Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Searching for inhibitors of the protein arginine methyl transferases : synthesis and characterisation… Knuhtsen, Astrid 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2016_may_knuhtsen_astrid.pdf [ 6.61MB ]
Metadata
JSON: 24-1.0228345.json
JSON-LD: 24-1.0228345-ld.json
RDF/XML (Pretty): 24-1.0228345-rdf.xml
RDF/JSON: 24-1.0228345-rdf.json
Turtle: 24-1.0228345-turtle.txt
N-Triples: 24-1.0228345-rdf-ntriples.txt
Original Record: 24-1.0228345-source.json
Full Text
24-1.0228345-fulltext.txt
Citation
24-1.0228345.ris

Full Text

   SEARCHING FOR INHIBITORS OF THE PROTEIN ARGININE METHYL TRANSFERASES: SYNTHESIS AND CHARACTERISATION OF PEPTIDOMIMETIC LIGANDS    by  ASTRID KNUHTSEN  B. Sc., Aarhus University, 2009 M. Sc., Aarhus University, 2012     A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF   DOCTOR OF PHILOSOPHY   in    THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Pharmaceutical Sciences)      THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)     March 2016  © Astrid Knuhtsen, 2016 ii  U N I V E R S I T Y  O F  C O P E N H A G E N  F A C U L T Y  O F  H E A L T H   A N D  M E D I C A L  S C I E N C E S       PhD Thesis Astrid Knuhtsen  Searching for Inhibitors of the Protein Arginine Methyl Transferases: Synthesis and Characterisation of Peptidomimetic Ligands          December 2015 This thesis has been submitted to the Graduate School of The Faculty of Health and Medical Sciences, University of Copenhagen iii  Thesis submission: 18th of December 2015 PhD defense:  11th of March 2016   Astrid Knuhtsen  Department of Drug Design and Pharmacology Faculty of Health and Medical Sciences University of Copenhagen  Universitetsparken 2 DK-2100 Copenhagen Denmark  and  Faculty of Pharmaceutical Sciences University of British Columbia 2405 Wesbrook Mall BC V6T 1Z3, Vancouver  Canada   Supervisors:  Principal Supervisor: Associate Professor Jesper Langgaard Kristensen Department of Drug Design and Pharmacology, University of Copenhagen, Denmark   Co-Supervisor: Associate Professor Daniel Sejer Pedersen Department of Drug Design and Pharmacology, University of Copenhagen, Denmark   Co-Supervisor: Assistant Professor Adam Frankel Faculty of Pharmaceutical Sciences, University of British Columbia, Canada   Assessment Committee: Chairperson: Assistant Professor Anders Bach Department of Drug Design and Pharmacology, University of Copenhagen, Denmark   Principal Scientist Christian Wenzel Tornøe, PhD Novo Nordisk A/S, Denmark  Associate Professor Louis Lefebvre Department of Medical Genetics, University of British Columbia, Canada iv  Abstract Within the last two decades research in the field of epigenetics has increased significantly as targeting the epigenetic enzymes has the potential to alter the transcription of genes. Aberrant regulation of transcription is seen in several disease states, and drugs targeting the epigenetic histone deacetylases and DNA methylases are already marketed for cancer treatment. The Protein Arginine Methyl Transferases (PRMTs) belong to an epigenetic enzyme family that is upregulated in several cancers. However, currently no inhibitors of the PRMTs have been marketed.    In this thesis several peptidomimetic strategies were utilised to modify the tryptophan residues in two peptide leads in order to discover new inhibitors of the PRMTs. One of these strategies involved constraining the side chain indole of tryptophan to the peptide backbone, thus producing a seven-membered azepinone mimetic, Aia.  The peptidomimetic efforts resulted in a structure-activity relationship study from which a constrained peptidomimetic containing two Aias was discovered to be a low micromolar inhibitor of several PRMTs. To characterise the inhibitor the conformation of the inhibitor was examined using solution-phase NMR and was shown to display an interesting turn-structure. The original peptide lead was fluorescently tagged and investigated in a cellular setting, but did not reveal any PRMT-specific localisation.  In an effort to study the binding of the discovered inhibitor with the PRMTs, protein expression in E. coli and purification was performed. This resulted in the optimisation of PRMT6 purification in order to obtain highly pure PRMT6 for isothermal titration calorimetry (ITC) studies. Unfortunately these ITC studies were unsuccessful.  Furthermore, as the constrained tryptophan mimetic had proven very useful in the peptidomimetic inhibitors of the PRMTs, we attempted to synthesise a lysine/arginine dipeptide mimetic using aziridine chemistry.         v  Abstract in Danish (Dansk Resumé) Forskningen indenfor epigenetik er i de seneste to årtier øget betragteligt idet stoffer rettet imod modulering af effekten af epigenetiske enzymer har vist potentiale til at ændre i gentranskriptionen. Abnormal regulering af gentranskription ses i adskillige sygdomme, og lægemidler målrettet som hæmmere af de epigenetiske histon deacetylaser og DNA methylaser er allerede markedsført til kræftbehandling. Protein Arginin Methyl Transferaserne (PRMTerne) er en epigenetisk enzymfamilie der er overudtrykt i forskellige kræftformer. På trods af dette findes der dog ikke på nuværende tidspunkt markedsførte hæmmere af PRMTerne.  I denne afhandling blev flere strategier benyttet til at lave peptid-lignende stoffer ved at modificere tryptophanerne i to basis peptider med det formål at finde nye hæmmere af PRMTerne. En af disse strategier involverede inkorporation af en ufleksibel tryptophan-lignende byggeblok, den 7-ledede azepinon, Aia, som har C2 i sidekæde indolen fra tryptophan bundet til peptidets rygrad. Disse peptid-modificerende strategier resulterede i et studie af struktur-aktivitet-forholdet for stoffernes effekt på PRMTerne. Således blev en rigid peptid-lignende forbindelse med to Aia’er identificeret som en lav mikromolær hæmmer af flere af PRMTerne. NMR blev brugt til at karakterisere strukturen af hæmmeren i opløsning, som viste sig at have en interessant drejnings-konformation. Det oprindelige peptid blev desuden mærket med en fluorofor og undersøgt i en cellekultur, men afslørede ikke nogen genkendelig PRMT-specifik lokalisering i cellerne.  I et forsøg på at studere bindingen af den opdagede PRMT-hæmmer blev PRMT proteiner udtrykt i E. coli og oprenset. Dette resulterede i optimering af PRMT6 oprensningen med henblik på at producere yderst rent PRMT6 protein til brug i isotermal titrationskalorimetriske (ITC) studier. Disse ITC studier endte desværre uden resultat.  Idet den ufleksible byggeblok Aia havde vist sig brugbar i udviklingen af de peptid-modificerede hæmmere af PRMTerne søgte vi ydermere at syntetisere en lysin/arginin dipeptid forbindelse ved hjælp af aziridin-kemi.            vi  Preface This PhD thesis discloses the majority of work performed during my joint PhD studies from December 2012 to December 2015 at the Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen (UCPH), Denmark and at the Department of Pharmaceutical Sciences, University of British Columbia (UBC), Canada. During the time spent at UCPH the project was supervised by Assoc. Professor Jesper L. Kristensen (main supervisor) and Daniel Sejer Pedersen (co-supervisor). At UBC the project was supervised by Assistant Professor Adam Frankel (co-supervisor).  Chapters 1 and 2 of the thesis provide a background to the project; Chapter 1 deals with epigenetics and the Protein Arginine Methyl Transferases (PRMTs) which were the targets for the medicinal chemistry efforts of the project and chapter 2 provides an overview of the use of peptides in drug discovery. Chapter 3 outlines the objectives of the project.  Chapters 4 and 5 comprise the work performed at UCPH; Chapter 4 deals with the synthesis of peptides and peptidomimetics as inhibitors of the PRMTs and chapter 5 is concerned with synthetic strategies using aziridines towards dipeptide mimetic tool compounds for use in the synthesis of peptidomimetics.  A manuscript on the work in chapter 4 concerning the peptidomimetic inhibitors may be found in Appendix 1. This manuscript will be submitted as soon as possible.   Chapter 6 gives an account of the work with protein expression of the PRMTs and isothermal titration calorimetry (ITC) experiments which was performed at UBC.  Chapter 7 provides a conclusion and perspectives to the PhD project. In addition, a review is under preparation in collaboration with researchers at Vrije Universiteit Brussel. The review is based on case studies involving constrained amino acid building blocks, such as the one used in the peptidomimetic inhibitors in chapter 4, and their use in biologically relevant peptidomimetic applications. The case studies are not included as an integral part of the thesis but may be found in Appendix 2.      This dissertation is formatted in accordance with the regulations of the University of Copenhagen and submitted in partial fulfillment of the requirements for a PhD degree awarded jointly by the University of Copenhagen and the University of British Columbia.  Versions of this dissertation will exist in the institutional repositories of both institutions. vii  Acknowledgements The past three years have been such a journey; full of ups and downs, good and less good ideas, scientific creativity as well as hard work. It has been a fantastic experience to be involved in a multifaceted project and I have learnt much more than I ever expected to. To this end I want to express my deepest gratitude to my supervisors Jesper L. Kristensen and Daniel Sejer Pedersen for the endless scientific as well as moral support, and for cheering me on from the very first day. It has been an honour to have two such insightful and knowledgeable medicinal chemists as supervisors to challenge me, push me forward and help me develop into a proper scientist.     I have truly loved to work at ILF, to interact with so many great, competent people and of course to have been a part of the Sejer and Kristensen groups. Over the years we have drunk countless cups of “coffee”, shared scientific endeavours and hardships, devoured copious amounts of delicacies and bake goods from around the world, listened to some great music as well as other sounds within the same classification, shared numerous bottles of rosé, been on adventures and have built some great teams. But most of all I have enjoyed laughing with all of you.     I also want to express my gratitude to my supervisor at UBC, Adam Frankel. Thank you for welcoming me to your lab and showing me the ropes within your branch of science. To all my friends at the Faculty of Pharmaceutical Sciences and from St. John’s College: It’s been such a blast!  To Steven Ballet and the guys at VUB: Thank you for introducing me to the azepinone chemistry as well as to Brussels, huge burgers and Belgian beers.     Dr. Brian Lohse is thanked for initially performing the phage display, as well as all kinds of biological assistance at KU. Birgitte Nielsen and Dr. Nils Nyberg are thanked for their readily available technical assistance. I am thankful to Sebastian Leth-Petersen for proof-reading this thesis.   Last but not least I want to thank all the fantastic people in my life outside the walls of KU, who have been there through all the good times as well as the bad. You mean the world to me.    Astrid Knuhtsen, December 2015  viii  Abbreviations 2-Ns 2-nitrobenzenesulfonyl moiety Aba 4-amino-1,2,4,5-tetrahydro-2-benzazepin-3-one AcOH acetic acid ADMA ω, ω-N-asymmetric dimethylarginine AEX anion exchange aHPLC analytical HPLC Aia 4-amino-1,2,4,5-tetrahydro-indolo[2,3-c]azepin-3-one Alloc allyloxycarbonyl protection group Ata (S)-7-amino-7,8-dihydro-4H-[1,2,3]triazolo[1,5-a][1,4]diazepin-6(5H)-one Atc 2-aminotetralin-2-carboxy moiety Boc tert-butyloxycarbonyl protection group Boc2O Di-tert-butyl dicarbonate CD circular dichroism COSY correlation spectroscopy DAOTA 4,8-diaza-12-dioxa-4,8,12,12c-tetrahydro-dibenzo[cd,mn]-pyrenylium Dap-OH 2,3-diaminopropionic acid DBU 1,8-Diazabicyclo[5.4.0]undec-7-ene DCC N,N'-dicyclohexylcarbodiimide DEPT-135 Distortionless Enhancement by Polarisation Transfer-135 DIPEA N,N-diisopropylethylamine DMF N,N-dimethylformamide DMSO dimethyl sulfoxide DNA deoxyribonucleic acid DTT dithiothreitol EDTA ethylenediaminetetraacetic acid ee enantiomeric excess eq. equivalents ERETIC Electronic reference to access in vivo concentrations Et2O diethyl ether EtOAc ethyl acetate EtOH ethanol FA  formic acid Fmoc fluorenylmethyloxycarbonyl protection group FPLC Fast Protein Liquid Chromatography g g-force/gravity GST Glutathione S-transferase H2A Histone 2A H2B Histone 2B H3 Histone 3 H4 Histone 4 HAT Histone Acetyl Transferase HATU 1-((dimethylamino)-(dimethyliminio)methyl)-1H-1,2,3-triazolo[4,5-b]-pyridinium 3-oxide hexafluorophosphate Hba 4-amino-8-hydroxy-1,2,4,5-tetrahydro-3H-2-benzazepin-3-one H-bond hydrogen bond ix  HBTU 1-((dimethylamino)-(dimethyliminio)methyl)-1H-benzo[d][1,2,3]triazole 3-oxide hexafluorophosphate HDAC Histone Deacetylase HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HMBC Heteronuclear Multiple-Bond Correlation spectroscopy hMCR human melanocortin receptor HPLC High Performance Liquid Chromatography hr  hour/hours HRMS High Resolution Mass Spectrometry HSQC Heteronuclear Single-Quantum Correlation spectroscopy IC50 half maximal inhibitory concentration iPrOH isopropanol IPTG isopropyl β-D-1-thiogalactopyranoside IR infrared ITC isothermal titration calorimetry J coupling constant  KD dissociation constant kDa kilodalton KDM Lysine Demethylase LB Luria-Bertani LC Liquid Chromatography LC-MS Liquid Chromatography Mass Spectroscopy LRMS Low Resolution Mass Spectrometry LSD1 Lysine Specific Demethylase 1 mCit-PRMT6 mCitrine-tagged PRMT6 MeCN acetonitrile MeOH methanol MES 2-(N-morpholino)ethanesulfonic acid min minutes MMA mono-methyl arginine mRNA messenger RNA MS mass spectrometry N stoichiometry NIR near infrared NMP N-methyl-2-pyrrolidone NMR nuclear magnetic resonance NOE Nuclear Overhauser Effect PAD Protein Arginine Deiminase PBS phosphate-buffered saline PG protection group pHPLC preparative reverse phase HPLC pI isoelectric point PKMT Protein Lysine Methyl Transferase PMT Protein Methyl Transferase POI Protein of interest ppm parts per million PRMT Protein Arginine Methyl Transferase PTM posttranslational modification PyBOP (Benzotriazol-1-yloxy) tripyrrolidinophosphonium hexafluorophosphate x  qNMR quantitative NMR ROESY Rotating-frame nuclear Overhauser Effect correlation Spectroscopy RP reversed phase rpm rounds per minute rt room temperature SAH S-Adenosyl Homocysteine SAM S-Adenosyl Methionine SCX strong cation-exchange SDMA ω, ω’-N-symmetric dimethylarginine SDS sodium dodecyl sulfate  SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis SEC size-exclusion chromatography sec seconds  SPA scintillation proximity assay SPPS solid phase peptide synthesis TFA trifluoroacetic acid Tic 1,2,3,4-tetrahydroisoquinoline-3-carboxy moiety TIS triisopropylsilane TLC thin-layer chromatography TMOF trimethyl orthoformate TMR tetramethylrhodamine TNBS trinitrobenzenesulfonic acid tR retention time UV ultraviolet Xxx random amino acid β3-hTrp β3-homotryptophan  ∆G Gibbs free energy ∆H change in enthalphy ∆S change in entropy δ chemical shift μ micro νmax wavenumber  The 3-letter and 1-letter IUPAC nomenclature for the 21 proteinogenic amino acids will be used in this thesis.  xi  TABLE OF CONTENTS Abstract ................................................................................................................................................ iv Abstract in Danish ................................................................................................................................ vii Preface ................................................................................................................................................ ivi Acknowledgements ................................................................................................................................ v Abbreviations ..................................................................................................................................... viii Chapter 1: Epigenetics and Protein Arginine Methyl Transferases ............................................................ 1 1.1 Defining Epigenetics ................................................................................................................................ 1 1.2 Chromatin and the Histone Code ............................................................................................................ 2 1.3 The Epigenetic Protein Enzyme Families ................................................................................................. 3 1.4 The Protein Arginine Methyl Transferases (PRMTs)................................................................................ 5 1.5 Inhibitors of the PRMTs ........................................................................................................................... 9 1.6 Phage Display ......................................................................................................................................... 13 Chapter 2: Peptides and Peptidomimetics ............................................................................................. 15 2.1 Introduction: Peptides in Nature ........................................................................................................... 15 2.2 Stability and Reactivity of Amides ......................................................................................................... 15 2.3 Peptides in Drug Discovery .................................................................................................................... 16 2.4 What are Peptidomimetics? .................................................................................................................. 16 2.5 Optimising Peptides into Peptidomimetics: Using the Peptidomimetic Toolbox ................................. 17 2.6 Conformational Constraints in the Peptide ........................................................................................... 19 Chapter 3: Research Objectives of the PhD Project ................................................................................ 23 Chapter 4: Developing PRMT Peptidomimetic Inhibitors from Phage Display Leads................................ 24 4.1 The Hits from Phage Display .................................................................................................................. 24 4.2 Aim of the Peptidomimetic Project ....................................................................................................... 25 4.3 Using Peptidomimetics in the Search for PRMT Inhibitors ................................................................... 25 4.3.1 Peptidomimetics based on P1 ........................................................................................................ 25 4.3.2 β-amino acids ................................................................................................................................. 26 4.3.3 Nα-methylation ............................................................................................................................... 28 4.3.4 The Aia scaffold .............................................................................................................................. 30 4.4 Evaluating the Peptidomimetics as Inhibitors of the PRMTs ................................................................ 36 4.4.1 Inhibitor Screenings using an Antibody-Based Assay ..................................................................... 36 4.4.2 SPA-based approach to determining inhibitor concentrations ...................................................... 40 xii  4.4.3 Comparing the two inhibition assays ............................................................................................. 41 4.5 Structural Studies of P21 ....................................................................................................................... 41 4.6 Cellular Studies of P16 ........................................................................................................................... 44 4.7 Conclusion to the PRMT Peptidomimetic Studies ................................................................................. 46 4.8 EXPERIMENTAL SECTION ....................................................................................................................... 48 Chapter 5: Expanding the Peptidomimetic Toolbox ............................................................................... 59 – Using Aziridine Chemistry in the Pursuit of a Lysine/Arginine Dipeptide-mimetic ................................ 59 5.1 Introduction ........................................................................................................................................... 59 5.2 Retrosynthetic Routes towards Diazepane 5.1 ..................................................................................... 62 5.3 The Chemistry of Aziridines ................................................................................................................... 63 5.4 Previous Efforts ..................................................................................................................................... 64 5.5 The Intermolecular Aziridine Ring-Opening Strategy ............................................................................ 65 5.6 The Intramolecular Aziridine Ring-Opening Strategy ............................................................................ 66 5.6.1 The Boc/N3 and Fmoc/N3 Protection Strategies ............................................................................. 67 5.6.2 The Fmoc/Boc Protection Strategy ................................................................................................. 70 5.6.3 The 2-Nitrosulfonylbenzene (2-Ns) Protection Strategy ................................................................ 71 5.7 Conclusion to the Aziridine Project ....................................................................................................... 74 5.8 Alternative Synthesis Routes to Diazepane 5.1 ..................................................................................... 75 5.9 EXPERIMENTAL SECTION ....................................................................................................................... 76 Chapter 6: Protein Purification and ITC ................................................................................................. 85 6.1 Introduction ........................................................................................................................................... 85 6.2 Protein Expression and Purification ...................................................................................................... 85 6.2.1 Lysis Buffer Optimisation ................................................................................................................ 86 6.2.2 PRMT6 Expression and Purification Optimisation .......................................................................... 87 6.3 Determining Concentration and Enzymatic Activity of PRMT6 ............................................................. 89 6.3.1 DTT Increases Enzymatic Activity of PRMT6 ................................................................................... 90 6.4 Isothermal Titration Calorimetry (ITC) .................................................................................................. 91 6.4.1 Dialysis ............................................................................................................................................ 92 6.4.2 Performing ITC on PRMT6 .............................................................................................................. 92 6.5 Discussion and Conclusion..................................................................................................................... 95 6.6 EXPERIMENTAL SECTION ....................................................................................................................... 96 Chapter 7: Conclusion and Perspectives .............................................................................................. 104 References ......................................................................................................................................... 105 1  Chapter 1: Epigenetics and Protein Arginine Methyl Transferases 1.1 Defining Epigenetics  Epigenetics refers to changes in a chromosome without alterations in the DNA nucleotide sequence. These alterations can be stable between cell divisions, and sometimes even between generations, but also subject to dynamic changes and may potentially be reversed.1,2 The word epigenome literally means “above the genome” and was coined by Waddington in 19423 to account for the gap between the inherited genotype and the expressed phenotype. Thus, in epigenetics modifications may alter the phenotype without changing the genotype, which explains why cells with the same genetic material may evolve differently.      Figure 1.1 Epigenetic enzymes may modify the histone tails thereby altering the state of chromatin. Writer and eraser enzymes add or remove marks on the histone tails, and readers interpret these. Reprinted with permission from Nature Reviews Drug Discovery, Arrowsmith et al. 2012.1    2  Whereas stable alterations in genetics often evolve over several generations, epigenetic modifications may be triggered within a much shorter time frame reflecting the dynamic environment of the cell. Epigenetic changes can result in cellular reprogramming in response to differentiation or quick adaptations due to environmental cues, such as early life experiences, lifestyle and diseases, and toxins in our environment. For this reason epigenetic modulators are recognised as important therapeutic targets for the treatment of lifestyle diseases such as cancer, inflammation and cardiovascular diseases, among many others.1,4   1.2 Chromatin and the Histone Code  In order for our genetic material to fit into each cell nucleus it is tightly packed into chromatin fibres, which contain the histone proteins around which the DNA is wound (figure 1.1). The units comprising the chromatin fibres are the nucleosomes; a histone octamer consisting of the four core histones, H2A, H2B, H3 and H4, with 147 base pairs of the DNA strand wrapped around it.5 The strong ionic interactions between the negatively charged phosphodiester backbone of the DNA and the positively charged histone proteins stabilise the nucleosome.  The chromatin fibres may be tightly or loosely packed. In the tightly packed state, also known as heterochromatin, the transcriptional machinery cannot bind to the DNA strands leading to gene silencing or repression. Euchromatin is the loosely packed form of chromatin which is associated with transcriptionally active genes. The state of the chromatin depends on post-translational modifications (PTMs) on the so-called histone tails, the charged N-terminal chains of the histone proteins protruding from the nucleosome.5   The epigenetic enzymes which catalyse these modifications are referred to as either ‘writers’, which add marks, or ‘erasers’, which remove marks. The interpretation of these marks may be facilitated by the ‘reader’ proteins, which may bind to the marks and trigger a response, often resulting in gene transcription or silencing.1 A large variety of PTMs, some report more than 60 different,6 exist for the histones including acetylation, methylation, phosphorylation, citrullination, crotonylation and sumoylation. Some of these are illustrated for the H3 and H4 histones in figure 1.2, reflecting the vast combinational scope. The complex interplay of these modifications is now known as “the histone code”.7   3  Figure 1.2 Possible posttranslational modifications for H3 and H4 tails. Only the 4 modifications in the framed box are shown. Based on data from uniprot.org for Homo Sapiens H3 (P68431) and H4 (P62805) as per November 3 2015.The N-terminal initiator methionine indicated in purple is cleaved.  Another important epigenetic mechanism is the direct methylation of the DNA nucleotides catalysed by the DNA methyltransferases. This always leads to transcriptional silencing, as steric bulk of the methyl group prevents binding of the transcriptional machinery.8 In contrast to this is the histone tail methylation of lysine and arginine which may either activate or repress gene expression, depending on a variety of factors such as location of the methylated residue in the sequence, degree of methylation (lysine residues can be non-, mono-, di- or tri-methylated, whereas arginines can be either non-, mono- or di-methylated, vide infra) and the so-called crosstalk between neighbouring modifications.7,9 Due to the abundance of modifications and combinations of these the histone code seems unlimited. Understanding the enzymes which modulate the histone code could thus be the key to understanding the histone code and its implication in physiology and pathophysiology.   1.3 The Epigenetic Protein Enzyme Families  So far the most studied epigenetic enzyme families involve acetylation and methylation, reflecting that these are also among the most abundant marks in the epigenome.1  The histone acetyltransferases (HATs) transfer an acetyl group onto lysine, thereby removing the charge on the side chain which weakens the binding of the DNA in the nucleosome. The histone deacetylases (HDACs) catalyse the reverse reaction also on both histone and non-histone targets. The importance of acetylation is highlighted by that fact that acetylation on lysine 16 in histone 4 (H4K16) is a crucial switch between heterochromatin and euchromatin.10 Furthermore, two HDAC inhibitors are already being used in cancer treatment of T cell lymphoma and several are in clinical development for various diseases.1,4,11  4  Lysine methylation state is controlled by the protein lysine methyltransferases (PKMTs) and the lysine demethylases (KDMs). The PKMTs are shown on the phylogenetic tree in figure 1.3 on the blue and green branches except DOT1L which is situated on the red branch with the Protein Arginine Methyl Transferases (PRMTs),1 signifying that DOT1L shares more sequence similarity with the PRMTs than the PKMTs. DOT1L is a unique PKMT as it does not contain a SET-domain, a property by which the other 50 known PKMTs were identified.12 This range of enzymes seems to be required as PKMTs differ in their effects by virtue of residue localisation selectivity and degree of methylation.13 Interestingly, the KDMs were first discovered a decade ago, and until then the lysine methylation was regarded as being a very stable modification.14   The Protein Arginine Methyl Transferases (PRMTs) transfer one or two methyl groups onto arginine, but curiously the arginine demethylases have not yet been identified. The dynamic appearance and disappearance of methylated arginines do however indicate that these exist.15,16 Instead the Protein Arginine Deiminases (PADs) are recognised to antagonise the PRMTs by converting non- or mono-methylated arginines into citrulline, which changes the chemical properties of the residue side chain entirely.15  As the PRMTs have been the target enzymes for this PhD-project they will be elaborated on further in the following section.       Figure 1.3 Phylogenetic tree of the Protein Methyl Transferases (PMTs) indicating sequence similarities and thus evolutionary background. The circled red branch contains the Protein Arginine Methyl Transferases (PRMTs) and DOT1L, a Protein Lysine Methyl Transferase (PKMT). The other branches all contain PKMTs. Reprinted with permission from Nature Reviews Drug Discovery, Arrowsmith et al. 2012.1      5  1.4 The Protein Arginine Methyl Transferases (PRMTs) Post-translational protein arginine methylation has been shown to impact several cellular processes such as gene transcription, mRNA splicing, DNA repair, cellular localisation of proteins, cell fate determination and signalling,17-25 thus emphasising the importance of the PRMTs.  Thus far 9 PRMTs have been broadly acknowledged to belong to the PRMT family. Two F-Box proteins, FBXO10 and FBXO11, were proposed to belong to the PRMT family, but did not show methyltransferase activity and can thus not be considered true PRMTs.26 Chemogenetic analysis proposes the existence of up to 44 PRMTs in the human proteome.12  Function The PRMTs may be divided into three types based on their preferred methylation products (Figure 1.4). All PRMTs can mono-methylate arginine residues on the ω-nitrogen. In a second step Type I PRMTs catalyse the formation of ω, ω-N-asymmetric dimethylarginine (ADMA) and PRMTs 1-4, 6 and 8 belong to this class.i PRMTs 5 and 9 belong to the Type II class, which catalyses the formation of ω, ω’-N-symmetric dimethylarginine (SDMA). PRMT7 is a Type III PRMT, and can only produce mono-methyl arginine (MMA).      Figure 1.4 The PRMTs may be divided into three types; Type I produces ADMA, Type II produces SDMA and Type III produces only MMA. All require the S-Adenosyl Methionine (SAM) co-factor in order to methylate the substrate.                                                           i PRMT4 is also known as CARM1 (Coactivator-Associated Arginine Methyltransferase 1), but for simplicity it will be referred to as PRMT4 in this thesis. 6  The methyl-group is transferred to the arginine-containing substrate from the co-factor S-Adenosyl Methionine (SAM, also referred to as AdoMet) yielding the methylated arginine and the byproduct S-Adenosyl Homocysteine (SAH, also referred to as AdoHCy).17,27   Structure Although the PRMTs vary in length they all contain a highly conserved catalytic core region of about 310 amino acids.28 This region consists of a SAM-binding domain, a β-barrel and a dimerisation arm (orange, green and red in figure 1.5A, respectively). The SAM-binding domain, which adopts a typical Rossmann fold, is highly conserved in the PRMTs as well as other SAM-dependent methyl-transferases.29 Being unique to the PRMT family, the β-barrel domain is thought to be important for substrate binding.28 The dimerisation arm is a short α-helix which connects PRMT protomers to form dimers.    Figure 1.5 A: The structure of the dimerised PRMT1 showing the structural regions that are conserved in most PRMTs. Figure reprinted with permission from Chemical Reviews, Fuhrmann et al. 201530 B: Active site of PRMT1 indicating the residues which hydrogen-bond to SAH or the substrate. (PDB code 1OR8, rat PRMT1, in PyMol v1.3 Schrodinger LLT).  Active site The active site is situated in a deep, narrow pocket inside the dimer between the SAM-binding domain and the β-barrel, which has the dimensions to fit the guanidinium side chain of an arginine.31-34 The active site contains several residues important for co-factor binding as well as a double E-loop (consisting in PRMT1 of E144 and E153), essential for recognition and alignment of the guanidinium group in order for proper catalysis to occur (Figure 1.5B).31,33 The methyl-transfer proceeds via a bimolecular SN2-mechanism aided by the double E-loop.31,35  7   Figure 1.6 Schematic representation of the 9 known PRMTs. The yellow region indicates the conserved core region, all situated C-terminally, except in PRMT4. PRMTs 1,2,3,4,6 and 8 all have the THW loop conserved. Figure reprinted with permission from Chemical Reviews, Fuhrmann et al. 2015.30  Dimerisation Each PRMT has a single catalytic core region except PRMT7 and PRMT9, which contain two (yellow bars, Figure 1.6). Crystal structures of several PRMTs have shown that these mainly exist as head-to-tail homodimers or even higher order multimers (Figure 1.5A).31-34,36 Dimerisation seems to be essential for the function, as deletion of the dimerisation arm in PRMT1 leads to a complete loss of enzymatic activity.33,37 PRMT7 exists as a monomer, but folds back onto itself forming a pseudodimer.38 In mammalian PRMT7 the second catalytic core is inactive, which could explain the propensity to produce MMA,38 however this was not seen in protozoa.39  Product Selectivity The importance of product selectivity is reflected in the fact that symmetric dimethylation of H4R3 leads to gene silencing whereas asymmetric dimethylation is associated with gene activation.18,19,40  The propensity for Type I PRMTs and Type II PRMTs to exclusively produce ADMA and SDMA, respectively, has been attributed to a few residues close to the active site. Mutagenesis of a methionine in PRMT1 to a phenylalanine (M48F) showed production of both dimethylated arginines,41 and the F327M mutation in PRMT5 likewise yielded both products.40 Also the THW loop, which is conserved in Type I PRMTs but not in Type II or III (Figure 1.6) has been suggested to affect product outcome.38 It thus seems that subtle alterations in active site architecture may affect the product switch between ADMA and SDMA.   8  Substrate Recognition The dimerisation of the enzymes likely restricts the use of larger folded proteins as substrates, and due to the location of the active site the arginines of the substrates for the PRMTs are often seen in unstructured regions such as loops or the two protein termini.42 For similar reasons the arginines undergoing methylation are often surrounded by glycine-rich regions.17,43 Substrate recognition by the enzymes also depends on residues distal from the arginines as the acidic surface of the enzymes favour binding to positively charged residues in the substrate sequence.44,45  Localisation Most PRMTs are house-keeping enzymes and thus ubiquitously expressed in cells.27,46 The one exception is PRMT8, which is only found in the brain.47 The subcellular localisation of most PRMTs is cell specific46 except PRMT6 which has a nuclear localisation signal.48 PRMT1 is said to account for >50% of the ADMA in vivo49,50 and exists in many isoforms, and these splicing variants have different localisations and thus different cellular functions.51    Regulation The cell regulates arginine methylation in several ways.27 Firstly, post-translational modification of the PRMTs may regulate activity, one example being phosphorylation of PRMT4.52 Secondly, modulator proteins may regulate function, the best example being MEP50-binding to PRMT5, which is necessary for catalytic activity.36,53 Thirdly, adaptor proteins may bridge the PRMT and the substrate which may tailor the effect of the enzyme.21 Fourth, microRNA may regulate the expression of the PRMTs endogeneously.54 Fifth, methylation events may be regulated by histone crosstalk, illustrated by the acetylation of H4K5 which results in a symmetric rather than asymmetric methylation of H4R3.55 Lastly, subcellular compartmentalisation will also provide control of methylation.46      Biological relevance The importance of the PRMTs is probably most evident by the fact that knocking out Prmt1, 4 and 5 in mice leads to embryonic lethality,27 indicating their pivotal role in cell growth, proliferation and differentiation.23,49,56,57 However, the enzymes are not essential for basal cellular metabolism.49,57 PRMT1 and PRMT5 are the major asymmetric and symmetric arginine methyltransferases,27,50 respectively, but the PRMTs seem to have distinct functions and substrates and cannot be substituted by other enzymes in the family in vivo,27,49,50,57 possibly due to the various additional domains of the enzymes27 (figure 1.6).   Mapping the effects of arginine methylation on the histone code is still a work in progress, but has thus far resulted in identification of genetically activating or silencing events caused by methylation of various histone-tail arginines using several of the PRMTs.18,19,27,58,59 9  The PRMTs have several other non-histone protein targets.16,17,20,22,25,27,53 They can indirectly affect transcription by methylating transcription factors, such as the tumor suppressor protein p53,20 or effector proteins responsible for transduction in signaling pathways, such as the estrogen receptor.16 PRMTs may also function in larger complexes with other proteins in vivo, complicating the understanding of substrate selectivity and the effects of methylation.17   In cancer the PRMTs are generally overexpressed rather than lost, and PRMTs 1-7 have thus far been associated with cancer.27 Significant data suggests that PRMT1 and PRMT4 are overexpressed in breast and prostate cancer, among others, as they are co-activators of the nuclear receptors,16,60-63 whereas overexpression of PRMT5 and PRMT6 suppress tumor suppressor genes associated with lung cancers and leukemia.54,60,64   1.5 Inhibitors of the PRMTs It is the current understanding that undruggable targets in diseases such as cancer, like the transcription factor MYC,65 may be targeted by inhibiting the epigenetic enzymes that regulate their binding partners and  are dysregulated in the pathophysiological state.1 Since several disease states are associated with dysregulated PRMTs1,11,27 much effort has been invested in discovering potent PRMT inhibitors. Preferably such inhibitors should have selectivity for the PRMTs over other epigenetic enzymes as well as subtype-selectivity within the PRMT family.  In the following examples of non-selective and selective PRMT inhibitors will be presented highlighting the various discovery methods and applications as well as important findings (figures 1.7-1.8). These efforts may be divided into two classes; the small molecule inhibitors (figure 1.7) and the bisubstrate/covalent inhibitors (figure 1.8).  Since 2004 a variety of small-molecule PRMT inhibitors have been discovered66-78 (figure 1.7). Generally, the discovery process started with a virtual library67,68,75 or an in vitro library screen.69-71,76,77 Some of these leads were later developed further to improve potency.66,72,74,78   AMI-1 (1.1) was discovered in an in vitro library screen containing 9000 structures. It showed selectivity for PRMT1 over PRMTs 3, 4 and 6 and other epigenetic enzymes, was cell permeable and showed inhibition in cells.76 However, AMI-1 was found to target the histone substrate rather than the enzyme.75 In two virtual screens the Jung group discovered several leads including stilbamidine (1.2).67,68 The compound inhibited PRMT1 in an antibody-based assay and was shown to block the transcriptional activation of the estrogen receptor in a cellular system at high doses. However, it was also shown to affect lysine methylation.68    10  Inspired by the resemblance of the amidine group in stilbamidine with the guanidinium moiety in arginine, Yan and coworkers synthesised a library of diamidines in which furamidine (1.3) was shown be a low micromolar inhibitor.66 It was >17-fold selective for PRMT1 compared to PRMTs 4, 5 and 6 and decreased proliferation in leukemia cells. However, it was also shown to bind in the DNA minor groove due to its flat structure and this could potentially explain the effects seen in cells. The allosteric PRMT3 inhibitor SGC707 (1.4)72 was developed based on a lead structure identified in an in vitro screen of 16,000 compounds.69 This inhibitor showed an IC50 of 31 nM without affecting any of the other 31 tested methyltransferases.     Figure 1.7 Small molecule inhibitors of the PRMTs: Non-selective PRMT inhibitors (1.1 and 1.2)68, 76 and selective inhibitors for PRMT1 (1.3),66 PRMT3 (1.4),72 PRMT4 (1.5),73 PRMT5 (1.6)77 and PRMT6 (1.7).71 11  Similarly to its predecessor SGC707 was found to bind in a novel allosteric site at the base of the dimerisation arm without disrupting dimerisation as determined by crystallography.69,72 Furthermore, the compound reduced methylation of the H4R3 mark comparably to catalytically inactive PRMT3.72 On the basis of an in vitro library screen Purandare et al. discovered the pyrazole scaffold as a selective inhibitor of PRMT4.70 This scaffold was further optimised in a ligand-based approach,74 ultimately leading to CMPD2 (1.5),73 a low nanomolar inhibitor of PRMT4. Interestingly, CMPD2 binds to the substrate binding cavity in the presence of SAH. The crystal structure revealed the L-alaninamide moiety to be interacting with the double E-loop thus mimicking the arginine substrate. Recently, scientists from Epizyme/GlaxoSmithKline published a paper on the potent, selective PRMT5 inhibitor EPZ015666 (1.6) discovered in a high-throughput screening of a library containing 370,000 compounds. EPZ015666 presented a low nanomolar inhibition with selectivity against a broad panel of methyltransferases and was furthermore orally available. In a crystal structure the 1,2,3,4-tetrahydroisoquinoline-moiety (Tic) was shown to form π-π stacking interactions with the F327, characteristic for PRMT5, possibly explaining the Type II PRMT enzyme selectivity.77 Shortly after, researchers from Epizyme also discovered the PRMT6-selective inhibitor EPZ020411 (1.7) in a library screen.71 The compound presented an IC50 of 10 nM, good dose-dependent inhibition in a cellular system and promising bioavailability in rats. The diamine side chain was shown to occupy the putative binding site of the arginine substrate side chain.    It is interesting to note that the most potent and selective small-molecule PRMT inhibitors to date were all discovered in in vitro-based library screenings containing a large number of compounds (although Purandare et al. do not disclose the size of their library70), whereas virtual screenings have not yet proven very efficient in the discovery of potent and selective PRMT inhibitors.  Being the by-product of the methylation reaction, SAH has inherent affinity and inhibitory activity for the PRMTs and may act as a low micromolar competitive inhibitor11 (figure 1.8, box). SAM is a highly utilised co-factor in the human body, only second to ATP.29 The SAM-homologue sinefungin (1.8) shows inhibitory activity comparable to SAH, but is also non-selective for the PRMTs, targeting all SAM-based methyltransferases.11  A bisubstrate inhibitor is a compound that resembles both the co-factor and the substrate of a reaction in order to convey added selectivity and potency to the inhibitor.79   The groups of Ward and Martin have synthesised the bisubstrate inhibitors 1.9 and 1.10, which have the guanidinium-moieties of the substrate arginines linked to the adenosine-ribose units of the co-factor, with or without the amino-chain of methionine, respectively (figure 1.8).80,81 Both groups investigated whether 12  chain length affected inhibitory activity. Compounds 1.9a-c were only tested on PRMT1 and showed comparable low micromolar inhibiton,80 whereas 1.10a was shown to be more than 100-fold selective for PRMT4 over PRMT1 and PRMT6, and 1.10b  less than 2-fold.81 This suggests that PRMT4 accommodates shorter chains better, whereas chain length in PRMT1-inhibitors may not be important. The in situ inhibitor AAI (1.11) contains an iodoethyl chain connected to the nitrogen,82 which under biological conditions is thought to form a reactive aziridinium moiety susceptible to nucleophilic attack83 (figure 1.8 bottom).      