UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Marine natural products : synthesis and isolation of bioactive analogues Pereira, Alban R. 2007

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

Item Metadata

Download

Media
24-ubc_2008_spring_pereira_alban.pdf [ 20.12MB ]
Metadata
JSON: 24-1.0061556.json
JSON-LD: 24-1.0061556-ld.json
RDF/XML (Pretty): 24-1.0061556-rdf.xml
RDF/JSON: 24-1.0061556-rdf.json
Turtle: 24-1.0061556-turtle.txt
N-Triples: 24-1.0061556-rdf-ntriples.txt
Original Record: 24-1.0061556-source.json
Full Text
24-1.0061556-fulltext.txt
Citation
24-1.0061556.ris

Full Text

MARINE NATURAL PRODUCTS: SYNTHESIS AND ISOLATION OF BIOACTIVE ANALOGUES by ALBAN R. PEREIRA B.Sc, University of Costa Rica, 2000 M.Sc, University of Costa Rica, 2002 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDDZS (Chemistry) THE UNIVERSITY OF BRITISH COLUMBIA December 2007 © Alban R. Pereira, 2007 ABSTRACT Tauramamide (2-12), a linear acylpentapeptide recently isolated from cultures of Brevibacillus laterosporus (PNG-276) collected in Papua New Guinea, was synthesized in 9 steps and 29% overall yield. Besides confirming the proposed structure, synthetic (2-12) allowed the antimicrobial assessment of this novel antibiotic. Additionally, a new analogue of the surfactin depsipeptides family named dealkylsurfactin (2-48), was prepared in 10 steps and 14% overall yield. The compound was employed as a biological tool in binding studies between the mitotic regulator isomerase Pinl and the microtubule-associated protein tau, a crucial interaction involved in Alzheimer's disease. Chemical exploration of Garveia annulata, a seasonal hydroid collected in Barkley Sound, British Columbia, led to the isolation of twelve secondary metabolites including four new compounds (3-53 to 3-56). Nine of these metabolites showed inhibition of indoleamine 2,3- dioxigenase (IDO), with the annulins among the most potent in vitro IDO inhibitors isolated to date. IDO plays a central role in immune escape, which prevents the immunological rejection of tumors or the allogeneic fetus. The ceratamine inspired antimitotic agent (4-142) and inactive analogue (4-157) were synthesized in no more than 8 steps, with overall yields of 20% and 15% respectively. Activity evaluation of these analogues suggested that potency improves with planarity and that the synthetically laborious imidazo[4,5,^/]azepine core heterocycle of ceratamines is not required for activity. Haplosamate A (5-62), isolated from the marine sponge Dasychalina fragilis collected in Papua New Guinea, was found to be the first member of a new family of cannabinoid-active sterols. Saturation transfer double-difference (STDD) NMR experiments confirmed that (5-62) ii specifically binds the cannabinoid human receptors CB1 and CB2 via the classical cannabinoid pharmacophore. A growing appreciation of the therapeutic potential of PI3K inhibitors has encouraged the development of new inhibitory compounds with enhanced potency, selectivity and pharmacological properties. Such substances are destined to the treatment of inflammatory and autoimmune disorders as well as cancer and cardiovascular diseases. An optimization program intended to develop more stable and isoform-selective PI3K inhibitors based on the marine- derived natural product liphagal (6-1), led to the preparation of a small library of synthetic analogues. HN / O H ^ r S ^ r V a V NH HNT^NH2 OH 2-12 O ^ O H Q ^ , N H u ^ ^ ^ NH H N ^ S D I H H 7 crr^xx OH 2-48 o \ 0 OH HO 3-54 3-55 3-56 4-142 4-157 CHO 111 TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vii List of Figures xi List of Abbreviations xvi Acknowledgements xix Dedication xx 1. Marine Natural Products Chemistry: Overview 1 1.1. Past and Present 1 1.2. Secondary metabolites? 2 1.3. Biological activity and current success in marine natural products research 5 1.4. Synthesis of marine natural products 9 2. Bioactive Marine Peptides: Total Synthesis of Tauramamide and Dealkylsurfactin 11 2.1. Peptides from the sea 11 2.2. Peptides from marine PNG bacterial isolates 12 2.2.1. New antibiotics from PNG-276 12 2.2.2. Anti-Alzheimer activity of PNG10A 15 2.3. Total synthesis of tauramamide 18 2.4. Preparation of dealkylsurfactin 29 2.5. Conclusions 41 2.6. Experimental 44 2.6.1. Total synthesis of tauramamide and tauramamide ethyl ester 48 2.6.2. Synthesis of dealkylsurfactin 62 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata. 71 3.1. IDO inhibition and Garveia annulata 71 3.2. IDO and the kynurenine pathway 72 iv 3.3. Current IDO inhibitors and the bioassay 75 3.4. Secondary metabolites from hydrozoans 77 3.5. Minor metabolites from Garveia annulata 80 3.6. Isolation of IDO-active minor metabolites 83 3.6.1. AnnulinC 86 3.6.2. 2-Hydroxygarveatin E 90 3.6.3. Garveatin E 93 3.6.4. Garvin C 96 3.7. Biological evaluation of G. annulata secondary metabolites 101 3.8. Conclusions 105 3.9. Experimental 108 4. Progress towards the Synthesis of Ceratamines 112 4.1. Microtubule-stabilizing agents from marine origin 112 4.2. Ceratamines A and B 114 4.2.1. Biosynthesis and similar marine metabolites 116 4.2.2. NMR data 119 4.2.3. Retrosynthetic analysis 121 4.3. Related syntheses and relevant synthetic methodologies: literature review 123 4.4. Initial synthetic proposals 128 4.5. Attempted preparation of ceratamines and analogues 130 4.5.1. Biomimetic approach to ceratamines 130 4.5.2. First generation approach to ceratamines 132 4.5.3. Second generation approach to ceratamines 151 4.6. Conclusions and future directions 169 4.7. Experimental 173 5. Cannabinoid activity of the Marine Sterol Haplosamate A: Direct Observation of Binding to Human Receptors by Saturation Transfer Double-Difference (STDD) NMR Spectroscopy 212 5.1. Brief history 212 5.2. Marijuana: chemistry and psychoactive effects 215 v 5.3. Phytocannabinoids 218 5.4. Cannabinoid receptors, endogenous cannabinoids and the endocannabinoid System 219 5.4.1. Cannabinoid human receptors 220 5.4.2. Endocannabinoids 222 5.4.3. The endocannabinoid system 223 5.5. Synthetic and patented cannabinoids 225 5.6. Cannabinoids and cancer 229 5.7. Biological evaluation of cannabinoid compounds 230 5.8. Isolation of haplosamate A 232 5.9. Haplosamate A: NMR analysis 234 5.10. Pharmacophoric requirements for cannabinoid activity 240 5.11. Evidence of haplosamate A binding to CB1 and CB2: STD NMR experiments... 246 5.12. Conclusions 255 5.13. Experimental 258 6. Synthesis of Liphagal Analogues 275 6.1. Liphagal, a selective PBKcc inhibitor 275 6.2. The phosphatidylinositol-3-kinase (PI3K) signaling pathway 276 6.3. First generation PI3K inhibitors 279 6.4. Isoform-selective second generation inhibitors 281 6.5. Marine natural products and the PI3K signaling pathway 284 6.6. Biogenesis and synthetic preparations of liphagal 286 6.7. Preparation of liphagal analogues 290 6.7.1. Variation atC 14 290 6.7.2. Variations at C15 and C16 303 6.8. Biological evaluation of the new synthetic liphagane derivatives 312 6.9. Conclusions and future directions 315 6.10. Experimental 319 7. References 368 vi LIST OF TABLES Table 1.1. Marine-derived compounds currently in clinical trials (Phase I and above) for the treatment of several human conditions 7 Table 2.1. NMR data for tauramamide (12) (recorded in DMSO-a?6) 24 Table 2.2. Antimicrobial activity (MIC's in |ig/mL) of synthetic tauramamide (12) and tauramamide ethyl ester (14) 26 Table 2.3. NMR data for dealkylsurfactin (48) (recorded in DMSO-d6) 36 Table 2.4. Specific rotation values for natural surfactin B2 (19), synthetic analogues (19a,b), "norsufactin" (19c), and dealkylsurfactin (48) 39 Table 2.5. NMR data for tauramamide (12) (recorded in DMSO-4) 55 Table 2.6. *H and 13C NMR data for natural and synthetic tauramamide ethyl ester (14) (recorded in DMSO-rf6) 58 Table 2.7. NMR data for dealkylsurfactin (48) (recorded in DMSO-cfc) 67 Table 3.1. Some IDO inhibitors with high in vitro activity 76 Table 3.2. NMR data for annulin C (53) (recorded in CDC13) 88 Table 3.3. NMR data for 2-hydroxygarveatin E (55) (recorded in CDCI3) 92 Table 3.4. NMR data for garveatin E (54) (recorded in CDCI3) 95 Table 3.5. NMR data for garvin C (56) (recorded in CDCI3) 99 Table 3.6. IDO-active secondary metabolites from G. annulata 102 Table 3.7. Annulin C (53) and commercially available IDO-active naphthoquinones 104 Table 3.8. In vitro inhibition of IDO by marine natural products, commercial naphthoquinones, P-carbolines, and tryptophan analogues 106 Table 4.1. Rearrangement of various 1,4-and 1,3-cyclohexadione derivatives 132 Table 4.2. NMR data for 3-benzylazepane-2,4-dione (78) (recorded in CD2C12) 135 Table 4.3. NMR data for 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3- benzyl-3-bromo-l-methylazepane-2,4-dione (99) (recorded in CDCI3) 140 vn Table 4.4. NMR data for 3-benzyl-4-(^n-butyldimethylsilanyloxy)-l-methyl-1,5,6,7- tetrahydroazepin-2-one (106) (recorded in CD2CI2) 144 Table 4.5. NMR data for 3-[(fer?-butyldimethylsilanyl)phenylmethyl]-3-chloro-l- methylazepane-2,4-dione (107) (recorded in CD2C12) 148 Table 4.6. NMR data for 3-benzyl-l,3,6,7-tetrahydroazepin-2-one (124) (recorded in CDCI3) 153 Table 4.7. NMR data for 3-(3,5-dibromo-4-methoxybenzyl)-lH-azepine-2,5-dione (141) (recorded in DMSO-<4) 157 Table 4.8. Reaction of 126,141-147 with several nitrogen-based nucleophiles 161 Table 4.9. NMR data for 19-demethyl-l,4,5,8,9,10-hexahydroceratamine B (157) (recorded in DMSO-4) 166 Table 4.10. NMR data for azepane-2,4-dione (50) (recorded in CD2C12) 176 Table 4.11. NMR data for 3-benzyl-3-chloroazepane-2,4-dione (96) (recorded in CD2C12). 179 Table 4.12. NMR data for 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) (recorded in CDCI3) 181 Table 4.13. NMR data for 3-[(^r?-butyldimethylsilanyl)phenylmethyl]-3-chloro-l- methylazepane-2,4-dione (107) (recorded in CD2CI2) 184 Table 4.14. NMR data for 3-benzylidene-azepane-2,4-dione (120) (recorded in CD2C12)... 185 Table 4.15. NMR data for 2-benzyl-8-oxa-4-aza-bicyclo[5.1.0]octan-3-one (125) (recorded in CDCI3) 193 Table 4.16. NMR data for toluene-4-sulfonic acid 6-benzyl-5-hydroxy-7-oxoazepan-4-yl ester (137) (recorded in CDCI3) 194 Table 4.17. NMR data for 4-bromo-3-(3,5-dibromo-4-methoxybenzyl)-azepane-2,5-dione (140) (recorded in CDCI3) 198 Table 4.18. NMR data for 3-(3,5-dibromo-4-methoxy-benzyl)-6,7-dihydro-lH-azepine- 2,5-dione (144) (recorded in CDCI3) 200 Table 4.19. NMR data for 5-bromo-6-(3,5-dibromo-4-methoxy-benzyl)-7-oxo-azepan-4- yl-cyanamide (156) (recorded in CD3OD) 204 Table 4.20. NMR data for 4,5-dibromo-3-(3,5-dibromo-4-methoxy-benzyl)-azepan-2-one (147) (recorded in CDCI3) 205 Table 4.21. NMR data for 17-Boc-1,4,5,8,9,10-hexahydroceratamine B (158) (recorded in CDCI3) 208 vin Table 4.22. NMR data for 7,17-dipivaloyl-l,4,5,8,9,10-hexahydroceratamine B amide (159) (recorded in CDCI3) 209 Table 5.1. Constituents in the resin of Cannabis sativa reported until 2005 216 Table 5.2. Pharmacological effects of smoked marijuana in humans 217 Table 5.3. Cannabinoid-based therapeutic agents, approved and in development 228 Table 5.4. Tumors exhibiting cannabinoid-induced inhibition 229 Table 5.5. NMR data for haplosamate A (47) (recorded in DzO) 237 Table 5.6. NMR data for haplosamate A triacetates (60) and (61) (recorded in CeD6) 242 Table 5.7. NMR data for haplosamate A (47) (recorded in DMSO-a?6) 263 Table 5.8. NMR data for haplosamate A (47) (recorded in DMSO-d6) 264 Table 5.9. NMR data for haplosamate A 3,4,7-triacetate derivative (60) (recorded in C6D6) 267 Table 5.10. NMR data for haplosamate A 3,4,7-triacetate derivative (60) (recorded in C6D6) 268 Table 5.11. NMR data for haplosamate A 3,6,7-triacetate derivative (61) (recorded in C6D6) 269 Table 5.12. NMR data for haplosamate A 3,6,7-triacetate derivative (61) (recorded in C6D6) 270 Table 5.13. Group epitope mapping (GEM) analysis for CB1/CB2 and haplosamate A (relative to H18/H21 at 0.96 ppm) 272 Table 6.1. Organization of the PDK's family 277 Table 6.2. NMR data for (+)-8-e/?j'-desformyl-14-bromoliphagane (46) (recorded in CeD6). 293 Table 6.3. NMR data for (+)-14-bromo-9,16-dihydroxyliphagane quinone (52) (recorded in C6D6) 295 Table 6.4. NMR data for (±)-desformyl-14-bromospiroliphagal (53) (recorded in CDC13).. 298 Table 6.5.13C-NMR assignments for (+)-16-hydroxyliphagane (68), (+)-15- hydroxyliphagane (69), and (+)-14-bromo-15-hydroxyliphagane (70) (recorded in CeD6).... 306 Table 6.6. 'H-NMR assignments for (+)-16-hydroxyliphagane (68), (+)-15- hydroxyliphagane (69), and (+)-14-bromo-15-hydroxyliphagane (70) (recorded in C^D^).... 308 Table 6.7. Optical activity measurements for liphagal (1) and its synthetic analogues 311 ix Table 6.8. Inhibition of PDKoc and PBKyby liphagal, LY294002, wortmannin, and several new liphagane derivatives 312 Table 6.9. NMR data for 14-bromo-15-methoxyliphagane (67) (recorded in C6D6) 337 Table 6.10. NMR data for 14-bromo-15-methoxyliphagane (67) (recorded in C6D6) 338 Table 6.11. NMR data for (+)-8-<?/?/-desformyl-14-bromoliphagal (46) (recorded in CDC13) 344 Table 6.12. NMR data for (+)-8-£7N-desformyl-14-bromoliphagal (46) (recorded in CDCI3) 345 Table 6.13. NMR data for (-)-8-ep/-desformyl-14-bromoliphagal (46) (recorded in CeD6).. 346 Table 6.14. NMR data for (±)-desformyl-14-bromospiroliphagane (53) (recorded in CDCI3) 347 Table 6.15. NMR data for (-)-14-bromo-9,16-dihydroxyliphagane quinone (52) (recorded in C6D6) 348 Table 6.16. NMR data for (±)-desformylspiroliphagane A (55) (recorded in CDC13) 350 Table 6.17. NMR data for (±)-desformylspiroliphagane A (55) (recorded in CDCI3) 351 Table 6.18. NMR data for (±)-desformylspiroliphagane B (56) (recorded in C6D6) 352 Table 6.19. NMR data for (±)-desformylspiroliphagane B (56) (recorded in C^D^) 353 Table 6.20. NMR data for (+)-14-bromo-15-hydroxyliphagane (70) (recorded in C6D6) 355 Table 6.21. NMR data for (+)-14-bromo-15-hydroxyliphagane (70) (recorded in C6D6) 356 Table 6.22. NMR data for (+)-15-hydroxyliphagane (69) (recorded in C6D6) 358 Table 6.23. NMR data for (+)-15-hydroxyliphagane (69) (recorded in C6D6) 359 Table 6.24. NMR data for (+)-16-hydroxyliphagane (68) (recorded in C6D6) 361 Table 6.25. NMR data for (+)-16-hydroxyliphagane (68) (recorded in C6D6) 362 x LIST OF FIGURES Figure 1.1. Some marine secondary metabolites with known ecological roles 2 Figure 1.2. Distribution of marine natural products by phylum (A) and biosynthetic origin (B), according to the MarinLit database 3 Figure 1.3. Distribution of biological activities evaluated in marine natural products in 2004 6 Figure 1.4. First marine-inspired drugs approved for human use 8 Figure 2.1. Prominent marine peptides currently in clinical trials 11 Figure 2.2. Structure of surfactin analogues 16 Figure 2.3. !H-NMR spectrum of tauramamide (12) (recorded in DMSO-J6 at 600 MHz). 22 Figure 2.4.13C-NMR spectrum of tauramamide (12) (recorded in DMSO-d6 at 150 MHz). 23 Figure 2.5. Partial HMBC spectrum of tauramamide (12) (recorded in DMSO-J6 at 600 MHz) 27 Figure 2.6. !H-NMR spectra of bacterial-derived and synthetic tauramamide ethyl ester (14) (recorded in DMSO-rf6 at 600 MHz) 28 Figure 2.7. !H-NMR spectrum of dealkylsurfactin (48) (recorded in DMSO-d6 at 600 MHz) 34 Figure 2.8.13C-NMR spectrum of dealkylsurfactin (48) (recorded in DMSO-J6 at 150 MHz) 35 Figure 2.9. Partial HMBC spectrum of dealkylsurfactin (48) (in DMSO-J6 at 600 MHz)... 40 Figure 2.10. 'H-NMR spectrum of tauramamide ethyl ester (14) (recorded in DMSO-ck at 600 MHz) 60 Figure 2.11.13C-NMR spectrum of tauramamide ethyl ester (14) (recorded in DMSO-dk at 150 MHz) 61 Figure 2.12. !H-NMR spectrum of dibenzyl dealkylsurfactin (47) (recorded in DMSO-J6 at 600 MHz) 69 Figure 2.13. 13C APT-NMR spectrum of dibenzyl dealkylsurfactin (47) (recorded in DMSO-J6atl50MHz) 70 xi Figure 3.1. Tryptophan metabolism in the kynurenine pathway 72 Figure 3.2. Reactions involved in the new high throughput IDO bioassay developed by Mauk and Vottero 77 Figure 3.3. Representative structures for secondary metabolites isolated from hydroids 79 Figure 3.4. Secondary metabolites isolated from G. annulata 80 Figure 3.5. a) A bush-like colony of G. annulata; b) Map of collection site 84 Figure 3.6. 'H-NMR spectrum of a) annulin A (39) (recorded in CD2C12 at 400 MHz), and b) annulin C (53) (recorded in CDC13 at 400 MHz) 87 Figure 3.7. Summary of COSY and HMBC correlations for annulin C (53) 89 Figure 3.8. 'H-NMR spectrum of 2-hydroxygarveatin E (55) (recorded in CDC13 at 600 MHz) 91 Figure 3.9. Summary of HMBC correlations for 2-hydroxygarveatin E (55) 93 Figure 3.10. Summary of HMBC correlations for garveatin E (54) 94 Figure 3.11. ' H-NMR spectrum of garveatin E (54) (recorded in CDCI3 at 400 MHz) 96 Figure 3.12. Summary of HMBC correlations for garvin C (56) 97 Figure 3.13. *H-NMR spectrum of garvin C (56) (recorded in CDC13 at 600 MHz) 98 Figure 3.14. Proposed pharmacophore involved in IDO inhibition 107 Figure 4.1. Prominent microtubule-stabilizing agents (MSA) from diverse sources 112 Figure 4.2. Representative pyrrole-imidazole alkaloids 119 Figure 4.3. JH and 13C-NMR spectra of ceratamine A (13) (recorded in DMSO-d6 at 500 and 125 MHz respectively) 120 Figure 4.4. :H and 13C-NMR spectra of 3-benzylazepane-2,4-dione (78) (recorded in CD2C12 at 400 and 100 MHz, respectively) 136 Figure 4.5. !H-NMR spectra of 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3- benzyl-3-bromo-l-methylazepane-2,4-dione (99) (recorded in CDCI3 at 300 MHz) 138 Figure 4.6. 13C-NMR spectra of 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3-benzyl-3-bromo-l-methylazepane-2,4-dione (99) (recorded in CDCI3 at 75 MHz) 139 Figure 4.7. MM2 energy minimizations for enols (100) and (101) 142 Figure 4.8. *H and 13C-NMR spectra of 3-benzyl-4-(tert-butyldimethylsilanyloxy)-l- methyl-l,5,6,7-tetrahydroazepin-2-one (106) (recorded in CD2CI2 at 400 and 100 MHz, respectively) 145 xii Figure 4.9. 'H and 13C-NMR spectra of 3-[(tert-butyldimethylsilanyl)phenylmethyl]-3- chloro-l-methylazepane-2,4-dione (107) (recorded in CD2CI2 at 400 and 100 MHz, respectively) 147 Figure 4.10. 'H and 13C-NMR spectra of 3-benzyl-l,3,6,7-tetrahydroazepin-2-one (124) (recorded in CDC13 at 400 and 100 MHz, respectively) 154 Figure 4.11. *H and 13C-NMR spectra of 3-(3,5-dibromo-4-methoxybenzyl)-lH-azepine- 2,5-dione (141) (recorded in DMSO-J6 at 400 and 100 MHz, respectively) 158 Figure 4.12. 'H-NMR spectra of a) 3-(3,5-dibromo-4-mefhoxybenzyl)-l-methyl-lH- azepine-2,5-dione (142); and ceratamine A (13) (recorded in CDCI3 and DMSO-^6 at 400 MHz) 159 Figure 4.13. Minimum energy conformations for 3-(3,5-dibromo-4-methoxybenzyl)-l- methyl-lH-azepine-2,5-dione (142), and ceratamine A (13) 160 Figure 4.14. !H and 13C-NMR spectra of 19-demethyl-l,4,5,8,9,10-hexahydroceratamine B (157) (recorded in DMSO-J6 at 600 and 150 MHz respectively) 165 Figure 4.15. Minimum energy conformations for 19-demethyl-1,4,5,8,9,10- hexahydroceratamine B (157); and ceratamine A (13) 167 Figure 4.16. *H and 13C-NMR spectra of 2-(3,5-dibromo-4-mefhoxybenzyl)-but-3-enoic acid but-3-enylamide (135) (recorded in CDCI3 at 400 and 100 MHz respectively) 210 Figure 4.17. lH and 13C-NMR spectra of 3-(3,5-dibromo-4-methoxy-benzyl)-1,3,6,7- tetrahydroazepin-2-one (136) (recorded in CDC13 at 300 and 75 MHz respectively) 211 Figure 4.18. H and C-NMR spectra of 5-bromo-6-(3,5-dibromo-4-methoxybenzyl)-7- oxo-azepan-4-yl-cyanamide (156) (recorded in CD3OD at 400 and 100 MHz respectively). 212 Figure 5.1. a) Mature C. sativa crop ready for harvesting; b) trichomes 215 Figure 5.2. Classification of the typical cannabinoids 218 Figure 5.3. Serpentine-like topology of the cannabinoid receptors CB1 and CB2 221 Figure 5.4. Representative endocannabinoids 223 Figure 5.5. Endogenous cannabinoid system 224 Figure 5.6. Events involved in the cell-based cannabinoid bioassay 231 Figure 5.7. Marine natural products combining phosphorus and sulfur 233 Figure 5.8. Structures for the steroidal glucoside (56) and two members of the withanolides family 234 xiii Figure 5.9. 'H-NMR spectrum of haplosamate A (47) (recorded in D 2 0 at 600 MHz) 235 Figure 5.10.13C-NMR spectrum of haplosamate A (47) (recorded in D20 at 150 MHz)... 236 Figure 5.11. a) 3IP decoupled and b) 31P coupled 'H-NMR spectra of haplosamate A (47) (recorded in CD3OD at 400 MHz) 239 Figure 5.12. The classical cannabinoid pharmacophore 244 Figure 5.13. Minimum energy conformations for the known cannabimimetic agents (-)- 9p-hydroxyhexahydrocannabinol (62) and WIN55212-2 (23) 245 Figure 5.14. Basic pulse sequence of a ID STD NMR spectrum 247 Figure 5.15. Events involved in STD NMR spectroscopy 249 Figure 5.16. Sample preparation for saturation transfer double-difference (STDD) NMR... 250 Figure 5.17. STDD experiments measured for an aqueous solution of each cannabinoid receptor supported in SF21 insect cells and haplosamate A (47) 252 Figure 5.18. Qualitative STD NMR group epitope mapping for the binding of haplosamate A (47) to the cannabinoid human receptors CB1 and CB2 253 Figure 5.19. STDD NMR-derived group epitope mapping for haplosamate A (47) 256 Figure 5.20. ^ -NMR spectrum of haplosamate A (47) (recorded in CD3OD at 600 MHz). 261 Figure 5.21. 13C-NMR spectrum of haplosamate A (47) (recorded in CD3OD at 150 MHz) 262 Figure 5.22. Integration of proton signals for the STD percentage measurement in a suspension of the cannabinoid human receptor CB1 supported in insect cells and haplosamate A (47) (relative to HI 8/H21 at0.96ppm) 273 Figure 5.23. Integration of proton signals for the STD percentage measurement in a suspension of the cannabinoid human receptor CB2 supported in insect cells and haplosamate A (47) (relative to HI 8/H21 at0.96ppm) 274 Figure 6.1. Enzymatic synthesis and degradation of PI-3,4,5-P3 (3) during the first steps of thePDK signaling pathway 276 Figure 6.2. First generation PI3K inhibitors 280 Figure 6.3. Some PI3K-inhibitory chemotypes: A) arylmorpholine and B) non-related structures 282 Figure 6.4. *H and l3C-NMR spectra of (+)-8-^/-desformyl-14-bromoliphagane (46) (recorded in C6D6 at 600 and 150 MHz respectively) 292 xiv 1 1 ^ Figure 6.5. H and C-NMR spectra of (+)-14-bromo-9,16-dihydroxyliphagane quinone (52) (recorded in C6D6 at 600 and 150 MHz respectively) 294 Figure 6.6. Representative HMBC (HH>C) and COSY correlations for 52 296 Figure 6.7. *H and 13C-NMR spectra of (±)-desformyl-14-bromospiroliphagal (53) (recorded in CDC13 at 600 and 150 MHz respectively) 297 Figure 6.8. Representative HMBC (H->C) and COSY correlations for 53 299 Figure 6.9. *H-NMR spectra of (±)-desformylspiroliphagal A (55) and B (56) (recorded in CDC13 and C6D6 respectively at 600 MHz) 301 Figure 6.10.13C-NMR spectra of (+)-16-hydroxyliphagane (68) and (+)-15- hydroxyliphagane (69) (recorded in C6D6 at 150 MHz) 307 Figure 6.11. 'H-NMR spectra of (+)-16-hydroxyliphagane (68) and (+)-15- hydroxyliphagane (69) (recorded in C6D6 at 600 MHz) 309 Figure 6.12. Key HMBC (H->C) and COSY correlations for 68 and 69 310 Figure 6.13. IgE-induced calcium influx in bone marrow-derived mast cells treated with DMSO (control), liphagal (1), LY294002 (7), 68 and 69 314 Figure 6.14. New synthetic liphagal-based PI3K inhibitors 315 Figure 6.15. a) 'H-NMR and b) 13C-NMR spectra of (2-hydroxy-4-methoxyphenyl)- methyltriphenyphosphonium bromide (78) (recorded in CDCI3 at 300 and 75 MHz respectively) 363 Figure 6.16. !H and 13C-NMR spectra of 2-[3-(7-bromo-6-methoxy-benzofuran-2-yl)- butyl]-l,3,3-trimethylcyclohexanol (90) (recorded in CDC13 at 400 and 100 MHz respectively) 364 Figure 6.17. 'H and 13C-NMR NMR spectra of desformyl-15,16-dimethoxyliphagal (54) (recorded in CeDeat 400 and 100 MHz respectively) 365 Figure 6.18. C-NMR spectrum of (±)-desformylspiroliphagane A (55) (recorded in CDCl3atl50MHz) 366 Figure 6.19. C-NMR spectrum of (±)-desformylspiroliphagane B (56) (recorded in CDCI3 at 150 MHz) 366 Figure 6.20. *H and 13C-NMR spectra of (+)-14-bromo-15-hydroxyliphagane (70) (recorded in C^D^ at 600 and 150 MHz respectively) 367 xv LIST OF ABBREVIATIONS 0 ± ID 2D [a](D Ac AcOH AICN Arg Asp b Bn Boc BOM Bu BuLi fBuLi bs 13C °c Calcd CB1 CB2 Cbz CD2C12 C6D6 CDI COSY 8 d D DCC dd DDQ DffiA DMAP DMF DMP DMSO-^6 dt e EDCI ELISA eq. Et - degree(s) - racemic - one-dimensional - two-dimensional - specific rotation at wavelength of sodium D line at temperature t (°C) - acetate - acetic acid - 1,1' -azobis(cyclohexanecarbonitrile) - arginine - aspartic acid - broad - benzyl - ?-butoxycarbonyl - butoxymethyl - butyl - n-butyllithium - 7-butyllithium - broad singlet - carbon-13 - degrees Celsius - calculated - cannabinoid receptor 1 - cannabinoid receptor 2 - benzyloxycarbonyl - deuterated dichloromethane - deuterated benzene - 1,1' -cabonyldiimidazole - two-dimensional correlation spectroscopy - chemical shift in parts per million - doublet - dextrorotatory - 1,3-dicyclohexylcarbodiimide - doublet of doublets - 2,3-dichloro-5,6-dicyano-1,4-benzoquinone - diisopropylethylamine - 4-dimethylaminopyridine - A^N-dimethylformamide - Dess-Martin periodinane - deuterated dimethyl sulphoxide - doublet of triplets - extinction coefficient - l-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride - enzyme linked immuno sorbant assay - equivalent(s) - ethyl XVI Et3N EtOAc EtOH Fmoc g Glu h JH HATU HC1 HMBC HMQC HPLC H20 HOAt HOBt HOSu HRESEVIS HSQC Hz IC50 J X L Leu LDA LRESIMS m M m-CPBA Me MeCN MeOH Mes mg MHz MIC min mL mm mmol MOM MRSA m/z nM NBS NCS - triethylamine - ethyl acetate - ethanol - 9-fluorenylmethoxycarbonyl - gram(s) - glutamic acid - hour(s) - proton - [0-(7-azabenzotriazolyl)-l,l,3,3-tetramethyluronium hexafluorophosphate] - hydrochloric acid - two-dimensional heteronuclear multiple bond coherence - two-dimensional heteronuclear multiple quantum coherence - high-performance liquid chromatography - water - 1 -hydroxy-7-azabenzotriazole - 1-hydroxybenzotriazole - Af-hydroxysuccinimide - high resolution electrospray mass spectrometry - two-dimensional heteronuclear single quantum coherence - hertz - inhibitory concentration (for 50% of a biological sample) - coupling constant in hertz - wavelength - levorotatory - leucine - lithium diisopropylamide - low resolution electrospray mass spectrometry - multiplet - molar concentration - meto-chloroperbenzoic acid - methyl - acetonitrile - methanol - mesitylene - milligram(s) - megahertz - minimum inhibitory concentration - minute - millilitre(s) - millimetre(s) - millimol(s) - methoxymethyl - methicillin-resistant Staphylococcus aureus - mass to charge ratio - nanomolar - ./V-bromosuccinimide - ./V-chlorosuccinimide NMR NOE NRPS Pac pH Ph PHF PMB PNG ppm PPTs Pro PyBOP Pyr q R Rf ROESY s S SAR SCUBA Ser sp. STD STDD t TBDSC1 TCA TFA Thr TLC TOCSY Tr Trp Ts Tyr UV Val VRE - nuclear magnetic resonance - nuclear Overhauser enhancement - non-ribosomal peptide synthase - phenacyl - -log10[H+] - phenyl - paired helical filaments - p-methylbenzoate - Papua New Guinea - parts per million - pyridinium p-toluensulfonate - proline - (benzotriazol-1 -yl)oxytripyrrolidinophosphonium hexafluorophosphate - pyridine - quartet - rectus - retention factor - rotating frame Overhauser enhancement spectroscopy - singlet - sinister - structure-activity relationship - self-contained underwater breathing apparatus - serine - species - saturation transfer difference - saturation transfer double difference - triplet - ?-butyldimethyl silyl chloride - trichloroacetic acid - trifluoroacetic acid - threonine - thin-layer chromatography - total correlation spectroscopy - trityl - tryptophan - tosyl, p-toluenesulfonyl - tyrosine - ultraviolet - valine - vancomycin-resistant Enterococci ACKNOWLEDGEMENTS I would like to thank my supervisor Professor Raymond J. Andersen for giving me the opportunity to do research in marine natural products, and get actively involved in different aspects of such a gratifying scientific field. From isolation to synthesis, including diving and organization of a collecting trip to Costa Rica, his enthusiasm, support and guidance facilitated this work and allow me to grow as a chemist and above all, as a person. Many thanks to David Williams for his advice in practical aspects of natural product isolation, and to Mike Leblanc for his assistance in both the laboratory and diving. The collaboration and guidance of Professor Nick Burlinson was fundamental during STD data acquisition and processing. The assistance of technical staff (MS and NMR facilities) in the Department of Chemistry at UBC is also gratefully acknowledged. I am very grateful for all the biological assessments of natural and synthetic materials carried out by Eduardo Vottero and Professor Grant Mauk, as well as Professor Michel Roberge and members of his research group (Biochemistry and Molecular Biology, UBC), Tom Pfeifer and Professor Tom Grigliatti (Zoology, UBC), Professor Gerald Krystal (B.C. Cancer Agency), Dr. Ian Hollander (Wyeth) and Helen Wright (Biological Services, UBC). These rewarding collaborations have taught me that the understanding among professionals in multiple fields is crucial for the success of any drug discovery program. To everyone in the Andersen research laboratory, my most sincere appreciation. Whether it was playing soccer with Harry and Emiliano, practicing Japanese with Kaoru and Chelsea, running reactions on Sunday mornings with Fred, Lu and Xin-Hui, sailing with Justin and Takashi, listening to heavy metal with Matt, or diving with Roger, Kelsey, Kate and Julie, all these experiences allowed me to truly enjoy my experience at UBC. In particular, thanks to Chris Gray and his wife Claire for opening the doors of their home during Christmas 2005, and to Rob Keyzers for his careful proof-reading and useful comments during the writing of this thesis. Finally, I would like to thank all the members of my family for their support and encouragement throughout my studies. xix DEDICATION For Nancy and Helena xx Chapter 1. Marine Natural Products Chemistry 1. Marine Natural Products Chemistry: Overview 1.1. Past and Present The field of marine natural products chemistry encompasses the study of chemical structures and biological activities of secondary metabolites produced by marine plants, animals and microorganisms.1 In contrast to the study of terrestrial natural products, which started in the 1800's with the characterization of alkaloids such as morphine, strychnine and quinine, chemical exploration of marine life only began in earnest in the early 1960's.1'2 Pioneers like Paul Scheuer and Richard Moore in the United States, Luigi Minale and Ernesto Fattorusso in Italy, as well as Yoshiro Hashimoto and Yoshimasa Hirata in Japan, began to examine sponges, algae, and other unfamiliar marine organisms.2 Undoubtedly, the invention of SCUBA by Jacques-Yves Cousteau and Emile Gagnan in 1943 dramatically increased accessibility of sessile marine invertebrates and algae, allowing natural products chemists to make their own field collections and establish extensive libraries of marine extracts. Today, after nearly 50 years of research, steady progress by both academic institutions and pharmaceutical industries has led to the characterization of ca. 17,100 new marine natural products, reported in ca. 6,800 publications.1'4"6 Another ca. 9,000 references deal with syntheses, biological activities, ecological studies and reviews of marine natural products.4 Over 300 patents have been issued on this field.4'6 Secondary metabolites of marine origin range from simple achiral molecules to highly complex structures, rich in stereochemistry, concatenated rings and reactive functional groups.1 These chemical entities are full of structural surprises and not limited by human imagination. The additional incorporation of covalently bound halogens (mainly CI and Br) is a typical marine 1 Chapter 1. Marine Natural Products Chemistry 1 1 variation, presumably due to the high concentration of both ions in seawater. ' Equally as diverse are the molecular modes of action by which these compounds impart their biological activity, making them an extraordinary resource for the development of new therapeutic agents. 1.2. Secondary metabolites? In the field of marine natural products chemistry, natural products are alternatively called secondary metabolites because they are apparently not essential for the primary metabolic activities involved in the growth of the producing organism. 1,8 from sponges of the genus Agellas HO N r from the hydroid Tridentata marginata HO, from the hydroid Corydendrium parasiticum OH from the hydroid Lytocarpus philippinus N H 2 HV H * 0 ^ 'O-OH 2N 5 from the marine bacteria Brevibacillus laterosporus (PNG-276) X>H I 1 H 2 N-G-C-C-SDPR-C-NYDHPEI—C-CONH 2 from the cone snail Conus victoriae H2N-GEXXLQXNQXLIRXKSN-CONH2 from the cone snail Conus geographus Figure 1.1. Some marine secondary metabolites with known ecological roles: oroidin (l),9 tridentatol A (2),10'11 corydendramine A (3),12 lytophilipinne A (4),13 bogorol A (5),14 ACV-1 (6) andCGX-1007(7) 15,16 2 Chapter 1. Marine Natural Products Chemistry Although in some cases there is evidence supporting antifeedant activity against predators (1-4), inhibitory effects on the growth of competitors (5, Figure l.l),9"14'17 and even immobilization of motile species prior to ingestion (6, 7) ;1516 for most secondary metabolites there is no rigorous proof of their actual ecological role. In contrast to omnipresent primary metabolites such as amino acids, fatty acids and nucleotides, the occurrence of a particular secondary metabolite is usually limited to one or a few species.8 Among the approximately 100,000 species of marine invertebrates described so far, most of the animals in the phyla Cnidaria, Porifera, Bryozoa, and Echinodermata are sessile, slow growing, brightly colored, and lack the physical protection of shells and spines.1 Their intense concentration in marine habitats makes them highly competitive and biologically complex. Additionally, nutrients, light, water current, and temperature represent growth limiting components, further fueling competition.7 Thus, in order to ensure their survival, a high percentage of species make extensive use of biologically active secondary metabolites (Figure 1.2). Eztinoderrrtfa Fungi and 5% Chordata i fiyozoa \ 1% QiorrDphyccrta B Porifera 37% Alkaloid 15% Terpenoid 48% Foljfctide 27% Figure 1.2. Distribution of marine natural products by phylum (A) and biosynthetic origin (B), according to the MarinLit database (Total: 17,068 structures up to 2005).5 3 Chapter 1. Marine Natural Products Chemistry Among marine invertebrates, sponges (Porifera), coelenterates (Cnidaria: soft corals and gorgonians), echinoderms, tunicates (Chordata), algae and mollusks are the most prolific sources of compounds.7'18 These metabolites take the form of terpenoids, polyketides, alkaloids, peptides, shikimic acid derivatives, sugars, steroids, and a large number of molecules with combined biogenetic origins (Figure 1.2B). ' ' Drug discovery programs in both the pharmaceutical industry and academia have focused special attention on marine sponges (Figure 1.2). The broad molecular diversity and the high cytotoxic activity exhibited by sponge metabolites makes them promising anticancer drug leads.1'2'18'20 Approximately 5,000 living sponge species exist in the phylum Porifera, which is composed of three distinct classes: the Hexactinellida (glass sponges), the Demospongiae (siliceous sponges), and the Calcarea (calcareous sponges).21 As expected from their phylogenetic position, marine sponges are among the oldest known animals, with fossils dating from the late Precambrian (542 million years ago). Since 1950, over 5,600 references have reported more than 6,500 secondary metabolites including ca. 3,100 nitrogen-containing compounds (such as pyrrole-imidazole alkaloids, Chapter 4), unique to marine sponge 6,20-23 species. An increasing amount of evidence suggests that many interesting compounds isolated from marine invertebrate extracts are actually biosynthetic products of symbiotic microorganisms, including microalgae, cyanobacteria, as well as heterotrophic bacteria and fungi.1'7,24 However, systematic studies of invertebrate-symbiont associations are usually accompanied by serious technical challenges, such as the general resistance of symbionts to culturing attempts separately from their hosts and the complexity of many microbial consortia. ' Nevertheless, since laboratory fermentation represents a sustainable supply of an interesting metabolite,1' 'l advances in genomics-based taxonomy and in new methods for 4 Chapter 1. Marine Natural Products Chemistry acquisition and cultivation of marine bacteria are already leading to the discovery of new classes of bacteria that produce unprecedented antibiotics and potential anticancer drugs.2'25 Deep ocean sediments have recently emerged as an immeasurable source (70% of Earth's surface) of actinomycete bacteria, biologically relevant since their terrestrial counterparts are responsible for producing most of the currently available antibiotics. When microbial life is considered, the total number of species inhabiting the world's oceans may approach 1 to 2 million.7 1.3. Biological activity and current success in marine natural products research Historically, the chemical exploration of oceanic invertebrate life has relied on assays to screen libraries of marine extracts in order to identify those ones capable of eliciting a desired cellular response with potential pharmaceutical applications.1 These range from fairly simple toxicity tests such as the sea urchin egg and brine shrimp assays, to more expensive and labor intensive in vitro screens using purified enzymes and receptors, as well as cell-based bioassays 1 9 7 that employ genetically engineered eukaryotic cells or microorganisms. ' ' Recently, the combination of robotics and target-based bioassays has given birth to high- throughput screening, a process by which large numbers of compounds are tested in an automated fashion for activity as inhibitors or activators of a particular molecular target. ' Having as a primary directive the identification of lead chemical structures and their structure- activity optimization, ' high-throughput screening programs in major drug companies can typically handle 3,000 enzyme inhibition or 4,800 cell-based bioassays per day, allowing the biological evaluation of up to 100,000 chemical entities in a reasonable time frame.1 With groups of academic researchers, major pharmaceutical companies and the National Cancer Institute (NCI) heavily investing resources in cancer research for decades now, it is not surprising that 5 Chapter 1. Marine Natural Products Chemistry such activity is a major driving force behind the discovery of new marine natural products (Figure 1.3.).23 AitBnflarrrrEtory 3% Antiviral ricultural 6% X \ \ \ 1 IrmunoriDdulatory 1% Various I 3% iilSMi^ i mi -^ fyfeJhodology 21% Figure 1.3. Distribution of biological activities evaluated in marine natural products in 2004. Various: neurological, blood pressure, fertility, allergy-based, laxative and enzyme activation bioassays. Methodology: mechanism of action, SAR studies, radio immunoassays, as well as sea urchin egg and brine shrimp toxicity assays. Consequently, in 2006 more than 30 marine-derived molecules were in preclinical development or clinical trials against a wide variety of cancers. Noteworthy, a significant number of these new drug candidates were initially developed with direct or indirect NCI assistance, and later licensed to pharmaceutical partners for clinical evaluation, manufacture and 2,7 sales/' ' Table 1.1 summarizes 18 marine-derived compounds currently in clinical trials. 6 Chapter 1. Marine Natural Products Chemistry Table 1.1. Marine-derived compounds currently in clinical trials (Phase I and above) for the treatment of several human conditions. 2,7 Compound Ecteinascidin 743 Bryostatin 1 ILX651a (Synthadotin) Kahalalide F Squalamine TZT-1027" (Soblidotin) E7389c Discodermolide ES-285 KRN-7000d NVP-LAQ8246 Salinosporamide A E-7974f AE-941 (Neovastat) GTS-21 (DMBX) Ziconotide (Prialt™) CGX-1160 ACV1 Synthetic derivatives of: Source Ecteinascidia turbinata (tunicate) Bugula neritina (bryozoan) Dolabella auricularia (mollusk) Elysia refescens (mollusk) Squalus acanthias (shark) Dolabella auricularia (mollusk) Halichondria okadai (sponge) Discodermia dissoluta (sponge) Mactromeris polynyma (mollusk) Agelas mauritianus (sponge) Psammaplysilla sp. (sponge) Salinospora sp. (bacterium) Cymbastella sp. (sponge) Scyliorhinus torazame (shark) Paranemertes peregrina (worm) Conus magus (mollusk) Conus geographus (mollusk) Conus victoriae (mollusk) aDolastatin 15, bDolastatin 10, Chemical class Tetrahydroisoqumolone alkaloid Polyketide Linear peptide Cyclic depsipeptide Aminosteroid Linear peptide Macrocyclic polyether Polyketide Alkylamino alcohol a-Galactosylceramide Indolic cinnamyl hydroxamate Bicyclic y-lactam-f} lactone Linear peptide Mixture from cartilage Pyridine alkaloid Cyclic peptide Cyclic peptide Cyclic peptide cHalichondrin B, dAgelasphin, Disease Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Alzheimer's Neuropathic pain Pain Pain Status II in w n II II II i i i i i i M i l I Appr. 2005 I I ePsammaplin, fHemiasterlin. 7 Chapter 1. Marine Natural Products Chemistry .TMN As seen in Table 1.1, the first modern marine drug was ziconotide (Prialt ), a potent calcium channel blocker used to provide relief from severe neuropathic pain.2 This pharmaceutical is based on the highly toxic small peptide co-conotoxin MVIIA (8) that was isolated in 1987 by Olivera and coworkers15 from the coneshell mollusk Conus magus. Although ziconotide is generally recognized as the first marine-derived drug, Werner Bergman, a pioneer in marine sterol chemistry, isolated two modified nucleosides (9 and 10) with unique antiviral activities from the sponge Cryptotethia crypta in the 1950's. ' ' These compounds were employed as a template that guided the development of the structurally related antiviral drugs Ara-A (11) and Ara-C (12), which have been in use since 1969 as chemotherapeutic agents for the treatment of herpes and leukemia.2'18 H 2 N - C - K G K G A K - C - S R L M Y D - C - C — T G S - C - R S G K - C - C O N H 2 I ' 1 ' HN | N H 2 O NH 2 H C S HO H C S HO Uoj^i ic-o-y OH OH 10 ^ ~N N OH 11 y XJ) O ^ N ^ H C S HO i0j OH 12 Figure 1.4. First marine-inspired drugs approved for human use. Two of the studies described herein focus in the isolation and characterization of marine natural products based on their biological activity. Considering the major bioassay areas summarized in Figure 1.3, Chapter 3 details how secondary metabolites from the hydroid Garveia annulata are capable of inhibiting human indoleamine 2,3-dioxygenase (IDO), an enzyme involved in immunomodulation and the ability of tumors to evade the host immune Chapter 1. Marine Natural Products Chemistry system. Chapter 5 features the first marine cannabinoid-active compound and provides conclusive NMR evidence about its binding mode to the cannabinoid human receptors (Methodology, Figure 1.3). 1.4. Synthesis of marine natural products Once a bioactive marine-derived lead structure has been identified, the issues of immediate supply and eventual manufacturing-scale availability become key factors in its further 118 progress towards pharmaceutical development. ' Evaluation of the biological activity in whole cells, in vivo efficacy in animal models, determination of maximum tolerated doses and minimum effective doses, as well as pharmacokinetic studies, demand large amounts of material that is in most cases unavailable from the source organism. In order for a large pharmaceutical company to invest financial and human resources in a drug development project (typically involving 12-15 years and more than US $350 million),1 a reliable and economical source of the target metabolite must be secured.1'2'7 Organic synthesis represents one possible supply option. The wealth of structurally challenging features and potent biological activities exhibited by some marine secondary metabolites have made them attractive targets for academic synthetic programs. Besides solving supply problems, synthesis plays an increasingly important role as a method for providing solutions to structural and stereochemical questions. Furthermore, the evaluation of biological 9Q mechanisms and targets connects synthetic organic chemistry with cell biology. Only the combination of both allows for broadly addressing target validation with suitable potency and specificity. Even in cases when the complexity of the natural product does not allow total synthesis on an industrial scale, a synthetic medicinal chemistry program using the marine 9 Chapter 1. Marine Natural Products Chemistry metabolite as a lead structure may provide simpler, synthetically accessible analogues containing the same pharmacophore. In 2003, Blunt, Munro and coworkers31 pointed out the necessity of a companion review to their annual Marine Natural Products report in view of the rapid growth in the area of marine 9 0 "iT -i-i natural product synthesis. In response, since 2005 Nicholas and Phillips ' ' have been summarizing first and improved total syntheses of marine natural products. These and other reviews reveal significant progress in the development of efficient asymmetric reactions for building multiple stereocenters in acyclic intermediates and for construction of ring systems,29'34 new general strategies and methodologies for efficient coupling of highly functionalized segments, ' development of highly stereoselective routes to E and Z, di- and trisubstituted alkenes,9'37 reduction of functional group protection and deprotection steps,38 and achieving efficient macrocyclization reactions to provide large ring sizes.34' The selection of marine natural products covered herein provides examples where synthesis was employed to: • Overcome shortages of supply from natural sources (Total synthesis of tauramamide, Chapter 2). • Generate modified natural products required as biological tools for target validation (Preparation of dealkylsurfactin, Chapter 2, and synthesis of liphagal analogues, Chapter 6). • Provide simplified analogues offering practical access to biologically active structures relevant for further drug development (Progress towards the synthesis of ceratamines, Chapter 4). 10 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2. Bioactive Marine Peptides: Total Synthesis of tauramamide and dealkylsurfactin 2.1. Peptides from the sea Cyclic and linear peptides isolated mainly from microorganisms and sponges represent an important class of marine-derived metabolites.40"42 They exhibit antibacterial, antifungal, cytotoxic, antimalarial, and antitumor activities, as well as incorporating novel structural features unseen in their terrestrial counterparts, making marine peptides important lead compounds for drug development research.43 For instance, over a dozen of antitumor compounds from marine origin are currently in various phases of human clinical trials.44 Among them, the peptides aplidin (l),45,46 dolastatin 10 (2)47'48 and its synthetic congeners ILX-65149 and cemadotin,50 hemiasterlin analogue HTI-286 (3),51"53 and kahalalide F (4),54"56 possess prominent inhibitory activity against different types of tumors. 44 <f7 i from the Mediterranean tunicate Aplidium albicans from the Indian Ocean mollusk Dolabella auricularia OH from several marine sponges including Cymbastella sp. H  \^H U I N H IXH XX H2hT ^ Y " ^ C HNL 4 6' from the Hawaiian mollusk Elysia rufescens HN ^ O HN Figure 2.1. Prominent marine peptides currently in clinical trials 44 11 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2.2. Peptides from marine PNG bacterial isolates Besides having a leading role in the isolation, synthesis, ' biological evaluation, and clinical development44 of hemiasterlin (5) and its analogues (such as HTI-286 3), the Andersen research group has reported in the last decade a considerable number of other marine- derived bioactive peptides.61"67 Most of these compounds have been isolated from laboratory cultures of bacterial isolates as part of an ongoing program aimed at discovering new antibiotics from the sea, which are desperately needed in the treatment of resistant strains of Gram positive human pathogens. Microorganisms isolated from marine habitats in Papua New Guinea (PNG) have proved to be an extremely rich source of new bioactive secondary metabolites.61'62' '65'68 OH 2.2.1. New antibiotics from PNG-276 PNG-276 was initially obtained from tissues of an unidentified tube worm collected off the coast of Loloata Island, Papua New Guinea, and later identified as Brevibacillus laterosporus by 16S RNA analysis.69 Crude MeOH extracts of B. laterosporus cells harvested from cultures grown as lawns on solid agar showed broad-spectrum antibiotic activity against several human pathogens including methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococcus (VRE), Mycobacterium tuberculosis, Candida albicans, and Escherichia coli. 2' ' To date, bioassay-guided fractionation of PNG-276 cultures has yielded five structurally unrelated families of antibiotics. 