UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthesis and evaluation of probes of RNA polymerase II : adaptation of the bicyclic scaffold of the… Dietrich, David John 2010

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_2010_fall_dietrich_david.pdf [ 122.01MB ]
Metadata
JSON: 24-1.0060398.json
JSON-LD: 24-1.0060398-ld.json
RDF/XML (Pretty): 24-1.0060398-rdf.xml
RDF/JSON: 24-1.0060398-rdf.json
Turtle: 24-1.0060398-turtle.txt
N-Triples: 24-1.0060398-rdf-ntriples.txt
Original Record: 24-1.0060398-source.json
Full Text
24-1.0060398-fulltext.txt
Citation
24-1.0060398.ris

Full Text

SYNTHESIS AND EVALUATION OF PROBES OF RNA POLYMERASE II: ADAPTATION OF THE BICYCLIC SCAFFOLD OF THE OCTAPEPTIDE AMANITIN  by DAVID JOHN DIETRICH B.Sc. (Honours), McGill University, 2001  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2010  © David John Dietrich, 2010  ABSTRACT This thesis covers the synthesis and evaluation of bicyclic octapeptides as intracellular probes of RNAP polymerase II. In Chapter 2, the synthetic methodology used to achieve the rigid bicyclic octapeptide and the introduction of modifications to gain spatio-temporal control and visualization is presented. This includes the synthesis of a nitrorveratryl protected derivative of hydroxyproline, and a diethylaminocoumarin labeled asparagine residue. These amino acids were incorporated into amatoxins through solid-phase peptide synthesis. The synthesis and properties of these amatoxins, including their photolability and fluorescent properties is discussed. Chapter 3 describes the initial evaluation of the amatoxin probes in various eukaryotic cell lines. The cytotoxicity of α-amanitin in a variety of cell lines was investigated, where Chinese hamster ovary cells proved to be most sensitive to the toxin. These cytotoxic effects were observed at ~1 µM, which is much higher than the reported binding constant (~ 3 nM). The cytotoxicity was slow, requiring 60-72 hours to achieve 100% cell death. This diminished activity was attributed to cell uptake. Cell permeabilizing agents digitonin and saponin were applied to improve αamanitin uptake and toxicity. All synthetic amatoxins were shown to have minimal cytotoxic effects. Confocal microscopy demonstrated no cell uptake of a fluorescent amatoxin unless the cells were treated with detergent, pointing to critical issues of cell permeability. The optimization of amatoxin synthesis is presented in Chapter 4. It is shown that up to 20% water can inhibit epimerization during macrolactamization of amatoxins. A convergent synthetic approach to the bicyclic octapeptide was also developed. This allowed for the incorporation of a variety of amino acids at position three. This new synthetic approach was applied to the synthesis of tryptathionine-containing analogs of the opioid receptor agonist enkephalin. The attempted synthesis of the unnatural amino acid (2S),(3R),(4R)-dihydroxyisoleucine is described in Chapter 5.  This was achieved through the use of a diastereoselective [3,3]  sigmatropic rearrangement followed by Sharpless asymmetric dihydroxylation (AD). Surprising selectivity was noted during the AD reaction using phthalazine-based ligands, but this selectivity was reversed using pyrimidine-based ligands. ii  TABLE OF CONTENTS Abstract.......................................................................................................................................... ii Table of contents .......................................................................................................................... iii List of tables................................................................................................................................... x List of figures.............................................................................................................................. xiii List of abbreviations and symbols .......................................................................................... xxiv Acknowledgements .................................................................................................................. xxix Co-authorship statement ......................................................................................................... xxxi Chapter 1: Introduction ............................................................................................................... 1 1.1 What is amanitin? ................................................................................................................. 1 1.2 The amatoxin producing mushrooms ................................................................................... 2 1.2.1 Structural variants of the cyclic peptides ....................................................................... 2 1.2.2 Physiological effects of amatoxin poisoning ................................................................. 5 1.3 Background information on transcription............................................................................. 6 1.3.1 Central dogma of molecular biology ............................................................................. 6 1.3.2 Different forms of RNA polymerase ............................................................................. 7 1.3.3 Biochemical mechanism of transcription....................................................................... 9 1.3.4 Molecular mechanism of transcription ........................................................................ 10 1.3.5 Structure of RNA polymerase II .................................................................................. 12 1.4 The naturally occurring amatoxins ..................................................................................... 15 1.4.1 Biosynthesis of amatoxins and phallotoxins................................................................ 16 1.4.2 Chemical modification of natural amatoxins............................................................... 17 1.4.2.1 Modification of the bicyclic ring structure ........................................................... 17 1.4.2.2 Modification of the sulfoxide................................................................................ 18 1.4.2.3 Modification of asparagine ................................................................................... 18 1.4.2.4 Modification of dihydroxyisoleucine.................................................................... 19 1.4.2.5 Modification of hydroxytryptophan...................................................................... 20 1.5 Chemical synthesis of amatoxins........................................................................................ 22 1.5.1 Tryptathionine formation ............................................................................................. 23 1.5.2 The Savige-Fontana reaction ....................................................................................... 27 1.5.3 Preparation of Hpi for the Savige-Fontana reaction .................................................... 29 1.5.4 Linear synthesis of amatoxins using Hpi ..................................................................... 32 iii  1.5.4.1 Effect of hydroxyproline and sulfoxide revisited ................................................. 33 1.5.4.2 Derivatives of the dihydroxyisoleucine residue.................................................... 34 1.5.4.3 Other backbone mutations .................................................................................... 36 1.6 Project goals........................................................................................................................ 37 Chapter 2: Synthesis of photolabile and fluorescent amatoxin probes.................................. 40 2.1 Introduction......................................................................................................................... 40 2.1.1 Light-induced spatio-temporal control in biology ....................................................... 41 2.1.1.1 Photolabile protecting groups ............................................................................... 42 2.1.1.2 Mechanism of photolysis ...................................................................................... 43 2.1.1.3 Control of transcription with photoactivation....................................................... 44 2.1.2 Fluorescent visualiztion of biological targets .............................................................. 46 2.1.2.1 Fluorophores in biology........................................................................................ 46 2.1.2.2 Fluorescent visualization of transcription............................................................. 48 2.1.3 Goals of this chapter .................................................................................................... 49 2.2 Preparation of amino acids for incorporation into probes .................................................. 51 2.2.1 Synthesis of hydroxyproline derivatives...................................................................... 52 2.2.1.1 Nitrobenzylation attempts using Williamson ether synthesis............................... 53 2.2.1.2 Preparation of nitroveratryl bromide .................................................................... 54 2.2.1.3 Alternative methods of nitrobenzylation .............................................................. 54 2.2.1.4 Protecting group manipulation.............................................................................. 57 2.2.2 Synthesis of fluorescent asparagine derivative ............................................................ 58 2.2.2.1 Diethylaminocoumarin synthesis.......................................................................... 58 2.2.2.2 preparation of a linker bearing a fluorophore ....................................................... 58 2.2.2.3 Coupling to asparagine and deprotection.............................................................. 59 2.3 Solid phase amatoxin synthesis .......................................................................................... 60 2.3.1 Linear solid phase peptide synthesis............................................................................ 61 2.3.2 Tryptathionine formation ............................................................................................. 66 2.3.2.1 Synthesis of Hpi containing dipeptide .................................................................. 66 2.3.2.2 Savige-Fontana cyclization of peptides ................................................................ 68 2.3.3 Macrolactamization...................................................................................................... 71 2.3.3.1 Epimerization........................................................................................................ 74 2.4 Properties of synthetic peptides .......................................................................................... 76 2.4.1 Absorption and emission properties............................................................................. 76 iv  2.4.2 Circular dichroism spectroscopy.................................................................................. 78 2.4.3 Deprotection of nitrobenzyl and nitroveratryl amatoxins............................................ 80 2.5 Conclusions......................................................................................................................... 82 2.6 Experimental section........................................................................................................... 84 2.6.1 General methods .......................................................................................................... 84 2.6.2 HPLC purification methods ......................................................................................... 84 2.6.3 Synthetic protocols....................................................................................................... 86 2.6.4 Peptide synthesis.......................................................................................................... 99 2.6.4.1 Resin preparation ................................................................................................ 100 2.6.4.2 Amino acid deprotection and coupling ............................................................... 100 2.6.4.3 Savige-Fontana reaction...................................................................................... 101 2.6.4.4 Peptide macrolactamization ................................................................................ 101 2.6.5 Peptide characterization ............................................................................................. 101 2.6.5.1 Monocyclic octapeptides .................................................................................... 101 2.6.5.1 Bicyclic octapeptides .......................................................................................... 102 Chapter 3: Biological evaluation of amatoxin probes ........................................................... 104 3.1 Introduction....................................................................................................................... 104 3.1.1 RNA Polymerase II Activity Assays ......................................................................... 104 3.1.1.1 Conditions for assaying mRNA production........................................................ 105 3.1.1.2 Quantification of RNA production ..................................................................... 106 3.1.1.3 Cell viability assays ............................................................................................ 108 3.1.1.4 Literature precedent ............................................................................................ 109 3.1.2 Cell membrane permeability of amatoxins ................................................................ 110 3.1.2.1 Previous amatoxin uptake studies....................................................................... 111 3.1.2.2 Methods of cell membrane permeabilization...................................................... 111 3.1.3 Fluorescent imaging of the cell.................................................................................. 113 3.1.3.1 Cell imaging techniques...................................................................................... 114 3.1.3.2 Literature precedent ............................................................................................ 116 3.1.4 Chapter goals ............................................................................................................. 117 3.2 Cell viability assays .......................................................................................................... 117 3.2.1 General cell culture protocols .................................................................................... 118 3.2.2 α-Amanitin cytotoxicity controls............................................................................... 119 3.2.2.1 Cell line sensitivitiy to α-amanitin ..................................................................... 120 v  3.2.2.2 Time dependence ................................................................................................ 122 3.2.3 Cell viability assays using modified amatoxins......................................................... 123 3.2.4 Conclusions................................................................................................................ 125 3.3 Improved amatoxin cell uptake......................................................................................... 126 3.3.1 Cell viability in the presence of saponin and digitonin.............................................. 127 3.3.2 Saponin and digitonin mediated α-amanitin uptake .................................................. 129 3.3.3 Conclusions................................................................................................................ 132 3.4 Application of fluorescent amatoxin in confocal microscopy .......................................... 133 3.4.1 Fixed cell uptake ........................................................................................................ 133 3.4.2 Live cell uptake.......................................................................................................... 135 3.4.3 Saponin mediated uptake ........................................................................................... 136 3.5 Conclusions....................................................................................................................... 138 3.6 Experimental section......................................................................................................... 140 3.6.1 Materials .................................................................................................................... 140 3.6.2 General cell culture .................................................................................................... 140 3.6.3 Cell viability assays ................................................................................................... 141 3.6.3.1 Specifications of each viability assay ................................................................. 142 3.6.4 Confocal microscopy ................................................................................................. 143 Chapter 4: Improved methodology for the preparation of amatoxins ............................... 145 4.1 Introduction....................................................................................................................... 145 4.1.1 Chapter goals ............................................................................................................. 145 4.2 Elimination of epimerization during macrolactamization ................................................ 146 4.2.1 Effect of coupling reagent.......................................................................................... 150 4.2.2 Effect of solvent......................................................................................................... 151 4.2.3 The effect of wet DMF as solvent.............................................................................. 153 4.2.4 Conclusions................................................................................................................ 155 4.3 Convergent synthesis of amatoxins varying at position three .......................................... 155 4.3.1 Methods of introduction of position three amino acid............................................... 157 4.3.2 Monocyclic heptapeptide precursor synthesis ........................................................... 161 4.3.3 Preparation of activated amino acids ......................................................................... 164 4.3.4 Addition of the activated esters to a heptapeptide ..................................................... 165 4.3.4.1 Optimization of acylation conditions.................................................................. 166 4.3.4.2 Initial macrolactamization attempts.................................................................... 169 vi  4.3.4.3 Improved macrolactamization conditions........................................................... 170 4.3.5 Cytotoxicity assay of amatoxins varying at position three ........................................ 174 4.3.6 Conclusions................................................................................................................ 175 4.4 Synthesis of enkephalin analogs bearing a tryptathionine link......................................... 175 4.4.1 Enkephalins as potential targets................................................................................. 177 4.4.2 Tryptathionine enkephalin synthesis.......................................................................... 180 4.4.3 Conclusions................................................................................................................ 184 4.5 Experimental section......................................................................................................... 185 4.5.1 General methods ........................................................................................................ 185 4.5.2 HPLC purification methods ....................................................................................... 185 4.5.3 Synthetic protocols..................................................................................................... 186 4.5.4 Peptide synthesis........................................................................................................ 190 4.5.4.1 Resin preparation ................................................................................................ 191 4.5.4.2 Amino acid deprotection and coupling ............................................................... 191 4.5.4.3 Savige-Fontana reaction...................................................................................... 192 4.5.4.4 Peptide macrolactamization ................................................................................ 192 4.5.4.5 Study of epimerization during macrolactamization............................................ 192 4.5.4.6 Solution-phase peptide acylation ........................................................................ 193 4.5.4.7 Enkephalin synthesis........................................................................................... 194 4.5.5 Peptide characterization ............................................................................................. 194 4.5.5.1 Heptapeptide precursors for dhIle3 mini-library:................................................ 194 4.5.5.2 Monocyclic octapeptides for dhIle3 mini-library:............................................... 195 4.5.5.3 Bicyclic octapeptide dhIle3 mini-library:............................................................ 196 4.5.5.4 Tryptathionine enkephalin peptides:................................................................... 197 4.5.6 Cell viability assay..................................................................................................... 198 Chapter 5: Studies towards the synthesis of (2S),(3R),(4R)-dihydroxyisoleucine............... 199 5.1 Introduction....................................................................................................................... 199 5.1.1 Previous syntheses of dihydroxyisoleucine ............................................................... 203 5.1.1.1 Non-selective synthesis of Georgi and Wieland................................................ 203 5.1.1.2 Stereoselective synthesis by Bartlett and Barstow.............................................. 204 5.1.2 Structurally related analogs of dihydroxyisoleucine.................................................. 209 5.1.2.1 Synthesis of γ-hydroxyvaline.............................................................................. 210 5.1.2.2 Synthesis of γ,δ-dihydroxynorvaline .................................................................. 211 vii  5.1.2.3 Synthesis of γ,δ-dihydroxyleucine...................................................................... 212 5.1.3 Proposed synthesis of γ,δ−dihydroxyisoleucine ........................................................ 214 5.1.3.1 Parallel kinetic resolution ................................................................................... 215 5.1.4 Chapter goals ............................................................................................................. 218 5.2 Sigmatropic rearrangement to generate alkene substrate ................................................. 218 5.3 Asymmetric dihydroxylation ............................................................................................ 222 5.3.1 Analysis of diastereomers fomred during AD reaction ............................................. 223 5.3.1.1 Absolute configuration of diastereomer.............................................................. 227 5.3.1.2 Modified asymmetric dihydroxylation conditions.............................................. 228 5.3.1.3 Effect of amine protecting group during AD reaction ........................................ 230 5.3.1.4 Proposed mechanism of AD selectivity.............................................................. 232 5.3.1.5 Effect of alternative chiral ligand on AD selectivity .......................................... 235 5.3.2 Analysis of enantiomers formed during AD reaction ................................................ 235 5.4 Protecting group strategies for the diol............................................................................. 240 5.4.1 Protection of the diol with trityl groups..................................................................... 240 5.4.2 Protection of the diol as a thiocarbonate.................................................................... 241 5.5 Conclusions....................................................................................................................... 244 5.6 Experimental section......................................................................................................... 245 5.6.1 General methods ........................................................................................................ 245 5.6.2 HPLC and CE methods.............................................................................................. 245 5.6.3 Synthetic protocols..................................................................................................... 246 References.................................................................................................................................. 256 Appendices................................................................................................................................. 282 Appendix A: Chapter 2 supplementary information............................................................... 282 A.1: HPLC chromatograms ................................................................................................ 282 A.2: NMR spectra ............................................................................................................... 290 Appendix B: Chapter 3 supplementary information............................................................... 309 B.1: Cell viability assay raw data ....................................................................................... 309 Appendix C: Chapter 4 supplementary information............................................................... 328 C.1: HPLC chromatograms................................................................................................. 328 C.2: NMR spectra ............................................................................................................... 346 C.3: Cell viability assay raw data ....................................................................................... 359 Appendix D: Chapter 5 supplementary information............................................................... 360 viii  D.1: NMR spectra ............................................................................................................... 360 D.2: HPLC and CE chromatograms.................................................................................... 374 D.3: Crystal structure data .................................................................................................. 378  ix  LIST OF TABLES Table 1.1: List of some fungal species shown to contain the toxin α−amanitin, and the average amount isolated from each, in mg/g dry weight of mushroom.12-18 ........................................ 2 Table 1.2: Nine different amatoxins isolated from Amanita mushrooms. The Ki values were determined against calf thymus RNAP II and the LD50 values listed are as observed in white mice.2 ...................................................................................................................................... 4 Table 1.3: Lethal dose of α−amanitin in various organisms, and the mode of administration. .... 5 Table 1.4: Comparison of RNA polymerase constructs from different origins, their respective products and relative sensitivity to α-amanitin....................................................................... 7 Table 1.5: Inhibitory effect of α−amanitin on eukaryotic RNA polymerases II and III. ............... 8 Table 1.6: Sequence homology of the bridge helix region of E. coli RNA polymerase and RNAP II and RNAP III of S. cervisisiae. Bold residues are conserved, while the yellow residues are structurally related amino acids. ....................................................................................... 8 Table 1.7: Genetic analysis of A. bisporigera identified two genes that are part of the biosynthetic pathway leading to amatoxins and phallotoxins. The italicized residues are conserved in both genes, and the bold residues relate to the toxin. The highlighted residues correspond to the Trp and Cys involved in the tryptathionine crosslink. ............................. 16 Table 1.8: Effect of the oxidation of the thioether of the tryptathionine of amatoxins on their toxicity in white mice............................................................................................................ 18 Table 1.9: Effect of modification at the β-position of asparagine. Reported as LD50 in values of mg/kg in the white mouse. .................................................................................................... 19 Table 1.10: The toxicity and inhibitory capacity of a variety of modifications of the hydroxyl group of hTrp4 of α-amanitin................................................................................................ 21 Table 1.11: Synthetic proline and hydroxyproline variants of amatoxins. The Ki values reported are relative to α-amanitin, which were obtained against RNAP II from Calf Thymus (αamanitin Ki ~ 3 nM).............................................................................................................. 33 Table 1.12: Amatoxin analogs that have been prepared to test the significance of dihydroxyisoleucine. The derivatives were compared to α-amanitin against calf thymus RNAP II in vitro, or else to γ-amanitin, tested in vitro against RNAP II isolated from D. melanogaster embryos. ......................................................................................................... 35  x  Table 1.13: Effect of changing various backbone amino acids in the amatoxin structure. Values reported are relative to the effect of α-amanitin on calf thymus or D. melanogaster (values in talics) RNAP II. (n.o. = no inhibition observed) .............................................................. 36 Table 1.14: Previously described probes of transcription and their specific target. ..................... 38 Table 2.1: Common lasers used in confocal microscopy, and their respective emission wavelengths........................................................................................................................... 47 Table 2.2: Results of attempted alkylation of protected hydroxyproline derivative 105, following typical benzylation conditions. ............................................................................................. 53 Table 2.3: Nitrobenzylation of hydroxyproline 105 using silver mediated conditions. ............... 55 Table 2.4: Results of nitrobenzylation reaction used to produce 110b under phase transfer conditions (also see Figure 2.13). ......................................................................................... 56 Table 2.5: Linear hexapeptide precursors of amatoxins. Synthetically prepared amino acid monomers are shown in bold type. ....................................................................................... 66 Table 2.6: Mass spectrometry analysis of the monocyclic peptides obtained after the SavigeFontana reaction. Unreported values were determined in our lab by Dr. Jonathan May. .. 71 Table 2.7: Mass spectrometry analysis of the amatoxin analogs obtained after macrolactamization. .............................................................................................................. 74 Table 2.8: Photolysis rate constant and half-life of amatoxins bearing a light-sensitive protection group. These were determined from the HPLC chromatograms in Figure 2.38.................. 82 Table 3.1: Use of α-amanitin to distinguish the activity of RNA polymerases in sea urchin embryo nuclei, as reported in Hames et al.181 ..................................................................... 106 Table 3.2: Cell viability measurements of a variety of cell lines exposed to various doses of αamanitin. The concentrations required to achieve death in 25% of the cells (ED25) was determined using trypan blue.24 .......................................................................................... 110 Table 3.3: Cell lines tested for relative α-amanitin cytotoxicity. .............................................. 121 Table 3.4: Relative survival of various cell lines treated with α-amanitin for 66 hours. Percent survival reflects the remaining viable cells following 66-hour treatment at the highest concentration (25 µM). The ED50 refers to the concentration required to achieve 50% cell death relative to the amount remaining following the exposure period.............................. 122 Table 3.5: Results of saponin mediated amanitin-induced cytotoxicity. .................................... 131 Table 4.1: Summary of effect of various coupling reagents on epimerization at Ile3 during macrolactamization of 122. (N.D. = not determined) (For structures see Figure 4.3)........ 151 xi  Table 4.2: Summary of the effect of solvent on epimerization at Ile3 during macrolactamization of 122. (N.D. = not determined) (For structures see Figure 4.3) ........................................ 153 Table 4.3: Summary of effect of added water in DMF on epimerization at Ile3 during macrolactamization of 122. (N.D. = not determined) (For structures see Figure 4.3)........ 154 Table 4.4: Mass spectrometric analysis of monocyclic heptapeptides bearing a tryptathionine.163 Table 4.5: Conditions tested for the acylation of the N-terminus of heptapeptide 162. [Solvent system A: dioxane-water (4:5), B: dioxane-water (3:1).] ................................................... 166 Table 4.6: Observed mass spectra of the various monocyclic octapeptides obtained following acylation of 168................................................................................................................... 168 Table 4.7: Observed mass spectra of the various monocyclic octapeptides obtained following acylation of 162 and removal of the N-terminal Fmoc protecting group. .......................... 171 Table 4.8: Conditions tested for the attempted macrolactamization of 177. .............................. 172 Table 4.9: Isolated amount of bicyclic octapeptides containing different amino acid residues at position three....................................................................................................................... 173 Table 4.10: The observed binding constant for various opioid agonists in the three most common opioid receptors (κ, µ, δ) can show selectivity for the δ receptor. The ED50 reflects the effectiveness in vivo.288 289 .................................................................................................. 179 Table 4.11: Mass spectrometric analysis of the four isolated tetrapeptide enkephalin precursors. ............................................................................................................................................. 182 Table 4.12: Mass spectrometric analysis of the t-butyl protected tryptathionine enkephalins.. 184 Table 5.1: Expected and observed diastereoselectivity of the dihydroxylation reaction of D/L-239. ............................................................................................................................................. 227 Table 5.2: Summary of the results obtained following dihydroxylation of racemic olefin substrates 239 and 255. ....................................................................................................... 231 Table 5.3: Observed selectivity of the dihydroxylation reaction of 260 under various reaction conditions, reported by Gardiner and Bruce.333 .................................................................. 234 Table 5.4: Summary of the various results obtained following dihydroxylation of olefins 240 and 255 using Upjohn conditions in the presence of different chiral ligands. The d.r. were determined from the HPLC, and the e.r. were determined from the CE data..................... 239 Table 5.5: Estimated formation of the diol (241) and the aldehyde byproduct (271) obtained following base-promoted hydrolysis of thiocarbonate 268. ............................................... 243  xii  LIST OF FIGURES Figure 1.1: The structure and amino acid numbering of the natural product α-amanitin. The structure on the right shows an abbreviation that will be used in this thesis to represent the amanitin scaffold..................................................................................................................... 1 Figure 1.2: Related cyclic heptapeptides isolated from Amanita species. A. Phallotoxins B. Virotoxins ............................................................................................................................... 3 Figure 1.3: The central dogma of molecular biology. The bold arrows imply the common genetic transfer elements, while the dashed lines are unique pathways found in certain organisms. ............................................................................................................................... 6 Figure 1.4: Cartoon depicting the process of transcription. A. Transcription factors are recruited to the TATA-box promoter site to form the initiation complex. B. Transcription bubble is formed, and the first few NTPs are incorporated to produce new RNA (red). C. Escape from initiation factors, and the recruitment of elongation factors renders the RNAP II complex into a fully elongating form.................................................................................... 10 Figure 1.5: Depiction of the transcription bubble (reproduced from PDB 2E2H). Newly synthesized RNA (red) forms nine base pairs with the DNA template strand (blue). The green non-coding strand of DNA melts approximately six base pairs from the active site, and rejoins the template stand past the transcription bubble (not shown). The next nucleotide to be incorporated is shown as a space-filling model in the active site. ............. 11 Figure 1.6: Two views of the active site of RNAP II (reproduced from PDB 2E2H). A. General depiction of the active site, based on crystal structure data and genetic analysis. B. Reproduction of the critical residues in the active site as viewed in recent crystal structures. These residues are shown in light blue, the 3’ end of the RNA is shown in pink. ............... 12 Figure 1.7: Overall structure of RNAP II in complex with substrate and products. A. Cartoon diagram showing side-view of the polymerase, including the transcription bubble (from Kornberg et al.).52 B. Top-down view, reproduced from the reported crystal structure (reproduced from PDB 2E2H). In both figures the new RNA is shown in red, the template DNA strand in blue, and the non-coding strand in green. .................................................... 13 Figure 1.8: Close-up view of the active site of RNAP II. A. The active site when co-crystallized with the natural GTP substrate, shown in blue. B. The active site when co-crystallized with  xiii  the inhibitor α-amanitin, shown in red. The orange depicts the Mg2+-binding domain, green shows the bridge helix domain, and the trigger loop is shown in magenta. ......................... 14 Figure 1.9: Close-up diagram of α-amanitin in the binding pocket of RNAP II (reproduced from PDB 1K83). The residues of RNAP II proposed to make critical contacts with the toxin are shown in light blue................................................................................................................ 15 Figure 1.10: Selective cleavage of one cycle of α−amanitin. A. Mild acid hydrolysis of the amide bond between dhIle3 and hTrp4. B. Raney nickel reduction of the tryptathionine crosslink. ............................................................................................................................... 17 Figure 1.11: Modification of the dhIle3 residue of α−amanitin. When 1 was treated with periodate, aldehyde 21 was generated. This was converted to alcohol 22 or hydrazone 23. The values below each derivative represent the observed Ki assayed against wheat germ RNAP II. ............................................................................................................................... 20 Figure 1.12: Introduction of chemical and biological reporters to amatoxins through the hydroxyl group of hTrp4. ...................................................................................................... 21 Figure 1.13: Modified amatoxins prepared via modification of the hydroxytryptophan residue. A. The hydroxyl group was removed by catalytic reduction. B. Iodination at the 7’position. C. Dialkyl amination at the 7’-position. D. Diazotization at the 7’-position........ 22 Figure 1.14: Wieland’s synthesis of Gly1-norphalloin using a sulfenyl chloride to generate the tryptathionine crosslink......................................................................................................... 24 Figure 1.15: Guy’s solid-phase synthetic approach to synthesize Ala7-phalloidin using an orthogonally protected tryptathionine dipeptide................................................................... 25 Figure 1.16: Lokey’s approach to synthesize Glu7-phalloidin, using iodine to generate the typtathionine crosslink. ......................................................................................................... 26 Figure 1.17: Savige’s original findings leading to the Savige -Fontana reaction........................ 27 Figure 1.18: Proposed mechanism of the Savige-Fontana reaction.76 .......................................... 28 Figure 1.19: Zanotti and Wieland’s application of the Savige-Fontana reaction to generate Sdeoxo-Ile3-amaninamide (53). .............................................................................................. 29 Figure 1.20: Natural products containing an Hpi moiety. A. Brevianamide E B. Phakellistatin 3 C. Gypsetin. .......................................................................................................................... 30 Figure 1.21: Various methods reported for the preparation of useful Hpi surrogates. ................ 31 Figure 1.22: General linear amino acid sequence that is required to generate amatoxins. In solution phase approaches, R represents a t-butyl protecting group, whereas in SPPS, this represents the resin................................................................................................................ 32 xiv  Figure 1.23: Elaboration of linear peptides containing Hpi into amatoxins. ............................... 33 Figure 1.24: Small molecules that have been used to inhibit transcription. A. Rifampicin. B. Actinomycin D. C. Dichlororibofuranosyl benzimidazole (DRB). ...................................... 38 Figure 2.1: Photoactivatible ATP derivative used for the assay of Na/K ATPase. ...................... 