Figure 1.8 Top: The co-factor SAM and byproduct/inhibitor SAH (box) and the non-specific inhibitor sinefungin (1.8).11 Bottom: Bisubstrate inhibitors (1.980 and 1.1081), the in situ inhibitor AAI (1.11)82 and the peptide inhibitors (1.12).84, 85 13  When incubated with H4 and PRMT1 AAI reacts with the H4R3 which generates a bisubstrate inhibitor that can inhibit PRMT1. This probe is therefore useful for identification of novel PRMT substrates.82 The same laboratory also synthesised 1.12a, a peptide based on the H4-tail sequence but with a chloroacetamidine moiety in place of the guanidinium on R3, thus producing a substrate-based rather than a cofactor-based inhibitor. This inhibitor can react with a nucleophile in the PRMT1 active site, and as with 1.11 the irreversible nature of the inhibitor makes it applicable in proteomics settings.84 However, as this sequence was only 4-fold selective for PRMT1 over PRMT4, the authors used a combinatorial approach to discover 1.12b, which differs by only one residue, but shows a 17-fold selectivity PRMT1/PRMT4.85 The group has also conjugated fluorescein and biotin to the N-terminal of 1.12a thereby producing probes to be used in biological studies of the PRMTs.86    Thus, several highly potent and selective small molecule inhibitors of PRMTs 3-6 exist, some of which might have the potential to advance to drug development. On the contrary, the bisubstrate approach has not yet resulted in any significant inhibitors. And as PRMT1 is the most utilised PRMT in human cells the need for good PRMT1 selective inhibitors is evident. Our PRMT inhibitor discovery process started with phage displays, which will be covered in the next section.   1.6 Phage Display Invented in 1985,87 phage display is a powerful technique which utilises bacteriophages to generate libraries of peptides and proteins of up to 1012 in size.88 A peptide- or protein-encoding DNA sequence is inserted into the gene coding for the coat protein of the phage, resulting in the encoded peptide/protein being expressed and presented on the surface of the phage. This provides an easy way of linking the genetic information in a library of DNA to the desired library of peptides/proteins. The phage library can easily be screened in vitro against a target in a so-called ‘panning’ process; several rounds of binding to the target, washing, elution and amplification of the binding phages will result in identification of the best binders88 (figure 1.9).  The targets can vary from antibodies, receptors and other proteins to whole cell preparations.89,90 Due to the simplicity of library generation and screening phage display is now regarded as a strong biological combinatorial method.91   The hits from a phage display are often shown to bind at a biologically relevant site on the target, often in or near the active site where the side chains of the residues protrude from the protein structure in order to interact. This may disrupt the normal conformation of the target and the peptide hits therefore likely function as inhibitors.89 As the active site is often found buried in an enzyme, this region is characterised by 14  a high proportion of aromatic residues. Also, the identified peptides may resemble the interacting residues on the target, something that has been termed ‘convergent evolution’.92  Due to the vigourous washing during panning it has been estimated that the hits must at least bind with a dissociation constant of 50 μM or lower.89 However, testing the peptides off the phages often results in lower binding affinities. The phages used for peptide discovery each display 3-5 peptide copies. This induces a better interaction with the target; an up to 45-fold increase in binding affinity has been seen when comparing a tetrameric presentation to the monomeric peptide.90   Despite this drawback, phage display is a simple yet strong tool for identification of peptide leads, which may be further developed to yield potent binders or even inhibitors of the target in question.     Figure 1.9 The panning process in phage display; A phage library is generated from a DNA library and screened against a target, the non-binding phages are washed away and the eluted phages are amplified and the cycle repeated in several rounds of panning. Finally the best binders are isolated.    15  Chapter 2: Peptides and Peptidomimetics 2.1 Introduction: Peptides in Nature  Peptides are found in all living systems and play vital roles in most biological processes.93-95 They are short chains of up to 50 amino acids in length – longer chains are by convention referred to as proteins.96 In the human body peptides primarily act as hormones such as bradykinin, a nonapeptide which dilates blood vessels and ghrelin, which regulates hunger signals, or as neurotransmitters, such as the enkephalin opioids which modulate pain perception. However, peptides can also function as growth factors or enzyme inhibitors/substrates etc.93 In Nature peptides and proteins are synthesised in the ribosomes of our cells by connecting the amino acid building blocks using peptide bonds, also known as amide bonds.97      Scheme 2.1: Equilibrium of amide bond formation and hydrolysis. Grey box shows resonance stabilisation through orbital overlap.   2.2 Stability and Reactivity of Amides Chemically amide bonds are formed in an exothermic reaction when a carboxylic acid reacts with an amine expelling water (scheme 2.1). To overcome the high activation barrier of the reaction an input of energy is required, which in Nature is circumvented through the use of enzymes.97 In laboratory peptide synthesis the coupling requires a coupling reagent to activate the carboxylic acid. Thermodynamically the hydrolysis is favoured by a factor of 103-104, however, this reaction is kinetically very slow and amide bonds are therefore highly stable at physiological pH and temperature on physiologically relevant time scales.96 The high kinetic stability and thus low reactivity of the amide bond is due to resonance stabilisation; the nitrogen lone pair in the p-orbital is perfectly aligned to overlap with the π*-orbital of the carbonyl, pushing the carbonyl electrons onto the oxygen and thus creating a partial double bond character of the C-N bond (Scheme 2.1). The high degree of delocalisation due to resonance therefore results in amides being neither nucleophilic nor basic. The double bond character furthermore restricts rotation around the bond making the amide planar, thus assuming either trans or cis double bond conformation. Due to the steric clash of the R and R’ groups extending on either side of the amide in longer peptide chains the trans conformation 16  is typically preferred.96 However, in strained amide bonds, such as the one involving a random amino acid and proline (Xxx-Pro), both conformations are strained and equally disfavoured, and thus both trans and cis isomers exist.  2.3 Peptides in Drug Discovery The great stability of amides makes them excellent functional groups in the chemical synthesis of small molecule compounds. In peptide drugs, however, the amide bond is the Achilles Heel, as proteases in the gut and blood plasma will recognise the characteristic structure of a peptide backbone and cleave it.  Peptides thus exhibit low oral bioavailability and fast excretion, and were therefore previously considered poor drug candidates.93,94,98,99 In addition, low lipophilicity and thus low tissue penetration as well as a large number of rotatable bonds (vide infra) also precluded their use as therapeutics.93,99 In the last few decades, however, peptides have had a revival in the medicinal industry, as they possess several advantageous properties to other drug candidates.99 Compared with the much larger proteins and antibodies, peptides are able to cross biological barriers more easily, they show a decreased risk of immunogenic reactions and have greater storage stability. Contrary to small molecule metabolites the amino acid break-down products of peptides are non-toxic and the short half-life of peptides reduces accumulation in tissues.99 Thus, by improving the proteolytic stability and the physico-chemical properties while preserving or even enhancing the biological potency and selectivity peptides may be further developed into the ideal drug candidates.  2.4 What are Peptidomimetics? A peptide can be chemically modified to improve its properties. Based on the degree of the modifications to the original peptide structure several definitions exist. Modified peptides contain only minor changes to the original structure while maintaining the amide bonds, whereas pseudopeptides have only partial peptide character due to alterations of the backbone. ‘Peptide mimetics’ or ‘peptidomimetics’ are terms used broadly – from defining any modification and thus also including modified peptides or pseudopeptides to a stricter definition in which the mimetics contain no peptide bonds.100 Nonetheless the widely accepted classification by Ripka and Rich divides peptidomimetics into three types:101 a) Type I or structural mimetics, which have a structure resembling the original peptide; b) Type II or functional mimetics, which interacts with the target in a similar fashion to the original peptide but without structural analogies; and c) Type III or functional-structural mimetics, which has a scaffold structurally different from the original but with all the functional groups necessary for the optimal interaction with the target. Thus, type III mimetics are the perfect drug candidates, as they contain all relevant functional groups for optimal efficacy and selectivity 17  towards the target while also possessing proteolytic stability due to ‘disguising’ of the original peptide structure. Henceforth for simplicity the term ‘peptidomimetic’ will be used about any of the aforementioned modifications to a peptide sequence.  2.5 Optimising Peptides into Peptidomimetics: Using the Peptidomimetic Toolbox When a biologically active peptide has been identified the first step in de novo peptidomimetic design is to identify the key amino acid residues and the spatial arrangement of these for optimal target binding. This is usually done by truncating the sequence both C- and N-terminally to determine the minimal active sequence of the peptide. Alanine scans are employed to identify key amino acids for receptor recognition and D-amino acid scans determine the importance of chirality.98  With the minimal active sequence in hand and the knowledge of which residues are essential for target binding the next step is to optimise the potency and selectivity towards the target and in addition increase proteolytic stability. Thus far a diverse range of peptide modifications have been developed allowing us to test a variety of backbone modifications as well as local and global cyclisation strategies, as can be seen in the peptidomimetic toolbox in figure 2.1.93,99   In retro-inverso-peptide mimetics the carbonyl and NH-functionalities in the backbone are interchanged. In contrast to these are peptoids, in which the side chain has been moved from the Cα to the NH-functionality thereby also removing chirality and thus avoiding the issue of racemisation. Replacement of the various functionalities in the backbone can lead to a multitude of isosteres with great similarity to the original peptide sequence; the carbonyl moiety may be exchanged for a thiocarbonyl or a sulfone functionality, or even deleted to yield a methylene amide; replacing the nitrogen for an oxygen or a sulphur atom yields a depsipeptide or a thiodepsipeptide, respectively; and changing the Cα for N will result in azapeptides etc.93,94,102 to name a few.  18    Figure 2.1: The Peptidomimetic Toolbox: Representation of the large variety of peptide backbone modifications. Grey boxes represent strategies employed in this thesis (see chapter 4) 93 99   The simplest backbone elongation is the addition of an extra methylene unit either between the nitrogen and the side chain (β2-homo amino acid) or the carbonyl and the side chain (β3-homo amino acid), however many others exist. Several peptide backbone extensions have shown a propensity to adopt ‘protein-like’ secondary structures through intramolecular hydrogen-bonding.93  Alkylating the nitrogen in the backbone results in an altered cis/trans ratio of the amide bond, as the alkyl-group can cause increased steric hindrance and force the two peptide chains onto the same side of the double bond (Z-conformation). Alkylation furthermore removes a hydrogen-bonding donor.103  As mentioned previously D-amino acid scans are used to investigate the importance of chirality.  Using conformationally constrained cyclised amino acids or larger building blocks limits the rotation locally in the peptidomimetic and may thereby impose an altered secondary structure (vide infra).98,104 Global cyclisation either between the two termini (head-to-tail), a side chain and a terminus (backbone-to-side chain) or two side chains (side chain-to-side chain) using a variety of cyclisation strategies,105 will like the 19  local cyclisation analogues result in reduced conformational flexibility. Moreover, using global restrictions also enables definition of how the side chain functionalities are presented.94  Ergo, most of these modifications will change the hydrogen-bonding properties either intramolecularly or in relation to the target. Furthermore, the conformation of the peptide may also be affected either through the altered hydrogen-bonding, cis/trans isomerism of the amide bond or local/global restrictions. Thus, these changes will also affect the potency and selectivity of the peptidomimetics to their targets. Finally, incorporation of all of the above mentioned modifications will likely result in an increased proteolytic stability as the normal peptide backbone pattern is disrupted making it difficult for proteases to recognise and degrade the drug.  2.6 Conformational Constraints in the Peptide        A peptide has several torsional angles present in the backbone (φ, ψ and ω) and side chains (χ1, χ2 etc.) (figure 2.2). As mentioned previously, the amide bond angle, ω, is usually 180o and in the trans conformation, except for the amide bond in Xxx-Pro which can also be 0o (cis). The other two backbone angles could in principle assume any angle. However, Ramachandran evaluated the possible conformations of the two and found that certain combinations of φ and ψ angles were not permitted, whereas some were preferred and would induce a specific secondary structure in the backbone such as an α-helix or a β-sheet.106      Figure 2.2: Definitions of the torsional angles ϕ, ψ, ω, and χ  Chi-space or χ-space refers to the space occupied by the amino acid side chains, and thus the χ-torsional angles are important for receptor recognition.  Each side chain χ-angle can assume three low energy staggered conformations also known as rotamers; gauche(-) with χ = - 60o; gauche(+) with χ = + 60o and anti with χ = 180o (figure 2.3, top, the angle is determined from the Nα to the side chain). The energy difference between these three rotamers is low and generally all conformations are accessible.98 20  Due to its large number of rotatable bonds a peptide has high flexibility and may adopt numerous conformations. It is therefore difficult to predict which of these conformations interact with the target and thus is the bioactive conformation - it is not necessarily the minimum energy conformation of the peptide. Moreover, the conformation of a peptide also depends on the surrounding environment and this could potentially lead to side effects if the favoured peptide conformation can bind to the wrong receptors.95 Hence, incorporation of conformational constraints and restricting several torsional angles in the peptide could improve target selectivity. Limiting the degrees of freedom also reduces the entropic cost of binding to the target.     Figure 2.3: A variety of conformational constraints. Top: Newman projections of the three staggered low energy conformations of L-amino acids. Bottom: Side chain constraints on L-phenylalanine resulting in selected conformations for Tic (pipecolic acid derivative), Atc (tetralin derivative) and Aba (benzazepinone derivative)  Much evidence suggests that it is mainly the side chain groups of the peptides that induce the binding to the target and trigger the biological response,107 and exchange or modification of a few residues may lead to loss of activity or even switching of bioactivity from agonism to antagonism.95  21  Constraining the χ angles allows us to investigate the conformational space of the peptidomimetic. A multitude of ring-constrained amino acid analogues with various connectivity, substitutions and ring sizes exist,98,104,108 but the focus here will be on 6- and 7-membered rings based on aromatic amino acids.  Figure 2.3 illustrates how constraining phenylalanine into the Tic (1,2,3,4-tetrahydroisoquinoline-3-carboxy moiety), Atc (2-aminotetralin-2-carboxy moiety) and Aba (4-aminobenzazepinone) analogues in each case excludes one of the three rotamer conformers. Connecting an ortho-carbon of the phenyl ring to the Nα in phenylalanine using a methylene bridge yields the pipecolic acid analogue Tic, constraining the χ and φ angles, which only allows the two gauche conformations. Using an ethylene bridge to connect the ortho-carbon to Cα and constrain the χ angles results in Atc, which excludes the gauche(+) conformation. The benzazepinone Aba is produced when a methylene bridge connects the ortho-carbon with the Nα+1, that is the nitrogen of the next amino acid, thus constraining the ψ, ω and χ angles and excluding the gauche(-) conformation. Evaluating how each of these modifications affects the target in question will provide structural information of the preferred binding conformation.  Thus, relating structural information from biophysical studies of the constrained peptidomimetic (using e.g. NMR) to the biological effects seen in assays or cellular studies can give us valuable information of not only the peptidomimetic, but also the target and their interactions.109,110   The Aba scaffold was first synthesised by Flynn et al.111 in 1987 and the concept of using azepinone-scaffolds has since been expanded to include a tyrosine (Hba),112 a tryptophan (Aia)113 and a histidine-like analogue (Ata)114 (figure 2.4). Ata, however, contains a triazole rather than the imidazole of a natural histidine. Due to the basic properties of the imidazole, synthesis of the true azepinone-mimic of histidine has proven difficult, and work is therefore still in progress.          Figure 2.4: The azepinone scaffolds  22  The pipecolic acids and azepinones have over the years been used in a large variety of peptidomimetics of natural peptide sequences, of which a few will be highlighted here.  In the development of opioid ligands Tic has been successfully used to produce both potent antagonists and agonists with sub-nanomolar binding affinities,110,115,116 and ligands containing Aba have proven to be μ-receptor subtype selective.117,118 Both Tic and Aba scaffolds have been utilised in bradykinin analogues,119-121 and has resulted in a marketed drug.119 Aba has been used in the development of selective human melanocortin receptor-3 (hMCR-3) ligands122 and recently Aia was used in potent hMCR-4 agonists.123 Aia has furthermore been applied in the development of the most potent and selective sst5-somatostatin mimetics to date.124    We are currently writing a review manuscript based on case stories in which the 5-, 6- and 7-membered ring-constrained amino acids have been used to modify biologically relevant peptides. The cases on angiotensin IV, bradykinin, farnesyl transferase and neurokinin 1 (NK-1) were written by this author and are being finalised for publication as soon as possible. The four cases (as of 15th of December 2015) can be found in Appendix 2 of this thesis.   23  Chapter 3: Research Objectives of the PhD Project Based on a phage display screening several lead peptides were identified as binders of the PRMTs.  The overall aim of this project was to identify peptide and peptidomimetic binders/inhibitors for the PRMTs. In order to achieve this goal we specifically aimed to:  1) Identify potent binders/inhibitors of the PRMTs based on the chosen lead peptide. For this we decided to employ various peptidomimetic strategies to modify the backbone and side chain flexibility of tryptophan residues within the lead sequence.  2) Express PRMT proteins for use in biophysical characterisation of the binders/inhibitors. As we did not know where the peptides and peptidomimetics would bind on the targets we intended to investigate the binding affinities of these by using isothermal titration calorimetry.  3) Characterise the most interesting binders/inhibitors further. Through collaborations we intended to perform further studies to elucidate the putative inhibition potencies, the solution structures, as well as cellular targeting of the binders/inhibitors.   4) Develop and synthesise a dipeptide mimetic for the basic amino acids lysine and arginine for use in future peptidomimetic studies.    24  Chapter 4: Developing PRMT Peptidomimetic Inhibitors from Phage Display Leads  4.1 The Hits from Phage Display This project was initiated as a collaboration between University of British Columbia (UBC) and University of Copenhagen (UCPH). The initial phage display screening to find lead compounds for the epigenetic enzymes was performed by Dr. Brian Lohse. Three peptide libraries were screened in search of binders/inhibitors of the PRMTs; a 7-mer and a 12-mer linear peptide library and a cyclic 12-mer library all purchased from New England Biolabs, Inc.125  Table 4.1 shows the identified hits from the phage display.   Table 4.1 Identified hits from the phage display performed by Dr. Brian Lohse  Name Sequence Target BL161 FPTLYGPVWGTM PRMT4 BL162 VKLSDDLMPGMV PRMT1 BL163 VKLYDDLMPGMV PRMT1 BL165 (P1) EFWDWGP PRMT2 BL166 SGLDWWETWINS PRMT2 BL167 QAAWVFPSQGYI PRMT3 BL168 YFDGDFAYWSPP PRMT3 BL169-C FAPSWSFNWKPT PRMT5 BL169-B TAGGLPKKRIAR PRMT9 BL169-7&A GYPAQKRSKPRP PRMT9 BL169-41&50 RSPSPKSKYRRL PRMT9 BL169-38&43 RSPSPKSKYGRL PRMT9   We were quite intrigued by the BL165 hit (from here on known as P1), as it was the only 7-mer peptide found to bind to the PRMTs and three out of the 7 residues are aromatic. Phage display can identify hits of variable length depending on the library screened, but often only 5-8 amino acids are involved in the specific interaction with the target.126 We therefore suspected that the full P1 sequence might be important for target interaction. Furthermore, as aromatic residues are frequently involved in protein-protein binding interactions127-130 and have been employed extensively in peptidomimetics104,107,110,114,115,117, 120,122-124,131  we chose P1 as our lead peptide for this study.       25  4.2 Aim of the Peptidomimetic Project Based on our lead P1 we aimed to  a) Synthesise peptidomimetics with modifications related to the tryptophans in the lead sequence.    b) Screen these peptidomimetics for inhibitory activity against available PRMT enzymes. c) Characterise the identified inhibitor(s) further; by using NMR, CD and through cellular studies.  4.3 Using Peptidomimetics in the Search for PRMT Inhibitors 4.3.1 Peptidomimetics based on P1 Due to the rotational flexibility in peptide chains the natural peptide hits are often identified as low affinity ligands, which can be optimised by exploring various methods to control flexibility. We thus aimed to investigate the effects of three peptidomimetic modifications to the tryptophan residues in P1 (scheme 4.1); backbone elongation through the use of β-amino acids (4.1), alteration of backbone conformation via amide N-methylation (4.2) and restricting conformational space of the side chains by incorporating azepinones (4.3; section 2.6).     Scheme 4.1 Modifications of the tryptophans employed in this thesis; backbone elongation, N-methylation, and incorporation of Aia-building blocks   Table 4.2 shows the identified peptide P1 and the 12 peptidomimetics produced by applying these three different modifications to P1. P2 and P3 have each had one of the tryptophans exchanged for a β3-26  tryptophan and P4 has both tryptophans exchanged. P5 and P6 contain a tertiary amide due to N-methylation of the amide bond in between a tryptophan and the succeeding amino acid, aspartic acid and glycine, respectively. P7 contain both of these N-methylations. P8-P10 have either one or both tryptophans exchanged for the L-Aia-building blocks, which are connected to the N-terminal of the succeeding amino acid (vide infra). This is similarly seen in P11-13 which contain D-Aia.      Table 4.2 P1 and the derived peptidomimetics P2-P13 P1 Ac-EFWDWGP-NH2 P2 Ac-EFWD(β3W)GP-NH2 P3 Ac-EF(β3W)DWGP-NH2 P4 Ac-EF(β3W)D(β3W)GP-NH2 P5 Ac-EFWDW(NMeG)P-NH2 P6 Ac-EFW(NMeD)WGP-NH2 P7 Ac-EFW(NMeD)W(NMeG)P-NH2 P8 Ac-EFWD(L-Aia)GP-NH2 P9 Ac-EF(L-Aia)DWGP-NH2 P10 Ac-EF(L-Aia)D(L-Aia)GP-NH2 P11 Ac-EFWD(D-Aia)GP-NH2 P12 Ac-EF(D-Aia)DWGP-NH2 P13 Ac-EF(D-Aia)D(D-Aia)GP-NH2   P1 and all of the derived peptidomimetics, P2-13, were synthesised using standard Fmoc-based solid phase peptide synthesis (SPPS; see experimental section) except when modifications were incorporated. In these cases the procedures are explained in detail in the relevant section (4.3.2, 4.3.3 or 4.3.4). In order to remove the terminal charges and mimic the backbone amide bonds in peptides the C- and N-terminals of all the peptides were amidated and acetylated, respectively. Rink amide MBHA resin was used for all solid phase peptide syntheses (SPPS), which yields an amidated C-terminal when the peptide is cleaved from the resin. Following coupling of the final amino acid the N-terminal was acetylated on resin prior to final cleavage.  4.3.2 β-amino acids In β-amino acids the amino group is bound to the β –carbon of the side chain instead of the α-carbon, thus elongating the peptide backbone by one methylene unit. The simplest β-amino acids are the mono-substituted β2- and β3-amino acids, which have the side chain next to the carbonyl or the amine functionalities, respectively (figure 4.1).    27   Figure 4.1 The general β2 and β3-amino acids as well as the β3-homotryptophan (β3-hTrp) used in this study  When binding to a target peptides may assume a secondary bioactive conformation, such as a helix. Peptidomimetics consisting of β-amino acids have been shown to induce secondary structures in both organic and aqueous solution, most notably helice or turn conformations. These secondary structures are stabilised by intramolecular hydrogen bonds (H-bonds) to various degrees and interestingly, the 14-helix conformation of a β-peptidomimetic has even proved to be more stable than the α-peptide helical counterpart.132,133 These secondary stabilisations may be seen in very short β-amino acid containing peptides of just 4 residues in length.132 The implementation of β-amino acids into peptides has in addition shown increased stability against proteolysis as mentioned previously.132,133  By exchanging a tryptophan residue for the β3-homotryptophan (β3-hTrp) we aimed to investigate the effect of the smallest possible backbone modification of P1 on PRMT binding/inhibition. By adding this extra methylene unit an additional rotatable bond is introduced which may thus increase the degrees of rotational freedom for the peptide, but on account of the altered hydrogen-bonding pattern, this could also lead to less conformational flexibility,132 and a more stable structure. Furthermore, as we do not know where the peptide will bind, the shift of the side chain indole could potentially furnish a better interaction with the target region in the PRMTs and thus lead to better binding/inhibition.  We chose to incorporate the β3-homotryptophan (4.4) rather than β2 simply due to the availability of the two. The availability most likely corresponds to ease of synthesis; β3-amino acids can be produced easily with retention of stereochemistry134 via the Arndt-Eistert homologation,133 whereas β2-amino acids are more difficult to synthesise. One way of synthesising β2-amino acids is the reaction of a carbon nucleophile with an asymmetric aziridine (see chapter 5).   The incorporation of β3-amino acids into P2-4 was performed using standard Fmoc-based SPPS, however only 2 equivalents (eq.) of Fmoc-β3-hTrp-OH were utilised for each coupling (normally 3 or 5 eq. were used, see experimental section). Yields of peptides containing natural amino acids and β3-amino acids were in the 25-50% range after preparative reverse phase HPLC (pHPLC).   28  4.3.3 Nα-methylation  Nα-methylated peptides have demonstrated a variety of properties which are advantageous in drug development, such as higher enzymatic stability and receptor subtype selectivity, as well as oral bioavailability and cell permeability due to the increased lipophilicity.103 This has so far been exploited to develop ligands for several natural peptide targets.135 Conformationally, N-methylation results in a tertiary amide in the peptide backbone and thus the removal of a hydrogen bonding donor as well as an altered cis/trans ratio of the amide bond involved. For homochiral peptides the imposed steric hindrance about the amide bond in question has been shown to result in a higher proportion of cis conformers which may alter the entire shape of the peptide possibly further aided by alteration in the hydrogen bonding motifs.103,136  Introduction of Nα-methylation in peptides can be done by preparing the protected Nα-methylated amino acids either in solution137 or on resin138 with stereochemical control. This would be advantageous if larger quantities of N-methylated amino acids would be needed for e.g. N-methylation scans. However, in our case we only needed a few select N-methylations and therefore this was done directly on resin following an optimised 3-step literature procedure139 (scheme 4.2).        Scheme 4.2 Three step procedure for N-methylation on resin.  One of the protons on the free N-terminal amine (4.5) may be protected with a 2-nitrobenzenesulfonyl (2-Ns) by reaction with 2-nitrobenzenesulfonyl chloride under basic conditions yielding 4.6. Subsequent deprotonation using DBU and methylation with the strong alkylating agent dimethyl sulphate renders the methylated amine (4.7), which may be deprotected using a thiol under basic conditions resulting in 4.8. Biron et al. used NMP as a solvent for all three steps and completed these in only 35 minutes (min).139       29  4.3.3.1 Synthesis of N-methylated peptidomimetics P5-P7 Synthesising P5-P7 on resin requires methylation of the N-terminal of either glycine (G) or aspartic acid (D). In both cases the N-methylation was done by performing the three step methylation procedure (scheme 4.2) twice followed by a test-cleavage which was analysed using liquid chromatography mass spectrometry (LC-MS) to ensure full N-terminal methylation (4.8).    Scheme 4.3 N-methylation and coupling of Fmoc-Trp(Boc)-OH in P5-P7.  Coupling of tryptophan to the free secondary amine on resin proved difficult but seemed to depend on the side chain of the preceding amino acid (4.9, scheme 4.3) and several coupling reagents were tested. Coupling to the N-methylated glycine (P5 and P7) was attempted twice using 3 eq. HBTU at room temperature (rt) for one hour (hr) each, but only resulted in full conversion after coupling with 5 eq. PyBOP also at rt for one hr. The coupling of tryptophan to a secondary amine succeeded by an aspartic acid side chain (P6 and P7) proved even more difficult; this was attempted with 3 eq. HBTU, 3 eq. HATU and 5 eq. PyBOP at rt for up to two hr leading to only 50-80% completion. Coupling using the Biotage Wave+ microwave was performed thrice using 5 eq. HBTU in DMF at 75 oC leading to roughly 80-90% coupling efficiency. This was deemed a sufficient conversion, since each coupling using the microwave generated traces of several side products.  Synthesis of the peptidomimetics was completed at rt using standard Fmoc-SPPS. After N-terminal acetylation and final cleavage the peptidomimetics were purified using pHPLC and resulted in yields of 12-45%.     30  4.3.4 The Aia scaffold As mentioned in section 2.6 the 4-amino-1,2,4,5-tetraindolo[2,3-c]azepin-3-one (Aia) scaffold (4.10) is the constrained azepinone of tryptophan, which has a methylene bridge from the C2 of the side chain indole to the Nα-of the succeeding amino acid (in blue, scheme 4.4), thus producing a rigid tryptophan mimetic controlled in chi-space. See section 2.6 and appendix 2 for other peptidomimetic applications.    Scheme 4.4 Retrosynthetic route towards the Aia-building block (4.10).  As the retro synthetic route indicates (scheme 4.4), Aia (4.10) may be formed by reductive amination of an amine with an Nα-protected tryptophan that has been formylated on the C2 of the indole (4.12) followed by an intramolecular coupling of the secondary amine with the carboxylic acid moiety in 4.11. The formyl-derivative 4.12 may be synthesised from tryptophan (4.13) in three steps with retention of stereochemistry (scheme 4.5).    4.3.4.1 Synthesis of Fmoc-2’-formyl-Trp for Aia synthesis Both enantiomers of Fmoc-2’-formyl-Trp (4.12a and 4.12b) were synthesised following literature procedures as indicated in scheme 4.5.140,141 The first step is a Pictet-Spengler reaction which uses formaldehyde to convert tryptophan (4.13) into the pipecolic acid derivative Tcc (4.14). The secondary amine is then Fmoc-protected to yield Fmoc-Tcc (4.15), which can be oxidised using selenium dioxide (SeO2) to yield Fmoc-2’-formyl-Trp (4.12).  Scheme 4.5 The three-step synthesis of the Fmoc-2’-formyl-Trp building block (4.12) to be used in Aia synthesis.*marks the stereocenter. 31  The putative mechanism of the SeO2 oxidation is shown in scheme 4.6. In a [4+2] cycloaddition SeO2 is added to the Fmoc-Tcc (4.16), and after undergoing a [2,3] sigmatropic rearrangement (4.17) the unstable intermediate (4.18) decomposes to yield alcohol 4.19 which reacts further to yield the formylated building block (4.12).     Scheme 4.6 The putative mechanism of oxidation of Fmoc-Tcc (4.15) using SeO2 to yield Fmoc-2’-formyl-Trp (4.12).  The three steps proceed with retention of stereochemistry, and thus both the L- and D-tryptophan (4.13a and 4.13b, respectively) were reacted yielding the L- and D-enantiomers of Fmoc-2’-formyl-tryptophan (4.12a and 4.12b) in 56% and 53% yield over three steps, respectively. The intermediates Tcc (4.14a and 4.14b) and Fmoc-Tcc (4.15a and 4.15b) were used in the following step without purification. Both Fmoc-2’-formyl-Trp building blocks (4.12a and 4.12b) were analysed by chiral column HPLC and found to have an enantiomeric excess (ee) of >97%. As incorporation of Aia into peptides using Fmoc-based SPPS in some cases resulted in incomplete reactions and messy crudes according to literature,140 the Boc-protected building blocks Boc-L-2’-formyl-Trp and Boc-D-2’-formyl-Trp were synthesised using the same methodology shown in scheme 4.5 (Fmoc substituted for Boc) in 55% yield and 59% yield over three steps, respectively. However, Boc-based SPPS requires the highly corrosive hydrogen fluoride (HF) for final cleavage and therefore the Fmoc-based synthesis was employed first. As the crude peptidomimetics from the Fmoc-based SPPS were of sufficient purity the Boc-protected building blocks were not employed in this project.         32  4.3.4.2 General on-resin synthesis of Aia The formylated building block (4.12) may be reacted with the free N-terminal of the peptide on resin (4.20) using a mild reducing agent, NaCNBH3, in a reductive amination to yield 4.21 (scheme 4.7).140 Once on resin the intramolecular cyclisation to facilitate the 7-membered azepinone ring may be done in a coupling of the newly formed secondary amine (in blue) with the free carboxylic acid (in red) using a coupling reagent such as HBTU thus rendering 4.22. Normal Fmoc-deprotection to produce 4.23 provides a free N-terminal where the peptide chain can be further extended.        Scheme 4.7 General on-resin synthesis of Aia.* marks the stereocenter in 4.12.  Several side reactions are possible in the on-resin Aia synthesis. A by-product resulting from double reductive amination could be produced (4.24, Scheme 4.8); that is two formylated building blocks (4.12) react with one amine on resin creating the tertiary amine product (4.24), which after Fmoc-deprotection may facilitate the generation of several additional by-products due to the free amine and carboxylic acid functionalities (shown in blue and red, respectively 4.25). On the other hand, if reductive amination does not go to completion and leaves a free primary amine on a peptide chain tethered to the same resin (4.26) this primary amine may in the subsequent coupling step react with the carboxylic acid on a neighbouring peptide chain thus producing by-product 4.27.140 The ensuing Fmoc-deprotection yields a product containing a primary and a secondary amine (4.28, amines coloured in blue) which will be points of extension for peptide chains.      33    Scheme 4.8 Side products observed in the synthesis of Aia.  These side reactions may be avoided by synthesising Aia-containing dipeptides (of the type Fmoc-Aia-Xxx-OH) in solution using a similar methodology to on-resin synthesis and coupling these onto the peptide by standard Fmoc-based SPPS. There are, however, several drawbacks to solution-phase synthesis of the Aia-dipeptide mimetics. Firstly, the C-terminal amino acid requires ester protection (Xxx-OR) during synthesis, and a final saponification step is required before coupling of the dipeptide on resin.  Secondly, purification of intermediates and the final dipeptide is needed. And thirdly, for each dipeptide combination of Aia and a different succeeding amino acid (Fmoc-Aia-Xxx-OR) this three-step synthesis followed by saponification and relevant purification(s) must be performed. Nonetheless solution-phase synthesis of Fmoc-Aia-Xxx-OH dipeptides provides an alternative synthesis strategy in cases where on-resin synthesis fails.  4.3.4.3 On-resin synthesis of Aia in Peptides P8-P13 The Aia-containing peptides in this thesis were all synthesised with the on-resin synthesis approach. Three colour tests were employed in an attempt to monitor reaction progress for the Aia synthesis. All three tests can be performed with the peptide on resin; the Kaiser test detects non-sterically hindered primary amines,142 the chloranil test is used specifically to detect secondary amines,143 but will also detect primary amines, and the TNBS (trinitrobenzene sulfonic acid) test is used specifically to detect primary 34  amines.144 As the reductive amination converts a primary amine to a secondary, and the coupling step converts the secondary amine to a tertiary amide we thought that these tests collectively would provide easy reaction progress monitoring. However, none of these tests produced consistent results in the reductive amination step, in the following coupling or even in the normal couplings following incorporation of Aia. Therefore test cleavages analysed by LC-MS were employed to monitor reaction progress in all steps of the Aia synthesis.  The reductive amination of the building block 4.12 with the resin-linked peptide was performed with a 2-fold excess of building block 4.12 and 4-fold excess reducing agent, NaCNBH3, at rt. The nature of the succeeding amino acid side chain seemed to be critical for the reaction rate; when the succeeding amino acid was a glycine (R1 = H in 4.20) the reductive amination was done in 15-45 min, whereas having a protected aspartic acid (R1 = CH2COOtBu in 4.20) as the neighbouring side chain the reaction time varied from 4-24 hr. In slow reactions another 2 eq. of NaCNBH3 was added. For all reactions involving Aia both side reactions mentioned in section 4.3.4.2 were observed. As 4.24 was mostly found only as trace, the reductive aminations were run until no starting material could be detected to minimise the production of side products originating from 4.28. It had previously been found that the unreacted building block 4.12 could not be washed away using the standard washing procedure possibly due to strong adsorption to the resin polymer, and that conversion of 4.12 to a semicarbazone by reaction with a semicarbazide would enable removal of 4.12 from resin.140 After completion of the reductive amination the resin was therefore treated twice with semicarbazide with subsequent washing. However, the formylated building block 4.12 could at times still be detected after test cleavages in the ensuing synthesis. If the remaining adsorbed 4.12 desorbs from the polymer resin in the following coupling steps this could potentially also result in additional side products.    The intramolecular coupling of the secondary amine with the carboxylic acid in 4.21 to yield 4.22 was tested thrice at rt, each time requiring 4-5 hr for completion. Microwave irradiation on the other hand usually required 2-3 couplings of 5 min at 75 oC using fresh reagents each time. Performing a test cleavage after 4 hr of coupling of Aia at rt during the synthesis of P8 (Ac-EFWD(Aia)GP-NH2) the major m/z detected was 590 and not the expected m/z of 592 (4.29a, scheme 4.9). We suspect that an elimination product, 4.30, is formed during longer periods of reaction, which could be favoured due to the extended conjugation in this product. Performing the coupling using microwave synthesis the resultant test cleavage indicated a peak with an m/z of 592 as the major product (4.29a) and only trace amounts of the peak with m/z 590 (4.30a). 35    Scheme 4.9 Product 4.33a was a putative major by-product when performing intramolecular Aia-coupling at rt in the synthesis of P8.  The only full peptidomimetic synthesised by both microwave Aia-coupling and Aia-coupling at rt was P9.  Generally comparing microwave coupling with coupling at rt for P9 it seemed that both produced by-products to the same extent. Considering the increased reaction rates and by-product formation microwave synthesis seemed superior to us. All couplings to form the azepinone-ring were therefore performed using microwave synthesis. Reactions were performed with 5 eq. HBTU and DIPEA for 5 min at 75 oC and reaction progress was checked by test cleavage. Generally, the synthesis of the azepinone ring required 2-3 couplings each time using fresh reagents. Yields of peptidomimetics containing a single Aia were in the 10-44% range, whereas peptidomimetics containing two Aias were isolated in 2-23% yield.    36  4.4 Evaluating the Peptidomimetics as Inhibitors of the PRMTs 4.4.1 Inhibitor Screenings using an Antibody-Based Assay All peptidomimetics (P2-P13) as well as the lead P1 were screened using a chemiluminescent antibody assay. The assay utilises the PRMT enzyme of interest to catalyse the transfer of a methyl group from the SAM co-factor onto an arginine in a histone tail substrate. By using a primary antibody highly selective for a specific arginine in the substrate which can bind a secondary antibody with a chemiluminescent probe attached, the methylated arginine can be detected.81 When an inhibitor is added a lowering in the binding of antibody is seen and thus a lowering in the chemiluminescent signal. PRMTs 1, 3, 4, 5, 6 and 8 were screened in this assay, as these were available at BPS Bioscience, the company which screened the inhibitors. PRMTs 2, 7 and 9 were not screened. PRMT2 possesses only very low enzymatic activity145 and is also seen in heterodimeric complexes with PRMT1.146 PRMT7 is a Type III PRMT, thus producing only MMA.147 PRMT9 has only recently been broadly accepted as a PRMT22 and was not available for screening.   For all assays histone substrates were used; for PRMTs 1, 3, 5, 6 and 8 residues 1-21 of the histone tail 4 substrate were used (H4, 1-21), and for PRMT4 the substrate was the histone 3 tail, residues 1-27 with an acetylated lysine in position 18 (H3, 1-27(acK18)). Kinetic studies have shown H4 1-21 as the minimal peptide substrate which has kinetics comparable to full length H4.44 The highly selective antibody detects dimethylated arginine on the histone substrate; for PRMTs 1, 3, 5, 6 and 8 assays the antibody recognises dimethylation on arginine 3 of the histone tail 4 substrate (H4R3me2), whereas the antibody used in the PRMT4 assay recognises dimethylation on arginine 26 of the histone tail 3 substrate (H3R26me2).   4.4.1.1 Inhibitor Screening of P1 mimetics The peptide (P1) and the derived peptidomimetics (P2-P13) were screened at a concentration of 20 μM. Tabel 4.3 shows the inhibition data, reported as %-inhibition. Inhibitory values of less than 20% were considered insignificant. The inhibitory values above 20% are highlighted.    37  Table 4.3 Inhibitory values of P1-P13 reported as % inhibition. The inhibitors were screened in duplicate at a concentration of 20 μM.   PRMT1 PRMT3 PRMT4 PRMT5 PRMT6 PRMT8 P1 Ac-EFWDWGP-NH2 4 6 7 10 13 5 P2 Ac-EFWD(β3W)GP-NH2 19 20 5 30 27 46 P3 Ac-EF(β3W)DWGP-NH2 13 22 7 15 31 16 P4 Ac-EF(β3W)D(β3W)GP-NH2 9 16 2 14 2 7 P5 Ac-EFWDW(NMeG)P-NH2 10 5 3 1 0 0 P6 Ac-EFW(NMeD)WGP-NH2 0 11 5 7 0 0 P7 Ac-EFW(NMeD)W(NMeG)P-NH2 9 9 11 0 5 5 P8 Ac-EFWD(L-Aia)GP-NH2 33 28 13 34 33 27 P9 Ac-EF(L-Aia)DWGP-NH2 13 19 8 28 28 18 P10 Ac-EF(L-Aia)D(L-Aia)GP-NH2 46 39 9 29 21 25 P11 Ac-EFWD(D-Aia)GP-NH2 1 1 5 0 1 5 P12 Ac-EF(D-Aia)DWGP-NH2 0 2 11 0 1 3 P13 Ac-EF(D-Aia)D(D-Aia)GP-NH2 4 5 9 0 1 4   The data revealed clear trends. Incorporation of one β3-homotryptophan (P2 and P3) showed an increase in inhibitory potency for PRMTs 3, 5, 6 and 8, whereas incorporation of two β3-homotryptophans (P4) was not as well tolerated for inhibition. This could indicate that either the elongation of the peptidomimetic chain by one methylene unit, or the conceivable conformational alteration due to the β3-amino acid facilitates this increase in potency. The data also clearly indicated that the N-methylated derivatives of P1 (P5-P7) did not significantly inhibit any of the PRMTs.  