12 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin The first group of PNG-276 antibacterial compounds is comprised of loloatins A-D, ' a family of cyclic decapeptides containing four aromatic amino acid residues, two of which have the unnatural D configuration. The presence of both ornithine and aspartic acid imparts zwitterionic character. Loloatins A (6) to C showed potent antibacterial activity against MRSA, VRE, and penicillin-resistant Streptococcus pneumoniae (MIC's 0.5-4.0 ug/mL).62 The bogorol family of linear peptides is also produced by PNG-276. At the time of its isolation, bogorol A (7)64 represented a new template for cationic antibiotic peptides. Cationic peptides are widespread in nature where they play an important role in innate immune systems 70 71 protecting living organisms from microbial infections. ' They seem to avoid rapid emergence of resistance by physically disrupting cell membranes, killing bacteria very quickly.64 13 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin The bogorols contain modified C- and iV-terminal residues, D amino acids and a dehydroamino acid.68 They exhibited strong activity against MRSA and VRE (MIC's of 2.0 and 10 |i.g/mL, respectively). Further investigations of PNG-276 culture extract fractions exhibiting potent inhibition of Candida albicans led to the isolation of basiliskamides A (8) and B (9) (MIC's 1.0-3.1 ug/mL),65 characterized by a linear a,(3,y,8-unsaturated amide backbone with a cinnamoyl substituent attached via an ester linkage. H2N H2N The same report included two Escherichia coli active components, tupuseleiamides A (10) and B (ll),6 5 new acyldipeptides in which the amino acids both have the non-proteinogenic D configuration. o  H H C N o HO^NvV = M H J* ° o  H n v ^ o H O A ^ N - r r A N A = M H JOT 10 11 The fifth and most recent family of PNG-276 derived peptide antibiotics is represented by tauramamide (12), a linear acylpentapeptide isolated as its methyl (13) and ethyl ester (14) derivatives. As with the tupuseleiamides, 12 is acylated at the TV-terminus and contains two D amino acids. Both esters showed potent (MIC's 0.1 ug/mL) and relatively selective activity against the important Gram positive human pathogen Enterococcus sp. 14 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 12R=H 13 R=CH3 14 R=CH2CH3 All attempts to convert the limited amounts of esters 13 and 14 obtained from B. laterosporus cultures to the natural product 12 failed to generate sufficient amounts for biological testing. Additional material was also necessary for an exhaustive antimicrobial assessment of 13 and 14. Such requirements prompted the total synthesis of tauramamide (12) and its ethyl analogue (14). Furthermore, a synthetic sample of both compounds would confirm the proposed structure for this new antibiotic. 2.2.2. Anti-Alzheimer activity of PNG10A In 1968, Arima and coworkers72 isolated surfactin Q (20) from Bacillus subtilis LAM 1213. The compound exhibited exceptional surfactant activity and its structure was elucidated as that of a cyclic depsipeptide having a hydroxyfatty acid residue. Further studies by Nagai and Okimura74 revealed that laboratory cultures of Bacillus natto KMD 2311 contained seven additional homologous lipopeptides (15-22), different only in the structure of their /^-hydroxy side chains (Figure 2.2). 15 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Leu ^ A s p O \ 1 DLeu Cf^Glu—Leif Analogue R R A, (15) / V ) A A2 (16) ^ A c > A 1  '6 A3 (17) > M ^ B, (18) / V ^ 1 1  '8 Analogue B2(19) > M ^ Q (20) •A-r^ 1  '9 C2 (21) \ ^ N s > ^ 1  '8 D(22) / V ^ 8 * '10 Figure 2.2. Structure of surfactin analogues.74 Surfactin Ci (20, or simply surfactin) remains the best known representative of the family, although other lipopeptides have since been discovered. " Several other physiological and biochemical properties have been demonstrated for surfactin, including antibacterial, QA oc of. 0*7 co antitumoral, antiviral, ' and antimycoplasmic activities. ' Additionally, 20 has been recognized as a hypocholesterolemic agent89 and exhibits valuable inhibition of biofilm and blood clot formation.72 While such properties qualify surfactin for potential applications in medicine or biotechnology, they have not been subsequently exploited.75'9 In our research group, surfactin Q (20) was found via bioassay-guided fractionation of cultures prepared using the marine microorganism PNG10A, also collected in Papua New Guinea. Marine organisms that have yielded similar cyclic depsipeptides, presumably through symbiotic collaborations with bacteria, include sponges, mollusks, chelicerata, crustaceans and 7 Q QO Qfi ascidians/ ' - '^The screening process, developed by Dr. M. Roberge and his research group in the Department of Biochemistry at UBC, was designed to detect natural products capable of inhibiting the binding of Pinl to the microtubule-associated protein tau. Such an interaction is believed to be crucial in the development of Alzheimer's disease symptoms.99"101 16 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Pinl is a mitotic regulator essential for the G2 to M phase transition during the eukaryotic cell cycle. ' This enzyme belongs to the peptidyl-prolyl cis-trans isomerases, a superfamily of ubiquitous catalysts responsible for enhancing the typically slow cis-trans isomerization of prolyl bonds.102 Pinl specifically isomerizes Ser/Thr-Pro bonds and regulates the function of mitotic phosphoproteins.101 The conformational changes induced by this enzyme play an important role in protein folding, signal transduction, trafficking, assembly and regulation of the cell cycle.101'102 Among others, Pinl targets the microtubule-associated protein tau.101 Cis Trans All cases of Alzheimer's disease are characterized by the formation of neurofibrillary tangles containing paired helical filaments (PHFs), one of the neuropathological hallmarks of this human condition." The main component of PHFs is hyperphosphorylated tau. In normal brains, tau stabilizes the internal microtubule structure of neurons that functions to transport proteins and other molecules through the cells. In the brains of Alzheimer's patients tau is hyperphosphorylated and unable to bind to microtubules and promote their assembly.101 Pinl has a high affinity for PHFs and is sequestered by the phosphorylated tau, leading to a depletion of soluble Pinl in the brains of Alzheimer's patients, which in turn induces mitotic arrest and apoptotic cell death. This translates to an acceleration of degenerative symptoms. Several studies99'101'102 suggest that disrupting the interaction between Pinl and phosphorylated tau in PHFs, might reverse the memory loss associated with the disease. Additionally, it has been shown that Pinl can restore the ability of phosphorylated tau to bind microtubules and promote microtubule assembly in vitro" Thus, Pinl might be useful as a 17 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin pharmacological target for the development of new therapies for Alzheimer's and related neurodegenerative diseases. Particularly interesting as a biological tool for the study of interactions between Pinl and phosphorylated tau, as suggested by the Roberge group, was an analogue of surfactin without its characteristic lipophilic chain. Such a compound would provide information regarding possible inhibitory mechanisms, and confirm whether the efficient surfactant capability of macrolide lipopeptides is in fact responsible for the observed disruption of Pinl/phosphorylated tau adducts. Additionally, if the macrocyclic peptide alone were active, variation of its amino acid sequence may yield more potent inhibitors attractive both as research tools and experimental drug candidates. 2.3. Total synthesis of tauramamide To date, the only peptidic PNG-276 metabolites synthetically prepared have been the loloatins, due mainly to an increasing interest in their biological mechanism (the non-peptidic basiliskamides have also been synthesized).104"107 An Fmoc solid phase strategy for peptide elongation using (benzotriazol-l-yl)oxytripyrrolidinophosphonium hexafluorophosphate (PyBOP) and diisopropylethylamine (DIEA) as coupling reagents, followed by a generally low yielding on-resin or after-cleavage cyclization step (4-74%) employing (9-(7-azabenzotriazolyl)- 1,1,3,3-tetramethyluronium hexafluorophosphate (HATU) and l-hydroxy-7-azabenzotriazole (HOAt), are common aspects in the three available syntheses of loloatins. The use of HOAt is reported to enhance coupling yield and suppress racemization. 4 1 Bogorols and tupuseleiamides have yet to be prepared. Our synthetic approach to tauramamide (12) was convergent, starting from both C- and iV-termini simultaneously (Scheme 2.1). 18 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin H C b z ^ ^ r ^ - Cbz OH H H C b z ^ ^ ^ ^ C b z C b z ^ ^ T ^ Cbz H N ^ HN > CH2CI2 ] 1. TFA, CH2CI2 Boc DCC, DMAP 97% 2. Boc-L-Trp, PyBOP Boc CH2CI2 , 85% BIT "TPH 23 24 O H -N. ^ N ^ C b z ^ ^ p ^ Cbz HN. PyBOP. CO "•" HN. 1. TFA, CH 2 CI 2 2. Boc-D-Leu, PyBOP CH 2CI 2 , 7 6 % H C b z ^ ^ r ^ Cbz HN "Boc AA ^ 1. TFA, CH 2CI 2 N ' "T~ "Boc 2. Boc-L-Ser(Bn), PyBOP " f ^ M 6 H 1 ^ CH 2CI 2 , 8 1 % ~ H OBn 27 Boc HO' Or- 1. TFA, CH 2 CI 2 O 98% 31 30 Scheme 2.1. Steps involved in the synthesis of tauramamide fragments (27) and (31). Thus, the commercially available TV-protected arginine derivative (23) was converted to 108 its benzyl ester (24) by reaction with benzyl alcohol, DCC, and DMAP. Deprotection of the a 19 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin amino nitrogen by treatment of 24 with TFA,109'110 followed by three cycles of standard PyBOP111 activated peptide coupling with Boc-L-Trp, Boc-D-Leu, and Boc-L-Ser(OBn) in sequence, generated the protected tetrapeptide (27). At the same time, the phenolic hydroxyl in D-Tyr methyl ester (28) was protected by 119 treatment with benzyl bromide and potassium carbonate in acetone. Exposure of the amino group in 29 by reaction with TFA,110 followed by DCC mediated amide formation108 with 7- i n methyloctanoic acid, and subsequent LiOH catalyzed ester hydrolysis gave acid (31). H Cbr^ ^y^~ CBz HN L I R 2 ^ ^ r = ^ R2 HN Boc OBn 1. TFA, CH 2CI 2 2. 3 1 , PyBOP, CH 2CI 2 , 8 1 % R i Q P>3' o " HN 32 R1=R3=Bn, R2=Cbz 12 R-]=R2=R3=H 14 R!=R2=H, R3=Et H2 , Pd/C, EtOH, 81 % HCI, EtOH, 8 6 % Scheme 2.2. Total synthesis of tauramamide (12) and its ethyl ester (14). , i n Treatment of the protected tetrapeptide (27) with TFA followed by PyBOP mediated amide coupling with acid (31) gave protected tauramamide (32) (Scheme 2.2). Hydrogenolysis 20 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin using Pd on C in EtOH removed the Bn and Cbz protecting groups to liberate tauramamide (12), which was purified using reversed-phase HPLC to give a pure sample. The structure of each reaction intermediate was confirmed by *H-NMR and HRESIMS. In most steps, the products afforded did not require exhaustive purification and were directly employed in the following reaction. From all nine reaction steps leading to tauramamide (12), only hydrogenation of precursor 32 needed some optimization, in order to minimize the amount of mono, di- and triprotected tauramamide. Several H2 pressures and Pd on C concentrations were tested, achieving best results when 10% Pd/C, wet Degussa type E101NE/W (-50% water, reported as an useful reagent for debenzylations), 14 was stirred with 32 under H2 atmosphere for one week at 20 atm. The free acid (12) gave a [M+H]+ ion at m/z 864.4981 in the HRESEVIS, consistent with a molecular formula of C44H65N9O9 (calculated for C44H66N9O9: 864.4984). Examination of its *H and 13C-NMR spectra (Figures 2.3 and 2.4, Table 2.1) revealed an evident similarity with the NMR data for both natural tauramamide esters. Detailed analysis of the COSY, HMQC and HMBC 2D spectra allowed a complete assignment of all protons and carbons in the peptide backbone and the 7-methyloctanoyl fragment (Table 2.1). The *H-NMR spectrum (Figure 2.3) displayed a broad singlet resonance at 8H 12.69 that was assigned to the OH group in the free arginine C-terminus. All cc-NH and oc-H resonances for the five amino acid residues in 12 can be easily recognized around 5H 7.0-8.5 and 5H 4.2-4.6, respectively. Five aliphatic methylene resonances between 8H 1.06 and 2.00, a methine resonance at 8H 1.45 (8c 27.6), and a pair of isochronous methyl doublets at 8H 0.82 (5c 22.5), were assigned to the 7-methyloctanoyl alkyl chain. 21 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin U H ? fi H 4 ti H H O - U Y N y ^ N A r N ^ ^ N A r N / 0 H NH H N ^ ^ N H 2 OH JJUUJuAiJ 1.0 7.5 7.0 6.5 3.0 2.5 2.0 1.5 1.0 L V \dJ\JWW JV[ 13 12 11 10 9 5 7 6 5 Chemical Shift (ppm) 4 3 2 1 0 Figure 2.3. H-NMR spectrum of tauramamide (12) (recorded in DMSO-rf6 at 600 MHz). 22 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin JO \=r OH P.  H \ Q H f fl H NH H N ^ N H 200 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 0 Figure 2.4. 13C-NMR spectrum of tauramamide (12) (recorded in DMSO-J6 at 150 MHz). 23 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.1. NMR data for tauramamide (12) (recorded in DMSO-d6). OH Amino Acid Arg Trp D-Leu Carbon No 1 2 3 4 5 6 1 2 3 3' 3a' 4' 5' 6' T 7a' 2' 1 2 ^ H tSVSnr?x H HN NH NH 2 13 C 8 (ppm)a 173.3 51.6 28.0 25.2 40.3 156.6 171.8 53.1 27.4 109.9 127.1 118.5 118.0 120.7 111.2 136.0 124.0 171.6 51.2 8(ppm)(mult,7(Hz))b'c OH 12.69 (s, broad) 4.22 (m) NH 8.32 (d, 7 = 7.8 Hz) 1.64 (m), 1.79 (m) 1.35 (m), 1.54 (m) 3.12 (m) NH 7.48 (dd,y = 5.8, 5.8 Hz) 4.54 (m) NH 8.14 (d, 7 = 8.3 Hz) 2.88 (m), 3.16 (m) 7.62 (d, 7 = 7.8 Hz) 6.95(dd,7 = 7A7.4Hz) 7.03 (dd, 7 = 7.1,7.1 Hz) 7.30 (d, 7=8.0 Hz) NH 10.76 (d, 7=1.5 Hz) 7.10 (d, 7 =1.9 Hz) 4.21 (m) NH 7.75 (d, 7 = 7.8 Hz) 3 40.7 1.08 (m), 1.17 (m) 4 23.7 1.24 (m) 5 22.8 0.68 (d, 7 = 6.8 Hz) 5' 21.5 0.66 (d, 7 = 6.5 Hz) "Recorded at 150 MHz.bRecorded at 600 MHz.c According to HMQC recorded at HMBC" (H-»C) CI, C3, C4 C2, C3,TrpCl C1,C2,C4,C5 C2, C3, C5 C3, C4, C6 C5 C1,C3,C3' C2,C3, D-Leu CI C2, C3\ C3a\ C2' C3',C3a',C6',C7a' C3a\ C7' C4', C7a' C3a', C5' C3, C3a',C7a',C2' C3, C3', C3a', C7a' C1,C3 C2, C3, SerCl C2 C3, C5, C5' C3, C4, C5' C3, C4, C5 600 MHz. 24 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.1. NMR data for tauramamide (12) (recorded in DMSO-^) (Continuation). Amino Acid Ser D-Tyr Acyl chain HO' Carbon No 1 2 1 2 r 2' 3' 4' 5' 6' 1 2 3 4 5 6 7 8 8' OH M ^ N ^ O - N J M ^ O " 3 \ O " NH 2 1 3 C 8(ppm)a 169.7 55.0 61.7 171.7 54.5 36.6 127.8 130.0 114.7 155.7 114.7 130.0 172.5 35.1 25.2 28.7 26.5 38.2 27.6 22.51 22.49 OH 8(ppra)(mult,y(Hz))b'c 4.23 (m) NH 8.03 (d, J = 7.8 Hz) 3.44 (m), 3.51 (m) OH 4.82 (s, broad) 4.42 (m) NH 7.99 (d, J = 7.7 Hz) 2.62 (dd, J = 10.2, 14.0 Hz) 2.86 (m) 7.02 (d, 7=8.3 Hz) 6.61 (d, J =8.4 Hz) OH 9.14 (s) 6.61 (d, J =8.4 Hz) 7.02 (d, J =8.3 Hz) 2.00 (m) 1.35 (m) 1.06 (m), 1.14 (m) 1.07 (m), 1.15 (m) 1.09 (m) 1.45 (m) 0.826 (d,y = 6.8 Hz) 0.823 (d, 7 = 6.8 Hz) HMBCb (H-»C) C1.C3 C2,C3, D-Tyr CI CI ci,c3, cr C2, C3, Acyl chain CI C1,C2,C1',C2' C3,C3',C4' C1',C4' C3',C4',C5' C1',C4' C3, C3\ C4' C1,C3,C4 CI, C2, C5 C2, C3, C6, C7 C4.C7 C4,C5 C6, C8, C8' C6, C7, C8' C6, CI, C8 "Recorded at 150 MHz.D Recorded at 600 MHz.c According to HMQC recorded at 600 MHz. 25 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin The second pair of methyl doublets at 5H 0.67 (dc 21.5, 22.8) was attributed to the isopropyl moiety of D-Leu. Six carbonyls between 8c 169.0-174.0 (Figure 2.4) account for five amide and one carboxylic acid (8c 173.3) functionalities, while the five methines bearing NH are grouped around Sc 51.0-55.0. The amino acid sequence of tauramamide (12) is clearly indicated by HMBC correlations (Table 2.1, Figure 2.5) observed between Arg NH (8H 8.32) and Trp CI (5C 171.8), Trp NH (8H 8.14) and D-Leu CI (5C 171.6), D-Leu NH (5H 7.75) and Ser CI (5C 169.7), Ser NH (5H 8.03) and D-Tyr CI (8C 171.7), and finally, between D-Tyr NH (8H 7.99) and carbonyl CI (5C 172.5) of the 7-methyloctanoyl residue, at the TV-terminus of the pentapeptide Tyr-Ser-Leu-Trp-Arg. Esterification of 12 with EtOH and catalytic HC1 yielded ethyl ester (14) (Scheme 2.2). This compound was in agreement by HPLC, MS, specific rotation, and NMR comparison (Figure 2.6, and Table 2.6 in Experimental section) with the material extracted from B. laterosporus cells using EtOH as extracting solvent, confirming the structure proposed for tauramamide (12). Table 2.2. Antimicrobial activity (MIC's in |i.g/mL) of synthetic tauramamide (12) and tauramamide ethyl ester (14).69 Pathogen 12 14 MRSA 200 9.4 C albicans 50 75 Enteroccus sp. 0.1 0.1 Both 12 and 14 showed potent (MIC's 0.1 ug/mL) and relatively selective activity against the important Gram positive human pathogen Enterococcus sp.69 Ethyl ester (14) exhibited stronger activity against MRSA, but neither compound is appreciably active against C. albicans. 26 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin H s I HN H-<" = OH 3 / °. H-^5 J! H-2 = Jj H-4 = II ~NH H N " ^ N H 2 ArgNH TrpC1 TrpNH D-Leu C1 OH D-Leu NH SerC1 : 169.0 169.5 170.0 170.5 171.0 6 * 2 on u 171.5 E 172.0 ; 172.5 173.0 8.3 .2 8.1 8.0 F2 Chemical Shift (ppm) 7.9 : 173.5 Figure 2.5. Partial HMBC spectrum of tauramamide (12) (recorded in DMSO-J6 at 600 MHz). 27 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 7 6 5 4 Chemical Shift (ppm) Figure 2.6. H-NMR spectra of bacterial-derived and synthetic tauramamide ethyl ester (14) (recorded in DMSO-d6 at 600 MHz). 28 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2.4. Preparation of dealkylsurfactin Large-scale production of surfactin is performed via fermentation,75'115'116 since its biosynthesis is a common feature of members of the genus Bacillus. From private and public collections, about 20 strains of Bacillus subtilis have been listed as producers.75 Considerable resources have been directed to enhance production efficiency and recovery bioprocesses, with the surfactant market as the major consumer of surfactin and other similar lipopeptides (an industry of around US $9.4 billion per annun).115 Several surfactin analogues have been obtained either by genetic engineering or by directed biosynthesis. ' ' To date however, surfactin has not been able to compete economically with petroleum-derived surfactants mainly due to poor strain productivity and the need for expensive substrates.75 The first laboratory synthesis of a surfactin-like cyclic depsipeptide was reported in 1976 by Morrison, Ciardelli and Husman, who employed DCC-HOBt-assisted coupling of Boc- and /?-nitrobenzyl (NB)-protected amino acids in their method. Cyclization was achieved under high dilution conditions using DCC-HOSu, and proceeded in a 41% yield. The only structural difference with the natural compound was the lack of the 13-methyl group in the (3-hydroxy acid chain, but their analogue (named norsurfactin) showed comparable hemolytic and anticoagulant activities. Twenty years later, Nagai and coworkers119 synthesized surfactin B2 (19), mainly using active ester and azide fragment condensation methods. Cyclization took place under the same conditions as reported by Morrison,118 but afforded the desired macrolide in 73% yield. 1 'JA More recently in 2002, Pagadoy, Peypoux and Wallach developed a solid-phase preparation of surfactin Ci (20) and four additional analogues using Fmoc protecting chemistry and HATU as coupling reagent. Once the linear depsipeptide had been constructed and cleaved from the resin, HATU-HOAt-mediated cyclization also in high dilution yielded the corresponding macrolides in 22-35% yields. 29 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Our initial attempt to synthesize a surfactin analogue lacking the characteristic lipophilic side chain involved protection of (/?)-(-)-3-hydroxybutyric acid sodium salt (33) as a phenacyl ester,74'121 followed by CDI mediated112 coupling with Fmoc-Leu-OH (Scheme 2.3). After deprotection,122 intermediate (36) was then coupled with hexapeptide (38), acquired from the Peptide Synthesis Laboratory at UBC. O Br- QH O U J QH Q _ . _.. = O ,Pac O   I' ^J O I M \ ^ I j f  p Fmoc-Leu-OH H II ^ ^ ^ X > N a + ~ E t O H / H , 0 . reflux ^ ^ ^ C T a C CDI. THF. 78% F m o c - L e u - O ^ ^ ^ ^ C O  E t / 2 , fl  "^ ^ ^ CT G I, ,  Fmoc-Leu-CT ^  CT 33 34 35 Fmoc i OH Cpiu(Bn) O DLeu DLeu Bu4NFATHF O^^0H Asp(Bn) Val CH2CI2, 66% 36 ___ -Glu(Bn) Asp(Bn)—Val 39 + o 37 - o ' P a c 72% Scheme 2.3. Initial approach to surfactin analogues. Two experimental factors made this approach impractical. First, the basic conditions required to remove the Fmoc group produced a substantial amount of side product (37), presumably via E2 elimination as outlined in Scheme 2.4. The same byproduct was obtained when piperidine was employed as deprotecting reagent.122123 Secondly, the sample of hexapeptide (38) provided was 20 mg of a peptide mixture containing only 56% of the desired compound. Repetitive reversed-phase HPLC (80% CH3CN/H20 + 0.01% TFA) afforded 9 mg of 38, which upon PyBOP111 coupling furnished 39 (0.011 g, 0.0077 mmol). Such a small amount of linear precursor made it extremely difficult to measure and handle appropriate proportions of 30 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin reagents in following steps, and although LRESIMS showed peaks evidencing phenacyl deprotection, the expected product was never detected by NMR. ~o = o o ~~*Y^ " work-up ^ * * 5 ^ \ ) ' P a c + \ ^ ~ F-NBu4 37 Scheme 2.4. Proposed mechanism for the formation of 37. According to the Peptide Synthesis Laboratory, the presence of D-Leu in 38 was the main reason for the reduced scale in which the hexapeptide was prepared. Resin-bound D amino acids were at that moment not as commercially accessible as they are nowadays and thus, constituted a limiting factor in solid phase peptide synthesis. Furthermore, since 38 was ordered Fmoc- protected, unwanted elimination reactions were envisioned for the eventual Fmoc cleavage of the final linear precursor prior to cyclization. Therefore, based on the apparent incompatibility of the (/?)-3-hydroxybutyrate fragment with basic conditions, and the impure and limited amounts of 38 provided by solid phase peptide synthesis, it was decided to change the entire synthetic proposal to Boc-based chemistry, and synthesize the required linear peptide chain by sequential coupling of amino acids starting with a Boc-protected version of 35. This compound was obtained in good yield by DCC-mediated esterification of Boc-Leu-OH employing 34 as the alcohol component (Scheme 2.5).124 Upon TFA deprotection109'110 of intermediate (40), six cycles of PyBOP111 activated peptide coupling with Boc-D-Leu,125"'27 Boc-Asp(OBn), Boc-Val, Boc-D-Leu, Boc-L-Leu and Boc-Glu(OBn) in sequence, afforded 0.63 g (0.49 mmol) of the desired linear precursor (46) in a very good overall yield (a 64-fold mol improvement compared with its Fmoc-protected counterpart). 31 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin OH o Boc-Leu-OH ^ B o c ' N X ^ O ^ ^ J " o ' P a C DCC, CH2CI2 , 90% 34 1. TFA, CH2CI2 2. Boc-Asp(OBn), DIEA, PyBOP, CH2CI2 1 . TFA, CH 2 CI 2 2. Boc-Val, PyBOP, DIEA, CH 2 CI 2 1. TFA, CH 2CI 2 2. Boc-D-Leu, PyBOP, DIEA, 1. TFA, CH 2CI 2 2. Boc-Leu, PyBOP, DIEA, CH 2CI 2 1. TFA, CH2CI2 2. Boc-Glu(OBn), PyBOP, DIEA, CH2CI2 Overall yield: 63% (5 steps) 40 0 = 0 1. TFA, CH 2CI 2 2. Boc-D-Leu, PyBOP, DIEA, CH 2CI 2 , 99% BOC -N-VN^A0^k- ' lVPaC * 41 BOC ,Pac y° OBn 42 Boc- Ok a ° y ° Y OBn ^ ,Pac BOC- y y -y yy-yH v^ rAAr p a c ~--~s;0 OBn 44 Boc^ N - A N A T N y | - N - y N y j - o ^ ^ o ' • « - - « o y K^Sr^V^K^T^T^0^^0' Pac ^ O OBn Scheme 2.5. Preparation of the linear intermediate (46). As with the synthesis of tauramamide, each coupling step proceeded smoothly and furnished reasonably pure products, not requiring exhaustive purification. C- and TV-termini 32 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin deprotections, followed by a last PyBOP-assisted amide formation yielded macrolide (47) (Scheme 2.6). This key cyclization step was performed under dilute conditions to prevent polymerization of the linear precursor, as suggested by literature precedents.74'118 Still, it may account for the low isolated yield of 47, comparable with those reported by Morrison118 and Pagadoy in their syntheses of surfactin analogues. Quantitative removal of benzyl groups on the aspartic and glutamic residues via hydrogenolysis112'119 gave final product (48), which was named dealkylsurfactin. N ^ N ^ X ^ ^ Pac H II s o  Y 1. Zn, HOAc 2. TFA, CH2CI2 3. PyBOP, DIEA, CH 2CI 2 , 30% OR O = O | H H | U 0<T ° / ° ^ ^ OR 47 R=Bn , H2 , Pd/C, HOAc, 99% 48 R=H -^ Scheme 2.6. Preparation of dealkylsurfactin (48). HRESEVIS of 48 displayed a [M+Na]+ peak at m/z 904.5015, consistent with the target molecular formula C42H71N7O13 (calculated for C42H7iN7Oi3Na: 904.5008). 33 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin O^^OH 0 = 0 O^^NH H HNk /LJ ^ f  H H HN^O 8.50 8.25 8.00 7.75 J I i . -> 5.00 4.75 4.50 4.25 4.00 O'' 2.5 2.0 1.5 U\iJ! 13 12 11 10 9 3 7 6 5 Chemical Shift (ppm) 4 3 2 1 0 Figure 2.7. 'H-NMR spectrum of dealkylsurfactin (48) (recorded in DMSO-cfo at 600 MHz). 34 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 174 173 172 171 170 169 64 56 48 40 32 O^^OH 180 160 140 120 100 80 60 Chemical Shift (ppm) 40 20 0 Figure 2.8.13C-NMR spectrum of dealkylsurfactin (48) (recorded in DMSO-cfc at 150 MHz). 35 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.3. NMR data for dealkylsurfactin (48) (recorded in DMSO-d6). , O H Amino Acid (fl)-3-HBA 'Leu D-'Leu Asp Carbon No 1 2 3 4 1 2 3 4 5 5' 1 2 3 4 5 5' 1 2 3 S^CA^NV 5 Q ^ N H " H N ^ > k / S ' ^ S l>IH  u u HN^O 5 1 3 C 5 (ppm)a 169.4 41.9 68.6 19.5 171.6 51.4 38.6 24.2 21.0 23.2 172.0 50.2 41.9 24.2 22.8 22.8 169.8 49.5 36.2 OH 4' l j j 1  8(ppm)(mult,7(Hz))b,c 2.37 (dd, 7 = 5.9, 13.8 Hz) 2.42 (dd, 7 = 7.9, 13.8 Hz) 5.01 (m) 1.11 (d, 7 = 5.9 Hz) 4.00 (m) NH 8.56 (d, 7 = 6.6 Hz) 1.45 (m), 1.61 (m) 1.63 (m) 0.82 (d, 7 = 6.4 Hz) 0.88 (d, 7 = 6.3 Hz) 4.48 (m) NH 7.84 (d, 7 = 9.5 Hz) 1.37 (m), 1.40 (m) 1.46 (m) 0.87 (d, 7 = 6.9 Hz) 0.87 (d, 7 = 6.9 Hz) 4.53 (m) NH 8.15 (d, 7 = 6.9 Hz) 2.56 (dd, 7 = 9.5,17.1Hz) 2.68 (dd, 7 = 4.3, 16.7 Hz) -5' HMBCb (H->C) CI, C3, C4, C1,C2,C4,'LeuCl C1,C2 CI, C3, C4 C2, C3, D-'LeuCl C2, C4, C5, C5' C2, C3, C5, C5' C3 C3, C4, C5 C1,C3, C4,AspCl C2, Asp CI C1,C2,C4 C2 C3, C4, C5' C3, C4, C5 CI, C3, C4 C2, C3, Val CI C1,C2,C4 4 171.6 OH 12.17d(s, broad) 'Recorded at 150 MHz.bRecorded at 600 MHz.c According to HMQC recorded at 600 MHz. d Interchangeable. 36 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.3. NMR data for dealkylsurfactin (48) (recorded in DMSO-Jg) (Continuation). ,OH Amino Acid Val D-2Leu 2Leu Glu Carbon No 1 2 3 4 4' 1 2 3 4 5 5' 1 2 3 4 5 5' 1 2 3 4 5 1 H 1 .1 5 ' ^ ^ * NH . . . . H N T ^ O 5 1 3 C 5 (ppm) 170.5 58.2 30.6 19.2 18.4 172.1 51.6 39.4 24.2 21.9 20.8 172.4 52.6 39.2 24.2 22.4 22.2 171.0 51.8 27.2 29.6 174.1 3  o4r N>^C'N^ l-"13 OH 4' *H B  8(ppm)(muIt,/(Hz))b'c 4.09 (dd, 7 = 8.6,8.9 Hz) NH 7.86 (d, 7 = 9.1 Hz) 1.97 (m) 0.89 (d, 7 = 6.9 Hz) 0.75 (d, 7 = 6.6 Hz) 4.13 (m) NH 8.38 (d, 7 = 5.6 Hz) 1.46 (m), 1.54 (m) 1.58 (m) 0.82 (d, 7 = 6.6 Hz) 0.78 (d, 7 = 6.2 Hz) 4.06 (m) NH 8.14 (d, 7 = 6.9 Hz) 1.46 (m) 1.48 (m) 0.85 (d, 7 = 6.5 Hz) 0.84 (d, 7 = 6.2 Hz) 4.21 (m) NH 7.79 (d, 7 = 6.3 Hz) 1.81 (ra), 1.96 (m) 2.20 (m) OH 12.28d (s, broad) . 5 ' HMBCb (H->C) Cl,C3,C4,C4',D-2LeuCl C2, D-2LeuCl C2, C4' C2, C3, C4' C2, C3, C4 C1,C3,C4 CI, C2, 2Leu CI C2,C4,C5',2LeuCl C3, C5, C5' C4 C4 C1,C3, C4 C1,C2,C4 C1,C5,C5' C3, C4, C5' C3, C4, C5 CI, C3, C4, 3-HBA CI C1,C2,C3, 3-HBAC1 CI, C2, C4, C5 C2, C3, C5 "Recorded at 150 MHz.b Recorded at 600 MHz.c According to HMQC recorded at 600 MHz.d Interchangeable. 37 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin The 'H-NMR spectrum of 48 (Figure 2.7, Table 2.3) showed four well defined regions typical of peptides. The most deshielded broad singlets around 6H 12.20 undoubtedly corresponded to both carboxylic functionalities in the aspartic (8c 171.6) and glutamic (8C 174.1) residues. Seven multiplets (two of them isochronous), all assigned to NH groups of 48 can be located between 7.75 and 8.55 ppm. Another set of multiplets between &H 4.00-5.05 were assigned to the seven NH-bearing a-methines, plus an oxygenated stereocenter in the (R)-3- hydroxybutyrate fragment (6H 5.01, 6c 68.6). Each of the diastereotopic methylene protons adjacent to this chiral carbon resonates as a doublet of doublets at 6H 2.42 and 6H 2.37 (6c 41.9), and were located in the fourth most shielded spectral region. The heavy Leu content in dealkylsurfactin is evidenced by a series of doublets close to 0.88 ppm, integrating for a total of 30 protons. This region also includes methyls for the Val residue, one of which is highly shielded and gives a doublet at 6H 0.75 ppm (6c 18.4). Relevant in this area of the 'H-NMR spectrum is a more deshielded doublet at 6H 1.11 (5c 19.5), assigned to the methyl of the (R)-3- hydroxybutyrate fragment. The C-NMR spectrum (Figure 2.8, Table 2.3) revealed five clear regions. Ten completely isolated resonances at the deshielded end of the spectrum (5c 169.4-174.1) account for all the carbonyls present in 48. These signals are followed by a group of eight methines attached to heteroatoms (5c 49.5-68.6). Obscured by the residual solvent peak and very typical in Leu-containing compounds, are the methylenes of its isobutyl side chain located between 5c 38.6-41.9. A fourth assembly of signals around 5c 27.2-36.2 includes a methine and three methylenes assigned to the side chains of Val, Asp and Glu residues. Finally, the most shielded cluster of peaks between 18.0 and 26.0 ppm reflects, for a second time, the high Leu content characteristic of the surfactin lipopeptide family. 38 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin As with tauramamide (12), the amino acid sequence of dealkylsurfactin (48) was determined using HMBC correlations (Figure 2.9, Table 2.3). The Glu NH and a-CH protons (8H 7.79 and 4.21 respectively) presented cross-peaks with the carbonyl of (ft)-3-hydroxybutyrate (5c 169.4), which in turn correlated through its 0-CH proton (5H 5.01) with 'Leu CI (8C 171.6). Likewise, 'Leu NH (5H 8.56) exhibited HMBC cross-peaks with D-'Leu CI (5C 172.0), D-!Leu NH (5H 7.84) with Asp CI (5C 169.8), Asp NH (5H 8.15) with Val CI (5C 170.5), Val NH (5H 7.86) with D-2Leu CI (5C 172.1), and finally, D-2Leu NH (5H 8.38) as well as p-CH2 (8H 1.46- 1.54) correlated with 2Leu CI (5c 172.4). Cross-peaks between this last residue and carbons in the Glu fragment were not evident. Dealkylsurfactin (48) was levorotatory (Table 2.4) in agreement with optical rotation values found in the products of other revised syntheses of surfactin analogues.118'119 Table 2.4. Specific rotation values for natural surfactin B2 (19), synthetic analogues (19a,b), "norsufactin" (19c), and dealkylsurfactin (48).118'119 Compound [ah 1 (c, MeOH) (t °C) Surfactin B2 (19) -36.5 (1 )(22) Surfactin B2 (19a)a -37.0 (1)(9.5) Surfactin B2 (19b)b -19.6 (1)(12.5) Norsurfactin (19c)a -35.2 (1 )(25) Dealkylsurfactin (48) -22.0 (3.8)(20) "Prepared using (+)-3-hydroxytetradecanoic acid. Prepared employing (-)-3-hydroxytetradecanoic acid. 39 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin fifi i.4 8.3 8.2 8.1 8.0 F2 Chemical Shift (ppm) 3-HBA H3 3-HBA C1 r 169.5 3-HBA H3 168.5 169.0 170.0 t 170.5 r y 171.0 S U 171.5 leu C1 172.0 t 172.5 173.0 5.025 Figure 2.9. Partial HMBC spectrum of dealkylsurfactin (48) (in DMSO-d6 at 600 MHz). 40 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2.5. Conclusions Tauramamide (12), a linear antibiotic acylpentapeptide recently isolated from cultures of Brevibacillus laterosporus (PNG-276) collected in Papua New Guinea, was synthesized in nine steps and 29% overall yield. Esterification of 12 (86%) yielded ethyl ester (14), which was identical by HPLC, MS, specific rotation, and NMR comparison with the material obtained from B. laterosporus cells using EtOH as the extracting solvent. Besides confirming the structure proposed for tauramamide (12), the synthetic material allowed an exhaustive antimicrobial assessment for this new family of PNG-276-derived antibiotics. Both 12 and 14 showed potent (MIC's 0.1 ug/mL) and relatively selective activity against the important Gram positive human pathogen Enterococcus sp.69 Ethyl ester (14) exhibited stronger activity against MRSA, but neither compound is appreciably active against C. albicans. Our convergent synthetic approach to tauramamide (12) (Schemes 2.1 and 2.2) started simultaneously from both ends of the molecule. Most reactions afforded relatively pure intermediates, which were used directly in the following reaction without any further 41 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin purification. Only the final deprotection leading to 12 required optimization in order to minimize amounts of partially benzylated tauramamide derivatives obtained as byproducts. A new analogue of the surfactin depsipeptide family, named dealkylsurfactin (48), has been prepared in 10 steps and 14% overall yield. The compound was employed as a biological tool in binding studies between the mitotic regulator isomerase Pinl and the microtubule- associated protein tau, a crucial interaction involved in the development of Alzheimer's disease symptoms. 9"101 The parent compound surfactin Q (20) had been previously isolated via bioassay-guided fractionation from cultures prepared using the microorganism PNG10A (also collected in Papua New Guinea). , O H ^ ^ VH H H "" ° cArN>-^NYJ'"'| OH 20 R=(CH2 )9CH(CH3 )2 48 R=H Our synthetic pathway to dealkylsurfactin (48) (Schemes 2.5 and 2.6) began with protection of the C-terminus, followed by sequential standard peptide elongation in solution using Boc-protecting chemistry and PyBOP as the coupling reagent for all amide bond formation steps. Such methodology furnished the desired linear precursor in a 64-fold yield improvement, when compared with a previously attempted preparation employing Fmoc solid phase synthesis. The key cyclization step was run under dilute conditions and proceeded in 30% yield, in 42 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin concordance with similar reported syntheses.105'106'118,120 Polymerization of the linear precursor may have been the main contributor to the low isolated yield. Biological evaluation of dealkylsurfactin (48) showed that the compound does not inhibit binding of Pinl to phosphorylated tau. Thus, the macrocyclic peptide fragment in surfactin Ci (20) does not play any role in activity, and more likely its efficient surfactant properties were responsible for the observed inhibition during assay-guided isolation. Such unspecific interaction is not ideal in the development of potential pharmaceuticals, since unwanted side effects are very likely to arise. In order to minimize the possibility of amino acid racemization in both syntheses, the amounts of base used were carefully controlled. This aspect was particularly important during the preparation to dealkylsurfactin (48), given the apparent susceptibility towards basic conditions shown by the (/?)-3-hydroxybutyrate fragment. Diisopropylethylamine (DIEA) was the only base employed, limited to assisting PyBOP during amide bond formation and is not i 198 1 9Q strong enough (Pr3NH /?Ka~l 1) to abstract a-protons in amino acids (/?Ka~15). ' The use of a urethane-based protecting strategy, with J-butyloxycarbonyl (Boc) amino acids as building blocks, as well as the fast and efficient coupling reagent PyBOP, is reported to be advantageous in avoiding racemization and dehydration side reactions. n ' Noteworthy, the levorotatory nature of the corresponding parent molecules was retained in both 14 and 48. Although the present status of linear peptide synthesis has reached a high degree of enantiomeric purity, macrocyclization of linear peptides still poses problems, such as low yields (as evidenced in the present work) and epimerization.130 After all, the process is entropically unfavored since it involves the loss of rotational freedom.105'131 Selection of a suitable cyclization point and the necessity of dilute conditions in order to inhibit intermolecular processes, are key aspects that must be considered when planning the synthesis of macrocyclic peptides. 43 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2.6. Experimental General experimental procedures All reactions described in this thesis were performed under dry nitrogen or argon using glassware previously oven dried (150°C), unless otherwise specified. Glassware was allowed to reach room temperature under a flow of inert gas. Likewise, glass syringes and stainless steel needles, used to handle anhydrous reagents and solvents, were oven dried, cooled in a desiccator, and flushed with inert gas prior to use. With the exception of THF and CH2CI2, which were distilled from sodium/benzophenone and CaH2 respectively, MeOH, benzene, toluene, DMF and pyridine were purchased anhydrous quality and used without further purification. HPLC grade solvents (MeCN, acetone, hexanes, CCI4, EtOAc and CH2CI2) were used without further purification. EtOH reagent grade was treated with activated molecular sieves prior to use. All chemical reagents were purchased in an analytical or higher grade from Aldrich or Fluka. Cold baths were prepared using ice/water, ice/NaCl/water, MeCN/dry ice and acetone/dry ice, for 0, -10, -40 and -78 °C respectively. Liquid nitrogen was employed for condensing ammonia. Flash chromatography was carried out with 70-230 and 230-400 mesh silica gel (Silicycle). For reverse phase column chromatography, Sep Pak® C18 columns (Waters) were used. Size exclusion chromatography was performed using lipophilic Sephadex® LH-20 (Sigma, bead size 25-100p.). Precoated silica gel plates (Merck, Kieselgel 60 F254, 0.25 mm and Whatman, MKC18F 60 A) were employed in normal and reversed-phase thin layer chromatography (TLC). TLC visualization was accomplish using ultraviolet light (254 nm), 44 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin followed by heating the plate after staining with vanillin in H2S04/EtOH (6% vanillin w/v, 4% H2S04 v/v, and 10% H20 v/v in EtOH), p-anisaldehyde in H2S04/EtOH (5% p-anisaldehyde v/v and 5% H2S04 v/v in EtOH) or 20% KMn04 w/v in H20. High performance liquid chromatography (HPLC) was carried out using a Waters 1500 Series pump system, equipped with Waters 2487 dual A absorbance detector and either a CSC-Inertsil 150A/ODS2 column, or an Alltech Econosil Silica 5u column. NMR spectra were recorded using chloroform-d (CDCI3), methylene chloride-^ (CD2C12), dimethylsulfoxide-^ (DMSO-d6), methanol-^ (CD3OD), benzene-d6 (C6D6), deuterium oxide (D20), acetonitrile-d? (CD3CN) or acetone-^ (CD3COCD3). Chemical shifts (8) are given in parts per million (ppm) relative to tetramethylsilane (5 0) and were calibrated internally to the signal of the solvent in which the sample was dissolved (CDC13: 5 7.24 *H- NMR; 5 77.0 13C NMR; CD2C12 5 5.32 'H-NMR; 5 54.0 13C NMR; DMSO-<4: 6 2.50 'H-NMR; 8 39.51 13C NMR; CD4OD: 8 3.31 'H-NMR; 8 49.15 13C NMR; C6D6: 8 7.16 'H-NMR; 8 128.39 13C NMR; D20: 8 4.80 'H-NMR; CD3CN: 8 1.94 !H-NMR; 8 118.69 13C NMR; CD3COCD3: 8 2.05 'H-NMR; 8 29.92 13C NMR). 'H-NMR spectral data are tabulated in the order: multiplicity (s, singlet; d, doublet; dd, doublet of doublets; t, triplet; q, quartet; m, multiplet), coupling constant, number of protons and proton assignment where applicable. !H-NMR data was acquired using Bruker spectrometers WH400 (400 MHz), Avance 300 (300 MHz), Avance 400 (400 MHz), or Avance 600 (600 MHz) equipped with a CRYOPROBE®. 13C-NMR spectra were recorded on Avance 300 (75 MHz), Avance 400 (100 MHz), or Avance 600 (150 MHz). 31P- NMR data was collected using the Avance 400 (100 MHz) spectrometer. 2D-NMR data was acquired using the following pulse sequences and parameters. 45 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin gs-COSY *H Field Gradients <ti pi tj  P2 r\ r\ aq !tl ? - pi : x, -x p2: x, -x, -x, -x aq: x, -x p i , p2: 90° 'H d l : 2 s t i : 3 us gs-HMQC 'H 13/- Field Gradients dl pi d2 tj/2 ' 1 - d3 p4 Ih /±L aq GARP nl 22 «3 p i , p2: x p3: x, -x p4: (x)2, (-x)2 aq: x, (-x)2, -x p i : 90° 'H p2: 180° 'H p3, p4: 90° 13C d l : 2 s d2: 3.57 ms tj: 3 us GARP: 19 dB, 70 us gs-HMBC 13 , Field Gradients dl pi d2 «l-'2 V2 ti-2 p3 d3 p4 ih a. gl «3 aq pl. p3 p4: p5. aq: pl: p2: p2:x X X, -X X X, -X 90°'H 180° 'H p3, p4, p5: 90° 13C dl: d2: d3: t,: 2s 3.57 ms 60 ms 3 us gs-TOCSY 'H Field Gradients dl pl t , p2 aq pl : x p2: spinlock of composite 180° pulses (MLEV16) aq: x p l : 9 0 o l H p2: composite 180° pulses (100 ms) d l : 2 s t,: 3 us 46 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin gs-HSQC 'H 13r x x x(y)^a-.v), x x x x X- (-X)-., X g dl p ld : p2 dZ P3 1)4 tj/2 p5 t •H ^ 1— ' ! • • d3pfid2 p~ <L>p8d2 «9 d2pl0d3pll x x, -x x > • ".v x pl2 p!3 pH pl5 pl6 p n pl8 Field Gradients rA pi, p4, p6, p8, plO: 90°'H p2, p5, p7,p9,pll: 180°'H p3:2ms'H pl3,pl5,pl7:90o l3C  P12,pl4,pl6,pl8: 180° 13C dl :2s d2: 1.8 ms d3: 1.6 ms ti:3 us GARP: 19 dB, 70 us gs-NOESY 'H Field Gradients dl pi !j p2 p3 0_ aq gl pi: x, -x p2, p3: x aq: x, -x pl,p2,p3:90o lH d l :2s ti:3us Low and high resolution electron impact (EI) mass spectra were recorded on Kratos MS50 or MS80 mass spectrometers at 70 eV. Low and high resolution electrospray (ESI) mass spectra were obtained with Bruker Esquire-LC and Micromass LCT mass spectrometers. Circular dichroism (CD) data was recorded with a JASCO J-810 CD spectrometer at 20.0°C, using a 2.0 mm micro cell. Optical rotations were measured with a JASCO P-1010 polarimeter at 20°C and 589 nm (sodium D line). UV spectra were acquired with a Waters 2487 Dual X Absorbance Detector, using a 1 cm cell. 47 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2.6.1. Total synthesis of tauramamide and tauramamide ethyl ester BOC:  i A X To a solution of Boc-Arg(Cbz)2-OH (2.5 g, 4.6 mmol) and benzyl alcohol (0.71 mL, 6.9 mmol) in CH2C12 (50 mL) at 0°C, DCC (1.04 g, 5.06 mmol) and DMAP (0.06 g, 0.46 mmol) were added and the resulting mixture was stirred overnight. The solution was poured into water and extracted with CH2CI2 (30 mL x 3). The organic extracts were dried (Na2SC>4) and concentrated in vacuo. Purification by Sephadex LH20 column chromatography (100% MeOH) afforded (24) as a colorless solid (2.91 g, 97%). lH NMR (CDC13, 600 MHz) £9.36 (IH, s), 9.12 (IH, s), 7.30-7.10 (15H, m), 5.30 (IH, d, / = 8.3 Hz), 5.10-4.90 (6H, m), 4.24 (IH, m), 3.83 (2H, m), 1.85-1.70 (4H, m), 1.31 (9H, s). 13C NMR (CDC13, 100 MHz) £171.9 (C), 163.3 (C), 159.9 (C), 155.2 (C), 155.0 (C), 136.5 (C), 135.0 (C), 134.3 (C), 128.3 (3CH), 128.0 (2CH), 127.9 (4CH), 127.8 (3CH), 127.7 (CH), 127.4 (CH), 127.3 (CH), 79.1 (C), 68.3 (CH2), 66.4 (2CH2), 52.9 (CH), 43.6 (CH2), 28.7 (CH2), 27.8 (3CH3), 24.4 (CH2). HRESIMS calcd for C34H4oN408Na ([M+Na]+): 655.2744; found 655.2743. Preparation of N^-Boc-Nx, A^-di-Z-arginme benzyl ester (24) H Boc"NY CbzN O II X 3 'OH DCC, DMAP  C b 2 ^ J l_ J N N . . H H 24 48 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Preparation of Boc-Ser(Bn)-DLeu-Trp-Arg(Cbz)2-OBn (27) 1 TFA Boc-Arg(Cbz)2-OBn — — — • Boc-Ser(Bn)-D-Leu-Trp-Arg(Cbz)2-OBn 2. Boc-X-OH 24 X: Tip, D-Leu, Ser(Bn) 27 3. PyBOP, DIEA A solution of Boc-Arg(Cbz)2-OBn (24) (1.87 g, 2.9 mmol, 1 eq) in CH2C12 (25 mL) was treated with TFA (3 mL, 40.3 mmol, 14 eq) and stirred for 4 h, whereupon it was concentrated in vacuo to a pale red oil. This oil was dissolved in CH2CI2 (20 mL) and added to a solution of Boc- Trp-OH (0.82 g, 2.7 mmol, 0.90 eq) and PyBOP (1.69 g, 3.24 mmol, 1.10 eq) in CH2C12 (10 mL); followed by addition of DIEA (1.55 mL, 8.90 mmol, 3 eq). After overnight stirring at 25 °C, NH4CI sat. was added and CH2C12 (3x20 mL) extractions performed. The organic extracts were combined and dried (Na2SC>4), to be then concentrated in vacuo. Column chromatography on silica (80% EtOAc/Hexanes) afforded (25) as a white amorphous solid (1.87 g, 85%). The procedure above was repeated successively with Boc-DLeu-OH (0.48 g, 2.07 mmol), and Boc- Ser(Bn)-OH (0.46 g, 1.57 mmol); to afford Boc-DLeu-Trp-Arg(Cbz)2-OBn (1.93 g, 76%) and Boc-Ser(Bn)-DLeu-Trp-Arg(Cbz)2-OBn (27) (1.75 g, 81%), respectively. *H NMR (CDC13, 400 MHz) £9.40 (IH, s), 9.28 (IH, s), 9.00 (IH, s), 7.26-6.90 (30H, m), 5.65 (IH, m), 5.20-4.90 (6H, m), 4.80 (IH, dd, J = 6.3, 7.4 Hz), 4.55-4.40 (2H, m), 4.40-4.25 (2H, m), 3.95-3.80 (2H, m), 3.67 (IH, m), 3.50 (IH, dd, J = 6.1, 2.8 Hz), 3.31 (IH, dd, 7= 6.3, 8.7 Hz), 3.15 (IH, dd, J = 5.7, 8.7 Hz), 1.85-1.75 (IH, m), 1.75-1.60 (2H, m), 1.60-1.35 (4H, m), 1.41 (9H, s), 0.80 (6H, m); 13C NMR (CDCI3, 100 MHz) £171.4 (C), 171.2 (C), 170.9 (C), 170.8 (C), 170.5 (C), 163.4 (C), 160.2 (C), 155.4 (C), 137.1 (C), 136.4 (C), 135.9 (C), 135.0 (C), 134.3 (C), 128.4 (4CH), 128.2 (4CH), 128.0 (3CH), 128.0 (3CH), 127.9 (3CH), 127.8 (CH), 127.8 (CH), 127.6 (CH), 127.4 (C), 123.4 (CH), 121.5 (CH), 119.0 (CH), 118.2 (CH), 111.1 (CH), 109.3 (C), 79.7 (C), 72.8 (CH2), 69.4 (CH2), 68.5 (CH2), 66.7 (CH2), 66.6 (CH2), 66.5 (CH2), 60.0 (CH2), 54.0 (CH), 53.6 (CH), 49 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 52.2 (CH), 51.9 (CH), 43.7 (CH2), 40.2 (CH2), 27.9 (3CH3), 24.1 (CH), 22.5 (CH2), 21.4 (CH3), 20.7 (CH3). HRESIMS calcd for C6iH72N80i2Na ([M+Na]+): 1131.5167; found 1131.5162. Preparation of Boc-DTyr(Bn)-OMe (29) HO' A solution of Boc-DTyr-OMe (2.0 g, 6.8 mmol), K2C03 (1.4 g, 10.1 eq) and BnBr (1 mL, 8.1 mmol) in acetone was refluxed for 3 h and stirred overnight at 25 °C. After solid filtration and solvent evaporation, column chromatography on silica (30% EtOAc/Hexanes) afforded (29) as an amorphous white solid (2.60 g, 99%). *H NMR (CDC13, 400 MHz) £7.35-7.20 (5H, m), 6.94 (2H, d, / = 5.6), 6.81 (2H, d, 7 = 5.7), 4.94 (2H, s), 4.89 (IH, d, / = 5.0 Hz), 4.45 (IH, dd, / = 4.4 Hz), 3.61 (3H, s), 2.95 (IH, dd, J = 3.64, 9.2 Hz), 2.91 (IH, dd, J = 3.6, 9.0 Hz), 1.33 (9H, s); 13C NMR (CDC13, 100 MHz) S 172.3 (C), 157.8 (C), 155.0 (C), 136.5 (C), 130.2 (2CH), 128.5 (2CH), 128.2 (C), 127.9 (CH), 127.4 (2CH), 114.2 (2CH), 79.8 (C), 69.9 (CH2), 54.5 (CH3), 52.1 (CH), 37.4 (CH2), 28.2 (3CH3). HRESIMS calcd for C22H27N05Na ([M+Na]+): 408.1787; found 408.1790. 50 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Preparation of (CH^CH^CH^jCONH-DTvrtBnVOMe (30) o o ^Y^Y^-cr^ 1 •T F A r^vr^Y"xr" 29 30 A solution of Boc-DTyr(Bn)-OMe (29) (1.50 g, 3.9 mmol) in CH2C12 (35 mL) was treated with TFA (2 mL, 27 mmol) and stirred for 4 h, whereupon it was concentrated in vacuo to a colorless oil. This oil was dissolved in CH2CI2 (10 mL) and added to a solution of 7- methyloctanoic acid (0.62 g, 3.9 mmol) and PyBOP (2.0 g, 3.8 mmol) in CH2C12 (40 mL); followed by addition of DIEA (2.0 mL, 12 mmol). After overnight stirring at 25 °C, NH4CI sat. was added and CH2CI2 (3x20 mL) extractions performed. The organic extracts were combined and dried (Na2SC>4), to be then concentrated in vacuo. Column chromatography on silica (80% EtOAc/Hexanes) afforded (30) as a yellowish powder (1.67 g, 88%). lH NMR (CDC13, 400 MHz) (J7.35-7.15 (5H, m), 6.98 (2H, d, 7 = 8.5), 6.80 (2H, d, 7 = 8.5), 6.71 (IH, d, 7 = 8.0), 4.87 (2H, s), 4.79 (IH, dd, 7 = 6.8, 7.2 Hz), 3.57 (3H, s), 3.01 (IH, dd, 7 = 5.6, 14.0 Hz), 2.89 (IH, dd, 7 = 7.0, 14.0 Hz), 2.10 (2H, t, 7 = 7.6 Hz), 1.51 (2H, quintet, 7 = 7.96 Hz), 1.43 (IH, septet, 7 = 6.5 Hz), 1.18 (4H, m), 1.08 (2H, quintet, 7 = 6.8 Hz), 0.80 (6H, d, 7 = 6.5 Hz); 13C NMR (CDC13, 100 MHz) £172.6 (C), 171.8 (C), 157.3 (C), 136.5 (C), 129.7 (2CH), 128.0 (C), 127.9 (2CH), 127.3 (CH), 126.8 (2CH), 114.2 (2CH), 69.2 (CH2), 52.8 (CH), 51.5 (CH3), 38.3 (CH2), 36.4 (CH2), 35.7 (CH2), 28.9 (CH2), 27.4 (CH), 26.6 (CH2), 25.1 (CH2), 22.1 (2CH3). HRESMS calcd for C26H36NO4 ([M+H]+): 426.2644; found 426.2645. 51 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Preparation of (Cft)2CH?(CH7)sCONH-DTvr(Bn)-Ser(Bn)-DLeu-Trp-Arg(Cbz)9-OBn (32) (CH3)2CH2(CH2)5CONH-D-Tyr(Bn)-OMe 30 L iOH.H 20 (CH3)2CH2(CH2)5CONH (CH3)2CH2(CH2)5CONH-D-Tyr(Bn)-OH i ONH-D-Tyr(Bn)-OH D-Tyr(Bn) PyBOP, DIEA 3 i J , D , NH2-Ser(Bn)-D-Leu-Trp-Arg(Cbz)2-OBn ber(Bn) T  T F A BnO-Arg(Cbz)2-Trp-D-Leu Boc-Ser(Bn)-D-Leu-Trp-Arg(Cbz)2-OBn 27 A solution of (CH3)2CH2(CH2)5CONH-DTyr(Bn)-OMe (30) (0.717 g, 1.68 mmol) and LiOH.H20 (0.705 g, 16.8 mmol) in l,4-dioxane/H20 (150 mL, 2:1) was stirred at 25 °C. After 1 h, TLC (80% EtOAc/Hexanes) showed absence of starting material. Solvents were evaporated off and the resulting residue was dissolved in H2O, acidified and extracted with EtOAc (3x30 mL). Combined organic extracts were dried (Na2S04) and concentrated to obtain a white solid (free acid, 0.68 g, 98%). In a separate reaction, Boc-Ser(Bn)-DLeu-Trp-Arg(Cbz)2-OBn (27) (1.42 g, 1.3 mmol) in CH2C12 (50 mL) was treated with TFA (3 mL, 40.3 mmol) and stirred for 5 h, whereupon it was concentrated in vacuo to a red oil. This oil was dissolved in CH2CI2 (20 mL) and added to a solution of free acid (0.58 g, 1.41 mmol) and PyBOP (0.73 g, 1.41 mmol) in CH2C12 (30 mL); followed by addition of DIEA (0.66 mL, 3.84 mmol). After stirring 16 h at 25 °C, reaction was quenched with NH4CI sat. and 1 M HC1 used to neutralize aqueous layer. CH2CI2 (3x30 mL) extractions were performed; the organic extracts were combined and dried (Na2S04), to be then concentrated in vacuo. Column chromatography on silica (80% EtOAc/Hexanes) afforded (32) as pale yellow leaves (1.45 g, 81%). 'H NMR (DMSO, 600 MHz) £10.74 (1H, s), 9.16 (1H, s, broad), 8.52, (1H, d, J = 6.9 Hz), 8.15, (1H, d, J = 8.0 Hz), 8.13, (1H, d, J = 8.6 Hz), 8.00 (2H, d, J = 6.9 Hz), 7.58 (1H, d, J = 7.7 Hz), 7.60-6.80 (25H, m), 7.13 (3H, m), 7.09 (1H, d, / = 2.0 Hz), 7.03 (1H, dd, J = 7.1, 7.1 Hz), 6.93 (1H, dd, J = 7.4, 7.4), 52 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 6.85 (3H, m), 5.20 (2H, m), 5.05-5.00 (6H, m), 4.55 (3H, m), 4.29 (2H, m), 3.90 (1H, m), 3.85 (1H, m), 3.61-3.45 (4H, m), 3.07 (1H, m), 2.85 (1H, m), 2.65 (1H, m), 1.98 (2H, m), 1.70 (1H, m), 1.43 (3H, m), 1.32 (4H, m), 1.25-1.00 (10H, m), 0.79 (3H, d, J = 6.2 Hz), 0.78 (3H, d, J = 6.3 Hz), 0.62 (6H, d, J = 5.5 Hz); 13C NMR (CDC13, 150 MHz) S 172.3 (C), 172.2 (C), 171.9 (C), 171.6 (C), 171.4 (C), 168.9 (C), 162.9 (C), 159.7 (C), 156.8 (C), 154.9 (C), 138.1 (C), 137.15 (C), 137.04 (C), 135.9 (C), 135.8 (C), 135.2 (C), 130.1 (CH), 130.0 (CH), 128.5 (2CH), 128.4 (4CH), 128.3 (2CH), 128.2 (4CH), 128.2 (2CH), 129.1 (2CH), 128.0 (2CH), 127.93 (CH), 127.9 (CH), 127.8 (CH), 127.7 (CH), 127.65 (CH), 127.64 (CH), 127.62 (CH), 127.5 (CH), 127.4 (C), 127.3 (C), 123.8 (CH), 120.6 (CH), 117.9 (CH), 114.18 (CH), 114.15 (CH), 111.1 (CH), 109.9 (C), 72.0 (CH2), 69.0 (CH2), 68.2 (CH2), 66.0 (CH2), 65.8 (CH2), 59.7 (CH2), 54.2 (CH), 53.8 (CH), 52.9 (CH), 52.1 (CH), 51.1 (CH), 44.0 (CH2), 40.2 (CH2), 38.2 (CH2), 36.7 (CH2), 35.1 (CH2), 28.8 (CH2), 28.7 (CH2), 27.7 (CH), 27.3 (CH2), 26.5 (CH2), 25.1 (CH2), 25.0 (CH2), 23.7 (CH), 22.8 (CH3), 22.4 (2CH3), 21.4 (CH3). HRESDVIS calcd for CgiH^NcAs ([M+H]+): 1402.7128; found 1402.7125. Synthesis of tauramamide (12) (CH3)2CH2(CH2)5CONH (CH3 )2CH2 (CH2 )5CONH D -y{Bn) H2 , 10% Pd/C D - T X r Ser(Bn) 20 atm  S e r BnO-Arg(Cbz)2-Trp-D-Leu HO-Arg-Trp-D-Leu 32 12 A solution of (CH3)2CH2(CH2)5CONH-DTyr(Bn)-Ser(Bn)-DLeu-Trp-Arg(Cbz)2-OBn (32) (0.21 g, 0.15 mmol) and 10 % Pd/C, wet Degussa type EIOINEAV; was stirred in the presence H2 at 25 °C and 20 atm, until TLC (RP, 90% CH3CN/H20) showed the absence of starting material (around 1 week). Catalyst filtration, followed by solvent evaporation and 53 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin reverse phase column chromatography (90% CH3CN/H2O); afforded a colorless solid residue. Further purification by reverse phase HPLC (50% CH3CN/H20, 0.1% TFA) yielded pure tauramamide (12) (0.132 g, 81%). [oc]25D -52 (c 0.9, MeOH). For a summary of !H and 13C NMR assignments, see Tables 2.1 and 2.5. !H NMR (DMSO, 600 MHz) £12.69 (IH, s, broad), 10.76 (IH, d, J = 1.5 Hz), 9.14 (IH, s), 8.32 (IH, d, J = 7.8 Hz), 8.14 (IH, d, J= 8.3 Hz), 8.03 (IH, d, J = 7.8 Hz), 7.99 (IH, d, J = 7.7 Hz), 7.75 (IH, d, J = 7.8 Hz), 7.62 (IH, d, J = 7.8 Hz), 7.48 (IH, dd, J = 5.8, 5.8 Hz), 7.30 (IH, d, J = 8.0 Hz), 7.10 (IH, d, J = 1.9 Hz), 7.03 (IH, dd, J = 7.1, 7.1 Hz), 7.02 (IH, d, J = 8.3 Hz), 7.02 (IH, d, / = 8.3 Hz), 6.95 (IH, dd, J = 7.4, 7.4 Hz), 6.61 (IH, d, J = 8.4 Hz), 6.61 (IH, d, J = 8.4 Hz), 4.82 (IH, s, broad), 4.54 (IH, m), 4.42 (IH, m), 4.23 (IH, m), 4.22 (IH, m), 4.21 (IH, m), 3.51 (IH, m), 3.44 (IH, m), 3.16 (IH, m), 3.12 (2H, m), 2.88 (IH, m), 2.86 (IH, m), 2.62 (IH, dd, / = 10.2, 14.0 Hz), 2.00 (2H, m), 1.79 (IH, m), 1.64 (IH, m), 1.54 (IH, m), 1.45 (IH, m), 1.35 (2H, m), 1.35 (IH, m), 1.24 (IH, m), 1.17 (IH, m), 1.15 (IH, m), 1.14 (IH, m), 1.09 (2H, m), 1.08 (IH, m), 1.07 (IH, m), 1.06 (IH, m), 0.826 (3H, d, / = 6.8 Hz), 0.823 (3H, d, J = 6.8 Hz), 0.68 (3H, d, / = 6.8 Hz), 0.66 (3H, d, J = 6.5 Hz); 13C NMR (CDCI3, 150 MHz) 8 173.3 (C), 172.5 (C), 171.8 (C), 171.7 (C), 171.6 (C), 169.7 (C), 156.7 (C), 155.7 (C), 136.0 (C), 130.0 (2CH), 127.8 (C), 127.1 (C), 124.0 (CH), 120.7 (CH), 118.5 (CH), 118.0 (CH), 114.7 (2CH), 111.2 (CH), 109.9 (C), 61.7 (CH2), 55.0 (CH), 54.5 (CH), 53.1 (CH), 51.6 (CH), 51.2 (CH), 40.7 (CH2), 40.3 (CH2), 38.2 (CH2), 36.6 (CH), 35.1 (CH2), 28.7 (CH2), 28.0 (CH2), 27.6 (CH2), 27.4 (CH2), 26.5(CH2), 25.2 (CH2), 25.1 (CH2), 23.7 (CH), 22.8 (CH3), 22.51 (CH3), 22.49 (CH3), 21.5 (CH3). HRESIMS calcd for C^HeeNgOs ([M+Na]+): 864.4984; found 864.4981. 