41 Figure 2.2: Common photolabile protecting groups used to control biological processes. In each case, the leaving group (LG) is released at the wavelength shown below. .......................... 42 Figure 2.3: Photolabile protecting groups based on the ortho-nitrobenzyl substitution pattern... 42 Figure 2.4: Products observed upon 254 nm photolysis of methyl nitrobenzyl ether, monitored by laser flash photolysis by Illitch et al.23............................................................................. 43 Figure 2.5: Light-sensitive ecdysone derivatives that have been used for photolytic control of gene expression..................................................................................................................... 44 Figure 2.6: Photocaged tamoxifen was shown to inhibit production of a specific mRNA product, upon exposure to 345 nm light.............................................................................................. 45 Figure 2.7: Fluorophores commonly employed in cell biology and their absorption and emission wavelengths: A. AlexaFluor 350, B. Fluorescein C. Tetramethyl rhodamine. (R= biological probe) .................................................................................................................................... 47 Figure 2.8: A. Fluorescent amanitin analog 28. B. Fluorescent image representing cellular localization of 28 within PtK1 cells, as reported by Wulf et al.44 ........................................ 48 Figure 2.9: Target molecules to be prepared in this chapter. A. Photolabile derivatives of amanitin. B. Fluorescently labeled amatoxin....................................................................... 50 Figure 2.10: Amino acids required for the synthesis of amatoxin probes. A. Hydroxyproline bearing a light-sensitive group (red) on the γ-OH group. B. Asparagine bearing a triethylene glycol (TEG)-linked fluorescent label (blue) on the β-carboxamide. ................ 51 Figure 2.11: Preparation of Boc protected hydroxyproline methyl ester 105............................... 53 Figure 2.12: Preparation of nitroveratryl bromide using adapted literature protocols. ................ 54 Figure 2.13: General scheme depicting alkylation of hydroxyproline derivative 105 with nitroveratryl bromide. ........................................................................................................... 55 Figure 2.14: Trichloroacetimidate protection of serine reported by the Lawrence lab. .............. 56 Figure 2.15: Protecting group manipulation of light-sensitive hydroyproline derivatives 110a/b to generate an SPPS compatible monomer. .............................................................................. 57 Figure 2.16: Use of Knoevenagel condensation followed by lactonization and saponification to yield diethylaminocoumarin (116)........................................................................................ 58  xv  Figure 2.17: Selective protection and coupling of a triethylene glycol (TEG) linker to the fluorophore diethylaminocoumarin (DEAC)........................................................................ 59 Figure 2.18: Synthesis and coupling of DEAC to an activated aspartate residue to yield the SPPS compatible amino acid 103. .................................................................................................. 60 Figure 2.19: Retrosynthetic approach used to target the bicyclic amanitin scaffold containing natural and modified amino acids. ........................................................................................ 61 Figure 2.20: Amanitin derivatives prepared as controls, containing no modification at asparagine residue one, and either a proline or hydroxyproline at position two. ................................... 62 Figure 2.21: Charging of chlorotrityl resin with the C-terminal amino acid Fmoc-Ile-OH (the black sphere represents the polystyrene resin)...................................................................... 63 Figure 2.22: Quantitation of the loading of the chlorotrityl resin is determined by monitoring dibenzofulvene (129) production after DBU treatment (the black sphere represents the entire resin, including the 2-chlorotrityl functionality)......................................................... 63 Figure 2.23: Modified laboratory apparatus used for manual solid-phase synthesis of peptides. A. Spin column, sealed with a plastic pipet tip. B. Spin column shaking on a vortexer. C. Filtration apparatus. .............................................................................................................. 64 Figure 2.24: A. Generic deprotection and coupling cycle that was used to generate linear peptides using SPPS. The black sphere represents the solid-phase. B. Structure of the coupling reagent HBTU (130). The specific substitutions of R1 and R2 are in Figure 2.29................ 65 Figure 2.25: The SPPS compatible monomer used for the introduction of Hpi into peptides for the generation of amatoxins. ................................................................................................. 67 Figure 2.26: Synthesis of a diastereomeric mixture of the desired Tr-Hpi-Gly-OMe dipeptide 58. ............................................................................................................................................... 67 Figure 2.27: Diastereomeric epoxides obtained during DMDO oxidation of 135 leads to the formation of the syn-cis and anti-cis isomers. ...................................................................... 68 Figure 2.28: The deprotected Hpi dipeptide 132 is suitable for introduction onto the solid-phase using similar coupling conditions as other modified amino acids........................................ 68 Figure 2.29: Savige-Fontana cyclization of the Hpi-containing octapeptides provides the desired monocyclic peptides.............................................................................................................. 69 Figure 2.30: Sample HPLC chromatogram of crude monocyclic peptide. Products containing a tryptathionine crosslink were identified by their unique UV absorbance spectrum (shown inset)...................................................................................................................................... 70  xvi  Figure 2.31: Macrolactamization of tryptathionine containing octapeptides produces the desired bicyclic amatoxin analogs. Shown inset is the structure of PyBOP. .................................. 72 Figure 2.32: A. Sample HPLC chromatogram obtained following macrolactamization reaction used to produce amatoxins.................................................................................................... 73 Figure 2.33: Possible isomeric products of 64. A. These isomers reflect an epimerization at the α-carbon of Ile3. B. These isomers cartoon the proposed basket atropisomers.................. 75 Figure 2.34: The absorption spectra of 100a (red) and 100b (blue) in methanol. ....................... 77 Figure 2.35: The absorption (blue) and emission (red) spectra of 101 in methanol..................... 78 Figure 2.36: This graph shows the unique CD spectra of the Ile3 epimeric amatoxins 64 and 138 obtained following macrolactamization of 122. Dr. Jonathan May obtained these spectra in previous work in our laboratory............................................................................................ 79 Figure 2.37: The CD spectra obtained with amatoxin probes 100a (shown in blue), and 100b (shown in red). ...................................................................................................................... 80 Figure 2.38: HPLC analysis of the photolysis of photoactivatible peptides. A. Photodeprotection profile following 366 nm exposure to 123a (150 µM in methanol) B. Photodeprotection profile following 254 nm exposure to 123b (67 µM in methanol). ...................................... 81 Figure 2.39: Gradient profiles used for HPLC analysis and purification of peptides.................. 85 Figure 3.1: Diagram depicting various conditions employed for the assay of mRNA production and the potential readout..................................................................................................... 105 Figure 3.2: Common dyes employed to detect the viability of cells A. Trypan blue is excluded from viable cells. B. MTT is reduced to the purple formazan. C. Resazurin is reduced to the purple resorufin. ............................................................................................................ 108 Figure 3.3: Structure of mild cell permeabilizing detergents. A. Lysolecithin B. Saponin C. Digitonin. ............................................................................................................................ 112 Figure 3.4: Approaches to fluorescent visualization of a desired target. A. Fluorescently labeled antibodies can recognize a protein of interest after cell permeabilization. B. Small molecules designed to bind to the target can be fluorescently labeled. C. Genetically modified cells produce target proteins that are covalently linked to a fluorescent protein. ....................... 114 Figure 3.5: A. Rhodamine labeled phalloidin. B. Confocal microscope image of a skin fibroblast cell imaged with rhodamine phalloidin (red). Image taken from Invitrogen catalog (Product #R415)................................................................................................................................. 115 Figure 3.6: Sample image of a 96-well plate that is obtained after incubation of cell lines in the presence of α-amanitin for 66 hours, followed by treatment with MTT............................ 120 xvii  Figure 3.7: Relative cytotoxicity of α-amanitin in various cell lines. ........................................ 121 Figure 3.8: Time dependent effect of 10 µM α-amanitin on Caco-2 cell viability. ................... 123 Figure 3.9: Amatoxin derivatives tested for initial toxicity studies in CHO cells..................... 124 Figure 3.10: Effect of proline derivatives 64 and 138 on CHO cell viability. (Structures shown in Figure 3.9)....................................................................................................................... 124 Figure 3.11: Treatment of CHO cells with modified amatoxins 100b and 53 (Structures shown in Figure 3.9)........................................................................................................................... 125 Figure 3.12: Effect of brief exposure to saponin on CHO cell viability..................................... 127 Figure 3.13: Effect of long-term exposure to low concentrations of saponin and digitonin on CHO and Cos-7 cell viability.............................................................................................. 128 Figure 3.14: CHO cells were treated with a 0.1 mg/mL saponin for a brief period of time, in the presence of α-amanitin to induce cell death. Cell viability was determined after 45-hours of growth in normal media. ..................................................................................................... 129 Figure 3.15: Effect of α-amanitin on CHO cell viability in the presence of detergents............. 130 Figure 3.16: Effect of α-amanitin on Cos-7 cell viability in the presence of detergents............ 130 Figure 3.17: Treatment of CHO cells with amatoxins in the presence of the mild detergent saponin. A. Viability of cells exposed to the toxin in medium containing 1.5 mg/mL saponin for three minutes. B. Viability of cells exposed to the toxin in the presence of 1.5 µg/mL saponin. (For structures, see Figure 3.9)................................................................ 132 Figure 3.18: Fluorescently labeled peptides used in confocal microscopy experiments........... 134 Figure 3.19: Confocal microscopic image of CHO cells treated with rhodamine-phalloidin (red) and DEAC-amatoxin 101 (blue). The cell membrane was permeabilized with Triton X-100 prior to staining. The white scale bar represents 85 µm. ................................................... 135 Figure 3.20: A. Bright field image of CHO cells, obtained with a 10X objective. B. Confocal microscopy image of CHO cells treated with 1 µM 101 in medium for a 24-period without membrane permeabilization. The white scale bar represents 85 µm. ................................. 136 Figure 3.21: Cellular uptake of DEAC-amatoxin in the presence of the mild detergent saponin. In both cases, cells were treated with 1 mg/mL saponin in PBS, containing 1 µM 101 for 5 min at 0 °C. The cells were then incubated with fresh medium containing 1 µM 101 at 37 °C for various times. A. Image obtained following a 1 h incubation period. B. Image obtained following a 24 h incubation period. In both images the white scale bar represents 85 µm. ................................................................................................................................. 137 xviii  Figure 4.1: Tautomerization and oxazolone formation are the generally accepted mechanisms that lead to scrambling of the configuration of the α-carbon. ............................................ 147 Figure 4.2: The structures of various coupling reagents and additives that are used in peptide synthesis.............................................................................................................................. 148 Figure 4.3: Macrolactamization control reaction carried out to test for conditions that reduce epimerization....................................................................................................................... 149 Figure 4.4: Section of the HPLC chromatograms (monitored at 292 nm) of the crude reaction following macrolactamization of 122 using various coupling agents. ............................... 150 Figure 4.5: Section of the HPLC chromatogram (monitored at 292 nm) obtained following macrolactamization of 122 in various solvents................................................................... 152 Figure 4.6: Section of the HPLC chromatogram (monitored at 292 nm) obtained following macrolactamization of 122 in DMF with added water. ...................................................... 154 Figure 4.7: Comparison of two possible synthetic approaches useful in the development of a position three library of amatoxins. In the Traditional Approach the modified amino acid is added to the resin in the first step. In the Convergent Approach, a common substrate is prepared, and the modified amino acid is added in the penultimate step. .......................... 156 Figure 4.8: Linear peptide synthesis in which an Hpi tripeptide was used to produce phakellistatin-like peptide 157............................................................................................ 157 Figure 4.9: Conversion of phakellistatin-like monocyclic octapeptide 157 into amatoxin 64 was achieved using TFA, based on previous work in our lab.................................................... 158 Figure 4.10: Preparation of the Fmoc-Ile-Hpi-Gly-OH monomer for use in solid-phase synthesis, previously described in our lab by Dr. May. ...................................................................... 159 Figure 4.11: In this retrosynthetic approach for the introduction of the position three amino acids, we would introduce an Hpi tripeptide (156) containing the variable position three amino acid. This could be made from dipeptide 158......................................................... 159 Figure 4.12: The tripeptide scaffold that would be required for the synthesis of a library of position three analogs of amatoxins following the protocol of May et al. ......................... 160 Figure 4.13: Proposed acylation of a generic heptapeptide with an N-hydroxysuccinimide (NHS) activated amino acid should provide a means to generate amatoxins. ............................... 160 Figure 4.14: Modified amino acids that were incorporated during the SPPS of the monocyclic heptapeptide amatoxin precursors....................................................................................... 162 Figure 4.15: Synthesis of heptapeptide precursors lacking a position three amino acid. .......... 163  xix  Figure 4.16: Various amino acids were activated as NHS esters, such that they could be coupled to the N-terminus of the heptapeptides 161-166 to generate a library of amatoxins that vary in their position three residue.............................................................................................. 165 Figure 4.17: HPLC analysis of acylation of 162 with activated threonine 168 (product is marked with an arrow). Shown inset is the UV absorption spectrum that was observed for the desired product.................................................................................................................... 167 Figure 4.18: HPLC analysis of deprotection of crude peptide 172 with NHEt2 followed by TFA shows that the desired product 177 is cleanly obtained following precipitation. ............... 169 Figure 4.19: Acylation and Fmoc-removal were carried out in one pot to generate octapeptides 177-181. Note the delayed retention times of 180 and 181 are due to the fact that they bear a side-chain trityl protecting group, as opposed to the t-butyl group of 177-179............... 170 Figure 4.20: HPLC analysis of macrolactamization reaction of 177 under various conditions. 173 Figure 4.21: CHO cell viability assay using amatoxins 53, 182, 183, 185 and 186 revealed that none of the derivatives induced cell death on a scale relative to α−amanitin (1)............... 174 Figure 4.22: Some examples of peptide cyclization methodologies that rely on side chain interactions, including: A. Disulfide bond. B. Lanthionine C. Alkene............................... 176 Figure 4.23: The structure of some opioid receptor agonists...................................................... 178 Figure 4.24: One of four proposed tryptathionine analogs of enkephlain, which were modeled after the cyclized lanthionine derivative 191. ..................................................................... 179 Figure 4.25: Synthetic approach that was used to prepare the tetrameric tryptathionineenkephalin precursors. ........................................................................................................ 181 Figure 4.26: HPLC analysis of the tetrapeptides 193a-d isolated following purification. ......... 181 Figure 4.27: Tetrapeptide precursors were treated with an activated ester of tyrosine, followed by removal of Fmoc to genreate trytptathionine enkephalins 194a-d................................. 183 Figure 4.28: HPLC analysis of pentapeptide enkephalins 194a-d.............................................. 183 Figure 5.1: Absolute stereochemistry of the dihydroxyisoleucine residue that is found in αamanitin, and the lactone that is produced upon total acid hydrolysis of 1. ....................... 199 Figure 5.2: Previously reported position three variants of α- and γ-amanitin (1 and 5), and their relative RNAP II inhibitory capabilities. (See Table 1.12 also) ........................................ 200 Figure 5.3: Problems associated with lactone formation of dhIle. A. Lactone formation of free dhIle can lead to epimerization of the α-carbon to generate the more stable trans-trans isomer. B. Alternatively, lactonization of a dhIle residue within a peptide can lead to strand scission................................................................................................................................ 202 xx  Figure 5.4: Non-selective synthesis of the lactone 202 reported by Georgi and Wieland.......... 203 Figure 5.5: Retrosynthetic analysis proposed by the Bartlett lab to achieve racemic dhIle lactone 202....................................................................................................................................... 204 Figure 5.6: Synthesis reported by Bartlett et al of intermediate 209 through a diastereoselective [3,3] sigmatropic rearrangement......................................................................................... 205 Figure 5.7: The chair-like transition state is proposed to lead to the diastereoselectivity of the ester-enolate Claisen-Ireland rearrangement. ..................................................................... 205 Figure 5.8: Proposed mechanism of iodolactonization. A. Formation of diastereomeric cyclic iodonium ions. B. Attack of carboxylate and deprotonation. Kinetic conditions are irreversible (green arrows), whereas thermodynamic conditions are (blue arrows), which leads to enrichment of the more stable cis-trans substituted isomer 214. .......................... 206 Figure 5.9: Bartlett’s successful generation of desired diastereomer 216 through iodolactonization of phthaloyl protected substrate 215. ..................................................... 207 Figure 5.10: Results of attempted hydrolysis of the iodolactone D/L-216, observed by Bartlett et al. ........................................................................................................................................ 208 Figure 5.11: The lactone of dihydroxyisoleucine (202) was obtained by displacement of the δ− iodo group followed by total acid hydrolysis of 216. ......................................................... 208 Figure 5.12: Structurally related analogs of dihydroxyisoleucine that have been successfully adapted to SPPS. ................................................................................................................. 210 Figure 5.13: Synthesis of 222 reported by Cudic, Marí and Fields.309 ...................................... 211 Figure 5.14: Synthesis of 223 reported by Spetzler and Hoeg-Jensen.307 .................................. 212 Figure 5.15: Synthesis of 224 reported by Edagwa and Taylor.................................................. 213 Figure 5.16: Proposed retrosynthetic approach to be employed in our synthesis of an SPPS compatible monomer of dhIle............................................................................................. 214 Figure 5.17: Schematic representation of kinetic resolution and parallel kinetic resolution. Reaction of a racemic substrate in the presence of a chiral additive (X*), can lead to an enrichment of a single product (Product A in kinetic resolution), or else to a mixture of nonenantiomeric products (Products A and B in parallel kinetic resolution). The expected major products are shown in red.316 .............................................................................................. 215 Figure 5.18: Examples of the different classes of parallel kinetic resolution. A. A chemodivergent process.320 B. A regiodivergent process.321 C. A stereodivergent process.322 ............................................................................................................................................. 216  xxi  Figure 5.19: Proposed stereoselective AD of each of the enantiomers of D/L-239. Effective parallel kinetic resolution (selective attack at the re face for both enantiomers) would generate L-241 and D-242 as the major products. ............................................................... 217 Figure 5.20: Synthesis of alkene precursor 238......................................................................... 219 Figure 5.21: Sigmatropic rearrangement of 238 produced the desired diastereomer 246 as the major product. ..................................................................................................................... 219 Figure 5.22: Diastereoselectivity obtained following the [3,3] sigmatropic rearrangement of 238 was determined by 1H-NMR analysis in D2O. The Cbz-group was first removed by hydrogenolysis. The arrows point to the α-proton of the expected products. ................... 220 Figure 5.23: Overall reaction that was used to generate the indoline amides D/L-239. The desired diastereomer was isolated by chromatography. .................................................................. 221 Figure 5.24: Solid-state molecular structure of D-239. The unit cell showed a racemic mixture of both the D- and L-isomers. .................................................................................................. 221 Figure 5.25: Desired parallel kinetic resolution of olefin D/L-239 using Sharpless asymmetric dihydroxylation should yield a diastereomeric mixture of products. ................................. 222 Figure 5.26: Chiral ligands supplied with AD-mix-β (253) and AD-mix-α (254).................... 223 Figure 5.27: Four possible products that can be formed followed dihydroxylation of racemic olefin 239. The table shows the theoretical yields of each stereoisomer under three different conditions, and the resultant diastereomeric ratio (d.r.) and enantiomeric ratio (e.r.)........ 224 Figure 5.28: 1H-NMR spectrum of the product obtained following AD reaction of 232 using ADmix-β. The spectrum shows the formation of mainly one diastereomeric product. .......... 225 Figure 5.29: The 1H-NMR analysis of the dihydroxylation reaction of D/L-239 using AD-mix-α or β showed unexpected diastereoselectivity...................................................................... 226 Figure 5.30: Solid-state molecular structure of D-242. The unit cell was a racemic mixture of both D- and L-242. ............................................................................................................... 228 Figure 5.31: Crude 1H-NMR analysis of products obtained following oxidation of D/L-239 using Upjohn conditions wherein chiral ligands 253 or 254 were added. Note that formation of diastereomer 242 is still favoured, regardless of the ligand used. ...................................... 229 Figure 5.32: Synthesis of Boc-protected olefin substrate D/L-255. ............................................ 230 Figure 5.33: The 1H-NMR spectrum obtained following Upjohn dihydroxylation of 255 in the presence of chiral ligand (DHQD)2Phal. The signature β-CH3 peaks (arrows) at δ=0.9 and 1.1 ppm show a 2.7:1 selectivity for diol D/L-257. ............................................................. 231 xxii  Figure 5.34: A. Corey’s proposed binding of allyl benzoate within the (DHQD)2Phal binding pocket. B. Sharpless’ proposed binding of styrene in the (DHQD)2Phal binding pocket. The features that lead to substrate binding are shown in blue (π-π stacking) and red (approach of OsO4).329,331.................................................................................................... 233 Figure 5.35: Dihydroxylation reaction of chiral alkene 260 attempted by Gardiner and Bruce.333 Shown inset is the pyrimidine linked chiral auxiliary (DHQD)2Pyr. ................................. 234 Figure 5.36: The 1H-NMR spectra obtained following dihydroxylation of olefins 239 and 255 in the presence of chiral ligand (DHQD)2Pyr. The signature β-CH3 protons are marked with a colour-coded arrow. ............................................................................................................ 235 Figure 5.37: The diastereomeric products obtained following deprotection of mixtures of 256 and 257, or 241 and 242 prepared using either (DHQD)2Phal or (DHQD)2Pyr under Upjohn dihydroxylation conditions were separated by RP-HPLC. This reaffirms the observed diastereoselectivity deteremined by 1H-NMR spectroscopy. ............................................. 236 Figure 5.38: Chiral CE analysis of the desired diastereomer of dhIle derivative D/L-264, produced from the Cbz- and Boc-protected diols 242 and 257 respectively. The peaks represent each of the enantiomers of this diastereomer (we are unsure of which is L-264 and which is D-264). .................................................................................................................. 238 Figure 5.39: Attempted trityl protection of a mixture of diols was unsuccessful....................... 241 Figure 5.40: Protection of a mixture of the various diols generated following the AD reaction with carbonyl diimidazole................................................................................................... 242 Figure 5.41: Test scale removal of thiocarbonate protecting group under basic conditions. .... 242 Figure 5.42: Solvent gradient utilized in all HPLC analyses in this chapter. ............................ 246  xxiii  LIST OF ABBREVIATIONS AND SYMBOLS φ °C [ox] A Å Abu Ac AD Ala (A) Ama1 Ar Arg (R) Asn (N) Asp (D) atm. ATP BBN Boc BOP Bu c Cbz CD CE CHO cm Cys (C) d d d.r. DABCO DBU DCC dd ddd DDQ DEAC dhIle dhLeu dhNva DHQ DHQD DIPEA DMAP DMDO DMEM  quantum yield degrees centigrade oxidation absorbance Angstrom (2S)-2-aminobutyric acid acetyl asymmetric dihydroxylation L-alanine amanitin gene code generic aromatic group (defined in text) L-arginine L-asparagine L-aspartic acid atmosphere adenosine triphosphate borabicyclo[3.2.2]nonane tert-butyloxycarbonyl benzotriazole-1-yl-tris-(dimethylamino)phosphoniumhexafluorophosphate 1-butyl concentration benzyloxycarbonyl circular dichroism capillary electrophoresis chinese hamptser ovary centimeter L-cysteine deuterio doublet diastereomeric ratio 1,4-diaminobicyclo[2.2.2]octane 1,8-diazabicyclo[5.4.0]undec-7-ene N,N’-dicycloehexylcarbodiimide doublet of doublets doublet of doublet of doublets 2,3-dichloro-5,6-dicyano-1,4-benzoquinone 7-diethylamino-3-carboxy-coumarin (2S),(3R),(4R)-γ,δ-dihydroxyisoleucine dihydroxyleucine dihydroxynorvaline dihydroquinine dihydroquinidine ethyldiisopropylamine 4-(N,N-dimethylamino)-pyridine dimethyldioxirane Dulbecco’s modified essential medium xxiv  DMF DMSO DNA DPDPE DPPA DRB DTT e.r. EcR ED25 EDC EDTA ELL enk. equiv. ERB ESI Et EtOAc FBS Fmoc g GFP Gln (Q) Glu (E) Gly (G) GTP h HATU HBTU HFIP hIle His (H) HOAt HOBt Hpi HPLC HRMS hTrp hVal Hyp Hz hν IC50 Ile (I) iPr IR J k  dimethylformamide dimethylsulfoxide deoxyribonucleic acid 2 5 D-Pen -D-Pen -enkephalin diphenylphosphoryl azide dichlororibofuranosyl benzimidazole (2S),(3S)-dithiothreitol enantiomeric ratio nuclear envelope ecdysone receptor effective dose (in 25% of the population) 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide ethylenediaminetetraaceitc acid RNAP II elongation factor enkephalin equivalents estrogen receptor B electrospray ionization ethyl ethyl acetate fetal bovine serum fluorenylmethoxycarbonyl gram green fluorescent protein L-glutamine L-glutamic acid glycine guanosine-5’-triphosphate hours N,N,N’,N’-tetramethyl-O-(7-azabenzotriazol-1-yl)-uronium hexafluorophosphate N,N,N′,N′-tetramethyl-O-(benzotriazol-1-yl)uronium hexafluorophosphate 1,1,1,3,3,3-hexafluoroisopropanol (2S),(3S)-γ-hydroxyisoleucine L-histidine 1-hydroxy-7-azabenzotriazole 1-hydroxy-benzotriazole 3a-hydroxy-1,2,3,3a,8,8a-hexahydropyrrolo[2,3-b]indole-2-carboxylate high performance liquid chromatography high-resolution mass spectrometry 6’-hydroxy-L-tryptophan (2S),(3S)-γ-hydroxyvaline γ(R)-hydroxy-L-proline hertz light inhibitory concentration (in 50% of the population) L-isoleucine isopropyl infrared coupling constant rate constant xxv  Kd kg Ki kobs l L LD50 LDA Leu (L) LG LR LRMS Lys (K) M m m mCPBA Me MEM Met (M) mg MHz min mL mM mol mRNA MTT n.o. NBn NCS NHS NIH nM nm NMO NMP NMR N-PSP NRPS Ns Ntcp NTP Nv OATP ODN PBS PCR PDB  dissociation constant kilogram inhibitor dissociation constant observed rate constant length litre lethal dose (in 50% of the population) lithium diisopropylamide L-leucine leaving group low resolution low-resolution mass spectrometry L-lysine molar meter multiplet meta-chloroperoxybenzoic acid methyl modified essential medium L-methionine milligram megahertz minutes milliliter millimolar mole messenger RNA (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide none observed 2-nitrobenzyl N-chlorosuccinimide N-hydroxysuccinimide National Institutes of Health nanomolar nanometer N-methylmorpholine-N-oxide N-methyl pyrolidinone nuclear magnetic resonance N-phenylseleno phthalimide non-ribosomal peptide synthesis 4-nitrobenzenesulfonyl sodium-taurocholate transporter protein nucleoside triphosphate 3,4-dimethoxy-2-nitrobenyl organic anion transporter protein oligodeoxynucleotide phosphate buffered saline polymerase chain reaction Protein Data Bank xxvi  PFA Ph Pha1 Phal Phe (F) Pi pKa Pr Pro (P) PyAOP PyBOP PyBrOP Pyr q R RCM rel RNA RNAP RP Rpb rRNA rt s s SAR SDS Ser (S) SN 2 SPPS Su t t½ TBDMS TBP t-Bu TEA TEG Tf TF TFA Tfa TfOH THF Thr (T) TIPS TLC TMP TMS  paraformaldehyde phenyl phalloidin gene code phthalimide L-phenylalanine inorganic phosphate (PO43-) acid dissociation constant (log of) 1-propyl L-proline (7-azabenzotriazol-1-yloxy)trispyrrolidinophosphonium hexafluorophosphate (benzotriazol-1-yloxy)trispyrrolidinophosphonium hexafluorophosphate (bromo)trispyrrolidinophosphonium hexafluorophosphate 2,5-diphenylpyrimidine quartet generic functional group (defined in text) ring closing metathesis relative ribonucleic acid RNA polymerase reversed phase RNA polymerase II gene code ribosomal RNA room temperature singlet seconds structure activity relationship sodium dodecylsulfate L-serine bimolecular nucleophilic substitution solid-phase peptide synthesis succinimidyl triplet half-life t-butyldimethylsilyl TATA-binding protein tert-butyl triethylamine triethyleneglycol trifluoromethylsulfonyl transcription factor trifluoroacetic acid trifluoroacetyl trifluoromethylsulfonic acid tetrahydrofuran L-threonine triisopropylsilyl thin layer chromatrography 2,4,6-trimethylpyridine trimethylsilyl xxvii  Tmse TOF tRNA Trp (W) Trt Tyr (Y) UV Val (V) Xaa α-CD ε λem λex λmax µM µm  2-(trimethylsilyl)-ethyl time-of-flight transfer RNA L-tryptophan trityl (triphenylmethyl) L-tyrosine ultraviolet L-valine generic amino acid (defined in text) α-cyclodextrin extinction coefficient emission wavelength excitation wavelength wavelength of maximum absorbance micromolar micrometer  xxviii  ACKNOWLEDGEMENTS There is no doubt that this dissertation could not have been accomplished without the appreciated assistance of many people. I will attempt to acknowledge these people here. First and foremost, I must acknowledge my supervisor Dr. David Perrin. He has patiently provided me with valued insight and scientific inspiration, and this thesis definitely would not have been possible without him. His comments and guidance during the preparation of this thesis were also extremely helpful. I appreciate the many scientific outlets that I have had the privilege of exploring while working with Dr. Perrin. I would also like to acknowledge Dr. Gregory Dake, for his pertinent reflections when reading this thesis. Several people were instrumental in developing this project before I began work on it, and I would like to mention them as well.  Former post-doctoral researcher Dr. Jonathan May,  performed yeoman’s work determining the basic chemistry and characterization of the amatoxins. He taught me the laboratory skills necessary for accomplishing the work presented here. Jon was also helpful in reading early drafts of this dissertation. I also would like to acknowledge Dr. Pierre Fournier, (soon-to-be) Dr. Antoine Blanc, Mr. Jonathan Pellecelli, (soon-to-be) Dr. Jack Huang and Mr. Liu Yu, for their previous contributions and continued efforts with respect to this project. The various lab members that I had the opportunity of working with not only helped me to grow scientifically, but were a critical part of enjoying the many hours spent in the lab. To all of you I say “Thank-you!” Specifically (and in no particular order), thanks to: Dr. Yoann Roupioz, Dr. Len Lermer, Dr. Marcel Hollenstein, Dr. Curtis Harwig, Dr. Rich Ting, Dr. Jay Thomas, Dr. Ali Asadi, (soon-to-be) Dr. Chris Hipolito, (soon-to-be) Dr. Curtis Lam, (soon-to-be) Dr. Ying Li, and the many undergraduate students to grace A352. There were also a number of people outside of the Perrin lab that contributed to the work presented in this thesis. Thank-you to Dr. Calvin Roskelley and Ms. Jane Cipollone, for their help with the confocal microscopy experiments. Thank-you to Dr. David Chen and his student (soon-to-be) Dr. Jane Maxwell, for their help with chiral capillary electrophoresis separations. Thank-you to Dr. Elena Polishchuk and Ms. Jie Chen, for their patience in helping me adapt my naïve cell biology ideas. Thank-you to Dr. Yun Ling and Mr. David Wong, for their assistance xxix  in mass spectrometry. Thank-you to Dr. Brian Patrick for his crystallographic expertise. Thankyou to Ms. Zorana Danilovic and Ms. Maria Ezhova for their assistance with NMR. Thank-you to John, Pat and Paul in chem stores. Thank-you to Judy, Sherri, Lani, Sandra and Bev in the front office. And finally, thanks to Dr. Chris Orvig, Dr. John Sherman and Dr. Glenn Sammis for patiently dealing with my many, many questions. Outside of the realm of UBC a number of people contributed to this thesis as well. Not the least of whom is my family. For standing by me, supporting me and patiently accepting my answer to the proverbial question “when are you going to graduate?” again I say “Thank-you!” Thanks so much to Mom, Dad, Paul and Megan. I have also had the great pleasure of developing many critical friendships over the years, and my time here at UBC could not have come to pass without them. To Jon, thanks for being so supportive, such a great friend, roommate and defenseman. Rob, Lisa, Julian and Mike, you too cannot know how much I appreciate your constant friendship. Alison – good luck this summer! Steve, Anne, Geoff and Ryan: you are always there when I need a beer or a trip to the mountains. Finally, and for mostly unfounded reasons, thanks to: the staff at the Pendulum, as well as the staff at Elwoods. To everyone listed above: THANK YOU!  xxx  CO-AUTHORSHIP STATEMENT The work presented in this thesis is a collaboration of efforts from many people, and as such, the terms “we” and “our” is used throughout. However, the ideas and design of the research program presented in this thesis were developed primarily by myself in conjunction with Dr. David Perrin. I performed the vast majority of the work presented in this thesis, and any exceptions are noted in the text. The details of work performed by other people are also presented in this statement. Manuscripts that have come from this thesis were co-written by Dr. Perrin and myself. Former students and postdoctoral researchers in the Perrin lab undertook a small number of the reactions presented as data in this thesis. Specifically, Dr. Jonathan May, who began this project, was involved in several features. He was responsible for the CD spectra described in Figure 2.36. His crystallization of 64 and 138 helped us to identify the epimeric products generated upon peptide macrolactamization. The study presented in Chapter 4, where we tested the effects responsible for epimerization (Section 4.2) was undertaken in collaboration with Dr. May. While he performed the reactions and HPLC separations, I provided all work up and analysis of the data. Other students involved in synthetic problems include Mr. Liu Yu (a former summer student), who synthesized Tfa-protected asparagine derivative 160, and Mr. Jack Huang (a former 449 student), who synthesized the D-analog of Tr-Trp-Gly-OMe (197). Help in the initial development of cell culture protocols for Chapter 3 was gained from Ms. Jie Chen and Dr. Elena Polishchuk, in the Biological Services section of the Department of Chemistry. I conducted all of cell viability experiments and analysis presented here. The confocal microscopic analysis of cell permeability was done in collaboration with Dr. Roskelley in the Department of Cell Science at UBC. His student Ms. Jane Cipollone taught me the techniques of preparing the slides, and she operated the confocal microscope. Finally, (soon-to-be) Dr. Jane Maxwell of Dr. David Chen’s lab in the Department of Chemistry was responsible for the chiral CE analysis that was performed in Chapter 5. Dr. Brian Patrick and Mr. David Wong of the Department of Chemistry at UBC performed the crystal structure analysis and high-resolution mass spectrometry respectively. xxxi  CHAPTER 1: INTRODUCTION 1.1 WHAT IS AMANITIN? α-Amanitin (1) is a highly toxic naturally occurring peptide that has been isolated from a variety of mushrooms of the Amanita family.1-5 It is known to be an extremely potent and specific inhibitor of eukaryotic RNA polymerase II (RNAP II),6-8 the chief enzyme responsible for messenger RNA (mRNA) production by the process of transcription. The affinity of RNAP II for α-amanitin is at least 103 times greater than that shown for other polymerases.9 In fact, several other polymerases display no sensitivity to the toxin, despite significant structural and functional homology. These key observations are what have led us to adopt this bicyclic peptide scaffold to develop unique and sensitive probes that can ultimately lead to the understanding and control of mRNA production through RNAP II inhibition. HO HO HN  H 4 N  3  O  O O 2  HO  O  S N H  N 1  O  H N  H2 O N C H 5 HN OH O H2 C NH  N  O  Hyp2  6  8 H  H2N  O  Trp4  dhIle3  Asn1  7  O  S  N H Cys8  Gly5  OH  Ile6  Gly7  1  O  Figure 1.1: The structure and amino acid numbering of the natural product α-amanitin. The structure on the right shows an abbreviation that will be used in this thesis to represent the amanitin scaffold. Transcription is a term used to describe the synthesis of ribonucleic acid (RNA) in a deoxyribonucleic acid (DNA) dependent fashion, including the synthesis of mRNA by RNAP II. Since transcription is a key function within a cell, the ability to understand, visualize and control this process is of great importance.10 With the development of amatoxin probes of RNAP II we hope to provide greater understanding of the structural basis for the toxicity of α-amanitin, which can in turn help to elucidate the chemical mechanisms involved in transcription. We also hope to 1  provide tools that can be used for the spatial localization and control of RNAP II and its associated processes. These key developments will lead to fundamental knowledge that can be applied to the understanding and control of disease at the genetic level.  1.2 THE AMATOXIN PRODUCING MUSHROOMS Approximately 90% of all reported fatal mushroom-related poisonings are related to αamanitin.11 Several fungal species have been shown to produce this toxin in varying amounts, including Amanita,12-15 Gallerina16,17 and Lepiota.18 The most infamous of this group are the Amanita, specifically A. phalloides (also known as the Green Death Cap), A. virosa and A. bisporigera (also known as the Destroying Angel). Species Amanita phalloides Amanaita bisporigera Amanita verna Amanita virosa Gallerina marginata Lepiota brunneo-incarnata  α-amanitin (mg/g) 1.6 1.8 1.55 0.1 1.1 1.3  Table 1.1: List of some fungal species shown to contain the toxin α−amanitin, and the average amount isolated from each, in mg/g dry weight of mushroom.12-18 It is clear that different mushroom species contain varying amounts of the toxin. Confusion with the identification of different Amanita mushrooms can have fatal effects. Aside from the main toxin, α-amanitin, other bicyclic peptides have been isolated from these fungal species, and their activities are also well documented.  1.2.1 STRUCTURAL VARIANTS OF THE CYCLIC PEPTIDES While α-amanitin is the most notorious of the peptides isolated from these fungi, other amatoxins as well as other structurally related peptides have been identified in fungal isolates. The phallotoxins (2)19 and the virotoxins (3)20 are two other classes of peptides described. These toxins all share the property of being cyclic peptides that contain highly oxidized side chains. In the case of the amatoxins and the phallotoxins, there is a second cyclization between a 2  tryptophan and cysteine residue, creating a tryptathionine linkage.  