogKhsa[l] HOA[m] %HOA [n] PSA[o] RO5 [p] RO3[q] 28z 0 0 301.366 2.238 5.000 4.500 1.804 -2.361 129.049 -0.653 3 -0.141 3 75.288 84.538 0 0 28a’ 0 0 301.366 1.612 5.000 4.500 1.733 -1.900 130.843 -0.521 3 -0.149 3 74.976 84.581 0 0 28b’ 0 0 301.366 2.304 5.000 4.500 1.937 -2.666 120.683 -0.696 3 -0.099 3 75.544 84.880 0 0 28c’ 0 0 317.820 2.007 5.000 4.500 1.765 -2.007 102.881 -0.614 3 -0.117 3 73.297 82.426 0 0 28d’ 0 0 317.820 1.708 5.000 4.500 1.978 -2.213 136.648 -0.469 3 -0.090 3 76.747 84.466 0 0 28e’ 0 0 317.820 2.286 5.000 4.500 2.185 -3.006 121.337 -0.650 3 -0.036 3 77.037 84.905 0 0 28f’ 0 0 352.266 2.959 5.000 4.500 2.206 -2.591 102.659 -0.468 3 -0.026 3 75.864 82.423 0 0 28g’ 0 0 352.266 4.683 5.000 4.500 2.078 -1.693 210.630 -0.080 3 -0.099 3 80.702 83.874 0 0 DHF 4 -2 441.402 11.864 6.250 12.750 -0.381 -3.944 0.033 -5.176 6 -1.066 1 0.000 249.823 2 1 WR99210 1 -1 394.687 5.114 4.000 6.450 3.030 -5.159 395.962 -0.968 1 -0.055 3 91.180 88.642 0 0 PYR 0 -2 248.714 1.972 3.000 4.000 1.775 -2.905 389.391 -0.784 1 -0.245 3 83.7 74.965 0 0 Key for Table 2: [a] Number of parameters falling outside of a 95% similarity range for known drugs, [b] Central nervous system activity [c] Molecular weight [d] Calculated dipole moment, [e] Number of H-bond donors, [f] Number of H-bond acceptors, [g] Octanol/water partition coefficient, [h] Water solubility, [i] Caco-2 cell permeability in nm/s (model for gut-blood barrier, only non-active transport is considered), [j] Blood – brain barrier permeability, [k] Number of likely metabolites, [l] Human serum albumin binding coefficient, [m] Human oral absorption: Measure of human oral absorption combing serval factors, [n] Percentage Human oral absorption: Predicted human oral absorptivity percentage scale, value is used in the calculation of HOA, [o] Polar surface area, [p] Number of violations of L ipinski’s rule of 5, [q] Number of violations of Jorgensen’s rule of 3. 37 Qikprop predictions suggested good pharmacokinetic properties for our series of compounds 28. #Stars is a descriptor value that demonstrates the number of properties of the compound that fall outside the 95 % range of similar values for known drugs, a recommended range is 0- 5. Our series of compounds therefore showed excellent results with only three compounds (28p, 28s and 28y) showing 1 star and the rest of the compounds showing 0 stars, meaning that very few properties fall outside the 95 % range of similar values for known drugs. These compounds compared well with the known inhibitors WR99210 (25) and PYR (7) which showed 0 and 1 star respectively and performed better than the natural substrate DHF which showed 4 stars. None of the compounds violated either Lipinski’s rule of 5 or Jorgensen’s rule of 3. Lipinski’s rule of 5 takes into account the following factors: molecular weight (MW) < 500, log P (QPlogPo/w) < 5, number of H-bond donors (donorHB) ≤ 5, number of H-bond acceptors (accptHB) ≤ 10; and Jorgensen’s rule of 3 considers these factors: predicted aqueous solubility (QPlogS) > -5.7, predicted apparent Caco-2-cell permeability (QP PCaco) > 22 nm/s and the number of primary metabolites < 7. This therefore means that the compounds were all predicted to be drug-like and more likely to be orally available. Other properties such as the predicted central nervous system activity (CNS) with -2 (inactive) and +2 (active) predicted that none of our compounds were active against the central nervous system and this data compared well with the known inhibitors WR99210 25 and PYR 7 and the natural substrate DHF 19. QPlogKhsa which predicts binding to the human serum albumin also predicted all our compounds to be in the recommended range -1.5 – 1.5. Percentage human absorption (% HOA) predicted compounds 28 to be in the recommended range (<25 % poor; >80 % high) with values in the range of 72.8 – 82.9 %, with compound 28k showing the highest % HOA (83.9 %). None of our analogues performed better than WR99210 (91.2 %). Generally the percentage human absorption is predicted to be the highest for the 6-cycohexyl series 28l-v (77.9 – 82.6 %), followed by the 6-phenyl series 28a-k (74.1 -83.9 %) then finally the 6-cyclopropyl series 28w-g’ (72.8 – 80.7 %). The mean value and percentage compliance for the series of analogues 28a-g’ was also calculated for each of these properties and the results are summarised in Table 3 below. 