Incorporation of L-Aia (P8-P10) proved  significantly effective; Replacing the tryptophan closest to the C-terminal (P8) showed good inhibition of PRMTs 1, 3, 5, 6 and 8, whereas replacement of the N-terminal tryptophan (P9) proved better at inhibiting PRMT5 and PRMT6 than the other PRMTs. Having two L-Aias in the sequence (P10) resulted in good inhibition of all tested PRMTs except PRMT4, with a maximum inhibition of 46% for PRMT1. On the contrary, all three peptidomimetics containing D-Aia in place of the tryptophans (P11-P13) did not show any inhibitory activity. Interestingly, as L-Aia only allows the gauche(+) or the anti-conformations (figure 2.3, section 2.6) and D-Aia locks the conformation in either the gauche(-) or the anti-conformation it seems likely that the gauche(+) conformation might be the bioactive conformation for the L-Aia residues in the P1 mimetics.      PRMT4 was not significantly inhibited by any of the peptidomimetics, leading us to hypothesise that the inhibition is substrate dependent, seeing as the PRMT4 assay is the only assay tested here that utilises H3 as a substrate rather than H4 (vide supra).      38  4.4.1.2 P1 is a truncated version of P16 When expressed on the phage all the peptides from the same construct exhibit the same two N-terminal and C-terminal sequences, SHS and GGG, respectively. N-terminally, SHS is a part of the leader sequence which truncates the mature fusion of the peptide with the coat protein of the phage. The N-terminal leader sequence except SHS is removed upon secretion from the E. coli (in which the phages are expressed) resulting in the randomised peptide positioned at the N-terminus of the mature phage. On the C-terminal GGG is a linker sequence between the phage and the expressed peptide.125  Thus, when screening the peptide libraries on the phages, these two terminal sequences were also present, and we speculated that these might possibly increase inhibitory activity by providing better binding to the PRMTs. We therefore synthesised the three native peptides P14-P16 and screened these in the same antibody assay as previously described (table 4.4). P16 contains both the N- and C-terminal residues, whereas P14 and P15 are analogues of P16 truncated either at the N-terminal or the C-terminal, respectively.  Table 4.4 Inhibitory values of P14-P16 reported as % inhibition. The inhibitors were screened in duplicate at a concentration of 20 μM. P1 is shown for comparison.   PRMT1 PRMT3 PRMT4 PRMT5 PRMT6 PRMT8 P1 Ac-EFWDWGP-NH2 4 6 7 10 13 5 P14 Ac-EFWDWGPGGG-NH2 0 4 11 0 7 9 P15 Ac-SHSEFWDWGP-NH2 1 9 13 0 7 14 P16 Ac-SHSEFWDWGPGGG-NH2 1 18 9 43 33 32   As P16 showed increased inhibition at PRMT5, 6 and 8 compared to P1 as well as the truncated peptides P14 and P15, we decided to investigate whether the modifications that had been employed in the potent P1 mimetics would produce even better inhibitors when employed in P16 mimetics.     4.4.1.3 Inhibitor Screening of P16 mimetics The peptidomimetics P17-P24 based on the P16 lead were synthesised using the same methodologies as for the P1 mimetics. The modifications which had not shown any inhibition for the P1 mimetics, the N-methylated mimetics and the peptidomimetic containing two β3-hTrps, were not used in the P16 mimetics, except for the D-Aia which could be important for elucidating the conformational preference for the side chains in chi space.  Table 4.5 shows the results from the antibody inhibitory screening of the P16 mimetics.   39  Table 4.5 Inhibitory values of P16-P24 reported as % inhibition. The inhibitors were screened in duplicate at a concentration of 20 μM. P16 is shown for comparison.   PRMT1 PRMT3 PRMT4 PRMT5 PRMT6 PRMT8 P16 Ac-SHSEFWDWGPGGG-NH2 1 18 9 43 33 32 P17 Ac-SHSEFWD(β3W)GPGGG-NH2 2 2 15 7 2 30 P18 Ac-SHSEF(β3W)DWGPGGG-NH2 0 0 13 0 0 10 P19 Ac-SHSEFWD(L-Aia)GPGGG-NH2 0 1 10 0 3 5 P20 Ac-SHSEF(L-Aia)DWGPGGG-NH2 0 0 10 4 1 18 P21 Ac-SHSEF(L-Aia)D(L-Aia)GPGGG-NH2 57 44 4 42 36 36 P22 Ac-SHSEFWD(D-Aia)GPGGG-NH2 0 0 3 0 12 11 P23 Ac-SHSEF(D-Aia)DWGPGGG-NH2 0 0 5 1 20 10 P24 Ac-SHSEF(D-Aia)D(D-Aia)GPGGG-NH2 0 0 3 9 16 7   The peptidomimetics containing one β3-homotryptophan (P17 and P18) showed lower inhibitory values compared to the P1 mimetics with the same modification (P2 and P3, respectively). This was likewise seen for the P16-derived mimetics containing one L-Aia (P19 and P20) when comparing these to the P1-derived peptidomimetics with the same modifications (P8 and P9). Interestingly, the peptidomimetic with both tryptophans replaced with L-Aias (P21) displayed inhibition of PRMTs 5, 6 and 8 comparable to the native sequence P16, but showed a significant increase in inhibition for PRMT1 and PRMT3. Compared to the P1-mimetic containing two L-Aias (P10) the longer sequence P21 presented slightly higher inhibition values at PRMTs 1, 3, 5, 6 and 8 (tables 4.3 and 4.5). Similarly to the 7mer mimetics, the 13mer mimetics containing the D-Aia scaffolds (P22-P24) did not show any significant inhibition, except for P23 at PRMT6 which showed 20% inhibition. Yet again inhibition of PRMT4 was insignificant for all tested peptidomimetics.   Figure 4.2 Structure of P21  40  4.4.1.4 IC50 values for P21 using the Antibody-Based Assay As P21 (figure 4.2) had indicated inhibition in the low micromolar range of PRMTs 1, 3, 5, 6 and 8, we decided to determine the IC50 values of this peptidomimetic. This was done using the same antibody-based assay as previously. The IC50 values for P21 along with the positive control SAH, the by-product of the methylation reaction (figure 1.4), are reported in table 4.6.  Table 4.6 IC50 values of P21 based on the antibody inhibition assay compared to SAH. Values are averages of 3 separate experiments.  IC50 values (μM) PRMT1 PRMT3 PRMT4 PRMT5 PRMT6 PRMT8 P21 Ac-SHSEF(L-Aia)D(L-Aia)GPGGG-NH2 23.1 42.0 - 22.4 17.8 16.1  SAH 2.49 6.67 0.112 0.836 0.561 1.52       For all PRMTs tested, except for PRMT4, the IC50 was in the low micromolar range (16-42 μM), whereas SAH showed slightly better inhibition. In order to determine the concentration of peptidomimetic in the sample tested in vitro, we performed quantitative NMR (qNMR). This is done by using the ERETIC2 method (“Electronic REference To access In vivo Concentrations”)148 in 1H NMR, where the intensity of a signal from the peptidomimetic is compared to the intensity of a signal from a reference compound, in our case dried potassium hydrogenphthalate. The qNMR analysis showed a 69% peptide content, and thus 31% of the sample consists of residue salts and water molecules. The inhibitory potency of P21 could therefore likely be even lower than reported in table 4.6. Note that the purity of P21 was found to be >99% at 210nm and >97% at 280nm using analytical HPLC (aHPLC; Method A, see experimental section).        4.4.2 SPA-based approach to determining inhibitor concentrations After the identification of P21 as a low micromolar inhibitor of several PRMTs, we initiated a collaboration with the Structural Genomics Consortium in Toronto to test the potency of P21 as a PRMT inhibitor using a different assay.  The Structural Genomics Consortium are experts in the scintillation proximity assay-based (SPA) inhibitor screening approach, which uses SPA beads. Briefly, using a biotin/streptavidin linkage the substrate is connected to an SPA bead, which emits a signal when triggered by a radioactive source in close proximity. Using a radiolabelled co-factor, in this case 13C-SAM, a radioactive methyl group can be transferred to the substrate when the enzyme is not inhibited.149 Testing P21 in the SPA-assay with PRMTs 1, 3, 5, 6, 7 and 8 revealed no inhibition, except for PRMT7, which showed an IC50 of 62 μM.     41  4.4.3 Comparing the two inhibition assays It is still quite unclear why the SPA-assay yielded negative results, while the antibody-based assay showed inhibition of several PRMTs using P21. One explanation could be the difference in substrate; for the antibody-based assay H4 residues 1-21 was used, whereas the SPA-assay used H4 residues 1-24. As mentioned before, however, both the 21mer and the 24mer substrates should be of a sufficient length to yield results comparable to full-length H4.44 The antibody-based assay can only detect the specific dimethylation events on H4R3 for PRMTs 1, 3, 5, 6 and 8, as the residues surrounding the arginine in the histone tail substrate are important for recognition by the antibody. On the contrary, the SPA-assay should detect all methylation events that involve transfer of the radiolabelled methyl group. This could indicate that other arginines on the histone 4 tail substrate are methylated and detected in the SPA assay. However, currently the only arginine which has been identified as a site of methylation/dimethylation in the first 24 residues of the histone 4 tail is arginine 3 (H4R3), according to data from uniprot.org (figure 1.2). Furthermore, it is interesting to note that PRMT7 is inhibited slightly by P21 in the SPA-assay, as PRMT7 is the only type III PRMT. This could indicate that P21 may inhibit the enzymes differently depending on the type; if P21 does not inhibit the methylation event leading to MMA, but does so for the methylation leading to ADMA/SDMA, the SPA-method would show no inhibition, whereas the antibody-based assay would. In order to investigate this it would be interesting to analyse the substrates by MS/MS to determine the site and level of methylation.   4.5 Structural Studies of P21  Due to the theory of ‘convergent evolution’ (section 1.6)92 we speculated that the P1 and the derivatives P2-P13 all with a high proportion of aromatic residues might possibly bind to a region rich in aromatic residues, such as the region around the dimerisation arm.30 Because P10 and P21 both contain the same central region containing two L-Aias and showed comparable inhibition trends in the antibody-based assay, we suspected that these two peptidomimetics might bind in the same region on these PRMTs. However, as we did not know and as crystallisation is often difficult for low-affinity binders, we decided to investigate the structure of P21.   In order to determine whether the peptides and peptidomimetics would adopt a preferred structure in solution, we initiated a collaboration with the Université de Montpellier, who conducted the following structural studies.  42  Using plane polarised light, circular dichroism (CD) can be used to identify secondary structures in peptides and proteins as known structural elements such as α-helices and β-sheets have distinct CD spectra.150 For the native peptide P16 (blue) as well as our inhibitor P21 (red) CD was performed in both MeOH (dashed lines) and phosphate buffer (pH 7.4, full lines) (figure 4.3).    Figure 4.3 Circular dichroism of P16 (blue) and P21 (red) at 20 oC in MeOH (dashed line) and phosphate buffer, pH 7.4 (full line).  For the native peptide P16 (blue bands) both solvents show a random coil signature structure with negative values around 200 nm indicating a mostly disordered conformation, which is often seen for short peptides with a large degree of flexibility.151 The peptidomimetic P21 (red bands) showed unusual spectra in both solvents with a positive maxima at 217 nm, a negative maxima at 236 nm and a broad absorption band around 250 nm. The same trends in maxima and minima are seen for the shorter peptidomimetic P10 in both solvents, although the extrema are not as pronounced as for P21 (data not shown).  As the CD data did not reveal any specific structural information, solution NMR was performed for the 13mers P16 and P21. NMR experiments were run in both MeOH-d3 and phosphate buffer (H2O/D2O 9:1, pH 6.5). In MeOH-d3 all 1H NMR peaks could be assigned using various NMR techniques, but in phosphate buffer spectra with signal overlaps were obtained. Due to the similarity of the CD profiles in both solvents for P16 and P21, the NMR data in MeOH-d3 was used for structural assignment of both 13mers.  Consistent with the CD data, peptide P16 exhibited inter-residue Nuclear Overhauser Effect (NOE) signals corresponding to large conformational flexibility and a random conformation in solution. In contrast, NOE 43  signals for P21 (figure 4.4) indicated that the conformational constraints due to the two L-Aia residues lead to a defined conformation in the central core of the peptidomimetic.         Figure 4.4 Inter-residue NOE correlations of P21. The typical NH-Hα correlations of neighbouring residues were omitted for clarity. The typical NOEs observed for the short L-Aia-AA-L-Aia-AA sequences are shown in red.   As can be seen in figure 4.5A the superposition of the Phe5-L-Aia6-Asp7-L-Aia8-Gly9 region in the 15 lowest energy NMR solution states for P21 indicate a structural preference of the peptidomimetic. This turn conformation is stabilised by π-stacking of the locked indole rings in the two L-Aias (figure 4.5B) and an H-bond between the carbonyl of L-Aia6 and the amide of L-Aia8, which forms an interresidue pseudocycle (C7; figures 4.5B and C). Due to the flexibility of the extremities in P21 these showed no convergent structural features.        Figure 4.5 A) Superimposition of the Phe5-Aia6-Asp7-Aia8-Gly9 region of the 15 lowest energy NMR solution structures of P21. B) The π-stacking and C7-stabilising H-bond. C) Indication of the C7 pseudocycle due to hydrogen bonding.  As both P10 and P21 contain the central five residues Phe5-L-Aia6-Asp7-L-Aia8-Gly9, we expect P10 to also adopt a similar turn structure. It thus seems likely that the increased inhibitory potency of P10 and P21 at both PRMT1 and PRMT3 could be due to this structural motif. For the peptidomimetics containing only one L-Aia (P8, P9, P19, P20) the inhibition is not as pronounced, indicating less stable conformations.       44  4.6 Cellular Studies of P16 Concurrently with our inhibitor screening we wanted to investigate the cellular localisation of our peptide lead P16. In order to study the native 13mer P16 in cells we needed to conjugate a fluorophore to the peptide. To create a peptide-fluorophore probe which could be used in a variety of applications, we decided to employ a near-infrared (NIR) small molecule fluorophore, as these have several advantageous properties in vivo, such as reduced light scattering, high photo stability and good tissue penetration.152  Linking a fluorophore to short peptides changes the physicochemical properties of the probes quite dramatically, and it is therefore important to consider which fluorophore to use and where the attachment point should be on the peptide. In this project we utilised two NIR fluorophores; DAOTA (4.31) and tetramethylrhodamine (TMR, 4.32) which are shown in figure 4.6. Fluorophores which function in the NIR part of the visible spectrum must have a high degree of conjugation, and are therefore often rather flat molecules. This may pose problems, as these molecules have a tendency to intercalate in the grooves of nucleic acid helices.153      Figure 4.6 The two utilised fluorophores DAOTA (4.31) and tetramethylrhodamine (4.32; TMR). Carboxylic acids marked in red were points of attachment to the N-terminal of P16.  The DAOTA fluorophore (4.31) was synthesised in the laboratories of Bo Laursen at University of Copenhagen.154 As DAOTA (4.31) contains a short carbon chain functionalised with a carboxylic acid, we decided to couple the DAOTA onto the N-terminal of P16. This was done under normal SPPS coupling conditions and the DAOTA-N-P16 was isolated in 34% yield. The probe was tested in HEK293 cells in the laboratory of Frank Fackelmeyer (FORTH/BRI, Greece), and gratifyingly the DAOTA-N-P16 was able to penetrate the cell membrane and the cells did not seem to suffer under the treatment. However, imaging showed that the probe localised mainly to the nucleoli and stained the cytoplasm slightly (figure 4.7). As this distribution does not fit with any known distribution of the PRMTs,46 we suspected that the probe might intercalate with DNA/RNA, even though this had not been seen with DAOTA-probes in other cellular 45  settings.154 Test experiments showed no binding of P16 to DNA, but as DAOTA without P16 (4.31) showed the same unspecific binding pattern as in figure 4.7, we turned our attention to the TMR fluorophore.     Figure 4.7 DAOTA-N-P16 stains nucleoli, presumably intercalating in the nucleotides in HEK293 cells  TMR-N-P16 was synthesised at Genscript, and also tested in HEK293 cells. Gratifyingly, in this case the negative control, TMR showed no staining. However, the probe was found to mainly localise in the nucleus, but also showed staining of the cytoplasm and the nucleoplasm, but none in the mitochondria (figure 4.8). This is not consistent with the previously seen localisations of the different PRMTs.46     Figure 4.8 TMR-N-P16 localises to the nucleus and cytoplasm in HEK293 cells  Initially, the phage expressing P16 was actually identified in a phage display screen for LSD1 (Lysine Specific Demethylase 1) and was then shown to also bind to PRMTs 1-4 and 9. However, P16 was later shown to bind to the glutathione S-transferase-tag (GST-tag) of the expressed LSD1 rather than the protein itself.155 The phage displaying P16 was also cross screened against the epigenetic lysine demethylases (LSD- and KDM-families) as well as HAT EP300 (an acetyl transferase) and did not show any affinity for these. In this cellular setting however, it seems that the probe may bind to other targets as well as the PRMTs. It is hard to draw conclusions from this study, as other factors such as compartmentalisation, proteolytic stability and ability to cross membranes among others should be considered. It would therefore be interesting to investigate the probe in other cell lines, as well as investigating a TMR-C-P16 probe, as the displayed peptide is attached to the phage using the C-terminal in the phage display screen.  46  4.7 Conclusion to the PRMT Peptidomimetic Studies Our approach of finding new PRMT inhibitor leads using phage display had to our knowledge not been attempted before, as all previous approaches are based on the bisubstrate approach or compound/in silico library screenings.  We designed and synthesised a small library of peptides and peptidomimetics based on two leads from phage display, P1 and P16, which were used in a structure-activity relationship study to develop inhibitors for the epigenetic PRMTs. Several modifications of or around the two tryptophans were tested on the 7mer (P1) and the 13mer (P16) leads; replacing tryptophan with β3-homotryptophans (P2-P4 and P17-P18), N-methylation of the succeeding amino acid (P5-P7), incorporation of one or two D-Aias (P11-P13 and P22-P24) and incorporation of one or two L-Aias (P8-P10 and P19-P21).   The study lead to the identification of P21, a 13mer peptide containing two L-Aias in place of tryptophans, which showed low micromolar inhibition of PRMTs 1, 3, 5, 6 and 8 in an antibody-based assay, thus having inhibition in the same range as the PRMT bisubstrate inhibitors.  As the assay for PRMT4 uses a different substrate and as P21 was not able to inhibit PRMT4, we speculate that the inhibition may be substrate specific. Furthermore, as the SPA-assay did not show any inhibition of PRMTs 1, 3, 5, 6 or 8 using P21, we suspect that this inhibition may in fact be even more specific – either due to the difference in H4 substrate used in the two assays (21mer for antibody-based assay and 24mer for the SPA-assay) or the extent to which methylation is inhibited in the two assays. It is, however, interesting to note that replacing the two tryptophans with L-Aias in P21 does not affect the inhibition of PRMT5, 6 and 8 much (~30-40% inhibition of P16 and P21), whereas this modification switches inhibition from non-significant (P16) to potent inhibition (P21) for PRMT1 and PRMT3. As PRMT1 has been shown to be the most expressed PRMT in vivo49,50 and as no potent PRMT1 inhibitors exist as of today (vide supra), it would be interesting to further study P21 for the purpose of developing a subspecific potent PRMT1 inhibitor. Studying PRMT1 is still difficult, as no human crystal structure for PRMT1 exists as of yet. The crystal structure for rat PRMT128 has been used to derive a homology model of human PRMT1,67,68,78 but this has thus far not lead to potent, selective PRMT1 inhibitors. Complicating the search even further is the fact that PRMT1 has been found to exist in various splicing isoforms,51 and thus isoforms lacking specific regions may not be targeted by a general PRMT1 inhibitor.  The structural studies revealed an interesting stabilised turn conformation when two L-Aias were incorporated into the peptide sequence. This conformation seems to be conferring inhibition towards PRMT1 and PRMT3 while not affecting the inhibitory effects on PRMTs 5, 6 and 8 much, as P10 and P21 show the same inhibition pattern and both contain the two L-Aias. Moreover, seeing as this structural 47  feature is stabilised through the L-Aia moieties and the peptide backbone only, we suspect that incorporation of –L-Aia-Xxx-L-Aia- into other peptide sequences could facilitate the same turn conformation.  The cellular studies indicated that TMR-N-P16 was a promiscuous binder in HEK293 cells, since it did not display a familiar localisation pattern. In order to be able to draw conclusions from this result it would be interesting to investigate several more probes, such as the TMR-C-P16, but also TMR-N-P21, which may be less promiscuous due to its constrained conformation. Also, as P21 inhibits PRMT1 whereas P16 does not, and as PRMT1 is the most expressed PRMT in the cells, P21 may display a more familiar PRMT localisation pattern in human cells.      As mentioned previously cells have several ways of regulating PRMT function. The PRMTs are often found to be active in larger protein complexes, and may behave differently when regulated by posttranslational modifications.17,27 Thus, when trying to modulate the activity of the PRMTs it may not be relevant to look at the isolated enzyme, but rather at a variety of active states using various interaction partners, preferably in a cellular setting, complicating the search even further. The first study on inhibitors of the PRMTs was published only 11 years ago, and research in the PRMTs as drug targets is still a newer field in research. In order to develop potent, selective PRMT inhibitors in the future it will be of utmost importance to gain a thorough understanding of the regulation, mechanisms, localisations and activity of these enzymes.                48  4.8 EXPERIMENTAL SECTION  4.8.1 GENERAL INFORMATION All natural amino acids, resin and coupling reagents were purchased from Iris Biotech GmbH, Marktredwitz, Germany or Chem-Impex International Inc., Wood Dale, Illinois, USA. β3-amino acids were purchased from Chem-Impex International Inc. Chemicals for preparation of 3.11a and 3.11b and for N-methylation were purchased from Sigma-Aldrich, Schnelldorf, Germany.  4.8.2 ANALYTICAL METHODS AND EQUIPMENT 4.8.2.1 Analytical High Performance Liquid Chromatography (aHPLC) Method A Analytical HPLC was performed on a Dionex UltiMate 3000 system equipped with a photodiode array detector and an RP Phenomenex Gemini NX-C18 column (250 x 4.6 mm, 3 μM) with eluents A (100% H2O + 0.1% trifluoroacetic acid (TFA)) and B (90% MeCN, 10% H2O + 0.1% TFA) with a flowrate of 1 mL/min.  Reported retention times (tR) were obtained using the following gradient: 0% B 0-5 min, 0-100% B 5-35 min, 100% B 35-40 min.   4.8.2.2 Low Resolution Mass Spectrometry (LC-LRMS or LC-MS) Method B LC-LRMS was performed on an Agilent 1200 series LC equipped with a diode array detector coupled to a Bruker Esquire3000plus ion trap mass spectrometer. The column (Waters Eclipse XDB-C18, 4.6 × 100 mm, 5 µm) was kept at 45°C. Ionisation of the eluting compounds was obtained by ESI in the positive mode with a nebulising pressure of 50 psi, a dry gas of 10 L/min kept at 240°C and a capillary potential of -2750 V. Eluents used: A (95% H2O, 5% MeCN + 0.1% FA) and B (95% MeCN, 5% H2O + 0.1% FA) with a flowrate of 0.5 mL/min. Method: 0-7 min 10-90% B, 7-9 min 90% B. Method C LC-LRMS was performed on an Agilent 1200 series LC with a diode-array detector coupled to an Agilent 6410 Triple Quad mass spectrometer with an ESI. The column was a Zorbax Eclipse XBD-C18 column (50 mm × 4.6 mm). Eluents used: A (95% H2O, 5% MeCN + 0.1% HCO2H) and B (95% MeCN, 5% H2O + 0.043% HCO2H) with a flowrate of 0.75 ml/min. Method: 0-5 min 0-50% B, 5-5.1 min 50-90% B, 5.1-5.5 min 90% B.  4.8.2.3 High Resolution Mass Spectrometry/Exact mass (HRMS) Performed at Vrije Universiteit Brussel. HRMS HPLC was performed on a C18 column (25 cm × 4.6 mm, 5 μm) with a linear gradient (10% to 100% MeOH in H2O containing 0.1% TFA (v/v) in 20 min) at a flow rate of 1 mL/min and UV detection at 215 nm.  High-resolution Mass Spectrometry was recorded on a Micromass Q-TOF 1.5, UB137 or on a time-of-flight (TOF) MS system coupled to an ESI detector.  49  4.8.2.4 Preparative Reverse-Phase High Performance Liquid Chromatography (pHPLC) Method D Preparative RP HPLC of the peptides was carried out on a Dionex UltiMate 3000 HPLC with a diode array detector using a preparative RP Phenomenex Gemini NX-C18 column (250 x 21.20 mm, 5 μM) using eluent A (100% H2O + 0.1% TFA) and eluent B (90% MeCN, 10% H2O + 0.1% TFA) and a flow rate of 20 mL/min. For each peptide a step program was made based on elution time of impurities and product in the crude mixtures.  4.8.2.5 Optical rotation Optical rotation was recorded on an Anton Paar Modular Circular Polarimeter (MCP300) with the temperature set to 25 oC. The calculated specific optical rotation was the average of three separate readings.  4.8.2.6 Chiral HPLC Chiral HPLC was run on a Dionex UltiMate 3000 HPLC with a diode array detector with a Chiral Cell OD-H chiral column (4.6 x 250 mm) from Diacell using an isocratic solvent system (70% heptane, 30% EtOH and 0.1% TFA). Enantiomeric excess was calculated using the following formula: ee% = ((R-S)/(R+S)) x 100.    4.8.2.7 Nuclear Magnetic Resonance Spectroscopy (NMR) NMR spectra were recorded on an Ascend 400 MHz Ultrashield Bruker instrument. Samples were dissolved in DMSO-d6 purchased from VWR with a purity of >99.8%. The FID-files were obtained and processed using standard parameters in Topspin v3.2 and analysed using the MestReNova software version 6.0. Signals are reported in ppm (δ) using the solvent as reference. Signal assignment was based on unambiguous chemical shifts. Coupling constants (J) are reported in Hertz (Hz) and rounded to the nearest 0.5 Hz.   4.8.2.8 Quantitative NMR (qNMR)  The net peptide content for peptide P21 was determined on a Bruker Avance IIIHD 600 MHz NMR spectrometer equipped with a 5 mm cryogenically cooled dual 1H/13C-probe for quantification and structural characterisation of the peptidomimetics. The software TopSpin 3.2 was used for data acquisition and analysis. Quantitative determinations was performed with the ERETIC2 method148 by measuring a separate sample of monobasic potassium phthalate at 0.67 mM (Merck) under identical conditions as the sample (P21 stock solution 0.67 mM in 20% D2O in DMSO-d6, 300 K, 12 kHz sweep width, acquisition time 2.73 sec., relaxation delay 4.27 sec.) with appropriate adjustments of tuning and matching of the probe, determination of the correct 90°-pulse, number of scans and receiver gain.  50  4.8.3 GENERAL SYNTHESIS PROCEDURES 4.8.3.1 Standard Fmoc-based SPPS Standard Fmoc-based SPPS was performed by hand using Rink amide resin (0.7 mmol/g) in a disposable preassembled plastic syringe reactor fitted with a Teflon filter. The standard protocol was followed except when performing N-methylation or Aia insertion.  1 Swell resin in CH2Cl2 30 min 2 Wash with DMF 3x 10 sec 3 Fmoc deprotection 5 min 4 Wash with DMF 3x 10 sec 5 Fmoc deprotection 20 min 6 Wash with DMF 3x 10 sec; 3x 1 min; 3x 10 sec 7 Coupling  60 min 8 Wash with DMF 3x 1 min 9 Colour test  Full conversion; repeat steps 3-9. If coupling incomplete; repeat steps 7-9    Fmoc-deprotection mixture 20% piperidine in DMF + 0.1 M HOBt.  The coupling mixture consisted of 1:1:1 amino acid, DIPEA and HBTU in DMF 3-fold and later in the project 5-fold compared to resin loading. The coupling mixture was preactivated for 5 min prior to addition.  N-terminal acetylation After coupling of the final amino acid steps 3-6 were repeated and resin treated with 8:1:1 DMF/DIPEA/acetic anhydride for 30 min followed by 3 x 10 sec DMF wash and drying of resin.  4.8.3.2 Microwave Fmoc-SPPS (MW Fmoc-SPPS) Microwave peptide synthesis was performed on a Biotage Initiator+ SP Wave in reaction vessels supplied with the SP Wave.  1 Swell resin in CH2Cl2 30 min 2 Wash with DMF 4x 30 sec 3 Fmoc deprotection 3 min 4 Fmoc deprotection 3 min 5 Wash with DMF 4x 30 sec 6 Coupling 5 min at 75 oC 7 Coupling  5 min at 75 oC 8 Wash with DMF 4x 30 sec 9 Colour test   Full conversion; repeat steps 3-9. If coupling incomplete; repeat steps 6-8    Fmoc-deprotection mixture, the coupling mixture and N-terminal acetylation similar to section 4.8.3.1 51  4.8.3.3 General procedure for N-methylation By an adaptation of the published procedure.139 Briefly, the resin was deprotected to yield the free N-terminal primary amine. To this 4 eq. 2-nitrobenzenesulfonyl chloride and 10 eq. collidine in NMP was added. The resin was shaken for 15 min at rt followed by 5x 30 sec washes with NMP. Then 3 eq. DBU in NMP was added and shaken for 3 min before 10 eq. dimethyl sulphate in NMP was added and shaken for another 2 min. The resin was washed for 10 sec with NMP and DBU/dimethyl sulphate treatment was repeated, followed by 5x 30 second NMP washes. The resin was treated twice with 10 eq. β-mercaptoethanol and 5 eq. DBU in NMP for 5 min, followed by 5x 30 second NMP washes. Test cleavage and LC-MS analysis (Method B or C) was evaluated before continuing standard Fmoc-based SPPS.  4.8.3.4 General procedure for reductive amination on resin By an adaptation of the published procedure.140 In short, the resin was deprotected to yield the free N-terminal primary amine followed by 3x 10 sec, 3x 1 min, 3x 10 sec DMF washes. Resin was shaken in 1:1 CH2Cl2/trimethyl orthoformate (TMOF) with 0.5% acetic acid (AcOH) for 30 min at rt. Then 2 eq. of Fmoc-2’-formyl-tryptophan (3.11) were suspended in 1:1 CH2Cl2/TMOF with 0.5% AcOH (1 mL/250 mg resin) and added to the resin. Subsequently 4 eq. NaCNBH3 was suspended in DMF (1 mL/250 mg resin) and then added to the resin and shaken at rt. The reaction was monitored closely; the reaction vessel was drained and a test cleavage was performed and analysed by LC-MS (Method B or C). If the reaction was incomplete the same reagents were added along with 2 eq. NaCNBH3 to the reactor and shaken further. If the reaction was complete the resin was washed 3x 10 sec with DMF, followed by treatment with 2 eq. semicarbazide hydrochloride and 2 eq. DIPEA in DMF for 15 min. This was repeated and finally the resin was washed with 3x 1 min DMF, 3x 1 min iPrOH and 3x 1 min CH2Cl2.    4.8.3.5 General procedure for coupling to yield Aia Following reductive amination the resin was washed with DMF for 3x 10 sec. Coupling was performed twice on the Biotage Initiator+ SP Wave by using 5 fold excess of HBTU and DIPEA in DMF reacted for 5 min at 75 oC (fresh reagents for each coupling). Resin was washed with DMF 4x 30 sec and a test cleavage was performed and analysed using LC-MS (Method B or C). If the reaction was incomplete another coupling was performed followed by a test cleavage as previously described. If the reaction was complete standard Fmoc-based SPPS was continued.    52  4.8.3.6 Colour tests a) The Kaiser test was performed by mixing a few drops of each of the solutions A, B and C with resin and heating to 110 oC for 5 min. Blue resin and solution indicate free primary amine.  Solution A: Phenol in EtOH (4:1, w/v); Solution B: 0.28 M ninhydrin in EtOH; Solution C: 0.2 mM KCN in pyridine.  b) The Chloranil test was performed by mixing a few drops of the solutions D and E and incubating with resin at rt for 5 min. Blue/green resin indicates free primary/secondary amines.  Solution D: 2% acetaldehyde in DMF (v/v%); Solution E: 2% chloranil in DMF (w/v%). c) The TNBS test was performed by mixing solutions F and G with resin, incubating at rt for 10 min followed by 3-4 thorough washes with DMF. Orange/red resin indicates free primary amines. Solution F: 10% DIPEA in DMF (v/v%); Solution G: 1M aqueous TNBS.    4.8.3.7 Test-cleavage A visible amount of resin was stirred in 1 mL cleavage cocktail (90:5:5 TFA/TIS/H2O) for 30 min. A stream of N2 was used to evaporate the solvent, leaving a film, which was taken into 1:1 H2O/MeCN, filtered and analysed by LC-MS (Method B or C).    4.8.3.8 Final cleavage The dried resin was cleaved in a cooled cleavage cocktail of 90:5:5 TFA/TIS/H2O with 3 mL/100 mg resin. The solution was stirred at rt for 3 hr and filtered into a 10-fold excess cold Et2O. The remaining resin was washed with 1x 1 mL TFA and 2x 1 mL CH2Cl2. To ensure full peptide precipitation the Et2O phase was stored at -20 oC for 30 min and subsequently filtered to yield the crude peptide, which was washed with 2x 10 mL cold Et2O and dried thoroughly. The residue was dissolved in a mixture of acetic acid and H2O, as slight basic solutions were shown to racemise the peptide, before concentration and thorough freezedrying.     The crude peptide/peptidomimetic was kept in the freezer until purification.  Prior to pHPLC purification (Method D) the peptides/peptidomimetics were dissolved in 60-80% DMF in H2O. (The crude peptide/peptidomimetics was dissolved in DMF, filtered and then H2O was added.)    53  4.8.4 COMPOUND CHARACTERISATION L-Tcc (4.14a) and D-Tcc  (4.14b)  Synthesised according to published procedure using 1.4 eq 37% formaldehyde and 0.0125M  H2SO4.141 The reactions were stirred overnight, monitored using aHPLC and the resultant white precipitates was used in the following step without further purification.  For both 4.14a and 4.14b: HPLC tR = 9.09 min (Method A).      Fmoc-L-Tcc  (4.15a) and Fmoc-D-Tcc  (4.15b) Synthesised according to published procedure140 and monitored using aHPLC. After work-up the yellow residue was triturated in CH2Cl2. The resultant precipitates were not purified further.   4.15a: HPLC tR = 18.64 min (Method A). 4.15b: HPLC tR = 18.71 min (Method A).   Fmoc-L-2’-formyl-Trp-OH (4.12a) Synthesised according to published procedure.140 HPLC tR = 16.83 minii (Method A). Purity at 210nm >96%; Purity at 280nm >97%. 1H NMR (400 MHz, DMSO) δ 12.91 (s, 1H, COOH), 11.76 (s, 1H, NH (indole)), 10.07 (s, 1H, CHO), 7.86 (d, J = 8 Hz, 2H, Fmoc-Ar), 7.82 (d, J = 8 Hz, 1H, indole-Ar), 7.61 (dd, J = 16.5, 7.5 Hz, 2H, Fmoc-Ar), 7.46 – 7.37 (m, 3H, Fmoc-Ar + indole-Ar), 7.37 – 7.24 (m, 3H, Fmoc-Ar + indole-Ar), 7.11 (t, J = 7.5 Hz, 1H, indole-Ar), 4.34 (m, 1H, CαH), 4.21 – 4.11 (m, 3H, Fmoc-CH2 + Fmoc-CH), 3.60 (dd, J = 14, 4.5 Hz, 1H, CHαHβ), 3.42 (dd, J = 14, 8.5 Hz, 1H, CHαHβ). 13C NMR (101 MHz, DMSO) δ 182.7 (COOH), 173.5 (CHO), 156.4 (-O-CO-NH-), 144.2 (Fmoc-Ar), 144.1 (Fmoc-Ar), 141.1 (indole-Ar), 138.1 (indole-Ar), 133.8 (Fmoc-Ar), 128.1 (Fmoc-Ar), 127.5 (Fmoc-Ar), 127.4 (indole-Ar), 127.0 (indole-Ar), 125.7 (Fmoc-Ar), 123.3 (indole-Ar), 121.5 (indole-Ar), 120.5 (indole-Ar), 113.3 (indole-Ar), 66.3 (CH-CH2), 55.7 (CαH), 47.0 (Fmoc-CH), 26.0 (CH2). [α]25D = +25.8o (c 1, dioxane).iii Chiral HPLC: ee% >97% (HPLC tR = 12.7 min).                                                            ii A mix of Fmoc-L-2’-formyl-Trp-OH and Fmoc-D-2’-formyl-Trp-OH co-eluted in a sharp peak using aHPLC (Method A).  iii Literature reports an optical rotation of [α]25D = +104o/-104 o  with c=1 in dioxane.140  54  Fmoc-D-2’-formyl-Trp-OH (4.12b) Synthesised according to published procedure.140 HPLC tR = 16.87 minii (Method A). Purity at 210nm: > 97%;  Purity at 280nm: > 97%. 1H NMR (400 MHz, DMSO) δ 12.92 (s, 1H, COOH), 11.74 (s, 1H, NH (indole)), 10.08 (s, 1H, CHO), 7.87 (d, J = 8 Hz, 2H, Fmoc-Ar), 7.82 (d, J = 8 Hz, 1H, indole-Ar), 7.61 (dd, J = 16.5, 7.5 Hz, 2H, Fmoc-Ar), 7.47 – 7.37 (m, 3H, Fmoc-Ar + indole-Ar), 7.37 – 7.25 (m, 3H, Fmoc-Ar + indole-Ar), 7.11 (t, J = 7.5 Hz, 1H, indole-Ar), 4.35 (td, J = 9, 5 Hz, 1H, CαH), 4.21 – 4.11 (m, 3H, Fmoc-CH2 + Fmoc-CH), 3.61 (dd, J = 14, 5 Hz, 1H, CHαHβ), 3.45 (dd, J = 14, 10 Hz, 1H, CHαHβ). 13C NMR (101 MHz, DMSO) δ 182.7 (COOH), 173.5 (CHO), 156.4 (-O-CO-NH-), 144.2 (Fmoc-Ar), 144.1 (Fmoc-Ar), 141.1 (indole-Ar), 138.1 (indole-Ar), 133.8 (Fmoc-Ar), 128.1 (Fmoc-Ar), 127.5 (Fmoc-Ar), 127.4 (indole-Ar), 127 (indole-Ar), 125.7 (Fmoc-Ar), 123.3 (indole-Ar), 121.5 (indole-Ar), 120.5 (indole-Ar), 113.3 (indole-Ar), 66.3 (CH-CH2), 55.7 (CαH), 47.0 (Fmoc-CH), 26.0 (CH2). [α]25D = -27.40◦ (c 1, dioxane).iii Chiral HPLC: ee% >99% (HPLC tR = 9.7 min).  PEPTIDE AND PEPTIDOMIMETIC CHARACTERISATIONS    Sequence  Formula  Calculated  HRMS (ESP+) m/z (deviation in ppm)  P1 Ac-EFWDWGP-NH2 C49H56N10O12 999.3972 [M + Na]+ 999.3976 (0,4)  P2 Ac-EFWD(β3W)GP-NH2 C50H58N10O12 1013.4133 [M + Na]+ 1013,4085 (4,2) P3 Ac-EF(β3W)DWGP-NH2 C50H58N10O12 1013.4133 [M + Na]+ 1013,4109 (2,4) P4 Ac-EF(β3W)D(β3W)GP-NH2 C51H60N10O12 1027.4258 [M + Na]+ 1027,4194 (-8,9) P5 Ac-EFWDW(NMeG)P-NH2 C50H58N10O12 1013.4128 [M + Na]+ 1013,4206 (7,7) P6 Ac-EFW(NMeD)WGP-NH2 C50H58N10O12 991.4308 [M + H]+ 991,4203 (-10,6) P7 Ac-EFW(NMeD)W(NMeG)P-NH2 C51H60N10O12 1027.4285 [M + Na]+ 1027,4276 (-0,9) P8 Ac-EFWD(L-Aia)GP-NH2 C50H56N10O12 1011.3972 [M + Na]+ 1011,4065 (9,2) P9 Ac-EF(L-Aia)DWGP-NH2 C50H56N10O12 1011.3972 [M + Na]+ 1011,3938 (-3,4) P10 Ac-EF(L-Aia)D(L-Aia)GP-NH2 C51H56N10O12 1023.3972 [M + Na]+ 1023,4064 (9,0) P11 Ac-EFWD(D-Aia)GP-NH2 C50H56N10O12 1011.3972 [M + Na]+ 1011,3977 (0,5) P12 Ac-EF(D-Aia)DWGP-NH2 C50H56N10O12 1011.3972 [M + Na]+ 1011,3962 (1,0) P13 Ac-EF(D-Aia)D(D-Aia)GP-NH2 C51H56N10O12 1023.3972 [M + Na]+ 1023,3984 (1,2) P14 Ac-EFWDWGPGGG-NH2 C55H65N13O15 1170.4615 [M + Na]+ 1170,4625 (0,9) P15 Ac-SHSEFWDWGP-NH2 C61H73N15O17 1310.5201 [M + Na]+ 1310,5178 (1,8) P16 Ac-SHSEFWDWGPGGG-NH2 C67H82N18O20 1481.5845 [M + Na]+ 1481,5929 (5,7) P17 Ac-SHSEFWD(β3W)GPGGG-NH2 C68H84N18O20 1473.6182 [M + H]+ 1473,6154 (1,9) P18 Ac-SHSEF(β3W)DWGPGGG-NH2 C68H84N18O20 1495.6001 [M + Na]+ 1495,6008 (0,5) P19 Ac-SHSEFWD(L-Aia)GPGGG-NH2 C68H82N18O20 1471.6025 [M + H]+ 1471,6047 (1,5) P20 Ac-SHSEF(L-Aia)DWGPGGG-NH2 C68H82N18O20 1493.5845 [M + Na]+ 1493,5867 (1,5) P21 Ac-SHSEF(L-Aia)D(L-Aia)GPGGG-NH2 C69H82N18O20 1505.5845 [M + Na]+ 1505,5814 (2,1) P22 Ac-SHSEFWD(D-Aia)GPGGG-NH2 C68H82N18O20 1493.5845 [M + Na]+ 1493,5870 (1,7) P23 Ac-SHSEF(D-Aia)DWGPGGG-NH2 C68H82N18O20 1493.5845 [M + Na]+ 1493,5895 (3,3) P24 Ac-SHSEF(D-Aia)D(D-Aia)GPGGG-NH2 C69H82N18O20 1505.5845 [M + Na]+ 1505,5876 (2,1)  55   Sequence tR (min) aHPLC purity (210nm/280nm) P1 Ac-EFWDWGP-NH2 22.1 >98%/>99% P2 Ac-EFWD(β3W)GP-NH2 22.0 >99%/>99% P3 Ac-EF(β3W)DWGP-NH2 21.8 >99%/>99% P4 Ac-EF(β3W)D(β3W)GP-NH2 21.6 100%/>99% P5 Ac-EFWDW(NMeG)P-NH2 22.1 >98%/>97% P6 Ac-EFW(NMeD)WGP-NH2 22.1 >95%/>96% P7 Ac-EFW(NMeD)W(NMeG)P-NH2 22.3 >99%/>99% P8 Ac-EFWD(L-Aia)GP-NH2 22.5 >97%/>98% P9 Ac-EF(L-Aia)DWGP-NH2 22.3 >97%/97% P10 Ac-EF(L-Aia)D(L-Aia)GP-NH2 23.0 >98%/>98% P11 Ac-EFWD(D-Aia)GP-NH2 22.8 >99%/>99% P12 Ac-EF(D-Aia)DWGP-NH2 23.5 >99%/>99% P13 Ac-EF(D-Aia)D(D-Aia)GP-NH2 23.5 >99%/>99% P14 Ac-EFWDWGPGGG-NH2 21.5 100%/100% P15 Ac-SHSEFWDWGP-NH2 19.7 >99%/>98% P16 Ac-SHSEFWDWGPGGG-NH2 19.0 >95%/96% P17 Ac-SHSEFWD(β3W)GPGGG-NH2 18.9 >96%/96% P18 Ac-SHSEF(β3W)DWGPGGG-NH2 18.9 >98%/>98% P19 Ac-SHSEFWD(L-Aia)GPGGG-NH2 19.4 >98%/>98% P20 Ac-SHSEF(L-Aia) DWGPGGG-NH2 19.4 >97%/>95% P21 Ac-SHSEF(L-Aia)D(L-Aia)GPGGG-NH2 20.0 >99%/>97% P22 Ac-SHSEFWD(D-Aia)GPGGG-NH2 19.6 >95%/>98% P23 Ac-SHSEF(D-Aia)DWGPGGG-NH2 20.4 >98%/>99% P24 Ac-SHSEF(D-Aia)D(D-Aia)GPGGG-NH2 20.2 >96%/>98%  DAOTA-N-P18  P18 was synthesised on solid support according to the standard Fmoc-based SPPS protocol (4.8.3.1). After coupling of the N-terminal serine the terminal was Fmoc-deprotected and DAOTA (4.31) was coupled onto the N-terminal using 2-fold excess of 4.31, HBTU and DIPEA for 90 min at rt. The resin was thoroughly washed with DMF until washing solvent was clear and the peptide was cleaved and purified according to protocol. HPLC tR = 20.49 min (Method A); purity >97%/>97% (210nm and 280nm, respectively). m/z (+ESI) found: MH+, 1782,6852. (C89H92N20O21+ requires M, 1782,7204).   P18-N-TMR The peptide was synthesised at Genscript, who reported a purity of >97% at 210nm, HPLC tR = 21.10 min and m/z of [M+2H]2+ 916.00 56  4.8.5 ASSAYS AND STRUCTURAL STUDIES  4.8.5.1 Antibody-Based Inhibition Assay  Performed at BPS Bioscience. The inhibitor screening of peptides/peptidomimetics was performed in duplicate for 60 min at rt. Stock solutions of 10 mM peptide/peptidomimetic in DMSO were prepared and diluted to a final concentration of 20 μM peptide/peptidomimetic (1% DMSO final) in the 50 μL reaction mixture containing methyltransferase assay buffer, SAM (0.2 μM for PRMT1 and 1 μM for the remaining PRMTs) and enzyme (either PRMT1 (5 ng), PRMT3 (30 ng), PRMT4 (200 ng), PRMT5 (100 ng) , PRMT6 (100 ng) or PRMT8 (50 ng)) in histone substrate-precoated wells (H4 1-21 for PRMT1, 3, 5, 6, 8 and H3 1-27 (acK18) for PRMT4). The reaction mixtures were discarded and the wells were washed 3 x with TBS-T buffer and then shaken with Blocking Buffer for 10 min. The wells were emptied and incubated with a primary antibody for 60 min. Wells were emptied, washed 3 x with TBS-T, shaken 10 min with Blocking Buffer and incubated with an HRP-secondary antibody for 30 min at rt. Wells were again emptied, washed 3 x with TBS-T and shaken 10 min with Blocking Buffer before a freshly prepared HRP chemiluminescent mixture was added. The plates were read immediately on the BioTek SynergyTM 2 plate reader.  The %-activity was calculated using: %-activity = (C-C0)/(Ce-C0), where C is the luminescence in the presence of compound, C0 is the luminescence in the absence of enzyme and Ce is the luminescence in the absence of peptide/peptidomimetic. SAH was used as a positive control. IC50 curves of P21 were produced using the same assay in triplicates.  4.8.5.2 SPA-based Inhibition Assay  Performed at SGC Toronto. The assay was performed in 20 mM Tris X HCl pH 7.5-8.5 (PRMTs 1, 3, 5, 7 and 8) or 20 mM Bis-Tris propane pH 7.5 (for PRMT6) all containing 0.01% detergent (either Tween-20 or Triton X-100) and 5-10 mM DTT. Concentration of enzyme used was 15 nM (PRMT1), 20 nM (PRMT3), 15 nM (PRMT5), 50 nM (PRMT6), 25 nM (PRMT7), 20 nM (PRMT8). For PRMTs 1, 3, 5, 6 and 8 H4 1-24 was used as substrate in the 120-700 nM range. 300 nM H2B 23-37 was the substrate for PRMT7. SAM was used as the co-factor in the 1.1-28 μM concentration range.         57  4.8.5.3 CD Studies Performed by Baptiste Legrand at Université de Montpellier.  Samples were dissolved in a spectrophotometric grade MeOH and 50 mM phosphate buffer (pH 7.4) at 100-200 μm. Circular dichroism (CD) experiments were carried out using a Jasco J815 spectropolarimeter. Spectra were recorded using a 1 mm path length CD cuvette, over a wavelength range of 190-340 nm at 20°C Continuous scanning mode was used with a response of 1.0 s with 0.1 nm steps and a scan speed of 100 nm/min. The signal to noise ratio was improved by acquiring each spectrum over an average of three scans. Baseline was corrected by subtracting background from the sample spectrum.  4.8.5.4 NMR Structural Studies Performed by Baptiste Legrand at Université de Montpellier.  The NMR samples contained 2 mM of P16 or P21 dissolved in CD3OH and 50 mM phosphate buffer (potassium phosphate, pH 6.5). All spectra were recorded on a Bruker Avance 600 AVANCE III spectrometer equipped with a 5 mm quadruple resonance probe (1H, 13C, 15N, 31P). Homonuclear 2-D spectra DQF-COSY, TOCSY (DIPSI2) and ROESY were typically recorded in the phase-sensitive mode using the States-TPPI method as data matrices of 300 real (t1) × 2048 (t2) complex data points; 8-64 scans per t1 increment with 1.0 s recovery delay and spectral width of 7210 Hz in both dimensions were used. The mixing times were 80 ms for TOCSY and 350 ms for the ROESY experiments. Sequences with a presaturation or a watergate block was applied to remove water signal. Spectra were processed with Topspin (Bruker Biospin) and visualised with Topspin or NMRView156 on a Linux station. The matrices were zero-filled to 1024 (t1) x 2048 (t2) points after apodisation by shifted sine-square multiplication and linear prediction in the F1 domain. Chemical shifts were referenced to the tetramethylsilane (TMS in CD3OH and TMSP in phosphate buffer). 1H chemical shifts were assigned according to classical procedures. NOE cross-peaks were integrated and assigned within the NMRView software.156  The volume of a ROE between methylene pair protons was used as a reference of 1.8 Å. The lower bound for all restraints was fixed at 1.8 Å and upper bounds at 2.7, 3.3 and 5.0 Å, for strong, medium and weak correlations, respectively. Pseudo-atom corrections of the upper bounds were applied for unresolved aromatic, methylene and methyl proton signals as described previously.157 Structure calculations were performed with AMBER 11158 in two stages: cooking and simulated annealing.  The cooking stage was performed at 1000 K to generate 100 initial random structures. SA calculations were carried during 20 ps (20000 steps, 1 fs long) as described elsewhere. First, the temperature was risen quickly and was maintained at 1000 K for the first 5000 steps, then the system was cooled gradually from 1000 K to 100 K from step 5001 to 18000 and finally the temperature was brought to 0 K during the 2000 remaining steps. For the 3000 first steps, the force constant of the distance 58  restraints was increased gradually from 2.0 kcal.mol-1.Å to 20 kcal.mol-1.Å. For the rest of the simulation (step 3001 to 20000), the force constant is kept at 20 kcal.mol-1.Å. The 20 lowest energy structures with no violations > 0.3 Å were considered as representative of the peptidomimetic structure. The representation and quantitative analysis were carried out using MOLMOL159 and PyMOL (Delano Scientific).   