54 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.5. NMR data for tauramamide (12) (recorded in DMSO-d6)- \ = < 3 ' OH c6 NH H N T ^ N H 2 Amino Proton Acid No Arg 10H 2 2NH 3 4 5 5NH Trp 2 2NH 3 4' 5' 6' T 7a'NH 2' D-Leu 2 2NH 3 4 5 5' Ser 2 2NH 3 30H "Recorded at 600 MHz. OH *H COSY" 5 (ppra) (mult, / (Hz))bc (H-»H) 12.69 (s, broad) 4.22 (m) 2NH, H3 8.32 (d, 7 = 7.8 Hz) H2 1.64 (m), 1.79 (m) H2, H4 1.35 (m), 1.54 (m) H3, H5 3.12 (m) 4H,5NH 7.48 (dd, 7 = 5.8,5.8 Hz) H5 4.54 (m) 2NH, H3 8.14 (d, 7=8.3 Hz) H2 2.88 (m), 3.16 (m) H2 7.62 (d, 7 = 7.8 Hz) H5' 6.95 (dd, 7 = 7.4, 7.4 Hz) H4' 7.03 (dd, 7 = 7.1, 7.1 Hz) H7' 7.30 (d, 7 =8.0 Hz) H6' 10.76 (d, 7 = 1.5 Hz) H2' 7.10 (d, 7 = 1.9 Hz) 7a'NH 4.21 (m) 2NH, H3 7.75 (d, 7 = 7.8 Hz) H2 1.08 (m), 1.17 (m) H2 1.24 (m) H5,H5' 0.68 (d, 7 = 6.8 Hz) H4 0.66 (d, 7 = 6.5 Hz) H4 4.23 (m) 2NH, H3 8.03 (d, J = 7.8 Hz) H2 3.44 (m), 3.51 (m) H2, 30H 4.82 (s, broad) H3 55 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.5. NMR data for tauramamide (12) (recorded in DMSO-Jg) (Continuation). HO' ^ = < 3 ' OH H N ^ NH 2 Amino Proton Acid No D-Tyr 2 2NH 3 T y 4'OH 5' 6' Acyl 2 chain 3 4 5 6 7 8 8' "Recorded at 600 MHz. *H COSY" 6 (ppm) (mult, J (Hz))b'c (H->H) 4.42 (m) 2NH, H3 7.99 (d, 7 = 7.7 Hz) H2 2.62 (dd, J = 10.2, 14.0 Hz), H2 2.86 (m) 7.02 (d, 7 = 8.3 Hz) H3' 6.61 (d, 7 =8.4 Hz) H2' 9.14 (s) 6.61 (d, 7 =8.4 Hz) H6' 7.02 (d, 7 = 8.3 Hz) H5' 2.00 (m) H3 1.35 (m) H2,H4 1.06 (m), 1.14 (m) H3 1.07 (m), 1.15 (m) 1.09 (m) H7 1.45 (m) H6,H8,H8' 0.826 (d, 7 = 6.8 Hz) H7 0.823 (d, 7 = 6.8 Hz) H7 Preparation of tauramamide ethyl ester (14) (CH3)2CH2(CH2)5CONH (CH3)2CH2(CH2)5CONH D Tv r HCI D T r Ser _ i t ° H " S.er HO-Arg-Trp-D-Leu CH3CH20-Arg-Trp-D-Leu 12 14 56 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin A solution of tauramamide (12) (0.0628 g, 0.073 mmol) and HC1 (1 mL HC1 2M in diethyl ether, 2 mmol) in ethanol was stirred at 25 °C until the TLC spot for starting material (RP, 90% CH3CN/H20) disappeared (around 5 days). Solvent evaporation, followed by reverse phase HPLC (50% CH3CN/H20, 0.1% TFA) yielded pure tauramamide ethyl ester (14) (0.0559 g, 86%). [a]25D -4.7 (c 1.8, MeOH). For a summary of !H and 13C NMR assignments, see Table 2.6. *H NMR (DMSO, 600 MHz) £10.77 (1H, d, 7 = 1.6 Hz), 9.14 (1H, s), 8.43 (1H, d, 7 = 7.4 Hz), 8.16 (1H, d, 7= 8.3 Hz), 8.04 (1H, d, 7= 7.7 Hz), 7.98 (1H, d, 7 = 7.7 Hz), 7.75 (1H, d, 7 = 7.7 Hz), 7.62 (1H, d, 7 = 7.7 Hz), 7.48 (1H, dd, 7 = 5.5, 5.6 Hz), 7.30 (1H, d, 7 = 8.0 Hz), 7.10 (1H, d, 7 = 2.2 Hz), 7.04 (1H, dd, 7 = 7.2, 7.5 Hz), 7.02 (1H, d, 7 = 8.6 Hz), 7.02 (1H, d, 7 = 8.6 Hz), 6.96 (1H, dd, 7 = 7.2, 7.5 Hz), 6.61 (1H, d, 7 = 8.6 Hz), 6.61 (1H, d, 7 = 8.6 Hz), 4.84 (1H, dd, 7 = 5.2, 5.5 Hz), 4.54 (1H, m), 4.42 (1H, m), 4.25 (1H, m), 4.24 (1H, m), 4.23 (1H, m), 4.09 (2H, m), 3.51 (1H, m), 3.45 (1H, m), 3.15 (1H, dd,7= 3.8, 14.6 Hz), 3.11 (2H, m), 2.88 (1H, m), 2.84 (1H, m), 2.62 (1H, dd, 7 = 10.0, 13.8 Hz), 2.00 (2H, m), 1.77 (1H, m), 1.66 (1H, m), 1.53 (2H, m), 1.45 (1H, m), 1.35 (2H, m), 1.26 (1H, m), 1.18 (3H, t, 7 = 7.2 Hz), 1.14 (2H, m), 1.08 (2H, m), 1.08 (2H, m), 1.06 (2H, m), 0.824 (3H, d, 7 = 6.7 Hz), 0.822 (3H, d, 7 = 6.6 Hz), 0.70 (3H, d, 7 = 6.6 Hz), 0.67 (3H, d, 7 = 6.6 Hz); 13C NMR (CDC13, 150 MHz) S 172.4 (C), 172.0 (C), 171.72 (C), 171.66 (C), 171.64 (C), 169.7 (C), 156.6 (C), 155.7 (C), 136.0 (C), 130.0 (2CH), 127.8 (C), 127.1 (C), 123.9 (CH), 120.7 (CH), 118.4 (CH), 118.0 (CH), 114.7 (2CH), 111.2 (CH), 109.8 (C), 61.7 (CH2), 60.5 (CH2), 55.0 (CH), 54.5 (CH), 53.1 (CH), 51.8 (CH), 51.2 (CH), 40.7 (CH2), 40.3 (CH2), 38.2 (CH2), 36.6 (CH), 35.1 (CH2), 28.8 (CH2), 27.8 (CH2), 27.6 (CH2), 27.4 (CH2), 26.5 (CH2), 25.15 (CH2), 25.08 (CH2), 23.7 (CH), 22.8 (CH3), 22.50 (CH3), 22.49 (CH3), 21.5 (CH3), 14.0 (CH3). HRESEVIS calcd for C46H70N9O9 ([M+Na]+): 892.5297; found 892.5295. 57 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.6. 'H and 13C NMR data for natural69 and synthetic tauramamide ethyl ester (14) (recorded in DMSO-cfc). Amino Acid Ethyl Arg Tip D-Leu No 1 2 1 2 2NH 3 4 5 5NH 6 1 2 2NH 3 3' 3a' 4' 5' 6' T 7a' 7a'NH 2* 1 2 2NH 3 4 2^<T \ = < 3 ' OH o  H \ o H 3 / 0 H A O H 3\ O H \3 8 J 6 HNT^NH2 13C 5 (ppm)' 60.5 14.1 172.0 51.8 27.8 25.1 40.3 156.6 - 53.1 27.6 109.9 127.1 118.4 118.1 120.7 111.2 136.0 123.9 - 51.2 40.5 23.7 Natural 'H ' 5(ppm)(mult,7(Hz))a 4.09 (m) 1.18 (t, 7 = 7.2 Hz) 4.25 (s, broad) 8.43 (d, 7 = 7.7 Hz) 1.67 (m) 1.77 (m) 1.53 (m) 3.11 (m) 7.49 (lH.m) 4.54 (s, broad) 8.17 (d, 7= 8.2 Hz) 2.88 (m) 3.16 (m) 7.62 (d, 7 = 7.7 Hz) 6.96 (dd, 7 = 7.2, 7.7 Hz) 7.04 (m) 7.30 (d, 7 = 8.2 Hz) 10.77 (s) 7.11 (d, 7= 1.5 Hz) 4.23 (s, broad) 7.76 (d, 7 =8.2 Hz) 1.08 (m) 1.23 (m) Recorded at 150 MHz.b Recorded at 600 MHz. 3 ^ O H 13C 8(ppm)a 60.5 14.0 172.0 51.8 27.8 25.08 40.3 156.6 171.72 53.1 27.6 109.8 127.1 118.4 118.0 120.7 111.2 136.0 123.9 171.64 51.2 40.7 23.7 ^ ^ ^ _ ^ 8 8' Synthetic 'H 8 (ppm) (mult, 7 (Hz))' 4.09 (m) 1.18 (t, 7 = 7.2 Hz) 4.25 (m) 8.43 (d, 7 = 7.4 Hz) 1.66 (m) 1.77 (m) 1.53 (m) 3.11 (m) 7.48 (dd, 7 = 5.5, 5.6 Hz) 4.54 (m) 8.16(d,7 = 8.3Hz) 2.88 (m) 3.15 (dd, 7 = 3.8, 14.6 Hz) 7.62 (d, 7 = 7.7 Hz) 6.96 (dd, 7 = 7.2, 7.5 Hz) 7.04 (dd, 7 = 7.2, 7.5 Hz) 7.30 (d, 7 =8.0 Hz) 10.77 (d, 7 =1.6 Hz) 7.10 (d, 7 =2.2 Hz) 4.23 (m) 7.75 (d, 7 = 7.7 Hz) 1.08 (m) 1.26 (m) 58 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.6. *H and 13C NMR data for natural69 and synthetic tauramamide ethyl ester (14) (recorded in DMSO-rfg) (Continuation). Amino Acid D-Leu Ser D-Tyr Acyl chain No 5 5' 1 2 2NH 3 30H 1 2 2NH 3 r 2' 3' 4' 4'OH 5' 6' 1 2 3 4 5 6 7 8 8' V = < 3 ' OH o  H i o H 3/ o H A O H 3 \ O H \3 o NH J6 H N ^ ^ N H 2 v_° 13 C 5 (ppm)a 22.8 21.5 169.7 55.0 61.7 54.5 36.6 127.9 130.0 114.7 155.7 114.7 130.0 35.1 25.2 28.8 26.5 38.3 27.4 22.5 22.5 Natural 8(ppm)(mult,7(Hz))a 0.70 (d, 7 = 6.6 Hz) 0.67 (d, 7 = 6.7 Hz) 4.24 (s, broad) 8.05(d,/ = 7.7Hz) 3.46 (m) 3.51 (m) 4.84 (t, 7 = 5.4 Hz) 4.42 (s, broad) 7.99 (d, J = 7.7 Hz) 2.63 (m) 2.84 (m) 7.02 (d, J = 8.2 Hz) 6.61 (d, 7 = 8.2 Hz) 9.14 (s) 6.61 (d, 7 = 8.2 Hz) 7.02 (d, 7 = 8.2 Hz) 2.00 (m) 1.35 (m) 1.06 (m) 1.14 (m) 1.08 (m) 1.45 (m) 0.83 (d, 7 = 6.7 Hz) 0.83 (d, 7 = 6.7 Hz) OH Synthetic 1 3 C 8 (ppm)a 22.8 21.5 169.7 55.0 61.7 171.67 54.5 36.6 127.8 130.0 114.7 155.7 114.7 130.0 172.4 35.1 25.15 28.8 26.5 38.2 27.4 22.49 22.50 'H 8(ppm)(mult,/(Hz))a 0.70 (d, 7 = 6.6 Hz) 0.67 (d, 7 = 6.6 Hz) 4.24 (m) 8.04 (d, 7 = 7.7 Hz) 3.45 (m) 3.51 (m) 4.84 (dd, 7 = 5.2,5.5 Hz) 4.42 (m) 7.98 (d, 7 = 7.7 Hz) 2.62 (dd, 7= 10.0,13.8 Hz) 2.84 (m) 7.02 (d, 7 = 8.6 Hz) 6.61 (d, 7 = 8.6 Hz) 9.14 (s) 6.61 (d, 7 = 8.6 Hz) 7.02 (d, 7 = 8.6 Hz) 2.00 (m) 1.35 (m) 1.06 (m) 1.14 (m) 1.08 (m) 1.45 (m) 0.822 (d, 7 = 6.6 Hz) 0.824 (d, 7 = 6.7 Hz) •Recorded at 150 MHz.D Recorded at 600 MHz. 59 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin H = N H = || H 0 " v O M \ o ULLIU IL 8.0 7.5 7.0 6.5 3.5 3.0 2.5 2.0 1.5 1.0 il ; ,i V ' ^ --~*J _J^'J»U-. 'U> JflJJlw 11.010.5 10.0 9.5 9.0 8.5 8,0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1,5 1.0 0.5 0 Chemical Shift (ppm) Figure 2.10. 'H-NMR spectrum of tauramamide ethyl ester (14) (recorded in DMSO-^6 at 600 MHz). 60 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin ° H H H J o \ o " \ o NH HN^*~NH2 h^^f^^^ 172 171 170 40 35 30 25 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Chemical Shift (ppm) 13, Figure 2.11. C-NMR spectrum of tauramamide ethyl ester (14) (recorded in DMSO-Jg at 150 MHz). 61 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 2.6.2. Synthesis of dealkylsurfactin Preparation of (7?)-(-)-3-hydroxybutyric phenacyl ester (34) OH Q " ~T7 ^ 1 OH O EtOH/H 20, reflux 6 34 To a solution of (7?)-(-)-3-hydroxybutyric acid sodium salt (0.76 g, 6.0 mmol) in EtOH/H20 (H2O added drop by drop until complete solubility was achieved, total volume: 18 mL), was added phenacyl bromide (1.20 g, 6.0 mmol) and the mixture refluxed for 4 h. The reaction crude was concentrated in vacuo and the resulting residue purified by silica gel column chromatography (50% EtOAc/hexanes), to afford (34) as a yellowish solid (1.10 g, 82%). !H NMR (CD3OD, 400 MHz) £7.97 (2H, d, J = 7.3 Hz), 7.65 (1H, dd, J = 7.3, 7.6 Hz), 7.52 (2H, dd, J = 7.9, 7.6 Hz), 5.45 (2H, m), 4.23 (1H, m), 2.60 (2H, m), 1.26 (3H, d, J = 6.1 Hz); 13C NMR (CD3OD, 100 MHz) £194.6 (C), 172.5 (C), 135.5 (C), 135.1 (CH), 130.0 (2CH), 128.9 (2CH), 67.4 (CH2), 65.6 (CH), 44.5 (CH2), 23.2 (CH3). HRESIMS calcd for Ci2Hi404Na ([M+Na]+): 245.0790; found 245.0784. 62 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Preparation of Boc-Glu(Bn)-Leu-DLeu-Val-Asp(Bn)-DLeu-Leu-OCH(CH^CH7COOCH2COPh(46) 34 ° 40 O 1.TFA 2. Boc-X-OH X: D-Leu, Asp(Bn), Val, D-Leu, Leu, Glu(Bn) y 3. PyBOP, DIEA = O Boc-Glu(Bn)-Leu-D-Leu-Val-Asp(Bn)-D-Leu-Leu- o 46 To a mixture of Boc-Leu-OH (0.91 g, 3.9 mmol, 1 eq) and (i?)-(-)-3-hydroxybutyric phenacyl ester (34) (0.79 g, 3.6 mmol) in CH2C12 (25 mL) at 0 °C, was added DCC (0.80 g, 3.9 mmol) dissolved in CH2CI2 (5 mL). The cooling bath was removed and the mixture was stirred overnight. Purification by silica gel column chromatography (50% EtOAc/hexanes) afforded (40) as a crystalline amorphous powder (1.39 g, 90%). Intermediate (40) (0.39 g, 0.89 mmol) was treated with TFA (1 mL, 12.5 mmol, 14 eq) and stirred for 5 h, whereupon it was concentrated in vacuo to a pale red oil. This oil was dissolved in CH2CI2 (10 mL) and added to a solution of Boc-DLeu-OH (0.19 g, 0.81 mmol, 0.90 eq) and PyBOP (0.51 g, 0.98 mmol, 1.10 eq) in CH2C12 (20 mL); followed by addition of DIEA (0.46 mL, 2.70 mmol, 3 eq). After overnight stirring at 25 °C, NH4CI sat. was added and CH2CI2 (3x20 mL) extractions performed. The organic extracts were combined and dried (Na2S04), to be then concentrated in vacuo. Column chromatography on silica (70% EtOAc/Hexanes) afforded desired product as a white amorphous solid (0.45 g, 99%). The procedure above was repeated successively with Boc-Asp(Bn)-OH (0.25 g, 0.77 mmol), Boc-Val-OH (0.16 g, 0.70 mmol), Boc-D-Leu-OH (0.15 g, 0.62 mmol), Boc-Leu-OH (0.16 g, 0.70 mmol) and Boc-Glu-OH (0.16 g, 0.49 mmol); to afford product (46) (0.63 g, overall yield for 5 steps: 63%). !H NMR (CDC13, 400 MHz) £8.36 (1H, d, J = 8.0 Hz), 63 - O Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 7.82 (2H, d, J = 7.3 Hz), 7.57 (IH, d, J = 9.5 Hz), 7.53 (IH, dd, J = 7.3, 7.6 Hz), 7.39 (2H, dd, J = 7.6, 7.9 Hz), 7.32-7.21 (10H, m), 7.07 (IH, d, J = 7.9 Hz), 6.99 (IH, s), 5.97 (IH, d, J = 7.3 Hz), 5.37 (IH, d, J= 16.8 Hz), 5.27 (IH, d, J= 16.5 Hz), 5.23 (IH, s), 5.07 (IH, d, J= 12.5 Hz), 5.04 (3H, m), 4.89 (IH, d, J= 12.5 Hz), 4.54 (IH, m), 4.46 (IH, m), 4.42 (IH, m), 4.19 (IH, m), 3.97 (IH, m), 3.81 (IH, dd, J = 5.2, 6.4 Hz), 3.08 (IH, dd, J = 3.7, 15.3 Hz), 2.93 (IH, dd, / = 10.1, 15.3 Hz), 2.86 (IH, dd, J = 7.3, 15.6 Hz), 2.57 (IH, dd, J = 6.1, 15.6 Hz), 2.54 (IH, m), 2.46 (IH, m), 1.90 (2H, m), 1.78 (2H, m), 1.70-1.45 (13H, m), 1.30 (9H, s), 1.23 (3H, d, J = 6.1 Hz), 0.95-0.75 (30H, m); 13C NMR (CDC13, 100 MHz) £192.2 (C), 173.7 (C), 173.5 (C), 173.2 (C), 173.1 (C), 172.3 (C), 172.1 (C), 171.7 (C), 171.3 (C), 170.2 (C), 169.4 (C), 155.5 (C), 135.5 (C), 135.2 (C), 134.0 (CH), 133.8 (C), 128.8 (2CH), 128.4 (2CH), 128.3 (2CH), 128.2 (CH), 128.1 (2CH), 128.0 (CH), 127.9 (2CH), 127.6 (2CH), 79.5 (C), 68.4 (CH), 66.4 (CH2), 66.4 (CH2), 66.1 (CH2), 61.8 (CH), 53.2 (CH), 52.8 (CH), 52.0 (CH), 50.7 (CH), 50.1 (CH), 49.2 (CH), 41.0 (CH2), 40.4 (CH2), 40.1 (CH2), 40.0 (CH2), 37.3 (CH2), 35.7 (CH2), 30.2 (CH2), 29.1 (CH2), 28.1 (3CH3), 24.6 (CH), 24.6 (CH), 24.5 (CH), 24.4 (CH), 22.9 (CH3), 22.8 (CH3), 22.7 (CH3), 22.6 (CH3), 22.0 (CH3), 21.7 (CH3), 21.5 (CH3), 21.3 (CH3), 19.3 (CH3), 19.2 (CH3), 18.5 (CH3). HRESIMS calcd for CegB^NvOnNa ([M+Na]+): 1320.6995; found 1320.6971. Preparation of dibenzyl dealkylsurfactin (47) 7 M 1.Zn/HOAc ^ 7 ™_i PyBOP ^ 7 T Leu O • Leu OH '- ^ - Leu Glu(Bn) D-Leu Boc 2- T F A D-Leu Glu(Bn) D I E A D-Leu D-Leu Asp(Bn)—Val—D-Leu-Glu(Bn) Asp(Bn)—Val—D-Leu Asp(Bn) Val 46 47 To a solution of (46) (0.30 g, 0.23 mmol) in HOAc (8 mL), was added Zn (2 g, 30 mmol) and the mixture was stirred overnight. After filtration through a reverse phase Sep Pak® (2 g) and 64 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin solvent evaporation, resulting residue was dissolved in EtOAc and washed with water, followed by a second wash with NaCl sat. The organic extracts were dried (Na2SC>4) and concentrated in vacuo. The resulting residue was dissolved in CH2CI2 (10 mL) and stirred in the presence of TFA (1 mL) for 24 h. A second Sep Pak® filtration and solvent evaporation allowed to isolate the desired deprotected intermediate, which was dissolved in CH2CI2 (20 mL) and reacted with PyBOP (0.056 g, 1.1 mmol) in the presence of DIEA (0.05 mL, 0.29 mmol). Overnight stirring followed by LH-20 (70% EtOAc/hexanes) and silica gel (gradient from 70% EtOAc/hexanes to 10% MeOH/EtOAc) column chromatography afforded (47) as a colorless solid (0.074 g, 30%). [a]20D -18 (c 1.9, MeOH); UV (MeOH) Xmax (log e) 226 nm (2.80). *H NMR (DMSO, 600 MHz) £8.54 (1H, d, 7 = 6.5 Hz), 8.34 (1H, d, 7 = 7.0 Hz), 8.17 (1H, d, 7 = 7.5 Hz), 8.11 (1H, d, 7 = 4.8 Hz), 7.90 (1H, d, 7= 8.9 Hz), 7.82 (1H, d, 7 = 8.7 Hz), 7.78 (1H, d, 7 = 6.3 Hz), 7.40-7.29 (10H, m), 5.08 (4H, m), 5.02 (1H, dd, 7 = 6.3, 7.0 Hz), 4.63 (1H, m), 4.48 (1H, dd, 7 = 6.5, 8.6 Hz), 4.23 (1H, dd, 7 = 5.9, 6.4 Hz), 4.14 (1H, dd, 7= 10.9, 14.9 Hz), 4.10 (1H, dd, 7= 12.6, 12.8 Hz), 4.08 (1H, m), 4.01 (1H, m), 2.84 (1H, dd, 7 = 4.3, 16.7 Hz), 2.72 (1H, dd, 7 = 9.8, 16.7 Hz), 2.38 (4H, m), 2.08-1.82 (3H, m), 1.68-1.32 (12H, m), 1.11 (3H, d, 7 = 6.2 Hz), 0.90-0.80 (24H, m), 0.78 (3H, d, 7 = 5.9 Hz), 0.74 (3H, d, 7 = 6.8 Hz); 13C NMR (DMSO, 150 MHz) S 172.4 (C), 172.3 (C), 172.1 (C), 171.9 (C), 1715 (C), 170.8 (C), 170.5 (C), 169.9 (C), 169.4 (C), 169.3 (C), 136.1 (C), 135.9 (C), 128.4 (2CH), 128.3 (2CH), 128.0 (CH), 127.9 (CH), 127.8 (2CH), 127.7 (2CH), 68.5 (CH), 65.6 (CH2), 65.4 (CH2), 58.2 (CH), 52.5 (CH), 51.7 (2CH), 51.4 (CH), 50.2 (CH), 49.3 (CH), 41.9 (CH2), 38.6 (CH2), 36.0 (CH2), 30.5 (CH), 29.5 (CH2), 26.9 (CH2), 24.2 (CH), 24.1 (3CH), 23.1 (CH3), 22.8 (CH3), 22.7 (CH3), 22.4 (CH3), 22.1 (CH3), 21.9 (CH3), 21.0 (CH3), 20.8 (CH3), 19.5 (CH3), 19.1 (CH3), 18.3 (CH3). HRESIMS calcd for CseH^NvOnNa ([M+Na]+): 1084.5947; found 1084.5957. 65 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Synthesis of dealkylsurfactin (48) = o = o ep^^Np cp-^^Sp ^.eu Cplu(Bn) H2 , Pd / C lj.eu <p.lu D-Leu D-Leu HOAc D-Leu D-Leu Asp(Bn)—Val Asp Val 47 48 Dibenzyl dealkylsurfactin (47) (0.030 g, 0.028 mmol) was dissolved in HOAc (5 mL), and stirred overnight in the presence of Pd/C under H2 atmosphere using a balloon. Filtration followed by solvent evaporation afforded a dark residue, which was purified by size exclusion column chromatography (100% EtOAc). A colorless residue, identified as dealkylsurfactin (0.025 g, 99%) was obtained. [a]20D -22 (c 3.8, MeOH); UV (MeOH) ^ (log e) 222 nm (2.85). For a summary of 'H and 13C NMR assignments, see Tables 2.3 and 2.7. !H NMR (DMSO, 600 MHz) 812.28 (IH, s, broad), 12.17 (IH, s, broad), 8.56 (IH, d, J = 6.6 Hz), 8.38 (IH, d, J = 5.6 Hz), 8.15 (IH, d, J = 6.9 Hz), 8.14 (IH, d, J = 6.9 Hz), 7.86 (IH, d, J = 9.1 Hz), 7.84 (IH, d, J = 9.5 Hz), 7.79 (IH, d, J = 6.3 Hz), 5.01 (IH, m), 4.53 (IH, m), 4.48 (IH, m), 4.21 (IH, m), 4.13 (IH, m), 4.09 (IH, dd, J = 8.6, 8.9 Hz), 4.06 (IH, m), 4.00 (IH, m), 2.68 (IH, dd, / = 4.3, 16.7 Hz), 2.56 (IH, dd, J = 9.5, 17.1 Hz), 2.42 (IH, dd, J = 7.9, 13.8 Hz), 2.37 (IH, dd, J = 5.9, 13.8 Hz), 2.20 (2H, m), 1.97 (IH, m), 1.96 (IH, m), 1.81 (IH, m), 1.63 (IH, m), 1.61 (IH, m), 1.58 (IH, m), 1.54 (IH, m), 1.48 (IH, m), 1.46 (2H, m), 1.46 (IH, m), 1.46 (IH, m), 1.45 (IH, m), 1.40 (IH, m), 1.37 (IH, m), 1.11 (3H, d, J = 5.9 Hz), 0.89 (3H, d, J = 6.9 Hz), 0.88 (3H, d, J = 6.3 Hz), 0.87 (3H, d, J = 6.9 Hz), 0.87 (3H, d, J = 6.9 Hz), 0.85 (3H, d, J = 6.5 Hz), 0.84 (3H, d, J = 6.2 Hz), 0.82 (3H, d, J = 6.6 Hz), 0.82 (3H, d, J = 6.5 Hz), 0.78 (3H, d, J = 6.2 Hz), 0.75 (3H, d, J = 6.6 Hz); 13C NMR (DMSO, 150 MHz) £174.1 (C), 172.4 (C), 172.1 (C), 172.0 (C), 171.6 (C), 171.6 (C), 171.0 (C), 170.5 (C), 169.8 (C), 169.4 (C), 68.6 (CH), 58.2 (CH), 52.6 (CH), 51.8 (CH), 51.6 (CH), 51.4 (CH), 50.2 (CH), 49.5 (CH), 41.9 (CH2), 41.9 (CH2), 39.4 66 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin (CH2), 39.2 (CH2), 38.6 (CH2), 36.2 (CH2), 30.6 (CH), 29.6 (CH2), 27.2 (CH2), 24.2 (CH), 24.2 (CH), 24.2 (CH), 24.2 (CH), 23.2 (CH3), 22.8 (CH3), 22.8 (CH3), 22.4 (CH3), 22.2 (CH3), 21.9 (CH3), 21.0 (CH3), 20.8 (CH3), 19.5 (CH3), 19.2 (CH3), 18.4 (CH3). HRESIMS calcd for C42H7iN7Oi3Na ([M+Na]+): 904.5008; found 904.5015. Table 2.7. NMR data for dealkylsurfactin (48) (recorded in DMSO-d6). 5 CU^NH i " ^ > - ' ^ N H  H H HNT^O 5 Q ^ 3 O W o ^ L , OH 4' 4 5 Amino Acid (/?)-3-HBA Leu D-Leu Asp Proton No 2 3 4 2 2NH 3 4 5 5' 2 2NH 3 4 5 5' 2 2NH 3 Recorded at 600 MHz. !H 8(ppm)(mult,./(Hz))a 2.37 (dd, 7 = 5.9, 13.8 Hz) 2.42 (dd, 7=7.9 , 13.8 Hz) 5.01 (m) 1.11 (d, 7 = 5.9 Hz) 4.00 (m) 8.56 (d, 7 = 6.6 Hz) 1.45 (m), 1.61 (m) 1.63 (m) 0.82 (d, 7 = 6.4 Hz) 0.88 (d, 7 = 6.3 Hz) 4.48 (m) 7.84 (d, 7 = 9.5 Hz) 1.37 (m), 1.40 (m) 1.46 (m) 0.87 (d, 7 = 6.9 Hz) 0.87 (d, 7 = 6.9 Hz) 4.53 (m) 8.15 (d, 7 = 6.9 Hz) 2.56 (dd, 7 = 9.5,17.1Hz) 2.68 (dd, 7 = 4.3,16.7 Hz) TOCSY* (H->H) H3,H4 H2.H4 H2, H3 H3, H5, 2NH H2,H3 H2, 2NH H5, H5' H2.H4 H4 H3, H4, 2NH H2,H3 H2.2NH H2, H5, H5' H4 H4 H3.2NH H2,H3 H2.2NH 67 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin Table 2.7. NMR data for dealkylsurfactin (48) (recorded in DMSO-J6) (Continuation). ,OH o = o Amino Acid Val D-Leu Leu Glu 5 Proton No 2 2NH 3 4 4' 2 2NH 3 4 5 5' 2 2NH 3 4 5 5' 2 2NH 3 4 corded at 600 MHz. 3 ^ Y ^ C T - ^ ^ N ' 2 5 Q ^ N H H ' ^ ^ ° NH o i H H 1 Q HNT^O 5 OH 4' >H 8(ppm)(mult,/(Hz))a 4.09 (dd, 7 =8.6, 8.9 Hz) 7.86 (d, 7 = 9.1 Hz) 1.97 (m) 0.89 (d, 7 = 6.9 Hz) 0.75 (d, 7 = 6.6 Hz) 4.13 (m) 8.38 (d, 7 = 5.6 Hz) 1.46 (m), 1.54 (m) 1.58 (m) 0.82 (d, 7 = 6.6 Hz) 0.78 (d, 7 = 6.2 Hz) 4.06 (m) 8.14 (d, 7 = 6.9 Hz) 1.46 (m) 1.48 (m) 0.85 (d, 7 = 6.5 Hz) 0.84 (d, 7 = 6.2 Hz) 4.21 (m) 7.79 (d, 7 = 6.3 Hz) 1.81 (m), 1.96 (m) 2.20 (m) TOCSY8 (H->H) H3, H4, H4\ 2NH H2, H3, H4, H4' H2, H4, H4\ 2NH H2, H3, 2NH H2, H3, 2NH H3,2NH H2,H3 H2, 2NH H5, H5' H4 H4 H4, 2NH H2,H3 2NH H2, H5, H5' H4 H4 H3, H4, 2NH H2, H3, H4 H2, H4, 2NH H2, H3, 2NH 68 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin O^ OBn 0 = 0 HN ^ O XX j l 8.50 8.25 8.007.75 JdU'U1 U J yjUdjilA 5.0 4.5 4.0 H V j W W > 9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Chemical Shift (ppm) Figure 2.12. 'H-NMR spectrum of dibenzyl dealkylsurfactin (47) (recorded in DMSO-rf6 at 600 MHz). 69 Chapter 2. Total Synthesis of Tauramamide and Dealkylsurfactin 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Chemical Shift (ppm) Figure 2.13.13C APT-NMR spectrum of dibenzyl dealkylsurfactin (47) (recorded in DMSO-d6 at 150 MHz). 70 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3.1. IDO inhibition and Garveia annulata Immune escape plays an important role in cancer progression132"135 and fetal development.136 Although the mechanism is still not completely understood, it has been proposed that indoleamine 2,3-dioxygenase (IDO) contributes to evasion of T-cell-mediated immune rejection. IDO catalyzes the oxidative cleavage of the 2,3 bond of tryptophan. ' T-cell lymphocytes are extremely sensitive to tryptophan shortage, which causes them to undergo cell cycle arrest in Gl. Degradation of tryptophan via IDO expressed by tumors or by the placenta inhibits T-cell proliferation and, as a result, prevents immunological rejection of the tumor or fetus. ' ' In addition, EDO present in the ocular lens has been implicated as a key factor in the development of senile cataracts.139140 Most of the known IDO inhibitors are tryptophan analogues active only at micromolar concentrations, making them marginal drug candidates.137 As part of an ongoing program designed to find potent IDO inhibitors belonging to new structural classes, a library of marine invertebrate extracts was screened for their ability to inhibit purified recombinant human IDO in vitro. A MeOH extract of the Northeastern Pacific hydroid Garveia annulata was among the first hits exhibiting potent enough IDO inhibitory activity to justify further investigation. Details of the isolation and structure elucidation of G annulata metabolites along with a discussion of their biological activity are described below. 71 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3.2. IDO and the kynurenine pathway Indoleamine 2,3-dioxygenase (IDO; EC 1.13.11.42) is a heme-containing intracellular enzyme that catalyses the initial and rate-limiting step in the metabolism of tryptophan, oxidation of tryptophan to Af-formylkynurenine (1), along the kynurenine pathway in mammalian cells (Figure 3.1). 141,142 or H Af-Formylkynurenine 1 OH N COOH Kynurenic acid COOH iH NH 2 ~NH2 Kynurenine 2 OH COOH Xanthurenic acid t OH 3-Hydroxykynurenine 3 oinr - o-Aminohippuric acid — O^0H ^ ^ N H 2 Anthranillic acid co2, 3- H 2 0 O Y ^ N H 2 OH Hydroxyanthranillic acid 1 ^ ^ ^ ~^- NAD Figure 3.1. Tryptophan metabolism in the kynurenine pathway. 141,143 All proposed mechanisms for this reaction initiate with O2 and tryptophan binding (the order is unknown) at the active site (Scheme 3.1).143 The active form of IDO has the heme iron in the ferrous (Fe2+) oxidation state and, although the enzyme is prone to auto-oxidation, the primary catalytic effect does not involve changes in oxidation state (the ferric form Fe3+ is inactive). Formation of intermediate (4) may occur via ionic (shown), pericyclic or radical 72 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata processes. The following step most likely proceeds through a concerted Criegee-type rearrangement to yield a labile cyclic hemiacetal (5), which is released to generate N- formylkynurenine (l).143 Scheme 3.1. Possible mechanism for the IDO-catalyzed formation of iV-formylkynurenine (1). IDO was originally discovered in 1967 in the rabbit intestine. It is found under basal conditions in the epididymis, thymus, gut, lung, placenta, and some subsets of dendritic cells.145"147 Recent studies indicate that IDO plays an important immunosuppressive function.148 T lymphocytes must divide to be activated, but this process is particularly sensitive to availability of tryptophan. Cells expressing IDO can deplete their microenvironment of tryptophan, promoting T lymphocyte arrest in the Gl phase of the cell cycle. Additionally, tryptophan metabolites from the kynurenine pathway have a strong T cell inhibitory action. IDO is highly expressed in the placenta where it plays a significant role in the immunosuppressive and tolerogenic mechanism contributing to maternal tolerance towards the 73 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata allogeneic fetus during pregnancy. 5'149 It has also been implicated in regulation of autoimmune disorders and suppression of transplant rejection.146 Furthermore, IDO was associated with depression as well as other neurological and psychiatric diseases through interference with serotonin production and accumulation of neurotoxic kynurenine metabolites.14 The immunosuppressive function of IDO can be exploited by tumor cells to escape immune detection and achieve tolerance, a hallmark in carcinogenesis and cancer progression.145 IDO has been found in cancer cells in a variety of human malignancies.133'148 EDO is also expressed by host antigen-presenting cells (APC's) at the periphery of tumors, and draining lymph nodes of breast cancer and melanoma, where appropriate T lymphocytes would otherwise be activated.1 ' In patients with malignant melanoma, the presence of these EDO-expressing cells (APC's) in sentinel lymph node biopsies was correlated with a significantly worse clinical outcome.146 In murine models, transfection of immunogenic tumor cell lines with recombinant EDO renders them immunosuppressive and lethally progressive in vivo.14 Expression of EDO by ovarian,151 endometrial142 and colorectal cancer152 cells has been found to be a significant predictor of poor prognosis. Thus, expression of EDO, either by host cells or by tumor cells, seems associated with poor outcome in a number of clinical settings. In addition, IDO is expressed in the ocular lens where it participates in the UV filter synthetic pathway. In 2001, Takikawa and coworkers demonstrated that EDO is the first enzyme in the UV filter biosynthesis.140 Some of the tryptophan degradation products (Scheme 3.1) bind to the lens major protein, crystalline, and accumulate with time in a process linked to the formation of age-related nuclear cataracts.139'154 It is also known that 3-hydroxykynurenine (3) is readily oxidized in the presence of just trace amounts of oxygen, generating mostly orange, 1 OQ red, brown or black pigments. Thus, oxidation of the protein-bound UV filters could possibly 74 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata lead to the range of lens colors that are characteristic of age-related nuclear cataracts. This opacification may be preventable by drug-induced suppression of lens IDO activity.1 From a therapeutic standpoint, drugs that inhibit IDO potentially represent a novel class of immunomodulatory agents, targeting a pathway that might be involved mechanistically in the ability of tumors to evade the host immune system. Their initial application is likely to be as adjuvants, used in combination with existing vaccination, conventional chemotherapy and . . . . 149 radiation strategies. 3.3. Current IDO inhibitors and the bioassay To date, the most potent competitive IDO inhibitors known are tryptophan analogues, active only at concentrations of -10 (J.M and greater (Table 3.1).134 Clearly, more potent and effective IDO inhibitors are required as lead compounds for drug therapy. The most widely studied IDO inhibitor is 1-methyltryptophan (6),132 which has been used to demonstrate proof-of- principle in cancer therapy.148 Recently, it was shown that administration of either 6 or thiohydantomtryptophan (10) potentiates the efficacy of DNA-damaging chemotherapeutic agents in the inhibition of tumor growth in mouse models.132'135 However, 6 is poorly water- soluble and difficult to administer to animals, ' and even though 10 is more potent and water- soluble than 6, its availability in plasma is low.148 In 1984, a group of P-carbolines (e.g. 13 and 14) were reported to exhibit IDO inhibition.155 They were classified as noncompetitive inhibitors upon evidence of direct interaction with the heme iron as an electron donor ligand (instead of O2), and remain as the most common structures for this type of IDO-active compounds.143 P-Carboline derivatives, however, possess neuroactivity as benzodiazepine receptor ligands, which could cause problematic side effects in cancer therapy.157 75 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.1. Some IDO inhibitors with high in vitro activity. 143,148 6 R,= CH3, R2=R3=R4=H 7 R,=R3=H, R2=R4=F o J\.|=i\2=H> K.3=K.4~r" 9 R1=R2=R3=H, R4=F NH 10 COOH (Xf^NH' V\ y,N 11 x=o 12 X=S 13 Rj= /i-C4H9> R2=H 14 R1= C02CH3, R2=NCS 15 16 Type Competitive Noncompetitive Compound jRTj (^M) (5)-Methyltryptophan (6) 34 (5)-4,7-Difluorotryptophan (7) 40 (5)-5,7-Difluorotryptophan (8) 24 (5)-7-Difluorotryptophan (9) 37 (#,S)-Thiohydantointryptophan (10) 11.4 (/?,5)-2-Amino-3-benzofuran-3-yl-propionic acid (11) 25 (/?,S)-2-Amino-3-benzo[£]thiophen-3-yl-propionic acid (12) 70 3-Butyl-9/7-p-carboline (13) 3.3 6-Isothiocyanato-9//-P-carboline-3-carboxylic acid methyl ester (14) 8.5 4-Phenylimidazole (15) 4.4 Brassilexin (16) 5.4 Recognizing the need for new IDO inhibitors belonging to different structural classes, the Mauk research group in the Department of Molecular Biology at UBC modified and adapted an IDO-activity assay published by Takikawa and colleagues in 1988,138158 in order to develop a high-throughput bioassay capable of screening libraries of marine invertebrate extracts in a multiwell plate format. A mixture containing phosphate buffer (pH 6.5), ascorbic acid, catalase, methylene blue, L-tryptophan and recombinant IDO, was added to a solution of the invertebrate extract and the resulting reaction crude was allowed to react at 37°C for 30 min (Figure 3.2). The process was stopped by addition of trichloroacetic acid, followed by heating at 65 °C to afford kynurenine (2). Treatment of 2 with p-dimethylaminobenzaldehyde in acetic acid generates the yellow adduct p-(A^-dimethylaminobenzylidene)kynurenine (17). The amount of this product 76 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata formed was monitored at 480 nm as a measure of IDO activity. As shown in Figure 3.2, extracts with inhibitory activity displayed uncolored wells, since the enzyme cannot oxidize tryptophan to jV-formylkynurenine (1), the precursor of 2 and 17. In total, approximately 4500 extracts from a marine invertebrate extract library were screened. Marine extract ( rr f ~* + Y s~\ /-%, f~\ <<->, . A _ f- f~* \ <r"' • ^ Yes Uncolored well or H TFA 60°C, 15 min Yellow well HOAc / », J t ,J \ / X J Figure 3.1. Reactions involved in the new high throughput IDO bioassay developed by Mauk and Vottero.156 Control column: (+), wells with IDO; (-), wells without IDO. 3.4. Secondary metabolites from hydrozoans The phylum Cnidaria encompasses a diverse collection of marine invertebrates that include hard and soft corals, gorgonians, sea pens, jellyfish, and sea anemones. Hydroids, a large class of cnidarians (around 3000 species), are generally small and somewhat inconspicuous.159 In the last three decades, marine natural products researchers have successfully isolated a large number of secondary metabolites from soft corals, sea pens and zooanthids,160"163 but hydroids 77 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata have received very little attention.159'164 According to MarinLit,137'165 from a total of 2281 structures reported for cnidarians, only 60 were found from hydroids. The polyhydroxylated steroid (18, Figure 3.3), isolated by Cimino166 in 1980 from Eudendrium sp. collected in the Bay of Naples, was the first publication on a natural compound from a hydroid. In the following years, Fattorusso and coworkers167"169 reported the distribution of mono and polyhydroxylated sterols such as (19), in four Mediterranean hydroids. The same source yielded three brominated (3-carbolines similar to (20).170 Fusetani171 reported in 1986 one of the first phosphate-containing marine natural products (21) along with two more phosphorylglycerylethers exhibiting hemolytic activity, from the hydroid Solanderia secunda collected in the Gulf of Sagami, Japan. Around a decade later in 1996, the same hydroid was collected offshore of Jaeju Island (Korea), and its DCM/MeOH extract was purified by Shin and coworkers172 to give solandelactones A (22) to I, nine lactonized oxylipins exhibiting moderate inhibitory activity against Farnesyl Protein Transferase. A study focused on predator-prey interactions in animals of the pelagic Sargassum communities of the western Atlantic Ocean, found that among four common hydroids growing on the Sargassum, only Tridentata marginata was not eaten by the most abundant predator, the Planehead fish. Bioassay-guided fractionation led Lindquist, Lobkovski and Clardy, to isolate tridentatols A (23) to C in 1996,174 and tridentatols D-H in 2002.175 Given the UV-absorbing characteristics and their high tissue concentrations, these compounds were also proposed to function as sunscreens, protecting T. marginata from intense levels of solar radiation typical of near-surface oceanic habitats. Potent antioxidant activity against human low density lipoprotein (LDL) was later reported for tridentatol A (23).176 Corydendramines A (24) and B, two piperidinol metabolites isolated from Corydendrium parasiticum L. by Lindquist, 1 77 Shigematsu and Pannel, showed similar predator deterrence activity. 78 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata OH OH "V to 4 HO,,, 18 21 24 19 Ho^r HO" 22 20 OH 23 OH 26 Figure 3.3. Representative structures for secondary metabolites isolated from hydroids: cholest- 4-en-4,16p\18,22/?-tetrol-3-one 16,18-diacetate (18),166 cholest-5-en-2oc,3a,7p\15p\18-pentol 2,7,15,18-tetraacetate (19),168 6-bromo-l-ethyl-P-carboline (20),170 l-hexadecyl-^n-glycerol-3- phosphorylcholine (21),171 solandelactone A (22),172 tridentatol A (23),174 corydendramine A 177 (24), lytophilippine A (25), and Campanularia sp. metabolite (26). 179 Recently, lytophilipinnes A (25) to C, three new chloro-containing macrolides, were found by Rezanka, Hanus and Dembitsky,178 in the Red Sea hydroid Lytocarpus philippinus. The compounds exhibited moderate antibacterial activity, as well as crown gall tumor inhibition and potent toxicity against Artemia salina (brine shrimp toxicity assay), suggesting a role in the hydroid's defense. Finally, Houssen and Jaspars179 examined the hydroid Campanularia sp. collected from New Zealand and isolated (26), with no relevant activity. 79 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3.5. Minor metabolites from Garveia annulata One more representative family of metabolites isolated from hydroids must be added to those ones depicted in Figure 3.3. Garveatins O OH OH X HO' 27 X = 0 28 X = H2 O OH OR O 29 R = H 30 R = CH3 OH OH X 32 X = 0 33 X = H2 Garvins Annulins Garvalones 159 Figure 3.4. Secondary metabolites isolated from G annulata: garveatins A (27), B (28), C (29),164 and D (30), garveatin A quinone (31), and 2-hydroxigarveatins A (32) and B (33); 164 garvin A (34), 2-hydroxygarvin A (35), garvin A quinone (36), garvin B (37), and 2 .180 181 hydroxygarvin B (38); U annulins A (39) and B (40); ' garvalones A (41) and B (42). 180 80 Chapter 3. Isolation and Structure Elucidation of new IDC* Inhibitors from Garveia annulata In 1985, Fahy and Andersen159 reported garveatin A (27, Figure 3.4), the first compound isolated from G. annulata. Metabolite (27), which exhibited mild antimicrobial activity, was obtained as orange needles via bioassay-guided fractionation of methanolic extracts. In the following months, additional reports on 16 new G. annulata polyketide secondary metabolites164'180'181 would follow. They can be organized into four families according to the number of carbons in the putative polyketide precursor and other structural similarities (Figure 3.4). 2-Hydroxygarvin A ^ C Garvin A ^ O O (• Ga ^, rvin O A quinone 2-Hydroxygarvin A O ^ A w Garvin B IT L i o A> A ^As-~ x o o o o Ao^^yy^^k^ x p p. H,CH3or s ^ J ^ U ^ a ^ ^ ^ \ . O ^ X R = H o r C H 3 2-Hydroxy garveatins A and B Garvalone A Garvalone B Scheme 3.2. Proposed biogenesis of G. annulata secondary metabolites. 81 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Garveatins, garvins, annulins and garvalones, are classical examples of polyketide derived secondary metabolites. At the time of their isolation, they represented the first examples of polyketide metabolism from coelenterates.164 All of them are proposed to be biosynthesized from similar nonaketide precursors exhibiting the same methylation pattern at C8 and CIO. Additional variability is introduced at CIO, limited to three substituents: a proton, a methyl, or a diketobutyl branch (Scheme 3.2). Decarboxylation at CI generates the only methyl substituent for the aromatic system in the garveatins. Different folding patterns give rise to an ethyl substituent or a 5-lactone in garvins and garvalones. Secondary biogenetic transformations, such as hydroxylation at CIO and oxidation of the central ring to a quinone, may be responsible for the formation of 2-hydroxygarveatin A (32), B (33) and 2-hydroxygarvin A (35), as well as garveatin A and garvin A quinones (31 and 36). 2-Hydroxygarveatin B (33) is possibly an intermediate in the biogenesis of annulins A (39) and B (40), considered to be degradation products of the garveatins.181 G. annulata metabolites are closely related to the cytotoxic abietinarins A (43) and B (44), also isolated by the Andersen research group shortly after the garveatin publications.182 Abietinaria sp., collected off the southern coast of Vancouver Island, became the second hydroid reported to produce l-[4//]-anthracenone-based polyketide-derived metabolites. Abietinarin A (43) exhibited significant in vitro cytotoxicity. o OH OH HO*- ^Y^ \ ^ \ ^ ^0^ H O < 43 44 Additional marine-derived naphthoquinone and anthraquinone compounds include a series of pigments (e.g. 45-47), presumably polyketide-derived and isolated from echinoderms,183'184 as well as a benz[a] anthraquinone antibiotic (48) from a marine actinomycete Chania sp. 82 Chapter 3. Isolation and Structure Elucidation of new EDO Inhibitors from Garveia annulata OH o OH 45 46 47 48 Condensed polycyclic aromatic systems are more common in terrestrial species of plants and fungi. Berries, leaves, bark and root bark in specimens of the genuses Vismia,m'm Cassiam'm and Gasteria,190 have afforded a series of mono and dimeric anthranoids similar to 49 and 50. Rare diterpene ortho-quinones such as pygmaeocine E (51), were found in the roots of several medicinal plants used as a folk remedy in China to reduce inflammation and cure malaria.191 Cytoskyrin A (52), a bisanthraquinone highly active in an anticancer bioassay192 was isolated from an endophytic fungus collected in Costa Rica. 49 50 51 52 3.6. Isolation of IDO-active minor metabolites Garveia annulata is a small orange hydroid encountered in rocky subtidal habitats from Alaska to Southern California during the winter and spring months.193 It is reasonably abundant in Barkley Sound, British Columbia, especially in winter. In its polyp stage, it possesses small tentacles supported by a hydrostatic skeleton that allows the different colonies to stay together in bush-like structures, and keep their shape in the moving tides (Figure 3.5). As a colonial hydrozoan, it is composed of a number of specialized polyps with feeding, reproductive and protective functions. Its medusa stage is sexually active, but has a limited lifespan. 83 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Cape Beak Figure 3.5. a) A bush-like colony of G annulata; b) Map of collection site. Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Bioassay-guided fractionation of a G. annulata methanolic extract (0.35 g) as detailed in the Experimental section, led to the purification and structure elucidation of two of the most potent IDO inhibitors known to date; the new metabolite annulin C (0.3 mg, 0.0006% wet wt) obtained as a yellow oil, and the known annulin B (40) (0.6 mg, 0.001% wet wt). Such promising results prompted the recollection and examination of additional Garveia specimens. Samples of G. annulata (424 g wet wt) were collected by hand using SCUBA at depths of 10-15 m off the coast surrounding the Broken Islands, Barkley Sound, on June 2003 (Figure 3.5). Specimens were frozen immediately upon collection and transported back to UBC in coolers packed with dry ice, where they were immediately extracted using MeOH. Repetitive extraction and solvent concentration for a week yielded an UDO-active brown gum (2.5 g), which was fractionated as described in the Experimental section. This second batch of specimens afforded the new metabolite garveatin E (0.5 mg, 0.0001%, wet wt) as a pale yellow oil, and the previously reported 2-hydroxygarveatin B (33) (1.1 mg, 0.0002%, wet wt), annulin A (39) (0.5 mg, 0.0001%, wet wt), garveatins A (27) (34.0 mg, 0.008%, wet wt) and C (29) (0.5 mg, 0.0001%, wet wt), 2-hydroxygarvin A (35) (3.7 mg, 0.0009, wet wt), and garvin A quinone (36) (0.4 mg, 0.00009%, wet wt). Once all IDO-active fractions were elucidated, it was decided to attempt isolating G. annulata metabolites with reduced or zero inhibition, in order to show that activity resides preferentially in the annulin naphthoquinone substructure, and confirm that garveatins, garvins and garvalones are less potent. Thus, exploration of the inactive fractions yielded the new compounds 2-hydroxygarveatin E (0.4 mg, 0.00009%, wet wt) and garvin C (0.3 mg, 0.00007%, wet wt) as pale yellow oils. Due to limited amounts of all the new G. annulata compounds, the l3C-NMR data was derived from HMQC and HMBC spectra (sections 3.6.1-3.6.4, and Experimental), and some quaternary carbons were not assigned. 85 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3.6.1. Annulin C Annulin C (53) was isolated as a yellow oil that gave an [M]+ ion at mlz 374.1367 in the HREIMS, appropriate for a molecular formula of C20H22O7. The ID and 2D NMR data obtained for 53 showed a strong resemblance to the data reported for annulin A (39),n indicating that the molecules were closely related (Figure 3.6). The major difference in the !H NMR spectra of the two compounds was the absence in the spectrum of annulin C (53) (Figure 3.6b, Table 3.2) of a singlet that could be assigned to the hemiketal C8-OH present in the spectrum of 39 (8 4.85), and its replacement by a methyl singlet at 3.45 ppm (HI8). Through HMQC, this new methyl resonance was shown to belong to a carbon at 51.7 ppm, whereas the HMBC spectrum showed a correlation (Table 3.2, Figure 3.7) from HI8 to a carbon resonance at 8 106.8 (C8), typical of a ketal functionality. Therefore, it was evident that the hemiketal at C 8 in annulin A (39) had been replaced by a methyl ketal in annulin C (53). The remaining HMBC correlations established two main fragments for the structure of annulin C: I. Ketal-gem dimethyl fragment: with methyl H10 (s, 8H 3.82, 8c 52.5) correlating to the carbonyl C9 at 167.2 ppm; and both methyl singlets H19/H20 showing cross peaks with quaternary carbons C l l (8c 88.3) adjacent to an oxygen atom, and C12 (8c 154.3) in an aromatic system (Figure 3.7). II. Branched phenol fragment: displaying very informative correlations from methylene H2 (8H 2.75, q, J = 7.5 Hz, 8C 19.4) and methyl H17 (s, 8H 2.41, 8c 19.7), that allowed assignment of their position in an aromatic ring and the relative location of methine HI5 (s, 8H 7.44, Sc 121.5) and sp2 carbon C4 (8c 160.3), attached to a hydroxyl group (s, 8H 12.2) (Figure 3.7). 86 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata a) H,0 H10 2.5 2 .0 1.5 40H H15 80H IH|< AMINIM III I IMIII IIHI<IIII«IIIIII'WIH n»iiimi»iMiiii|iiiiiniiniiii I IWKMI 'H 'J I \ H I liminiin nmnii imimii fU \JmJ H17 H2 »i> cJiiMii i^^ni ii IP4 '*»*f' H19 H18 12 11 10 9 HI b) H17 "20 H19 HO HI , HI 5 H10 H18 !uiy H17 H20 H2 XJ H19 HI V \ X 12 11 10 7 6 5 Chemical Shift (ppm) Figure 3.6.' H-NMR spectrum of a) annulin A (39) (recorded in CD2C12 at 400 MHz), and b) annulin C (53) (recorded in CDC13 at 400 MHz). 87 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.2. NMR data for annulin C (53) (recorded in CDC13). Carbon No 1 2 3 4 5b 6b 7 b 8 9 10 11 12 13b 14b 15 16 17 18 19 20 1 3 C 8 (ppm)a 12.8 19.4 140.0 160.3 106.8 167.2 52.5 88.3 154.3 121.5 145.3 19.7 51.7 26.4 26.4 *H 8(ppm)(mult,/(Hz))a 1.12 {t,J = 2.75 (2q,J OH 12.2 (s) 3.82 (s) 7.44 (s) 2.41 (s) 3.45 (s) 1.68 (s) 1.62 (s) 7.5 Hz) = 7.5 Hz) 1 HMBCa (H-»C) C2,C3 C1,C3,C4,C16 C4 C9 C3,C15, C16 C8 C11,C12,C20 C11,C12,C19 COSYa (H->H) H2 HI "According to HMQC and HMBC recorded at 400 MHz. Not assigned. After accounting for the atoms already assigned from the NMR data of substructures I and II, a C5O2 fragment remained to be allocated. Based on these atoms, and by comparing the chemical shift of C12 with NMR data for annulin A (39), it was evident that a para-quinone was connecting I and II. 88 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 10 , 1 8 \ / \ 9 \ - (/ \ "̂ ^ / \ 19 „ / > ' C^ 2 0 - HO \4S ^- A 13„ 14 16 15>v 17 COSY HMBC Figure 3.7. Summary of COSY and HMBC correlations for annulin C (53). It is important to point out that annulin C (53) may be an isolation artifact in which the C8 methyl ketal has been formed by reaction of the co-occurring metabolite annulin A (39) with MeOH, according to Scheme 3.3. Scheme 3.3. Proposal for the formation of annulin C (53) from annulin A (22). 89 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3.6.2. 2-Hydroxygarveatin E 2-Hydroxygarveatin E (55), isolated as a yellow oil, gave a [M+Na]+ ion at mlz 379.1160 in the positive ion HRESDVIS, consistent with a molecular formula of C20H20O6 (calculated for C2oH2o06Na: 379.1158), requiring 11 sites of unsaturation. The 'H-NMR spectrum of 55 (Figure 3.8, Table 3.3) contained resonances that could be assigned to an ethyl fragment (8H 1.12, t, J - 7.7 Hz, HI; 8H 2.67, q, J = 7.7 Hz, H2/H2'), four methyl groups (all singlets, 6H 1.44, H19; 5H 1.56, H18; 5H 1.76, H20; 5H 2.21, H17), an aromatic methine (SH 7.72, s, H13), and a phenol OH (5H 13.37, s, 6OH), accounting for 19 of a total of 20 protons in the molecule. HMBC correlations (Table 3.3, Figure 3.9) observed between methyl HI and a carbon resonance at 149.2 ppm (C3), between methylene H2 and quaternary sp2 carbon resonances at 5c 144.3 (C16), 149.2 (C3) and 188.6 (C4), and between H17 and carbon resonances at 5C 144.3 (C16), 149.2 (C3), and 183.5 (C15), suggested that the methyl and ethyl residues were vicinal substituents on a /?ara-quinone substructure. The aromatic methine H13 showed an HMBC correlation to a quinone carbonyl resonance at 183.5 ppm (C15), which situated C13 (8c 115.2) two bonds away from the carbonyl. Additional HMBC correlations between HI3 and sp2 carbons at 5C 114.5 (C5) and 5C 122.3 (C14), and between 60H (6H 13.27) and carbons at 5c 114.5 (C5), 5c 122.3 (C7), and 8c 160.4 (C6) were consistent with placement of the phenol-bearing carbon C6 two bonds away from the second quinone carbonyl C4. The two methyl resonances H19 and H20 displayed HMBC correlations to each other as well as to carbon resonances at 8c 48.4 (Cll), 154.5 (C12), and 207.0 (C10), which demonstrated the methyls were geminal substituents on a quaternary carbon (Cll) attached on one side to an aromatic ring and on the other side to a saturated ketone. 90 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 19 20 H 1 7  H19 H18)^H20 H2 J SJ#U»*ll**#fcbfyJ d4J Kmi4tfW 2.5 2.0 60H W»>|»H«»IW 1.5 H1 H13 PJM H18 H17 H19 i ^ l l l ^ ^ l W ^ i l H ' M ^ W ' H2 L^jy H,0 H20 H1 ^ 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 Chemical Shift (ppm) Figure 3.8. 'H-NMR spectrum of 2-hydroxygarveatin E (55) (recorded in CDC13 at 600 MHz). 91 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.3. NMR data for 2-hydroxygarveatin E (55) (recorded in CDC13). Carbon No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 13/ 8 (ppm)a 8 (ppm) (mult, / (Hz))a 12.7 19.8 149.2 188.6 114.5 160.4 122.2 194.3 83.9 207.0 48.4 154.5 115.2 122.3 183.5 144.3 12.3 26.4 26.2 30.0 1.12 (t, 7 = 7.7 Hz) 2.67 (q, 7 = 7.7 Hz) OH 13.3 (s) 7.72 (s) 2.21 (s) 1.56 (s) 1.76 (s) 1.44 (s) HMBCa (H->C) C2,C3 CI, C3, C4, C16 C5, C6, C7 C5.C14, C15 C3, C15, C16 C8, C9, C10 C10,C11,C12, C20 C10,C11,C12,C19 According to HMQC and HMBC recorded at 600 MHz. Not assigned. HMBC cross peaks from methyl H18 to carbon resonances at 8C 83.9 (C9), 194.3 (C8), and 207.0 (C10) showed that H18 was attached to a quaternary oxygenated carbon (C9), which was in turn flanked by two ketones. The oxygen atom on C9 had to be part of a hydroxyl functionality to account for the molecular formula of 55. 92 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata O QH\- o 19 20 O Figure 3.9. Summary of HMBC correlations for 2-hydroxygarveatin E (55). There are two possible ways to attach the gem-dimethyl-bearing C l l and ketone termini C8 of the aliphatic fragment, to the unsatisfied valences of the naphthoquinone fragment (C7 and CI2). The HMBC data showed that the gem-dimethyl-bearing carbon is linked to a carbon with a chemical shift of 154.5 ppm. This deshielded carbon (C12) could only be meta to the phenol (C6), leading to the structure 55 for 2-hydroxygarveatin E. Compound (55) is a C4/C15 para- quinone, analogue of the previously described metabolite 2-hydroxygarveatin B (33). 3.6.3. Garveatin E Garveatin E (54), a very minor component of the extract, was obtained as a pale yellow oil that gave an [M+Na]+ ion at mlz 363.1200 in the HRESIMS, consistent with a molecular formula of C20H20O5 (calculated for Cio^oOsNa: 363.1208), which differed from the molecular formula of 2-hydroxygarveatin E (55) simply by loss of one oxygen atom. Analysis of the NMR data obtained for 54 indicated that it differed from 2-hydroxygarveatin E (55) simply by loss of the hydroxyl functionality at C9. 