In the virotoxins, the  tryptophan moiety is oxidized to a sulfone, but it is not attached to the peptide backbone forming a second cycle.  A. O HN  2  H 3 N  R2 O  O 1  N O  N 7 H  O  5  O O 6  R6  R1  O  OH  N H 4 HN  S N H  B.  R3  HN R4  2  1  NH  N HO  R5  2  N H 4  Me O  HO  H 3 N  R1 O  CH2OH OH O HN  R2 5 S N O2 H O O 7 H O 6 N NH  HO  HO  3  Figure 1.2: Related cyclic heptapeptides isolated from Amanita species. A. Phallotoxins B. Virotoxins Of the amatoxins, nine related structures have been identified in Amanita fungi. These different peptides represent the same amino acid sequence, with varying degrees of oxidation of the side chains.  Different mushroom species produce different quantities of each of the amatoxins in  question, which leads to the diverse toxicological effects of the various mushrooms.  3  R4 R3 HN O O O  R2  N H N O  O  H2 O N C H H2N R5 S N O H H2 C NH N H O O H N  R1 O  #  Toxin  R1  R2  R3  R4  R5  1 4 5 6 7 8 9 10 11  α-amanitin β-amanitin γ-amanitin ε-amanitin amanin amaninamide amanullin amanullinic acid proamanullin  NH2 OH NH2 OH OH NH2 NH2 OH OH  OH OH OH OH OH OH OH OH H  OH OH OH OH OH OH H H H  OH OH H H OH OH H H H  OH OH OH OH H H OH OH OH  Ki (nM) 2.3 2.5 5 5 5 10 -  LD50 (mg/kg) 0.3 0.5 0.2 0.3-0.6 0.3 20 >20 >20 >20  Table 1.2: Nine different amatoxins isolated from Amanita mushrooms. The Ki values were determined against calf thymus RNAP II and the LD50 values listed are as observed in white mice.2 Although structurally related, both of the other two cyclic peptide classes isolated in the amatoxin producing fungi display completely different biological activity. Neither phallotoxins nor virotoxins are known to inhibit RNAP II (or any RNA polymerase for that matter) to any relevant degree. The phallotoxins do bind very specifically and potently to filamentous actin (Factin).21 This binding has been exploited heavily in cell biology for the visualization of the cellular structure that is imparted by F-actin. The activity of the virotoxins is related to that of the phallotoxins.20 They also bind F-actin, and have no activity against RNAP II. This result is surprising, due to the virotoxins’ lack of the rigidifying tryptathionine linkage. Regardless, the phallotoxins and virotoxins, both heptapeptides, are presumably derived from the same biochemical pathway (vide infra).  4  1.2.2 PHYSIOLOGICAL EFFECTS OF AMATOXIN POISONING Research has been dedicated to the physiological mechanism of action of these poisonous mushrooms for over a century. This effort was greatly enhanced by the isolation of the toxins. Several physiological effects have been recorded in animal models upon the administration of αamanitin. The timeline of these effects, based on the time of ingestion of the toxin, is relatively slow.22 Generally a period of five to seven days pass before the affected subject will die. There is a reported six- to twelve-hour latency period before any symptoms develop. These symptoms are recognized as gastrointestinal discomfort. Following this phase, hepatic lesions develop, as noted by the increased detection of key liver enzymes in serum. This can lead to reduced blood coagulation and internal bleeding. Finally liver failure leads to complications and death. The lethal dose of α−amanitin is different among species. Based on postmortem estimates, humans appear to be quite susceptible to the toxins, with an approximate LD50 of 0.1 mg/kg body weight. There is also a noted variation of susceptibility between species and the method of administration. For example, the rat and mouse are not affected when dosed orally with α− amanitin, but show strong effects when it is administered intravenously. Organism Human White mouse White rat Pig Snail Frog Table 1.3:  Administration oral intra-peritoneal intra-peritoneal intra-peritoneal intra-peritoneal intra-peritoneal  LD50 (mg/kg) ~0.1 0.4 - 0.8 2.8 - 3.5 0.1 - 0.2 > 20 2-5  Lethal dose of α−amanitin in various organisms, and the mode of  administration. After years of research, it was recognized that α−amanitin induces nuclear fragmentation of liver cells, which was shown to be an effect of reduced RNA production. This coupled with the discovery of distinct eukaryotic RNA polymerase structures helped to define the exact mechanism of α−amanitin poisoning. The effect of α−amanitin on the various forms of RNA polymerase was shown to be unique, with its most susceptible target being RNAP II.  5  1.3 BACKGROUND INFORMATION ON TRANSCRIPTION This thesis focuses on the synthesis of derivatives of the natural product α-amanitin with an eye towards probing the mechanism of inhibition of transcription, and providing spatio-temporal control of this process. While the main goal here lies in the chemistry of this molecule, it is important to provide a general introduction to the process of transcription, and to develop understanding of the molecular basis for this biochemical process. With that, I will try to introduce the topic, and provide relevant references for those who desire further understanding of the process.  1.3.1 CENTRAL DOGMA OF MOLECULAR BIOLOGY The central dogma of molecular biology details the flow of genetic information in living organisms. Francis Crick originally proposed the central dogma in 1958, and it remains valid more than sixty years later.23,24 The discovery of novel pathways has increased its complexity, yet the fundamental concept remains constant. The basis of this dogma implies that DNA is the carrier of genetic information.  DNA has the ability to undergo replication to transfer its  information to daughter cells. Transcription is the process of reading a DNA sequence and preparing an RNA analog of it. While DNA stores the genetic information, RNA provides a readable message within a cell. This process produces all of the unique forms of RNA that are required for cellular reproduction and growth.  The final process of the central dogma is  translation. Translation produces proteins from mRNA, with the help of the ribosome. It is these proteins that comprise the vast majority of catalytic and structural elements of a cell. Replication  DNA TRANSCRIPTION  RNA Translation  Protein Figure 1.3: The central dogma of molecular biology. The bold arrows imply the common genetic transfer elements, while the dashed lines are unique pathways found in certain organisms. 6  It is clear from the central dogma that the process of transcription is key to cellular growth and function. The role RNAP II plays in the production of mRNA is vital, and therefore provides an excellent target for developing understanding of a variety of cellular pathways.  1.3.2 DIFFERENT FORMS OF RNA POLYMERASE While transcription refers to the process of preparing RNA from a DNA template, there are a variety of different RNA species that are synthesized. The three most common forms of RNA generated during transcription include: ribosomal RNA (rRNA), transfer RNA (tRNA) and mRNA.25 In prokaryotes, all forms of RNA are transcribed by a single protein complex, termed RNA polymerase.26 In eukaryotic species, however, each unique form is expressed by different RNA polymerase constructs. The specific polymerases responsible for their transcription are termed RNAP I, RNAP II and RNAP III.27,28 Each of these constructs shows different sensitivity to the toxin α-amanitin. Origin prokaryote eukaryote eukaryote eukaryote Table 1.4:  Polymerase RNA polymerase RNAP I RNAP II RNAP III  Product mRNA, tRNA, rRNA rRNA mRNA tRNA  α-amanitin Ki no effect no effect ~1 nM ~1 µM  Comparison of RNA polymerase constructs from different origins, their  respective products and relative sensitivity to α-amanitin. Each of these RNA polymerase complexes described above is comprised of a series of peptide subunits, each of which are required to form an active transcribing enzyme.29 RNAP II is comprised of twelve different peptide fragments, which are labeled Rpb1 through Rpb12. The two largest units (Rpb1 and Rpb2) comprise the catalytic core of the enzyme. The action of α-amanitin on similar RNA polymerase constructs from different organisms is also variable. While RNAP I is not affected by α−amanitin at concentrations as high as 100 µg/mL, RNAP II and RNAP III show varying effects and sensitivities amongst species.30-33  7  Organism Mammalian tissue C. elegans D. melanogaster X. laevis S. cerevisiae A. phalloides  RNAP II IC50 (nM) 10 70 30 50 80 20 x 104  RNAP III IC50 (nM) ~10 x 103 80 x 103 20 x 103 none -  selectivity (RNAP III/II) 100 114 40 very high -  Table 1.5: Inhibitory effect of α−amanitin on eukaryotic RNA polymerases II and III. Although each RNA polymerase complex provides a different function, they share a high degree of homology.  The three eukaryotic polymerases are comprised of several similar peptide  fragments, including Rpb1 and Rpb2.  Remarkably, this homology does not confer cross-  inhibition by α-amanitin. Although the toxin is a potent inhibitor of RNAP II, its effect on RNAP I and RNAP III is drastically reduced. This fact is often taken advantage of to monitor the level of transcription induced by the various polymerases within a eukaryotic cell.9 The bridge helix is a common structural element to all of these polymerases, and is implicated in various important aspects of transcription. It is located near the active site of the enzyme, and is located on the largest peptide subunit. The sequence homology of this region in polymerases of prokaryotic and eukaryotic origin is relatively high. Enzyme RNA polymersase RNAP II RNAP III  Bridge helix amino acid sequence 768 NVLQYFISTHGARKGLADTALKTANSGYLTRRLVDVAQDLVV 809 809 TPQEFFFHAMGGREGLIDTAVKTAETGYIQRRLVKALEDIMV 850 857 SPPEFLFHAISGREGLVDTAVKTAETGYMSRRLMKSLEDLSC 895  Table 1.6: Sequence homology of the bridge helix region of E. coli RNA polymerase and RNAP II and RNAP III of S. cervisisiae.  Bold residues are conserved, while the yellow  residues are structurally related amino acids. While several structural and sequence similarities exist in the various polymerases,34 the exact relationships these exert on the process and mechanism are not fully understood.35 Despite these various similarities, α-amanitin displays unique selectivity for RNAP II. A set of discrete amatoxin analogs could potentially lead to greater understanding of this phenomenon. 8  1.3.3 BIOCHEMICAL MECHANISM OF TRANSCRIPTION The biochemical features of transcription vary between the unique polymerase constructs. Reviews of the biochemical mechanism of prokaryotic transcription are provided elsewhere.36-39 In eukaryotic transcription, genes are often regulated through complex mechanisms, and these can vary gene-to-gene and species-to-species.40 Regardless of the signaling process used to initiate gene production, there are four general phases of transcription: 1) initiation, 2) promoter escape, 3) elongation and 4) termination. Initiation of mRNA production in eukaryotes is complex, but has perhaps been studied in the greatest detail. Upon receiving a signal that gene synthesis is required, RNAP II is recruited to the promoter region of the DNA sequence. This region contains a thymine and adenine rich sequence that occurs upstream of the transcription start site that is referred to as the TATA-box. At first, this region binds the TATA-binding protein (TBP), which acts as a platform for the recruitment of a series of transcription factors (TF) including TFIIB, TFIID, TFIIE and TFIIH.41 This collection of TBP and transcription factors in-turn recruit RNAP II to the DNA, to complete formation of the initiation complex. Upon formation of the initiation complex, the process of transcription is begun, signaled by the TFIIH catalyzed hydrolysis of ATP. In this stage nucleoside triphosphates (NTPs), which are complimentary to the template DNA strand, are added in succession, to the growing 3’ end of the newly synthesized RNA. This continues for approximately ten to fifteen bases. At some point during this process, the core RNAP II unit dissociates from the initiation factors and becomes a fully elongating species capable of extremely rapid and processive incorporation of complimentary NTPs. During elongation, RNAP II rapidly synthesizes mature RNA products. There are a series of different proteins that have been associated with the elongation complex, including Elongin and ELL.42 A series of RNA and DNA processing activities have been observed during elongation, including DNA repair and mRNA capping. Other events that have been described during this stage include pausing and backtracking.43 Various amatoxin probes could be designed that would help to further elucidate the underlying mechanisms of elongation. 9  A. RNAP II B TBP  H  D  E  promoter (including TATA-box)  B. RNAP II B TBP  H  D  E  promoter (including TATA-box) SII  C.  RNAP II B TBP  Elongin  D  promoter (including TATA-box)  Figure 1.4: Cartoon depicting the process of transcription. A. Transcription factors are recruited to the TATA-box promoter site to form the initiation complex. B. Transcription bubble is formed, and the first few NTPs are incorporated to produce new RNA (red). C. Escape from initiation factors, and the recruitment of elongation factors renders the RNAP II complex into a fully elongating form. Upon completion of RNA synthesis, the mature RNA is processed by a series of proteins, some of which are most likely associated with the RNAP II complex. The process of termination of transcription is unclear, and many details are still being studied.44  1.3.4 MOLECULAR MECHANISM OF TRANSCRIPTION The process of transcription is highly processive, with the polymerase traversing across the template DNA to yield mRNA. During transcription, RNA is generated in the 5’ → 3’ direction, and the sequence of the incorporated nucleotides is dependent on the DNA sequence. A small 10  section of the double stranded DNA is melted to produce what is termed the transcription bubble.45 This “bubble” contains a chimeric DNA-RNA double stranded helix of about nine nucleotides in length, and approximately six unassociated DNA bases. 3’  5’  Direction of transcription Figure 1.5: Depiction of the transcription bubble (reproduced from PDB 2E2H). Newly synthesized RNA (red) forms nine base pairs with the DNA template strand (blue). The green non-coding strand of DNA melts approximately six base pairs from the active site, and rejoins the template stand past the transcription bubble (not shown).  The next  nucleotide to be incorporated is shown as a space-filling model in the active site. During transcription, a new nucleotide is added to the 3’ end of the growing RNA transcript. This occurs via nucleophilic attack of the 3’-hydroxyl group onto the α-phosphate of the incoming nucleotide. The exact mechanism has not been conclusively determined, but empirical evidence shows that two Mg2+ ions are involved in the process.34,46 One of these ions remains associated with the polymerase, and is involved with the nucleophilic attack of the hydroxyl group, while the other is only transiently associated with the enzyme, and is involved in binding the pyrophosphate leaving group.35 A series of aspartate residues located in the active site are responsible for binding the magnesium ions. Recently, a histidine residue in the trigger loop has been identified to be involved in the chemical step, and is proposed to activate the leaving group.47 Together, these data provides an accurate picture of the active site and its mechanism.  11  A.  B.  RNA R  O O P  O  O  O  O Asp481 O  OH  Mg2+  O  Asp483  N  O Asp485 O  Mg2+  O  O  P  O P O  N  NH2  O  O O  NH  N  OH  O  O OH  O O  OH  P O  O  HN N H  His1085  Figure 1.6: Two views of the active site of RNAP II (reproduced from PDB 2E2H). A. General depiction of the active site, based on crystal structure data and genetic analysis. B. Reproduction of the critical residues in the active site as viewed in recent crystal structures. These residues are shown in light blue, the 3’ end of the RNA is shown in pink. After the addition of the nucleotide to the growing RNA chain, RNAP II translocates along the DNA chain.48 The polymerase slides one base pair upstream on the DNA template, which is achieved by a ratcheting mechanism induced by the bridge helix. This is accompanied by the elimination of pyrophosphate from the active site. At this point, the polymerase is ready to incorporate the subsequent RNA nucleotide in the sequence.  1.3.5 STRUCTURE OF RNA POLYMERASE II Structural data obtained by the Kornberg lab over the last decade has significantly contributed to understanding the chemical interactions of RNAP II.34,49-51 They have prepared and diffracted a series of crystals of ten out of the twelve possible peptide subunits of the polymerase. They have achieved this in its free form, and in the presence of unique substrates and products. They also describe a co-crystal structure of RNAP II in complex with α-amanitin.52 The original structure published in this series gives a definitive three-dimensional shape to the polymerase, and details the interaction of the various subunits that comprise this complex.  It  also shows key components of the active site, including the Mg2+-binding region, the bridge helix and the trigger loop. 12  A.  B.  Figure 1.7: Overall structure of RNAP II in complex with substrate and products. A. Cartoon diagram showing side-view of the polymerase, including the transcription bubble (from Kornberg et al.).52  B. Top-down view, reproduced from the reported crystal  structure (reproduced from PDB 2E2H). In both figures the new RNA is shown in red, the template DNA strand in blue, and the non-coding strand in green. Important insight into the mechanism of inhibition of transcription by RNAP II induced by α− amanitin exposure comes from the reported co-crystal structure of the toxin and its substrate. The crystal structure confirmed that the toxin binds at a site away from the active site, as had been proposed based on biochemical data. They showed that the toxin interacts with the bridge helix region. It was proposed that this tight interaction of amanitin removes the flexibility of the bridge helix, thus inhibiting translocation. The interaction of amanitin with the polymerase also shows that the trigger loop region of Rpb1 cannot move into the position that it adopts during the trans-phosphorylation step.  13  A.  B.  Figure 1.8: Close-up view of the active site of RNAP II. A. The active site when cocrystallized with the natural GTP substrate, shown in blue. B. The active site when cocrystallized with the inhibitor α-amanitin, shown in red. The orange depicts the Mg2+binding domain, green shows the bridge helix domain, and the trigger loop is shown in magenta. The structures reproduced in Figure 1.8 clearly demonstrate that the binding location of αamanitin in RNAP II is close but distinct from the active site. The altered conformation of the bridge helix is visible, and the trigger loop is oriented in a different environment when exposed to the toxin. A series of conclusions regarding the chemical requirements for binding of amanitin to RNAP II were drawn from the co-crystal structure. While some of these were in agreement with previous structure activity relationships (SAR), others were not.  The most notable interaction that  confirmed several SAR data was that of the hydroxyl group of Hyp2 of 1, which was shown to form a hydrogen bond with the Glu 822 side chain located on Rpb1. This could serve to explain the drastic loss in potency when this hydroxyl group is removed (vide infra).  14  dhIle3  Glu 822 – Hyp2 Figure 1.9: Close-up diagram of α-amanitin in the binding pocket of RNAP II (reproduced from PDB 1K83). The residues of RNAP II proposed to make critical contacts with the toxin are shown in light blue. Perhaps the most confusing aspect derived from this crystal structure is based on the relative flexibility that is observed for the dhIle3 region of the toxin in its binding site. Many studies have investigated the function of this amino acid within the amatoxin framework. These SAR data detail some important requirements of this modified amino acid for the toxicity of various amatoxins (vide infra). However, the function and importance of these dhIle3 modifications is not entirely clear from this co-crystal structure. This discrepancy further details the need to generate a discrete library of amatoxins to further develop understanding of the key requirements for inhibition of transcription by α-amanitin.  1.4 THE NATURALLY OCCURRING AMATOXINS The development of amatoxin probes for RNAP II to increase the understanding of transcription relies on the synthesis of these cyclic peptides. It has been demonstrated that the bicyclic nature of the octapeptide is critical for inhibition of RNAP II. This rigid framework also induces a degree of hydrolytic and proteolytic stability. A complete synthesis of the natural product 1 has never been reported, based on the high degree of unnatural amino acids that exist in the structure, however several analogs of this structure have been prepared. These analogs have been prepared both through modifications of the isolated natural product, as well as through completely 15  synthetic methodologies. The initial steps of the biosynthetic pathway have also recently been described.  All of these features combined will allow for the identification of the most useful  approaches to develop probes of RNAP II.  1.4.1 BIOSYNTHESIS OF AMATOXINS AND PHALLOTOXINS The toxic cyclic peptides of the Amanita fungi were proposed to be biosynthetically prepared by a non-ribosomal peptide synthesis (NRPS) pathway. The incorporation of modified amino acids and the bicyclic nature of these toxins all pointed to this theory. Recently, researchers attempted to determine the biosynthetic pathway of amatoxins and phallotoxins in A. bisporigera.53 Assuming an NRPS pathway, the researchers searched the fungus’ genome for sequences related to other known fungal and bacterial NRPS genes with no success. They then scanned the genome for genes encoding for the cognate amino acid sequence of the amatoxins or phallotoxins. Surprisingly, they detected a gene encoding for an Ile-Trp-Gly-Ile-Gly-Cys-AsnPro octapeptide – a possible sequence of amanitin. The genome sequence for this octapeptide did not display any matches to other species, implicating its role in amatoxin specificity. Upon PCR isolation of the gene, they discovered that the original gene product contains 35 amino acids. A similar gene was isolated which corresponded to the biosynthesis of phallotoxins. Gene Ama1 Pha1  Peptide product MSDINATRLPIWGIGCNPCIGDDVTTLLTRALC MSDINATRLPAWLVDC-PCVGDDVNRLLTRSLC  Table 1.7: Genetic analysis of A. bisporigera identified two genes that are part of the biosynthetic pathway leading to amatoxins and phallotoxins. The italicized residues are conserved in both genes, and the bold residues relate to the toxin. The highlighted residues correspond to the Trp and Cys involved in the tryptathionine crosslink. The genes encoding for the biosynthesis of amatoxins and phallotoxins contain an invariable consensus sequence up-stream of the toxin. The proline residues that bookend the toxin are likely involved in signaling specific proteases to liberate a linear version of the toxin. These key findings will lead to further elucidation of the enzymes required for the biosynthesis of these peptides into mature toxins.53 16  1.4.2 CHEMICAL MODIFICATION OF NATURAL AMATOXINS Aside from data obtained from the nine naturally occurring amatoxins, little was known about the structural elements required for binding to RNAP II. Based on the highly modified backbone of the peptide, the synthetic challenge of preparing derivatives was daunting. The first set of SARs that were completed on amatoxins was based on chemically modified derivatives of the isolated natural products. 1.4.2.1 MODIFICATION OF THE BICYCLIC RING STRUCTURE Confirmation that the rigidity imparted by the bicyclic nature of the amatoxins is crucial to activity was determined through two experiments. Mild acid hydrolysis of α−amanitin leads to preferential hydrolysis of the dhIle3 residue, via formation of a γ-lactone, which yields monocyclic derivative 12 that retains the tryptathionine linkage. HO  A. Trp4  dhIle3  O  O  Gly5  Trp4  H3N  Gly5  NH Hyp2  O  S  Asn1  B.  Ile6  N H  mild acid  Hyp2  Asn1  O  O  Cys8  Gly7  Asn1  Trp4  Gly5  dhIle3  1 dhIle3  Hyp2  S  Ile6  N H Cys8  Raney Ni  Hyp2  Asn1  Gly7  S  Cys8  12  Trp4  Gly7  Gly5  Ile6  N H Ala8 CH3  1  Ile6  N H  Gly7  13  Figure 1.10: Selective cleavage of one cycle of α−amanitin. A. Mild acid hydrolysis of the amide bond between dhIle3 and hTrp4. B. Raney nickel reduction of the tryptathionine crosslink. Alternatively, exhaustive treatment of amanitin with Raney nickel,54 cleaves the tryptathionine linkage, yielding desulfurized macrolactam 13. 17  Both of these derivatives were shown to  completely abolish the activity, proving the importance of the conformational rigidity of the bicyclic peptides. 1.4.2.2 MODIFICATION OF THE SULFOXIDE The function of the sulfoxide was probed through reduction with Raney nickel.55 Controlled reduction conditions were able to revert the sulfoxide of 1 to thioether 14. This was then reoxidized to produce a separable mixture of the natural (R) sulfoxide, the unnatural (S) sulfoxide 15, as well as sulfone 16. The toxic effect imparted in white mice based on each of these modified amatoxins was tested. Surprisingly, the reduced thioether and the sulfone retained most of the activity of the natural (R) sulfoxide while the (S) isomer was shown to be nearly ten times less effective at inhibiting the activity of RNAP II.55  dhIle3 Hyp2 Asn1  Trp4  X  Gly5 Ile6  N H Cys8  Gly7  # 1 14 15 16  Linkage (X) (R) SO S (S) SO SO2  LD50 (mg/kg) 0.3 0.25 20 0.5  Table 1.8: Effect of the oxidation of the thioether of the tryptathionine of amatoxins on their toxicity in white mice. The difference in activity that was observed in the (S) sulfoxide relative to the other forms was thought to be related to a conformational change in the peptide.  This was confirmed by CD  spectroscopy and X-ray crystal structure analysis. In this study, the (S) form seemed to induce a global conformational change relative to the natural (R) form, based on the CD spectra. 1.4.2.3 MODIFICATION OF ASPARAGINE Two natural amatoxins that display similar activity against RNAP II are α-amanitin (1) and βamanitin (4). The difference between these two derivatives is a formal Asn to Asp mutation at position 1. Derivatives of β-amanitin are easily accessible using coupling reagents to form esters 18  or amides.56 Introduction of a methyl or thiophenyl ester at this position did not have a drastic effect on the observed LD50 in white mice. Amide modification of the side chain showed a loss of activity of approximately one order of magnitude.  The modified peptides that did not  drastically affect toxicity (esters 17 and 18) were proposed to be physiologically hydrolyzed upon administration,2 which would generate the potent derivative 4. dhIle3 O  Hyp2  O  Trp4  S H N  N H Cys8  Gly5 OH  # 1 4 17 18 19 20  Ile6  Gly7  O  R1 NH2 OH OCH3 SPh NHPh NH(CH2)11CH3  LD50 0.39 0.97 0.8 0.6 4.45 4.0  R1  Table 1.9: Effect of modification at the β-position of asparagine. Reported as LD50 in values of mg/kg in the white mouse. The crystal structure of β-amanitin displays a hydrogen bond between the carbonyl of the γcarboxylate with the peptide backbone.57,58 This could serve to explain why both Asn and Asp are nearly equally active in the natural product.  Since they both possess this carbonyl  functionality, they possess the ability to form this hydrogen bond. 1.4.2.4 MODIFICATION OF DIHYDROXYISOLEUCINE Of the naturally occurring amatoxins, three derivatives contain an unmodified isoleucine at position three, and they all show reduced activity (9 – 11). Two more derivatives are lacking the δ-OH group, but still have the γ-OH functionality (5 and 6). These derivatives behave similar to the parent compound. The dihydroxylated isoleucine residue at position three provides a handle for chemical manipulation.  The 1,2-diol has been selectively oxidized in the presence of  periodate to yield the aldehyde derivative 21.59 This aldehyde was reduced to the alcohol by borohydride treatment to yield S-hydroxyvaline ((S)-hVal) derivative 23. Alternatively, 21 reacted with dinitrophenylhydrazine to yield hydrazone 22.  19  Ph  H N  H N  H N  H H  Trp4  Gly5  O Hyp2  O  H N  HN  Trp4  Gly5  Asn1  O  S  Cys8  Gly7  0.26 µM  Ile6  N H  Ile6  N H Cys8  Asn1  O Hyp2  S  O  22  HO  H N  HN  Gly7  Trp4  Gly5  O Hyp2  12 µM  S  O  21  Cys8  Asn1  Ile6  N H Gly7  3.3 µM  23 Figure 1.11: Modification of the dhIle3 residue of α−amanitin. When 1 was treated with periodate, aldehyde 21 was generated. This was converted to alcohol 22 or hydrazone 23. The values below each derivative represent the observed Ki assayed against wheat germ RNAP II. Of these derivatives, the aldehyde showed a drastic reduction in activity, while the hydroxyvaline (hVal) derivative (23) re-gained some of the activity. The hydrazone adduct also displayed greater activity than the aldehyde. At this point, the importance of the γ-hydroxyl group seemed to be implicated in the activity of amatoxins, which was further explored through synthetic derivatives (vide infra). 1.4.2.5 MODIFICATION OF HYDROXYTRYPTOPHAN The 6’hydroxy group of hTrp4 presents a unique handle for the chemical modification of amatoxins.  Natural products lacking this hydroxyl group (8, for example) do not display  significantly different inhibition of RNAP II compared to α-amanitin. The Faulstich group converted the phenolic hydroxyl group into methyl ether 24 using diazomethane.60 Alternatively, they used the phenolate ion to nucleophilically displace alkyl halides to yield a variety of ethers 20  (25 - 27). When the activity of these ethers was tested, a fine balance between LD50 and IC50 was noted. The ethyl ether behaved similarly in vitro, but was much less effective in vivo. With the decyl ether, the in vivo effect was on par with the parent compound, but it was much less active in vitro. These results were postulated to be a reflection of the balance between binding ability and pharmacokinetics (including gut absorption, liver accumulation, cell-membrane translocation and metabolic clearance).2,60  dhIle3 Hyp2  Gly5  Trp4  O  Asn1  Table 1.10:  S  Ile6  R5  N H  # 1 8 24 25 26 27  Gly7  Cys8  R5 OH H OCH3 OCH2CH3 OCH2Ph O(CH2)9CH3  LD50 (rel) 1 1.1 1 6.7 5 1  IC50 (rel) 1 2 1.2 1.3 10 4  The toxicity and inhibitory capacity of a variety of modifications of the  hydroxyl group of hTrp4 of α-amanitin. Alternatively, this hydroxyl group was alkylated with bifunctional linkers, such that they displayed a free carboxylic acid or amine group. These bifunctional linkers were useful for the introduction of reporter groups to amatoxins. Some reporters prepared include: a fluorescent moiety,61 biotin (for avidin capture),62 proteins,63 or a sepharose resin (for the purification of RNAP II).64 dhIle3 Hyp2  Trp4  O  Asn1  S  N H Cys8  Gly5 O  H N  Ile6 O  Gly7 fluorophore = biotin protein sepharose resin  (28) (29) (30) (31)  Figure 1.12: Introduction of chemical and biological reporters to amatoxins through the hydroxyl group of hTrp4. Other modifications that have been introduced onto the tryptophan moiety include the catalytic reduction of a tetrazole derivative, to yield 8.65 Alternatively, hydroxytryptophan was iodinated 21  to give 32.66 In the presence of N-iodinated dialkyl amines, the 7’ position was modified with dialkyl amines (33).67,68 Treatment with diazotized aromatics led to coupling at the 7’ position as well (34).69,70 These derivatives displayed a varying degree of activity in toxicity and inhibitory studies.  A.  dhIle3 Hyp2  Trp4  O S  Asn1  N H Cys8  B.  Gly5 H  dhIle3 Hyp2  Ile6  Trp4  O S  Asn1  Gly7  dhIle3 Hyp2  Trp4  O S  Asn1  N H Cys8  OH  Ile6  I Gly7  32  8  C.  Gly5  D.  Gly5 OH  N H  NR2  Cys8  Gly7  Ile6  dhIle3 Hyp2  Trp4  O  Asn1  33  S  N H Cys8  Gly5 OH Ile6 Ph N N Gly7  34  Figure 1.13: Modified amatoxins prepared via modification of the hydroxytryptophan residue. A. The hydroxyl group was removed by catalytic reduction. B. Iodination at the 7’-position. C. Dialkyl amination at the 7’-position. D. Diazotization at the 7’-position. The limited information gained from these studies enforced the importance of the development of a reliable method to synthesize these peptides. An approach that would allow the preparation of amatoxins from their constituent amino acids would permit the introduction of novel functionality, and in so doing would allow further elucidation of the chemical mechanism of RNAP II inhibition.  1.5 CHEMICAL SYNTHESIS OF AMATOXINS There are several difficulties that present themselves in the attempted total synthesis of α− amanitin. The synthesis of peptides in a stepwise fashion and head-to-tail cyclization are the least of these challenges. Three of the most significant challenges lie in the preparation of the tryptathionine  crosslink,  the  γ,δ-dihydroxyisoleucine 22  (dhIle)  residue  and  the  6’hydroxytryptophan (hTrp). The final challenge is of the least importance, since derivatives lacking this hydroxyl group have been shown several times over to have a minimal effect on the inhibition induced by the toxin. Preparation of the dhIle3 residue presents several problems. The stereochemistry of the amino acid is important, and acid catalyzed formation of a γ-lactone must be avoided. To date, no amatoxin has been prepared that incorporates a synthetic version of this residue. Since the co-crystal structure of α−amanitin and RNAP II does not show any key contacts of the hydroxyl groups of this amino acid - and in fact even shows a fair amount of disorder in this region - implies that they are not necessary for binding to the yeast polymerase. This provides encouraging evidence that the dhIle can be mutated to a commercially available Ile residue without significant loss of activity. The tryptathionine linkage has been shown to be absolutely critical for the toxicity of amatoxins, due to the conformational rigidity that it imposes on the peptide. Conditions that can induce this oxidative crosslink while maintaining the chemical integrity of the rest of the peptide are crucial. Several efforts over the years have appeared in the literature, which outline methodology to introduce this crosslink into peptides.  1.5.1 TRYPTATHIONINE FORMATION Three main approaches have been utilized to generate the tryptathionine crosslink common to the amatoxins and phallotoxins. The original approach took advantage of an activated cysteine residue that was nucleophilically attacked by the indole of tryptophan.  This has been  demonstrated in the early synthetic approach to prepare Gly1-norphalloin (38).71 In this approach two peptide fragments were prepared in solution, one bearing an embedded tryptophan and the other a cysteine residue (35). The free thiol moiety of the cysteine-containing fragment was treated with N-chlorosuccinimide (NCS) to generate sulfenyl chloride 36. This was reacted with the tryptophan-containing peptide to generate the desired crosslinked peptide (37). Following tryptathionine formation, two N- to C-terminal cyclization events were induced in a stepwise fashion to generate the desired phallotoxin.  23  Cl  HS O HO  O N O  N H  H N O  NCS H2 CH3CO2H C N Boc H  O  O N  HO  O  HO  OH  H2N  tripeptide  O HO  O N H  O OH  OCH3 O  cyclization deprotection cyclizaion  HO  O H N  N H  H N O  HO  N H O NH  O S H2 C N Boc H  O  HN  N H O  N  O  H2 C N Boc H  36  O S  H N  N H  OH  35  H N  S  N H  H2C O  O N O  HO  NH  N H HO  37  38  Figure 1.14: Wieland’s synthesis of Gly1-norphalloin using a sulfenyl chloride to generate the tryptathionine crosslink. More recently a solid-phase approach to the synthesis of a phallotoxin was achieved using similar crosslinking strategies.72 In this approach, orthogonally protected tryptathionine dipeptide 41 was generated via sulfenyl chloride activation of cysteine. This dipeptide was added to a hydroxyproline residue anchored to a solid support (42). The rest of the amino acids were introduced on the solid phase using standard methodology to generate 44.  Finally, the  orthogonal protection of the tryptathionine moiety allowed for control of two lactamization events to generate Ala7-phalloidin (45).  24  H Ns-Trp-OAllyl Ns N  Cl S  O OAllyl steps  NaHCO3  tBu-O  S  NH Boc O  tBu-O  39  S  N H NH Boc  O  H Ns N  HO  OAllyl N H NH Fmoc  O  O  41  40 O Tmse  HN O O  HO  NH  OAllyl  S  +  O  42  O Tmse  O  H Ns N  O  N H NH Fmoc  O  H Ns N  HN O O  N O  HN O O  Fmoc NH OAllyl  N H tBu O  OAllyl  S  N  N H NH Fmoc  43  O  deprotection x2 cyclizaion  O  H N  H N O  NH  O HN  O  O  H Ns N  O  O  S N H  PyAOP HOAt 2,4,6-collidine  41  O Tmse SPPS  O  S  O  O  O NH  N H O  N  N H tBu O  O  44  HN O NH  45  Figure 1.15: Guy’s solid-phase synthetic approach to synthesize Ala7-phalloidin using an orthogonally protected tryptathionine dipeptide. There are drawbacks to generating peptides containing a tryptathionine crosslink using activated thiols. The conditions required for formation of the sulfenyl chloride (NCS), and the reactivity and limited stability of this species limits its usefulness. In the approach detailed by Wieland, the sulfenyl chloride is generated on a peptide substrate. This severely limits the substrate scope, because other amino acids may be sensitive to the conditions required to generate the sulfenyl chloride. In Guy’s synthesis of Ala7-phalloidin, the problem of exposing the peptide to these conditions was averted through the generation of the Trp-Cys dipeptide. The problem with this approach is the high degree of protecting group manipulation that was required to fully elaborate the peptide into a phallotoxin.  25  A complementary approach to generate the tryptathionine crosslink has been described recently by Lokey et al. in their preparation of Glu7-phalloidin.73 In this approach, linear peptide 46 containing underivatized tryptophan and S-trityl protected cysteine residues was synthesized on the solid-phase. The peptide was macrolactamized on resin to afford 47, which was then treated with ten equivalents of iodine to sequentially deprotect the S-trityl group and generate sulfenyl iodide in situ, which was attacked by the indole to generate the Glu7-phalloidin 48. TIPSO NH  O tBu O  H N  Fmoc  O  H N  N H  N O  S  O  H N  Trt  O  O  H N  N H  OAllyl  O  46  deprotection cyclizaion  O  O  O  O  N H  HN  O  O 10 eq I2 DMF  HN O TrtS  TIPSO  O  O  H N  O  N H  O  NH  N O  47  H N  TIPSO  S  N H  O  NH  N O  tBu O  O HN  O NH  N H  HN  O  O  O  48  O NH  tBu O  Figure 1.16: Lokey’s approach to synthesize Glu7-phalloidin, using iodine to generate the typtathionine crosslink. The initially reported yields for this approach were high (>90%), with disulfide formation between two neighboring molecules being the main byproduct. While this approach is mild and occurs with good yield, it does have limitations. The first problem is that a very low loading of the resin was required (generally 90% of the bead capacity was blocked). In this case much more resin must be used to generate an equivalent amount of peptide, as compared to other strategies. In this approach, one is limited to a peptide that contains only one single tryptophan residue and one cysteine residue, while all other amino acids must be inert to iodine treatment. The main approach that has been employed to generate amatoxins and phallotoxins relies on the Savige-Fontana reaction. This approach reverses the polarity of the reaction, and instead relies 26  on the nucleophilicity of the cysteine thiol. It has been readily applied to solid-phase techniques, does not require extra protecting group manipulation of the peptide, and is compatible with a wide variety of amino acids.  1.5.2 THE SAVIGE-FONTANA REACTION During an early study of the byproducts obtained upon oxidation of tryptophan, an unexpected product was observed.74 This was shown to be a mixture of syn-cis (49a) and anti-cis (49b) diastereomers of 3a-hydroxy-1,2,3,3a,8,8a-hexahydropyrrolo[2,3-b]indole-2-carboxylate (from here on referred to as Hpi). Hpi proved to be susceptible to nucleophilic attack by thiols under acidic conditions, which generated tryptophan moieties substituted at the 2’ position of the indole ring.75 It was quickly shown that the thiol of cysteine is a suitable nucleophile for this reaction, generating the tryptathionine linkage shown in 50. H2N  CO2H CH CO H 3 3 0° C  H3N CO2H  HO  NH HN  CO2H  HO  NH  +  HSR H+  N H H  N H H syn-cis  CO2H  anti-cis  N H  S R  Hpi 49a  49b  50  Figure 1.17: Savige’s original findings leading to the Savige -Fontana reaction. The exact mechanistic details of the Savige-Fontana have not been studied, but a general mechanism has been postulated.76 Under acidic conditions, the N-terminal protecting group is removed, and simultaneous activation of C-2’ of tryptophan occurs through a preformed imminium ion. Subsequently, this group is susceptible to attack by free thiol. Elimination of water to rearomatize the indole ring yields the tryptathionine linkage.  27  O HO N  Trt  N H  O  O HO  HO  H+  R  R  R  NH3  NH2  N H  N H  HS  H+ HO  O R  NH3 S N H H R  R O R N H  NH3 S R  Figure 1.18: Proposed mechanism of the Savige-Fontana reaction.76 This process is generalizable: any number of sulfur nucleophiles can be used and the reaction can happen either inter- or intramolecularly. One important feature is that the diastereomers that are formed upon oxidation of tryptophan (49a/b) are destroyed in the Savige-Fontana reaction. This implies that separation of these two isomers during Hpi preparation is not necessary. Zanotti and Wieland recognized the potential of the Savige-Fontana reaction to prepare members of the amatoxins and phallotoxins, including S-deoxo-Ile3-amaninamide (53).77 In their new approach, the Hpi was generated in solution, then added to a linear peptide. In the presence of acid, an internal cysteine residue was poised to nucleophilically attack the Hpi, to yield monocyclic intermediate 52. An N- to C- terminal macrolactamization reaction yielded the desired amatoxin.  28  NH2 O  Trt N  N H  HN  O  H N  N H  O  OH  H N O  O  O N H  STrt  51 O H2N  O HO  O N H  O N O OH  H2 O N C H HN  TFA  O  N  O tBu  N H  O  O  O tBu  macrolactamization  H N  HN  O  O O  S N O H H2 H N N C NH H O O  HO  N O H2N  NH2  52  H N  S N H O    N H  H2 O N C H H N 2 O H2 C NH O  O  53  Figure 1.19: Zanotti and Wieland’s application of the Savige-Fontana reaction to generate S-deoxo-Ile3-amaninamide (53). This approach has proven to be useful in the synthesis of unnatural amatoxins and phallotoxins.77-81 It is amenable to solid phase peptide synthesis, and relies on the relatively facile preparation of a robust amino acid surrogate.  