4.8.5.5 Cell assay for DAOTA-N-P18 and P18-C-TMR Performed by Frank Fackelmeyer, FORTH/BRI, Greece. Human kidney HEK293 cells were split onto Alcian-blue coated coverslips the day before imaging. Subconfluent cells were washed twice with ice-cold PBS, and fixed with 4% paraformaldehyde in PBS for 20 min at rt. After one wash in PBS, remaining reactive sites were quenched by incubation in 100 mM Tris-HCl pH 7.5 for 30 min at rt, if needed cells were permeabilised by incubation in 0.5% Triton X100 in PBS for 10 min on ice, and washed twice in PBS at rt. Unspecific binding sites were blocked by incubation in PBS with 6% BSA for 60 min, before cells were incubated with TMR-labelled peptides (diluted 1:100 into PBS with 1% BSA, from 4 mg/mL stocks in DMSO) for 1 hr at rt in the dark. In some experiments, Sytox Green was added to the labelling mixture at a 1:20.000 dilution (from a 5 mM stock, Molecular Probes) as a nuclear counterstain. Cells were washed five times with PBS with 0.1% Triton X100, and finally mounted on microscope slides. Imaging was performed on a Leica SP5 confocal microscope with 63x oil immersion objective.     59  Chapter 5: Expanding the Peptidomimetic Toolbox  – Using Aziridine Chemistry in the Pursuit of a Lysine/Arginine Dipeptide-mimetic  5.1 Introduction The aromatic and basic residues are the most utilised in target receptor recognition in Nature.127-130 This is due to the high propensity of the side chains of these residues to interact with the receptor residues through aromatic π-π interactions, hydrogen bonding to aromatic π-systems or cation-π-interactions. At physiological pH the side chains of the basic residues, arginine and lysine, will be fully protonated and therefore positively charged. The ammonium or guanidinium side chains of lysine and arginine, respectively, can thus participate in H-bond-to-π interactions as well as cation-π interactions, the latter interactions being common in the interpretation of the histone code by the epigenetic methyl-lysine and methyl-arginine ‘reader’ proteins (section 1.2) which contain aromatic cage structures.160,161        Scheme 5.1 Diazepane 5.1 as a building block in peptide synthesis which may be modified and functionalised further. 60  In order to study the importance of these interactions in a variety of biological targets, we propose the synthesis of a lysine dipeptide mimetic 5.1. We envisioned that 5.1 would be a valuable tool just as the azepinone-scaffolds mimicking phenylalanine (Aba), tyrosine (Hba), tryptophan (Aia) and “histidine” (Ata) have already proven useful in several biological applications (see section 2.6 and appendix 2), and that the synthesis of 5.1 would expand the peptidomimetic toolbox for studying protein interactions with basic residues.  The diazepane 5.1 mimics a lysine side chain, shortened by three methylene units, connected to the neighbouring α-carbon via a methylene bridge on either the C- or N-terminal side of the residue. 5.2a and 5.2b indicate how this would impose constraints in the extending peptide chain, respectively, by highlighting the connectivity with arrows.  In order to incorporate 5.1 as a lysine dipeptide mimetic into a peptide chain, the side chain secondary amine should be Boc-protected and the methyl ester hydrolysed, enabling the use of the building block in standard Fmoc-based SPPS.  The building block may also be converted into an arginine dipeptide mimetic 5.3 by using a guanylation reagent such as N,N′-di-Boc-N′′-triflylguanidine.162 The resulting dipeptide mimetic, 5.3 would have the ε-nitrogen of the arginine side chain tethered to the neighbouring α-carbon similarly using a methylene bridge, but with the side chain shortened by 2 methylene units in comparison to native arginine. In analogue to 5.2b, 5.4 indicates the connectivity, which could also be produced with a C-terminal methylene bridge (in analogue to 5.2a). Interestingly, the guanylation reaction may be done on resin, thus allowing for 5.1 to be guanylated during SPPS if suitable orthogonal protection groups are used.163   Moreover, the scope for the 5.1 building block may be even further extended as the secondary amine may be functionalised to alter potential binding interactions with the target. Alkylations to yield a 5.6a type building block or Buchwald-Hartwig aminations to produce aromatically functionalised amines such as in 5.6b will remove an H-bond donor and furthermore change the steric space around the diazepane ring. Depending on the aryl-substitutent the positive charge may be preserved or removed thus influencing the cation-π interaction potential. Acylation of the secondary amine in 5.1 will result in a teriary amide, 5.6c. This removes an H-bond donor but adds an H-bond acceptor due to the carbonyl functionality. Furthermore the cationic charge is removed. When the R-group in 5.6c is an amine functionality, the resulting dipeptide mimetic will resemble the diazepane version of a citrulline residue (grey box) - a non-essential amino acid produced in the body or converted from arginines in peptide or protein chains by the PAD enzymes (section 1.3).30 61  Thus, with the diazepane 5.1 in hand, we would be able to investigate not only conformational space due to the side chain constraints, but also H-bond as well as cation-π bonding interactions. Alkylation or aromatisation could reveal the steric limitations in the binding pocket of the target. Finally, the secondary amine in 5.1 could be functionalised further by providing a site for extending a peptide chain.     Figure 5.1 shows the overall configuration of the Boc-Nα-protected diazepane building block 5.1 which has been energy minimised in Maestro (v10.1.013, Schrödinger LLT). As can be seen this building block with its (S,S)-stereocenters presents a rather flat conformation with the chains extending in opposite directions (figure 5.1B). Producing the diastereomers (S,R) and (R,S) as well as the enantiomer (R,R) would thus yet again expand the scope of this building block in peptide chemistry, as these rigidified structures would produce bends in the peptide chain which may induce secondary structural motifs such as turns.  The diazepane 5.1 building block would present us with an excellent tool for future peptidomimetic studies.      Figure 5.1 Boc-Nα-protected diazepane building block 3.1 minimised in Maestro Version 10.1.013, MMshare Version 2.9.013 using LigPrep, Schrödinger LLT. A shows the front view and B shows the side view.    62  5.2 Retrosynthetic Routes towards Diazepane 5.1 We developed our synthetic route towards 5.1 using aziridine chemistry, as the following retrosynthetic strategy seemed feasible and facile to us. Furthermore, we had easy access to copious amounts of methyl (S)-1-tritylaziridine-2-carboxylate (5.7), an aziridine synthesised from (S)-serine in 3 steps by Jan Bornholdt, former PhD-student at the department (scheme 5.2).    Scheme 5.2 The trityl-protected aziridine 5.7, synthesised from L-serine, as starting material for the synthesis of 5.1.  In the pursuit of diazepane 5.1 two retrosynthetic strategies using aziridine chemistry can be envisaged (scheme 5.3), an intermolecular (purple route) and an intramolecular aziridine ring-opening (green route).   In the intermolecular ring-opening the final step is the coupling of the primary amine with the carboxylic acid (both in 5.8). The intermediate for the coupling (5.8) is produced by ring-opening of a substituted, activated aziridine (5.10) with the amino acid 2,3-diaminopropionic acid (Dap-OH) protected at the α-nitrogen (5.9). On the contrary, in the intramolecular ring-opening route the aziridine is not reacted until the last step. The intermediate for the intramolecular ring-opening (5.11) is the product of the coupling of Dap-OH protected on both nitrogens (5.12) with the free-base aziridine carboxylate (5.13).    Scheme 5.3 Retrosynthetic routes towards diazepane 5.1 using aziridine chemistry.   63  Both routes require the use of 3 orthogonal protection groups, as a total of three amine functionalities and two carboxylic acids are implicated in the synthesis of a basic dipeptide mimetic. In both routes the methyl ester originating from the aziridine starting material (5.7) is utilised to protect the C-terminal carboxylic acid and should be preserved all through the synthesis. The intermolecular ring-opening route requires an amine protection group as well as a protecting group on the aziridine nitrogen. In the intramolecular ring-opening two orthogonal amine protection groups are required. As these protecting groups all alter the reactivity of the compounds, several combinations of amine protection groups have been tested in the intramolecular route (vide infra). Finally, as the desired dipeptide mimetic (5.1) will be employed as a building block in peptide synthesis it is crucial that the synthesis may be done on a gram-scale in order to be applicable. For this reason both routes were investigated.   5.3 The Chemistry of Aziridines Aziridines are saturated three-membered heterocycles which contain one nitrogen atom. The simplest aziridine, ethylene imine (5.14 with Pg = R = H, Scheme 5.4) is a colourless liquid with a pKa of 7.98. It thus has weaker basicity than alkylamines but stronger than arylamines. It is stable to bases, but reacts violently with acids in an exothermic polymerisation reaction, and is therefore a powerful alkylating reagent in vivo.164  Aziridines are highly strained due to the bond angles in the molecule. The ring strain energy is estimated to be 111 kJ/mol – comparable to that of oxirane, the simplest unsubstituted epoxide. It is due to this strain and the electronegativity of nitrogen that aziridines readily undergo ring-opening under relatively mild conditions.    Scheme 5.4 Nucleophilic attack at C2 or C3 yields two regioisomers with different stereochemistry.  Nucleophiles attack the ring either at C2 or C3 (Scheme 5.4). Most nucleophiles will attack at the less substituted carbon (C3, 5.15), however opening of 2-aryl-substituted aziridines (R = Ar) often shows reversed regioselectivity resulting in the C2 ring-opened 5.16 as the major product.164,165 The ring-opening 64  most often proceeds via an SN2 mechanism and thus will result in inversion of stereochemistry if the reaction proceeds at a stereocenter.165 Thus, attacking the asymmetric aziridine 5.14 in scheme 5.4 at C3 will retain the stereochemistry as no stereocenters are involved in the reaction, whereas an attack at C2 will result in inversion. In both the intermolecular and the intramolecular routes the aziridine is opened by a primary amine which should attack at C3. Generally, amine nucleophiles have been shown to favour the C3 ring-opened products.166-168 As stereochemistry is highly controlled in the ring-opening, aziridines are used as intermediates in a variety of syntheses.164,165 Reacting enantiomerically pure aziridine-2-carboxylates (5.14; R = COOR1) with an appropriate nucleophile provides a method for the synthesis of novel α- and β2-amino acids by reaction at either C3 or C2, respectively.165,166,169    Aziridines are classified as either activated or non-activated.170 Activated aziridines have an electron-withdrawing substituent on the nitrogen (Pg in scheme 5.4), which is able to stabilise the negative charge developing on nitrogen during ring-opening. Most common activating groups are carbonyls, sulfonyls and phosphoryls. In the case of carbonyl-activated aziridines resonance stabilisation in the resulting amide bond will further augment ring-opening.164 The starting material 5.7 which is stable at rt is a non-activated aziridine.   5.4 Previous Efforts This project was embarked upon by M.Sc. student Andoni Fidelereno. Andoni synthesised the compounds 5.17-5.19 (figure 5.2) and attempted to deprotect the side chain amines with subsequent attack and ring-opening of the aziridine. Compounds 5.17 and 5.18 both had the amine side chain allyloxycarbonyl-protected (Alloc-protected). Deprotecting the Alloc-group in 5.17 showed a promising crude NMR, but purification yielded no product. Therefore 5.18 was synthesised, in which the aziridine 5.7 was ester-deprotected and coupled to the esterified Dap-OEt.     Figure 5.2 Aziridine-containing compounds synthesised by Andoni Fidelereno  The Alloc-deprotection of 5.18 yielded the primary amine, and the product could be purified. Unfortunately the resulting compound did not ring-open the aziridine to yield the cyclised product even when heated 65  overnight. Furthermore, Andoni synthesised aziridine 5.19 which had an azide in place of the side chain primary amine. An attempt to unmask this amine by performing a Staudinger reaction using PPh3 was unsuccessful as crude NMR showed no identifiable structures, but rather compound decomposition.   5.5 The Intermolecular Aziridine Ring-Opening Strategy The intermolecular ring-opening route has the advantage that the aziridine is opened in the first step, thus avoiding complications due to handling of the aziridine as well as potential side reactions (Scheme 5.5).   Scheme 5.5 The intermolecular ring-opening route  However, the drawback of this route is the possible separation of the C2 and C3 aziridine-opening products (5.20 and 5.22). As both products contain amine and carboxylic acid functionalities purification by pHPLC is required if the nucleophilic attack during aziridine opening does not favour C3-regioselectivity entirely (5.22). Furthermore, the aziridine protecting/activating-group should be easily cleaved in step 2, without affecting the other protecting groups or the free functionalities. Finally the 7-membered diazepane (5.1) should be formed in a coupling reaction of the primary amine and the carboxylic acid. Formation of the 4-membered ring in an intramolecular reaction between the secondary amine and the carboxylic acid was not considered likely due to ring strain, but in order to avoid intermolecular coupling products involving both the primary and secondary amines the reaction should be run dilute. The success of this route for building block synthesis thus is dependent on the regioselectivity and the side products of the reaction in step 1.    First, the trityl-protected aziridine (5.7) was converted from a non-activated to an activated sulfonyl-protected aziridine (5.10) (Scheme 5.6). 2-nitrobenzenesulfonyl (2-Ns) was chosen as the 66  protecting/activating group for the aziridine nitrogen, as it is easily removed under mild conditions using thiols.171    Scheme 5.6 Conditions for intermolecular aziridine ring-opening  The opening of the aziridine 5.10 using the Boc-Dap-OH (5.9a, scheme 5.6) was attempted in 2:1 CH2Cl2/MeOH and MeCN with the addition of 1.05 eq. DIPEA. In both cases the reactions were heated to 45 oC, but no product could be identified using LC-MS. However, using DMF with 2 eq. DIPEA resulted in full conversion of the starting material after overnight stirring at rt, but resulted in several peaks. Purification by pHPLC was complicated by the fact that 5.23 and 5.24 seemed to co-elute, as two peaks with identical m/z were seen on LC-MS. Yields were quite poor (~10%) due to the many side products and thus optimisation of reaction and purification conditions would be necessary for this route to be feasible in gram-scale synthesis. Protecting the carboxylic acid functionality in Boc-Dap-OH (5.9a) with an orthogonal and easily removed protection group such as a benzyl ester would enable large-scale purification by column chromatography. Changing solvent and aziridine protection/activation group could potentially increase yield and preferred regioselectivity,167,168 but this would require testing of several orthogonal protection-strategies.  We therefore decided to focus our attention on the intramolecular route.  5.6 The Intramolecular Aziridine Ring-Opening Strategy The intramolecular strategy (scheme 5.3, green) has the advantage that coupling of the protected Dap-OH amino acid (5.12) with the aziridine methyl ester (5.13) in the first step (scheme 5.7) activates the aziridine due to acylation, and as the acyl group is part of the final diazepane ring (5.1), removal of an activation/protection group in a later step is avoided. The intramolecular aziridine ring-opening has proven regioselective depending on the ring size.165 In this strategy both the C2- and C3-products (5.21 and 5.1, respectively) produce a 7-membered ring and are allowed according to the Baldwin rules (7-exo-dig and 7-exo-trig, respectively).172 However, as the activated aziridine should be accommodated until ring-opening in the final step, reaction conditions for the deprotection of the amine need to be relatively mild.  67  Furthermore, the free amine of the activated aziridine (5.11) could potentially react before isolation of the compound and thus produce a mixture of products (5.21, 5.11, 5.1).   Acyl aziridines have been shown to rearrange to oxazolines under thermal, acidic or nucleophilic conditions.173-177 Also, the activated aziridine 5.10 was shown to be unstable at temperatures above 60 oC.178 For these reasons all reactions were run at rt and when needed heating was applied with caution. Only non-nucleophilic bases were utilised.     Scheme 5.7 Overview of the intramolecular ring-opening route  5.6.1 The Boc/N3 and Fmoc/N3 Protection Strategies Starting where Andoni left off, we decided to investigate the use of an azide as amine-protection on the Dap-OH side chain, as we expected this could be reduced back to the amine under mild conditions.      68    Scheme 5.8 Synthesis of Boc- and Fmoc-protected Dap(N3)-aziridine methyl esters 5.27a and 5.27b  Both the Boc- and the Fmoc-protected Dap-OH amino acids were azide-protected using the diazotransfer reagent 5.25 (scheme 5.8). Subsequently three coupling reagents were tested for the coupling of 5.26a with 5.13. Yields were observed to increase with the potency of coupling reagent; DCC was slow and yielded no identifiable product, HBTU produced 23% yield and PyBOP 43%. All following couplings were performed with PyBOP as the coupling reagent. Coupling of 5.26a or 5.26b with 5.13 yielded the two azide-protected aziridines, 5.27a and 5.27b.     Scheme 5.9 Reduction of the azide functionality followed by possible cyclisation  Next the azide in 5.27a and 5.27b should be reduced to the corresponding amine. As the free amine functionality may spontaneously ring-open the aziridine, 5.11, 5.1 and 5.21 were all potential products (Scheme 5.9). Several conditions were tested (Table 5.1) and all of the reactions were monitored using IR and worked-up after disappearance of the characteristic N3-stretch.     69  Table 5.1 Reaction conditions tested for 5.27a (Boc-protected) and 5.27b (Fmoc-protected). All reactions were run at rt.  Entry PG Solvent Conditions Reaction Time Conclusion A Fmoc THF P(Bu)3 then H2O 2 h Complex mixture B Boc THF H2 and Pd/C o/n Complex mixture C Boc THF H2 and Pd/C, H+ 3 h Complex mixture D Fmoc THF H2 and Pd/C, H+ 6.5 h Possible degradation during purification  E Fmoc MeOH H2 and Pd/C 3.5 h Small scale reaction; promising 1H NMR F Boc MeOH H2 and Pd/C 3 h Small scale reaction; promising 1H NMR G Fmoc MeOH H2 and Pd/C 2 h Possible degradation during purification H Boc MeOH H2 and Pd/C 2 h Possible degradation during purification   Submitting 5.27b to another Staudinger reagent, P(Bu)3, yielded a complex mixture (entry A) and we turned our attention to the palladium on carbon (Pd/C) catalysed hydrogenation. As an overnight hydrogenation of 5.27a using Pd/C and H2 yielded a complex mixture (entry B), all the following reactions were monitored closely and worked-up immediately after full conversion of starting material.  We suspected that by adding 1 eq. of weak acid we would be able to detect 5.11 as the protonated species avoiding a mixture of the three potential products. For both 5.27a and 5.27b hydrogenation resulted in messy but promising crude 1H NMRs, however, separation of the hydrogenated product mixture of 5.27b by column chromatography yielded only degradation products.     Switching solvent to MeOH proved a good choice, and test reactions with both 5.27a and 5.27b (entries E and F) showed clear signs of aziridine peaks in 1H NMR. It thus seemed that MeOH was a good solvent as it was effective for azide reduction, while being a poor solvent for nucleophilic attack on the aziridine. Thus, both reactions were scaled up (entries G and H). Submitting the Fmoc-protected aziridine (entry G) to column chromatography yielded several products, but none contained all three aziridine signals in 1H NMR, suggesting that this compound may be unstable on silica gel. On the contrary, purification of the Boc-protected aziridine (entry H) provided a promising, but impure 1H NMR with clear aziridine signals. But due to the low yield (<10%) and the problematic purification it was decided to abandon the azide-protection strategy. It should be mentioned that catalytic hydrogenation with Pearlman’s catalyst (Pd(OH)2/C) has been reported to hydrogenate (ie. ring-open) simple aziridines at rt.179 However, as several of the reactions reported herein showed clear signs of aziridine peaks when employing the less active Pd/C catalyst we believe that this was not the source of the encountered problems.  70  5.6.2 The Fmoc/Boc Protection Strategy Next we turned our attention to an Fmoc/Boc protection strategy; using Fmoc-Dap(Boc)-OH (5.28). Fmoc-Dap-OH underwent Boc-protection of the amine side chain and was then coupled to 5.13 and finally purified to yield 5.29 (Scheme 5.10).     Scheme 5.10 Synthesis of the Fmoc-protected Dap(Boc)-aziridine methyl ester (5.29)  The Boc-protected amine in 5.29 was deprotected in dry CH2Cl2 with 10% trifluoroacetic acid (TFA; v/v%). In general this reaction lead to two products according to LC-MS, here referred to as peak A and peak B, which were submitted to full NMR-analysis (utilising 1H NMR, 13C NMR, COSY, DEPT-135, HSQC and HMBC). Peak B was identified as the protected dipeptide Fmoc-Dap-Ser-OMe, a by-product most likely formed by reaction with residual water under the reaction conditions.       Scheme 5.11 Reactions with Boc-deprotection in TFA produced the dipeptide (peak B) from the putative intermediate, peak A.   It was not possible to determine the identity of peak A. However, several experiments indicated that peak A was converted to peak B in the presence of TFA. Potentially peak A represented the Boc-deprotected version of 5.29 that could be encouraged to cyclise to produce the desired 7-membered ring structure 5.1. Various conditions were tested to promote the desired reaction; replacing the CH2Cl2/TFA solvent after Boc-protection with pure dry CH2Cl2, neutralising the TFA with a dry DIPEA, heating to 50 oC in dry CH2Cl2 in a sealed vial, changing solvent to dry DMF and heating to 45 oC. None of these conditions were productive; generally peak A was the only product. Thus we decided to abandon the Fmoc/Boc strategy.       71  5.6.3 The 2-Nitrosulfonylbenzene (2-Ns) Protection Strategy In keeping with previous aziridine synthesis the side chain amine of the Nα-protected Dap-OH amino acid was protected with 2-Ns (5.30a or 5.30b) and then coupled to the aziridine 5.13. Both the Fmoc- and the Boc-versions were purified by column chromatography to yield 5.31a and 5.31b (Scheme 5.12).    Scheme 5.12 Synthesis of the Fmoc- and Boc-protected Dap(2-Ns)-aziridine methyl esters (5.31a and 5.31b)  Protecting the side chain amine with 2-Ns has the advantage of not needing removal in order for the amine to function as a nucleophile, as relatively weak bases can deprotonate the protected amine which may then attack the aziridine (scheme 5.13). Moreover, we anticipated that the bulkiness of the 2-Ns would favour C3-regioselectivity in the aziridine-opening.   Scheme 5.13 Reaction of 5.31 with base  5.6.3.1 The Fmoc/2-Ns Protection Strategy However, this strategy is not compatible with the Fmoc-protected 5.31b, as Fmoc is base-labile. Instead we were hoping that 5.31b would ring-close by simply heating the compound in an appropriate solvent. Thus compound 5.31b was heated to 45 oC in both DMF and DMSO and the reaction was monitored by aHPLC. After 48 hr of heating in DMF 95% of the starting material had been converted, but only undesired degradation products were detected. On the other hand the reaction in DMSO produced a new product eluting close to the starting material. The NMR spectrum of the product seemed promising but did however display a few signals which could not be assigned to the desired product, the 2-Ns-protected 5.1. Moreover 72  the reaction in DMSO only resulted in conversion of 15% of the starting material after 24 hr, thus further investigation into this route was abandoned.    5.6.3.2 The Boc/2-Ns Protection Strategy As mentioned previously, we were anticipating the 2-Ns protected amine in 5.31a to ring-open the aziridine regioselectively at C3 when treated with a non-nucleophilic base. Several bases, temperatures and two solvents were screened (table 5.2) and reactions were monitored by aHPLC.   Table 5.2 Test reactions of aziridine 5.31a with various bases, temperatures and dry solvents. Entry Base Solvent Temperature Reaction time Conclusion 1 Cs2CO3 (2 eq) DMF 75 oC <2 h Degradation 2 DBU (1.2 eq) DMF 45 oC 6 h Product pattern A 3 DBU (1.1 eq) DMSO rt 5 days Product pattern A 4 Molecular sieves DMF rt > 6 days Product pattern A 5 Molecular sieves DMSO 45 oC 48 h Product pattern A 6 DIPEA (1.1 eq) DMSO 45 oC > 5 days Product pattern B 7 Cs2CO3 (1.5 eq) DMF 45 oC > 6 days Product pattern B 8 Cs2CO3 (1.5 eq) DMSO 45 oC > 6 days Product pattern B 9 K2CO3 (1.1 eq) DMSO 45 oC 5 days Product pattern B   Ring-opening of acyl-activated aziridines with an amine nucleophile in DMF at 75 oC has been reported previously180 and this was attempted in this study (entry 1). However in our hands it resulted in rapid degradation, highlighting the instability of acyl aziridines at temperatures > 50 oC. The reaction in entry 2 was started at 45 oC, whereas the reactions in entries 3-9 were stirred overnight at rt followed by heating to 45 oC if no reaction had occured. All reactions were stopped when less than 5% starting material remained, with the exception of entries 4 and 6-8 that were terminated when reaction progress ceased. The non-nucleophilic bases DBU (entries 2, 3) and DIPEA (entry 6) were tested along with cesium carbonate (entries 1, 7, 8) and potassium carbonate (entry 9). Moreover, it has been reported by Monfregola et al. that 2-Ns-protected amines can be deprotonated by molecular sieves (4Å) in DMF181 and this was likewise tested in both solvents (entries 4, 5).  73   Figure 5.3 Product patterns A and B when 5.31a is reacted with base, exemplified by entries 3 and 8, respectively.   In general all reactions initially produced a large peak eluting around 14 min in aHPLC (Method E, figure 5.3). This was identified as a degradation product containing the 2-Ns-group, and in most of the reactions this product dominated. All reactions resulted in 3-8 major products with clear product patterns. Figure 5.3 shows a representative product pattern of each type. The red arrows indicate the starting material and the blue/green arrows the peaks that were significant for each product pattern. Interestingly, the weaker bases, DIPEA and carbonate (entries 6-9, product pattern B) had a very different product pattern to the more basic DBU as well as molecular sieves (entries 2-5, product pattern A). Purification of entry 2 yielded a product (green arrow) which by full NMR-analysis resembled the expected diazepane product, but as with the Fmoc/2-Ns strategy contained shifts which could not be assigned. Having tested several conditions which all resulted in complex mixtures and a dominating degradation product, we decided to abandon the Boc/2-Ns-strategy.         74  5.7 Conclusion to the Aziridine Project Unfortunately we were unable to synthesise the desired diazepane 5.1 using aziridine chemistry (scheme 5.14). The many orthogonal protection group strategies in addition to the complex reactivity of aziridines made this endeavour quite troublesome. These investigations were by no means exhaustive, but represent the strategies we thought viable after getting acquainted with the chemistry of activated aziridines.     Scheme 5.14 Overview of the tested routes for the synthesis of 5.1.  The acyl-activated aziridines in the intramolecular ring-opening route were difficult to handle, and flash chromatographic purification of products containing aziridines showed slow but significant degradation visualised by thin-layer chromatography (TLC). The investigations involving deprotection of the activated aziridine compounds in this project showed that low heat, acid, non-nucleophilic bases and mild hydrogenation all resulted in degraded compounds as the major products (scheme 5.14). The intermolecular ring-opening route seemed to be inefficient for large scale synthesis. We thus concluded that the large scale synthesis of 5.1 could not be achieved in our hands using an aziridine strategy.      75  5.8 Alternative Synthesis Routes to Diazepane 5.1  Two alternative and feasible synthesis strategies towards 5.1 may be envisioned (scheme 5.15). Once again these rely on an amide coupling and the use of an amine as a nucleophile.     Scheme 5.15 Alternative retrosynthetic routes to 5.1  In the purple synthesis route the final step is an intramolecular coupling of an amine and a carboxylic acid (5.32). This is preceded by an intermolecular attack of a primary amine (5.9b) on a Boc-protected serine-methyl ester with the alcohol functionality substituted for a good leaving group (5.33). Mesylates/tosylates are excellent leaving groups used with amine nucleophiles as they are not acidic. Following this nucleophilic attack is the deprotection of the Boc-protected amine with subsequent intramolecular coupling  in 5.32. Alternatively, the nucleophilic substitution may be the final step in the synthesis of 5.1 as outlined in the green route. This synthesis starts with the coupling of two tert-Butyl-protected serines; one Nα-Fmoc protected (5.34), and one with an ester protection (5.35). After the intermolecular coupling the side chains are deprotected to yield 5.36 and a suitable primary amine-nucleophile (5.37) is used to form the ring by double nucleophilic attack on the leaving-group activated side chains. As highlighted in the grey box, benzylic moieties may be used for the amine nucleophile (5.39) as well as the amide functionality (5.38). The bulkiness of the benzyl will hopefully force the two side chains in 5.38 into close proximity on the same side of the amide bond, thus facilitating reaction. Subsequently the benzyl groups may be removed by mild hydrogenation resulting in the final building block 5.1.  It is our intention to investigate these strategies towards synthesis of diazepane 5.1 in the future.76  5.9 EXPERIMENTAL SECTION 5.9.1 GENERAL INFORMATION All reactions run under dry conditions were carried out in flame-dried glassware which was evacuated prior to running the reaction under a nitrogen (N2) atmosphere. Dry CH2Cl2, THF, DMF were dried using an SG Water solvent purification system. DMSO was dried using 3 Å molecular sieves from VWR. Amino acids Boc-Dap-OH (5.9a) and Fmoc-Dap-OH (5.9b) were purchased from Chem-Impex International Inc., Wood Dale, Illinois, USA and methyl-(S)-1-tritylaziridine-2-carboxylate (5.7) was kindly synthesised by Jan Bornholt. The diazotransfer reagent 5.25 was kindly synthesised by former PhD student Tue Heesgaard Jepsen. All other reagents and solvents were of analytical grade or higher purity. Sulfate buffer was made of 0.75 M Na2SO4 and 0.25 M H2SO4 in deionised H2O.  Flash column chromatography was performed using silica gel 60 (40–63 µm) from Merck.   5.9.2 ANALYTICAL METHODS AND EQUIPMENT 5.9.2.1 Thin-Layer Chromatography (TLC) TLC was carried out on pre-coated silica gel 60 F254 plates from Merck and visualised using UV (254 nm) and ninhydrin stain (1.5 g in 100 mL EtOH with 3 mL AcOH).   5.9.2.2 Infrared Spectroscopy (IR) IR spectroscopy was recorded on a Perkin-Elmer Spectrum One IR spectrometer using Spectrum One version 3.02 software. All compounds were dissolved in CH2Cl2 or CHCl3 and the solvent was allowed to evaporate before recording the spectrum. Signals (νmax) are reported in wavenumbers (cm-1).   5.9.2.3 Nuclear Magnetic Resonance Spectroscopy (NMR) NMR spectra were recorded on either an Ascend 400 or a 600 MHz Ultrashield Bruker instrument. Samples were dissolved in deuterated solvents (CDCl3, DMSO-d6, MeCN-d3, D2O or MeOD) purchased from Cambridge Isotope Laboratories or VWR, all with a purity of >99.8%. The FID-files were obtained and processed using standard parameters in Topspin v3.2 and analysed using the MestReNova software version 6.0. Signals are reported in ppm (δ) using the solvent as reference. Signal assignment was based on unambiguous chemical shifts in combination with DEPT-135, and 2D-methods such as COSY, HSQC and HMBC experiments. Coupling constants (J) are reported in Hertz (Hz) and rounded to the nearest 0.5 Hz.      77  5.9.2.4 Analytical High Performance Liquid Chromatography (aHPLC) Method E:  Analytical HPLC was performed on a Dionex UltiMate 3000 system equipped with a photodiode array detector and a RP Phenomenex Gemini NX-C18 column (250 x 4.6 mm, 3 μM) with eluents A (100% H2O + 0.1% TFA) and B (90% MeCN, 10% H2O + 0.1% TFA) with a flowrate of 1 mL/min.  Reported retention times (tR) were obtained using the following gradient: 0% B 0-5min, 0-100% B 5-20min, 100% B 20-25min.   5.9.2.5 Low Resolution Mass Spectrometry (LC-LRMS or LC-MS) Method F:  LC-LRMS was performed on an Agilent 1200 series LC equipped with a diode array detector coupled to a Bruker Esquire3000plus ion trap mass spectrometer. The column (Waters Eclipse XDB-C18, 4.6 × 100 mm, 5 µm) was kept at 45°C. Ionisation of the eluting compounds was obtained by ESI in the positive mode with a nebulising pressure of 50 psi, a dry gas of 10 L/min kept at 240°C and a capillary potential of -2750 V. Eluents used: A (95% H2O, 5% MeCN + 0.1% FA) and B (95% MeCN, 5% H2O + 0.1% FA) with a flowrate of 0.5 mL/min. Program: 0% B 0-5min, 0-100% B 5-25min, 100% B 20-25min.   5.9.2.6 High Resolution Mass Spectrometry/Exact mass (HRMS) Performed by Christian Janfeldt, Department of Pharmacy, UCPH Accurate mass analysis was performed in positive ion mode with MALDI ionisation on a Thermo Q Exactive Orbitrap mass spectrometer (Thermo Scientific, Bremen, Germany) equipped with an AP-SMALDI 10 ion source (TransmitMIT, Giessen, Germany) and operated at mass resolving power 140,000 @ m/z 200. DHB was used as matrix and lock-mass for internal mass calibration, providing mass accuracy of 3 ppm or better.  5.9.2.7 Preparative Reverse-Phase High Performance Liquid Chromatography (pHPLC) Method D: Preparative RP HPLC was carried out on a Dionex UltiMate 3000 HPLC with a diode array detector using a preparative RP Phenomenex Gemini NX-C18 column (250 x 21.20 mm, 5 μM) using eluent A (100% H2O + 0.1% TFA) and eluent B (90% MeCN, 10% H2O + 0.1% TFA) and a flowrate of 20 mL/min.  Program: 0% B 0-5 min, 0-100% B 5-35 min, 100% B 35-40 min.     78  5.9.3 GENERAL PROCEDURE 5.9.3.1 General coupling procedure The appropriate 2,3-diaminopropionic acid with both amine functionalities protected (1.0 eq)  was dissolved in CH2Cl2 (50 mL/ 1 g 2,3-diaminopropionic acid), followed by the addition of PyBOP (1.0 eq) and methyl aziridine-2-carboxylate (5.13, ~0.75 eq for 5.27a and 5.27b; ~0.83 eq for 5.29; 1.0 eq for 5.31a and 5.31b) and finally DIPEA (1.3 eq) was added dropwise. The reaction was stirred for 15-24 hr. The reaction was diluted 3-fold with CH2Cl2, washed with 2x 3-fold sulphate buffer, 2x 3-fold NaHCO3 (sat), 1x 3-fold brine, dried with Na2SO4, filtered and concentrated in vacuo. Purified as specified for each compound.      5.9.4 COMPOUND CHARACTERISATIONS Methyl (S)-1-((2-nitrophenyl)sulfonyl)aziridine-2-carboxylate (5.10)    By an adaptation of the published procedure178 aziridine 5.7 (2.00 g, 5.84 mmol) was dissolved in MeOH/CHCl3 (1:1, 10 mL) and cooled to 0 oC. TFA (7.2 mL, 94.1 mmol) was added dropwise to give a yellow/green solution. The reaction was left to stir for 3.5 hr and then concentrated in vacuo. (3 x 15 mL) Et2O was added with ensuing in vacuo concentration. The residue was dissolved in H2O (120 mL) and Et2O (120 mL) and the Et2O phase was extracted with H2O (80 mL). The aqueous phase was basified with NaHCO3 (s) and EtOAc (225 mL) was added. The biphasic reaction mixture was cooled to 0 oC and 2-nitrobenzenesulfonyl chloride (1.30 g, 5.86 mmol) was added and the reaction was stirred vigorously overnight at rt. The phases were separated and the aqueous phase extracted with EtOAc (120 mL). The organic phases were combined, dried with Na2SO4, filtered and concentrated in vacuo. Purification by flash column chromatography (2:1 n-heptane/EtOAc, v/v) gave nosylated 5.10 (0.72 mg, 43%) as a clear colourless oil.  1H NMR (400 MHz; CDCl3) δ 8.17 (d, J = 7 Hz, 1H, Ar), 7.77 – 7.65 (m, 3H, Ar), 3.73 (s, 3H, CH3), 3.54 (dd, J = 7, 4.5 Hz, 1H, aziridine-CH), 3.00 (d, J = 7 Hz, 1H, aziridine-CHaHb), 2.72 (d, J = 4.5 Hz, 1H, aziridine-CHaHb). 13C NMR (101 MHz; CDCl3) δ 167.1 (C=O), 148.5 (Ar), 135.1 (Ar), 132.7 (Ar), 131.7 (Ar), 131.6 (Ar), 124.8 (Ar), 53.1 (CH3), 37.7 (aziridine-CH), 34.3 (aziridine-CH2). IR νmax(CH2Cl2)/cm-1 3099 (C-H, sp2), 2958 (C-H, sp3) 1748 (C=O ester), 1543 (NO2), 1343 (NO2), 1167 (S=O).  All analytical data was consistent with that previously reported.182   79  Methyl (S)-aziridine-2-carboxylate (5.13) By an adaptation of the published procedure,178 aziridine 5.7 (2.0 g, 5.84 mmol) was dissolved in MeOH/CHCl3 (1:1, 10 mL) and cooled to 0 oC. TFA (7.2 mL, 94.1 mmol) was added dropwise to give a yellow/green solution. The reaction was left to stir for 3.5 hr and then concentrated in vacuo. Et2O (3 x 15 mL) was added with ensuing in vacuo concentration. The residue was dissolved in H2O (120 mL) and Et2O (120 mL) and the Et2O phase was extracted with H2O (80 mL). The aqueous phase was basified with NaHCO3 (s) and extracted with CH2Cl2 (3 x 100 mL). The combined organic phases were dried with Na2SO4, filtered and concentrated in vacuo to give 5.13 (0.41 g, 69%) as a yellow oil, which required no further purification.  1H NMR (400 MHz, CDCl3) δ 3.77 (s, 3H, CH3), 2.54 (dd, J = 5.5, 3 Hz, 1H, Aziridine-CH), 2.01 (dd, J = 3, 1.5 Hz, 1H, Aziridine-CHaHb), 1.88 (d, J = 5.5 Hz, 1H, Aziridine-CHaHb).  13C NMR (101 MHz, CDCl3) δ 173.6 (C=O), 52.7 (CH3), 29 (Aziridine-CH), 27.4 (Aziridine-CH2).  IR νmax(CH2Cl2)/cm-1 3290 (N-H), 3001 (C-H), 2956 (C-H, sp3), 1728 (C=O). 1H NMR of the compound was consistent with that previously reported in the literature.183    3-Azido-N-(tert-butoxycarbonyl)-L-alanine (5.26a)  Diaminopropionic acid 5.9a (1.00 g, 4.91 mmol) was dissolved in MeOH/H2O (4:1 v/v, 25 mL). CuSO4 x 5 H2O (0.2 g, 0.80 mmol), NaHCO3 (1.43 g, 17.1 mmol) and imidazole-1-sulfonyl azide x HCl (1.25 g, 5.93 mmol) were added and the pH was adjusted to 9 with sat. NaHCO3. After stirring for 15 hr the solution was diluted with sulfate buffer (75 mL) and extracted with EtOAc (3 x 50 mL). The combined organic phases were washed with brine (2 x 50 mL), dried with Na2SO4, filtered and concentrated in vacuo to give the crude azide 5.26a (1.46 g, 129%) as a white foam. The product was used with no further purification.  1H NMR (400 MHz, CD3OD) δ 4.31 (t, J = 5 Hz, 1H, CH), 3.65 – 3.63 (m, 2H, CH2), 1.46 (s, 9H, C-(CH3)3). 13C NMR (101 MHz, CD3OD) δ 173.0 (COOH), 157.7 (NH-(C=O)-O), 80.9 (C-(CH3)3), 56.4 (CH2), 53.2 (CH), 28.7 (C-(CH3)3). IR νmax(CH2Cl2)/cm-1 3281 (br, O-H), 2981 (C-H, sp3), 2106 (N3), 1693 (C=O). All analytical data was consistent with that previously reported in the literature.184   80  Methyl (S)-1-(3-azido-2-((tert-butoxycarbonyl)amino)propanoyl) aziridine-2-carboxylate (5.27a)  Synthesised and worked-up according to the general coupling procedure using amino acid 5.26a (0.54 g, 2.35 mmol) and aziridine 5.13 (174 mg, 1.73 mmol). The product was purified by flash column chromatography (3:2 n-heptane/EtOAc, v/v) to afford aziridine 5.27a (0.23 g, 43%) as a clear oil.  TLC Rf = 0.25 (3:2 n-heptane/EtOAc, v/v). 1H NMR (400 MHz, CDCl3) δ 5.37 (d, J = 7.5 Hz, 1H, CO-NH-CH), 4.47 – 4.41 (m, 1H, NH-CH), 3.79 (s, 3H, CH3), 3.74 (dd, J = 6, 5 Hz, 2H, CH2), 3.27 (dd, J = 6, 3 Hz, 1H, Aziridine-CH), 2.65 (dd, J = 3, 1.5 Hz, 1H, Aziridine-CHaHb), 2.62 (dd, J = 6, 1.5 Hz, 1H, Aziridine-CHaHb), 1.45 (s, 9H, (CH3)3). 13C NMR (101 MHz, CDCl3) δ 180.1 (CαH-(C=O)), 168.4 (COOCH3), 155.2 (NH-(C=O)), 80.7 (C(CH3)3), 55.4 (NH-CαH), 53.1 (CH3), 52.9 (CH2), 34.9 (Aziridine-CH), 30.8 (Aziridine-CH2), 28.4 ((CH3)3). IR νmax(CH2Cl2)/cm-1 3220 (N-H), 2970 (C-H, sp3), 2870 (C-H, sp3), 2100 (N3), 1745 (C=O, ester), 1707 (C=O carbamate), 1646 (C=O, tertiary amide). HPLC tR = 17.19 min (Method E). m/z: Compound could not ionise.   3-Azido-N-(fluorenylmethyloxycarbonyl)-L-alanine (5.26b) Diaminopropionic acid 5.9b (1.45 g, 4.44 mmol) was dissolved in MeOH/H2O (4:1 v/v 25 mL). CuSO4 x 5 H2O (0.18 g, 0.71 mmol), NaHCO3 (1.31 g, 15.5 mmol) and imidazole-1-sulfonyl azide x HCl (1.12 g, 5.33 mmol) were added and the pH was adjusted to 9 with sat. NaHCO3. Work-up procedure was equivalent to that of 5.26a and gave a crude azide 5.26b (1.57 g, 100%) as a white foam. The product was used with no further purification. 1H NMR (400 MHz, CD3OD) δ 7.79 (d, J = 7.5 Hz, 2H, Ar), 7.66 (dd, J = 9.5, 6 Hz, 2H, Ar-H), 7.38 (t, J = 7.5 Hz, 2H, Ar-H), 7.30 (t, J = 7.5 Hz, 2H, Ar-H), 4.43 – 4.39 (m, 1H, CαH), 4.36 (d, J = 7.0 Hz, 2H, CH-CH2-O), 4.24 (t, J = 7 Hz, 1H, CH-CH2-O), 3.69 (d, J = 6 Hz, 2H, CH2-N3). 13C NMR (101 MHz, CD3OD) δ 172.7 (COOH), 158.5 ((C=O)-NH), 145.3 (Ar), 142.6 (Ar), 128.8 (Ar), 128.2 (Ar), 126.3 (Ar), 120.9 (Ar), 68.2 (CH-CH2-O), 56.4 (CH-CH2-O), 55.4 (CαH), 53.1 (CH2-N3). IR νmax(CH2Cl2)/cm-1 3322 (O-H/N-H), 3067 (C-H), 2955 (C-H), 2107 (N3), 1706 (C=O). HPLC tR = 18.74 min (Method E). m/z (+ESI) found: MNa+, 375.1066. (C18H16N4O4Na requires M, 375.1063) 1H and 13C NMR were consistent with that previously reported in the literature, other data not reported.185    81  Methyl (S)-1-(3-azido-2-((Fluorenylmethyloxycarbonyl)amino)propanoyl) aziridine-2-carboxylate (5.27b)  Synthesised and worked-up according to the general coupling procedure using amino acid 5.26b (0.81 g, 2.30 mmol) and aziridine 5.13 (170 mg, 1.68 mmol). The product was purified by flash column chromatography (1:1 n-heptane/EtOAc, v/v) to give aziridine 5.27b (0.28 g, 38%) as a white solid. 1H NMR (400 MHz, CDCl3) δ 7.77 (d, J = 7.5 Hz, 2H, Ar), 7.60 (d, J = 7 Hz, 2H, Ar), 7.41 (t, J = 7.5 Hz, 2H, Ar), 7.32 (t, J = 7.5 Hz, 2H, Ar), 5.71 (d, J = 7.5 Hz, 1H, CO-NH-CH), 4.55 – 4.48 (m, 1H, CαH), 4.48 – 4.34 (m, 2H, CH-CH2-O), 4.22 (t, J = 7 Hz, 1H, CH-CH2-O), 3.78 (t, J = 5 Hz, 2H, CH2-N3), 3.74 (s, 3H, CH3), 3.32 – 3.25 (m, 1H, (Aziridine-CH)), 2.68-2.62 (m, 1H, (Aziridine-CHaHb)), 2.59 (d, J = 5.5 Hz, 1H, (Aziridine-CHaHb)).  13C NMR (101 MHz, CDCl3) δ 179.6 (CαH -(C=O)), 168.4 (COOCH3), 155.9 ((C=O)-NH), 143.8 (Ar), 141.4 (Ar), 127.9 (Ar), 127.3 (Ar), 125.2 (Ar), 120.2 (Ar), 67.5 (CH-CH2-O), 55.8 (CαH), 53.1 (CH3), 52.8 (CH2-N3), 47.2 (CH-CH2-O), 35.1 (Aziridine-CH), 30.7 (Aziridine-CHaHb). IR νmax(CH2Cl2)/cm-1 3333 (N-H), 2954 (C-H, sp3), 2105 (N3), 1705 (br, C=O). TLC Rf = 0.33 (1:1 n-heptane/EtOAc, v/v). HPLC tR = 19.58 min (Method E). m/z (+ESI) found: MNa+, 458.1434. (C22H21N5O5Na requires M, 458.1434).    (S)-2-FIuorenyl-9-methoxycarbonyl-3-(N-tert-butoxycarbonyl) propionic acid (5.28)   By an adaptation of the published procedure,186 diaminopropionic acid 5.9b (0.25 g, 0.77 mmol) was dissolved in DMF (2.5 mL). Boc2O (0.34 g, 1.57 mmol) and DIPEA (267 μL, 1.53 mmol) were added and the reaction was stirred for 15 hr. EtOAc (40 mL) was added and the organic phase was washed with sulfate buffer (30 mL), H2O (20 mL), dried over MgSO4, filtered and concentrated in vacuo. The solid was triturated with heptane to yield amino acid 5.28 (0.31 g, 95%) as a white sticky residue.  1H NMR (400 MHz, CDCl3) δ 9.37 (br s, 1H, OH), 7.75 (d, J = 7.5 Hz, 2H, Ar), 7.59 (d, J = 5.5 Hz, 2H, Ar), 7.39 (t, J = 7.5 Hz, 2H, Ar), 7.29 (t, J = 7.5 Hz, 2H, Ar), 4.58 – 4.43 (m, 1H, CαH), 4.43 – 4.28 (m, 2H, CH-CH2-O), 4.22 (t, J = 6 Hz, 1H, CH-CH2-O), 3.72 – 3.60 (m, 1H, CHaHb-NH), 3.60 – 3.49 (m, 1H, CHaHb-NH), 1.44 (s, 9H, (CH3)3).13C NMR (101 MHz, CDCl3) δ 173.0 (COOH), 163.3 (CO-NH), 146.9 (CO-NH), 143.9 (Ar), 141.4 (Ar), 127.9 (Ar), 127.2 (Ar), 125.3 (Ar), 120.1 (Ar), 85.3 (C(CH3)3), 47.2 (CH-CH2-O), 36.9 (CαH), 31.8 (CH-CH2-O), 28.4 (CH2-NH), 27.6 ((CH3)3). Analytical data was comparable to that previously reported.187  82   Methyl (S)-1-(3-(tert-butoxycarbonyl)-2-((fluorenylmethyloxycarbonyl) amino)propanoyl) aziridine-2-carboxylate (5.29)   Synthesised and worked-up according to the general coupling procedure using amino acid 5.28 (0.29 g, 0.68 mmol) and aziridine 5.13 (56 mg, 0.56 mmol). The product was purified by flash column chromatography (1:1 n-heptane/EtOAc, v/v) to afford aziridine 5.29 (0.18 g, 63%) as a white solid. TLC Rf = 0.28 (1:1 n-heptane/EtOAc, v/v). 1H NMR (600 MHz, CDCl3) δ 7.76 (d, J = 7.5 Hz, 2H, Ar), 7.62 – 7.58 (m, 2H, Ar), 7.40 (t, J = 7.5 Hz, 2H, Ar), 7.31 (t, J = 7.5 Hz, 2H, Ar), 6.11 (d, J = 4.5 Hz, 1H, CO-NH-CαH), 4.98 – 4.92 (m, 1H, CH2-NH-CO), 4.45 – 4.39 (m, 1H, CαH), 4.36 (d, J = 7 Hz, 2H, CH-CH2-O), 4.21 (t, J = 7 Hz, 1H, CH-CH2-O), 3.74 (s, 3H, CH3), 3.72 – 3.64 (m, 1H, CHaHb-NH), 3.60 – 3.52 (m, 1H, CHaHb-NH), 3.35 – 3.31 (m, 1H, aziridine-CH), 2.70 – 2.66 (m, 1H, aziridine-CHaHb), 2.64 – 2.60 (m, 1H, aziridine-CHaHb), 1.45 (s, 9H, (CH3)3). 13C NMR (151 MHz, CDCl3) δ 180.7 (CαH-(C=O)), 168.7 (COOCH3), 157.0 ((C=O)-NH), 156.4 ((C=O)-NH), 143.9 (Ar), 141.4 (Ar), 127.9 (Ar), 127.2 (Ar), 125.3 (Ar), 120.1 (Ar), 80.4 (C(CH3)3), 67.4 (CH-CH2-O), 57.4 (CαH), 53.0 (CH3), 47.3 (CH-CH2-O), 42.6 (CH2-NH), 34.6 (Aziridine-CH), 30.8 (Aziridine-CHaHb), 28.4 ((CH3)3). IR νmax(CH2Cl2)/cm-1 3355 (N-H), 2978 (C-H), 1696 (br, C=O). HPLC tR = 20.03 min (Method E). m/z (+ESI) found: MNa+, 532.2068. (C27H31N3O7Na requires M, 532.2054).   (2S)-2-((tert-Butoxycarbonyl)amino)-3-(((2-nitrophenyl)sulfonyl)amino)  propionic acid (5.30a) By an adaptation of the published procedure,188 diaminopropionic acid 5.9a (0.50 g, 2.46 mmol) was dissolved in a K2CO3-solution (10%, 4.5 mL) and cooled to 0 oC.  2-nitrobenzenesulfonyl chloride (0.52 g, 2.35 mmol) was dissolved in dioxane (150 μL) and added over 10 min. The solution was stirred for 15 hr at rt. To the solution was added Et2O (80 mL) and H2O (40 mL) and the phases were separated. The organic phase was extracted with NaHCO3 (sat, 2 x 30 mL), the combined aqueous phases were acidified with 6M HCl and extracted with EtOAc (3 x 100 mL). The organic phases were dried with MgSO4, filtered and concentrated in vacuo to give amino acid 5.30a (704 mg, 77%) as a clear oil. The product was used with no further purification. 