93 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata H19 H17 H18 H20 H2 W J H1 2.5 2.0 1.5 H13 10OH 60H «i< ili.iiiiiiniJiiiiirt..m»iii,ln »> umii imi»,iMi ill \\t mi mni myK'-^M i i i n .» H18 H17 H2 H19 H20 H1 14 12 10 8 6 Chemical Shift (ppm) Figure 3.10. 'H-NMR spectrum of garveatin E (54) (recorded in CDC13 at 400 MHz). 94 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.4. NMR data for garveatin E (54) (recorded in CDC13). 9 10, Carbon No 1 2 3 4 5b 6b 7 b 8 9 10 11 12 13 14 15 16 17 18 19 20 1 3 C 8 (ppm)a 12.5 19.5 148.9 188.9 200.0 113.3 160.6 48.3 114.4 117.5 122.2 183.3 145.3 12.2 7.18 30.0 27.7 According to HMQC and HMBC recorded at 400 MHz. Not assigned. The lH NMR spectrum of (54) (Figure 3.11, Table 3.4) contained resonances that could be assigned to an ethyl (8H 1.14, t, J = 7.6 Hz, HI; 2.66, q, / = 7.6 Hz; H2/H2') and two olefmic methyl (SH 2.21, s, H17; 1.95, s, H18) groups, an aromatic methine (8H 7.76, s, H13), and a pair of geminal methyls (6H 1.46, s, H19/H20). HMBC correlations (Figure 3.10, Table 3.4) between methylene H2 and a carbonyl resonance at 188.9 (C4) ppm, and between methyl H17 and a 95 8(ppm)(muIt,7(Hz))a 1.14 (t, J = 7.6 Hz) 2.66 (q, 7 = 7.6 Hz) HMBC8 (H->C) C2, C3 C1,C3, C4, C16 OH 10.15 (s) OH 14.6 (s) 7.76 (s) C11,C12,C14 2.21 (s) 1.95 (s) 1.46 (s) 1.46 (s) C3,C15, C16 C8, C9, C10 Cll Cll Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata second carbonyl resonance at 183.3 (CI5) ppm, confirmed that the ethyl and one of the methyl residues in 54 were also vicinal substituents on a/?ara-quinone, as in 55. Figure 3.11. Summary of HMBC correlations for garveatin E (54). The olefinic methyl HI8 displayed HMBC correlations to carbon resonances at 6c 200.0 (C8), 113.3 (C9), and 160.6 (CIO), which were assigned to an a,P-unsaturated ketone, a quaternary olefinic methine and an oxygenated olefinic carbon, respectively. Furthermore, the aromatic methine H13 showed an HMBC cross peak with the quaternary carbon CI 1 (48.3 ppm), all in agreement with the proposed structure 54. Fortuitously, attachment of the gem-dimethyl- bearing CI 1 and ketone termini C8 of the aliphatic fragment, to the naphthoquinone fragment at C12 is completely unambiguous, as evidenced by HMBC correlations from HI3 to C12 and CI 1. 3.6.4. Garvin C Garvin C (56) was isolated as a pale yellow solid that gave a [M-H]" ion at mlz 327.0866 in the negative ion HRESIMS, appropriate for a molecular formula of CisH^Oe (calculated for CigHisOe: 327.0869), requiring 11 sites of unsaturation. Detailed analysis of the NMR data collected for 56 showed that it was closely related to the known metabolite garvin B (37). In 96 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata particular, HMQC and HMBC (Figure 3.12) data confirmed the presence of a 8-lactone (CI to C4, C15, and C16) and dihydroxy naphthalene (C4 to C6 and C l l to C15) substructures in 56 that were identical to the corresponding substructures in 37. Figure 3.12. Summary of HMBC correlations for garvin C (56). The 'H-NMR spectrum of 56 (Figure 3.13, Table 3.5) contained resonances suitable for a methyl (HI, 8H 1.59, d, J = 6.4 Hz), an oxygenated methine (H2, 8H 4.75, m), and a benzylic methylene (H3, SH 4.31, dd, J = 4.3, 17.9 Hz, 8H 3.39, dd, J = 3.4, 18.2 Hz) in the 5-lactone. For the dihydroxy naphthalene fragment, a singlet for 140H at 10.9 ppm is evident, whereas the aromatic region exhibits methines H l l (5H 7.06, m) and H13 (5H 7.15, m). Finally, the characteristic geminal methyls H17 and H18 present singlets at 1.576 and 1.582 ppm, respectively. Although 'H-NMR resonances for impurities can be observed in compounds (53) to (55), metabolite (56) in particular possesses an important number of additional peaks exhibiting integration values well below those measured for its constituting protons. Clearly, the chemical complexity of the invertebrate extract and the reduced amounts of material available limited the degree of purity of these new Garveia compounds. 97 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 140H H13 \ gJJt H3 H2 H11 "V_ J\ H20 H3' H1 -j——-i 1 | [—•—| r- ' l i l l i j n ' T l 11rr | ! in11Mr 4.50 4.25 3.40 H3 H2 JJLJJLL H18 H17 H1 i 1.60 1.58 Ldj r'-j\jj^_ 11 10 7 6 5 Chemical Shift (ppm) 1 0 Figure 3.13. !H-NMR spectrum of garvin C (56) (recorded in CDC13 at 600 MHz). 98 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.5. NMR data for garvin C (56) (recorded in CDC13). Carbon No 13/ 8 (ppm)8 XH 8(ppm)(mult,/(Hz))a 1 2 3 4 5 6b 7 8 9 10 11 12 13 14 15 16 17 18 20.6 75.4 33.7 140.3 110.5 134.1 178.5 43.9 137.8 112.6 136.8 111.2 155.7 116.1 169.9 25.1 25.1 1.59 (d, 7 = 6.4 Hz) 4.71 (m) 3.39 (dd, / = 3.4, 18.2 Hz), 4.31 (dd, J = 4.3, 17.9 Hz) 7.06 (s) 7.15 (s) OH 10.9 (s) 1.576 (s) 1.582 (s) HMBC8 (H->C) C2,C3 C1,C2, C4, C5, C15 C7, C9, C12, C13 C5, C11,C12, C14, C15,C16 C12.C13.C14 C8,C9, C10 C8.C9.C10 "According to HMQC and HMBC recorded at 600 MHz.b Not assigned. Long-range HMBC correlations (Table 3.5, Figure 3.12) between the aromatic methine HI3 and the ester carbonyl resonance at 5H 169.9 (CI6), and between 140H and the aromatic carbon C12 (136.8 ppm), both attributed to W coupling pathways, provided additional support for the substitution pattern on the naphthalenic substructure. HMBC correlations observed 99 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata between methyl singlets H17/H18, and carbon resonances at 8C 43.9 (C9), 137.8 (CIO), and 178.5 (C8), identified a quaternary carbon bearing a pair of geminal methyl groups, linked to an ester/acid carbonyl as one of the two remaining substituents at C7 or CIO of the naphthalenic fragment in 56. In order to account for the molecular formula of 56, the carbonyl functionality had to be part of a y-lactone fused to the naphthalenic fragment. An HMBC correlation between the aromatic methine H l l and the quaternary carbon C9 demonstrated that C9 was ortho to HI 1, as shown in (56). Such positioning was consistent with the relatively shielded chemical shift observed for the oxygenated aromatic carbon C7 (134.1 ppm), which forms part of the y-lactone. The new metabolites annulin C (53), garveatin E (54) and 2-hydroxygarveatin E (55), represent minor structural variations of the known compounds annulin A (39) and garveatin B (28). Garvin C (56) on the other hand, while obviously related to garvin B (37), has a new carbon framework not previously found among G. annulata polyketides. The garvin C skeleton might arise from degradation of garvin B (37), which requires excision of the equivalent of one polyketide residue comprising C8/C9 (60) along with the associated substituents (Scheme 3.4). Alternatively, garvin C (56) might arise from a dimethylated octaketide (57) instead of the putative nonaketide precursor (58) to garvin B (37). Cyclization of the octaketide (57) to give a naphthalenic core fused to a 5-lactone on one end and a cyclobutanone on the other (compound 59), followed by a biological Baeyer-Villiger reaction on the cyclobutanone, would lead directly to garvin C (56). 100 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata O O O O O O O O O O O O O O O O O 59 Baeyer-Vi Niger OH Scheme 3.4. Possible biogenesis of garvin C (56). 3.7. Biological evaluation of G. annulata secondary metabolites Table 3.6 summarizes all the chemical components isolated from Garveia annulata that exhibit IDO inhibition ordered by their potency. Compared with those listed in Table 3.1, annulins B and C are 100 and 30 times more potent than (/?,S)-thiohydantointryptophan (10) and 3-butyl-9//-P-carboline (13), two IDO inhibitors (competitive and noncompetitive) exhibiting the highest activity to date. All Garveia compounds showed noncompetitive inhibition against 156 tryptophan, suggesting that they may bind IDO at a site other than the tryptophan active site. 101 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.6. IDO-active secondary metabolites from G. annulata. Entry Name 1 Annulin B (40) 2 Annulin C (53) 3 Annulin A (39) 4 Garveatin C (29) 5 2-Hydroxygarveatin B (33) 6 2-Hydroxygarvin A (35) 7 Garveatin E (54) 8 Garveatin A (27) 9 2-Hydroxygarvin B (38) Structure K% (uM)1 Remarks a Q \ O OH O OH O*^ O O OH OH H O H O 0.12 Known 0.14 New 181 0.69 Known 181 164 1.2 Known 1.4 Known180 164 2.3 Known 3.1 New 3.2 Known159 >10 Known [EI] =st: [Enzyme][Inhibitor]. Missing in Table 3.6 are garvin A quinone (36), and the new 2-hydroxygarveatin E (55) and garvin C (56), isolated from G annulata fractions exhibiting low IDO inhibition. 102 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 36 55 56 A sharp decay in inhibitory activity (more than 10 times) is observed when the naphthoquinone substructure of the annulins is replaced by the anthracenoid skeleton of the remaining metabolites (garveatins, garvins, garvalones), with the exception of garveatin E (54). Additionally, large substituents in the C ring, such as w-propyl in 2-hydroxygarvin A (35) and garvin A quinone (36), or a y-lactone in 2-hydroxygarvin B (38) and garvin C (56), decrease inhibition. The relative position of a para-quinone within the polycyclic ring system seems to be crucial for potency, since activity falls as the quinone moves from ring B in annulins, to ring C in garveatin E (54) and 2-hydroxygarveatin E (55). This trend is overshadowed by large substituents in ring C, which explains the inactivity shown by garvin A quinone (36). The high inhibitory activity of G. annulata compounds exhibiting a naphthoquinone core, and its dependence to the position of the quinone, led Mauk and Vottero156 to evaluate IDO inhibition in a small library of commercially available naphthoquinones (Table 3.7). Among them, a water-soluble analogue of vitamin Ki (69) called menadione (66), exhibited nanomolar inhibition making it a candidate for in vivo examination, and further drug development. Annulins A (39), B (40), and C (53) were all inactive in a recently developed yeast-based IDO inhibition assay,148 suggesting that they may not cross the yeast cell wall. Conversely, menadione (66) and 2-bromo-2,4-naphthoquinone (63) are able to cross the yeast membrane and inhibit human IDO expressed in the yeast cytoplasm.156 103 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Table 3.7. Annulin C (53) and commercially available IDO-active naphthoquinones. 156 fl II 61 62 63 R=Br 64 R=H 65 R=OCH3 66 R=CH3 70 R=OH 67 Entry 1 2 2 4 5 6 7 8 9 10 11 K, Name Dichlone (61) Juglone (62) Annulin C (53) 2-Bromo-l,4-naphthoquinone (63) 1,4-Naphthoquinone (64) 2- Methoxynaphthoquinone (65) Vitamin K3 (Menadione, 66) 1-Naphthol (67) 1,2-Naphthoquinone (68) Vitamin Ki (69) Lawsone (70) [EI] ^^- [Enzyme][Inhibitor]. 68 69 Kt (nM)1 45 48 144 215 334 530 580 1800 3400 >40000 >100000 Besides their mild antimicrobial and high IDO inhibitory activities, no additional biological roles have been reported for the secondary metabolites of G. annulata. Menadione (66) causes oxidative stress due to the formation of reactive oxygen species such as H2O2, 'C^, and "OH during metabolism,195"197 and has been shown to induce lung198 and liver199 cancer cell death by either apoptosis or necrosis. Computer-assisted docking studies and kinetic experiments indicate that interaction of menadione (66) with IDO occurs primarily through hydrophobic contacts above the distal side of the heme in the binding site.156 The high potential of 66 to become a drug was recognized by NewLink Genetics Corporation, a biopharmaceutical company in Iowa that recently acquired patent rights for the IDO technology from UBC 200 104 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Recently, the same IDO bioassay detailed in Section 3.3 led to the isolation of exiguamine A (71) from Neopetrosia exigua, a sponge collected in Papua New Guinea.134 This new potent IDO inhibitor {K\ = 210 nM) exhibited a complex hexacyclic alkaloid skeleton, without precedent among known natural products. Noteworthy, the novel carbon backbone combines elements from tryptophan and naphthoquinones in one structure. 3.8. Conclusions Bioassay-guided fractionation of MeOH extracts obtained from Garveia annulata, a seasonal hydroid collected in Barkley Sound, British Columbia, led to the isolation of twelve secondary metabolites including four new compounds: annulin C (53), garveatin E (54), 2- hydroxygarveatin E (55), and garvin C (56). 53 54 55 56 Annulin C (53), garveatin E (54), and 2-hydroxygarveatin E (55), are analogues of the previously reported annulin A (39) and garveatin B (28),164'181 while garvin C (56) has a new carbon framework not previously found among G. annulata polyketides. Nine of these 105 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata metabolites showed inhibition of indoleamine 2,3-dioxygenase (IDO), with the annulins among the most potent in vitro EDO inhibitors isolated to date (Table 3.8). Table 3.8. In vitro inhibition of EDO by marine natural products, commercial naphthoquinones, P-carbolines, and tryptophan analogues. Entry 1 2 3 4 5 6 7 8 Name AnnulinB (40) AnnulinC (53) Exiguamine A (71) Menadione (66) AnnulinA (39) 3-Butyl-9//-P-carboline (13) (/?,5)-Thiohydantointryptophan (10) (5)-Methyltryptophan (6) [EI] =i= [Enzyme] [Inhibitor]. K, (nM)1 120 144 210 580 690 3300 11400 34000 The term pharmacophore refers to the ensemble of steric and electronic features required to ensure optimal interactions with a specific biological target structure and to trigger (or to block) its biological response.201 The annulin family of Garveia metabolites share a common 5- hydroxy-6-ethyl-7-methyl-l,4-naphthoquinone core substructure, which by comparison with other inhibitors, seems to be a requisite for potent EDO inhibition. Based on the initial inhibitory activity shown by tryptophan derivatives (Table 3.1), and the high potency exhibited by annulins and similar naphthoquinones (Tables 3.6 and 3.7), a possible pharmacophore for human EDO can be proposed (Figure 3.14). 106 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata \ 11 = 0,1 X = O, N, C Y = C=0, CH2 E = NH2, OCH3 o Figure 3.14. Proposed pharmacophore involved in IDO inhibition. The core for this proposal is the quinone ring B. Some flexibility is seen for size and heteroatom positioning within ring A, as well as for length and substituents in its side arm. An additional third ring, or substituents on quinone B are not required for activity. IDO inhibition improves as planarity in the A/B ring system decreases from an indole in tryptophan derivatives, to dihydroheteronuclear five or six-membered A rings fused to ort/zo-quinones as in the annulins. 107 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 3.9. Experimental General Experimental Procedures For general experimental procedures see Section 2.6. Isolation Procedure The screening of a library of marine invertebrates for their ability to inhibit human IDO in vitro, revealed very promising activity in crude extracts of the Northeastern Pacific hydroid Garveia annulata. Samples of the red/orange hydroid Garveia annulata (200 g wet wt) were collected from Barkley Sound, British Columbia, on April of 1999, by hand using SCUBA at depths of 10-15 m (48° 0.50' N, 125° 12.50' W). The collected material was frozen immediately upon collection and transported back to the University of British Columbia in coolers packed with dry ice. A portion of the frozen material (-50 g) was extracted in MeOH (3 x 200 mL) and the combined extracts concentrated to dryness in vacuo to give a brownish solid (0.35 g). This residue was dissolved in H2O to obtain a light brown colored solution, and extracted sequentially with hexanes (3 x 50 mL), CH2CI2 (3 x 50 mL) and EtOAc (3 x 50 mL), followed by concentration in vacuo of each partition. The CH2CI2 bioactive fraction (0.1275 g) was subjected to silica gel column chromatography (20% EtOAc/Hexanes), followed by normal phase HPLC (15% EtOAc/Hexanes), to afford the new analogue annulin C (53) (0.0003 g, 0.80 ^mol, 0.0006% wet wt) as a yellow oil, and the known annulin B (40) (0.0006 g, 1.5 \ymo\, 0.001% wet wt). 108 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata A second batch of specimens (424 g wet wt), collected on June of 2003, was processed as above to generate 2.50 g of active CH2CI2 extract. Silica gel gradient column chromatography, followed by gradient (20% CH3CN/H2O to 83% CH3CN/H20) or isocratic (70% CH3CN/H20) reversed phase HPLC, yielded the new garveatin E (54) (0.0005 g, 1.5 |amol, 0.0001%, wet wt) as a pale yellow oil, and the previously reported 2-hydroxygarveatin B (33) (0.0011 g, 3.2 umol, 0.0002%, wet wt), annulin A (39) (0.0005 g, 1.4 umol, 0.0001%, wet wt), garveatins A (27) (0.034 g, 0.1 mmol, 0.008%, wet wt) and C (29) (0.0005 g, 1.3 umol, 0.0001%, wet wt), 2- hydroxygarvin A (35) (0.0037 g, 8.9 umol, 0.0009, wet wt), and garvin A quinone (36) (0.0004 g, 0.97 umol, 0.00009%, wet wt). Additional purification of IDO inactive fractions by reverse phase HPLC (70% CH3OH/H20) allowed the identification of the new compounds 2- hydroxygarveatin E (55) (0.0004 g, 0.1 umol, 0.00009%, wet wt) as a yellow oil, and garvin C (56) (0.0003 g, 0.91 umol, 0.00007%, wet wt) as a pale yellow oil. Due to limited amounts of all G. annulata compounds, the C-NMR data was derived from HMQC and HMBC spectra. Therefore, some quaternary carbons were not assigned. Annulin B (40): spectral data was in accord with that previously reported for the natural product.181 Annulin C (53): for a summary of lH and 13C NMR assignments based on HMQC and HMBC data, see Table 3.2. Yellow oil; lH NMR (CDC13, 400 MHz) £12.2 (1H, s), 7.44 (1H, s), 3.82 (3H, s), 3.45 (3H, s), 2.75 (2H, q, J = 7.5 Hz), 2.41 (3H, s), 1.68 (3H, s), 1.62 (3H, s), 1.12 (3H, t, J = 7.5 Hz); 13C NMR (CDC13, 100 MHz) £167.2 (C), 160.3 (C), 154.3 (C), 145.3 (C), 140.0 (C), 121.5 (CH), 106.8 (C), 88.3 (C), 52.5 (CH3), 51.7 (CH3), 26.4 (CH3), 26.4 (CH3), 19.7 (CH3), 19.4 (CH2), 12.8 (CH3); HREEV1S calcd for C20H22O7 (M+): 374.13655; found 374.13670. 109 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata Garveatin E (54): for a summary of [H and 13C NMR assignments based on HMQC and HMBC data, see Table 3.4. Pale yellow oil; lH NMR (CDC13, 400 MHz) £14.6 (1H, s), 10.15 (1H, s), 7.76 (1H, s), 2.66 (2H, q,J = 7.6 Hz), 2.21 (3H, s), 1.95 (3H, s), 1.46 (3H, s), 1.46 (3H, s), 1.14 (3H, t, J = 7.6 Hz); 13C NMR (CDC13, 100 MHz) £200.0 (C), 188.9 (C), 183.3 (C), 160.6 (C), 148.9 (C), 145.3 (C), 122.2 (C), 117.5 (CH), 114.4 (C), 113.3 (C), 48.3 (C), 30.0 (CH3), 27.7 (CH3), 19.5 (CH2), 12.5 (CH3), 12.2 (CH3), 7.18 (C); HRESIMS calcd for C2oH2o05Na ([M+Na]+): 363.1208; found 363.1200. 2-Hydroxygarveatin B (33): spectral data was in accord with that previously reported for the 1 OA natural product. Annulin A (39): spectral data was in accord with that previously reported for the natural product.181 Garveatin A (27): spectral data was in accord with that previously reported for the natural product.159 Garveatin C (29): spectral data was in accord with that previously reported for the natural product.164 2-Hydroxygarvin A (35): spectral data was in accord with that previously reported for the natural product.164 Garvin A quinone (36): spectral data was in accord with that previously reported for the natural product.164 110 Chapter 3. Isolation and Structure Elucidation of new IDO Inhibitors from Garveia annulata 2-Hydroxygarveatin E (55): for a summary of [H and 13C NMR assignments based on HMQC and HMBC data, see Table 3.3. Yellow oil; 'H NMR (CDC13, 600 MHz) 8 13.3 (IH, s), 7.72 (IH, s), 2.67 (2H, q, 7 = 7.7 Hz), 2.21 (3H, s), 1.76 (3H, s), 1.56 (3H, s), 1.44 (3H, s), 1.12 (3H, t, 7 = 7.7 Hz); 13C NMR (CDC13, 150 MHz) £207.0 (C), 194.3 (C), 188.6 (C), 183.5 (C), 160.4 (C), 154.5 (C), 149.2 (C), 144.3 (C), 122.3 (C), 122.2 (C), 115.2 (CH), 114.5 (C), 83.9 (C), 48.4 (C), 30.0 (CH3), 26.4 (CH3), 26.2 (CH3), 19.8 (CH2), 12.7 (CH3), 12.3 (CH3); HRESIMS calcd for C2oH2o06Na ([M+Na]+): 379.1158; found 379.1160. Garvin C (56): for a summary of JH and 13C NMR assignments based on HMQC and HMBC data, see Table 3.5. Pale yellow oil; !H NMR (CDC13, 600 MHz) S 10.9 (IH, s), 7.15 (IH, s), 7.06 (IH, s), 4.71 (IH, m), 4.31 (IH, dd, 7 = 4.3, 17.9 Hz), 3.39 (IH, dd, 7 = 3.4, 18.2 Hz), 1.59 (3H, d, 7 = 6.4 Hz), 1.582 (3H, s), 1.576 (3H, s); 13C NMR (CDC13, 150 MHz) £178.5 (C), 169.9 (C), 155.7 (C), 140.3 (C), 137.8 (C), 136.8 (C), 134.1 (C), 116.1 (C), 112.6 (CH), 111.2 (CH), 110.5 (C), 75.4 (C), 43.9 (C), 33.7 (C), 25.1 (CH3), 25.1 (CH3), 20.6 (CH3); HRESIMS calcd for Ci8H1506 ([M-H]"): 327.0869; found 327.0866. I l l Chapter 4. Progress towards the Synthesis of Ceratamines 4. Progress towards the Synthesis of Ceratamines 4.1. Microtubule-stabilizing agents from marine origin The microtubule-stabilizing properties of taxol (1) were first reported in 1979,202 and for the next 16 years 1 and its analogues were the only compounds known to exhibit such activity. However, since 1995 a variety of natural products from plant (2, 3), bacterial (4-5) and marine (6-12) origins have been recognized as microtubule-stabilizing agents (MSA). 203-209 Plant origin OAc \ HO 6 ° v OH  0=( y- o 1 Taxol O A C ; 2 Taccalonolide A - H -s BzO OAc 3 Jatrophane ester H O ^ °̂ Jd A„ k^k^ ^ > 0 ^ 0 O OH O 4 Epothilone A Bacterial N origin = H O s / V ^ O H''y'" ' \ u u I MY 5 Cyclostreptin Marine origin OH O / ^ O ^ O A c f O H OH 8 Eleutherobin D — 9 R=CH3 , Sarcodictyin A 10 R=C 2H 5 , Sarcodictyin B 11 Laulimalide 12 Peloruside A Figure 4.1. Prominent microtubule-stabilizing agents (MSA) from diverse sources. 203,210 112 Chapter 4. Progress towards the Synthesis of Ceratamines It is noteworthy that all potent microtubule stabilizers known to date are either natural products, or derived from natural products leads.203 Furthermore, the majority of these prominent compounds are marine-derived. Discodermolide (6), a polyhydroxy-8-lactone of polypropionate origin exhibiting immunosuppressive activity, was isolated from the sponge Discodermia dissoluta in 1990 by Gunasekera.211'212 The polyketide-derived macrolide dictyostatin-1 (7), was first reported by Pettit and coworkers213 in 1988 from a marine sponge of the genus Spongia sp. collected in the Republic of Maldives. The compound showed potent growth inhibition of lymphocytic leukemia cells. The soft coral Eleutherobia sp. found in waters of Western Australia yielded in 1997 a tricyclic triterpene christened eleutherobin (8) by Lindel and colleagues. This compound was re-isolated in 2000 by Andersen and coworkers215'216 from the octocoral Erythropodium caribaeorum, along with several analogues. E. caribaeorum is more abundant and represents a better source of 8 than Eleutherobia sp. Its potent tubulin polymerization activity was recognized shortly after the first isolation.217 The discovery of microtubule-stabilizing properties of eleutherobin (8) triggered a more serious and intense search for non-taxane-based MSA's, which eventually led to sarcodictyins A (9) and B (10). Both metabolites had been isolated 10 years before 8, in 1987 from the Mediterranean stolonifer Sarcodictyon roseum by Pietra and coworkers.218 As in the case of sarcodictiyns, it would take several years (even decades) before the initially reported biological activities of discodermolide (6) and dictyostatin-1 (7) could be traced to their interaction with the tubulin-microtubule system. ' The last member of this group of rediscovered secondary metabolites is laulimalide (11), reported in 1988 by two research groups working independently. 99] Crews isolated 11 from the sponge Spongia mycofijiensis collected in Vanuatu, whereas Moore and Scheuer222 found it in extracts of the Indonesian sponge Hyatella sp. The compound was reported as a potent antiproliferative agent, and in 1999 confirmed as a MSA by Mooberry 113 Chapter 4. Progress towards the Synthesis of Ceratamines and colleagues.223 Laulimalide (11) was found unable to displace radiolabeled taxol, and therefore does not bind to the taxol site on |3-tubulin.224 More recently, another polyketide-based macrolide, peloruside A (12), was isolated by Northcote and colleagues from the New Zealand sponge Mycale sp. Miller and coworkers reported that 12 promotes tubulin polymerization in vitro with essentially the same activity as taxol (l).226 It is the second MSA, after laulimalide (11), demonstrated to bind at a site different from taxol (1). 227 13 R=CH3, Ceratamine A 14 R=H, Ceratamine B OH 15R1=H, Ft2=\- 16 R,=H, R2= V ^ \ ^ ^ - ORi , Nigricanoside A , Nigricanoside B The latest additions to the expanding group of potent MSA's from marine organisms are metabolites (13)-(16), discovered in the Andersen research laboratory. The heterocyclic alkaloids ceratamines A (13) and B (14) were isolated in 2003 from the sponge Pseudoceratina sp. collected in Papua New Guinea. ' Similarly, specimens of the green algae Avrainvillea nigricans harvested from reef flats near Portsmouth, Dominica, yielded in 2007 the antimitotic glycoglycerolipids nigricanosides A (15) and B (16) as the corresponding dimethyl esters.228 Herein, details regarding several attempts to synthesize ceratamines will be described. 4.2. Ceratamines A and B The heterocyclic alkaloids ceratamines A (13) and B (14) are the first examples of a new family of MSA's for which tubulin-polymerizing activity has been firmly established. ' ' 114 Chapter 4. Progress towards the Synthesis of Ceratamines Both compounds are achiral and present significantly less elaborate structures than other polyketide and terpenoid-based marine MSA's (Figure 4.1). In addition, the imidazo[4,5,<i]azepine core heterocycle of the ceratamines has no precedent among known natural and synthetic compounds to date.210 The metabolites were discovered in a cell-based screen for mitosis inhibitors that uses human breast carcinoma MCF-7 cells. In the assay, arrested cells are detected by ELISA using the monoclonal antibody TG-3.230 Treatment of MCF-7 cells with a 20 (iM concentration of 13 caused readily detectable changes in the interphase microtubule network, leading to accumulation of microtubules around the nucleus.230 The same dose also led to cell cycle arrest in mitosis, where arrested cells displayed pillar-like tubulin structures extending vertically from the basal surface of the cells. These observations are qualitatively different from those described for taxol (1), and confirmed ceratamines as microtubule-stabilizers that target microtubules directly. The polymerizing power of ceratamine A (13), the most active analogue, is around 10-fold lower than taxol (l).203'229 Additionally, 13 was found not to compete with 1 for binding to microtubules, suggesting a different binding site.229 Such unusual characteristics make ceratamines attractive as research tools and experimental drug candidates. They might act differently on normal and cancer cells than other MSA's, and show a distinctive spectrum of toxicity and antitumor activity. A limiting factor in the further development of ceratamines A (13) and B (14) as leads for new anticancer drugs is their availability, since only 8 and 14 mg were isolated, respectively, from a relatively rare sponge. Clearly, total synthesis of the natural product and preparation of analogues is imperative to solve the supply issue. 115 Chapter 4. Progress towards the Synthesis of Ceratamines 4.2.1. Biosynthesis and similar marine metabolites The biosynthesis of ceratamines may involve bromination and methylation of a histidine- tyrosine dipeptide (17), followed by amination and oxidation to generate common precursors in the biogenesis of pseudoceratinine B (18), ianthelline (19), 5-bromoverongamide (20), and pseudoceratinine A (21),231 also isolated from Pseudoceratina sp. sponges collected off New Caledonia, Bahamas and Curacao. 18 Br 19 R=NH2 20R=H 13 R=CH3 14R=H Scheme 4.1. Proposed biogenesis of Pseudoceratina sp. secondary metabolites. 116 Chapter 4. Progress towards the Synthesis of Ceratamines Oxidative decarboxylation and oxime hydrolysis lead to intermediates 22 and 23, which may eliminate H2O or NH2OH via nucleophilic addition to a carbonyl (or oxime) in an aldol-like fashion (Scheme 4.1). The driving force for this last reaction is a complete aromatization of the core heterocycle in the final product. Besides 18-21, the ceratamines share a biogenetic relationship with at least 30 bromotyrosine/histidine-derived metabolites isolated from marine sponges belonging to the order Verongida, which include among others, the genera Aplysina, Ianthella, Psammaplysilla, Pseudoceratina, and Verongula. ' These compounds are associated with the large number of chemical variations possible within the aromatic ring systems and side chains of the tyrosine and histidine moieties. 24 verongamine 25 R=NH2> purealidin J 26 R=H, aerophobin-1 27 purealidin K 28 R=NH2, purealidin M 29 R=H, purealidin N 30 R=H, lipopurealidin H 31 R=(CH2)3NHCO(CH2)12CH3, lipopurealidin A 32 R=(CH2)3NH2, purealidin A 33 R=(CH2)3C5H5N+, purealidin D 34 R=(CH2)3N+(CH3)3, purealidin E 35 purealin NH 2 117 Chapter 4. Progress towards the Synthesis of Ceratamines The aromatic system can be either oxygenated (28, 29),236 monobrominated (24)237 or dibrominated. The bromotyrosine moiety can undergo rearrangement to a spirohexadienylisoxazole system (25-27, 35), " or link to form linear chains through ether bonds (31-35).236'239"243 The imidazole ring in histidine may be left unchanged (24),237 but 'jACi OMO 04"^ 01f t amination is commonly observed (25, 28, 30-35) ' ' and oxidation is also possible (27). Bromotyrosine-derived metabolites have long been noted as distinct markers for marine sponges belonging to the order Verongida, which are rather variable in color, shape, consistency, and skeletal arrangement.234 These compounds exhibit a wide range of interesting biological activities including antiviral, antibiotic, Na+/K+ ATPase inhibition, anti-HIV, antifouling, anti- inflammatory, and anticancer.234'235'238,244 The 2-aminoimidazole moiety in ceratamines is also a characteristic structural element of the pyrrole-imidazole alkaloids in the oroidin family, some of the most common metabolites isolated from marine sponges (Figure 4.2).245 They range in complexity from the chlorohydrin derivative girolline (36)246 to the tetrameric stylissadine A (40).247'248 The non-cyclized precursors for this small (approximately 100 members) but highly diverse family of sponge alkaloids are oroidin (37) and congeners (hymenidin 38, clathrodine 39).249 In addition to their ornate structures, these molecules also possess a broad range of biological activities. Oroidin (37) is the major fish feeding deterrent agent of sponges of the genus Agellas, and thus, secures their survival.249 Sceptrin (41) is a potent antibacterial, as well as an antiviral, antihistaminic, and antimuscarinic agent.250 Ageliferin (43) also possesses antibiotic and antiviral activities, and is an useful compound for the study of actin-myosin contractile systems.250 Palau'amine (42) exhibits potent immunosuppressive activity,249 while grossularines A (44) and B (45) possess pronounced effects against solid human tumor cell lines.251 118 Chapter 4. Progress towards the Synthesis of Ceratamines OH H 2 N - ^ . N - ^ ^ Y ^ N H 2 H N - ^ CI 36 37 X=Y=Br 38 X=Br, Y=H 39 X=Y=H Br Br Br Br >=rBr NH N H , 40 41 N H 2 Br O^ . N ^ _ \ H V N < f ^ H N ^ NH 2 N H 2 42 OH H2N HNL A> HN ) = N B, ° ^ 43 44 R=CH3 45R=H Figure 4.2. Representative pyrrole-imidazole alkaloids. 4.2.2. NMR data The NMR data in Figure 4.3 shows evidence of two conformers for ceratamine A (13). This is particularly apparent for the resonances assigned to H13/H17 (5H 7.67, 5c 133.2), H9 (5H 6.42, d, J = 10.0 Hz, 5c 100.4), and H8 (6H 7.73, d, J = 10.0 Hz, 5c 142.9), all in a peak ratio of 4:1. Scalar coupling between 18NH (5H 8.69) and H19 (5H 3.07, 5c 29.2) can be appreciated in both rotational isomers. 119 Chapter 4. Progress towards the Synthesis of Ceratamines H21 19\ N; H N - ( N 3 L ™- H13 H17 H20 H8 «W£ H18 9.0 Hll H9 I T | i 1 n 10 yffyjiyy J«U f T I I I I I I ' l l 1 .0 6.0 *|W^lltM*>WW«WIW>MW llftM"WM1>*'*****1 J // H19 5.0 4.0 (ppm) t~' ) • r- rT-T-T-r-r-r i i" i T - n T " ' '' » -r-T-T-T-T-f""r,"t",T""r-"T"T"T" " l"TTT"]""l I' )-t-|-T""r"T"f-j""f"T""r,"T",r"T",TgT 3.0 2.0 1.0 W _,—|—,—| ,—,—,—, ,—! ,—! ,—,—| 1 — | 1 — , ,—, !—,  r — I — i — i — i — i — i — i — i — i — i — i — i — i — i — r 60 40 20 0 220 200 180 160 140 120 100 80 (ppm) Figure 4.3. 'H and 13C-NMR spectra of ceratamine A (13) (recorded in DMSO-d6 at 500 and 125 MHz respectively). 210 120 Chapter 4. Progress towards the Synthesis of Ceratamines Evidently, the C2-N18 imine bond possesses a high double bond character which gives rise to restricted rotation around this bond (Scheme 4.2). The case is comparable to a vinylogous amide. Br / Scheme 4.2. Double bond character of the C2-N18 imine bond of ceratamine A (13). 210 The NMR data for ceratamine B (14) also presents two sets of NMR resonances in the same ratio as 13. When methyl HI9 is removed as in recent synthetic analogues, only one set of NMR peaks can be observed. Nevertheless, both amine protons are distinct in such derivatives, suggesting again restriction in the free rotation of this functionality. 4.2.3. Retrosynthetic analysis Considering the proposed biogenetic origin of ceratamines (Section 4.2.1), disconnection I (Scheme 4.3) seemed a logical starting point. Synthones (46) and (47) may be prepared from phenylpyruvic acid and either histidine or histamine. Disconnection II leads to guanidine (48) and synthon (49), which may be obtained via alkylation of azepanediones (50) or (51) with a benzyl bromide derivative (52). Preparation of azepanediones (50) and (51) may be possible via Beckmann rearrangement of a suitable cyclic diketone. 121 Chapter 4. Progress towards the Synthesis of Ceratamines H \ N— HN—4 1 N " 7 o HN—<^ N- Br / H N - H2N—^ I NH 2 46 H N 4 i N N H 2 NH 2 OH OH H 2 N ^ , N H 2 48 NH Xc- Br / Br / 49 50 or Q = \ y ^ O 51 Br Scheme 4.3. Retrosynthetic analysis of ceratamine A (13). Clearly, synthetic methodologies leading to the preparation of 2-aminoimidazoles, either from guanidine or starting with a preformed imidazole ring, will determine the viability of any proposal for the preparation of ceratamines. Besides possessing the common aspects of an efficient synthetic pathway (minimum amount of steps, good yields, commercially accessible starting materials, inexpensive), the synthesis of ceratamines should be designed to allow generation of analogues required for structure-activity relationship studies. 122 Chapter 4. Progress towards the Synthesis of Ceratamines 4.3. Related syntheses and relevant synthetic methodologies: literature review Among the bromotyrosine/histidine-derived metabolites biogenetically related to the ceratamines (Scheme 4.1), only verongamine (24) has been synthesized. In 1988, Wasserman and Wang244 developed an efficient method for producing a-keto amido residues, precursors of the oc-oximido units common in these compounds (Scheme 4.4). The procedure involves conversion of a carboxylic acid to an acyl cyano phosphorane (53) which may be oxidized to a oc,P-diketonitrile, and afforded 24 in 6 steps with an overall yield of 31%. "xb 1. Br2, AICI3, CH2CI2 " l j" ^ Ph3P=CHCN, EDCI 2. NaOH, MeOH ^ T CH2CI2 , 8 2 % O ^ 73% <2 s t e e s > k^OH . c CN o 53 o I 1. 0 3 , CH2CI2 , -78 °C 2. histamine, f-BuOH 3. NH2OH.HCI, NaOAc, EtOH, 52% (3 steps) Scheme 4.4. Total synthesis of verongamine (24).244 Several syntheses of the 1,4-disubstituted imidazole alkaloid girolline (36) have been reported.252"254 Although 36 is structurally simple, the density of functionalities and rapid decomposition of intermediates provided an additional challenge during its preparation. In one of these methodologies, Al-Mourabit and coworkers exploited the fact that 2-aminoimidazoles can undergo nucleophilic addition in reactions analogous to those of enamides with aldehydes (Scheme 4.5).256 123 Chapter 4. Progress towards the Synthesis of Ceratamines C ' \ = =  + H O ^ N 1 " 1 _H2§Q«_ CI y ^ ^ 9 3 % o Scheme 4.5. Synthesis of girolline (36) by Al-Mourabit and coworkers. Synthetic transformations directed to the preparation of the 2-aminoimidazole motif can be divided in two categories, de novo imidazole construction and imidazole elaboration. In the first category, classical approaches include the condensation of a-aminocarbonyl compounds with cyanamide (I), ~ combination of a-diketones with guanidine followed by reduction (II), ' and reaction of a-haloketones (III) and cycloheptatrienones (IV) with guanidine j • .. 264-266 derivatives. I) R 1 T ° _H2NCN » R V ^ N H 2 R / ^ N H , H20,pH1-4 p ^ r / -12 ™«2 • " 2 reflux NH R 2 ° rt, 0.5-24h R 2 ^ H R 2 NH U R _ H 2 N ^ N H R R H Ml) J R: Ac. B o c _ ^ > f ^ - N H R R ^ B r EtOH/DMF/H20 p ^ t f rt or reflux 2 24-96h NH  N H R 2 R2HN ,V) I, fl Rg LHoM !h_^ / T V + Y^T u l l  KOH/EtOH, reflux \ _ _ / V tf or C 6 H 6 , reflux R,=OMe, SMe, H Scheme 4.6. Classical de novo preparation of 2-aminoimidazoles. 124 Chapter 4. Progress towards the Synthesis of Ceratamines Recently, methodologies I and III have been successfully applied in the total synthesis of sceptrin (41),250'267 ageliferin (43),268 grossularines A (44) and B (45),251 naamidine A (54),269 970 971 977 ageladine A (55), and other similar marine metabolites. o 11 / N, N - •N H *>—NH2 H OH 278 54 55 A different approach was developed by Molina, Fresneda and Sanz/'  who synthesized isonaamine A (58), dorimidazole A (59), and preclathridine A (60), via a Staudinger/aza- Wittig/carbodiimide-mediated cyclization process (Scheme 4.7). The one-pot conversion of oc- azido ester (56) to imidazolone (57) proceeded in a 50-60% yield. Ar cr 56 N3 PPh3 Et 20, rt nnQ MOM A r ^ p A Ts-NCO , A = P P h 3 EEtzO, rt Ar RNH2 Ar N=C=N—Ts ElzO, 0 °C N = C C ,N—Ts N—R H I Et 2 0, rt, 57% MeS02CI r V H I y— N—Ts ^ N CH2CI2 , R ^ M L A r ^ N H ^ ^  p | B A L A r ^ T - N H ^ rt, 98%  H r A N T H F . reflux rr^N H O  h 57% U * BBr3 / CH2CI2 , reflux, 98% 57 -Ts HO S m l 2 / T H F HO' reflux, 76% ^ ^ \ 59 N HO 58 Y__ L > - N H 2 60 Scheme 4.7. Construction of 2-aminoimidazoles according to Molina, Fresneda and Sanz. 278 125 Chapter 4. Progress towards the Synthesis of Ceratamines Another procedure was reported by Al-Mourabit and colleagues279 in their synthesis of (Z)-3-amino-l-(2-aminoimidazol-4-yl)prop-l-ene methyl carbamate (64), closely related to metabolite (65) isolated in 1991 from Axinellidae sponges.280 Addition of Boc-guanidine (62) to ./V-carbomethoxy-l,2-dihydropyridine (61) in the presence of Br2 afforded aminal (63), which upon deprotection was cleaved under basic conditions to afford 64. The yield for this last step is dramatically time and temperature dependant, and the instability exhibited by 64 under basic conditions limits its preparation in large quantities. Recently, an improved version of this method was applied in the synthesis of hymenidin (38).281 NH U H2N NHBoc 6 2  R2 O NaBH4 , CICQ2Me ^ | f ^ | DMF/CH3CN ^ R H N / J ^ T ' ^ | MeOH, -78 °C, 2h k M ^ Br2, rt, 15min , 67% 1 N - " ^ N ^ 6 4 % COoMe CO J 2 I 61 63 — M e R ^ B o c , R2=H R i=H, R2=Boc 2M HCI ,100 °C, 5 min, quant. H x . N ~ 7 r ^ ^ 1 1M N a O H / N * ~ r ^ ^ l H H N 23-85% H 7 C 0 2 M e C 0 2 M e 1M NaOH / \ 1M NaOH 25 min, reflux, 47% / x \ 4 8 h , rt, 30% JM3 H 2N. ...^^ , , ^ ^-- - N ' NH H Scheme 4.8. Synthesis of (Z)-3-amino-l-(2-aminoimidazol-4-yl)prop-l-ene methyl carbamate (64) according to Al-Mourabit and colleagues.279 Only three methods, starting from preformed imidazole rings, have been reported for the direct introduction of an amino functionality at position 2: coupling with arene diazonium salts and further reduction (I, Scheme 4.9),282 bromide-mediated or direct metalation followed by 126 Chapter 4. Progress towards the Synthesis of Ceratamines sequential treatment with aryl azide, acid or hydrogenation (II), " and oxidation of a sulfur substituent at position 2 followed by treatment with ammonia (III), or azidation-hydrogenation sequence as in II.287 N <' N H R NHAc N 2 Br R NHAc H20, 0 °C, 2h R=COOCH3 R=H H N=N. 1 3 % Br N " N H N ^ N Br AcHN R 82% 0' Br Br N" N HN X N N = N ' AcHN' "R 5% I H2, Pt H 2 N-^ N I N ' H R NHAc II) N - ^ M 1 N " ^ R , R2 P=H, Tr, S0 2 Me 2 p' 1. BuLi, -78 ° C , T H F ^ — ^ ^ ^ , - " 2 . TsN3 ~ \ N B S T H F " \ _ 0 ° C r f T ^ ^ c ^ H 2 N^- N T 1. NBuLi,-78 °C, THF 2. TsN3 N ^ - R i III) B r ^ - T N - ^ R 2 P / ^ T r " 0 1 Oxone/MeOH+H20 CU/? 2J~7T'Rl S—Cf II *- ^S—<' II M - ^ r , or m-CPBA/CHpClp / N--^-r, NaN3 DMF or DMSO N - ^ R l N 3 ^ ' T / P P=BOM, PMB N H 3 \ MeOH R 2 H2, Pd/C EtOH / ^ R 2 P Scheme 4.9. Synthesis of 2-aminoimidazoles via elaboration of a preexisting imidazole ring. The initial preparations of oroidin (37) ' and its methyl analogue keramadine (66), as well as the more recent total syntheses of naamidine A (54), ageladine A (55), ' chartelline C (67),293 naamine C (68), and pyronaamidine (69),294 were executed using methods II and III. 127 Chapter 4. Progress towards the Synthesis of Ceratamines 66 67 4.4. Initial synthetic proposals The first route envisioned to build the new imidazo[4,5,<flazepine core of ceratamines follows very closely their proposed biogenesis (Scheme 4.1), and employs L-histidine methyl ester (70) and phenylpyruvic acid (71) as starting materials (Scheme 4.10). N H O C H 3 + NH 2 " 70 BrC 6 H 4 — N •  ft N—^ NH O OH DCC or PyBOP I H OCH3 NH 71 72 BrCgH4— JSJ p -BrC 6 H 4 N 2 + O LiOH, H zO, THF ™~~-<tlf^\^ OCH3 N - ^ NH 74 I 1) Pb(OAc)4, Cu(OAc)2 , Pyr, THF I 2) UCIO4, DIEA, THF o- 73 H-1* NH H 2 / P d / H '  W  n II . N ^ H Q 75 Scheme 4.10. Biomimetic synthetic pathway for the preparation of ceratamine analogue (77). 128 Chapter 4. Progress towards the Synthesis of Ceratamines Thus, standard peptide coupling between starting materials (70) and (71) was expected to afford intermediate (72), which under treatment with an aryldiazonium salt,282 would form 73. Ester hydrolysis and Pb(OAc)4/Cu(OAc)2 mediated oxidative decarboxylation lead to 75, while hydrogenation conditions cleaves the diazo linkage to afford a 4-substituted 2- aminoimidazole ring. In the last step, an intramolecular base-promoted condensation as in the synthesis of girolline (36) by Marchais and coworkers,255 would afford ceratamine analogue (77). As mentioned in Section 4.2.1, the driving force for the formation of 77 is achievement of aromaticity. Parallel to the biomimetic route above, a more classical pathway was also implemented (Scheme 4.11). Lithiation of azepane-2,4-dione (50) and treatment with benzyl bromide was expected to form 78, which upon a-halogenation with excess of base and NBS would afford dibrominated intermediate (79). The bromine in C5 sets the stage for condensation with N- acetylguanidine (80) according to Little and Weber, whereas bromine at C3 will serve as a leaving group assisting the DDQ-mediated aromatization of the entire carbon backbone. / \ , H 1)LDA/THF o ^ ^ A o 2) BzBr 50 oc-Halogenation LDA excess / T H F NBS excess Jl HCI H o 80 U NH 2 NH 2 EtOH, reflux 77 Scheme 4.11. A more classical first generation approach to analogue 77. 129 Chapter 4. Progress towards the Synthesis of Ceratamines Both synthetic sequences can be applied in the preparation of analogues with several substitution patterns in the benzylic side arm, or even completely different alkyl groups at C3, by employing other commercially available pyruvic acids (Scheme 4.10) or alkyl halides (Scheme 4.11). The second route is more flexible if ring size or heteroatom content is to be varied. For example, the use of a six-membered analogue of 50 can be easily implemented. Replacement of TV-acetylguanidine with acetamide (CH3CONH2) and thioacetamide (CH3CSNH2) should incorporate oxygen and sulfur into the bicyclic system. 4.5. Attempted preparation of ceratamines and analogues 4.5.1. Biomimetic approach to ceratamines Initial DCC, CDI and PyBOP-mediated coupling reactions between L-histidine methyl ester and phenylpyruvic acid (Scheme 4.10) were fruitless. Examination of NMR data revealed in all reaction crudes the presence of an enolized phenylpyruvic acid, also detected as the major constituent in the starting material. Unlike pyruvic acid, enolization of phenylpyruvic acid leads to a stable conjugated system, which seems to prevail over the ketone form depending on the solvent.296299 O OH O O ^ 5 ^ - ^ ° CHCI3 Such conjugation decreases reactivity of the carboxylic group, which cannot be activated by standard peptide coupling reagents. Additionally, the presence of a P-hydroxyl group may not be compatible with their action. It was thus necessary to build a carbonyl-protected version of phenylpyruvic acid, capable of coupling with L-histidine methyl ester. Bates and Ramaswamy300 130 Chapter 4. Progress towards the Synthesis of Ceratamines prepared several protected cc-keto acids, by alkylating the lithium dianion derived from glyoxilic thioketal (82) with various alkyl halides (Scheme 4.12). This methodology afforded analogues (83) and (84) with high purity, in a simple, fast and high yielding fashion. Furthermore, it added flexibility in the future preparation of ceratamine variants via the biomimetic approach. The carbonyl can later be deprotected by acid hydrolysis.300'301 o o HO. JL H S — ' SH 9 1)LDA, THF-78°C \ ~ - £ S \ JL. I H TsOH, C6H6, reflux V - S - ^ ^ ^ O H 2) BnCI or Mel (^ 80%  8 2 R 83 R=Ph 89% 84 R=H 72% o <*NTT ^ r "ocH3 H N - " NH2 PyBOP, DIEA, CH2CI2 H N - ^ NH ^ NaOH H Y J J / ^ H 0<KJtsC7 Dioxane 0<KJtsZ7 FT FT 87 R=Ph 0% 85 R=Ph 50% 88 R=H traces 86 R=H 53% Scheme 4.12. Reactions involved in the biomimetic approach to ceratamine analogues. "levy PyBOP-mediated coupling of L-histidine methyl ester with 83 or 84 afforded the desired adducts in 50% and 53% yields, respectively, which were confirmed by HRESIMS and NMR analysis. The thioketal motif as a protecting group would be incompatible with the strong acidic conditions involved in preparing 2-aminoimidazoles via arene diazoniun salts (I, Scheme 4.9.), hence it was decided to proceed with the decarboxylation step. However, basic hydrolysis of the methyl ester in the histidine component also cleaved the amide bond, apparently weakened by steric as well as electronic effects from both sulfurs and the six- membered ring constituting the thioketal functionality (Scheme 4.12). Such surprising events, 131 Chapter 4. Progress towards the Synthesis of Ceratamines together with several failed attempts to separately functionalize the imidazole of L-histidine methyl ester, prompted immediate implementation of an alternative route to ceratamines. 4.5.2. First generation approach to ceratamines According to Scheme 4.11, access to an azepanedione similar to 50 was the first task that needed to be achieved. As mentioned in Section 4.2.3, Beckmann rearrangement of an appropriate oxime may be a viable method to generate 50. Several starting materials and reaction conditions were therefore tested and are summarized in Table 4.1. In principle, azepane-2,4- dione (50) and azepane-l,4-dione (51) can both act as a suitable precursor for constructing 2- aminoimidazoles. Table 4.1. Rearrangement of various 1,4- and 1,3-cyclohexadione derivatives. No S.M.' NOH NOH 6. Reagents and conditions CH3SO3H, NaN3, 25°C, 3 h.303"306 NH2OH.HCl, (COOH)2, 80°C, 12 h.303,304-307 PC15,25°C, 16 h.308'309 AlClj, 60°C, 30 min.309'310 PPA,b 100°C, 4 h.309 0.5 eq. BF3.Et20, 25°C, 12 h.311 1 eq. BF3.Et20, 25°C, 12 h.3" leq.BF3.Et20,250C,48h.311 1.5eq.BF3.Et20,25°C, 12 h.311 2 eq. BF3.Et20,25°C, 12 h.311 3 eq. BF3.Et20, 25°C, 12 h. 311 311 4 eq. BF3.Et20,25°C, 12 h. 32 mmol, 120 g PPA115°C, 1 h, CH2C12 extractions. 312,313 6.6 mmol, 30 g PPA, 115°C, 1 h, CH2C12 extractions.312*313 37.0 mmol, 100 g PPA, 115°C, 1 h, EtOAc extractions.312,313 o--~-v. 24.5 mmol, 100 g PPA, 115°C, 1 h, BuOH extractions 312,313 Observations0 Multiple byproducts Multiple byproducts Traces (51) 1,4-cyclohexadione Traces (51) S.M. S.M. Deprotection S. M., deprotection S. M., deprotection Multiple byproducts S. M., deprotection 17% yield (50) 21% yield (50) 60% yield (50) 89% yield (50) aS. M: starting material. PPA: polyphosphoric acid. cAccording to NMR and MS data, or isolated yield. 132 Chapter 4. Progress towards the Synthesis of Ceratamines Entry 1 (Table 4.1) actually corresponds to Schmidt305'306 rearrangement conditions (I, Scheme 4.13), which employ NaN3 and CH3SO3H to generate hydrazoic acid (HN3) in situ. Entries 2-5 represent several reaction conditions for Beckmann rearrangement (II). " The best results were obtained when polyphosphoric acid (PPA) was used to catalyze the process. Clearly, azepane-2,4-dione (50) is quite water-soluble and must be extracted from the neutralized aqueous layer during work-up with very polar organic solvents, in order to minimize losses. All - 7 1 0 NMR and MS data were in accord with that previously reported for 50. I \ H ^\ \ H C X j ^ N - y = N 15 31 90 P= 0 N^-NSN Qp3 p + H 2 0 : O H 2 • N ^ \ - H * V5 37 P N 2 ,0 P 91 II PPA work-up 9 2 ^ ^6^bN work-up PPA 94  :OH 2 « t "- ^ ^ : O H 2 93 A - - E t O H » ^ N H NH 50 Scheme 4.13. Mechanisms for Schmidt (I) and Beckmann (II) rearrangements. Noteworthy, only one rearrangement product was obtained when cyclohexan-l,3-dione was employed as starting material (II). According to the classical Beckmann mechanism, any of the alkyl groups flanking the carbonyl can migrate (Scheme 4.13). Moreover, Tamura and T I T coworkers reported that the migratory aptitude of both alkyl substituents during the Beckmann 133 Chapter 4. Progress towards the Synthesis of Ceratamines rearrangement of oc,(3-unsaturated ketones is difficult to predict confidently. The symmetric cyclohexan-l,4-dione derivative (90) was expected to furnish only one rearrangement product (I). A possible explanation for the observed regioselectivity leading to 50 may involve the coordination of both the oxime's oxygen and the methoxy substituent at C3 with PPA, which acts as an anchor favoring the oxime's yyrc-isomer (94) over the anri-isomer (92). The former possesses an appropriate alignment between the acting nucleophile C6 and the leaving group attached to the nitrogen atom, for an intramolecular SN2-like process. Lewis acids assist in the departure of the leaving group. Lithiation of azepane-2,4-dione (50) and treatment with benzyl chloride yielded the key intermediate 3-benzylazepane-2,4-dione (78) in 35% yield. The C3-dibenzylated byproduct (95) was also detected in the reaction crude, and accounts for the low yields obtained in this alkylation. When benzyl bromide was used as electrophile, the amount of dibenzyl byproduct increased. Evidently, after monobenzylation takes place, proton H3 of the newly formed 3- benzylazepane-2,4-dione (78) (pKa~ll) is almost as acidic as the corresponding H3 of azepane- 2,4-dione (50) (/?Ka~10)298'299, and reacts easily with LDA giving rise to dibenzylation. When a more reactive electrophile such as BnBr is employed, the process is accelerated and the consumption of 78 increases. 