1.5.3 PREPARATION OF HPI FOR THE SAVIGE-FONTANA REACTION The Hpi moiety that is required for the induction of the Savige-Fontana has been prepared in a series of different laboratories. In fact, the Hpi moiety itself is a common motif that is observed in a variety of isolated natural products. Some of these products, shown in Figure 1.20, include: brevianamide E (54),82 phaketllistatin 3(55)83 and gypsetin (56).84  29  O  O N H  O  HO  N H  N NH O  H N  HN  O  H N  O  HO  N  HO O  O  N  O  N  N  HN  N  N N H  OH O  O HO  54  55  56  Figure 1.20: Natural products containing an Hpi moiety. A. Brevianamide E B. Phakellistatin 3 C. Gypsetin. The core tricyclic Hpi unit is obtained from the mild oxidation of a suitable tryptophan surrogate. The oxidation of tryptophan with mild oxidants induces the formation of epimeric epoxides across the 2’-3’ bond of the indole ring, which are subsequently attacked by the α-nitrogen to form the core Hpi unit. The epoxidation can lead to two diastereomers of Hpi. Over the years, a variety of methods have been developed and refined to maximize the yield of Hpi formation (see Figure 1.21).79 Danishefsky et al. have shown that the t-butyl protecting group can be used to block the α-carboxylate and generate 57. Previous work in our lab by Mr. Pierre Fournier85 showed that the addition of this protecting group was not trivial, and that the acid conditions required for the hydrolysis of this protecting group led to poor yields of Hpi. Other work in our lab carried out by Dr. Jonathan May, showed that changing the α-carboxylate protecting group can increase the yield and ability to manipulate Hpi.78 In this case the methyl ester of the C-terminal amino acid (glycine) in the peptide sequence is used to protect the αcarboxylate of the tryptophan, to produce Hpi dipeptide 58. The advantages of using an amino acid as a carboxylate protecting group include the stability of the derivative, the diastereoselectvity that can be obtained with some amino acids, and the base labile protecting group (methyl ester) that presents itself (58). 30  Trt  O  O  H N  O  tBu  DMDO -78 °C  HO N  H N  Trt  N H H N  O  O N H  CO2Me DMDO -78 °C  HO N H H  O N H  1. O2, hν Rose bengal 2. PPh3  CO2Me  N H Trt  N  58  O HO  N H  N  59  N H H  N H Cbz  57  Trt  N H H  N H  tBu  O  H N  O  1. N-PSP 2. mCPBA O  O HO  O N  N H Boc  60  Cbz  N Boc  Figure 1.21: Various methods reported for the preparation of useful Hpi surrogates. Other methods to generate Hpi have been developed over the years. One method involved the generation of Hpi after tryptophan was introduced into the peptide, to prepare phakellistatin analogs (59).83 This was carried out photochemically using oxygen, in the presence of Rose Bengal as a sensitizer.  Alternatively, when tryptophan was treated with N-phenylseleno  phthalimide (N-PSP), a 3a-phenylseleno Hpi substrate was produced. This was deselenated using m-chloroperbenzoic acid (mCPBA) to generate 60.86-88 A commonly employed approach involves the oxidation of a suitably protected tryptophan with the mild oxidant dimethyldioxirane (DMDO).82 This generates the desired Hpi as a mixture of diastereomers, generally in good yield (>70%).  31  1.5.4 LINEAR SYNTHESIS OF AMATOXINS USING HPI The majority of amatoxin analogs reported in the literature have been generated through the use of the Savige-Fontana reaction. This approach takes advantage of a protected derivative of tryptophan, yet the natural product bears a 6’-hydroxy tryptophan group. This hydroxyl group can be omitted without concern since derivatives lacking this group display nearly equal potency. Amatoxins have been prepared both by solid-phase techniques and in solution using this approach.  Regardless, they share a certain degree of similar requirements.  In the linear  synthesis of amatoxins, the Hpi moiety must be added at the N-terminus of the peptide. This is particularly important in syntheses using acid sensitive protecting groups on the α-nitrogen. The requirement of an N-terminal Hpi necessitates the following amino acid sequence: O O Trt H Xaa1 Xaa2 Xaa3 O N Xaa5 Xaa6 Xaa7 N  HN  S  OH  O  Trt 61  Figure 1.22: General linear amino acid sequence that is required to generate amatoxins. In solution phase approaches, R represents a t-butyl protecting group, whereas in SPPS, this represents the resin. In this sequence the C-terminal amino acid relates to the dihydroxisoleucine residue of the natural product. Since this amino acid is not readily available, synthetic endeavors have mostly focused on the mutation of this residue to isoleucine. In solution phase syntheses the carboxylate of this amino acid is generally protected as a t-butyl ester, whereas in solid-phase peptide synthesis (SPPS), it is directly attached to the resin. To elaborate linear sequence 61 into a functional amatoxin, two cyclization steps are required. Linear peptides are exposed to neat trifluoroacetic acid (TFA), which concomitantly removes all acid labile protecting groups, and invokes the Savige-Fontana reaction. Following purification, monocyclic peptide 62 is obtained, which bears both a free N-terminal amine and a C-terminal 32  carboxylate. These functional groups are coupled together in the second cyclization step by a macrolactamization reaction, to yield the desired amatoxin (63). CO2H Xaa3 H2N Trp  61  Xaa5  Xaa3  Trp  Xaa5  lactamization  TFA  Xaa2  S  Xaa1  N H Cys  Xaa6  Xaa2  Xaa7  S  Xaa1  62  N H Cys  Xaa6 Xaa7  63  Figure 1.23: Elaboration of linear peptides containing Hpi into amatoxins. This process is general, and can accept almost any amino acid in most positions, so long as a cysteine residue exists within the linear peptide, and Hpi exists at the N-terminus. A series of derivatives of amatoxins have been made following this general protocol, which have provided important SAR data regarding amatoxin inhibition of RNAP II. 1.5.4.1 EFFECT OF HYDROXYPROLINE AND SULFOXIDE REVISITED One of the first sets of synthetic amatoxins prepared using the Savige-Fontana reaction focused on observing the effect of the hydroxyl group of hydroxyproline. In this case, peptides bearing an isoleucine in position three, and either hydroxyproline (53) or proline (64) at position two were synthesized.77,80,89 These were synthesized in solution phase according to the protocols outlined in Figures 1.22 and 1.23. Trp  Ile  Gly  # 53 64 65 66 67 68  O R  X  N Asn  N H Cys  Ile Gly  R OH H OH OH H H  X S S (R)-SO (S)-SO (R)-SO (S)-SO  Ki (rel) 32 880 36 230 3880 8000  Table 1.11: Synthetic proline and hydroxyproline variants of amatoxins. The Ki values reported are relative to α-amanitin, which were obtained against RNAP II from Calf Thymus (α-amanitin Ki ~ 3 nM). 33  The purified bicyclic compounds were also exposed to mild oxidizing conditions to generate the sulfoxide, as a separable mixture of the (R) and (S) isomers (65 – 68). All of these derivatives were tested for their ability to inhibit RNAP II isolated from calf thymus. Confirming earlier work, the derivatives lacking the hydroxyl group at position two were nearly four orders of magnitude less active than the natural product. Also, the (R) sulfoxides were nearly equally effective at inhibition of RNAP II as the thioether derivatives, however the (S) sulfoxides were less effective. This mimics the results obtained with the natural product, as outlined in Section 1.4.2.2. 1.5.4.2 DERIVATIVES OF THE DIHYDROXYISOLEUCINE RESIDUE One crucial difference between the available synthetic amatoxins and the natural amatoxins is the amino acid incorporated at the dihydroxyisoleucine position (dhIle3). Various amatoxin derivatives have been prepared through the Savige-Fontana approach that incorporate variations of dhIle, but none has been described that include a genuine version of the modified amino acid. Modifications have focused on two classes of substitutions: the first class tested the effect of various hydrophobic amino acids in this position,89 while the other was involved in testing the importance and stereochemistry of the γ-hydroxyl group.90,91 The ability for these derivatives to inhibit RNAP II in vitro was compared to α-amanitin (1) or γ-amanitin (5) (Ki = 2.3 and 5 nM respectively). The structure and data obtained with these derivatives are displayed in Table 1.12.  34  Trp  Xaa Hyp  S  Asn  R2 R1 Xaa =  R3  N H Cys  R4 R5  N H O  Table 1.12:  # 53 69 70 71 72 73  Gly Ile Gly  # 1 5 74 75 23  Xaa dhIle (4S)-hIle (4R)-hIle (3S)-hVal (3R)-hVal  R1 R2 CH2OH H CH3 H CH3 OH H H H H  Xaa Ile D-Ile allo-Ile Val Leu Ala R3 OH OH H OH OH  Ki rel (α) 32 128 1400 2700 R4 H H H CH3 H  R5 CH3 CH3 CH3 H CH3  Ki rel (γ) 100 100 Ki rel (γ) 1 1 10-50 >200 10-20  Amatoxin analogs that have been prepared to test the significance of  dihydroxyisoleucine. The derivatives were compared to α-amanitin against calf thymus RNAP II in vitro, or else to γ-amanitin, tested in vitro against RNAP II isolated from D. melanogaster embryos. While it is capricious to directly compare these data, since the amatoxin derivatives were tested in different contexts, the results nonetheless show that dhIle3 is relatively sensitive to modification. Many of the modifications that were introduced at this position showed drastically reduced activity compared to the natural products. In particular, three factors seem to play a crucial role in the in vitro inhibition of RNAP II. First of all chain branching at the β-position, as evidenced by derivative 53 compared to 72 is important. The stereochemistry at this position is also important, based on the results obtained with 75 relative to 23. Finally, the stereochemistry of the γ-hydroxyl group is also important, based on derivative 74 as compared to 5. To date, no synthetic amatoxin derivative has been prepared that incorporates dhIle into an amatoxin, although the synthesis of this amino acid has been described. Greater focus on the requirements and synthesis of dhIle3 will be discussed in the fifth chapter of this thesis.  35  1.5.4.3 OTHER BACKBONE MUTATIONS A variety of amatoxins have been synthesized that test the function of some of the other side chain residues with respect to RNAP II inhibition.81,89,92,93 These include mutations at the asparagine in position one, glycine at position five, isoleucine at position six, and glycine at position seven. These derivatives are summarized below.  Ile Hyp Xaa1  Trp S N H Cys  Xaa5 Xaa6 Xaa7  # 53 76 77 78 79 80 81  Xaa1 Asn Asn Asn Asn Asn Asn Abu  Xaa5 Gly Ala Gly Gly D-Ala Gly Gly  Xaa6 Ile Ile Ala Ile Ile Ile Ile  Xaa7 Gly Gly Gly Ala Gly D-Ala Gly  Ki (rel) 32 7200 22400 110 870-1740 12-52 n.o.  Table 1.13: Effect of changing various backbone amino acids in the amatoxin structure. Values reported are relative to the effect of α-amanitin on calf thymus or D. melanogaster (values in talics) RNAP II. (n.o. = no inhibition observed) This table shows that each of these positions plays a crucial role in the inhibition of RNAP II. Most notably, substitution of isoleucine six for an alanine residue, as in peptide 77, effects the inhibition by over five orders of magnitude. CD spectroscopy revealed that this derivative resulted in a severely altered conformation, relative to the toxic analogs. Also important was the modification of asparagine in position one to α-amino butyric acid (Abu) to give peptide 81. This represents a formal mutation of the β-carboxamide with a methyl group. This eliminates the hydrogen bond to back bone amides that is implicated for the carboxamide. Sure enough, this derivative displayed no activity. The altered conformation of this derivative was also evident by CD spectroscopy. It is clear that a wide range of factors within the natural amatoxins contribute to a toxic molecule. Perhaps the most important feature that these toxins provide is their extremely rigid conformation. Modifications of the backbone that alter the three-dimensional structure lead to an extremely diminished capacity to inhibit RNAP II. Hydroxyproline and the γ-hydroxyl group of dihydroxyisoleucine are also critical. Loss of these residues results in derivatives that are 1036  1000 times less effective inhibitors. It appears that the most effective place for modifications of the synthetic amatoxins that should not greatly affect its inhibitory capacity stems from the carboxylate of the asparagine.  1.6 PROJECT GOALS This project aims to build on this existing SAR data to develop further understanding of the structural requirements for the inhibition of RNAP II by amatoxins. Using a combination of previously reported structures with novel amatoxins designed by our lab, we hope to use this knowledge to create probes of eukaryotic transcription. Our goal is to synthesize amatoxins that will provide us with the capability to expand understanding of the process of transcription. With these probes we hope to gain the ability to visualize and control this process on a fundamental level. This can prove to be beneficial in the control of human disease.94,95 The NIH has recently identified the need for the development of chemical probes of biological systems.96 They have developed a number of initiatives for “New Pathways to Discovery”, including a “Molecular Libraries and Imaging” roadmap. The goal of this initiative is to: “… identify small molecules that can be optimized as chemical probes to study the functions of genes, cells, and biochemical pathways. This will lead to new ways to explore the functions of genes and signaling pathways in health and disease.” Selective chemical probes have the ability to provide understanding of chemical responses to disease through spatial visualization, metabolic control and the identification of interacting cellular components. As transcription is a central process in biology, the development of probes of RNAP II would be extremely useful. A series of small molecules that bind bacterial RNA polymerase have been identified, with rifampicin (82) perhaps being one of the most famous.97 Current methods of inhibiting eukaryotic transcription rely heavily on targeting enzymes that affect the activity of RNAP II but do not directly act on this target.  37  AcO H3CO  OH  OH OH OH  82 NH N N  O O  N  OH  O  O  O N O  O N O O  N  N  N  N  O  O NH  NH  O  O N  NH2  O  O  N N HO  O  O  O  HN  NH  83  O O  Cl  Cl  84 OH  HO  Figure 1.24: Small molecules that have been used to inhibit transcription. A. Rifampicin. B. Actinomycin D. C. Dichlororibofuranosyl benzimidazole (DRB). Actinomycin (83) is used to control transcription through DNA intercalation. This inhibits RNAP II’s ability to elongate RNA due to the stable duplex DNA formation.98 Another molecule, dichlororibofuranosyl benzimidazole (DRB - 84), inhibits an RNAP II associated kinase,99-101 which stalls the elongation process. Probe IIA Antibody IIB Antibody IIo Antibody Rifampicin (82) Actinomycin (83) DRB (84) α-amanitin (1)  Target RNAP II C-terminal domain RNAP II C-terminal domain RNAP II C-terminal domain prokaryotic RNA polymerase double stranded DNA RNAP II associated kinase RNAP II  Table 1.14: Previously described probes of transcription and their specific target. 38  Antibodies have been isolated that selectively recognize different domains of RNAP II. These have been used to control and visualize various aspects of transcription.102 While the antibodies are extremely selective for RNAP II, they are of limited use in the study of dynamic functions. The application of antibodies as probes is hampered by their ease of preparation and delivery. The peptide α-amanitin provides a great scaffold for the development of probes of transcription. The bicyclic core is accessible via synthetic means, and many amino acids, both natural and synthetic, are commercially available. Specifically we hope to prepare amatoxins that will allow for the visualization and control of transcription on a cellular level.  We aim to achieve  visualization through the synthesis of a fluorescent amatoxin, and we hope to gain control of inhibition through the use of a light dependent protection group. The synthesis of these probes, as well as the development of relevant assays, will be the main focus of this thesis. It will also cover improved methodology in the preparation of amatoxins. Finally, the synthesis of stereochemically-defined dihydroxyisoleucine was undertaken, with the hopes of providing more efficient inhibitors of RNAP II.  39  CHAPTER 2: SYNTHESIS OF PHOTOLABILE AND FLUORESCENT AMATOXIN PROBES  2.1 INTRODUCTION The selective binding of α-amanitin to RNAP II provides us with an accessible scaffold for the preparation of probes of transcription.  Chemical probes that specifically act on the  transcriptional machinery can provide insight into the mechanism of this process. The ability to spatially and temporally control transcription can also have important consequences.10 Since mRNA production is vital for cell growth, controlled inhibition of this process could lead to selective modulation of disease proliferation. Further control and understanding of transcription and related processes could be achieved with appropriate inhibitors of RNAP II. Preceding structure activity relationships5 provide a basis for identification of the flexible positions of the bicyclic peptide.  It has been demonstrated that the 6’-OH group of  hydroxytryptophan (hTrp4) and substitution at asparagine (Asn1) are relatively plastic and provide a good target for adding chemical labels.3 Hydroxyproline (Hyp2) also bestows a unique position for labeling amatoxins. Due to the key hydrogen bond that this residue has been shown to form with Glu A822 of RNAP II,52 the hydroxyl group offers the possibility to modulate the activity of an amatoxin. We hope to apply techniques of probe development currently described to achieve RNAP II inhibition.103 We hope to provide understanding of this fundamental process, and the consequences of inhibition of this process, through discreet probes that display spatiotemporal control. This chapter will focus on our approach to synthesizing amatoxin-based probes of RNAP II. One set of these probes will bear a photolabile protecting group at Hyp2, which should prove to be useful in generating a light-activatible inhibitor.  Another probe was designed to bear a  fluorescent label on Asn1. This label was intended to provide the ability to visualize RNAP II within a cell, and help to study the cell permeability of amatoxins.  40  2.1.1 LIGHT-INDUCED SPATIO-TEMPORAL CONTROL IN BIOLOGY The ability to chemically control biological function is a significant challenge.104 For years, scientists have developed chemicals that possess the ability to modulate enzymatic activity. However, the ability to induce this modulation in a space- and time-dependent fashion provides an added layer of difficulty. Several approaches to achieve this have been attempted.105-110 Perhaps one of the most effective methods for achieving this control is through the use of photolabile functionalities.  Light provides a non-invasive method of triggering inhibition  through a process sometimes referred to as photoactivation.  This process (also called  photocaging or photo-deprotection) relies on the use of a protecting group which blocks the action of a molecule of interest, but is reverted to an active form upon exposure to a controlled light source. The concept of photoactivation in biology is one that has been in practice for several decades. While the presence of light sensitive protecting groups had been described for use in chemical synthesis,111,112 Kaplan et al. were the first to demonstrate that light could be used to selectively activate a biological process.113,114 This was demonstrated using 85, a nitrobenzyl protected derivative of adenosine triphosphate (ATP). This was assayed against an ATP-hydrolyzing enzyme that is responsible for cellular sodium and potassium influx. It was shown that the enzyme did not produce the expected inorganic phosphate (Pi) when exposed to 85 in the dark. Upon irradiation of this system, centered at 342 nm, the protecting group was removed to yield 86, and Pi formation was observed.  NH2  NH2 O O O O P O P O P O O O O  N O  N  N 340 nm N  O  N  N N  HO OH  HO OH NO2  O O O O P O P O P O O O O  N  hydrolyzed by ATPase  no reaction with ATPase  85 86 Figure 2.1: Photoactivatible ATP derivative used for the assay of Na/K ATPase. Kaplan et al. confirmed that this was a light dependent process, and that the applied light itself did not have a dramatic effect on cell function. This seminal discovery opened the door for  41  research into a variety of groups suitable to photoactivation and their subsequent use in the control of biological processes. 2.1.1.1 PHOTOLABILE PROTECTING GROUPS A variety of molecular skeletons have been identified that possess the ability to cleave upon exposure to light.110,115 The use of such a moiety as a biologically relevant protecting group requires that it is rapidly released in the desired environment, and it yields non-toxic byproducts. Figure 2.2 shows some of the more common photoactivating groups used for control of biological processes.116-119 LG O O  LG O2N  O  O  LG O  Br HO  254 nm  O  LG  O  365 nm  87  > 300 nm  340 nm  89  90  88  Figure 2.2: Common photolabile protecting groups used to control biological processes. In each case, the leaving group (LG) is released at the wavelength shown below. Some commonly employed photoactivating groups based on the nitrobenzyl moiety are shown in Figure 2.3. These require at least an ortho-nitro substituted benzyl functionality. This generic skeleton has been used to protect molecules bearing various functional groups, including phosphates, alcohols, amines and amides. LG O2N  LG  LG O2N  O2N  OCH3 OCH3  nitrobenzyl (NBn) 87  nitroveratryl (Nv) 92  91  Figure 2.3: Photolabile protecting groups based on the ortho-nitrobenzyl substitution pattern.  42  Various derivatives of this group have been developed, as is shown in Figure 2.3. They are based on either a benzyl alcohol, as in 87 and 92, or a 2-hydroxy phenyl ethyl alcohol, as in 91. The extra methyl group of 91 has been shown to improve the rate and yield of deprotection. Importantly, when the phenyl ring of the protecting group is modified, it can affect the wavelength of light required for removal. Functionality that increases this wavelength renders the group more useful in a biological setting, as cells can be sensitive to shorter wavelengths (< 300 nm). 2.1.1.2 MECHANISM OF PHOTOLYSIS The mechanism of photolytic nitrobenzyl ether cleavage has been studied in detail.120,121 Nanosecond laser flash photolysis is one method that has been applied to distinguish between the kinetically unique transient species formed in this process. Specifically, methyl nitrobenzyl ether (93) was used as a model compound in one study. Following photolysis of 93, the end products were identified as methanol and nitrosobenzaldehyde (94). MeO  MeO NO2  93 Figure 2.4:  hν  N  93b  MeO  O  MeO O N  OH  OH  NO +  NO  OH  93c  O  93d  MeOH  94  Products observed upon 254 nm photolysis of methyl nitrobenzyl ether,  monitored by laser flash photolysis by Illitch et al.23 This photolysis proceeds through three distinct intermediates (93b-93d), as evidenced by the species detected by both UV and IR spectroscopy. The generation of photo-product aci-nitro intermediate 93b is the rate limiting step, with metastable species 93c and 93d only transiently observed. The overall quantum yield of this reaction was demonstrated to be φ = 0.49 ± 0.05 in an acetate buffer.  The overall rate of decomposition of 93 proved to be dependent on several  factors, including solvent, pH, buffer and leaving group. While these intermediates are of chemical relevance on a short time-scale (< 100 s), we are more interested in the release of the caged compound, and the biological fate of byproduct 94. This 43  byproduct can react deleteriously with proteins, however, free thiol scavengers such as DTT or 2-mercaptoethanol have been added to quench the nitrosobenzaldehyde in in vitro experiments.109,113,122,123 When photolysis of a 2-nitrobenzyl protecting group is performed within a cellular environment, the nitroso byproduct 94 is thought to be quenched by cellular quantities of glutathione or cysteine.122 Despite the limited information available on the intracellular fate of this species, the prominence of this photolabile protecting group in the biochemical realm renders it useful. 2.1.1.3 CONTROL OF TRANSCRIPTION WITH PHOTOACTIVATION Photolytic control of biological processes has proven to be extremely useful for agonist release,124,125 but not been fully explored in regards to transcription. Control of transcription has generated some interest, with the ability to spatially and temporally control the production of mRNA (and hence a specific gene product) being the goal. The experiments that have been described affect transcription in an indirect fashion.126 There are no known light-triggered inhibitors that directly inhibit the activity of RNAP II. An example of indirect control of gene expression through photoactivation, has been described using nitroveratryl protected ecdysone derivatives 95a and 95b.127-129 Ecdysone is a steroid involved in insect metamorphosis that binds to the nuclear envelope receptor EcR. Once EcR recognizes ecdysone, it forms a complex that initiates gene expression. When cells that express EcR were treated with 95a/b, no gene expression was observed.  Following a one-minute  exposure to 348 nm light, 60% of the transcription activity expected from exposure to free ecdysone was obtained.  HO HO  H3CO H3CO  OH  OH  OH  NO2  HO OH  O OH HO  OH  H  O  O H3CO H3CO  95a  H  O  NO2  95b  Figure 2.5: Light-sensitive ecdysone derivatives that have been used for photolytic control of gene expression. 44  Another example130 of photolytic control of gene expression uses estrogen receptor B (ERB) tamoxifen based antagonist 96. When this derivative was photolyzed, a marked decrease in the expression of certain genes was observed, as the released molecule inhibits estrogen response elements. Selective inhibition of transcription was observed with a luciferase reporter gene when cells growing in the presence of 96 were exposed to light at wavelengths greater than 345 nm.  O  O  N  O2N  OCH3 OCH3  96 Figure 2.6: Photocaged tamoxifen was shown to inhibit production of a specific mRNA product, upon exposure to 345 nm light. The control of transcription through a photolytic process has perhaps been most often demonstrated with photo-caged derivatives of oligonucleotides (ODNs).131-134 The sequence specificity that oligonucleotides confer can target these polymers to distinct genes. Several examples of ODNs that bear light-sensitive protecting groups have been shown to target specific cellular targets only after they have been exposed to light. These have been applied in both cellular assays and whole organisms. While the systems described above possess the capability to control transcription, the effect is indirect.  They do not have the ability to control transcription at the level of the RNA  polymerase, since they do not act directly on the transcriptional machinery. The ability to control transcription by directly affecting RNAP II activity in a photolytically dependent fashion would prove to be advantageous. It would allow for the study of discrete transcription events in a spatial and temporal manner.  45  2.1.2 FLUORESCENT VISUALIZTION OF BIOLOGICAL TARGETS The development of fluorescent probes that enable the visualization of biological processes on a cellular level has revolutionized the field of cellular and molecular biology. Over thirty years ago Taylor and Wang noted that:135 “…the high sensitivity… and versatility of fluorescence techniques can be fully utilized in living cells to yield information at the molecular level.” Through covalent modification of a biologically relevant molecule with a fluorophore, its target can be visualized within a cell.  This can provide information about its localization and  interaction with other cellular components.135-138 Fluorescent markers have been appended to a whole host of biological probes, including nucleic acids,139,140 antibodies,141 and small molecules.142 The introduction of fluorescent markers within a cell through genetic engineering has been extremely important and this was recognized with the 2008 Nobel Prize in Chemistry. Fluorescent probes have been used to detect structural elements, organelles, and individual proteins.  Other fluorescent probes have been designed that can detect changes in the  intracellular conditions. Some of the approaches used to prepare fluorescent probes for use in biology will be briefly presented here. 2.1.2.1 FLUOROPHORES IN BIOLOGY When selecting a fluorophore for biological labeling, the possibilities seem endless.  The  Molecular Probes Handbook143 outlines the process by which a fluorescent label should be selected for any given probe. The two most important questions that are necessary to answer are: 1) At what wavelength should the probe be active? And, 2) how will the fluorophore be covalently attached to the probe?  The answer to these questions will greatly narrow the  selection. For the purposes of confocal microscopy, a limited number wavelength tuned lasers are available (see Table 2.1), so the fluorophore should be active in one of these ranges.  46  Laser Source  Emission wavelengths (nm)  Argon  457, 488, 514  Helium-Neon  543, 594, 612, 633, 1523  Krypton-Argon  488, 568, 647  Diode  405, 440  Table 2.1: Common lasers used in confocal microscopy, and their respective emission wavelengths. The fluorescent label should have an absorption maximum that centers near the excitation wavelength, and should have ideal fluorescent properties. These are reflected in the measure of two physical quantities: the absorption coefficient and the quantum yield. The former refers the intensity of light that is absorbed by the molecule, while the latter refers to the efficiency of absorption versus emission. Figure 2.7 shows some of the more common fluorophores employed in cell biology applications. A.  H2N  O  HO O  O  97 343 nm 442 nm  C.  N O  OH  R  HO3S  λex λem  B.  O  N  O R  O  O R  O  98  99  497 nm 517 nm  557 nm 578 nm  Figure 2.7: Fluorophores commonly employed in cell biology and their absorption and emission wavelengths: A. AlexaFluor 350, B. Fluorescein C. Tetramethyl rhodamine. (R= biological probe) With the selection of a tag bearing the ideal fluorescent properties, it can then be decided what its point of conjugation to the probe of interest should be. The most important feature is that the position of attachment does not dramatically affect the activity of the probe, which can be determined a priori or through literature precedence. The majority of methods used for labeling 47  a probe rely on reactive functional groups. For the most part, these groups comprise thiols or amines. Electrophilic derivatives of many fluorescent probes are well documented and in many cases commercially available. Some of the most common electrophilic moieties used are isocyanates, alkyl halides and activated esters. These react smoothly with a desired target molecule that bears a free nucleophile - under a variety of compatible conditions - to yield the desired fluorescent probe. 2.1.2.2 FLUORESCENT VISUALIZATION OF TRANSCRIPTION The development of a fluorescent probe that specifically targets RNAP II can help to further elucidate its mechanism, on both a molecular and cellular level. Since amanitin binds RNAP II tightly and selectively, a fluorescently labeled analog could help to provide insight into its cellular localization, and potentially separate the binding from inhibition. To the best of our knowledge, only one such analog has been described.61 HO  A.  O OH S  NH  B.  O  NH O  HO  HN O  HO  H N  HN O O HO  N  O  H2N O  H H2 N C  O  HN O  H N  O  O  S  N H N H  O O O  C NH H2  28  Figure 2.8: A. Fluorescent amanitin analog 28. B. Fluorescent image representing cellular localization of 28 within PtK1 cells, as reported by Wulf et al.44 The fluorescent derivative 28 was prepared from the natural product α-amanitin, which was directly isolated from mushroom cultures. It was modified on the 6’-hydroxy functional group of tryptophan to contain a free amine group on the end of a spacer. This was then labeled with 48  fluorescein, and tested in cells. Panel B in Figure 2.8 shows the nuclear localization of 28 in rat kangaroo PtK1 cells, which was shown to be amanitin dependent. While this probe proved useful in the intended use (observation of RNAP II in different cell phases), it has some drawbacks. It was prepared directly from the isolated natural product, the synthesis cannot be easily applied to derivatives of amanitin, since these lack the 6’-hydroxyl group. For use in understanding the mechanism of RNAP II inhibition and function, it would be beneficial to have a series of easily prepared amanitin analogs that bear a fluorescent tag. An alternative approach that has been heavily employed to visualize transcription and related processes has relied on molecular biology techniques.144 In these cases, variants of green fluorescent protein (GFP) are encoded through genetic manipulation, such that they are covalently linked to peptides of interest. Most of the reports that use this strategy have focused on proteins relating to RNAP II or the products of transcription.145 An engineered RNAP II has been prepared in Chinese hamster ovary (CHO) cells that bear a GFP derivative on the largest subunit of the polymerase.146 The introduction of fluorescent functionality into biological molecules through genetic engineering approaches seems to have enormous potential. This approach essentially guarantees the required link between the observed fluorescent output (phenotype) and the target to which it is covalently attached to (genotype). The several unique GFP variants that have been described can allow for the detection of several proteins in the same sample, which are distinguishable by wavelength. One of the drawbacks of this approach is the genetic manipulation that is required. On a cellular level, this is relatively easily accomplished in a host of cell lines, however, in larger organisms this can prove to be more problematic. Another issue with this approach is that the addition of protein bulk to a cellular enzyme has the very real possibility of altering or even abolishing its wild-type activity. Synthetic molecules that possess the ability to bind specifically and tightly to a desired target and also bear a fluorescent functionality can prove to be extremely useful. These could be complimentary to the genetic techniques currently in practice.  2.1.3 GOALS OF THIS CHAPTER The remarkable binding of α-amanitin (1) to RNAP II renders this toxin useful for the preparation of probes that can be used to provide understanding, control and visualization of 49  transcription. It is a chemically accessible bicyclic octapeptide, with previous SAR data that outlines the key interactions that are required for its function.2 This chapter will focus on the preparation of two classes of amatoxin probes of RNAP II, bearing either a photolabile protecting group or a fluorescent label. These are shown in Figure 2.9. O  A. HN  R  H2 O N C H HN  H N O O  R  O  N  NO2  S N H  H N  O  C NH H2  N H  O  O  O  O  a, R = H b, R = OCH3  100a/b  N O  B.  H2 O N C H HN  H N  HN O O HO  N  S N H  H N  O  C NH H2  N H  O  O  O  O NH O  O O  101 O  NEt2  NH O  Figure 2.9: Target molecules to be prepared in this chapter. A. Photolabile derivatives of amanitin. B. Fluorescently labeled amatoxin. Our first generation of probes of RNAP II will reflect two approaches to transcription control and visualization. The first class of probe will bear a 2-nitrobenzyl-derived protecting group on the critical hydroxyproline residue. This group was introduced since it can be photolyzed intracellulary, with the hopes of generating an active probe of transcription. The second class of amatoxin prepared will bear a fluorescent label on asparagine for the purpose of cellular visualization of RNAP II and amanitin binding studies. These proposed probes display a few mutations relative to the natural product. Aside from the novel functionality that will be added to the peptide scaffold at positions two or three, there are  50  several deletions relative to the natural product that we chose to introduce. These modifications were chosen based on the previously reported SAR data. Based on the SAR data described in the introduction, it is clear that amatoxins that are lacking the 6’-OH group of hTrp4 are nearly equally as potent.77 The (R)-sulfoxide of the natural product can also be reverted to a thioether without significant reduction in activity.147 The oxidized dhIle3 residue poses the greatest challenge to synthesizing amatoxins. Since this residue shows disorder in yeast RNAP II co-crystal structure, we predicted that a dhIle3→Ile3 mutation would be suitable to maintain some of the inhibitory capacity of the natural product.52 We chose to use a synthetic approach to achieve these amatoxin probes based on previously reported methodologies that take advantage of the Savige-Fontana reaction.75  2.2 PREPARATION OF AMINO ACIDS FOR INCORPORATION INTO PROBES In order to produce desired probes 100a/b and 101, a series of modified amino acids must be prepared. While fully protected standard L-amino acids compatible with solid-phase peptide synthesis (SPPS) are commercially available, ones specifically designed for photoactivation or fluorescence techniques are not. For our goals, we require a hydroxyproline derivative that bears a light-sensitive protecting group on the hydroxyl moiety, and an asparagine residue bearing a fluorophore attached through a linker to the β-carboxamide. Fmoc O  A.  R  N  R  B. OH  Fmoc  H N  O OH O  O  HN  NO2  O  O  H N  O  O  NEt2  O a, R = H Fmoc-Hyp(NBn)-OH b, R= OCH3 Fmoc-Hyp(Nv)-OH  Fmoc-Asn(TEG-DEAC)-OH  102a/b  103  Figure 2.10: Amino acids required for the synthesis of amatoxin probes. A. Hydroxyproline bearing a light-sensitive group (red) on the γ-OH group. B. Asparagine bearing a triethylene glycol (TEG)-linked fluorescent label (blue) on the β-carboxamide.  51  For the solid-phase compatible hydroxyproline monomer, we chose to prepare two analogs (102a and 102b), bearing two different nitrobenzyl ethers. We chose the nitrobenzyl-based protecting group based on its predominance in the literature, its tunability and commercial availability. The nitrobenzyl (NBn) derivative 102a is less suitable for cell biology application due to the 254 nm light required for maximal photolysis, but could prove to be useful in in vitro experiments. The nitroveratryl (Nv) derivative 102b can be photolyzed at the longer wavelength of 366 nm, which is more compatible with living cells. One requirement of these monomers is that the photolabile protecting groups withstand the standard Fmoc-based SPPS conditions used to generate the linear peptide, as well as the conditions required to induce the Savige-Fontana reaction. While the incorporation of a fluorophore into an amatoxin has been achieved previously (28), this report appended a fluorescein moiety onto the 6’-OH position of tryptophan of the natural product. This is not adaptable to the synthesis of modified amatoxins. We chose to add the fluorophore through a triethylene glycol (TEG) linker attached to the β-carboxamide of Asn. This should maintain the required back-bone hydrogen bond which is commonly observed at this position. The amino acid required for the synthesis of the fluorescent amatoxin (103), also must maintain stability to the standard conditions required for amatoxin preparation. Our choice of fluorophores was based on its ability to be prepared easily in the lab, and its ability to be conjugated to amatoxins through an amide bond. Previous reports of diethylaminocoumarin (DEAC) labeled amino acids and their subsequent use led us to use this same fluorophore.148  2.2.1 SYNTHESIS OF HYDROXYPROLINE DERIVATIVES The installation of a light-sensitive benzyl protecting group onto the commercially available trans-hydroxyproline was visualized to proceed via Williamson ether synthesis.149  This  required the preparation of suitably protected starting material 105, which has been described (Figure 2.11).150 Hydroxyproline was converted to the methyl ester using SOCl2 and methanol. The α-nitrogen was protected as a t-butyl carbamate with Boc2O, to yield the suitably protected derivative 105.  52  H N  O  SOCl2 MeOH  Cl  O  H2 N  OH  OCH3  (73 %)  1 equiv. Boc2O 3 equiv. NaHCO3 dioxane:H2O (2:1)  Boc N  O OCH3  (86 %)  HO  HO  HO  104  105  Figure 2.11: Preparation of Boc protected hydroxyproline methyl ester 105. These protecting groups were chosen to provide the desired stability during alkylation of the γOH group. Conveniently, they also begin to provide important information on the stability of the photolabile protecting group. The conditions required for the removal of the Boc protecting group mimic those used to achieve tryptathionine formation. If the light-sensitive protecting groups can withstand removal of the Boc protecting group, then it stands to reason that they will be stable to Savige-Fontana cyclization. 2.2.1.1 NITROBENZYLATION ATTEMPTS USING WILLIAMSON ETHER SYNTHESIS We sought to alkylate the γ-OH group using standard benzylation methodology.149 Hydroxyproline was treated with sodium hydride, followed by the addition of commercially available nitrobenzyl bromide (106) or nitroveratryl bromide (107).  Unfortunately several  attempts were unsuccessful at producing any amount of product.  Varying the solvent,  temperature and equivalents of base and/or electrophile were tested all with disappointing results (Table 2.2). Entry  107  Solvent  Conditions  Yield  1  1 equiv.  THF  30 min. rt  0%  2  1.5 equiv.  DMF  1h, 0 °C → 5 °C  0%  3  1.1 equiv.  DMF  3h, 0 °C → rt.  0%  Table 2.2: Results of attempted alkylation of protected hydroxyproline derivative 105, following typical benzylation conditions.  53  Following these results we consulted the literature.  Many of the examples in which a  nitroveratryl or nitrobenzyl ether were installed, did so onto a phenolic hydroxyl group. These were prepared under milder conditions (K2CO3, THF) than those employed in Williamson ether synthetic conditions. We also found a report where they claim that ortho-nitro substituted benzyl bromide systems are much more vulnerable to decomposition under basic conditions and addition of this protecting group is not practical in conditions using sodium hydride.151 Hence, we had to determine an alternative approach to achieve our desired product. 2.2.1.2 PREPARATION OF NITROVERATRYL BROMIDE During the course of these studies we encountered two difficulties with the nitrobenzyl and nitroveratryl bromide electrophiles. We found that these were not stable for long periods of time on the bench (especially the nitroveratryl derivative), and their commercial availability was decreased. We then prepared nitroveratryl bromide fresh, starting from nitroveratryl aldehyde, adapting known procedures (Figure 2.12).152 HO  O NO2  NaBH4 MeOH  NO2  (85 %)  H3CO  H3CO  OCH3  OCH3  108  Br PPh3 Br2 CH2Cl2  109  NO2 H3CO OCH3  107  Figure 2.12: Preparation of nitroveratryl bromide using adapted literature protocols. Aldehyde 108 was reduced with NaBH4 in methanol to yield the alcohol 109. The product was stored as the alcohol, and bromination was always performed directly before it was required. The bromide was prepared by triphenylphosphine-mediated bromination in good yield. This proved to be a more feasible method of obtaining nitroveratryl bromide than directly purchasing it from commercial suppliers, since it was prepared fresh, and was readily available. 2.2.1.3 ALTERNATIVE METHODS OF NITROBENZYLATION Once we determined that SN2 displacement of nitroveratryl bromide did not proceed smoothly using sodium hydride promoted alkylating conditions, we sought to test some other techniques. 