38 Table 3: Average values of pharmacokinetic properties as well as the acceptable range (95 % range for similar drugs), and the percentage compliance our series of compounds 28a-g’ assessed. Property Mean value Acceptable range Percentage compliance (%) #stars[a] 0.09 0 – 5 100 CNS[b] -0.58 -2 (inactive), +2 (active) 100 MW[c] 356.38 130 - 725 100 dipole[d] 2.51 1 – 12.5 90.9 donor HB[e] 5.00 0 - 6 100 accept HB[f] 4.70 2 - 20 100 QPlogPo/w[g] 2.43 -2 – 6.5 100 QPlogS[h] -2.90 -6.5 - -0.5 100 QPPCaco[i] 123.23 <25 poor; >500 great 100 QPlogBB[j] -0.70 -3 – 1.2 100 #metab[k] 3.60 1 – 8 100 QPlogKhsa[l] 0.070 -1.5 – 1.5 100 HOA[m] 2.97 1 low; 2 medium; 3 high 100 %HOA[n] 78.33 % <25 % poor; >80 % high 100 PSA[o] 86.18 7-200 100 RO5[p] 0 Max 4 100 RO3[q] 0 Max 5 100 Key for Table 3: [a] Number of parameters falling outside of a 95% similarity range for known drugs, [b] Central nervous system activity [c] Molecular weight [d] Calculated dipole moment, [e] Number of H-bond donors, [f] Number of H-bond acceptors, [g] Octanol/water partition coefficient, [h] Water solubility, [i] Caco-2 cell permeability in nm/s (model for gut-blood barrier, only non-active transport is considered), [j] Blood – brain barrier permeability, [k] Number of likely metabolites, [l] Human serum albumin binding coefficient, [m] Human oral absorption: Measure of human oral absorption combing serval factors, [n] Percentage Human oral absorption: Predicted human oral absorptivity percentage scale, value is used in the calculation of HOA, [o] Polar surface area, [p] Number of violations of Lipinski’s rule of 5, [q] Number of violations of Jorgensen’s rule of 3. As illustrated by Table 3 above, our compounds are predicted to display pharmacokinetic properties within acceptable ranges of known drugs. This is promising for the potential antiplasmodial activity of these compounds which will be assessed. The results of this study therefore confirmed the rationale in synthesising the 2,4- diaminopyrimidine compounds bearing a 6-cyclohexyl, 6-cyclopropyl or 6-phenyl substituent for biological testing. 39 2.3. Approaches to the synthesis of 2,4-diaminopyrimidine core The 2,4-diaminopyrimidine core has an essential role in the binding of the analogues to the active site, as previously shown with the in silico molecular model, Figure 24. There have been various synthetic protocols reported to synthesise 2,4-diaminopyrimidine derivatives bearing an aromatic substituent at the C-6 position. These protocols however, do not always involve the synthesis of the 2,4-diaminopyrimidine core as the first step. Previous work reported by Seanego et al. included a four step synthesis, (Scheme 3) to synthesise 2,4-diaminopyrimidines 32.45 Initially, commercially available substituted phenols 34 were alkylated with 1,4- dibromobutane to afford substituted ethers 35. The resulting ethers 35 undergo a functional- group interconversion by reaction with potassium cyanide to generate pentanenitriles 36. An α-cyanoketone 37 is then generated by reacting the nitrile 36 with ethyl benzoate in the presence of base. The formation of the 2,4-diaminopyrimidine was achieved in the final step, where 37 was reacted with diazomethane to form the intermediate enol ether, which was subsequently reacted with guanidine hydrochloride to afford the 2,4-diaminopyrimidine derivatives 32 in low yields (9-11 %).45 Scheme 3: Reagents and conditions: (a) 1,4-dibromobutane, K2CO3, CH3CN, reflux, 20 h; (b) KCN, EtOH/H2O, reflux, 48 h; (c) ethyl benzoate, KOtBu, THF, rt, 18 h; (d) diazomethane, CH2Cl2, rt, 18 h; (e) guanidine HCl, DMSO, NaOMe, 90 °C, 24 h.45 We could employ the same method for the synthesis for our desired compounds 28 by starting with substituted anilines instead of phenols 34 in step a. However, the same method was attempted for the synthesis of 2,4-diaminopyrimidines 32 bearing a non-aromatic substituent (cyclohexyl and cyclopropyl) at C-6, and was not successful. The formation of the essential α- 40 cyanoketones 38 by reacting the nitriles 36 with either methyl cyclohexane carboxylate or methyl cyclopropane carboxylate, was unsuccessful as mainly starting material was recovered (Scheme 4). The reaction therefore could not be continued to afford the desired pyrimidine derivatives 33. We therefore did not attempt this method in the synthesis of our desired compounds 28. Scheme 4: Attempted synthesis of analogues 33 bearing non-aromatic substituents at C- 6 using previously reported methodology. The limitations posed by the methodology in Scheme 4 on the synthesis of analogues bearing smaller, more flexible cycloalkyl R groups, encouraged further investigation into alternative approaches for the synthesis of these compounds, including the synthesis of the pyrimidine ring as the first step. To synthesise the 2,4-diaminopyrimidine core in the first step, multicomponent coupling reactions (MCRs) are commonly used. MCRs are one-pot reactions where three or more compounds are reacted together to form one target compound. 48 These reactions are commonly used in heterocyclic and combinatorial chemistry. MCR methodologies to form the pyrimidine ring reported by Wan