50 pKa-10 1. LDA, -78 °C, THF 2. BnBr Slow Fast 95 134 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.2. NMR data for 3-benzylazepane-2,4-dione (78) (recorded in CD2C12). Carbon No Nl 2 3 4 5 6 7 8 9 10 11 12 1 3 C 8 (ppm)a 169.5 59.4 204.2 46.3 30.7 41.9 31.8 140.5 129.6 128.8 126.7 XH 8(ppm)(mult,7(Hz))b'c 6.11 (s, broad) 4.24 (dd, 7 = 6.7, 6.8 Hz) 2.60 (m) 1.92 (m), 2.10 (m) 3.31 (m), 3.75 (m) 3.15 (m) 7.16 7.24 7.22 HMBCb (H->C) C2, C4, C8, C9 C3, C4, C7 C4.C5 C2,C5 C2, C3, C4, C9, C10 C8, C12 C9 C10 Recorded at 100 MHz.b Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. Positive HRESIMS provided a [M+Na]+ ion at m/z 240.0996 consistent with a molecular formula of Ci3H15N02 (calculated for C13H15N02Na: 240.1000). NMR data (Table 4.2, Figure 4.4) shows the same coupled spin system H5/H6/H7/NH as in the starting material (50), and a new methine H3 (5H 4.24, dd, J = 6.7, 6.8 Hz, 5c 59.4) coupled to the benzylic methylene H8 (5H 3.15, m, 8c 31.8). The new stereocenter makes methylenes H5/H6/H7 diastereotopic, with protons in H6 (8H 1.92/2.10, 5C 30.7) and H7 (5H 3.31/3.75, Sc 41.9) showing important differences in chemical shifts. H3 displays HMBC correlations with both carbonyls C2 (5c 169.5) and C4 (5C 204.2), as well as C8 and the aromatic quaternary carbon C9 (5c 140.5). H8 in turn showed HMBC cross-peaks with neighboring carbons C3 and C9 (5c 140.5), carbonyls C2 and C4, as well as aromatic C10 (5c 129.6). 135 Chapter 4. Progress towards the Synthesis of Ceratamines H10 I H15 H3 H7' H7 H8 4.00 3.75 3.50 3.25 H3 H5 H5' H1 <l / H7' H7 H6' H6 _^_«L 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 L5 L0 0.5 Chemical Shift (ppm) C10 C10' C11 C11' C12 C4 C2 C9 iifciiiniiiMiniwitiHiiirtuiM wimm 0* C3 t\m*4*mmw*»m\tmtMtmt C5 C6 C7 C8 HlMtH^milliHHillilMIMIlUI 200 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 0 Figure 4.4. !H and 13C-NMR spectra of 3-benzylazepane-2,4-dione (78) (recorded in CD2C12 at 400 and 100 MHz, respectively). 136 Chapter 4. Progress towards the Synthesis of Ceratamines The next chemical modifications aimed to sequentially eliminate acidic protons and install halogens at C3 and C5, having in mind the procedure reported by Little and Weber for building 2-aminoimidazoles (Scheme 4.6). cc-Bromination at C3 was achieved by treatment of 78 with 2,4,4,6-tetrabromoquinone in acidic media.314'315 Likewise, lithiation using LDA and addition of NCS316'317 furnished the chlorinated intermediate 96 (Scheme 4.14), which was then methylated under standard conditions318 to afford 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97). Scheme 4.14. Synthesis of intermediates 97 and 98. The ^ -NMR data for 97 shows absence of the amide proton resonance (around 6 ppm), which was replaced by a new singlet for methyl H13 (5H 3.00, 8C 36.5) (Figures 4.5 and 4.6). More importantly, the now isolated methylene H8 (5c 44.0) displays two doublet resonances at SH 3.29 (J - 12.8 Hz) and 8H 3.83 (J = 12.8 Hz), typical for diastereotopic benzylic protons. 137 Chapter 4. Progress towards the Synthesis of Ceratamines H8 H8' H7' /"L H13 H7 H6 H6' H5' H5 3.5 3.0 2.5 H10 I H15 H7' H13 H7 H5' I \ J \ 3.0 H10 I H15 2.5 2.0 H13 H6 H6' H7 H5' H5I 2.0 H8' H8 H13 ii H6 H7JH7H5 H5Hb 5 4 3 Chemical Shift (ppm) Figure 4.5. H-NMR spectra of 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3- benzyl-3-bromo-l-methylazepane-2,4-dione (99) (recorded in CDC13 at 300 MHz). 138 Chapter 4. Progress towards the Synthesis of Ceratamines C10l C7  r 5 C6 C8 ° 5 C13 45 C4 40 35 30 C2 25  c LW C11 C12 C3 n C7 C6 C 8 ° 5 U L J C13 J i C4 C5 J C13 36.75 C2 C10 C11 C12 C9 JJJ C3 C5 C13 C6 C7 |C8 200 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 Figure 4.6.13C-NMR spectra of 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3- benzyl-3-bromo-l-methylazepane-2,4-dione (99) (recorded in CDCI3 at 75 MHz). 139 Chapter 4. Progress towards the Synthesis of Ceratamines Methylenes H5 (5H 1.90, 2.33; 5C 37.8), H6 (5H 1.75, 8C 24.2), and H7 (8H 2.85, 3.28; 5C 46.7) are also diastereotopic, but the difference in chemical shifts is less pronounced for H6 (flanked by methylenes, similar chemical and magnetic environments for both protons). Table 4.3 summarizes all the NMR assignments for 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97). Methylation and chlorination was also confirmed by positive HRESIMS, which produced a [M+H]+ ion at m/z 288.0764, in agreement with the molecular formula C14H16NO2CI (calculated for Ci4Hi6N02Na35Cl: 288.0767). Table 4.3. NMR data for 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3-benzyl-3- bromo-l-methylazepane-2,4-dione (99) (recorded in CDCI3). 97 Carbon No 2 3 4 5 6 7 8 9 10 11 12 13 13C 5(ppm)a 166.7 74.1 203.6 37.8 24.2 46.7 44.0 134.2 131.0 128.4 127.6 36.5 XH 8(ppm)(mult,/(Hz))b'c 1.90 (m), 2.33 (m) 1.75 (m) 2.85 (dt, J = 5.2, 15.7 Hz), 3.28 (m) 3.29 (d, 7=12.8 Hz), 3.83 (d,/= 12.8 Hz) 7.21 (m) 7.19 (m) 7.17 (m) 3.00 (s) 13C 8(ppm)a 167.2 69.1 203.1 37.0 24.3 47.3 44.4 135.2 131.1 128.6 127.8 36.9 "Recorded at 75 MHz.b Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. 99 lU 8(ppm)(mult,/(Hz)) b,c 1.95 (m), 2.48 (m) 1.75 (m) 2.76 (m), 3.14 (m) 3.50 (d,J= 12.7 Hz), 3.90 ( d , / = 12.7 Hz) 7.25 7.19 7.18 2.99 (s) 140 Chapter 4. Progress towards the Synthesis of Ceratamines Unfortunately, intermediates 97 and 98 proved remarkably intransigent, since attempts to functionalize C5 all failed. When 97 was treated with LDA and NCS a second time, only starting material was recovered. Replacing NCS by NBS furnished 3-benzyl-3-bromo-l-methylazepane- 2,4-dione (99) in 79% yield, easily identified by comparing its NMR data with that previously obtained for 97 (Figures 4.5 and 4.6, Table 4.3). Additionally, positive HRESDVIS afforded a [M+H]+ peak at 310.0439, suitable for the molecular formula Ci4Hi6N02Br (calculated for Ci4Hi7N0279Br: 310.0443). Reaction of this compound with f-BuLi (2 eq.) and NCS yielded back the initial chlorinated starting material (97). These last reactions showed that instead of proton abstraction at C5, 97 was undergoing lithium-halogen exchange at C3, and revealed the particular affinity of any lithium-based reagent towards this center. It is widely known that heteroatoms are able to direct the attack of lithium bases and promote regioselective deprotonation. They can coordinate with the organolithium reagent and increase kinetic basicity (A), or simply make nearby protons more acidic via inductive and/or resonance effects (B). A combination of both mechanisms is also possible ( Q319-321 ( T ^ r ^ N ^ „ R-Li r f ^ N ^ R-Li/TMEDA _ " " ' ^ f ^ ^ V ^ N ^ Upon LDA treatment, it is therefore reasonable to assume that the influence of both oxygen atoms at C2 and C4 leads preferentially to halogen-lithium exchange at C3, than the expected H5 abstraction. Furthermore, formation of an intermediate similar to enol (101) seems to be greatly disfavored in comparison with the conjugated and more stable enol (100), due to an 141 RH Chapter 4. Progress towards the Synthesis of Ceratamines increment in eclipsed interactions (H5<->H6<-»H7, Cl<->0) and ring tension, product of the additional induced planarity (Figure 4.7). 100 101 Figure 4.7. MM2 energy minimizations for enols (100) and (101). Arrows indicate eclipsed atoms (CS Chem3D Ultra 7.0, minimum RMS gradient 0.100). A bulkier bromine at C3 would provoke even more strain in the putative reaction intermediate (101), favoring again halogen-lithium exchange. In order to avoid this effect, an 0*7T ^ 0 1 T7A alternative route to functionalize C5 was evaluated. Several publications ' " suggest the use of hypervalent iodine-based reagents to modify asymmetric ketones at their cc-carbon. ^99 977 Particularly, Koser and coworkers, as well as Nicolaou and his group, achieved a-tosylation of ketones using the commercially available [hydroxyl(tosyloxy)iodo]benzene (102) and iodobenzene diacetate PhI(OAc)2, respectively. Presumably, reaction is initiated by electrophilic addition of (PhIOH)+OTs" (102) (Scheme 4.15) to the enol tautomer yielding oc-phenyliodonio ketones (104) as intermediates, which undergo nucleophilic displacement of iodobenzene by the tosylate ion, yielding the desired product.322 142 Chapter 4. Progress towards the Synthesis of Ceratamines T s O - ^ ; o ljL ° HO ^ B , _ « L . HO R3 _ ^ d % * _ ^ R A ^ P h + T s ? ^ H R2 R 1 ) R 2 " 1 _ | R 2 R2R3 * — ' f Ph— I—OH ! TsO" 102 1 0 3 TsCT- O p Ri^X°Ts + P h l -" R i - ^ uv / j P t l + H 2 0 R2 R3 R2 R3 104 322,325 Scheme 4.15. oc-Tosylation of ketones with hypervalent iodine reagent (102) However, neither 97 nor 99 underwent tosylation at C5 using the previous methodologies. It was then decided to trap enol (100) and use a strong base to removed H5 in the enolate itself. In principle, r-BuLi (pKa~53 for 2-methylpropane) would be a good candidate for proton abstraction at C5 (pKa~43 for propene).299 However, from the intermediate viewpoint and as seen previously, two more sp2 centers (five in total for the new enolate) in a seven-membered ring may not be viable. Nevertheless, the existence in nature of the ceratamine core itself was the main argument supporting such a forceful measure. After several failed attempts to produce enolate (100) from intermediate (97) using various methodologies, " the target compound was synthesized from 3-benzylazepane-2,4- dione (78) using TBDSCl, according to Scheme 4.16. Table 4.4 outlines all the NMR data collected for compound (106). NH ° 1 • LDA, -78 °C, THF / ^ S i - Q ^ ^ ^ * 0 1. LDA, -78 °C, T H F ^ ~ / " ~ S i - 0 ' 2. TBDSCl, 96% ~~ ' ) 2. Mel, 40% ' 105 106 Scheme 4.16. Preparation of 3-benzyl-4-(?m-butyldimethylsilanyloxy)-l -methyl-1,5,6,7- tetrahydroazepin-2-one (106). 143 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.4. NMR data for 3-benzyl-4-(fer?-butyldimethylsilanyloxy)-l-methyl-1,5,6,7- tetrahydroazepin-2-one (106) (recorded in CD2CI2). irbon No 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 3 C 5(ppm)a 172.6 117.2 154.1 30.8 29.0 48.4 32.9 142.0 128.9 128.7 126.0 34.5 -3.23 18.6 26.0 XH 8(ppm)(mult , / (Hz))b ' c 2.36 (t, J = 7.2 Hz) 2.00 (m) 3.24 (m) 3.67 (s) 7.25-7.15 (m) 7.25-7.15 (m) 7.12 (m) 2.94 (s) 0.22 (s) 0.96 (s) HMBCb (H-*C) C3, C4, C6, C7 C4, C5, C7 C2, C5, C13 C2,C3,C4,C9,C10,C11 C8 C8 C2.C7 C15 C15 "Recorded at 100 MHz." Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. In comparison with Figures 4.5 and 4.6, NMR data for 106 was significantly simplified when planarity was induced in the molecule. The !H-NMR spectrum (Figure 4.8, Table 4.4) displayed four singlet resonances, two of them typical for methyls H14 (8H 0.22, 5c -3.23) and H16 (§H 0.96, 6c 26.0) in the J-butyldimethylsilanyloxy moiety. The remaining ones can be assigned to methyl HI3 (SH 2.94, 5c 34.5) and methylene H8 (5H 3.67, 8C 32.9). HI3 presented HMBC correlations with amide carbonyl C2 (5C 172.6) and methylene H7 (5H 3.24, 5C 48.4), whereas H8 correlates with carbons in the aromatic and seven-membered rings, among them the vinylic C3 (5C 117.2) and C4 (5C 154.1), attached to an oxygen. 144 Chapter 4. Progress towards the Synthesis of Ceratamines H7' H7 M M H10, H10' H11,H11' H12 3.0 2.5 2.0 H13 H8 H7' H7 H16 H14 5 4 3 Chemical Shift (ppm) C2 C4 |IMHMHI>lliHlliiWu'liHI>l1l>lllWl|ii||ll m»#l^li|Ml<llilH' C14 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 Figure 4.8. *H and 13C-NMR spectra of 3-benzyl-4-(/er?-butyldimethylsilanyloxy)-l-methyl- l,5,6,7-tetrahydroazepin-2-one (106) (recorded in CD2CI2 at 400 and 100 MHz, respectively). Chapter 4. Progress towards the Synthesis of Ceratamines Finally, HRESIMS displayed a [M+Na]+ ion at m/z 368.2028, which was assigned the molecular formula C2oH31N02Si (calculated for C2oH3iN02NaSi: 368.2022). Interestingly enough, treatment of 106 with LDA at low temperature followed by NCS addition furnished only one main reaction product displaying a [M+Na]+ ion at m/z 402.1633, which was in agreement with a molecular formula of CaotboNC^SiCl (calculated for C2oH3oN02NaSi CI: 402.1632). Clearly, a proton had been replaced by chlorine as expected, but not with the desired regiochemistry. Examination of the apparently simple ^ -NMR spectrum for this reaction product (Figure 4.9) showed a new singlet H8 (5H 3.02, 8c 45.6) and a more disperse aromatic region (5H 7.75-6.75 ppm) than the starting material (Figure 4.8). HMQC correlations (Table 4.5) allow assignment of protons for methylenes H5 (SH 1.63/1.80, 5c 39.2), H6 (6H 1.60/1.66, 5C 24.2) and H7 (8H 3.11/3.79, 8C 47.8) in the seven-membered ring, the now disperse aromatic proton resonances, especially H10 (8H 7.58, 8c 132.5) and H14 (8H 6.37, 8c 134.7), as well as methyls H16 (8H 0.14, 8C -2.89), H17 (8H 0.33, 8C -1.41) and H19 (SH 0.80, 8C 27.8) in the f-butyldimethylsilyl group. The remaining carbons had chemical shifts typical for a ketone (C4, 8c 206.4), an amide (C2, 8c 168.1), a chlorinated quaternary carbon (C3, Sc 77.6) as in 96 (Figure 4.6, Table 4.3), a quaternary sp2 (C9, 8c 138.6) and a quaternary sp3 (C18, 8C 19.4). Thus, the /-butyldimethylsilyl motif was linked to neither the seven-membered nor the aromatic ring, leaving methine C8 as the only possible option. This was confirmed by HMBC correlations between H8 and carbons C16/C18, as well as between methyls H16/H17 and C8 (Table 4.5). H8 also gave HMBC cross peaks with numerous carbons in both rings, including carbonyls C2 and C4, as well as the ortho C10 and C14. The proximity between benzyl and t- butyldimethylsilyl moieties in 107 explains the newly acquired differences in chemical shifts for methyls H16 and H17 usually indistinguishable by NMR, as well as the pronounced dispersion of the aromatic resonances. 146 Chapter 4. Progress towards the Synthesis of Ceratamines H19 H10 H12 H11 H13 H7' J I H14 ^w~^-^-r^\U*«J»iNf^«** H6' H6 H5' H5 H15 7.50 7.25 7.00 3.75 3.50 3.25 3.00 1.9 1.8 1.7 1.6 1.5 H8 H17 H16 _AAU \1 i C12 C9 iili till i C13 C10 \ ° 1 1 C14 ill, iiill Ik Jjlli L I , I 5 4 3 Chemical Shift (ppm) 6 X N ^ 1 5 W.™ J ' I I, J " " i ' " I T f ' iHI uLiy,Mnijiu,iL,Ji,, 135 130 125 C 1 2 C4 C9 C2 illMtuLii. , JIIUJIILU. ji ibji, . li.iiuitfkiliitlUMUi i uijj.i'.iiiiiiLiiiUt C3 :JkiUijlJjjlUilljllli4ULiiik<l C7, C5 C8 C15 i d , lUlilll C19 C6 C16 C18 C17.I 1,J| I I ,• ll.iliU.I ..iiililMh1, ^i,,,i,iiLLLu^Jihli^illuilibUMliLtiilUjiiiU. lJltUl....,.ii.U>illiillltLjUll KjUjuHil̂ JiiiijlUiJidJ i jjJa,ti.iui iiuJa 200 180 160 140 120 100 80 60 40 20 Chemical Shift (ppm) 0 Figure 4.9. 'H and 13C-NMR spectra of 3-[(ferf-butyldimethylsilanyl)phenylmethyl]-3-chloro-l- methylazepane-2,4-dione (107) (recorded in CD2CI2 at 400 and 100 MHz, respectively). 147 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.5. NMR data for 3-[(te^butyldimethylsilanyl)phenylmethyl]-3-chloro-l- methylazepane-2,4-dione (107) (recorded in CD2CI2). Carbon No 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ' 1 3 C 8 (ppm)a 168.1 77.6 206.4 39.2 24.2 47.8 45.6 138.6 132.5 128.4 127.0 129.0 134.7 36.8 -2.89 -1.41 19.4 27.8 C N"15 \ 4 2^ 1 6 ~ - s f N T i ^ | 19 X JH 8(ppm)(mult,7(Hz))bc 1.63 (m), 1.80 (m) 1.60 (m), 1.66 (m) 3.11 (m), 3.79 (m) 3.02 (s) 7.58 (d, J = 7.6 Hz) 7.18 (m) 7.14 (m) 7.12 (m) 6.87 (d, 7 = 6.7 Hz) 3.08 (s) 0.14 (s) 0.33 (s) 0.80 (s) HMBCb (H->C) C3, C4, C6, CI C4, C5 C2, C5, C6, C15 C2, C3, C4, C9, C10, C14, C16, CI C12, C14 C9, C13 C9, C10, C14 C14 C10,C12 C2,C7 C8, CI7, CI8 C8,C16,C18 C18 'Recorded at 100 MHz.bRecorded at 400 MHz.c According to HMQC recorded at 400 MHz. I ' l l wy Rearrangements of organosilanes have been found to take place under acidic, ' basic ' " or thermal ' conditions. During their studies about the stereochemistry of such rearrangements, Hijji, Hudrlik and Hudrlik338 treated silyl ether (108) with fBuLi at low temperature, to isolate after work-up, oxasilacyclopentane (111) as the only reaction product (Scheme 4.17). It was proposed that 111 arises via rearrangement to y-oxidosilane (109), followed by methyl migration (presumably through intermediate 110). 148 Chapter 4. Progress towards the Synthesis of Ceratamines cc ,CI 2 eq. t-BuLi THF, -78 °C 30 min 108 CC5 s ^ c i Br N - — - S i R ^ z R g - 1.1 eq. n-BuLi THF, -78 °C, 2h 112 ^ N s Q - S i ^ R a R g - Si / \ 109 ci II -PyCI /vV ) n o " 1 Si / 111 113 S i R ^ ^ 114 H2Q SiR-| ^2*^3 115 Ri=R2=R3=Me, 87% 116 R,=R2=R3=iPr, 89% 117 R!=R2=Me, R3=fflu, 85% 118 Ri=R2=Ph, R3=fiu, 80% Scheme 4.17. Some examples for the rearrangement of organosilanes induced by alkyl-lithium reagents.338'339 In another example, Turnbull and Krein339 synthesized in excellent yields (80-89%) a series of new sydnones (115-118) upon treatment of 112 with n-butyllithium in the absence of electrophiles. They suggested a very facile rearrangement of 113 to 114, unaffected by steric hindrance around the silicon atom, with the greater thermodynamic stability of the sydnone anion (114) as the driving force for the whole process. Based on these reports, a similar mechanism can be proposed for the formation of 107 (Scheme 4.18). Thus, upon exposure to alkyl-lithium, starting material 106 undergoes deprotonation at C8 to form a stable benzyl/allyl anion (119) which attacks the 4-silyl moiety and liberates the trapped enol. Ketone regeneration leads to a nucleophilic attack on N- chlorosuccinimide, giving rise to chlorination at C3. Preferential deprotonation at C8 correlates with relative acidities exhibited by methylene protons of 3-benzylpropan-2-one (pKa~20) and 299 propene (pKa~43), as well as with the formation of a stable benzyl/allyl anion. 149 Chapter 4. Progress towards the Synthesis of Ceratamines 106 ci—N' 119 8) 107 Scheme 4.18. Proposed mechanism for the rearrangement of 106 to 107. As a last resort, 3-benzylideneazepane-2,4-dione (120) (Scheme 4.19) was synthesized via a classical aldol condensation,340'341 with the hope of facilitating the deprotonation of C5 given the highly conjugated character of the molecule. Furthermore, absence of acidic protons at C3 makes 1NH and H5 the most likely targets for non-nucleophilic bases such as LDA. However, 120 proved to be non-reactive under either acidic or neutral conditions, as well as highly prone to polymerization under basic conditions, generating bakelite-like solids completely insoluble in non-polar or polar solvents, including water. At this point, it became apparent that a completely new strategy was required. N.R. t Br- Br Br Br HCI, rt, overnight TsO—I—OH 6 • x N.R. CH2CI2 reflux Polymerization / \ l H PhCHO / PhH 50 piperidine/HOAc cat. reflux, 86% NaH, Mel DMF, rt LDA/THF, -78 °C NBS ^K2C03 , Mel / MeOH ~~^~-~^_ reflux Polymerization 120 LDA THF, -78 °C Polymerization Et3N TMSCI Mel Polymerization Polymerization Scheme 4.19. Preparation of 3-benzylideneazepane-2,4-dione (120). 150 Chapter 4. Progress towards the Synthesis of Ceratamines 4.5.3. Second generation approach to ceratamines A new strategy was devised with 3-(3,5-dibromo-4-methoxybenzyl)-l,3,6,7-tetrahydro- azepin-2-one (124) as the key intermediate (Scheme 4.20). From 124, at least four logical pathways could lead to reduced ceratamine analogues. 134 133 Scheme 4.20. Second generation approach to ceratamine analogues 128 and 133. Once again, routes A and B aim to prepare a 2-aminoimidazole ring by condensation of oc-functionalized ketones with ./V-acetylguanidine (Scheme 4.6). As starting material, 124 was considered more suitable than 78 (Scheme 4.11) for the preparation of intermediates 126 or 127, either by epoxidation, TsOH-mediated ring opening and oxidation (Route A), or through 151 Chapter 4. Progress towards the Synthesis of Ceratamines oxidation of halohydrin (129) (Route B). The remaining pathways make use of available methodologies to form vicinal syn diamines342"346 as intermediates to facilitate the construction of a dihydro-2-aminoimidazole moiety. Route C was encouraged by the commercially accessible guanilating reagent (132), 7 and involves a hypervalent iodine-mediated vicinal diazidation recently employed by Austin and coworkers348 in their synthesis of (±)-dibromophakellstatin. Route D introduces the use of cyanamide (NH2CN) to build 2-aminoimidazoles (Scheme 4.6) and is based on a method used by Jung and Kohn349'350 to prepare vicinal diamines, combined with the Pinner synthesis of amidines through imidoesters. 351-353 A final dehydrogenation/aromatization step354"356 on either 128 or 133 would afford the imidazo[4,5,<f|azepine core of ceratamines. OH 1. LDA, THF, -78 °C J^k Br 1. Boc2NH, CH3CN C s 2 C 0 3 , reflux 2 . TFA, 6 7 % (2 steps) 122 R!=R2=R3=H 68% 134 R1=R3=Br, R2=OCH3 58% (COCI)2, CH 2 CI 2 "NH2 .HCI 121 Et3N, 0 °C r~\ M e s - N ^ - N PCy, CH2CI2 , reflux 123 R,=R2=R3=H, 87% 135 R,=R3=Br, R2=OCH3, 69% 124 R1=R2=R3=H, 61% 136 R,=R3=Br, R2=OCH3, 84% Scheme 4.21. Preparation of key intermediates 124 and 136. N357 Thus, coupling of but-3-enylamine (121) hydrochloride salt and 2-benzylbut-3-enoic acid (122), 358,359 or the more elaborate 2-(3,5-dibromo-4-methoxybenzyl)-but-3-enoic acid 152 Chapter 4. Progress towards the Synthesis of Ceratamines (134),360"362 was best accomplished by activating both carboxylic acids as their respective acyl chlorides (Scheme 4.21).358 Ring closing metathesis of 123 to produce the desired seven- membered cyclolactam (124) was carried out in good yields using second generation Grubbs catalyst.363-368 Table 4.6. NMR data for 3-benzyl-l,3,6,7-tetrahydroazepin-2-one (124) (recorded in CDC13). bon No Nl 2 3 4 5 6 7 8 9 10 11 12 l3C8(ppmT 176.8 42.8 126.4 128.9 30.1 38.8 36.5 139.9 129.2 128.4 126.1 <; N H \w A "~r~ o 8V^ 10 \^> 1 2 'H 8 (ppm) (mult, J (Hz))"* 6.44 (s, broad) 3.85 (m) 5.32 (m) 5.56 (m) 2.24 (m) 3.16 (m), 3.71 (m) 2.73 (dd, 7 = 9.4, 14.2 Hz), 3.29 (dd, J = 5.8,14.2 Hz) 7.28 (m) 7.25 (m) 7.18 (m) HMBCb(H-»C) C2, C4, C8 C2, C3, C6 C3, C6 C5,C7 C2.C5 C2, C3, C9, CIO C8, C12 C9 CIO a  Recorded at 100 MHz. D Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. The NMR data for 124 (Table 4.6, Figure 4.10) shows both vinylic methines H4 (6H 5.32, 6c 127.1) and H5 (6H 5.56, 6C 129.6) as multiplets due to coupling with H3 (6H 3.85, 5c 42.8) and methylene H6 (5H 2.24, 5c 30.1), respectively. As previously seen with similar intermediates (96 and 98 Table 4.3), the strategic bridge position of methylene H8 (SH 2.73/3.29, 6c 36.5) allows it to display HMBC correlations with carbons in both the seven-membered (C2, 6c 176.8; C3, 5c 42.8) and the aromatic rings (C9, 6c 139.9; C10, 5c 129.2). The amide proton resonance is found as a broad singlet at 6.44 ppm. HRESIMS gave an [M+Na]+ peak at m/z 224.1050, in accordance with the molecular formula C13H15NO (mass calculated for CnHisNONa: 224.1051). 153 Chapter 4. Progress towards the Synthesis of Ceratamines H5 H10" H10 H4 i 11 \ H11' H11 5.6 5.5 H3 A A 5.4 5.3 H7' H8' H7 3.5 3.0 IH6 H12 1NH H8' H8 H3 H7 LJL H8 C10' C10 C5 • * $ \ C11' C11 C10' C10 C12 C4 Vifc \nM#v^AV^WV 5 4 3 Chemical Shift (ppm) C11' C11 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 Figure 4.10. *H and 13C-NMR spectra of 3-benzyl-l,3,6,7-tetrahydroazepin-2-one (124) (recorded in CDCI3 at 400 and 100 MHz, respectively). 154 Chapter 4. Progress towards the Synthesis of Ceratamines Extensive effort was spent in developing streamlined conditions to facilitate the large scale synthesis of key intermediates (124) and (136). Epoxidation, tosylation ' and oxidation methodologies " starting from 124 according to route A (Scheme 4.20), required small scales and exhausting column chromatography after each step, in order to eliminate excess reagents and expected byproducts (benzoic, iodobenzoic and p-toluenesulfonic acids exhibited similar TLC Rf values with the corresponding reaction products). Normally, aqueous NaHCCh extractions provide good separation, but products (125) and (137) proved to be quite water- soluble. The low yields obtained during this sequence reflect a combination of these experimental factors and a generalized low efficiency (Scheme 4.22). Furthermore, attempts to produce a 2-aminoimidazole ring from 126 ' ' were fruitless and afforded very complicated mixtures. Neither NMR nor LRESIMS were able to confirm the presence of the desired addition products. 124 m-CPBA CH2CIZ , 39% Multiple products N.R. TsO TsOH.HpO H 2 0 /CH 3 COCH 3 25% N H 2 H a r - T ^ N H H C I C 6 H 6 , Et3N, PPTs —-~__ reflux NH 2 O H 2 N ^ N ^ \ CH 3 CN, rt HO 137 A c ° OAc AcOj,y 1 CH2CI2 DMP O 5 4 % TsO 126 Scheme 4.22. Preparation of toluene-4-sulfonic acid 6-benzyl-5,7-dioxoazepan-4-yl ester (126) and its reaction with guanidine derivatives. 155 Chapter 4. Progress towards the Synthesis of Ceratamines Execution of pathway B (Scheme 4.20) led to the preparation of halohydrin (138) in good yields, whereas its oxidation using Dess-Martin penodinane (1.5 eq.) yielded a 1:1 mixture of products (Scheme 4.23). When 3 eq. of DMP were employed, only 4-bromo-3-(3,5-dibromo-4- methoxybenzyl)-3,4-dihydro-lH-azepine-2,5-dione (139) was obtained. The surprising formation of this compound was initially considered very positive since it already has a C6-C7 double bond, required at some point to produce the imidazo[4,5,rf]azepine core of ceratamines. Nicolaou, Zhong and Baran354 had proposed the use of another hypervalent iodine-based reagent, iodoxybenzoic acid (IBX), in an efficient method for the one-pot conversion of alcohols, ketones and aldehydes to a,(3-unsaturated carbonyl compounds. Thus, it is reasonable for DMP to yield 139 when used in excess. H °V° A c 2 0 , HOAc IBX & A c Q O A C ~ - l ^ -OAc DMP o HO- NBS, H 2 0 CH3COCH3 Br Br  ia%" *~ 139 Br 138 A: 1.5 eq. DMP 27% B: 3 eq. DMP 98% N H 2 A c N = < N H 2 DMF, 25 °C 7 1 % , N H 2 H N = < . HCI N H 2 DMF, Et 3N, 70 °C 1 1 % Mel, K 2CQ 3 CH3COCH3, 8 6 % Scheme 4.23. Chemical transformations involved in the synthesis of 141 and 142. "yf/i 'yfi.'l 0*7^* Interestingly, the conditions used to react 139 with guanidine derivatives ' ' afforded compound (141) as the main product. Clearly, bromide elimination to constitute a stable 156 Chapter 4. Progress towards the Synthesis of Ceratamines conjugated heterocycle was favored over addition of guanidine to the non-reactive carbonyl in 139. Furthermore, guanidine is one of the best organic bases, with strength comparable to NaOH OKa of guanidinium salt in water at 25°C: 13.6).352 The NMR spectrum of 141 (Table 4.7, Figure 4.11) showed a coupled spin system comprised of H4 (5H 6.94, 5c 141.9), H6 (6H 5.67, 5c 111.5), and H7 (5H 6.81, 5C 136.1). The remaining resonances in the 'H-NMR are all singlets. Methylene H8 (5H 3.76, 5c 38.5) and the broad 1NH signal (5H 11.10), H4, H8 and H10 (5H 7.58, 5C 133.33) all showed HMBC correlations with all possible carbons located two and three bonds apart, evidencing the highly conjugated character of 141. HRESIMS gave a [M+H]+ peak at m/z 399.9189, corresponding to a molecular formula of Ci4HiiN03Br2 (calculated for C14H12N0379Br2: 399.9184). Table 4.7. NMR data for 3-(3,5-dibromo-4-methoxybenzyl)-lH-azepine-2,5-dione (141) (recorded in DMSO-Je). Carbon No 13C8(ppm)a *H 8 (ppm) (mult, / (Hz))b'c HMBCb (H-»C) Nl 11.10 (s, broad) 2 165.1 3 142.7 4 141.9 6.94 (d, J = 2.2 Hz) C2, C3, C5, C6, C8 5 185.4 6 111.5 5.67 (dd, J = 2.2,10.0 Hz) C4,C5,C7 7 136.1 6.81 (d, 7=10.0 Hz) C2, C5, C6 8 38.5 3.76 (s) C2, C3, C4, C9, C10 9 138.0 10 133.3 7.58 (s) C3, C8, C9, C11,C12 11 117.2 12 152.0 13 60.3 3.76 (s) C12 a  Recorded at 100 MHz.b Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. 157 Chapter 4. Progress towards the Synthesis of Ceratamines H7 H4 I 7.0 1NH H6 6.5 6.0 5.5 J_J H10 H10' H8 H4 H7 H6 H13 UJJ L 11 C5 10 7 6 5 4 Chemical Shift (ppm) C10' C10 C13 i|C8 ! I " ; ' I ' 60 55 50 45 40 rJ \—•—' C4 C3 C7 C9 C2 C12 l*ia#aM«M«IM#|*«M*#«*NiMdqw« uL C11' C 6 C11 C13 I I M H I I W " mwnMliiilni C8 U**m«mmmmmMm***mmimm 200 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 Figure 4.11. 'H and 13C-NMR spectra of 3-(3,5-dibromo-4-methoxybenzyl)-lH-azepine-2,5- dione (141) (recorded in DMSO-d6 at 400 and 100 MHz, respectively). 158 Chapter 4. Progress towards the Synthesis of Ceratamines Methylation of 141 furnished 3-(3,5-dibromo-4-methoxybenzyl)-l-methyl-lH-azepine- 2,5-dione (142) in good yield. Noteworthy, analysis of the 'H-NMR spectra for 142 and ceratamine A (13) (Figure 4.12) generates quite similar structural information. A) H7 H4 H6 6.5 6.0 H10 H10' H7 H4 H6 CH2CI2 H13 H8 H14 6 5 4 Chemical Shift (ppm) 19\ N-  a H13 H17 (ppm) Figure 4.12. H-NMR spectra of a) 3-(3,5-dibromo-4-methoxybenzyl)-l-methyl-lH-azepine- 2,5-dione (142); and ceratamine A (13) (recorded in CDC13 and DMSO-rf6 at 400 MHz). 159 Chapter 4. Progress towards the Synthesis of Ceratamines Lacking only a 2-aminoimidazole motif, compound (142) possesses instead a carbonyl group that can establish hydrogen bond interactions in a similar fashion as Nl in ceratamines A (13) or B (14). Some planarity in the seven-membered ring is also shared, particularly for the area surrounding carbonyl C5, but carbonyl C2 may project outside the plane outlined by carbons C3 to C7. Evidently, the absence of a fused five-membered aromatic ring concedes additional flexibility to the amide bond (Figure 4.13). 142 H 13 Figure 4.13. Minimum energy conformations for 3-(3,5-dibromo-4-methoxybenzyl)-l-methyl - lH-azepine-2,5-dione (142), and ceratamine A (13). 2C=0 in 142 points out of the page (MM2, CS Chem3D Ultra 7.0, minimum RMS gradient 0.100). Such structural resemblance may explain the activity exhibited by 142 in the same bioassay used to test ceratamines, clearly showing cell arrest in mitosis at a starting concentration of 33 |ig/mL (IC50 for ceratamines A and B: 10 u.g/mL). ' However, some of the unusual features observed for the natural product concerning microtubule arrangements around the cell nucleus were not observed with this analogue. 160 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.1. Reaction of 126,141-147 with several nitrogen-based nucleophiles. 126 141 R,=H, R2=R4=Br, R3=OMe 144 R1=H, R2=R4=Br, R3=OMe 147 142 R1=Me, R2=R4=Br, R3=OMe 145 Ri=Me, R2=R4=Br, R3=OMe 143 R,=Me, R2=R3=R4=H 146 R,=R2=R3=R4=H N 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 SMa 126 141 142 143 144 145 146 147 Nucleophile Guanidine.HCl Guanidine.HCl yV-Acetylguanidine Af-Methylguanidine Guanidine.HCl Guanidine.HCl Guanidine.HCl Guanidine.HCl N-Methylguanidine NH3 NH3 NH3 NH2CN GuanidineJHCl N-Methylguanidine Guanidine.HCl Guanidine.HCl iV-Boc-guanidine Guanidine.HCl Guanidine.HCl N-Methylguanidine GuanidincHCl Af-Boc-guanidine iV-Boc-guanidine Guanidine.HCl iV-Boc-guanidine N-Boc-guanidine Af-Boc-guanidine Conditions263'267'275'349'350'374 Et3N, PPTs, C6H6, reflux, 18 h Et3N, C6H6, reflux, 16 h CH3CN, rt, 96 h Et3N, DMF, 70 °C, 24 h EtONa, EtOH, reflux 2 h EtONa, EtOH, reflux 12 h EtONa, EtOH, rt, 48 h DMF, 70 °C, 48 h Et3N, DMF, 70 °C, 24 h CH3OH, 80 °C, sealed tube, 5 h CH3OH, 80 °C, sealed tube, 18 h 1,4-Dioxane, 80 °C, sealed tube, 5 h 1,4-Dioxane, 80 °C, sealed tube, 18 h CH3CN, reflux, 24 h DMF, reflux, 48 h CH3CN, reflux, 24 h DMF, reflux, 24 h DMF, rt, 4 days DMF, rt, 3 days Et3N, DMF, 70 °C, 18 h Et3N, DMF, 70 °C, 24 h DMF, 70 °C, 5 days DMF, rt, 4 days EtONa, EtOH, rt, 3 h DMF, 70 °C, 2 days DMF, 70 °C, 2 days THF, Et3N, reflux, 24 h THF, NaOMe, 60 °C, 24 h Outcome15 Complex mixture Product not found S. M. Complex mixture S. M. Product not found Complex mixture S.M. S.M. Addition? Addition? S.M. S.M. S.M. Addition? S.M. S.M. S.M. S.M. S.M. S.M. S.M. S.M. Complex mixture Br" elimination Br" elimination S.M. Product not found aS.M: starting material. bAccording to 'H-NMR and LRESIMS. 161 Chapter 4. Progress towards the Synthesis of Ceratamines Compounds 141-143 showed an extraordinary stability towards condensation with guanidine derivatives (Table 4.8, entries 5-9 and 14-18). Therefore, in order to avoid forming a C6-C7 double bond in precursor (139) that eventually leads to elimination and complete conjugation in 141, starting material (138) was oxidized exclusively to 4-bromo-3-(3,5-dibromo- 4-methoxybenzyl)-azepane-2,5-dione (140) employing B X (Scheme 4.24).354,374"378 Nevertheless, condensation attempts of 140 with guanidine produced only elimination product (144), which in turn also proved to be inert (Table 4.8, entries 19-24). N H 2 N H 2 DMF, 25 °C , 8 7 % Mel , K g C Q 3 Br CH3COCH3, 9 9 % 145 Scheme 4.24. Reactions leading to the synthesis of 3-(3,5-dibromo-4-methoxybenzyl)-l-methyl - 6,7-dihydro-lH-azepine-2,5-dione (145). According to Table 4.8, only strong basic media and heating seemed to produce changes in the starting materials, sometimes accompanied by minimum amounts of products usually detected in the 'H-NMR baseline, exhibiting a new coupling pattern for H4, H6 and H7 in the seven-membered ring (entries 10, 11 and 15). However, scarce amounts made it impossible to verify nucleophile addition in these compounds. Starting material 147 (entries 25-28) was always isolated as a minor byproduct in reactions involving NBS, or directly prepared by bromination of 136 in 99% yield. Even small nucleophiles like NH3 required extreme reaction conditions to only 162 Chapter 4. Progress towards the Synthesis of Ceratamines give a faint indication of addition (entries 10-14). In general, all the data in Table 4.8 reflects the high stability of the different starting materials, and poor performance of guanidine and its derivatives as nucleophiles. Route C (Scheme 4.20) rapidly led to another dead end when, upon reaction with ~IAQ "570 ^ 4 8 "380 IN3 'M,J" or -r(N3)2, ,JOU only byproducts were isolated. Additionally, hydrogenation conditions250'348,381 intended to form a primary amine at C5 were not successful (Scheme 4.25). NH O A 1) Phl(0°c)2, TMSfvfe, 2) B4M, CH^32 B: 1) NaNfc,, MeCN; 2) IQ Ffe 124 R,=R2=R3=H A 136 R,=R3=Br, R2=OMe B * - N.R Ffe R3 148 40% 150 0% Ffe R3 149 21% 151 90% Scheme 4.25. Chemical transformations of 124 and 136 involved in route C. Contrary to the guanidine case, cyanamide (NH2CN) is highly nucleophilic and can be added to double bonds through a bromoniun ion.382 In their methodology to stereoselectively prepare vicinal diamines, Jung and Kohn349'350'374 generated a stable cyclic imido ester (154) by reaction of a P-bromoalkyl cyanamide (152) with ethanol in acidic media (Scheme 4.26), as the precursor for the final product 155. <,* NH2CN NBS <*7 \ HCI NHCN EtOH 152 c i OEt * , N - C = N - H OEt 153 ft V H *^ NvN H H " N = < d)Et 154 Ba(OH)2 , Heat H2N NH 2 155 Scheme 4.26. Synthesis of vicinal diamines by Jung and Kohn.374 163 Chapter 4. Progress towards the Synthesis of Ceratamines In analogy to this process, it was envisioned that a similar five-membered cyclic compound might be formed by treatment of (3-bromoalkyl cyanamide (156) with ammonia, giving rise to a reduced version of the targeted imidazo[4,5,J]azepine core of ceratamines (Scheme 4.27). Addition of amines to cyanamide derivatives is just one of the common methods to synthesize substituted guanidines,352 whereas the preparation of imido esters followed by addition of primary amines (including NH3) is known as the Pinner method for synthesis of amidines [RC(=NH)NH2].351 Thus, heating a sealed pyrex tube containing intermediate (156) and NH3 in CH3OH, yielded the reduced ceratamine B analogue (157) in good yield. HN NH3/CH3OH ^ H 2 N " N heat, 96% B0C2O/CH3OH Et3N, 24% 159 158 Scheme 4.27. Synthesis of 19-demethyl-l,4,5,8,9,10-hexahydroceratamine B (157). The 'H and 13C-NMR data for 157 (Figure 4.14, Table 4.9) shows resonances for NH protons at 7.66, 7.90 and 8.20 ppm, whereas signals for the vinylic protons of 136 were substituted by methines H4 (SH 3.80, 5C 60.7) and H10 (5H 3.82, 5C 58.5). Diastereotopic methylenes H8 (5H 3.10/3.42, 5c 38.3), H9 (5H 1.57/2.15, 5c 31.7) and HI 1 (5H 2.92, 5c 29.0) do not exhibit important changes in their chemical shifts, when compared with 156 (Experimental). 164 Chapter 4. Progress towards the Synthesis of Ceratamines H16 H20 H8 H11 U DMSO H9' H9 H9' H9 1 2.00 1.75 1.50 5 4 3 Chemical Shift (ppm) C13 C13 C6 C15 C12 «*•*- C14' C14 C16 C16 C4 C10 60 59 C5 KfHWMMMMa C9 £8 / c i 1 ; \ mm*. 180 160 140 120 100 80 Chemical Shift (ppm) 60 40 20 Figure 4.14. *H and 13C-NMR spectra of 19-demethyl-1,4,5,8,9,10-hexahydroceratamine B (157) (recorded in DMSO-J6 at 600 and 150 MHz respectively). 165 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.9. NMR data for 19-demethyl-l,4,5,8,9,10-hexahydroceratamine B (157) (recorded in DMSO-d6). Carbon 13C *H HMBCb No 8(ppm)a 8(ppm)(mult,7(Hz))b'c (H->C) Nl 7.90 (s, broad) C2, C4, CIO 2 160.4 NH2 8.20 (s, broad) C2, C4, CIO N3 4 60.7 3.80 (m) C6, C9, C10, C l l 5 48.5 3.12 (m) C6, C11,C12 6 172.9 N7 7.66 (d, J = 7.4 Hz) C5, C6, C9 8 38.3 3.10 (m), 3.42 (m) C9, C10 9 31.7 1.57 (m), 2.15 (m) C4, C8,C10 10 58.5 3.82 (m) C4 11 29.0 2.92 (m) C5, C6, C12, C13 12 139.1 13 133.1 7.53 (s) C11,C14, C15 14 117.0 15 151.6 16 60.4 3.77 (s) C15 "Recorded at 150 Hz.b Recorded at 600 MHz.c According to HMQC recorded at 600 MHz. More evidence for the presence of a fused five-membered ring was given by HMBC correlations from the primary amine at C2 to carbons C4 and C10. The secondary amine INH also correlates with these carbons, whereas 7NH does it with methine C5 (5c 48.5), carbonyl C6 (8c 172.9) and methylene C9 (5c 31.7), allowing an unambiguous assignment of these groups to the dihydro-2-aminoguanidine and amide motifs, respectively. As usual, the linking methylene HI 1 displayed HMBC cross-peaks with carbons in both rings. HRESIMS provided a [M+H]+ peak at m/z 444.9871, suggesting a molecular formula CisHisN^Br^ (calculated for 166 Chapter 4. Progress towards the Synthesis of Ceratamines C15H19N4O2 Br2: 444.9875), and confirmed that upon ammonia addition, displacement of bromide by guanidine installed the new ring. Compared with active analogue (142), compound (157) is completely non-planar and its low-energy conformation presents folding of the seven-membered ring in a boat-like conformation (Figure 4.15). The lack of antimitotic activity exhibited by 157 is undoubtedly due to this factor. Additionally, in the case of inactive compounds (158) and (159), the presence of bulky groups must interfere with binding and generation of hydrogen bond interactions with the active site. 157 13 Figure 4.15. Minimum energy conformations for 19-demethyl-1,4,5,8,9,10- hexahydroceratamine B (157); and ceratamine A (13). 6C=0 in 157 is pointing out of the page (MM2, CS Chem3D Ultra 7.0, minimum RMS gradient 0.100). So far, a noticeable regioselectivity has been obtained regarding addition of nucleophiles to the double bond in 136. Whether the electrophilic species was an epoxide or a halonium cation, TsOH, H2O, Br" and NH2CN all attacked the same carbon C5 in the tricyclic intermediate (160) (Scheme 4.28). An important steric effect from the benzyl substituent at C3, which blocks 167 Chapter 4. Progress towards the Synthesis of Ceratamines access to C4 and consequently favors nucleophile attack at C5, may be the main reason behind this tendency. 136a R v : Nu I ^ - N ^ - l \ H X 160a X=0, Br+, I+ 161a Nu=OTs, X=OH 138a Nu=OH, X=Br 151a Nu=N3, X=I 156a Nu=NHCN, X=Br Scheme 4.28. Preferential nucleophile attack at C5 observed during addition reactions to (136). Only one stereoisomer is shown (MM2, CS Chem3D Ultra 7.0, minimum RMS gradient 0.100). Attempts to manipulate 157 revealed the intractable nature of 2-aminoimidazole containing compounds. The analogue does not react with common staining reagents and is immovable on normal phase silica gel. The dihydro-2-aminoimidazole motif is quite labile, and exhibited ring opening during reversed-phase chromatography. Such effect was more evident with the imido ester version (154) (Scheme 4.26), only detectable in traces by LRESIMS. Likewise, when methylamine and benzylamine replaced ammonia in the reaction with oc- bromoalkyl cyanamide (156), the corresponding methyl and benzyl analogues were obtained in minimum amounts. Reactions of 157, as well as protected derivatives (158) and (159), were characterized by opening of the five-membered ring and total loss of the resulting substituted guanidine group, as evidenced by LRESIMS. Normal procedures intended to remove protons and assist with aromatization of the carbocyclic skeleton, including treatment with DDQ, 354-356 , : i270 387-390 , IBX,^ J J 0 chloranilz/u or Pd/C,J8/jyu were all fruitless 168 Chapter 4. Progress towards the Synthesis of Ceratamines 4.6. Conclusions and future directions The ceratamine inspired antimitotic agent (142) and inactive analogue (157) were synthesized in no more than 8 steps, with overall yields of 20% and 15% respectively (Scheme 4.29). X>H 1. LDA, THF, -78 °C 58% Br Boc2NH, CH3CN C s 2 C 0 3 , reflux TFA, 67% (COCI)2, CH 2 CI 2 ~NH2.HCI 121 Et3N, 0 °C 69% r~\. C l - I=<p h C I * " I >H PCy3 Br CH2CI2 , reflux, 84% 136 NBS H 2 0 /CH 3 COCH 3 98% AcN= , N H 2 \H2 Br Br 141 R=H 142 R=CH3 DMF, 25 °C 7 1 % Mel, K 2 C 0 3 CH 3COCH 3 , 86% 139 138 Scheme 4.29. Syntheses of 19-demethyl-l,4,5,8,9,10-hexahydroceratamine B (157) and 3-(3,5- dibromo-4-methoxy-benzyl)-1 -methyl-1 H-azepine-2,5-dione (142). 169 Chapter 4. Progress towards the Synthesis of Ceratamines Preparation of key intermediate 136 was achieved in only four steps, via standard peptide coupling of 121 and 134, followed by ring closing metathesis using second generation Grubbs catalyst. This efficient and condensed synthetic sequence proceeded with acceptable yields and in scale large enough to ensure an appropriate stock of material for further reactions. The double bond in 136 is quite versatile, and exhibited high regioselectivity towards nucleophilic addition. NBS mediated addition of cyanamide and condensation of the resulting precursor with ammonia yielded the reduced ceratamine B analogue (157), in a total of six steps. Alternatively, preparation of bromohydrin (138), followed by Dess-Martin periodinane oxidative dehydrogenation afforded intermediate (139) in excellent yields. The azepine core of 141 was best furnished by basic treatment with iV-acetylguanidine, initially intended as a methodology to produce 2-aminoimidazole motifs. Lastly, a simple methylation procedure provided 142 in a total of eight steps. In general, the chemistry involved preparing 142 and 157 proved to be quite robust and repeatable. Some aspects that may be considered as shortcomings include the high costs of second generation Grubbs catalysts, and dangers associated with ammonia reactions in closed systems. Furthermore, addition of cyanamide to 136 through a bromonium cation was always accompanied by a considerable amount of dibrominated product. Thus, variation of reaction conditions to optimize this step may be in order. The incredible stability exhibited by 142 and its demethyl analogue 141 made attempts to elaborate both compounds fruitless. On the other hand, the dihydro-2-aminoimidazole motif in 157 turned out to be very labile, and efforts intended to aromatize its bicyclic carbon skeleton were characterized by ring opening and loss of the resulting guanidine moiety. When antimitotic activity ' was evaluated for these compounds, 142 exhibited cell arrest in mitosis at a starting concentration of 33 (Xg/mL and thus, it is significantly less potent 170 Chapter 4. Progress towards the Synthesis of Ceratamines than both natural ceratamines A and B, sharing an IC50 value of 10 (ig/mL. Furthermore, some of the attractive unusual effects in microtubule polymerization described for ceratamine A that suggest its binding to a different active site than taxol, 29 were not observed for 142. Insights about the ceratamines pharmacophore earned with the preparation and activity evaluation of both 142 and 157, include: I. Requirement for flat carbon backbones: both 142 and ceratamine A share some planarity, and while the natural product is completely flat, absence of a fused imidazole ring in 142 concedes additional mobility to the azepine core, which folds to project the amide carbonyl outside the plane outlined by other ring components (Figure 4.13). Analogue (157) on the other hand is non-planar and completely inactive. II. The unique imidazo[4,5,d]azepine core heterocycle proper of ceratamines is not required for activity. A 2-aminoimidazole ring improves potency and efficiency, but it is not crucial for cell cycle arrest. Replacement by just a carbonyl as in 142 provided a biologically active compound. This second statement is very important for future synthetic attempts, since modification or replacement of the 2-aminoimidazole substructure in ceratamine analogues may increase their antimitotic power. Additionally, the reported ' ' intractable nature of compounds containing this functionality suggests the use of more synthetically accessible heteroatomic arrangements, capable of structurally emulating substituted imidazole rings. For this purpose, availability of active analogues for structure-activity relationship studies is clearly vital. The preparation of tetrahydroazepin-2-one (136) provides a good starting point to synthesize such compounds. 171 Chapter 4. Progress towards the Synthesis of Ceratamines In case of a total synthesis of ceratamines, deferring installation of the 2-aminoimidazole ring until the final stages of synthesis is definitely advantageous. Most of the recently published total syntheses of compounds bearing this substructure, including sceptrin (41), ' ageliferin (43),268 naamidine A (54),269 and ageladine A (55)270 prepare it in the last reaction step via condensation of a-haloketones with guanidine derivatives as proposed by Little and Weber263 in 1994. According to observations made during the first and second generation approaches implemented in the present work, available methods to prepare 2-substituted imidazoles and their compatibility with acidic protons, steric constraints of seven-membered rings, and the presence of other functionalities such as amides, will determine the outcome of future ceratamine synthetic proposals. A thorough exploration of biomimetic approaches, using a preformed imidazole ring as starting material is also suggested. Literature procedures to sequentially functionalize imidazoles are readily available, " and as evidenced when handling 157, the additional aromatic stability of imidazole derivatives is essential to successfully elaborate synthetic intermediates without ring opening or decomposition. Contrary to the 7+5 approach followed in the present work, a 5+7-membered ring construction may give better results in installing both the amide and the primary amine moieties in imidazole. Above all, the aromatic element present in the imidazole ring provides a solid ground that may facilitate dehydrogenation steps down the road. 172 Chapter 4. Progress towards the Synthesis of Ceratamines 4.7. Experimental General experimental procedures For general experimental procedures see Section 2.6. Preparation of 2-benzyl-ri,31dithiane-2-carboxylic (83) and 2-methyl-ri.31dithiane-2-carboxylic (84) acids p o o i H . H 2 0 T S ° H L J - 2.BzClorMel [ ^ | - J 83 R=Ph 84 R=H A mixture of dihydroxyacetic acid monohydrate (9.2 g, 100 mmol), 1,3-propanedithiol (11.0 mL, 110 mmol) and p-toluenesulfonic acid (1.9 g, 10 mmol), in benzene (250 mL), was refluxed for 1 h. Once at 25 °C, the crude was extracted with NaHCCh sat. (3 x 100 mL), acidified with HC1 cone, and extracted again with EtOAc (4 x 100 mL). The organic layer was dried (Na2SC>4) and concentrated in vacuo, to afford desired [l,3]dithiane-2-carboxylic acid intermediate as a yellowish solid (13.15 g, 80%). Without further purification, this intermediate (1.64 g, 10 mmol) was dissolved in THF (75 mL) and treated with LDA (12 mL, 21.6 mmol, 1.8 M) at 0 °C. After stirring for 30 min at the same temperature, benzyl chloride (1.27 mL, 11 mmol) was added and the resulting mixture stirred overnight. Aqueous hydrolysis, followed by HC1 acidification and Et^O extractions, provided after solvent evaporation a pale yellow amorphous solid identified as (83) (2.54 g, 89%). When Mel (0.70 mL, 11 mmol) was used to quench the dianion intermediate under the same conditions, 2-methyl-[l,3]dithiane-2-carboxylic acid (84) was obtained (1.78 g, 72%). 173 Chapter 4. Progress towards the Synthesis of Ceratamines 2-Benzyl-[l,3]dithiane-2-carboxylic acid (83): *H NMR (CDC13, 400 MHz) £7.36-7.08 (5H, m), 3.37 (2H, s), 3.25 (2H, ddd, J = 2.8, 12.2, 13.7 Hz), 2.68 (2H, ddd, J = 3.7, 4.0, 13.7 Hz), 2.12 (1H, m), 1.83 (1H, m); 13C NMR (CDC13, 100 MHz) £176.3 (C), 134.0 (C), 130.8 (2CH), 128.1 (2CH), 127.6 (CH), 46.5 (C), 44.3 (CH2), 28.1 (2CH2), 24.1 (CH2). HRESEVIS calcd for Ci2H1402S2Na ([M+Na]+): 277.0333; found 277.0328. 2-Methyl-[l,3]dithiane-2-carboxylic (84): lH NMR (CDC13, 400 MHz) £3.39 (2H, ddd, J = 2.8, 13.6, 14.4 Hz), 2.63 (2H, ddd, J = 3.8, 4.3, 13.4 Hz), 2.16 (1H, m), 1.82 (1H, m), 1.67 (3H, s); 13C NMR (CDCI3, 100 MHz) 5 176.2 (C), 45.9 (C), 28.0 (2CH2), 25.4 (CH2), 23.8 (CH3). HREEVIS calcd for C6Hi0O2S2 (M+): 178.01222; found 178.01228. Coupling of 2-benzyl-ri,31dithiane-2-carboxyric (83) and 2-methvl-ri,31dithiane-2-carboxvhc (84) acids with histidine methyl ester ^ S ^ l P^BOP, DIEA * " N X ^ Y * ' 0 R 83 R=Ph 85 R=Ph 84 R=H 86 R=H To a solution of (83) (0.25 g, 1 mmol) in CH2C12 (10 mL), was added PyBOP (0.54 g, 1 mmol), histidine methyl ester (0.27 g, 1.1 mmol) and DIEA (0.50 mL, 2.8 mmol). After overnight stirring at 25 °C, NaCl sat. was added and CH2C12 extractions performed. The organic extracts were dried (Na2S04), concentrated in vacuo, and the resulting residue was purified by a gradient silica gel column chromatography (50% EtOAc/hexanes to 20% MeOH/EtOAc), to afford (85) as a colorless powder (0.20 g, 50%). When compound (84) (0.18 g, 1 mmol) was used as starting material, the procedure above provided desired product (86) also as a colorless solid (0.18 g, 53%). 