54  The two main alternative routes tested were: silver mediated alkylation153,154 and phase transfer catalysis.155 Boc N  Br  O  HO  NO2  +  OCH3  Boc  solvent additive rt, dark, 72 h  OCH3 H3CO  H3CO  OCH3  O OCH3  105  N  O  NO2  107  110b  Figure 2.13: General scheme depicting alkylation of hydroxyproline derivative 105 with nitroveratryl bromide. We first tested the ability for silver to aid in this substitution reaction. The protected Hyp derivative 105 was reacted with nitroveratryl bromide (107) in the dark, in the presence of soluble and insoluble silver salts. Entry  Solvent  Additive  Yield  1  CH2Cl2  10 eq Ag2O  17%  2  MeCN  10 eq Ag2O  21%  3  CH2Cl2  10 eq AgCO2  ~ 15%  4  toluene  10 eq AgClO4  ~ 15%  Table 2.3: Nitrobenzylation of hydroxyproline 105 using silver mediated conditions. These reactions produced clean product, with no observable byproducts. However, the yields observed were still not ideal (~ 15-20%) and the reaction was extremely slow, taking up to 72 hours to obtain these yields. While this set of conditions was promising, based on its ability to produce the desired product, we set out to find alternative approaches to improve the yield and reaction time required. During the course of these studies, a report surfaced where a nitroveratryl serine derivative was prepared using a trichloroacetimidate derivative of nitroveratrol (Figure 2.14).156,157  55  O Fmoc  H N  Cl3C OAllyl  OH  +  O  O NO2  NH H3CO  TfOH  H N  OAllyl  Fmoc  OCH3  O H3CO  111  112  H3CO  113 NO2  Figure 2.14: Trichloroacetimidate protection of serine reported by the Lawrence lab. While this protocol descrives the synthesis of a light-sensitive serine derivative, in modest yields, it is undertaken on a primary alcohol. In our case we are working with a more sterically hindered secondary alcohol, which would likely produce a lower yield. We did not follow up on this approach, since we developed an alternative methodology that proved to be fruitful. The best approach to undertake this benzylation reaction proved to be using phase transfer catalysis. In this reaction, both the nucleophile and electrophile are soluble in the organic phase, but the base is soluble in the aqueous phase. To optimize the reaction, various phase transfer catalysts, base concentrations and solvents were tested.  1  Organic solvent CH2Cl2  Aqueous solvent 5 M NaOH  Phase transfer additive 0.4 equiv. Bu4NHSO4  2  CH2Cl2  2.5 M NaOH  3  CH2Cl2  4  Entry  Other additive  Yield  -  41%  0.4 equiv. Bu4NHSO4  -  49%  5 M NaOH  0.4 equiv. Bu4NI  -  46%  CH2Cl2  2.5 M NaOH  0.2 equiv. Bu4NHSO4  -  52%  5  CH2Cl2  1 M NaOH  0.4 equiv. Bu4NHSO4  -  62%  6  toluene  1 M NaOH  0.4 equiv. Bu4NHSO4  -  ~ 0%  7  CH2Cl2  1 M NaOH  0.4 equiv. Bu4NOH  0.1 equiv. NaI  ~ 35%  Table 2.4: Results of nitrobenzylation reaction used to produce 110b under phase transfer conditions (also see Figure 2.13). It was found that an excess of the hydroxyproline starting material 105, when reacted with freshly prepared nitroveratryl bromide 107 (0.3 equiv.) in 1 M NaOH and tetrabutylammonium 56  hydrogen sulfate produced our desired product. Additives such as NaI, or alternative phase transfer catalysts were not as effective. Switching CH2Cl2 for toluene in the reaction also gave lower yields. This represents some of the best yields obtained for the alkylation of secondary aliphatic alcohols with nitroveratryl bromide. This approach was successful for the preparation of the nitrobenzyl derivative as well. 2.2.1.4 PROTECTING GROUP MANIPULATION To convert the benzylated products into SPPS compatible monomers, a series of protecting group manipulations were carried out. The methyl ester was cleaved with an excess of LiOH, and the product was extracted into organic solvent. This was directly treated with a 50% solution of TFA in CH2Cl2. Upon thorough evaporation, the TFA-salt of the amino acid was obtained.  Boc R  N  R O  1. LiOH dioxane:H2O (2:1) 2.TFA:CH2Cl2 (1:1) O 3. Fmoc-OSu NaHCO3 OCH3 dioxane :H2O (2:1)  Fmoc O R  N  OH  R  (65 %)  O  NO2  NO2  110a/b  102a/b  a, R = H b, R= OCH3  Figure 2.15: Protecting group manipulation of light-sensitive hydroyproline derivatives 110a/b to generate an SPPS compatible monomer. The final product was obtained via reaction of the TFA salt with the commercially available Fmoc protecting reagent, fluorenylmethoxycarbonyl N-hydroxysuccinimide (Fmoc-OSu) in a 2:1 dioxane-water mixture. Upon purification by silica gel chromatography, the desired products (102a and b) were obtained with an average 65% yield, over three steps. An important observation made during these protecting group manipulations was that the nitroveratryl and nitrobenzyl protecting groups were stable through the course of the synthesis of 102a and 102b. Most importantly these groups were stable to the deprotection conditions used to remove the Boc group. This provides good evidence that the nitroveratryl and nitrobenzyl protecting groups will be stable to the conditions used to prepare amatoxins.  57  2.2.2 SYNTHESIS OF FLUORESCENT ASPARAGINE DERIVATIVE There exists a variety of fluorescently labeled peptides scattered throughout the literature.158-160 Many of these peptides were labeled post-synthetically, using an appropriate electrophilic derivative of the fluorophore. There are also synthetic amino acid analogs for the incorporation into peptides during SPPS.161-163 The inspiration for the synthesis of a diethylaminocoumarin (DEAC) appended amino acid came from a previous report where this functionality was appended to the side chain of a lysine residue.148 The fluorophore was stable to the synthetic conditions that were used, and incorporated into a peptide on solid phase. 2.2.2.1 DIETHYLAMINOCOUMARIN SYNTHESIS We chose to synthetically prepare the DEAC fluorophore based on literature reports.148 It is a stable molecule, and is easily prepared and purified. The synthesis is based on a Knoevenagel condensation, and uses readily available starting materials (Figure 2.16).  OHC +  HO  EtO2C  1. piperidine toluene, 80 °C O 2. NaOH EtOH/H2O HO CO2Et  NEt2  114  (62 %)  115  O  O  NEt2  116  Figure 2.16: Use of Knoevenagel condensation followed by lactonization and saponification to yield diethylaminocoumarin (116). Diethyl malonate was smoothly reacted with diethylaminosalicylaldehyde in the presence of piperidine to afford the ethyl ester of the desired product. Following silica gel chromatography, the ester was saponified in a water-ethanol mixture with sodium hydroxide. The desired product was cleanly isolated as an orange solid during acid precipitation in 62% yield. 2.2.2.2 PREPARATION OF A LINKER BEARING A FLUOROPHORE We chose to attach this fluorophore with a short triethylene glycol-based spacer (TEG). This linker provides good solubility in polar solvents, and is readily available. One amine of 2,2′58  (ethylenedioxy)bis(ethylamine) (117) was selectively protected with a Boc protecting group, under dilute conditions using a large excess of the diamine to yield 118. H2N  O  O  NH2  0.1 eq Boc2O CH2Cl2 Boc  H N  O  O  Boc  H N  O  115  (87 %) O HO  NH2 +  O  O  EDC NaHCO3 DMF  118  117  (73 %)  116  O  O  NEt2  N H O  O  NEt2  119 Figure 2.17: Selective protection and coupling of a triethylene glycol (TEG) linker to the fluorophore diethylaminocoumarin (DEAC). The mono-protected diamine was extracted and used without further purification. This was coupled to diethylaminocoumarin using EDC in the presence of NaHCO3, to afford derivative 119. Conjugate 119 was stored with the Boc protecting group on the amino group until directly before the compound was required, based on the ease of manipulation of 119. 2.2.2.3 COUPLING TO ASPARAGINE AND DEPROTECTION Commercially available protected aspartate residue 120 was activated at the β-carboxylate as the N-hydroxysuccinimide (NHS) ester164 to yield 121, as shown in Figure 2.18. The protected coumarin surrogate 119 was treated with trifluoroacetic acid, to remove the Boc protecting group.  Upon evaporation of the TFA, it was exposed to 121 in the presence of excess  diisopropylethylamine (DIPEA) in DMF. Following purification by silica gel chromatography, the t-butyl ester intermediate was obtained.  59  Fmoc  H N  DCC NHS EtOAc  O OtBu OH  120  Fmoc  O  H N  OtBu O O N  (87 %)  O  O O  1. 50% TFA:CH2Cl2 2. 121 DIPEA DMF 3. 50% TFA:CH2Cl2  119  121  Fmoc  H N  O OH H N  (78 %) O  O O  103  O  N H O  O  NEt2  Figure 2.18: Synthesis and coupling of DEAC to an activated aspartate residue to yield the SPPS compatible amino acid 103. Prior to incorporation into peptides, the t-butyl ester was hydrolyzed to yield the fluorescent amino acid derivative 103, which will be referred to as Fmoc-Asp(TEG-DEAC)-OH. This deprotection was achieved through exposure of the fully protected amino acid to a mixture of TFA in CH2Cl2. Once the starting material was consumed (as determined by TLC), the solvent was evaporated. The product was co-evaporated several times with toluene and ether to remove excess acid, and was used without further purification or characterization.  2.3 SOLID PHASE AMATOXIN SYNTHESIS In order to synthesize the desired probes shown in Figure 2.9, we must have a facile method of obtaining this amatoxin scaffold. We need to be able to routinely prepare cyclic peptides that contain the tryptathionine crosslink, and which can host a variety of natural and modified amino acids. Solid phase peptide synthesis (SPPS) is a very well established method that provides quick access to peptides.165 Our lab has recently improved upon the methodology to prepare amatoxins using solid-phase, and to induce tryptathionine cyclization of peptides using the Savige-Fontana reaction.76 The retrosynthetic scheme that was followed to synthesize the desired probes is provided in Figure 2.19.  60  O H N  HN O  N H  HN  O S  R1  N  N H  H N  O  O  O N H  C NH H2 Macrolactamization  R2  O  O H2 O N C H HN  H2N  HO  #  R1  R2  64 53 100a 100b 101  H OH ONv ONBn OH  H H H H TEG-DEAC  O  HN  O  O  H2 C  O  O N  N H  S N H  H N  O O C NH N H2 H  O  #  R1  R2  122 52 123a 123b 124  H OH ONv ONBn OH  H H H H TEG-DEAC  #  R1  R2  125a 125b 125c 125d 125e  H OH ONv ONBn OH  H H H H TEG-DEAC  HN R1  R2  Trt  N  O H N C N H H2 O  O  N H  Savige-Fontana cyclization  Trt S  OH HN  O  H2 C O  N H  O  H N  O H N  N  O O  O NH O  R1 = H, OH, ONv, ONBn R2 = H, PEG-DEAC  R2  R1  Figure 2.19: Retrosynthetic approach used to target the bicyclic amanitin scaffold containing natural and modified amino acids. This approach follows the previously reported synthetic approaches used to generate amatoxins.77,166 It requires the synthesis of a linear octapeptide containing the Hpi moiety and the modified amino acids 102a/b or 103 needed to furnish our targets.  2.3.1 LINEAR SOLID PHASE PEPTIDE SYNTHESIS The advent of solid-phase peptide synthesis techniques, developed by Merrifield over a quarter of a century ago, has proven to expedite the synthesis of both naturally occurring and synthetic peptides.167-169 Each step in the process has been highly optimized, and the majority of the starting materials are commercially available at low cost. This includes enantiomerically pure amino acids, in fully protected form. 61  We chose to build our peptides on the mild-acid labile 2-chlorotrityl chloride resin. This is a commercially available polystyrene based resin, with 1% divinylbenzene added as a crosslinking reagent. The resin contains a free 2-chlorotrityl chloride anchor for attachment of the C-terminus of amino acids, with an average loading of 1-2 mmol/g. Cleavage of the peptide from this resin using TFA also removes acid labile side chain protecting groups. An advantage of this resin is that the very mild acid hexafluoroisopropanol (HFIP, pKa 9.3) will remove the peptide from the resin, while maintaining most acid sensitive protecting groups. This is particularly useful for the removal of peptides that include the acid sensitive Hpi moiety. Previous work carried out in our laboratory has established that this is an efficient approach to prepare amatoxin derivatives. This was demonstrated through the synthesis of 64 and 53. Steps to prepare the required linear precursor, and the introduction of the Hpi moiety were developed when this project began. The peptides 64 and 53 were re-synthesized to gain understanding of the synthetic methodology, and also to provide important control peptides in this project. O  O H N  HN O  N H  O S  N  H N  O  N H  O N H  H2 C  HN O  HN O C NH H2  O  N  O  H N  O  N H  O N H  H2 C  O  HN O C NH H2  O  H2N  64  N H S  HO  O  H2N  H N  O  O  53  Figure 2.20: Amanitin derivatives prepared as controls, containing no modification at asparagine residue one, and either a proline or hydroxyproline at position two. The first step in this linear SPPS approach involves charging the resin with the C-terminal amino acid.  The Fmoc-protected derivative of isoleucine (126) was added to the resin using  manufacturer’s protocols. The amino acid was dissolved in dry CH2Cl2, to which was added DIPEA. The mixture was added to a dry flask containing the resin, and the mixture was stirred for thirty minutes.  The resin was filtered, and washed with a capping mixture, to block  unreacted resin positions with methanol.  62  Cl  +  Cl  Fmoc  H N  O  DIPEA CH2Cl2 OH  Fmoc  126  H N  O  Cl O  127  Figure 2.21: Charging of chlorotrityl resin with the C-terminal amino acid Fmoc-Ile-OH (the black sphere represents the polystyrene resin). The loading of the resin with isoleucine was determined by quantitation of the dibenzofulvene byproduct 129 released upon Fmoc-removal when the resin is treated with a solution of 2% DBU.170 This protocol uses DBU as opposed to the more commonly used piperidine, to avoid side reactions with 129, which leads to a lower-than-actual loading value.  Fmoc  H N  O  2% DBU DMF O  127  O H2N  O  128  + CO2  +  129  Figure 2.22: Quantitation of the loading of the chlorotrityl resin is determined by monitoring dibenzofulvene (129) production after DBU treatment (the black sphere represents the entire resin, including the 2-chlorotrityl functionality). At this point the linear peptide could easily be generated on the resin, using standard techniques. Since an automated synthesizer was not available to us, we used a manual technique, in which addition of the reagents, decoupling and washings were performed by hand. This was performed using commercially available spin columns that were fitted with a frit and a screw-cap seal. For shaking operations, the sealed tube was fixed to a standard bench-top vortexer and shaken at the lowest possible speed. Filtrations were carried out using a water-aspirator attached to a flask with an adaptor to fit the spin column.  63  A.  B.  C.  Figure 2.23: Modified laboratory apparatus used for manual solid-phase synthesis of peptides. A. Spin column, sealed with a plastic pipet tip. B. Spin column shaking on a vortexer. C. Filtration apparatus. To synthesize the peptides, the isoleucine-loaded resin was placed in the spin column, and swollen in DMF. This was to help maximize the yield by exposing all reactive sites to solvent. Following swelling, the N-terminal Fmoc protecting group was removed by washing the resin three times with 20% piperidine in DMF. Each wash lasted three minutes, and the resin was washed with solvent between deprotection cycles. Prior to coupling of the following amino acid, the resin was washed thoroughly with DMF, CH2Cl2, and DMF again. The next amino acid (Hyp) was coupled to the isoleucine residue (anchored to the resin) in DMF, with the N-terminus protected as Fmoc, and varying side chain protections. This was done using a four-fold excess of the amino acid as well as the coupling agent HBTU (130) and 8 equivalents of DIPEA. Coupling was carried out for 20-40 minutes. The progress of coupling was often monitored using a few beads of resin. This was done either by a Kaiser test, or by mass spectrometry. Following complete coupling, the resin was filtered and washed with copious amounts of DMF and CH2Cl2.  When coupling to hindered amino acids (proline or  hydroxyproline), the coupling reaction was repeated, to maximize yield and minimize sequence deletion products. When synthetic amino acids 102a/b and 103 were used, the amounts of reagents in the coupling were decreased, and the coupling times increased. In general, 1.5 equivalents of the amino acid and coupling reagent were used, with three equivalents of DIPEA.  64  This process was iteratively conducted to generate linear hexapeptides bearing the appropriate modifications at positions two and three. A.  Fmoc  H N  O  O N H  1. 20% piperidine/DMF three x 3 min 2. wash w/DMF+CH2Cl2  H2N N H  R1  3. 4 equiv. Fmoc-Xaa-OH 4 equiv. HBTU 8 equiv. DIPEA 4. wash w/DMF+CH2Cl2  O  H N  N Fmoc  O  H N  O  N H  Trt S repeat 1-4  H Fmoc N  O  H2 N C H  O  N H  O N  O H N  O  O NH  B.  O  131a-e  R2  R1  N N  .  N O  N N  130 PF6  Figure 2.24: A. Generic deprotection and coupling cycle that was used to generate linear peptides using SPPS. The black sphere represents the solid-phase. B. Structure of the coupling reagent HBTU (130). The specific substitutions of R1 and R2 are in Figure 2.29. These deprotection and coupling protocols were repeated several times, to yield five different peptide precursors. Some of these included the use of proline, hydroxyproline, or protected hydroxyproline at position two, and asparagine or fluorescent asparagine at position one. The mass of the desired hexapeptides was confirmed by low-resolution mass spectrometry upon treatment with 25% HFIP in CH2Cl2 for thirty minutes. The linear hexapeptide sequences representing the various probes are provided in Table 2.5.  65  #  Representative hexapeptide sequences  131a  Fmoc-Ile-Gly-Cys(Trt)-Asn(Trt)-Pro-Ile-resin  131b  Fmoc-Ile-Gly-Cys(Trt)-Asn(Trt)-Hyp(tBu)-Ile-resin  131c  Fmoc-Ile-Gly-Cys(Trt)-Asn(Trt)-Hyp(Nv)-Ile- resin  131d  Fmoc-Ile-Gly-Cys(Trt)-Asn(Trt)-Hyp(NBn)-Ile- resin  131e  Fmoc-Ile-Gly-Cys(Trt)-Asn(TEG-DEAC)-Hyp(tBu)-Ile- resin  Table 2.5: Linear hexapeptide precursors of amatoxins. Synthetically prepared amino acid monomers are shown in bold type. Upon confirmation of the desired hexapeptide mass by low resolution mass spectrometry, we were in a position to add the Hpi moiety to the remaining peptide in the chlorotrityl resin.  2.3.2 TRYPTATHIONINE FORMATION As was reviewed in the introduction, the synthesis of peptides containing a tryptophan crosslink is best achieved through the incorporation of an Hpi moiety. This route has proven successful both in our lab and others.78,79,171 Focus in our lab has revolved around the facile introduction of tryptathionine linkages into peptides, using an adaptation of the Savige-Fontana reaction. We have found that this is best achieved with introduction of an Hpi moiety onto the solid support as a dipeptide, with glycine as a C-terminal “protecting group” of tryptophan. To introduce the tryptathionine linkage into peptides 131a-e, we must first synthesize the Hpi-Gly dipeptide, couple it to these peptides and then induce the Savige-Fontana reaction. 2.3.2.1 SYNTHESIS OF HPI CONTAINING DIPEPTIDE For the purposes of making amatoxins and phallotoxins, we have seen that 132 (Figure 2.25) is an ideal precursor to be introduced onto the solid-phase. This monomer is generated from a suitably protected tryptophanyl-glycine dipeptide. Former graduate student Dr. Pierre Fournier first described the synthesis of 132, which is conducted in solution phase.85  66  O O  HO  NHEt3  N H O  N Trt  N H  132  Figure 2.25: The SPPS compatible monomer used for the introduction of Hpi into peptides for the generation of amatoxins. The N-terminus of L-tryptophan (133) was protected with a trityl group using excess trityl chloride under basic conditions. The reaction resulted in a bis-protected intermediate, but the indole nitrogen trityl group was removed upon workup to yield 134. This was coupled to the methyl ester of glycine to obtain the dipeptide 135. The oxidation to form Hpi was carried out using the oxidant dimethyldioxirane (DMDO)172 under anhydrous conditions.  H2N  1. Trt-Cl (2.2 equiv.) NEt3 (4 equiv.) H CHCl3:DMF (3:1) CO2H N 2. MeOH Trt 55 °C  CO2H  134  (49 %) N H  CO2Me  Trt  O  HO N H  N H  N H  135  O  HO  O  HN  N H  133  DMDO CH2Cl2 -78 °C  H Trt N  (28 %)  (57 %) N H  H-Gly-OMe EDC DIPEA CH2Cl2  +  O  N H  N N H  O  H  syn-cis  58a  Trt anti-cis  O O  58b  Figure 2.26: Synthesis of a diastereomeric mixture of the desired Tr-Hpi-Gly-OMe dipeptide 58. The product resulting from DMDO oxidation was a mixture of diastereomers (syn-cis and anticis). This is a result of the first step in the DMDO oxidation mechanism, in which an epoxide across the 2’-3’ bond of the indole is formed (Figure 2.27).  There is no expected facial  selectivity in this case, which is why no diasteroselectivity was observed. If glycine is swapped for a chiral amino acid, some diastereoselectivity can be induced.78 67  O  H N  Trt  HN  CO2CH3  Trt  O  H N  O HN  CO2CH3  O  syn-cis  anti-cis  HN  HN  136a  136b  Figure 2.27: Diastereomeric epoxides obtained during DMDO oxidation of 135 leads to the formation of the syn-cis and anti-cis isomers. The diastereomers 58a and 58b have been characterized individually, but for the purposes of amatoxin formation, this is not necessary. Upon Savige-Fontana cyclization, these diastereomers are destroyed. Therefore, procedures always used the Hpi containing dipeptides as a mixture of diastereoemers. 2.3.2.2 SAVIGE-FONTANA CYCLIZATION OF PEPTIDES The diastereomeric mixture 58 was saponified directly before coupling to the solid phase, using excess LiOH. The product was isolated as the triethylammonium salt 132 (Figure 2.28). The Hpi dipeptide 132 was introduced onto the solid phase using HBTU and DIPEA, with fewer equivalents and longer reaction times than used in the coupling of the commercial amino acids. The deprotected residue was coupled to each of the solid-phase anchored hexapeptides based on the sequences outlined in Table 2.5. O HO  O  N H N N H  O Trt  1. LiOH (10 equiv.) dioxane:H2O (2:1) 2. NEt3 work-up  O O  HO  (88 %)  N H O  N Trt  N H  58  NHEt3  132  Figure 2.28: The deprotected Hpi dipeptide 132 is suitable for introduction onto the solidphase using similar coupling conditions as other modified amino acids.  68  As mentioned several times, the critical tryptathione crosslink that is characteristic of amatoxins is obtained via the Savige-Fontana reaction.  This takes advantage of the presence of the  embedded cysteine residue and the N-terminal Hpi. Following successful coupling of the HpiGly unit onto the hexapeptides, the Savige-Fontana reaction was induced. This was achieved by the addition of neat TFA to an amount of resin in a round bottom flask. The exact amount of resin used varied batch-to-batch, and was difficult to quantify due to the added weight and trapped solvent. When possible, the mass of resin used was calculated as a fraction, to help determine yields. Trt S  OH HN Trt  N  O H N C N H H2 O  O  N H  H2 C O  O  H N  N H  N  O H N  O O  O NH  125a-e  O TFA 5h  HO  O N H  O  H2 O N C H HN S N H  H N  N  O  O O C NH N H2 H  HN R1  R2  R1  O H2N  O  R2  O  #  R1  R2  122 52 123a 123b 124  H OH ONv ONBn OH  H H H H TEG-DEAC  Figure 2.29: Savige-Fontana cyclization of the Hpi-containing octapeptides provides the desired monocyclic peptides. The resin was stirred at room temperature for five hours, after which point a red-black mixture was observed. The crude mixture was filtered through cotton wool and washed with acetonitrile. The solvent was evaporated, followed by co-evaporation with acetonitrile or toluene to aid in the removal of excess TFA. The residue was then taken up in methanol or acetonitrile and ten volumes of water were added. The precipitated by-products were removed in a centrifuge and then filtered through a 0.2 µm filter. The peptides were purified on C18 reversed-phase HPLC with an increasing gradient of acidic acetonitrile in water. The tryptathionine linkage displays a characteristic UV absorbance signature, with a λmax of 290 nm. This was used to identify the products on HPLC. 69  Figure 2.30:  Sample HPLC chromatogram of crude monocyclic peptide.  Products  containing a tryptathionine crosslink were identified by their unique UV absorbance spectrum (shown inset). The product was collected from HPLC and lyophilized. The quantity of the peptide obtained was determined by UV spectroscopic analysis.  As previously reported, peptides containing an  unoxidized (thioether) tryptathionine display a signature absorbance at 292 nm. Peptides with this functional group are assumed to have the same molar absorption co-efficient of ε290 =12600 M-1cm-1, based on previous reports.2 All isolated products were also characterized by high-resolution mass spectrometry. Characterization by 1H-NMR was performed when quantities permitted. The following table lists the monocyclic octapeptides obtained.  70  O H2 O N C H HN  H2N  O HO  O N H  O  S N H  H N  N  O O C NH N H2 H  O HN R1  R2  O  #  Name  R1  R2 H  Expected Mass -  Observed Mass -  122  Pro-monocycle  H  52  Hyp-monocycle  OH  H  -  -  123a  Nv-monocycle  ONv  H  1030.4069  1030.4050  123b  NBn-monocycle  ONBn  H  1068.4461  1068.4464  124  DEAC-monocycle  OH  (TEG-DEAC)  1247.5771  1247.5757  Table 2.6: Mass spectrometry analysis of the monocyclic peptides obtained after the Savige-Fontana reaction. Unreported values were determined in our lab by Dr. Jonathan May. Peptides obtained via this methodology were stored in the monocyclic form due to the potential toxicity of the bicyclic analogs. The macrolactamization to obtain the desired probes was carried out on small quantities, with the hopes of yielding enough peptide for biological assays.  2.3.3 MACROLACTAMIZATION Plenty of research has been dedicated to the development of efficient macrolactamization techniques.173 This reaction has proven useful in the preparation of a variety of cyclic molecules, including the synthesis of cyclic peptides.174 Cyclic peptides confer not only an element of increased binding efficiency (through ridigification), but impose resistance to protease degradation.175,176 Both straight-forward and elaborate approaches have been developed to produce these cyclic structures.  In any case, the main problems associated with 71  macrocyclization stems from entropic issues. The conditions used to impose this coupling often lead to dimerization or multimerization in place of cyclization. This problem is most often controlled through reaction at very low concentrations. In the case of amatoxins, macrolactamization generates an 18- and a 24-membered bicyclic ring structure. The preformed tryptathionine ring induces some rigidity into the peptide, which should help the second cyclization event.  Based on previous syntheses of amatoxins,  macrolactamization using a mixed anhydride method, or standard peptide coupling reagents is sufficient for achieving the desired product, with low (~ 5-15%) yields. Indeed, we found that treating monocyclic peptides with PyBOP under dilute conditions produced the best results. O H N  O H2N  O HO  O N H  O  PyBOP DIPEA DMF  S N H  H N  N  HN  H2 O N C H HN  O  O O C NH N H2 H  O  R1  R2  N  H N  O  H2 C  N H  O  O N H  O  HN  O S  R1  HN  N H  C NH H2  O  HN O  R2  122, 52, 123a/b, 124  O  64, 53, 100a/b, 101 N N N O N P N N  137 PF6  Figure 2.31: Macrolactamization of tryptathionine containing octapeptides produces the desired bicyclic amatoxin analogs. Shown inset is the structure of PyBOP. When fully cyclized probes were required for biological testing, they were marcolactamized in this fashion. The purified monocyclic peptides (Table 2.6) were dissolved in DMF in dilute conditions (1-5 mM), and three equivalents of PyBOP and six equivalents of DIPEA were added. The reaction was allowed to stir for 12-18 hours, at which point the solvent was removed in vacuo. The samples were dissolved in a 1:1 acetonitrile-water mixture, passed through a 0.2 µm 72  syringe filter and purified on reverse-phase C  18  HPLC. The products were again identified by  their distinct UV absorbance at 290 nm.  Figure 2.32: A. Sample HPLC chromatogram obtained following macrolactamization reaction used to produce amatoxins. The products were isolated by HPLC and lyophilized. spectrometry and UV absorbance.  These were characterized by mass  Typically, 50 - 500 nmol of material was isolated, as  determined by UV spectroscopy.  73  O H N  HN O  N H  O S  R1  N  H N  O  N H  O N H  H2 C  O  HN O C NH H2  O  HN R2  O  #  Name  R1  R2 H  Expected Mass 861.3694  Observed Mass 861.3680  64  Pro-bicycle  H  53  Hyp-bicycle  OH  H  877.3643  877.3629  100a  Nv-bicycle  ONv  H  1012.3965  1012.3963  100b  NBn-bicycle  ONBn  H  1072.4174  1072.4187  101  DEAC-bicycle  OH  (TEG-DEAC)  1251.5485  1251.5494  Table 2.7: Mass spectrometry analysis of the amatoxin analogs obtained after macrolactamization. As expected, the yields of peptide obtained after macrolactamization were quite low. However, this afforded enough material to conduct initial biological assays.  During the process of  cyclization, a byproduct was often observed. When formed, the byproduct was separable on HPLC, with approximately one-minute difference in retention times. 2.3.3.1 EPIMERIZATION The product and by-product formed during macrolactamization were produced in nearly equal quantities. The by-product was shown to be an isomer of the desired product, as each isomer displayed the expected characteristic UV and mass spectra. One possibility that could have lead to these isomers is an inversion of the amino acid sequence, where one (or more) of the amino acids occur out of order. This is highly unlikely due to the synthetic approach used. More likely, these isomers were suggested to be borne of one of two possibilities: 1) epimerization of the α-carbon of an amino acid or 2) basket atropisomerisation of the bicyclic ring system (see 74  Figure 2.33). If the two products reflected an epimerization of the α-carbon, it most likely occurred during the macrolactamization reaction. Had there been epimerization during the synthesis of the linear peptide, it would have been observed during the purification of the monocyclic peptides listed in Table 2.6. It is well documented that macrolactamization reactions are prone to epimerization,177,178 based on both the lifetime of the activated carboxylate and slow reaction times. The alternative explanation of these isomers is based on atropisomers. The monocyclic peptides 122, 52, 123a/b and 124 containing the tryptathionine, could have directed the formation of the second macrolactam cycle “above” or “below” the plane of the first ring. This would result in atropisomers that are related through a high-energy barrier of inversion of the tryptathionine functionality through the center of peptide cyclus (see Figure 2.33 B.).  A.  O  H N  HN O O  S N  H N  O  O  H2N  H2 O N C H HN  HN O O  epimers  S  N H O O N C NH H H2  N  H2 O N C H HN  N H O O N C NH H H2  H N  O  O  H2N  64  O  O  H N  138  O  B. O O  NH  NH S  basket atropisomers  NH  S HN  139a  139b  Figure 2.33: Possible isomeric products of 64. A. These isomers reflect an epimerization at the α-carbon of Ile3. B. These isomers cartoon the proposed basket atropisomers. Previous work was carried out in our lab to determine the identity of these isomers.166 Neither NMR nor CD spectroscopy could conclusively determine whether the isomeric products were epimers or atropisomers. Finally, it was shown through X-ray crystal structure analysis that the two isomers were, in fact, the epimers 64 and 138.  75  2.4 PROPERTIES OF SYNTHETIC PEPTIDES The synthesis of these amatoxin probes was undertaken to test our hypothesis that we could develop further understanding of the inhibition of transcription with derivatives of α-amanitin. We were also hoping to have the ability to control and visualize RNAP II within a cell. The hydroxyl functionality of Hyp2 in 1 seems critical for tight binding to RNAP II through a key hydrogen bond to Glu A822. Therefore, we propose that the blockage of this hydroxyl group (as seen in our nitrobenzyl and nitroveratryl amatoxins 100a and 100b) would eliminate the ability to make this hydrogen bond. Ideally, these protecting groups can be removed intracellularly upon exposure to light, which will allow us to induce RNAP II inhibition in a spatio-temporal fashion. We have also successfully synthesized a fluorescent amatoxin probe (101), with the hopes of providing a method of visualization of amatoxin binding within cells. Prior to testing these probes within a biological setting, we needed to verify that they provided the desired chemical characteristics. Several key elements of amatoxins can be characterized by spectroscopic techniques.  A  tryptathionine residue can be distinguished from a tryptophan, oxindole and an oxidized tryptathionine by the shape and λmax of the UV absorption spectrum.  Alternatively, CD  spectroscopy can provide qualitative information regarding the three dimensional shape of the peptide, based on literature reports of other amatoxin derivatives. It was also important to ensure that the protected hydroxyproline containing probes underwent smooth photolysis to generate the desired amatoxins.  2.4.1 ABSORPTION AND EMISSION PROPERTIES The absorption spectra of amatoxins and phallotoxins are characteristic. The tryptathionine chromophore has a unique absorption spectrum that does not overlap with any of the other residues that are found in the common amatoxins. Also, the UV spectrum of the oxidized form that is found in the natural products as a sulfoxide is distinct from the thioether form.  As  mentioned earlier in the text, it is this unique UV spectrum that is used to detect the amatoxins. Each of the amatoxin probes produced in this thesis was quantified by this UV absorbance. The UV spectrum of a solution of known volume of each peptide was recorded, and the absorbance 76  (A) at 292 nm was noted.  The molar extinction coefficient (ε) of peptides that bear an  unoxidized (thioether) tryptathionine linkage has been shown to be 12600 M-1cm-1.  The  concentration (c) of the peptide solution was determined from the Beer-Lambert equation [c = A/εl], where l is the cell pathlength, in cm. For probes that did not contain any added chromophores (53 and 64, for example), the expected UV absorption spectrum was observed.  In the case of the modified probes containing  photolabile protecting groups or fluorescent moieties, the expected absorbance was seen at 292 nm, however some extra characteristics were observed.  280 nm  290 nm  358 nm  Figure 2.34: The absorption spectra of 100a (red) and 100b (blue) in methanol. The UV spectra of probes 100a and 100b are shown in Figure 2.34. Each of these spectra show the expected peak representing the tryptathionine linkage, but also show secondary absorption peaks. There was an additional absorption band of lower intensity in the nitroveratryl derivative 100b centered at 358 nm, while the nitrobenzyl derivative 100a broadened the absorption band of the tryptathionine (at 292 nm) towards the blue region. It is important to note that there is no significant overlap of the λmax of the nitroveratryl protecting group with that of amatoxins. For quantification purposes, we assumed that there was no significant absorption overlap of these protecting groups at the characteristic 292 nm peak. While this assumption may affect the quantitiation of the peptide, it was still deemed to be most appropriate, as peptides were generally prepared on small scale. The process of lyophilization and weighing on this scale is as likely to introduce significant error. 77  The fluorescently labeled amatoxin 101 also displayed a unique absorption spectrum. The emission spectrum of the DEAC-appended aspartate was also determined, to confirm its applicability in confocal microscopy techniques (Figure 2.35).  Figure 2.35: The absorption (blue) and emission (red) spectra of 101 in methanol. The coumarin appended probe displayed the characteristic 292 nm peak of tryptathionine, but also absorbed at a λmax of 465 nm. This is an acceptable wavelength for confocal microscopy experiments equipped with lasers that emit at 450 or 488 nm. The λmax of emission of the peptide was determined to be 490 nm. These spectral properties should prove to render this probe useful in biological experiments. This amatoxin should also be useful in experiments in conjunction with other fluorescently labeled probes that emit at different wavelengths. Fluorophores such as tetramethylrhodamine (99) absorb and emit at wavelengths that should not overlap with the DEAC-labeled amatoxin.  2.4.2 CIRCULAR DICHROISM SPECTROSCOPY Each of the amatoxins synthesized was characterized by mass spectrometry and UV absorption spectroscopy. For the monocyclic derivatives 123a, 123b and 124 in Table 2.6, we were also able to obtain 1H-NMR spectra. While this helps to confirm the presence of the desired amino 78  acids, their sequence, and the presence of the bicyclic scaffold, it does not speak to the overall conformation. Previous work dedicated to the synthesis of amatoxins has shown that the toxicity of these peptides is strictly tied to their conformation.57,58 Effective methods for determining the global structure of amatoxins are achieved through X-ray crystal structure analysis or circular dichroism (CD) spectroscopy.  Events such as epimerization can be revealed through CD  spectroscopy, as we have showed previously in the determination of the isomeric products formed from macrolactamization.  Figure 2.36: This graph shows the unique CD spectra of the Ile3 epimeric amatoxins 64 and 138 obtained following macrolactamization of 122.  Dr. Jonathan May obtained these  spectra in previous work in our laboratory. These spectra show that an epimerization at Ile3 leads to an alternative conformation of the peptides. The literature states that the CD spectra of amatoxins and their toxicity are directly related.68,90,91,179 These are reflected as non-RNAP II binding conformations of the peptide backbone. The amatoxin probes synthesized here were subjected to CD spectroscopy analysis. In order to obtain decent spectra, a concentration of at least 1 mM is ideal. Due to the size of the cuvette, this required a fairly large amount of the peptide, therefore some spectra look better than others. Regardless, we attempted to obtain CD spectra of the bicyclic peptides. 79  Figure 2.37: The CD spectra obtained with amatoxin probes 100a (shown in blue), and 100b (shown in red). Peptides 53 and 64 had been previously reported in our lab so they were not retested, however, 100a and 100b showed a positive Cotton effect at ~ 260 nm. This implies that the amatoxin probes prepared adopt the conformation consistent with observed RNAP II inhibition.  2.4.3 DEPROTECTION OF NITROBENZYL AND NITROVERATRYL AMATOXINS Before we set out to test these peptides in a biological setting, we also needed to confirm the rapid deprotection of the photolabile derivatives 100a and 100b. Based on previous studies using these protecting groups, we were aware that both of these could be photolyzed at wavelengths available with a simple hand-held UV source. In fact, the nitroveratryl group has successfully been reported to undergo photoactivation in cells using a hand-held UV lamp. The efficient photodeprotection of peptides was determined by exposure to light at either 254 or 366 nm. We chose to study the kinetics of this deprotection using the monocyclic derivatives 123a and 123b, as we had larger quantities of these peptides, and were unsure of their toxicity as bicyclic derivatives. The peptides were dissolved in methanol at a given concentration, and placed in a quartz cuvette with a cap. The cell was exposed to light from a hand-held UV lamp 80  approximately 1 cm away from the window. Aliquots were taken from the sample at various times, and diluted in a water-acetonitrile mixture. These samples were directly applied to HPLC to identify the extent of deprotection. The product peak on HPLC was verified by the injection of pure standard 52. A.  B.  Figure 2.38: HPLC analysis of the photolysis of photoactivatible peptides.  A.  Photodeprotection profile following 366 nm exposure to 123a (150 µM in methanol) B. Photodeprotection profile following 254 nm exposure to 123b (67 µM in methanol). The desired photolysis product (52) was produced in both situations.  These HPLC  chromatograms show that no significant byproduct with absorbance at 292 nm (region of tryptathionine absorbance) was formed. The relative area under the curve for each of these peaks was analyzed to approximate a rate of photolysis using the hand-held lamp. While the numbers obtained provide a good guideline, they are not firm, as multiple repetitions were not performed.  81  #  Wavelength Concentration  kobs  t1/2  (min-1)  (min)  123a  366 nm  150 µM  0.07  2.44  123b  254 nm  67 µM  0.28  9.95  Table 2.8: Photolysis rate constant and half-life of amatoxins bearing a light-sensitive protection group. These were determined from the HPLC chromatograms in Figure 2.38. To ensure that this deprotection was not an artifact that is only applicable to the monocyclic system, bicyclic amatoxins 100a and 100b were exposed to similar conditions.  In this  experiment, the photolysis was not done to yield kinetic data, but instead solely to observe the formation of the desired product. The products obtained following exposure of 100a or 100b to UV light from a hand-held source were monitored by HPLC, and the final products were submitted for mass spectrometric analysis. Peptide 100a in methanol was exposed to 254 nm light in a quartz cell for thirty minutes, and peptide 100b was exposed to 366 nm light, also in methanol, for one hour. The crude reactions were directly analyzed by mass spectrometry and HPLC. The results showed that the bicyclic peptides also produce the desired product 53 upon exposure to the required wavelength of UV light. Peptides bearing the nitroveratryl group on the hydroxyproline can be used in cell-based techniques, since the wavelength of deprotection has been shown to be compatible with cell culture techniques. While peptides bearing the nitrobenzyl protected hydroxyproline are not compatible in these same conditions, they could potentially be applied to in vitro assays, or even as a purely synthetic precursor where orthogonal protection of hydroxyproline is required.  2.5 CONCLUSIONS The first goal of this project was to prepare a series of amatoxin probes that would allow us to understand, control and visualize RNAP II inhibition. We chose to prepare derivatives that contained a few mutations relative to the natural product α-amanitin. These modifications were chosen based on a combination of literature data compiled from previous SAR data and crystallographic information. We also introduced other modifications that will help us with our 82  goal of synthesizing probes of transcription. We prepared derivatives that bear light-sensitive protecting groups on hydroxyproline, or a fluorescent marker at asparagine. To prepare these derivatives we first synthesized monomers compatible with solid-phase peptide synthesis that bore the desired functionality on hydroxyproline and asparagine. Though some steps were not trivial, these derivatives were all effectively prepared and characterized. The most challenging step was the addition of a nitrobenzyl derivative to the secondary alcohol of hydroxyproline. Through a series of trials it was found that phase transfer catalysis conditions were most effective for the introduction of this style of protecting group. The Lewis-acid catalyzed approach described by Lawrence et al. (Figure 2.14) may prove to also be an effective method of making these derivatives, but was not pursued here. It was also found that freshly prepared nitroveratryl bromide provided superior results to commercially obtained material. These monomers were successfully incorporated into a linear peptide on the solid-phase, without any deviation from the standard approach. Upon introduction of an Hpi moiety onto the solidphase at the N-terminus, octapeptides bearing the cognate amino acid sequence of amatoxins were completed. The peptides were exposed to neat TFA to induce the Savige-Fontana reaction. The added functionality appeared to be stable to these conditions, as the isolated monocyclic peptides contained the desired properties. These were macrolactamized on small scale to afford the fully elaborated bicyclic amatoxin probes. The probes were characterized by UV spectroscopy and mass spectrometry, and when applicable NMR and CD spectroscopy. When the light-sensitive amatoxins were exposed to light of the appropriate wavelength, the desired products were obtained, rendering them potentially useful in a biological setting. With the effective preparation of these probes bearing the desired chemical properties, we needed to find the ideal biological system to test these peptides for their use as probes of transcription.  