1H NMR (600 MHz, CDCl3) δ 8.27 (br s, 1H, OH), 8.13 – 8.08 (m, 1H, Ar), 7.86 – 7.80 (m, 1H, Ar), 7.77 – 7.68 (m, 2H, Ar), 6.17 – 6.00 (m, 1H, NH-SO2), 5.58 (d, J = 6 Hz, 1H, (C=O)-NH), 4.43 – 4.28 (m, 1H, CαH), 3.63 – 3.49 (m, 2H, CH2), 1.43 (s, 9H, (CH3)3). 13C NMR (151 MHz, CDCl3) δ 173.3 (COOH), 156.0 ((C=O)-NH), 148.0 (Ar), 134 (Ar), 133.5 (Ar), 133.2 (Ar), 131.2 (Ar), 125.6 (Ar), 81.2 (C(CH3)3), 53.5 (CαH), 44.8 (CH2), 28.4 83  ((CH3)3).iv IR νmax(CH2Cl2)/cm-1 3333 (N-H), 2980 (C-H, sp3), 1695 (br, C=O), 1540 (NO2), 1366 (NO2), 1162 (S=O).    Methyl (S)-1-(3-(((2-nitrophenyl)sulfonyl)amino))-2-((tert-Butoxycarbonyl) amino)propanoyl) aziridine-2-carboxylate (5.31a)   Synthesised and worked-up according to the general coupling procedure using amino acid 5.30a (0.22 g, 1.218 mmol) and aziridine 5.13 (123 mg, 1.218 mmol). The product was purified by flash column chromatography (45% n-heptane in EtOAc) to give aziridine 5.31a (0.14 g, 51%) as a clear oil. TLC Rf = 0.38 (2:3 n-heptane/EtOAc, v/v). 1H NMR (400 MHz, CDCl3) δ 8.14 – 8.08 (m, 1H, Ar), 7.88 – 7.83 (m, 1H, Ar), 7.76 – 7.70 (m, 2H, Ar), 6.03 (t, J = 6.5 Hz, 1H, CH2-NH-SO2), 5.40 (d, J = 7.5 Hz, 1H, (C=O)-NH), 4.44 – 4.35 (m, 1H, CαH), 3.76 (s, 3H, CH3), 3.58 (ddd, J = 12, 7, 5 Hz, 1H, CHaHb-NH-SO2), 3.45 (dt, J = 13.5, 5.5 Hz, 1H, CHaHb-NH-SO2), 3.25 (dd, J = 5, 3.5 Hz, 1H, aziridine-CH), 2.65-2.59 (m, 2H, aziridine-CH2), 1.41 (s, 9H, ((CH3)3). 13C NMR (101 MHz, CDCl3) δ 180.0 (CαH-(C=O)), 168.5 (COOCH3), 155.4 ((C=O)-NH), 148.1 (Ar-C-NO2), 133.9 (Ar), 133.6 (Ar), 133.0 (Ar), 131.1 (Ar), 125.6 (Ar), 80.7 (C(CH3)3), 55.0 (CαH), 53.1 (CH3), 45.4 (CH2-NH-SO2), 34.9 (Aziridine-CH), 30.8 (Aziridine-CH2), 28.3 (C(CH3)3). IR νmax(CH2Cl2)/cm-1 3351 (N-H), 2984 (C-H), 1742 (C=O ester), 1704 (br, C=O), 1541 (NO2), 1366 (NO2), 1165 (S=O). HPLC tR = 17.76 min (Method E). m/z (+ESI) found: MNa+, 495.1134. (C18H24N4O9SNa requires M, 495.1156).   (2S)-2-((Fluorenylmethyloxycarbonyl)amino)-3-(((2-nitrophenyl)sulfonyl)amino) propionic acid (5.30b)    By an adaptation of the published procedure,188 similar to the procedure of 5.30a using diaminopropionic acid 5.9b (0.25 g, 0.77 mmol) and 2-nitrobenzenesulfonyl chloride (0.16 g, 0.73 mmol). The reaction yielded amino acid 5.30b (0.20 g, 56%) as a slightly yellow solid. The product was used with no further purification.  1H NMR (400 MHz, DMSO) δ 12.90 (br s, 1H, OH), 8.18 – 8.10 (m, 1H, O-NH-CH), 8.02 – 7.99 (m, 1H, Ar), 7.99 – 7.95 (m, 1H, Ar), 7.87 – 7.82 (m, 2H, Ar), 7.71 (d, J = 7.5 Hz, 2H, Ar), 7.58 (d, J = 8.5 Hz, 2H, Ar), 7.42 (t, J = 7.5 Hz, 2H, Ar), 7.33 (t, J = 7.5 Hz, 2H, Ar), 4.30 – 4.25 (m, 2H, CH-CH2-O), 4.24 – 4.19 (m, 1H, CH-CH2-O), 4.12 (td, J = 8, 5 Hz, 1H, CαH), 3.36 – 3.27 (m, 2H, CH-CH2-NH). 13C NMR (101 MHz, DMSO) δ 171.3                                                           iv NMR not reported in the literature. 84  (COOH), 155.8 ((C=O)-NH), 147.6 (Ar), 143.7 (Ar), 140.7 (Ar), 134.1 (Ar), 132.8 (Ar), 132.6 (Ar), 129.5 (Ar), 127.6 (Ar), 127.1 (Ar), 125.2 (Ar), 124.5 (Ar), 120.1 (Ar), 65.8 (CH-CH2-O), 53.9 (CαH), 46.6 (CH-CH2-O), 43.6 (CH2-NH). IR νmax(CH2Cl2)/cm-1 3062 (C-H, sp2), 1717 (br, C=O), 1540 (NO2), 1342 (NO2), 1165 (S=O). 1H NMR data of the compound was consistent with the literature.188    Methyl (S)-1-(3-(((2-nitrophenyl)sulfonyl)amino))-2-((Fluorenylmethyloxy carbonyl)amino)propanoyl) aziridine-2-carboxylate (5.31b)  Synthesised and worked-up according to the general coupling procedure using amino acid 5.30b (0.19 g, 0.375 mmol) and aziridine 5.13 (38 mg, 0.372 mmol). The product was purified by flash column chromatography (45% n-heptane in EtOAc) to give aziridine 5.31b (0.10 g, 45%) as a slightly yellow solid. TLC Rf = 0.37 (2:3 n-heptane/EtOAc, v/v). 1H NMR (600 MHz, CDCl3) δ 8.14 – 8.09 (m, 1H, Ar), 7.85 – 7.81 (m, 1H, Ar), 7.77 (d, J = 7.5 Hz, 2H, Ar), 7.72 – 7.67 (m, 2H, Ar), 7.60 (d, J = 7.5 Hz, 2H, Ar), 7.40 (t, J = 7.5 Hz, 2H, Ar), 7.32 (t, J = 7.5 Hz, 2H, Ar), 6.02 (t, J = 5.5 Hz, 1H, CH2-NH-SO2), 5.73 (d, J = 7.5 Hz, 1H, CO-NH-CαH), 4.49 (dd, J = 12, 5 Hz, 1H, CαH), 4.37 (d, J = 7 Hz, 2H, CH-CH2-O), 4.20 (t, J = 7 Hz, 1H, CH-CH2-O), 3.74 (s, 3H, CH3), 3.65 – 3.59 (m, 1H, CHaHb-NH-SO2), 3.58 – 3.52 (m, 1H, CHaHb-NH-SO2), 3.28 – 3.24 (m, 1H, Aziridine-CH), 2.64 – 2.61 (m, 1H, Aziridine-CHaHb), 2.58 (d, J = 5 Hz, 1H, Aziridine-CHaHb). 13C NMR (151 MHz, CDCl3) δ 179.6 (CαH-(C=O)), 168.6 (COOCH3), 156.1 ((C=O)-NH), 148.1 (Ar), 143.8 (Ar), 141.4 (Ar), 133.9 (Ar), 133.6 (Ar), 133.0 (Ar), 131.1 (Ar), 127.9 (Ar), 127.3 (Ar), 125.6 (Ar), 125.3(Ar), 120.2 (Ar), 67.6 (CH-CH2-O), 55.6 (CαH), 53.2 (CH3), 47.2 (CH-CH2-O), 45.3 (CH2-NH-SO2), 35.2 (Aziridine-CH), 30.8 (Aziridine-CHaHb). IR νmax(CH2Cl2)/cm-1 3336 (N-H), 2956 (C-H, sp3), 1709 (br, C=O), 1540 (NO2), 1346 (NO2), 1167 (S=O). HPLC tR = 19.58 min (Method E). m/z (+ESI) found: MNa+, 617.1315. (C28H26N4O9SNa requires M, 617.1312).    85  Chapter 6: Protein Purification and ITC 6.1 Introduction This part of the thesis was conducted in the Frankel laboratory, Department of Pharmaceutical Sciences, University of British Columbia (UBC). It was the intention to produce large quantities of highly pure (>95%) PRMT proteins to be used for isothermal titration calorimetry (ITC) studies of the putative peptidomimetic binders/inhibitors (chapter 4).     6.2 Protein Expression and Purification The plasmids for expression of PRMTs 1-6 were available in DH5α E. coli cells kept in glycerol stock (1:1 ratio of 40 g cells/L Luria-Bertani (LB) broth and glycerol) at -80 oC. PRMTs 1, 4, 5 and 6 were expressed from pET28a(+) vectors that code for the N-terminal addition of six histidine residues, whereas PRMT2 and PRMT3 were expressed from pGEX vectors that provide an N-terminal GST-tag.   The PRMTs were expressed in BL21(DE3)pLysS E. coli cells, as these have a lower basal expression and show increased stability of the expressed protein.189 The BL21(DE3)pLysS E. coli cells were made competent by treatment with rubidium chloride as previously described.190 Competent cells were then transformed applying the heat-shock method. To verify successful transformation cells were grown on agar plates containing relevant antibiotics.   All proteins were expressed on large scale (in 10-12 L LB broth) as we anticipated having to perform several purification steps in order to produce enough protein of sufficient purity for ITC studies.191 In order to further enhance cell growth the recommended concentration of LB broth was doubled (from 20 g/L to 40 g/L) and 10 g/L D-glucose was added to the media, as this procedure had previously been employed in the Frankel laboratory to successfully increase protein yield. Attempts to express all PRMTs 1-6 were made in parallel. PRMTs 1, 4 and 6 were expressed with an N-terminal 6x His-tag,192 allowing for purification using a Ni2+ affinity column (HisTrap FF, GE Healthcare). PRMT1 expression yielded several co-purifying proteins and low yield of the protein of interest (POI). PRMT4 yielded high quantities of protein, but co-purified with several other proteins. Production of human PRMT5 was also attempted, but we were not able to express this protein in E. coli cells. This is most likely because the E. coli expression system does not facilitate posttranslational modifications.192 PRMT5 activity requires the co-expression of the protein MEP50, and the human PRMT5 has thus far only been expressed in insect cells.36,77      86  PRMT2 and PRMT3 proteins were expressed with a GST-tag and were therefore purified using an affinity column containing glutathione sepharose (GSTrap FF, GE Healthcare). For both PRMT2 and PRMT3 the expression and purification yielded low quantities (estimated by SDS-PAGE), and needed optimisation in order to yield significant quantities of protein. In general all the proteins needed further purification to reach a purity applicable for ITC experiments.193 As the Frankel laboratory had previously worked extensively with PRMT6194-197 we decided to start by optimising conditions for protein expression to obtain sufficient material to perform ITC on PRMT6.  6.2.1 Lysis Buffer Optimisation       After growing the transformed E. coli cells and inducing expression of the POI using isopropyl β-D-1-thiogalactopyranoside (IPTG), the cells were harvested and lysed in a lysis buffer. The lysis buffer composition is crucial for good protein isolation, as it provides stability for the POI, while also degrading the unwanted components of the E. coli, such as the cell wall and DNA.  The lysis buffer used in the Frankel laboratory at the start of this project contained several components; to digest the E. coli cell wall after expression lysozyme was added. The integrity of the cell wall was further compromised by freeze-thawing of the cell pellet as well as osmotic breakage due to a high salt concentration. Phenylmethanesulfonyl fluoride (PMSF) and Roche protease inhibitor tablets were added to inhibit several native proteases. DNase I digested the DNA from the cell and MgCl2 in the primary buffer acted as a chaotropic agent. β-mercaptoethanol maintained the protein in a reduced state, and detergent was added to solubilise the protein while glycerol was used to stabilise it. After lysis buffer treatment the cells were sheared by sonication 8x 30s and spun down at 45,000 g and 4 oC. Several issues with the lysis buffer were faced during this project, which are summarised in table 6.1.  Table 6.1 Lysis buffer was optimised during expression of the PRMTs. Left column indicates the changes made to the buffer and the right column indicates the result of these changes.  Entry Addition/removal of component Result 1 Original lysis buffer Clumpy cell debris pellet; loss of 50-65% lysate 2 Removal of DNase I (Thermo Fischer), lowering of detergent concentration, doubling of sonication Very viscous lysate, difficult to filter and load onto the affinity chromatography column 3 Addition of MgCl2 to lysate No change 4 Dilution of lysate with wash buffer No change 5 Changed primary buffer to 50 mM HEPES and 300 mM NaCl, removal of glycerol Lysate easier to filter 6 Addition of DNase I (Sigma) Quantitative lysate volume, easy to filter 7 Change back to primary buffer Quantitative lysate volume, more difficult to filter, but better protein yields after affinity chromatography column. 87  The first several protein purifications resulted in large clumpy precipitations after centrifugation with a resultant loss in lysate volume of 50-65% (entry 1). As a consequence the subsequent lysis buffers were prepared without DNase I, the detergent concentration was lowered from 0.5% to 0.1% and the sonication time was doubled (entry 2). This resulted in quantitative recovery of lysate after centrifugation. However, this lysate was very viscous and almost impossible to filter. Attempts to alleviate this problem by adding more MgCl2 (entry 3) or diluting the lysate with wash buffer (entry 4) were futile, and loading the small amount of filtered lysate onto the affinity column resulted in high backpressure and precipitation on the column. These lysate preparations were discarded as a result. To reduce viscosity of the lysate primary buffer, which contained high concentrations of salts (1M NH4Cl and 10mM MgCl2), was exchanged for 50 mM HEPES pH 7.6 with 300 mM NaCl to reduce salt contents, and glycerol was taken out of the lysis buffer composition, in both cases resulting in less viscous lysates (entry 5). A different DNase I was then employed (Sigma Aldrich) and found to yield quantitative yields of easily filtered lysate (entry 6). As the following purifications yielded less protein than before, the primary buffer containing high salt concentrations was employed in the following purifications (entry 7). The lysate was more difficult to filter, but yielded larger quantities of protein after affinity chromatography. Thus, this lysis buffer was employed in all following purifications.  6.2.2 PRMT6 Expression and Purification Optimisation PRMT6 was expressed several times according to the protocols in the experimental section 6.6. This was followed by purification on a 1.0-mL HisTrap FF affinity column (GE Healthcare) per 2.0-L bacterial culture, after which several further purification steps were attempted, in order to reach a purity of >95%. In an attempt to minimise nonspecific binding without severe loss of the protein of interest, 10-20 mM imidazole was added to the wash buffer.   Table 6.2 PRMT6 purification optimisation. The lane numbers correspond to the lane numbers in figure 6.1. Lane 1 is the protein standard. MES is 2-(N-morpholino)ethanesulfonic acid.   Lane  Purification Result 1 Precision Plus Protein™ Dual Color Standard  2 HisTrap FF affinity column Product band is major 3 Size exclusion column (SEC): 50 mM HEPES pH 7.5 No additional purification 4 Anion-exchange (AEX): 20 mM HEPES pH 8.0 No additional purification - Cation-exchange (SCX): 50 mM formic acid (FA) pH 4.0 Precipitation of protein 5 Cation-exchange (SCX): 50 mM NaH2PO4 pH 7.5 Only some protein binds but product is major band 6 Cation-exchange (SCX): 50 mM MES pH 6.0 Very poor binding of wrong protein  88  Table 6.2 shows the attempted purifications and the result. After Ni2+-affinity purification the band seen at the expected PRMT6 monomeric mass (42.8 kDa) in the SDS-PAGE (figure 6.1) is the major band (lane 2). Size exclusion chromatography did not result in an increased purity of the POI (lane 3).    Ion-exchange chromatography was then employed. As the isoelectric point (pI) of the PRMT6 sequence is 5.33, it would seem that a pH of 6.5 or above would render the protein negatively charged overall, and thus would favour purification by anion-exchange (AEX) chromatography. Performing AEX chromatography with 20 mM HEPES at pH 8.0 resulted in binding of the protein to the column, but the impurities likewise bound to the column (lane 4). A pI of 5.33 would favour the use of cationic-exchange (SCX) purification at pH 4 or below. As we feared that this low pH would denature the protein to a state that we would not be able to refold it, this was tested on a small fraction of protein. Not surprising to us, the protein precipitated when added to the 50 mM formic acid buffer at pH 4.0, and the remaining supernatant contained almost no protein.      Figure 6.1 SDS-PAGE showing different purification attempts of PRMT6 (see table 6.2). Box and arrow show the protein of interest (POI).  In its folded state the protein does not necessarily display the residues on the surface that correlates with the charges which a pI of 5.33 would produce; if residues that are prone to protonation (basic residues) are displayed on the protein surface, strong-cation exchange (SCX) purification could be feasible. Therefore an SCX-column was employed using 50 mM NaH2PO4 at pH 7.5, which had been applied previously with success.196 This resulted in significantly improved purification with elution of the protein of interest (POI) in >95% purity (lane 5). Unfortunately only a fraction of the loaded protein bound. Using 50 mM MES at pH 6.0 did not improve the result as only a small fraction of impurities bound to the column (lane 6). It thus seemed that using the SCX column with a 50 mM NaH2PO4 buffer at pH 7.5 gave the best result.      89   Figure 6.2 SDS-PAGE showing cationic purification (SCX) of PRMT6 using 50 mM NaH2PO4, pH 7.5. Lane 7 is the Precision Plus Protein™ Dual Color Standard. Lanes 8, 10 and 12 (in blue) are the non-binding fractions of the loaded protein from purification 1, 2 and 3, respectively, and 9, 11 and 13 are the eluted fractions from purification 1, 2 and 3, respectively. Box and arrow show the protein of interest.   To limit the time spent on purification in order to minimise potential denaturation and thus loss of enzymatic activity, SCX purification was performed after 6x His-tag affinity purification with a desalting step in between. As mentioned, only a part of the protein bound under these conditions. The non-binding fractions were reloaded onto the column after elution of the binding fractions, as we suspected that this could be due to a monomer/dimer equilibrium in the concentrated protein solution, and that one of these states had a higher affinity for the stationary phase than the other. Indeed we were able to recover several fractions of our POI (figure 6.2). Lanes 8, 10 and 12 show the non-binding fractions of the first, second and third loading, respectively, and 9, 11 and 13 show the eluted fractions from the corresponding loadings. Concentrating the fractions from each elution yielded a protein band that was of sufficient purity for ITC experiments.   6.3 Determining Concentration and Enzymatic Activity of PRMT6 In order to ensure good results from the ITC experiments it is critical to determine the correct concentration of protein as well as the enzymatic activity to ensure correctly folded protein. When this project was initiated the determination of protein concentration was performed by densitometry of an SDS-PAGE of the purified protein by comparing with a quantified stock of the protein.145 This method has some drawbacks; the many pipetting steps in the preparation of the SDS-PAGE could result in errors, the analysis of the gel relies on equal staining of the bands as well as precision of the software, and protein overloading will cause erroneous results. Finally, the concentration of the quantified 90  stock must be reliable. Determining protein concentration with this method unfortunately resulted in high standard deviations (>30%). We therefore decided to use the Edelhoch method,198 in which a protein sample is denatured using 6 M guanidinium chloride and the concentration is determined using the Lambert-Beer law. The concentration could be determined using only very little sample and in general this resulted in standard deviations of ~5%, thus producing reliable concentration results which could be used in the ITC experiments.    Measuring enzymatic activity can be done using several methods.199 The methods at hand were based on spectroscopic quantification of the by-product SAH using a SAH standard curve,199 or the P81 filter assay,66 which utilises 13C-SAM as a cofactor and measures transferred radioactivity. In both assays auto-methylation can alter the signal output. The P81 filter assay became the method of choice as the project progressed due to the ease of the preparation and sample reading. The activity for PRMT6 after Ni2+ affinity column and SCX-column purification as described was found to be in the same range as the mCitrine-tagged PRMT6 (mCit-PRMT6) previously expressed in the group.195   6.3.1 DTT Increases Enzymatic Activity of PRMT6 As the lysis buffer contained a reducing agent, we suspected that this was of importance to the enzymatic activity. Adding 1 mM dithiothreitol (DTT) to the enzyme before performing the activity assay showed a >2-fold increase in activity compared to the activity without DTT. We therefore decided to maintain the reducing agent in all steps during lysis, purification and running of the assays. However, it seemed that SCX purification with 50 mM NaH2PO4 with 5 mM DTT at pH 7.5 resulted in co-elution of another protein (~60-65 kDa) (figure 6.3). As this protein did not co-elute to the same degree under the identical conditions without reducing agent, it seemed that this protein was not binding to the column but to the protein of interest. We thus suspected that the reducing agent increased the affinity of this contaminant protein for the expressed PRMT6.    91   Figure 6.3 SDS-PAGE showing cationic purification (SCX) of PRMT6 using 50 mM NaH2PO4, pH 7.5 with 5 mM DTT. Lane 13 is the Precision Plus Protein™ Dual Color Standard. Lanes 14-17 are the eluted fractions, and the red box and arrow show the co-eluting protein. The black box and arrow show the protein of interest.   6.4 Isothermal Titration Calorimetry (ITC) ITC is a highly sensitive and accurate method to determine the affinity and thermodynamics of binding between binding partners.200 The binding reaction is performed in a cell containing one of the binding partners in a suitable buffer into which the other binding partner is titrated, most often a protein and a ligand, respectively. By measuring the heat absorbed or released during binding the dissociation constant for the binding (KD) may be calculated. A typical exothermic ITC experiment yields titration curves as those in figure 6.4, top. Integrating these will most often yield a sigmoidal curve such as the one seen in the bottom of figure 6.4. Three parameters can be easily determined from the resulting curve;193,200 the enthalphy of binding (∆H, green) is the difference in energy between no binding and 100% binding, the incline on the steepest part of the curve gives the binding dissociation constant (KD, red) and the midpoint of the titration reveals the stoichiometry of binding (N, blue).193 From these parameters the free energy of binding (∆G) and the entropy of binding (∆S) may be determined. As the heat absorbed from or released to the buffer is very small, the buffer of the two binding partners must be identical to avoid heats of mixing. This can be achieved by performing dialysis of one or both binding partners and dissolving the other binding partner in the resulting buffer (if relevant). As mentioned previously, it is also necessary to employ highly pure protein and ligand in order to avoid interference of contaminants with the binding reaction or the buffering.191   92   Figure 6.4 A typical ITC experiment yields the titration curves (top), which can be integrated to the sigmoidal curve (bottom) from which several parameters (∆H, N, KD) can be read directly. The data is taken from Pochetti et al. Nature Protocol Exchange 2012 (doi:10.1038/protex.2012.063)201 with permission and adapted with parameter indications.    6.4.1 Dialysis  Dialysis was performed using a cellulose ester dialysis tubing with a 50 mM phosphate buffer containing 100-200 mM NaCl and later also a reducing agent (DTT or β-mercaptoethanol). The dialysis was run for at least 24 hr with 2-3 buffer exchanges, the last exchange done at least 12 hr prior to the ITC experiment.   6.4.2 Performing ITC on PRMT6 Sack et al. performed ITC studies on PRMT4 and showed that co-factor binding was essential for binding of their inhibitor to the protein.73 The intention of this study was to determine the binding affinity of our inhibitors to the PRMTs. And as we did not know whether co-factor binding to the protein would increase binding or possibly even be essential for binding of the inhibitor, we started by titrating SAH into the cell containing the purified PRMT6. The KD of SAH or SAM with PRMT6 had been previously determined in the low micromolar range (1.4 μM195 and 16.5 μM,196 respectively) and we therefore performed several ITC experiments using the MicroCal™ Auto-iTC 200 (GE Healthcare) with a PRMT6 concentration in the low micromolar range. The general outcome of the ITC experiments can be seen in figure 6.5.  93   Figure 6.5 The general outcome of an ITC experiment with PRMT6 and SAH in the low micromolar concentration range.    Several separate experiments with several different batches of PRMT6 were performed with PRMT6 concentrations varying from 10-150 μM and SAH concentrations 1.5-10-fold higher. Ligand-into-buffer control experiments were performed for each ITC experiment. Fitting a curve to the data in most cases resulted in a curve as seen in figure 6.5, resembling the upper half of the desired sigmoidal curve. As we expect SAH to bind to PRMT6 in a 1:1 fashion, the stoichiometry (N) should be 1, but was always significantly lower, indicating too low concentrations or inactive/denatured protein. We started out by using a 10-fold concentration of ligand compared to the protein as this is the general recommendation.193, 200 However, after several experiments yielding similar poor results and getting the aid of the technical support from GE Healthcare, the sigmoidal curve seen in figure 6.6 was obtained with a PRMT6 concentration of 150 μM and 250 μM SAH titrated into the enzyme. Unfortunately the stoichiometry of this experiment yet again was poor.  We suspected that the enzyme might be aggregating in the phosphate buffer, but running the enzyme through the size-exclusion column in the phosphate buffer yielded only one eluted peak. The poor results would therefore most likely arise from the use of wrong concentrations of ligand or protein, or from low activity of the protein.  94   Figure 6.6 ITC experiment performed with 150μM PRMT6 and 250 μM SAH titrated into the enzyme.    The concentration of our ligand stock was verified spectroscopically and as the concentration of all protein preparations had been determined with low standard deviations we did not suspect concentration error to be the problem. Measuring the concentration of protein in either of the two methods (section 6.3) requires denaturation of the protein,145, 198 and thus we did not know if all of the protein was active or if it was partially degraded based on concentration measurements. During the course of the project the Fast Protein Liquid Chromatography (FPLC) system experienced high backpressure and problems with the pump heads. Although thorough cleaning of the system was performed regularly before each purification, on several occasions the purification had to be interrupted and the system cleaned before being able to continue. This of course slowed the purification and activity may have been lost. However, as mentioned previously the activity of the purified PRMT6 was comparable to mCit-PRMT6 previously purified in the group.  The experience gained in working with the MicroCal™ Auto-iTC 200 during the course of the project, and the fact that the only experiment which had yielded good results on the ITC was a test experiment in which Ca2+ was titrated into EDTA, led to the performance of a few test experiments. These clearly showed heats of mixing during titrations of water into water, which led to the conclusion that detergent was left behind in the cell after the thorough automated wash of the system before each batch of experiments. This would 95  explain the low stoichiometry, as the protein would be denatured in the cell, while also explaining why the only successful experiment to date was the Ca2+/EDTA test experiment. This realisation concluded the experiments performed at UBC.        6.5 Discussion and Conclusion Having set the bar quite high, we realised during this part of the project that performing ITC with our inhibitor on all PRMTs would not be feasible within the timeframe at UBC. We did however manage to optimise lysis buffer composition for large scale purification of the PRMTs.  Purification of PRMT6 also proved challenging. As the bottleneck for running ITC experiments is the large amount of active, highly pure protein required, various purification methods were tested. The addition of 10-20 mM imidazole to the wash buffer was done in an attempt to avoid nonspecific binding while not significantly affecting the yield. Several ion-exchange buffers were tested in an effort to minimise handling time of the enzyme to ensure a highly active enzyme. In the end we found that the best purification consisted of Ni2+affinity chromatography followed by several rounds of loading and eluting of the SCX column using NaH2PO4 pH 7.5. Ironically, when adding a reducing agent to all purification steps, as this was shown to be beneficial for the activity of the enzyme, a contaminant protein most likely bound with a higher affinity to our protein. None-the-less we were able to produce protein of a high purity which had an enzymatic activity in the same range as the mCit-PRMT6 previously expressed in the group.195 This PRMT6 was used for the subsequent dialysis and ITC experiments. Unfortunately the preliminary ITC experiments in which we titrated SAH into a phosphate buffer containing PRMT6 did not yield any useful results, as we were not able to reach the expected stoichiometry of N=1 most likely due to the denaturation of protein in the ITC cell.  In retrospect, the most feasible strategy would have been to use a simpler assay to screen for the most potent binding affinities between the PRMT inhibitors (chapter 4) and the PRMTs. One such assay could be the fluorescence polarisation assay, based on the principle that a fluorescently labelled compound bound to a macromolecule will emit a different polarisation value compared to a non-bound molecule.202 The fluorescence polarisation assay is easier to perform, and does not require the same amount or purity of the protein. The most interesting ligand-protein interactions determined in this screen could subsequently have been submitted to further biophysical studies such as ITC.    96  6.6 EXPERIMENTAL SECTION 6.6.1 GENERAL INFORMATION Luria-Bertani (LB) broth, antibiotics, DNase I (Sigma: DN25) and various salts/chemicals were all purchased at Thermo-Fisher or Sigma Aldrich. Protogel (30%) was purchased from National Diagnostics. Protease inhibitor tablets (“cOmplete Protease Inhibitor Cocktail Tablets”) were purchased from Roche. HisTrap FF, GSTrap FF, HiTrap SP HP, HiTrap Q HP were all purchased from GE Healthcare (either 1 mL or 5 mL) and used singly or in series during FPLC purification. HisTrap FF columns were stripped and reloaded according to vendor protocol recommendations.  HiLoad 26/600 Superdex 200 pg (Size exclusion column, SEC) was purchased from GE Healthcare.  6.6.2 GENERAL SOLUTIONS AND BUFFERS 6.6.2.1 LB broth for E. coli growth and expression 40 g/L in H2O, sterilised.   6.6.2.2 Antibiotics stocks Kanamycin (Kan):  50 mg/mL (dissolved in MQ); used for pET28a(+) plasmids Chloramphenicol (CAM):  35 mg/mL (dissolved in MeOH); used for BL21 (DE3) pLysS E. coli Ampicillin (Amp):  100 mg/mL (dissolved in MQ); used for pGEX plasmids Antibiotic stocks were diluted 1:1000 in the final solutions  6.6.2.3 Solutions for competent cells TfbI  100mM RbCl, 50mM MnCl2, 30mM KOAc, 10mM CaCl2 x 2H2O, 15% glycerol.  Sterilised and stored in a dark bottle at 4 oC. TfbII 10 mM 3-(N-morpholino)propanesulfonic acid, 10 mM RbCl, 75 mM CaCl2 x 2 H2O, 15% glycerol. Sterilised and stored in a dark bottle at 4 oC.    97  6.6.2.4 Solutions for sodium dodecyl sulfate-gels (SDS-PAGE) Separating Buffer :  4x Tris x Cl/SDS, pH 8.8 (1.5 M Tris. Cl) Stacking Buffer:  4x Tris x Cl/SDS, pH 6.8 (0.5 M Tris. Cl) 5x SDS-PAGE sample buffer: 3 mL 0.5M Tris x HCl pH 6.8, 6 g glycerol, 5 mL 10% sodium dodecyl sulfate (SDS), 1.25 mL β-mercaptoethanol, 0.8 mL 1% bromophenolblue. 5x SDS running buffer: 0.125 M Tris Base, 0.96 M glycine, 0.5 w/v % SDS. Fixing Solution: 50% MeOH, 10% AcOH, 40% H2O Coomassie Brilliant Blue Stain solution: 1 g/L in fixing solution Destaining Solution: 10% MeOH, 10% AcOH, 80% H2O  6.6.2.5 1x Primary buffer 50mM HEPES-KOH pH 7.6, 1M NH4Cl and 10mM MgCl2  6.6.2.6 Lysis buffer (Note: This is the final version, see section 6.2.1) 1x primary buffer, 1 mg/mL lysozyme, 7 mM β-mercaptoethanol, 1 mM phenylmethanesulfonyl fluoride (PMSF), DNase I, 1 protease inhibitor tablet/ 50 mL lysis buffer, 10 mM imidazole (only for 6x His-tagged proteins)   6.6.2.7 Buffers for 6x His-tagged proteins (HisTag FF) Wash buffer A:  1x primary buffer, 7 mM β-mercaptoethanol, 10-20 mM imidazole, 200 μM PMSF or 50 mM HEPES pH 7.5, 300 mM NaCl, 7 mM β-mercaptoethanol, 10-20 mM imidazole, 200 μM PMSF  Elution buffer B: 1x primary buffer, 7 mM β-mercaptoethanol, 400 mM imidazole or 50 mM HEPES pH 7.5, 300 mM NaCl, 7 mM β-mercaptoethanol, 400 mM imidazole   98  6.6.2.8 Buffers for GST-tagged proteins (GSTrap FF) Wash buffer A: PBS, pH 7.3 (140 mM NaCl, 2.7 mM KCl, 10 mM  Na2HPO4, 1.8 mM KH2PO4, pH 7.3), 200 mM PMSF  Elution buffer B: 50 mM Tris-HCl, 10 mM reduced glutathione, 5 mM DTT  6.6.2.9 Size exclusion (SEC) buffer 50 mM HEPES pH 7.5 (Add 300 mM NaCl only if the SEC column is not followed by a SCX or AEX column)  6.6.2.10 Mono-basic phosphate buffer (HiTrap SP HP purification (SCX)) Wash buffer A: 50 mM NaH2PO4 pH 7.5 (tested with/without 200 mM NaCl) Elution buffer B: 50 mM NaH2PO4 pH 7.5, 1 M NaCl  6.6.2.11 Other tested buffers (AEX and SCX chromatography) 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (AEX): Wash buffer: 20 mM HEPES pH 8.0; Elution buffer: 20 mM HEPES pH 8.0 + 1 M NaCl.  2-(N-morpholino)ethanesulfonic acid (MES) (SCX): Wash buffer: 50 mM MES pH 6.0; Elution buffer: 50 mM MES pH 6.0 + 1 M NaCl.   Formic acid (FA) (SCX):  Wash buffer: 50 mM FA pH 4.0; Elution buffer: 50 mM FA pH 4.0 + 1 M NaCl.    6.6.2.12 Storage buffer (pH 8.0) 100 mM HEPES-KOH, 200 mM NaCl, 1 mM DTT, 10% glycerol, 2 mM EDTA  6.6.2.13 Buffer for ITC (pH 7.5) 50 mM phosphate buffer (6.3 mL 1M NaH2PO4 + 43.7 mL 1M Na2HPO4) pH 7.5, with 100-200 mM NaCl (and later a reducing agent; DTT or β-mercaptoethanol)     99  6.6.3 GENERAL PROCEDURES 6.6.3.1 Making competent BL21(DE3)pLysS (E. coli) or DH5α (E. coli) cells   Cells were thawed on ice and 7 mL LB broth containing the appropriate antibiotics (if needed) was inoculated and shaken overnight at 37 oC and 260 rpm. This suspension was then transferred to 500 mL LB broth containing the appropriate antibiotics (if needed) and incubated at 37 oC and 260rpm until OD600 reached 0.5-0.9. The culture was then spun down at 3100 x g and 4 oC for 5 min. The supernatant was discarded and the pellet resuspended in 30 mL TfbI and cooled to 0 oC for 15 min. The suspension was then spun down at 3100 x g and 4 oC for 5 min and the supernatant discarded. The pellet was resuspended in 6 mL TfbII at 0 oC and aliquoted into chilled Eppendorf tubes. The cells were quick frozen using liquid N2 and stored at -80 oC.  6.6.3.2 Mini plasmid preparation: Extraction of plasmids from DH5α stock Plasmids were extracted from the DH5α stocks using the PureLink® Quick Plasmid Miniprep Kits (Invitrogen).203  6.6.3.3 LB Agar plates Premade LB agar (20 g/L agar, 20 g/L LB broth, sterilised) was heated until obtaining a homogeneous liquid, and cooled to 40-50 oC. For each plate 20 mL premade agar and relevant antibiotic(s) were added to a petri dish. Bubbles were removed using a bunsenburner and the agar plates were allowed to set for 20 min at rt. The plates were heated to 37 oC 10 min prior to use.  6.6.3.4 Transformation of plasmids into BL21(DE3)pLysS (E. coli) or DH5α (E. coli)   Competent cells were thawed on ice for 5 min. Maximum 20 ng plasmid was aliquoted into a pre-chilled Eppendorf tubes and 50 μL cell suspension was added and gently mixed.  The mixtures were chilled for 30-60 min and heat shocked at 41-43 oC for 45 sec. The mixtures were then iced for two min and 450 μL LB broth containing no antibiotics was added. The cultures were shaken at 300 rpm at 37 oC for 60 min. The suspensions were then spun down at 800 x g and 2/3 of the supernatant was discarded. The remaining supernatant was used to resuspend the cells and these were plated onto agar plates containing the appropriate antibiotics. The plates were inverted and incubated at 37 oC overnight for colony growth. When visible colonies were seen the plates were taken out and stored at 4 oC.    100  6.6.3.5 Protein expression  A starter culture was made by inoculating a single colony from the agar plate of BL21(DE3)pLysS E. coli cells transformed with the appropriate plasmid into 200mL LB broth with the required antibiotic(s) in a 500 mL Erlenmeyer flask. The starter culture was shaken overnight at 300rpm and 37 oC.  The next day the OD600 of the starter culture was determined at a 1:9 dilution (0.1 mL of the culture + 0.9 mL MQ) and 1 L cultures were inoculated to a final OD600 of 0.05 of E. coli with addition of the required antibiotics and 10 g/L D-glucose. The cultures were incubated at 37 oC, 260rpm until reaching an OD600 of 0.6-0.8. The culture flasks were chilled at 0 oC while incubators were cooled, and protein expression was induced by addition of IPTG to a final concentration of 0.5 mM. The cultures were incubated at ambient temperature (16-22oC) at 260 rpm for 16 hr. Cultures were centrifuged at 5000 g for 10 min at 4 oC, the supernatant was discarded and the pellets were stored at -80 oC until lysis.      6.6.3.6 Lysis of E. coli and protein purification using tag-affinity columns Pellets were thawed for 15-20 min on ice, and freshly prepared lysis buffer (2 mL/g pellet) was added. The pellets were broken up using mechanical strength and homogenised on ice. The suspension was then incubated on ice for 20-30 min, and sonicated on ice for 8 x 30 sec with 1 minute intervals of lightly shaking (Ultrasonic Liquid Processor Sonicator; power level 3, 50% output).  The suspension was kept on ice for 20 min and then centrifuged for 45 min at 4 oC and 45,000 g. The supernatant was isolated, 500 μL 1M MgCl2 per 50 mL lysate was added and everything was filtered using 0.45 μM syringe filters.  All proteins were purified on a Fast Protein Liquid Chromatography (FPLC) system from Äkta.  The lysate was loaded onto a tag-affinity column (HisTrap FF (GE Healthcare) for 6x His-tagged proteins; PRMT1, PRMT4, PRMT5 and PRMT6 or a GSTrap FF (GE Healthcare) for GST-tagged proteins; PRMT2 and PRMT3 using a superloop at a loading flow rate of 1 mL/min. The column was then washed using wash buffer A (for either 6x His-tagged or GST-tagged proteins) until a stable baseline was attained and then for an additional 5 column volumes. Proteins were eluted using a step gradient with wash buffer A and elution buffer B (for either 6x His-tagged or GST-tagged proteins). During the entire lysis and tag-affinity purification 0.5 mL 200 mM PMSF was added every 30 min.      6.6.3.7 Size exclusion (SEC) column The collected fractions were collected from the tag-affinity columns and loaded onto the SEC column. This was run at a flow rate of 1-2 mL depending on the backpressure of the FPLC system.    101  6.6.3.8 Storage of the purified protein For long term storage after FPLC purification the proteins were concentrated, and spun down thrice in 10 mL storage buffer at 4 oC and 3100 x g using a 10 or 30 kDa molecular weight cut-off (MWCO) centrifugal filter (Amicon). The proteins were stored at -80 oC in cryotubes.  6.6.3.9 Making and running an SDS-PAGE (10%) The gels were cast using the Mini-PROTEAN Tetra Handcast system (BioRAD). The glass slides were cleaned thoroughly using MeOH and set up in the holders.   For the making of four gels: First mix (5.00 mL Protogel, 3.75 mL separating buffer, 6.25 mL H2O, 0.05 mL of 10% w/v ammonium persulfate, 0.01 mL tetramethylethylenediamine) was distributed between the four casting frames, butanol was added to the top to remove air bubbles and the gels were allowed to set for 45 min. Butanol was removed and second mix (0.75 mL Protogel, 3.05 mL H2O, 1.25 mL stacking buffer, 38 μL of 10% w/v ammonium persulfate, 7 μL tetramethylethylenediamine) was distributed between the four casting frames and combs were inserted. The gels were allowed to set for 60 min.  The gels were stored at 4 oC for up to 7 days.  Samples were prepared using 5x SDS-PAGE sample buffer and the desired protein to a final concentration of 1x SDS-PAGE sample buffer with protein in a final volume of 15 μL (15-well PAGE) or 20 μL (10-well PAGE) . A marker was prepared in the same volume as the samples (2 μL marker (Precision Plus Protein™ Dual Color Standards, BioRAD) in 1x SDS-PAGE sample buffer). The samples were heated to 90 oC for 10 min, cooled on ice and spun down.  The gels were run using the mentioned BioRAD system in a 1x SDS running buffer (5x dilution of 5x SDS running buffer). The wells were rinsed with 1x SDS running buffer and the samples loaded. The gel was run at 100V until the SDS-PAGE sample buffer was at the bottom of the gel.  The gel was taken out and shaken in fixing solution for 10 min. The fixing solution was removed and the gel was microwaved for 10 sec in Coomassie Brilliant Blue Stain solution, the solution discarded and shaken in fixing solution for 5 min or until bands were visible. The fixing solution was removed. Destaining solution was added to the gel and microwaved for 20 sec. Kim wipes were put on top of the gels and shaken overnight. The gel was dried onto a filter paper for 45 min at 80oC using a gel slab dryer.     102  6.6.3.10 Protein quantification: SDS-PAGE densitometry The concentrations of protein were determined by separation of purified proteins on SDS-PAGE with concurrent densitometry of Coomassie blue-stained bands as described previously.145 Using a quantified stock of the protein in question in several concentrations a standard curve was constructed by which the protein could be quantified. Densitometry was performed using ImageJ (free online software).   6.6.3.11 Protein quantification: Edelhoch method198 The protein was denatured in a 6 M guanidinium chloride in 50 mM NaH2PO4 in a 1:9 ratio. The concentration of the protein was determined using a spectrophotometer (Nanodrop, Thermo Scientific) by measuring absorption at 280 nm and using the Lambert-Beer law as well as the extinction coefficient for the protein in question. For each concentration determination two samples were prepared, from each of these 3-4 separate samples were measured 6 times each.   6.6.3.12 Activity assay: SAH Quantification Samples were made of 200 nM protein, 20 μM SAM, 20 μM substrate (H3 tail) in a 1x methylation buffer (50 mM HEPES-KOH, pH 8.0, 10 mM NaCl, and 1.0 mM DTT) in a total volume of 80 μL. No enzyme and no substrate controls were run, as well as mCit-PRMT6, which was previously synthesised in the group.195 Samples were incubated for 120 min. To stop methylation, the samples were filtered using a NanoSep 3K Omega filters (Pall Corporation) in order to remove the enzyme. A standard curve using 5-10,000 nM SAH was produced, which was used for quantifying SAH from Q Exactive Orbitrap LC-MS/MS data.  6.6.3.13 Activity assay: P81 filter binding assay66 Samples were made with PRMT6 (final concentration of 400/800 nM), 50 μM 14C-SAM, 5 μg substrate (GST-GAR used) in a 1x methylation buffer (50 mM HEPES-KOH, pH 8.0, 10 mM NaCl, and 1.0 mM DTT) in a total volume of 35 μL. No enzyme and no substrate controls were run, as well as mCit-PRMT6, which was previously synthesised in the group.195 Samples were incubated for 60 or 120 min. To stop methylation, the samples were flash-frozen and the entire reaction volume spotted onto a Whatman P81 (phosphocellulose) filter paper (1.2 cm x 1.2 cm square) placed in an Eppendorf tube. After allowing the sample to dry for at least 20 min, the filter paper was washed five times with 1-2 mL of 50 mM NaHCO3 buffer (pH = 9.0) (the filter paper was soaked for 5 min during each wash). The filter papers were dried in the Eppendorf tubes at 37 oC overnight, transferred to scintillation vials and then 5 mL scintillation fluid (SX16-4 ScintiVerse, Fisher Scientific) was added. The vials were capped and shaken prior to counting radioactivity in a scintillation counter (Tri-Carb 3110TR Liquid Scintillation Counter - Beta Counter).  103  6.6.3.14 Dialysis and ITC The protein was dialysed in a Biotech CE Dialysis Tubing, 0.5-1.0 kD MWCO membrane (Spectrum Labs) using NaH2PO4 pH 7.5 with 200mM NaCl (and 5 mM DTT/ 1 mM β-mercaptoethanol) at 4 oC with light stirring. Prior to dialysis the membrane was soaked in H2O to get rid of the NaN3 in which the membrane was preserved. The dialysis always lasted for 24 hr or more, with 2-3 buffer exchanges; the last buffer exchange performed at least 12 hr prior to ITC experiments.  The protein was then quantified using the Edelhoch method198 and diluted to the desired concentration using the buffer from the final buffer exchange. SAH was diluted using this same final buffer.   The samples were loaded into a tray, and samples were kept at 4 oC until reaction and run on the MicroCal™ Auto-iTC 200 System.193 The data was analysed using the Origin 7.0 software (OriginLab Corporation). Prior to the first run of each batch the “Plates Clean” automated washing programme was performed and the “Plates” automation programme was performed in between each run. For every batch two “ligand into buffer” controls were run.    104  Chapter 7: Conclusion and Perspectives In conclusion we performed a structure-activity relationship study using three peptidomimetic modifications at the tryptophans in the P1 and P16 leads and identified a low micromolar inhibitor of the PRMTs, P21. The inhibitor is likely substrate specific and may depend on methylation level as inhibition was only seen for PRMTs using a histone tail 4 substrate in the antibody-based inhibition assay.  The peptidomimetic P21 was found to have an interesting turn-conformation in the Phe-L-Aia-Asp-L-Aia-Gly motif, and we suspect that this motif could confer the inhibitory potency towards PRMT1 and PRMT3 without affecting inhibition levels of PRMTs 5, 6 and 8 significantly. Furthermore, as the turn motif was stabilised by π-stacking of the Aia indoles and an H-bond interaction in the backbone we expect the general –L-Aia-Xxx-L-Aia- motif to induce this turn conformation, and that this strategy may therefore be utilised in other peptidomimetic applications.  The 13mer peptide P16 displayed a localisation pattern in HEK293 cells, which did not correspond to any known PRMT, and may therefore be binding to other targets in the cell. In the future it would be interesting to investigate P21 in a cellular setting, due to its inhibitory potency for PRMT1, the most expressed PRMT in human cells. The efforts in expression of PRMT6 resulted in the production of highly pure active protein, which was used in ITC experiments. Unfortunately we were not able to produce a binding curve using PRMT6 and SAH with the right stoichiometry using ITC.  Furthermore, as the azepinone-mimics of the aromatic amino acids have proven valuable in peptidomimetics, and given that the best PRMT-inhibitor of this study contained two L-Aia, we attempted to synthesise a lysine/arginine dipeptide mimetic scaffold (5.1) for use in future peptidomimetic studies. The synthesis of 5.1 using aziridine chemistry proved more difficult than expected, and we hope to accomplish synthesis of 5.1 using other strategies as described in chapter 5.  The interest in epigenetics in general and specifically in the inhibition of the epigenetic enzymes will likely grow in the coming years, as development of potent selective inhibitors of these enzymes will provide tools for deciphering the epigenetic code as well as drugs for future disease treatment.             