174 Chapter 4. Progress towards the Synthesis of Ceratamines 2-Benzyl-[l,3]dithiane-2-carboxylic acid-histidine methyl ester adduct (85): 2H NMR (CD2CI2, 300 MHz) £8.34 (IH, d, J = 7.4 Hz), 7.58 (IH, s), 7.20-7.14 (5H, m), 6.70 (IH, s), 4.68 (IH, ddd, J= 5.5, 6.6, 6.6 Hz), 3.63 (3H, s), 3.17 (2H, m), 3.05 (IH, dd, J = 9.9, 14.9 Hz), 2.94 (IH, dd, J= 9.4, 14.9 Hz), 2.91 (IH, m), 2.79 (IH, m), 2.57 (2H, m), 1.98 (IH, m), 1.83 (IH, m); 13C NMR (CD2C12, 75 MHz) J171.4 (C), 169.5 (C), 134.7 (C), 134.4 (CH), 133.5 (C), 130.5 (2CH), 127.7 (2CH), 127.2 (CH), 115.4 (CH), 59.9 (CH), 52.1 (CH3), 49.5 (C), 42.2 (CH2), 28.6 (CH2), 28.15 (CH2), 28.07 (CH2), 24.4 (CH2). HRESIMS calcd for C19H24N3O3S2 ([M+H]+): 406.1259; found 406.1251. 2-Methyl-[l,3]dithiane-2-carboxylic acid-histidine methyl ester adduct (86): !H NMR (CD2C12, 300 MHz) S8.53 (IH, d, J = 7.6 Hz), 7.57 (IH, s), 6.80 (IH, s), 4.67 (IH, ddd, J = 5.7, 6.2, 6.9 Hz), 3.64 (3H, s), 3.15 (IH, m), 3.07 (IH, m), 3.00 (IH, m), 2.86 (IH, m), 2.57 (2H, m), 2.00 (IH, m), 1.80 (IH, m), 1.55 (3H, s); 13C NMR (CD2C12, 75 MHz) S 172.0 (C), 171.4 (C), 135.2 (CH), 134.6 (C), 115.6 (CH), 53.9 (CH), 52.5 (CH3), 42.7 (C), 29.0 (2CH2), 28.9 (CH2), 27.7 (CH2), 24.7 (CH3). HRESIMS calcd for C B ^ O ^ O J S Z ([M+H]+): 330.0946; found 330.0954. Preparation of azepane-2,4-dione (50) via Beckmann rearrangement O NOH r ^ l NH2OH.HCI ( J ^ PPA / ^NH I ^ o ~MeOHorE,OH *" l ^ J L O R 120~̂C "~ J ^ ^ O R: Me or Et 50 A mixture of 1,3-cyclohexadione (22.4 g, 200 mmol) and hydroxylamine hydrochloride (14.0 g, 200 mmol) in EtOH (200 mL) was refluxed for 1.5 h. Once at room temperature, the crude was neutralized using 10% K2CO3, extracted with CH2CI2 and dried (Na2SC<4). Solvent evaporation provided the corresponding 3-ethoxy-2-cyclohexen-l-one oxime quantitatively. This intermediate (7.6 g, 49 mmol) was heated under mechanical stirring and solvent less conditions 175 Chapter 4. Progress towards the Synthesis of Ceratamines in the presence of polyphosphoric acid (PPA, 100 g), for 3 h at 110-120 °C. Upon reaction time completion, the viscous final crude was added to ice/H20 (100 mL), neutralized with NaOH pellets and extracted with BuOH. Solvent concentration in vacuo, followed by silica gel column chromatography (100 EtOAc), afforded azepane-2,4-dione (50) as a reddish amorphous solid (5.26 g, 85%). For a summary of *H and 13C NMR assignments, see Table 4.10. lH NMR (CD2C12, 400 MHz) J7.37 (IH, s, broad), 3.50 (2H, s), 3.45 (2H, m), 2.62 (2H, t, J = 6.9 Hz), 1.98 (2H, m); 13C NMR (CD2C12, 100 MHz) £203.1 (C), 169.8 (C), 52.6 (CH2), 44.3 (CH2), 41.6 (CH2), 28.6 (CH2). HRESIMS calcd for C6Hi0NO2 ([M+H]+): 128.0712; found 128.0710. Table 4.10. NMR data for azepane-2,4-dione (50) (recorded in CD2C12). Carbon No "C8(ppm)a *H 5 (ppm) (mult, / (Hz))"-1 HMBCb(H-»C) Nl 7.37 (s, broad) 2 169.8 3 52.6 3.50 (2H,s) C2, C4, C5 4 203.1 5 44.3 2.62 (t, 7=6.9 Hz) C3, C4, C6, C7 6 28.6 1.98 (m) C3, C5, C7 7 41.6 3.45 (m) C2, C5, C6 "Recorded at 100 MHz. " Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. Preparation of 3-benzylazepane-2,4-dione (78) / NH s?—'^O 2. BnBr ) 78 Azepane-2,4-dione (50) (0.70 g, 5.5 mmol) was dissolved in THF (25 mL) and cool down to -78°C. After 15 min, LDA (3.7 mL, 6.6 mmol, 1.8 M) was added and the crude allowed to gradually warmed up to -30°C for lh. Once at this temperature, benzyl bromide (0.90 mL, 7.5 176 Chapter 4. Progress towards the Synthesis of Ceratamines mmol) was added and stirring continued overnight. The reaction crude was treated with NH4CI sat., extracted with EtOAc and dried (Na2SC>4). Solvent concentration, followed by gradient silica gel column chromatography (80% EtOAc/hexanes to 10% MeOH/EtOAc), provided desired 3-benzylazepane-2,4-dione (78) (0.43 g, 36%) as a white solid. For a summary of !H and I3C NMR assignments, see Table 4.2. lH NMR (CD2C12, 400 MHz) J7.29-7.13 (5H, m), 6.11 (1H, s, broad), 4.24 (1H, dd, J = 6.7, 6.8 Hz), 3.75 (1H, m), 3.31 (1H, m), 3.15 (2H, m), 2.60 (2H, m), 2.10 (1H, m), 1.92 (1H, m); 13C NMR (CD2C12, 100 MHz) 8 204.2 (C), 169.5 (C), 140.5 (C), 129.6 (2CH), 128.8 (2CH), 126.7 (CH), 59.4 (CH), 46.3 (CH2), 41.9 (CH2), 31.8 (CH2), 30.7 (CH2). HRESIMS calcd for Ci3Hi5N02Na ([M+Na]+): 240.1000; found 240.0996. Halogenation of 3-benzylazepane-2,4-dione (78) Br Br To a solution of 3-benzyl-azepane-2,4-dione (78) (0.040 g, 0.18 mmol) in Et20 (5 mL), was added HC1 in Et20 (0.4 mL, 0.4 mmol, 1M) and 2,4,4,6-tetrabromoquinone (0.30 g, 0.73 mmol) diluted also in Et20 (15 mL). The resulting mixture was stirred at 25 °C for 45 h refluxing periodically, whereupon 10% Na2C03 was added and THF extractions performed. The organic fractions were dried with Na2S04 and concentrated in vacuo to give a yellowish residue, which was additionally purified by silica gel column chromatography (80% EtOAc/hexanes) to afford 3-benzyl-3-bromo-azepane-2,4-dione (98) as a white amorphous powder (0.047 g, 87%). On the other hand, a solution of starting material (78) (0.13 g, 0.6 mmol) in THF (5 mL) at -78 °C was treated with LDA (0.36 mL, 0.66 mmol, 1.8 M), following shortly after the 177 Chapter 4. Progress towards the Synthesis of Ceratamines addition of Af-chlorosuccinimide (0.12 g, 0.9 mmol) in THF (5 mL), and overnight stirring at 25 °C. Crude work-up and further purification was done following the same procedure as above, to obtain desired 3-benzyl-3-chloro-azepane-2,4-dione (96) (0.12 g, 83%). 3-Benzyl-3-bromoazepane-2,4-dione (98): XH NMR (CD2C12, 300 MHz) Si.25 (5H, m), 6.65 (1H, s, broad), 3.84 (1H, d, J = 13.1 Hz), 3.50 (1H, d, J= 13.1 Hz), 3.12 (1H, m), 2.84 (1H, m), 2.56 (1H, m), 2.12 (1H, m), 1.80 (2H, m); 13C NMR (CD2C12, 75 MHz) £203.2 (C), 169.8 (C), 135.4 (C), 131.5 (2CH), 129.0 (2CH), 128.1 (CH), 68.6 (C), 44.2 (CH2), 39.6 (CH2), 37.6 (CH2), 26.7 (CH2). HRESIMS calcd for Ci3H,4N02Na79Br ([M+Na]+): 318.0106; found 318.0099. 3-Benzyl-3-chloroazepane-2,4-dione (96): for a summary of !H and 13C NMR assignments, see Table 4.11. 'H NMR (CD2C12, 300 MHz) Si.25 (5H, m), 6.69 (1H, s, broad), 3.78 (1H, d, J = 13.1 Hz), 3.33 (1H, d, J = 13.1 Hz), 3.23 (1H, m), 2.95 (1H, m), 2.50 (1H, m), 2.13 (1H, m), 1.83 (2H, m); 13C NMR (CD2C12, 75 MHz) £203.8 (C), 169.3 (C), 134.6 (C), 131.6 (2CH), 128.9 (2CH), 128.1 (CH), 74.1 (C), 43.6 (CH2), 39.3 (CH2), 37.8 (CH2), 27.0 (CH2). HRESIMS calcd for C13Hi4N02NaCl ([M+Na]+): 274.0611; found 274.0609. 178 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.2. NMR data for 3-benzyl-3-chloroazepane-2,4-dione (96) (recorded in CD2C12). irbon No N1 2 3 4 5 6 7 8 9 10 11 12 1 3 C 8 (ppm)" 169.3 74.1 203.8 37.8 27.0 39.3 43.6 134.6 131.6 128.9 128.1 A NH 1  °/=\. \L-P 12 "H 8(ppm) (mult,/(Hz))"* 6.69 (s, broad) 2.13 (m), 2.50 (m) 1.83 (m) 2.95 (m), 3.23 (m) 3.33 (d, 7=13.1 Hz), 3.78 (d,y= 13.1 Hz) 7.25 (m) 7.25 (m) 7.25 (m) HMBC" (H->C) C3, C4, C6, C7 C4, C5, C7 C2, C5 C2, C3, C4, C5, C9, CIO C8, C12 C9 CIO " Recorded at 75 MHz.b Recorded at 300 MHz.c According to HMQC recorded at 300 MHz. Preparation of 3-benzvl-4-(fer?-butvldimethvlsilanvloxy)-l,5,6,7-tetrahydroazepin-2-one (105) To a solution of 3-benzylazepane-2,4-dione (78) (0.13 g, 0.58 mmol) in THF (3 mL) at - 78°C, was added LDA (0.32 mL, 0.58 mmol, 1.8 M). After stirring for 30 min., tert- butyldimethylsilyl chloride (TBDSCl) (0.13 g, 0.87 mmol) was added and the mixture stirred overnight. The reaction was quenched with NH4CI sat., extracted with EtOAc and dried (Na2SC>4). Solvent concentration in vacuo provided a colorless solid identified as pure (105) (0.18 g, 96%). No further purification was required according to TLC (100% EtOAc) and 'H NMR. *H NMR (CD2C12, 400 MHz) £7.28-7.19 (5H, m), 6.92 (IH, s, broad), 3.69 (2H, s), 3.10 179 Chapter 4. Progress towards the Synthesis of Ceratamines (2H, m), 2.48 (2H, t, J = 7.2 Hz), 1.96 (2H, m), 0.97 (9H, s), 0.24 (6H, s); UC NMR (CD2C12, 100 MHz) £175.7 (C), 157.0 (C), 141.9 (C), 128.8 (2CH), 128.6 (2CH), 126.1 (CH), 116.4 (C), 40.0 (CH2), 32.7 (CH2), 31.5 (CH2), 30.5 (CH2), 26.0 (3CH3), 18.6 (C), -3.19 (2CH3). HRESIMS calcd for a C19H29N02Na28Si ([M+Na]+): 354.1865; found 354.1859. Preparation of 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) and 3-benzvl-4-(fert- butyldimethylsilanyloxy)-l-methyl-l,5,6,7-tetrahydroazepin-2-one (106) 96 105 97 106 To a solution of 3-benzyl-3-chloroazepane-2,4-dione (96) (0.26 g, 1.0 mmol) in THF (5 mL) at -78°C, was added LDA (0.57 mL, 1.0 mmol, 1.8 M). After stirring for 30 min., methyl iodide (0.10 mL, 1.5 mmol) was added and the mixture stirred overnight. The reaction was quenched with NH4CI sat., extracted with EtOAc and dried (Na2SC>4). Solvent concentration in vacuo and silica gel column chromatography (100% EtOAc), provided (97) as a white powder (0.22 g, 79%). When 3-benzyl-4-(tert-butyldimethylsilanyloxy)-l,5,6,7-tetrahydroazepin-2-one (105) (0.051 g, 0.15 mmol), LDA (0.22 mL, 0.4 mmol, 1.8 M) and Mel (0.071 g, 0.5 mmol), were used following the same procedure as above (column chromatography with 80% EtOAc/hexanes as eluent), a colorless solid identified as desired (106) (0.021 g, 40%) was obtained. 3-Benzyl-3-chloro-l-methylazepane-2,4-dione (97): for a summary of H and C NMR assignments, see Tables 4.3 and 4.12. ]H NMR (CDC13, 400 MHz) £7.24-7.16 (5H, m), 3.83 (IH, d, J = 12.8 Hz), 3.29 (IH, d, J = 12.8 Hz), 3.28 (IH, m), 3.00 (3H, s), 2.85 (IH, ddd, J 180 Chapter 4. Progress towards the Synthesis of Ceratamines = 5.2, 5.2, 15.7 Hz), 2.33 (1H, m), 1.90 (1H, m), 1.75 (2H, m); 13C NMR (CDC13, 75 MHz) 8 203.6 (C), 166.7 (C), 134.2 (C), 131.0 (2CH), 128.4 (2CH), 127.6 (CH), 74.1 (C), 46.7 (CH2), 44.0 (CH2), 37.1 (CH2), 36.5 (CH3), 24.2 (CH2). HRESIMS calcd for Ci4H16N02Na35Cl ([M+Na]+): 288.0767; found 288.0764. 3-Benzyl-4-(te7t-butyldimethylsilanyloxy)-l-methyl-l,5,6,7-tetrahydroazepin-2-one (106): for a summary of *H and 13C NMR assignments, see Table 4.4. *H NMR (CD2C12, 400 MHz) £7.25- 7.15 (5H, m), 3.67 (2H, s), 3.24 (2H, m), 2.94 (3H, s), 2.36 (2H, t, / = 7.2 Hz), 2.00 (2H, m), 0.96 (9H, s), 0.22 (6H, s); 13C NMR (CD2C12, 100 MHz) £172.6 (C), 154.1 (C), 142.0 (C), 128.9 (2CH), 128.7 (2CH), 126.0 (CH), 117.2 (C), 48.4 (CH2), 34.5 (CH3), 32.9 (CH2), 30.8 (CH2), 29.0 (CH2), 26.0 (3CH3), 18.6 (C), -3.23 (2CH3). HRESIMS calcd for C20H31NO2NaSi ([M+Na]+): 368.2022; found 368.2028. Table 4.12. NMR data for 3-benzyl-3-chloro-l-methylazepane-2,4-dione (97) (recorded in CDC13). 12 Carbon No 2 3 4 5 6 7 8 9 10 11 12 13 1 3 C 5 (ppm)" 166.7 74.1 203.6 37.8 24.2 46.7 44.0 134.2 131.0 128.4 127.6 36.5 'H 5 (ppm) (mult, J (Hz))"* 1.90 (m), 2.33 (m) 1.75 (m) 2.85 (ddd, / = 5.2,5.2,15.7 Hz), 3.28 (m) 3.29 (d, J = 12.8 Hz), 3.83 (d, J = 12.8 Hz) 7.21 (m) 7.19 (m) 7.17 (m) 3.00 (s) HMBCb (H-»C) C3, C4, C6, C7 C5, C4, C7 C2, C5, C6 C2, C3, C4, C9, C10 C8, C12 C9 C10 C2,C7 'Recorded at 75 MHz.b Recorded at 400 MHz.' According to HMQC recorded at 400 MHz. 181 Chapter 4. Progress towards the Synthesis of Ceratamines Bromination of 3-benzyl-3-chloro-l-methvlazepane-2,4-dione (97) 1. LDA 2. NBS 99 A solution of 3-benzyl-3-chloro-l-methyl-azepane-2,4-dione (97) (0.061 g, 0.2 mmol) in THF (5 mL) at -78 °C was treated with LDA (0.27 mL, 0.50 mmol, 1.8 M), following shortly after the addition of A -̂bromosuccinimide (0.13 g, 0.74 mmol) in THF (5 mL), and overnight stirring at 25 °C. NH4CI was added and EtOAc extractions performed. The organic fractions were dried with Na2SC<4 and concentrated in vacuo to give a white powder, which was additionally purified by silica gel column chromatography (80% EtOAc/hexanes) to afford 3- benzyl-3-bromo-l-methyl-azepane-2,4-dione (99) as a white amorphous powder (0.056 g, 79%). *H NMR (CDCI3, 300 MHz) £7.28-7.15 (5H, m), 3.90 (1H, d, J = 12.7 Hz), 3.50 (1H, d, J = 12.7 Hz), 3.14 (1H, m), 2.99 (3H, s), 2.76 (1H, m), 2.48 (1H, m), 1.95 (1H, m), 1.75 (2H, m); 13C NMR (CDCI3, 75 MHz) £203.1 (C), 167.2 (C), 135.2 (C), 131.1 (2CH), 128.6 (2CH), 127.8 (CH), 69.1 (C), 47.3 (CH2), 44.4 (CH2), 37.0 (CH2), 36.9 (CH3), 24.3 (CH2). HRESMS calcd for Ci4Hi7N0279Br ([M+H]+): 310.0443; found 310.0439. Chlorination of 3-benzvl-4-(fert-butyldimethylsilanvloxy)-1 -methyl-1,5,6,7-tetrahvdroazepin-2- one(106) A solution of 3-benzyl-4-(fcrr-butyldimethylsilanyloxy)-l-methyl-1,5,6,7- tetrahydroazepin-2-one (106) (0.013 g, 0.038 mmol) in THF (5 mL) at -78 °C was treated with 182 Chapter 4. Progress towards the Synthesis of Ceratamines LDA (0.023 raL, 0.041 mmol, 1.8 M), following shortly after the addition of N- chlorosuccinimide (0.0076 g, 0.057 mmol) in THF (5 mL), and overnight stirring at 25 °C. NH4CI was added and EtOAc extractions performed. The organic fractions were dried with Na2SC>4 and concentrated in vacuo to give a colorless solid, which was additionally purified by silica gel column chromatography (100% EtOAc) to afford rearranged product (107) as a colorless solid (0.0065 g, 45%). For a summary of !H and 13C NMR assignments, see Tables 4.5 and 4.13. !H NMR (CD2C12, 400 MHz) J7.58 (1H, d, J = 7.6 Hz), 7.18 (1H, m), 7.14 (1H, m), 7.12 (1H, m), 6.87 (1H, d, J = 6.7 Hz), 3.79 (1H, m), 3.11 (1H, m), 3.08 (3H, s), 3.02 (1H, s), 1.80 (1H, m), 1.66 (1H, m), 1.63 (1H, m), 1.60 (1H, m), 0.80 (9H, s), 0.33 (3H, s), 0.14 (3H, s); 13C NMR (CD2CI2, 100 MHz) £206.4 (C), 168.1 (C), 138.6 (C), 134.7 (CH), 132.5 (CH), 129.0 (CH), 128.4 (CH), 127.0 (CH), 77.6 (C), 47.8 (CH2), 45.6 (CH), 39.2 (CH2), 36.8 (CH3), 27.8 (3CH3), 24.2 (CH2), 19.4 (C), -1.41 (CH3), -2.89 (CH3). HRESIMS calcd for C2oH3oN02NaSi35Cl ([M+Na]+): 402.1632; found 402.1633. 183 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.13. NMR data for 3-[(?ert-butyldimethylsilanyl)phenylmethyl]-3-chloro-l- methylazepane-2,4-dione (107) (recorded in CD2CI2). Proton *H No 5(ppm)(mult,/(Hz))» H5 1.63 (m) H5' 1.80 (m) H6 1.60 (m) H6' 1.66 (m) H7 3.1 l(m) H7' 3.79 (m) 8 3.02 (s) 10 7.58 (d, 7 = 7.6 Hz) 11 7.18 (m) 12 7.14 (m) 13 7.12 (m) 14 6.87 (d, 7 = 6.7 Hz) 15 3.08 (s) 16 0.14 (s) 17 0.33 (s) 19 0.80 (s) ordedat400MHz. COSY" (H-»H) H7, H7' H6, H6\ H7, H7' H5\ H7, H7' H5', H7, H7' H5, H5', H6, H6', H7' H5, H5', H6, H6\ H7 H11,H14 H10, H12 H11,H13 H12.H14 H10,H13 Preparation of 3-benzylidene-azepane-2,4-dione (120) / % J H NH J—^O PhCHO 50 HOAc cat., XJ piperidine cat. r= 120 A mixture of azepane-2,4-dione (50) (1.27 g, 10 mmol) and benzaldehyde (1.01 mL, 10 mmol) in benzene (50 mL), was refluxed for 3 h in the presence of catalytic amounts of HOAc and piperidine. Once at 25 °C, H2O (25 mL) was added and Et20 extractions performed. The organic extracts were combined and washed with H2O, HCl IM and NaHCOs sat., to be then 184 Chapter 4. Progress towards the Synthesis of Ceratamines dried (Na2SC>4) and concentrated in vacuo. The resulting yellow solid was additionally purified by washing with petroleum ether, and later identified as pure (120) (1.85 g, 86%). For a summary of !H and 13C NMR assignments, see Table 4.14. lH NMR (CD2C12, 400 MHz) £7.78 (1H, s), 7.66 (2H, m), 7.45-7.30 (4H, m), 6.45 (1H, s, broad), 3.38 (2H, m), 2.75 (2H, m), 2.01 (2H, m); 13C NMR (CD2C12, 100 MHz) £195.7 (C), 171.9 (C), 143.2 (CH), 134.2 (C), 133.8 (C), 131.4 (2CH), 131.3 (CH), 129.1 (2CH), 39.8 (CH2), 38.8 (CH2), 27.4 (CH2). HRESIMS calcd for Ci3Hi3N02Na ([M+Na]+): 238.0844; found 238.0844. Table 4.14. NMR data for 3-benzylidene-azepane-2,4-dione (120) (recorded in CD2C12). Carbon No Nl 2 3 4 5 6 7 8 9 10 11 12 1 3 C 8 (ppm)" 171.9 133.8 195.7 38.8 27.4 39.8 143.2 134.2 131.4 129.1 131.3 >H 8 (ppm) (mult, J (Hz))"-11 6.45 (s, broad) 2.75 (m) 2.01 (m) 3.38 (m) 7.78 (s) 7.66 (m) 7.45-7.30 (m) 7.45-7.30 (m) HMBCb (H->C) C3, C4, C6, C7 C4, C5, C7 C2, C5, C6 C2, C3, C4, C10 C8, C12 C12 Cll 'Recorded at 100 MHz. b Recorded at 400 MHz.' According to HMQC recorded at 400 MHz. 185 Chapter 4. Progress towards the Synthesis of Ceratamines Preparation of 2-(3,5-dibromo-4-methoxy-benzyl)-but-3-enoic (134) and 2-benzylbut-3-enoic (122) acids OH B i \ ^ L ^Br 1) K 2 C 0 3 , Mel, reflux 2) NBS, AICN, reflux ' Br Br Br 52 1. LDA "OH  2^ B z B r OH 1 .LDA, 2. 52 122 A mixture of 2,6-dibromo-4-methylphenol (15 g, 56 mmol) and K2CO3 (11.6, 84 mmol) was stirred at 25 °C for 30 min, whereupon Mel (5.2 mL, 84 mmol) was added and the resulting slurry refluxed for 3 h. Once at room temperature, the excess K2CO3 was filtrated off and the crude concentrated in vacuo, to give a white solid. This solid was dissolved in CCI4, and NBS (10 g, 56 mmol) followed by l,l'-azobis(cyanocyclohexane) (AICN) (0.5 g, 2 mmol) were added to the solution. The resulting mixture was refluxed for 12 h, to be then separated from an insoluble precipitate (JV-succinimide) by filtration and concentrated. The obtained crystalline white solid (19.8 g, 98%) was pure by TLC and identified by 'H-NMR as l,3-dibromo-5- bromomethyl-2-methoxybenzene (52). In a second reaction, a mixture of diisopropylamine (21.8 mL, 154 mmol) and BuLi (96 mL, 154 mmol, 1.6 M) in THF (150 mL) was stirred at 0 °C for 10 min, to be then treated with vinylacetic acid (6.4 mL, 75 mmol). After stirring for 45 min at 0 °C, l,3-dibromo-5- bromomethyl-2-methoxybenzene (52) (27 g, 75 mmol) was added and the solution stirred for 30 min at 0 °C and 2 h at 25 °C. H2O was slowly added to the reaction crude and the pH for the aqueous layer was adjusted to 2.5 using HC1 0.1 M. After EtOAc extractions, the organic extracts were combined, dried (Na2SC>4) and concentrated in vacuo. The resulting residue was 186 Chapter 4. Progress towards the Synthesis of Ceratamines poured into a silica gel column and eluted using a gradient of EtOAc/hexanes mixtures (from 30% to 100%), to afford after solvent evaporation a white amorphous solid (16.0 g, 58%) of desired product (134). When the previous procedure was followed using LDA (13.4 mL, 24.1 mmol, 1.8 M), vinylacetic acid (1.0 mL, 11.8 mmol) in THF (10 mL), and benzyl bromide (1.43 mL, 12.0 mmol), 2-benzyl-but-3-enoic acid (122) was obtained as a colorless liquid (1.4 g, 68%). l,3-Dibromo-5-bromomethyl-2-methoxybenzene (52): lH NMR (CDC13, 300 MHz) £7.51 (2H, s), 4.34 (2H, s), 3.86 (3H, s); 13C NMR (CDC13, 75 MHz) £155.7 (C), 137.6 (C), 134.6 (2CH), 119.6 (2C), 62.1 (CH3), 32.0 (CH2). 2-(3,5-Dibromo-4-methoxy-benzyl)-but-3-enoic acid (134): !H NMR (CDC13, 300 MHz) £10.84 (1H, s, broad), 7.30 (2H, s), 5.79 (1H, m), 5.15 (2H, m), 3.83 (3H, s), 3.27 (1H, ddd, / = 7.3, 7.7, 8.1 Hz), 3.00 (1H, dd, J = 7.3, 13.9 Hz), 2.74 (1H, dd, J = 7.3, 13.9 Hz); 13C NMR (CDCI3, 75 MHz) £180.0 (C), 154.2 (C), 138.5 (C), 135.3 (CH), 134.6 (2CH), 120.5 (CH2), 119.3 (2C), 62.0 (CH3), 52.7 (CH2), 37.9 (CH). HRESIMS calcd for C12Hn0379Br2 ([M+H]+): 360.9075; found 360.9071. 2-Benzylbut-3-enoic acid (122): 'H NMR (CDC13, 400 MHz) £7.36-7.12 (5H, m), 5.87 (1H, m), 5.15 (2H, m), 3.36 (1H, ddd, / = 7.6, 7.7, 7.8 Hz), 3.14 (1H, dd, J = 7.3, 13.7 Hz), 2.88 (1H, dd, J = 7.5, 13.7 Hz); 13C NMR (CDCI3, 100 MHz) £ 179.9 (C), 138.2 (C), 134.6 (CH), 129.0 (2CH), 128.3 (2CH), 126.5 (CH), 118.3 (CH2), 51.7 (CH), 38.0 (CH2). HRESIMS calcd for CnHnCb ([M+H]+): 175.0759; found 175.0758. 187 Chapter 4. Progress towards the Synthesis of Ceratamines Preparation of 2-(3,5-dibromo-4-methoxvbenzvl)-but-3-enoic acid but-3-envlamide (135) and 2- benzvlbut-3-enoic acid but-3-enylamide (123) 134 135 122 123 A solution of di-f-butyl iminodicarboxilate (44.0 g, 202 mmol) in CH3CN (200 mL) was treated with CS2CO3 (65 g, 200 mmol), followed by addition of 4-bromobutene (26 mL, 256 mmol) while stirring. The resulting slurry was refluxed for 18 h, whereupon the insoluble salts were removed by filtration and the crude concentrated in vacuo. Dilution of the obtained residue in CH2C12 and addition of TFA (60 mL, 300 mmol) with stirring for 4 h at 25 °C provided the trifluoroacetic salt of but-3-enylamine, which was precipitated as a gray powder by addition of HCl/Et20 (120 mL, 240 mmol). This precipitate was characterized by NMR and MS as but-3- enylamine (121) hydrochloric salt (14.5 g, 67%). To a solution of 2-(3,5-dibromo-4-methoxy-benzyl)-but-3-enoic acid (134) (32.3 g, 88 mmol) in CH2CI2 (150 mL) at 0 °C, was added oxalic chloride (8.38 mL, 97 mmol) and ten drops of DMF. The mixture was stirred for 16 h at 25 °C, to be then concentrated in vacuo under N2 atmosphere. The resulting red and oily residue was diluted in CH2CI2 (30 mL) and added at 0 °C to a slurry of but-3-enylamine (121) hydrochloric salt (14.2 g, 132 mmol) and triethylamine (37 mL, 264 mmol), with the consequent formation of HCl gas. After stirring at 25 °C for 12 h, a H2O partition was made and EtOAc extractions performed. Drying (Na2SC»4) and concentration 188 Chapter 4. Progress towards the Synthesis of Ceratamines of the organic extracts provided a reddish residue, which was purified by gradient silica gel column chromatography (40% to 100% EtOAc/hexanes) to afford 2-(3,5-dibromo-4- methoxybenzyl)-but-3-enoic acid but-3-enylamide (135) as a gray powder (25.5 g, 69%). Following the second procedure above, 2-benzylbut-3-enoic acid but-3-enylamide (123) (3.6 g, 87%) was prepared using the following amounts of reagents: 2-benzylbut-3-enoic acid (122) (3.1 g, 17.9 mmol), oxalic chloride (5.53 mL, 63.4 mmol), but-3-enylamine.HCl (121) (4.55 g, 42.3 mmol) and triethylamine (8.84 mL, 63.4 mmol). But-3-enylamine.HCl (121): lH NMR (CD3OD, 300 MHz) £5.81 (IH, m), 5.18 (2H, m), 4.87 (2H, s, broad), 3.00 (2H, t, J = 7.4 Hz), 2.42 (2H, q, J = 7.1 Hz); 13C NMR (CD3OD, 75 MHz) £ 134.3 (CH), 119.3 (CH2), 40.1 (CH2), 32.8 (CH2). HRESEVIS calcd for C4H10N ([M+H]+): 72.0813; found 72.0812. 2-(3,5-Dibromo-4-methoxybenzyl)-but-3-enoic acid but-3-enylamide (135): lH NMR (CDCI3, 400 MHz) £7.29 (2H, s), 5.79 (IH, m), 5.64 (IH, m), 5.49 (IH, s, broad), 5.12 (2H, m), 4.99 (2H, m), 3.83 (3H, s), 3.26 (2H, m), 3.09 (IH, dd, / = 7.4, 13.5 Hz), 2.94 (IH, ddd, J = 7.4, 7.5, 8.0 Hz), 2.69 (IH, dd, J= 6.7, 13.5 Hz), 2.16 (2H, m); 13C NMR (CDC13, 100 MHz) £171.9 (C), 152.2 (C), 138.1 (C), 136.0 (CH), 134.9 (CH), 133.1 (2CH), 118.0 (CH2), 117.6 (2C), 116.9 (CH2), 60.4 (CH3), 53.0 (CH), 38.5 (CH2), 36.7 (CH2), 33.6 (CH2). HRESMS calcd for Ci6H19N02Na79Br81Br ([M+Na]+): 439.9660; found 439.9651. 2-Benzylbut-3-enoic acid but-3-enylamide (123): *H NMR (CDCI3, 400 MHz) £7.29-7.11 (5H, m), 5.85 (IH, m), 5.62 (IH, m), 5.46 (IH, s, broad), 5.10 (2H, m), 4.96 (2H, m), 3.24 (2H, m), 3.16 (IH, dd, J = 7.4, 13.6 Hz), 3.01 (IH, ddd, J = 7.5, 7.6, 7.8 Hz), 2.79 (2H, dd, J = 7.0, 13.4 189 Chapter 4. Progress towards the Synthesis of Ceratamines Hz), 2.09 (2H, m); 13C NMR (CDC13, 100 MHz) £172.2 (C), 139.2 (C), 136.6 (CH), 135.0 (CH), 129.0 (2CH), 128.2 (2CH), 126.2 (CH), 117.8 (CH2), 117.1 (CH2), 54.0 (CH), 38.20 (CH2), 38.17 (CH2), 33.5 (CH2). HRESIMS calcd for Ci5H19NONa ([M+Na]+): 252.1364; found 252.1368. Preparation of 3-(3,5-dibromo-4-methoxv-benzyl)-l,3,6,7-tetrahvdroazepin-2-one (136) and 3- benzyl-l,3,6,7-tetrahydroazepin-2-one (124) C I . I C l ' ^ P h PCy3 CH2OI2 I \ M e s - N ^ N ~ M e s ClJ „ , , R u = ^  t. C I ' J Ph Yk PCy3 C H gC I2 135 136 123 124 A solution of 2-(3,5-dibromo-4-methoxybenzyl)-but-3-enoic acid but-3-enylamide (135) (0.70 g, 1.68 mmol) in degassed CH2C12 (110 mL), was treated with Grubbs catalyst 2nd generation (0.1 g, 0.12 mmol) and refluxed for 18 h. Once at room temperature, the crude was concentrated and poured into a silica gel column (70% EtOAc/hexanes) followed by TLC. Concentration of the fractions containing product afforded (136) as white amorphous powder (0.69 g, 84%). This same procedure, using 2-benzyl-but-3-enoic acid but-3-enylamide (123) (0.33 g, 1.45 mmol) together with Grubbs catalyst (0.050 g, 0.06 mmol), provided 3-benzyl-1,3,6,7- tetrahydro-azepin-2-one (124) (0.13 g, 61%) as a colorless solid. 3-(3,5-Dibromo-4-methoxy-benzyl)-l,3,6,7-tetrahydroazepin-2-one (136): :H NMR (CDCI3, 300 MHz) £7.33 (2H, s), 7.19 (IH, dd, J = 5.8, 6.5 Hz), 5.51 (IH, m), 5.16 (IH, m), 3.75 (3H, s), 190 Chapter 4. Progress towards the Synthesis of Ceratamines 3.73 (1H, m), 3.60 (1H, m), 3.11 (1H, dd, J = 6.9, 14.2 Hz), 3.10 (1H, m), 2.54 (1H, dd, J = 7.7, 14.5 Hz), 2.16 (2H, m); 13C NMR (CDC13, 75 MHz) S 177.8 (C), 153.7 (C), 140.5 (C), 134.8 (2CH), 131.1 (CH), 127.2 (CH), 119.1 (2C), 62.0 (CH3), 44.0 (CH2), 40.1 (CH), 36.9 (CH2), 31.5 (CH2). HRESEVIS calcd for Ci4Hi5N02Na79Br81Br ([M+Na]+): 411.9347; found 411.9343. 3-Benzyl-l,3,6,7-tetrahydroazepin-2-one (124): for a summary of !H and 13C NMR assignments, see Table 4.6. 'H NMR (CDC13, 400 MHz) J7.33-7.15 (5H, m), 6.44 (1H, s, broad), 5.56 (1H, m), 5.32 (1H, m), 3.85 (1H, m), 3.71 (1H, m), 3.29 (1H, dd, J = 5.8, 14.2 Hz), 3.16 (1H, m), 2.73 (1H, dd, J = 9.4, 14.2 Hz), 2.24 (2H, m); 13C NMR (CDC13, 100 MHz) S 176.8 (C), 139.9 (C), 129.6 (CH), 129.2 (2CH), 128.4 (2CH), 127.1 (CH), 126.1 (CH), 42.8 (CH), 38.8 (CH2), 36.5 (CH2), 30.1 (CH2). HRESEVIS calcd for Ci3Hi5NONa ([M+Na]+): 224.1051; found 224.1050. Preparation of 2-(3,5-dibromo-4-methoxvbenzyl)-8-oxa-4-aza-bicvclor5.1.01octan-3-one (162) and 2-benzyl-8-oxa-4-aza-bicvclor5.1.01octan-3-one (125) Br Br 136 162 124 125 A solution of 3-(3,5-dibromo-4-methoxybenzyl)-l,3,6,7-tetrahydro-azepin-2-one (136) (0.1 g, 0.26 mmol) in CH2C12 (10 mL) was stirred in the presence of m-chloroperbenzoic acid (0.11 g, 0.51 mmol, 77%), until TLC analysis showed the absence of starting material (12-72 h). Water partition and EtOAc extraction gave a crystalline solid after solvent evaporation, which was purified by silica gel column chromatography (70% EtOAc/hexanes) to obtain desired product (162) (0.043 g, 42%) as a white powder. With 3-benzyl-l,3,6,7-tetrahydroazepin-2-one 191 Chapter 4. Progress towards the Synthesis of Ceratamines (124) (0.95 g, 4.7 mmol) and m-CPBA (2.0 g, 9.5 mmol) as starting materials, 2-benzyl-8-oxa- 4-aza-bicyclo[5.1.0]octan-3-one (125) (0.36 g, 39%) was successfully prepared. 2-(3,5-Dibromo-4-methoxybenzyl)-8-oxa-4-aza-bicyclo[5.1.0]octan-3-one (162): !H NMR (CDC13, 400 MHz) J7.44 (2H, s), 6.50 (IH, m, broad), 3.83 (3H, s), 3.44 (IH, m), 3.30 (IH, m) 3.18 (IH, m), 3.11 (IH, m) 3.00 (IH, m), 2.95 (IH, m), 2.91 (IH, m) 2.20 (IH, m), 1.96 (IH, m); 13C NMR (CDCI3, 100 MHz) S 172.8 (C), 152.6 (C), 137.9 (C), 133.3 (2CH), 117.8 (2C), 60.5 (CH3), 55.7 (CH), 53.6 (CH), 43.8 (CH), 36.1 (CH2), 34.5 (CH2), 28.2 (CH2). HRESIMS calcd for Ci4Hi5N03Na79Br2 ([M+Na]+): 425.9316; found 425.9315. 2-Benzyl-8-oxa-4-aza-bicyclo[5.1.0]octan-3-one (125): for a summary of !H and 13C NMR assignments, see Table 4.15. *H NMR (CDC13, 400 MHz) £7.29 (5H, m), 6.42 (IH, m, broad), 3.38 (IH, m), 3.35 (IH, m), 3.18 (IH, m), 3.08 (IH, m), 3.01 (IH, m), 2.98 (IH, m), 2.93 (IH, m), 2.17 (IH, m), 1.94 (IH, m); 13C NMR (CDC13, 100 MHz) J 173.7 (C), 138.9 (C), 129.0 (2CH), 128.4 (2CH), 126.3 (CH), 55.8 (CH), 53.5 (CH), 43.8 (CH), 36.1 (CH2), 35.6 (CH2), 28.2 (CH2). HRESIMS calcd for C13H15N02Na ([M+Na]+): 240.1000; found 240.1001. 192 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.15. NMR data for 2-benzyl-8-oxa-4-aza-bicyclo[5.1.0]octan-3-one (125) (recorded in CDC13). arbon No 1 2 3 N4 5 6 7 08 9 10 11 12 13 "c 6 (ppm)" 55.8 43.8 173.7 36.1 28.2 53.5 35.6 138.9 129.0 128.4 126.3 5 8 <C^ H VTj° ^==^13 lH 8(ppm)(mult,7(Hz))bc 2.93 (m) 3.08 (m) 6.42 (s, broad) 2.98 (m), 3.35 (m) 1.94 (m), 2.17 (m) 3.18 (m) 3.01 (m), 3.38 (m) 7.28 (m) 7.29 (m) 7.20 (m) HMBCb (H-»C) C7 C2 C7 C1,C7 C6 C1,C2,C3,C10,( C10,C13 C10.C13 Cll "Recorded at 100 MHz." Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. Preparation of toluene-4-sulfonic acid 6-benzvl-5-hvdroxy-7-oxoazepan-4-vl ester (137) XSL-o. TsOH.H zO ° CH2CI2/acetone 137 2-Benzyl-8-oxa-4-aza-bicyclo[5.1.0]octan-3-one (125) (0.55 g, 2.5 mmol) was diluted in CH2Cl2/acetone 1:4 (40 mL), and stirred at room temperature in the presence of p-toluensulfonic acid monohydrate (2.7 g, 14.2 mmol), until TLC analysis (100% EtOAc) indicated the absence of starting material (48 h). Silica gel column chromatography (100% EtOAc) provided (137) as a white solid (0.24 g, 25%). For a summary of 'H and 13C NMR assignments, see Table 4.16. lU NMR (CDCI3, 300 MHz) £7.77 (2H, m), 7.33 (2H, m), 7.20 (2H, m), 7.20 (2H, m), 7.13 (IH, 193 Chapter 4. Progress towards the Synthesis of Ceratamines m), 6.69 (IH, s, broad), 2.43 (3H, s), 4.47 (IH, m), 4.20 (IH, m), 3.55 (IH, m), 3.20 (2H, m), 3.18 (IH, m), 3.11 (IH, m), 3.00 (IH, m), 2.43 (IH, m); 13C NMR (CDC13, 100 MHz) S 173.9 (C), 145.2 (C), 138.9 (C), 133.4 (C), 129.9 (2CH), 129.1 (2CH), 128.3 (2CH), 127.7 (2CH), 126.2 (CH), 78.6 (CH), 67.4 (CH), 41.4 (CH), 36.3 (CH2), 33.7 (CH2), 31.5 (CH2), 21.6 (CH3). HRESIMS calcd for C2oH24N0532S ([M+H]+): 390.1375; found 390.1372. Table 4.16. NMR data for toluene-4-sulfonic acid 6-benzyl-5-hydroxy-7-oxoazepan-4-yl ester (137) (recorded in CDC13). 15 17  tOi-Wr-A2 HO ; r \ > 12 1 3 C 8 (ppm)a 36.3 31.5 67.4 78.6 41.4 173.9 33.7 138.9 128.3 129.1 126.2 133.4 127.7 129.9 145.2 21.6 lB 5(ppm)(muIt,/(Hz))hc 6.69 (s, broad) 3.20 (m) 3.11 (m), 3.18 (m) 4.20 (m) 4.47 (m) 3.55 (m) 2.43 (m), 3.00 (m) 7.20 (m) 7.20 (m) 7.13 (m) 7.77 (m) 7.33 (m) 2.43 (s) HMBCb (H->C) C7 C3, C4, CI, C8 C7,C8 C5, C6, C7, C9 C8 C9, C10,C12 Cll C13,C15 C13, C14, C16, C17 C15, C16 b  Recorded at 300 MHz.c According to HMQC recorded at 400 MHz. Carbon No Nl 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 "Recorded at 100 MHz. 194 Chapter 4. Progress towards the Synthesis of Ceratamines Preparation of toluene-4-sulfonic acid 6-benzyl-5J-dioxoazepan-4-yl ester (126) 137 , DMP °"o ° CH2CI2/acetone ° 126 Toluene-4-sulfonic acid 6-benzyl-5-hydroxy-7-oxoazepan-4-yl ester (137) (0.088 g, 0.22 mmol) was diluted in C^C^/acetone 1:1 (4 mL), and stirred at room temperature in the presence of Dess-Martin periodinane (0.19 g, 0.45 mmol), until TLC analysis (100% EtOAc) indicated the absence of starting material (2 h). Silica gel column chromatography (80% EtOAc/hexanes) provided toluene-4-sulfonic acid 6-benzyl-5,7-dioxo-azepan-4-yl ester (126) as a colorless solid (0.047 g, 54%). *H NMR (CDC13, 300 MHz) £7.76 (2H, m), 7.48 (IH, m), 7.38-7.15 (4H, m), 7.07 (2H, m), 6.35 (IH, s, broad), 5.12 (IH, dd, J = 7.0, 11.4 Hz), 3.99 (IH, dd, J = 6.7, 6.9 Hz), 3.71 (IH, m), 3.30 (IH, dd, J = 6.7, 14.0 Hz), 2.94 (IH, dd, J = 6.6, 14.2 Hz), 2.45 (2H, m), 2.43 (3H, s), 2.05 (IH, m); 13C NMR (CDC13, 100 MHz) £198.7 (C), 167.7 (C), 145.7 (C), 139.0 (C), 131.7 (C), 130.2 (2CH), 129.3 (2CH), 128.4 (2CH), 128.0 (2CH), 126.3 (CH), 82.5 (CH), 54.0 (CH), 36.7 (CH2), 35.6 (CH2), 30.9 (CH2), 21.7 (CH3). HRESMS calcd for C2oH2iN05Na32S ([M+Na]+): 410.1038; found 410.1033. Preparation of 4-bromo-3-(3,5-dibromo-4-methoxybenzvl)-5-hydroxvazepan-2-one (138) NBS Acetone/H20 138 195 Chapter 4. Progress towards the Synthesis of Ceratamines 3-(3,5-Dibromo-4-methoxybenzyl)-l,3,6,7-tetrahydroazepin-2-one (136) (0.19 g, 0.5 mmol) was dissolved in a mixture acetone/H20 (3:1, 2 mL) and stirred in the presence of NBS (0.12 g, 0.65 mmol), until TLC analysis (70% EtOAc/hexanes) showed the absence of starting material (72 h). The crude was concentrated in vacuo and purified by silica gel column chromatography (50, then 70% EtOAc/hexanes) to afford a white crystalline solid (0.24 g, 98%) identified as desired halohydrine (138). lH NMR (CDC13, 300 MHz) £7.33 (2H, s), 6.84 (1H, dd, J = 5.4, 6.5 Hz), 4.14 (1H, m), 3.83 (1H, m), 3.76 (3H, s), 3.70 (1H, m), 3.60 (1H, dd, J = 5.8, 9.2 Hz), 3.08 (1H, dd, J = 5.4, 14.6 Hz), 2.90 (1H, m), 2.63 (1H, dd, J = 5.4, 14.6 Hz), 2.15 (1H, m), 1.72 (1H, m); 13C NMR (CDC13, 100 MHz) £174.3 (C), 152.6 (C), 138.1 (C), 133.4 (2CH), 118.0 (2C), 70.8 (CH), 60.6 (CH3), 51.0 (CH), 42.6 (CH), 34.8 (CH2), 33.9 (CH2), 30.6 (CH2). HRESIMS calcd for Ci4H16N03Na79Br3 ([M+Naf): 505.8578; found 505.8577. Preparation of 4-bromo-3-(3,5-dibromo-4-methoxybenzvl)-azepane-2,5-dione (140) and 4- bromo-3-(3,5-dibromo-4-methoxybenzyl)-3,4-dihvdro-lH-azepine-2,5-dione (139) 138 Dess-Martin periodinane (0.2 g, 0.47 mmol) was added to a solution of 4-bromo-3-(3,5- dibromo-4-methoxybenzyl)-5-hydroxyazepan-2-one (138) (0.13, 0.28 mmol) in CH2CI2 (5 mL), and the mixture stirred at 25 °C until TLC revealed the absence of starting material (24 h). After solvent removal, the residue was purified by silica gel column chromatography (70% EtOAc/hexanes) to provide a 1:1 mixture (0.13 g, 54%) of desired (140) and byproduct (139), as a white powder. 196 Chapter 4. Progress towards the Synthesis of Ceratamines 4-Bromo-3-(3,5-dibromo-4-methoxybenzyl)-azepane-2,5-dione (140) was more efficiently prepared by adding IBX (0.12 g, 0.42 mmol) to a solution of starting material (138) (0.16 g, 0.3 mmol) in DMSO (3 mL), and stirring for 5 h at 25 °C. The solvent was removed using high vacuum, and the resulting residue purified as above, to afford (140) (0.13 g, 84%) as a colorless powder. 4-Bromo-3-(3,5-dibromo-4-methoxy-benzyl)-azepane-2,5-dione (140): for a summary of !H and 13C NMR assignments, see Table 4.17. !H NMR (CDC13, 300 MHz) £7.35 (1H, s, broad), 7.34 (2H, s), 4.11 (1H, m), 3.82 (3H, s), 3.45 (1H, m), 3.37 (1H, m), 3.32 (1H, m), 3.23 (1H, m), 3.11 (1H, m), 2.71 (1H, dd, J= 8.1, 10.9 Hz), 2.50 (1H, m); 13C NMR (CDCI3, 75 MHz) £200.6 (C), 172.3 (C), 153.0 (C), 136.7 (C), 133.4 (2CH), 118.2 (2C), 60.5 (CH3), 52.0 (CH), 44.5 (CH), 40.1 (CH2), 36.8 (CH2), 33.4 (CH2). HRESIMS calcd for Ci4Hi5N0379Br3 ([M+H]+): 481.8602; found 481.8607. 4-Bromo-3-(3,5-dibromo-4-methoxy-benzyl)-3,4-dihydro-lH-azepine-2,5-dione (139): }H NMR (CDCI3, 300 MHz) £8.91 (1H, d, J = 6.9 Hz), 7.34 (2H, s), 6.58 (1H, dd, J = 6.9, 10.4 Hz), 5.53 (1H, d, J= 10.0 Hz), 4.19 (1H, m), 3.84 (3H, s), 3.33 (1H, dd, J = 4.6, 11.5 Hz), 3.22 (1H, m), 2.74 (1H, dd, J = 6.5, 11.6 Hz); 13C NMR (CDC13, 75 MHz) £199.2 (C), 170.2 (C), 154.7 (C), 136.67 (C), 134.7 (2CH), 136.58 (CH), 120.0 (2C), 110.0 (CH), 62.1 (CH3), 53.5 (CH), 44.7 (CH), 34.8 (CH2). HRESIMS calcd for Ci4Hi2N03Na79Br28,Br ([M+Na]+): 503.8245; found 503.8243. 197 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.17. NMR data for 4-bromo-3-(3,5-dibromo-4-methoxybenzyl)-azepane-2,5-dione (140) (recorded in CDC13). Carbon No Nl 2 3 4 5 6 7 8 9 10 11 12 13 8 (ppm)' 172.3 44.5 52.0 200.6 40.1 36.8 33.4 136.7 133.4 118.2 153.0 60.5 W " Br  < ^ ff\-Br Br 13 'H 5 (ppm) (mult,/(Hz))"-0 7.35 (s, broad) 3.11 (m) 4.11 (m) C7 C2, C2, C5, C3, HMBC" (H-»C) C8, C9 C5, C6, C8 2.50 (m), 3.32 (m) C4, C5, C8 3.45 (m), 3.37 (m) C2, C5, C6 2.71 (dd, J = 8.1,10.9 Hz), 3.23 (m) C2, C3, C4, C9, C10 7.34 (s) 3.82 (s) C3,C8,C11,C12 C12 " Recorded at 75 MHz. " Recorded at 300 MHz.c According to HMQC recorded at 300 MHz. Reaction of 4-bromo-3-(3.5-dibromo-4-methoxybenzyl)-azepane-2.5-dione (140) and 4-bromo- 3-(3,5-dibromo-4-methoxvbenzvl)-3,4-dihydro-lH-azepine-2,5-dione (139) with guanidine reagents o s \ / N H 2 NH 2 + . N H 2 H 2 N = ^ _ , B or C CI" NH 2 A: DMF, 25 °C B: EtONa, EtONa, 25 °C C: DMF, 70 °C 144 ^ ^ N = < , DMF, 25 °C, or NH2 1/ " V B r T ^ N H . ~ H2N=«C ' D M F ' E , 3 N ' 7 0 C ci- NH 2 198 Chapter 4. Progress towards the Synthesis of Ceratamines 4-Bromo-3-(3,5-dibromo-4-methoxybenzyl)-azepane-2,5-dione (140) (0.058 g, 0.12 mmol) and acetylguanidine (0.36 mmol) were stirred in DMF at 25 °C, until TLC analysis showed total consumption of starting material (5 days). After solvent removal, the residue was purified by silica gel chromatography (70% EtOAc/hexanes), to afford 3-(3,5-dibromo-4- methoxy-benzyl)-6,7-dihydro-lH-azepine-2,5-dione (144) as a colorless solid (0.042 g, 87%). This product was also obtained by reacting (140) (0.075 g, 0.15 mmol) with guanidine hydrochloride (0.1 g, 1 mmol) in DMF at 70 °C (0.044 g, 70%). When the same starting material (0.010 g, 0.021 mmol) was diluted in dry EtOH and added to a mixture of EtONa/EtOH (0.11 g of Na in 5 mL EtOH) and guanidine hydrochloride (0.5 g, 5 mmol), the procedure above afforded 3-(3,5-dibromo-4-methoxybenzyl)-lH-azepine-2,5-dione (141) (0.0016 g, 18%) also as a colorless solid. In another reaction, byproduct (139) (0.10 g, 0.22 mmol) was diluted in DMF (5 mL) and stirred in the presence of acetylguanidine (0.082 g, 0.81 mmol) at 25 °C. The previous work-up and isolation procedure provided more 3-(3,5-dibromo-4-methoxybenzyl)-lH-azepine-2,5-dione (141) (0.062 g, 71%). When byproduct (139) (0.34 g, 0.7 mmol) was reacted with guanidine hydrochloride (0.20 g, 2.1 mmol) in DMF at 70 °C for 18 h, the only product isolated was also (141) (0.032 g, 11%). 3-(3,5-Dibromo-4-methoxybenzyl)-6,7-dihydro-lH-azepine-2,5-dione (144): for a summary of lH and 13C NMR assignments, see Table 4.18. lH NMR (CDC13, 300 MHz) ST.33 (2H, s), 7.30 (1H, s, broad), 6.17 (1H, s), 3.84 (3H, s), 3.71 (2H, s), 3.44 (2H, m), 2.73 (2H, m); 13C NMR (CDCI3, 75 MHz) £200.2 (C), 168.2 (C), 153.1 (C), 145.1 (C), 135.7 (C), 133.3 (2CH), 132.6 (CH), 118.3 (2C), 60.6 (CH3), 45.0 (CH2), 39.2 (CH2), 36.8 (CH2). HRESIMS calcd for Ci4Hi3N03Na79Br2 ([M+Na]+): 423.9160; found 423.9164. 199 Chapter 4. Progress towards the Synthesis of Ceratamines 1 1 "\ 3-(3,5-Dibromo-4-methoxybenzyl)-lH-azepine-2,5-dione (141): for a summary of H and C NMR assignments, see Table 4.7. JH NMR (DMSO, 400 MHz) S\ 1.10 (IH, s, broad), 7.58 (2H, s), 6.94 (IH, d, / = 2.2 Hz), 6.81 (IH, d, J = 10.0 Hz), 5.67 (IH, dd, J = 2.2, 10.0 Hz), 3.76 (2H, s), 3.76 (3H, s); 13C NMR (DMSO, 100 MHz) J 185.4 (C), 165.1 (C), 152.0 (C), 142.7 (C), 141.9 (CH), 138.0 (C), 136.1 (CH), 133.3 (2CH), 117.2 (2C), 111.5 (CH), 60.3 (CH3), 38.5 (CH2). HRESIMS calcd for Ci4Hi2N0379Br2 ([M+H]+): 399.9184; found 399.9189. Table 4.18. NMR data for 3-(3,5-dibromo-4-methoxy-benzyl)-6,7-dihydro-lH-azepine-2,5- dione (144) (recorded in CDC13). Carbon No Nl 2 3 4 5 6 7 8 9 10 11 12 13 8 (ppm)* 168.2 145.1 132.6 200.2 45.0 36.8 39.2 135.7 133.3 118.3 153.1 60.6 Br 13 5(ppm)(mult,/(Hz))b,c 7.30 (s, broad) 6.17 (s) 2.73 (m) 3.44 (m) 3.71 (s) 7.33 (s) 3.84 (s) HMBCb (H-*C) C2, C3, C8 C2,C5 C2, C5, C6 C2, C3, C4, C10 C8,C11,C12 C12 "Recorded at 75 MHz." Recorded at 300 MHz.c According to HMQC recorded at 300 MHz. 200 Chapter 4. Progress towards the Synthesis of Ceratamines Synthesis of 3-(3,5-dibromo-4-methoxybenzvl)-l-methvl-6,7-dihydro-lH-azepine-2,5-dione (145) and 3-(3,5-dibromo-4-methoxybenzvl)-l-methyl-lH-azepine-2,5-dione (142) 144 141 145 142 The reflux of starting materials (144) (0.044 g, 0.10 mmol) and (141) (0.025 g, 0.063 mmol), with K2CO3 (0.026 g, 0.19 mmol) and iodomethane (0.020 mL, 0.31 mmol) in acetone (10 mL) for 1 h, afforded after silica gel column chromatography (70% EtOAc/hexanes) desired products (145) (0.045 g, 99%) and (142) (0.023 g, 86%) as colorless solids. 3-(3,5-Dibromo-4-methoxybenzyl)-l-methyl-6,7-dihydro-lH-azepine-2,5-dione (145): *H NMR (CDCI3, 400 MHz) £7.33 (2H, s), 7.19 (IH, s, broad), 6.16 (IH, s), 3.83 (3H, s), 3.70 (2H, s), 3.44 (2H. m), 2.95 (3H, s), 2.71 (2H, m); 13C NMR (CDCI3, 100 MHz) £200.1 (C), 167.9 (C), 153.1 (C), 145.1 (C), 135.7 (C), 133.2 (2CH), 132.5 (CH), 118.2 (2C), 60.6 (CH3), 45.0 (CH2), 39.3 (CH2), 36.8 (CH2), 24.7 (CH3). 3-(3,5-Dibromo-4-methoxybenzyl)-l-methyl-lH-azepine-2,5-dione (142): !H NMR (CDC13, 300 MHz) £7.35 (2H, s), 6.815 (IH, d, J= 10.4 Hz), 6.806 (IH, d, / = 2.7 Hz), 5.76 (IH, dd, J = 2.7, 10.8 Hz), 3.83 (3H, s), 3.77 (2H, s), 3.43 (3H, s); 13C NMR (CDC13, 75 MHz) £185.4 (C), 164.8 (C), 153.0 (C), 143.2 (C), 140.9 (CH), 139.7 (C), 136.3 (CH), 133.2 (2CH), 118.2 (2C), 112.4 (CH), 60.6 (CH3), 42.2 (CH2), 41.0 (CH3). HRESIMS calcd for Ci5Hi4N0379Br2 ([M+H]+): 413.9340; found 413.9335. 201 Chapter 4. Progress towards the Synthesis of Ceratamines Preparation of 5-azido-3-(3,5-dibromo-4-methoxv-benzyl)-4-iodoazepan-2-one (151) To a slurry of NaN3 (0.15 g, 2.3 mmol) in CH3CN (5 mL) at 0 °C, was added IC1 (2 mL, 2 mmol, 1 M) over a period of 20 min, and the mixture was stirred for 10 additional min, to be then treated drop by drop with 3-(3,5-dibromo-4-methoxybenzyl)-l,3,6,7-tetrahydro-azepin-2- one (136) (0.19 g, 0.5 mmol) in CH3CN and stirred overnight. NH4CI sat. was added and EtOAc extractions performed. The combined organic extracts were washed with Na2S2C<3, H2O and NaCl sat, followed by Na2S04 treatment and solvent concentration. The residue was purified by silica gel column chromatography (70% EtOAc/hexanes), to afford white crystals (0.50 g, 90%) of desired product (151). lH NMR (CDC13, 400 MHz) Si.34 (2H, s), 4.11 (1H, m), 3.97 (1H, m), 3.76 (3H, s), 3.50 (1H, m), 3.00 (3H, m), 2.48 (1H, m), 2.29 (1H, m), 1.76 (1H, m); 13C NMR (CDCI3, 100 MHz) £174.3 (C), 152.6 (C), 137.2 (C), 132.9 (2CH), 118.1 (2C), 65.0 (CH), 60.3 (CH3), 43.4 (CH), 35.2 (CH2), 34.4 (CH2), 29.4 (CH2), 28.0 (CH2). HRESIMS calcd for Ci4H16N402Na79Br2l ([M+Na]+): 556.8685; found 556.8683. Preparation of 5-bromo-6-(3.5-dibromo-4-methoxvbenzyl)-7-oxo-azepan-4-vl-cyanarnide (156) and4,5-dibromo-3-(3,5-dibromo-4-methoxybenzyl)-azepan-2-one (147) NH O ^ ^ - - B r s-ss/ lp°- NBS NH2CN H N C ' V Br NH l"^° Vs Br + —Br ' O — B r — / Br NH / X " > /T<V~Br V ^ / T 0— Br 136 156 147 202 Chapter 4. Progress towards the Synthesis of Ceratamines To a solution of 3-(3,5-dibromo-4-methoxybenzyl)-l,3,6,7-tetrahydroazepin-2-one (136) (1.0 g, 2.6 mmol) in CH2C12 (20 mL) at 0 °C, was added NBS (0.50 g, 2.8 mmol) followed by cyanamide (0.43 g, 10.3 mmol), until TLC analysis (70% EtOAc/hexanes) showed the absence of starting material (72 h). The crude was concentrated in vacuo and purified by silica gel column chromatography (100% EtOAc) to afford first 4,5-dibromo-3-(3,5-dibromo-4- methoxybenzyl)-azepan-2-one (147) (0.69 g) as white crystals, and then desired 5-bromo-6-(3,5- dibromo-4-methoxybenzyl)-7-oxo-azepan-4-yl-cyanamide (157) (0.62 g, 47%) as a colorless powder. Compound (147) was also prepared by Br2 (0.2 mL, 3.9 mmol) addition at -78 °C, to a solution of 3-(3,5-dibromo-4-methoxybenzyl)-l,3,6,7-tetrahydro-azepin-2-one (136) (0.31 g, 0.80 mmol) in CH2CI2 (5 mL). After stirring for 3 h, the crude was added to Na2S2C>3 sat. and extracted with CH2C12, to be then dried (Na2SC>4) and concentrated in vacuo. This procedure provided (147) as a foamy white solid (0.44 g, 99%). 5-Bromo-6-(3,5-dibromo-4-methoxybenzyl)-7-oxo-azepan-4-yl-cyanamide (156): for a summary of lH and 13C NMR assignments, see Table 4.19. lH NMR (CD3OD, 400 MHz) Si.48 (2H, s), 4.00 (IH, m), 3.81 (IH, m), 3.80 (3H, s), 3.53 (IH, m), 3.52 (IH, m), 3.16 (IH, dd, J = 6.1, 14.4 Hz), 3.01 (IH, m), 2.64 (IH, dd, J = 8.7, 14.2 Hz), 2.25 (IH, m), 1.85 (IH, m); 13C NMR (CD3OD, 100 MHz) £175.4 (C), 154.4 (C), 139.5 (C), 134.8 (2CH), 119.3 (2C), 116.0 (C), 61.2 (CH3), 60.0 (CH), 50.2 (CH), 45.0 (CH), 35.9 (CH2), 35.4 (CH2), 29.3 (CH2). HRESEVIS calcd for Ci5Hi7N30279Br3 ([M+H]+): 507.8871; found 507.8874. 4,5-Dibromo-3-(3,5-dibromo-4-methoxybenzyl)-azepan-2-one (147): for a summary of ]H and 13C NMR assignments, see Table 4.20. !H NMR (CDC13, 400 MHz) £7.39 (2H, s), 6.21 (IH, s, 203 Chapter 4. Progress towards the Synthesis of Ceratamines broad), 4.77 (IH, m), 4.15 (IH, m), 3.85 (3H, s), 3.84 (IH, m), 3.72 (IH, m), 3.21 (IH, dd, 7 = 5.4, 14.6 Hz), 3.13 (IH, m), 2.69 (IH, dd, J = 9.4, 14.6 Hz), 2.61 (IH, m), 2.08 (IH, m); 13C NMR (CDC13, 75 MHz) £173.0 (C), 152.9 (C), 136.9 (C), 133.4 (2CH), 118.1 (2C), 60.6 (CH3), 55.6 (CH), 50.0 (CH), 43.9 (CH), 36.6 (CH2), 34.0 (CH2), 32.1 (CH2). HRESIMS calcd for Ci4Hi5N02Na79Br381Br ([M+Na]+): 569.7713; found 569.7709. Table 4.19. NMR data for 5-bromo-6-(3,5-dibromo-4-methoxy-benzyl)-7-oxo-azepan-4-yl- cyanamide (156) (recorded in CD3OD). Carbon "C *H HMBCb No 8(ppm)a 8(ppm)(mult,/(Hz))b'c (H-»C) IN 2 35.9 3.01 (m), 3.52 (m) C4 3 29.3 1.85 (m), 2.25 (m) C4, C5 4 60.0 3.81 (m) C2,C3,C5,C6,C14 5 50.2 4.00 (m) C3, C4, C6, C7, C8 6 45.0 3.53 (m) C4,C7,C8,C9 7 . . . . 175.4 35.4 139.5 134.8 119.3 154.4 61.2 116.0 .  ( ), .  ( ) . .  ( ) .  ( ) 3.16 (dd, 7 = 6.1,14.4 Hz), 2.64 (dd, J =8.7, 14.2 Hz) 7.48 (s) 3.80 (s) C5, C6, C7, C9, C10 9 10 .  .   C8,C11,C12 11 12 13 .  .  ( ) C12 14 "Recorded at 100 MHz.b Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. 204 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.20. NMR data for 4,5-dibromo-3-(3,5-dibromo-4-methoxy-benzyl)-azepan-2-one (147) (recorded in CDC13). Carbon 13, *H HMBC" No Nl 2 3 4 5 6 7 8 9 10 11 12 13 5 (ppm)a 173.0 43.9 50.0 55.6 32.1 36.6 34.0 136.9 133.4 118.1 152.9 60.6 5(ppm)(mult,/(Hz))b,c 6.21 (s, broad) 3.72 (m) 4.15 (m) 4.77 (m) 2.08 (m), 2.61 (m) 3.13 (m), 3.84 (m) 2.69 (dd, .7 = 9.4,14.6 Hz), 3.21 (dd, J =5.4,14.6 Hz) 7.39 (s) 3.85 (s) (H-*C) C2, C4, C5, C8 C2, C6, C5, C8 C7 C2.C5 C2,C3,C4,C9,C10 C8,C11,C12 C12 'Recorded at 75 MHz.b Recorded at 400 MHz.c According to HMQC recorded at 400 MHz. Synthesis of 19-demethvl-l,4,5,8,9,10-hexahydroceratamine B (157) NC H 2 N ^ ; 5-Bromo-6-(3,5-dibromo-4-methoxybenzyl)-7-oxo-azepan-4-yl-cyanamide (156) (0.23 g, 0.45 mmol) was dissolved in NH3/MeOH (2 mL, 4 mmol, 2 M), and heated to 80 °C for 5 h in a sealed tube. Once at room temperature, the reaction crude was concentrated in vacuo, to give desired 19-demethyl-1,4,5,8,9,10-hexahydroceratamine B (157) (0.19 g, 96%) as a pale yellow powder. For a summary of *H and 13C NMR assignments, see Table 4.9. !H NMR (DMSO, 600 205 Chapter 4. Progress towards the Synthesis of Ceratamines MHz) £8.20 (2H, s, broad), 7.90 (1H, s, broad), 7.66 (1H, d, J = 7.4 Hz), 7.53 (2H, s), 3.82 (1H, m), 3.80 (1H, m), 3.77 (3H, s), 3.42 (1H, m), 3.12 (1H, m), 3.10 (1H, m), 2.92 (2H, m), 2.15 (1H, m), 1.57 (1H, m); 13C NMR (DMSO, 150 MHz) 8172.9 (C), 160.4 (C), 151.6 (C), 139.1 (C), 133.1 (2CH), 117.0 (2C), 60.7 (CH), 60.4 (CH3), 58.5 (CH), 48.5 (CH), 38.3 (CH2), 31.7 (CH2), 29.0 (CH2). HRESIMS calcd for Ci5Hi9N40279Br2 ([M+H]+): 444.9875; found 444.9871. Protection 19-demethvl-l,4,5,8,9,10-hexahydroceratamine B (157) H N 158 157 159 To a solution of 19-demethyl-l,4,5,8,9,10-hexahydroceratamine B (157) (0.14 g, 0.32 mmol) in MeOH (10 mL), was added triethylamine (0.13 mL, 0.96 mmol) and di-tert- butyldicarbonate (0.14 g, 0.64 mmol). After stirring for 30 min. at 50 °C and then at room temperature, H2O partition followed by EtOAc extractions, Na2SC>4 treatment, and solvent concentration, provided a colorless residue. This crude was purified by gradient silica gel column chromatography (30% EtOAc/hexanes to 100% EtOAc), to afford (158) (0.043 g, 24%) as a colorless solid. In another reaction, the same starting material (157) (0.2 g, 0.45 mmol) was dissolved in pyridine (15 mL) and stirred in the presence of 2,2-dimethylpropanoic acid chloride (5 mL, 40 mmol), first at 80 °C for 2 h, and then at room temperature overnight. Silica gel column chromatography (10%MeOH/EtOAc) provided diprotected product (159) (0.033 g, 12%) as a colorless solid. 206 Chapter 4. Progress towards the Synthesis of Ceratamines 18-Boc-l,4,5,8,9,10-hexahydroceratamine B (158): for a summary of !H and 13C NMR assignments, see Table 4.21. *H NMR (CDC13, 400 MHz) £7.44 (2H, s), 7.04 (1H, m, broad), 3.88 (1H, m), 3.81 (3H, s), 3.73 (1H, m), 3.29 (1H, dd, / = 2.8, 13.7 Hz), 3.04 (1H, dd, / = 6.1, 13.7 Hz), 2.94 (1H, m), 2.83 (1H, m), 2.80 (1H, m), 2.75 (1H, m), 2.61 (1H, m), 1.48 (9H, s); 13C NMR (CDC13, 100 MHz) 8174.6 (C), 156.7 (C), 152.8 (C), 151.7 (C), 137.9 (C), 133.8 (2CH), 117.9 (2C), 85.0 (C), 61.7 (CH), 60.6 (CH3), 58.4 (CH), 53.2 (CH), 36.6 (CH2), 34.2 (CH2), 31.5 (CH2), 28.0 (3CH3). HRESIMS calcd for C2oH27N40479Br2 ([M+H]+): 545.0399; found 545.0397. 7,17-Dipivaloyl-l,4,5,8,9,10-hexahydroceratamine B amide (159): for a summary of *H and 13C NMR assignments, see Table 4.22. !H NMR (CDCI3, 600 MHz) £7.43 (2H, s), 3.84 (3H, s), 3.74 (1H, m), 3.56 (1H, m), 3.52 (2H, m), 3.15 (1H, m), 3.00 (2H, m), 2.28 (1H, m), 1.75 (1H, m), 1.19 (9H, s), 1.18 (9H, s); 13C NMR (CDC13, 150 MHz) £190.2 (C), 189.0 (C), 173.9 (C), 162.9 (C), 153.0 (C), 137.4 (C), 133.4 (2CH), 118.2 (2C), 63.1 (CH), 60.6 (CH3), 60.5 (CH), 51.6 (CH), 44.2 (C), 42.5 (CH2), 40.8 (C), 33.3 (CH2), 30.9 (CH2), 27.9 (3CH3), 27.4 (3CH3). HRESIMS calcd for C25H35N40479Br2 ([M+H]+): 613.1025; found 613.1024. 207 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.21. NMR data for 17-Boc-l,4,5,8,9,10-hexahydroceratamine B (158) (recorded in CDC13). Carbon »H HMBC" No Nl 2 N3 4 5 6 N7 8 9 10 11 12 13 14 15 16 N17 18 19 20 Recorded at 100 Hz. 8 (ppm)a 156.7 61.7 53.2 174.6 36.6 31.5 58.4 34.2 137.9 133.8 117.9 152.8 60.6 151.7 85.0 28.0 5(ppm)(muIt,/(Hz))b>c 3.88 (m) 2.83 (m) 7.04 (m, broad) 2.80 (m), 2.94 (m) 2.61 (m), 2.75 (m) 3.73 (m) 3.04 (dd, 7 = 6.1, 13.7 Hz), 3.29 (dd, 7=2.8, 13.7 Hz) 7.44 (s) 3.81 (s) 1.48 (s) (H->C) C5.C10 Cll C4,C9 C5, C6, C12, C13 C11,C14,C15 C15 C19 " Recorded at 400 MHz.' According to HMQC recorded at 400 MHz. 208 Chapter 4. Progress towards the Synthesis of Ceratamines Table 4.22. NMR data for 7,17-dipivaloyl-l,4,5,8,9,10-hexahydroceratamine B amide (159) (recorded in CDC13). Carbon No Nl 2 N3 4 5 6 N7 8 9 10 11 12 13 14 15 16 N17 18 19 20 21 22 23 1 3 C 8 (ppm)a 162.9 60.5 51.6 173.9 42.5 30.9 63.1 33.3 137.4 133.4 118.2 153.0 60.6 189.0 44.2 27.4 190.2 40.8 27.9 »H 5(ppm)(mult,/(Hz))bc 3.56 (m) 3.15 (m) 3.52 (m) 1.75 (m), 2.28 (m) 3.74 (m) 3.00 (m) 7.43 (s) 3.84 (s) 1.18 (s) 1.19 (s) HMBCb (H-»C) C6.C9 C4, C6, C12 C6, C10, C21 C8, C10 C4.C5 C4,C6,C12,C13 C12.C14, C15 C15 C18.C19 C21, C22 "Recorded at 150 Hz.b Recorded at 600 MHz.' According to HMQC recorded at 600 MHz. 209 Chapter 4. Progress towards the Synthesis of Ceratamines 5.75 i.5(> 3.00 4.75 , JJJLL Chemical Shift (ppm) Chemical Shift (ppm) Figure 4.16. *H and 13C-NMR spectra of 2-(3,5-dibromo-4-methoxybenzyl)-but-3-enoic acid but-3-enylamide (135) (recorded in CDCI3 at 400 and 100 MHz respectively). 210 Chapter 4. Progress towards the Synthesis of Ceratamines S.5 5.4 5.3 5.2 5.1 _A_J JLJ Chemical Shift (ppm) Chemical Shift (ppm) Figure 4.17. JH and 13C-NMR spectra of 3-(3,5-dibromo-4-methoxy-benzyl)-1,3,6,7- tetrahydroazepin-2-one (136) (recorded in CDCI3 at 300 and 75 MHz respectively). 211 Chapter 4. Progress towards the Synthesis of Ceratamines Jli/J'ULJL Chemical Shift (ppm) 4 wy uiiiifLiM'*!" Chemical Shift (ppm) Figure 4.18. 'H and 13C-NMR spectra of 5-bromo-6-(3,5-dibromo-4-methoxybenzyl)-7-oxo- azepan-4-yl-cyanamide (156) (recorded in CD3OD at 400 and 100 MHz respectively). 