83  2.6 EXPERIMENTAL SECTION 2.6.1 GENERAL METHODS Commercially available chemicals were purchased from Sigma-Aldrich or Novabiochem. Solvents were purchased form Fisher Scientific and used without further purification unless otherwise noted. When required, solvents were dried as follows: THF was dried over Na, with benzophenone indicator; CH2Cl2 was dried over CaH2; DMF was dried over 3Å molecular sieves (activated 24 h at 110 °C). Deuterated solvents for NMR were purchased from Cambridge Isotope Laboratories. TLC analysis was performed on aluminium-backed silica gel-60 plates from EMD Chemicals. Flash chromatography was carried out on SiliaFlash F60 (230 – 400 mesh) from SiliCycle. Low-resolution ESI mass spectrometry was performed on a Waters ZQ with a single quadrupole detector, attached to a Waters 2695 HPLC. High-resolution ESI mass spectra were obtained on a Waters-Micromass LCT with a time-of-flight (TOF) detector. All NMR were recorded on Bruker Avance instruments at room temperature, with results reported as chemical shift (δ) in ppm. The UV spectra reported were all obtained on a Beckman Coulter DU800 in 1 cm quartz cuvettes.  CD spectroscopy was performed on a Jasco J-710 in 1 mm quartz cuvettes.  Fluorescence measurements were carried out on a PTI Quantmaster. For photolysis of peptides, a hand-held UVGL-58 Mineralight UV lamp from UVP equipped with 254 nm and 366 nm filters was used.  2.6.2 HPLC PURIFICATION METHODS All HPLC chromatograms were obtained on an Agilent 1100 system equipped with an auto injector, a fraction collector and a diode array detector. Analytical injections were performed on either an Agilent Eclipse XDB C-18 (4.6 x 250 mm) column (Column I) or a Phenomenex Jupiter 10U C-18 300A (4.6 x 250 mm) column (Column II) with a flow rate of 1 mL/min. Semi-preparative injections were performed on a Agilent Eclipse XDB C-18 (9.4 x 250 mm) column (Column III), with a flow rate of 5 mL/min. All columns were fit with a column guard. Chromatograms were obtained with a solvent gradient of 0.1% TFA in water (Solvent A) and  84  0.05% TFA in acetonitrile (Solvent B). Four different gradients were used, dependent on the nature of the peptide. These are shown graphically in Figure 2.39. Gradient System A:  Gradient System B:  Gradient System C:  Gradient System D:  Figure 2.39: Gradient profiles used for HPLC analysis and purification of peptides.  85  2.6.3 SYNTHETIC PROTOCOLS trans-L-hydroxyproline methyl ester hydrochloride salt (104) Cl H2 N  O O  HO  In a 100 mL flask, MeOH (25 mL) was cooled on an ice-water bath. To the cooled solvent was added SOCl2 (4 mL, 56 mmol) over a 15-minute period. After ten minutes of stirring, trans-Lhydroxyproline (4.0 g, 30 mmol) was added at once, and the reaction was left at 0 °C for 5 minutes. The reaction was then fitted with a reflux condenser and heated to 55 °C for 1.5 h. The reaction was monitored by TLC (CH2Cl2:MeOH, 3:2) and visualized by ninhydrin.  Upon  completion, the mixture was cooled back to room temperature and the solvent was evaporated. The residue was taken up in MeOH (15 mL) and re-evaporated. This process was repeated three times. The colourless residue was crystallized from a mixture of MeOH and Et2O. The product was isolated as a white solid (4.1g, 73% yield). ESI-HRMS (C6H11NO3H)+: 416.0819 (observed), 416.0817 (expected) 1  H-NMR (300 MHz, d6-DMSO) δ: 10.01-9.87 (br s, 2H, NH2), 5.62 (br s, 1H, OH), 4.47-4.40  (m, 2H, α-CH, γ-CH), 3.73 (s, 3H, OCH3), 3.35 (dd, 1H, J=12.0, 4.4 Hz, δ-CH), 3.05 (d, 1H, J=12.0 Hz, δ-CH), 2.22-2.02 (m, 2H, β-CH2). 13  C-NMR (75 MHz, d 6-DMSO) δ: 169.0, 68.4, 57.4, 53.0, 36.9.  N-t-butyloxycarbonyl-trans-L-hydroxyproline methyl ester (105) Boc O N O HO  The HCl salt 104 (1.0 g, 5.5 mmol) was added to a dioxane:water mixture (1:1, 55 mL) and was cooled on ice. Solid NaHCO3 (0.6 g, 7 mmol) was added in two portions over 15 minutes, at which point di-t-butyl dicarbonate (2.6 g, 6 mmol) was added over 5 minutes. The reaction was warmed to room temperature and stirred for 1.5 h. When TLC (9:1 CH2Cl2:MeOH) indicated 86  that the reaction was complete, the dioxane was evaporated, and the remaining aqueous slurry was extracted four times with CH2Cl2 (30 mL). The organic solvent was dried on Na2SO4 and evaporated to yield 105 as a colourless oil (1.16 g, 86%). ESI-HRMS (C11H19NO5Na)+: 268.1159 (observed), 268.1161 (expected) 1  H-NMR (300 MHz, CDCl3) shows a mixture of carbamate rotamers, δ: 4.43-4.34 (m, 2H), 3.7  (s, 3H, OCH3), 3.61-3.31 (m, 3H), 2.31-2.20 (m, 1H), 2.08-1.95 (m, 1H), 1.44-1.36 (m, 9H, Boc). 13  C-NMR (75 MHz, CDCl3) with rotamer shifts in brackets, δ: 173.8 (173.6), 154.1 (154.7),  80.5 (80.3), 69.9 (69.2), 58.0 (57.6), 54.7 (54.7), 52.3 (52.1), 39.1 (38.4), 28.3 (28.4). Nitroveratryl bromide (107) Br NO2 H3CO OCH3  To a stirred solution of triphenylphosphine (3.62 g, 13.8 mmol) in dry CH2Cl2 (140 mL) at room temperature, was added Br2 (0.71 mL, 13.8 mmol) dropwise over 5 minutes. The light orange solution was allowed to stir under a stream of N2 for another 10 minutes, at which point, dimethoxynitrobenzyl alcohol (2.94 g, 13.8 mmol) was added at once. The reaction was allowed to proceed for two hours at room temperature, at which point the contents were transferred to a separatory funnel containing saturated NaHSO3 (50 mL), and the organic phase was separated. The organic phase was washed once with H2O (50 mL), and the combined aqueous phases were then extracted three times with CH2Cl2 (50 mL). All organic extractions were combined and dried over anhydrous Na2SO4. The crude orange solid was purified on silica gel chromatography (7:2 hexanes:EtOAc) to yield a light yellow solid (3.25 g, 85%). The product was found to decompose over time, and was best used immediately, and was best stored in the dark. LR-ESI (C9H10NBrO4H)+: 274.9, 276.9 (observed), 275.9, 277.9 (expected) 1  H-NMR (400 MHz ,CDCl3) δ: 7.67 (s, 1H, Ar-H), 6.95 (s, 1H, Ar-H), 4.87 (s, 2H, Ar-CH2),  4.00 (s, 3H, OCH3), 3.96 (s, 3H, OCH3). 13  C-NMR (100 MHz ,CDCl3) δ: 153.4, 149.1, 140.4, 127.6, 113.8, 108.7, 56.7, 56.6, 30.2. 87  N-t-butyloxycarbonyl-O-nitrobenzyl-trans-L-hydroxyproline methyl ester (110a) N-t-butyloxycarbonyl-O-nitroveratryl-trans-L-hydroxyproline methyl ester (110b) Boc O N O  O2N  OCH3 H3CO  Boc N  O O  O  O O2N  a) NaH method A flame dry 25 mL flask was charged with a dispersion of 60% NaH on mineral oil (46.4 mg, 1.16 mmol). The powder was washed with hexanes and dried under high vacuum. In a separate 25 mL flask, 105 (189 mg, 0.77 mmol) was dissolved in dry DMF (7.7 mL) and cooled to 0 °C on an ice-water bath, and stirred under a steady flow of nitrogen. The dried NaH was added at once, and the reaction was allowed to stir at 0 – 6 °C for one hour. At this time, nitrobenzyl bromide (250 mg, 1.16 mmol) was added at once, and the reaction immediately turned purple, which then quickly disappeared. After stirring for several hours, no product was observed by TLC. b) Ag2O method In a 20 mL vial, 105 (213 mg, 0.87 mmol) was added to CH2Cl2 (2 mL). Freshly prepared Ag2O (0.81g, 3.5 mmol) was added to the solution, giving a dark brown mixture. Finally, 107 (240 mg, 0.87 mmol) was added. The reaction was sealed, and allowed to stir at room temperature for 72 hours.  The product was isolated via filtration of the Ag2O followed by silica gel  chromatography (EtOAc-hexanes, 1:1). Average yields as approximated by TLC analysis were 10-25% and average isolated yields were 8-15%. c) Phase transfer method This is a representative protocol, which was applied for the preparation of both 110a and 110b. The suitably protected hydroxyproline derivative 105 (5.4 g, 19.6 mmol) was dissolved in CH2Cl2 (50 mL). This was added to a solution containing Bu4NHSO4 (0.82 g, 2.4 mmol) in CH2Cl2 (50 mL) and 2.5 M NaOH (50 mL), and was shielded from external light sources. A solution of freshly prepared 107 (2 g, 8.2 mmol) in CH2Cl2 (50 mL) was added, and the reaction was vigorously stirred for 2.5 hours. The layers were separated in a separatory funnel, and the 88  aqueous phase was extracted with three times with CH2Cl2 (100 mL). The combined organic layers were washed with brine and dried over anhydrous Na2SO4. The desired product 110b was isolated (2.06 g, 57%) as a light orange foam by silica gel chromatography (hexanes:EtOAc, 2:1). 110a: (49% yield) HRMS-ESI (C18H24N2O7Na)+: 403.1492 (observed), 403.1481 (expected) 1  H-NMR (300 MHz, CD2Cl2) δ: 8.04 (d, 1H, J=8.1 Hz), 7.71 (d, 1H, J=7.8 Hz), 7.64 (t, 1H,  J=7.8 Hz), 7.44 (1H, t, J=7.7 Hz), 4.90-4.80 (m, 2H. Ar-CH2), 4.38.4.30 (m, 1H, α−H), 4.244.20 (m, 1H, γ−H), 3.69 (s, 3H, OCH3), 3.64-3.53 (m, 2H, δ−H), 2.48-2.35 (m, 1H, β−H), 2.08 (ddd, 1H, J=13.3, 7.9, 5.1 Hz, β−H), 1.42-1.35 (m, 9H, Boc). 13  C-NMR (75 MHz, CD2Cl2) δ: 173.8, 173.5, 134.0, 129.1, 128.6, 124.9, 80.3, 78.3, 68.1, 58.3,  52.3, 51.8, 36.8, 28.3. 110b: (57% yield) HRMS-ESI (C20H28N2O9Na)+: 463.1695 (observed), 463.1693 (expected) 1  H-NMR (300 MHz, CD2Cl2) δ: 7.67 (s, 1H), 7.19 (s, 1H), 4.93-4.83 (m, 2H, Ar-CH2), 4.43-  4.31 (m, 1H, α−Η), 4.28-4.25 (m, 1H, γ−H), 3.93 (s, 3H, OCH3), 3.89 (s, 3H, OCH3), 3.723.56 (m, 5H), 2.47-2.44 (m, 1H, β−H), 2.16-2.06 (m, 1H, β−H), 1.43-1.32 (m, 9H, Boc). 13  C-NMR (75 MHz, CD2Cl2) δ: 154.0, 80.3, 78.4, 77.6, 68.3, 58.5, 58.2, 56.6, 53.4, 52.2, 36.8,  36.1, 28.3. O-nitrobenzyl-trans-L-hydroxyproline (140a) O-nitroveratryl-trans-L-hydroxyproline (140b) H2 N  O2 N  O  O OCH3  OH O O  H3CO CF3  H2 N  O  O OH O O  CF3  O2N  This is a representative protocol, used for the deprotection of both derivatives 110a and 110b. Compound 110b (400 mg, 0.91 mmol) was dissolved in a dioxane-water mixture (2:1, 9 mL), and stirred at room temperature. Excess LiOH•H2O (360 mg, 10 eq) was added and the reaction stirred for 1.25 h. Water (10 mL) was added to the reaction, and was extracted three times with 89  CH2Cl2 (15 mL). The aqueous phase was acidified with 1M HCl to pH 1, and extracted three times with EtOAc (15 mL).  The organic phase was dried over anhydrous Na2SO4 and  evaporated to yield a clear oil. The oil was dissolved in CH2Cl2 (10 mL) and TFA (3 mL) was added and the reaction was stirred for 1 h at room temperature.  The crude reaction was  evaporated, and co-evaporated three times with toluene (3 mL) and twice with Et2O (3 mL). The product was dried under reduced pressure, to yield the TFA salt 140a as a light yellow solid (320 mg, 80%). 140a (73% yield) HRMS-ESI (C12H14N2O5H)+: 267.0985 (observed), 267.0981 (expected) 1  H-NMR (300 MHz, d4-MeOH) δ: 8.04 (dd, 1H, J=8.2, 0.9 Hz), 7.77-7.68 (m, 2H), 7.55 (dt, 1H,  J=7.6, 1.6 Hz), 4.94 (q, 2H, J=13.4 Hz, Ar-CH2), 4.53-4.47 (m, 2H), 3.58-3.47 (m, 2H, δ-CH2), 2.74 (ddt, 1H, J=14.2, 7.6, 1.3 Hz, β-H), 2.29 (ddd, 1H, J=14.4, 10.5, 4.1, β-H). 13  C-NMR (75 MHz, d4-MeOH) δ: 171.0, 149.0, 138.8, 134.5, 130.5, 129.8, 125.5, 79.2, 68.9,  59.4, 52.1, 35.5. 140b (80% yield) HRMS-ESI (C14H18N2O7H)+: 327.1182 (observed), 327.1192 (expected) 1  H-NMR (300 MHz, d4-MeOH) δ: 10.15-9.76 (br s, 1H), 9.16-8.70 (br s, 1H), 7.68 (s, 1H), 7.26  (s, 1H), 4.85 (q, 2H, J=13.0 Hz, Ar-CH2), 4.47-4.40 (m, 2H), 3.92 (s, 3H, OCH3), 3.86 (s, 3H, OCH3), 3.48-3.39 (m, 1H, β-H), 2.16 (ddd, 1H, J=14.2, 10.5, 4.1 Hz, β-CH). 13  C-NMR (75 MHz, d4-MeOH) δ: 170.1, 153.2, 147.5, 139.4, 128.6, 110.8, 108.0, 77.2, 67.0,  57.8, 56.2, 56.2, 50.7, 34.0. N-fluorenylmethoxycarbonyl-O-nitrobenzyl-trans-L-hydroxyproline (102a) N-fluorenylmethoxycarbonyl-O-nitroveratryl-trans-L-hydroxyproline (102b) Fmoc N  O2N  O  O OH  H3CO  Fmoc OCH3 N  O OH  O O2N  Representative protocol: The protected TFA salt of hydroxyproline derivative 140b (300 mg, 0.68 mmol) was dissolved in a dioxane-water mixture (2:1, 7 mL) and NaHCO3 (126 mg, 1.5 90  mmol) was added at once. The reaction was stirred at room temperature for 10 min, followed by the addition of Fmoc-OSu (250 mg, 0.75 mmol). The reaction was initially cloudy, but clarified over the period of 1 h, and was determined to be complete after 2 h. Water (10 mL) was added to the reaction, which was then extracted three times with EtOAc (10 mL). The EtOAc was back extracted twice with water (10 mL), and all aqueous extractions were pooled, and acidified with 1 M HCl to a pH of 1. This was extracted five times with CH2Cl2 (15 mL), and the organic phases dried over anhydrous Na2SO4 to yield 102b. 102a (84% yield) HRMS-ESI (C27H24N2O7Na)+: 511.1476 (observed), 511.1481 (expected) 1  H-NMR (400 MHz, d6-DMSO, as a mixture of Fmoc-rotamers) δ: 8.03 (t, 1H, J=8.3 Hz), 7.90-  7.85 (m, 2H), 7.78-7.54 (m, 5H), 7.42-7.36 (m, 2H), 7.31-7.25 (m, 2H), 4.90-4.76 (m, 2H), 4.424.13 (m, 5H), 3.64-3.47 (m, 2H), 2.53-2.38 (m, 1H), 2.17-1.98 (m, 1H). 13  C-NMR (100 MHz, d6-DMSO, as a mixture of Fmoc-rotamers) δ: 173.9, 154.6, 148.3, 144.4  (144.3), 141.1 (141.2), 134.3, 129.8 (129.7), 129.3, 128.3 (128.2), 127.8 (127.7), 125.9, 125.7 (125.6), 125.1 (125.0), 120.7 (120.7), 78.0 (77.1), 67.8, 67.5 (67.4), 67.2, 58.4 (58.0), 52.7 (52.0), 47.2 (47.1). 102b (312 mg, 82% yield) HRMS-ESI (C29H28N2O9Na)+: 571.1710 (observed), 571.1693 (expected) 1  H-NMR (300 MHz, d6-DMSO, as a mixture of Fmoc-rotamers) δ: 7.90-7.83 (m, 2H), 7.69-7.56  (m, 3H), 7.43-7.18 (m, 5H), 4.89-4.76 (m, 2H), 4.48-4.10 (m, 5H), 3.87-3.79 (m, 6H), 3.66-3.46 (m, 2H), 2.53-2.37 (m, 1H), 2.20-1.99 (m, 1H). 13  C-NMR (75 MHz, d6-DMSO, as a mixture of Fmoc-rotamers) δ: 174.5 (174.0), 154.7 (154.6),  148.1 (148.0), 144.4 (144.1), 141.3 (141.2), 140.2 (140.0), 139.9 (138.0), 129.9, 129.6 (129.5), 128.3 (128.2), 127.9 (127.8), 127.7 (127.6), 125.9 (125.9), 125.7 (125.6), 122.0, 130.7 (120.6), 111.3, 110.9, 110.4, 108.7, 78.1, 77.2, 67.8 (67.8), 67.6 (67.3), 58.5 (58.2), 56.7, 52.7 (52.0), 47.2 (47.1), 36.8 (35.2).  91  7-diethylamino-coumarin-3-carboxylic acid (116) O OH Et2N  O  O  A 250 mL round bottom flask was charged with diethylaminosalicylaldehyde (114) (2.66 g, 13.8 mmol) and a CH3CN:toluene mixture (2:1, 150 mL).  The mixture was stirred, and then  diethylmalonate (3.12 mL, 20.7 mmol) and piperidine (4.1 mL, 41.5 mmol) were added at once. The reaction was heated to reflux for 8 hours, until the aldehyde was no longer observed during routine TLC analysis (EtOAc:hexanes, 1:1).  The mixture was evaporated on a rotary  evaporator, and run through a silica gel plug (EtOAc:hexanes, 11:7) to isolate the ethyl ester of the desired product as an orange oil. The oil was dissolved in EtOH (150 mL) and heated to reflux, at which point 3 M NaOH (9 mL) was added. During the course of the reaction, a large amount of yellow-orange precipitate formed. The precipitate was filtered, but the filtrate showed a large amount of starting material. Solid NaOH (100 mg) was added to the filtrate, and the reaction was stirred at 65 °C for 2 hours, until no more starting material was observed. The reaction was cooled to room temperature and acidified with 6 N HCl to pH 1, at which point orange precipitate formed. This was filtered and combined with the rest of the precipitate. The combined precipitates were dissolved in CH2Cl2 and dried over anhydrous Na2SO4. The product was isolated via evaporation, to yield 116 a bright orange solid (2.24 g, 62%). HRMS-ESI (C14H14NO4)-: 260.0917 (observed), 260.0923 (expected) 1  H-NMR (300 MHz, CDCl3) δ: 8.64 (s, 1H), 7.45 (d, 1H, J=9.1 Hz), 6.73 (dd, 1H, J=9.1, 2.5  Hz), 6.52 (d, 1H, J=2.3 Hz), 3.49 (q, 4H, J=7.2), 1.26 (t, 6H, J=7.2 Hz). 13  C-NMR (75 MHz, CDCl3) δ: 165.7, 164.6, 158.2, 153.9, 150.4, 132.1, 111.0, 108.7, 105.8,  97.0, 45.5, 12.5.  92  N-t-butyloxycarbonyl-2,2′-(ethylenedioxy)bis(ethylamine) (118) Boc  N H  O  NH2  O  To a solution of diamine 2,2′-(ethylenedioxy)bis(ethylamine) (13.6 g, 92 mmol) in CH2Cl2 (272 mL) at room temperature was slowly added di-t-butyldicarbonate (2.0 g, 9.2 mmol) in CH2Cl2 (92 mL) via a dropper funnel over 4 hours. The mixture was stirred at room temperature overnight. The following morning, the reaction was concentrated to 100 mL and extracted with H2O (100 mL) and with saturated NaHCO3 (100 mL). The organic layer was dried over anhydrous Na2SO4 to yield 118 as a colourless oil (2.5 g, 87%). 1  H-NMR (300 MHz, d4-MeOH) δ: 3.62 (br s, 4H), 3.52 (td, 4H, J=5.49, 2.8 Hz), 3.23 (t, 2H,  J=5.7), 2.79 (t, 2H, J=5.3 Hz), 1.44 (s, 9H, Boc). 13  C-NMR (75 MHz, d4-MeOH) δ: 158.5, 80.1, 73.7, 71.4, 71.4, 71.1, 42.2, 41.4, 28.9.  Boc-TEG-DEAC (119) O Boc  N H  O  O  O  NEt2  H N O  In a 100 mL round bottom flask containing DMF (24 mL), was added mono-protected amine 118 (0.50 g, 2 mmol), and stirred at room temperature. Next, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (0.46 g, 2.4 mmol), hydroxybenzotriazole (324 mg, 2.4 mmol), and NEt3 (0.7 mL, 4.8 mmol) were all added at once. Finally, diethylamino coumarin 116 (0.52 g, 2 mmol) was added.  The entire mixture was stirred overnight at room temperature.  The DMF was  evaporated, followed by co-evaporation two times with toluene (10 mL) and twice with CH2Cl2 (10 mL). The residue was purified on silica gel chromatography using MeOH:CH2Cl2 (4:96) to yield pure 119 (0.73 g, 73%).  93  HRMS-ESI (C25H37N3O7Na)+: 514.2540 (observed), 514.2529 (expected) 1  H-NMR (300 MHz, CDCl3) δ: 9.03 (br s, 1H), 8.68 (s, 1H), 7.41 (d, 1H, J=8.9 Hz), 6.63 (dd,  1H, J=8.9, 2.4 Hz), 6.48 (d, 1H, J=2.3 Hz), 5.21 (br s, 1H), 3.68-3.63 (m, 8H), 3.56 (t, 2H, J=5.1 Hz), 3.44 (q, 4H, J=7.1 Hz), 3.35 – 3.30 (m, 2H), 1.42 (s, 9H, Boc), 1.23 (t, 6H, J=7.1 Hz). 13  C-NMR (75 MHz, CDCl3) δ: 163.4, 162.7, 157.8, 152.6, 148.1, 131.2, 110.5, 110.0, 96.7,  70.6, 70.4, 69.9, 45.2, 39.5, 28.5, 12.5. N-fluorenylmethoxycarbonyl-γ-N-hydroxysuccinimidyl-L-aspartate-t-butyl ester (121) Fmoc O O N  H N  O O  tBu  O O  Solid Fmoc-L-aspartate t-butyl ester (1.0 g, 2.4 mmol) was dissolved in EtOAc (24 mL), to which was added dicyclohexylcarbodiimide (0.55 g, 2.6 mmol) and N-hydroxysuccininmide (0.28 g, 2.4 mmol). The reaction was stirred at room temperature for 1.5 h, during which point a white precipitate formed. The precipitated dicyclohexylurea was filtered, and the filtrate was concentrated and purified on silica gel (hexanes:EtOAc, 2:1) to yield the product 121 as a white foam (1.1 g, 87%). 1  H-NMR (300 MHz, CDCl3) δ: 7.79 (d, 2H, J=7.4 Hz) 7.64 (d, 2H, J=7.3 Hz), 7.42 (t, 2H, J=7.3  Hz), 7.34 (t, 2H, J=7.8 Hz), 5.93 (d, 1H, J=7.8 Hz), 4.68 (dt, 1H, J=8.0, 4.1 Hz), 4.42-4.35 (m, 2H), 4.28 (t, 1H, J=7.2 Hz, α-CH), 3.28 (ddd, 2H, J=17.4, 8.3, 4.6 Hz, β-CH2), 2.86 (s, 4H, NHS), 1.49 (s, 9H, OtBu). 13  C-NMR (75 MHz, CDCl3) δ: 168.8, 168.5, 166.4, 156.0, 143.9, 141.4, 127.8, 127.2, 125.4,  125.4, 120.1, 83.7, 67.6, 50.7, 47.2, 34.3, 34.0, 27.9, 25.7, 25.0.  94  Fmoc-Asp(TEG-DEAC)-OtBu (103)  Fmoc O  H N  HN  O O  tBu  O  O O  O  NEt2  H N O  The Boc-protected diethylaminocoumarin derivative 119 (0.29 g, 0.59 mmol) was dissolved in CH2Cl2:TFA (1:1, 6 mL) and stirred at room temperature until the starting material disappeared by TLC analysis (95:5, CH2Cl2:MeOH). The solvent was evaporated, and then co-evaporated with toluene (2 mL) and Et2O (2 mL) to remove excess TFA. The residue was taken up in dioxane:water (1:1, 6 mL) to which was added 121 (0.30 g, 0.6 mmol) and DIPEA (0.3 mL, 1.8 mmol). The reaction was stirred at room temperature for 2 hours, at which point TLC indicated reaction completion.  The dioxane was evaporated, and the remaining aqueous phase was  extracted three times with CH2Cl2 (20 mL each). The combined organic phases were dried over anhydrous Na2SO4 and evaporated to yield an orange oil. The oil was purified by silica gel chromatography (5-10% MeOH in CH2Cl2) to yield the product as a yellow foam (0.36 g, 78%). HRMS-ESI (C43H52N4O10Na)+: 807.3569 (observed), 805.3581 (expected) 1  H-NMR (300 MHz, CDCl3) δ: 9.15 (br s, 1H), 8.74 (s, 1H), 7.75 (d, 2H, J=7.5 Hz), 7.62 (d, 2H,  J=7.4 Hz), 7.47-7.29 (m, 5H), 7.16 (br s, 1H), 6.66-6.59 (m, 2H), 6.28 (d, 1H, J=8.7 Hz), 4.56 (dt, 1H, J=8.7, 4.40 Hz), 4.42-4.20 (m, 3H), 3.74 – 3.52 (m, 12H), 3.44 (q, 9H, J=7.1 Hz), 3.142.83 (m, 2H), 1.48 (s, 9H), 1.23 (t, 6H J=7.1 Hz). 13  C-NMR (75 MHz, CDCl3) δ: 170.4, 170.3, 163.2, 163.0, 162.6, 157.5, 156.2, 152.8, 148.4,  144.0, 143.9, 131.2, 127.1, 125.3, 125.2, 119.9, 110.1, 109.9, 96.6, 81.9, 70.4, 70.4, 70.2, 69.6, 67.1, 51.6, 47.2, 45.0, 37.8, 36.5, 31.5, 28.0, 12.5.  Nα -trityl-L-tryptophan (134) Trt  H N  CO2H  N H  To a 250 mL round bottom flask containing CHCl3:DMF (65 mL, 2:1) was added L-tryptophan (2 g, 9.67 mmol). Trityl chloride (5.95 g, 21.3 mmol) was then added to the suspension and the 95  mixture was stirred at room temperature for 45 min, over which time, ~95% of the tryptophan entered solution. Triethylamine (5.4 mL, 38.7 mmol) was added in small (~1 mL) portions over a 15-minute period, such that no change in temperature was observed. The reaction was stirred for 3.5 h at room temperature, at which point MeOH (50 mL) was added, and the solution became cloudy. The reaction was stirred at 55-60 °C for 4.5 h further, to remove the indole trityl group. During this time the reaction turned orange and clarified. The whole mixture was evaporated to a brown sludge, and partitioned across Et2O (100 mL) and 5% aqueous citrate (100 mL). The organic layer was continuously washed with portions of 5% citrate (50 mL), until no more solid persisted. The ether layer was dried on Na2SO4, and concentrated to 50 mL. The title product was precipitated from ether as the diethylammonium salt via the addition of 4 mL HNEt2, and leaving the mixture at room temperature overnight. The precipitated salt was filtered and converted to its free acid via extraction in CH2Cl2 against 5% citrate. The CH2Cl2 layer was washed with brine, dried on Na2SO4 and evaporated to yield the desired product as a tan foam, yielding 2.44 g (5.47 mmol, 57% yield). ESI-HRMS (C30H26N2O4Na)+: 469.1889 (observed), 469.1892 (expected) 1  H-NMR (300 MHz, d6-DMSO) δ: 10.69 (br s, 1H, NH), 7.41-7.33 (m, 7H), 7.28 (d, 1H, J=8.1  Hz), 7.22-7.11 (m, 9H), 7.07 (d, 1H, J=2.0 Hz), 6.99 (t, 1H, J=7.5 Hz), 6.87 (t, 1H, J=7.4 Hz), 3.25 (t, 1H, J = 5.2 Hz, α-H), 2.80 (dd, 2H, J=18.3, 4.4 Hz, β-CH2), 2.60 (q, 4H, J=7.2 Hz), 1.00 (t, 6H, J=7.2 Hz). 13  C-NMR (75 MHz, d 6-DMSO) δ: 176.5, 146.8, 135.8, 128.7, 128.2, 127.5, 126.0, 123.7, 120.3,  118.9, 117.7, 111.3, 110.9, 71.0, 57.5, 41.8, 30.1. Nα -trityl-L-tryptophanyl glycine methyl ester (135) Trt  H N  O N H  O O  HN  To an oven dry 100 mL flask containing dry CH2Cl2 (50 mL), Nα-trityl-tryptophan 134 (3 g, 5.1 mmol) was added. The reaction was stirred at room temperature, and triethylamine (1.6 mL, 11.2 mmol), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (1.08 g, 5.6 mmol) and hydroxybenzotriazole (0.76 g, 5.6 mmol) were all added at once. Finally, the HCl salt of H-Gly96  OMe (0.71 g, 5.6 mmol) was added, and the reaction was stirred at room temperature. The reaction was monitored by TLC (95:5 CH2Cl2:MeOH), and was stopped after the tryptophan starting material disappeared (~2 h). The reaction was transferred to a separatory funnel and washed two times with 5% aqueous citrate (15 mL), two times with saturated NaHCO3 (15 mL) and twice with brine (15 mL). The solvent was dried on Na2SO4 and evaporated. The residue was purified by silica gel chromatography using 45% EtOAc in hexanes, to yield 0.75 g of a tan solid (1.45 mmol, 28%). ESI-HRMS (C33H31N3O3H)+: 518.2072 (observed), 518.2080 (expected) 1  H-NMR (300 MHz, CDCl3) δ: 8.15 (br s, 1H, NH), 7.55 (d, 1H, J=7.9 Hz), 7.38-7.29 (m, 9H),  7.23-7.20 (m, 7H), 7.09 (t, 1H, J=7.5), 7.00 (t, 1H, J=5.3), 6.93 (d, 1H, J=2.1 Hz), 3.70 (s, 3H, OCH3), 3.63-3.59 (m, 3H), 3.15 (dd, 1H, J=14.5, 6.1 Hz, β-CH2), 2.87 (br s, 1H, NH), 2.59 (dd, 1H, J=14.5, 5.6 Hz, β-CH2). 13  C-NMR (75 MHz, CDCl3) δ: 175.4, 170.1, 145.6, 136.4, 128.9, 128.0, 127.8, 126.7, 123.7,  12.4, 119.7, 119.6, 11.1, 110.9, 71.6, 57.8, 52.2, 41.0, 30.1. Dimethyldioxirane O  O  Dimethyldioxirane (DMDO) was prepared following literature protocol.172 Briefly, a threenecked 2 L flask was charged with H2O (254 mL), acetone (192 mL) and NaHCO3 (58 g). The mixture was stirred on an ice-water bath at 5 °C.  To the flask was added a distillation arm  adaptor, which was fixed with a cold finger and a clean, dry 100 mL round bottom collection flask. The cold finger was chilled to -78 °C with a dry ice/acetone mixture, and the collection flask was also cooled to -78 °C. Oxone was added to the 2 L flask in 5 portions (24 g each), 3 minutes apart. After the final oxone addition, the system was sealed, and vacuum was applied via the cold finger. The reaction was slowly warmed to room temperature through removal of the ice bath by siphoning. The vacuum was maintained at a pressure to ensure that there was no bumping (generally 45-60 mmHg). The distillate was collected as a light yellow liquid, until the rate of collection slowed to only a few drops per minute. The collected distillate was dried on K2CO3 and filtered into a dry flask containing 3 Å molecular sieves, and stored at -20 °C. The 97  collected distillate containing DMDO was used directly without titration. Typically 50-75 mL of the mixture were collected. The DMDO mixture was not used after a storage period of two weeks. 3a-hydroxy-1-trityl-1,2,3,3a,8,8a-hexahydropyrrolo[2,3-b]indole-2-carboxamido glycine methyl ester (58) Trt N HN  O N H  O O  OH  The oxidation of 135 to yield the title product was performed using un-titrated DMDO. The amounts of DMDO required varied from batch to batch.  The reaction was consistently  monitored after 1 mL additions of DMDO to determine reaction progress. A flame dried 100 mL flask was charged with 137 (140 mg, 0.27 mmol) and dissolved in dry CH2Cl2 (27 mL). A stream of nitrogen was applied, and the solution was cooled to -78 °C. Fresh DMDO was added in 1 mL portions every 15 minutes until less than 5% starting material remained, as observed by TLC (2:1 EtOAc:hexanes). When the reaction was complete, the solvent was evaporated, and the residue directly applied to silica gel chromatography. The product was purified using a gradient of hexanes and EtOAc (from 2:1 to 1:1), which yielded 71.2 mg 58 as a mixture of syncis and anti-cis diastereomers (0.13 mmol, 49%). ESI-HRMS (C33H31N3O4Na)+: 556.2211 (observed), 556.2212 (expected) 1  H-NMR (300 MHz, CDCl3) δ: 9.19 (dd, 1H, J=6.8, 2.9 Hz), 7.72 (d, 6H, J=7.7 Hz), 7.54 (d,  6H, J=7.1 Hz), 7.40-7.16 (m, 20H), 7.12-7.05 (m, 3H), 6.75 (td, 2H, J=7.3, 2.2 Hz), 6.65 (d, 1H, J=7.3 Hz), 5.96 (t, 1H, J=4.7 Hz), 5.59 (d, 1H, J=3.8 Hz), 5.55 (s, 1H), 5.31 (s, 1H), 4.17-4.08 (m, 2H), 3.90 (d, 1H, J=9.2 Hz), 3.82 (s, 3H), 3.72 (s, 3H), 3.65-3.57 (m, 2H), 3.21-3.06 (m, 2H), 2.57-2.45 (m, 1H), 2.38-2.24 (m, 2H), 1.10 (dd, 1H, J=13.4, 10.1). 13  C-NMR (75 MHz, CDCl3) δ: 176.9, 174.5, 170.6, 170.1, 148.7, 147.2, 145.0, 144.0, 131.3,  130.3, 129.5, 129.1, 128.5, 128.1, 127.8, 127.3, 127.1, 126.8, 124.8, 123.0, 120.9, 119.0, 110.6, 110.3, 92.5, 90.8, 89.8, 87.1, 78.9, 75.8, 66.9, 64.8, 52.7, 52.5, 45.1, 43.1, 41.5, 41.2.  98  3a-hydroxy-1-trityl-1,2,3,3a,8,8a-hexahydropyrrolo[2,3-b]indole-2-carboxamido glycine triethylammonium salt (132) Trt N HN  NHEt3  O  O  N H  O  OH  In a 5 mL round bottom flask containing a dioxane:water mixture (2:1, 1.5 mL) was added 58 (45 mg, 0.075 mmol), followed by LiOH•H2O (32 mg, 0.75 mmol). The reaction was stirred at room temperature for 1.5 h at which point TLC (9:1 CH2Cl2:MeOH) indicated reaction completion. The solvent was evaporated, and the orange residue was purified by silica gel chromatography using a mixture of CH2Cl2:MeOH:TEA (9:1:0.1). The triethylammonium salt 132 was isolated as an orange solid as a mixture of diastereoisomers (41 mg, 88% yield). ESI-HRMS (C32H28N3O4)-: 518.2088 (observed), 518.2080 (expected) 1  H-NMR (300 MHz, d4-MeOH) δ: 9.41 (d, 1H, J=5.6 Hz), 7.72 (d, 6H, J=7.8 Hz), 7.55-7.36 (m,  6H), 7.40-7.11 (m, 20H), 7.03-6.94 (m, 3H), 6.69-6.58 (m, 2H), 6.32 (d, 1H, J=7.9 Hz), 5.50 (d, 1H, J=1.9 Hz), 5.29 (s, 1H), 4.23 (d, 1H, J=9.4 Hz), 3.90 (d, 1H, J=9.3 Hz), 3.73-3.67 (m, 5H), 3.02 (q, 12H, J=7.3 Hz), 2.80-2.73 (m, 1H), 2.56-2.48 (m, 1H), 2.37-2.30 (m, 1H), 2.14 (d, 1H, J=12.5 Hz), 1.29 (t, 18H, J=7.3 Hz). 13  C-NMR (75 MHz, d4-MeOH) δ: 176.5, 174.5, 174.4, 174.0, 150.0, 147.3, 145.9, 145.7, 145.1,  143.7, 131.8, 131.7, 131.4, 129.6, 129.5, 129.3, 129.2, 128.9, 128.8, 127.9, 127.6, 127.5, 127.2, 126.9, 126.4, 123.8, 122.2, 118.4, 118.2, 109.9, 09.7, 92.0, 89.4, 88.9, 86.7, 78.8, 75.9, 66.7, 64.8, 63.1, 46.3, 8.4.  2.6.4 PEPTIDE SYNTHESIS Linear peptide sequences were prepared via standard Fmoc-SPPS protocols using a chloro-trityl resin (Novabiochem). Synthesis was carried out in polyethylene Zeba spin columns with a teflon frit (Pierce) in 5 mL or 10 mL sizes, which were fitted with a plastic frit to allow for quick filtration. Shaking was carried out gently using an adapted laboratory vortexer (VWR). Filtering and washing were carried out with a water aspirator. 99  2.6.4.1 RESIN PREPARATION The following protocol describes the addition of isoleucine to the resin. The protocols are identical for other amino acids. To a flame dried flask was added chloro trityl resin (0.459 g, 0.59 mmol), which was then suspended in dry CH2Cl2 (4 mL). To this flask was added Fmoc-Lisoleucine (239.4 mg, 0.66 mmol) and DIPEA (461 µL, 2.64 mmol). The reaction was stirred gently at room temperature for 2 h and transferred to a Buchner funnel. The resin was washed three times with a mixture of CH2Cl2:MeOH:DIPEA (17:2:1, 7 mL each), and then washed with CH2Cl2 (5-10 resin volumes), then DMF (5-10 resin volumes) and finally CH2Cl2 again (5-10 resin volumes). The resin was dried on high vacuum over KOH to remove residual solvent. Loading of the resin was determined using manufacturer’s protocols. Briefly, a weighed amount of resin was treated with a solution of 1,8-diazabicyclo[5.4.0]undec-7-ene in DMF (2%) for 30 minutes. The solution was diluted and the UV absorbance of the liberated dibenzofulvene was measured at 304 nm (ε310 = 7624 M-1cm-1). An average of two experiments was used to determine the loading. 2.6.4.2 AMINO ACID DEPROTECTION AND COUPLING Resin was placed in a Zeba spin column (up to 200 mg in a 5 mL column) and swollen in DMF (3 mL) for 30 min, and then the solvent was drained. The N-terminal Fmoc protecting group was removed by washing the resin three times with 20% piperidine in DMF (3 - 5 mL), three minutes per wash. Following deprotection, the resin was washed three times with DMF (2 - 4 mL), three times with CH2Cl2 (3 - 5 mL) and again three times with DMF (2 - 4 mL). The next amino acid was coupled to the resin in the generic form Fmoc-Xaa-OH (2-5 eq), in the presence of HBTU (2-5 eq) and DIPEA (5-10 eq) in DMF (3 - 6 mL). The resin was mixed on the vortexer at minimum speed for 30 min. A Kaiser test was performed to check for complete couplings, following Novabiochem protocols. Occasionally, a small amount of resin (5-10 beads) was deprotected with HFIP:CH2Cl2 (1:3, 1 mL), and checked by low resolution ESI mass spectrometry. When the reaction was complete, the coupling mixture was drained, and the resin was washed with DMF (15-20 mL). Procedures in which a non-commercially available amino acid was used, fewer equivalents and longer coupling times were employed. Double coupling  100  was performed when the free N-terminus on the resin was derived from a proline/hydroxyproline residue. 2.6.4.3 SAVIGE-FONTANA REACTION Linear peptides on resin containing an Hpi moiety were transferred to a round bottom flask and stirred in TFA (700 µL) in the dark for 5 h. The resin was filtered over cotton wool, and washed with CH2Cl2 (0.5 mL) and twice with MeCN (1 mL per wash). The filtrate was evaporated, followed by co-evaporation with MeCN (1 mL) and dried under reduced pressure. The residue was brought up in MeOH (200 µL) and the organic by-products were precipitated with H2O (2 mL). The mixture was centrifuged and the supernatant was filtered through a 2 µm Acrodisk, and purified on HPLC using Gradient System A or B. Fractions containing the desired peak were detected at 229 and 292 nm and collected. The product was lyophilized and re-suspended in water. Concentrations of the peptide were determined by their UV absorbance at 290 nm, with an assumed extinction co-efficient of 12600 M-1cm-1. 2.6.4.4 PEPTIDE MACROLACTAMIZATION Linear monocyclic peptides were dissolved in dry DMF (final concentration of 1-5 mM). The coupling agent PyBOP (3 equivalents) and DIPEA (3 – 10 equivalents) were added, and the reaction was stirred at room temperature overnight.  The solvent was evaporated and re-  suspended in a H2O:MeCN (1:1, 1 mL). The bicyclic peptides were purified on HPLC using Gradient System C or D, detected at 229 and 292 nm. Collected product was lyophilized, and dissolve in H2O:MeCN (1:1) to determine the concentration by UV spectroscopy. HPLC, UV and ESI-HRMS were employed to characterize these peptides. When quantities permitted, NMR analysis was also performed, with water suppression if necessary.  2.6.5 PEPTIDE CHARACTERIZATION 2.6.5.1 MONOCYCLIC OCTAPEPTIDES Hyp(NBn)-monocycle (123a) Retention time (HPLC System A, Column I): 31.29 min 101  ESI-HRMS (C46H61N11O13SNa)+: 1030.4050 (observed), 1030.4069 (expected) 1  H-NMR (300 MHz, D2O) δ: 7.61 (d, 1H, J=8.1 Hz), 7.52 (d, 1H, J=8.0 Hz), 7.47 (t, 1H, J=7.5  Hz), 7.38 (d, 1H, J=7.4 Hz), 7.33 (d, 1H, J=7.9 Hz), 7.26-7.18 (m, 2H), 7.11 (t, 1H, J=7.6 Hz), 4.94 (t, 1H, J=7.1 Hz), 4.68-4.60 (m, 1H), 4.41 (t, 1H, J=8.6 Hz), 4.32 (m, 1H), 4.31-4.29 (m, 1H), 4.22 (t, 1H, J=7.2 Hz), 4.14 (d, 1H, J=6.6 Hz), 4.06-3.96 (m, 4H), 3.86 (d, J=16.9 Hz), 3.72-3.64 (m, 4H), 3.43-3.37 (m, 1H), 3.24-3.37 (m, 4H), 2.70 (dd, 1H, J=6.4, 15.2 Hz), 2.58 (dd, 1H, J=7.8, 15.2 Hz), 2.41-2.35 (m, 1H), 2.01-1.94 (m, 1H), 1.87-1.78 (m, 2H), 1.47-1.35 (m, 2H), 1.17-1.09 (m, 2H), 0.87-0.80 (m, 12H). Hyp(Nv)-monocycle (123b) Retention time (HPLC System B, Column II): 18.31 min ESI-HRMS (C48H65N11O15SH)+: 1068.4464 (observed), 1068.4461 (expected) 1  H-NMR (300 MHz, D2O) δ: 7.27 (d, 1H, J=8.1 Hz), 7.12 (d, 1H, J=8.2 Hz), 7.07 (s, 1H), 7.04  (t, 1H, J=7.7 Hz), 6.96 (t, 1H, J=7.6 Hz), 6.69 (s, 1H), 5.04 (t, 1H, J=7.2 Hz), 4.63-4.54 (m, 2H), 4.46 (t, 1H, J=8.5 Hz), 4.36 (t, 1H, J=5.9 Hz), 4.33-4.28 (m, 1H), 4.16-4.08 (m, 3H), 3.95 (d, 1H, J=17.2 Hz), 3.82-3.69 (m, 5H), 3.67 (s, 3H), 3.58 (s, 3H), 3.42 (d, 1H, J=16.3 Hz), 3.19-2.96 (m, 4H), 2.70 (dd, 1H, J=6.8, 15.5 Hz), 2.57 (dd, 1H, J=7.3, 15.3 Hz), 2.41-2.34 (m, 1H), 1.98-1.87 (m, 1H), 1.85-1.76 (m, 1H), 1.66-1.57 (m, 1H), 1.42-1.28 (m, 2H), 1.17-1.01 (m, 2H), 0.85-0.69 (m, 12H). Asn(TEG-DEAC)-monocycle (124) Retention time (HPLC System B, Column II): 21.06 min ESI-HRMS (C48H65N11O15SH)+: 1247.5757 (observed), 1247.5771 (expected) 1  H-NMR (300 MHz, D2O) δ: 8.08 (s, 1H), 7.35 (d, 1H, J=9.0 Hz), 7.23 (d, 1H, J=7.9 Hz), 7.12  (d, 1H, J=7.8 Hz), 7.07-6.96 (m, 2H), 6.77 (d, 1H, J=9.0 Hz), 6.30 (s, 1H), 4.98-4.90 (m, 1H), 4.52-4.47 (m, 1H), 4.33 (t, 1H, J=8.2 Hz), 4.16 (d, 1H, J=7.5 Hz), 3.98-3.91 (m. 4H), 3.75-3.21 (m, 17H), 3.08-2.87 (m, 3H), 2.79-2.40 (m, 5H), 2.00-1.87 (m, 2H), 1.83-1.54 (m, 4H), 1.39-1.02 (m, 7H), 0.88-0.70 (m, 4H), 0.68-0.55 (m, 3H). 2.6.5.1 BICYCLIC OCTAPEPTIDES Pro-bicycle (64) Retention time (HPLC System B, Column II): 17.55 min 102  ESI-HRMS (C39H54N10O9SNa)+: 861.3680 (observed), 861.3694 (expected) Hyp-bicycle (53) Retention time (HPLC System B, Column II): 16.26 min ESI-HRMS (C39H54N10O10SNa)+: 877.3629 (observed), 877.3643 (expected) Hyp(NBn)-bicycle (100a) Retention time (HPLC System D, Column I): 14.39 min ESI-HRMS (C46H59N11O12SNa)+: 1012.3965 (observed), 1012.3963 (expected) Hyp(Nv)-bicycle (100b) Retention time (HPLC System B, Column II): 22.03 ESI-HRMS (C48H63N11O14SNa)+: 1072.4187 (observed), 1072.4174 (expected) Asn(TEG-DEAC)-bicycle (101) Retention time (HPLC System B, Column II): 23.99 min ESI-HRMS (C69H80N11O14SNa)+: 1251.5494 (observed), 1251.5485 (expected)  103  CHAPTER 3: BIOLOGICAL EVALUATION OF AMATOXIN PROBES 3.1 INTRODUCTION The biological target of α-amanitin, RNAP II, is a multisubunit protein complex localized to the nucleus of eukaryotic cells.28 It is responsible for the transcription of DNA sequences to yield mRNA.27,180 As was reviewed in the introduction, this is an intricate process, requiring the cooperation of a variety of protein and enzyme cofactors. The assay of RNAP II activity is not trivial; there are many protein components that are required for transcription. Despite this challenge, the importance of creating the ability to study and control this process was the driving force for this project. There are two environments in which we sought to achieve this control: in vitro, and within a eukaryotic cell. An in vitro approach to assaying the activity of RNAP II is useful to provide further information about the structural requirements of the amatoxins with respect to inhibition. However, to understand the fate of various cellular processes that are affected by transcriptional arrest, an intracellular approach is required. Based on the probes that were designed in the previous chapter, we will need methods of testing their activity not only directly on the target protein complex, but also in the context of a cell. This next chapter will deal with the approaches that were undertaken in an attempt to develop reliable and reproducible assays of the effects of the amatoxin probes on RNAP II activity. The set of amatoxin-based probes of RNAP II that were developed in the previous chapter (100a/b and 101) were based on two unique types of desired activity: photoactivation and fluorescent visualization. While both of these techniques would be useful in a direct “enzyme-substrate” style assay, their usefulness was targeted for whole cell applications.  3.1.1 RNA POLYMERASE II ACTIVITY ASSAYS A variety of techniques that provide the ability to monitor transcription and the production of mRNA in eukaryotic cells have been described.181-186 The choice of assay is based on several factors. These include whether the researcher chooses to study the mechanism of transcription or wishes to develop control of the process. Other features that are important are the type of readout that is desired and the medium in which transcription is to be studied. These choices will affect the type of assay that is used. 104  Many enzyme-based assays are performed in a straightforward manner, where the purified protein is used to monitor substrate turnover.  In the case of RNAP II, both the purified  polymerase complex and template DNA are not generally sufficient to achieve transcription; various transcription factors must also be present.183,184 Alternatively, when studying mRNA production in a cellular context, there are issues with competing transcription of other RNA species by RNAP I and RNAP III. This leads to many unique RNA transcripts whose presence can interfere with the specific assay of RNAP II activity. Both the experimental conditions and the RNA product readout must be taken into consideration when choosing an appropriate RNAP II assay. 3.1.1.1 CONDITIONS FOR ASSAYING MRNA PRODUCTION The generation of mRNA by RNAP II can be studied in both complex and minimized situations. Four of the more common environments that are employed to study eukaryotic transcription include whole-cell systems, isolated nuclei, purified nuclear extract or purified polymerase.181  Figure 3.1:  Diagram depicting various conditions employed for the assay of mRNA  production and the potential readout. The first two methods depicted in Figure 3.1 are useful for the study of several aspects of transcription, from expression levels to regulation. Whole-cell based applications are useful when the effect of inhibition of transcription on the overall cellular fate is to be studied. In this approach one can monitor either the production of RNA,187 or else a down-stream effect, related to altered levels of mRNA expression due to RNAP II inhibition.188 Whole nuclei offer the advantage that the RNA species produced in this setting (run-on transcription) are representative of the transcription levels of the cell. Processes related to transcription, however, cannot be monitored in this case, as the RNA is not translated into protein. 105  Both nuclear extract and purified RNAP II are useful for the study of the specific mechanisms of transcription.147,189 In these cases, a discrete RNA product is formed, based on an exogenous DNA substrate. This is referred to as run-off transcription. The nuclear extract provides the various transcription factors that are associated with the process, while the purified polymerase complex can function on certain DNA fragments, in the absence of these co-factors. Each of these techniques provides different levels of reproducibility and error. The approach chosen is dependent on the readout method that is chosen, and the desired information. While fundamental understanding of the chemical relationship required for amatoxin binding to RNAP II are best served through a run-off transcription assay, these do not prove information on the cellular effect of inhibition. The probes described in the previous chapter are designed to be effective in cellular conditions, where visualization and control of transcription is the goal. 3.1.1.2 QUANTIFICATION OF RNA PRODUCTION There are various methods used to quantify mRNA production.190 The first decision is whether to study bulk RNA or a single RNA transcript. The biggest problem with studying bulk RNA is separating the mRNA products from those generated through the action of RNAP I, and III. The most common approach that is used to distinguish the activity of these polymerases is to incubate with various concentrations of α-amanitin.7,191 Since it is a potent inhibitor of RNAP II, the decrease in RNA production observed upon incubation with low concentrations of the toxin (0.5 µM) is attributed to a lack of mRNA production. At higher concentrations (100 µM), the decreased amount of RNA produced is reflective of inhibition of RNAP II and RNAP III activity. This experiment gives information about the basal transcription level induced by RNAP I. Assay Conditions no α-amanitin 0.5 µg/mL α-amanitin 100 µg/mL α-amanitin  Observed Products mRNA, tRNA, rRNA tRNA, rRNA rRNA  % of total RNA 100% 50% 5%  Table 3.1: Use of α-amanitin to distinguish the activity of RNA polymerases in sea urchin embryo nuclei, as reported in Hames et al.181 106  The largest hurdle to overcome when monitoring RNA production involves the isolation and detection of the product. In each case, the total RNA produced must first be isolated from the rest of the components of the reaction. This procedure can be cumbersome, as the sensitivity of the product to nucleases requires strict control of the experimental conditions. Following isolation of the bulk RNA, the transcription related to RNAP II must be determined. There are many approaches that have been developed to monitor the production of discrete mRNA sequences, including nuclease protection assays and northern blots.190 The observation of the generation of a specific mRNA transcript reduces the problems associated with background transcription. After exposure to the experimental conditions the bulk RNA is isolated, but only a specific gene product is detected. This approach is useful in each of the assay conditions mentioned in the previous section, and provides the opportunity to avoid the use of radioactive labels in the transcription experiment. More recent approaches have allowed for the array of a variety of mRNA transcripts, such that the relative expression levels of several products can be detected in one set of experimental conditions. There are also indirect methods used for the identification of RNAP II inhibition. In these experiments, the effect of inhibition is monitored through down-stream effects, such as protein production or even cell death.192,193 This approach is useful in cell-based systems, where cells are permitted to grow under the desired experimental conditions in the presence of the potential inhibitor. The results of such an experiment are generally advantageous in that no intermediate purification steps are required, and the effect of decreased levels of transcription can be measured spectrophotometrically. The effect of amatoxins on cell viability is a convenient method for the observation of RNAP II inhibition capacity without the need for isolation and separation of distinct RNA species. Regardless, there are a few important associated risks when monitoring the effect of inhibited RNA production in an indirect fashion. While the effects that are measured are likely related to decreased levels of transcription, it is imprudent to state unequivocally that this is a direct effect of RNAP II inhibition. Several cellular factors could also be inhibited under these experimental conditions. In the case of amatoxin inhibition of RNAP II, we anticipate decreased levels of mRNA will ultimately lead to cell death.  