105  References   1.  Arrowsmith, C. H.; Bountra, C.; Fish, P. V.; Lee, K.; Schapira, M. Epigenetic protein families: a new frontier for drug discovery. Nature reviews Drug discovery 2012, 11, 384-400.  2.  Berger, S. L.; Kouzarides, T.; Shiekhattar, R.; Shilatifard, A. An operational definition of epigenetics. Genes & Development 2009, 23, 781-783.  3.  Waddington, C. The Epigenotype. Endeavour 1942, 18-20.  4.  Tough, D. F.; Lewis, H. D.; Rioja, I.; Lindon, M. J.; Prinjha, R. K. Epigenetic pathway targets for the treatment of disease: accelerating progress in the development of pharmacological tools: IUPHAR review 11. British journal of pharmacology 2014, 171, 4981-5010.  5.  Luger, K.; Mader, A. W.; Richmond, R. K.; Sargent, D. F.; Richmond, T. J. Crystal structure of the nucleosome core particle at 2.8 Å resolution. Nature 1997, 389, 251-260.  6.  Tan, M.; Luo, H.; Lee, S.; Jin, F.; Yang, J.; Montellier, E.; Buchou, T.; Cheng, Z.; Rousseaux, S.; Rajagopal, N.; Lu, Z.; Ye, Z.; Zhu, Q.; Wysocka, J.; Ye, Y.; Khochbin, S.; Ren, B.; Zhao, Y. Identification of 67 Histone Marks and Histone Lysine Crotonylation as a New Type of Histone Modification. Cell 2011, 146, 1016-1028.  7.  Strahl, B. D.; Allis, C. D. The language of covalent histone modifications. Nature 2000, 403, 41-45.  8.  Cedar, H. DNA methylation and gene activity. Cell 1988, 53, 3-4.  9.  Fischle, W.; Wang, Y.; Allis, C. D. Histone and chromatin cross-talk. Current Opinion in Cell Biology 2003, 15, 172-183.  10.  Shahbazian, M. D.; Grunstein, M. Functions of Site-Specific Histone Acetylation and Deacetylation. Annu. Rev. Biochem. 2007, 76, 75-100.  11.  Copeland, R. A.; Solomon, M. E.; Richon, V. M. Protein methyltransferases as a target class for drug discovery. Nat Rev Drug Discov 2009, 8, 724-732.  12.  Richon, V. M.; Johnston, D.; Sneeringer, C. J.; Jin, L.; Majer, C. R.; Elliston, K.; Jerva, L. F.; Scott, M. P.; Copeland, R. A. Chemogenetic Analysis of Human Protein Methyltransferases. Chemical Biology & Drug Design 2011, 78, 199-210.  13.  Qian, C.; Zhou, M. M. SET domain protein lysine methyltransferases: Structure, specificity and catalysis. Cell. Mol. Life Sci. 2006, 63, 2755-2763.  14.  Kubicek, S.; Jenuwein, T. A Crack in Histone Lysine Methylation. Cell 2004, 119, 903-906.  15.  Cuthbert, G. L.; Daujat, S.; Snowden, A. W.; Erdjument-Bromage, H.; Hagiwara, T.; Yamada, M.; Schneider, R.; Gregory, P. D.; Tempst, P.; Bannister, A. J.; Kouzarides, T. Histone Deimination Antagonizes Arginine Methylation. Cell 2004, 118, 545-553.  16.  Le Romancer, M.; Treilleux, I.; Leconte, N.; Robin-Lespinasse, Y.; Sentis, S. p.; Bouchekioua-Bouzaghou, K.; Goddard, S.; Gobert-Gosse, S. p.; Corbo, L. Regulation of Estrogen Rapid Signaling through Arginine Methylation by PRMT1. Molecular Cell 2008, 31, 212-221.  17.  Bedford, M. T.; Clarke, S. G. Protein Arginine Methylation in Mammals: Who, What, and Why. Molecular Cell 2009, 33, 1-13.  18.  Zhao, Q.; Rank, G.; Tan, Y. T.; Li, H.; Moritz, R. L.; Simpson, R. J.; Cerruti, L.; Curtis, D. J.; Patel, D. J.; Allis, C. D.; Cunningham, J. M.; Jane, S. M. PRMT5-mediated methylation of histone H4R3 recruits DNMT3A, coupling histone and DNA methylation in gene silencing. Nat Struct Mol Biol 2009, 16, 304-311.  19.  Wang, H.; Huang, Z. Q.; Xia, L.; Feng, Q.; Erdjument-Bromage, H.; Strahl, B. D.; Briggs, S. D.; Allis, C. D.; Wong, J.; Tempst, P.; Zhang, Y. Methylation of Histone H4 at Arginine 3 Facilitating Transcriptional Activation by Nuclear Hormone Receptor. Science 2001, 293, 853-857.  20.  Jansson, M.; Durant, S. T.; Cho, E. C.; Sheahan, S.; Edelmann, M.; Kessler, B.; La Thangue, N. B. Arginine methylation regulates the p53 response. Nat Cell Biol 2008, 10, 1431-1439. 106   21.  Pal, S.; Vishwanath, S. N.; Erdjument-Bromage, H.; Tempst, P.; Sif, S. d. Human SWI/SNF-Associated PRMT5 Methylates Histone H3 Arginine 8 and Negatively Regulates Expression of ST7 and NM23 Tumor Suppressor Genes. Molecular and Cellular Biology 2004, 24, 9630-9645.  22.  Yang, Y.; Hadjikyriacou, A.; Xia, Z.; Gayatri, S.; Kim, D.; Zurita-Lopez, C.; Kelly, R.; Guo, A.; Li, W.; Clarke, S. G.; Bedford, M. T. PRMT9 is a Type II methyltransferase that methylates the splicing factor SAP145. Nat Commun 2015, 6.  23.  Torres-Padilla, M. E.; Parfitt, D. E.; Kouzarides, T.; Zernicka-Goetz, M. Histone arginine methylation regulates pluripotency in the early mouse embryo. Nature 2007, 445, 214-218.  24.  Chen, D.; Ma, H.; Hong, H.; Koh, S. S.; Huang, S. M.; Schurter, B. T.; Aswad, D. W.; Stallcup, M. R. Regulation of Transcription by a Protein Methyltransferase. Science 1999, 284, 2174-2177.  25.  Cheng, D.; Coté, J.; Shaaban, S.; Bedford, M. T. The Arginine Methyltransferase CARM1 Regulates the Coupling of Transcription and mRNA Processing. Molecular Cell 2007, 25, 71-83.  26.  Wang, Y. C.; Li, C. Evolutionarily conserved protein arginine methyltransferases in non-mammalian animal systems. FEBS Journal 2012, 279, 932-945.  27.  Yang, Y.; Bedford, M. T. Protein arginine methyltransferases and cancer. Nature Reviews Cancer 2013, 13, 37-50.  28.  Cheng, X.; Collins, R. E.; Zhang, X. Structural and Sequence Motifs of Protein (Histone) Methylation Enzymes. Annu Rev Biophys Biomol Struct 2005, 34, 267-294.  29.  Schubert, H. L.; Blumenthal, R. M.; Cheng, X. Many paths to methyltransfer: a chronicle of convergence. Trends in Biochemical Sciences 2003, 28, 329-335.  30.  Fuhrmann, J.; Clancy, K. W.; Thompson, P. R. Chemical Biology of Protein Arginine Modifications in Epigenetic Regulation. Chem. Rev. 2015, 115, 5413-5461.  31.  Yue, W. W.; Hassler, M.; Roe, S. M.; Thompson-Vale, V.; Pearl, L. H. Insights into histone code syntax from structural and biochemical studies of CARM1 methyltransferase. The EMBO Journal 2007, 26, 4402-4412.  32.  Zhang, X.; Zhou, L.; Cheng, X. Crystal structure of the conserved core of protein arginine methyltransferase PRMT3. The EMBO Journal 2000, 19, 3509-3519.  33.  Zhang, X.; Cheng, X. Structure of the Predominant Protein Arginine Methyltransferase PRMT1 and Analysis of Its Binding to Substrate Peptides. Structure 2003, 11, 509-520.  34.  Troffer-Charlier, N.; Cura, V.; Hassenboehler, P.; Moras, D.; Cavarelli, J. Functional insights from structures of coactivator-associated arginine methyltransferase 1 domains. The EMBO Journal 2007, 26, 4391-4401.  35.  Rust, H. L.; Zurita-Lopez, C. I.; Clarke, S.; Thompson, P. R. Mechanistic Studies on Transcriptional Coactivator Protein Arginine Methyltransferase 1. Biochemistry 2011, 50, 3332-3345.  36.  Antonysamy, S.; Bonday, Z.; Campbell, R. M.; Doyle, B.; Druzina, Z.; Gheyi, T.; Han, B.; Jungheim, L. N.; Qian, Y.; Rauch, C. Crystal structure of the human PRMT5: MEP50 complex. Proceedings of the National Academy of Sciences 2012, 109, 17960-17965.  37.  Lee, D. Y.; Ianculescu, I.; Purcell, D.; Zhang, X.; Cheng, X.; Stallcup, M. R. Surface-Scanning Mutational Analysis of Protein Arginine Methyltransferase 1: Roles of Specific Amino Acids in Methyltransferase Substrate Specificity, Oligomerization, and Coactivator Function. Molecular Endocrinology 2007, 21, 1381-1393.  38.  Cura, V.; Troffer-Charlier, N.; Wurtz, J.-M.; Bonnefond, L.; Cavarelli, J. Structural insight into arginine methylation by the mouse protein arginine methyltransferase 7: a zinc finger freezes the mimic of the dimeric state into a single active site. Acta Crystallographica Section D: Biological Crystallography 2014, 70, 2401-2412.  39.  Wang, C.; Zhu, Y.; Caceres, T.; Liu, L.; Peng, J.; Wang, J.; Chen, J.; Chen, X.; Zhang, Z.; Zuo, X.; Gong, Q.; Teng, M.; Hevel, J.; Wu, J.; Shi, Y. Structural Determinants for the Strict Monomethylation Activity by Trypanosoma brucei Protein Arginine Methyltransferase 7. Structure 2014, 22, 756-768. 107   40.  Sun, L.; Wang, M.; Lv, Z.; Yang, N.; Liu, Y.; Bao, S.; Gong, W.; Xu, R. M. Structural insights into protein arginine symmetric dimethylation by PRMT5. Proceedings of the National Academy of Sciences 2011, 108, 20538-20543.  41.  Gui, S.; Gathiaka, S.; Li, J.; Qu, J.; Acevedo, O.; Hevel, J. M. A Remodeled Protein Arginine Methyltransferase 1 (PRMT1) Generates Symmetric Dimethylarginine. Journal of Biological Chemistry 2014, 289, 9320-9327.  42.  Kölbel, K.; Ihling, C.; Kühn, U.; Neundorf, I.; Otto, S.; Stichel, J.; Robaa, D.; Beck-Sickinger, A. G.; Sinz, A.; Wahle, E. Peptide Backbone Conformation Affects the Substrate Preference of Protein Arginine Methyltransferase I. Biochemistry 2012, 51, 5463-5475.  43.  Thandapani, P.; O-Connor, T. R.; Bailey, T. L.; Richard, S. p. Defining the RGG/RG motif. Molecular Cell 2013, 50, 613-623.  44.  Osborne, T. C.; Obianyo, O.; Zhang, X.; Cheng, X.; Thompson, P. R. Protein Arginine Methyltransferase 1: Positively Charged Residues in Substrate Peptides Distal to the Site of Methylation Are Important for Substrate Binding and Catalysis. Biochemistry 2007, 46, 13370-13381.  45.  Wang, M.; Fuhrmann, J.; Thompson, P. R. Protein Arginine Methyltransferase 5 Catalyzes Substrate Dimethylation in a Distributive Fashion. Biochemistry 2014, 53, 7884-7892.  46.  Herrmann, F.; Pably, P.; Eckerich, C.; Bedford, M. T.; Fackelmayer, F. O. Human protein arginine methyltransferases in vivo  distinct properties of eight canonical members of the PRMT family. Journal of Cell Science 2009, 122, 667-677.  47.  Lee, J.; Sayegh, J.; Daniel, J.; Clarke, S.; Bedford, M. T. PRMT8, a New Membrane-bound Tissue-specific Member of the Protein Arginine Methyltransferase Family. Journal of Biological Chemistry 2005, 280, 32890-32896.  48.  Frankel, A.; Yadav, N.; Lee, J.; Branscombe, T. L.; Clarke, S.; Bedford, M. T. The Novel Human Protein Arginine N-Methyltransferase PRMT6 Is a Nuclear Enzyme Displaying Unique Substrate Specificity. Journal of Biological Chemistry 2002, 277, 3537-3543.  49.  Pawlak, M. R.; Scherer, C. A.; Chen, J.; Roshon, M. J.; Ruley, H. E. Arginine N-Methyltransferase 1 Is Required for Early Postimplantation Mouse Development, but Cells Deficient in the Enzyme Are Viable. Molecular and Cellular Biology 2000, 20, 4859-4869.  50.  Tang, J.; Frankel, A.; Cook, R. J.; Kim, S.; Paik, W. K.; Williams, K. R.; Clarke, S.; Herschman, H. R. PRMT1 Is the Predominant Type I Protein Arginine Methyltransferase in Mammalian Cells. Journal of Biological Chemistry 2000, 275, 7723-7730.  51.  Goulet, I.; Gauvin, G.; Boisvenue, S.; Coté, J. Alternative Splicing Yields Protein Arginine Methyltransferase 1 Isoforms with Distinct Activity, Substrate Specificity, and Subcellular Localization. Journal of Biological Chemistry 2007, 282, 33009-33021.  52.  Feng, Q.; He, B.; Jung, S. Y.; Song, Y.; Qin, J.; Tsai, S. Y.; Tsai, M. J.; O'Malley, B. W. Biochemical control of CARM1 enzymatic activity by phosphorylation. Journal of Biological Chemistry 2009, 284, 36167-36174.  53.  Friesen, W. J.; Wyce, A.; Paushkin, S.; Abel, L.; Rappsilber, J.; Mann, M.; Dreyfuss, G. A Novel WD Repeat Protein Component of the Methylosome Binds Sm Proteins. Journal of Biological Chemistry 2002, 277, 8243-8247.  54.  Wang, L.; Pal, S.; Sif, S. d. Protein Arginine Methyltransferase 5 Suppresses the Transcription of the RB Family of Tumor Suppressors in Leukemia and Lymphoma Cells. Molecular and Cellular Biology 2008, 28, 6262-6277.  55.  Feng, Y.; Wang, J.; Asher, S.; Hoang, L.; Guardiani, C.; Ivanov, I.; Zheng, Y. G. Histone H4 acetylation differentially modulates arginine methylation by an in Cis mechanism. Journal of Biological Chemistry 2011, 286, 20323-20334.  56.  Gkountela, S.; Li, Z.; Chin, C.; Lee, S.; Clark, A. PRMT5 is Required for Human Embryonic Stem Cell Proliferation But Not Pluripotency. Stem Cell Rev and Rep 2014, 10, 230-239. 108   57.  Yadav, N.; Lee, J.; Kim, J.; Shen, J.; Hu, M. C. T.; Aldaz, C. M.; Bedford, M. T. Specific protein methylation defects and gene expression perturbations in coactivator-associated arginine methyltransferase 1-deficient mice. Proceedings of the National Academy of Sciences 2003, 100, 6464-6468.  58.  Waldmann, T.; Izzo, A.; Kamieniarz, K.; Richter, F.; Vogler, C.; Sarg, B.; Lindner, H.; Young, N. L.; Mittler, G. M. J.; Garcia, B. A. Methylation of H2AR29 is a novel repressive PRMT6 target. Epigenetics & Chromatin 2011, 4.  59.  Tsai, W. W.; Niessen, S.; Goebel, N.; Yates, J. R.; Guccione, E.; Montminy, M. PRMT5 modulates the metabolic response to fasting signals. Proceedings of the National Academy of Sciences 2013, 110, 8870-8875.  60.  Yoshimatsu, M.; Toyokawa, G.; Hayami, S.; Unoki, M.; Tsunoda, T.; Field, H. I.; Kelly, J. D.; Neal, D. E.; Maehara, Y.; Ponder, B. A. J.; Nakamura, Y.; Hamamoto, R. Dysregulation of PRMT1 and PRMT6, Type I arginine methyltransferases, is involved in various types of human cancers. Int. J. Cancer 2011, 128, 562-573.  61.  Cheung, N.; Chan, L. C.; Thompson, A.; Cleary, M. L.; So, C. W. E. Protein arginine-methyltransferase-dependent oncogenesis. Nat Cell Biol 2007, 9, 1208-1215.  62.  Hong, H.; Kao, C.; Jeng, M. H.; Eble, J. N.; Koch, M. O.; Gardner, T. A.; Zhang, S.; Li, L.; Pan, C. X.; Hu, Z.; MacLennan, G. T.; Cheng, L. Aberrant expression of CARM1, a transcriptional coactivator of androgen receptor, in the development of prostate carcinoma and androgen-independent status. Cancer 2004, 101, 83-89.  63.  Kim, Y. R.; Lee, B.; Park, R. Y.; Nguyen, N.; Bae, J.; Kwon, D.; Jung, C. Differential CARM1 expression in prostate and colorectal cancers. BMC Cancer 2010, 10, 197.  64.  Wei, T. Y.; Juan, C. C.; Hisa, J. Y.; Su, L. J.; Lee, Y. C.; Chou, H. Y.; Chen, J. M.; Wu, Y. C.; Chiu, S. C.; Hsu, C. P.; Liu, K. L.; Yu, C. T. Protein arginine methyltransferase 5 is a potential oncoprotein that upregulates G1 cyclins/cyclin-dependent kinases and the phosphoinositide 3-kinase/AKT signaling cascade. Cancer Sci 2012, 103, 1640-1650.  65.  Darnell, J. E. Transcription factors as targets for cancer therapy. Nat Rev Cancer 2002, 2, 740-749.  66.  Yan, L.; Yan, C.; Qian, K.; Su, H.; Kofsky-Wofford, S. A.; Lee, W. C.; Zhao, X.; Ho, M. C.; Ivanov, I.; Zheng, Y. G. Diamidine Compounds for Selective Inhibition of Protein Arginine Methyltransferase 1. Journal of medicinal chemistry 2014, 57, 2611-2622.  67.  Spannhoff, A.; Machmur, R.; Heinke, R.; Trojer, P.; Bauer, I.; Brosch, G.; Schüle, R.; Hanefeld, W.; Sippl, W.; Jung, M. A novel arginine methyltransferase inhibitor with cellular activity. Bioorganic & medicinal chemistry letters 2007, 17, 4150-4153.  68.  Spannhoff, A.; Heinke, R.; Bauer, I.; Trojer, P.; Metzger, E.; Gust, R.; Schüle, R.; Brosch, G.; Sippl, W.; Jung, M. Target-Based Approach to Inhibitors of Histone Arginine Methyltransferases. Journal of medicinal chemistry 2007, 50, 2319-2325.  69.  Siarheyeva, A.; Senisterra, G.; Allali-Hassani, A.; Dong, A.; Dobrovetsky, E.; Wasney, G.; Chau, I.; Marcellus, R.; Hajian, T.; Liu, F.; Korboukh, I.; Smil, D.; Bolshan, Y.; Min, J.; Wu, H.; Zeng, H.; Loppnau, P.; Poda, G.; Griffin, C.; Aman, A.; Brown, P.; Jin, J.; Al-awar, R.; Arrowsmith, C.; Schapira, M.; Vedadi, M. An Allosteric Inhibitor of Protein Arginine Methyltransferase 3. Structure 2012, 20, 1425-1435.  70.  Purandare, A. V.; Chen, Z.; Huynh, T.; Pang, S.; Geng, J.; Vaccaro, W.; Poss, M. A.; Oconnell, J.; Nowak, K.; Jayaraman, L. Pyrazole inhibitors of coactivator associated arginine methyltransferase 1 (CARM1). Bioorganic & medicinal chemistry letters 2008, 18, 4438-4441.  71.  Mitchell, L. H.; Drew, A. E.; Ribich, S. A.; Rioux, N.; Swinger, K. K.; Jacques, S. L.; Lingaraj, T.; Boriack-Sjodin, P. A.; Waters, N. J.; Wigle, T. J.; Moradei, O.; Jin, L.; Riera, T.; Porter-Scott, M.; Moyer, M. P.; Smith, J. J.; Chesworth, R.; Copeland, R. A. Aryl Pyrazoles as Potent Inhibitors of Arginine Methyltransferases: Identification of the First PRMT6 Tool Compound. ACS medicinal chemistry letters 2015, 6, 655-659. 109   72.  Kaniskan, H. Ü.; Szewczyk, M. M.; Yu, Z.; Eram, M. S.; Yang, X.; Schmidt, K.; Luo, X.; Dai, M.; He, F.; Zang, I.; Lin, Y.; Kennedy, S.; Li, F.; Dobrovetsky, E.; Dong, A.; Smil, D.; Min, S. J.; Landon, M.; Lin-Jones, J.; Huang, X. P.; Roth, B. L.; Schapira, M.; Atadja, P.; Barsyte-Lovejoy, D.; Arrowsmith, C. H.; Brown, P. J.; Zhao, K.; Jin, J.; Vedadi, M. A Potent, Selective and Cell-Active Allosteric Inhibitor of Protein Arginine Methyltransferase3 (PRMT3). Angewandte Chemie International Edition 2015, 54, 5166-5170.  73.  John, S. S.; Sandrine, T.; Tiziano, B.; Marina, F.; Gerald, J. D.; Lata, J.; Kevin, F. K.; Herbert, E. K.; Ashok, V. P.; Pamela, R. Structural basis for CARM1 inhibition by indole and pyrazole inhibitors. Biochemical Journal 2011, 436, 331-339.  74.  Huynh, T.; Chen, Z.; Pang, S.; Geng, J.; Bandiera, T.; Bindi, S.; Vianello, P.; Roletto, F.; Thieffine, S.; Galvani, A.; Vaccaro, W.; Poss, M. A.; Trainor, G. L.; Lorenzi, M. V.; Gottardis, M.; Jayaraman, L.; Purandare, A. V. Optimization of pyrazole inhibitors of Coactivator Associated Arginine Methyltransferase 1 (CARM1). Bioorganic & medicinal chemistry letters 2009, 19, 2924-2927.  75.  Feng, Y.; Li, M.; Wang, B.; Zheng, Y. G. Discovery and Mechanistic Study of a Class of Protein Arginine Methylation Inhibitors. Journal of medicinal chemistry 2010, 53, 6028-6039.  76.  Cheng, D.; Yadav, N.; King, R. W.; Swanson, M. S.; Weinstein, E. J.; Bedford, M. T. Small Molecule Regulators of Protein Arginine Methyltransferases. Journal of Biological Chemistry 2004, 279, 23892-23899.  77.  Chan-Penebre, E.; Kuplast, K. G.; Majer, C. R.; Boriack-Sjodin, P. A.; Wigle, T. J.; Johnston, L. D.; Rioux, N.; Munchhof, M. J.; Jin, L.; Jacques, S. L. A selective inhibitor of PRMT5 with in vivo and in vitro potency in MCL models. Nature chemical biology 2015.  78.  Bissinger, E. M.; Heinke, R.; Spannhoff, A.; Eberlin, A.; Metzger, E.; Cura, V.; Hassenboehler, P.; Cavarelli, J.; Schüle, R.; Bedford, M. T.; Sippl, W.; Jung, M. Acyl derivatives of p-aminosulfonamides and dapsone as new inhibitors of the arginine methyltransferase hPRMT1. Bioorganic & Medicinal Chemistry 2011, 19, 3717-3731.  79.  Broom, A. D. Rational design of enzyme inhibitors: multisubstrate analog inhibitors. Journal of medicinal chemistry 1989, 32, 2-7.  80.  Dowden, J.; Hong, W.; Parry, R. V.; Pike, R. A.; Ward, S. G. Toward the development of potent and selective bisubstrate inhibitors of protein arginine methyltransferases. Bioorganic & medicinal chemistry letters 2010, 20, 2103-2105.  81.  van Haren, M.; van Ufford, L. Q.; Moret, E. E.; Martin, N. I. Synthesis and evaluation of protein arginine N-methyltransferase inhibitors designed to simultaneously occupy both substrate binding sites. Organic & biomolecular chemistry 2015, 13, 549-560.  82.  Osborne, T.; Weller Roska, R. L.; Rajski, S. R.; Thompson, P. R. In Situ Generation of a Bisubstrate Analogue for Protein Arginine Methyltransferase 1. J. Am. Chem. Soc. 2008, 130, 4574-4575.  83.  Weller, R. L.; Rajski, S. R. Design, Synthesis, and Preliminary Biological Evaluation of a DNA MethyltransferaseDirected Alkylating Agent. ChemBioChem 2006, 7, 243-245.  84.  Obianyo, O.; Causey, C. P.; Osborne, T. C.; Jones, J. E.; Lee, Y. H.; Stallcup, M. R.; Thompson, P. R. A Chloroacetamidine-Based Inactivator of Protein Arginine Methyltransferase 1: Design, Synthesis, and In Vitro and In Vivo Evaluation. ChemBioChem 2010, 11, 1219-1223.  85.  Bicker, K. L.; Obianyo, O.; Rust, H. L.; Thompson, P. R. A combinatorial approach to characterize the substrate specificity of protein arginine methyltransferase 1. Molecular bioSystems 2011, 7, 48-51.  86.  Obianyo, O.; Causey, C. P.; Jones, J. E.; Thompson, P. R. Activity-Based Protein Profiling of Protein Arginine Methyltransferase 1. ACS Chem. Biol. 2011, 6, 1127-1135.  87.  Smith, G. P. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science 1985, 228, 1315-1317. 110   88.  Sidhu, S. S.; Lowman, H. B.; Cunningham, B. C.; Wells, J. A. Phage display for selection of novel binding peptides. In Methods in Enzymology Applications of Chimeric Genes and Hybrid Proteins - Part C: Protein-Protein Interactions and Genomics, Volume 328 ed.; Jeremy Thorner, S. D. E. Ed.; Academic Press: 2000; p 333-IN5.  89.  Kay, B. K.; Hamilton, P. T. Identification of Enzyme Inhibitors from Phage-Displayed Combinatorial Peptide Libraries. Combinatorial Chemistry & High Throughput Screening 2001, 4, 477-496.  90.  Gray, B. P.; Li, S.; Brown, K. C. From Phage Display to Nanoparticle Delivery: Functionalizing Liposomes with Multivalent Peptides Improves Targeting to a Cancer Biomarker. Bioconjugate Chem. 2013, 24, 85-96.  91.  Rodi, D. J.; Makowski, L. Phage-display technology  finding a needle in a vast molecular haystack. Current Opinion in Biotechnology 1999, 10, 87-93.  92.  Kay, B. K.; Kasanov, J.; Knight, S.; Kurakin, A. Convergent evolution with combinatorial peptides. FEBS Letters 2000, 480, 55-62.  93.  Avan, I.; Hall, C. D.; Katritzky, A. R. Peptidomimetics via modifications of amino acids and peptide bonds. Chem. Soc. Rev. 2014, 43, 3575-3594.  94.  Grauer, A.; König, B. Peptidomimetics  A Versatile Route to Biologically Active Compounds. Eur. J. Org. Chem. 2009, 2009, 5099-5111.  95.  Stewart, J. M. The design of peptide hormone analogs. Trends in Pharmacological Sciences 1982, 3, 300-303.  96.  Berg, J. M.; Tymoczko, J. L.; Stryer, L. Biochemistry., Sixth ed.; Freeman: 2006; pp 34-40.  97.  Alberts, B.; Johnson, A.; Lewis, J.; Raff, M.; Roberts, K.; Walter, P. The Cell, Molecular Biology of; Garland Science: 2008.  98.  Maes, V.; Tourwe, D. Aspects of Peptidomimetics. Peptide and Protein Design for Biopharmaceutical Applications 2009, 49.  99.  Vlieghe, P.; Lisowski, V.; Martinez, J.; Khrestchatisky, M. Synthetic therapeutic peptides: science and market. Drug Discovery Today 2010, 15, 40-56.  100.  Moore, G. J. Designing peptide mimetics. Trends in Pharmacological Sciences 1994, 15, 124-129.  101.  Ripka, A. S.; Rich, D. H. Peptidomimetic design. Current Opinion in Chemical Biology 1998, 2, 441-452.  102.  Gante, J. PeptidomimeticsTailored enzyme inhibitors. Angewandte Chemie International Edition in English 1994, 33, 1699-1720.  103.  Chatterjee, J.; Rechenmacher, F.; Kessler, H. NMethylation of Peptides and Proteins: An Important Element for Modulating Biological Functions. Angewandte Chemie International Edition 2013, 52, 254-269.  104.  Gibson, S. E.; Guillo, N.; Tozer, M. J. Towards control of Chi-space: conformationally constrained analogues of Phe, Tyr, Trp and His. Tetrahedron 1999, 55, 585-615.  105.  White, C. J.; Yudin, A. K. Contemporary strategies for peptide macrocyclization. Nature chemistry 2011, 3, 509-524.  106.  Ramachandran, G. N.; Ramakrishnan, C. T.; Sasisekharan, V. Stereochemistry of polypeptide chain configurations. Journal of molecular biology 1963, 7, 95-99.  107.  Hruby, V. J. Conformational restrictions of biologically active peptides via amino acid side chain groups. Life Sciences 1982, 31, 189-199.  108.  Cowell, S. M.; Lee, Y. S.; Cain, J. P.; Hruby, V. J. Exploring Ramachandran and Chi space: conformationally constrained amino acids and peptides in the design of bioactive polypeptide ligands. Current medicinal chemistry 2004, 11, 2785-2798.  109.  Kazmierski, W. M.; Yamamura, H. I.; Hruby, V. J. Topographic design of peptide neurotransmitters and hormones on stable backbone templates: relation of conformation and dynamics to bioactivity. J. Am. Chem. Soc. 1991, 113, 2275-2283. 111   110.  Tourwé, D.; Verschueren, K.; Frycia, A.; Davis, P.; Porreca, F.; Hruby, V. J.; Toth, G.; Jaspers, H.; Verheyden, P.; Van Binst, G. Conformational restriction of Tyr and Phe side chains in opioid peptides: Information about preferred and bioactive sidechain topology. Biopolymers 1996, 38, 1-12.  111.  Flynn, G. A.; Giroux, E. L.; Dage, R. C. An acyl-iminium ion cyclization route to a novel conformationally restricted dipeptide mimic: applications to angiotensin-converting enzyme inhibition. J. Am. Chem. Soc. 1987, 109, 7914-7915.  112.  Casimir, J. R.; Tourwé, D.; Iterbeke, K.; Guichard, G.; Briand, J. P. Efficient Synthesis of (S)-4-Phthalimido-1,3,4,5- tetrahydro-8-(2,6-dichlorobenzyloxy)-3-oxo-2H-2-benzazepin-2-acetic Acid (Pht-Hba(2,6-Cl2-Bn)-Gly-OH)1. J. Org. Chem. 2000, 65, 6487-6492.  113.  Pulka, K.; Feytens, D.; Van den Eynde, I.; De Wachter, R.; Kosson, P.; Misicka, A.; Lipkowski, A.; Chung, N. N.; Schiller, P. W.; Tourwé, D. Synthesis of 4-amino-3-oxo-tetrahydroazepino [3, 4-b] indoles: new conformationally constrained Trp analogs. Tetrahedron 2007, 63, 1459-1466.  114.  Buysse, K.; Farard, J.; Nikolaou, A.; Vanderheyden, P.; Vauquelin, G.; Sejer Pedersen, D.; Tourwé, D.; Ballet, S. Amino triazolo diazepines (Ata) as constrained histidine mimics. Organic letters 2011, 13, 6468-6471.  115.  Schiller, P. W.; Weltrowska, G.; Berezowska, I.; Nguyen, T. M.-D.; Wilkes, B. C.; Lemieux, C.; Chung, N. N. The TIPP opioid peptide family: development of  antagonists,  agonists, and mixed  agonist/ antagonists. Peptide Science 1999, 51, 411-425.  116.  Guerrini, R.; Capasso, A.; Marastoni, M.; Bryant, S. D.; Cooper, P. S.; Lazarus, L. H.; Temussi, P. A.; Salvadori, S. Rational design of dynorphin A analogues with -receptor selectivity and antagonism for - and -receptors. Bioorganic & Medicinal Chemistry 1998, 6, 57-62.  117.  Ballet, S.; Salvadori, S.; Trapella, C.; Bryant, S. D.; Jinsmaa, Y.; Lazarus, L. H.; Negri, L.; Giannini, E.; Lattanzi, R.; Tourwé, D.; Balboni, G. New 2',6'-Dimethyl-l-tyrosine (Dmt) Opioid Peptidomimetics Based on the Aba-Gly Scaffold. Development of Unique -Opioid Receptor Ligands. Journal of medicinal chemistry 2006, 49, 3990-3993.  118.  Van den Eynde, I.; Laus, G.; Schiller, P. W.; Kosson, P.; Chung, N. N.; Lipkowski, A. W.; Tourwé, D. A New Structural Motif for -Opioid Antagonists. Journal of medicinal chemistry 2005, 48, 3644-3648.  119.  Hock, F. J.; Wirth, K.; Albus, U.; Linz, W.; Gerhards, H. J.; Wiemer, G.; Henke, S. T.; Breipohl, G.; König, W.; Knolle, J. Hoe 140 a new potent and long acting bradykininantagonist: in vitro studies. British journal of pharmacology 1991, 102, 769-773.  120.  Kyle, D. J.; Martin, J. A.; Farmer, S. G.; Burch, R. M. Design and conformational analysis of several highly potent bradykinin receptor antagonists. Journal of medicinal chemistry 1991, 34, 1230-1233.  121.  Ballet, S.; De Wachter, R.; Van Rompaey, K.; Tomboly, C.; Feytens, D.; Toth, G.; Quartara, L.; Cucchi, P.; Meini, S.; Tourwé, D. Bradykinin analogs containing the 4amino-2-benzazepin-3-one scaffold at the C-terminus. Journal of Peptide Science 2007, 13, 164-170.  122.  Ballet, S.; Mayorov, A. V.; Cai, M.; Tymecka, D.; Chandler, K. B.; Palmer, E. S.; Van Rompaey, K.; Misicka, A.; Tourwé, D.; Hruby, V. J. Novel selective human melanocortin-3 receptor ligands: use of the 4-amino-1, 2, 4, 5-tetrahydro-2-benzazepin-3-one (Aba) scaffold. Bioorganic & medicinal chemistry letters 2007, 17, 2492-2498.  123.  Van der Poorten, O.; Fehér, K.; Buysse, K.; Feytens, D.; Zoi, I.; Schwartz, S. D.; Martins, J. C.; Tourwé, D.; Cai, M.; Hruby, V. J.; Ballet, S. Azepinone-Containing Tetrapeptide Analogues of Melanotropin Lead to Selective h MC4R Agonists and h MC5R Antagonist. ACS medicinal chemistry letters 2014, 6, 192-197. 112   124.  Feytens, D.; De Vlaeminck, M.; Cescato, R.; Tourwé, D.; Reubi, J. C. Highly Potent 4-Amino-Indolo [2, 3-c] Azepin-3-One-Containing Somatostatin Mimetics With a Range of sst Receptor Selectivities. Journal of medicinal chemistry 2008, 52, 95-104.  125.  New England Biolabs. Ph.D. Phage Display Libraries Manual.  2014.   126.  Hyde-DeRuyscher, R.; Paige, L. A.; Christensen, D. J.; Hyde-DeRuyscher, N.; Lim, A.; Fredericks, Z. L.; Kranz, J.; Gallant, P.; Zhang, J.; Rocklage, S. M.; Fowlkes, D. M.; Wendler, P. A.; Hamilton, P. T. Detection of small-molecule enzyme inhibitors with peptides isolated from phage-displayed combinatorial peptide libraries. Chemistry & Biology 2000, 7, 17-25.  127.  Burley, S. K.; Petsko, G. A. Aromatic-aromatic interaction: a mechanism of protein structure stabilization. Science 1985, 229, 23-28.  128.  Burley, S. K.; Petsko, G. A. Amino-aromatic interactions in proteins. FEBS letters 1986, 203, 139-143.  129.  Gallivan, J. P.; Dougherty, D. A. Cation- interactions in structural biology. Proceedings of the National Academy of Sciences 1999, 96, 9459-9464.  130.  Ma, J. C.; Dougherty, D. A. The cation- interaction. Chemical reviews 1997, 97, 1303-1324.  131.  Hruby, V. J. Designing peptide receptor agonists and antagonists. Nat Rev Drug Discov 2002, 1, 847-858.  132.  Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. -Peptides: from structure to function. Chem. Rev. 2001, 101, 3219-3232.  133.  Seebach, D.; Gardiner, J. -Peptidic Peptidomimetics. Acc. Chem. Res. 2008, 41, 1366-1375.  134.  Linder, M. R.; Steurer, S.; Podlech, J. (S)-3-(tert-Butyloxycarbonylamino)-4-Phenylbutanoic Acid. Organic syntheses 2003, 154.  135.  Sagan, S.; Karoyan, P.; Lequin, O.; Chassaing, G.; Lavielle, S. N-and C-methylation in biologically active peptides: synthesis, structural and functional aspects. Current medicinal chemistry 2004, 11, 2799-2822.  136.  Sikorska, E.; Slusarz, M. J.; Slusarz, R.; Kowalczyk, W.; Lammek, B. Investigation of cis/trans ratios of peptide bonds in AVP analogues containing N-methylphenylalanine enantiomers. Journal of Peptide Science 2006, 12, 13-24.  137.  Biron, E.; Kessler, H. Convenient synthesis of N-methylamino acids compatible with Fmoc solid-phase peptide synthesis. J. Org. Chem. 2005, 70, 5183-5189.  138.  Yang, L.; Chiu, K. Solid phase synthesis of Fmoc N-methyl amino acids: Application of the Fukuyama amine synthesis. Tetrahedron letters 1997, 38, 7307-7310.  139.  Biron, E.; Chatterjee, J.; Kessler, H. Optimized selective Nmethylation of peptides on solid support. Journal of Peptide Science 2006, 12, 213-219.  140.  Feytens, D.; De Vlaeminck, M.; Tourwé, D. A novel solid phase approach to Aia-containing peptides. Journal of Peptide Science 2009, 15, 16-22.  141.  Iterbeke, K.; Laus, G.; Verheyden, P.; Tourwé, D. Side reactions in the preparation of 1,2,3,4-tetrahydro-carboline-3-carboxylic acid. Lett Pept Sci 1998, 5, 121-123.  142.  Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Color test for detection of free terminal amino groups in the solid-phase synthesis of peptides. Analytical Biochemistry 1970, 34, 595-598.  143.  CHRISTENSEN, T. H. O. R. Qualitative test for monitoring coupling completeness in solid phase peptide synthesis using chloranil. Acta Chem Scand Ser B 1979, 33.  144.  Hancock, W. S.; Battersby, J. E. A new micro-test for the detection of incomplete coupling reactions in solid-phase peptide synthesis using 2,4,6-trinitrobenzene-sulphonic acid. Analytical Biochemistry 1976, 71, 260-264.  145.  Lakowski, T.; Frankel, A. Kinetic analysis of human protein arginine N-methyltransferase 2: formation of monomethyl-and asymmetric dimethyl-arginine residues on histone H4. Biochem. J 2009, 421, 253-261. 113   146.  Pak, M. L.; Lakowski, T. M.; Thomas, D.; Vhuiyan, M. I.; Hüsecken, K.; Frankel, A. A protein arginine N-methyltransferase 1 (PRMT1) and 2 heteromeric interaction increases PRMT1 enzymatic activity. Biochemistry 2011, 50, 8226-8240.  147.  Zurita-Lopez, C. I.; Sandberg, T.; Kelly, R.; Clarke, S. G. Human Protein Arginine Methyltransferase 7 (PRMT7) Is a Type III Enzyme Forming -NG-Monomethylated Arginine Residues. Journal of Biological Chemistry 2012, 287, 7859-7870.  148.  Wider, G.; Dreier, L. Measuring Protein Concentrations by NMR Spectroscopy. J. Am. Chem. Soc. 2006, 128, 2571-2576.  149.  Wu, J.; Xie, N.; Feng, Y.; Zheng, Y. G. Scintillation Proximity Assay of Arginine Methylation. Journal of Biomolecular Screening 2012, 17, 237-244.  150.  Greenfield, N. J. Using circular dichroism spectra to estimate protein secondary structure. Nat. Protocols 2007, 1, 2876-2890.  151.  Maupetit, J.; Derreumaux, P.; Tuffery, P. PEP-FOLD: an online resource for de novo peptide structure prediction. Nucleic Acids Research 2009, 37, W498-W503.  152.  Escobedo, J. O.; Rusin, O.; Lim, S.; Strongin, R. M. NIR dyes for bioimaging applications. Current Opinion in Chemical Biology 2010, 14, 64-70.  153.  Arndt-Jovin, D. J.; Jovin, T. M. Fluorescence labeling and microscopy of DNA. Methods Cell Biol 1989, 30, 417-448.  154.  Maliwal, B. P.; Fudala, R.; Raut, S.; Kokate, R.; Sørensen, T. J.; Laursen, B. W.; Gryczynski, Z.; Gryczynski, I. Long-lived bright red emitting azaoxa-triangulenium fluorophores. PLoS ONE 2013.  155.  Leurs, U.; Lohse, B.; Ming, S.; Cole, P. A.; Clausen, R. P.; Kristensen, J. L.; Rand, K. D. Dissecting the Binding Mode of Low Affinity Phage Display Peptide Ligands to Protein Targets by Hydrogen/Deuterium Exchange Coupled to Mass Spectrometry. Analytical chemistry 2014, 86, 11734-11741.  156.  Johnson, B. A.; Blevins, R. A. NMR View: A computer program for the visualization and analysis of NMR data. Journal of biomolecular NMR 1994, 4, 603-614.  157.  Wuthrich, K. NMR of proteins and nucleic acids; Wiley: 1986.  158.  Case, D. A.; Darden, T. A.; Cheatham, T. E.; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M. Amber 11.  2010.  University of California.   159.  Koradi, R.; Billeter, M.; W++thrich, K. MOLMOL: a program for display and analysis of macromolecular structures. Journal of molecular graphics 1996, 14, 51-55.  160.  Taverna, S. D.; Li, H.; Ruthenburg, A. J.; Allis, C. D.; Patel, D. J. How chromatin-binding modules interpret histone modifications: lessons from professional pocket pickers. Nature structural & molecular biology 2007, 14, 1025-1040.  161.  Tripsianes, K.; Madl, T.; Machyna, M.; Fessas, D.; Englbrecht, C.; Fischer, U.; Neugebauer, K. M.; Sattler, M. Structural basis for dimethylarginine recognition by the Tudor domains of human SMN and SPF30 proteins. Nat Struct Mol Biol 2011, 18, 1414-1420.  162.  Feichtinger, K.; Zapf, C.; Sings, H. L.; Goodman, M. Diprotected triflylguanidines: a new class of guanidinylation reagents. The Journal of organic chemistry 1998, 63, 3804-3805.  163.  Feichtinger, K.; Sings, H. L.; Baker, T. J.; Matthews, K.; Goodman, M. Triurethane-protected guanidines and triflyldiurethane-protected guanidines: new reagents for guanidinylation reactions. The Journal of organic chemistry 1998, 63, 8432-8439.  164.  Sweeney, J. B. Aziridines: epoxides' ugly cousins? Chemical Society Reviews 2002, 31, 247-258.  165.  McCoull, W.; Davis, F. A. Recent synthetic applications of chiral aziridines. Synthesis 2000, 2000, 1347-1365.  166.  Farthing, C. N.; Baldwin, J. E.; Russell, A. T.; Schofield, C. J.; Spivey, A. C. Syntheses of (S)-pyrazolylalanine and (S)-quisqualic acid from a serine-derived aziridine. Tetrahedron Letters 1996, 37, 5225-5226. 114   167.  Harada, H.; Morie, T.; Suzuki, T.; Yoshida, T.; Kato, S. Efficient synthesis of (R)-6-benzyloxycarbonylamino-1-methyl-4-(3-methyl-benzyl) hexahydro-1, 4-diazepine. 2. Novel formation of hexahydro-1, 4-diazepine ring using 1, 2, 3-trisubstituted aminopropane derivative and glyoxal. Tetrahedron 1998, 54, 10671-10676.  168.  Kato, S.; Harada, H.; Morie, T. Efficient synthesis of (6R)-6-amino-1-methyl-4-(3-methylbenzyl)hexahydro-1H-1,4-diazepine from methyl (2R)- and (2S)-1-benzyloxycarbonylaziridine-2-carboxylates. J. Chem. Soc. , Perkin Trans. 1 1997, 3219-3226.  169.  Davis, F. A.; Reddy, G. V.; Liang, C. H. Aziridine 2-carboxylate ester mediated asymmetric synthesis of -alkyl -amino acids. Tetrahedron Letters 1997, 38, 5139-5142.  170.  Ham, G. E. Activated Aziridines. I. Reaction of Anilines with O-Ethyl-N, N-Ethyleneurethane. Mechanism and Hammett -Constant. The Journal of organic chemistry 1964, 29, 3052-3055.  171.  Fukuyama, T.; Jow, C. K.; Cheung, M. 2- and 4-Nitrobenzenesulfonamides: Exceptionally versatile means for preparation of secondary amines and protection of amines. Tetrahedron Letters 1995, 36, 6373-6374.  172.  Baldwin, J. E. Rules for ring closure. J. Chem. Soc. , Chem. Commun. 1976, 734-736.  173.  Cardillo, G.; Gentilucci, L.; Tolomelli, A.; Tomasini, C. Ring expansion of N-acyl aziridine-2-imides to oxazoline-4-imides, useful precursors of pure -Hydroxy -aminoacids. Tetrahedron Letters 1997, 38, 6953-6956.  174.  Cardillo, G.; Gentilucci, L.; Tolomelli, A. Synthesis of optically pure threonine-containing dipeptides by regio- and stereo-controlled ring expansion of aziridine-2-imide derivatives. Chem. Commun. 1999, 167-168.  175.  Cardillo, G.; Gentilucci, L.; Tolomelli, A. Dipeptides containing D-serine or D-isoserine from the same (R)-aziridine-2-imide by a simple reversal of the synthetic procedure. Tetrahedron 1999, 55, 15151-15158.  176.  Eastwood, F. W.; Perlmutter, P.; Yang, Q. Preparation of new 24-disubstituted oxazoles. Tetrahedron Letters 1994, 35, 2039-2042.  177.  Eastwood, W.; Perlmutter, P.; Yang, Q. Preparation of new 2,4-disubstituted oxazoles from N-acylaziridines. J. Chem. Soc. , Perkin Trans. 1 1997, 35-42.  178.  Schjoeth-Eskesen, C.; Hansen, P. R.; Kjaer, A.; Gillings, N. Efficient Regioselective Ring Opening of Activated Aziridine-2-Carboxylates with [18F]Fluoride. ChemistryOpen 2015, 4, 65-71.  179.  Lim, Y.; Lee, W. K. Regiospecific reductive ring cleavage of N-substituted aziridine-2-carboxylates and an aziridine-2-methanol via catalytic hydrogenation using Pd as a catalyst. Tetrahedron Letters 1995, 36, 8431-8434.  180.  Wissner, A.; Floyd, M. B.; DUSHIN, R.; Fraser, H. L.; Hu, Y.; Maderna, A.; Nittoli, T.; Wang, Y. D. Benzo[c][2,7]naphthyridine derivatives, and their use as kinase inhibitors.  Page 64. 12-9-2008.  Google Patents.   181.  Monfregola, L.; De Luca, S. Synthetic strategy for side chain mono-N-alkylation of Fmoc-amino acids promoted by molecular sieves. Amino acids 2011, 41, 981-990.  182.  Kim, S. K.; Jacobsen, E. N. General Catalytic Synthesis of Highly Enantiomerically Enriched Terminal Aziridines from Racemic Epoxides. Angewandte Chemie 2004, 116, 4042-4044.  183.  Bucciarelli, M.; Forni, A.; Moretti, I.; Prati, F.; Torre, G. Candida cylindracea lipase-catalysed hydrolysis of methyl aziridine-2-carboxylates and -2,3-dicarboxylates. J. Chem. Soc. , Perkin Trans. 1 1993, 3041-3045.  184.  Stanley, N. J.; Pedersen, D. S.; Nielsen, B.; Kvist, T.; Mathiesen, J. M.; Bräuner-Osborne, H.; Taylor, D. K.; Abell, A. D. 1,2,3-Triazolyl amino acids as AMPA receptor ligands. Bioorganic & Medicinal Chemistry Letters 2010, 20, 7512-7515.  185.  Lau, Y. H.; Spring, D. R. Efficient Synthesis of Fmoc-Protected Azido Amino Acid. Synlett 2011, 13, 1917-1919. 115   186.  Akaji, K.; Aimoto, S. Synthesis of MEN11420, a glycosylated bicyclic peptide, by intramolecular double cyclization using a chloroimidazolinium coupling reagent. Tetrahedron 2001, 57, 1749-1755.  187.  Rao, R. V. R.; Tantry, S. J.; Babu, V. V. S. Practical and Efficient Synthesis of Orthogonally Protected Alpha-2,3-Diaminopropionic Acid (2,3Dap), 2,4-Diaminobutanoic Acid (2,4Dab), and their NMethylated Derivatives. Synthetic Communications 2006, 36, 2901-2912.  188.  De Luca, S.; Moglie, R. D.; De Capua, A.; Morelli, G. New synthetic strategy for o-NBS protected amino acids and their use in synthesis of mono-benzylated peptides. Tetrahedron Letters 2005, 46, 6637-6640.  189.  Studier, F. W.; Moffatt, B. A. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. Journal of molecular biology 1986, 189, 113-130.  190.  Maniatis, T.; Fritsch, E. F.; Sambrook, J. Molecular cloning: a laboratory manual; Cold Spring Harbor Laboratory Cold Spring Harbor, NY: 1982.  191.  Grüner, S.; Neeb, M.; Barandun, L. J.; Sielaff, F.; Hohn, C.; Kojima, S.; Steinmetzer, T.; Diederich, F.; Klebe, G. Impact of protein and ligand impurities on ITC-derived proteinligand thermodynamics. Biochimica et Biophysica Acta (BBA) - General Subjects 2014, 1840, 2843-2850.  192.  Tolia, N. H.; Joshua-Tor, L. Strategies for protein coexpression in Escherichia coli. Nat Meth 2006, 3, 55-64.  193.  GE Healthcare. MicroCalTM Auto-iTC200 System User Manual.  2012.  GE Healthcare.   194.  Hart, P.; Lakowski, T. M.; Thomas, D.; Frankel, A.; Martin, N. I. Peptidic Partial Bisubstrates as Inhibitors of the Protein Arginine NMethyltransferases. ChemBioChem 2011, 12, 1427-1432.  195.  Thomas, D.; Lakowski, T. M.; Pak, M. L.; Kim, J. J.; Frankel, A. Förster resonance energy transfer measurements of cofactordependent effects on protein arginine Nmethyltransferase homodimerization. Protein Science 2010, 19, 2141-2151.  196.  Lakowski, T. M.; Frankel, A. A kinetic study of human protein arginine N-methyltransferase 6 reveals a distributive mechanism. Journal of Biological Chemistry 2008, 283, 10015-10025.  197.  Thomas, D.; Koopmans, T.; Lakowski, T. M.; Kreinin, H.; Vhuiyan, M. I.; Sedlock, S. A.; Bui, J. M.; Martin, N. I.; Frankel, A. Protein Arginine NMethyltransferase Substrate Preferences for Different NSubstituted Arginyl Peptides. ChemBioChem 2014, 15, 1607-1613.  198.  Pace, C. N.; Vajdos, F.; Fee, L.; Grimsley, G.; Gray, T. How to measure and predict the molar absorption coefficient of a protein. Protein science: a publication of the Protein Society 1995, 4, 2411.  199.  Luo, M. Current chemical biology approaches to interrogate protein methyltransferases. ACS Chem. Biol. 2012, 7, 443-463.  200.  Pierce, M. M.; Raman, C. S.; Nall, B. T. Isothermal Titration Calorimetry of ProteinProtein Interactions. Methods 1999, 19, 213-221.  201.  Pochetti, G.; Montanari, R. Isothermal titration calorimetry to determine the association constants for a ligand bound simultaneously to two specific protein binding sites with different affinities. Nature Protocol Exchange 2012.  202.  Owicki, J. C. Fluorescence polarization and anisotropy in high throughput screening: perspectives and primer. Journal of Biomolecular Screening 2000, 5, 297-306.  203.  Invitrogen. PureLink® Quick Plasmid Miniprep Kits.  25-4-2011. 1-12-2015.   A1        Appendix 1: Manuscript 1   A2  Conformationally Constrained Peptidomimetics as Pan-Selective Inhibitors of the Protein Arginine Methyl Transferases  Astrid Knuhtsen,† Baptiste Legrand,‡ Olivier Van der Poorten,§ Muriel Amblard,‡ Jean Martinez,‡ Steven Ballet,§ Jesper L. Kristensen,† Daniel Sejer Pedersen,†,*  † Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100, Copenhagen, Denmark ‡ IBMM, UMR 5247 CNRS, Université de Montpellier, ENSCM, 15 Avenue Charles Flahault, 34000 Montpellier, France § Research Group of Organic Chemistry, Departments of Chemistry and Bio-engineering Sciences, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium  Abstract Protein arginine N-methyl transferases (PRMTs) belong to a family of enzymes that modulate the epigenetic code via modifications of histones. In the present study, peptides emerging from a Phage Display screening were modified in the search for promiscuous PRMT inhibitors via substitution with non-proteinogenic amino acids, N-alkylation of the peptide backbone and incorporation of constrained dipeptide mimics. One of the modified peptides (P18) showed an increased inhibitory activity towards several PRMTs in the low µM range and the conformational preference of this peptide was investigated and compared with the original hit using CD and NMR spectroscopy. Introducing two constrained tryptophan residue mimics (L-Aia) spaced by a single A3  amino acid was found to induce a unique turn structure stabilized by a hydrogen bond and aromatic π-staking interaction between the two L-Aia residues. Introduction Epigenetic enzymes regulate the on/off switching of genes by adding or removing marks on DNA and the histones, which regulate the binding of the transcriptional machinery to the DNA. An epigenetic trait has been defined as a stably heritable phenotype resulting from changes in a chromosome without alterations in the DNA sequence.1 The epigenetic enzymes can be divided into three classes; writers, that leave marks, erasers, that remove marks, and readers, that interpret marks and recruit effector proteins. The interplay of epigenetic marks thus plays an essential role in the translation of genetic information into phenotypes. Consequently, dysregulated epigenetic enzymes can cause unwanted expression or repression of genes, which may be seen in pathophysiological states. Indeed, dysregulation of these enzymes is implicated in many cancerous states.2-4 The family of epigenetic enzymes include the protein arginine N-methyl transferases (PRMTs), lysine methyl transferases (KMTs), histone acetyltransferases (HATs), and histone deacetylases (HDACs) among others.2 In recent years, considerable resources have been invested in exploring these enzymes as targets for drug development.2 To date this has resulted in the marketing of two HDAC inhibitors for the treatment of cutaneous T cell lymphoma and more drug candidates are currently in clinical trials.2, 3, 5-7 The PRMTs are a class of writer enzymes that catalyze the transfer of up to two methyl-groups from the co-factor SAM to arginine residues on the substrate target (Fig. . To date nine PRMTs have been discovered. The PRMTs can be divided into three classes that produce mono-methylated arginines (MMA, PRMT7), asymmetrical dimethylated arginines (ADMA, PRMTs 1-4, 6 and 8) or symmetrical dimethylated arginines (SDMA, PRMTs 5 and 9) (Fig. 1A). A4   Figure 1. (A) Reactions catalyzed by the PRMT enzymes producing MMA, SDMA and ADMA. (B) Examples of known PRMT1 inhibitors based on a co-factor,8 A5  substrate,9 or bi-substrate approach.10 (C) Subtype-selective inhibitors of PRMT311 and PRMT4.12 To date, successful lead discovery of PRMT inhibitors has been based on developing ligands that mimic compounds that bind to the active site of the PRMTs (Fig. 1B). PRMT inhibitors are often mimics of either the substrate (1),9 the cofactor SAM (2),8 or bi-substrate inhibitors resembling both the substrate and the cofactor (3).10 Common for the majority of these inhibitors is that they inhibit the PRMTs in the low micromolar range and display little to no selectivity for the PRMT family of epigenetic enzymes. Poor selectivity is particularly problematic for mimics of the co-factor SAM and stems from the fact that it is the second most utilized endogenous co-factor in vivo, resulting in undesired off-target effects. A few potent subtype-selective PRMT inhibitors have been reported all of which are small molecules discovered using library or in silico screening (e.g. 4 and 5, Fig. 1C).11, 12 However, to date no pan-selective ligands for the PRMT family of epigenetic enzymes have been developed. To address the shortcoming of PRMT pan-selective ligands we decided to explore an approach based on phage display screening.13 Hits from a phage display screening would likely result in the identification of allosteric binders, thus increasing the likelihood for ligands that display selectivity for the PRMTs over the other epigenetic enzyme classes. Phage display screening of two peptide libraries resulted in the identification of several peptide hits that displayed good selectivity and binding to the PRMTs.14 Two peptides from that screening (P1 and P2, Table 1) were selected for further investigation and optimization. In order to increase the inhibitory potency of these hit compounds, several different modifications were undertaken: substitution with non-proteinogenic amino acids, N-alkylation of the peptide backbone as well as incorporation of constrained dipeptide mimics. Based on those investigations we now A6  wish to disclose the identification of the first peptide-based PRMT inhibitor P18 that is a promiscuous PRMT inhibitor in the micromolar range.  