212 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5. Cannabinoid activity of the Marine Sterol Haplosamate A: Direct Observation of Binding to Human Receptors by Saturation Transfer Double- Difference (STDD) NMR Spectroscopy 5.1. Brief history Hemp, Cannabis sativa L., and its derivatives (such as marijuana, hashish and hash oil) have a long history of use for therapeutic, ritual and recreational purposes. The therapeutic potential of ginseng, ephedra and cannabis was documented in Chinese medicine as early as 3000 B.C. when, during the rule of Emperor Shen Nung, the Chinese used marijuana for the treatment of malaria, constipation, rheumatic pains, absentmindedness, and gynecological disorders.396"399 The use of cannabis would spread west from early China to ancient India, where the anxiety relieving effect of bhang (the Indian term for marijuana ingested as food) was recorded more than 3000 years ago.399 Other civilizations used cannabis to treat pain, nausea, fever, infections, and to stimulate appetite.400 The use of cannabis as a psychoactive substance reached Europe and the Americas through the Arab world at the end of the 18th century, where anecdotal information about its health benefits was finally put to scientific scrutiny. Early in the 19l century, the therapeutic value of cannabis was recognized by British physicians. Sir John Russell Reynolds, one of Queen Victoria's physicians, was a proponent of the use of cannabis in medicine. In 1839 W. B. O'Shaughnessy at the Medical College of Calcutta observed its use in the indigenous treatment of various disorders and found that tincture of hemp was an effective analgesic, anticonvulsant, and muscle relaxant.400'401 Over the next several decades, many papers on cannabis appeared in European medical literature.398 Marijuana extracts were widely used for medicinal purposes in the first decades of 213 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A the 20th century until 1937, when concern about the dangers of abuse led to the banning of marijuana for further medicinal use in United States (Marijuana Tax Act).399 Canada had made it illegal more than a decade before, in 1923, under the Opium and Drug Narcotic Act.402 A major breakthrough in the esoteric field of cannabis research was the isolation and elucidation in 1964, by Gaoni and Mechoulam, of the main psychoactive ingredient of marijuana, A9-tetrahydrocannabinol (1), and the following demonstration that bioactivity resides in its (-)-enantiomer. ' This discovery stimulated the generation of a structurally diverse range of synthetic compounds in the 1970's (classical and non-classical cannabinoids, aminoalkylindoles). More potent and selective new cannabinoids constituted the key for a second milestone in cannabis research, the discovery in 1988 and 1991 of specific cannabinoid receptors. ' Soon, reports on the first endogenous cannabinoids would follow. Over the last few years there has been an active debate regarding the medicinal aspects of cannabis. Currently cannabis products are classified as Schedule I drugs under the U. S. Drug Enforcement Administration (DEA) Controlled Substances Act, which means that the drug is only available for research purposes.406 In 1999, the Court of Appeal for Ontario ruled it unconstitutional to enforce the rule of law with respect to cannabis. Since 2001, the Marijuana Medical Access Regulations (MMAR) have made cannabis possession legal for authorized patients in Canada. Herbal cannabis, cultivated by Prairie Plant Systems Inc., licensee to Health Canada, is distributed to patients for CAN $5/g.408 214 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.2. Marijuana: chemistry and psychoactive effects The popular term marijuana refers to the dried material of the Cannabis sativa L. plant {Cannabaceae), a hemp which grows throughout temperate and tropical climates in almost any soil condition400 Dried flowering tops and leaves are commonly smoked in hand-rolled cigarettes. All the different cannabinoids are concentrated in a viscous resin produced in glandular structures known as trichomes (Figure 5.1).406 The content ranges from 5% to 16% m/m, depending on latitude, weather, and soil conditions.400 Figure 5.1. a) Mature C sativa crop ready for harvesting; b) trichomes.406 Cannabis constituents represent almost all kinds of chemical compounds, including mono- and sesquiterpenes, sugars, hydrocarbons, steroids, flavonoids, nitrogenous compounds 215 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A and amino acids (Table 5.1),406 with A9-tetrahydrocannabinol (1) and similar terpenophenolic components as the most psychoactive. 403,405 Table 5.1. Constituents in the resin of Cannabis sativa reported until 2005 406 General Class Cannabinoids Nitrogenous compounds Amino acids Proteins, enzymes and glycoproteins Sugars and related compounds Hydrocarbons Alcohols Aldehydes Ketones Carboxylic acids Fatty acids Esters and lactones Steroids Terpenes Non-cannabinoid phenols Flavonoids Vitamins Pigments Elements Total Amount 70 27 18 11 34 50 7 12 13 20 23 13 11 120 25 23 1 2 9 489 Table 5.2 summarizes the most typical effects of cannabis in humans. Many of them are described as biphasic,409 showing increased activity with acute or smaller doses and decreased response with larger doses or chronic use. They also vary greatly among individuals and may be more severe in ill and elderly patients. 216 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.2. Pharmacological effects of smoked marijuana in humans. 409 Body System Nervous (CNS) Cardiovascular Respiratory Eyes Immune Reproductive Perception Sedative Effect Description Psychological Euphoria (high), dysphoria, anxiety, precipitation or aggravation of psychosis, depersonalization. Heightened sensory perception, distortion of space and time, sense, hallucinations, misperceptions. Generalized CNS depression, drowsiness, somnolence; additive with other CNS depressants. Fragmentation of thoughts, mental clouding, memory impairment, global impairment of performance especially in complex demanding tasks. Increased motor activity followed by inertia and in coordination, ataxia, dysarthria, tremulousness, weakness, muscle twitching. Currently available oral cannabinoids are similar in potency to codeine. With acute doses; the effect is reversed with larger doses or chronic use (tolerance). To most behavioral and somatic effects, including the "high". Produced experimentally following prolonged intoxication: symptoms include disturbed sleep, decreased appetite, restlessness, irritability and sweating. Tachycardia with acute dosage, bradycardia with chronic use. Vasodilation, conjunctival redness, postural hypotension. Increased output and myocardial oxygen demand. Increased with acute dose, decreased with chronic use. Small doses stimulate; larger doses depress. Coughing, but tolerance develops. From chronic smoking. Decreased intraocular pressure. Chronic use: impaired bactericidal activity of macrophages in lung and spleen. Males Anti-androgenic, decreased sperm count and sperm motility (chronic use, but tolerance may develop). Females Suppression of ovulation, complex effects on prolactin secretion; chronic use: increased obstetric risk. Cognition, psychomotor performance Motor function Analgesic Anti-emetic, increased appetite Tolerance Dependence, abstinence syndrome Heart rate Peripheral circulation Cardiac output Cerebral blood flow Ventilation Bronchodilation Airways obstruction Currently, there are three general types of cannabinoids: herbal cannabinoids or phytocannabinoids, which occur only in the cannabis plant; endogenous cannabinoids, produced in humans and other animals; and synthetic cannabinoids. 217 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.3. Phytocannabinoids Introduced in the late 1990's, the term phytocannabinoids is used to describe the natural and most psychoactive terpenophenolic components of Cannabis sativa.397'406 There are around 70 known phytocannabinoids (Table 5.1), which are in turn grouped into 11 different classes according to common carbon backbones.406 Figure 5.2 shows the most representative member for each type. HO Cannabigerol (CBG) type Cannabichromene (CBC) type (-)-Cannabidiol (CBD) type (-)-A -mzn.s-Tetrahydrocannabinol (A9-THC) type (-)-A -Jra/w-Tetrahydrocannabinol (A8-THC) type Cannabicyclol (CBL) type HO 8 Cannabielsoin (CBE) type HO. 11 trans-Cannabitiiol (CBT) type Cannabinol (CBN) type 12 Miscellaneous type HO 10 Cannabinodiol (CBND) type 13 Miscellaneous type Figure 5.2. Classification of the typical cannabinoids. 406 218 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Cannabigerol (2) was the first compound isolated from the C. sativa resin. Although psychologically inactive when compared with A9-THC compounds, CBG-type cannabinoids exhibit antibacterial activity.406 The absolute configuration for (-)-cannabidiol (4) was assigned as (-)-trans-(lR,6R) by synthesis, being the first compound synthetically prepared.410 There is no clear evidence for the natural origin of cannabielsoin (8)-type compounds, since they can be considered as oxidation products of naturally occurring CBD metabolites. Likewise, the fully aromatized CBN and CBND compounds are also thought to be artifacts, since their concentration increases during storage and exposure to light and air, with a subsequent decrease in the amount of A9-THC and CBD cannabinoids respectively.406 In fresh plants, the amount of CBE, CBN and CBND is usually minimal.400 5.4. Cannabinoid receptors, endogenous cannabinoids and the endocannabinoid system In the 1980s, scientific research on cannabis became significantly more interdisciplinary, attracting the attention of more chemists, biochemists and pharmacologists. Before then, it was often speculated that cannabinoids produced their physiological and behavioral effects via nonspecific interaction with cell membranes, instead of interacting with specific membrane- bound receptors.397 The synthesis of less lipophilic and more potent cannabinoid ligands than A -THC (1) improved the observations obtained from traditional receptor binding techniques, making "iQf. "ion possible the discovery of authentic cannabinoid binding sites. ' In 1988 the Pfizer compound CP-55,940 (14), a more potent and polar agonist, was used as the first probe of cannabinoid receptors.397"399 Autoradiography and competitive radioligand binding assays using [3H]CP- 55,940, confirmed the presence of cannabinoid binding sites in the brain. ' 219 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A OH HO' 14 In 1990, the CBl receptor was cloned from rats, and later from humans and mice.411 Its distribution has been well characterized in rat and human brain. ' Three years later, a second receptor subtype CB2 was found by sequence homology. Exhibiting a low overall homology with the CBl receptor (44%, with 68% in the helical regions), CB2 was cloned from humans, mice and rats, and found to be restricted to the immune system, including macrophages from the spleen.397'399'411 5.4.1. Cannabinoid human receptors CBl receptors exhibit a widespread distribution in the brain that correlates well with the known effects of cannabinoids on memory, perception, and the control of movement.412 Specifically, they are found in the basal ganglia and the cerebellum (involved in motor activity), in the limbic system, including the cortex and hippocampus (related with memory and cognition), amygdala (emotion), thalamus (sensory perception), hypothalamus, pons and medulla •7QT 'XQQ (regulation of the autonomic and endocrine functions). ' The CBl receptor is also expressed in peripheral nerve terminals and various extraneural sites such as the reproductive systems (both male and female), eyes, vascular endothelium and spleen. 9 In addition, CBl receptor mRNA has been described in the adrenal gland, heart, lung, prostate, bone marrow, thymus, and tonsils.398'411 In contrast, CBl receptors are essentially absent in the medulla oblongata, the part of the brain stem responsible for respiratory and cardiovascular functions. This might explain the general lack of serious acute effects, including respiratory or cardiovascular failure, associated 220 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A with cannabis abuse when compared with other drugs. Both central and many of the peripheral effects of cannabinoids depend on the activation of CB1 receptors. In humans, CB1 is responsible for the "high" and anticonvulsive effects produced by smoked marijuana. CB2 receptors are almost exclusively found in the immune system, with the greatest density in the spleen.398 They are expressed by cells, including B and T lymphocytes, macrophages, and by tissues, like tonsils and lymph nodes besides the spleen. CB2 receptors appear to be responsible for the anti-inflammatory and possibly other therapeutic effects of cannabis. Thus, compared to CB1, CB2 represents a more attractive pharmacological target in developing cannabinoid-based therapeutic agents.399 Both CB1 and CB2 have amino acid sequences characteristic of G-protein-coupled receptors (GPCR's), in which a single polypeptide structure spans the plasma membrane seven times in a serpentine-like topology (Figure 5.3).397'411 They are common in animals, and have been found in mammals, birds, fish, and reptiles. The term cannabinoid has been extended to any molecule that binds to one of the cannabinoid receptors. Figure 5.3. Serpentine-like topology of the cannabinoid receptors CB1 and CB2. CB1 is 126 amino acids longer and thus, possesses more units outside the cellular membrane. 221 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Recent studies have demonstrated a number of biological activities that are not considered to be mediated via CB1 or CB2 receptors. These include endothelium-dependent vasodilator effects of certain cannabinoids, and presynaptic inhibition of glutamatergic neurotransmission in the hippocampus.411'413'414 However, to date the NC-IUPHAR (International Union of Pharmacology Committee on Receptor Nomenclature and Drug Classification) has not designated additional cannabinoid receptors. 5.4.2. Endocannabinoids The discovery of specific cannabinoid receptors launched the quest for their corresponding endogenous ligands. Assuming that the endogenous agonist would have hydrophobic properties similar to those of current cannabinoid agonists, Devane, Mechoulam, and coworkers415 isolated in 1992 the lipid derivative anandamide (15, from the Sanskrit ananda meaning "bliss and tranquility"),397'411 from organic solvent extracts of porcine brain. It is about as potent as A9-THC (1), binds to both CB1 and CB2 receptors, and is found in nearly all tissues -an*7 in a wide range of animals. Besides anandamide (15), numerous endogenous polyunsaturated compounds capable of binding one or both cannabinoid receptors have been reported (Figure 5.4).399'416 Oleamide (16) shares with 15 some pharmacological properties, including sleep induction, anticonvulsant effects and modulation of appetite.397 Arachidonoyldopamine (19) presented high selectivity as •^Q-y "7QQ a CB1 agonist and inhibited the proliferation of human breast cancer cells. ' 2- Arachidonoylglycerol (20) is more abundant and potent than 15,399'417 but as a monoglyceride, it takes part in several pathways of lipid metabolism which minimize its availability in endocannabinoid signaling. This metabolite inhibited the proliferation of breast and prostate cancer cells, and was found to induce hypotension, contractile action on colon muscles, and neuroprotection after brain injury.397 222 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Endocannabinoid Amides o OH Endocannabinoid Esters o -OH -OH 20 o 21 , N H 2 Endocannabinoid Ether OH OH Figure 5.4. Representative endocannabinoids: anandamide (15), oleamide (16), homo-y- linolenoylethanolamide (17), docosatetraenoylethanolamide (18), arachidonoyldopamine (19), 2- arachidonoylglycerol (20), O-arachidonoylethanolamine (21), and 2-arachidonylglyceryl ether (noladin ether 22). 397 5.4.3. The endocannabinoid system Phytocannabinoids, as well as their synthetic analogues, act in the organism by activating -3QT "3QQ CB1 and/or CB2, normally engaged by the endogenous cannabinoids. ' The cannabinoid receptors, endocannabinoid ligands, and specific processes of synthesis, uptake and degradation, constitute the endogenous cannabinoid system (Figure 5.5), essential in brain modulation and therefore, almost every major life function in the human body.396'399'418 After neurotransmitters bind their receptors (iR, mR) in a postsynaptic neuron, they synthesize membrane-bound endocannabinoid precursors and cleave them to release active 223 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A ,2+ endocannabinoids following an increase of cytosolic free Ca concentrations. Endocannabinoids subsequently act as retrograde messengers, traveling backwards against the synaptic flow to bind presynaptic CB1 cannabinoid receptors, which downregulate voltage-sensitive Ca2+ channels and activate K+ channels.418 This blunts membrane depolarization and exocytosis, which in turn inhibits the release of neurotransmitters such as glutamate, dopamine and y- aminobutyric acid (GABA), involved in learning, movement and memory processes, respectively. ' The endocannabinoid neuromodulatory signaling is terminated by a membrane-transport system419 (T) and a family of intracellular degradative enzymes. The most studied member of this family is a fatty acid amide hydrolase (FAAH), in charge of degrading anandamide (15) to ethanolamine and arachidonic acid.3 '397418'420 Newotransiuitter* © © \ © © © © [&*) L° ° @ © C&)\ © Q ° ° • ^ ( © © © A W © © o ^ ' © © T R , : < 3 — * i c 3 i * % ' oo o o ^ 7 / Degradation products Figure 5.5. Endogenous cannabinoid system. Presynaptic and postsynaptic designate the sending and receiving sides of a synapse, respectively. iR: ionotropic receptors, channel-like receptors opened by agonist binding, and through which ions (Na+, K+ and Ca2+) can pass. mR: metabotropic receptors: seven-transmembrane heptahelical receptors coupled to G proteins.397-399'4lf 224 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A The endogenous cannabinoid system might also exert modulatory functions outside the brain, both in the peripheral nervous system and in extraneural sites, controlling processes such as peripheral pain, vascular tone, intraocular pressure and immune function.39 ' Although they share an intercellular signaling role with traditional water-soluble neurotransmitters, endocannabinoids are hydrophobic molecules with restricted mobility in the aqueous media an*7 surrounding cells and, therefore, act locally on nearby cells. 5.5. Synthetic and patented cannabinoids Historically, synthetic cannabinoids were often based on the structure of phytocannabinoids and a large number of analogues have been produced and tested.421"426 The first total synthesis of A9-THC (1) was reported in 1965 by Gaoni and Mecholaun,427 the same researchers who had elucidated and reported its structure just one year before. As mentioned earlier, more potent and selective synthetic cannabinoids were the key for the discovery of both cannabinoid receptors currently known. Newer synthetic compounds are no longer related to natural cannabinoids or based on the structure of the endogenous cannabinoids. The massive and growing spectrum of cannabinoid- active synthetic compounds is organized according to their efficacy and affinity for the CBl and CB2 receptors, as follows: 397>399'411 a) High efficacy compounds without remarkable selectivity between CB receptors: including 2-arachidonoylglycerol (20), CP-55,940 (14), (R)-WJN 55,212-2 (23), HU-210 (24) and BAY 38-7271 (25). 225 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A o N- HO HO \....< 1 o ^ ^ ^ o , C F 3 23 24 25 b) Partial agonists without remarkable selectivity between CB receptors: for example A - tetrahydrocannabinol (1), anandamide (15), (R)-methanandamide (26) and BAY 59-3074 (27). F3C , O H NC in , C F 3 26 27 c) Selective CB1 agonists: like ACEA (28) and 0-1812 (29). ,ci OH 28 29 d) Selective CB2 agonists: HU308 (30), AM1241 (31), GW405833 (32), JWH015 (33) and JWH133 (34), among others. 426 226 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A s428 e) Selective CB1 antagonists/inverse agonists: including SR141716A (35, rimonabant) , AM251 (36), AM281 (37) and 0-2654 (38). ci X=CH2, 36 X=0, 37 H O 38 f) Selective CB2 antagonists/inverse agonists: with JTE907 (39), SR144528 (40) and AM630 (41) as main representatives. 429 x H^Q:0> 39 Table 5.3 presents commercial cannabinoid-based therapeutic agents currently available or in development.412'428'430 Marinol® was the first cannabinoid with approval for marketing in "2Q£ "3QO the United States. ' Manufactured as a capsule containing dronabinol (42) in sesame oil, it is taken orally. It was approved by the FDA in 1985 for the treatment of nausea and vomiting associated with cancer chemotherapy, and anorexia associated with weight loss in AIDS patients.396'400 Annual sales of Marinol® are estimated at US $20 million.412'430 Cesamet® was also endorsed in 1985 under the same therapeutic applications as Marinol®, but only became marketed in 2006.400'412'430 Sativex® was accepted by Health Canada for prescription use in 2005.398'412 It is an oromucosal (mouth) spray developed by the British GW Pharmaceuticals for 227 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A multiple sclerosis patients, to alleviate neuropathic pain and spasticity.399'400 It is also the only cannabinoid preparation currently available produced from botanical material, rather than 412 organic synthesis. Rimonabant (35) is the first CB1 antagonist to be approved worldwide as an anorectic anti-obesity drug. 412 Table 5.3. Cannabinoid-based therapeutic agents, approved and in development. 398,412,428,430 HO^ 42 Name Marinol® Cesamet® Acomplia® Sativex® None None None Marijuana Cannabis sativa Company Unimed Valeant Sanofi-Aventis GW Pharmos Corp. Atlantic Unimed Prairie Plant Systems Inc., Health Canada HortaPharm, GW Donald Abrams, MD Ethan Russo, MD 43 Active Ingredient Dronabinol (42) Nabilone (43) Rimonabant (35) A9-THC(1), cannabidiol (4) HU-211 (44) CT-3 (45) A9-THC (1) Phyto- cannabinoids Phyto- cannabinoids 44 Use Anti-emetic, appetite stimulant1 Anti-emetic1 Anti-obesity Neuropathic pain (MS) Neuroprotecti on2 Anti-inflammatory, analgesic Neuropathic pain (MS) Numerous Development of new strains Appetite stimulant Migraine 45 Approved in: USA, CAN USA, CAN, UK Worldwide CAN, Spain, CT III USA CT II Israel Preclinical USA CT I USA CAN Clinical UK CTIUSA Chemotherapy relief. Neurotrauma, stroke, Parkinson's, Alzheimer's. MS: multiple sclerosis. CT: clinical trials. Some cannabinoids are being developed for new therapeutic applications like neuroprotection, since they are able to rescue neurons from cell death associated with trauma, ischemia, and neurological diseases. The synthetic HU-211 (dexanabinol, 44) is an antioxidant which protects neurons from neurotoxicity induced by excess glutamate concentrations. This compound is employed in the treatment of severe head trauma. 430 228 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.6. Cannabinoids and cancer The palliative effects of cannabinoids in cancer patients have been well known since the 1970's, and include appetite stimulation, inhibition of nausea and emesis associated with chemo- -3QO A] Q or radiotherapy, pain relief, mood elevation, and relief from insomnia. ' As shown in Table 5.3, both dronabinol (42) and nabilone (43) are being currently prescribed to cancer patients. In addition, numerous studies have suggested that cannabinoids might directly inhibit cancer growth through several mechanisms, including induction of apoptosis in tumor cells, antiproliferative action, and an antimetastatic effect by inhibiting angiogenesis and tumor cell migration. ' 18 Table 5.4 shows the cannabinoid-induced inhibition of several tumors treated with A9-tetrahydrocannabinol (1), cannabidiol (4), WIN-55,212-2 (23), HU-210 (24), anandamide (15) and 2-arachidonoylglycerol (20), members of the three basic cannabinoid types.418 Table 5.4. Tumors exhibiting cannabinoid-induced inhibition. 399,418 Type Lung carcinoma Glioma Thyroid epithelioma Lymphoma, leukemia Skin carcinoma Uterus carcinoma Breast carcinoma Prostate carcinoma Neuroblastoma Colon carcinoma Lymphoid tumors N.D.: not determined. VR1: System In vivo (mouse), in vitro In vivo (mouse, rat), in vitro In vivo (mouse), in vitro In vivo (mouse), in vitro In vivo (mouse), in vitro In vitro In vitro In vitro In vitro In vitro In vitro type 1 vanilloid receptor. Effect Decreased tumor size, growth inhibition Decreased tumor size, apoptosis Decreased tumor size, cycle arrest Decreased tumor size, apoptosis Decreased tumor size, apoptosis Cell-growth inhibition Cell-cycle arrest Apoptosis Apoptosis Apoptosis Apoptosis cell- cell Receptor N.D. CB1/CB2 CB1 CB2 CB1/CB2 N.D. CB1 CB1 VR1 CB1/CB2/VR1 CB1/CB2/VR1 229 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.7. Biological evaluation of cannabinoid compounds Before the identification of cannabinoid receptors, in vivo screening assays for cannabinoid activity were largely behavioral in nature.411 Among them, the cannabinoid tetrad test397'431 is still used to provide valuable quantitative information on the central cannabimimetic activity of test compounds. This assay comprises four different behavioral tests performed mostly in mice, evaluating hypothermia, reduction of locomotion, analgesia, and catalepsy.411'431 Modern in vitro bioassays include displacement assays of [H ]CP-55,940 by new unlabeled ligands, inhibition of cyclic AMP (adenosine-3',5'-monophosphate) production, inhibition of electrically evoked contractions of isolated smooth muscle preparations (high Off sensitivity), and [ S]guanosine-5'-0-(3-thiophosphate) binding assays, which exploits the coupling of CB1 and CB2 to G proteins.411 In contrast to biochemical binding assays, functional cell-based bioassays directly identify ligands that activate receptors, and not simply bind them. ' They detect and quantify the interactions between G-protein-coupled receptors (GPCR's) and ligands in a faster, nonradioactive, convenient and more efficient way, due to its adaptability to multiwell plate formats.432'433 Cell-based bioassays clearly depend on the nature of the host cells, in which GPCR's are expressed.433 To date, expression of CB1 and CB2 (two GPCR's, Section 5.4.1) has been reported in mammalian cells, Escherichia coli, yeast, and insect cells.434 Mammalian and amphibian cells possess a high amount of endogenous GPCR's which can lead to false positives, while yeast cells in turn can fail to express traffic receptors properly, and generate false negative results.432'435 Insect cells fall in the middle, combining many of the best features of mammalian and yeast cells as host systems for any GPCR assay.432'433 This last host was used by the At/) Grigliatti research group (Department of Zoology, UBC) to design a cell-based bioassay capable of detecting CB1 and CB2 ligands selectively, according to their binding affinity. 230 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Lepidopteran cell lines, specifically SF21 cells from pupal ovaries of the army fallworm Spodoptera frugiperda, were engineered to express high levels of both cannabinoid receptors, using the Ca+ -sensitive luminescent protein aequorin as a reporter. Obtained from the jellyfish Aequoria victoria, aequorin forms a bioluminescent complex when linked to the chromophore coelenterazine and O2. When the agonist binds CB1 or CB2, the signal is channeled to the insect phospholipase (PLCp\ Figure 5.6), which generates inositol triphosphate (IP3). In the next step, Ca+2 is released from intercellular stores via IP3 receptor-mediated response. Upon Ca+2 binding to aequorin, a conformational change takes place, resulting in the oxidation of bound coelenterazine to coelenteramide, with subsequent CO2 and blue light (k^a 470 nm) production. Since cells do not spontaneously produce light, background noise is extremely low and the emitted blue light is easily detectable with a luminometer. 432 Camiabkoids Insect cell membrane + Coelenterazine GTP CJOP 1P3 Ca~ stores f Aeq0oria 1 2 + f A e q U ° r i n 1 I Coelenteramide + CO-> 5 Lwninometer Figure 5.6. Events involved in the cell-based cannabinoid bioassay. 432 231 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.8. Isolation of haplosamate A As part of an ongoing program designed to isolate new biologically active secondary metabolites from marine sources, a series of methanolic extracts of marine invertebrates were evaluated for cannabinoid activity in the bioassay previously described. These extracts resulted inactive in other bioassays run by different collaborators at UBC, including antimitotic, angiogenesis inhibition, anti-invasion, antibacterial, IDO and PI3K inhibition assays. Screening afforded just a few hits, among them the methanolic extract of the marine sponge Dasychalina fragilis (Ridley & Dendy, 1886), collected from Keviang in Papua New Guinea. Previous reports on isolation of secondary metabolites from members of the Dasychalina genus include only adenosine (46) (D. cyathina), exhibiting cardiovascular activity.436 Similar nucleosides inspired the first marine-derived drugs, as mentioned in Chapter 1. NH2 OH OH 46 Bioassay-guided fractionation of the D. fragilis extract led to the isolation of the known phosphorylated sterol sulfate haplosamate A (47), as a white amorphous powder. This compound was initially isolated together with haplosamate B (48) by Qureshi and Faulkner437 in 1999, from two sponges collected in the Philippines, one a Xestospongia sp. and the other an unidentified haplosclerid sponge. Reported as sulfamate esters (Figure 5.7), haplosamates A and B were shown to inhibit HIV-1 integrase with ICso's of 50 and 15 flg/mL, respectively. A'lQ In 2001, Fusetani and coworkers isolated two inhibitors of membrane-type metalloproteinase (MTl-MMP, a key enzyme in tumor metastasis), from a marine sponge 232 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Cribrochalina sp., collected in western Japan. The NMR data for the major metabolite was almost identical to the one acquired for haplosamate A two years before, leading to a revision and correction of the haplosamate structures, as shown in Figure 5.7. Both compounds exhibited moderate inhibition, with ICso's of 150 ji,g/mL for haplosamate A (47) and 160 fig/mL for the minor metabolite (48). Na03SO° 'OH 0 < % OH OH "Haplosamate A" O H O H  SOsNa "Haplosamate B" Na0 3SO° o - p * OH OH 47 Haplosamate A (revised) Na03SO° ^~f^C "'P OH OH o—P! NaO7 D— Na03SO' 50 R = R! = H, Tremasterol A 51 R = Ri = Ac, Tremasterol B 52 R=H, R,=Ac, Tremasterol C HN bH 54 P03Na2 48 Haplosamate B (revised) Na03SP OH 53 s HN N ,N 55 NH Figure 5.7. Marine natural products combining phosphorus and sulfur. 233 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Although steroidal sulfates are widely distributed in marine sponges, steroids containing both sulfate and phosphate groups have only been reported from marine sources once before, by De Riccardis and coworkers440 in 1992. Christened as tremasterols A (50), B (51) and C (52), they were isolated from the starfish Tremaster novaecaledoniae (Jangoux, 1982) collected at a depth of 530 m off New Caledonia. A year later, the same authors added compound (53). 441 Figure 5.7 also shows one more relevant P-S combination found in marine organisms: (£)-2-(l-methyl-2-oxopropylidene)phosphorohydrazidothioate (£) oxime (54), an ichthyotoxic metabolite from the dinoflagelate Gymnodinium breve (implicated in the production of toxic red tides in Florida),442 and a minor cytotoxic metabolite (55) from the fungus Lignincol laevis. Furthermore, an additional heteroatomic 6-membered ring in the basic sterol structure sets haplosamates A (47), B (48) as well as the minor metabolite (49), apart from other marine and terrestrial sterol sulfates. Such a feature has been reported before only in three steroids isolated from aerial parts of plants (Figure 5.8).444" 7 OH OH 57 R = H 56 58 R=CH3 Figure 5.8. Structures for the steroidal glucoside (56) (from Vernonia hindii S. Moore, Asteraceae), M5 and two members of the withanolides family (from Physalis philadelphica Lam, Solanaceae).446,447 5.9. Haplosamate A: NMR analysis Figures 5.9, 5.10 and Table 5.5 show the NMR data for haplosamate A (47) in D2O (see Experimental section for NMR spectra in CD3OD as in the original report,437 and DMSO-cfe)- 234 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Na03SO% H6 H3 H4 H15 !! H16 OH OH H29 H23 H7 ' ' ' k^'' J ^ -J I H8 H2E H2a JL k t H5 H22a H20 | H 1 4 H1a H2p H19 H26 H27 H28 H21 H1f H24 j| |H11 |i H18 iH12a I \l\ l l H 2 2 P I1 ' V.: i 4.5 4.0 3.5 3.0 2.5 Chemical Shift (ppm) 2.0 1.5 1.0 0.5 Figure 5.9. H-NMR spectrum of haplosamate A (47) (recorded in D20 at 600 MHz). 235 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A C15 My % 81.4 C4 C6 C23 C7 C3 C16 C15 liiMiAitklfal Na03SO' "OH K1 J t>- NaO CH3CN OH OH C14 C29 I C26 C24 W \ J I V 56.9 53.55 C17 C9 C5 C14 C29 fW C25 C1 C13 C 1 0 C 8 C20 C22 C12 T"Wr WW C21 C2 C19 C28 C27 C18 V 11 96 88 80 72 64 56 48 40 Chemical Shift (ppm) 32 24 16 8 0 Figure 5.10.13C-NMR spectrum of haplosamate A (47) (recorded in D20 at 150 MHz). 236 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.5. NMR data for haplosamate A (47) (recorded in D20). Na0 3 SO' OH OH Carbon No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 "C 5(ppm)8 34.2 23.4 78.5 77.1 44.6 75.3 79.0 34.1 53.1 36.5 19.4 39.6 42.6 57.6 82.0 91.2 61.9 15.4 17.5 33.4 20.5 40.1 83.5 44.4 27.9 21.7 17.2 10.6 54.0 'H 5(ppm)(mult,7(Hz))bc HP 1.27 (m), Ha 1.51 (m) Ha 2.07 (dd, 7 = 14.2,14.3 Hz) HP 1.86 (m) 4.40 (d, 7 = 2.6 Hz) 4.06 (s, broad) 1.54 (m) 4.15 (s, broad) 3.45 (dd, 7 = 3.8, 10.6 Hz) 2.21 (ddd, 7 = 10.6, 10.6, 10.7 Hz) 0.99 (m) 1.42 (m), 1.54 (m) Ha 1.77 (d, 7 = 13.4 Hz) HP 0.88 (m) 1.49 (m) 4.64 (s, broad) 3.95 (dd, 7 = 4.0, 10.1 Hz) 0.79 (m) 0.95 (s) 1.25 (s) 1.88 (m) 0.97 (d, 7 = 6.3 Hz) Ha 1.87 (m) HP 1.21 (dd, 7 = 4.1,12.7 Hz) 3.54 (dd, 7 = 8.9, 9.1 Hz) 1.42 (m) 2.00 (m) 0.88 (d, 7 =6.9 Hz) 0.80 (d, 7 = 6.9 Hz) 0.75 (d, 7= 6.9 Hz) 3.62 (d, 7= 10.9 Hz) HMBC" (H->C) CI C1,C3,C10 C4,C6,C10, C19 C7,C10 C8 C7, C9, C14 C7,C8,C9,C13,C17,C18 C13.C16 CI5, C20 C13.C16, C18,C20,C22 C12,C13,C14,C17 C1,C5,C9,C10 CI7, C20, C22 C23 C24, C26, C27, C28 C23, C24, C25, C27 C24, C25, C26 C23, C24, C25 P ecorded at 150 MHz. b Recorded at 600 MHz.c According to HMQC recorded at 600 MHz. 237 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A A number of shielded methyl doublets H28 (5H 0.75, 5C 10.6), H27 (5R 0.80, 5C 17.2), H26 (5H 0.88, 5C 21.7) and H21 (5H 0.97, Sc 20.5), as well as singlets H18 (5H 0.95, 5C 15.4) and H19 (5H 1.25, 8c 17.5), are very characteristic of a basic sterol structure. The most striking feature is the additional methyl doublet H29 (5H 3.62, 8c 54.0), assigned to the O- methylphosphate functionality. Flanking this resonance are methines H23 (8H 3.54, 8c 83.5) and H16 (8H 3.95, 8c 91.2), connected through an oxygen atom. The remaining methines in this area of the 'H-NMR spectrum are either hydroxylated, as in H7 (8H 3.45, 8C 79.0), H4 (5H 4.06, Sc 77.1) and H6 (5H 4.15, 8C 75.3), sulfated as H3 (8H 4.40, Sc 78.5), or phosphorylated as H15 (8H 4.64, 8C 82.0). The presence of a phosphate group also generates splitting in neighboring carbon resonances (Figure 5.10), particularly C29 (8C 54.0, 2/c,p = 6.0 Hz), C14 (8C 57.6, Vc,p = 9.0 Hz) and C15 (8c 82.0, 2/c,p = 7.5 Hz). Figure 5.11 presents the dramatic change in multiplicity observed when the ] H-NMR spectrum is acquired while decoupling 31P. The presence of one phosphorous atom was also confirmed by a singlet at 0.84 ppm in the 31P-NMR spectrum. HRESIMS gave a [M+Na]+ peak at m/z 721.2377, consistent with the elemental composition C29H490i2Na2PS (calculated for C29H490i2Na3PS: 721.2375). In general, bioassaying cannabinoids both in vivo and in vitro has always faced water solubility problems, due to the high lipophilic character of these ligands.411 Therefore, water- miscible vehicles (such as EtOH, DMSO, bovine serum albumin) have been used for the administration and test of cannabinoids, requiring additional control experiments to evaluate any possible pharmacological change produced by the vehicle itself. The present cell-based bioassay was also run in aqueous media, but haplosamate A (47) is completely water-soluble, eliminating the use of additional solvents. 238 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A a) Na03SO' b) H3 H4 H6 H16 : ( i H29 H23 & i'J'Ji ! ! i ! -> .». /->. 4.5 4.0 3.5 Chemical Shift (ppm) Figure 5.11. a) 31P decoupled and b) 31P coupled 'H-NMR spectra of haplosamate A (47) (recorded in CD3OD at 400 MHz). 239 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Sulfur is the fourth more abundant element in sea-water, after chlorine, sodium and magnesium. Its sulfate ion follows chloride in importance and is the most stable sulfur combination in sea-water.439 This correlates with the fact that, according to MarinLit,448 there are approximately 1838 marine-derived sulfated compounds mostly isolated from Porifera and Echinodermata phyla, whereas reports on organophosphorus are close to 97 structures. 5.10. Pharmacophoric requirements for cannabinoid activity A general trend for most steroidal sulfates isolated via bioassay-guided fractionation, as evidenced by Qureshi and Faulkner,437 is the loss of activity after the natural product is desulfated. In general, sulfated compounds are known to irreversibly bind proteins, since the sulfate functionality is able to establish multiple hydrogen bonds with amino acids in the binding sites, or simply adhere to phospholipids in cell membranes.449 In most cases, such interactions are completely unspecific and the usually displayed high activity disappears upon sulfate removal. When haplosamate A (47) was refluxed in acidic media and acetylated using standard conditions, the triacetyl-derivatives (60) and (61) were obtained. Surprisingly, the cannabinoid activity was retained in both triacetates, suggesting that the phosphate and sulfate groups in haplosamate A (47) are not required for activity, and possible do not even interact with CBl or CB2. 61 240 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A NMR data for both triacetates (Table 5.6) showed an absence of the methylphosphate group, and the presence of two new alkene carbons C14 (8c -151) and C15 (5c -124). HMQC data assigned only a proton (H15, 6H 5.64) to the second carbon. For derivative (60), HMBC correlations between H3 (5H 5.12) and C29 (8C 169.2), H4 (8H 5.47) and C31 (5C 169.9), as well as H7 (8H 4.93) and C33 (SH 169.9), allow unambiguous assignment of three acetate groups (see Experimental section). HMBC data for compound (61) does not show cross peaks between H3 (5H 4.77) or H4 (SH 3.51) and acetate carbons, but the deshielding effect commonly observed in oxygenated methines when attached to an acetate suggested that H3 (5H 4.77) was also acetylated. Water loss and the presence of only three acetates were additionally confirmed by HRESEVIS, which yielded [M+Na]+ ions at m/z 611.3558 for 60 and 611.3564 for 61, in concordance with the molecular formula C34H52O8 (calculated for C34H520sNa: 611.3560). Given the right geometry in the starting material, as well as bulkiness and high stability in the leaving group, it is not surprising for an El elimination to occur under hot acidic conditions (Scheme 5.1). Scheme 5.1. El elimination in haplosamate A (47). 241 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.6. NMR data for haplosamate A triacetates (60) and (61) (recorded in CeDe). c No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 1 3 C 8 (ppm)a 34.8 23.1 70.5 73.2 44.0 72.0 76.3 36.5 53.3 36.0 22.0 42.3 46.3 151.0 123.9 86.6 65.9 17.1 18.0 32.0 20.8 1 v — ^*vJ--'" 1 llJL AcO'' ^ T i TT OAc 1 H 1 OR, OR2 \ d 15 60 R,= OAc, R2=H 61 R,= H, R2=OAc 60 8(ppm)(muIt,7(Hz))bc HP 1.13 (td, 3.3, 13.6) Ha 1.34 (m) Hal.94(ddt, 3.0, 14.4, 14.6) HP 1.72 (m) 5.12 (d, 2.8) 5.47 (s, broad) 1.54 (m) 4.43 (s, broad) 4.93 (dd, 3.3, 10.8) 2.63 (t, 11.3) 0.65 (td, 2.5, 11.9) 1.31 (m) Ha 1.74 (m) Hp 1.16 (td, 3.9, 14.1) 5.64 (s, broad) 4.13 (dd, 2.2,7.1) 1.23 (dd, 8.0,9.4) 0.90 (s) 1.38 (s) 1.63 (m) 0.90 (d, 6.9) 22 39.3 Ha 1.54 (m) HP 0.99 (dd, 11.2, 11.6) 'Recorded at 150 MHz.bRecorded at 600 MHz.cAccording to I 13 C 8 (ppm)a 35.1 22.5 72.8 73.3 43.3 72.6 73.8 36.2 53.3 37.2 21.8 42.3 46.3 150.9 124.3 86.5 65.8 17.1 18.0 32.0 20.8 V" / 26 61 JH 8(ppm)(mult,7(Hz))bc Ha 1.11 (td, 3.3, 13.8) Hp 1.31 (m) Ha 1.96 (ddt, 3.3, 14.3,14.4) HP 1.54 (m) 4.77 (d, 2.5) 3.51 (m) 1.42 (m) 5.83 (d, 2.4) 5.06 (dd, 3.8,11.0) 2.65 (t, 11.3) 0.68 (td, 2.2, 11.6) 1.40 (m), 1.31 (m) Ha 1.73 (dd, 3.1, 3.6) Hp 1.17 (td, 3.3, 13.3) 5.72 (s, broad) 4.12 (dd, 2.5,9.4) 1.23 (dd, 9.7,10.8) 0.90 (s) 1.33 (s) 1.62 (m) 0.89 (d, 6.4) 39.4 Ha 1.54 (m) Hp0.98(dd, 11.6, 12.0 Hz) 3MQC recorded at 600 MHz.dInterchangeable for 73. 242 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.6. NMR data for haplosamate A triacetates (60) and (61) (recorded in CeDeXCont.) O R T O R 2 60 Rj= OAc, R2=H 61 R,= H, R2=OAc c No l^lU 23 24 25 26 27 28 29 30 31 32d 33 34d 1 3 C 8 (ppm)a 80.7 44.7 28.2 18.4 22.1 11.3 169.2 21.0 169.9 21.5 169.9 20.9 XH 8(ppm)(mult,/(Hz))bc 3.38 (ddd, J = 2.2,1.2, 9.7 Hz) 1.66 (m) 2.16 (m) 0.87 (d, J = 6.6 Hz) 0.94 (d, 7 = 6.9 Hz) 0.88 (d, 7 = 7.2 Hz) 1.66 (s) 1.76 (s) 1.42 (s) 1 3 C 8(ppm)a 80.7 44.7 28.2 18.3 22.1 11.3 169.4 21.2 170.6 21.4 170.2 21.0 ln 8(ppm)(muIt,/(Hz))bc 3.37 (ddd, J = 2.2, 7.2, 9.3 Hz) 1.66 (m) 2.17 (m) 0.87 (d, 7 = 6.4 Hz) 0.95 (d, 7 =6.7 Hz) 0.88 (d, 7 = 6.4 Hz) 1.72 (s) 1.76 (s) 1.67 (s) "Recorded at 150 MHz.bRecorded at 600 MHz.cAccording to HMQC recorded at 600 MHz ̂ Interchangeable for 73. Retention of the cannabinoid activity in triacetates (60) and (61) led to an extensive examination of the structural requirements for their cannabinoid activity. Traditional cannabinoid structure-activity relationships (SAR)411'450"454 indicate three general molecular requirements for activity (Figure 5.12): I. Bonding availability (A) at one end (H bridge, % electron interactions). II. An appropriately oriented carbocyclic ring system (B) in the central part of the ligand. III. A lipophilic alkyl side chain (C) at the opposite end. 243 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Figure 5.12. The classical cannabinoid pharmacophore. In addition, multiple computer-assisted studies have shown superimposition of electronegative regions (associated with carbonyl or hydroxyl oxygen atoms), linear hydrocarbon side chains, and ^-electron rich areas, between low-energy conformations for A9- tetrahydrocannabinol (1) and anandamide (15),454 as well as (-)-9(5-hydroxyhexahydrocannabmol (62) and (/?)-WIN55212-2 (23).455 This suggests common pharmacophore elements among phytocannabinoids, endocannabinoids and aminoalkylindoles. When the minimum energy conformation of haplosamate A (47) was estimated through molecular mechanics (RMS = 0.100, CS Chem3D Pro® 9.0) following Xie, Eissenstat and Makriyannis publication,455 the 1,2-dimethylpropyl side chain and the pentacyclic ring system show acceptable alignment with similar features in (-)-9|3-hydroxyhexahydrocannabinol (62) and (/?)-WIN55212-2 (23). Due to its polyhydroxylated character, haplosamate A (47) possesses several possibilities for hydrogen bonding interactions (Figure 5.13). 244 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Figure 5.13. Minimum energy conformations for the known cannabimimetic agents (-)-9(3- hydroxyhexahydrocannabinol (62) and WIN55212-2 (23). 18 Properly oriented, haplosamate A (47) shows acceptable alignment. 245 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A In summary, Figure 5.13 predicts that haplosamate A (47) may interact with CB1 and CB2 in the same fashion as classical and synthetic cannabinoids, and not unspecifically via sulfate or phosphate functionalities. This observation correlates with the cannabinoid activity observed for triacetate derivatives (60) and (61), suggesting that neither sulfate nor phosphate groups are required for receptor binding. 5.11. Evidence of haplosamate A binding to CB1 and CB2: STDD NMR experiments Because the cannabinoid receptor's structure and shape are still being characterized and no crystallographic data is available showing receptor-bound ligands, knowledge regarding the nature of ligand/receptor interactions is limited. Specific characterization of those parts of a ligand in direct contact with a protein is mostly left to X-ray analyses of cocrystallized ligand- receptor complexes, 5 but crystallization of membrane receptor proteins is extremely difficult, and even if a truncated form is available, its binding affinity or specificity may differ from the native receptor embedded into a membrane.457 In 1999, Mayer and Meyer458 introduced saturation transfer difference (STD) NMR, developed to characterize binding interactions at an atom level, or group epitope mapping (GEM). Several NMR methods such as transferred nuclear Overhauser effect (trNOE),459 SAR by NMR,460 NOE pumping,461 and competitive binding spectroscopy,462 are also employed to screen and study binding processes, but these techniques do not allow the observation of binding ligands to membrane proteins in their natural environment. '458 STD NMR has been used in the last seven years as an efficient tool to study protein-ligand recognition events in a variety of systems.463^167 Advantages include high sensitivity, requiring small amounts of both protein and ligand. The methodology enables mapping of the ligand's binding epitope, since those parts of the ligand having the strongest contact to the protein exhibit the most intense NMR signals, and 246 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A the binding component can be identified even from mixtures, allowing it to be used as a screening method for ligands with dissociation constants ranging 10-10" M. '4 8 Very valuable and related to this last advantage is that the STD pulse sequence (Figure 5.14) can be applied to most ID or nD NMR experiments already in use, to generate STD versions of TOCSY, COSY, NOESY and inversely detected 13C or 15N spectra, for example.458 This makes structure elucidation conceivable even before the binding ligand has been purified.468 The basis for STD NMR is the transfer of saturation from the protein to the ligand.456'458'4 9 Selective saturation of a protein is possible because its resonances are often anisotropically shifted as well as broadened, and can be irradiated outside the spectral window of low-molecular-weight ligands (usually 0-10 ppm).456 In practice, this is done using the pulse sequence detailed in Figure 5.14. ^ S T D : Figure 5.14. Basic pulse sequence of a ID STD NMR spectrum, di: relaxation time; 8: delay between hard pulses (usually 1 ms); n: number of pulses (around 40 is recommended); T/p: filter, consisting of a spin-lock pulse, to eliminate broad resonances of the protein. 456,458 The ID STD NMR corresponds to a modified ID NOE difference pulse sequence. The experiment is performed by selectively saturating the protein with a train of Gauss-shaped pulses separated by 8.456 In the first scan, these hard pulses are centered on a region of the !H NMR that contains only resonances for the protein envelop: any frequency ranging -0.4 to -5 ppm, or 8 to 15 ppm if no aromatic or acid moieties are present in the ligands is usually chosen {on-resonance irradiation) 469 247 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Once the protein is saturated, the remainder of the pulse sequence corresponds to a ID NOE experiment. The saturated protein will transfer its magnetization to the bound ligand through spin diffusion. Then, the experiment is repeated aiming the train of hard pulses this time to another region of the H NMR where no protein signals are present: anywhere between 30-100 ppm (off-resonance irradiation).457 Subtraction of the spectrum without saturation (off-resonance) from the spectrum with saturation of the protein (on-resonance), yields the final STD NMR spectrum that cancels all resonances, except those from species with binding affinity. The difference between STD spectra is obtained by internal alternated subtraction with appropriate phase cycling using a frequency list for on-resonance and off-resonance irradiations. 