107  3.1.1.3 CELL VIABILITY ASSAYS The cellular toxicity of any given molecule can be determined through the observation of its effect on cell viability.193,194 These experiments rely on the introduction of the proposed toxin to growing cells, followed by the detection of the viable cells remaining after a given incubation time. The viable cells are most often detected through a coloured output. A variety of agents have been described which are useful for selectively visualizing and quantifying living and/or dead cells. A dye-exclusion assay is one of the most fundamental methods of determining the viability of a cell.195 When living cells are exposed to the organic dye trypan blue (141), they will exclude it from the cytosol. However, non-living cells will actively incorporate the dye into the cell. If a pool of cells exposed to the dye is visualized on a light microscope the dead cells can be distinguished from the living cells through their distinct blue appearance. Simply counting the ratio of living to dead cells will yield a numerical reflection of the toxicity of a given molecule. NH2  H2N  OH  A.  HO  HO3S  SO3H  N  N  N SO3H  N HO3S  141 N N N N  B. N  mitochondial reductase Br  S  yellow  N  C. O  O  blue  N  N  purple  metabolic reduction O  N  S  142  O N  N H  143  N O  144  O purple  O  145  Figure 3.2: Common dyes employed to detect the viability of cells A. Trypan blue is excluded from viable cells. B. MTT is reduced to the purple formazan. C. Resazurin is reduced to the purple resorufin.  108  Often these assays are performed on multiple samples in a 96-well plate. The viability of cells can be determined in a high-throughput manner using a coloured indicator that can be measured without the use of a microscope. Various molecules have been identified that are reductively converted into a visible dye by viable cells. Non-viable cells do not generate the dye, therefore the intensity of the dye in various conditions represents the ratio of living cells. Two of the more commonly employed reagents for the visualization of viable cells are dimethylthiazolyl diphenyltetrazolium bromide (MTT) (142) and resazurin (144) (commercially available as alamarBlue).196 Each of these reagents are a substrate for active cellular reductases. When viable cells are treated with these reagents they are reduced to form either a purple formazan product (in the case of MTT) (143) or purple resorufin (in the case of alamarBlue) (145). The relative concentration of viable cells can be detected using a spectrophotometer, observing the absorption of the various cell stocks at 595 nm. 3.1.1.4 LITERATURE PRECEDENT The majority of amatoxins that have appeared in the literature (both natural and synthetic) have been studied within two different contexts.2,197 Many original pursuits studied the toxicity of these molecules in animals, and reported their relative activity in terms of the lethal dose in 50% of the population (LD50). Developments of RNAP II identification and isolation led to the ability to obtain data using run-off transcription assays to obtain 50% inhibitory concentrations (IC50). Radioactively labeled amatoxin derivatives have been used in the determination of dissociation constants (Kd). The activity of these toxins in cultured cell assays is not as commonly reported. Most cell-based evaluations have relied on the measurement of RNA production; however, there are a few instances where the cell viability has been recorded.  109  Source  Cell line  human cervical adenocarcinoma oral carcinoma larynx epidermoid carcinoma regulatory T-cells Madin Darby bovine kidney mouse kidney fibroblasts (transformed) human embryo fibroblasts (transformed) human embryo fibroblasts Baby hamster kidney African green monkey kidney  HeLa KB HEp-2 RTC MDBK MKS-Bu 100 HEF-SV HEF BHK VERO  ED25 (µg/mL) 3 5 10 5 10 5 2.5 5 2.5 10  Table 3.2: Cell viability measurements of a variety of cell lines exposed to various doses of α-amanitin. The concentrations required to achieve death in 25% of the cells (ED25) was determined using trypan blue.24 In one particular study a variety of cultured cell lines from both human and animal sources were assayed for their cell-viability in the presence of α-amanitn (Table 3.2).198 In this case, the cells were exposed to the toxin for 24 hours, and then the number of remaining viable cells was determined using trypan blue. The effective dose required to achieve 25% cell death was reported for each cell line (ED25). Of all the cells tested, they each showed a similar response to the toxin, varying with up to a five-fold difference. One thing to note is the much higher concentrations of the toxin required to achieve this cell death with respect to the reported IC50’s. This implies that the cellular response to α-amanitin exposure may have competing mechanisms that diminish the effect of the toxin.  3.1.2 CELL MEMBRANE PERMEABILITY OF AMATOXINS One of the mechanisms that may lead to the diminished toxic effect of amanitin within a cell could be related to cell permeability. Since the probes that were developed in the previous chapter were designed to be functional in living cells, the cell permeability of the probes is a critical aspect. Previous studies have been undertaken in an attempt to determine the cellular uptake of this class of toxin.199-201  110  3.1.2.1 PREVIOUS AMATOXIN UPTAKE STUDIES The cellular uptake of α-amanitin has been studied in a few contexts.198-200 These uptake studies were centered around the use of radiolabeled derivatives, prepared by modification of the isolated natural product.22 The most common cell lines that have been used are derived from the liver, as these show prominent effects of amatoxin poisoning. The amatoxins were shown to be taken up by mechanisms analogous to the uptake of phallotoxins.199 This process was shown to be mediated by membrane-associated transporters responsible for bile-acid uptake. Cells that do not express these transporters are much less sensitive to the toxin.200,202 Two of the main transporters associated with this process are the sodium-taurocholate co-transporter polypeptide (Ntcp), and organic anion transporter 1B3 (OATP1B3). Cells that were stably transfected with plasmids that encode for the production of OATP1B3 were shown to display improved amanitin uptake. In one particular study it was shown that other substrates for this transporter (paclitaxel, rifampicin, cyclosporin A) inhibited the uptake of α-amanitin.200 One study was even able to determine the relative rate of uptake of various amatoxin derivatives.201 The fact that all of the amanitin derivatives were substrates for the transporter implies that the receptor OATP1B3 is capable of importing various analogs of this bicyclic octapeptide scaffold. Cells that do not naturally express these transporters are also susceptible to the toxin, but over longer periods of time, implying the important effect of cell membrane permeabilization on amatoxin induced inhibition of RNAP II. 3.1.2.2 METHODS OF CELL MEMBRANE PERMEABILIZATION The identification of the intracellular effects that are exerted through the application of small molecules to cell culture relies on the ability for the molecule to efficiently enter the cell. Since the amatoxins were shown to enter cells at a much slower rate when specific transporters are not expressed, cell uptake could prove to be crucial in the application of our amatoxin probes. Various approaches have been designed that are meant to improve the passive uptake of molecules across a cell membrane.203-208 These approaches involve the introduction of an additive that has a minimal effect on cell function, but renders the cell membrane permeable. These can range from solvents to small molecules to complex polypeptides. Important features 111  that must be satisfied when permeabilizing a cell membrane include the retention of native activity, and the reversibility and reproducibility of this approach. A common additive that is employed to modify the structure of the cell membrane is dimethyl sulfoxide (DMSO).209 This polar aprotic additive is known to render membranes “leaky”, and increase the passive uptake of pharmacologically relevant molecules. Lower concentrations (≤ 5%) of this solvent when added to the growth medium do not greatly affect the cell viability, but higher concentrations will begin to exert negative cellular effects. Other organic solvents such as diethyl ether and toluene have been used to extract cell membrane proteins. The extraction process has been shown to increase porosity, but these effects are irreversible, and lead to cell death. Detergents have also been used to varying degrees to affect the permeability of cell membranes.210 While detergents such as Triton X-100 and Nonidet P-40 lead to irreversible destruction of the membrane, others have been identified that diminish this undesired side effect. Specifically, lysolecithin 146, digitonin 147 and saponin 148 are commonly employed to aid in the introduction of foreign molecules into the cellular environment.206 O  A.  O  P  N  OH O  O 16  O O  146  OR CHO  B.  OH O RO  OH  O  O O  HO  HO  O  HO OH O  OH HO HO  OH OH  O  O  O  O  O  147  OH H O  C. HO HO  OH O OH  OH  OH O  O OH HO HO  OH OH O  HO O  O O O  OH O  HO  H O  O  H H  H OH  H  OH OH  148  Figure 3.3: Structure of mild cell permeabilizing detergents. A. Lysolecithin B. Saponin C. Digitonin. 112  The mechanism of action of these reagents is proposed to be through an intercalative effect in the cell membrane.206 This process is though to introduce pores through which molecules can flow into the cellular environment. The size of the pores can roughly be controlled through the concentration of detergent added. The main problem with this approach is that the components of the cell are equally free to leave the cytosol. Other methodologies that have been developed to improve the cellular permeability have relied on mimicking natural methods.  Discrete peptide sequences have been identified that are  responsible for the introduction of cargo into a cell.207 These sequences are generally highly positively charged, such as in the case of octaarginine and a fragment of the HIV-TAT peptide.211 Synthetic variants of these peptides have been appended to molecular cargo to introduce material that otherwise would not penetrate the cell membrane. There is still debate as to whether this process is passive or energy dependent. It is clear that cell permeability of the amatoxins may pose a real concern in the efficient application of the probes we have designed for use within a cellular context. If this does prove to be an issue, there are a variety of techniques that can be applied to render these molecules more relevant, based on these above-mentioned techniques.  3.1.3 FLUORESCENT IMAGING OF THE CELL The development of confocal microscopy and fluorescent labeling techniques has allowed for the rapid expansion of imaging in cell biology.136,138 It allows for the visualization of cellular targets, interactions and processes through a fluorescent readout, with remarkable resolution.  This  technique can allow for the imaging of various cellular components, ranging from small molecules to organelles. A wide range of fluorophores (both organic and inorganic) have been developed for the purpose of cell imaging in confocal microscopy, and entire companies exist which are devoted to this application.143 The sensitivity that is afforded by this microscopy technique coupled to the relative ease of the preparation of fluorescent probes has rendered this field significant.  113  3.1.3.1 CELL IMAGING TECHNIQUES There are a variety of approaches that have been developed that permit the visualization of cellular function. The wide host of targets that exist within the milieu of the cell requires that the fluorescent labeling must be extremely specific. Three main approaches have been developed that address this specificity requirement. A.  1. Permeabilize 2. Add fluorescent antibody  B.  1. Add fluorescent inhibitor  C.  Induce mutant target protein growth  Figure 3.4: Approaches to fluorescent visualization of a desired target. A. Fluorescently labeled antibodies can recognize a protein of interest after cell permeabilization. B. Small molecules designed to bind to the target can be fluorescently labeled.  C. Genetically  modified cells produce target proteins that are covalently linked to a fluorescent protein. One method that is used to guarantee specific labeling of a particular biological function relies on antigen-antibody recognition.212 Antibodies can be generated such that they bind selectively to a specific target. If these antibodies are labeled with a fluorophore, they can permit the visualization of intracellular binding. Since these labeled antibodies do not readily possess the ability to penetrate across the cell membrane, most microscopy experiments using this approach require irreversible and consequently lethal permeabilization of the cell membrane. Based on this fact, dynamic studies of cellular function are not readily achieved.  114  As an alternative to antibody-targeted labeling of a desired cellular target, one can use a smallmolecule approach.142 Inhibitors and other small molecules that bind to specific enzymes and proteins have been identified. If these molecules are labeled with a fluorophore, the distribution and localization of their interaction within a cell can be realized. Molecules that interact with proteins away from the active site are particularly promising, since the function of their target is less likely to be affected through binding. Phalloidin, a close relative of α-amanitin, is a bicyclic heptapeptide containing a tryptathionine crosslink that displays remarkable ability to selectively bind filamentous actin (F-actin). Since F-actin is a key structural component of cells, the binding of fluorescently labeled phalloidin 149 to F-actin is commonly used to achieve visual representation of cell structure.179,213 N  A.  B. CO2  O  S N H  N  CH2  H3C H N HN  O N H  O HO  HN CH3  S N H  N  OH  O  O O  NH  O  N H HO  O NH CH3  149 Figure 3.5: A. Rhodamine labeled phalloidin. B. Confocal microscope image of a skin fibroblast cell imaged with rhodamine phalloidin (red). Image taken from Invitrogen catalog (Product #R415). Another approach that is used to image cellular targets relies on the genetic manipulation of the host cell.144 Target proteins and enzymes can be genetically encoded to covalently append a fluorescent protein, without significantly affecting its function.144 This method does not require post-biosynthetic addition of a fluorescent label or probe. This has proven to be useful in a  115  number of situations, especially in dynamic cellular studies, since the manipulation of cells is not required for imaging. Each of these techniques that is used for the fluorescent labeling of cellular targets has advantages but also disadvantages.  In each situation, the added molecular bulk that the  fluorescent probe adds to the desired target can potentially affect its inherent activity. This diminished activity can be minimized, however, through rational design. The many advantages that this technique provides, improved labeling methodology and increased programming in the microscope will only improve this approach to cell imaging. Each of these techniques can be used interchangeably, and in cooperation to generate the most relevant data. 3.1.3.2 LITERATURE PRECEDENT Fluorescent visualization of transcription has been reported using each of the techniques mentioned above.61,214-216 Antibodies have been raised that specifically target various functioning aspects of RNAP II.217 These have been used for many applications, including the visualization of the enzyme complex in fixed cells.218 Transcription has been visualized in live cells using general transcription factors and elongation factors that genetically encode a fluorescent protein.219-221 The production of RNA generated by transcription can be directly visualized as well.145 Certain mRNA sequences are known to form structural elements that are recognized by specific proteins. These elements can be visualized using a fluorescent variant of the protein that is directed to it. The MS2 coat protein is an example, which recognizes a 24-base pair stem loop RNA sequence.222 The application of amatoxins to RNAP II imaging has been reported in the literature.61 In this example, the researchers appended fluorescein onto a functionalized derivative of α-amanitin (149) as described in the previous chapter. With this labeled amanitin, researchers were able to show nuclear localization of the fluorescent amatoxin to cell nuclei. The localization was dependent on the presence of amanitin, and was also shown to be active in various cell cycles. In their study, the imaging was performed on fixed and permeabilized cells; no cell uptake or membrane permeability was mentioned.  116  A generalized approach to the development of fluorescently labeled amatoxins would greatly increase the ability to visualize and understand the underlying mechanisms of inhibition of transcription brought about by exposure to amatoxins. These derivatives could be used to gain further information regarding the structural requirements for cell localization as well as cell permeability. It has been proposed that less active derivatives (specifically ones lacking the hydroxyl group of Hyp2), still possess the ability to bind the polymerase, but do not affect its function. Fluorescently labeled amatoxins could provide detailed information regarding this proposal.  The development of amatoxins that bear fluorescent labels would prove to be  complimentary to the many techniques that have already been described.  3.1.4 CHAPTER GOALS In this chapter we set out to develop methodology to assay the activity of the amatoxin probes prepared in the previous chapter. Previous work in our lab has undertaken in vitro methods for the analysis of eukaryotic transcription with varying results. These include run-on and run-off transcription assays, using commercial and in-house prepared material. Based on the desired intracellular activity that we hope to achieve with the photoactivatible and fluorescent probes 100a, 100b and 101, we chose to focus on cell-based assays. We propose that the nitroveratryl-protected amatoxin 100b will not be toxic in the cell if deprived of light. The exposure of cells treated with 100b to 366 nm light will then hopefully induce cell death following photolysis of the nitroveratryl protecting group, thus revealing a potent amatoxin. The proline and hydroxyproline derivatives 53 and 64 were designed to provide controls for this activity. The fluorescent derivative 101 was designed to observe and confirm the cellular localization of this probe in the nucleus. It will also be useful in the study of the cellular permeability of these probes.  We will use cell viability assays and confocal  microscopy to test these hypotheses, and attempt to improve the cellular permeability of the probes prepared in Chapter 2.  3.2 CELL VIABILITY ASSAYS Our goal is to induce inhibition of RNAP II in a controlled manner, to develop further understanding and control of transcriptional arrest. Since the amatoxins are designed to inhibit 117  all mRNA production, they should lead to cell death. The most direct method to ensure that our probes bear the desired activity is to determine their effect on cell viability.  3.2.1 GENERAL CELL CULTURE PROTOCOLS Assaying the cell viability of cultured adherent mammalian cells is a well-established process, and the equipment and chemicals that are required to do so are commercially available.193,196,223 There is a wide variety of cell lines that are available, and the ideal growth medium for these has been established. For detailed approaches to establishing cell culture techniques there exist excellent reference texts.223-225 I will briefly provide some general details here to orient the uninitiated reader, and describe the specific processes that were undertaken to assay cell viability. Adherent cell lines are grown on the surface of coated plastic flasks in a humidified environment containing 5% CO2. The growth medium is supplemented with deactivated fetal bovine serum (FBS) and also contains the antibiotics penicillin and streptomycin. Due to the adherent nature of the cells, the redistribution of cells grown to confluence cannot be done directly by dilution. Cells are first detached from the growth flask following a brief exposure to a solution of the protease trypsin, in EDTA. Upon detachment, the protease is quenched with fresh growth medium and the cells are centrifuged at low speed. The stock of cells can then be counted with a hemacytometer and diluted to a desired concentration. For cell viability assays, cells are diluted to a desired concentration (generally 103-105 cells/mL), and then distributed evenly in a 96-well plate. The cells grow uninhibited for a 24-hour period, during which point they attach to the surface of the wells. At this point, the growth medium is aspirated, and is replaced with fresh medium containing the required concentration of toxin. After growth in the presence of toxin for a set period of time, the medium is supplemented with a solution of the indicator dye. Further incubation in the presence of this dye allows for the formation of the coloured product indicative of viable cells. In the case of MTT-based assays, the purple formazan product precipitates, and the dye must be dissolved in DMSO prior to measurement. The cell viability is then directly assessed by the relative absorbance values of the various wells at 595 nm. 118  Each of the conditions is run in triplicate (unless otherwise noted) to obtain statistically relevant data. Statistical analysis of these data were performed using GraphPad Prism software. Based on the heterogeneous nature of the cells that were pipetted into the 96-well plates, a large degree of error was sometimes noted. In each of the cases presented in this theis, the error bars in the data presented represent the standard deviation of the replicates for that point. No further statistical analysis was performed, since our goal primarily focused on the identification of trends in the data. We sought to achieve an overall protocol that would represent a reproducible assay of cell viability induced by amatoxins, but not to provide accurate measurement of the ED50.  3.2.2 α-AMANITIN CYTOTOXICITY CONTROLS We first set out to determine the ideal conditions for cell culture viability assays using the potent toxin α-amanitin. Based on the availability of reagents and the accessible expertise, our assays were based on the MTT methodology (Figure 3.2-B). The cell lines that were used in this study were previously used in the Department of Chemistry at UBC and were generously made available by previous users (as credited in the experimental section).  We began these studies using the human colorectal adrenocacinoma cell line Caco-2. The cells were cultured in α-MEM medium containing 10% FBS and the antibiotics penicillin and streptomycin. To assay the effect of α-amanitin, cells were seeded in 96-well plates at an initial concentration of 5000 cells per well.  After a 24-hour growth period, we applied varying  concentrations (from 0.1 – 25 µM) of commercially obtained α-amanitin in medium containing 0.5% DMSO. Following a 66-hour period of exposure to 1, the cells were treated with MTT (142) for three hours to allow for bioreduction to generate formazan 143. The fraction of viable cells remaining was determined using a 96-well plate UV spectrophotometer. The absorbance data is used to generate relative concentrations of the viable cells that remain following treatment with the toxin relative to control cells that are not treated with α-amanitin.  A visual  representation of the plate (obtained with a computer scanner) is also generated to provide a relative idea of the toxicity.  119  Figure 3.6: Sample image of a 96-well plate that is obtained after incubation of cell lines in the presence of α-amanitin for 66 hours, followed by treatment with MTT. This initial assay showed that Caco-2 cells were relatively insensitive to commercially obtained α−amanitin. Following 66 hours of exposure to the toxin, approximately 30% of the cells were still viable relative to untreated cells at a concentration of 10 µM α−amanitin. This result was surprising, based on the reported binding constant of α-amanitin to its target RNAP II of approximately 1 nM. The fact that ~30% of the cells remained viable at concentrations 104 times greater than the binding constant after such a long period of exposure, led us to believe that other factors affect the observed cytotoxic effects that are imparted by 1. Specifically, we proposed two different possible scenarios to account for this diminished level of activity. Our first hypothesis was that different cell types display unique sensitivity to αamanitin exposure. Different organs of live animals have been shown to accumulate greater amounts of the toxin, especially the liver and the kidney. Similar effects were observed in different cultured cell lines as well. Alternatively, the cellular uptake of α-amanitin could be a cause for this low level of activity. The cell permeability of any given chemical species is not a predictable process.  It is possible that poor cell uptake (leading to a lower intracellular  concentration) could be responsible for the observed effect of the toxin in Caco-2 cells. 3.2.2.1 CELL LINE SENSITIVITIY TO α-AMANITIN Following the poor toxicity of α-amanitin observed in Caco-2 cells, we tested the effect on cell viability of this compound in a series of different cell types. We chose cell lines that represented both cancerous and healthy human cells from different organs, as well as cells originating from different species. 120  Cell Line Caco-2 HepG2 CHO Cos-7 Hfl-1  Cell line source Human colorectal adrenocarcinoma Human hepatocellular carcinoma Chinese hamster ovary African green monkey kidney Human lung fibroblast  Table 3.3: Cell lines tested for relative α-amanitin cytotoxicity. Each of these cell lines was grown in the appropriate media and assayed for viability in the presence of α-amanitin. These cells were cultured in 96-well plates and exposed to varying concentrations of the toxin (100 nM to 25 µM) in the presence of 0.5% DMSO. Following a 66hour incubation period, the relative cell viability of each of the cell lines was determined using MTT methodology. 120  Caco-2 Hep-G2 CHO Hfl-1 Cos-7  100 80 60 40 20 0 -7.0  -6.5  -6.0  -5.5  -5.0  -4.5  log concentration (M)  Figure 3.7: Relative cytotoxicity of α-amanitin in various cell lines. The data were plotted as a sigmoidal dose-response curve and fit to the Hill equation with a variable slope.226,227 While the numerical results were crude, they represented a clear trend; of all the cell lines tested, only one of them (CHO) achieved complete inhibition of cell viability over the 66-hour period. Of the human derived cell lines, HepG2 (a liver derived cell line) was the most sensitive to α-amanitin. This was not surprising, since it is known that the toxin targets liver cells. Fitting the data to the Hill equation allowed for the determination of an approximate ED50. In this case, the value was defined as the concentration that was required to achieve 50% of the observed activity. Since most of the cell lines did not reach 100% death at the end of the experiment, it is difficult to compare these data. The most important feature that is required for our studies, is that the cell line achieves 100% cell death following the incubation period. It is clear that the non-human derived cell lines CHO and Cos-7 were the most susceptible to the cytotoxic effects of 1. 121  Cell Line Hfl-1 Caco-2 HepG2 Cos-7 CHO  % survival 47% 28% 23% 16% 0%  ED50 (µM) 0.67 ± 0.10 1.35 ± 0.38 2.00 ± 0.22 1.14 ± 0.08 0.67 ± 0.01  Fit (R2) 0.8811 0.8655 0.9645 0.9710 0.9899  Table 3.4: Relative survival of various cell lines treated with α-amanitin for 66 hours. Percent survival reflects the remaining viable cells following 66-hour treatment at the highest concentration (25 µM). The ED50 refers to the concentration required to achieve 50% cell death relative to the amount remaining following the exposure period. It is important to note that all the data curves shown in Figure 3.7 are all sigmoidal, yet they do not all approach 0% cell viability at the higher concentration values. This seemed to imply that the lack of observed toxicity was related to the time of exposure to the toxin, and that higher concentrations of α-amanitin would not have led to increased toxicity over this 66-hour period. If merely higher concentrations were required to achieve 0% viability, we would expect to see only half of a sigmoidal curve, where the bottom limit has not been reached. Had some of these poorly behaving cell lines been treated with amanitin for longer times, it is possible that 100% cell death would have been achieved. The slow toxic response to amanitin shown by these cells could be indicative of factors such as delayed response to transcriptional inhibition or poor cell uptake. 3.2.2.2 TIME DEPENDENCE To further investigate the effect of the time of exposure to amanitin on cell viability, an MTTbased cell viability assay was undertaken where the toxin was added at various times. Caco-2 cells were grown in a 96-well plate, and stocks of α-amanitin were added to the cells such that a final concentration of 10 µM was achieved. After various exposure times, the viability of the cells was determined.  122  120 100 80 60 40 20 0  0  25  50  75  100  Time (h)  Figure 3.8: Time dependent effect of 10 µM α-amanitin on Caco-2 cell viability. The results in Figure 3.8 indicate that there is a linear relationship between the time of exposure to amanitin and Caco-2 cell viability. There is a rate of decrease in cell viability of 1.2 ± 0.1% per hour in the presence of 10 µM α-amanitin. This provides further evidence that the toxic effect exerted on the cells is slow. However, the underlying mechanisms that influence this slow rate cannot be determined from this experiment.  3.2.3 CELL VIABILITY ASSAYS USING MODIFIED AMATOXINS Even though the toxicity of the natural product α-amanitin in CHO and Cos-7 cell lines was lower than expected, we decided to test the toxic effects achieved through exposure to some of the synthetically prepared amatoxin derivatives. This was deisgned to provide a proof-ofconcept that we would observe noticeable different effects between drastically different amatoxins. Specifically, we first decided to test the cytotoxic effect of a series of three amatoxins, which vary either in the hydroxyl group of the side chain of hydroxyproline (53 and 64) or the stereochemistry of the α-carbon at Ile3, in the case of 64 and 138. Their potentially toxic behaviour in CHO cells was tested using an MTT cell viability assay.  123  O H N  HN O O  S  N  H N  O H2N  N H  N  O  C NH H2  O  O O  H N  O  O  N H N H  O  O  H N O O  O  S  N  H N  O2N  H2N O  O OCH3 H 2N OCH3  53  O O  O  C NH H2  138  HN  O C NH H2  O  O  H2 O N C H HN  H2 O N C H HN  N H N H  O  O  S  S H N  H2N  H N  HN  N  O  H N O O  64  O  HO  HN  N H  O  O  H2 O N C H HN  H2 O N C H HN  N H N H  O O  C NH H2  100b  Figure 3.9: Amatoxin derivatives tested for initial toxicity studies in CHO cells. These relative cytotoxicity of these compounds were tested in a similar fashion to the tests conducted on α-amanitin. First, proline derivative 64 and its Ile3 epimer 138 were tested in CHO cells.  These were grown in a 96-well plate and exposed for 48 hours to three different  concentrations of the peptides in medium containing 0.5% DMSO, at which point the remaining viable cells were assessed. These derivatives were compared to α-amanitin. 150  α-amanitin (1) Pro2-Ile3-amaninamide (64)  100  Ile3-Cα-epimer of 64 (138)  50  0  Concentration (µM)  Figure 3.10: Effect of proline derivatives 64 and 138 on CHO cell viability. (Structures shown in Figure 3.9) As can be observed from these results (Figure 3.10), the addition of either Ile3 epimer of the proline containing amatoxins did not show much of an effect on cytotoxicity. At concentrations 124  as high as 100 µM, these molecules still only inhibited less than 25% cell growth over this time period, with a high degree of error. This result was not surprising though, since these derivatives do not bear a hydroxyl group on Hyp2. To compare these results with a peptide bearing this critical hydroxyl group, we tested our lightsensitive amatoxin derivative. In this initial assay, we treated cells with the amatoxin that contained the nitroverartyl protecting group (100b), and compared it to the deprotected version (53). In this experiment we hoped to observe conclusive evidence that the protected derivative would have a much less pronounced effect on cell toxicity.  The assay was conducted in a  similar fashion to that used for the proline derivatives described earlier. In this experiment, the cells were incubated with 100b in the dark, or else with a stock of 53. 140  Hyp(Nv)2-Ile3-amaninamide (100b)  120  Hyp2-Ile3-amaninamide (53)  100 80 60 40 20 0  Concentration (µM)  Figure 3.11: Treatment of CHO cells with modified amatoxins 100b and 53 (Structures shown in Figure 3.9) As expected, there was no observable loss of cell viability when CHO cells were treated with the protected amatoxin derivative 100b. Unfortunately, the deprotected version 53 also did not appear to have any cytotoxic effect. These disappointing results point to the fact that our goal of light sensitive control of transcription in whole cells using the probes 100a and 100b will not be readily attainable under the current assay conditions. Based on the extensive SAR data that has previously been compiled for synthetic amatoxins, the problem likely lies in the assay methodology and not in the structure of the probes.  3.2.4 CONCLUSIONS The cellular toxicity of α-amanitin was determined using an MTT-based cell viability assay. We showed that the effect induced by the toxin was dependent on the identity of the cell line. We 125  determined that Chinese hamster ovary (CHO) cells, African Green monkey kidney (Cos-7) cells and human hepatocytes (HepG2) were the most sensitive of the cell lines tested. Even in these most sensitive cell lines, however, we observed that the concentrations of the toxin required to achieve cell death were much greater than the reported Ki of amanitin against RNAP II (~ 3nM).. The cytotoxic effect of 1 in cultured cells was shown to be slow in Caco-2 cells, requiring three days to obtain 80% cell death at the relatively high concentration of 10 µM. Based on the known tight interaction of some of these peptides with their target RNAP II, these results point towards cell penetration of the toxin being the limiting step in the observation of amatoxin induced cytotoxicity. Derivatives lacking a hydroxyl group at Hyp2 (64 and 138), the nitroveratryl derivative 100b and its deprotected counterpart 53 all displayed virtually no effect on CHO cell viability.  3.3 IMPROVED AMATOXIN CELL UPTAKE The slow observation of toxicity that was noted when cells were treated with α-amanitin led us to believe that cell membrane permeability was a limiting factor in achieving efficient inhibition of RNAP II. We postulated that we would be able to improve the sensitivity of various cell lines to 1 (and hence our synthetic amatoxins) if we could determine a method of improving cell uptake. A variety of techniques have been developed for the introduction of foreign molecules into a cell.  These have been developed for several applications, including transfection of  plasmid DNA and drug delivery. Several molecules have been identified that possess the ability to affect the integrity of the cell membrane. Detergents can disrupt the amphipathic nature of the membrane, and permit the free transfer of reagents across this barrier.  Strong detergents such as Triton X-100, sodium  dodecylsulfate (SDS) and Nonidet P-40 irreversibly disrupt membranes, and thereby induce cell death. Milder detergents such as saponin (147) and digitonin (148) have been shown to improve the cell permeability to various reagents, yet still maintain the integrity of the cell membrane. These milder detergents were tested for their ability to improve cell uptake of amatoxins in cell viability assays.  126  3.3.1 CELL VIABILITY IN THE PRESENCE OF SAPONIN AND DIGITONIN The chemical nature of these detergents is very similar to the glycolipid bilayer of the cell. Saponin and digitonin have been proposed to disrupt the cell membrane through the introduction of pores. The size and frequency of these pores is dependent on the time of exposure and concentration of the detergent. When these compounds are removed from the cell media, the pores will close, regenerating an intact membrane. Before either of these detergents was used to improve the uptake of amatoxins, the ideal exposure conditions were determined, such that they did not in themselves lead to cell death. The viability of CHO and/or Cos-7 cell lines treated with these detergents was determined using the MTT assay as described earlier.  The first attempts that were undertaken involved a brief  exposure to the permeabilizing reagents at higher concentrations. This mimics conditions that have been described in the application of fluorescent molecules to live cells. CHO cells were grown in 96-well plates in standard media to approximately 60% confluence. The media was aspirated from the cells, and they were washed with cold phosphate-buffered saline (PBS) and chilled on ice.  The cells were treated with stocks of PBS containing saponin at five  concentrations (from 0.05 to 5 mg/mL) for one, five or ten minute periods. The saponin containing-PBS was removed, the cells were washed with cold PBS, and fresh growth medium was added. The cells were then incubated for two days, following which the number of viable cells was determined. 1 min 5 min 10 min  100 80 60 40 20 0  Concentration (mg/mL)  Figure 3.12: Effect of brief exposure to saponin on CHO cell viability. The data in Figure 3.12 imply that high concentrations and long exposure periods of saponin to CHO cells can affect their viability. Since the goal is to test cell viability of the amatoxins, we 127  must use permeabilization conditions that do not disrupt normal cell function. This requires concentrations lower than 0.1 mg/mL with exposure times of less than five minutes. The practical limitations of this approach in a 96-well plate format led us to test the effect of dilute concentrations of the detergent over longer periods of time. Long-term exposure to low concentrations of these detergents may also prove to be an effective method of improving cell membrane penetration. To ensure this was compatible with our assay system, we set out to determine the concentration ranges most acceptable with both saponin and digitonin. We grew CHO and Cos-7 cells in 96-well plates. The cells were incubated with growth media containing sapoinin or digitonin at concentrations ranging from 2 – 50 µg/mL for a 67-hour period, and the viability of the remaining cells was determined.  CHO  Cos-7  125  120  100  100 80  75  60 50  40  25  20  0  0  Concentration (µg/mL)  Concentration (µg/mL) saponin  digitonin  Figure 3.13: Effect of long-term exposure to low concentrations of saponin and digitonin on CHO and Cos-7 cell viability. These results in Figure 3.13 show that Cos-7 cells seem to be slightly more sensitive to both detergents. In the CHO cell line, there appears to be a greater tolerance for these compounds, and there appears to be no affect on the cell viability at concentrations up to 19 µg/mL. These experiments provide a general guideline as to the concentrations that should be useful for the improved uptake of amatoxins in cell viability assays.  128  3.3.2 SAPONIN AND DIGITONIN MEDIATED α-AMANITIN UPTAKE After determination of the approximate non-toxic operating range of saponin and digitonin, we tested whether or not they could be used to improve the uptake of α-amanitin in CHO and Cos-7 cells. The first attempt involved shocking the cells with higher concentrations of the detergent in the presence of α-amanitin, followed by a 45-hour incubation in medium only. Cells chilled on ice were exposed to 0.1 mg/mL saponin in the presence of varying concentrations of the toxin in PBS for various periods of time. Following one, three or five minute incubation times, the permeabilization mixture was aspirated, and the cells were washed with fresh media, and then incubated in the presence of fresh media at 37 °C for 45 hours. The remaining viable cells were determined using MTT methodology. 120  10 min PBS 1 min saponin 5 min saponin  100 80 60 40 20 0  Concentration (µM)  Figure 3.14: CHO cells were treated with a 0.1 mg/mL saponin for a brief period of time, in the presence of α-amanitin to induce cell death. Cell viability was determined after 45hours of growth in normal media. The results shown in Figure 3.14. As expected, when cells were treated with amanitin in PBS for ten minutes, then replaced with fresh medium, there was no cell death observed. However, when cells were treated with α-amanitin in the presence of saponin in PBS for one or five minutes, some cell death is observed. The degree of loss of viability is about 50% and 35% respectively, at the highest tested concentration of 1. The difficult and irreproducible experimental setup led us to study alternative avenues for the saponin-mediated uptake of α-amanitin.  