Results and Discussion Peptide P2 is a truncated version of P1 where the three N- and C-terminal residues have been deleted to provide a 7- vs. a 13-mer peptide, the latter being a significantly better PRMT inhibitor (Table 1). Due to high degree of similarity between P1 and P2 it is likely that both peptides bind to the same region and inhibit the PRMTs by a similar mechanism, and both peptides were explored as a source of a pan-selective PRMT inhibitor. To gain further insight into the importance of the three N- and C-terminal amino acid residues the 10-mers P3 and P4 were synthesized by deleting the three N- or the three C-terminal residues, respectively. For both P3 and P4 this resulted in a significant drop in the inhibitory activity compared to P1 suggesting that both the three N- and three C-terminal amino acid residues play an important role for binding/inhibition in the full length proteinogenic peptide P1. Due to the lack of information regarding the binding site and the bioactive conformation for P1 and P2 a conformational study was undertaken. The importance of the tryptophan residue in receptor recognition is well described.15-17 Thus, the two centrally placed tryptophan residues observed in peptide P1 were targeted for further ligand optimization. It was envisaged that systematic replacement of the tryptophan residues in P1 and P2 with β3homotryptophan (β3hW), D- and L-Aia (D- and L- 4-amino-1,4,5,10-tetrahydroazepino[3,4-b]indol-3(2H)-one), as well as N-methylation (NMe) of the succeeding amino acid would provide insights into the bioactive conformation of the peptides and guide the search for a potent PRMT inhibitor (Scheme 1). Extension of the backbone using β3-amino acids is known to favor the gauche conformation in amino acid side chains, which can induce turns in peptides.18 Insertion of the A7  constrained D- and L-Aia-scaffold forces the backbone into either the anti-conformation or gauche-(‒)/gauche-(+) conformation, respectively.19 Finally, N-methylation affects the hydrogen bonding capability and cis/trans ratio of the backbone amides.20 In addition to influencing the secondary structure of peptides, these modifications have the added benefit of improving the proteolytic stability and cell-penetration of peptides.21  Scheme 1. Modifications of the peptide backbone employed in this study.  Peptides incorporating proteinogenic and β3-amino acid residues were synthesized using standard Fmoc-based microwave solid phase peptide synthesis (SPPS) on Rink amide resin as described in the experimental section. N-methylation of peptides was performed on resin as follows (Scheme 2): After deprotection of the N-terminus the terminal amine was activated with o-nitrobenzenesulfonyl chloride (o-Ns-Cl), followed by methylation using DBU and dimethyl sulfate.22 Finally, the sulfonamide group was cleaved using 2-mercaptoethanol and standard SPPS was continued. A8   Scheme 2. N-methylation of peptides on solid support.22 For the introduction of L- and D-Aia both Fmoc-2’-formyl-tryptophan building blocks (R)-6 and (S)-6 were synthesized (Scheme 3).23 Aia was introduced on solid support in a two-step procedure by reductive amination of the N-terminally deprotected peptide with either (R)-6 for the introduction of D-Aia or (S)-6 in the case of L-Aia, followed by intramolecular lactam formation (Scheme 3).23  Scheme 3. Incorporation of the Aia moiety on solid support.  The reaction rate for the reductive amination step of Fmoc-2’-formyl-tryptophan (6) was highly dependent on the peptide sequence and in particular on the bulkiness of the preceding amino acid. When glycine was the preceding amino acid the reductive A9  amination went to completion within 15 minutes, whereas aspartic acid required reaction times of up to 24 hours. Subsequent intramolecular amide coupling of the secondary amine with the carboxylic acid of the Fmoc-2’-formyl-tryptophan building block (6) was carried out under microwave irradiation at 75 ºC using a 5-fold excess of HBTU and DIPEA and repeated twice for 5 minutes. The reaction progress was monitored by LC-MS and the coupling repeated when necessary. Finally, peptides were acetylated on the N-terminus, cleaved and deprotected (TFA cleavage cocktail) and purified by preparative RP-HPLC.  Inhibitor screening Peptides were tested in duplicate against PRMTs 1, 3-6 and 8 at a concentration of 20 μM (Table 1). The inhibitor assay utilized antibodies recognizing dimethylation events on histone 3 or 4 as specified in the experimental section. Inhibitory values below 20% were considered insignificant. PRMTs 2, 7 and 9 were not included. PRMT2 has very low enzyme activity and mainly hetero-dimerizes with PRMT1.24 PRMT7 only serves to mono-methylate its substrates. PRMT9 has only recently been broadly accepted as a member of the PRMT-family and was not available for screening.25  Table 1. Inhibitory activity of peptides against PRMTs 1, 3, 4, 5, 6 and 8 Peptidea PRMTb (%-inhibition) 1 3 4 5 6 8 P1 SHSEFWDWGPGGG 1 18 9 43 33 32 P2 EFWDWGP 4 6 7 10 13 5 P3 EFWDWGPGGG 0 4 11 0 7 9 P4 SHSEFWDWGP  1 9 13 0 7 14 P5 EFWD(β3hW)GP  19 20 5 30 27 46 P6 EF(β3hW)DWGP  13 22 7 15 31 16 P7 EF(β3hW)D(β3hW)GP  9 16 2 14 2 7 A10  P8 SHSEFWD(β3hW)GPGGG  2 2 15 7 2 30 P9 SHSEF(β3hW)DWGPGGG  0 0 13 0 0 10 P10 EFWDW(NMeG)P 10 5 3 1 0 0 P11 EFW(NMeD)WGP 0 11 5 7 0 0 P12 EFW(NMeD)W(NMeG)P 9 9 11 0 5 5 P13 EFWD(L-Aia)GP 33 28 13 34 33 27 P14 EF(L-Aia)DWGP 13 19 8 28 28 18 P15 EF(L-Aia)D(L-Aia)GP 46 39 9 29 21 25 P16 SHSEFWD(L-Aia)GPGGG 0 1 10 0 3 5 P17 SHSEF(L-Aia)DWGPGGG 0 0 10 4 1 18 P18 SHSEF(L-Aia)D(L-Aia)GPGGG 57 44 4 42 36 36 P19 EFWD(D-Aia)GP  1 1 5 0 1 5 P20 EF(D-Aia)DWGP 0 2 11 0 1 3 P21 EF(D-Aia)D(D-Aia)GP 4 5 9 0 1 4 P22 SHSEFWD(D-Aia)GPGGG 0 0 3 0 12 11 P23 SHSEF(D-Aia)DWGPGGG 0 0 5 1 20 10 P24 SHSEF(D-Aia)D(D-Aia)GPGGG 0 0 3 9 16 7 a Peptides were amidated at the C-terminus and acetylated at the N-terminus. b %-inhibition at a 20 µM concentration. See SI for standard deviations.  Incorporation of one β3-homotryptophan (β3hW) in the 7-mer sequence (P5 and P6) improved inhibitory activity on all tested PRMTs with the exception of PRMT4. However, incorporation of two β3hW residues in the 7-mer (P7) had no effect indicating that extension of the chain length by two methylene units was not well-tolerated in the binding sites of the enzymes. Introduction of a single β3hW in the 13-mer analogues to give peptides P8 and P9 abolished the activity for both peptides in contrast to 7-mers (P5 and P6), thus 13-mers incorporating two β3hW were not explored. Mono- and bis-N-methylation in the 7-mer peptide sequence (P10-12) had no effect on the activity and thus this modification was not investigated further. Next, introducing the constrained L- and D-Aia building blocks was explored. Substituting one tryptophan residue with L-Aia (P13 and P14) or both (P15) in the 7-mer peptide increased inhibition significantly (13-46% inhibition) on all tested A11  PRMTs, except PRMT4. It is noteworthy that the 13-mer peptide analogues containing a single L-Aia (P16 and P17) were almost entirely devoid of activity. In contrast, the 13-mer analogue containing two L-Aia residues (P18) displayed good properties on PRMT1, 3, 5, 6 and 8 (36-57% inhibition). Unlike the L-Aia containing peptides incorporation of a single or two D-Aia residues in the 7-mers (P19-21) and the 13-mers (P22-24) abolished all inhibitory activity. In summary, the inhibitory data shows that the introduction of a single or two β3hW residues can be beneficial in the case of the 7-mer peptides (P5 and P6) but not for the 13-mer analogues (P8 and P9). On the other hand, introduction two L-Aia residues in both the 7- and 13-mer peptides (P15 and P18, respectively) significantly improved the inhibitory activity, with the 13-mer peptide P18 displaying the best overall properties. Based on the screening results we decided to move forward with peptide P18 and determine the IC50 values at all 6 PRMTs (Table 2). Peptidomimetic P18 was able to inhibit PRMT1, 3, 5, 6 and 8 with comparable IC50 in the micromolar range (16-42 µM). As observed previously, PRMT4 was not inhibited by peptide P18 or any other peptides reported herein.  Tabel 2. IC50 values for peptide P18 and SAH against PRMTs 1, 3-6 and 8.a Cmpd. PRMT IC50±SD (µM)b 1 3 4 5 6 8 P18 23.1±1.2 42.0±1.1 No inhib. 22.4±1.2 17.8±1.1 16.1±1.5 SAH 2.49±1.1 6.67±1.1 0.12±1.1 0.84±1.1 0.56±1.1 1.52±1.2 a SAH was used as a positive control, because the by-product from the methyl-transfer reaction may inhibit further activity by blocking the co-factor binding site of the enzyme.8, 26  b Three independent experiments. See SI for full curves and experimental details.  A12  The Aia dipeptide-motif forces the peptide backbone into either the gauche-(+)/gauche-(‒) or anti-conformation,19 a principle that has been exploited successfully in the design of bioactive peptidomimetics.27, 28 However, the structural implications of having two L-Aia residues spaced by one residue as seen in peptidomimetics P15 and P18 are unknown. To gain further insight into the bioactive conformation of these peptidomimetics that could serve as a tool in the future design of bioactive peptides, including PRMT inhibitors, a series of CD and NMR experiments were carried out.  CD and NMR studies First, the structural preferences of the lead peptidomimetics P18 and truncated P15 containing two L-Aia residues were studied by CD and compared to their tryptophan-containing peptide counterparts P1 and P2 in methanol and phosphate buffer at pH 7.4 (Fig. 2, Fig. S3.a-b).   Figure 2. CD spectra of P1 (in blue) and P18 (in red) at 20 °C in methanol (dashed line) and in phosphate buffer, pH 7.4 (plain line). CD spectra of the truncated peptides P2 and P15 are reported in supplementary material (Fig. S3.a-b).  A13  P1 and P2 exhibited random coil signatures with a strong negative cotton effect around 200 nm in both solvent indicating that they are mostly disordered while P15 and P18 showed similar unusual spectra with a positive maximum at 217 nm and two negative maxima at 236 nm and a broad maximum at approx. 250 nm. However, structural details for P15 or P18 cannot be derived from the CD spectra. Subsequently, a NMR study to compare the structures of the longer sequences, P1 and P18 was performed to gain detailed structural insights. NMR experiments were carried out in methanol-d3 and phosphate buffer at 20 °C (H2O/D2O 9:1, pH 6.5). Well-resolved spectra were obtained for both compounds in methanol-d3 and all proton chemical shifts were assigned using COSY, TOCSY and ROESY experiments (Fig. S4.a). In contrast, P1 and P18 exhibited poor spectra in phosphate buffer with numerous signal overlaps and broad line widths preventing the assignment of most resonances (Fig. S4.d-e). However, considering the very similar CD profiles for P1 and P18 when recorded in methanol or phosphate buffer (Fig. 2) it is reasonable to assume that the secondary structure does not change markedly between the two solvent systems. Consequently, further detailed NMR studies for both compounds were carried out in methanol-d3. It was observed that the 3J(HN,Hα) coupling constant values for both P1 and P18 were mostly distributed between 6 and 8 Hz indicative of averaged ϕ backbone angle values (Table S4.g). Moreover, for P1 very few and only sequential NOEs were observed, underlining the high conformational flexibility of P1, in accordance with the observed random coil CD signature (Fig. 2). The lack of secondary structure for P1 is consistent with a short peptide sequence (13 amino acids), which includes four glycine residues and a proline residue. With the exception of a single glutamic acid residue, P1 contains no strong secondary structure inducers.29 A14   Figure 2. Inter-residue NOE correlations for P18. Sequential HN-H(i, i-1) correlations were omitted for clarity. Typical NOEs observed for the Aia-Xxx-Aia-Xxx sequence are shown in red. In contrast, the backbone of the central moiety of P18 was constrained by the L-Aia residues at positions 6 and 8, which imposed extended conformations. Interestingly, while only a few NOEs for the four N- and C-terminal residues were detected, five (i, i+2)-NOEs between the two Aia residues were observed, indicating more defined conformations in the central part of P18 (Fig. 2). This is consistent with the HN and H backbone proton chemical shifts variations seen between P1 and P18, which mostly involved the residues adjacent to the two L-Aia residues (Fig. S4.f). Thus, NOEs were used as constraints to elucidate the NMR solution structure of P18 using a simulated annealing procedure with AMBER 11.30 As expected, the two extremities of P18 did not converged due to the lack of distance restraints. In contrast, the central hydrophobic Phe5-Gly9 fragment forms a turn stabilized by the aromatic π-staking interaction between the two L-Aia residues. Moreover, a hydrogen bonding interaction between the intra-cyclic carbonyl group of L-Aia6 and the amide proton of L-Aia8 results in the formation of a C7 pseudo-cycle (Fig. 3). The root mean square deviation was reduced from 3.54 Å2 to 2.00 Å2 when the four N- and C-terminal residues were omitted. A15   Figure 3. A) Superimposition of the Phe5-Gly9 region of the 15 lowest energy NMR solution structures of P18. B) Enlargement of the Phe5-Gly9 moiety. C) C7-membered H-bond pseudo-cycle between the CO(L-Aia6) and NH(L-Aia8).  In agreement with its low occurrence in peptides and proteins, the seven-membered pseudo-ring was not observed for peptide P1. This structural feature observed in the bis-L-Aia-containing peptides P15 and P18 could explain the marked difference in inhibitory activity at PRMT 1 and 3 when compared to the disordered peptide P1. Moreover, no inhibition of the PRMTs were detected for the P19-P24 peptides incorporating D-Aia residues, which affect the overall structure and prevents the typical fold observed for P15 and P18. Notably, weaker or no inhibition of PRMT1 and 3 were measured for the truncated P13-14 and the full-length P16 and P17 peptides containing a single L-Aia residue, which indicates that the stability of the turn structure is increased by the presence of two L-Aia residues.  Conclusion In the present study, a peptide hit emerging from a phage display screening was optimized to produce the first peptide based pan-selective PRMT inhibitor (P18) in the 16-42 µM range. The effects of different modifications to the original hit were investigated, including substitution with non-proteinogenic amino acids (β3hW), N-alkylation of the peptide backbone and incorporation of constrained dipeptide mimics A16  (D- and L-Aia). It was found that incorporation of two L-Aia residues spaced by a single amino acid residue in peptides P15 and P18 lead to a significant increase in the inhibitory activity. Using CD and NMR the conformational preferences of the peptidomimetics were investigated. A unique seven-membered pseudo-ring turn structure was observed for P15 and P18, which is not present in the native peptide counterparts (P1 and P2). The distinct conformation observed for P15 and P18 provides an explanation for the increased potency observed for this class of peptidomimetic ligands and serves as a template for further development of PRMT inhibitors. We are in the process of investigating the use of this new turn inducing motif in other bioactive peptides as well as exploring the structural implications of introducing multiple repeats of L-Aia-Xxx-L-Aia units in peptides, which will be reported in due course. Experimental section. General information All natural amino acids, resin and coupling reagents were purchased from Iris Biotech GmbH, Marktredwitz, Germany or Chem-Impex International Inc., Wood Dale, Illinois, USA. β3-amino acids were purchased from Chem-Impex International Inc. Chemicals for preparation of (R)- and (S)-6 and for N-methylation were purchased from Sigma-Aldrich, Schnelldorf, Germany. Microwave peptide synthesis was performed on a Biotage Initiator+ SP Wave. High-resolution Mass Spectrometry was recorded on a Micromass Q-TOF 1.5, UB137 or on a time-of-flight (TOF) MS system, coupled to an analytical HPLC and ESI detector. HRMS HPLC was performed on a C18 column (25 cm × 4.6 mm, 5 μm) with a linear gradient (10% to 100% MeOH in H2O containing 0.1% TFA (v/v) in 20 min) at a flow rate of 1 ml/min and UV detection at 215 nm. LC-LRMS was performed on an Agilent 1200 series LC equipped with a Zorbax Eclipse XBD-C18 column (50 mm × 4.6 mm), A17  eluent A (95% H2O, 5% MeCN + 0.1% HCO2H) and eluent B (95% MeCN, 5% H2O + 0.043% HCO2H) at a flow rate of 0.75 ml/min, coupled to an Agilent 6410 Triple Quad mass spectrometer with an ESI and a diode-array detector. Method: 0-5 min 0-50% B, 5-5.1 min 50-90% B, 5.1-5.5 min 90% B. Preparative RP-HPLC of peptides was carried out on a Dionex UltiMate 3000 HPLC with a diode array detector using a preparative RP Phenomenex Gemini NX-C18 column (250 × 21.20 mm, 5 μM). Peptides were eluted using a step gradient from 0-100% B, with eluent A (0.1% TFA in H2O) and eluent B (90:10 MeCN:H2O + 0.1% TFA) at a flow rate of 20 ml/min. Analytical RP-HPLC was performed on a Dionex UltiMate 3000 system equipped with a photodiode array detector and a RP Phenomenex Gemini NX-C18 column (250 × 4.6 mm, 3 μM) with the same eluents A and B used for preparative HPLC at a flow rate of 1 ml/min. Reported retention times (tR) were obtained using the following gradient: 0% B 0-5 min, 0-100% B 5-35 min, 100% B 35-40 min.  Peptide synthesis Standard SPPS: Peptides were synthesized by Nα-Fmoc solid phase methodology on Rink amide MBHA resin (0.7 mmol/g) using HBTU as coupling reagent with DIPEA as base. 3- or 5-fold excess of Fmoc-protected amino acids were used. DIPEA and coupling reagents were used for coupling of natural amino acids using DMF as solvent. β3-amino acids were coupled using a 2-fold excess of reagents. For P2 and P5-7 all syntheses were performed manually with one coupling step per residue for 60 minutes. For peptides P1, P3-4 and P8-24 manual synthesis was employed in general (as above) and microwave synthesis was used for couplings involving the modifications. Microwave synthesis was generally performed twice at 75 oC for 2 × 5 min, with a fresh amino acid/HBTU/DIPEA coupling mixture. A18  Fmoc deprotection was carried out by treating the resin twice (5 min + 20 min) with 20% piperidine in DMF + 0.1M HOBt. The synthesis progress was checked at regular intervals using LC-MS after a test cleavage on a few resin beads using a cleavage cocktail consisting of 95:2.5:2.5 TFA:H2O:TIS (v/v/v). After coupling of the final amino acid, the N-terminus of the peptide was deprotected and acetylated using a 1:1:8 acetic anhydride:DIPEA:DMF mixture (v/v/v). Final cleavage and deprotection from the resin was carried out for 3 hours using the cleavage cocktail described above. The crude product was precipitated, washed in cold ether and filtered to yield the crude peptide, which was purified using preparative RP-HPLC. Yields of peptides containing natural amino acids and β3-amino acids were in the 25-50% range after preparative RP-HPLC. N-methylation of peptides was carried out on the deprotected N-terminus of the specified amino acid using the methodology of Kessler et al.22 Standard SPPS was continued after N-methylation. Yields of peptides containing one or two N-methylations were in the 12-45% range after preparative RP-HPLC. Aia-containing peptides: Fmoc-L-2’-formyl-Trp-OH ((S)-6) and Fmoc-D-2’-formyl-Trp-OH ((R)-6) building blocks for insertion of L- and D-Aia, respectively, were synthesized according to the published procedure.23 Resin containing the N-terminally deprotected peptide on resin was shaken for 30 minutes in 1:1 TMOF/DCM + 0.5% AcOH mixture. Reductive amination was carried out using a 2-fold excess of the building block and a 4-fold excess of NaBH3CN in a 1:1 TMOF/DCM + 0.5% AcOH mixture at room temperature and monitored using LC-MS after performing a test cleavage as described above. Coupling of the building block with the secondary amine in the peptide backbone was carried out under microwave irradiation at 75 oC for 2-3 × 5 min and repeated when necessary (monitored by LC-MS after test cleavage). Standard SPPS was continued after introduction of Aia. Yields of peptides containing A19  a single Aia were in the 10-44% range. Peptides containing two Aia’s were isolated in 2-23% yields.  Histone Methyltransferase Inhibitor Assay The inhibitor screening of peptides was performed in duplicate at room temperature for 60 minutes. Stock solutions of 10 mM peptide in DMSO were prepared and diluted to a final concentration of 20 μM peptide (1% DMSO final) in the 50 μl reaction mixture containing methyltransferase assay buffer, S-adenosyl methionine (0.2 μM for PRMT1 and 1 μM for the remaining PRMTs) and enzyme (either PRMT1 (5 ng), PRMT3 (30 ng), PRMT4 (200 ng), PRMT5 (100 ng) , PRMT6 (100 ng) or PRMT8 (50 ng)) in histone substrate-precoated wells (H4 1-21 for PRMT1, 3, 5, 6, 8 and H3 1-27 (acK18) for PRMT4). The reaction mixtures were discarded and the wells were washed 3× with TBS-T buffer and then shaken with Blocking Buffer for 10 minutes. The wells were emptied and incubated with a primary antibody for 60 minutes. Wells were emptied, washed 3× with TBS-T, shaken 10 minutes with Blocking Buffer and incubated with an HRP-secondary antibody for 30 minutes at room temperature. Wells were again emptied, washed 3× with TBS-T and shaken 10 minutes with Blocking Buffer before a freshly prepared HRP chemiluminescent mixture was added. The plates were read immediately on the BioTek SynergyTM 2 plate reader.  The %-activity was calculated using %-activity = (C-C0)/(Ce-C0), where C is the luminescence in the presence of compound, C0 is the luminescence in the absence of enzyme and Ce is the luminescence in the absence of peptide. SAH was used as a positive control. IC50 curves of P18 were produced using the same assay in triplicates.  Associated content. A20  Supporting information. Equipment and general experimental information, for all synthesized peptides, including MS data and HPLC chromatograms. This material is available free of charge via the internet at http://pubs.acs.org.  Author information. Corresponding author.  *(D.S.P.) E-mail: daniel.pedersen@sund.ku.dk. The authors declare no competing financial interest.  Acknowledgements. The Danish Chemical Society, The AP Møller Foundation and the Lundbeck Foundation are gratefully acknowledged for financial support. The work of O.V.D.P. was supported by the Agency for Innovation by Science and Technology (IWT Vlaanderen).  Abbreviations used. 2-NBS-Cl = 2-Nitrobenzenesulfonyl chloride; ADMA = Asymmetrical dimethylated arginine; Aia = 4-Amino-1,4,5,10-tetrahydroazepino[3,4-b]indol-3(2H)-one; DBU = 1,8-Diazabicycloundec-7-ene; HBTU = 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate; HDAC = Histone deacetylases; HOBt = Hydroxybenzotriazole; MMA = Mono-methylated arginine; NOE = Nuclear Overhauser effect; o-Ns-Cl = o-Nitrobenzenesulfonyl chloride; PRMT = Protein arginine N-methyl transferases; SAH = S-adenosyl-homocysteine; SAM = S-Adenosyl methionine; SDMA = Symmetrical dimethylated arginine; SPPS = Solid phase peptide synthesis; TMOF = Trimethyl orthoformate; Xxx = Unspecified amino acid residue. A21   References  1.  Berger, S. L.; Kouzarides, T.; Shiekhattar, R.; Shilatifard, A. An operational definition of epigenetics. Genes & Development 2009, 23, 781-783.  2.  Arrowsmith, C. H.; Bountra, C.; Fish, P. V.; Lee, K.; Schapira, M. Epigenetic protein families: a new frontier for drug discovery. Nat. Rev. Drug Discov. 2012, 11, 384-400.  3.  Bojang, J.; Ramos, K. S. The promise and failures of epigenetic therapies for cancer treatment. Cancer Treatment Rev. 2014, 40, 153-169.  4.  Yang, Y.; Bedford, M. T. Protein arginine methyltransferases and cancer. Nat. Rev. Cancer 2013, 13, 37-50.  5.  Dhanak, D.; Jackson, P. Development and classes of epigenetic drugs for cancer. Biochem. Biophys. Res. Commun. 2014, 455, 58-69.  6.  Falkenberg, K. J.; Johnstone, R. W. Histone deacetylases and their inhibitors in cancer, neurological diseases and immune disorders. Nat. Rev. Drug Discov. 2014, 13, 673-691.  7.  Tough, D. F.; Lewis, H. D.; Rioja, I.; Lindon, M. J.; Prinjha, R. K. Epigenetic pathway targets for the treatment of disease: accelerating progress in the development of pharmacological tools: IUPHAR review 11. Br. J. Pharmacol. 2014, 171, 4981-5010.  8.  Cheng, D.; Yadav, N.; King, R. W.; Swanson, M. S.; Weinstein, E. J.; Bedford, M. T. Small Molecule Regulators of Protein Arginine Methyltransferases. J. Biol. Chem. 2004, 279, 23892-23899.  9.  Obianyo, O.; Causey, C. P.; Osborne, T. C.; Jones, J. E.; Lee, Y. H.; Stallcup, M. R.; Thompson, P. R. A Chloroacetamidine-Based Inactivator of Protein Arginine Methyltransferase 1: Design, Synthesis, and In Vitro and In Vivo Evaluation. ChemBioChem 2010, 11, 1219-1223.  10.  van Haren, M.; van Ufford, L. Q.; Moret, E. E.; Martin, N. I. Synthesis and evaluation of protein arginine N-methyltransferase inhibitors designed to simultaneously occupy both substrate binding sites. Org. Biomol. Chem. 2015, 13, 549-560.  11.  Kaniskan, H. Ü.; Szewczyk, M. M.; Yu, Z.; Eram, M. S.; Yang, X.; Schmidt, K.; Luo, X.; Dai, M.; He, F.; Zang, I.; Lin, Y.; Kennedy, S.; Li, F.; Dobrovetsky, E.; Dong, A.; Smil, D.; Min, S. J.; Landon, M.; Lin-Jones, J.; Huang, X. P.; Roth, B. L.; Schapira, M.; Atadja, P.; Barsyte-Lovejoy, D.; Arrowsmith, C. H.; Brown, P. J.; Zhao, K.; Jin, J.; Vedadi, M. A Potent, Selective and Cell-Active Allosteric Inhibitor of Protein Arginine Methyltransferase 3 (PRMT3). Angew. Chem. Int. Ed. 2015, 54, 5166-5170.  12.  Sack, J. S.; Thieffine, S.; Bandiera, T.; Fasolini, M.; Duke, G. J.; Jayaraman, L.; Kish, K. F.; Klei, H. E.; Purandare, A. V.; Rosettani, P.; Troiani, S.; Xie, D.; Bertrand, J. A. Structural basis for CARM1 inhibition by indole and pyrazole inhibitors. Biochem. J. 2011, 436, 331-339.  13.  New England Biolabs. Ph.D. Phage Display Libraries Instruction Manual.  2011.   14.  Unpublished results. Manuscript in preparation.  15.  Gallivan, J. P.; Dougherty, D. A. Cation-π interactions in structural biology. Proc. Nat. Acad. Sci. 1999, 96, 9459-9464.  16.  Burley, S. K.; Petsko, G. A. Aromatic-aromatic interaction: a mechanism of protein structure stabilization. Science 1985, 229, 23-28. A22   17.  Burley, S. K.; Petsko, G. A. Amino-aromatic interactions in proteins. FEBS letters 1986, 203, 139-143.  18.  Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. Peptides: From Structure to Function. Chem. Rev. 2001, 101, 3219-3232.  19.  Maes, V.; Tourwe, D. Aspects of Peptidomimetics. In Peptide and protein design for biopharmaceutical applications; Jensen; K Eds.; Wiley: 2009; pp 49-131.  20.  Sikorska, E.; Slusarz, M. J.; Slusarz, R.; Kowalczyk, W.; Lammek, B. Investigation of cis/trans ratios of peptide bonds in AVP analogues containing N-methylphenylalanine enantiomers. J. Pep. Sci. 2006, 12, 13-24.  21.  Tömböly, C.; Ballet, S.; Feytens, D.; Kövér, K. E.; Borics, A.; Lovas, S.; Al-Khrasani, M.; Fürst, Z.; Tóth, G.; Benyhe, S.; Tourwé, D. Endomorphin-2 with a β-Turn Backbone Constraint Retains the Potent µ-Opioid Receptor Agonist Properties. J. Med. Chem. 2008, 51, 173-177.  22.  Biron, E.; Chatterjee, J.; Kessler, H. Optimized selective N-methylation of peptides on solid support. J. Pep. Sci. 2006, 12, 213-219.  23.  Feytens, D.; De Vlaeminck, M.; Tourwé, D. A novel solid phase approach to Aia-containing peptides. J. Pep. Sci. 2009, 15, 16-22.  24.  Pak, M. L.; Lakowski, T. M.; Thomas, D.; Vhuiyan, M. I.; Hüsecken, K.; Frankel, A. A Protein Arginine N-Methyltransferase 1 (PRMT1) and 2 Heteromeric Interaction Increases PRMT1 Enzymatic Activity. Biochem. 2011, 50, 8226-8240.  25.  Yang, Y.; Hadjikyriacou, A.; Xia, Z.; Gayatri, S.; Kim, D.; Zurita-Lopez, C.; Kelly, R.; Guo, A.; Li, W.; Clarke, S. G.; Bedford, M. T. PRMT9 is a Type II methyltransferase that methylates the splicing factor SAP145. Nat. Commun. 2015, 6.  26.  Bedford, M. T.; Richard, S. p. Arginine Methylation: An Emerging Regulator of Protein Function. Mol. Cell 2005, 18, 263-272.  27.  Feytens, D.; Cescato, R.; Reubi, J. C.; Tourwé, D. New sst4/5-Selective Somatostatin Peptidomimetics Based on a Constrained Tryptophan Scaffold. J. Med. Chem. 2007, 50, 3397-3401.  28.  Ballet, S.; Feytens, D.; Buysse, K.; Chung, N. N.; Lemieux, C.; Tumati, S.; Keresztes, A.; Van Duppen, J.; Lai, J.; Varga, E.; Porreca, F.; Schiller, P. W.; Vanden Broeck, J.; Tourwé, D. Design of Novel Neurokinin 1 Receptor Antagonists Based on Conformationally Constrained Aromatic Amino Acids and Discovery of a Potent Chimeric Opioid Agonist-Neurokinin 1 Receptor Antagonist. J. Med. Chem. 2011, 54, 2467-2476.  29.  Chou, P. Y.; Fasman, G. D. Prediction of protein conformation. Biochem. 1974, 13, 222-245.  30.  Case, D. A.; Darden, T. A.; Cheatham, T. E.; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M. Amber 11. 2010. University of California.  Table of Contents Graphic. A23   B1  Appendix 2:  Use of Constrained Amino Acids and Peptidomimetics in Biological Applications: Angiotensin IV, Bradykinin, Farnesyl Transferase, Neurokinin-1  The following four cases have been written as part of a review on chi-space constrained peptidomimetics presented in section 2.6, to which Aia used in this thesis belongs.   The review is currently being written in collaboration with researchers at Vrije Universiteit Brussel and will contain 8 cases on the use of constrained amino acids and peptidomimetics in biologically relevant peptide applications. The following four cases have been written by the author of this thesis. The review will contain an introduction to using constrained amino acids and peptidomimetics such as the one in section 2.6. On the following page amino acids used in these cases which have not been introduced in section 2.6 are shown. We expect to publish the review in a relevant journal in due course.    Angiotensin IV ............................................................................................................................................ C1-C7 Bradykinin ................................................................................................................................................. D1-D9 Farnesyl Transferase ...................................................................................................................................E1-E8 Neurokinin-1 ............................................................................................................................................... F1-F9     B2  List of amino acids used in the four review cases, which have not been introduced in the PhD dissertation   C1  ANGIOTENSIN IV The hexapeptide hormone angiotensin IV (H-Val-Tyr-Ile-His-Pro-Phe-OH, AngIV, 1) is derived from the bioactive precursor angiotensin II by cleavage of the two N-terminal amino acids (figure 1). AngIV has been shown to improve memory acquisition, protect against ischemic stroke as well as hyperglycemia but also have functions in the vascular and renal systems, just like its precursors.1-5 It binds to the putative AT4-receptor, also known as insulin regulated aminopeptidase (IRAP), but the exact mechanism of how the biological functions are exerted is still not fully understood.6, 7 One hypothesis is that it inhibits IRAP and thereby also the catalytic breakdown of other endogenous peptides such as vasopressin, oxytocin and bradykinin.1, 8 As native AngIV also binds to and is degraded by aminopeptidase N (AP-N), is a weak but full agonist at the AT1-receptor, which has angiotensin II as its primary ligand, and has a very short half-life in vivo (t½ < 1min), more stable and selective analogues are required in order to fully understand the mechanisms of action of the hormone.9, 10   Figure 1: Sequences of the angiotensins and angiotensinogen   In order to improve stability of AngIV Lukaszuk et al. performed a β-homo amino acid scan which identified H-β2hVal-Tyr-Ile-His-Pro-β3hPhe-OH (AL-11, pKi(HEK293+IRAP) = 7.56, 1a) as a potent, selective  and stable analogue.11  Due to the high stability and selectivity of 1a in vitro studies of the IRAP receptor were undertaken several years ago.12, 13 Using a [3H]1a it was shown that IRAP undergoes semi-continuous cycling between the surface of the cell and endosomal compartments.13      The β-homo amino acid scan inspired the authors to investigate the effects of a conformational constraining strategy for the aromatic amino acids Tyr2 and Phe6, as well as Pro5 on IRAP inhibition (figure 2).14  Tyr2 was exchanged for both stereoisomers of Hat (2a), 6-Htc (2b) and 7-Htc (2c) (figure 2). All analogues showed a large drop in IRAP inhibition indicating that none of the constrained amino acids used were able to fix the required χ1 and χ2 angles in a favourable orientation. Thus, structure and orientation of Tyr2 seem C2  to be essential for the biological function of AngIV, further demonstrated by the detrimental effect of even simple methyl substitutions at the β-carbon or the aromatic ring (βMeTyr (2d) and Dmt (2e), respectively).  Exchanging Pro5 for its “β-amino acid analogues”, the trans-2-aminocyclohexane-1-carboxylic acids (Acpcs) and cis-Acpcs, resulted in analogues with similar potency to AngIV (2f and 2g, respectively). Comparing this with the results of the replacement by 2- and 3-homoPro (1b and 1c, respectively)11 indicates that neither the amino acid length nor the conformational constraints induced by the pyrrolidine ring are essential for binding to IRAP.  The previous -homo amino acid scan had shown that modification of Phe6 had a positive influence on IRAP versus AP-N activity and abolished AT1 affinity (1a, 1d-f).11 Although IRAP potency was reduced in the Phe-modified analogues 2h-2i, these compounds showed good selectivity.  As the Atc- (2h) and Aic-analogues (2i), which prefer the g+/g- conformations, and the Tic-analogue (2j), which adopts the g+ conformation,15 showed a similar drop in IRAP inhibition compared to AngIV it was speculated that the bioactive conformation at 1 of this residue is g-, even though the constraint at 2 is probably not optimal. This is further supported by the fact that the best Phe6 analogue is the e-βMePhe (2k), which is known to disfavour the trans rotamer and prefer the g+/g- conformations. However, compared to AngIV all 2a-2i compounds had decreased or similar IRAP inhibition.    Figure 2: Conformational constraining strategy for Tyr2, Pro5 and Phe6 and the pKi (IRAP) values of the most potent stereoisomers. C3  The most interesting results were obtained by constraining the His4 or His4-Pro5 fragment (Figure 3). His4 was exchanged by the constrained amino acids Spi (3a) and Tic (3b) or the His4-Pro5 fragment by Ata-Gly (3c),16 Aba-Gly (3d) and Aia-Gly (3e).17 Compounds 3a and 3b, which adopt either a g+ or a g- conformation at His4, showed a significant decrease in inhibitory potency for both IRAP and AP-N compared to native AngIV. Compounds 3c-e16, 17 are either in the g+ or anti orientations. 3c and 3d showed inhibitory activities similar to AngIV, indicating that the imidazole moiety could be exchanged for the triazole (3c) or the phenyl (3d). Incorporation of the indole moiety, however, led to a lowering of inhibition (3e). Taken together, this suggests that the anti conformation at position 4 is preferred, whereas the g+ conformation is not. All analogues 3c-e showed good IRAP selectivity.  As a β2h-/ β3h-amino acid scan of AngIV had previously shown that exchanging Val1 by β2hVal1 resulted in selectivity for IRAP versus AP-N and stability against the enzymatic degradation of the peptide,11 this modification was applied to the phenyl and indole analogues, 3d and 3e, resulting in 3f and 3g, respectively. Both compounds showed good binding and stability in cell preparations as well as high IRAP versus AP-N selectivity. 3f was, however, shown to also bind to the AT1-receptor, and thus with an IRAP pKi of 8.07 3g seemed to be the superior analogue yet.  As a previous study had shown that exchanging Pro5 for Gly5 in AngIV resulted in a 3-fold loss in affinity18 Nikolaou et al. replaced Gly5 in 3g for residues with sidechains of various lengths19. Neither Ala (4a), D-Ala (4b), propargylglycine (Pra, 4c) or deletion of the residue altogether (4d) improved affinity for IRAP. Adding a saturated alkyl chain in the form of L-norvaline (Nva, 4e), however, improved both specifity (no AT1 activation) and affinity compared to previous analogues and had a 40-fold higher affinity for IRAP compared to AngIV (figure 3).  The analogue 4e was shown to be stable in human plasma contrary to native AngIV and did not bind to the AT1-receptor. Moreover, the [3H]-analogue of 4e could easily be obtained by catalytic hydrogenation of 4d using tritium gas, providing an excellent compound for in vitro and in vivo mechanistic studies. The authors demonstrated this by showing that pre-treatment of intact CHO-K1 cells with 4e could completely abolish IRAP-availability at the cell surface in vitro as no [3H]4e-IRAP complexes were formed when the radioactive analogue was added to the cells.19  A SAR on cyclic tripeptides found 5 to be a very potent and selective IRAP-inhibitor, more selective but slightly less potent than 4e.20  Accordingly, modifying Val1, His4 and Pro5 proved beneficial in discovering good IRAP-selective, stable compounds, and produced better IRAP-inhibitors than exchanging Tyr2, Pro5 and Phe6 individually.  The analogues 4e and 5 are therefore still the best constrained peptidomimetics IRAP inhibitors to date.         C4      Figure 3: His4 constraints: a key to the most potent, selective and stable AngIV analoguesC5  Angiotensin analogues: Studies were performed in HEK293 cell membrane homogenates transiently expressing either IRAP or AP-N. a) Ref 14+17, b) Ref 19, c) Different stereoisomers eluted first and last, respectively. These compounds were not assigned an absolute configuration as none of these analogues were better than previous ones. d) racemic mixture  Dmt = 2’,6’-dimethyltyrosine; Pra = propargylglycine; Nva = norvaline; Hcy = homocysteine; 2-AMP = 2-(aminomethyl)phenylacetic acid; * = side chain cyclisation  1 2 3 4 5 6 HEK293 + IRAP pKi HEK293 + AP-N pKi Selectivity Ki (AP-N)/Ki (IRAP) Ref Angiotensin IV V Y I H P F 7.25a/7.20b 6.08a/5.76b 14.8a/27b 14,17,19 1a (AL-11) (R)-β2-hV Y I H P β3-hF 7.56 5.23 214 11 1b V Y I H (R)-β2-hP F 7.09 5.53 36 11 1c V Y I H β3-hP F 6.84 5.27 37 11 1d V Y I H P β2-hF 6.96 d 5.61 d 22 11 1e V Y I H P β3-hF 7.69 5.61 120 11 1f β2-hL Y I H P β2-hF 6.78 d 4.16 d 417 11 2a V Hat I H P F 4.73/5.24 c 3.70/4.35  c 10.7/7.8  c 12 2b V 6-Htc I H P F 4.90/4.40  c 3.48/3.42  c 26.3/9.5  c 12 2c V 7-Htc I H P F 5.06/4.81  c 4.21/3.60  c 7.1/16.2  c 12 2d V βMeTyr I H P F 5.69/5.20  c 3.84/3.66 c 70.8/34.7 c 12 2e V Dmt I H P F 5.51 c 4.34 c 14.8 c 12 2f V Y I H Trans-Acpc F 7.08/6.78  c 5.14/5.12  c 87.1/45.7  c 12 2g V Y I H Cis-Acpc F 7.30/6.89  c 5.76/5.43  c 34.7/28.8  c 12 2h V Y I H P Atc 6.73 5.41 20.9 12 2i V Y I H P Aic 6.67 5.55 13.2 12 2j V Y I H P Tic 6.76 6.20 3.6 12 2k V Y I H P e-βMePhe 7.12 5.18 87.1 12 3a V Y I Spi P F 6.29 5.52 5 14 3b V Y I Tic P F 6.66 5.61 11 14 3c V Y I Ata G F 7.09 5.08 102 15 3d V Y I Aba G F 7.30 5.69 50 14 3e V Y I Aia G F 6.74 5.07 50 14 3f D-hβ2-V Y I Aba G F 7.90 5.53 250 14 3g (AL-40) D-hβ2-V Y I Aia G F 8.07 6.10 100 14 4a D-hβ2-V Y I Aia A F 7.56 5.66 79 18 4b D-hβ2-V Y I Aia D-A F 7.81 5.39 265 18 4c D-hβ2-V Y I Aia Pra F 7.59 5.29 200 18 4d D-hβ2-V Y I Aia - F 7.37 5.20 148 18 4e (IVDE-77) D-hβ2-V Y I Aia Nva F 8.77 5.93 687 18 4f D-hβ2-V Y I Aia D-Nva F 7.92 5.12 622 18 5 Hcy* β3-hY C* 2-AMP   8.48 5.14 2195 19 C6  Reference List   1.  Chai, S. Y.; Fernando, R.; Peck, G.; Ye, S. Y.; Mendelsohn, F. A. O.; Jenkins, T. A.; Albiston, A. L. What's new in the renin-angiotensin system? CMLS, Cell. Mol. Life Sci. 2004, 61, 2728-2737.  2.  Faure, S.; Chapot, R.; Tallet, D.; Javellaud, J.; Achard, J. M.; Oudart, N. CEREBROPROTECTIVE EFFECT OF ANGIOTENSIN IV IN EXPERIMENTAL ISCHEMIC STROKE IN THE RAT MEDIATED BY AT 4 RECEPTORS. J. Physiol. Pharmacol. 2006.  3.  Pham, V.; Albiston, A. L.; Downes, C. E.; Wong, C. H.; Diwakarla, S.; Ng, L.; Lee, S.; Crack, P. J.; Chai, S. Y. Insulin-regulated aminopeptidase deficiency provides protection against ischemic stroke in mice. Journal of neurotrauma 2012, 29, 1243-1248.  4.  Wong, Y. C.; Sim, M. K.; Lee, K. O. Des-aspartate-angiotensin-I and angiotensin IV improve glucose tolerance and insulin signalling in diet-induced hyperglycaemic mice. Biochemical pharmacology 2011, 82, 1198-1208.  5.  Wright, J. W.; Krebs, L. T.; Stobb, J. W.; Harding, J. W. The angiotensin IV system: functional implications. Frontiers in neuroendocrinology 1995, 16, 23-52.  6.  Vanderheyden, P. M. From angiotensin IV binding site to AT 4 receptor. Molecular and cellular endocrinology 2009, 302, 159-166.  7.  Albiston, A. L.; McDowall, S. G.; Matsacos, D.; Sim, P.; Clune, E.; Mustafa, T.; Lee, J.; Mendelsohn, F. A. O.; Simpson, R. J.; Connolly, L. M.; Chai, S. Y. Evidence That the Angiotensin IV (AT4) Receptor Is the Enzyme Insulin-regulated Aminopeptidase. Journal of Biological Chemistry 2001, 276, 48623-48626.  8.  Stragier, B.; De Bundel, D.; Sarre, S.; Smolders, I.; Vauquelin, G.; Dupont, A.; Michotte, Y.; Vanderheyden, P. Involvement of insulin-regulated aminopeptidase in the effects of the reninangiotensin fragment angiotensin IV: a review. Heart failure reviews 2008, 13, 321-337.  9.  Chansel, D.; Czekalski, S.; Vandermeersch, S.; Ruffet, E.; Fournié-Zaluski, M. C.; Ardaillou, R. Characterization of angiotensin IV-degrading enzymes and receptors on rat mesangial cells. American Journal of Physiology-Renal Physiology 1998, 275, F535-F542.  10.  Handa, R. K. Metabolism alters the selectivity of angiotensin-(1-7) receptor ligands for angiotensin receptors. Journal of the American Society of Nephrology 2000, 11, 1377-1386.  11.  Lukaszuk, A.; Demaegdt, H.; Szemenyei, E.; Töth, G.; Tymecka, D.; Misicka, A.; Karoyan, P.; Vanderheyden, P.; Vauquelin, G.; Tourwé, D. -Homo-amino Acid Scan of Angiotensin IV. J. Med. Chem. 2008, 51, 2291-2296.  12.  Demaegdt, H.; Lukaszuk, A.; De Buyser, E.; De Backer, J. P.; Szemenyei, E.; Töth, G.; Chakravarthy, S.; Panicker, M.; Michotte, Y.; Tourwé, D. Selective labeling of IRAP by the tritiated AT 4 receptor ligand [3 H] Angiotensin IV and its stable analog [3 H] AL-11. Molecular and cellular endocrinology 2009, 311, 77-86. C7   13.  Demaegdt, H.; Gard, P.; De Backer, J. P.; Lukaszuk, A.; Szemenyei, E.; Tóth, G. +.; Tourwé, D.; Vauquelin, G. Binding of "AT 4 receptor" ligands to insulin regulated aminopeptidase (IRAP) in intact Chinese hamster ovary cells. Molecular and cellular endocrinology 2011, 339, 34-44.  14.  Lukaszuk, A.; Demaegdt, H.; Van den Eynde, I.; Vanderheyden, P.; Vauquelin, G.; Tourwé, D. Conformational constraints in angiotensin IV to probe the role of Tyr2, Pro5 and Phe6. Journal of Peptide Science 2011, 17, 545-553.  15.  Ruzza, P.; Cesaro, L.; Tourwé, D.; Calderan, A.; Biondi, B.; Maes, V.; Menegazzo, I.; Osler, A.; Rubini, C.; Guiotto, A. Spatial conformation and topography of the tyrosine aromatic ring in substrate recognition by protein tyrosine kinases. J. Med. Chem. 2006, 49, 1916-1924.  16.  Buysse, K.; Farard, J.; Nikolaou, A.; Vanderheyden, P.; Vauquelin, G.; Sejer Pedersen, D.; Tourwé, D.; Ballet, S. Amino triazolo diazepines (Ata) as constrained histidine mimics. Organic letters 2011, 13, 6468-6471.  17.  Lukaszuk, A.; Demaegdt, H.; Feytens, D.; Vanderheyden, P.; Vauquelin, G.; Tourwé, D. The replacement of His (4) in angiotensin IV by conformationally constrained residues provides highly potent and selective analogues. J. Med. Chem. 2009, 52, 5612-5618.  18.  Sardinia, M. F.; Hanesworth, J. M.; Krebs, L. T.; Harding, J. W. AT 4 receptor binding characteristics: D-amino acid-and glycine-substituted peptides. Peptides 1993, 14, 949-954.  19.  Nikolaou, A.; Van den Eynde, I.; Tourwé, D.; Vauquelin, G.; Töth, G.; Mallareddy, J. R.; Poglitsch, M.; Van Ginderachter, J. A.; Vanderheyden, P. M. [3 H] IVDE77, a novel radioligand with high affinity and selectivity for the insulin-regulated aminopeptidase. European journal of pharmacology 2013, 702, 93-102.  20.  Andersson, H.; Demaegdt, H.; Vauquelin, G.; Lindeberg, G.; Karlén, A.; Hallberg, M.; Erdélyi, M.; Hallberg, A. Disulfide cyclized tripeptide analogues of angiotensin IV as potent and selective inhibitors of insulin-regulated aminopeptidase (IRAP). J. Med. Chem. 2010, 53, 8059-8071.   D1  BRADYKININ The nonapeptide bradykinin (H-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg-OH, BK) is involved in various physiological processes including vasodilatation and vascular permeability as well as pathophysiological pathways causing inflammation and pain.1, 2 BK exerts its biological effects by activating the constitutively expressed B2 receptors2 as well as the B1 receptors produced mainly during tissue injury,3 and therefore blocking these receptors would allow modulation of these effects.     Figure 1: Comparison of the amino acid sequences of Bradykinin and the B2 antagonist Hoe140  The search for stable, selective peptidomimetic antagonists has culminated in the discovery of Hoe140, the thus far most potent B2 antagonist. Marketed as Icatiban for the treatment of angioedema, Hoe 140 has been thoroughly characterised both in vitro and in vivo, with IC50 values in the low nanomolar range in several preparations as well as high potency and stability in vivo in various animal models.4, 5    D2  Due to the superiority of Hoe140, almost all efforts in the discovery of improved antagonists have been based on this compound.  