7' From a macromolecular viewpoint, proteins are constituted by a large system of proton spins tightly coupled by dipole-dipole interactions due to restricted mobility (slow tumbling rates). Thus, selective saturation of a single protein resonance (See Figure 5.15,1) will result in a rapid spread of magnetization over the entire molecule via spin diffusion (II). Intermolecular transfer of magnetization from protein to ligand (also by spin diffusion) leads to progressive saturation of the bound ligands (III). This saturation is then transferred into solution through fast exchange of ligand molecules from the bound to the free state (IV), where the saturation-transfer effect is detected. The difference mode ensures that only nuclei of molecules that were at one time bound to the receptor contribute to the STD spectrum.456'468 Resonances of non-binding compounds are canceled out, since they do not become saturated. Additionally, the ligand's binding epitope can be determined from the STD spectra because protons in close proximity to the protein surface carry a much larger saturation (V, Figure 5.15), and therefore more STD signal intensity than other nuclei situated far away from 248 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A the binding site, which obtain saturation (if any) only through spin diffusion within the ligand 456,457 { Non-binding / I molecule jr f -Non-binding / 1 molecule jf \7u-i Free ligand - * « - \e Insect cell membrane (-**> Selective Sk saturation Mr l«»ke Cannabinoid recepror-ligand complex Increasing Figure 5.2. Events involved in STD NMR spectroscopy: I) selective protein saturation; II) magnetization transfer within the protein and, III) to the bound ligand via spin diffusion; IV) receptor-ligand complex dissociation that carries saturation into solution; V) differential saturation degree in protons of the ligand correlates to their proximity to the protein.456'458 The intensity of the STD signal increases as a function of ligand excess, provided that ligand molecules with zero or little saturation bind to the receptor. ' Since the method relies on transferred saturation from protein to ligand, the larger the number of ligands that pick up magnetization, the greater the build-up in STD signal intensity. Therefore, faster turnover rates at high ligand excess results in a larger STD effect.457 The application of a T/^rfilter in the STD pulse sequence (Figure 5.14) is a common procedure to suppress NMR signals of large molecules. However, sometimes even such a spin 249 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A lock field is not enough to eliminate signals originating from the biological components in the system (buffers, glucose, host cells), which produce hump regions with extreme proton overlap. To eliminate these disturbing background resonances, Meyer and coworkers introduced in their last update an additional filter, or STDD-filter (Figure 5.16). cell suspension STD A __ STDB = STDD spectrum (cells+ligand) (cells only) Figure 5.16. Sample preparation for saturation transfer double-difference (STDD) NMR.469 The idea behind the STDD-filter is to divide the cell suspension into two NMR tubes, one of which has been preloaded with ligand. This tube will provide a STD spectrum containing signals generated by binding of the ligand to the receptor. The second NMR tube, prepared under the same conditions but without ligand, affords another STD spectrum with background resonances. A further subtraction of these two STD's yields the new saturation transfer double- difference spectrum, showing only signals of the binding ligand. As a result, acquisition times of STDD spectra are shorter, and spin lock pulses (Tlp) which often result in loss of saturation from the ligand due to Ti and T2 relaxation processes, are not required (better S/N ratio).469 250 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Figure 5.17 displays STDD spectra obtained for buffered aqueous solutions of each cannabinoid receptor embedded in insect cells, and haplosamate A (47) in an approximate 24000-fold excess of ligand/receptor. In order to detect unspecific interactions that may take place between ligand and host cells, two negative controls were prepared by adding haplosamate A (47) to SF21 insect cells incubated in absence of receptors, and to F8 (coagulation factor VIII) cells. As shown in Figure 5.17 (Controls 1 and 2), no STD effects were observed for these control samples, indicating that haplosamate A (47) does not bind unspecifically to the cell membranes. The STD spectra prove that haplosamate A (47) specifically binds to the cannabinoid receptors CB 1 and CB2, since signals of the ligand can be observed. Sucrose resonances between 3.25-4.25 ppm in the reference spectra, are evidenced by a series of dispersion peaks in both STDD spectra, which obscure any STD effect given by methine protons in this region. The high concentration of sucrose in the media (0.080 mM) generates very intense NMR signals that cannot be completely subtracted in the STDD spectrum. Another practical aspect that contributes to deficient signal subtraction is the use of two NMR tubes for each sample (Figure 5.16), which introduces shimming differences. Fortunately, some of the shielded protons between 0.50 and 2.25 ppm can be observed without any interference. Signal broadening is a common issue in STD experiments and it results from ligand- receptor exchange processes or fast Ti relaxation in the protein. The protein envelop is basically absent in the reference spectra, which show a flat baseline. Although early references456'458 show broad hump regions reaching negative chemical shift values, this is not always the case. ' '4 On- and o/f-resonances were chosen as -1.1 and 114 ppm as suggested by Mayer and Meyer.456'469 251 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A STDD Insect cells F8 + Haplosamate A (Control 1) hPlff 111" i>H^»M^*»*^»<>liP'^i»tl^' l>'Mi^t^ STDD Insect cells SF21 + Haplosamate A (Control 2) STDD Insect cells + CB1 + Haplosamate A H-NMR Insect cells + CB1 + Haplosamate A (Reference) J STDD Insect cells + CB2 + Haplosamate A ^Aj -Mj i^X- id^ lv* -^ i Will \ I 'H-NMR Insect cells + CB2 + Haplosamate A (Reference) iUU Uv j—<__ A J A A J \J\ H-NMR Haplosamate A in D20 H29 / V H18 H «  H3 H g H 4 15 , , , H16 H22a H14 H23 H7 H 8H2a H20 2P H12a \>, H1a H19 H21 u o jH2P H5 H24 H23 , H11 HH22P H26 H27 H28 H9i "  ; j v 1 * ; 4.5 4.0 3.5 3.0 2.5 2.0 Chemical Shift (ppm) 1.5 1.0 0.5 Figure 5.17. STDD experiments measured for an aqueous solution of each cannabinoid receptor supported in SF21 insect cells and haplosamate A (47). 252 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 3.5 3.0 2.5 Chemical Shift (ppm) 0.5 Figure 5.18. Qualitative STD NMR group epitope mapping for the binding of haplosamate A (47) to the cannabinoid human receptors CB 1 and CB2. 253 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Only those protons of the ligand which are nearest to the binding site of the protein can easily be identified from STDD spectra, since they are saturated to the highest degree. Hence, if both STDD and the *H-NMR of 47 are compared (Figures 5.17, 5.18), some resonances are missing in the STDD spectra. The different signal intensities are best analyzed by integrating the broad proton resonances and referencing them to the most intense signal.456'458 However, the reduced S/N ratio obtained for the STDD spectra makes peak integration impractical. Additionally, just a few NMR signals are dispersed enough to be separately integrated. Therefore, only those STD resonances of significant intensity were considered for group epitope mapping (Figure 5.18). Clearly, the strongest STD effect is localized in the (l,2-dimethyl)propyl side chain (H26-H28) and protons of methyls H18/H21 as well as methine H6 of the ring system. The latter indicates a possible hydrogen bond between C6-OH and amino acids in the receptor. The STD NMR data points out the same binding epitope in 47 for both cannabinoid receptors. These results match and confirm the prediction established in Section 5.10, where through molecular mechanics the low energy conformation of 47 was estimated. Therefore, haplosamate A (47) binds CB1 and CB2 in the same fashion as some phytocannabinols, endocannabinols, and synthetic analogues. Since the sulfate and phosphate bulky groups are not required for activity and protons nearby do not exhibit strong STD effects, it is reasonable to assume that they are not in close proximity to the binding site. 254 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.12. Conclusions Haplosamate A (47) constitutes the first member of a new family of cannabinoid-active compounds. Structurally different than phytocannabinoids, endocannabinoids and synthetic analogues known to date, it is the only marine-derived sterol that exhibits such activity. Natural occurring steroids in Cannabis sativa, such as stigmasterol (63), ergosterol (64), P-sitosterol (65) and campesterol (66), are not psychoactive (Table 5.1, Section 5.2).4 '4 The only active steroid-like compound reported so far is the synthetic analogue (67), prepared by Razdan, Pars and Granchelli423 in 1968. 66 67 The presence of sulfate, phosphate and several hydroxyl groups bestow haplosamate A (47) with high water solubility, a very desirable characteristic when studying cannabinoid receptor inhibition in vivo or in vitro. Historically, lack of water solubility has always presented a challenge during bioassay due to the lipophilic character of typical cannabinoid ligands. Based on the activity of 47, it is reasonable to propose the use of phosphate and sulfate groups to modify traditional inhibitors and improve their solubility in water without interference in activity, since as proven by STDD NMR and chemical degradation, they do not play any crucial role in the binding process. 255 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Saturation transfer double-difference (STDD) NMR spectroscopy is a fast and versatile methodology to screen and characterize binding processes. Selective saturation of CB1 and CB2 yielded difference spectra where only those protons of haplosamate A (47) close enough to the binding site in the protein were visible, confirming that 47 specifically binds to these receptors. Additionally, STDD-derived group epitope mapping correlates with computer-assisted predictions about the binding mode of 47 to the cannabinoid receptors. For both CB1 and CB2, this natural ligand interacts via the classical cannabinoid pharmacophore, which suggests that the (l,2-dimethyl)propyl side chain, the ring system, and hydrogen-bonding groups, are required for activity (Figure 5.19). 18 21 H 3 C CH OS CH- 28 Figure 5.19. STDD NMR-derived group epitope mapping for haplosamate A (47). Only protons displaying STD effects are numbered. OS: sulfate; OP: methylphosphate. Although group epitope mapping characterization used to be left mostly to X-ray crystal analyses, the present investigation proves that STDD NMR can also provide binding epitope information in a short amount of time. On ligands, the only limitation imposed is that they must not be affected by the selective saturation pulse. On the biological system, this must be fully operational during the whole acquisition time. In our case, since CB1 and CB2 are integrated 256 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A into the insect cell membranes, efficient protein saturation was not a limiting factor. Moreover, the use of STDD NMR allowed a complete data acquisition within the life span of the host cells. From the biomedical viewpoint, most relevant proteins in drug discovery are membrane- bound, with about 7% of the human genome coding for seven helix GPCR's,457 such as CBl and CB2. Since their binding specificity and affinity is very susceptible to parameters including ionic strength, ion composition, and electric field strength, it is desirable to study membrane proteins in living cells, avoiding the use of truncated receptor forms. STDD NMR has made it possible to investigate both cannabinoid receptors and their binding events directly in a cell-based bioassay, which provides the closest conditions to a natural environment. 257 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 5.13. Experimental General experimental procedures For general experimental procedures see Section 2.6. Isolation procedure Samples of a red/purple irregular tube sponge (350 g wet wt) were collected from Keviang in Papua New Guinea on September of 2003, by hand using SCUBA at depths of 15 m (2° 45.33' S, 150° 41.23' E). The sponge was later identified as Dasychalina fragilis (Ridley & Dendy, 1886) by Dr. R. van Soest (University of Amsterdam), and a voucher specimen was deposited at the Zoologisch Museum, Amsterdam (ref. no. ZMA POR 19111). The collected material was frozen immediately upon collection and transported back to the University of British Columbia in coolers packed with dry ice. A portion of the frozen material (100 g) was extracted in MeOH (3 x 200 mL) and the combined extracts concentrated to dryness in vacuo to give a brown solid (0.60 g). This residue was treated with H20 to obtain a light yellow colored solution and an insoluble precipitate. The aqueous layer was extracted sequentially with hexanes (3 x 50 mL), CH2CI2 (3 x 50 mL) and EtOAc (3 x 50 mL), followed by concentration in vacuo of each partition. The water insoluble brown gum (major fraction, 0.38 g) was placed on a small reversed- phase column (20 g) and eluted with MeOH/tkO (7:3, 200 mL), to afford three major fractions: 183Mel (0.250 g), 183Me2 (0.032 g) and 183Me3 (0.056 g). ID and 2D NMR data confirmed the presence of the known compound haplosamate A in fraction 183Me2, in high purity. Further 258 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A purification by reversed-phase HPLC using CH3CN/H2O (7:3), afforded haplosamate A (47) (0.030 g, 0.043 mmol, 0.008% wet wt) as a white amorphous powder. [cc]20D -5.8 (c 0.8, MeOH); UV (MeOH) K^ (log e) 206 nm (2.98). For a summary of !H and 13C NMR assignments based on HMQC and HMBC data, see Tables 5.5, 5.7-5.8. *H NMR (CD3OD, 600 MHz) £4.73 (IH, m, broad), 4.39 (IH, d, 7 = 2.6 Hz), 4.15 (IH, s, broad), 4.01 (IH, s, broad), 3.88 (IH, d, 7 = 9.8 Hz), 3.61 (3H, d, J = 11.0 Hz), 3.44 (IH, m), 3.34 (IH, m), 2.30 (IH, dt, 7 = 10.5, 10.6 Hz), 2.09 (IH, m), 2.07 (IH, m), 1.89 (IH, d, 7 = 14.2 Hz), 1.85 (IH, d, 7 = 12.5 Hz), 1.83 (IH, m), 1.64 (IH, d, 7= 13.4 Hz), 1.52 (IH, m), 1.50 (IH, m), 1.45 (IH, m), 1.43 (IH, m), 1.40 (IH, m), 1.33 (2H, m), 1.32 (3H, s), 1.20 (IH, td, 7= 3.9, 12.9 Hz), 1.00 (3H, s), 0.98 (3H, d, 7 = 6.5 Hz), 0.90 (IH, m), 0.89 (3H, d, 7 = 6.8 Hz), 0.86 (IH, m), 0.82 (3H, d, 7 = 6.8 Hz), 0.77 (3H, d, 7 = 7.0 Hz), 0.72 (IH, dd, 7 = 10.2, 10.3 Hz); 13C NMR (CD3OD, 150 MHz) £92.4 (CH), 82.7 (CH), 82.0 (d, 7 = 7.5 Hz, CH), 80.1 (CH), 78.2 (CH), 77.7 (CH), 76.2 (CH), 63.0 (CH), 58.9 (d, 7 = 7.5 Hz, CH), 54.4 (CH), 53.4 (d, 7 = 6.0 Hz, CH3), 45.6 (CH), 45.3 (CH), 43.1 (C), 41.2 (CH2), 40.0 (CH), 36.7 (C), 35.1 (CH2), 34.8 (CH), 34.4 (CH), 28.4 (CH), 23.9 (CH2), 22.0 (CH3), 21.0 (CH3), 20.0 (CH2), 17.9 (CH3), 17.7 (CH3), 15.9 (CH3), 10.8 (CH3). HRESMS calcd for C29H490i2Na3PS ([M+Na]+): 721.2375; found 721.2377. lH NMR (DMSO-d6, 600 MHz) £5.68 (IH, s, broad), 5.08 (IH, s, broad), 4.72 (IH, s, broad), 4.61 (IH, s, broad), 4.08 (IH, d, 7 = 2.0 Hz), 3.89 (IH, s, broad), 3.79 (IH, s, broad), 3.53 (IH, d, 7= 9.0 Hz), 3.31 (3H, d, 7= 11.0 Hz), 3.26 (IH, dd, 7 = 9.0, 9.4 Hz), 3.13 (IH, s, broad), 2.14 (IH, ddd, 7 = 10.5, 10.5, 10.6 Hz), 2.03 (IH, m), 1.81 (IH, dd, 7 = 13.8, 13.8 Hz), 1.732 (IH, d, 7 = 12.0 Hz), 1.727 (IH, m), 1.68 (IH, dd, 7 = 13.0, 13.0 Hz), 1.58 (IH, d, 7 = 12.9 Hz), 1.36 (2H, m), 1.33 (IH, m), 1.27 (IH, m), 1.23 (IH, m) 1.23 (3H, s), 1.20 (IH, m), 1.12 (IH, dd,7 = 12.9, 13.6 Hz), 1.10 (IH, dd, 7 = 13.0, 13.6 Hz), 0.91 (3H, d, 7 = 6.3 Hz), 0.86 (3H, s), 0.83 (3H, 259 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A d, 7 = 7.0 Hz), 0.78 (IH, dd, 7 = 11.2, 12.0 Hz), 0.76 (IH, m) 0.70 (3H, d, 7 = 5.9 Hz), 0.69 (3H, d, 7 = 5.7 Hz), 0.60 (IH, dd, 7 = 10.1, 10.3 Hz); I3C NMR (DMSO-d6, 150 MHz) £92.5 (CH), 80.4 (CH), 78.6 (CH), 78.5 (d, 7 = 13.5 Hz, CH), 77.2 (CH), 74.3 (CH), 73.6 (CH), 61.0 (CH), 56.5 (d, 7 = 6.0 Hz, CH), 52.4 (CH), 51.0 (d, 7 = 6.0 Hz, CH3), 44.2 (CH), 43.4 (CH), 41.2 (C), 38.9 (CH2), 38.6 (CH2), 35.2 (C), 33.5 (CH2), 32.7 (CH), 32.5 (CH), 26.3 (CH), 22.4 (CH2), 21.3 (CH3), 20.3 (CH3), 18.3 (CH2), 17.0 (CH3), 16.7 (CH3), 15.2 (CH3), 10.1 (CH3). 'H NMR (D20, 600 MHz) £4.64 (IH, s, broad), 4.40 (IH, d, 7 = 2.6 Hz), 4.15 (IH, s, broad), 4.06 (IH, s, broad), 3.95 (IH, dd, 7 = 4.0, 10.1 Hz), 3.62 (3H, d, 7 = 10.9 Hz), 3.54 (IH, dd, 7 = 8.9, 9.1 Hz), 3.45 (IH, dd, 7 = 3.8, 10.6 Hz), 2.21 (IH, ddd, 7 = 10.6, 10.6, 10.7 Hz), 2.07 (IH, dd, 7= 14.3, 14.3 Hz), 2.00 (IH, m), 1.88 (IH, m), 1.87 (IH, m), 1.86 (IH, m), 1.77 (IH, d, 7 = 13.4 Hz), 1.54 (IH, m), 1.54 (IH, m), 1.51 (IH, m), 1.49 (IH, m), 1.42 (IH, m), 1.42 (IH, m), 1.27 (IH, m), 1.25 (3H, s), 1.21 (IH, dd, 7 = 4.1, 12.7 Hz), 0.99 (IH, m), 0.97 (3H, d, 7 = 6.3 Hz), 0.95 (3H, s), 0.88 (3H, d, 7 = 6.9 Hz), 0.88 (IH, m), 0.80 (3H, d, 7 = 6.9 Hz), 0.79 (IH, m), 0.75 (3H, d, 7= 6.9 Hz); 13C NMR (D20, 150 MHz) £91.2 (CH), 83.5 (CH), 82.0 (d, 7= 7.5 Hz, CH), 79.0 (CH), 78.5 (CH), 77.1 (CH), 75.3 (CH), 61.9 (CH), 57.6 (d, 7= 9.0 Hz, CH), 54.0 (d, 7 = 6.0 Hz, CH3), 53.1 (CH), 44.6 (CH), 44.4 (CH), 42.6 (C), 40.1 (CH2), 39.6 (CH2), 36.5 (C), 34.2 (CH2), 34.1 (CH), 33.4 (CH), 27.9 (CH), 23.4 (CH2), 21.7 (CH3), 20.5 (CH3), 19.4 (CH2), 17.5 (CH3), 17.2 (CH3), 15.4 (CH3), 10.6 (CH3). 31P NMR (CD3OD, 81 MHz) £0.84 (IP, s). 260 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A H29 H4 H6 H3 H15 H16 H23 Na03SO" OH OH i —-J"" <J Vj_.  u H11 H5 H20 H12a H25 HfP H 8  H2a | H22a !I i n . • H24 H14 h H19 H21 H1 H18 H27 H26 H12p H28 H17 »' '« 1r. 5.0 4.5 4.0 3.5 3.0 2.5 2.0 Chemical Shift (ppm) 1.5 1.0 0.5 Figure 5.20. 'H-NMR spectrum of haplosamate A (47) (recorded in CD3OD at 600 MHz). 261 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 96 88 80 72 64 56 48 40 32 24 Chemical Shift (ppm) 16 8 0 Figure 5.21. 13C-NMR spectrum of haplosamate A (47) (recorded in CD3OD at 150 MHz). 262 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.7. NMR data for haplosamate A (47) (recorded in DMSO-d6). Carbon No N a 0 3 S O " OH OH ,3C8(ppm)" 33.5 22.4 73.6 74.3 44.2 77.2 78.6 1H8(ppm)(muU,7(Hz))b'c Ha 1.23 (m) Hp 1.12 (dd, 7=12.9, 13.6 Hz) Ha 1.81 (dd, 7 =13.8, 13.8 Hz) HP 1.68 (dd, 7 =13.0,13.0 Hz) 4.08 (d, 7 = 2.0 Hz) 3.89 (s, broad) OH 5.08 (s, broad) 1.27 (m) 3.79 (s, broad) OH 4.72 (s, broad) 3.13 (s, broad) OH 5.68 (s, broad) HMBC" (H-»C) C2, C10, C19, C4, C10 C1.C4, C5 C2, C3, C10 C4, C6, C7, C9, C10.C19 C10 C9 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 ;ordei 32.7 52.4 35.2 18.3 38.9 41.2 56.5 78.5 92.5 61.0 15.2 17.0 32.5 20.3 38.6 80.4 43.4 26.3 21.3 16.7 10.1 51.0 d at 150 MHz. b] 2.14 (ddd, 7= 10.5, 10.5, 10.6 Hz) 0.76 (m) 1.36 (m) Ha 1.73 (d, 7 =12.0 Hz) Hp 1.10 (dd, 7=13.0, 13.6 Hz) 1.20 (m) 4.61 (s, broad) 3.53 (d, 7 = 9.0 Hz) 0.60 (dd, 7 =10.1, 10.3 Hz) 0.86 (s) 1.23 (s) 1.73 (m) 0.91 (d, 7 = 6.3 Hz) Ha 1.58 (d, 7 =12.9 Hz) Hp 0.78 (dd, 7 =11.2,12.0 Hz) 3.26 (dd, 7 = 9.0, 9.4 Hz) 1.33 (m) 2.03 (m) 0.83 (d, 7 = 7.0 Hz) 0.70 (d, 7 = 5.9 Hz) 0.69 (d, 7 = 5.7 Hz) 3.31 (d, 7= 11.0 Hz) C7,C9,C10,C14 C8,C11 C12 C9,C11,C13,C14,C17,C18 C7,C8,C12,C13,C17,C18 C13, C16 C15,C17,C20,C23 C12,C13, C14,C16,C18, C20 C12,C13,C14,C17 C1,C2, C4, C5,C10 C13 C13,C17, C20,C22,C23 C17, C20, C23, C24 C24, C28 C22, C23, C25, C26, C27, C28 C24, C26, C27, C28 C24, C25, C27 C24, C26 C23, C25 Recorded at 600 MHz.c According to HMQC recorded at 600 MHz. 263 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.8. NMR data for haplosamate A (47) (recorded in DMSO-rf6)- Na0 3 SO' OH OH Proton No la IP 2a 2P 3 4 40H 5 6 60H 7 70H 8 9 11 12a 12P 14 15 16 17 18 19 20 21 22a 22P 23 24 25 26 27 28 29 lH&(ppm)(mulUJ(Hz))' 1.23 (m) 1.12 (dd, J = 12.9,13.6 Hz) 1.81 (dd,7=13.8, 13.8 Hz) 1.68 (dd, 7 =13.0, 13.0 Hz) 4.08 (d, 7 = 2.0 Hz) 3.89 (s, broad) 5.08 (s, broad) 1.27 (m) 3.79 (s, broad) 4.72 (s, broad) 3.13 (s, broad) 5.68 (s, broad) 2.14 (ddd, 7=10.5, 10.5, 10.6 Hz) 0.76 (m) 1.36 (m) 1.73 (d, 7=12.0 Hz) 1.10 (dd, 7=13.0, 13.6 Hz) 1.20 (m) 4.61 (s, broad) 3.53 (d, 7 = 9.0 Hz) 0.60 (dd, 7 =10.1, 10.3 Hz) 0.86 (s) 1.23 (s) 1.73 (m) 0.91 (d, 7 = 6.3 Hz) 1.58 (d, 7 =12.9 Hz) 0.78 (dd, 7= 11.2, 12.0 Hz) 3.26 (dd, 7 = 9.0,9.4 Hz) 1.33 (m) 2.03 (m) 0.83 (d, 7 = 7.0 Hz) 0.70 (d, 7 = 5.9 Hz) 0.69 (d, 7 = 5.7 Hz) 3.31 (d,7= 11.0 Hz) Recorded at 600 MHz. COSY* H2a, H2P H2a Hla,Hip,H3 Hla,H3,H4,H5 H2a, H2P, H4 H2P,H3,40H,H5 H4 H4.H6 H5, 60H H6 H6, H8 H7, H9, H14 H8 H9, H12a, H12P H11.H12P Hll,H12a H8,H15 H14,H16 H15,H17 H16, H20 H17,H21,H22 H20 H22P, H23 H20, H22a, H23 H22a, H22P, H24 H23, H25, H27 H24, H26, H28 H25 H24 H25 P NOES H19 H11.H18.H19 H8 H18 H15,H17 H14 H14, H22p H8,H12a H2a, H8 H17 264 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Acetylation procedure A solution of haplosamate A (0.0074, 0.0106 mmol) was dissolved in CH3OH (5 mL) and refluxed for 2 h in the presence of HC1 (2 mL, 4 mmol, 2M). After solvent evaporation, the obtained residue was dissolved in pyridine (2 mL, 24.8 mmol) and treated with acetic anhydride (3 mL, 31.7 mmol). Upon stirring overnight at 25°C, the solvent was concentrated in vacuo and the resulting solid purified using normal phase column chromatography (20 g, EtOAc/hexanes 3:7) to afford two different acetylated compounds: haplosamate A 3,4,7-(60) (0.0029 g, 0.0049 mmol, 46%) and 3,6,7-(61) triacetates (0.0031 g, 0.0052 mmol, 49%). Haplosamate A 3,4,7-triacetate (60): for a summary of !H and 13C NMR assignments based on HMQC and HMBC data, see Tables 5.6, 5.9 and 5.10. *H NMR (C6D6, 600 MHz) S5.64 (IH, s, broad), 5.47 (IH, s, broad), 5.12 (IH, dd, J= 1.3, 2.8 Hz), 4.93 (IH, dd, J= 3.3, 10.8), 4.43 (IH, s, broad), 4.13 (IH, dd, J = 2.2, 7.1 Hz), 3.38 (IH, m), 2.63 (IH, dd, J = 11.3, 11.3 Hz), 2.16 (IH, m), 1.94 (IH, ddd, J= 3.0, 14.4, 14.6 Hz), 1.76 (3H, s), 1.74 (IH, m), 1.72 (IH, m), 1.66 (3H, s), 1.66 (IH, m), 1.63 (IH, m), 1.54 (IH, m), 1.54 (IH, m), 1.42 (3H, s), 1.38 (3H, s), 1.34 (IH, m), 1.31 (2H,m), 1.23 (IH, dd, J= 8.0, 9.4 Hz), 1.16 (IH, ddd, J = 3.9, 14.1, 14.6 Hz), 1.13 (IH, ddd, 7= 3.3, 13.6, 14.1 Hz), 0.99 (IH, ddd, 7= 1.2, 11.6 Hz), 0.94 (3H, d, J = 6.9 Hz), 0.90 (3H, s), 0.90 (3H, d, J = 6.9 Hz), 0.88 (3H, d, J = 7.2 Hz), 0.87 (3H, d, J = 6.6 Hz), 0.65 (IH, ddd, J =2.5, 11.9, 11.9 Hz); 13C NMR (QA;, 150 MHz) £169.9 (C), 169.9(C), 169.2 (C), 151.0 (C), 123.9 (CH), 86.6 (CH), 80.7 (CH), 76.3 (CH), 73.2 (CH), 72.0 (CH), 70.5 (CH), 65.9 (CH), 53.3 (CH), 46.3 (C), 44.7 (CH), 44.0 (CH), 42.3 (CH2), 39.9 (CH2), 36.6 (CH), 36.0 (C), 34.8 (CH2), 32.0 (CH), 28.2 (CH), 23.1 (CH2), 22.1 (CH3), 22.0 (CH2), 21.5 (CH3), 21.0 (CH3), 20.9 (CH3), 20.8 (CH3), 18.4 (CH3), 17.9 (CH3), 17.1 (CH3), 11.3 (CH3). HRESIMS calcd for C34H5208Na ([M+Na]+): 611.3560; found 611.3558. 265 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Haplosamate A 3,6,7-triacetate (61): for a summary of H and C NMR assignments based on HMQC and HMBC data, see Tables 5.6, 5.11 and 5.12. !H NMR (C6D6, 600 MHz) S5.S3 (IH, d, 7 = 2.4 Hz), 5.72 (IH, s, broad), 5.06 (IH, dd, 7 = 3.8, 11.0), 4.77 (IH, d, 7 = 2.5 Hz), 4.12 (IH, dd, 7 = 2.5, 9.4 Hz), 3.51 (IH, m), 3.37 (IH, ddd, 7 = 2.2, 7.2, 9.3 Hz), 2.65 (IH, dd, 7 = 11.3, 11.3 Hz), 2.17 (IH, m), 1.73 (IH, dd, 7 = 3.1, 3.6 Hz), 1.54 (IH, m), 1.40 (IH, m) 1.31 (IH, m), 1.31 (IH, m), 1.96 (IH, dddd, 7 = 3.1, 3.3, 14.3, 14.4 Hz), 1.76 (3H, s), 1.72 (3H, s), 1.67 (3H, s), 1.66 (IH, m), 1.62 (IH, m), 1.54 (IH, m), 1.42 (IH, m), 1.33 (3H, s), 1.23 (IH, dd, 7=9 .7 , 10.8 Hz), 1.17 (IH, ddd, 7 =3.3, 13.0, 13.3 Hz), 1.11 (IH, ddd, 7 = 3.3, 13.8, 13.8 Hz), 0.98 (IH, dd, 7= 11.6, 12.0 Hz), 0.95 (3H, d, 7 = 6.7 Hz), 0.90 (3H, s), 0.89 (3H, d, 7 = 6.4 Hz), 0.88 (3H, d, 7 = 6.4 Hz), 0.87 (3H, d, 7 = 6.4 Hz), 0.68 (3H, ddd, 7 = 2.2, 11.6, 11.6 Hz); 13C NMR (C6D6, 150 MHz) £170.6 (C), 170.2 (C), 169.4 (C), 150.7 (C), 124.3 (CH), 86.5 (CH), 80.7 (CH), 73.8 (CH), 73.3 (CH), 72.8 (CH), 72.6 (CH), 65.8 (CH), 53.3 (CH), 46.3 (C), 44.7 (CH), 43.3 (CH), 42.3 (CH2), 39.4 (CH2), 37.2 (CH), 36.2 (C), 35.1 (CH2), 32.0 (CH), 28.2 (CH), 22.2 (CH2), 22.1 (CH3), 21.8 (CH2), 21.4 (CH3), 21.2 (CH3), 21.0 (CH3), 20.8 (CH3), 18.3 (CH3), 18.0 (CH3), 17.1 (CH3), 11.3 (CH3). HRESIMS calcd for C34H5208Na ([M+Na]+): 611.3560; found 611.3564. 266 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.9. NMR data for haplosamate A 3,4,7-triacetate derivative (60) (recorded in CeD6). Carbon No 1 13C 8 (ppm)" 34.8 23.1 *H 5 (ppra) (mult, / (Hz))^ Hp 1.13 (ddd, 7= 3.3, 13.6, 14.1 Hz) Ha 1.34 (m) Ha 1.94 (ddd, 7 = 3.0,14.4,14.6 Hz) Hpl.72(m) HMBCh (H- C3, C9, CIO •C) 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32d 33 34d 'Recorded at 150 MHz. 70.5 73.2 44.0 72.0 76.3 36.5 53.3 36.0 22.0 42.3 46.3 151.0 123.9 86.6 65.9 17.1 18.0 32.0 20.8 39.3 80.7 44.7 28.2 18.4 22.1 11.3 169.2 21.0 169.9 21.5 169.9 5.12 (dd, 7=1.3, 2.8 Hz) 5.47 (s, broad) 1.54 (m) 4.43 (s, broad) 4.93 (dd, 7 = 3.3, 10.8) 2.63 (dd,7= 11.3, 11.3 Hz) 0.65 (ddd, 7 = 2.5, 11.9, 11.9 Hz) 1.31 (m) Ha 1.74 (m) H|3 1.16 (ddd, 7 = 3.9, 14.1,14.6 Hz) 5.64 (s, broad) 4.13(dd,7=2.2,7.1Hz) 1.23 (dd, 7 = 8.0, 9.4 Hz) 0.90 (s) 1.38 (s) 1.63 (m) 0.90 (d, 7 = 6.9 Hz) Ha 1.54 (m) HP 0.99 (ddd, 7 =1.2,11.6 Hz) 3.38 (m) 1.66 (m) 2.16 (m) 0.87 (d, 7 = 6.6 Hz) 0.94 (d, 7 = 6.9 Hz) 0.88 (d, 7 = 7.2 Hz) 1.66 (s) 1.76 (s) 20.9 1.42 (s) "Recorded at 600 MHz. 'According to HMQC recorded at 600 MHz. C1.C4, C5,C29 C2, C3, C10, C31/C33 C4, C6, C9,C10 C5, C7, C10 C6,C8,C14,C31/C33 C14,C15 C19 C8, C9,C10,C12 C9,C11,C13,C14 C8,C13,C14,C16,C17 C14,CI5,C20,C23 C8, C12, C13, C16, C18, C20, C22 C12.C13 C1,C5,C9,C10 C22, C23, C25, C28 C17, C20, C22 C21 CI6, C24, C28 C23, C24, C26, C27, C28 C24,C25,C27 C24, C25, C26 C23, C27 C29 C31/C33 C31/C33 •"interchangeable. 267 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.10. NMR data for haplosamate A 3,4,7-triacetate derivative (60) (recorded in C^). Proton No la IP 2a 2P 3 4 5 6 7 8 9 11 12a 12P 15 16 17 18 19 20 21 22a 22P 23 24 25 26 27 28 30 32b 34" ACO%,> ^ r = IT 1 H 1 OAc OH lH 8 (ppm) (mult, J (Jttz))' 1.34 (m) 1.13 (ddd, 7=3.3,13.6,14.1 Hz) 1.94 (ddd, 7 = 3.0, 14.4, 14.6 Hz) 1.72 (m) 5.12 (dd, 7 =1.3, 2.8 Hz) 5.47 (s, broad) 1.54 (m) 4.43 (s, broad) 4.93 (dd, 7 = 3.3, 10.8) 2.63 (dd, 7= 11.3,11.3 Hz) 0.65 (ddd, 7 = 2.5, 11.9, 11.9 Hz) 1.31 (m) 1.74 (m) 1.16 (ddd, 7 = 3.9,14.1,14.6 Hz) 5.64 (s, broad) 4.13 (dd, 7 = 2.2, 7.1 Hz) 1.23 (dd, 7 =8.0, 9.4 Hz) 0.90 (s) 1.38 (s) 1.63 (m) 0.90 (d, 7 = 6.9 Hz) 1.54 (m) 0.99 (ddd, 7 =1.2, 11.6 Hz) 3.38 (m) 1.66 (m) 2.16 (m) 0.87 (d, 7 = 6.6 Hz) 0.94 (d, 7 = 6.9 Hz) 0.88 (d, 7 = 7.2 Hz) 1.66 (s) 1.76 (s) 1.42 (s) 'Recorded at 600 MHz. ^ O A c COSY* Hip, H2ct Hla,H2a,H2P Hla,Hip,H2p, H3 Hip,H2a,H3,H4 H2a, H2P, H4 H2p, H3, H5 H4, H6 H5.H7 H6, H8 H7,H9,H15,H16 H8, Hll H9 H12P H12a H8, H16 H8, HI 5, HI 7 HI6, H20 H17,H21,H22P H20 H23, H22p H20, H23 H22a, H22P, H24 H23, H25, H28 H24,H26,H27 H25 H25 H24 NOESY* H19 H6 H7 H4 H5 H18.H19 H12P H9 H23, H28 H8 H2a, H8 HI 6, 28 H16, H23 268 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.11. NMR data for haplosamate A 3,6,7-triacetate derivative (61) (recorded in C^Ds). Carbon No 1 13CS(ppm)" 35.1 22.5 lK 8 (ppm) (mult, J (Hz))1* Ha 1.11 (ddd, 7 = 3.3, 13.8, 13.8 Hz) HP 1.31 (m) Ha 1.96 (dddd, 7 = 3.1,3.3,14.3, 14.4 Hz) Hpl.54(m) HMBC" (H-»C) C3,C5 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 "Recorded at 150 MHz 72.8 73.3 43.3 72.6 73.8 36.2 53.3 37.2 21.8 42.3 46.3 150.9 124.3 86.5 65.8 17.1 18.0 32.0 20.8 39.4 80.7 44.7 28.2 18.3 22. 1 11.3 169.4 21.2 170.6 21.4 170.2 21.0 4.77 (d, J = 2.5 Hz) 3.51 (m) 1.42 (m) 5.83 (d, 7 = 2.4 Hz) 5.06 (dd, 7 = 3.8, 11.0) 2.65 (dd, 7= 11.3, 11.3 Hz) 0.68 (ddd, 7= 2.2, 11.6, 11.6 Hz) 1.40(m), 1.31 (m) Ha 1.73 (dd, 7= 3.1,3.6 Hz) HP 1.17 (ddd, 7 = 3.3,13.0,13.3 Hz) 5.72 (s, broad) 4.12 (dd, 7 = 2.5,9.4 Hz) 1.23 (dd, 7 = 9.7, 10.8 Hz) 0.90 (s) 1.33 (s) 1.62 (m) 0.89 (d, 7 = 6.4 Hz) Ha 1.54 (m) HP 0.98 (dd, 7= 11.6,12.0 Hz) 3.37 (ddd, 7 = 2.2, 7.2, 9.3 Hz) 1.66 (m) 2.17 (m) 0.87 (d, 7 = 6.4 Hz) 0.95 (d, 7 =6.7 Hz) 0.88 (d, 7 = 6.4 Hz) 1.72 (s) 1.76 (s) 1.67 (s) ;. Recorded at 600 MHz. According to HMQC recorded at 600 MHz. CI C7,C9,C10,C19 C5,C7,C8,C31 C6, C8, C14, C33 C7, C9, C14 C5,C8,C10,C19 C9,C11,C17 C8,C13,C14,C16,C17 C14,C15,C20,C23 C12,C13,C16,C18,C20,C21, C10,C12,C14,C17 C1.C5, C9, C10 C17, C22 C20, C22, C23 C13, C20, C23 C16, C24, C25, C28 C22, C23, C25, C26, C28 C23, C24, C26, C27, C28 C24, C25, C27 C24, C25, C26 C23, C25 C29 C31 C33 C22 269 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Table 5.12. NMR data for haplosamate A 3,6,7-triacetate derivative (61) (recorded in CeD6). AcO" ' OH OAC Proton No la IP 2a 2(J 3 4 5 6 7 8 9 11 12a 12P 15 16 17 18 19 20 21 22a 22P 23 24 25 26 27 28 30 32 34 ln 8(ppm)(mu!t,/(Hz))a 1.31 (m) 1.11 (ddd, 7 = 3.3,13.8, 13.8 Hz) 1.96 (dddd, 7 = 3.1,3.3, 14.3, 14.4 Hz) 1.54 (m) 4.77 (d, 7 = 2.5 Hz) 3.51 (m) 1.42 (m) 5.83 (d, 7 =2.4 Hz) 5.06 (dd, 7=3.8, 11.0) 2.65 (dd, 7= 11.3, 11.3 Hz) 0.68 (ddd, 7 = 2.2, 11.6, 11.6 Hz) 1.40 (m), 1.31 (m) 1.73 (dd, 7= 3.1, 3.6 Hz) 1.17 (ddd, 7= 3.3,13.0,13.3 Hz) 5.72 (s, broad) 4.12 (dd, 7 = 2.5,9.4 Hz) 1.23 (dd, 7 = 9.7, 10.8 Hz) 0.90 (s) 1.33 (s) 1.62 (m) 0.89 (d, 7 = 6.4 Hz) 1.54 (m) 0.98 (dd, 7 =11.6, 12.0 Hz) 3.37 (ddd, 7 = 2.2,7.2,9.3 Hz) 1.66 (m) 2.17 (m) 0.87 (d, 7 = 6.4 Hz) 0.95 (d, J = 6.7 Hz) 0.88 (d, 7 = 6.4 Hz) 1.72 (s) 1.76 (s) 1.67 (s) "Recorded at 600 MHz. COSY8 NOESY8 Hip, H2a, H2P Hla, H2a, H2P Hla, Hip,H3,H2P Hla,Hip,H3,H2a Hl,H4a, H4p H3.H5 H4, H6 H5.H7 H6, H8 H7, H9, H15 H8,H11 H9, H12a H11,H12P H12a H8 H17 H16,H20 H17.H21 H20 H23, H22P H23, H22a H22a, H22P, H24 H23, H28 H26, H27 H25 H25 H24 H9 H19 H6 H1,H6, H7, H9 H1,H5,H7 H5, H6, H9, HI5 H18.H19 Hip, H5, H7, H12p H18 H9 H7 H18.H23 H8, HI 2a, HI 6 H2a, H8 H23 H23 H16,H20,H22a,H: H23 270 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A Saturation transfer double-difference (STDD) NMR sample preparation CB1 or CB2 cell lines were grown to confluency in T75 cell culture flasks and harvested by centrifugation at 500 g. The cell pellet was resuspended in 10 mL of a deuterated phosphate- buffered modified saline (PMS), prepared in D2O (99.9%) using 11 mM phosphate (pH 6.2), 40 mM NaCl, 40 mM KC1 and 80 mM sucrose. All preparations were done at room temperature. The suspension was centrifuged at 500 g for 10 min, the supernatant discarded, and the pellet resuspended in 5 mL of PMS. This wash was repeated two more times and after the last centrifugation, cell pellets were resuspended in 1.0 ml aliquots of PMS buffer which provided a cell concentration of approximately 5 x 106 cells/mL. Typical of GPCR expressing cell lines is a receptor density of about 10 per cell,4 which gives a total concentration of 5 x 101 receptors/mL. The NMR samples for each receptor were prepared in pairs of tubes, by previously adding haplosamate A (1.7 mg, 2.0 |Limol) to only one NMR tube, and splitting the above suspension into both tubes (total volume of each tube: 1.0 mL). This is equivalent to an approximate 24000-fold excess ligand/receptor. Saturation transfer double-difference (STDD) NMR measurements STD NMR spectra were recorded according to the procedure described by Claasen, Axmann, Meinecke and Meyer.4 9 All measurements were made at 298 K with a spectral width of 10 ppm on a Bruker Avance 600 MHz spectrometer, equipped with a 5 mm inverse double- resonance cryoprobe. Selective saturation of the protein was achieved by a train of Gaussian pulses of 50 ms length, truncated at 1% and separated by 1 ms delay. A total of 40 selective 271 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A pulses were applied, leading to a saturation train of 2.04 s. The on-resonance irradiation of the protein was performed at -1.1 ppm, while the off-resonance irradiation was set at 114 ppm, where no protein signals were present. The total number of scans was 64, 256 or 512, preceded by 8 dummy scans. The spectra were subtracted internally by phase cycling after every scan using different memory buffers for on- and off-resonance. A second manual subtraction of the STD spectrum for cells+receptor from the STD spectrum for cells+receptor+(47) sample, afforded a double-difference saturation transfer (STDD) spectrum. Spectra processing was performed on Silicon Graphics Octane workstations using XWinnmr 3.1 software (Bruker). Dr. Wolfgang Bermel from Bruker Analytik (Germany) kindly provided the automation routine commonly used for the additional difference in STDD. Group epitope mapping analysis The group epitope mapping was accomplished by referencing the STDD integrals for the individual protons to the strongest STD signal in each spectrum, which was assigned as 100%. Table 5.13. Group epitope mapping (GEM) analysis for CB1/CB2 and haplosamate A (relative toH18/H21 at 0.96 ppm). N 1 2 3 4 5 6 7 8 9 Proton H28 H27 H26 H18, H21 H22p\H19,Hlcc H11,H24 H14,Hlp \H5,Hl l H2p\ H22a, H20 H6 8 (ppm) 0.75 0.80 0.88 0.95,0.97 1.21, 1.25, 1.27 1.420, 1.423 1.49, 1.51, 1.5421, 1.5422 1.86,1.87, 1.88 4.15 CB1STD(%) 76 72 91 100 59 30 37 37 21 CB2STD(%) 49 59 76 100 39 - - 26 50 272 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A 0.11 0.12 0.12 0.12 0.12 4.5 4.0 3.5 3.0 2.5 2.0 Chemical Shift (ppm) 1.5 1.0 Figure 5.22. Integration of proton signals for the STD percentage measurement in a suspension of the cannabinol human receptor CB 1 supported in insect cells and haplosamate A (47) (relative toH187H21at0.96ppm). 273 Chapter 5. Cannabinoid Activity of the Marine Sterol Haplosamate A O.II 0.12 0.12 0.12 0.12 4.5 4.0 3.5 3.0 2.5 2.0 Chemical Shift (ppm) 1.5 1.0 Figure 5.23. Integration of proton signals for the STD percentage measurement in a suspension of the cannabinol human receptor CB2 supported in insect cells and haplosamate A (47) (relative toH18/H21 at 0.96 ppm). 274 Chapter 6. Synthesis of Liphagal Analogues 6. Synthesis of Liphagal Analogues 6.1. Liphagal, a selective PI3Koc inhibitor The phosphatidylinositol-3-kinase (PI3K) signaling pathway plays a central role in regulating cell proliferation and survival, adhesion, membrane trafficking, movement, differentiation, glucose transport, neurite outgrowth, as well as superoxide production.471"47 There are several closely related PI3K isoforms exhibiting different biological activities,471'474"476 and over the past five years a growing appreciation of the therapeutic potential of PI3K inhibitors has encouraged significant efforts within the pharmaceutical industry to identify new inhibitory compounds with enhanced potency, selectivity and pharmacological properties. 7 ' 77 Such drugs are destined for the treatment of inflammatory and autoimmune disorders as well as cancer and cardiovascular diseases.472,475'478 When a library of marine invertebrate extracts was screened as part of a program designed to find new isoform-selective PI3K inhibitors, the meroterpenoid liphagal (1) was identified via bioassay-guided fractionation of extracts from the sponge Aka coralliphaga collected in Dominica.472 Liphagal (1), which has an unprecedented carbon skeleton, inhibited PI3Kcc with an IC50 of 100 nM and approximately a 10-fold selectivity compared to PDKyin a fluorescent polarization enzyme bioassay.472 Liphagal (1) provided one more chemotype 75 for the development of new synthetic PI3K inhibitors useful as biological tools and potential drug candidates. 275 Chapter 6. Synthesis of Liphagal Analogues 6.2. The phosphatidylinositol-3-kinase (PI3K) signaling pathway The term PI3K is applied to a large family of lipid signaling kinases that catalyze the phosphorylation of phosphatidylinositol-4,5-bisphosphate (PI-4,5-P2 2) giving rise to the second messenger phosphatidylinositol-3,4,5-trisphosphate (PI-3,4,5-P3 3).471.475.478 Although undetectable in resting cells, this compound is synthesized by PBK's in response to a wide array of extracellular stimuli,472' 74 and provides a critical signal for a downstream cascade of events that control diverse cellular processes (see Section 6.1 and Figure 6.1).471'472 Class IA PI3K Ti'n'nWrfrtrriiwr U U ii ii U11 Xl& 111 1 t >• ( l( Ii !(J( U U Uiililili'iiiUUUcURilJll HO 2 PI-4,5-P2 Growth Metabolism Survival Proliferation Differentiation Development Apoptosis Adhesion Motility Migration Phagocytosis Figure 6.1. Enzymatic synthesis and degradation of PI-3,4,5-P3 (3) during the first steps of the PBK signaling pathway 473,476 In order to terminate PBK signaling and ensure a tight regulation, cells degrade PI-3,4,5- P3 (3) back to PI-4,5-P2 (2) via PTEN (a 3'-phosphatase termed phosphatase and tensin 276 Chapter 6. Synthesis of Liphagal Analogues homology deleted on chromosome ten), or to phosphatidylinositol-3,4-bisphosphate (PI-3,4-P2 4) through at least three phosphatases: SHIP, sSHIP and SHIP2 (Src homology 2-containing inositol 5-phophatases).471'479'480 Found only in blood cells, SHIP is a negative regulator of mast cell and macrophage activation, osteoclast formation and resorptive function.479'481 Its product PI-3,4-P2 (4) can also mediate PI3K-dependent responses. 471 Table 6.1. Organization of the PDK's family 471,475 Class IA IB IIa m Catalytic pi 10a pllOp pll0p8 pi lay subunit PI3K-C2a/p/y PI3K Adaptor/Binding partner p85a, p50a, p55a p85p p55a pl01/p84 Clathrinb pl50 Distribution Broad Broad Leukocytes Leukocytes Broadc Broad Three isoforms, y is liver specific. A protein involved in endocytosis. cResistant to wortmannin. Different types of PBK's have been identified and grouped into three classes according to their primary and secondary structures, mode of regulation and substrate specificity (Table 6.1).471'475'478'480 Class I PDK's have been the most extensively studied, and consist of two 478 subgroups of heterodimeric proteins (Figure 6.1). The class IA members are composed of a catalytic 110 kDa subunit exhibiting three known isoforms (pi 10a, pliop, pi 108), and a tightly associated 85 kDa regulatory subunit that controls their expression, activation and subcellular localization (five isoforms: p85a, p85p\ p55y, p55a and p50a).471'480 They are often activated by growth factor and cytokine receptors through a tyrosine-kinase-dependent mechanism, and their main role seems to be the direction of energy into cell growth and proliferation.476'478 Unlike the omnipresent PI3Ka and PI3K[} isoforms, PI3K5 is predominantly expressed in the 277 Chapter 6. Synthesis of Liphagal Analogues haematopoietic (blood) system and possesses important roles in T and B-cell signaling, the neutrophil oxidative burst, and mast-cell-mediated allergic responses.475'478 The only class IB isoform described so far, PDKy, possesses its 110 kDa-catalytic subunit associated with one of two regulatory subunits, plOl and p84 (Figure 6.1).475'482 PDKyis mostly activated by seven-transmembrane-spanning G-protein-coupled receptors (GPCR's) through direct interaction with the G protein Py-subunits.475'478'480 Like PI3K5, it is also highly expressed in blood cells and controls processes in inflammation and allergy. 7 '478 The PI3K family is completed by the class II C2-domain-containing PBK's and the class III phosphatidylinositol-specific 3-kinases (Table 6.1), of which little is known.476,482 The differential tissue distribution of PI3K isoforms is a key factor in the distinct 478 biological functions of PI3Ks. Complete genetic inactivation of the omnipresent PI3Ka or PI3KP resulted in embryonic lethality, attributing essential and non-redundant roles to these isoforms. 7 '47 Therefore, from a drug development perspective pharmacological inhibition of PI3Ka and PI3KP activities is likely to be associated with significant toxicity.477 By contrast, PI3K8 and PI3Ky expression is mainly restricted to the haematopoietic system, and mutant mice deprived of their expression or function, either by deletion of the whole gene (pi 108 and pi lOy- knockout mice) or by mutation of the kinase domain (pi 108 and pi lOy kinase-inactive knock-in mice), are viable, fertile and apparently healthy with a normal life span.475'476'478'483 Thus, both PI3K8 and PDKy isoforms represent promising therapeutic targets to intervene signaling pathways involved in inflammatory and auto immune diseases473'476'478'483'484 such as rheumatoid arthritis, systemic lupus erythematosus (SLE), multiple sclerosis, asthma, chronic obstructive pulmonary disease, and psoriasis. Recent interest in PDK signaling has also been fuelled by evidence that the PI3K pathway is among the most commonly activated signaling pathways in cancer.477'480 For instance, 278 Chapter 6. Synthesis of Liphagal Analogues the PBKa isoform was found to be activated by mutation in colon, gastric and breast carcinomas,476 and is likely to be the most commonly mutated kinase in the human genome.477 The presence of PI3Kp mutations in colon cancer and its role in metastasis of a prostate mouse model was recently demonstrated.476 An increase of PI-3,4,5-P3 (3) has also proved to promote tumor progression in mutant mice lacking functional PTEN, a well-characterized tumor suppressor frequently inactivated in human cancer by mutation, gene deletion or epigenetic silencing.471'476'480 There is also evidence that SHIP acts as a tumor suppressor in both acute myelogenous leukemia (AML) and chronic myelogenous leukemia (CML). Furthermore, many tyrosine kinases that activate PI3K are themselves the target of mutations or amplification in cancer. Together, these observations clearly reveal a connection of genetic alterations in cancer that stimulate PI3K signaling, suggesting that PI3K activation is likely to be an essential step in tumorigenesis.476'480 Such an impressive variety of potential therapeutic applications has led some authors to compare a pure hypothetical isoform-selective PI3K inhibitor with classical cyclooxygenase- inhibitor drugs like aspirin.475 6.3. First generation PI3K inhibitors The early availability of PI3K inhibitors played an essential role in understanding PI3K signaling processes.472'477 The fungal natural product wortmannin (5),487"489 isolated from Penicillium wortmanni, was originally described as a potent inhibitor of the respiratory burst in neutrophils and monocytes,477'490 and later shown to target PDK's via nucleophilic attack of Lys833 (pi 10a within the ATP-binding site) at the highly electrophilic C20 position of the furan ring yielding an enamine (Figure 6.2).471'477 The concentration of wortmannin (5) required for 279 Chapter 6. Synthesis of Liphagal Analogues •471 PI3K inhibition ranges between 1-100 nM. Equivalent nucleophilic residues can be found in all PI3 and protein kinases, and probably account for the poor selectivity of 5. 473,476,478 HN V ^ H+ Lys833 HN; Lys833 OH O OH OH Figure 6.2. First generation PI3K inhibitors. x491-494 The widely distributed flavonoid quercetin (6) " showed inhibition of PI3K with an IC50 of 3.8 jiM, but also exhibited poor selectivity by acting on PI4K's as well as several tyrosine and serine/threonine kinases.471'477 This compound was used as a template for SAR studies, from which the synthetic inhibitor LY294002 (7) (IC50 100 tlM) was identified.494,495 Unlike quercetin (6), LY294002 (7) had no detectable effect on other ATP-requiring enzymes. Both compounds are competitive and inhibit PDK's in a reversible fashion at the ATP binding site. 77' Wortmannin (5) and LY294002 (7) have been extensively used for more than a decade to analyze PI3K-driven pathways.477' 7 ' These molecules, however, do not exhibit any degree of selectivity for individual PI3K isoforms and moreover, they have been shown to also block class II and class III PBKs, as well as other closely related enzymes such as mammalian target of 280 Chapter 6. Synthesis of Liphagal Analogues rapamycin (MTOR), and unrelated enzymes such as casein kinase 2 (CK2), myosin light chain kinase (MLCK) and polo-like kinase (PLK).471'474'478'480 6.4. Isoform-selective second generation inhibitors Although the multiple roles of PDK's initially raised concerns about potential target- related toxicity and unwanted side effects, the benign phenotypes of the PI3K8 and PDKy mutant mice (knock-out and knock-in), as well as their confined expression pattern (blood cells) gained interest from pharmaceutical companies to pursue the development of orally active and selective small-molecule inhibitors of PI3K.475'477'496 As a result, over the past five years there has been a rapid increase in patenting activity disclosing new isoform-selective PI3K-inhibitor chemotypes, for the treatment of various human diseases including inflammatory autoimmune conditions (rheumatoid arthritis, SLE and psoriasis), allergic diseases (asthma), and cardiovascular disorders (thrombosis, atherosclerosis, cardiac hypertrophy).473"475'478'483 In 2006 alone, eight different lead compounds (including liphagal 1) entered the patent literature.475 From a structural viewpoint, all PI3K inhibitors available to date can be classified in two groups: derivatives of LY294002 (arylmorpholine compounds) and non-related structures.474*475 The available co-crystal structure of PI3Ky with LY294002 (7) provided unique and valuable insights into the general binding mode and key interactions involved in PDKy binding (Figure 6.3).471'475'477 The oxygen in the morpholino ring forms a hydrogen bond with Val-882, whereas another hydrogen bridge links the chromone keto-oxygen and Lys-833. The first interaction is shared by all PI3K inhibitors and ATP, and a clamp of approximately 8 A spanned between these two oxygen atoms is known to be essential for PI3K binding, making it a common motif throughout many chemotypes of PI3K inhibitors (Figure 6.3).477 281 Chapter 6. Synthesis of Liphagal Analogues Lys833 £ o ° O = 3  -cr 7  Val882 R 2 ^ R 3 ^ R, O rT^r^x R4 I ^ O Eli Lilly, 1991 PI3Ka Kinacia, 2001 PI3KP HCX cw, T2 rQ o k^o J/ F 2 HC^^N ' Chiron, 2004 c < ^ ^ — - ° PI3Ka N ^ N CHF2 °^Y J R2 £K Zenyaku Kogyo, 2005 PI3Ka B OH H2N ICOS, 2001 PI3K5 CHO UBC/Wyeth, 2006 PI3Ka Figure 6.3. Some PI3K-inhibitory chemotypes: A) arylmorpholine and B) non-related structures. The quinazolinone purine IC87114 (8) patented by ICOS 2001 is one of the most remarkable isoform-selective inhibitors described to date.474'477'497 The compound exhibited nanomolar inhibition of PI3K8 and a 100-1000-fold selectivity against the other class I PI3K's.497 In the following years, a series of aminothiazoles (9-11)498'499 and amino-bis-thiazoles (12)500 have been disclosed by Novartis, Boehringer and Serono, with a primary focus on respiratory diseases. Their linked aromatic and heteroaromatic cycles have been found to interact with the inositol binding site of PI3Ky.475 282 Chapter 6. Synthesis of Liphagal Analogues N VL -R1 Novartis, 2003 PI3Ky R2 R i ~ ^ , N - T " R 2 10 11 " 4 Novartis, 2005 Boehringer, 2006 12 PI3Ky PI3Kv Serono, 2005 PI3Ky Almost simultaneously, Pfizer and Serono presented in 2004 thiazolidinediones derivatives (13 and 14)501'502 for the treatment of rheumatoid disease. In these compounds, the 475 slightly acidic imide-NH of the thiazolidinedione motif was shown to bind PI3Ky. A third 503 class of inhibitors is based on the alkaloid pteridine (15), and the lead compound of the series, 504 • TG100-115 (16), is currently in preclinical development for acute myocardial infarction. OH X. ~ N R I S\ AA I =W >~" 13 Serono, 2004 PI3Ky NH vo 14 Pfizer, 2004 PI3Ky N I* CO N N' 15 NH 2 H N ^ - A N -2N N N 16 OH Targegen, 2004 PI3Ky Another structurally unrelated class of PI3K inhibitors was released by Bayer in 2004, and consists of imidazo[l,2-c]quinazolines (17)505 claiming a broad coverage of inflammatory and immunoregulatory disorders. 17 Bayer, 2004 PI3Ky In general, all these molecules are reversible inhibitors that bind the ATP-binding pocket 477 of PI3K and exhibit nanomolar affinity for their primary target. While no absolute preference 283 Chapter 6. Synthesis of Liphagal Analogues between PI3K isoforms has been achieved, some of them present a remarkable ~100-fold selectivity. Co-crystal structures for some of these inhibitors and pllOy are also available.496'506 6.5. Marine natural products and the PI3K signaling pathway A year before the reported isolation of liphagal (1) in 2006, the Andersen research group published the isolation and synthesis of another inositolphosphatase-active meroterpenoid. Identified from the sponge Dactylospongia elegans collected in Papua New Guinea, pelorol (18) 7 ' 7' °8 exhibited selective in vitro activation of SHIP, whereas its C20 methyl analogue (19) showed promising in vivo activity in two mouse models of inflammation, confirming SHIP- activators as a new class of anti-inflammatory agents. 479 18 19 20 21 509 Also in 2005, Clardy and coworkers isolated a previously unreported bromotyrosine derivative from the sponge Psammaplysilla sp., collected in the Indian Ocean. Named psammaplysene A (20), the compound was shown to compensate for loss of PTEN (PTEN deficiencies have been observed in several human malignancies, Section 6.2) by relocalizing the transcription factor FOXOla, one of its downstream targets. In another example, Xie and researchers510 demonstrated that scalaradial (21),511'512 isolated from the sponge Cacospongia 284 Chapter 6. Synthesis of Liphagal Analogues sp., binds to epidermal growth factor receptors (EGFR) and inhibits phosphorylation. EGFR's are one of the multiple activators of the PI3K signaling pathway. ff^l fl Y H UO H fl . H fi S ^ H I H  \/< H I X T ^ » YX» XX 22 o HN- CI -5 Janmaat and colleagues confirmed that inhibition of the PI3K pathway is an important determinant for the in vitro cytotoxic activity of kahalalide F (22), a known antitumor agent originally isolated from the Hawaiian marine mollusk Elysia rufescens514'516 and currently undergoing phase II clinical trials (Chapter 2, Section 2.1). Similarly, Ohizumi and his research group proposed the inhibition of PI3K activity as a mechanism for the observed halenaquinone517"519 (23)-induced apoptosis of nerve-growth factor PC12 cells. 23 >CHO 24 More recently in 2007, Sun and coworkers reported that hyrtiosal (24), ' isolated from the marine sponge Hyrtios erectus, is a noncompetitive inhibitor of protein tyrosine phosphatase IB (PTP1B), a negative modulator of insulin signaling and one of the many downstream enzymes involved in the PI3K signaling pathway. Likewise, Kwon and Nam523 found that a PI3K-related insulin growth factor mediates apoptosis in gastric cancer cells when 285 Chapter 6. Synthesis of Liphagal Analogues they are treated with an unidentified polysaccharide extracted from the marine algae Capsosiphon fulvescens, commonly used as foodstuff in Korea. 6.6. Biogenesis and synthetic preparations of liphagal Liphagal (1) represented the first example of the "liphagane" meroterpenoid carbon 479 skeleton. Its biosynthesis may involve a proton initiated polyene cyclization, with C2' of farnesylated trihydroxybenzaldehyde (25) acting as a nucleophilic center, followed by two Wagner-Meerwein hydride shifts to produce siphonodictyal B (27) (Scheme 6.1).472 This compound was first isolated in 1981 by Faulkner and coworkers524'525 from specimens of Aka 595 59fi coralliphaga collected in Belize. In 2003, Schmitz and his group found a sulfated version in A. coralliphaga sponges taken from the coasts of Micronesia, whereas more recently Kock and 597 Assmann reisolated 27 from the same organism together with eleven previously unreported analogues, including several orr/zo-quinones with radical-scavenging properties. CHO CHO OH OH 25 26 27 28 n j L OH OH HO Scheme 6.1. Proposed biogenesis of (+)-liphagal (1).' 286 Chapter 6. Synthesis of Liphagal Analogues Continuing from siphonodictyal B (27), a seven-membered ring could be generated via ring expansion of epoxide (28) to afford ketone (29), which undergoes epimerization at C8 and hemiketal formation to give 30. In the last step, dehydration of 30 would yield liphagal (1) 472 - ° W C H ° 1 . B B r 3 , C H 2 C I 2 , 8 7 % > / ° ^ ^ C H O L T B S C I J m . CH 2CI 2 ^ ^ C X ^ ^ O H ^Q^K^f^0^ 2. Br2, NaOAc, HOAc ^ c r ' ^ X l H 2 ' N a B H - » . MeOH, 94% ^ c r ^ S ^ X I T B S Br Br 31 32 33 i I 35 MeOCH2PPh3CI U (f (BuOK, THF, 100% 1. PPh3HBr, CH3CN 2. HF/pyridine, THF 9 4 % 36 1. PPTs, MeOH, 90% 2. PPTs, H 2 0 , Acetone, 8 4 % ' ° > p * ^ P P h 3 B r i I Br 34 CHO 37 NaCI0 2 , N a H 2 P 0 4 amylene, acetone, H 2 0 , 60% —Q COOH 38 OH or- CHO Bl 3 , -78°C to rt CH2CI2 , 64% CHO 1. DCC, DMAP, 34 CH 2 CI 2 2. Et 3N, THF, 8 0 % CI