129  Based on the fact that both CHO and Cos-7 cells were viable when treated long-term with low concentrations of either saponin or digitonin, we decided to use the approach to assay αamanitin. Cells were grown in 96-well plates, in media that was supplemented with 25 µg/mL of saponin or digitonin (for CHO) or 5 µg/mL saponin and 2.5 µg/mL digitonin (for Cos-7). The media also contained varying concentrations of the toxin 1 (5 nM – 1 µM). After a 64-hour incubation, MTT was added to the cells to determine the remaining viable cells.  150  120  125  100  100  80  75  60  50  40  25  20  0  0  Concentration (nM)  Concentration (nM) without saponin  without digitonin  with 25 µg/mL saponin  with 25 µg/mL digitonin  Figure 3.15: Effect of α-amanitin on CHO cell viability in the presence of detergents. 120  120  100  100  80  80  60  60  40  40  20  20  0  0  Concentration (nM)  Concentration (nM)  without digitonin  without saponin with 5 µg/mL saponin  with 2.5 µg/mL digitonin  Figure 3.16: Effect of α-amanitin on Cos-7 cell viability in the presence of detergents. Based on these results, it appears that both saponin and digitonin have the ability to improve the toxicity that is imparted by α-amanitin. The results in Figure 3.15 show that a high degree of 130  background cell death was induced by the detergent alone, making it difficult to interpret these data. However, the data in Figure 3.16, particularly in the case of saponin, a trend begins to appear. In this case, a clear effect on cell viability is observed when the growth medium is supplemented with 5 µg/mL saponin. This implies that the weak cytotoxicity that was obtained previously (Figure 3.7) is at least partially reflective of poor cell uptake of the toxin.  Cell Line  Detergent  CHO Cos-7  Saponin Saponin  Concentration [µg/mL] 25 5  Background cytotoxicity 70% 23%  Estimated ED50 (nM)* 60 60  *not enough data to accurately calculate.  Table 3.5: Results of saponin mediated amanitin-induced cytotoxicity. The data in Table 3.5 identify both the usefulness and the flaws of this approach. When saponin was added to the cell media, we observed an effect in increased amanitin cytotoxicity. However, this comes at the cost of the background cytotoxicity of the detergent itself. It has been reported that sometimes a drug and the permeabilizing agent (both individually non-toxic), can display a synergistic toxic effect when added in conjunction. This may render it difficult to separate the activity of each species. Regardless, it appears that the introduction of the cell permeabilizing reagents saponin and digitonin to cells might lead to improved cytotoxicity of synthetic toxins. Based on the fact that we observed improved toxic effects of α-amanitin in CHO cells in the presence of mild detergents, we attempted to observe a pronounced effect in the toxicity of some amatoxin derivatives under these conditions as well. Two approaches were again used for the introduction of saponin; one was based on the quick treatment of the detergent at higher concentrations (1.5 mg/mL, 5 min), while the other was with prolonged exposure at low concentrations (1.5 µg/mL) for the entire incubation period (66-hours). In both cases, the expected effect on the toxicity of the control toxin α-amanitin was observed, however, both proline-containing derivative 53 and hydroxyproline-containing derivative 64 did not respond to these conditions.  131  A.  B. 150  120 100  100  80 60  50  40 20  0  0  Concentration (µM)  Concentration (µM) α-amanitin (1)  α-amanitin (1) 2  Hyp2-Ile3-amaninamide (53) Pro2-Ile3-amaninamide (64)  3  Hyp -Ile -amaninamide (53) Pro2-Ile3-amaninamide (64)  Figure 3.17: Treatment of CHO cells with amatoxins in the presence of the mild detergent saponin. A. Viability of cells exposed to the toxin in medium containing 1.5 mg/mL saponin for three minutes. B. Viability of cells exposed to the toxin in the presence of 1.5 µg/mL saponin. (For structures, see Figure 3.9) Unfortunately, Figure 3.17 shows there was no observed cytotoxic effect of these derivatives when exposed to CHO cells in the presence of saponin, at peptide concentrations as high as 10 µM. It is possible that at higher concentrations of the derivatives an effect could be observed, but these were not tested.  3.3.3 CONCLUSIONS The use of saponin and digitonin in cell viability assays seemed useful in improving the observed toxicity. This also provided further evidence that cell uptake could be a limiting factor in using amatoxins to inhibit transcription on a cellular level. The practical difficulties in using these detergents to help with this problem, unfortunately make this approach unreliable. The accurate and reproducible application of saponin or digitonin is challenging, and there are noted cases where the detergent and the drug can work synergistically to induce effects not expected by either component individually. There are other approaches that can be used to improve the cellular uptake of amatoxins, which could prove to be more useful and reliable. As noted earlier, specific transporters have been 132  identified that are responsible for the cellular import of amatoxins (OATP1B3 and Ntcp). If a derived cell line that over-expressed these transporters was available, it is reasonable to assume that they would improve the cell uptake of amatoxin derivatives. Alternatively, cell penetrating peptide sequences (CPPs) that are known to cross the cell membrane, and carry cargo with them could be useful in the import of amatoxins.  3.4 APPLICATION OF FLUORESCENT AMATOXIN IN CONFOCAL MICROSCOPY We chose to visualize cellular uptake of the fluorescent amatoxin derivative 101, to provide further evidence for our hypothesis that the cellular uptake of our amanitin derivatives is the limiting step in toxicity. Confocal microscopy techniques should allow for the determination of whether or not the fluorescently labeled amatoxin gains access to the cell. This would hopefully also provide evidence that the probe reaches its target inside of the nucleus.  3.4.1 FIXED CELL UPTAKE First, we needed to determine whether or not the amatoxin 101 truly was a viable probe for visualization of RNAP II binding within a cell. This was undertaken using conditions that are commonly employed in immunostaining applications. Cells are rendered irreversibly permeable prior to treatment with the fluorescent amatoxin. Under these conditions, cell components are no longer active, but their structural integrity is maintained. In this experiment, CHO cells were grown on a glass cover slip for a 24-hour period, to ensure adherence. Cells were fixed by a 15-minute incubation with a 4% solution of paraformaldehyde (PFA). This treatment ensured that all of the cellular components were fixed in place, in their native form. The cover slip was washed with a solution of phosphate-buffered saline (PBS) and could be stored for extended periods of time at 4 °C. To permeabilize the cell membrane, the cover slip was extracted with a 1% solution of Triton X-100. This extraction protocol allows for several types of probes to freely enter the cellular environment, including enzymes and antibodies. The permeabilized cells were thoroughly washed, and then exposed to a cocktail of two different fluorophores, shown in Figure 3.18.  133  N  CO2  O  N  H N  O HO  O N H  O  N H  O O  HN  HO  HO  N  O NH  O  H N  HN  CH3  S N H  N  OH  O  O O  NH CH2  H3C  HN  S N H  H N  O  CH3  S N H  O O N C NH H H2  O  O NH O  H2 O N C H HN  O O  O  NEt2  NH O  149  101  Figure 3.18: Fluorescently labeled peptides used in confocal microscopy experiments. Rhodamine-labeled phallotoxin 149 is often used as a co-stain in confocal microscopy experiments to provide a detailed view of the cell. As mentioned earlier, it binds to F-actin, and when fluorescently labeled, it details cell structure. The diethylaminocoumarin-labeled amatoxin (DEAC-amatoxin) 101 has absorbance and emissions properties that are compatible with rhodamine. This means that we are able to stain both the F-actin and RNAP II simultaneously, detecting each at different wavelengths. The only requirement is that they do not bind the same target. The cells were exposed to these fluorophores for 45 minutes, and then washed with PBS to remove excess and loosely bound dye. To visualize the cells, the stained coverslip is mounted onto a microscope slide, and an antifade reagent (DABCO or a commercially available derivative) is applied to minimize photo-bleaching. The microscope slide was visualized at 488 and 580 nm and images were digitally collected.  134  Figure 3.19: Confocal microscopic image of CHO cells treated with rhodamine-phalloidin (red) and DEAC-amatoxin 101 (blue). The cell membrane was permeabilized with Triton X-100 prior to staining. The white scale bar represents 85 µm. The image in Figure 3.19 represents an overlap of two individually obtained images, where one was taken during exposure to 580 nm light, and the other was taken during exposure to 488 nm light. The colours are added for visual representation, and do not reflect the actual image. The longer wavelength renders the rhodamine-phalloidin active, which is represented as red in the image. The lower wavelength stimulates the DEAC-amatoxin, which is represented in blue. This image shows that when CHO cells are fixed and irreversibly permeabilized, the DEACamatoxin 101 localizes to the nucleus. This is what we expected for binding to RNAP II, as was observed previously in the literature (see Figure 2.8). When treated with the free fluorophore (DEAC), cells do not stain, suggesting that the observed localization is dependent on the amatoxin. This provides good evidence that the DEAC-amatoxin 101 is a suitable probe for the visualization of RNAP II within permeabilized cells.  3.4.2 LIVE CELL UPTAKE The results presented in the last section demonstrated nuclear localization of the fluorescently labeled amatoxin under cell permeabilization conditions. To further show that cell uptake is a 135  limiting factor in amanitin toxicity, we exposed growing cells to 101. CHO cells were adhered to glass coverslips and were exposed to media containing DEAC-amatoxin for one hour. After this time period, the cells were fixed with PFA and directly mounted onto glass slides. A.  B.  Figure 3.20: A. Bright field image of CHO cells, obtained with a 10X objective. B. Confocal microscopy image of CHO cells treated with 1 µM 101 in medium for a 24-period without membrane permeabilization. The white scale bar represents 85 µm. In this study, rhodamine-phalloidin staining of the cell structure could not be achieved based on the requirement of cell membrane permeabilization. The image obtained following exposure to 488 nm light shows nothing but slight background fluorescence in the cell, and no clear nuclear localization of the amatoxin. This implies that no uptake of the fluorescent amatoxin in CHO cells occurred during exposure to media containing the probe when no permeabilization steps were employed.  3.4.3 SAPONIN MEDIATED UPTAKE Based on the improved toxicity of α-amanitin that was observed when the cells were treated with the mild detergent saponin, we chose to test this approach to improving uptake of the fluorescent probe. We exposed adherent CHO cells to DEAC-amatoxin in the presence of saponin using two approaches. In one case, we treated the cells with PBS containing the amatoxin (1 µM) and a 136  high concentration of saponin (1 mg/mL) for a short five-minute exposure at 0 °C, followed by one hour with medium and amatoxin only (1 µM) at 37 °C. We repeated this experiment with a 24-hour exposure to 101 (1 µM) in the cell medium at 37 °C. A.  B.  Figure 3.21: Cellular uptake of DEAC-amatoxin in the presence of the mild detergent saponin. In both cases, cells were treated with 1 mg/mL saponin in PBS, containing 1 µM 101 for 5 min at 0 °C. The cells were then incubated with fresh medium containing 1 µM 101 at 37 °C for various times. A. Image obtained following a 1 h incubation period. B. Image obtained following a 24 h incubation period. In both images the white scale bar represents 85 µm. As is demonstrated in Figure 3.21, saponin was able to improve the cellular uptake of the fluorescent DEAC-amatoxin 101. In the case where the cells were exposed only briefly to the amatoxin (1 hour, following permeabilization), the localization of the probe to the nucleus can clearly be seen, along with the auto-fluorescence of the cell.  When cells were exposed to the  fluorescent probe for a full day following saponin treatment, the image obtained showed that the cell integrity was lost. In this case, it appears that cellular necrosis had taken place. While it is impossible to tell whether the cells were dead based on the inhibition of RNAP II by 101 or else due to exposure to the detergent, this is a promising result.  137  3.5 CONCLUSIONS Here, we investigated the ability to study the toxic effects of amanitin and its derivatives through a cell-based viability assay. The readout of this assay is based on the ability of viable cells to produce mitochondrial reductase, which in turn converts the dye MTT to a quantifiable fomazan product. We were able to show that the toxic effect induced by exposure to α-amanitin is dependent on the cell line examined. Cells derived from Chinese hamster ovaries (CHO) and African green monkey kidneys (Cos-7) proved to be the most sensitive to the toxin. The effect of α-amanitin on these cell lines was shown to be slow, and the observed cytotoxic concentration ranges were much higher than the toxin’s reported RNAP II inhibition constants (Ki). The effects of some of our synthetic amatoxin derivatives did not display any cytotoxicity. We attempted to achieve improved cell uptake of α-amanitin in CHO and Cos-7 cells using the mild detergents saponin and digitonin in the MTT cell viability assay. We determined the effective concentration ranges where the cells maintain their integrity in the presence of longterm exposure to these detergents, finding that 5-25 µg/mL is acceptable, depending on the detergent and cell line. When α-amanitin was applied to cells that were also treated with these detergents, we were able to see a marked increase in the cytotoxic effects of 1 of over an order of magnitude. To further verify the hypothesis that cellular uptake of amatoxins is limited, we used a fluorescently labeled amatoxin to visualize cell penetration. Images that were obtained using a confocal microscope showed that there was not any cellular uptake of fluorescent amatoxin derivative 101 unless the cell membrane was first permeabilized. This was observed in fixed cells, using the detergent Triton X-100, and also in live cells, when they were first treated with a solution of 1 mg/mL of the mild detergent saponin at 0 °C. When cells were treated with 101 for a 24-hour period following brief permeabilization of the cell membrane with saponin (1 mg/mL), the cells seemingly had undergone necrosis. The limiting factor in these cell viability studies appeared to be the cellular uptake of the amatoxins.  While we were able to improve the uptake and toxic effects of α-amanitin to a  respectable level, we were not able to achieve the same levels using the synthetically prepared amatoxins. We always expected that these derivatives would be less active than the natural 138  product, yet we did expect to be able to see some level of inhibition. Some of the derivatives tested are known to be functional at higher concentrations in in vitro assays. For example, 64 has a Ki that is 32 times worse than 1, while 53 is 880 times less effective. Since 1 was shown to be 50% active at concentrations of 0.5 – 1 µM, we would expect the same level of activity at 1632 µM for 64. Unfortunately, this was not observed in our study. Of all the derivatives that were tested here, the dihydroxyisoleucine residue in position three of the natural product was introduced as an isoleucine (dhIle3 → Ile3). This was chosen for two main reasons. The first is reason is based on the fact that the co-crystal structure of amanitin binding site of RNAP II showed flexibility for this side chain. This implies that the residue does not provide critical binding energy. Secondly, the synthesis and incorporation of dhIle is not trivial. We propose that the side chain at dhIle3 may provide an important feature related to the solubility and uptake of amatoxins. There are other methods by which we could improve the cellular uptake of synthetic amatoxins. While the addition of detergents to the cell growth medium proved to be unreliable and difficult to obtain satisfactory results, we could use other carriers. Specifically, we could synthesize amatoxins that are conjugated to cell penetrating peptides or other cell permeable structures. Alternatively, cells expressing the transporters that are associated with uptake of α-amanitin could prove to increase the cytotoxic effects of various synthetic amatoxins. The pursuit of these directions is required to achieve reliable spatio-temporal control of RNAP II inhibition.  139  3.6 EXPERIMENTAL SECTION 3.6.1 MATERIALS Cells were cultured in α-MEM or high-sucrose DMEM, purchased from Gibco. Fetal bovine serum (FBS), 0.25% trypsin (with 1.3 mM EDTA), 0.85% Trypan blue, and the antibiotic mixture Pen/Strep (10K U/mL penicillin, 10K µg/mL streptomycin) were also purchased from Gibco. All cell culture plasticware was obtained from Corning or Falcon. Cells were cultured at 37 °C in a humidified chamber with 5% CO2. Cell membrane permeabilizing agents Triton X100, saponin and digitonin were obtained from Sigma-Aldrich.  When used in cell culture,  DMSO was purified by filtration through a 0.2 µm filter. Paraformaldehyde was obtained as a 4% solution from Biorad. The anti-fade reagent ProLong Gold and rhodamine-phalloidin were from Invitrogen. Colourless nail polish was obtained from Revlon. All experiments were carried out in a laminar flow culture cabinet, unless otherwise noted. Absorbance measurements of the 96-well plates were obtained using a Beckman-Coulter DTX 880 multimode detector, equipped with an excitation filter of 595 nm. Bright field images were obtained on an Olympus 1X70 inverted microscope. Fluorescent images were obtained on an Olympus FV1000 confocal microscope fitted with a 405 nm blue-violet diode laser, 488 nm Argon-ion laser and a 543 nm He-Ne laser. Microscope slides and coverslips were obtained from Fisher Scientific. The various cell lines used in this study were gifts from various researchers at UBC. Cells grown in α-MEM were: CHO cells, from the research group of Dr. Brian James, Caco-2 cells, from Dr. Chris Orvig, and HepG2 cells from Dr. David Chen. Cells grown in DMEM were: Hfl1 cells, from Dr. Darryl Knight (UBC iCapture), and Cos-7 cells, from Dr. Robert Molday (UBC biochemistry).  3.6.2 GENERAL CELL CULTURE All of the cell lines that were used were immortalized, and stored in liquid nitrogen. To revive cells, a 1 mL tube of the frozen cells in medium containing 10% DMSO was warmed in a 30 °C 140  water bath, and diluted with 9 mL of fresh media. All fresh media contained 10% FBS, 100 U/mL penicillin and 100 µg/mL streptomycin, unless otherwise indicated.  The cells were  incubated in a T-25 flask at 37 °C at 5% CO2. After 24 hours, the medium was aspirated and replaced with fresh medium. When cells reached a level of 90 -100% confluence, they were subcultured. The medium was removed, and the cells were treated in the incubator with 0.25% trypsin containing 1.3 mM EDTA. Once the cells were detached from the tissue culture flask, 35 mL media was added to quench the trypsin, and transferred to a 10 mL centrifuge tube. The mixture was centrifuged for 5 min at 800 rpm, and the supernatant was discarded. The cells were suspended in fresh medium, diluted as required, and transferred to new culture flasks. This process was repeated every 3-5 days, until approximately 100 passages, at which point a new vial of frozen cells was revived.  3.6.3 CELL VIABILITY ASSAYS To assay cell viability, a nearly confluent tissue culture flask was trypsinized, and the cells were counted following treatment with Trypan blue, using a hemacytometer. The cells were then diluted to the appropriate stock concentrations in fresh medium, and transferred in 150 µL to a 96-well plate using a multi-channel pipette. The number of cells plated varied from experimentto-experiment. These were incubated a 37 °C and 5% CO2 for a 24-hour period to allow for plate adherence. The medium was aspirated, and fresh medium was added, which contained the desired additives (amatoxin, permeabilizing agents, etc). The cells were then re-incubated for the desired time. At the completion of the experiment, a 50 µL aliquot of 2.5 mg/mL MTT in PBS was added to each well. The plate was incubated three hours further, to allow for the formation of the formazan product in viable cells. The media was carefully aspirated, and the purple precipitate was dissolved in DMSO. The absorbance of each well was recorded at 595 nm, and an image of the plate was generated using a scanner. Data was processed in Microsoft Excel and GraphPad Prism. Experiments were performed in triplicate, unless otherwise noted, and the error was calculated as the standard error of the mean.  141  3.6.3.1 SPECIFICATIONS OF EACH VIABILITY ASSAY Effect of cell line identity on the cytotoxic effect of α-amanitin: Trypsinized cells were diluted to a concentration of 3.3 x 104 cells/mL for the Caco-2, CHO, HepG2 and Hfl cell lines, and 3.3 x 103 cells/mL for Cos-7. Each cell line was plated in a 96well plate, with 150 µL of the stock solution per well (5000 or 500 cells per well), and incubated for 24 hours.  Stocks of α-amanitin were prepared in α-MEM or DMEM at various  concentrations, containing 0.5% DMSO, and added to various wells, according to the desired final concentration (25, 10, 7.5, 5, 2, 1, 0.75, 0.5, 0.2, 0.1 µM). The cells were incubated for 66 hours, at which point viability was assessed as described. Effect of exposure time of α-amanitin on Caco-2 cell viability: A confluent flask of modified Caco-2 cells was trypsinized and diluted to a final concentration of 2.5 x 105 cells/mL. These were plated in a 96-well plate such that each well contained 50000 cells, and grown for 72-hours in α-MEM, supplemented with the antibiotic G418. At varying time points, the media was diluted with a 2X stock of fresh media containing 100 µM α-amanitin and 1% DMSO. Following completion of the assay, the cell viability was assessed with MTT. The various incubation times in the presence of the toxin tested were: 0, 535, 1130, 2035, 2815, 3465 and 4350 minutes. Effect of amatoxins on cell viability in CHO cells: A 96-well plate was seeded with CHO cells to a final concentration of 50000 cells per well. The cells were left in the incubator overnight to become adherent. The Pro-amatoxin derivatives (64 and 138) were dissolved in α-MEM with 0.5% DMSO to final concentrations of 100, 10, 1 and 0.1 µM. The Hyp-amatoxins (53 and 100b) were also dissolved in media containing 0.5% DMSO, to a final concentration of 50 and 5 µM. These compounds were applied to the CHO cells, and incubated for 48 hours, followed by MTT evaluation of cell viability. Effect of saponin and digitonin on CHO and Cos-7 cell viability: 1. Quick saponin treatment: CHO cells were grown in a 96-well plate to 80% confluence in αMEM. The media was aspirated and the cells were treated with stock solutions of saponin in PBS at concentrations of 5, 1, 0.5, 0.1 and 0.05 mg/mL at room temperature. At times of 1, 5 or 10 minutes, the saponin was removed and the cells were washed three times with PBS, and fresh 142  medium was added. As a control, cells were treated 10 minutes with PBS. The cells were incubated 48 hours at 37 °C, and then assessed for viability. 2. Prolonged saponin and digitonin treatment: Separate 96-well plates were seeded with 5000 CHO or 500 Cos-7 cells per well. After 24 hours, the cells were treated with saponin or digitonin in α-MEM or DMEM, at final concentrations of 50, 25, 10, 5 and 2 µg/mL. The cells were incubated for 67 hours, and cell viability was determined using MTT. Use of saponin and digitonin to improve cytotoxicity of α-amanitin in CHO and Cos-7: CHO and Cos-7 cells were seeded in 96-well plates at final concentrations of 5000 and 500 cells/well respectively. Following a 24-hour incubation, 2X α-amanitin stocks were prepared in suitable media with concentrations of 2, 1, 0.2, 0.1 0.02 and 0.01 µM. Stocks of saponin (50 µg/mL for CHO, 10 µg/mL for Cos-7) and digitonin (50 µg/mL for CHO, 5 µg/mL for Cos-7) were also prepared at a 2X concentration. The appropriate mixture of permeabilizing agent and amanitin were mixed in each well to obtain the desired final concentrations, and the cells were placed back in the incubator. After 64 hours, the viability of the cells was determined. Use of saponin to improve cell uptake of amatoxins in CHO cells: CHO cells were grown to 85% confluence in a 96-well plate, at which point the media was replaced with 90 µL α-MEM supplemented with 1.5 µg/mL saponin. Following five minutes at room temperature, 10 µL of 10X stocks of the various amatoxins in 2.5% DMSO were added to produce final concentrations of 10, 5, 1, 0.5 and 0.1 mM of the peptide and 1.35 µg/mL saponin. After 44 hours at 37 °C, the viability of the cells was assessed using MTT.  3.6.4 CONFOCAL MICROSCOPY All confocal cell imaging was carried out on the CHO cell line, cultured in α-MEM media. Cells were grown in 12-well plates on glass cover slips that were sterilized with ethanol and air-dried. Assays were undertaken once the cells reached approximately 70% confluence. Cell fixation: To fix cells, the media was removed and they were washed with cold PBS. The cells were then exposed to 1 mL of a solution of 4% paraformaldehyde at 4 °C for fifteen  143  minutes. The cover slips were washed with cold PBS, and stored immersed in PBS until further used. Rhodamine-phalloidin and DEAC-amatoxin staining: Fixed cells were rendered permeable by a 15 min exposure to 0.1% Triton X-100 in PBS at room temperature. The cells were then washed three times with cold PBS. Stock solutions of the dye (17 nM for phalloidin 149, 600 nM for amatoxin 101) in PBS were spotted onto parafilm (70 µL final). The coverslip with the fixed cells was then placed over the dye solutions, which created a seal. These were incubated in a dark humidified chamber for thirty minutes at room temperature. The coverslips were then washed four times with PBS and mounted. Mounting: Glass microscope slides were spotted with a drop of glycerol containing an antifade reagent (DABCO or ProLongGold). The coverslip was placed on this drop, face down, and sealed onto the slide using clear nail polish. Slides were stored at 4 °C. Saponin uptake assay: CHO cells were grown on sterilized glass coverslips in 12-well plates to 60% confluence. The media was removed, and the cells were treated with either cold PBS or cold PBS containing 1 mg/mL saponin and 1 µM DEAC-amatoxin for five minutes on ice. The cells were then aspirated, and washed three times with cold PBS. Media containing 1 µM DEAC-amatoxin was added back to the slides, and then incubated for one hour at 37 °C. In some cases, the media also contained 1.5 µg/mL saponin. The cells were then fixed and directly mounted, as described above.  144  CHAPTER 4: IMPROVED METHODOLOGY FOR THE PREPARATION OF AMATOXINS  4.1 INTRODUCTION Our approach to generate synthetic amatoxin derivatives for their use as intracellular probes of transcription thus far has proven to be limited. While the synthesis and desired chemical properties of these probes were demonstrated in Chapter 2, their functional application in cellbased assays was found to be hampered by cell-uptake in Chapter 3. At this point we were in a position where we needed to address the cell permeability of the amatoxins. We chose to address this from a chemical standpoint, wherein we focused on the amino acid composition of the amatoxins. In this chapter we revisit the synthesis of these peptides from a variety of standpoints, focusing on improved methodology as well as an alternative synthetic approach to this peptide scaffold. We also attempted to apply the tryptathionine crosslinking approach to alternative peptide targets.  Specifically, this chapter is split into three sections:  the first  concentrates on the minimization of epimerization during macrolactamization, the second describes our efforts to efficiently prepare a small library of position three modified amatoxins, and the third relates our application of tryptathionine crosslinking to a small family of peptides known as enkephalins.  4.1.1 CHAPTER GOALS The goals in this chapter reflect our desire to improve on the synthetic methodology that is currently employed to generate cyclic peptides bearing a tryptathionine crosslink. Recall that the most inefficient step experienced in the synthesis of the amatoxin probes described in Chapter 2 was macrolactamization. This reaction could yield upwards of 50% of the undesired epimer at the C-terminal isoleucine residue. We wanted to minimize the epimerization that is obtained upon macrolactamization of the amatoxin peptides.166 This unwanted side reaction reduced the yield of an already low yielding process, and produced structures that are not active against RNAP II. We will approach this problem through a trial-and-error approach, modifying the conditions used to generate the macrolactam. 145  We will then attempt a modified synthetic route to the amatoxins.  Based on the lack of  intracellular activity of our probes designed in Chapter 2, we decided to pursue the generation of a small library of amatoxins that vary in the amino acid identity at position three. To do this efficiently, we needed to develop a new convergent approach to the generation of the amatoxin scaffold. We will explore the possibility of adding functionality to a shorter version of the amatoxin scaffold following solid-phase synthesis. If successful, this would allow for a rapid route by which we could generate several amatoxin analogues. Finally, we will lay down some of the groundwork that will contribute to our hypothesis that the tryptathionine linkage could be used as an important pharmacophore in SAR studies of generic peptides. Based on the synthetic methodology developed to generate the library of position three analogs of amanitin, we will prepare four analogues of the pentapeptide enkephalin. These are designed based on the structure of the opioid agonist enkephalin, and synthetic derivatives known as DPDPE and lanthionine enkephalin (vide infra).  4.2 ELIMINATION OF EPIMERIZATION DURING MACROLACTAMIZATION During the synthesis of a peptide bond, it is important that the stereochemical integrity of the αcarbon be maintained.228-230 An epimerization (or enantiomerization)231 event during peptide coupling will generate undesired diastereomeric byproducts. Several improvements in peptide bond formation have been made such that epimerization can be minimized.  This has been  largely achieved through the development of better coupling reagents,232 protecting groups,233 and the use of solid-phase synthesis.234,235 In most generic peptide-bond forming reactions, the carboxylic acid function is activated such that it bears a good leaving group. It is this activated species that is sensitive to scrambling of the stereochemistry of the α-carbon.236 There are two accepted routes that contribute to epimerization of this fragment (Figure 4.1): one is through direct tautomerization of the activated amino acid,233,237,238 and the other involves formation of an oxazolone intermediate.239-241  146  O  H N  R O  OH  activation  R1 O  H N  R  STARTING MATERIAL  O  R1  O  NH2-R O R  R1 N H  H N  OH  H N  R  OLg  OLg R1  O  H N  R O  OLg NH2-R  R1  tautomerization pathway O  intramolecular cyclization  R  R  O  DESIRED PRODUCT  R  R O  NH2-R N  N  R1  N H  O OH  R1  N  H N  R  O EPIMERIC PRODUICT  R  O O  R1  NH2-R O  R1  oxazolone pathway  Figure 4.1: Tautomerization and oxazolone formation are the generally accepted mechanisms that lead to scrambling of the configuration of the α-carbon. A variety of factors increase the likelihood of epimerizing the α-carbon, and this has led to a well-established series of conditions that are useful in suppressing this undesired side reaction. If the rate of amide bond formation is slow, the lifetime of the activated species increases. Since it is the activated species that leads to epimerization, it is best to have a short lifetime. This has been largely overcome through the use of solid-phase synthesis, where multiple equivalents mimic pseudo-first order conditions, and the rate of the reaction is increased. The use of carbamates as N-terminal protecting groups (as opposed to amides), has gained popularity based on their ability to suppress epimerization. These protecting groups are thought to lower the pKa of the resultant α-carbamate, thereby favouring deprotonation of the NH over tautomerization. Finally, the use of sterically bulky bases, such as DIPEA or 2,4,6,trimethylpyridine reduce epimerization.242 Another important factor that can affect the extent of epimerization during peptide bond formation is the coupling reagent.232,243-246 Numerous coupling reagents have been designed to increase the efficiency of peptide coupling, and simultaneously reduce epimerization. Unfortunately, there does not seem to be a clear trend in the choice of the ideal coupling reagent to minimize this side reaction, and each individual situation requires a trial-and-error approach. 147  N N C N N  EDC (150)  N  C  N  N  N N O  N PF6  N  DCC (151)  HATU (152)  PF6 Br N  N  PyBrOP (153)  N  O  N P  O  P  N3  O  DPPA (154)  N  N N OH  HOBt (155)  Figure 4.2: The structures of various coupling reagents and additives that are used in peptide synthesis. In the case of macrolactamization reactions, such as the one discussed here, epimerization of the α-carbon is more frequently observed than during the synthesis of linear peptides.173,247 Since this is an intramolecular reaction, the pseudo-first order rate conditions obtained through multiple equivalents of one reacting species cannot be achieved. The formation of the large ring size is also complicated by the competing formation of dimers, and the reaction must be performed at high dilution.178,248 The strain imposed upon formation of these large rings leads to a relatively long lifetime of the activated carboxylate, and therefore pronounced epimerization during cyclization reactions is often observed. The most common factors that are varied in an attempt to reduce epimerization during macrolactamization include the coupling reagent, the solvent and the choice of organic base.177,247,249 During the course of preparation of the amatoxin derivatives described in Chapter 2, the formation of the undesired epimer at Ile3 was observed following macrolactamization. We also noted that the extent of epimerization was not consistent. Products that bore functionality on the hydroxyproline group (as in derivatives 100a and 100b) showed less epimerized product relative to the proline analog 64. Originally, this was thought to be a steric influence during the cyclization reaction, however, upon further review it was noted that the source of the solvent (DMF) during the reaction seemed to have an effect on the extent of epimerization. The inconsistent formation of the undesired epimer encouraged us to investigate methods to reduce this side reaction. Generally, this reaction was performed at high dilution (1 mM peptide) 148  in the presence of three equivalents of PyBOP and excess DIPEA in dry DMF. There were three sets of conditions that we first decided to probe to achieve the most efficient macrolactamization. To test this, we conducted this reaction varying: 1) the identity of the coupling agent, 2) the reaction solvent, or 3) the water content in the reaction solvent. O H2N  O  O N H  HO  O  H N  N  O  H2 O N C H HN  122  S N H O O N C NH H H2  H2N O conditions  O  H N  HN O O  S N  H N  O  O  H2N O  N H N H  H2 O N C H HN  HN  O  N  O  H N O O S  O  C NH H2  H N  O  O  H2N  64  O  N H N H  H2 O N C H HN  O  O C NH H2  138  Figure 4.3: Macrolactamization control reaction carried out to test for conditions that reduce epimerization. The control reaction shown in Figure 4.3 was used to test these conditions. This reaction uses the monocyclic peptide 122, containing minimal modification in the side chain residues. This substrate was chosen since our lab has previously identified the desired bicyclic product (64) and its epimer (138) through crystal structure analysis.166 This allowed for the rapid and accurate detection of the formation of the two products using reversed-phase HPLC analysis. The work in this section of the thesis (Section 4.2) was performed as a collaborative effort in our lab. Former post-doctoral researcher Dr. May performed the macrolactamization reactions and HPLC separations in this section, and the author performed the HPLC analysis and work-up of the data presented herein.  149  4.2.1 EFFECT OF COUPLING REAGENT One of the most common methods used to eliminate epimerization during peptide bond formation is to change the coupling reagent. A number of the different reagents have proven to be useful in a variety of situations.232,243 Unfortunately, there is no clear trend as to which reagents are better or worse, so we must use a trial and error approach. We first chose to compare the extent of epimerization noted with PyBOP (137) to that obtained with EDC (150), DPPA (154) and PyBrOP (153). The macrolactamization of 122 was carried out under dilute conditions (1 mM peptide) in dry DMF with DIPEA (3 equiv.) added as a base. The coupling reagent (3 equiv.) was added and the reaction was stirred at room temperature overnight (~ 18 hours). To analyze the extent of epimerization, the solvent was evaporated, and the products were directly resuspended in a water-acetonitrile mixture (1:1), and analyzed on reversed-phase (C18) HPLC. epimer product  138  0  5  10  15  64  20  25  PyBOP DPPA PyBrOP EDC  30  Retention Time (min)  Figure 4.4: Section of the HPLC chromatograms (monitored at 292 nm) of the crude reaction following macrolactamization of 122 using various coupling agents. The desired product and its epimer were identified by their retention times (20.6 min and 18.7 min respectively), which were determined using previously crystallized and verified products. The HPLC traces shown in Figure 4.4 imply that of the various coupling agents tested, there was no dramatic inhibition of epimerization. The relative area under each peak representing the two products was integrated and normalized to obtain a numerical reflection of the results. 150  122  Coupling agent Additive Base DMF  64 (desired)  + 138 (epimer)  Product ratio  Coupling Agent  Additive  Base  PyBOP (3 equiv.)  --  DIPEA (3 equiv.)  1.1 : 1  DPPA (3 equiv.)  --  TEA (3 equiv.)  1.6 : 1  PyBrOP (3 equiv.)  --  DIPEA (3 equiv.)  N.D.  EDC (1.25 equiv.)  HOBt (1.25 equiv.)  NaHCO3 (1.25 equiv.)  1.8 : 1  (64 : 138)  Table 4.1: Summary of effect of various coupling reagents on epimerization at Ile3 during macrolactamization of 122. (N.D. = not determined) (For structures see Figure 4.3) Of all of the coupling reagents tested, our original choice of PyBOP was shown to be the most ideal. While EDC and DPPA did display slightly better selectivity for the desired epimer (approximately 2:1 versus 1:1), they did pose some issues. A number of unidentified byproducts were observed when DPPA was used, and only small amounts of products were obtained when EDC was used. Performing the macrolactamization reaction using PyBrOP did not produce any significant amount of the desired product, and yielded mainly unidentified byproducts. The additive hydroxybenzotriazole (HOBt) has also been reported to effect epimerization in coupling reactions with varying results. When the macrolactamization of 122 was performed in the presence of HOBt and PyBOP, no difference was observed compared to the reaction that did not contain the additive (not shown). It is clear from these results that PyBOP in the absence of HOBt generates the least amount of byproducts, cleanly forms the desired bicyclic peptide, yet produces a 1:1 mixture of the desired product (64) and its epimer (138). Of the alternative coupling reagents tested, no significant reduction of epimerization was noted, and more byproducts were observed when DPPA and PyBrOP were used.  4.2.2 EFFECT OF SOLVENT The reaction solvent has also been reported to have an effect on the formation of epimeric products during peptide bond formation.250,251 Most often, this reaction is carried out in polar aprotic solvents.165 This allows for the dissolution of all of the components of the reaction, and is optimal for nucleophilic attack of the amine group. The most commonly employed solvent in 151  peptide bond forming reactions is DMF, but N-methylpyrrolidone (NMP) has generated interest as well. This solvent has similar properties to DMF, but is thought to provide better solvation of the peptide chain. One drawback of NMP is that it has a much higher boiling point than DMF (202 °C vs 154 °C), making its removal through evaporation more challenging. We chose to test the extent of epimerization in NMP, and also to compare it to acetonitrile (MeCN), and methylene chloride (CH2Cl2).  epimer product 138  0  5  10  15  64  20  25  DMF MeCN NMP CH2Cl2  30  Retention Time (min)  Figure 4.5: Section of the HPLC chromatogram (monitored at 292 nm) obtained following macrolactamization of 122 in various solvents. In this set of reactions, the macrolactamization was carried out at high dilution, using PyBOP (3 equiv.) and DIPEA (3 equiv.). Everything was added to the reaction at once, and the reactions were stirred overnight (~18 hours). The solvents were evaporated (as best possible, in the case of NMP), and the crude mixture was resuspended in a water-acetonitrile mixture (1:1) and analyzed by HPLC. Based on the HPLC traces shown in Figure 4.5, it appears that the reaction was not successful in the solvents NMP and CH2Cl2. However, in the case of acetonitrile, the reaction worked, and the degree of epimerization was reduced.  152  122  PyBOP (3 equiv.) DIPEA (3 equiv.) Solvent  Solvent  64 (desired)  + 138 (epimer)  Product ratio (64:138)  DMF  1.1 : 1  MeCN  6.7 : 1  NMP  N.D.  DCM  N.D.  Table 4.2: Summary of the effect of solvent on epimerization at Ile3 during macrolactamization of 122. (N.D. = not determined) (For structures see Figure 4.3) It is important to note that even though some of these reactions repressed formation of the undesired epimer 138, the formation of byproducts was much more significant. The difficulty of removing NMP, and the undesired byproducts observed rendered this solvent less useful. It is possible that optimization of the macrolactamization in NMP and CH2Cl2 could yield the desired product with minimal epimerization, but this was not pursued. Acetonitrile proved to be quite a suitable solvent for this reaction, as it did provide a suppression of epimerization. This solvent system also did not show the formation of any of the byproducts generated in other solvents.  4.2.3 THE EFFECT OF WET DMF AS SOLVENT While the use of acetonitrile as solvent did produce promising results, we wanted to investigate another possibility. During the course of the preparation of various amatoxin analogs in our lab, we noticed that there was a discrepancy in the formation of the epimer during macrolactamization. It was shown that the DMF that w