Lembeck et al.6 exchanged Thi5 for Phe5, the fifth residue in native BK, which, however, produced a weaker antagonist (2a) than Hoe140 highlighting the importance of each substitution in the short sequence. This was further evidenced, as removal of the C-terminal Arg9 from Hoe140 (1, desArg9-[Hoe140]) caused selectivity to shift to B1 receptors.7 N-terminal modifications of Hoe140 by addition of a tosyl or a guanidinobenzoyl moiety produced B2 antagonists 2b and 3, respectively, with lower or similar potencies.6, 8 In fact, the N-terminal D-Arg0 was shown to stabilise the peptide towards carboxypeptidase N,4, 5 and the substitution of Pro7-Phe8 in BK for D-Tic7-Oic8 in Hoe140 prevented cleavage of the antagonist by angiotensin-converting enzyme (ACE).9   An alanine scan of Hoe140 had shown that all residues except Ser7 and to a lesser extent D-Arg0 and Hyp3 were important for activity, the most important residues being Pro3 and D-Tic7-Oic8.10 This led Amblard et al. to hypothesise that exchanging residues 7 and 8 in BK and Hoe140 by the core of various ACE inhibitors would stabilise the peptidomimetic towards ACE.11, 12 Incorporation of the S-benzothiazepinone (D-BT) moiety produced the best B2 receptor affinities for both the BK- and the Hoe140-variants (4a and 4b, respectively, figures 2 and 3), superior to benzoxazepinone (4c), a methylated benzothiazepinone (4d) as well as various lactams. Exchanging the N-terminal D-Arg0 for Lys even improved affinity 10-fold (4e). Surprisingly, all compounds were shown to be full agonists with good potencies.11, 12  Amblard furthermore showed that the D-BT motif adopts a type II´ β-turn in both solution and solid state.13 These results are consistent with previous research, which indicates that potent binding to the B2 receptor correlates with the degree of a C-terminal type II´ β-turn conformation,14-17 seen in the residues Ser6-Pro7-Phe8-Arg9 in BK.    Kyle et al. produced five Hoe140-mimics with the intention of showing the correlation between binding and structure (compounds 5, figure 3).15 Using contour plots they were able to show a trend in which the compounds with the best binding and antagonism (5c and 5a) (sub-nanomolar affinity and pA2 of 9.4 and 9.1, respectively) at the B2 receptor had the highest degree of a type II´ β-turn structure imposed by the bulky residues, whereas the compounds covalently cyclised using cysteines (5d and 5e) showed a higher propensity to adopt a different β-turn conformation, and therefore lacked binding and biological activity at B2.   D3   Figure 2: Bradykinin analogues  Based on molecular mechanics calculations Alcaro et al. introduced dipeptide isosteres in place of D-Tic7-Oic8 in Hoe140 to induce a β-turn conformation (6a and 6b, figure 3).18 However, these compounds lacked affinity and potency, with models showing poor overlap between both 6a and 6b and thus indicating the importance of the overall conformation of the peptide.   Similarly Ballet et al. incorporated the various forms of Aba-Gly or Aba-Ala in place of Pro7-Phe8 in BK (7a-7f, figure 2) or D-Tic7-Oic8 in Hoe140 (7g-7j, figure 3).19 As the researchers had previously shown Aba-dipeptides to adopt extended conformations, but spiro-Aba-dipeptides to prefer turn conformations, a distinct difference when incorporating these scaffolds was expected, to support the β-turn hypothesis. Unfortunately all the BK- and Hoe140-mimics showed significantly lowered affinities compared with the parent compounds. However, the spiro-Aba-Gly-Hoe140-mimics (7i and 7j) showed superior antagonism to 7g and 7h supporting the hypothesis. Molecular modelling indicated the positioning of the aromatic ring of Aba to be the problem.19 Interestingly, comparing the 7-membered rings produced by Amblard, Alcaro and Ballet (compound groups 4, 6 and 7) it seems the orientation of the aromatic rings plays a major role in determining biological activity including agonism/antagonism.11, 12, 18, 19 D4    Figure 3: Hoe140 analogues   D5  Meini et al. exchanged the dispensable Ser6 residue in Hoe140 for Dab6 (diaminobutyric acid) and cyclised the C-terminal of the peptide into a 14-membered lactam-ring (c(Dab- D-Tic-Oic-Arg)c(7γ-10α))(8, figure 3), as previous work in the group with a hexapeptide containing a 14-membered lactam had been shown to induce a β-turn conformation.20 The cyclised Hoe140-mimic showed B2-receptor affinity and pA2 comparable to Hoe140, which lead the authors to conclude that they had in fact stabilised a β-turn conformation.  The importance of the four C-terminal residues of Hoe140 was further demonstrated by Daffix et al. who synthesised a series of dimers from C-terminal fragments (compounds 9, figure 3).21 The (Ser- D-Tic-Oic-Arg)-dimer linked N-terminally by a succinyl (9a) showed poor binding and an IC50 of 247 nM, but was shown to be orally active in the rat. Optimised compounds, in which succinyl was exchanged for Arg or D-Arg (9b and 9c, respectively) showed improved binding and antagonist activity. Interestingly, none of the equivalent monomers showed binding to the B2 receptor, thus indicating the potential in dimerising the antagonists while also emphasising the importance of the C-terminal of Hoe140 for biological activity.  Potent B1-receptor antagonists have been synthesised by Gera et al.  based on the desArg9-[Hoe140] (1) sequence by exchanging D-Arg0 for two Lys (addition of a Lys was also shown by Amblard12 to improve affinity on B2-receptors as mentioned previously).22 The best of these was shown to have subnanomolar receptor affinity and high antagonistic activity (10a).  Newer research has suggested BK as a growth factor in prostate and lung carcinomas.23, 24 Adding pentafluorocinnamic acid (f5c) to the N-terminal of 10a resulted in a compound (10b) that was able to inhibit lung cancer in vivo by 86% as well as prostate cancer (PC3) by 43% in a xenotransplantation system.22    Summing up, it seems the four C-terminal amino acids of bradykinin/Hoe140 are essential for the biological activity. And the excellent affinity and potency of Hoe140 towards B2 receptors are most likely due to the backbone conformation induced by the 6-membered constrained amino acid D-Tic7, whereas 7-membered mimics constrain the backbone in the wrong conformation.    D6           Bradykinin analogues: Made by Amblard et al and Ballet et al. 4a showed agonism (pD2) rather than antagonism. 7a-7f were poor binders and not tested for antagonism. f) Human B2 receptors, g) Human umbilical vein; B2 receptors, presence of inhibitors BT = [amino]-5-(carbonylmethyl)-2,3-dihydro-1,5-benzothiazepin-4(5H)-one      1 2 3 4 5 6 7 8 9 Rec Ki (nM)  pD2 Ref BK R P P G F S P F R  - Compared to BK - - 4a R P P G F S D-BT G R B2 13.0f Ki = 0.65 nMf pD2 =7.9f 6.6g 11 7a R P P G F S D-Aba G R B2 pKi=5.6f    pKi=10f - 19 7b R P P G F S D/L-MeAba G R B2 pKi=6.1f - 19 7c R P P G F S D-Aba A R B2 54% @ 10mMf - 19 7d R P P G F S D-Aba D-A R B2 pKi=6.3f - 19 7e R P P G F S D/L-spiro-Aba G R B2 pKi=5.9f - 19 7f R P P G F S D/L-spiro-Aba G R B2 pKi=6.6f - 19 D7   Hoe140 analogues: All compounds were B2 selective, except 1, 10a and 10b which were B1 selective. All compounds were antagonists except 4b-4e. a) Guinea-pig ileum preparation; B2 receptors, b) Guinea-pig ileum organ preparation; B2 receptors c) Rat uterus preparation; B2 receptors, d) Rabbit aorta; B1 receptors, e) Rabbit jugular vein; B2 receptors, f) Human B2 receptors, g) Human umbilical vein; B2 receptors, presence of inhibitors, h) Bradykinin-stimulated prostaglandin synthesis, i) human umbilical artery smooth muscle; B1 receptors Hyp = Hydroxy-proline; Thi = β-thienyl-L-alanine; Oic = Octahydroindole-2-carboxylic acid p-GB=p-Guanidinobenzoyl; Tos = tosylate; BT = [amino]-5-(carbonylmethyl)-2,3-dihydro-1,5-benzothiazepin-4(5H)-one; Suc = succinyl; Boe =benzoxazepinone; β1/β2 = β-turn inducer; ψcyclised via cysteine bonds; Cpg = cyclopentylglycine; F5c = pentafluorocinnamic acid.  0 1 2 3 4 5 6 7 8 9 10 Rec IC50 (nM) Ki (nM)  pA2 pD2 Ref Hoe140 D-R R P Hyp G Thi S D-Tic Oic R  B2 1.1a/11b/4.9c 0.798a Comp. Hoe140 8.42b - 5 1 D-R R P Hyp G Thi S D-Tic Oic   B1 12d/29000a - IC50=Inactd/11nMa  - - 8 2a D-R R P Hyp G F S D-Tic Oic R  B2 -log(7.9)a/-log(9.2)c - IC50 -log(8.1)a/-log(9.3)c - - 7 2b  R(Tos) P Hyp G Thi S D-Tic Oic R  B2 -log(7.8)a/-log(8.4)c - - - 7 3 p-GB R P Hyp G Thi S D-Tic Oic R  B2 - - pA2 = 9.04f 9.58f - 9 4b D-R R P Hyp G Thi S D-BT G R  B2 - 0.7f  Ki=0.06 nMf pA2 =8.2g - 6.8g 11 4c D-R R P Hyp G Thi S Boe G R  B2 - 2.5f - ND 12 4d D-R R P Hyp G Thi S D-BT(Me) G R  B2 - 6f - 7.4g 12 4e K R P Hyp G Thi S D-BT G R  B2 - 0.07f - 7.1g 12 5a D-R R P Hyp G Thi S D-Tic Tic R  B2 - 0.381a  Not compared to Hoe140 in the article 9.1h - 15 5b D-R R P Hyp G Thi S D-Tic D-Tic R  B2 - 571a 5.6h - 15 5c D-R R P Hyp G Thi S D-Tic Aoc R  B2 - 0.176a 9.4h - 15 5d D-R R P Hyp G Thi C ψ D-Tic F R C ψ B2 - Inacta NDh - 15 5e D-R R P Hyp G Thi C ψ D-F F C ψ R B2 - 876a 5.4h - 15 6a D-R R P Hyp G Thi S β1 R   B2 - pKi=7.0f pKi=10.6f pA2=9.55a 5.6a - 18 6b D-R R P Hyp G Thi S β2 R   B2 - pKi=5.5f <5a - 18 7g D-R R P Hyp G Thi S D-Aba G R  B2 - pKi=6.5f  pKi=10.1f pA2 = 9.5i 5.6a - 19 7h D-R R P Hyp G Thi S L-Aba G R  B2 - pKi=6.7f - - 19 7i D-R R P Hyp G Thi S D-spiro-Aba G R  B2 - pKi=7.6f 5.5a - 19 7j D-R R P Hyp G Thi S L-spiro-Aba G R  B2 - pKi=8.5f 5.7a - 19 8 D-R R P Hyp G Thi (Dab D-Tic Oic R) 7γ-10α B2 - pKi=10.3f pKi=10.6f 8.14g - 20 9a    Suc D-D (S D-Tic Oic R)2  B2 - 6609f Not compared to Hoe140 in the article IC50=247e - 21 9b    R D-D (S D-Tic Oic R)2  B2 - 90f 5.49g - 21 9c    D-R D-D (S D-Tic Oic R)2  B2 - 76f 6.45g - 21 10a K-K R P Hyp G Cpg S D-Tic Cpg   B1 - 0.089i/0.29d Not compared to Hoe140 in article 9.11d - 22 10b F5c-K R P Hyp G Cpg S D-Tic Cpg   B1 - 0.169d 8.08d - 22 D8  Reference List   1.  Hall, J. M. Bradykinin receptors: Pharmacological properties and biological roles. Pharmacology & Therapeutics 1992, 56, 131-190.  2.  Regoli, D.; Barabe, J. Pharmacology of bradykinin and related kinins. Pharmacological reviews 1980, 32, 1-46.  3.  Marceau, F. Kinin B 1 receptors: a review. Immunopharmacology 1995, 30, 1-26.  4.  Wirth, K.; Hock, F. J.; Albus, U.; Linz, W.; Alpermann, H. G.; Anagnostopoulos, H.; Henke, S. T.; Breipohl, G.; König, W.; Knolle, J. Hoe 140 a new potent and long acting bradykinin-antagonist: in vivo studies. British journal of pharmacology 1991, 102, 774-777.  5.  Hock, F. J.; Wirth, K.; Albus, U.; Linz, W.; Gerhards, H. J.; Wiemer, G.; Henke, S. T.; Breipohl, G.; König, W.; Knolle, J. Hoe 140 a new potent and long acting bradykinin-antagonist: in vitro studies. British journal of pharmacology 1991, 102, 769-773.  6.  Lembeck, F.; Griesbacher, T.; Eckhardt, M.; Henke, S.; Breipohl, G.; Knolle, J. New, longacting, potent bradykinin antagonists. British journal of pharmacology 1991, 102, 297-304.  7.  Wirth, K.; Breipohl, G.; Stechl, J.; Knolle, J.; Henke, S.; Schölkens, B. DesArg 9 D-Arg [Hyp 3, Thi 5, D-Tic 7, Oic 8] bradykinin (desArg 10-[Hoe140]) is a bradykinin B 1 receptor antagonist. European journal of pharmacology 1991, 205, 217-218.  8.  Thurieau, C.; Félétou, M.; Canet, E.; Fauchére, J. L. p-Guanidinobenzoyl-[Hyp 3, Thi 5, D-Tic 7, Oic 8] bradykinin is almost completely devoid of the agonist effect of HOE140 on the endothelium-free femoral artery of sheep. Bioorganic & Medicinal Chemistry Letters 1994, 4, 781-784.  9.  Dorer, F. E.; Ryan, J. W.; Stewart, J. M. Hydrolysis of bradykinin and its higher homologues by angiotensin-converting enzyme. Biochemical Journal 1974, 141, 915.  10.  Galoppini, C.; Patacchini, R.; Meini, S.; Viganó, S.; Tancredi, M.; Quartara, L.; Triolo, A.; Maggi, C. A.; Rovero, P. A structure-activity study on the bradykinin B1 antagonist desArg10-HOE 140: The alanine scan. Letters in Peptide Science 1999, 6, 123-127.  11.  Amblard, M.; Daffix, I.; Bedos, P.; Bergé, G.; Pruneau, D.; Paquet, J. L.; Luccarini, J. M.; Bélichard, P.; Dodey, P.; Martinez, J. Design and synthesis of potent bradykinin agonists containing a benzothiazepine moiety. J. Med. Chem. 1999, 42, 4185-4192.  12.  Amblard, M.; Daffix, I.; Bergé, G.; Calmés, M.; Dodey, P.; Pruneau, D.; Paquet, J. L.; Luccarini, J. M.; Bélichard, P.; Martinez, J. Synthesis and characterization of bradykinin B2 receptor agonists containing constrained dipeptide mimics. J. Med. Chem. 1999, 42, 4193-4201.  13.  Amblard, M.; Raynal, N.; Averlant-Petit, M. C.; Didierjean, C.; Calmés, M.; Fabre, O.; Aubry, A.; Marraud, M.; Martinez, J. Structural elucidation of the -turn inducing (S)-[3-amino-4-oxo-2, 3-dihydro-5H-benzo [b][1, 4] thiazepin-5-yl] acetic acid (DBT) motif. Tetrahedron letters 2005, 46, 3733-3735. D9   14.  Chakravarty, S.; Wilkins, D.; Kyle, D. J. Design of potent, cyclic peptide bradykinin receptor antagonists from conformationally constrained linear peptides. J. Med. Chem. 1993, 36, 2569-2571.  15.  Kyle, D. J.; Martin, J. A.; Farmer, S. G.; Burch, R. M. Design and conformational analysis of several highly potent bradykinin receptor antagonists. J. Med. Chem. 1991, 34, 1230-1233.  16.  Kyle, D. J.; Martin, J. A.; Burch, R. M.; Carter, J. P.; Lu, S.; Meeker, S.; Prosser, J. C.; Sullivan, J. P.; Togo, J. Probing the bradykinin receptor: mapping the geometric topography using ethers of hydroxyproline in novel peptides. J. Med. Chem. 1991, 34, 2649-2653.  17.  Kyle, D. J.; Blake, P. R.; Smithwick, D.; Green, L. M.; Martin, J. A.; Sinsko, J. A.; Summers, M. F. NMR and computational evidence that high-affinity bradykinin receptor antagonists adopt C-terminal. beta.-turns. J. Med. Chem. 1993, 36, 1450-1460.  18.  Alcaro, M. C.; Vinci, V.; Anna, M. D.; Scrima, M.; Chelli, M.; Giuliani, S.; Meini, S.; Di Giacomo, M.; Colombo, L.; Papini, A. M. Bradykinin antagonists modified with dipeptide mimetic -turn inducers. Bioorganic & Medicinal Chemistry Letters 2006, 16, 2387-2390.  19.  Ballet, S.; De Wachter, R.; Van Rompaey, K.; Tömböly, C.; Feytens, D.; Töth, G.; Quartara, L.; Cucchi, P.; Meini, S.; Tourwe, D. Bradykinin analogs containing the 4-amino-2-benzazepin-3-one scaffold at the C-terminus. Journal of Peptide Science 2007, 13, 164-170.  20.  Meini, S.; Quartara, L.; Rizzi, A.; Patacchini, R.; Cucchi, P.; Giolitti, A.; Caló, G.; Regoli, D.; Criscuoli, M.; Maggi, C. A. MEN 11270, a novel selective constrained peptide antagonist with high affinity at the human B2 kinin receptor. Journal of Pharmacology and Experimental Therapeutics 1999, 289, 1250-1256.  21.  Daffix, I.; Amblard, M.; Bergé, G.; Dodey, P.; Pruneau, D.; Paquet, J.-L.; Fouchet, C.; Franck, R.-M.; Defréne, E.; Luccarini, J.-M. Synthesis and pharmacological evaluation of dimer derivatives of the bradykinin receptor antagonist HOE-140. The Journal of peptide research 1998, 52, 1-14.  22.  Gera, L.; Stewart, J. M.; Fortin, J. P.; Morissette, G.; Marceau, F. Structural modification of the highly potent peptide bradykinin B1 receptor antagonist B9958. International immunopharmacology 2008, 8, 289-292.  23.  Stewart, J. M.; Gera, L.; Chan, D. C.; York, E. J.; Simkeviciene, V.; Bunn, P. A.; Taraseviciene-Stewart, L. Combination cancer chemotherapy with one compound: pluripotent bradykinin antagonists. Peptides 2005, 26, 1288-1291.  24.  Chan, D.; Gera, L.; Helfrich, B.; Helm, K.; Stewart, J.; Whalley, E.; Bunn, P. Novel bradykinin antagonist dimers for the treatment of human lung cancers. Immunopharmacology 1996, 33, 201-204.                                                             i From Kyle et al. J Med Chem 1993, 36, 1450-1460 E1  FARNESYL TRANSFERASE The GTP-binding proteins of the Ras family are involved in the regulation of cellular growth and differentiation in mammalian cells.1 In order to function the proteins must be anchored to the cellular membrane, which requires a series of post-translational modifications to the C-terminal CA1A2X sequence (A being an aliphatic residue and X any residue), the first being a transfer of farnesyl to the cysteine residue (when X = Met or Ser) catalysed by the enzyme farnesyl transferase (FT).2-5 Mutated ras genes are found in 15% of all human cancers, mostly occurring in pancreas (90%), colon (50%) and lung (30%) carcinomas, and inhibition of FT in these cancers may therefore potentially block the oncogenic Ras signalling,1, 6 and finding inhibitors of FT has thus attracted much attention, yielding both non-peptide7 and constrained peptidomimetic inhibitors (vide infra).      A systematic study replacing A1 and A2 in the CA1A2X sequence with various amino acids revealed the CVFM sequence to be the best native tetrapeptide inhibitor of FT (IC50 = 25 nM).8, 9  Based on these results Clerc et al. performed molecular modelling studies on KCVFM showing a preference for extended conformations.10 The authors added the lysine N-terminally to confer water solubility, while retaining full biological activity, as KCVIM is the C-terminal pentapeptide in the K-RasB subtype. As in vitro studies showed that constraining the peptide by replacing Val3-Phe4 with Val3-Tic4 (1a) led to a 50-fold improvement in FT inhibition (1 μM and 20 nM, respectively) a series of compounds with constrained amino acids was synthesised (figure 1) to investigate the correlation between the conformation and inhibition properties of the peptidomimetics.  Replacing Phe4 with D-Tic4 (1b, 1c and 1d), which favours a turn-like conformation in the pentapeptides, results in IC50s in the high micromolar range. Similarly for the pentapeptides containing D-conformers of the third residue, whether it is Val3 or the constrained Pro3 (1d, 1e and 1f), FT inhibition is only weak.  In contrast, incorporating residues which lead to extended conformations of the pentapeptides show low nanomolar inhibition (1a, 1g and 1h), except in the case of the Pro-Tic pair (1i) proving that the CVFM sequence is the superior inhibitor.8, 9 Comparing Phe4 and Tic4-containing peptides (1j vs. 1g and 1k vs. 1h) confirms the superiority of Tic for inhibition of FT as well as inducing extended conformations, the best inhibitor being 1g with and IC50 of 5 nM and 95% of the conformers with an extended conformation.           E2   Figure 1: Modifications to the sequence KCA1A2M  Marsters et al. similarly showed that replacing Phe3 with Tic3 in the N-terminally acetylated CVFM tetrapeptide (2a, figure 2) improved inhibitory activity more than 20-fold.11 Furthermore, a small SAR indicated that replacing either A1 or A2 in CA1A2X by aromatic residues was favoured, whereas replacing both with aromatic residues was not, leading the authors to suggest that these residues might occupy the same confined pocket in FT.  Based on the CVFM sequence researchers at Bristol-Myers Squibb conducted a sequential SAR. First they examined the effect of exchanging the Phe3 for constrained phenylalanine derivatives Tic (3a), 2-amino-3-phenylacrylic acid (Paa, 3b), (S)-indoline-2-carboxylic acid (Ica, 3c) and 2-aminoindane-2-carboxylic acid (Aic, 3d) (figure 2).12    E3   Figure 2: Modifications to the CVFM-sequence  E4  Since CV(Tic)M proved to be the best inhibitor scaffold for FT (3a, IC50 = 1 nM) the researchers  then examined the effects of replacing the amide bonds with the reduced amide isotere, as previous efforts had led to isosteres with improved inhibition effects and enhanced whole cell activity for CAAX-based FT inhibitors.13-16 Furthermore, the modifications were expected to lead to more stable analogues as proteolytic degradation would be prevented.  As had also been shown previously on the CVFM-scaffold15 incorporation of the methyleneamine in place of the amide functionality was well tolerated between Cys and Val (3e) and Val and Tic (3f), whereas the methyleneamine between Tic and Met (3g) led to a detrimental loss in in vitro inhibition. Having methyleneamine between both Cys and Val as well as Val and Tic also led to a potent inhibitor, 3h. Both 3e and 3h showed good selectivity, as they did not to a large extend inhibit geranylgeranyltransferase 1 (GGT1, an enzyme sharing a common subunit with FT and catalysing similar reactions). The importance of the Tic3 residue was highlighted by the fact that removing the phenyl ring from the residue, thereby only incorporating the heterocyclic 1,2,3,6-tetrahydropyridine-ring (THP) (3i, both stereoisomers) markedly reduced inhibitory potency. The authors concluded that the role of the heterocyclic ring in Tic serves to orient the phenyl-ring of Tic in the optimal binding orientation, rather than orienting the tetrapeptide into a bioactive conformation. This was supported the fact that N-methylation or incorporation of Aic3 (3d) in place of Tic3 did not produce potent inhibitors.12, 15  Both 3e and 3h showed 80% or more ras transformation inhibition in the RTI assay in NIH-3T3 cells, an assay which evaluates the ability of the compound to inhibit the transformation of cells transfected with oncogenic H-ras DNA. Both compounds were also able to inhibit the anchorage-independent growth of transfected cells in soft agar (SAG assay) with an EC50 of 5.0 μM. However, data indicated that both 3e and 3h were poorly cell permeable leading the authors to develop another SAR using 3e as a lead scaffold, but varying the sidechains of residues 2 and 4. Interestingly, exchanging Met4 for a norleucine (Nle4) (3j), which is often used as a stable isosteric replacement for Met, changes selectivity towards GGT1. Exchanging Val2 to Tbg2 (in effect adding a methyl group, compounds 3k and 3l) weakens the in vitro FT inhibition compared to the parent compound 3e, but increases RTI inhibition to 100% at 10 μM and also improves SAG EC50 to 0.19 and 0.41 μM, respectively. As both compounds were membrane permeable and showed no cytotoxicity, 3k and 3l were evaluated in vivo in mice implanted with H-ras-transformed tumor cells. I.p. injections of 45mg/kg twice a day for 11 days of either 3k or 3l prolonged survival time in treated vs. control mice significantly.         However, thiol-containing compounds may be oxidised in vivo, a disadvantage when developing therapeutic agents. The same research group therefore replaced the Cys1 with an imidazole linked to Val2 E5  using a short, saturated chain,17 as it is known that FT requires both zinc and magnesium ions to function.18 Exchanging the Cys1 for 5-ethyl-1H-imidazole in 3e and 3k resulted in 4a and 4b, respectively (figure 2). Both 4a and 4b showed low in vitro FT inhibition and good selectivity. Furthermore both compounds were active at low micromolar concentrations in the SAG assay without displaying cytotoxicity.    The development of peptide FT inhibitors for use in the clinic to treat various cancers linked with ras mutations therefore seems promising, however, Bristol-Myers Squibb have apparently so far only moved forward into clinical trials with a non-peptidic benzodiazepine-based inhibitor.19        E6                     Farnesyl transferase inhibitors: Based on the KCVFM or CVFM sequences. Paa = 2-amino-3-phenylacrylic acid; Ica = (S)-indoline-2-carboxylic acid; Thp = 1,2,3,6-tetrahydropyridine-2-carboxylic acid, Tbg = tert-butyl glycine; Im-Et- = 5-Ethyl-1H-imidazole.     1 2 3 4 5  IC50 (nM) FT % extended structure IC50 (nM) GGT1 RTI inhib (%) 10 μM SAG EC50 (μM) Ref Native seq. K C V F M  1,000 67    10 1a K C V Tic M  20 80 - - - 10 1b K C V D-Tic M  100,000 6 - - - 10 1c K C P D-Tic M  50,000 15 - - - 10 1d K C D-P D-Tic M  100,000 34 - - - 10 1e K C D-P Tic M  100,000 10 - - - 10 1f K C D-V Tic M  15,000 20 - - - 10 1g K C (N-Me)V Tic M  5 95 - - - 10 1h K C Tbg Tic M  40 98 - - - 10 1i K C P Tic M  2,000 91 - - - 10 1j K C (N-Me)V F M  1,000 55 - - - 10 1k K C Tbg F M  1,000 56 - - - 10 2a Ac- C V Tic M  11 - - - - 11   C V F M  37 - 8100 - - 12 3a  C V Tic M  1.0 - - - - 12 3b  C V Paa M  160 - - - - 12 3c  C V Ica M  120 - - - - 12 3d  C V Aic M  2,900 - - - - 12 3e  C* (X1 = H,H) V Tic M  0.6 - 110 >80 5.0 12 3f  C V* (X2 = H,H) Tic M  0.37 - - 0.0 - 12 3g  C V Tic*(X3 = H,H) M  5300 - - - - 12 3h  C* (X1 = H,H) V* (X2 = H,H) Tic M  0.75 - 4100 80 5.0 12 3i  C* (X1 = H,H) V L- or D-Thp M  270/3900 - - - - 12 3j  C* (X1 = H,H) V Tic Nle  770 - 1.3 20 26 12 3k  C* (X1 = H,H) Tbg Tic Q  2.8 - 1400 100 0.19 12 3l  C* (X1 = H,H) Tbg Tic M -OMe 85 - 200 100 0.41 12 4a  Im-Et- V  Tic M  0.79 - 234 - 3.8 17 4b  Im-Et- Tbg  Tic Q  3.4 - 30,500 - 3.2 17 E7  Reference List   1.  Bos, J. L. Ras oncogenes in human cancer: a review. Cancer research 1989, 49, 4682-4689.  2.  Willumsen, B. M.; Norris, K.; Papageorge, A. G.; Hubbert, N. L.; Lowy, D. R. Harvey murine sarcoma virus p21 ras protein: biological and biochemical significance of the cysteine nearest the carboxy terminus. The EMBO journal 1984, 3, 2581.  3.  Hancock, J. F.; Magee, A. I.; Childs, J. E.; Marshall, C. J. All ras proteins are polyisoprenylated but only some are palmitoylated. Cell 1989, 57, 1167-1177.  4.  Jackson, J. H.; Cochrane, C. G.; Bourne, J. R.; Solski, P. A.; Buss, J. E.; Der, C. J. Farnesol modification of Kirsten-ras exon 4B protein is essential for transformation. Proceedings of the National Academy of Sciences 1990, 87, 3042-3046.  5.  Kato, K.; Cox, A. D.; Hisaka, M. M.; Graham, S. M.; Buss, J. E.; Der, C. J. Isoprenoid addition to Ras protein is the critical modification for its membrane association and transforming activity. Proceedings of the National Academy of Sciences 1992, 89, 6403-6407.  6.  Barbacid, M. Ras genes. Annual review of biochemistry 1987, 56, 779-827.  7.  Buss, J. E.; Marsters Jr, J. C. Farnesyl transferase inhibitors: the successes and surprises of a new class of potential cancer chemotherapeutics. Chemistry & Biology 1995, 2, 787-791.  8.  Goldstein, J. L.; Brown, M. S.; Stradley, S. J.; Reiss, Y.; Gierasch, L. M. Nonfarnesylated tetrapeptide inhibitors of protein farnesyltransferase. Journal of Biological Chemistry 1991, 266, 15575-15578.  9.  Reiss, Y.; Stradley, S. J.; Gierasch, L. M.; Brown, M. S.; Goldstein, J. L. Sequence requirement for peptide recognition by rat brain p21ras protein farnesyltransferase. Proceedings of the National Academy of Sciences 1991, 88, 732-736.  10.  Clerc, F. F.; Guitton, J. D.; Fromage, N.; Leliévre, Y.; Duchesne, M.; Tocqué, B.; James-Surcouf, E.; Commerçon, A.; Becquart, J. Constrained analogs of KCVFM with improved inhibitory properties against farnesyl transferase. Bioorganic & Medicinal Chemistry Letters 1995, 5, 1779-1784.  11.  Marsters, J. C.; McDowell, R. S.; Reynolds, M. E.; Oare, D. A.; Somers, T. C.; Stanley, M. S.; Rawson, T. E.; Struble, M. E.; Burdick, D. J.; Chan, K. S. Benzodiazepine peptidomimetic inhibitors of farnesyltransferase. Bioorganic & medicinal chemistry 1994, 2, 949-957.  12.  Leftheris, K.; Kline, T.; Vite, G. D.; Cho, Y. H.; Bhide, R. S.; Patel, D. V.; Patel, M. M.; Schmidt, R. J.; Weller, H. N.; Andahazy, M. L. Development of highly potent inhibitors of Ras farnesyltransferase possessing cellular and in vivo activity. J. Med. Chem. 1996, 39, 224-236.  13.  Garcia, A. M.; Rowell, C.; Ackermann, K.; Kowalczyk, J. J.; Lewis, M. D. Peptidomimetic inhibitors of Ras farnesylation and function in whole cells. Journal of Biological Chemistry 1993, 268, 18415-18418.  14.  Graham, S. L.; deSolms, S. J.; Giuliani, E. A.; Kohl, N. E.; Mosser, S. D.; Oliff, A. I.; Pompliano, D. L.; Rands, E.; Breslin, M. J. Pseudopeptide Inhibitors of Ras Farnesyl-Protein Transferase. J. Med. Chem. 1994, 37, 725-732.  15.  Leftheris, K.; Kline, T.; Natarajan, S.; DeVirgilio, M. K.; Cho, Y. H.; Pluscec, J.; Ricca, C.; Robinson, S.; Seizinger, B. R.; Manne, V.; Meyers, C. A. Peptide based P21RAS farnesyl transferase inhibitors: systematic modification of the tetrapeptide CA1A2X motif. Bioorganic & Medicinal Chemistry Letters 1994, 4, 887-892. E8   16.  Kohl, N. E.; Mosser, S. D.; deSolms, S. J.; Giuliani, E. A.; Pompliano, D. L.; Graham, S. L.; Smith, R. L.; Scolnick, E. M.; Oliff, A.; Gibbs, J. B. Selective inhibition of ras-dependent transformation by a farnesyltransferase inhibitor. Science 1993, 260, 1934-1937.  17.  Hunt, J. T.; Lee, V. G.; Leftheris, K.; Seizinger, B.; Carboni, J.; Mabus, J.; Ricca, C.; Yan, N.; Manne, V. Potent, cell active, non-thiol tetrapeptide inhibitors of farnesyltransferase. J. Med. Chem. 1996, 39, 353-358.  18.  Reiss, Y.; Brown, M. S.; Goldstein, J. L. Divalent cation and prenyl pyrophosphate specificities of the protein farnesyltransferase from rat brain, a zinc metalloenzyme. Journal of Biological Chemistry 1992, 267, 6403-6408.  19.  Cortes, J.; Faderl, S.; Estey, E.; Kurzrock, R.; Thomas, D.; Beran, M.; Garcia-Manero, G.; Ferrajoli, A.; Giles, F.; Koller, C. Phase I study of BMS-214662, a farnesyl transferase inhibitor in patients with acute leukemias and high-risk myelodysplastic syndromes. Journal of clinical oncology 2005, 23, 2805-2812.  F1  NEUROKININ-1  Substance P (SP, Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2), neurokinin A (NK-A) and neurokinin B (NK-B) are members of the tachykinin family, which are endogenous neuropeptides that bind to the G-protein coupled receptors NK-1, NK-2 and NK-3 with highest affinity, respectively.1, 2 The NK-1 receptor is expressed widely in CNS and peripheral tissues and release of its ligand, SP, is involved in pain transmission and noxious stimuli,1, 2 which makes NK-1 antagonists potential therapeutics in pathological states such as postoperative pain,3 arthiritis,4 asthma2 and emesis.5.  To investigate the size of the binding pockets for Phe7 and Phe8 in the NK-1 receptor and the binding conformation of the Phe7 and Phe8 side chains  in SP by replacing these with constrained amino acids Josien et al. performed a peptidomimetic study on the substance P sequence (figure 1).6     Figure 1: Substance P (SP) and the modified agonists 1a-1d  F2  When Phe7 was replaced with Tic7 (1a) the g+ conformation of the Tic7 distorted the α-helical structure that the core of SP (Lys3(CO)-Phe8(NH)) normally adopts and binding affinity to all NK-receptors decreased substantially. The same was seen when replacing Phe8 with Tic8 (1b). This was confirmed in the GPI assay, which utilises the guinea pig ileum that contains only NK1 receptors, as 1a and 1b were shown to be poor agonists of the NK-1 receptor. Furthermore the studies included the side chain constrained Phe analogues such as indenylglycine (Ing) and fluorenylglycine (Flg). Using these in place of Phe7 or Phe8 the authors concluded that the binding pocket for Phe7 was quite small and allowed for one aromatic ring, since [(2S,3S)Ing7]SP (1c) was the most potent analogue and Phe7 most likely preferably adopts the g- orientation. In contrast, the activity of [(2S)Flg8]SP (1d) indicated that the binding pocket for Phe8 was large enough to fit two aromatic rings in the g- conformation and one in the anti conformation.6 These conclusions were confirmed in CHO cells expressing the NK-1 receptor, where cAMP and inositol phosphate (IP) production were significantly lowered when stimulating with 1a and 1b, whereas compounds with residues 7 and 8 in the g- or anti conformation (1c and 1d) were able to increase cAMP/IP production slightly compared to endogeneous SP.7  These strategies, however, did not result in a substantially more potent agonist.  The most interesting results were obtained in the development of peptidomimetic  antagonists.   Based on the NK-1 antagonist lead L-732,138 (Ac-Trp-O-3’5’(CF3)2-Bn), which showed low nanomolar binding (IC50 = 1.6 nM8/ 0.73 nM9) and inhibition (Kb = 25 nM)8, Millet et al. developed constrained analogues  2a-2d containing either a tetrahydrocarboline or tetrahydro-carbazole skeleton (figure 2).10 Unfortunately, the constrained derivatives 2a-2d all showed weaker NK-1 binding.  Compounds 2b and 2d were, however, 6- and 4-fold better NK-1 binders than 2a and 2c, respectively, indicating that short peptide sequences may produce better NK-1 ligands than small molecule compounds. A superpositioning of 2a and 2c with L-732,138 after energy minimisation indicated that the rigidity in 2a and 2c due to the constraints was preventing the compounds from attaining the bioactive conformation.  Based on the same lead, Ballet et al. prepared the N-methyl amide Tic-analogue 2e, the Aia-constrained amide 2f as well as the Tic analogue 2g. Neither 2e nor 2f showed any NK-1 antagonism.  Surprisingly, the Tic analogue 2g (hNK-1 Ki = 32 nM) was more potent than ester 2a (hNK-1 Ki = 390 nM)  and showed antagonism with pA2 = 7.5.11 F3   Figure 2: Constraining the Trp residue in the NK-1 antagonist L-732,138.  The latter result indicated that in such constrained peptidomimetics, an indole ring is not essential for NK-1 affinity, and can be replaced by a benzene ring as had already been demonstrated in a variety of non-peptide NK-1 antagonists.12-14 The benzodiazepinone antagonist 3a served as an inspiration to design the 1-phenyl-substituted Aba-containing analogue 3b, which unfortunately turned out to be inactive (figure 3).11  However, removal of the 1-phenyl substituent led to a potent NK-1 antagonist 3c (hKN-1 Ki = 27 nM, pA2 = 8.4).  Since the Aba-Gly constraint was also present in potent opioid tetrapeptide analogues, Ballet et al. suspected that these two pharmacophores could be merged into a compact opioid agonist/NK-1 antagonist.  F4   Figure 3: Non-peptide NK-1 antagonist 3a as inspiration for potent peptidomimetic NK-1 antagonists  As prolonged usage of opioids in pain treatment has been shown to increase the secretion SP and upregulate the expression of tachykinin receptors counteracting the effect of the analgesics,15, 16 bifunctional ligands that act as opioid agonists and NK-1 antagonists have been developed.15, 17-22  There are several advantages to such chimeras, including single biodistribution, simple pharmacokinetic profile and higher efficacy.19 Hruby and Lipkowski designed chimeras containing the enkephalin-based opioid pharmacophore Tyr-D-Ala-Gly-Phe-Met- in the N-terminal, fused with the C-terminal NK-1 pharmacophore -Met-Pro-Leu-Trp-NH/O-3’,5’(CF3)2-Bn (figure 4, top). Both the ester and amide chimeras showed enhanced antinociception in acute pain models but also prevented opioid-induced tolerance in chronic trials.18, 22 F5   Figure 4: Opioid agonist/NK-1 antagonist chimeras. 3g developed by Ballet et al. and the derivatives thereof (compound group 4) by Guillemyn et al.  For the discovery of 3c Ballet et al. had synthesised several potential NK-1 antagonists containing the constrained amino acids 1-phenyl-Aba, Aba, Tic, Aia or Tcc. Interestingly, the analogues containing the tryptophane mimics Aia or Tcc showed no NK-1 antagonism, whereas the Tic- and Aba-containing analogues 2g, 3c and 3d showed nanomolar affinity and good NK-1 antagonism. As 3c (blue box in figure 4) proved to have the best affinity and was most potent antagonist for NK-1, it was fused with the opioid pharmacophore 3e (red box in figure 4), which showed subnanomolar affinity and agonism at both μ and δ receptors.11 This yielded the opioid/NK-1-chimera, 3f, which showed a slightly reduced antagonism activity, but even better NK-1 affinity than 2g, 3c and 3d. The opioid binding and activity were lowered at both μ and δ receptors but were still in the low nanomolar range. Even with these minor losses in affinity and potency 3f proved to be a potent candidate for an opioid agonist/NK-1 antagonist chimera. An in vivo study in mice comparing 3f (2 mg/kg or 4 mg/kg i.v.) with morphine (4 mg/kg i.v.) unfortunately showed 3f to have a slightly lowered acute antinociceptive profile with a maximum effect after 2 hours.23 Comparison F6  with the ‘pure’ opioid agonist 3e, which was shown to be a potent antinociceptive with a longer lasting acute effect of up to 7 hours, highlighted the importance of strong μ- and δ-receptor affinity and activation for the analgesic effect. Unfortunately, the 3f chimera did not show any effect on the opioid-induced tolerance, which the NK-1 antagonism was expected to counteract, and showed a complete loss in antinociception on the fifth day of administration.  Recently a substantial SAR was performed in order to improve the affinity and potency of the chimera 3f (figure 4).24 In an attempt to avoid steric clashes between the two pharmacophores Gly was exchanged for β-Ala (4a), which resulted in improved affinity at both δ- and μ-receptors, improved δ-receptor activation and only slightly lowered μ-receptor activation. However, the extra methylene unit decreased NK-1 binding 120-fold and lowered antagonism by almost two log units. Furthermore, D-Arg was replaced with D-Cit in order to remove a charge in the chimera (4c) producing a compound with good binding and potency at all receptors. As molecular modelling had shown the cis conformer to be present in one of the three lowest energy conformations,11 the N-methyl amide was exchanged for an N-iBu amide in several compounds (4e-4h) in order to increase the amount of cis amide conformer relative to trans conformer. This did, however, not improve NK-1 affinity, probably due to the bulkiness of the N-iBu group. The best analogue of the SAR, 4b, had a lowered molecular weight as the trifluoromethyl groups were removed, which improved opioid binding and agonism as well as NK-1 binding and antagonism, however not surpassing the affinity and potency of 3f at the NK-1 receptor in vitro. In vivo testing, however, proved 4b to be more active than 3f in neuropathic pain models testing allodynia and hyperalgesia, as well as having an improved toxicity profile. In acute pain 3f and 4b had similar potency profiles inferior to ‘pure’ opioid agonists (3e and 4d) but superior to morphine. Unfortunately, unlike 3f, 4b was not able to pass the blood-brain barrier.  With these advances in finding a opioid agonist/NK-1 antagonist chimera it seems likely that a ligand which may perform well in vivo may be obtained in due course. Furthermore, knowing more of the binding pockets in the NK-1 receptor may help improve the success in the efforts.                F7  Neurokinin-1 agonist study a) Rat brain synaptosomes [125I]BHSP, b) Rat duodenum membranes [3H]NKA, c) Rat brain synaptosomes [125I]BHELE, d) Guinea pig membranes [125I]BHSP,  e) CHO cells expressing hNK-1, f) [3H][Pro9]SP in CHO cells expressing hNK-1 Ing = indenylglycine; Flg = fluorenylglycine.   1 2 3 4  5 hNK-1 Ki (nM)a NK-1 pA2 hNK-2 Ki (nM)b GPI (μ) IC50 (nM)c MVD (δ) IC50 (nM)d MOR Ki (nM)c DOR Ki (nM)d KOR Ki (nM)e Selectivity DOR Ki/MOR Ki Ref 2a   Ac- Tcc -O- 3’5’(CF3)2-Bn 390 - >10000 - - - - - - 10 2b Cbz- G L Tcc -O- 3’5’(CF3)2-Bn 63 - >10000 - - - - - - 10 2c   Ac- Thc -O- 3’5’(CF3)2-Bn 795 - >10000 - - - - - - 10 2d Cbz- G L Thc -O- 3’5’(CF3)2-Bn 200 - >10000 - - - - - - 10 2g   Ac- Tic -NMe- 3’5’(CF3)2-Bn 32 7.5 - - - - - - - 11 3c  Ac- Aba G -NMe- 3’5’(CF3)2-Bn 27 8.4 - - - - - - - 11,24 3d  Ac- Aba G  3’5’(CF3)2-Bn 387 6.2 - - - - - - - 11 3e Dmt D-R Aba G -NH2  - - - 0.32 0.42 0.15 0.60 118 4.0 11,20,24 3f Dmt D-R Aba G -NMe- 3’5’(CF3)2-Bn 0.5 7.8 - 8.51 43.3 0.416 10.4 445 25.0 11,20,24 4a Dmt D-R Aba β-Ala -NMe- 3’5’(CF3)2-Bn 59.7 6.04 - 11.3 6.20 0.08 2.14 - 25.3 24 4b Dmt D-R Aba β-Ala -NMe- Bn 13.0 6.44 - 1.86 2.16 0.08 0.28 - 3.5 24 4c Dmt D-Cit Aba β-Ala -NMe- 3’5’(CF3)2-Bn 34.7 6.27 - 2.03 1.06 0.37 0.55 - 1.5 24 4d Dmt D-R Aba β-Ala -NH2  - - - 0.80 0.24 1.34 17.0 - 12.7 24 4e Dmt D-R Aba G -NiBu- 3’5’(CF3)2-Bn 35 - - 19.7 14.5 - - - - 24 4f Dmt D-R Aba G -NiBu- Bn 1020 - - P.A.* 28.7 - - - - 24 4g Dmt D-R Aba β-Ala -NiBu- 3’5’(CF3)2-Bn 743 - - - - - - - - 24 4h Dmt D-R Aba β-Ala -NiBu- Bn 308 - - - - - - - - 24 Neurokinin-1 antagonist studies: Studies of NK-1 antagonists and NK-1 antagonist/opioid agonist chimeras a) CHO cells [3H]SP, b) CHO cells [3H]NKA, c) Rat brain membrane [3H]DAMGO, d) Rat brain membrane [3H]DSLET, e) Guinea pig brain membrane [3H]U69,593  Thc = 2-amino-2,3,4,9-tetrahydro-1H-carbazole-2-carboxylic acid; Cit = citrulline.  * P.A. = partial agonist. Values for 2e, 2f and 3b are not reported.  1 2 3 4 5 6 7 8 9 10 11 NK-1  IC50 (nM)a NK-2  IC50 (nM)b NK-3  IC50 (nM)c GPI  IC50 (nM)d EC50  (nM)d EC50 (nM) cAMPe Ki  (nM)f EC50 (nM) IPse Ref SP R P K P Q Q F F G L M-NH2 0.64 200 130 1.6 2.5 8 1.6 1.0 6,7 1a  R P K P Q Q Tic F G L M-NH2 2400 520 1900 n.d. 417 >10000 365 373 6,7 1b  R P K P Q Q F Tic G L M-NH2 1200 18000 20000 2500 83 6000 374 193 6,7 1c R P K P Q Q (2S,3S)Ing F G L M-NH2 3.1 380 2200 n.d. 1.2 5.3 3.7 3.0 6,7 1d R P K P Q Q F (2S)Flg G L M-NH2 2.5 41 91 n.d. 5.7 4.8 2.1 1.8 6,7 F8  Reference List   1.  Longmore, J.; Hill, R. G.; Hargreaves, R. J. Neurokinin-receptor antagonists: pharmacological tools and therapeutic drugs. Canadian journal of physiology and pharmacology 1997, 75, 612-621.  2.  Leroy, V.; Mauser, P.; Gao, Z.; Peet, N. P. Neurokinin receptor antagonists. Expert opinion on investigational drugs 2000, 9, 735-746.  3.  Moskowitz, M. A. Neurogenic versus vascular mechanisms of sumatriptan and ergot alkaloids in migraine. Trends in pharmacological sciences 1992, 13, 307-311.  4.  Lotz, M.; Carson, D. A.; Vaughan, J. H. Substance P activation of rheumatoid synoviocytes: neural pathway in pathogenesis of arthritis. Science 1987, 235, 893-895.  5.  Rupniak, N. M.; Tattersall, F. D.; Williams, A. R.; Rycroft, W.; Carlson, E. J.; Cascieri, M. A.; Sadowski, S.; Ber, E.; Hale, J. J.; Mills, S. G. In vitro and in vivo predictors of the anti-emetic activity of tachykinin NK 1 receptor antagonists. European journal of pharmacology 1997, 326, 201-209.  6.  Josien, H.; Lavielle, S.; Brunissen, A.; Saffroy, M.; Torrens, Y.; Beaujouan, J. C.; Glowinski, J.; Chassaing, G. Design and synthesis of side-chain conformationally restricted phenylalanines and their use for structure-activity studies on tachykinin NK-1 receptor. J. Med. Chem. 1994, 37, 1586-1601.  7.  Sagan, S.; Josien, H.; Karoyan, P.; Brunissen, A.; Chassaing, G.; Lavielle, S. Tachykinin NK-1 receptor probed with constrained analogues of substance P. Bioorganic & medicinal chemistry 1996, 4, 2167-2178.  8.  MacLeod, A. M.; Merchant, K. J.; Cascieri, M. A.; Sadowski, S.; Ber, E.; Swain, C. J.; Baker, R. N-Acyl-L-tryptophan benzyl esters: potent substance P receptor antagonists. J. Med. Chem. 1993, 36, 2044-2045.  9.  Yamamoto, T.; Nair, P.; Vagner, J.; Largent-Milnes, T.; Davis, P.; Ma, S. w.; Navratilova, E.; Moye, S.; Tumati, S.; Lai, J.; Yamamura, H. I.; Vanderah, T. W.; Porreca, F.; Hruby, V. J. A StructureActivity Relationship Study and Combinatorial Synthetic Approach of C-Terminal Modified Bifunctional Peptides That Are / Opioid Receptor Agonists and Neurokinin 1 Receptor Antagonists. J. Med. Chem. 2008, 51, 1369-1376.  10.  Millet, R.; Goossens, J. F.; Bertrand-Caumont, K.; Chavatte, P.; Houssin, R.; Hénichart, J. P. Synthesis and biological evaluation of conformationally restricted derivatives of tryptophan as NK1/NK2 ligands. Letters in Peptide Science 1999, 6, 221-233.  11.  Ballet, S.; Feytens, D.; Buysse, K.; Chung, N. N.; Lemieux, C.; Tumati, S.; Keresztes, A.; Van Duppen, J.; Lai, J.; Varga, E. Design of novel neurokinin 1 receptor antagonists based on conformationally constrained aromatic amino acids and discovery of a potent chimeric opioid agonist-neurokinin 1 receptor antagonist. J. Med. Chem. 2011, 54, 2467-2476.  12.  Young, J. R.; Eid, R.; Turner, C.; DeVita, R. J.; Kurtz, M. M.; Tsao, K. L.; Chicchi, G. G.; Wheeldon, A.; Carlson, E.; Mills, S. G. Pyrrolidine-carboxamides and oxadiazoles as potent hNK1 antagonists. Bioorganic & Medicinal Chemistry Letters 2007, 17, 5310-5315. F9   13.  Desai, M. C.; Lefkowitz, S. L.; Thadeio, P. F.; Longo, K. P.; Snider, R. M. Discovery of a potent substance P antagonist: recognition of the key molecular determinant. J. Med. Chem. 1992, 35, 4911-4913.  14.  Armour, D. R.; Aston, N. M.; Morriss, K. M. L.; Congreve, M. S.; Hawcock, A. B.; Marquess, D.; Mordaunt, J. E.; Richards, S. A.; Ward, P. 1,4-Benzodiazepin-2-one derived neurokinin-1 receptor antagonists. Bioorganic & Medicinal Chemistry Letters 1997, 7, 2037-2042.  15.  Yamamoto, T.; Nair, P.; Jacobsen, N. E.; Kulkarni, V.; Davis, P.; Ma, S. w.; Navratilova, E.; Yamamura, H. I.; Vanderah, T. W.; Porreca, F. Biological and conformational evaluation of bifunctional compounds for opioid receptor agonists and neurokinin 1 receptor antagonists possessing two penicillamines. J. Med. Chem 2010, 53, 5491-5501.  16.  Kalso, E. Improving opioid effectiveness: from ideas to evidence. European Journal of Pain 2005, 9, 131-135.  17.  Bonney, I. M.; Foran, S. E.; Marchand, J. E.; Lipkowski, A. W.; Carr, D. B. Spinal antinociceptive effects of AA501, a novel chimeric peptide with opioid receptor agonist and tachykinin receptor antagonist moieties. European journal of pharmacology 2004, 488, 91-99.  18.  Hruby, V. J. Organic chemistry and biology: chemical biology through the eyes of collaboration. The Journal of organic chemistry 2009, 74, 9245-9264.  19.  Morphy, R.; Rankovic, Z. Designing multiple ligands-medicinal chemistry strategies and challenges. Current pharmaceutical design 2009, 15, 587-600.  20.  Yamamoto, T.; Nair, P.; Davis, P.; Ma, S. w.; Navratilova, E.; Moye, S.; Tumati, S.; Lai, J.; Vanderah, T. W.; Yamamura, H. I. Design, synthesis, and biological evaluation of novel bifunctional C-terminal-modified peptides for / opioid receptor agonists and neurokinin-1 receptor antagonists. J. Med. Chem. 2007, 50, 2779-2786.  21.  Yamamoto, T.; Nair, P.; Jacobsen, N. E.; Davis, P.; Ma, S. w.; Navratilova, E.; Moye, S.; Lai, J.; Yamamura, H. I.; Vanderah, T. W. The importance of micelle-bound states for the bioactivities of bifunctional peptide derivatives for / opioid receptor agonists and neurokinin 1 receptor antagonists. J. Med. Chem. 2008, 51, 6334-6347.  22.  Yamamoto, T.; Nair, P.; Ma, S. w.; Davis, P.; Yamamura, H. I.; Vanderah, T. W.; Porreca, F.; Lai, J.; Hruby, V. J. The biological activity and metabolic stability of peptidic bifunctional compounds that are opioid receptor agonists and neurokinin-1 receptor antagonists with a cystine moiety. Bioorganic & medicinal chemistry 2009, 17, 7337-7343.  23.  Guillemyn, K.; Kleczkowska, P.; Novoa, A.; Vandormael, B.; Van den Eynde, I.; Kosson, P.; Asim, M. F.; Schiller, P. W.; Spetea, M.; Lipkowski, A. W. In vivo antinociception of potent mu opioid agonist tetrapeptide analogues and comparison with a compact opioid agonist-neurokinin 1 receptor antagonist chimera. Molecular brain 2012, 5, 1-11.  24.  Guillemyn, K.; Kleczkowska, P.; Lesniak, A.; Dyniewicz, J.; Van der Poorten, O.; Van den Eynde, I.; Keresztes, A.; Varga, E.; Lai, J.; Porreca, F. Synthesis and biological evaluation of compact, conformationally constrained bifunctional opioid agonist-Neurokinin-1 antagonist peptidomimetics. European journal of medicinal chemistry 2015, 92, 64-77.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0228345/manifest

Comment

Related Items